diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,297152 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 0.9999941101981918, + "eval_steps": 500, + "global_step": 42446, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "grad_norm": 1.2233918720371746, + "learning_rate": 1.569858712715856e-08, + "loss": 1.0054, + "step": 1 + }, + { + "epoch": 0.0, + "grad_norm": 2.774883504800584, + "learning_rate": 3.139717425431712e-08, + "loss": 1.1368, + "step": 2 + }, + { + "epoch": 0.0, + "grad_norm": 2.549451882854754, + "learning_rate": 4.709576138147567e-08, + "loss": 1.0416, + "step": 3 + }, + { + "epoch": 0.0, + "grad_norm": 2.803962883878213, + "learning_rate": 6.279434850863424e-08, + "loss": 1.2588, + "step": 4 + }, + { + "epoch": 0.0, + "grad_norm": 2.624965400694845, + "learning_rate": 7.849293563579278e-08, + "loss": 1.0941, + "step": 5 + }, + { + "epoch": 0.0, + "grad_norm": 2.956669302933711, + "learning_rate": 9.419152276295134e-08, + "loss": 1.2385, + "step": 6 + }, + { + "epoch": 0.0, + "grad_norm": 3.7295310434703124, + "learning_rate": 1.098901098901099e-07, + "loss": 1.0, + "step": 7 + }, + { + "epoch": 0.0, + "grad_norm": 2.7174260700152213, + "learning_rate": 1.2558869701726847e-07, + "loss": 1.0737, + "step": 8 + }, + { + "epoch": 0.0, + "grad_norm": 2.5781498691052955, + "learning_rate": 1.4128728414442702e-07, + "loss": 1.1172, + "step": 9 + }, + { + "epoch": 0.0, + "grad_norm": 2.58915701901822, + "learning_rate": 1.5698587127158556e-07, + "loss": 1.2309, + "step": 10 + }, + { + "epoch": 0.0, + "grad_norm": 2.5600383264080238, + "learning_rate": 1.726844583987441e-07, + "loss": 1.1556, + "step": 11 + }, + { + "epoch": 0.0, + "grad_norm": 2.785299577342246, + "learning_rate": 1.8838304552590267e-07, + "loss": 1.2105, + "step": 12 + }, + { + "epoch": 0.0, + "grad_norm": 3.666175409979818, + "learning_rate": 2.0408163265306121e-07, + "loss": 1.0583, + "step": 13 + }, + { + "epoch": 0.0, + "grad_norm": 4.554411447188799, + "learning_rate": 2.197802197802198e-07, + "loss": 1.1477, + "step": 14 + }, + { + "epoch": 0.0, + "grad_norm": 2.581599127236383, + "learning_rate": 2.3547880690737838e-07, + "loss": 1.1262, + "step": 15 + }, + { + "epoch": 0.0, + "grad_norm": 2.8006924128974657, + "learning_rate": 2.5117739403453695e-07, + "loss": 1.0938, + "step": 16 + }, + { + "epoch": 0.0, + "grad_norm": 3.648402194862145, + "learning_rate": 2.6687598116169546e-07, + "loss": 1.1267, + "step": 17 + }, + { + "epoch": 0.0, + "grad_norm": 2.956522575169481, + "learning_rate": 2.8257456828885403e-07, + "loss": 1.1758, + "step": 18 + }, + { + "epoch": 0.0, + "grad_norm": 1.0981740275467426, + "learning_rate": 2.982731554160126e-07, + "loss": 0.9695, + "step": 19 + }, + { + "epoch": 0.0, + "grad_norm": 2.4836218386921964, + "learning_rate": 3.139717425431711e-07, + "loss": 1.107, + "step": 20 + }, + { + "epoch": 0.0, + "grad_norm": 2.934234704679603, + "learning_rate": 3.296703296703297e-07, + "loss": 1.0285, + "step": 21 + }, + { + "epoch": 0.0, + "grad_norm": 3.5985294435890496, + "learning_rate": 3.453689167974882e-07, + "loss": 1.1793, + "step": 22 + }, + { + "epoch": 0.0, + "grad_norm": 2.634905465715372, + "learning_rate": 3.610675039246468e-07, + "loss": 1.2512, + "step": 23 + }, + { + "epoch": 0.0, + "grad_norm": 2.9549774690442194, + "learning_rate": 3.7676609105180534e-07, + "loss": 1.1237, + "step": 24 + }, + { + "epoch": 0.0, + "grad_norm": 3.0703527431847575, + "learning_rate": 3.924646781789639e-07, + "loss": 1.2955, + "step": 25 + }, + { + "epoch": 0.0, + "grad_norm": 3.0423096119208113, + "learning_rate": 4.0816326530612243e-07, + "loss": 1.2439, + "step": 26 + }, + { + "epoch": 0.0, + "grad_norm": 2.643830343663285, + "learning_rate": 4.2386185243328105e-07, + "loss": 1.0324, + "step": 27 + }, + { + "epoch": 0.0, + "grad_norm": 3.0472236042624252, + "learning_rate": 4.395604395604396e-07, + "loss": 1.0963, + "step": 28 + }, + { + "epoch": 0.0, + "grad_norm": 4.70317448387221, + "learning_rate": 4.5525902668759813e-07, + "loss": 1.1516, + "step": 29 + }, + { + "epoch": 0.0, + "grad_norm": 2.6193932609480783, + "learning_rate": 4.7095761381475676e-07, + "loss": 1.1417, + "step": 30 + }, + { + "epoch": 0.0, + "grad_norm": 2.7913964294060123, + "learning_rate": 4.866562009419152e-07, + "loss": 1.249, + "step": 31 + }, + { + "epoch": 0.0, + "grad_norm": 4.127412454071907, + "learning_rate": 5.023547880690739e-07, + "loss": 1.0855, + "step": 32 + }, + { + "epoch": 0.0, + "grad_norm": 2.916594331630571, + "learning_rate": 5.180533751962324e-07, + "loss": 1.2459, + "step": 33 + }, + { + "epoch": 0.0, + "grad_norm": 2.369100393636127, + "learning_rate": 5.337519623233909e-07, + "loss": 1.1498, + "step": 34 + }, + { + "epoch": 0.0, + "grad_norm": 3.009414481560615, + "learning_rate": 5.494505494505495e-07, + "loss": 1.126, + "step": 35 + }, + { + "epoch": 0.0, + "grad_norm": 2.716883971287457, + "learning_rate": 5.651491365777081e-07, + "loss": 1.0694, + "step": 36 + }, + { + "epoch": 0.0, + "grad_norm": 2.419907145728605, + "learning_rate": 5.808477237048666e-07, + "loss": 1.1291, + "step": 37 + }, + { + "epoch": 0.0, + "grad_norm": 3.962903543294247, + "learning_rate": 5.965463108320252e-07, + "loss": 0.8793, + "step": 38 + }, + { + "epoch": 0.0, + "grad_norm": 3.3196410117224473, + "learning_rate": 6.122448979591837e-07, + "loss": 1.2793, + "step": 39 + }, + { + "epoch": 0.0, + "grad_norm": 1.1604121063334434, + "learning_rate": 6.279434850863422e-07, + "loss": 0.9604, + "step": 40 + }, + { + "epoch": 0.0, + "grad_norm": 1.0806949484508648, + "learning_rate": 6.436420722135009e-07, + "loss": 0.9971, + "step": 41 + }, + { + "epoch": 0.0, + "grad_norm": 2.48117068376233, + "learning_rate": 6.593406593406594e-07, + "loss": 1.1449, + "step": 42 + }, + { + "epoch": 0.0, + "grad_norm": 2.2681322126214414, + "learning_rate": 6.750392464678179e-07, + "loss": 1.19, + "step": 43 + }, + { + "epoch": 0.0, + "grad_norm": 4.462358553824224, + "learning_rate": 6.907378335949764e-07, + "loss": 1.0042, + "step": 44 + }, + { + "epoch": 0.0, + "grad_norm": 2.2803151037015867, + "learning_rate": 7.064364207221351e-07, + "loss": 1.0777, + "step": 45 + }, + { + "epoch": 0.0, + "grad_norm": 2.441619010523774, + "learning_rate": 7.221350078492937e-07, + "loss": 1.1, + "step": 46 + }, + { + "epoch": 0.0, + "grad_norm": 1.0798508124018094, + "learning_rate": 7.378335949764521e-07, + "loss": 0.9538, + "step": 47 + }, + { + "epoch": 0.0, + "grad_norm": 2.400676100014351, + "learning_rate": 7.535321821036107e-07, + "loss": 1.1456, + "step": 48 + }, + { + "epoch": 0.0, + "grad_norm": 2.492666465581212, + "learning_rate": 7.692307692307694e-07, + "loss": 1.0806, + "step": 49 + }, + { + "epoch": 0.0, + "grad_norm": 2.6029181649820408, + "learning_rate": 7.849293563579278e-07, + "loss": 1.1357, + "step": 50 + }, + { + "epoch": 0.0, + "grad_norm": 2.651965092496999, + "learning_rate": 8.006279434850864e-07, + "loss": 1.0655, + "step": 51 + }, + { + "epoch": 0.0, + "grad_norm": 2.407516603425903, + "learning_rate": 8.163265306122449e-07, + "loss": 1.0737, + "step": 52 + }, + { + "epoch": 0.0, + "grad_norm": 2.4682565196630586, + "learning_rate": 8.320251177394035e-07, + "loss": 1.2563, + "step": 53 + }, + { + "epoch": 0.0, + "grad_norm": 3.0479370321850476, + "learning_rate": 8.477237048665621e-07, + "loss": 1.1382, + "step": 54 + }, + { + "epoch": 0.0, + "grad_norm": 2.3857238018726536, + "learning_rate": 8.634222919937206e-07, + "loss": 1.1595, + "step": 55 + }, + { + "epoch": 0.0, + "grad_norm": 2.351166922681625, + "learning_rate": 8.791208791208792e-07, + "loss": 1.0541, + "step": 56 + }, + { + "epoch": 0.0, + "grad_norm": 2.2064391806281662, + "learning_rate": 8.948194662480378e-07, + "loss": 1.0044, + "step": 57 + }, + { + "epoch": 0.0, + "grad_norm": 2.519906214971554, + "learning_rate": 9.105180533751963e-07, + "loss": 1.0502, + "step": 58 + }, + { + "epoch": 0.0, + "grad_norm": 2.5215202939528236, + "learning_rate": 9.262166405023548e-07, + "loss": 0.9641, + "step": 59 + }, + { + "epoch": 0.0, + "grad_norm": 2.4919381868053296, + "learning_rate": 9.419152276295135e-07, + "loss": 1.2044, + "step": 60 + }, + { + "epoch": 0.0, + "grad_norm": 2.1817776629269026, + "learning_rate": 9.57613814756672e-07, + "loss": 1.1276, + "step": 61 + }, + { + "epoch": 0.0, + "grad_norm": 2.4302025190495518, + "learning_rate": 9.733124018838304e-07, + "loss": 1.1176, + "step": 62 + }, + { + "epoch": 0.0, + "grad_norm": 2.579824371081719, + "learning_rate": 9.890109890109891e-07, + "loss": 1.1728, + "step": 63 + }, + { + "epoch": 0.0, + "grad_norm": 2.858291157389531, + "learning_rate": 1.0047095761381478e-06, + "loss": 1.314, + "step": 64 + }, + { + "epoch": 0.0, + "grad_norm": 2.1564426081263597, + "learning_rate": 1.0204081632653063e-06, + "loss": 1.1145, + "step": 65 + }, + { + "epoch": 0.0, + "grad_norm": 3.5481921350673162, + "learning_rate": 1.0361067503924647e-06, + "loss": 1.2308, + "step": 66 + }, + { + "epoch": 0.0, + "grad_norm": 2.375500430226011, + "learning_rate": 1.0518053375196232e-06, + "loss": 1.0357, + "step": 67 + }, + { + "epoch": 0.0, + "grad_norm": 2.304623618291232, + "learning_rate": 1.0675039246467819e-06, + "loss": 0.982, + "step": 68 + }, + { + "epoch": 0.0, + "grad_norm": 2.2217214270223553, + "learning_rate": 1.0832025117739405e-06, + "loss": 1.1608, + "step": 69 + }, + { + "epoch": 0.0, + "grad_norm": 2.2943720603317828, + "learning_rate": 1.098901098901099e-06, + "loss": 1.1551, + "step": 70 + }, + { + "epoch": 0.0, + "grad_norm": 2.5209664912520413, + "learning_rate": 1.1145996860282575e-06, + "loss": 1.1879, + "step": 71 + }, + { + "epoch": 0.0, + "grad_norm": 2.2717720438555093, + "learning_rate": 1.1302982731554161e-06, + "loss": 1.0285, + "step": 72 + }, + { + "epoch": 0.0, + "grad_norm": 1.1089957246446518, + "learning_rate": 1.1459968602825746e-06, + "loss": 0.8977, + "step": 73 + }, + { + "epoch": 0.0, + "grad_norm": 2.1045490666555837, + "learning_rate": 1.1616954474097333e-06, + "loss": 1.0562, + "step": 74 + }, + { + "epoch": 0.0, + "grad_norm": 2.4405252280353382, + "learning_rate": 1.1773940345368917e-06, + "loss": 1.1593, + "step": 75 + }, + { + "epoch": 0.0, + "grad_norm": 2.074691575114656, + "learning_rate": 1.1930926216640504e-06, + "loss": 0.9963, + "step": 76 + }, + { + "epoch": 0.0, + "grad_norm": 2.443397239331446, + "learning_rate": 1.2087912087912089e-06, + "loss": 0.9573, + "step": 77 + }, + { + "epoch": 0.0, + "grad_norm": 2.320416922780358, + "learning_rate": 1.2244897959183673e-06, + "loss": 1.1326, + "step": 78 + }, + { + "epoch": 0.0, + "grad_norm": 2.5714172370452464, + "learning_rate": 1.240188383045526e-06, + "loss": 1.1014, + "step": 79 + }, + { + "epoch": 0.0, + "grad_norm": 1.1111724327328796, + "learning_rate": 1.2558869701726845e-06, + "loss": 0.9486, + "step": 80 + }, + { + "epoch": 0.0, + "grad_norm": 2.103236380786169, + "learning_rate": 1.2715855572998431e-06, + "loss": 1.0746, + "step": 81 + }, + { + "epoch": 0.0, + "grad_norm": 2.4540623291900685, + "learning_rate": 1.2872841444270018e-06, + "loss": 1.1535, + "step": 82 + }, + { + "epoch": 0.0, + "grad_norm": 2.180414852691283, + "learning_rate": 1.30298273155416e-06, + "loss": 1.2345, + "step": 83 + }, + { + "epoch": 0.0, + "grad_norm": 2.2674342348222045, + "learning_rate": 1.3186813186813187e-06, + "loss": 1.0608, + "step": 84 + }, + { + "epoch": 0.0, + "grad_norm": 1.181040191805885, + "learning_rate": 1.3343799058084774e-06, + "loss": 0.9533, + "step": 85 + }, + { + "epoch": 0.0, + "grad_norm": 2.1980216438054745, + "learning_rate": 1.3500784929356359e-06, + "loss": 1.062, + "step": 86 + }, + { + "epoch": 0.0, + "grad_norm": 1.0816177955817146, + "learning_rate": 1.3657770800627946e-06, + "loss": 0.9639, + "step": 87 + }, + { + "epoch": 0.0, + "grad_norm": 1.9511358377825037, + "learning_rate": 1.3814756671899528e-06, + "loss": 0.8569, + "step": 88 + }, + { + "epoch": 0.0, + "grad_norm": 2.034215749049191, + "learning_rate": 1.3971742543171115e-06, + "loss": 1.0519, + "step": 89 + }, + { + "epoch": 0.0, + "grad_norm": 2.07349327557405, + "learning_rate": 1.4128728414442702e-06, + "loss": 1.0018, + "step": 90 + }, + { + "epoch": 0.0, + "grad_norm": 2.8193300200553186, + "learning_rate": 1.4285714285714286e-06, + "loss": 1.1879, + "step": 91 + }, + { + "epoch": 0.0, + "grad_norm": 2.1437621024978073, + "learning_rate": 1.4442700156985873e-06, + "loss": 1.1944, + "step": 92 + }, + { + "epoch": 0.0, + "grad_norm": 2.203764927470732, + "learning_rate": 1.459968602825746e-06, + "loss": 1.0641, + "step": 93 + }, + { + "epoch": 0.0, + "grad_norm": 2.077746581591978, + "learning_rate": 1.4756671899529042e-06, + "loss": 1.0696, + "step": 94 + }, + { + "epoch": 0.0, + "grad_norm": 1.1319217469360052, + "learning_rate": 1.491365777080063e-06, + "loss": 0.8957, + "step": 95 + }, + { + "epoch": 0.0, + "grad_norm": 2.7451420080256375, + "learning_rate": 1.5070643642072214e-06, + "loss": 0.9567, + "step": 96 + }, + { + "epoch": 0.0, + "grad_norm": 2.798574437382459, + "learning_rate": 1.52276295133438e-06, + "loss": 1.0332, + "step": 97 + }, + { + "epoch": 0.0, + "grad_norm": 1.062616373120263, + "learning_rate": 1.5384615384615387e-06, + "loss": 0.9806, + "step": 98 + }, + { + "epoch": 0.0, + "grad_norm": 2.4337726943599325, + "learning_rate": 1.554160125588697e-06, + "loss": 1.2502, + "step": 99 + }, + { + "epoch": 0.0, + "grad_norm": 2.378540508631424, + "learning_rate": 1.5698587127158556e-06, + "loss": 1.105, + "step": 100 + }, + { + "epoch": 0.0, + "grad_norm": 2.1676327076577473, + "learning_rate": 1.5855572998430143e-06, + "loss": 1.1495, + "step": 101 + }, + { + "epoch": 0.0, + "grad_norm": 2.27198839684319, + "learning_rate": 1.6012558869701728e-06, + "loss": 1.2003, + "step": 102 + }, + { + "epoch": 0.0, + "grad_norm": 2.170436339118517, + "learning_rate": 1.6169544740973315e-06, + "loss": 1.12, + "step": 103 + }, + { + "epoch": 0.0, + "grad_norm": 2.149656421632478, + "learning_rate": 1.6326530612244897e-06, + "loss": 1.103, + "step": 104 + }, + { + "epoch": 0.0, + "grad_norm": 2.7196260666557595, + "learning_rate": 1.6483516483516484e-06, + "loss": 0.9662, + "step": 105 + }, + { + "epoch": 0.0, + "grad_norm": 2.131869421122366, + "learning_rate": 1.664050235478807e-06, + "loss": 1.0918, + "step": 106 + }, + { + "epoch": 0.0, + "grad_norm": 4.7104081361852845, + "learning_rate": 1.6797488226059655e-06, + "loss": 1.104, + "step": 107 + }, + { + "epoch": 0.0, + "grad_norm": 1.8521250397000681, + "learning_rate": 1.6954474097331242e-06, + "loss": 0.7746, + "step": 108 + }, + { + "epoch": 0.0, + "grad_norm": 2.065750599678774, + "learning_rate": 1.7111459968602829e-06, + "loss": 0.9518, + "step": 109 + }, + { + "epoch": 0.0, + "grad_norm": 2.264491907568744, + "learning_rate": 1.7268445839874411e-06, + "loss": 1.0501, + "step": 110 + }, + { + "epoch": 0.0, + "grad_norm": 2.097917006111737, + "learning_rate": 1.7425431711145998e-06, + "loss": 0.9951, + "step": 111 + }, + { + "epoch": 0.0, + "grad_norm": 1.0692342166551623, + "learning_rate": 1.7582417582417585e-06, + "loss": 0.9558, + "step": 112 + }, + { + "epoch": 0.0, + "grad_norm": 2.120707997788842, + "learning_rate": 1.773940345368917e-06, + "loss": 1.084, + "step": 113 + }, + { + "epoch": 0.0, + "grad_norm": 2.2268423452527193, + "learning_rate": 1.7896389324960756e-06, + "loss": 1.182, + "step": 114 + }, + { + "epoch": 0.0, + "grad_norm": 1.0720946235527253, + "learning_rate": 1.8053375196232339e-06, + "loss": 0.9612, + "step": 115 + }, + { + "epoch": 0.0, + "grad_norm": 2.5704654598087493, + "learning_rate": 1.8210361067503925e-06, + "loss": 1.0699, + "step": 116 + }, + { + "epoch": 0.0, + "grad_norm": 2.1083208309898667, + "learning_rate": 1.8367346938775512e-06, + "loss": 1.1106, + "step": 117 + }, + { + "epoch": 0.0, + "grad_norm": 2.3923121561408744, + "learning_rate": 1.8524332810047097e-06, + "loss": 1.1902, + "step": 118 + }, + { + "epoch": 0.0, + "grad_norm": 2.1229013170030195, + "learning_rate": 1.8681318681318684e-06, + "loss": 1.0513, + "step": 119 + }, + { + "epoch": 0.0, + "grad_norm": 2.2014418874715616, + "learning_rate": 1.883830455259027e-06, + "loss": 1.1637, + "step": 120 + }, + { + "epoch": 0.0, + "grad_norm": 2.304952426914051, + "learning_rate": 1.8995290423861853e-06, + "loss": 1.1123, + "step": 121 + }, + { + "epoch": 0.0, + "grad_norm": 1.9497552350335985, + "learning_rate": 1.915227629513344e-06, + "loss": 1.0838, + "step": 122 + }, + { + "epoch": 0.0, + "grad_norm": 2.229415981646212, + "learning_rate": 1.9309262166405024e-06, + "loss": 1.0625, + "step": 123 + }, + { + "epoch": 0.0, + "grad_norm": 2.732446505027916, + "learning_rate": 1.946624803767661e-06, + "loss": 0.9025, + "step": 124 + }, + { + "epoch": 0.0, + "grad_norm": 2.4680308450019064, + "learning_rate": 1.9623233908948198e-06, + "loss": 1.2296, + "step": 125 + }, + { + "epoch": 0.0, + "grad_norm": 1.958449358633352, + "learning_rate": 1.9780219780219782e-06, + "loss": 1.1205, + "step": 126 + }, + { + "epoch": 0.0, + "grad_norm": 2.014200738671417, + "learning_rate": 1.9937205651491367e-06, + "loss": 1.0862, + "step": 127 + }, + { + "epoch": 0.0, + "grad_norm": 2.3436275360776286, + "learning_rate": 2.0094191522762956e-06, + "loss": 1.1155, + "step": 128 + }, + { + "epoch": 0.0, + "grad_norm": 2.0871864816247228, + "learning_rate": 2.0251177394034536e-06, + "loss": 1.1495, + "step": 129 + }, + { + "epoch": 0.0, + "grad_norm": 2.0529303013727436, + "learning_rate": 2.0408163265306125e-06, + "loss": 1.2526, + "step": 130 + }, + { + "epoch": 0.0, + "grad_norm": 2.1376796322003946, + "learning_rate": 2.056514913657771e-06, + "loss": 1.0421, + "step": 131 + }, + { + "epoch": 0.0, + "grad_norm": 2.046897729718335, + "learning_rate": 2.0722135007849294e-06, + "loss": 1.0234, + "step": 132 + }, + { + "epoch": 0.0, + "grad_norm": 2.0323808748711145, + "learning_rate": 2.0879120879120883e-06, + "loss": 1.077, + "step": 133 + }, + { + "epoch": 0.0, + "grad_norm": 2.2697506503504465, + "learning_rate": 2.1036106750392464e-06, + "loss": 1.1608, + "step": 134 + }, + { + "epoch": 0.0, + "grad_norm": 2.124605916275208, + "learning_rate": 2.1193092621664052e-06, + "loss": 1.1522, + "step": 135 + }, + { + "epoch": 0.0, + "grad_norm": 2.0780203842687617, + "learning_rate": 2.1350078492935637e-06, + "loss": 1.0075, + "step": 136 + }, + { + "epoch": 0.0, + "grad_norm": 1.9427518834396533, + "learning_rate": 2.150706436420722e-06, + "loss": 1.0145, + "step": 137 + }, + { + "epoch": 0.0, + "grad_norm": 2.1477827433413017, + "learning_rate": 2.166405023547881e-06, + "loss": 1.1006, + "step": 138 + }, + { + "epoch": 0.0, + "grad_norm": 2.321510619618211, + "learning_rate": 2.1821036106750395e-06, + "loss": 1.0449, + "step": 139 + }, + { + "epoch": 0.0, + "grad_norm": 2.2298566990828808, + "learning_rate": 2.197802197802198e-06, + "loss": 1.0291, + "step": 140 + }, + { + "epoch": 0.0, + "grad_norm": 2.3888412298193065, + "learning_rate": 2.2135007849293564e-06, + "loss": 1.0459, + "step": 141 + }, + { + "epoch": 0.0, + "grad_norm": 2.4934550706460303, + "learning_rate": 2.229199372056515e-06, + "loss": 1.0977, + "step": 142 + }, + { + "epoch": 0.0, + "grad_norm": 2.125484914029918, + "learning_rate": 2.244897959183674e-06, + "loss": 1.1276, + "step": 143 + }, + { + "epoch": 0.0, + "grad_norm": 2.2529600161646974, + "learning_rate": 2.2605965463108323e-06, + "loss": 1.1124, + "step": 144 + }, + { + "epoch": 0.0, + "grad_norm": 2.2357158376574433, + "learning_rate": 2.2762951334379907e-06, + "loss": 1.1011, + "step": 145 + }, + { + "epoch": 0.0, + "grad_norm": 1.998548168980678, + "learning_rate": 2.291993720565149e-06, + "loss": 1.1931, + "step": 146 + }, + { + "epoch": 0.0, + "grad_norm": 2.304038304184638, + "learning_rate": 2.307692307692308e-06, + "loss": 1.1747, + "step": 147 + }, + { + "epoch": 0.0, + "grad_norm": 2.310152945308108, + "learning_rate": 2.3233908948194665e-06, + "loss": 0.916, + "step": 148 + }, + { + "epoch": 0.0, + "grad_norm": 2.1394383893446673, + "learning_rate": 2.339089481946625e-06, + "loss": 1.1961, + "step": 149 + }, + { + "epoch": 0.0, + "grad_norm": 1.191434416477073, + "learning_rate": 2.3547880690737835e-06, + "loss": 1.0145, + "step": 150 + }, + { + "epoch": 0.0, + "grad_norm": 2.0492504386246004, + "learning_rate": 2.370486656200942e-06, + "loss": 0.9844, + "step": 151 + }, + { + "epoch": 0.0, + "grad_norm": 2.20200147159523, + "learning_rate": 2.386185243328101e-06, + "loss": 1.168, + "step": 152 + }, + { + "epoch": 0.0, + "grad_norm": 2.2533581961353364, + "learning_rate": 2.4018838304552593e-06, + "loss": 1.0593, + "step": 153 + }, + { + "epoch": 0.0, + "grad_norm": 1.1132274436914669, + "learning_rate": 2.4175824175824177e-06, + "loss": 0.9992, + "step": 154 + }, + { + "epoch": 0.0, + "grad_norm": 2.9779483465233243, + "learning_rate": 2.4332810047095766e-06, + "loss": 1.0505, + "step": 155 + }, + { + "epoch": 0.0, + "grad_norm": 1.0722720171176792, + "learning_rate": 2.4489795918367347e-06, + "loss": 0.9693, + "step": 156 + }, + { + "epoch": 0.0, + "grad_norm": 2.766017287761176, + "learning_rate": 2.4646781789638936e-06, + "loss": 1.0163, + "step": 157 + }, + { + "epoch": 0.0, + "grad_norm": 2.072746880275649, + "learning_rate": 2.480376766091052e-06, + "loss": 1.0878, + "step": 158 + }, + { + "epoch": 0.0, + "grad_norm": 1.9789604159408853, + "learning_rate": 2.4960753532182105e-06, + "loss": 1.0267, + "step": 159 + }, + { + "epoch": 0.0, + "grad_norm": 2.1026758005214727, + "learning_rate": 2.511773940345369e-06, + "loss": 1.0572, + "step": 160 + }, + { + "epoch": 0.0, + "grad_norm": 1.982410755502118, + "learning_rate": 2.5274725274725274e-06, + "loss": 1.0054, + "step": 161 + }, + { + "epoch": 0.0, + "grad_norm": 2.4819869545658815, + "learning_rate": 2.5431711145996863e-06, + "loss": 1.1154, + "step": 162 + }, + { + "epoch": 0.0, + "grad_norm": 2.126434154376635, + "learning_rate": 2.5588697017268448e-06, + "loss": 1.1127, + "step": 163 + }, + { + "epoch": 0.0, + "grad_norm": 1.0814918876278992, + "learning_rate": 2.5745682888540036e-06, + "loss": 0.961, + "step": 164 + }, + { + "epoch": 0.0, + "grad_norm": 2.1078478308232302, + "learning_rate": 2.5902668759811617e-06, + "loss": 1.0894, + "step": 165 + }, + { + "epoch": 0.0, + "grad_norm": 2.160175893467477, + "learning_rate": 2.60596546310832e-06, + "loss": 1.0489, + "step": 166 + }, + { + "epoch": 0.0, + "grad_norm": 2.305177899189796, + "learning_rate": 2.621664050235479e-06, + "loss": 0.9078, + "step": 167 + }, + { + "epoch": 0.0, + "grad_norm": 2.1396822236813495, + "learning_rate": 2.6373626373626375e-06, + "loss": 1.0184, + "step": 168 + }, + { + "epoch": 0.0, + "grad_norm": 2.0024131841166026, + "learning_rate": 2.6530612244897964e-06, + "loss": 1.0897, + "step": 169 + }, + { + "epoch": 0.0, + "grad_norm": 2.12706089097061, + "learning_rate": 2.668759811616955e-06, + "loss": 1.0728, + "step": 170 + }, + { + "epoch": 0.0, + "grad_norm": 2.0053798095062314, + "learning_rate": 2.684458398744113e-06, + "loss": 1.0647, + "step": 171 + }, + { + "epoch": 0.0, + "grad_norm": 2.1088502071043647, + "learning_rate": 2.7001569858712718e-06, + "loss": 1.0022, + "step": 172 + }, + { + "epoch": 0.0, + "grad_norm": 2.102833988520245, + "learning_rate": 2.7158555729984302e-06, + "loss": 1.085, + "step": 173 + }, + { + "epoch": 0.0, + "grad_norm": 2.244831292457436, + "learning_rate": 2.731554160125589e-06, + "loss": 1.0358, + "step": 174 + }, + { + "epoch": 0.0, + "grad_norm": 2.325264626286564, + "learning_rate": 2.7472527472527476e-06, + "loss": 1.0899, + "step": 175 + }, + { + "epoch": 0.0, + "grad_norm": 2.4789516578902737, + "learning_rate": 2.7629513343799056e-06, + "loss": 1.0436, + "step": 176 + }, + { + "epoch": 0.0, + "grad_norm": 2.1419247782568234, + "learning_rate": 2.7786499215070645e-06, + "loss": 1.1183, + "step": 177 + }, + { + "epoch": 0.0, + "grad_norm": 1.098028223362932, + "learning_rate": 2.794348508634223e-06, + "loss": 0.9021, + "step": 178 + }, + { + "epoch": 0.0, + "grad_norm": 3.1028654814882404, + "learning_rate": 2.810047095761382e-06, + "loss": 1.0382, + "step": 179 + }, + { + "epoch": 0.0, + "grad_norm": 2.4192932883112954, + "learning_rate": 2.8257456828885403e-06, + "loss": 1.1305, + "step": 180 + }, + { + "epoch": 0.0, + "grad_norm": 2.5881148428321503, + "learning_rate": 2.8414442700156984e-06, + "loss": 1.1674, + "step": 181 + }, + { + "epoch": 0.0, + "grad_norm": 2.1112010212441508, + "learning_rate": 2.8571428571428573e-06, + "loss": 1.2737, + "step": 182 + }, + { + "epoch": 0.0, + "grad_norm": 2.739918414781274, + "learning_rate": 2.8728414442700157e-06, + "loss": 0.9343, + "step": 183 + }, + { + "epoch": 0.0, + "grad_norm": 2.0120519962722043, + "learning_rate": 2.8885400313971746e-06, + "loss": 1.0704, + "step": 184 + }, + { + "epoch": 0.0, + "grad_norm": 2.52838025624281, + "learning_rate": 2.904238618524333e-06, + "loss": 1.1384, + "step": 185 + }, + { + "epoch": 0.0, + "grad_norm": 2.273689114253073, + "learning_rate": 2.919937205651492e-06, + "loss": 1.1443, + "step": 186 + }, + { + "epoch": 0.0, + "grad_norm": 1.9930727986613561, + "learning_rate": 2.93563579277865e-06, + "loss": 1.0449, + "step": 187 + }, + { + "epoch": 0.0, + "grad_norm": 2.1876641108695054, + "learning_rate": 2.9513343799058085e-06, + "loss": 1.0734, + "step": 188 + }, + { + "epoch": 0.0, + "grad_norm": 2.3827345499202295, + "learning_rate": 2.9670329670329673e-06, + "loss": 1.0074, + "step": 189 + }, + { + "epoch": 0.0, + "grad_norm": 2.399611446042683, + "learning_rate": 2.982731554160126e-06, + "loss": 1.0779, + "step": 190 + }, + { + "epoch": 0.0, + "grad_norm": 1.8213140815775481, + "learning_rate": 2.9984301412872847e-06, + "loss": 0.9869, + "step": 191 + }, + { + "epoch": 0.0, + "grad_norm": 2.158139880254828, + "learning_rate": 3.0141287284144427e-06, + "loss": 1.2164, + "step": 192 + }, + { + "epoch": 0.0, + "grad_norm": 1.9924001195756273, + "learning_rate": 3.029827315541601e-06, + "loss": 1.1084, + "step": 193 + }, + { + "epoch": 0.0, + "grad_norm": 2.1838659115182932, + "learning_rate": 3.04552590266876e-06, + "loss": 1.0305, + "step": 194 + }, + { + "epoch": 0.0, + "grad_norm": 2.0909351110679903, + "learning_rate": 3.0612244897959185e-06, + "loss": 1.1452, + "step": 195 + }, + { + "epoch": 0.0, + "grad_norm": 1.164386088943603, + "learning_rate": 3.0769230769230774e-06, + "loss": 0.9014, + "step": 196 + }, + { + "epoch": 0.0, + "grad_norm": 1.1346858734397869, + "learning_rate": 3.092621664050236e-06, + "loss": 0.9739, + "step": 197 + }, + { + "epoch": 0.0, + "grad_norm": 2.013373290595924, + "learning_rate": 3.108320251177394e-06, + "loss": 1.0527, + "step": 198 + }, + { + "epoch": 0.0, + "grad_norm": 1.1499325477982547, + "learning_rate": 3.124018838304553e-06, + "loss": 0.939, + "step": 199 + }, + { + "epoch": 0.0, + "grad_norm": 2.2820467844438905, + "learning_rate": 3.1397174254317113e-06, + "loss": 1.0396, + "step": 200 + }, + { + "epoch": 0.0, + "grad_norm": 2.122623461534409, + "learning_rate": 3.15541601255887e-06, + "loss": 1.0363, + "step": 201 + }, + { + "epoch": 0.0, + "grad_norm": 1.1231557512373829, + "learning_rate": 3.1711145996860286e-06, + "loss": 0.9478, + "step": 202 + }, + { + "epoch": 0.0, + "grad_norm": 1.7683517765803995, + "learning_rate": 3.1868131868131867e-06, + "loss": 1.0692, + "step": 203 + }, + { + "epoch": 0.0, + "grad_norm": 2.065100515290166, + "learning_rate": 3.2025117739403456e-06, + "loss": 1.1406, + "step": 204 + }, + { + "epoch": 0.0, + "grad_norm": 2.2764827972210795, + "learning_rate": 3.218210361067504e-06, + "loss": 1.0795, + "step": 205 + }, + { + "epoch": 0.0, + "grad_norm": 2.4021360044865565, + "learning_rate": 3.233908948194663e-06, + "loss": 1.0068, + "step": 206 + }, + { + "epoch": 0.0, + "grad_norm": 2.0003201710482275, + "learning_rate": 3.2496075353218214e-06, + "loss": 1.1096, + "step": 207 + }, + { + "epoch": 0.0, + "grad_norm": 2.280599825206697, + "learning_rate": 3.2653061224489794e-06, + "loss": 1.0562, + "step": 208 + }, + { + "epoch": 0.0, + "grad_norm": 1.0935685572457474, + "learning_rate": 3.2810047095761383e-06, + "loss": 0.9387, + "step": 209 + }, + { + "epoch": 0.0, + "grad_norm": 2.446601679998678, + "learning_rate": 3.2967032967032968e-06, + "loss": 1.1275, + "step": 210 + }, + { + "epoch": 0.0, + "grad_norm": 2.1650355155771237, + "learning_rate": 3.3124018838304557e-06, + "loss": 1.1366, + "step": 211 + }, + { + "epoch": 0.0, + "grad_norm": 2.1651979893738362, + "learning_rate": 3.328100470957614e-06, + "loss": 1.1144, + "step": 212 + }, + { + "epoch": 0.01, + "grad_norm": 1.8330295148132778, + "learning_rate": 3.343799058084773e-06, + "loss": 0.9644, + "step": 213 + }, + { + "epoch": 0.01, + "grad_norm": 2.37701290687433, + "learning_rate": 3.359497645211931e-06, + "loss": 1.0178, + "step": 214 + }, + { + "epoch": 0.01, + "grad_norm": 1.9013090065987381, + "learning_rate": 3.3751962323390895e-06, + "loss": 1.0437, + "step": 215 + }, + { + "epoch": 0.01, + "grad_norm": 2.2575628740135327, + "learning_rate": 3.3908948194662484e-06, + "loss": 1.2109, + "step": 216 + }, + { + "epoch": 0.01, + "grad_norm": 2.108307750354133, + "learning_rate": 3.406593406593407e-06, + "loss": 1.1318, + "step": 217 + }, + { + "epoch": 0.01, + "grad_norm": 1.9932609720982355, + "learning_rate": 3.4222919937205657e-06, + "loss": 1.0738, + "step": 218 + }, + { + "epoch": 0.01, + "grad_norm": 2.262238130123094, + "learning_rate": 3.4379905808477238e-06, + "loss": 0.9719, + "step": 219 + }, + { + "epoch": 0.01, + "grad_norm": 2.0453568443868995, + "learning_rate": 3.4536891679748822e-06, + "loss": 1.0447, + "step": 220 + }, + { + "epoch": 0.01, + "grad_norm": 2.6455757445530304, + "learning_rate": 3.469387755102041e-06, + "loss": 1.0743, + "step": 221 + }, + { + "epoch": 0.01, + "grad_norm": 2.1819039347466993, + "learning_rate": 3.4850863422291996e-06, + "loss": 1.0062, + "step": 222 + }, + { + "epoch": 0.01, + "grad_norm": 1.0437406717600255, + "learning_rate": 3.5007849293563585e-06, + "loss": 0.9714, + "step": 223 + }, + { + "epoch": 0.01, + "grad_norm": 2.093582639481813, + "learning_rate": 3.516483516483517e-06, + "loss": 1.2473, + "step": 224 + }, + { + "epoch": 0.01, + "grad_norm": 2.1601348046166087, + "learning_rate": 3.532182103610675e-06, + "loss": 1.0494, + "step": 225 + }, + { + "epoch": 0.01, + "grad_norm": 2.3984398237063367, + "learning_rate": 3.547880690737834e-06, + "loss": 1.0003, + "step": 226 + }, + { + "epoch": 0.01, + "grad_norm": 2.204312768002084, + "learning_rate": 3.5635792778649923e-06, + "loss": 1.0707, + "step": 227 + }, + { + "epoch": 0.01, + "grad_norm": 2.1086561566501194, + "learning_rate": 3.5792778649921512e-06, + "loss": 1.0517, + "step": 228 + }, + { + "epoch": 0.01, + "grad_norm": 1.964919058551467, + "learning_rate": 3.5949764521193097e-06, + "loss": 1.0522, + "step": 229 + }, + { + "epoch": 0.01, + "grad_norm": 2.4201233398383004, + "learning_rate": 3.6106750392464677e-06, + "loss": 1.1182, + "step": 230 + }, + { + "epoch": 0.01, + "grad_norm": 2.1352996522498646, + "learning_rate": 3.6263736263736266e-06, + "loss": 1.0281, + "step": 231 + }, + { + "epoch": 0.01, + "grad_norm": 1.1324254026108005, + "learning_rate": 3.642072213500785e-06, + "loss": 0.9549, + "step": 232 + }, + { + "epoch": 0.01, + "grad_norm": 2.4211631228098534, + "learning_rate": 3.657770800627944e-06, + "loss": 1.0653, + "step": 233 + }, + { + "epoch": 0.01, + "grad_norm": 1.1721597300769662, + "learning_rate": 3.6734693877551024e-06, + "loss": 0.892, + "step": 234 + }, + { + "epoch": 0.01, + "grad_norm": 2.1090337340366814, + "learning_rate": 3.6891679748822605e-06, + "loss": 1.1752, + "step": 235 + }, + { + "epoch": 0.01, + "grad_norm": 2.0857240032073494, + "learning_rate": 3.7048665620094194e-06, + "loss": 1.1869, + "step": 236 + }, + { + "epoch": 0.01, + "grad_norm": 2.251857010691829, + "learning_rate": 3.720565149136578e-06, + "loss": 1.0373, + "step": 237 + }, + { + "epoch": 0.01, + "grad_norm": 2.3574102876705543, + "learning_rate": 3.7362637362637367e-06, + "loss": 1.1853, + "step": 238 + }, + { + "epoch": 0.01, + "grad_norm": 1.0776534437861638, + "learning_rate": 3.751962323390895e-06, + "loss": 0.9316, + "step": 239 + }, + { + "epoch": 0.01, + "grad_norm": 2.0471844364186866, + "learning_rate": 3.767660910518054e-06, + "loss": 0.9385, + "step": 240 + }, + { + "epoch": 0.01, + "grad_norm": 1.054229164095773, + "learning_rate": 3.783359497645212e-06, + "loss": 0.9862, + "step": 241 + }, + { + "epoch": 0.01, + "grad_norm": 2.043137493467671, + "learning_rate": 3.7990580847723706e-06, + "loss": 1.0916, + "step": 242 + }, + { + "epoch": 0.01, + "grad_norm": 2.0434887521257448, + "learning_rate": 3.814756671899529e-06, + "loss": 1.0105, + "step": 243 + }, + { + "epoch": 0.01, + "grad_norm": 2.1677563166928238, + "learning_rate": 3.830455259026688e-06, + "loss": 1.233, + "step": 244 + }, + { + "epoch": 0.01, + "grad_norm": 2.170512822493901, + "learning_rate": 3.846153846153847e-06, + "loss": 1.0952, + "step": 245 + }, + { + "epoch": 0.01, + "grad_norm": 1.9781307269010837, + "learning_rate": 3.861852433281005e-06, + "loss": 1.0883, + "step": 246 + }, + { + "epoch": 0.01, + "grad_norm": 2.1519039097530017, + "learning_rate": 3.877551020408164e-06, + "loss": 1.0732, + "step": 247 + }, + { + "epoch": 0.01, + "grad_norm": 2.1915602785010413, + "learning_rate": 3.893249607535322e-06, + "loss": 1.0544, + "step": 248 + }, + { + "epoch": 0.01, + "grad_norm": 3.2551884582071726, + "learning_rate": 3.908948194662481e-06, + "loss": 1.1744, + "step": 249 + }, + { + "epoch": 0.01, + "grad_norm": 2.1944155993726913, + "learning_rate": 3.9246467817896395e-06, + "loss": 1.0791, + "step": 250 + }, + { + "epoch": 0.01, + "grad_norm": 2.0647476085728913, + "learning_rate": 3.940345368916798e-06, + "loss": 1.0524, + "step": 251 + }, + { + "epoch": 0.01, + "grad_norm": 2.238142204792449, + "learning_rate": 3.9560439560439565e-06, + "loss": 1.0356, + "step": 252 + }, + { + "epoch": 0.01, + "grad_norm": 1.911691041375555, + "learning_rate": 3.9717425431711145e-06, + "loss": 1.1419, + "step": 253 + }, + { + "epoch": 0.01, + "grad_norm": 2.1883005210139594, + "learning_rate": 3.987441130298273e-06, + "loss": 0.96, + "step": 254 + }, + { + "epoch": 0.01, + "grad_norm": 2.2608752113696076, + "learning_rate": 4.003139717425432e-06, + "loss": 1.0237, + "step": 255 + }, + { + "epoch": 0.01, + "grad_norm": 1.9090079201628307, + "learning_rate": 4.018838304552591e-06, + "loss": 0.9172, + "step": 256 + }, + { + "epoch": 0.01, + "grad_norm": 2.602840498989665, + "learning_rate": 4.034536891679749e-06, + "loss": 1.0232, + "step": 257 + }, + { + "epoch": 0.01, + "grad_norm": 2.0772871025872073, + "learning_rate": 4.050235478806907e-06, + "loss": 1.0383, + "step": 258 + }, + { + "epoch": 0.01, + "grad_norm": 2.1277393553665735, + "learning_rate": 4.065934065934066e-06, + "loss": 0.9092, + "step": 259 + }, + { + "epoch": 0.01, + "grad_norm": 2.1826876870019456, + "learning_rate": 4.081632653061225e-06, + "loss": 1.0495, + "step": 260 + }, + { + "epoch": 0.01, + "grad_norm": 1.9628639609627703, + "learning_rate": 4.097331240188384e-06, + "loss": 0.8803, + "step": 261 + }, + { + "epoch": 0.01, + "grad_norm": 1.0881761074878462, + "learning_rate": 4.113029827315542e-06, + "loss": 0.9359, + "step": 262 + }, + { + "epoch": 0.01, + "grad_norm": 2.3416564009827723, + "learning_rate": 4.1287284144427e-06, + "loss": 1.0468, + "step": 263 + }, + { + "epoch": 0.01, + "grad_norm": 2.068569689962263, + "learning_rate": 4.144427001569859e-06, + "loss": 1.0695, + "step": 264 + }, + { + "epoch": 0.01, + "grad_norm": 1.8351317685140809, + "learning_rate": 4.160125588697018e-06, + "loss": 1.0089, + "step": 265 + }, + { + "epoch": 0.01, + "grad_norm": 2.108757577710367, + "learning_rate": 4.175824175824177e-06, + "loss": 1.0291, + "step": 266 + }, + { + "epoch": 0.01, + "grad_norm": 2.179906468988929, + "learning_rate": 4.191522762951335e-06, + "loss": 1.2089, + "step": 267 + }, + { + "epoch": 0.01, + "grad_norm": 2.1568144746388014, + "learning_rate": 4.207221350078493e-06, + "loss": 1.1494, + "step": 268 + }, + { + "epoch": 0.01, + "grad_norm": 2.1546682700596786, + "learning_rate": 4.222919937205652e-06, + "loss": 1.0444, + "step": 269 + }, + { + "epoch": 0.01, + "grad_norm": 2.5548484138023113, + "learning_rate": 4.2386185243328105e-06, + "loss": 1.0993, + "step": 270 + }, + { + "epoch": 0.01, + "grad_norm": 1.0816173386891343, + "learning_rate": 4.254317111459969e-06, + "loss": 0.9151, + "step": 271 + }, + { + "epoch": 0.01, + "grad_norm": 2.447854136357803, + "learning_rate": 4.270015698587127e-06, + "loss": 1.0594, + "step": 272 + }, + { + "epoch": 0.01, + "grad_norm": 2.0024313647388876, + "learning_rate": 4.2857142857142855e-06, + "loss": 1.0416, + "step": 273 + }, + { + "epoch": 0.01, + "grad_norm": 2.086777648307459, + "learning_rate": 4.301412872841444e-06, + "loss": 1.104, + "step": 274 + }, + { + "epoch": 0.01, + "grad_norm": 2.333969098903032, + "learning_rate": 4.317111459968603e-06, + "loss": 1.0936, + "step": 275 + }, + { + "epoch": 0.01, + "grad_norm": 2.213926715622632, + "learning_rate": 4.332810047095762e-06, + "loss": 1.0892, + "step": 276 + }, + { + "epoch": 0.01, + "grad_norm": 2.229507546804309, + "learning_rate": 4.34850863422292e-06, + "loss": 1.0809, + "step": 277 + }, + { + "epoch": 0.01, + "grad_norm": 1.9848347781482014, + "learning_rate": 4.364207221350079e-06, + "loss": 1.1134, + "step": 278 + }, + { + "epoch": 0.01, + "grad_norm": 2.207565250505493, + "learning_rate": 4.379905808477237e-06, + "loss": 1.0572, + "step": 279 + }, + { + "epoch": 0.01, + "grad_norm": 2.834704608610131, + "learning_rate": 4.395604395604396e-06, + "loss": 1.1179, + "step": 280 + }, + { + "epoch": 0.01, + "grad_norm": 2.168238083223041, + "learning_rate": 4.411302982731555e-06, + "loss": 1.0882, + "step": 281 + }, + { + "epoch": 0.01, + "grad_norm": 2.1062528499534823, + "learning_rate": 4.427001569858713e-06, + "loss": 1.0805, + "step": 282 + }, + { + "epoch": 0.01, + "grad_norm": 2.2919678454128425, + "learning_rate": 4.442700156985872e-06, + "loss": 1.055, + "step": 283 + }, + { + "epoch": 0.01, + "grad_norm": 4.193221711163101, + "learning_rate": 4.45839874411303e-06, + "loss": 0.9759, + "step": 284 + }, + { + "epoch": 0.01, + "grad_norm": 2.1041923531733513, + "learning_rate": 4.474097331240189e-06, + "loss": 1.2202, + "step": 285 + }, + { + "epoch": 0.01, + "grad_norm": 2.100538911485933, + "learning_rate": 4.489795918367348e-06, + "loss": 1.1558, + "step": 286 + }, + { + "epoch": 0.01, + "grad_norm": 2.260123545059462, + "learning_rate": 4.505494505494506e-06, + "loss": 1.0446, + "step": 287 + }, + { + "epoch": 0.01, + "grad_norm": 2.1353449202893664, + "learning_rate": 4.5211930926216645e-06, + "loss": 1.1356, + "step": 288 + }, + { + "epoch": 0.01, + "grad_norm": 2.21325551380608, + "learning_rate": 4.5368916797488226e-06, + "loss": 1.0412, + "step": 289 + }, + { + "epoch": 0.01, + "grad_norm": 2.0435205693547345, + "learning_rate": 4.5525902668759815e-06, + "loss": 0.959, + "step": 290 + }, + { + "epoch": 0.01, + "grad_norm": 2.5361294114755086, + "learning_rate": 4.56828885400314e-06, + "loss": 0.9986, + "step": 291 + }, + { + "epoch": 0.01, + "grad_norm": 2.0611037711954405, + "learning_rate": 4.583987441130298e-06, + "loss": 0.9326, + "step": 292 + }, + { + "epoch": 0.01, + "grad_norm": 2.332906927723471, + "learning_rate": 4.599686028257457e-06, + "loss": 1.1253, + "step": 293 + }, + { + "epoch": 0.01, + "grad_norm": 1.956352476419587, + "learning_rate": 4.615384615384616e-06, + "loss": 1.1149, + "step": 294 + }, + { + "epoch": 0.01, + "grad_norm": 2.22447608183365, + "learning_rate": 4.631083202511774e-06, + "loss": 1.0135, + "step": 295 + }, + { + "epoch": 0.01, + "grad_norm": 1.10979247601586, + "learning_rate": 4.646781789638933e-06, + "loss": 0.9387, + "step": 296 + }, + { + "epoch": 0.01, + "grad_norm": 4.023377443011198, + "learning_rate": 4.662480376766091e-06, + "loss": 0.9559, + "step": 297 + }, + { + "epoch": 0.01, + "grad_norm": 2.0596982940813278, + "learning_rate": 4.67817896389325e-06, + "loss": 1.1572, + "step": 298 + }, + { + "epoch": 0.01, + "grad_norm": 1.9409499076398655, + "learning_rate": 4.693877551020409e-06, + "loss": 1.0673, + "step": 299 + }, + { + "epoch": 0.01, + "grad_norm": 2.1331250708170577, + "learning_rate": 4.709576138147567e-06, + "loss": 1.1911, + "step": 300 + }, + { + "epoch": 0.01, + "grad_norm": 2.123751761721371, + "learning_rate": 4.725274725274726e-06, + "loss": 0.9492, + "step": 301 + }, + { + "epoch": 0.01, + "grad_norm": 2.3408569835663773, + "learning_rate": 4.740973312401884e-06, + "loss": 1.0935, + "step": 302 + }, + { + "epoch": 0.01, + "grad_norm": 2.171504902738344, + "learning_rate": 4.756671899529043e-06, + "loss": 0.957, + "step": 303 + }, + { + "epoch": 0.01, + "grad_norm": 2.2535068338422386, + "learning_rate": 4.772370486656202e-06, + "loss": 1.0501, + "step": 304 + }, + { + "epoch": 0.01, + "grad_norm": 2.33715213438924, + "learning_rate": 4.7880690737833605e-06, + "loss": 0.9709, + "step": 305 + }, + { + "epoch": 0.01, + "grad_norm": 2.521782730443791, + "learning_rate": 4.8037676609105186e-06, + "loss": 1.1, + "step": 306 + }, + { + "epoch": 0.01, + "grad_norm": 2.2513851017420015, + "learning_rate": 4.819466248037677e-06, + "loss": 1.1297, + "step": 307 + }, + { + "epoch": 0.01, + "grad_norm": 2.765893140384496, + "learning_rate": 4.8351648351648355e-06, + "loss": 0.9796, + "step": 308 + }, + { + "epoch": 0.01, + "grad_norm": 2.2462420285042803, + "learning_rate": 4.850863422291994e-06, + "loss": 1.1403, + "step": 309 + }, + { + "epoch": 0.01, + "grad_norm": 2.6916509140429343, + "learning_rate": 4.866562009419153e-06, + "loss": 0.9922, + "step": 310 + }, + { + "epoch": 0.01, + "grad_norm": 2.4142643206549312, + "learning_rate": 4.882260596546311e-06, + "loss": 0.9625, + "step": 311 + }, + { + "epoch": 0.01, + "grad_norm": 2.303474065408322, + "learning_rate": 4.897959183673469e-06, + "loss": 1.0892, + "step": 312 + }, + { + "epoch": 0.01, + "grad_norm": 2.450563002380457, + "learning_rate": 4.913657770800628e-06, + "loss": 0.9908, + "step": 313 + }, + { + "epoch": 0.01, + "grad_norm": 1.153925791667279, + "learning_rate": 4.929356357927787e-06, + "loss": 0.9436, + "step": 314 + }, + { + "epoch": 0.01, + "grad_norm": 2.051271581614704, + "learning_rate": 4.945054945054946e-06, + "loss": 1.0101, + "step": 315 + }, + { + "epoch": 0.01, + "grad_norm": 2.110095093587488, + "learning_rate": 4.960753532182104e-06, + "loss": 0.9123, + "step": 316 + }, + { + "epoch": 0.01, + "grad_norm": 2.0894129062780187, + "learning_rate": 4.976452119309262e-06, + "loss": 0.9551, + "step": 317 + }, + { + "epoch": 0.01, + "grad_norm": 2.45606791339142, + "learning_rate": 4.992150706436421e-06, + "loss": 0.9778, + "step": 318 + }, + { + "epoch": 0.01, + "grad_norm": 1.91172297950389, + "learning_rate": 5.007849293563579e-06, + "loss": 1.115, + "step": 319 + }, + { + "epoch": 0.01, + "grad_norm": 2.002502234678459, + "learning_rate": 5.023547880690738e-06, + "loss": 1.1698, + "step": 320 + }, + { + "epoch": 0.01, + "grad_norm": 2.3875047353151744, + "learning_rate": 5.039246467817897e-06, + "loss": 1.1469, + "step": 321 + }, + { + "epoch": 0.01, + "grad_norm": 2.1809156810162986, + "learning_rate": 5.054945054945055e-06, + "loss": 1.1168, + "step": 322 + }, + { + "epoch": 0.01, + "grad_norm": 2.454168531116619, + "learning_rate": 5.070643642072214e-06, + "loss": 0.9972, + "step": 323 + }, + { + "epoch": 0.01, + "grad_norm": 1.9087017825838286, + "learning_rate": 5.086342229199373e-06, + "loss": 0.9748, + "step": 324 + }, + { + "epoch": 0.01, + "grad_norm": 2.631530477706715, + "learning_rate": 5.1020408163265315e-06, + "loss": 0.9236, + "step": 325 + }, + { + "epoch": 0.01, + "grad_norm": 2.163703526322027, + "learning_rate": 5.1177394034536895e-06, + "loss": 1.21, + "step": 326 + }, + { + "epoch": 0.01, + "grad_norm": 1.8924508804297797, + "learning_rate": 5.133437990580848e-06, + "loss": 1.0441, + "step": 327 + }, + { + "epoch": 0.01, + "grad_norm": 2.36193883350647, + "learning_rate": 5.149136577708007e-06, + "loss": 1.0558, + "step": 328 + }, + { + "epoch": 0.01, + "grad_norm": 1.1187855940040101, + "learning_rate": 5.164835164835166e-06, + "loss": 0.9451, + "step": 329 + }, + { + "epoch": 0.01, + "grad_norm": 2.1291982209937648, + "learning_rate": 5.180533751962323e-06, + "loss": 1.2423, + "step": 330 + }, + { + "epoch": 0.01, + "grad_norm": 1.9336392665813584, + "learning_rate": 5.196232339089482e-06, + "loss": 1.0355, + "step": 331 + }, + { + "epoch": 0.01, + "grad_norm": 1.9765457501992774, + "learning_rate": 5.21193092621664e-06, + "loss": 1.118, + "step": 332 + }, + { + "epoch": 0.01, + "grad_norm": 1.908530393788132, + "learning_rate": 5.227629513343799e-06, + "loss": 1.0793, + "step": 333 + }, + { + "epoch": 0.01, + "grad_norm": 2.0190083328929505, + "learning_rate": 5.243328100470958e-06, + "loss": 1.1425, + "step": 334 + }, + { + "epoch": 0.01, + "grad_norm": 1.0224900223312507, + "learning_rate": 5.259026687598117e-06, + "loss": 0.9237, + "step": 335 + }, + { + "epoch": 0.01, + "grad_norm": 2.196230048921403, + "learning_rate": 5.274725274725275e-06, + "loss": 0.9818, + "step": 336 + }, + { + "epoch": 0.01, + "grad_norm": 2.2758757771894236, + "learning_rate": 5.290423861852434e-06, + "loss": 1.0908, + "step": 337 + }, + { + "epoch": 0.01, + "grad_norm": 1.1526278806203896, + "learning_rate": 5.306122448979593e-06, + "loss": 0.9513, + "step": 338 + }, + { + "epoch": 0.01, + "grad_norm": 1.9732829814482373, + "learning_rate": 5.321821036106752e-06, + "loss": 1.2437, + "step": 339 + }, + { + "epoch": 0.01, + "grad_norm": 2.5333115826354384, + "learning_rate": 5.33751962323391e-06, + "loss": 1.0498, + "step": 340 + }, + { + "epoch": 0.01, + "grad_norm": 1.1687133671682295, + "learning_rate": 5.353218210361068e-06, + "loss": 0.9453, + "step": 341 + }, + { + "epoch": 0.01, + "grad_norm": 1.0929315965408375, + "learning_rate": 5.368916797488226e-06, + "loss": 0.9353, + "step": 342 + }, + { + "epoch": 0.01, + "grad_norm": 1.0905193302729157, + "learning_rate": 5.384615384615385e-06, + "loss": 0.9707, + "step": 343 + }, + { + "epoch": 0.01, + "grad_norm": 2.3579560670552016, + "learning_rate": 5.4003139717425436e-06, + "loss": 1.0604, + "step": 344 + }, + { + "epoch": 0.01, + "grad_norm": 2.1874471203019654, + "learning_rate": 5.4160125588697024e-06, + "loss": 1.0079, + "step": 345 + }, + { + "epoch": 0.01, + "grad_norm": 1.9036921111841558, + "learning_rate": 5.4317111459968605e-06, + "loss": 1.0507, + "step": 346 + }, + { + "epoch": 0.01, + "grad_norm": 1.2034496616899248, + "learning_rate": 5.447409733124019e-06, + "loss": 0.9257, + "step": 347 + }, + { + "epoch": 0.01, + "grad_norm": 2.208728822931854, + "learning_rate": 5.463108320251178e-06, + "loss": 0.9462, + "step": 348 + }, + { + "epoch": 0.01, + "grad_norm": 1.927586290721344, + "learning_rate": 5.478806907378337e-06, + "loss": 1.1666, + "step": 349 + }, + { + "epoch": 0.01, + "grad_norm": 2.302975878536575, + "learning_rate": 5.494505494505495e-06, + "loss": 1.065, + "step": 350 + }, + { + "epoch": 0.01, + "grad_norm": 2.3988433849363062, + "learning_rate": 5.510204081632653e-06, + "loss": 1.0159, + "step": 351 + }, + { + "epoch": 0.01, + "grad_norm": 2.427951043511996, + "learning_rate": 5.525902668759811e-06, + "loss": 1.0395, + "step": 352 + }, + { + "epoch": 0.01, + "grad_norm": 2.3755770720602087, + "learning_rate": 5.54160125588697e-06, + "loss": 0.9112, + "step": 353 + }, + { + "epoch": 0.01, + "grad_norm": 2.005602728084362, + "learning_rate": 5.557299843014129e-06, + "loss": 1.0445, + "step": 354 + }, + { + "epoch": 0.01, + "grad_norm": 2.36655390537115, + "learning_rate": 5.572998430141288e-06, + "loss": 1.048, + "step": 355 + }, + { + "epoch": 0.01, + "grad_norm": 2.12173307680076, + "learning_rate": 5.588697017268446e-06, + "loss": 1.0216, + "step": 356 + }, + { + "epoch": 0.01, + "grad_norm": 1.9926251778560455, + "learning_rate": 5.604395604395605e-06, + "loss": 1.0387, + "step": 357 + }, + { + "epoch": 0.01, + "grad_norm": 2.0906402141950453, + "learning_rate": 5.620094191522764e-06, + "loss": 1.0377, + "step": 358 + }, + { + "epoch": 0.01, + "grad_norm": 2.0912768353525295, + "learning_rate": 5.635792778649923e-06, + "loss": 0.978, + "step": 359 + }, + { + "epoch": 0.01, + "grad_norm": 2.1626377823328946, + "learning_rate": 5.651491365777081e-06, + "loss": 0.9733, + "step": 360 + }, + { + "epoch": 0.01, + "grad_norm": 2.238607832582167, + "learning_rate": 5.6671899529042395e-06, + "loss": 1.1581, + "step": 361 + }, + { + "epoch": 0.01, + "grad_norm": 2.0053516179865047, + "learning_rate": 5.682888540031397e-06, + "loss": 1.1009, + "step": 362 + }, + { + "epoch": 0.01, + "grad_norm": 2.179029372415266, + "learning_rate": 5.698587127158556e-06, + "loss": 1.0565, + "step": 363 + }, + { + "epoch": 0.01, + "grad_norm": 1.1506914114328377, + "learning_rate": 5.7142857142857145e-06, + "loss": 0.9445, + "step": 364 + }, + { + "epoch": 0.01, + "grad_norm": 2.5202592168641305, + "learning_rate": 5.729984301412873e-06, + "loss": 1.1014, + "step": 365 + }, + { + "epoch": 0.01, + "grad_norm": 2.265320617272743, + "learning_rate": 5.7456828885400314e-06, + "loss": 1.0014, + "step": 366 + }, + { + "epoch": 0.01, + "grad_norm": 2.0351257426612137, + "learning_rate": 5.76138147566719e-06, + "loss": 0.9803, + "step": 367 + }, + { + "epoch": 0.01, + "grad_norm": 2.197621326212689, + "learning_rate": 5.777080062794349e-06, + "loss": 0.9836, + "step": 368 + }, + { + "epoch": 0.01, + "grad_norm": 1.1559932224666563, + "learning_rate": 5.792778649921508e-06, + "loss": 0.9717, + "step": 369 + }, + { + "epoch": 0.01, + "grad_norm": 2.467262677942051, + "learning_rate": 5.808477237048666e-06, + "loss": 1.0636, + "step": 370 + }, + { + "epoch": 0.01, + "grad_norm": 2.2326511547330288, + "learning_rate": 5.824175824175825e-06, + "loss": 1.1735, + "step": 371 + }, + { + "epoch": 0.01, + "grad_norm": 1.9017788751626192, + "learning_rate": 5.839874411302984e-06, + "loss": 1.0103, + "step": 372 + }, + { + "epoch": 0.01, + "grad_norm": 1.9230864534788807, + "learning_rate": 5.855572998430141e-06, + "loss": 1.0484, + "step": 373 + }, + { + "epoch": 0.01, + "grad_norm": 1.060661995507511, + "learning_rate": 5.8712715855573e-06, + "loss": 0.931, + "step": 374 + }, + { + "epoch": 0.01, + "grad_norm": 2.236223919806113, + "learning_rate": 5.886970172684459e-06, + "loss": 0.8707, + "step": 375 + }, + { + "epoch": 0.01, + "grad_norm": 1.9482795613750636, + "learning_rate": 5.902668759811617e-06, + "loss": 1.125, + "step": 376 + }, + { + "epoch": 0.01, + "grad_norm": 1.1040337318953692, + "learning_rate": 5.918367346938776e-06, + "loss": 0.9496, + "step": 377 + }, + { + "epoch": 0.01, + "grad_norm": 2.0845703342040656, + "learning_rate": 5.934065934065935e-06, + "loss": 0.9118, + "step": 378 + }, + { + "epoch": 0.01, + "grad_norm": 2.3236989688281295, + "learning_rate": 5.949764521193094e-06, + "loss": 1.109, + "step": 379 + }, + { + "epoch": 0.01, + "grad_norm": 2.0577510703959003, + "learning_rate": 5.965463108320252e-06, + "loss": 1.0668, + "step": 380 + }, + { + "epoch": 0.01, + "grad_norm": 2.3561771393770603, + "learning_rate": 5.9811616954474105e-06, + "loss": 1.0266, + "step": 381 + }, + { + "epoch": 0.01, + "grad_norm": 2.2833745377508285, + "learning_rate": 5.996860282574569e-06, + "loss": 0.8858, + "step": 382 + }, + { + "epoch": 0.01, + "grad_norm": 1.990538328926779, + "learning_rate": 6.012558869701728e-06, + "loss": 1.1291, + "step": 383 + }, + { + "epoch": 0.01, + "grad_norm": 2.0704955326000807, + "learning_rate": 6.0282574568288855e-06, + "loss": 1.0194, + "step": 384 + }, + { + "epoch": 0.01, + "grad_norm": 2.306248106978319, + "learning_rate": 6.043956043956044e-06, + "loss": 0.981, + "step": 385 + }, + { + "epoch": 0.01, + "grad_norm": 2.228358789578704, + "learning_rate": 6.059654631083202e-06, + "loss": 0.8984, + "step": 386 + }, + { + "epoch": 0.01, + "grad_norm": 2.3209435894346164, + "learning_rate": 6.075353218210361e-06, + "loss": 1.1242, + "step": 387 + }, + { + "epoch": 0.01, + "grad_norm": 1.8440660257872221, + "learning_rate": 6.09105180533752e-06, + "loss": 1.0493, + "step": 388 + }, + { + "epoch": 0.01, + "grad_norm": 2.4607375719398203, + "learning_rate": 6.106750392464679e-06, + "loss": 1.0193, + "step": 389 + }, + { + "epoch": 0.01, + "grad_norm": 2.149312197291877, + "learning_rate": 6.122448979591837e-06, + "loss": 0.9993, + "step": 390 + }, + { + "epoch": 0.01, + "grad_norm": 1.0886271046350913, + "learning_rate": 6.138147566718996e-06, + "loss": 1.0644, + "step": 391 + }, + { + "epoch": 0.01, + "grad_norm": 2.780342285769652, + "learning_rate": 6.153846153846155e-06, + "loss": 1.0035, + "step": 392 + }, + { + "epoch": 0.01, + "grad_norm": 2.1424614723288777, + "learning_rate": 6.169544740973314e-06, + "loss": 1.0596, + "step": 393 + }, + { + "epoch": 0.01, + "grad_norm": 2.59826170275896, + "learning_rate": 6.185243328100472e-06, + "loss": 1.0184, + "step": 394 + }, + { + "epoch": 0.01, + "grad_norm": 2.295159455034261, + "learning_rate": 6.20094191522763e-06, + "loss": 1.0726, + "step": 395 + }, + { + "epoch": 0.01, + "grad_norm": 2.113732601492172, + "learning_rate": 6.216640502354788e-06, + "loss": 0.9431, + "step": 396 + }, + { + "epoch": 0.01, + "grad_norm": 1.8392934920398916, + "learning_rate": 6.232339089481947e-06, + "loss": 0.8936, + "step": 397 + }, + { + "epoch": 0.01, + "grad_norm": 2.438471102641589, + "learning_rate": 6.248037676609106e-06, + "loss": 1.1077, + "step": 398 + }, + { + "epoch": 0.01, + "grad_norm": 2.2280834103372014, + "learning_rate": 6.2637362637362645e-06, + "loss": 1.0545, + "step": 399 + }, + { + "epoch": 0.01, + "grad_norm": 1.9711397088410376, + "learning_rate": 6.279434850863423e-06, + "loss": 0.9549, + "step": 400 + }, + { + "epoch": 0.01, + "grad_norm": 2.3003521148261736, + "learning_rate": 6.2951334379905815e-06, + "loss": 1.0045, + "step": 401 + }, + { + "epoch": 0.01, + "grad_norm": 2.1219795329860163, + "learning_rate": 6.31083202511774e-06, + "loss": 1.0888, + "step": 402 + }, + { + "epoch": 0.01, + "grad_norm": 2.2538963147515467, + "learning_rate": 6.326530612244899e-06, + "loss": 0.9717, + "step": 403 + }, + { + "epoch": 0.01, + "grad_norm": 2.4108355769817593, + "learning_rate": 6.342229199372057e-06, + "loss": 1.0679, + "step": 404 + }, + { + "epoch": 0.01, + "grad_norm": 2.592819416047953, + "learning_rate": 6.357927786499215e-06, + "loss": 0.9759, + "step": 405 + }, + { + "epoch": 0.01, + "grad_norm": 2.223928522656806, + "learning_rate": 6.373626373626373e-06, + "loss": 1.1307, + "step": 406 + }, + { + "epoch": 0.01, + "grad_norm": 1.9957201362167838, + "learning_rate": 6.389324960753532e-06, + "loss": 1.1108, + "step": 407 + }, + { + "epoch": 0.01, + "grad_norm": 2.616153477012478, + "learning_rate": 6.405023547880691e-06, + "loss": 1.0218, + "step": 408 + }, + { + "epoch": 0.01, + "grad_norm": 2.1245373237976417, + "learning_rate": 6.42072213500785e-06, + "loss": 1.1108, + "step": 409 + }, + { + "epoch": 0.01, + "grad_norm": 2.3660456245035792, + "learning_rate": 6.436420722135008e-06, + "loss": 1.1425, + "step": 410 + }, + { + "epoch": 0.01, + "grad_norm": 2.1213967596879835, + "learning_rate": 6.452119309262167e-06, + "loss": 0.9398, + "step": 411 + }, + { + "epoch": 0.01, + "grad_norm": 2.1141300352217653, + "learning_rate": 6.467817896389326e-06, + "loss": 1.0817, + "step": 412 + }, + { + "epoch": 0.01, + "grad_norm": 1.9671863044412448, + "learning_rate": 6.483516483516485e-06, + "loss": 1.1123, + "step": 413 + }, + { + "epoch": 0.01, + "grad_norm": 2.98388075004746, + "learning_rate": 6.499215070643643e-06, + "loss": 1.1022, + "step": 414 + }, + { + "epoch": 0.01, + "grad_norm": 1.114500262101427, + "learning_rate": 6.514913657770802e-06, + "loss": 0.9545, + "step": 415 + }, + { + "epoch": 0.01, + "grad_norm": 1.9802949717350953, + "learning_rate": 6.530612244897959e-06, + "loss": 1.0366, + "step": 416 + }, + { + "epoch": 0.01, + "grad_norm": 2.0223044544892645, + "learning_rate": 6.546310832025118e-06, + "loss": 1.2307, + "step": 417 + }, + { + "epoch": 0.01, + "grad_norm": 2.0611969239397103, + "learning_rate": 6.562009419152277e-06, + "loss": 1.0416, + "step": 418 + }, + { + "epoch": 0.01, + "grad_norm": 1.8917341198777906, + "learning_rate": 6.5777080062794355e-06, + "loss": 1.0785, + "step": 419 + }, + { + "epoch": 0.01, + "grad_norm": 2.0588650124327565, + "learning_rate": 6.5934065934065935e-06, + "loss": 1.1738, + "step": 420 + }, + { + "epoch": 0.01, + "grad_norm": 1.1980040045551794, + "learning_rate": 6.609105180533752e-06, + "loss": 0.9561, + "step": 421 + }, + { + "epoch": 0.01, + "grad_norm": 2.4359288262203034, + "learning_rate": 6.624803767660911e-06, + "loss": 1.1026, + "step": 422 + }, + { + "epoch": 0.01, + "grad_norm": 2.416275000565128, + "learning_rate": 6.64050235478807e-06, + "loss": 0.8955, + "step": 423 + }, + { + "epoch": 0.01, + "grad_norm": 2.319396813748169, + "learning_rate": 6.656200941915228e-06, + "loss": 1.0056, + "step": 424 + }, + { + "epoch": 0.01, + "grad_norm": 2.091757050648472, + "learning_rate": 6.671899529042387e-06, + "loss": 1.0575, + "step": 425 + }, + { + "epoch": 0.01, + "grad_norm": 2.5060270157485074, + "learning_rate": 6.687598116169546e-06, + "loss": 0.9926, + "step": 426 + }, + { + "epoch": 0.01, + "grad_norm": 2.5337837496103197, + "learning_rate": 6.703296703296703e-06, + "loss": 0.9976, + "step": 427 + }, + { + "epoch": 0.01, + "grad_norm": 2.9460559445496854, + "learning_rate": 6.718995290423862e-06, + "loss": 1.1333, + "step": 428 + }, + { + "epoch": 0.01, + "grad_norm": 2.074797018761096, + "learning_rate": 6.734693877551021e-06, + "loss": 0.9395, + "step": 429 + }, + { + "epoch": 0.01, + "grad_norm": 2.2118524249122418, + "learning_rate": 6.750392464678179e-06, + "loss": 0.9958, + "step": 430 + }, + { + "epoch": 0.01, + "grad_norm": 2.253736583899509, + "learning_rate": 6.766091051805338e-06, + "loss": 1.0467, + "step": 431 + }, + { + "epoch": 0.01, + "grad_norm": 2.561411389860909, + "learning_rate": 6.781789638932497e-06, + "loss": 1.0141, + "step": 432 + }, + { + "epoch": 0.01, + "grad_norm": 2.026674309746684, + "learning_rate": 6.797488226059656e-06, + "loss": 1.0024, + "step": 433 + }, + { + "epoch": 0.01, + "grad_norm": 2.4088866875575388, + "learning_rate": 6.813186813186814e-06, + "loss": 0.9114, + "step": 434 + }, + { + "epoch": 0.01, + "grad_norm": 2.145589759752312, + "learning_rate": 6.828885400313973e-06, + "loss": 0.9613, + "step": 435 + }, + { + "epoch": 0.01, + "grad_norm": 1.0611930749179133, + "learning_rate": 6.8445839874411315e-06, + "loss": 0.9033, + "step": 436 + }, + { + "epoch": 0.01, + "grad_norm": 1.136239877997087, + "learning_rate": 6.8602825745682895e-06, + "loss": 0.9174, + "step": 437 + }, + { + "epoch": 0.01, + "grad_norm": 2.2197552517372414, + "learning_rate": 6.8759811616954476e-06, + "loss": 1.2024, + "step": 438 + }, + { + "epoch": 0.01, + "grad_norm": 2.5233124086174517, + "learning_rate": 6.8916797488226065e-06, + "loss": 1.0505, + "step": 439 + }, + { + "epoch": 0.01, + "grad_norm": 2.3770217000117126, + "learning_rate": 6.9073783359497645e-06, + "loss": 1.1727, + "step": 440 + }, + { + "epoch": 0.01, + "grad_norm": 2.045028360792787, + "learning_rate": 6.923076923076923e-06, + "loss": 1.04, + "step": 441 + }, + { + "epoch": 0.01, + "grad_norm": 2.1119573498698343, + "learning_rate": 6.938775510204082e-06, + "loss": 1.2327, + "step": 442 + }, + { + "epoch": 0.01, + "grad_norm": 2.55337975872774, + "learning_rate": 6.954474097331241e-06, + "loss": 0.9934, + "step": 443 + }, + { + "epoch": 0.01, + "grad_norm": 2.3025253695882415, + "learning_rate": 6.970172684458399e-06, + "loss": 0.9927, + "step": 444 + }, + { + "epoch": 0.01, + "grad_norm": 2.3587688175941013, + "learning_rate": 6.985871271585558e-06, + "loss": 1.1588, + "step": 445 + }, + { + "epoch": 0.01, + "grad_norm": 1.9902128286524354, + "learning_rate": 7.001569858712717e-06, + "loss": 0.9408, + "step": 446 + }, + { + "epoch": 0.01, + "grad_norm": 1.917413549370171, + "learning_rate": 7.017268445839875e-06, + "loss": 1.0764, + "step": 447 + }, + { + "epoch": 0.01, + "grad_norm": 2.3441250330166947, + "learning_rate": 7.032967032967034e-06, + "loss": 1.0939, + "step": 448 + }, + { + "epoch": 0.01, + "grad_norm": 3.6297050133111024, + "learning_rate": 7.048665620094192e-06, + "loss": 0.9898, + "step": 449 + }, + { + "epoch": 0.01, + "grad_norm": 2.3617132604827327, + "learning_rate": 7.06436420722135e-06, + "loss": 1.1409, + "step": 450 + }, + { + "epoch": 0.01, + "grad_norm": 2.0831689787894674, + "learning_rate": 7.080062794348509e-06, + "loss": 1.0609, + "step": 451 + }, + { + "epoch": 0.01, + "grad_norm": 2.202398523014605, + "learning_rate": 7.095761381475668e-06, + "loss": 1.1286, + "step": 452 + }, + { + "epoch": 0.01, + "grad_norm": 2.021791741078468, + "learning_rate": 7.111459968602827e-06, + "loss": 1.1358, + "step": 453 + }, + { + "epoch": 0.01, + "grad_norm": 2.1253488674795964, + "learning_rate": 7.127158555729985e-06, + "loss": 1.1508, + "step": 454 + }, + { + "epoch": 0.01, + "grad_norm": 2.152528673958084, + "learning_rate": 7.1428571428571436e-06, + "loss": 1.1322, + "step": 455 + }, + { + "epoch": 0.01, + "grad_norm": 2.297649734386336, + "learning_rate": 7.1585557299843024e-06, + "loss": 1.1153, + "step": 456 + }, + { + "epoch": 0.01, + "grad_norm": 2.1255164090338803, + "learning_rate": 7.174254317111461e-06, + "loss": 0.9817, + "step": 457 + }, + { + "epoch": 0.01, + "grad_norm": 1.1695024920019501, + "learning_rate": 7.189952904238619e-06, + "loss": 0.9284, + "step": 458 + }, + { + "epoch": 0.01, + "grad_norm": 2.0312467273584924, + "learning_rate": 7.205651491365777e-06, + "loss": 0.998, + "step": 459 + }, + { + "epoch": 0.01, + "grad_norm": 2.217196726132943, + "learning_rate": 7.2213500784929355e-06, + "loss": 1.1487, + "step": 460 + }, + { + "epoch": 0.01, + "grad_norm": 1.1329377117510064, + "learning_rate": 7.237048665620094e-06, + "loss": 0.9704, + "step": 461 + }, + { + "epoch": 0.01, + "grad_norm": 2.0756243723324843, + "learning_rate": 7.252747252747253e-06, + "loss": 1.0153, + "step": 462 + }, + { + "epoch": 0.01, + "grad_norm": 1.9315122451659887, + "learning_rate": 7.268445839874412e-06, + "loss": 1.0242, + "step": 463 + }, + { + "epoch": 0.01, + "grad_norm": 2.225152155443336, + "learning_rate": 7.28414442700157e-06, + "loss": 1.0143, + "step": 464 + }, + { + "epoch": 0.01, + "grad_norm": 2.136232281711354, + "learning_rate": 7.299843014128729e-06, + "loss": 1.0761, + "step": 465 + }, + { + "epoch": 0.01, + "grad_norm": 2.3660893882814262, + "learning_rate": 7.315541601255888e-06, + "loss": 0.869, + "step": 466 + }, + { + "epoch": 0.01, + "grad_norm": 2.110229749045679, + "learning_rate": 7.331240188383047e-06, + "loss": 1.023, + "step": 467 + }, + { + "epoch": 0.01, + "grad_norm": 1.9336061828868616, + "learning_rate": 7.346938775510205e-06, + "loss": 1.008, + "step": 468 + }, + { + "epoch": 0.01, + "grad_norm": 2.341194001059622, + "learning_rate": 7.362637362637364e-06, + "loss": 1.1043, + "step": 469 + }, + { + "epoch": 0.01, + "grad_norm": 2.2421057008939944, + "learning_rate": 7.378335949764521e-06, + "loss": 1.0283, + "step": 470 + }, + { + "epoch": 0.01, + "grad_norm": 2.235235754940411, + "learning_rate": 7.39403453689168e-06, + "loss": 1.1132, + "step": 471 + }, + { + "epoch": 0.01, + "grad_norm": 2.15364513045007, + "learning_rate": 7.409733124018839e-06, + "loss": 1.1257, + "step": 472 + }, + { + "epoch": 0.01, + "grad_norm": 1.8753273199526148, + "learning_rate": 7.425431711145998e-06, + "loss": 0.9715, + "step": 473 + }, + { + "epoch": 0.01, + "grad_norm": 2.901806670559612, + "learning_rate": 7.441130298273156e-06, + "loss": 1.1408, + "step": 474 + }, + { + "epoch": 0.01, + "grad_norm": 1.9130845189150136, + "learning_rate": 7.4568288854003145e-06, + "loss": 0.8827, + "step": 475 + }, + { + "epoch": 0.01, + "grad_norm": 1.0890963504610516, + "learning_rate": 7.472527472527473e-06, + "loss": 0.9396, + "step": 476 + }, + { + "epoch": 0.01, + "grad_norm": 1.8717283180037085, + "learning_rate": 7.488226059654632e-06, + "loss": 1.0552, + "step": 477 + }, + { + "epoch": 0.01, + "grad_norm": 2.213233394497612, + "learning_rate": 7.50392464678179e-06, + "loss": 1.1561, + "step": 478 + }, + { + "epoch": 0.01, + "grad_norm": 2.1626712060274014, + "learning_rate": 7.519623233908949e-06, + "loss": 1.0005, + "step": 479 + }, + { + "epoch": 0.01, + "grad_norm": 2.1099475110153056, + "learning_rate": 7.535321821036108e-06, + "loss": 1.093, + "step": 480 + }, + { + "epoch": 0.01, + "grad_norm": 2.199843088106302, + "learning_rate": 7.551020408163265e-06, + "loss": 1.0491, + "step": 481 + }, + { + "epoch": 0.01, + "grad_norm": 2.22441308620407, + "learning_rate": 7.566718995290424e-06, + "loss": 0.9206, + "step": 482 + }, + { + "epoch": 0.01, + "grad_norm": 2.2078541457985725, + "learning_rate": 7.582417582417583e-06, + "loss": 1.1393, + "step": 483 + }, + { + "epoch": 0.01, + "grad_norm": 2.938231020880739, + "learning_rate": 7.598116169544741e-06, + "loss": 1.0015, + "step": 484 + }, + { + "epoch": 0.01, + "grad_norm": 1.0954890116331226, + "learning_rate": 7.6138147566719e-06, + "loss": 0.9341, + "step": 485 + }, + { + "epoch": 0.01, + "grad_norm": 2.961466317713113, + "learning_rate": 7.629513343799058e-06, + "loss": 1.281, + "step": 486 + }, + { + "epoch": 0.01, + "grad_norm": 2.523742542009909, + "learning_rate": 7.645211930926217e-06, + "loss": 0.9785, + "step": 487 + }, + { + "epoch": 0.01, + "grad_norm": 2.2579953102537482, + "learning_rate": 7.660910518053376e-06, + "loss": 1.0534, + "step": 488 + }, + { + "epoch": 0.01, + "grad_norm": 2.2763571766592703, + "learning_rate": 7.676609105180535e-06, + "loss": 1.0848, + "step": 489 + }, + { + "epoch": 0.01, + "grad_norm": 1.1018785952531645, + "learning_rate": 7.692307692307694e-06, + "loss": 0.8975, + "step": 490 + }, + { + "epoch": 0.01, + "grad_norm": 3.741058668882971, + "learning_rate": 7.708006279434852e-06, + "loss": 1.0915, + "step": 491 + }, + { + "epoch": 0.01, + "grad_norm": 2.3410630329066904, + "learning_rate": 7.72370486656201e-06, + "loss": 1.0937, + "step": 492 + }, + { + "epoch": 0.01, + "grad_norm": 2.0678606409126203, + "learning_rate": 7.739403453689169e-06, + "loss": 1.1495, + "step": 493 + }, + { + "epoch": 0.01, + "grad_norm": 3.0697338708482445, + "learning_rate": 7.755102040816327e-06, + "loss": 1.0935, + "step": 494 + }, + { + "epoch": 0.01, + "grad_norm": 1.1118510322137853, + "learning_rate": 7.770800627943486e-06, + "loss": 0.9274, + "step": 495 + }, + { + "epoch": 0.01, + "grad_norm": 2.140059448956104, + "learning_rate": 7.786499215070644e-06, + "loss": 1.1132, + "step": 496 + }, + { + "epoch": 0.01, + "grad_norm": 2.330366658966381, + "learning_rate": 7.802197802197802e-06, + "loss": 0.8192, + "step": 497 + }, + { + "epoch": 0.01, + "grad_norm": 2.32055413438159, + "learning_rate": 7.817896389324961e-06, + "loss": 1.1372, + "step": 498 + }, + { + "epoch": 0.01, + "grad_norm": 2.218318991159986, + "learning_rate": 7.83359497645212e-06, + "loss": 1.0882, + "step": 499 + }, + { + "epoch": 0.01, + "grad_norm": 1.9963537608361763, + "learning_rate": 7.849293563579279e-06, + "loss": 1.1023, + "step": 500 + }, + { + "epoch": 0.01, + "grad_norm": 2.1216104010604586, + "learning_rate": 7.864992150706438e-06, + "loss": 1.2294, + "step": 501 + }, + { + "epoch": 0.01, + "grad_norm": 2.425329813476702, + "learning_rate": 7.880690737833597e-06, + "loss": 0.9573, + "step": 502 + }, + { + "epoch": 0.01, + "grad_norm": 2.1605978668691037, + "learning_rate": 7.896389324960754e-06, + "loss": 1.1317, + "step": 503 + }, + { + "epoch": 0.01, + "grad_norm": 2.1482461699991555, + "learning_rate": 7.912087912087913e-06, + "loss": 1.1026, + "step": 504 + }, + { + "epoch": 0.01, + "grad_norm": 1.1221098077411742, + "learning_rate": 7.927786499215072e-06, + "loss": 0.9087, + "step": 505 + }, + { + "epoch": 0.01, + "grad_norm": 2.390635561296409, + "learning_rate": 7.943485086342229e-06, + "loss": 1.0577, + "step": 506 + }, + { + "epoch": 0.01, + "grad_norm": 2.349603121607284, + "learning_rate": 7.959183673469388e-06, + "loss": 1.0807, + "step": 507 + }, + { + "epoch": 0.01, + "grad_norm": 2.0015647079321033, + "learning_rate": 7.974882260596547e-06, + "loss": 1.0748, + "step": 508 + }, + { + "epoch": 0.01, + "grad_norm": 2.299453173944459, + "learning_rate": 7.990580847723706e-06, + "loss": 1.0518, + "step": 509 + }, + { + "epoch": 0.01, + "grad_norm": 1.974880149827705, + "learning_rate": 8.006279434850865e-06, + "loss": 1.0749, + "step": 510 + }, + { + "epoch": 0.01, + "grad_norm": 2.0886923120229, + "learning_rate": 8.021978021978023e-06, + "loss": 1.0731, + "step": 511 + }, + { + "epoch": 0.01, + "grad_norm": 1.2968650873685794, + "learning_rate": 8.037676609105182e-06, + "loss": 0.9405, + "step": 512 + }, + { + "epoch": 0.01, + "grad_norm": 1.9818641018861085, + "learning_rate": 8.05337519623234e-06, + "loss": 1.0441, + "step": 513 + }, + { + "epoch": 0.01, + "grad_norm": 2.0908744291422754, + "learning_rate": 8.069073783359498e-06, + "loss": 1.1332, + "step": 514 + }, + { + "epoch": 0.01, + "grad_norm": 1.1194528732671234, + "learning_rate": 8.084772370486657e-06, + "loss": 0.987, + "step": 515 + }, + { + "epoch": 0.01, + "grad_norm": 3.362185764176234, + "learning_rate": 8.100470957613814e-06, + "loss": 0.9942, + "step": 516 + }, + { + "epoch": 0.01, + "grad_norm": 2.224449750441655, + "learning_rate": 8.116169544740973e-06, + "loss": 1.3296, + "step": 517 + }, + { + "epoch": 0.01, + "grad_norm": 1.930925950742801, + "learning_rate": 8.131868131868132e-06, + "loss": 1.1336, + "step": 518 + }, + { + "epoch": 0.01, + "grad_norm": 2.408868255015578, + "learning_rate": 8.147566718995291e-06, + "loss": 0.9612, + "step": 519 + }, + { + "epoch": 0.01, + "grad_norm": 2.4985997827374895, + "learning_rate": 8.16326530612245e-06, + "loss": 1.0, + "step": 520 + }, + { + "epoch": 0.01, + "grad_norm": 2.0892835575382445, + "learning_rate": 8.178963893249609e-06, + "loss": 1.1121, + "step": 521 + }, + { + "epoch": 0.01, + "grad_norm": 2.092569788531129, + "learning_rate": 8.194662480376768e-06, + "loss": 0.9633, + "step": 522 + }, + { + "epoch": 0.01, + "grad_norm": 2.0427791501096744, + "learning_rate": 8.210361067503925e-06, + "loss": 1.031, + "step": 523 + }, + { + "epoch": 0.01, + "grad_norm": 1.1717910322266778, + "learning_rate": 8.226059654631084e-06, + "loss": 0.8943, + "step": 524 + }, + { + "epoch": 0.01, + "grad_norm": 1.1048813292673954, + "learning_rate": 8.241758241758243e-06, + "loss": 0.8912, + "step": 525 + }, + { + "epoch": 0.01, + "grad_norm": 1.0907886256343042, + "learning_rate": 8.2574568288854e-06, + "loss": 0.884, + "step": 526 + }, + { + "epoch": 0.01, + "grad_norm": 2.1373088672744873, + "learning_rate": 8.273155416012559e-06, + "loss": 1.084, + "step": 527 + }, + { + "epoch": 0.01, + "grad_norm": 2.077756790721361, + "learning_rate": 8.288854003139718e-06, + "loss": 0.9581, + "step": 528 + }, + { + "epoch": 0.01, + "grad_norm": 2.508902219122578, + "learning_rate": 8.304552590266877e-06, + "loss": 1.0186, + "step": 529 + }, + { + "epoch": 0.01, + "grad_norm": 2.097764089567714, + "learning_rate": 8.320251177394036e-06, + "loss": 0.9757, + "step": 530 + }, + { + "epoch": 0.01, + "grad_norm": 2.390492819929443, + "learning_rate": 8.335949764521194e-06, + "loss": 1.0833, + "step": 531 + }, + { + "epoch": 0.01, + "grad_norm": 2.355660785965016, + "learning_rate": 8.351648351648353e-06, + "loss": 1.1341, + "step": 532 + }, + { + "epoch": 0.01, + "grad_norm": 2.3580926487604557, + "learning_rate": 8.36734693877551e-06, + "loss": 1.215, + "step": 533 + }, + { + "epoch": 0.01, + "grad_norm": 3.5283143540761532, + "learning_rate": 8.38304552590267e-06, + "loss": 1.1403, + "step": 534 + }, + { + "epoch": 0.01, + "grad_norm": 2.028861693319721, + "learning_rate": 8.398744113029828e-06, + "loss": 1.0403, + "step": 535 + }, + { + "epoch": 0.01, + "grad_norm": 2.71525179897585, + "learning_rate": 8.414442700156985e-06, + "loss": 1.1117, + "step": 536 + }, + { + "epoch": 0.01, + "grad_norm": 2.131746943641586, + "learning_rate": 8.430141287284144e-06, + "loss": 1.0728, + "step": 537 + }, + { + "epoch": 0.01, + "grad_norm": 2.075131521703083, + "learning_rate": 8.445839874411303e-06, + "loss": 1.07, + "step": 538 + }, + { + "epoch": 0.01, + "grad_norm": 2.1960865603110107, + "learning_rate": 8.461538461538462e-06, + "loss": 1.0703, + "step": 539 + }, + { + "epoch": 0.01, + "grad_norm": 2.260050431015787, + "learning_rate": 8.477237048665621e-06, + "loss": 1.0041, + "step": 540 + }, + { + "epoch": 0.01, + "grad_norm": 2.3600117162949323, + "learning_rate": 8.49293563579278e-06, + "loss": 1.0951, + "step": 541 + }, + { + "epoch": 0.01, + "grad_norm": 2.2199300753673397, + "learning_rate": 8.508634222919939e-06, + "loss": 1.1057, + "step": 542 + }, + { + "epoch": 0.01, + "grad_norm": 2.3622998327841676, + "learning_rate": 8.524332810047096e-06, + "loss": 1.0082, + "step": 543 + }, + { + "epoch": 0.01, + "grad_norm": 2.333807517677162, + "learning_rate": 8.540031397174255e-06, + "loss": 1.17, + "step": 544 + }, + { + "epoch": 0.01, + "grad_norm": 2.44755507315464, + "learning_rate": 8.555729984301414e-06, + "loss": 1.1194, + "step": 545 + }, + { + "epoch": 0.01, + "grad_norm": 2.8153571950793594, + "learning_rate": 8.571428571428571e-06, + "loss": 1.1131, + "step": 546 + }, + { + "epoch": 0.01, + "grad_norm": 2.1112078784406494, + "learning_rate": 8.58712715855573e-06, + "loss": 1.0338, + "step": 547 + }, + { + "epoch": 0.01, + "grad_norm": 1.992906435217898, + "learning_rate": 8.602825745682889e-06, + "loss": 1.1793, + "step": 548 + }, + { + "epoch": 0.01, + "grad_norm": 1.8938808757795667, + "learning_rate": 8.618524332810048e-06, + "loss": 1.0676, + "step": 549 + }, + { + "epoch": 0.01, + "grad_norm": 2.4608126804509216, + "learning_rate": 8.634222919937206e-06, + "loss": 1.1187, + "step": 550 + }, + { + "epoch": 0.01, + "grad_norm": 2.237632609212455, + "learning_rate": 8.649921507064365e-06, + "loss": 1.0076, + "step": 551 + }, + { + "epoch": 0.01, + "grad_norm": 2.26692471620718, + "learning_rate": 8.665620094191524e-06, + "loss": 1.0056, + "step": 552 + }, + { + "epoch": 0.01, + "grad_norm": 1.1283633639000277, + "learning_rate": 8.681318681318681e-06, + "loss": 0.958, + "step": 553 + }, + { + "epoch": 0.01, + "grad_norm": 2.3678640981824075, + "learning_rate": 8.69701726844584e-06, + "loss": 1.1305, + "step": 554 + }, + { + "epoch": 0.01, + "grad_norm": 1.12450167272141, + "learning_rate": 8.712715855573e-06, + "loss": 0.9848, + "step": 555 + }, + { + "epoch": 0.01, + "grad_norm": 2.282799165463684, + "learning_rate": 8.728414442700158e-06, + "loss": 1.0662, + "step": 556 + }, + { + "epoch": 0.01, + "grad_norm": 2.1973291399570813, + "learning_rate": 8.744113029827315e-06, + "loss": 1.1774, + "step": 557 + }, + { + "epoch": 0.01, + "grad_norm": 2.2931268649077103, + "learning_rate": 8.759811616954474e-06, + "loss": 1.1704, + "step": 558 + }, + { + "epoch": 0.01, + "grad_norm": 1.113566745888374, + "learning_rate": 8.775510204081633e-06, + "loss": 0.9589, + "step": 559 + }, + { + "epoch": 0.01, + "grad_norm": 2.318470130940604, + "learning_rate": 8.791208791208792e-06, + "loss": 0.9662, + "step": 560 + }, + { + "epoch": 0.01, + "grad_norm": 2.1299089447540838, + "learning_rate": 8.80690737833595e-06, + "loss": 0.9172, + "step": 561 + }, + { + "epoch": 0.01, + "grad_norm": 2.3846388324449626, + "learning_rate": 8.82260596546311e-06, + "loss": 1.0947, + "step": 562 + }, + { + "epoch": 0.01, + "grad_norm": 2.337790048637974, + "learning_rate": 8.838304552590269e-06, + "loss": 1.0888, + "step": 563 + }, + { + "epoch": 0.01, + "grad_norm": 1.996725342319174, + "learning_rate": 8.854003139717426e-06, + "loss": 1.1314, + "step": 564 + }, + { + "epoch": 0.01, + "grad_norm": 2.2211430779271013, + "learning_rate": 8.869701726844585e-06, + "loss": 1.0073, + "step": 565 + }, + { + "epoch": 0.01, + "grad_norm": 1.0604913856855327, + "learning_rate": 8.885400313971744e-06, + "loss": 0.9565, + "step": 566 + }, + { + "epoch": 0.01, + "grad_norm": 2.539921275307718, + "learning_rate": 8.9010989010989e-06, + "loss": 1.0332, + "step": 567 + }, + { + "epoch": 0.01, + "grad_norm": 2.097609845772326, + "learning_rate": 8.91679748822606e-06, + "loss": 0.9778, + "step": 568 + }, + { + "epoch": 0.01, + "grad_norm": 2.407798933882745, + "learning_rate": 8.932496075353219e-06, + "loss": 1.1149, + "step": 569 + }, + { + "epoch": 0.01, + "grad_norm": 2.20971738966817, + "learning_rate": 8.948194662480377e-06, + "loss": 0.9982, + "step": 570 + }, + { + "epoch": 0.01, + "grad_norm": 1.1639751420148376, + "learning_rate": 8.963893249607536e-06, + "loss": 1.0252, + "step": 571 + }, + { + "epoch": 0.01, + "grad_norm": 2.5925227424775104, + "learning_rate": 8.979591836734695e-06, + "loss": 1.1294, + "step": 572 + }, + { + "epoch": 0.01, + "grad_norm": 2.30122612075121, + "learning_rate": 8.995290423861854e-06, + "loss": 0.9815, + "step": 573 + }, + { + "epoch": 0.01, + "grad_norm": 3.5875602672869547, + "learning_rate": 9.010989010989011e-06, + "loss": 1.0604, + "step": 574 + }, + { + "epoch": 0.01, + "grad_norm": 2.4254274740912627, + "learning_rate": 9.02668759811617e-06, + "loss": 1.0992, + "step": 575 + }, + { + "epoch": 0.01, + "grad_norm": 2.918798575058367, + "learning_rate": 9.042386185243329e-06, + "loss": 0.9043, + "step": 576 + }, + { + "epoch": 0.01, + "grad_norm": 2.2298360343050097, + "learning_rate": 9.058084772370488e-06, + "loss": 1.1319, + "step": 577 + }, + { + "epoch": 0.01, + "grad_norm": 2.172663040340845, + "learning_rate": 9.073783359497645e-06, + "loss": 1.0715, + "step": 578 + }, + { + "epoch": 0.01, + "grad_norm": 2.3018497311928146, + "learning_rate": 9.089481946624804e-06, + "loss": 1.0839, + "step": 579 + }, + { + "epoch": 0.01, + "grad_norm": 1.1856235374491055, + "learning_rate": 9.105180533751963e-06, + "loss": 1.026, + "step": 580 + }, + { + "epoch": 0.01, + "grad_norm": 2.1282265188265868, + "learning_rate": 9.120879120879122e-06, + "loss": 1.1014, + "step": 581 + }, + { + "epoch": 0.01, + "grad_norm": 3.1665667328377545, + "learning_rate": 9.13657770800628e-06, + "loss": 1.0389, + "step": 582 + }, + { + "epoch": 0.01, + "grad_norm": 2.2450066042899652, + "learning_rate": 9.15227629513344e-06, + "loss": 1.0974, + "step": 583 + }, + { + "epoch": 0.01, + "grad_norm": 2.0586400948606287, + "learning_rate": 9.167974882260597e-06, + "loss": 1.0569, + "step": 584 + }, + { + "epoch": 0.01, + "grad_norm": 2.299194771888394, + "learning_rate": 9.183673469387756e-06, + "loss": 0.9995, + "step": 585 + }, + { + "epoch": 0.01, + "grad_norm": 2.345294910122874, + "learning_rate": 9.199372056514915e-06, + "loss": 1.0601, + "step": 586 + }, + { + "epoch": 0.01, + "grad_norm": 2.011327013165279, + "learning_rate": 9.215070643642073e-06, + "loss": 1.0596, + "step": 587 + }, + { + "epoch": 0.01, + "grad_norm": 2.5403693925111637, + "learning_rate": 9.230769230769232e-06, + "loss": 1.0665, + "step": 588 + }, + { + "epoch": 0.01, + "grad_norm": 2.480594370976335, + "learning_rate": 9.24646781789639e-06, + "loss": 1.0927, + "step": 589 + }, + { + "epoch": 0.01, + "grad_norm": 2.774226977919718, + "learning_rate": 9.262166405023548e-06, + "loss": 1.0315, + "step": 590 + }, + { + "epoch": 0.01, + "grad_norm": 2.830563198599085, + "learning_rate": 9.277864992150707e-06, + "loss": 0.8911, + "step": 591 + }, + { + "epoch": 0.01, + "grad_norm": 2.1381091732995556, + "learning_rate": 9.293563579277866e-06, + "loss": 1.2075, + "step": 592 + }, + { + "epoch": 0.01, + "grad_norm": 2.1113623180496637, + "learning_rate": 9.309262166405025e-06, + "loss": 0.9986, + "step": 593 + }, + { + "epoch": 0.01, + "grad_norm": 2.077510960994402, + "learning_rate": 9.324960753532182e-06, + "loss": 0.9184, + "step": 594 + }, + { + "epoch": 0.01, + "grad_norm": 2.2297414674974307, + "learning_rate": 9.340659340659341e-06, + "loss": 1.174, + "step": 595 + }, + { + "epoch": 0.01, + "grad_norm": 1.9552127180158714, + "learning_rate": 9.3563579277865e-06, + "loss": 0.9552, + "step": 596 + }, + { + "epoch": 0.01, + "grad_norm": 2.8339980142023307, + "learning_rate": 9.372056514913659e-06, + "loss": 1.075, + "step": 597 + }, + { + "epoch": 0.01, + "grad_norm": 2.5737401856796436, + "learning_rate": 9.387755102040818e-06, + "loss": 0.9247, + "step": 598 + }, + { + "epoch": 0.01, + "grad_norm": 2.466581483103445, + "learning_rate": 9.403453689167977e-06, + "loss": 1.0391, + "step": 599 + }, + { + "epoch": 0.01, + "grad_norm": 2.5685050746254503, + "learning_rate": 9.419152276295134e-06, + "loss": 1.0425, + "step": 600 + }, + { + "epoch": 0.01, + "grad_norm": 2.0146736956116493, + "learning_rate": 9.434850863422293e-06, + "loss": 1.0301, + "step": 601 + }, + { + "epoch": 0.01, + "grad_norm": 2.2636992556345303, + "learning_rate": 9.450549450549452e-06, + "loss": 1.0586, + "step": 602 + }, + { + "epoch": 0.01, + "grad_norm": 2.5637262339033278, + "learning_rate": 9.46624803767661e-06, + "loss": 0.9005, + "step": 603 + }, + { + "epoch": 0.01, + "grad_norm": 2.468052589205153, + "learning_rate": 9.481946624803768e-06, + "loss": 0.9693, + "step": 604 + }, + { + "epoch": 0.01, + "grad_norm": 2.120393348061165, + "learning_rate": 9.497645211930927e-06, + "loss": 1.0771, + "step": 605 + }, + { + "epoch": 0.01, + "grad_norm": 1.9131434070154145, + "learning_rate": 9.513343799058085e-06, + "loss": 1.031, + "step": 606 + }, + { + "epoch": 0.01, + "grad_norm": 1.8227546125002376, + "learning_rate": 9.529042386185244e-06, + "loss": 1.0855, + "step": 607 + }, + { + "epoch": 0.01, + "grad_norm": 2.3077915502524213, + "learning_rate": 9.544740973312403e-06, + "loss": 0.9979, + "step": 608 + }, + { + "epoch": 0.01, + "grad_norm": 2.2691678190760673, + "learning_rate": 9.560439560439562e-06, + "loss": 1.0949, + "step": 609 + }, + { + "epoch": 0.01, + "grad_norm": 1.9924551486575468, + "learning_rate": 9.576138147566721e-06, + "loss": 1.1278, + "step": 610 + }, + { + "epoch": 0.01, + "grad_norm": 2.2955295855855113, + "learning_rate": 9.591836734693878e-06, + "loss": 1.0543, + "step": 611 + }, + { + "epoch": 0.01, + "grad_norm": 2.1264571298137422, + "learning_rate": 9.607535321821037e-06, + "loss": 1.0184, + "step": 612 + }, + { + "epoch": 0.01, + "grad_norm": 2.1095867049641024, + "learning_rate": 9.623233908948196e-06, + "loss": 1.0935, + "step": 613 + }, + { + "epoch": 0.01, + "grad_norm": 2.7634618983518977, + "learning_rate": 9.638932496075353e-06, + "loss": 0.9637, + "step": 614 + }, + { + "epoch": 0.01, + "grad_norm": 2.1222039144233955, + "learning_rate": 9.654631083202512e-06, + "loss": 1.0634, + "step": 615 + }, + { + "epoch": 0.01, + "grad_norm": 2.6017868772001114, + "learning_rate": 9.670329670329671e-06, + "loss": 1.3079, + "step": 616 + }, + { + "epoch": 0.01, + "grad_norm": 2.2252200847200427, + "learning_rate": 9.68602825745683e-06, + "loss": 0.9874, + "step": 617 + }, + { + "epoch": 0.01, + "grad_norm": 2.1037013760853136, + "learning_rate": 9.701726844583989e-06, + "loss": 1.1931, + "step": 618 + }, + { + "epoch": 0.01, + "grad_norm": 2.159981435545174, + "learning_rate": 9.717425431711148e-06, + "loss": 0.9743, + "step": 619 + }, + { + "epoch": 0.01, + "grad_norm": 1.9537117069104248, + "learning_rate": 9.733124018838307e-06, + "loss": 1.0394, + "step": 620 + }, + { + "epoch": 0.01, + "grad_norm": 1.1277736628153205, + "learning_rate": 9.748822605965464e-06, + "loss": 0.934, + "step": 621 + }, + { + "epoch": 0.01, + "grad_norm": 2.474219608491177, + "learning_rate": 9.764521193092623e-06, + "loss": 1.0829, + "step": 622 + }, + { + "epoch": 0.01, + "grad_norm": 2.0028464054962862, + "learning_rate": 9.780219780219781e-06, + "loss": 1.0616, + "step": 623 + }, + { + "epoch": 0.01, + "grad_norm": 2.0458144565589365, + "learning_rate": 9.795918367346939e-06, + "loss": 1.0105, + "step": 624 + }, + { + "epoch": 0.01, + "grad_norm": 2.037709173960045, + "learning_rate": 9.811616954474098e-06, + "loss": 1.0611, + "step": 625 + }, + { + "epoch": 0.01, + "grad_norm": 2.2617006081366586, + "learning_rate": 9.827315541601256e-06, + "loss": 1.1528, + "step": 626 + }, + { + "epoch": 0.01, + "grad_norm": 2.3322794101696025, + "learning_rate": 9.843014128728415e-06, + "loss": 0.9584, + "step": 627 + }, + { + "epoch": 0.01, + "grad_norm": 2.763945256989919, + "learning_rate": 9.858712715855574e-06, + "loss": 0.9212, + "step": 628 + }, + { + "epoch": 0.01, + "grad_norm": 2.3209015617503668, + "learning_rate": 9.874411302982733e-06, + "loss": 1.046, + "step": 629 + }, + { + "epoch": 0.01, + "grad_norm": 2.5382360253094647, + "learning_rate": 9.890109890109892e-06, + "loss": 0.9813, + "step": 630 + }, + { + "epoch": 0.01, + "grad_norm": 2.0641197631972568, + "learning_rate": 9.90580847723705e-06, + "loss": 1.0865, + "step": 631 + }, + { + "epoch": 0.01, + "grad_norm": 2.3353874568540784, + "learning_rate": 9.921507064364208e-06, + "loss": 0.986, + "step": 632 + }, + { + "epoch": 0.01, + "grad_norm": 2.567181344205934, + "learning_rate": 9.937205651491367e-06, + "loss": 0.9925, + "step": 633 + }, + { + "epoch": 0.01, + "grad_norm": 1.8968970989463219, + "learning_rate": 9.952904238618524e-06, + "loss": 0.9722, + "step": 634 + }, + { + "epoch": 0.01, + "grad_norm": 2.7295842483602093, + "learning_rate": 9.968602825745683e-06, + "loss": 0.9932, + "step": 635 + }, + { + "epoch": 0.01, + "grad_norm": 2.1742942845555375, + "learning_rate": 9.984301412872842e-06, + "loss": 1.063, + "step": 636 + }, + { + "epoch": 0.02, + "grad_norm": 2.0613877128889344, + "learning_rate": 1e-05, + "loss": 1.0711, + "step": 637 + }, + { + "epoch": 0.02, + "grad_norm": 2.2041562997471704, + "learning_rate": 1.0015698587127158e-05, + "loss": 0.8989, + "step": 638 + }, + { + "epoch": 0.02, + "grad_norm": 2.4935513103014344, + "learning_rate": 1.0031397174254319e-05, + "loss": 1.0556, + "step": 639 + }, + { + "epoch": 0.02, + "grad_norm": 2.1139614017481634, + "learning_rate": 1.0047095761381476e-05, + "loss": 0.869, + "step": 640 + }, + { + "epoch": 0.02, + "grad_norm": 2.254722104286999, + "learning_rate": 1.0062794348508635e-05, + "loss": 1.0769, + "step": 641 + }, + { + "epoch": 0.02, + "grad_norm": 3.433662321509383, + "learning_rate": 1.0078492935635794e-05, + "loss": 1.058, + "step": 642 + }, + { + "epoch": 0.02, + "grad_norm": 2.2097274013792085, + "learning_rate": 1.0094191522762952e-05, + "loss": 1.0908, + "step": 643 + }, + { + "epoch": 0.02, + "grad_norm": 2.815108188938875, + "learning_rate": 1.010989010989011e-05, + "loss": 1.0117, + "step": 644 + }, + { + "epoch": 0.02, + "grad_norm": 2.6568520342299893, + "learning_rate": 1.012558869701727e-05, + "loss": 1.0194, + "step": 645 + }, + { + "epoch": 0.02, + "grad_norm": 2.2567532533299772, + "learning_rate": 1.0141287284144427e-05, + "loss": 1.0419, + "step": 646 + }, + { + "epoch": 0.02, + "grad_norm": 2.07759922108056, + "learning_rate": 1.0156985871271588e-05, + "loss": 1.1711, + "step": 647 + }, + { + "epoch": 0.02, + "grad_norm": 2.2800894689580002, + "learning_rate": 1.0172684458398745e-05, + "loss": 0.9869, + "step": 648 + }, + { + "epoch": 0.02, + "grad_norm": 2.8019229457818216, + "learning_rate": 1.0188383045525902e-05, + "loss": 1.0465, + "step": 649 + }, + { + "epoch": 0.02, + "grad_norm": 1.9531863019978877, + "learning_rate": 1.0204081632653063e-05, + "loss": 1.0785, + "step": 650 + }, + { + "epoch": 0.02, + "grad_norm": 2.0091325010096943, + "learning_rate": 1.021978021978022e-05, + "loss": 1.144, + "step": 651 + }, + { + "epoch": 0.02, + "grad_norm": 3.036593520491222, + "learning_rate": 1.0235478806907379e-05, + "loss": 1.1074, + "step": 652 + }, + { + "epoch": 0.02, + "grad_norm": 2.054749843019664, + "learning_rate": 1.0251177394034538e-05, + "loss": 1.0987, + "step": 653 + }, + { + "epoch": 0.02, + "grad_norm": 1.0745647165908583, + "learning_rate": 1.0266875981161697e-05, + "loss": 0.9287, + "step": 654 + }, + { + "epoch": 0.02, + "grad_norm": 2.056368830690897, + "learning_rate": 1.0282574568288854e-05, + "loss": 1.1706, + "step": 655 + }, + { + "epoch": 0.02, + "grad_norm": 1.9305644144088185, + "learning_rate": 1.0298273155416015e-05, + "loss": 1.1287, + "step": 656 + }, + { + "epoch": 0.02, + "grad_norm": 2.385416022348612, + "learning_rate": 1.0313971742543172e-05, + "loss": 1.0216, + "step": 657 + }, + { + "epoch": 0.02, + "grad_norm": 2.255390649873031, + "learning_rate": 1.0329670329670332e-05, + "loss": 1.0511, + "step": 658 + }, + { + "epoch": 0.02, + "grad_norm": 2.2564567423151045, + "learning_rate": 1.034536891679749e-05, + "loss": 1.1292, + "step": 659 + }, + { + "epoch": 0.02, + "grad_norm": 2.001787351106439, + "learning_rate": 1.0361067503924647e-05, + "loss": 1.0731, + "step": 660 + }, + { + "epoch": 0.02, + "grad_norm": 2.3007343314573028, + "learning_rate": 1.0376766091051806e-05, + "loss": 0.9474, + "step": 661 + }, + { + "epoch": 0.02, + "grad_norm": 2.189758583186002, + "learning_rate": 1.0392464678178965e-05, + "loss": 1.1368, + "step": 662 + }, + { + "epoch": 0.02, + "grad_norm": 1.9796102219965228, + "learning_rate": 1.0408163265306123e-05, + "loss": 0.9997, + "step": 663 + }, + { + "epoch": 0.02, + "grad_norm": 1.9637912016375807, + "learning_rate": 1.042386185243328e-05, + "loss": 1.1047, + "step": 664 + }, + { + "epoch": 0.02, + "grad_norm": 2.1490337024668222, + "learning_rate": 1.0439560439560441e-05, + "loss": 1.0811, + "step": 665 + }, + { + "epoch": 0.02, + "grad_norm": 2.0167077523783794, + "learning_rate": 1.0455259026687598e-05, + "loss": 0.9347, + "step": 666 + }, + { + "epoch": 0.02, + "grad_norm": 2.150970436144793, + "learning_rate": 1.0470957613814759e-05, + "loss": 1.0801, + "step": 667 + }, + { + "epoch": 0.02, + "grad_norm": 2.6423345676997045, + "learning_rate": 1.0486656200941916e-05, + "loss": 1.0127, + "step": 668 + }, + { + "epoch": 0.02, + "grad_norm": 2.210607584948613, + "learning_rate": 1.0502354788069075e-05, + "loss": 1.0585, + "step": 669 + }, + { + "epoch": 0.02, + "grad_norm": 2.35962634009039, + "learning_rate": 1.0518053375196234e-05, + "loss": 1.1717, + "step": 670 + }, + { + "epoch": 0.02, + "grad_norm": 2.117973716888141, + "learning_rate": 1.0533751962323391e-05, + "loss": 1.0319, + "step": 671 + }, + { + "epoch": 0.02, + "grad_norm": 2.264825496366871, + "learning_rate": 1.054945054945055e-05, + "loss": 1.0487, + "step": 672 + }, + { + "epoch": 0.02, + "grad_norm": 2.601269952009485, + "learning_rate": 1.0565149136577709e-05, + "loss": 1.0725, + "step": 673 + }, + { + "epoch": 0.02, + "grad_norm": 2.111415083132811, + "learning_rate": 1.0580847723704868e-05, + "loss": 1.0369, + "step": 674 + }, + { + "epoch": 0.02, + "grad_norm": 1.9928673596257664, + "learning_rate": 1.0596546310832025e-05, + "loss": 0.9186, + "step": 675 + }, + { + "epoch": 0.02, + "grad_norm": 1.20940089214994, + "learning_rate": 1.0612244897959186e-05, + "loss": 1.0282, + "step": 676 + }, + { + "epoch": 0.02, + "grad_norm": 1.9385974946603293, + "learning_rate": 1.0627943485086343e-05, + "loss": 1.1416, + "step": 677 + }, + { + "epoch": 0.02, + "grad_norm": 2.4819293370155036, + "learning_rate": 1.0643642072213503e-05, + "loss": 1.1557, + "step": 678 + }, + { + "epoch": 0.02, + "grad_norm": 2.3537168594848037, + "learning_rate": 1.065934065934066e-05, + "loss": 1.092, + "step": 679 + }, + { + "epoch": 0.02, + "grad_norm": 1.9942775343876251, + "learning_rate": 1.067503924646782e-05, + "loss": 1.0178, + "step": 680 + }, + { + "epoch": 0.02, + "grad_norm": 1.9801319527091996, + "learning_rate": 1.0690737833594977e-05, + "loss": 1.1373, + "step": 681 + }, + { + "epoch": 0.02, + "grad_norm": 2.248252018509441, + "learning_rate": 1.0706436420722135e-05, + "loss": 0.9736, + "step": 682 + }, + { + "epoch": 0.02, + "grad_norm": 2.047938202115324, + "learning_rate": 1.0722135007849294e-05, + "loss": 1.072, + "step": 683 + }, + { + "epoch": 0.02, + "grad_norm": 2.160758124883826, + "learning_rate": 1.0737833594976452e-05, + "loss": 1.0925, + "step": 684 + }, + { + "epoch": 0.02, + "grad_norm": 1.9356336138735775, + "learning_rate": 1.0753532182103612e-05, + "loss": 1.1053, + "step": 685 + }, + { + "epoch": 0.02, + "grad_norm": 2.155003226738246, + "learning_rate": 1.076923076923077e-05, + "loss": 1.0559, + "step": 686 + }, + { + "epoch": 0.02, + "grad_norm": 2.2989838249863315, + "learning_rate": 1.078492935635793e-05, + "loss": 0.8375, + "step": 687 + }, + { + "epoch": 0.02, + "grad_norm": 1.8625360260439838, + "learning_rate": 1.0800627943485087e-05, + "loss": 1.0727, + "step": 688 + }, + { + "epoch": 0.02, + "grad_norm": 1.9129983838494375, + "learning_rate": 1.0816326530612246e-05, + "loss": 1.0925, + "step": 689 + }, + { + "epoch": 0.02, + "grad_norm": 2.174594098778036, + "learning_rate": 1.0832025117739405e-05, + "loss": 1.0139, + "step": 690 + }, + { + "epoch": 0.02, + "grad_norm": 2.1030489578580047, + "learning_rate": 1.0847723704866562e-05, + "loss": 1.0783, + "step": 691 + }, + { + "epoch": 0.02, + "grad_norm": 3.7873548471909166, + "learning_rate": 1.0863422291993721e-05, + "loss": 1.1974, + "step": 692 + }, + { + "epoch": 0.02, + "grad_norm": 1.169257536200449, + "learning_rate": 1.087912087912088e-05, + "loss": 1.0182, + "step": 693 + }, + { + "epoch": 0.02, + "grad_norm": 2.2970781405748295, + "learning_rate": 1.0894819466248039e-05, + "loss": 1.1625, + "step": 694 + }, + { + "epoch": 0.02, + "grad_norm": 2.1193224726718443, + "learning_rate": 1.0910518053375196e-05, + "loss": 1.0556, + "step": 695 + }, + { + "epoch": 0.02, + "grad_norm": 2.2322713218883186, + "learning_rate": 1.0926216640502357e-05, + "loss": 0.9128, + "step": 696 + }, + { + "epoch": 0.02, + "grad_norm": 2.0880843466089973, + "learning_rate": 1.0941915227629514e-05, + "loss": 1.1412, + "step": 697 + }, + { + "epoch": 0.02, + "grad_norm": 2.2963743450306744, + "learning_rate": 1.0957613814756674e-05, + "loss": 1.0499, + "step": 698 + }, + { + "epoch": 0.02, + "grad_norm": 1.9917914810376953, + "learning_rate": 1.0973312401883831e-05, + "loss": 0.9872, + "step": 699 + }, + { + "epoch": 0.02, + "grad_norm": 1.1227319194910057, + "learning_rate": 1.098901098901099e-05, + "loss": 0.951, + "step": 700 + }, + { + "epoch": 0.02, + "grad_norm": 2.3891139784177655, + "learning_rate": 1.1004709576138148e-05, + "loss": 1.0717, + "step": 701 + }, + { + "epoch": 0.02, + "grad_norm": 2.0255012196921713, + "learning_rate": 1.1020408163265306e-05, + "loss": 1.1795, + "step": 702 + }, + { + "epoch": 0.02, + "grad_norm": 2.1474098599133122, + "learning_rate": 1.1036106750392465e-05, + "loss": 0.9438, + "step": 703 + }, + { + "epoch": 0.02, + "grad_norm": 2.6425612263903964, + "learning_rate": 1.1051805337519623e-05, + "loss": 1.0524, + "step": 704 + }, + { + "epoch": 0.02, + "grad_norm": 1.1086789031547226, + "learning_rate": 1.1067503924646783e-05, + "loss": 0.904, + "step": 705 + }, + { + "epoch": 0.02, + "grad_norm": 2.003160610771393, + "learning_rate": 1.108320251177394e-05, + "loss": 1.0214, + "step": 706 + }, + { + "epoch": 0.02, + "grad_norm": 1.8880000162073174, + "learning_rate": 1.1098901098901101e-05, + "loss": 1.0838, + "step": 707 + }, + { + "epoch": 0.02, + "grad_norm": 2.5133594807977144, + "learning_rate": 1.1114599686028258e-05, + "loss": 0.9869, + "step": 708 + }, + { + "epoch": 0.02, + "grad_norm": 2.2448171743997745, + "learning_rate": 1.1130298273155417e-05, + "loss": 1.0992, + "step": 709 + }, + { + "epoch": 0.02, + "grad_norm": 2.3571932101223205, + "learning_rate": 1.1145996860282576e-05, + "loss": 0.9882, + "step": 710 + }, + { + "epoch": 0.02, + "grad_norm": 2.6463191671465918, + "learning_rate": 1.1161695447409735e-05, + "loss": 0.9855, + "step": 711 + }, + { + "epoch": 0.02, + "grad_norm": 1.058720049326792, + "learning_rate": 1.1177394034536892e-05, + "loss": 0.9139, + "step": 712 + }, + { + "epoch": 0.02, + "grad_norm": 2.1944254341048484, + "learning_rate": 1.119309262166405e-05, + "loss": 1.0567, + "step": 713 + }, + { + "epoch": 0.02, + "grad_norm": 2.159053280329375, + "learning_rate": 1.120879120879121e-05, + "loss": 1.1232, + "step": 714 + }, + { + "epoch": 0.02, + "grad_norm": 2.701690119348469, + "learning_rate": 1.1224489795918367e-05, + "loss": 1.1087, + "step": 715 + }, + { + "epoch": 0.02, + "grad_norm": 1.9882052493465345, + "learning_rate": 1.1240188383045527e-05, + "loss": 1.0173, + "step": 716 + }, + { + "epoch": 0.02, + "grad_norm": 1.1583077597956641, + "learning_rate": 1.1255886970172685e-05, + "loss": 0.9262, + "step": 717 + }, + { + "epoch": 0.02, + "grad_norm": 2.5772188922536032, + "learning_rate": 1.1271585557299845e-05, + "loss": 0.8938, + "step": 718 + }, + { + "epoch": 0.02, + "grad_norm": 1.094249591445517, + "learning_rate": 1.1287284144427002e-05, + "loss": 0.8605, + "step": 719 + }, + { + "epoch": 0.02, + "grad_norm": 1.1055999072951226, + "learning_rate": 1.1302982731554161e-05, + "loss": 0.8784, + "step": 720 + }, + { + "epoch": 0.02, + "grad_norm": 2.36635653976299, + "learning_rate": 1.131868131868132e-05, + "loss": 1.1444, + "step": 721 + }, + { + "epoch": 0.02, + "grad_norm": 2.2739347068008566, + "learning_rate": 1.1334379905808479e-05, + "loss": 1.0108, + "step": 722 + }, + { + "epoch": 0.02, + "grad_norm": 1.1384019145209565, + "learning_rate": 1.1350078492935636e-05, + "loss": 0.8914, + "step": 723 + }, + { + "epoch": 0.02, + "grad_norm": 2.676984433031132, + "learning_rate": 1.1365777080062793e-05, + "loss": 0.9551, + "step": 724 + }, + { + "epoch": 0.02, + "grad_norm": 1.1940222458442384, + "learning_rate": 1.1381475667189954e-05, + "loss": 0.8936, + "step": 725 + }, + { + "epoch": 0.02, + "grad_norm": 2.1032392704157763, + "learning_rate": 1.1397174254317111e-05, + "loss": 1.0954, + "step": 726 + }, + { + "epoch": 0.02, + "grad_norm": 2.080217384212294, + "learning_rate": 1.1412872841444272e-05, + "loss": 0.8854, + "step": 727 + }, + { + "epoch": 0.02, + "grad_norm": 2.165143608942416, + "learning_rate": 1.1428571428571429e-05, + "loss": 1.15, + "step": 728 + }, + { + "epoch": 0.02, + "grad_norm": 1.976777595451746, + "learning_rate": 1.1444270015698588e-05, + "loss": 1.0283, + "step": 729 + }, + { + "epoch": 0.02, + "grad_norm": 2.2819965734927607, + "learning_rate": 1.1459968602825747e-05, + "loss": 0.9947, + "step": 730 + }, + { + "epoch": 0.02, + "grad_norm": 2.0194416051610062, + "learning_rate": 1.1475667189952906e-05, + "loss": 1.0972, + "step": 731 + }, + { + "epoch": 0.02, + "grad_norm": 2.2247318028160614, + "learning_rate": 1.1491365777080063e-05, + "loss": 1.1307, + "step": 732 + }, + { + "epoch": 0.02, + "grad_norm": 2.020144016145253, + "learning_rate": 1.1507064364207223e-05, + "loss": 0.9986, + "step": 733 + }, + { + "epoch": 0.02, + "grad_norm": 2.132364358954378, + "learning_rate": 1.152276295133438e-05, + "loss": 1.1042, + "step": 734 + }, + { + "epoch": 0.02, + "grad_norm": 2.3027403370745274, + "learning_rate": 1.1538461538461538e-05, + "loss": 0.9903, + "step": 735 + }, + { + "epoch": 0.02, + "grad_norm": 2.2004359903523123, + "learning_rate": 1.1554160125588698e-05, + "loss": 1.077, + "step": 736 + }, + { + "epoch": 0.02, + "grad_norm": 2.11483028086191, + "learning_rate": 1.1569858712715856e-05, + "loss": 1.1357, + "step": 737 + }, + { + "epoch": 0.02, + "grad_norm": 1.150161214813921, + "learning_rate": 1.1585557299843016e-05, + "loss": 0.9032, + "step": 738 + }, + { + "epoch": 0.02, + "grad_norm": 2.1655931135939572, + "learning_rate": 1.1601255886970173e-05, + "loss": 1.0779, + "step": 739 + }, + { + "epoch": 0.02, + "grad_norm": 2.2334426849228857, + "learning_rate": 1.1616954474097332e-05, + "loss": 0.9353, + "step": 740 + }, + { + "epoch": 0.02, + "grad_norm": 2.138715050644722, + "learning_rate": 1.1632653061224491e-05, + "loss": 1.0349, + "step": 741 + }, + { + "epoch": 0.02, + "grad_norm": 2.2967798696251274, + "learning_rate": 1.164835164835165e-05, + "loss": 1.1195, + "step": 742 + }, + { + "epoch": 0.02, + "grad_norm": 2.270321030983876, + "learning_rate": 1.1664050235478807e-05, + "loss": 1.124, + "step": 743 + }, + { + "epoch": 0.02, + "grad_norm": 2.27499980865107, + "learning_rate": 1.1679748822605968e-05, + "loss": 1.0397, + "step": 744 + }, + { + "epoch": 0.02, + "grad_norm": 2.21963432322176, + "learning_rate": 1.1695447409733125e-05, + "loss": 1.1271, + "step": 745 + }, + { + "epoch": 0.02, + "grad_norm": 2.390443080091919, + "learning_rate": 1.1711145996860282e-05, + "loss": 1.0416, + "step": 746 + }, + { + "epoch": 0.02, + "grad_norm": 2.032460656814261, + "learning_rate": 1.1726844583987443e-05, + "loss": 1.0329, + "step": 747 + }, + { + "epoch": 0.02, + "grad_norm": 2.1337159989939827, + "learning_rate": 1.17425431711146e-05, + "loss": 1.1082, + "step": 748 + }, + { + "epoch": 0.02, + "grad_norm": 1.9857380478977043, + "learning_rate": 1.1758241758241759e-05, + "loss": 1.1932, + "step": 749 + }, + { + "epoch": 0.02, + "grad_norm": 1.1388437053952933, + "learning_rate": 1.1773940345368918e-05, + "loss": 0.9127, + "step": 750 + }, + { + "epoch": 0.02, + "grad_norm": 2.3866394320690754, + "learning_rate": 1.1789638932496077e-05, + "loss": 1.1685, + "step": 751 + }, + { + "epoch": 0.02, + "grad_norm": 2.102701985996273, + "learning_rate": 1.1805337519623234e-05, + "loss": 0.9811, + "step": 752 + }, + { + "epoch": 0.02, + "grad_norm": 2.2298231487022173, + "learning_rate": 1.1821036106750394e-05, + "loss": 0.9405, + "step": 753 + }, + { + "epoch": 0.02, + "grad_norm": 3.9567012535115933, + "learning_rate": 1.1836734693877552e-05, + "loss": 1.0548, + "step": 754 + }, + { + "epoch": 0.02, + "grad_norm": 2.067706373448108, + "learning_rate": 1.1852433281004712e-05, + "loss": 1.0719, + "step": 755 + }, + { + "epoch": 0.02, + "grad_norm": 2.127907177614354, + "learning_rate": 1.186813186813187e-05, + "loss": 1.0014, + "step": 756 + }, + { + "epoch": 0.02, + "grad_norm": 2.2155070382884006, + "learning_rate": 1.1883830455259027e-05, + "loss": 0.9326, + "step": 757 + }, + { + "epoch": 0.02, + "grad_norm": 2.7747908810983413, + "learning_rate": 1.1899529042386187e-05, + "loss": 1.1893, + "step": 758 + }, + { + "epoch": 0.02, + "grad_norm": 2.117582407615235, + "learning_rate": 1.1915227629513344e-05, + "loss": 1.0026, + "step": 759 + }, + { + "epoch": 0.02, + "grad_norm": 2.216087476197426, + "learning_rate": 1.1930926216640503e-05, + "loss": 1.2345, + "step": 760 + }, + { + "epoch": 0.02, + "grad_norm": 2.5706986073715252, + "learning_rate": 1.1946624803767662e-05, + "loss": 1.14, + "step": 761 + }, + { + "epoch": 0.02, + "grad_norm": 2.1334779199367544, + "learning_rate": 1.1962323390894821e-05, + "loss": 1.0552, + "step": 762 + }, + { + "epoch": 0.02, + "grad_norm": 2.3244863350087295, + "learning_rate": 1.1978021978021978e-05, + "loss": 0.9734, + "step": 763 + }, + { + "epoch": 0.02, + "grad_norm": 2.122759613160707, + "learning_rate": 1.1993720565149139e-05, + "loss": 1.1752, + "step": 764 + }, + { + "epoch": 0.02, + "grad_norm": 2.885530325975881, + "learning_rate": 1.2009419152276296e-05, + "loss": 0.9872, + "step": 765 + }, + { + "epoch": 0.02, + "grad_norm": 2.481429198222887, + "learning_rate": 1.2025117739403457e-05, + "loss": 1.1377, + "step": 766 + }, + { + "epoch": 0.02, + "grad_norm": 2.0972251474625856, + "learning_rate": 1.2040816326530614e-05, + "loss": 1.1725, + "step": 767 + }, + { + "epoch": 0.02, + "grad_norm": 2.482788272555507, + "learning_rate": 1.2056514913657771e-05, + "loss": 1.0933, + "step": 768 + }, + { + "epoch": 0.02, + "grad_norm": 1.950823524860713, + "learning_rate": 1.207221350078493e-05, + "loss": 0.9641, + "step": 769 + }, + { + "epoch": 0.02, + "grad_norm": 1.9598353709923595, + "learning_rate": 1.2087912087912089e-05, + "loss": 0.9301, + "step": 770 + }, + { + "epoch": 0.02, + "grad_norm": 1.9930307023852523, + "learning_rate": 1.2103610675039248e-05, + "loss": 1.0872, + "step": 771 + }, + { + "epoch": 0.02, + "grad_norm": 2.4643391400229535, + "learning_rate": 1.2119309262166405e-05, + "loss": 0.9632, + "step": 772 + }, + { + "epoch": 0.02, + "grad_norm": 2.164487052741329, + "learning_rate": 1.2135007849293565e-05, + "loss": 1.0495, + "step": 773 + }, + { + "epoch": 0.02, + "grad_norm": 2.1154363659920667, + "learning_rate": 1.2150706436420723e-05, + "loss": 1.0278, + "step": 774 + }, + { + "epoch": 0.02, + "grad_norm": 3.9258990308662782, + "learning_rate": 1.2166405023547883e-05, + "loss": 1.0258, + "step": 775 + }, + { + "epoch": 0.02, + "grad_norm": 2.2519420017015697, + "learning_rate": 1.218210361067504e-05, + "loss": 1.0763, + "step": 776 + }, + { + "epoch": 0.02, + "grad_norm": 2.197919980418764, + "learning_rate": 1.21978021978022e-05, + "loss": 1.1673, + "step": 777 + }, + { + "epoch": 0.02, + "grad_norm": 2.0991110134282023, + "learning_rate": 1.2213500784929358e-05, + "loss": 1.0536, + "step": 778 + }, + { + "epoch": 0.02, + "grad_norm": 2.3866797692195187, + "learning_rate": 1.2229199372056515e-05, + "loss": 1.013, + "step": 779 + }, + { + "epoch": 0.02, + "grad_norm": 2.0292659996738602, + "learning_rate": 1.2244897959183674e-05, + "loss": 0.9711, + "step": 780 + }, + { + "epoch": 0.02, + "grad_norm": 1.9711452847377406, + "learning_rate": 1.2260596546310833e-05, + "loss": 1.1059, + "step": 781 + }, + { + "epoch": 0.02, + "grad_norm": 2.285103798737076, + "learning_rate": 1.2276295133437992e-05, + "loss": 1.073, + "step": 782 + }, + { + "epoch": 0.02, + "grad_norm": 2.445741409019786, + "learning_rate": 1.229199372056515e-05, + "loss": 1.1042, + "step": 783 + }, + { + "epoch": 0.02, + "grad_norm": 2.5110225923373957, + "learning_rate": 1.230769230769231e-05, + "loss": 0.9771, + "step": 784 + }, + { + "epoch": 0.02, + "grad_norm": 2.2957875327611923, + "learning_rate": 1.2323390894819467e-05, + "loss": 1.0552, + "step": 785 + }, + { + "epoch": 0.02, + "grad_norm": 2.194224860066982, + "learning_rate": 1.2339089481946628e-05, + "loss": 1.1722, + "step": 786 + }, + { + "epoch": 0.02, + "grad_norm": 1.9893408312888006, + "learning_rate": 1.2354788069073785e-05, + "loss": 1.0492, + "step": 787 + }, + { + "epoch": 0.02, + "grad_norm": 2.1447724117069478, + "learning_rate": 1.2370486656200944e-05, + "loss": 1.0102, + "step": 788 + }, + { + "epoch": 0.02, + "grad_norm": 3.370079756196991, + "learning_rate": 1.23861852433281e-05, + "loss": 1.0639, + "step": 789 + }, + { + "epoch": 0.02, + "grad_norm": 2.443555622201271, + "learning_rate": 1.240188383045526e-05, + "loss": 1.0385, + "step": 790 + }, + { + "epoch": 0.02, + "grad_norm": 2.3350527771031815, + "learning_rate": 1.2417582417582419e-05, + "loss": 1.1614, + "step": 791 + }, + { + "epoch": 0.02, + "grad_norm": 2.1721539991406607, + "learning_rate": 1.2433281004709576e-05, + "loss": 1.1632, + "step": 792 + }, + { + "epoch": 0.02, + "grad_norm": 2.211441727231417, + "learning_rate": 1.2448979591836736e-05, + "loss": 1.0552, + "step": 793 + }, + { + "epoch": 0.02, + "grad_norm": 1.1299027109001278, + "learning_rate": 1.2464678178963894e-05, + "loss": 0.9457, + "step": 794 + }, + { + "epoch": 0.02, + "grad_norm": 1.7946512474354575, + "learning_rate": 1.2480376766091054e-05, + "loss": 1.0165, + "step": 795 + }, + { + "epoch": 0.02, + "grad_norm": 2.4470966458492676, + "learning_rate": 1.2496075353218211e-05, + "loss": 1.1242, + "step": 796 + }, + { + "epoch": 0.02, + "grad_norm": 2.199328971312565, + "learning_rate": 1.251177394034537e-05, + "loss": 1.1373, + "step": 797 + }, + { + "epoch": 0.02, + "grad_norm": 2.454113296878447, + "learning_rate": 1.2527472527472529e-05, + "loss": 1.1583, + "step": 798 + }, + { + "epoch": 0.02, + "grad_norm": 2.1855172582687206, + "learning_rate": 1.2543171114599686e-05, + "loss": 1.0808, + "step": 799 + }, + { + "epoch": 0.02, + "grad_norm": 1.8865132683956678, + "learning_rate": 1.2558869701726845e-05, + "loss": 1.0881, + "step": 800 + }, + { + "epoch": 0.02, + "grad_norm": 2.1928566226341593, + "learning_rate": 1.2574568288854004e-05, + "loss": 1.0572, + "step": 801 + }, + { + "epoch": 0.02, + "grad_norm": 2.1619273432593293, + "learning_rate": 1.2590266875981163e-05, + "loss": 1.1207, + "step": 802 + }, + { + "epoch": 0.02, + "grad_norm": 2.3052771442383206, + "learning_rate": 1.260596546310832e-05, + "loss": 1.1175, + "step": 803 + }, + { + "epoch": 0.02, + "grad_norm": 2.423372366842809, + "learning_rate": 1.262166405023548e-05, + "loss": 1.1739, + "step": 804 + }, + { + "epoch": 0.02, + "grad_norm": 2.194237863465144, + "learning_rate": 1.2637362637362638e-05, + "loss": 1.1706, + "step": 805 + }, + { + "epoch": 0.02, + "grad_norm": 2.241455600718936, + "learning_rate": 1.2653061224489798e-05, + "loss": 0.978, + "step": 806 + }, + { + "epoch": 0.02, + "grad_norm": 2.0670805311648777, + "learning_rate": 1.2668759811616956e-05, + "loss": 0.9952, + "step": 807 + }, + { + "epoch": 0.02, + "grad_norm": 2.2279052274546274, + "learning_rate": 1.2684458398744115e-05, + "loss": 1.0503, + "step": 808 + }, + { + "epoch": 0.02, + "grad_norm": 2.2497303212496926, + "learning_rate": 1.2700156985871272e-05, + "loss": 1.1136, + "step": 809 + }, + { + "epoch": 0.02, + "grad_norm": 2.244386717269416, + "learning_rate": 1.271585557299843e-05, + "loss": 1.0167, + "step": 810 + }, + { + "epoch": 0.02, + "grad_norm": 2.4665150303938486, + "learning_rate": 1.273155416012559e-05, + "loss": 1.1813, + "step": 811 + }, + { + "epoch": 0.02, + "grad_norm": 2.353202061520077, + "learning_rate": 1.2747252747252747e-05, + "loss": 0.9896, + "step": 812 + }, + { + "epoch": 0.02, + "grad_norm": 2.3737304893952227, + "learning_rate": 1.2762951334379907e-05, + "loss": 1.155, + "step": 813 + }, + { + "epoch": 0.02, + "grad_norm": 2.3070387703580515, + "learning_rate": 1.2778649921507064e-05, + "loss": 0.9417, + "step": 814 + }, + { + "epoch": 0.02, + "grad_norm": 2.014118570881036, + "learning_rate": 1.2794348508634225e-05, + "loss": 1.1778, + "step": 815 + }, + { + "epoch": 0.02, + "grad_norm": 2.1304946241077074, + "learning_rate": 1.2810047095761382e-05, + "loss": 1.0043, + "step": 816 + }, + { + "epoch": 0.02, + "grad_norm": 5.38842795785393, + "learning_rate": 1.2825745682888541e-05, + "loss": 0.9078, + "step": 817 + }, + { + "epoch": 0.02, + "grad_norm": 1.9047337231491963, + "learning_rate": 1.28414442700157e-05, + "loss": 1.0441, + "step": 818 + }, + { + "epoch": 0.02, + "grad_norm": 2.1203185780486584, + "learning_rate": 1.2857142857142859e-05, + "loss": 0.9212, + "step": 819 + }, + { + "epoch": 0.02, + "grad_norm": 2.246325896954263, + "learning_rate": 1.2872841444270016e-05, + "loss": 1.0396, + "step": 820 + }, + { + "epoch": 0.02, + "grad_norm": 2.129380156327021, + "learning_rate": 1.2888540031397175e-05, + "loss": 1.1062, + "step": 821 + }, + { + "epoch": 0.02, + "grad_norm": 2.5531664031626122, + "learning_rate": 1.2904238618524334e-05, + "loss": 0.9881, + "step": 822 + }, + { + "epoch": 0.02, + "grad_norm": 2.3368203399011658, + "learning_rate": 1.2919937205651491e-05, + "loss": 0.9897, + "step": 823 + }, + { + "epoch": 0.02, + "grad_norm": 2.199548954316752, + "learning_rate": 1.2935635792778652e-05, + "loss": 1.0641, + "step": 824 + }, + { + "epoch": 0.02, + "grad_norm": 2.0864885602193044, + "learning_rate": 1.2951334379905809e-05, + "loss": 0.9858, + "step": 825 + }, + { + "epoch": 0.02, + "grad_norm": 2.275119642158089, + "learning_rate": 1.296703296703297e-05, + "loss": 1.1603, + "step": 826 + }, + { + "epoch": 0.02, + "grad_norm": 2.084075895894492, + "learning_rate": 1.2982731554160127e-05, + "loss": 0.9923, + "step": 827 + }, + { + "epoch": 0.02, + "grad_norm": 2.434595085910085, + "learning_rate": 1.2998430141287286e-05, + "loss": 1.1164, + "step": 828 + }, + { + "epoch": 0.02, + "grad_norm": 2.189693007846231, + "learning_rate": 1.3014128728414443e-05, + "loss": 1.0534, + "step": 829 + }, + { + "epoch": 0.02, + "grad_norm": 2.6281576697406073, + "learning_rate": 1.3029827315541603e-05, + "loss": 1.019, + "step": 830 + }, + { + "epoch": 0.02, + "grad_norm": 2.3506822116646955, + "learning_rate": 1.304552590266876e-05, + "loss": 1.1334, + "step": 831 + }, + { + "epoch": 0.02, + "grad_norm": 2.1441046764109175, + "learning_rate": 1.3061224489795918e-05, + "loss": 1.0061, + "step": 832 + }, + { + "epoch": 0.02, + "grad_norm": 2.234478469652428, + "learning_rate": 1.3076923076923078e-05, + "loss": 0.9864, + "step": 833 + }, + { + "epoch": 0.02, + "grad_norm": 2.457957007504857, + "learning_rate": 1.3092621664050235e-05, + "loss": 1.1574, + "step": 834 + }, + { + "epoch": 0.02, + "grad_norm": 2.192848537038557, + "learning_rate": 1.3108320251177396e-05, + "loss": 1.0835, + "step": 835 + }, + { + "epoch": 0.02, + "grad_norm": 2.113060318518013, + "learning_rate": 1.3124018838304553e-05, + "loss": 1.1835, + "step": 836 + }, + { + "epoch": 0.02, + "grad_norm": 2.0612798827131074, + "learning_rate": 1.3139717425431712e-05, + "loss": 0.9176, + "step": 837 + }, + { + "epoch": 0.02, + "grad_norm": 2.4349339257051117, + "learning_rate": 1.3155416012558871e-05, + "loss": 1.0007, + "step": 838 + }, + { + "epoch": 0.02, + "grad_norm": 2.2632261739303337, + "learning_rate": 1.317111459968603e-05, + "loss": 1.0065, + "step": 839 + }, + { + "epoch": 0.02, + "grad_norm": 2.1043756639146487, + "learning_rate": 1.3186813186813187e-05, + "loss": 1.0547, + "step": 840 + }, + { + "epoch": 0.02, + "grad_norm": 2.711811223585446, + "learning_rate": 1.3202511773940348e-05, + "loss": 1.0968, + "step": 841 + }, + { + "epoch": 0.02, + "grad_norm": 1.1080295807093565, + "learning_rate": 1.3218210361067505e-05, + "loss": 0.9935, + "step": 842 + }, + { + "epoch": 0.02, + "grad_norm": 2.0978840255520743, + "learning_rate": 1.3233908948194662e-05, + "loss": 1.0421, + "step": 843 + }, + { + "epoch": 0.02, + "grad_norm": 1.9894962067286432, + "learning_rate": 1.3249607535321823e-05, + "loss": 1.1165, + "step": 844 + }, + { + "epoch": 0.02, + "grad_norm": 1.9114631715488313, + "learning_rate": 1.326530612244898e-05, + "loss": 1.1833, + "step": 845 + }, + { + "epoch": 0.02, + "grad_norm": 2.37980715209133, + "learning_rate": 1.328100470957614e-05, + "loss": 1.02, + "step": 846 + }, + { + "epoch": 0.02, + "grad_norm": 2.690858313865376, + "learning_rate": 1.3296703296703298e-05, + "loss": 1.0807, + "step": 847 + }, + { + "epoch": 0.02, + "grad_norm": 2.1877869163471693, + "learning_rate": 1.3312401883830456e-05, + "loss": 0.9606, + "step": 848 + }, + { + "epoch": 0.02, + "grad_norm": 2.087297876320505, + "learning_rate": 1.3328100470957614e-05, + "loss": 1.0345, + "step": 849 + }, + { + "epoch": 0.02, + "grad_norm": 2.340953790474061, + "learning_rate": 1.3343799058084774e-05, + "loss": 0.9694, + "step": 850 + }, + { + "epoch": 0.02, + "grad_norm": 2.230325415222141, + "learning_rate": 1.3359497645211931e-05, + "loss": 1.1096, + "step": 851 + }, + { + "epoch": 0.02, + "grad_norm": 2.0468550957624707, + "learning_rate": 1.3375196232339092e-05, + "loss": 1.1068, + "step": 852 + }, + { + "epoch": 0.02, + "grad_norm": 2.1334971533771956, + "learning_rate": 1.339089481946625e-05, + "loss": 1.0407, + "step": 853 + }, + { + "epoch": 0.02, + "grad_norm": 2.3588279036918305, + "learning_rate": 1.3406593406593406e-05, + "loss": 0.9026, + "step": 854 + }, + { + "epoch": 0.02, + "grad_norm": 1.9652153324049288, + "learning_rate": 1.3422291993720567e-05, + "loss": 1.0652, + "step": 855 + }, + { + "epoch": 0.02, + "grad_norm": 2.5678606625192932, + "learning_rate": 1.3437990580847724e-05, + "loss": 1.0616, + "step": 856 + }, + { + "epoch": 0.02, + "grad_norm": 2.1524688637607614, + "learning_rate": 1.3453689167974883e-05, + "loss": 1.0393, + "step": 857 + }, + { + "epoch": 0.02, + "grad_norm": 2.7794804385942604, + "learning_rate": 1.3469387755102042e-05, + "loss": 0.9492, + "step": 858 + }, + { + "epoch": 0.02, + "grad_norm": 1.9569114745486351, + "learning_rate": 1.34850863422292e-05, + "loss": 0.9983, + "step": 859 + }, + { + "epoch": 0.02, + "grad_norm": 2.0476182082476457, + "learning_rate": 1.3500784929356358e-05, + "loss": 0.9506, + "step": 860 + }, + { + "epoch": 0.02, + "grad_norm": 2.404917025169772, + "learning_rate": 1.3516483516483519e-05, + "loss": 1.109, + "step": 861 + }, + { + "epoch": 0.02, + "grad_norm": 1.3009950672222979, + "learning_rate": 1.3532182103610676e-05, + "loss": 0.9571, + "step": 862 + }, + { + "epoch": 0.02, + "grad_norm": 2.0263223014594094, + "learning_rate": 1.3547880690737836e-05, + "loss": 1.033, + "step": 863 + }, + { + "epoch": 0.02, + "grad_norm": 2.118269029147181, + "learning_rate": 1.3563579277864994e-05, + "loss": 0.8848, + "step": 864 + }, + { + "epoch": 0.02, + "grad_norm": 2.2205005730985654, + "learning_rate": 1.357927786499215e-05, + "loss": 0.9616, + "step": 865 + }, + { + "epoch": 0.02, + "grad_norm": 2.1221503025103576, + "learning_rate": 1.3594976452119311e-05, + "loss": 0.922, + "step": 866 + }, + { + "epoch": 0.02, + "grad_norm": 2.3249321270520875, + "learning_rate": 1.3610675039246469e-05, + "loss": 1.0404, + "step": 867 + }, + { + "epoch": 0.02, + "grad_norm": 2.141897006231913, + "learning_rate": 1.3626373626373627e-05, + "loss": 0.9233, + "step": 868 + }, + { + "epoch": 0.02, + "grad_norm": 2.412337936738671, + "learning_rate": 1.3642072213500786e-05, + "loss": 1.1951, + "step": 869 + }, + { + "epoch": 0.02, + "grad_norm": 2.1567484424888015, + "learning_rate": 1.3657770800627945e-05, + "loss": 1.0162, + "step": 870 + }, + { + "epoch": 0.02, + "grad_norm": 2.014318033908868, + "learning_rate": 1.3673469387755102e-05, + "loss": 0.9757, + "step": 871 + }, + { + "epoch": 0.02, + "grad_norm": 2.7681513708677294, + "learning_rate": 1.3689167974882263e-05, + "loss": 1.0129, + "step": 872 + }, + { + "epoch": 0.02, + "grad_norm": 1.9969991274880363, + "learning_rate": 1.370486656200942e-05, + "loss": 1.0653, + "step": 873 + }, + { + "epoch": 0.02, + "grad_norm": 2.4222705327728487, + "learning_rate": 1.3720565149136579e-05, + "loss": 1.1182, + "step": 874 + }, + { + "epoch": 0.02, + "grad_norm": 2.2679051989699746, + "learning_rate": 1.3736263736263738e-05, + "loss": 1.1281, + "step": 875 + }, + { + "epoch": 0.02, + "grad_norm": 2.188375905521005, + "learning_rate": 1.3751962323390895e-05, + "loss": 1.1036, + "step": 876 + }, + { + "epoch": 0.02, + "grad_norm": 2.1226274636264404, + "learning_rate": 1.3767660910518054e-05, + "loss": 1.1187, + "step": 877 + }, + { + "epoch": 0.02, + "grad_norm": 2.033988843074651, + "learning_rate": 1.3783359497645213e-05, + "loss": 1.1304, + "step": 878 + }, + { + "epoch": 0.02, + "grad_norm": 2.1130689496264443, + "learning_rate": 1.3799058084772372e-05, + "loss": 1.0253, + "step": 879 + }, + { + "epoch": 0.02, + "grad_norm": 2.0626055005709456, + "learning_rate": 1.3814756671899529e-05, + "loss": 1.0759, + "step": 880 + }, + { + "epoch": 0.02, + "grad_norm": 2.4872961285671895, + "learning_rate": 1.383045525902669e-05, + "loss": 0.9525, + "step": 881 + }, + { + "epoch": 0.02, + "grad_norm": 2.1811742771882776, + "learning_rate": 1.3846153846153847e-05, + "loss": 1.0805, + "step": 882 + }, + { + "epoch": 0.02, + "grad_norm": 1.163060696685913, + "learning_rate": 1.3861852433281007e-05, + "loss": 0.9009, + "step": 883 + }, + { + "epoch": 0.02, + "grad_norm": 2.099588210864189, + "learning_rate": 1.3877551020408165e-05, + "loss": 1.1568, + "step": 884 + }, + { + "epoch": 0.02, + "grad_norm": 2.512138990856189, + "learning_rate": 1.3893249607535323e-05, + "loss": 1.0436, + "step": 885 + }, + { + "epoch": 0.02, + "grad_norm": 2.0677867559193417, + "learning_rate": 1.3908948194662482e-05, + "loss": 1.0496, + "step": 886 + }, + { + "epoch": 0.02, + "grad_norm": 2.40218682745791, + "learning_rate": 1.392464678178964e-05, + "loss": 1.14, + "step": 887 + }, + { + "epoch": 0.02, + "grad_norm": 2.190785984176926, + "learning_rate": 1.3940345368916798e-05, + "loss": 1.0942, + "step": 888 + }, + { + "epoch": 0.02, + "grad_norm": 2.685780312189392, + "learning_rate": 1.3956043956043957e-05, + "loss": 1.0386, + "step": 889 + }, + { + "epoch": 0.02, + "grad_norm": 1.9222044917939611, + "learning_rate": 1.3971742543171116e-05, + "loss": 0.9975, + "step": 890 + }, + { + "epoch": 0.02, + "grad_norm": 2.0616416391854413, + "learning_rate": 1.3987441130298273e-05, + "loss": 1.0838, + "step": 891 + }, + { + "epoch": 0.02, + "grad_norm": 2.2110952027896813, + "learning_rate": 1.4003139717425434e-05, + "loss": 1.0456, + "step": 892 + }, + { + "epoch": 0.02, + "grad_norm": 2.171942359708213, + "learning_rate": 1.4018838304552591e-05, + "loss": 1.1886, + "step": 893 + }, + { + "epoch": 0.02, + "grad_norm": 2.1815666876270687, + "learning_rate": 1.403453689167975e-05, + "loss": 1.064, + "step": 894 + }, + { + "epoch": 0.02, + "grad_norm": 2.0870093587516894, + "learning_rate": 1.4050235478806909e-05, + "loss": 1.0286, + "step": 895 + }, + { + "epoch": 0.02, + "grad_norm": 2.193998909131481, + "learning_rate": 1.4065934065934068e-05, + "loss": 0.9622, + "step": 896 + }, + { + "epoch": 0.02, + "grad_norm": 6.932062075824738, + "learning_rate": 1.4081632653061225e-05, + "loss": 1.0599, + "step": 897 + }, + { + "epoch": 0.02, + "grad_norm": 2.191753709139767, + "learning_rate": 1.4097331240188384e-05, + "loss": 1.0417, + "step": 898 + }, + { + "epoch": 0.02, + "grad_norm": 2.1940748957527085, + "learning_rate": 1.4113029827315543e-05, + "loss": 1.156, + "step": 899 + }, + { + "epoch": 0.02, + "grad_norm": 2.078154924005163, + "learning_rate": 1.41287284144427e-05, + "loss": 1.1596, + "step": 900 + }, + { + "epoch": 0.02, + "grad_norm": 3.0321442516107777, + "learning_rate": 1.414442700156986e-05, + "loss": 1.0771, + "step": 901 + }, + { + "epoch": 0.02, + "grad_norm": 2.2221986770184508, + "learning_rate": 1.4160125588697018e-05, + "loss": 1.032, + "step": 902 + }, + { + "epoch": 0.02, + "grad_norm": 2.3389015976678107, + "learning_rate": 1.4175824175824178e-05, + "loss": 1.0721, + "step": 903 + }, + { + "epoch": 0.02, + "grad_norm": 1.0862548619741168, + "learning_rate": 1.4191522762951335e-05, + "loss": 0.9939, + "step": 904 + }, + { + "epoch": 0.02, + "grad_norm": 2.1740640687225734, + "learning_rate": 1.4207221350078494e-05, + "loss": 1.117, + "step": 905 + }, + { + "epoch": 0.02, + "grad_norm": 2.2451984760822223, + "learning_rate": 1.4222919937205653e-05, + "loss": 1.0373, + "step": 906 + }, + { + "epoch": 0.02, + "grad_norm": 2.3025928799438264, + "learning_rate": 1.423861852433281e-05, + "loss": 1.0891, + "step": 907 + }, + { + "epoch": 0.02, + "grad_norm": 2.0057479020468443, + "learning_rate": 1.425431711145997e-05, + "loss": 1.1612, + "step": 908 + }, + { + "epoch": 0.02, + "grad_norm": 1.9041932240423134, + "learning_rate": 1.4270015698587128e-05, + "loss": 0.9401, + "step": 909 + }, + { + "epoch": 0.02, + "grad_norm": 2.2463976460616317, + "learning_rate": 1.4285714285714287e-05, + "loss": 1.0895, + "step": 910 + }, + { + "epoch": 0.02, + "grad_norm": 1.9711142483063824, + "learning_rate": 1.4301412872841444e-05, + "loss": 0.996, + "step": 911 + }, + { + "epoch": 0.02, + "grad_norm": 2.2883102326492564, + "learning_rate": 1.4317111459968605e-05, + "loss": 0.9138, + "step": 912 + }, + { + "epoch": 0.02, + "grad_norm": 2.154554094701734, + "learning_rate": 1.4332810047095762e-05, + "loss": 1.0535, + "step": 913 + }, + { + "epoch": 0.02, + "grad_norm": 1.8585499663124532, + "learning_rate": 1.4348508634222923e-05, + "loss": 0.8364, + "step": 914 + }, + { + "epoch": 0.02, + "grad_norm": 1.9766206156179174, + "learning_rate": 1.436420722135008e-05, + "loss": 1.0084, + "step": 915 + }, + { + "epoch": 0.02, + "grad_norm": 2.2121638400161037, + "learning_rate": 1.4379905808477239e-05, + "loss": 0.9567, + "step": 916 + }, + { + "epoch": 0.02, + "grad_norm": 2.059025845678127, + "learning_rate": 1.4395604395604396e-05, + "loss": 0.9556, + "step": 917 + }, + { + "epoch": 0.02, + "grad_norm": 2.7088045708269237, + "learning_rate": 1.4411302982731555e-05, + "loss": 1.1148, + "step": 918 + }, + { + "epoch": 0.02, + "grad_norm": 2.1272271649052126, + "learning_rate": 1.4427001569858714e-05, + "loss": 1.0574, + "step": 919 + }, + { + "epoch": 0.02, + "grad_norm": 2.2835976825287756, + "learning_rate": 1.4442700156985871e-05, + "loss": 1.1046, + "step": 920 + }, + { + "epoch": 0.02, + "grad_norm": 2.1233322401190886, + "learning_rate": 1.4458398744113031e-05, + "loss": 1.1968, + "step": 921 + }, + { + "epoch": 0.02, + "grad_norm": 2.1139909426067174, + "learning_rate": 1.4474097331240189e-05, + "loss": 1.0216, + "step": 922 + }, + { + "epoch": 0.02, + "grad_norm": 1.992796944155056, + "learning_rate": 1.448979591836735e-05, + "loss": 1.0747, + "step": 923 + }, + { + "epoch": 0.02, + "grad_norm": 2.135891815808333, + "learning_rate": 1.4505494505494506e-05, + "loss": 1.1634, + "step": 924 + }, + { + "epoch": 0.02, + "grad_norm": 2.295393532719372, + "learning_rate": 1.4521193092621665e-05, + "loss": 1.0663, + "step": 925 + }, + { + "epoch": 0.02, + "grad_norm": 2.0330472461794797, + "learning_rate": 1.4536891679748824e-05, + "loss": 1.1131, + "step": 926 + }, + { + "epoch": 0.02, + "grad_norm": 1.1424184351369533, + "learning_rate": 1.4552590266875983e-05, + "loss": 0.948, + "step": 927 + }, + { + "epoch": 0.02, + "grad_norm": 2.1752391089248237, + "learning_rate": 1.456828885400314e-05, + "loss": 1.0407, + "step": 928 + }, + { + "epoch": 0.02, + "grad_norm": 1.1389801747362185, + "learning_rate": 1.45839874411303e-05, + "loss": 0.9111, + "step": 929 + }, + { + "epoch": 0.02, + "grad_norm": 2.7383956552820874, + "learning_rate": 1.4599686028257458e-05, + "loss": 0.9842, + "step": 930 + }, + { + "epoch": 0.02, + "grad_norm": 2.1580777617871263, + "learning_rate": 1.4615384615384615e-05, + "loss": 1.0505, + "step": 931 + }, + { + "epoch": 0.02, + "grad_norm": 2.2736426340585236, + "learning_rate": 1.4631083202511776e-05, + "loss": 1.1466, + "step": 932 + }, + { + "epoch": 0.02, + "grad_norm": 2.017452817027823, + "learning_rate": 1.4646781789638933e-05, + "loss": 1.1311, + "step": 933 + }, + { + "epoch": 0.02, + "grad_norm": 2.104762853400645, + "learning_rate": 1.4662480376766094e-05, + "loss": 1.1609, + "step": 934 + }, + { + "epoch": 0.02, + "grad_norm": 1.1346328857970918, + "learning_rate": 1.467817896389325e-05, + "loss": 0.8851, + "step": 935 + }, + { + "epoch": 0.02, + "grad_norm": 2.264027509367937, + "learning_rate": 1.469387755102041e-05, + "loss": 1.0667, + "step": 936 + }, + { + "epoch": 0.02, + "grad_norm": 2.081702825988145, + "learning_rate": 1.4709576138147567e-05, + "loss": 1.0834, + "step": 937 + }, + { + "epoch": 0.02, + "grad_norm": 1.1210001395932054, + "learning_rate": 1.4725274725274727e-05, + "loss": 1.0127, + "step": 938 + }, + { + "epoch": 0.02, + "grad_norm": 2.3198535981964543, + "learning_rate": 1.4740973312401885e-05, + "loss": 1.0741, + "step": 939 + }, + { + "epoch": 0.02, + "grad_norm": 1.944403741076784, + "learning_rate": 1.4756671899529042e-05, + "loss": 1.0012, + "step": 940 + }, + { + "epoch": 0.02, + "grad_norm": 1.2103940888484366, + "learning_rate": 1.4772370486656202e-05, + "loss": 1.0451, + "step": 941 + }, + { + "epoch": 0.02, + "grad_norm": 2.289058220870735, + "learning_rate": 1.478806907378336e-05, + "loss": 1.0785, + "step": 942 + }, + { + "epoch": 0.02, + "grad_norm": 1.2580611028801258, + "learning_rate": 1.480376766091052e-05, + "loss": 0.8419, + "step": 943 + }, + { + "epoch": 0.02, + "grad_norm": 2.04916694921862, + "learning_rate": 1.4819466248037677e-05, + "loss": 1.0668, + "step": 944 + }, + { + "epoch": 0.02, + "grad_norm": 2.1552817104035635, + "learning_rate": 1.4835164835164836e-05, + "loss": 0.994, + "step": 945 + }, + { + "epoch": 0.02, + "grad_norm": 2.3301139287803836, + "learning_rate": 1.4850863422291995e-05, + "loss": 1.05, + "step": 946 + }, + { + "epoch": 0.02, + "grad_norm": 2.148232892979023, + "learning_rate": 1.4866562009419154e-05, + "loss": 1.0368, + "step": 947 + }, + { + "epoch": 0.02, + "grad_norm": 2.322935893138868, + "learning_rate": 1.4882260596546311e-05, + "loss": 1.1941, + "step": 948 + }, + { + "epoch": 0.02, + "grad_norm": 2.2510933796076795, + "learning_rate": 1.4897959183673472e-05, + "loss": 0.9941, + "step": 949 + }, + { + "epoch": 0.02, + "grad_norm": 2.832065671884234, + "learning_rate": 1.4913657770800629e-05, + "loss": 0.9517, + "step": 950 + }, + { + "epoch": 0.02, + "grad_norm": 1.1334850292170269, + "learning_rate": 1.4929356357927786e-05, + "loss": 0.9888, + "step": 951 + }, + { + "epoch": 0.02, + "grad_norm": 2.142659110479108, + "learning_rate": 1.4945054945054947e-05, + "loss": 1.0916, + "step": 952 + }, + { + "epoch": 0.02, + "grad_norm": 2.2513480519149764, + "learning_rate": 1.4960753532182104e-05, + "loss": 1.075, + "step": 953 + }, + { + "epoch": 0.02, + "grad_norm": 2.1244457170399516, + "learning_rate": 1.4976452119309265e-05, + "loss": 1.1422, + "step": 954 + }, + { + "epoch": 0.02, + "grad_norm": 2.403327524045706, + "learning_rate": 1.4992150706436422e-05, + "loss": 1.1236, + "step": 955 + }, + { + "epoch": 0.02, + "grad_norm": 2.1746543623346466, + "learning_rate": 1.500784929356358e-05, + "loss": 1.0773, + "step": 956 + }, + { + "epoch": 0.02, + "grad_norm": 2.0996801224699024, + "learning_rate": 1.5023547880690738e-05, + "loss": 0.9854, + "step": 957 + }, + { + "epoch": 0.02, + "grad_norm": 2.0334200545235412, + "learning_rate": 1.5039246467817898e-05, + "loss": 1.1026, + "step": 958 + }, + { + "epoch": 0.02, + "grad_norm": 2.0226340590482996, + "learning_rate": 1.5054945054945056e-05, + "loss": 1.0447, + "step": 959 + }, + { + "epoch": 0.02, + "grad_norm": 2.0551553970118133, + "learning_rate": 1.5070643642072216e-05, + "loss": 1.1543, + "step": 960 + }, + { + "epoch": 0.02, + "grad_norm": 2.101082310006078, + "learning_rate": 1.5086342229199373e-05, + "loss": 1.1325, + "step": 961 + }, + { + "epoch": 0.02, + "grad_norm": 2.489701653654552, + "learning_rate": 1.510204081632653e-05, + "loss": 1.0568, + "step": 962 + }, + { + "epoch": 0.02, + "grad_norm": 2.240262397738014, + "learning_rate": 1.5117739403453691e-05, + "loss": 1.1982, + "step": 963 + }, + { + "epoch": 0.02, + "grad_norm": 2.2306338161895907, + "learning_rate": 1.5133437990580848e-05, + "loss": 1.0029, + "step": 964 + }, + { + "epoch": 0.02, + "grad_norm": 1.170863098322321, + "learning_rate": 1.5149136577708007e-05, + "loss": 0.9571, + "step": 965 + }, + { + "epoch": 0.02, + "grad_norm": 1.1181659776554613, + "learning_rate": 1.5164835164835166e-05, + "loss": 0.9341, + "step": 966 + }, + { + "epoch": 0.02, + "grad_norm": 2.1113731100925603, + "learning_rate": 1.5180533751962325e-05, + "loss": 1.002, + "step": 967 + }, + { + "epoch": 0.02, + "grad_norm": 2.453258356927402, + "learning_rate": 1.5196232339089482e-05, + "loss": 1.0394, + "step": 968 + }, + { + "epoch": 0.02, + "grad_norm": 2.31377057078818, + "learning_rate": 1.5211930926216643e-05, + "loss": 0.9848, + "step": 969 + }, + { + "epoch": 0.02, + "grad_norm": 1.993994995202197, + "learning_rate": 1.52276295133438e-05, + "loss": 0.9161, + "step": 970 + }, + { + "epoch": 0.02, + "grad_norm": 1.1432035845375843, + "learning_rate": 1.524332810047096e-05, + "loss": 0.9962, + "step": 971 + }, + { + "epoch": 0.02, + "grad_norm": 2.0089006129870977, + "learning_rate": 1.5259026687598116e-05, + "loss": 1.0074, + "step": 972 + }, + { + "epoch": 0.02, + "grad_norm": 1.971344312027117, + "learning_rate": 1.5274725274725277e-05, + "loss": 1.0273, + "step": 973 + }, + { + "epoch": 0.02, + "grad_norm": 2.2710117911831986, + "learning_rate": 1.5290423861852434e-05, + "loss": 1.1396, + "step": 974 + }, + { + "epoch": 0.02, + "grad_norm": 2.213132543294457, + "learning_rate": 1.530612244897959e-05, + "loss": 1.0356, + "step": 975 + }, + { + "epoch": 0.02, + "grad_norm": 7.308518696694729, + "learning_rate": 1.532182103610675e-05, + "loss": 1.1037, + "step": 976 + }, + { + "epoch": 0.02, + "grad_norm": 2.2561680952013603, + "learning_rate": 1.533751962323391e-05, + "loss": 0.9149, + "step": 977 + }, + { + "epoch": 0.02, + "grad_norm": 1.182199244039412, + "learning_rate": 1.535321821036107e-05, + "loss": 1.0389, + "step": 978 + }, + { + "epoch": 0.02, + "grad_norm": 2.66673172243174, + "learning_rate": 1.5368916797488227e-05, + "loss": 1.0695, + "step": 979 + }, + { + "epoch": 0.02, + "grad_norm": 2.542607076623358, + "learning_rate": 1.5384615384615387e-05, + "loss": 1.1216, + "step": 980 + }, + { + "epoch": 0.02, + "grad_norm": 2.5817353483742482, + "learning_rate": 1.5400313971742544e-05, + "loss": 1.1532, + "step": 981 + }, + { + "epoch": 0.02, + "grad_norm": 2.061560339080131, + "learning_rate": 1.5416012558869705e-05, + "loss": 1.0002, + "step": 982 + }, + { + "epoch": 0.02, + "grad_norm": 2.0379887884672137, + "learning_rate": 1.5431711145996862e-05, + "loss": 1.0133, + "step": 983 + }, + { + "epoch": 0.02, + "grad_norm": 1.160588048243416, + "learning_rate": 1.544740973312402e-05, + "loss": 0.9363, + "step": 984 + }, + { + "epoch": 0.02, + "grad_norm": 2.2046330505817844, + "learning_rate": 1.546310832025118e-05, + "loss": 1.132, + "step": 985 + }, + { + "epoch": 0.02, + "grad_norm": 2.16983637740033, + "learning_rate": 1.5478806907378337e-05, + "loss": 1.07, + "step": 986 + }, + { + "epoch": 0.02, + "grad_norm": 2.2419226602992275, + "learning_rate": 1.5494505494505498e-05, + "loss": 1.1141, + "step": 987 + }, + { + "epoch": 0.02, + "grad_norm": 2.307492284768501, + "learning_rate": 1.5510204081632655e-05, + "loss": 0.9576, + "step": 988 + }, + { + "epoch": 0.02, + "grad_norm": 2.0965567579642457, + "learning_rate": 1.5525902668759812e-05, + "loss": 1.0674, + "step": 989 + }, + { + "epoch": 0.02, + "grad_norm": 2.322584131181688, + "learning_rate": 1.5541601255886973e-05, + "loss": 0.9997, + "step": 990 + }, + { + "epoch": 0.02, + "grad_norm": 2.3348135448954803, + "learning_rate": 1.555729984301413e-05, + "loss": 1.0506, + "step": 991 + }, + { + "epoch": 0.02, + "grad_norm": 1.173111147585373, + "learning_rate": 1.5572998430141287e-05, + "loss": 0.9049, + "step": 992 + }, + { + "epoch": 0.02, + "grad_norm": 2.4200581207437484, + "learning_rate": 1.5588697017268448e-05, + "loss": 0.9865, + "step": 993 + }, + { + "epoch": 0.02, + "grad_norm": 2.031625317037022, + "learning_rate": 1.5604395604395605e-05, + "loss": 0.9995, + "step": 994 + }, + { + "epoch": 0.02, + "grad_norm": 2.151802144072626, + "learning_rate": 1.5620094191522762e-05, + "loss": 0.9936, + "step": 995 + }, + { + "epoch": 0.02, + "grad_norm": 2.2334136978006693, + "learning_rate": 1.5635792778649923e-05, + "loss": 1.0789, + "step": 996 + }, + { + "epoch": 0.02, + "grad_norm": 2.260737583493515, + "learning_rate": 1.565149136577708e-05, + "loss": 1.057, + "step": 997 + }, + { + "epoch": 0.02, + "grad_norm": 2.788833107225131, + "learning_rate": 1.566718995290424e-05, + "loss": 1.0743, + "step": 998 + }, + { + "epoch": 0.02, + "grad_norm": 1.1572208680040952, + "learning_rate": 1.5682888540031398e-05, + "loss": 1.0113, + "step": 999 + }, + { + "epoch": 0.02, + "grad_norm": 2.2652997312665777, + "learning_rate": 1.5698587127158558e-05, + "loss": 1.1129, + "step": 1000 + }, + { + "epoch": 0.02, + "grad_norm": 2.238913065657337, + "learning_rate": 1.5714285714285715e-05, + "loss": 1.0245, + "step": 1001 + }, + { + "epoch": 0.02, + "grad_norm": 3.174545781280836, + "learning_rate": 1.5729984301412876e-05, + "loss": 1.0719, + "step": 1002 + }, + { + "epoch": 0.02, + "grad_norm": 1.0626511056560537, + "learning_rate": 1.5745682888540033e-05, + "loss": 0.9717, + "step": 1003 + }, + { + "epoch": 0.02, + "grad_norm": 2.392663830273092, + "learning_rate": 1.5761381475667194e-05, + "loss": 1.0222, + "step": 1004 + }, + { + "epoch": 0.02, + "grad_norm": 1.1533345418225076, + "learning_rate": 1.577708006279435e-05, + "loss": 0.9878, + "step": 1005 + }, + { + "epoch": 0.02, + "grad_norm": 2.1140343614621586, + "learning_rate": 1.5792778649921508e-05, + "loss": 0.9977, + "step": 1006 + }, + { + "epoch": 0.02, + "grad_norm": 2.078197530192415, + "learning_rate": 1.580847723704867e-05, + "loss": 1.0099, + "step": 1007 + }, + { + "epoch": 0.02, + "grad_norm": 2.2716484536396746, + "learning_rate": 1.5824175824175826e-05, + "loss": 1.2378, + "step": 1008 + }, + { + "epoch": 0.02, + "grad_norm": 2.228305969512687, + "learning_rate": 1.5839874411302983e-05, + "loss": 1.0199, + "step": 1009 + }, + { + "epoch": 0.02, + "grad_norm": 2.468519913840236, + "learning_rate": 1.5855572998430144e-05, + "loss": 1.1123, + "step": 1010 + }, + { + "epoch": 0.02, + "grad_norm": 1.9966529823439245, + "learning_rate": 1.58712715855573e-05, + "loss": 1.1357, + "step": 1011 + }, + { + "epoch": 0.02, + "grad_norm": 2.0171348619389295, + "learning_rate": 1.5886970172684458e-05, + "loss": 1.0079, + "step": 1012 + }, + { + "epoch": 0.02, + "grad_norm": 2.24426684043321, + "learning_rate": 1.590266875981162e-05, + "loss": 1.0211, + "step": 1013 + }, + { + "epoch": 0.02, + "grad_norm": 2.236807456557601, + "learning_rate": 1.5918367346938776e-05, + "loss": 1.0928, + "step": 1014 + }, + { + "epoch": 0.02, + "grad_norm": 2.6997024325143437, + "learning_rate": 1.5934065934065933e-05, + "loss": 1.0217, + "step": 1015 + }, + { + "epoch": 0.02, + "grad_norm": 2.1545389739505074, + "learning_rate": 1.5949764521193094e-05, + "loss": 0.9611, + "step": 1016 + }, + { + "epoch": 0.02, + "grad_norm": 2.0967856135215706, + "learning_rate": 1.596546310832025e-05, + "loss": 1.1381, + "step": 1017 + }, + { + "epoch": 0.02, + "grad_norm": 2.67686796579279, + "learning_rate": 1.598116169544741e-05, + "loss": 1.1003, + "step": 1018 + }, + { + "epoch": 0.02, + "grad_norm": 2.108027891423765, + "learning_rate": 1.599686028257457e-05, + "loss": 1.229, + "step": 1019 + }, + { + "epoch": 0.02, + "grad_norm": 2.3047640896856687, + "learning_rate": 1.601255886970173e-05, + "loss": 1.1365, + "step": 1020 + }, + { + "epoch": 0.02, + "grad_norm": 2.0669157828931097, + "learning_rate": 1.6028257456828886e-05, + "loss": 1.0904, + "step": 1021 + }, + { + "epoch": 0.02, + "grad_norm": 1.8647791579122772, + "learning_rate": 1.6043956043956047e-05, + "loss": 0.9957, + "step": 1022 + }, + { + "epoch": 0.02, + "grad_norm": 2.176451281532121, + "learning_rate": 1.6059654631083204e-05, + "loss": 1.0744, + "step": 1023 + }, + { + "epoch": 0.02, + "grad_norm": 2.3700108312714105, + "learning_rate": 1.6075353218210365e-05, + "loss": 1.1258, + "step": 1024 + }, + { + "epoch": 0.02, + "grad_norm": 2.2095344893618267, + "learning_rate": 1.6091051805337522e-05, + "loss": 1.0757, + "step": 1025 + }, + { + "epoch": 0.02, + "grad_norm": 2.3458312501465444, + "learning_rate": 1.610675039246468e-05, + "loss": 1.1558, + "step": 1026 + }, + { + "epoch": 0.02, + "grad_norm": 2.4333109523431102, + "learning_rate": 1.612244897959184e-05, + "loss": 1.1336, + "step": 1027 + }, + { + "epoch": 0.02, + "grad_norm": 2.0786860701702494, + "learning_rate": 1.6138147566718997e-05, + "loss": 1.0854, + "step": 1028 + }, + { + "epoch": 0.02, + "grad_norm": 1.9461065040217924, + "learning_rate": 1.6153846153846154e-05, + "loss": 1.0756, + "step": 1029 + }, + { + "epoch": 0.02, + "grad_norm": 2.0706898918293013, + "learning_rate": 1.6169544740973315e-05, + "loss": 1.0645, + "step": 1030 + }, + { + "epoch": 0.02, + "grad_norm": 2.1572858734059213, + "learning_rate": 1.6185243328100472e-05, + "loss": 1.0728, + "step": 1031 + }, + { + "epoch": 0.02, + "grad_norm": 2.071851982423785, + "learning_rate": 1.620094191522763e-05, + "loss": 1.0832, + "step": 1032 + }, + { + "epoch": 0.02, + "grad_norm": 2.035362619418128, + "learning_rate": 1.621664050235479e-05, + "loss": 1.0708, + "step": 1033 + }, + { + "epoch": 0.02, + "grad_norm": 2.1486633995893403, + "learning_rate": 1.6232339089481947e-05, + "loss": 1.0891, + "step": 1034 + }, + { + "epoch": 0.02, + "grad_norm": 1.2440115347600451, + "learning_rate": 1.6248037676609107e-05, + "loss": 0.9386, + "step": 1035 + }, + { + "epoch": 0.02, + "grad_norm": 2.2480029889818467, + "learning_rate": 1.6263736263736265e-05, + "loss": 1.0494, + "step": 1036 + }, + { + "epoch": 0.02, + "grad_norm": 2.0278961302172247, + "learning_rate": 1.6279434850863422e-05, + "loss": 1.0048, + "step": 1037 + }, + { + "epoch": 0.02, + "grad_norm": 2.3738685056783773, + "learning_rate": 1.6295133437990582e-05, + "loss": 1.03, + "step": 1038 + }, + { + "epoch": 0.02, + "grad_norm": 2.428867654840401, + "learning_rate": 1.631083202511774e-05, + "loss": 1.0511, + "step": 1039 + }, + { + "epoch": 0.02, + "grad_norm": 2.6479435704702285, + "learning_rate": 1.63265306122449e-05, + "loss": 1.0578, + "step": 1040 + }, + { + "epoch": 0.02, + "grad_norm": 2.0233706052227602, + "learning_rate": 1.6342229199372057e-05, + "loss": 1.0485, + "step": 1041 + }, + { + "epoch": 0.02, + "grad_norm": 2.1344102077631186, + "learning_rate": 1.6357927786499218e-05, + "loss": 1.0585, + "step": 1042 + }, + { + "epoch": 0.02, + "grad_norm": 2.3876809732127477, + "learning_rate": 1.6373626373626375e-05, + "loss": 1.074, + "step": 1043 + }, + { + "epoch": 0.02, + "grad_norm": 1.9998332042311349, + "learning_rate": 1.6389324960753536e-05, + "loss": 1.0548, + "step": 1044 + }, + { + "epoch": 0.02, + "grad_norm": 2.073704818412457, + "learning_rate": 1.6405023547880693e-05, + "loss": 1.1121, + "step": 1045 + }, + { + "epoch": 0.02, + "grad_norm": 1.955692678653819, + "learning_rate": 1.642072213500785e-05, + "loss": 1.1676, + "step": 1046 + }, + { + "epoch": 0.02, + "grad_norm": 2.2399354391645954, + "learning_rate": 1.643642072213501e-05, + "loss": 1.1507, + "step": 1047 + }, + { + "epoch": 0.02, + "grad_norm": 2.20957819670784, + "learning_rate": 1.6452119309262168e-05, + "loss": 1.0554, + "step": 1048 + }, + { + "epoch": 0.02, + "grad_norm": 2.093317739080999, + "learning_rate": 1.6467817896389325e-05, + "loss": 1.0219, + "step": 1049 + }, + { + "epoch": 0.02, + "grad_norm": 1.9057079585421892, + "learning_rate": 1.6483516483516486e-05, + "loss": 1.0557, + "step": 1050 + }, + { + "epoch": 0.02, + "grad_norm": 2.2049712053488135, + "learning_rate": 1.6499215070643643e-05, + "loss": 1.0573, + "step": 1051 + }, + { + "epoch": 0.02, + "grad_norm": 2.5505135698419865, + "learning_rate": 1.65149136577708e-05, + "loss": 1.1808, + "step": 1052 + }, + { + "epoch": 0.02, + "grad_norm": 2.8332197773843224, + "learning_rate": 1.653061224489796e-05, + "loss": 0.9717, + "step": 1053 + }, + { + "epoch": 0.02, + "grad_norm": 2.335007271512359, + "learning_rate": 1.6546310832025118e-05, + "loss": 0.9658, + "step": 1054 + }, + { + "epoch": 0.02, + "grad_norm": 2.0785972641867003, + "learning_rate": 1.6562009419152278e-05, + "loss": 1.119, + "step": 1055 + }, + { + "epoch": 0.02, + "grad_norm": 1.9984531986343803, + "learning_rate": 1.6577708006279435e-05, + "loss": 1.0192, + "step": 1056 + }, + { + "epoch": 0.02, + "grad_norm": 1.1360228760343403, + "learning_rate": 1.6593406593406596e-05, + "loss": 0.9456, + "step": 1057 + }, + { + "epoch": 0.02, + "grad_norm": 2.232237111624172, + "learning_rate": 1.6609105180533753e-05, + "loss": 1.0597, + "step": 1058 + }, + { + "epoch": 0.02, + "grad_norm": 2.044855692848191, + "learning_rate": 1.662480376766091e-05, + "loss": 1.0015, + "step": 1059 + }, + { + "epoch": 0.02, + "grad_norm": 2.1215612210744546, + "learning_rate": 1.664050235478807e-05, + "loss": 1.1016, + "step": 1060 + }, + { + "epoch": 0.02, + "grad_norm": 2.1320419786430724, + "learning_rate": 1.6656200941915228e-05, + "loss": 1.0597, + "step": 1061 + }, + { + "epoch": 0.03, + "grad_norm": 1.9276883502440167, + "learning_rate": 1.667189952904239e-05, + "loss": 1.1089, + "step": 1062 + }, + { + "epoch": 0.03, + "grad_norm": 2.2026158303082366, + "learning_rate": 1.6687598116169546e-05, + "loss": 1.1113, + "step": 1063 + }, + { + "epoch": 0.03, + "grad_norm": 2.1498034933403485, + "learning_rate": 1.6703296703296707e-05, + "loss": 1.2138, + "step": 1064 + }, + { + "epoch": 0.03, + "grad_norm": 1.113617502010469, + "learning_rate": 1.6718995290423864e-05, + "loss": 0.9722, + "step": 1065 + }, + { + "epoch": 0.03, + "grad_norm": 2.2581694546223234, + "learning_rate": 1.673469387755102e-05, + "loss": 0.9327, + "step": 1066 + }, + { + "epoch": 0.03, + "grad_norm": 2.0248789273111476, + "learning_rate": 1.675039246467818e-05, + "loss": 0.9662, + "step": 1067 + }, + { + "epoch": 0.03, + "grad_norm": 2.132827759716967, + "learning_rate": 1.676609105180534e-05, + "loss": 0.864, + "step": 1068 + }, + { + "epoch": 0.03, + "grad_norm": 1.2209389422567194, + "learning_rate": 1.6781789638932496e-05, + "loss": 1.0305, + "step": 1069 + }, + { + "epoch": 0.03, + "grad_norm": 2.0713856148281846, + "learning_rate": 1.6797488226059656e-05, + "loss": 0.9662, + "step": 1070 + }, + { + "epoch": 0.03, + "grad_norm": 2.1851436670886892, + "learning_rate": 1.6813186813186814e-05, + "loss": 1.014, + "step": 1071 + }, + { + "epoch": 0.03, + "grad_norm": 2.2710425602406326, + "learning_rate": 1.682888540031397e-05, + "loss": 1.0623, + "step": 1072 + }, + { + "epoch": 0.03, + "grad_norm": 2.647763944106019, + "learning_rate": 1.684458398744113e-05, + "loss": 1.029, + "step": 1073 + }, + { + "epoch": 0.03, + "grad_norm": 2.35636441370145, + "learning_rate": 1.686028257456829e-05, + "loss": 1.1086, + "step": 1074 + }, + { + "epoch": 0.03, + "grad_norm": 2.0847496806775654, + "learning_rate": 1.687598116169545e-05, + "loss": 1.0589, + "step": 1075 + }, + { + "epoch": 0.03, + "grad_norm": 2.0459062165314994, + "learning_rate": 1.6891679748822606e-05, + "loss": 0.9185, + "step": 1076 + }, + { + "epoch": 0.03, + "grad_norm": 1.9240277430041155, + "learning_rate": 1.6907378335949767e-05, + "loss": 1.0066, + "step": 1077 + }, + { + "epoch": 0.03, + "grad_norm": 2.2219048625416202, + "learning_rate": 1.6923076923076924e-05, + "loss": 0.999, + "step": 1078 + }, + { + "epoch": 0.03, + "grad_norm": 2.274030173272554, + "learning_rate": 1.6938775510204085e-05, + "loss": 1.002, + "step": 1079 + }, + { + "epoch": 0.03, + "grad_norm": 2.099276501084259, + "learning_rate": 1.6954474097331242e-05, + "loss": 1.0294, + "step": 1080 + }, + { + "epoch": 0.03, + "grad_norm": 2.219125229460298, + "learning_rate": 1.69701726844584e-05, + "loss": 1.1112, + "step": 1081 + }, + { + "epoch": 0.03, + "grad_norm": 1.2238812039406306, + "learning_rate": 1.698587127158556e-05, + "loss": 1.0236, + "step": 1082 + }, + { + "epoch": 0.03, + "grad_norm": 2.3503212649014333, + "learning_rate": 1.7001569858712717e-05, + "loss": 1.1512, + "step": 1083 + }, + { + "epoch": 0.03, + "grad_norm": 2.1076781566274163, + "learning_rate": 1.7017268445839878e-05, + "loss": 1.0346, + "step": 1084 + }, + { + "epoch": 0.03, + "grad_norm": 2.209132286114992, + "learning_rate": 1.7032967032967035e-05, + "loss": 0.9953, + "step": 1085 + }, + { + "epoch": 0.03, + "grad_norm": 1.165439662344062, + "learning_rate": 1.7048665620094192e-05, + "loss": 0.9663, + "step": 1086 + }, + { + "epoch": 0.03, + "grad_norm": 4.853950399511229, + "learning_rate": 1.7064364207221352e-05, + "loss": 0.9485, + "step": 1087 + }, + { + "epoch": 0.03, + "grad_norm": 2.364957436198004, + "learning_rate": 1.708006279434851e-05, + "loss": 1.116, + "step": 1088 + }, + { + "epoch": 0.03, + "grad_norm": 2.0404494202927546, + "learning_rate": 1.7095761381475667e-05, + "loss": 0.9881, + "step": 1089 + }, + { + "epoch": 0.03, + "grad_norm": 2.1372808768744767, + "learning_rate": 1.7111459968602827e-05, + "loss": 1.0538, + "step": 1090 + }, + { + "epoch": 0.03, + "grad_norm": 1.8767823603446177, + "learning_rate": 1.7127158555729985e-05, + "loss": 0.9938, + "step": 1091 + }, + { + "epoch": 0.03, + "grad_norm": 2.0437168926873426, + "learning_rate": 1.7142857142857142e-05, + "loss": 1.0923, + "step": 1092 + }, + { + "epoch": 0.03, + "grad_norm": 2.009722629607147, + "learning_rate": 1.7158555729984302e-05, + "loss": 1.1195, + "step": 1093 + }, + { + "epoch": 0.03, + "grad_norm": 1.913020986563892, + "learning_rate": 1.717425431711146e-05, + "loss": 0.973, + "step": 1094 + }, + { + "epoch": 0.03, + "grad_norm": 2.4210309634417837, + "learning_rate": 1.718995290423862e-05, + "loss": 1.0507, + "step": 1095 + }, + { + "epoch": 0.03, + "grad_norm": 2.201607740340456, + "learning_rate": 1.7205651491365777e-05, + "loss": 1.1776, + "step": 1096 + }, + { + "epoch": 0.03, + "grad_norm": 2.083385740852809, + "learning_rate": 1.7221350078492938e-05, + "loss": 1.0544, + "step": 1097 + }, + { + "epoch": 0.03, + "grad_norm": 2.2877255739866316, + "learning_rate": 1.7237048665620095e-05, + "loss": 1.129, + "step": 1098 + }, + { + "epoch": 0.03, + "grad_norm": 2.1361791093374407, + "learning_rate": 1.7252747252747256e-05, + "loss": 1.1269, + "step": 1099 + }, + { + "epoch": 0.03, + "grad_norm": 1.9365468526358567, + "learning_rate": 1.7268445839874413e-05, + "loss": 1.2329, + "step": 1100 + }, + { + "epoch": 0.03, + "grad_norm": 2.283352274098108, + "learning_rate": 1.7284144427001574e-05, + "loss": 1.061, + "step": 1101 + }, + { + "epoch": 0.03, + "grad_norm": 1.9774698775163035, + "learning_rate": 1.729984301412873e-05, + "loss": 1.0521, + "step": 1102 + }, + { + "epoch": 0.03, + "grad_norm": 2.1233437264483306, + "learning_rate": 1.7315541601255888e-05, + "loss": 0.9886, + "step": 1103 + }, + { + "epoch": 0.03, + "grad_norm": 1.971810142370792, + "learning_rate": 1.733124018838305e-05, + "loss": 1.221, + "step": 1104 + }, + { + "epoch": 0.03, + "grad_norm": 2.7045582807165744, + "learning_rate": 1.7346938775510206e-05, + "loss": 1.1099, + "step": 1105 + }, + { + "epoch": 0.03, + "grad_norm": 1.1076381058484215, + "learning_rate": 1.7362637362637363e-05, + "loss": 0.9269, + "step": 1106 + }, + { + "epoch": 0.03, + "grad_norm": 2.1163954299009102, + "learning_rate": 1.7378335949764523e-05, + "loss": 1.0361, + "step": 1107 + }, + { + "epoch": 0.03, + "grad_norm": 2.8273730313160925, + "learning_rate": 1.739403453689168e-05, + "loss": 1.1062, + "step": 1108 + }, + { + "epoch": 0.03, + "grad_norm": 3.0966275928654343, + "learning_rate": 1.7409733124018838e-05, + "loss": 1.1547, + "step": 1109 + }, + { + "epoch": 0.03, + "grad_norm": 2.314406350021274, + "learning_rate": 1.7425431711146e-05, + "loss": 1.112, + "step": 1110 + }, + { + "epoch": 0.03, + "grad_norm": 2.0741886970305883, + "learning_rate": 1.7441130298273156e-05, + "loss": 1.1086, + "step": 1111 + }, + { + "epoch": 0.03, + "grad_norm": 2.1382969897843886, + "learning_rate": 1.7456828885400316e-05, + "loss": 0.9384, + "step": 1112 + }, + { + "epoch": 0.03, + "grad_norm": 9.211352805792131, + "learning_rate": 1.7472527472527473e-05, + "loss": 1.0722, + "step": 1113 + }, + { + "epoch": 0.03, + "grad_norm": 2.1286740915770848, + "learning_rate": 1.748822605965463e-05, + "loss": 1.1258, + "step": 1114 + }, + { + "epoch": 0.03, + "grad_norm": 1.162464421695455, + "learning_rate": 1.750392464678179e-05, + "loss": 0.9869, + "step": 1115 + }, + { + "epoch": 0.03, + "grad_norm": 2.1671821580592727, + "learning_rate": 1.751962323390895e-05, + "loss": 1.2053, + "step": 1116 + }, + { + "epoch": 0.03, + "grad_norm": 2.116722039822376, + "learning_rate": 1.753532182103611e-05, + "loss": 1.0902, + "step": 1117 + }, + { + "epoch": 0.03, + "grad_norm": 2.147049949543882, + "learning_rate": 1.7551020408163266e-05, + "loss": 0.8299, + "step": 1118 + }, + { + "epoch": 0.03, + "grad_norm": 2.2096982587846568, + "learning_rate": 1.7566718995290427e-05, + "loss": 1.1075, + "step": 1119 + }, + { + "epoch": 0.03, + "grad_norm": 2.1575113780249593, + "learning_rate": 1.7582417582417584e-05, + "loss": 0.9784, + "step": 1120 + }, + { + "epoch": 0.03, + "grad_norm": 1.9491003488843097, + "learning_rate": 1.7598116169544744e-05, + "loss": 1.1675, + "step": 1121 + }, + { + "epoch": 0.03, + "grad_norm": 1.9930507300844167, + "learning_rate": 1.76138147566719e-05, + "loss": 1.0174, + "step": 1122 + }, + { + "epoch": 0.03, + "grad_norm": 1.1829866634217954, + "learning_rate": 1.762951334379906e-05, + "loss": 0.9308, + "step": 1123 + }, + { + "epoch": 0.03, + "grad_norm": 2.3483303209355153, + "learning_rate": 1.764521193092622e-05, + "loss": 1.1438, + "step": 1124 + }, + { + "epoch": 0.03, + "grad_norm": 1.065141176597575, + "learning_rate": 1.7660910518053377e-05, + "loss": 0.9572, + "step": 1125 + }, + { + "epoch": 0.03, + "grad_norm": 2.840275154918177, + "learning_rate": 1.7676609105180537e-05, + "loss": 1.139, + "step": 1126 + }, + { + "epoch": 0.03, + "grad_norm": 2.3054210531952446, + "learning_rate": 1.7692307692307694e-05, + "loss": 1.1002, + "step": 1127 + }, + { + "epoch": 0.03, + "grad_norm": 1.1162840847601607, + "learning_rate": 1.770800627943485e-05, + "loss": 0.9485, + "step": 1128 + }, + { + "epoch": 0.03, + "grad_norm": 2.1859436229541838, + "learning_rate": 1.772370486656201e-05, + "loss": 0.9617, + "step": 1129 + }, + { + "epoch": 0.03, + "grad_norm": 1.9970466929486976, + "learning_rate": 1.773940345368917e-05, + "loss": 1.0555, + "step": 1130 + }, + { + "epoch": 0.03, + "grad_norm": 2.2886257707060733, + "learning_rate": 1.7755102040816327e-05, + "loss": 1.0941, + "step": 1131 + }, + { + "epoch": 0.03, + "grad_norm": 2.766892273604456, + "learning_rate": 1.7770800627943487e-05, + "loss": 0.9553, + "step": 1132 + }, + { + "epoch": 0.03, + "grad_norm": 1.916848582963223, + "learning_rate": 1.7786499215070644e-05, + "loss": 1.1034, + "step": 1133 + }, + { + "epoch": 0.03, + "grad_norm": 2.316847028838244, + "learning_rate": 1.78021978021978e-05, + "loss": 0.9107, + "step": 1134 + }, + { + "epoch": 0.03, + "grad_norm": 2.735889293435497, + "learning_rate": 1.7817896389324962e-05, + "loss": 1.1786, + "step": 1135 + }, + { + "epoch": 0.03, + "grad_norm": 1.1539105419945621, + "learning_rate": 1.783359497645212e-05, + "loss": 0.9341, + "step": 1136 + }, + { + "epoch": 0.03, + "grad_norm": 2.984737038614325, + "learning_rate": 1.784929356357928e-05, + "loss": 1.1399, + "step": 1137 + }, + { + "epoch": 0.03, + "grad_norm": 2.3567368320739943, + "learning_rate": 1.7864992150706437e-05, + "loss": 1.0169, + "step": 1138 + }, + { + "epoch": 0.03, + "grad_norm": 2.0312892252087416, + "learning_rate": 1.7880690737833598e-05, + "loss": 1.0023, + "step": 1139 + }, + { + "epoch": 0.03, + "grad_norm": 2.4230892385977887, + "learning_rate": 1.7896389324960755e-05, + "loss": 1.0843, + "step": 1140 + }, + { + "epoch": 0.03, + "grad_norm": 2.6278146510743983, + "learning_rate": 1.7912087912087915e-05, + "loss": 1.0905, + "step": 1141 + }, + { + "epoch": 0.03, + "grad_norm": 1.1123480200176505, + "learning_rate": 1.7927786499215073e-05, + "loss": 0.9082, + "step": 1142 + }, + { + "epoch": 0.03, + "grad_norm": 2.6033810314607577, + "learning_rate": 1.7943485086342233e-05, + "loss": 1.0673, + "step": 1143 + }, + { + "epoch": 0.03, + "grad_norm": 2.6556733631859966, + "learning_rate": 1.795918367346939e-05, + "loss": 0.9009, + "step": 1144 + }, + { + "epoch": 0.03, + "grad_norm": 2.4083438974849547, + "learning_rate": 1.7974882260596548e-05, + "loss": 1.0032, + "step": 1145 + }, + { + "epoch": 0.03, + "grad_norm": 2.198338616644362, + "learning_rate": 1.7990580847723708e-05, + "loss": 0.998, + "step": 1146 + }, + { + "epoch": 0.03, + "grad_norm": 2.2266472547277907, + "learning_rate": 1.8006279434850865e-05, + "loss": 1.0893, + "step": 1147 + }, + { + "epoch": 0.03, + "grad_norm": 2.667439893989676, + "learning_rate": 1.8021978021978023e-05, + "loss": 1.1744, + "step": 1148 + }, + { + "epoch": 0.03, + "grad_norm": 2.2365076819636016, + "learning_rate": 1.803767660910518e-05, + "loss": 1.139, + "step": 1149 + }, + { + "epoch": 0.03, + "grad_norm": 2.3493525627513594, + "learning_rate": 1.805337519623234e-05, + "loss": 1.1432, + "step": 1150 + }, + { + "epoch": 0.03, + "grad_norm": 2.2180860685188213, + "learning_rate": 1.8069073783359498e-05, + "loss": 1.0133, + "step": 1151 + }, + { + "epoch": 0.03, + "grad_norm": 2.0923964323946356, + "learning_rate": 1.8084772370486658e-05, + "loss": 1.0166, + "step": 1152 + }, + { + "epoch": 0.03, + "grad_norm": 2.280789507743777, + "learning_rate": 1.8100470957613815e-05, + "loss": 1.2348, + "step": 1153 + }, + { + "epoch": 0.03, + "grad_norm": 2.2028867469293756, + "learning_rate": 1.8116169544740976e-05, + "loss": 0.9041, + "step": 1154 + }, + { + "epoch": 0.03, + "grad_norm": 2.501341565415611, + "learning_rate": 1.8131868131868133e-05, + "loss": 1.051, + "step": 1155 + }, + { + "epoch": 0.03, + "grad_norm": 1.9879062377153922, + "learning_rate": 1.814756671899529e-05, + "loss": 0.9854, + "step": 1156 + }, + { + "epoch": 0.03, + "grad_norm": 2.04532559177454, + "learning_rate": 1.816326530612245e-05, + "loss": 1.2351, + "step": 1157 + }, + { + "epoch": 0.03, + "grad_norm": 2.9515856793412554, + "learning_rate": 1.8178963893249608e-05, + "loss": 1.0431, + "step": 1158 + }, + { + "epoch": 0.03, + "grad_norm": 2.2013202039089275, + "learning_rate": 1.819466248037677e-05, + "loss": 1.175, + "step": 1159 + }, + { + "epoch": 0.03, + "grad_norm": 2.0674061314961962, + "learning_rate": 1.8210361067503926e-05, + "loss": 1.2818, + "step": 1160 + }, + { + "epoch": 0.03, + "grad_norm": 2.8216964957138706, + "learning_rate": 1.8226059654631086e-05, + "loss": 1.0218, + "step": 1161 + }, + { + "epoch": 0.03, + "grad_norm": 2.568038775737139, + "learning_rate": 1.8241758241758244e-05, + "loss": 1.155, + "step": 1162 + }, + { + "epoch": 0.03, + "grad_norm": 3.039489361858671, + "learning_rate": 1.8257456828885404e-05, + "loss": 1.1114, + "step": 1163 + }, + { + "epoch": 0.03, + "grad_norm": 1.9862889066906462, + "learning_rate": 1.827315541601256e-05, + "loss": 1.1078, + "step": 1164 + }, + { + "epoch": 0.03, + "grad_norm": 2.575620881532997, + "learning_rate": 1.828885400313972e-05, + "loss": 1.0869, + "step": 1165 + }, + { + "epoch": 0.03, + "grad_norm": 2.1505357567628507, + "learning_rate": 1.830455259026688e-05, + "loss": 1.1537, + "step": 1166 + }, + { + "epoch": 0.03, + "grad_norm": 2.05155564378383, + "learning_rate": 1.8320251177394036e-05, + "loss": 1.0596, + "step": 1167 + }, + { + "epoch": 0.03, + "grad_norm": 2.1603821097242943, + "learning_rate": 1.8335949764521194e-05, + "loss": 1.0294, + "step": 1168 + }, + { + "epoch": 0.03, + "grad_norm": 2.5515819343689925, + "learning_rate": 1.835164835164835e-05, + "loss": 1.0346, + "step": 1169 + }, + { + "epoch": 0.03, + "grad_norm": 2.0810615941313007, + "learning_rate": 1.836734693877551e-05, + "loss": 1.0631, + "step": 1170 + }, + { + "epoch": 0.03, + "grad_norm": 2.188221811517598, + "learning_rate": 1.838304552590267e-05, + "loss": 0.9023, + "step": 1171 + }, + { + "epoch": 0.03, + "grad_norm": 2.0352969911485728, + "learning_rate": 1.839874411302983e-05, + "loss": 1.0135, + "step": 1172 + }, + { + "epoch": 0.03, + "grad_norm": 2.1182523076329938, + "learning_rate": 1.8414442700156986e-05, + "loss": 1.0652, + "step": 1173 + }, + { + "epoch": 0.03, + "grad_norm": 2.0122756664550647, + "learning_rate": 1.8430141287284147e-05, + "loss": 0.8829, + "step": 1174 + }, + { + "epoch": 0.03, + "grad_norm": 2.5623201676577225, + "learning_rate": 1.8445839874411304e-05, + "loss": 1.0434, + "step": 1175 + }, + { + "epoch": 0.03, + "grad_norm": 2.30665245808143, + "learning_rate": 1.8461538461538465e-05, + "loss": 0.9843, + "step": 1176 + }, + { + "epoch": 0.03, + "grad_norm": 2.5576079640799336, + "learning_rate": 1.8477237048665622e-05, + "loss": 1.1072, + "step": 1177 + }, + { + "epoch": 0.03, + "grad_norm": 2.3345561416529077, + "learning_rate": 1.849293563579278e-05, + "loss": 1.04, + "step": 1178 + }, + { + "epoch": 0.03, + "grad_norm": 2.1080590060907367, + "learning_rate": 1.850863422291994e-05, + "loss": 1.1004, + "step": 1179 + }, + { + "epoch": 0.03, + "grad_norm": 2.164752068272359, + "learning_rate": 1.8524332810047097e-05, + "loss": 1.0366, + "step": 1180 + }, + { + "epoch": 0.03, + "grad_norm": 2.2007758631384715, + "learning_rate": 1.8540031397174257e-05, + "loss": 0.9727, + "step": 1181 + }, + { + "epoch": 0.03, + "grad_norm": 2.2218438626507186, + "learning_rate": 1.8555729984301415e-05, + "loss": 0.9888, + "step": 1182 + }, + { + "epoch": 0.03, + "grad_norm": 2.7672983529885173, + "learning_rate": 1.8571428571428575e-05, + "loss": 0.9932, + "step": 1183 + }, + { + "epoch": 0.03, + "grad_norm": 2.303529426550099, + "learning_rate": 1.8587127158555732e-05, + "loss": 1.1264, + "step": 1184 + }, + { + "epoch": 0.03, + "grad_norm": 2.2631965271622394, + "learning_rate": 1.860282574568289e-05, + "loss": 1.172, + "step": 1185 + }, + { + "epoch": 0.03, + "grad_norm": 2.3260431115908236, + "learning_rate": 1.861852433281005e-05, + "loss": 0.9944, + "step": 1186 + }, + { + "epoch": 0.03, + "grad_norm": 1.9573606291952361, + "learning_rate": 1.8634222919937207e-05, + "loss": 0.9931, + "step": 1187 + }, + { + "epoch": 0.03, + "grad_norm": 2.0087692876852237, + "learning_rate": 1.8649921507064364e-05, + "loss": 0.9916, + "step": 1188 + }, + { + "epoch": 0.03, + "grad_norm": 2.041590459312246, + "learning_rate": 1.866562009419152e-05, + "loss": 1.0526, + "step": 1189 + }, + { + "epoch": 0.03, + "grad_norm": 1.1617123249101713, + "learning_rate": 1.8681318681318682e-05, + "loss": 0.9045, + "step": 1190 + }, + { + "epoch": 0.03, + "grad_norm": 2.1602634716757687, + "learning_rate": 1.869701726844584e-05, + "loss": 1.06, + "step": 1191 + }, + { + "epoch": 0.03, + "grad_norm": 2.2048265140015326, + "learning_rate": 1.8712715855573e-05, + "loss": 1.1219, + "step": 1192 + }, + { + "epoch": 0.03, + "grad_norm": 1.1407641099827712, + "learning_rate": 1.8728414442700157e-05, + "loss": 0.9703, + "step": 1193 + }, + { + "epoch": 0.03, + "grad_norm": 2.2557983719433965, + "learning_rate": 1.8744113029827318e-05, + "loss": 1.024, + "step": 1194 + }, + { + "epoch": 0.03, + "grad_norm": 2.0977563490062474, + "learning_rate": 1.8759811616954475e-05, + "loss": 1.0422, + "step": 1195 + }, + { + "epoch": 0.03, + "grad_norm": 2.704778914308905, + "learning_rate": 1.8775510204081636e-05, + "loss": 1.0562, + "step": 1196 + }, + { + "epoch": 0.03, + "grad_norm": 1.195928448606241, + "learning_rate": 1.8791208791208793e-05, + "loss": 1.0551, + "step": 1197 + }, + { + "epoch": 0.03, + "grad_norm": 2.1185588257760535, + "learning_rate": 1.8806907378335953e-05, + "loss": 1.1002, + "step": 1198 + }, + { + "epoch": 0.03, + "grad_norm": 2.384405752939235, + "learning_rate": 1.882260596546311e-05, + "loss": 1.0646, + "step": 1199 + }, + { + "epoch": 0.03, + "grad_norm": 2.068280672632866, + "learning_rate": 1.8838304552590268e-05, + "loss": 1.0056, + "step": 1200 + }, + { + "epoch": 0.03, + "grad_norm": 2.0175883459301285, + "learning_rate": 1.885400313971743e-05, + "loss": 1.0085, + "step": 1201 + }, + { + "epoch": 0.03, + "grad_norm": 2.206194366602631, + "learning_rate": 1.8869701726844586e-05, + "loss": 1.1596, + "step": 1202 + }, + { + "epoch": 0.03, + "grad_norm": 2.256440411811413, + "learning_rate": 1.8885400313971746e-05, + "loss": 1.0541, + "step": 1203 + }, + { + "epoch": 0.03, + "grad_norm": 2.538128690939282, + "learning_rate": 1.8901098901098903e-05, + "loss": 1.2248, + "step": 1204 + }, + { + "epoch": 0.03, + "grad_norm": 2.5291716377079765, + "learning_rate": 1.891679748822606e-05, + "loss": 0.9634, + "step": 1205 + }, + { + "epoch": 0.03, + "grad_norm": 2.3852951536253153, + "learning_rate": 1.893249607535322e-05, + "loss": 1.0248, + "step": 1206 + }, + { + "epoch": 0.03, + "grad_norm": 2.0565188562946934, + "learning_rate": 1.8948194662480378e-05, + "loss": 0.9989, + "step": 1207 + }, + { + "epoch": 0.03, + "grad_norm": 1.9959937476976715, + "learning_rate": 1.8963893249607535e-05, + "loss": 1.2235, + "step": 1208 + }, + { + "epoch": 0.03, + "grad_norm": 2.4577384285062913, + "learning_rate": 1.8979591836734696e-05, + "loss": 1.1572, + "step": 1209 + }, + { + "epoch": 0.03, + "grad_norm": 2.203739584323531, + "learning_rate": 1.8995290423861853e-05, + "loss": 1.0659, + "step": 1210 + }, + { + "epoch": 0.03, + "grad_norm": 2.3488845888510714, + "learning_rate": 1.901098901098901e-05, + "loss": 1.1889, + "step": 1211 + }, + { + "epoch": 0.03, + "grad_norm": 1.9256950012357426, + "learning_rate": 1.902668759811617e-05, + "loss": 1.1693, + "step": 1212 + }, + { + "epoch": 0.03, + "grad_norm": 2.0634697675593077, + "learning_rate": 1.9042386185243328e-05, + "loss": 1.0866, + "step": 1213 + }, + { + "epoch": 0.03, + "grad_norm": 2.1301301485054074, + "learning_rate": 1.905808477237049e-05, + "loss": 1.0786, + "step": 1214 + }, + { + "epoch": 0.03, + "grad_norm": 3.3418846850393535, + "learning_rate": 1.9073783359497646e-05, + "loss": 1.0977, + "step": 1215 + }, + { + "epoch": 0.03, + "grad_norm": 2.0914414137863084, + "learning_rate": 1.9089481946624807e-05, + "loss": 1.0233, + "step": 1216 + }, + { + "epoch": 0.03, + "grad_norm": 2.239133011088858, + "learning_rate": 1.9105180533751964e-05, + "loss": 1.0866, + "step": 1217 + }, + { + "epoch": 0.03, + "grad_norm": 2.4056202334141576, + "learning_rate": 1.9120879120879124e-05, + "loss": 1.1697, + "step": 1218 + }, + { + "epoch": 0.03, + "grad_norm": 2.126555539199654, + "learning_rate": 1.913657770800628e-05, + "loss": 0.9969, + "step": 1219 + }, + { + "epoch": 0.03, + "grad_norm": 2.1852830228149878, + "learning_rate": 1.9152276295133442e-05, + "loss": 1.2974, + "step": 1220 + }, + { + "epoch": 0.03, + "grad_norm": 2.176210916346904, + "learning_rate": 1.91679748822606e-05, + "loss": 0.9676, + "step": 1221 + }, + { + "epoch": 0.03, + "grad_norm": 2.290835779701157, + "learning_rate": 1.9183673469387756e-05, + "loss": 1.1372, + "step": 1222 + }, + { + "epoch": 0.03, + "grad_norm": 2.178501708578087, + "learning_rate": 1.9199372056514917e-05, + "loss": 1.0962, + "step": 1223 + }, + { + "epoch": 0.03, + "grad_norm": 2.1664764281330005, + "learning_rate": 1.9215070643642074e-05, + "loss": 1.1042, + "step": 1224 + }, + { + "epoch": 0.03, + "grad_norm": 1.1335044094029456, + "learning_rate": 1.923076923076923e-05, + "loss": 1.006, + "step": 1225 + }, + { + "epoch": 0.03, + "grad_norm": 2.067772765338306, + "learning_rate": 1.9246467817896392e-05, + "loss": 1.2259, + "step": 1226 + }, + { + "epoch": 0.03, + "grad_norm": 2.186503581795575, + "learning_rate": 1.926216640502355e-05, + "loss": 0.9032, + "step": 1227 + }, + { + "epoch": 0.03, + "grad_norm": 2.4340717173886808, + "learning_rate": 1.9277864992150706e-05, + "loss": 1.0314, + "step": 1228 + }, + { + "epoch": 0.03, + "grad_norm": 1.232133021411674, + "learning_rate": 1.9293563579277867e-05, + "loss": 0.9528, + "step": 1229 + }, + { + "epoch": 0.03, + "grad_norm": 2.304745140001457, + "learning_rate": 1.9309262166405024e-05, + "loss": 0.995, + "step": 1230 + }, + { + "epoch": 0.03, + "grad_norm": 2.36750717569727, + "learning_rate": 1.932496075353218e-05, + "loss": 1.0551, + "step": 1231 + }, + { + "epoch": 0.03, + "grad_norm": 2.5579091091559465, + "learning_rate": 1.9340659340659342e-05, + "loss": 1.0755, + "step": 1232 + }, + { + "epoch": 0.03, + "grad_norm": 2.312069947638379, + "learning_rate": 1.93563579277865e-05, + "loss": 1.0794, + "step": 1233 + }, + { + "epoch": 0.03, + "grad_norm": 2.1079468672000896, + "learning_rate": 1.937205651491366e-05, + "loss": 1.0053, + "step": 1234 + }, + { + "epoch": 0.03, + "grad_norm": 2.5274788720522365, + "learning_rate": 1.9387755102040817e-05, + "loss": 1.0732, + "step": 1235 + }, + { + "epoch": 0.03, + "grad_norm": 2.323683054781559, + "learning_rate": 1.9403453689167977e-05, + "loss": 1.0077, + "step": 1236 + }, + { + "epoch": 0.03, + "grad_norm": 2.344781954197227, + "learning_rate": 1.9419152276295135e-05, + "loss": 1.1437, + "step": 1237 + }, + { + "epoch": 0.03, + "grad_norm": 2.2759374419899405, + "learning_rate": 1.9434850863422295e-05, + "loss": 1.1598, + "step": 1238 + }, + { + "epoch": 0.03, + "grad_norm": 1.1924118328233968, + "learning_rate": 1.9450549450549452e-05, + "loss": 0.8442, + "step": 1239 + }, + { + "epoch": 0.03, + "grad_norm": 2.417066585667301, + "learning_rate": 1.9466248037676613e-05, + "loss": 1.0996, + "step": 1240 + }, + { + "epoch": 0.03, + "grad_norm": 2.0283239088637273, + "learning_rate": 1.948194662480377e-05, + "loss": 1.1426, + "step": 1241 + }, + { + "epoch": 0.03, + "grad_norm": 2.0001701676304533, + "learning_rate": 1.9497645211930927e-05, + "loss": 1.0425, + "step": 1242 + }, + { + "epoch": 0.03, + "grad_norm": 2.219474418119233, + "learning_rate": 1.9513343799058088e-05, + "loss": 1.0413, + "step": 1243 + }, + { + "epoch": 0.03, + "grad_norm": 1.1022836342420583, + "learning_rate": 1.9529042386185245e-05, + "loss": 0.965, + "step": 1244 + }, + { + "epoch": 0.03, + "grad_norm": 1.9914400297842434, + "learning_rate": 1.9544740973312402e-05, + "loss": 1.1362, + "step": 1245 + }, + { + "epoch": 0.03, + "grad_norm": 2.620935854583238, + "learning_rate": 1.9560439560439563e-05, + "loss": 1.0021, + "step": 1246 + }, + { + "epoch": 0.03, + "grad_norm": 2.2495429993916503, + "learning_rate": 1.957613814756672e-05, + "loss": 0.9802, + "step": 1247 + }, + { + "epoch": 0.03, + "grad_norm": 2.142815999203034, + "learning_rate": 1.9591836734693877e-05, + "loss": 1.1441, + "step": 1248 + }, + { + "epoch": 0.03, + "grad_norm": 2.1576831058439234, + "learning_rate": 1.9607535321821038e-05, + "loss": 1.0799, + "step": 1249 + }, + { + "epoch": 0.03, + "grad_norm": 2.3075790008617263, + "learning_rate": 1.9623233908948195e-05, + "loss": 1.0085, + "step": 1250 + }, + { + "epoch": 0.03, + "grad_norm": 1.0799763438182841, + "learning_rate": 1.9638932496075356e-05, + "loss": 0.9334, + "step": 1251 + }, + { + "epoch": 0.03, + "grad_norm": 2.2929086637199236, + "learning_rate": 1.9654631083202513e-05, + "loss": 1.1218, + "step": 1252 + }, + { + "epoch": 0.03, + "grad_norm": 2.1261664669562754, + "learning_rate": 1.967032967032967e-05, + "loss": 1.0897, + "step": 1253 + }, + { + "epoch": 0.03, + "grad_norm": 2.0920072969730263, + "learning_rate": 1.968602825745683e-05, + "loss": 0.9595, + "step": 1254 + }, + { + "epoch": 0.03, + "grad_norm": 2.517390594959475, + "learning_rate": 1.9701726844583988e-05, + "loss": 1.0859, + "step": 1255 + }, + { + "epoch": 0.03, + "grad_norm": 2.2736238002025018, + "learning_rate": 1.971742543171115e-05, + "loss": 1.0401, + "step": 1256 + }, + { + "epoch": 0.03, + "grad_norm": 2.878107791993771, + "learning_rate": 1.9733124018838306e-05, + "loss": 1.0211, + "step": 1257 + }, + { + "epoch": 0.03, + "grad_norm": 2.1362624242844808, + "learning_rate": 1.9748822605965466e-05, + "loss": 1.0816, + "step": 1258 + }, + { + "epoch": 0.03, + "grad_norm": 2.323464791239283, + "learning_rate": 1.9764521193092623e-05, + "loss": 1.0533, + "step": 1259 + }, + { + "epoch": 0.03, + "grad_norm": 2.0393397182301536, + "learning_rate": 1.9780219780219784e-05, + "loss": 1.2087, + "step": 1260 + }, + { + "epoch": 0.03, + "grad_norm": 2.173573515463089, + "learning_rate": 1.979591836734694e-05, + "loss": 1.0209, + "step": 1261 + }, + { + "epoch": 0.03, + "grad_norm": 1.1255489070310936, + "learning_rate": 1.98116169544741e-05, + "loss": 1.0256, + "step": 1262 + }, + { + "epoch": 0.03, + "grad_norm": 2.132913305010335, + "learning_rate": 1.982731554160126e-05, + "loss": 1.1437, + "step": 1263 + }, + { + "epoch": 0.03, + "grad_norm": 2.3195258070375813, + "learning_rate": 1.9843014128728416e-05, + "loss": 1.135, + "step": 1264 + }, + { + "epoch": 0.03, + "grad_norm": 2.000382706344901, + "learning_rate": 1.9858712715855573e-05, + "loss": 1.0425, + "step": 1265 + }, + { + "epoch": 0.03, + "grad_norm": 2.268762777348503, + "learning_rate": 1.9874411302982734e-05, + "loss": 1.1266, + "step": 1266 + }, + { + "epoch": 0.03, + "grad_norm": 2.656416629287638, + "learning_rate": 1.989010989010989e-05, + "loss": 1.0956, + "step": 1267 + }, + { + "epoch": 0.03, + "grad_norm": 1.1914577665486639, + "learning_rate": 1.990580847723705e-05, + "loss": 0.9352, + "step": 1268 + }, + { + "epoch": 0.03, + "grad_norm": 2.1103256295485036, + "learning_rate": 1.992150706436421e-05, + "loss": 1.1061, + "step": 1269 + }, + { + "epoch": 0.03, + "grad_norm": 2.1194076059523685, + "learning_rate": 1.9937205651491366e-05, + "loss": 1.1066, + "step": 1270 + }, + { + "epoch": 0.03, + "grad_norm": 2.064205501797094, + "learning_rate": 1.9952904238618527e-05, + "loss": 1.0607, + "step": 1271 + }, + { + "epoch": 0.03, + "grad_norm": 1.1291196380757684, + "learning_rate": 1.9968602825745684e-05, + "loss": 0.9774, + "step": 1272 + }, + { + "epoch": 0.03, + "grad_norm": 2.351560695325121, + "learning_rate": 1.9984301412872844e-05, + "loss": 1.0557, + "step": 1273 + }, + { + "epoch": 0.03, + "grad_norm": 2.0683319797583763, + "learning_rate": 2e-05, + "loss": 1.0557, + "step": 1274 + }, + { + "epoch": 0.03, + "grad_norm": 2.3035595310670516, + "learning_rate": 1.9999999970888418e-05, + "loss": 1.0105, + "step": 1275 + }, + { + "epoch": 0.03, + "grad_norm": 6.308787570956558, + "learning_rate": 1.999999988355367e-05, + "loss": 1.1632, + "step": 1276 + }, + { + "epoch": 0.03, + "grad_norm": 2.7864144702178173, + "learning_rate": 1.9999999737995757e-05, + "loss": 1.046, + "step": 1277 + }, + { + "epoch": 0.03, + "grad_norm": 1.230252864987822, + "learning_rate": 1.9999999534214676e-05, + "loss": 0.9548, + "step": 1278 + }, + { + "epoch": 0.03, + "grad_norm": 2.2113773336850087, + "learning_rate": 1.9999999272210435e-05, + "loss": 1.0213, + "step": 1279 + }, + { + "epoch": 0.03, + "grad_norm": 2.324841314270198, + "learning_rate": 1.9999998951983033e-05, + "loss": 1.0056, + "step": 1280 + }, + { + "epoch": 0.03, + "grad_norm": 1.926571361720726, + "learning_rate": 1.999999857353247e-05, + "loss": 1.0656, + "step": 1281 + }, + { + "epoch": 0.03, + "grad_norm": 2.45569463246395, + "learning_rate": 1.9999998136858748e-05, + "loss": 1.1057, + "step": 1282 + }, + { + "epoch": 0.03, + "grad_norm": 2.040806332607944, + "learning_rate": 1.9999997641961873e-05, + "loss": 1.0545, + "step": 1283 + }, + { + "epoch": 0.03, + "grad_norm": 2.3363602610123757, + "learning_rate": 1.9999997088841844e-05, + "loss": 1.0119, + "step": 1284 + }, + { + "epoch": 0.03, + "grad_norm": 2.1477301694673336, + "learning_rate": 1.9999996477498667e-05, + "loss": 1.2024, + "step": 1285 + }, + { + "epoch": 0.03, + "grad_norm": 2.0113383107609044, + "learning_rate": 1.9999995807932346e-05, + "loss": 1.0137, + "step": 1286 + }, + { + "epoch": 0.03, + "grad_norm": 1.915280558849914, + "learning_rate": 1.999999508014288e-05, + "loss": 1.104, + "step": 1287 + }, + { + "epoch": 0.03, + "grad_norm": 2.11713917283824, + "learning_rate": 1.999999429413028e-05, + "loss": 1.059, + "step": 1288 + }, + { + "epoch": 0.03, + "grad_norm": 2.229155358060873, + "learning_rate": 1.9999993449894542e-05, + "loss": 0.8786, + "step": 1289 + }, + { + "epoch": 0.03, + "grad_norm": 2.283471405805508, + "learning_rate": 1.9999992547435683e-05, + "loss": 1.2333, + "step": 1290 + }, + { + "epoch": 0.03, + "grad_norm": 1.1310108203644913, + "learning_rate": 1.99999915867537e-05, + "loss": 0.8824, + "step": 1291 + }, + { + "epoch": 0.03, + "grad_norm": 2.349766025747535, + "learning_rate": 1.9999990567848595e-05, + "loss": 1.0622, + "step": 1292 + }, + { + "epoch": 0.03, + "grad_norm": 5.365906061809128, + "learning_rate": 1.9999989490720384e-05, + "loss": 1.1375, + "step": 1293 + }, + { + "epoch": 0.03, + "grad_norm": 2.185530362480177, + "learning_rate": 1.9999988355369067e-05, + "loss": 1.0028, + "step": 1294 + }, + { + "epoch": 0.03, + "grad_norm": 2.45412820546556, + "learning_rate": 1.999998716179465e-05, + "loss": 1.1117, + "step": 1295 + }, + { + "epoch": 0.03, + "grad_norm": 2.8998107087969562, + "learning_rate": 1.9999985909997145e-05, + "loss": 1.0865, + "step": 1296 + }, + { + "epoch": 0.03, + "grad_norm": 2.6647496782289863, + "learning_rate": 1.9999984599976555e-05, + "loss": 1.0346, + "step": 1297 + }, + { + "epoch": 0.03, + "grad_norm": 2.045638847042334, + "learning_rate": 1.9999983231732888e-05, + "loss": 1.0944, + "step": 1298 + }, + { + "epoch": 0.03, + "grad_norm": 2.1685249838288834, + "learning_rate": 1.9999981805266152e-05, + "loss": 1.1838, + "step": 1299 + }, + { + "epoch": 0.03, + "grad_norm": 3.589684465213933, + "learning_rate": 1.9999980320576356e-05, + "loss": 1.0602, + "step": 1300 + }, + { + "epoch": 0.03, + "grad_norm": 2.292798096465458, + "learning_rate": 1.999997877766351e-05, + "loss": 1.2599, + "step": 1301 + }, + { + "epoch": 0.03, + "grad_norm": 2.048046806070973, + "learning_rate": 1.9999977176527625e-05, + "loss": 1.1412, + "step": 1302 + }, + { + "epoch": 0.03, + "grad_norm": 2.1411330219998606, + "learning_rate": 1.99999755171687e-05, + "loss": 1.2467, + "step": 1303 + }, + { + "epoch": 0.03, + "grad_norm": 2.0757217095257854, + "learning_rate": 1.9999973799586754e-05, + "loss": 0.9399, + "step": 1304 + }, + { + "epoch": 0.03, + "grad_norm": 2.1115193957465745, + "learning_rate": 1.9999972023781797e-05, + "loss": 1.0543, + "step": 1305 + }, + { + "epoch": 0.03, + "grad_norm": 2.1404233539714546, + "learning_rate": 1.999997018975383e-05, + "loss": 1.2242, + "step": 1306 + }, + { + "epoch": 0.03, + "grad_norm": 2.3403131522824516, + "learning_rate": 1.9999968297502878e-05, + "loss": 1.0276, + "step": 1307 + }, + { + "epoch": 0.03, + "grad_norm": 2.2046430363791107, + "learning_rate": 1.9999966347028943e-05, + "loss": 0.9922, + "step": 1308 + }, + { + "epoch": 0.03, + "grad_norm": 2.2283919471018168, + "learning_rate": 1.9999964338332037e-05, + "loss": 1.0813, + "step": 1309 + }, + { + "epoch": 0.03, + "grad_norm": 2.7453657762551646, + "learning_rate": 1.9999962271412173e-05, + "loss": 0.9861, + "step": 1310 + }, + { + "epoch": 0.03, + "grad_norm": 2.103535155530633, + "learning_rate": 1.9999960146269365e-05, + "loss": 1.2244, + "step": 1311 + }, + { + "epoch": 0.03, + "grad_norm": 2.58662534371993, + "learning_rate": 1.999995796290362e-05, + "loss": 0.9644, + "step": 1312 + }, + { + "epoch": 0.03, + "grad_norm": 2.5300846494575393, + "learning_rate": 1.9999955721314954e-05, + "loss": 1.1427, + "step": 1313 + }, + { + "epoch": 0.03, + "grad_norm": 4.52951808662103, + "learning_rate": 1.999995342150338e-05, + "loss": 0.9848, + "step": 1314 + }, + { + "epoch": 0.03, + "grad_norm": 2.1666063585832664, + "learning_rate": 1.999995106346891e-05, + "loss": 1.0345, + "step": 1315 + }, + { + "epoch": 0.03, + "grad_norm": 2.2860546043686956, + "learning_rate": 1.9999948647211562e-05, + "loss": 1.1268, + "step": 1316 + }, + { + "epoch": 0.03, + "grad_norm": 1.1931087225589356, + "learning_rate": 1.9999946172731347e-05, + "loss": 1.0017, + "step": 1317 + }, + { + "epoch": 0.03, + "grad_norm": 1.9935201407900274, + "learning_rate": 1.999994364002828e-05, + "loss": 1.1096, + "step": 1318 + }, + { + "epoch": 0.03, + "grad_norm": 2.072851530658438, + "learning_rate": 1.999994104910237e-05, + "loss": 0.9998, + "step": 1319 + }, + { + "epoch": 0.03, + "grad_norm": 2.1371090676748996, + "learning_rate": 1.9999938399953645e-05, + "loss": 1.1113, + "step": 1320 + }, + { + "epoch": 0.03, + "grad_norm": 2.3809031849398696, + "learning_rate": 1.9999935692582107e-05, + "loss": 0.9895, + "step": 1321 + }, + { + "epoch": 0.03, + "grad_norm": 2.3478513362274493, + "learning_rate": 1.9999932926987778e-05, + "loss": 1.145, + "step": 1322 + }, + { + "epoch": 0.03, + "grad_norm": 2.2318357490226672, + "learning_rate": 1.9999930103170676e-05, + "loss": 1.1816, + "step": 1323 + }, + { + "epoch": 0.03, + "grad_norm": 1.1759727154212498, + "learning_rate": 1.9999927221130814e-05, + "loss": 0.9421, + "step": 1324 + }, + { + "epoch": 0.03, + "grad_norm": 2.4098585300790596, + "learning_rate": 1.9999924280868208e-05, + "loss": 1.063, + "step": 1325 + }, + { + "epoch": 0.03, + "grad_norm": 2.552528894515688, + "learning_rate": 1.999992128238288e-05, + "loss": 0.983, + "step": 1326 + }, + { + "epoch": 0.03, + "grad_norm": 3.3511645109537267, + "learning_rate": 1.9999918225674843e-05, + "loss": 0.9817, + "step": 1327 + }, + { + "epoch": 0.03, + "grad_norm": 2.182449417152129, + "learning_rate": 1.9999915110744116e-05, + "loss": 1.1046, + "step": 1328 + }, + { + "epoch": 0.03, + "grad_norm": 2.4122233784719618, + "learning_rate": 1.9999911937590715e-05, + "loss": 1.1317, + "step": 1329 + }, + { + "epoch": 0.03, + "grad_norm": 2.2819039780338928, + "learning_rate": 1.999990870621466e-05, + "loss": 1.1591, + "step": 1330 + }, + { + "epoch": 0.03, + "grad_norm": 2.3991174243778293, + "learning_rate": 1.9999905416615976e-05, + "loss": 1.1131, + "step": 1331 + }, + { + "epoch": 0.03, + "grad_norm": 2.227002493463014, + "learning_rate": 1.9999902068794674e-05, + "loss": 0.908, + "step": 1332 + }, + { + "epoch": 0.03, + "grad_norm": 1.9768702280192438, + "learning_rate": 1.9999898662750778e-05, + "loss": 1.032, + "step": 1333 + }, + { + "epoch": 0.03, + "grad_norm": 2.59767599385151, + "learning_rate": 1.9999895198484303e-05, + "loss": 1.0861, + "step": 1334 + }, + { + "epoch": 0.03, + "grad_norm": 2.050226545245868, + "learning_rate": 1.9999891675995274e-05, + "loss": 0.9441, + "step": 1335 + }, + { + "epoch": 0.03, + "grad_norm": 2.1187638479598885, + "learning_rate": 1.9999888095283707e-05, + "loss": 0.9847, + "step": 1336 + }, + { + "epoch": 0.03, + "grad_norm": 2.3941127618905433, + "learning_rate": 1.999988445634963e-05, + "loss": 1.1266, + "step": 1337 + }, + { + "epoch": 0.03, + "grad_norm": 2.3788823079954247, + "learning_rate": 1.999988075919306e-05, + "loss": 1.1375, + "step": 1338 + }, + { + "epoch": 0.03, + "grad_norm": 2.4149735758471875, + "learning_rate": 1.9999877003814015e-05, + "loss": 1.017, + "step": 1339 + }, + { + "epoch": 0.03, + "grad_norm": 1.1034582381602518, + "learning_rate": 1.999987319021252e-05, + "loss": 0.9226, + "step": 1340 + }, + { + "epoch": 0.03, + "grad_norm": 2.1330883654234585, + "learning_rate": 1.99998693183886e-05, + "loss": 1.0053, + "step": 1341 + }, + { + "epoch": 0.03, + "grad_norm": 2.043363732996191, + "learning_rate": 1.9999865388342274e-05, + "loss": 1.1556, + "step": 1342 + }, + { + "epoch": 0.03, + "grad_norm": 2.7411317381546114, + "learning_rate": 1.9999861400073564e-05, + "loss": 1.1992, + "step": 1343 + }, + { + "epoch": 0.03, + "grad_norm": 2.20684677076715, + "learning_rate": 1.9999857353582496e-05, + "loss": 0.9495, + "step": 1344 + }, + { + "epoch": 0.03, + "grad_norm": 1.22830166435583, + "learning_rate": 1.9999853248869094e-05, + "loss": 1.0101, + "step": 1345 + }, + { + "epoch": 0.03, + "grad_norm": 2.6119226139737024, + "learning_rate": 1.999984908593338e-05, + "loss": 0.9714, + "step": 1346 + }, + { + "epoch": 0.03, + "grad_norm": 2.3439868350372564, + "learning_rate": 1.999984486477538e-05, + "loss": 1.1192, + "step": 1347 + }, + { + "epoch": 0.03, + "grad_norm": 3.6240586371885177, + "learning_rate": 1.9999840585395113e-05, + "loss": 1.0825, + "step": 1348 + }, + { + "epoch": 0.03, + "grad_norm": 2.2062720108977363, + "learning_rate": 1.9999836247792614e-05, + "loss": 1.0662, + "step": 1349 + }, + { + "epoch": 0.03, + "grad_norm": 2.2136243225073784, + "learning_rate": 1.99998318519679e-05, + "loss": 1.0491, + "step": 1350 + }, + { + "epoch": 0.03, + "grad_norm": 2.345209301591147, + "learning_rate": 1.9999827397920993e-05, + "loss": 1.0982, + "step": 1351 + }, + { + "epoch": 0.03, + "grad_norm": 1.984368105326337, + "learning_rate": 1.9999822885651935e-05, + "loss": 1.0812, + "step": 1352 + }, + { + "epoch": 0.03, + "grad_norm": 2.5640784893882715, + "learning_rate": 1.9999818315160736e-05, + "loss": 1.0716, + "step": 1353 + }, + { + "epoch": 0.03, + "grad_norm": 1.9901815087800094, + "learning_rate": 1.999981368644743e-05, + "loss": 0.9786, + "step": 1354 + }, + { + "epoch": 0.03, + "grad_norm": 2.57885970719834, + "learning_rate": 1.9999808999512044e-05, + "loss": 1.017, + "step": 1355 + }, + { + "epoch": 0.03, + "grad_norm": 2.0552008157847146, + "learning_rate": 1.99998042543546e-05, + "loss": 1.1334, + "step": 1356 + }, + { + "epoch": 0.03, + "grad_norm": 2.0292916198025415, + "learning_rate": 1.9999799450975136e-05, + "loss": 0.9349, + "step": 1357 + }, + { + "epoch": 0.03, + "grad_norm": 2.1675004887284706, + "learning_rate": 1.999979458937367e-05, + "loss": 0.9507, + "step": 1358 + }, + { + "epoch": 0.03, + "grad_norm": 2.2620106640056403, + "learning_rate": 1.9999789669550235e-05, + "loss": 1.0011, + "step": 1359 + }, + { + "epoch": 0.03, + "grad_norm": 3.0920415374538557, + "learning_rate": 1.999978469150486e-05, + "loss": 0.981, + "step": 1360 + }, + { + "epoch": 0.03, + "grad_norm": 2.451848262099848, + "learning_rate": 1.999977965523757e-05, + "loss": 1.0295, + "step": 1361 + }, + { + "epoch": 0.03, + "grad_norm": 5.236354428790487, + "learning_rate": 1.9999774560748398e-05, + "loss": 1.0845, + "step": 1362 + }, + { + "epoch": 0.03, + "grad_norm": 1.9845355153869546, + "learning_rate": 1.9999769408037377e-05, + "loss": 1.1012, + "step": 1363 + }, + { + "epoch": 0.03, + "grad_norm": 2.4016209877783945, + "learning_rate": 1.9999764197104526e-05, + "loss": 1.1973, + "step": 1364 + }, + { + "epoch": 0.03, + "grad_norm": 2.6022508125206563, + "learning_rate": 1.9999758927949882e-05, + "loss": 1.0746, + "step": 1365 + }, + { + "epoch": 0.03, + "grad_norm": 2.7157749786420315, + "learning_rate": 1.999975360057348e-05, + "loss": 1.159, + "step": 1366 + }, + { + "epoch": 0.03, + "grad_norm": 2.503674422112463, + "learning_rate": 1.9999748214975345e-05, + "loss": 1.1373, + "step": 1367 + }, + { + "epoch": 0.03, + "grad_norm": 2.0622333194241063, + "learning_rate": 1.999974277115551e-05, + "loss": 1.1482, + "step": 1368 + }, + { + "epoch": 0.03, + "grad_norm": 2.2196851949701286, + "learning_rate": 1.9999737269114008e-05, + "loss": 1.0685, + "step": 1369 + }, + { + "epoch": 0.03, + "grad_norm": 2.6477353670701373, + "learning_rate": 1.9999731708850868e-05, + "loss": 1.1011, + "step": 1370 + }, + { + "epoch": 0.03, + "grad_norm": 2.1467648811865327, + "learning_rate": 1.9999726090366122e-05, + "loss": 1.0644, + "step": 1371 + }, + { + "epoch": 0.03, + "grad_norm": 2.1558839281636244, + "learning_rate": 1.9999720413659808e-05, + "loss": 1.0675, + "step": 1372 + }, + { + "epoch": 0.03, + "grad_norm": 2.1978641827496497, + "learning_rate": 1.9999714678731955e-05, + "loss": 1.0729, + "step": 1373 + }, + { + "epoch": 0.03, + "grad_norm": 2.059758586403996, + "learning_rate": 1.99997088855826e-05, + "loss": 1.0013, + "step": 1374 + }, + { + "epoch": 0.03, + "grad_norm": 1.1967931905527636, + "learning_rate": 1.9999703034211772e-05, + "loss": 0.933, + "step": 1375 + }, + { + "epoch": 0.03, + "grad_norm": 2.066468699287715, + "learning_rate": 1.9999697124619505e-05, + "loss": 1.0092, + "step": 1376 + }, + { + "epoch": 0.03, + "grad_norm": 3.7476940284946223, + "learning_rate": 1.9999691156805836e-05, + "loss": 1.1286, + "step": 1377 + }, + { + "epoch": 0.03, + "grad_norm": 2.5145043347558893, + "learning_rate": 1.99996851307708e-05, + "loss": 1.069, + "step": 1378 + }, + { + "epoch": 0.03, + "grad_norm": 2.3552664889177697, + "learning_rate": 1.999967904651443e-05, + "loss": 1.0908, + "step": 1379 + }, + { + "epoch": 0.03, + "grad_norm": 2.3143919547363216, + "learning_rate": 1.9999672904036768e-05, + "loss": 1.1242, + "step": 1380 + }, + { + "epoch": 0.03, + "grad_norm": 2.787900392568153, + "learning_rate": 1.9999666703337838e-05, + "loss": 0.903, + "step": 1381 + }, + { + "epoch": 0.03, + "grad_norm": 2.2226292734444875, + "learning_rate": 1.9999660444417687e-05, + "loss": 1.103, + "step": 1382 + }, + { + "epoch": 0.03, + "grad_norm": 1.142691017242793, + "learning_rate": 1.9999654127276345e-05, + "loss": 0.9542, + "step": 1383 + }, + { + "epoch": 0.03, + "grad_norm": 2.3032719032872873, + "learning_rate": 1.999964775191385e-05, + "loss": 1.0187, + "step": 1384 + }, + { + "epoch": 0.03, + "grad_norm": 2.4670525050851286, + "learning_rate": 1.999964131833024e-05, + "loss": 1.1589, + "step": 1385 + }, + { + "epoch": 0.03, + "grad_norm": 2.2061157413366046, + "learning_rate": 1.9999634826525555e-05, + "loss": 1.1566, + "step": 1386 + }, + { + "epoch": 0.03, + "grad_norm": 2.1136887543291065, + "learning_rate": 1.999962827649983e-05, + "loss": 0.8762, + "step": 1387 + }, + { + "epoch": 0.03, + "grad_norm": 2.2010903552800203, + "learning_rate": 1.9999621668253103e-05, + "loss": 1.0113, + "step": 1388 + }, + { + "epoch": 0.03, + "grad_norm": 2.6213690194641357, + "learning_rate": 1.9999615001785412e-05, + "loss": 0.9994, + "step": 1389 + }, + { + "epoch": 0.03, + "grad_norm": 2.108455048894776, + "learning_rate": 1.9999608277096796e-05, + "loss": 1.1454, + "step": 1390 + }, + { + "epoch": 0.03, + "grad_norm": 1.9884532360383191, + "learning_rate": 1.99996014941873e-05, + "loss": 1.1426, + "step": 1391 + }, + { + "epoch": 0.03, + "grad_norm": 2.227331109982808, + "learning_rate": 1.9999594653056952e-05, + "loss": 1.0156, + "step": 1392 + }, + { + "epoch": 0.03, + "grad_norm": 2.31941947395065, + "learning_rate": 1.99995877537058e-05, + "loss": 1.1004, + "step": 1393 + }, + { + "epoch": 0.03, + "grad_norm": 2.347887606872079, + "learning_rate": 1.999958079613388e-05, + "loss": 1.0023, + "step": 1394 + }, + { + "epoch": 0.03, + "grad_norm": 2.1395311820756895, + "learning_rate": 1.9999573780341236e-05, + "loss": 1.0546, + "step": 1395 + }, + { + "epoch": 0.03, + "grad_norm": 2.002202963009829, + "learning_rate": 1.9999566706327908e-05, + "loss": 0.9226, + "step": 1396 + }, + { + "epoch": 0.03, + "grad_norm": 2.982730498986444, + "learning_rate": 1.999955957409394e-05, + "loss": 0.9799, + "step": 1397 + }, + { + "epoch": 0.03, + "grad_norm": 2.511216409570812, + "learning_rate": 1.9999552383639368e-05, + "loss": 1.0178, + "step": 1398 + }, + { + "epoch": 0.03, + "grad_norm": 2.302059982297438, + "learning_rate": 1.9999545134964235e-05, + "loss": 1.1747, + "step": 1399 + }, + { + "epoch": 0.03, + "grad_norm": 2.388259707778139, + "learning_rate": 1.9999537828068586e-05, + "loss": 1.0012, + "step": 1400 + }, + { + "epoch": 0.03, + "grad_norm": 2.310072550492587, + "learning_rate": 1.999953046295246e-05, + "loss": 1.0595, + "step": 1401 + }, + { + "epoch": 0.03, + "grad_norm": 2.0923982823345795, + "learning_rate": 1.9999523039615904e-05, + "loss": 1.0438, + "step": 1402 + }, + { + "epoch": 0.03, + "grad_norm": 2.5048318401340803, + "learning_rate": 1.9999515558058958e-05, + "loss": 1.1443, + "step": 1403 + }, + { + "epoch": 0.03, + "grad_norm": 2.2757064713704565, + "learning_rate": 1.9999508018281667e-05, + "loss": 1.0602, + "step": 1404 + }, + { + "epoch": 0.03, + "grad_norm": 2.286607598495656, + "learning_rate": 1.999950042028407e-05, + "loss": 0.975, + "step": 1405 + }, + { + "epoch": 0.03, + "grad_norm": 2.054493280984004, + "learning_rate": 1.9999492764066225e-05, + "loss": 1.0787, + "step": 1406 + }, + { + "epoch": 0.03, + "grad_norm": 2.553523013984823, + "learning_rate": 1.999948504962816e-05, + "loss": 1.0648, + "step": 1407 + }, + { + "epoch": 0.03, + "grad_norm": 2.4715039325640884, + "learning_rate": 1.9999477276969932e-05, + "loss": 0.9473, + "step": 1408 + }, + { + "epoch": 0.03, + "grad_norm": 2.1334264044424587, + "learning_rate": 1.9999469446091578e-05, + "loss": 1.0301, + "step": 1409 + }, + { + "epoch": 0.03, + "grad_norm": 2.171791158981531, + "learning_rate": 1.999946155699315e-05, + "loss": 1.0965, + "step": 1410 + }, + { + "epoch": 0.03, + "grad_norm": 2.1325239716786832, + "learning_rate": 1.999945360967469e-05, + "loss": 1.0037, + "step": 1411 + }, + { + "epoch": 0.03, + "grad_norm": 2.1814381456159198, + "learning_rate": 1.999944560413624e-05, + "loss": 1.159, + "step": 1412 + }, + { + "epoch": 0.03, + "grad_norm": 1.975126257784818, + "learning_rate": 1.999943754037786e-05, + "loss": 1.1225, + "step": 1413 + }, + { + "epoch": 0.03, + "grad_norm": 2.144259665464612, + "learning_rate": 1.9999429418399585e-05, + "loss": 1.0611, + "step": 1414 + }, + { + "epoch": 0.03, + "grad_norm": 2.5832608082836472, + "learning_rate": 1.9999421238201465e-05, + "loss": 1.1556, + "step": 1415 + }, + { + "epoch": 0.03, + "grad_norm": 2.5281855256998234, + "learning_rate": 1.9999412999783555e-05, + "loss": 1.286, + "step": 1416 + }, + { + "epoch": 0.03, + "grad_norm": 2.2608164878071504, + "learning_rate": 1.999940470314589e-05, + "loss": 0.9847, + "step": 1417 + }, + { + "epoch": 0.03, + "grad_norm": 2.4440475833678694, + "learning_rate": 1.9999396348288526e-05, + "loss": 1.117, + "step": 1418 + }, + { + "epoch": 0.03, + "grad_norm": 2.232093003659318, + "learning_rate": 1.9999387935211515e-05, + "loss": 1.0711, + "step": 1419 + }, + { + "epoch": 0.03, + "grad_norm": 2.6000524579628363, + "learning_rate": 1.99993794639149e-05, + "loss": 1.1557, + "step": 1420 + }, + { + "epoch": 0.03, + "grad_norm": 2.128338677534579, + "learning_rate": 1.999937093439873e-05, + "loss": 1.0012, + "step": 1421 + }, + { + "epoch": 0.03, + "grad_norm": 1.9036432905874325, + "learning_rate": 1.9999362346663056e-05, + "loss": 0.9792, + "step": 1422 + }, + { + "epoch": 0.03, + "grad_norm": 2.22057686933721, + "learning_rate": 1.999935370070793e-05, + "loss": 1.1922, + "step": 1423 + }, + { + "epoch": 0.03, + "grad_norm": 2.024337710184214, + "learning_rate": 1.9999344996533402e-05, + "loss": 1.2019, + "step": 1424 + }, + { + "epoch": 0.03, + "grad_norm": 2.3379963306759914, + "learning_rate": 1.9999336234139522e-05, + "loss": 1.1037, + "step": 1425 + }, + { + "epoch": 0.03, + "grad_norm": 2.24061601145606, + "learning_rate": 1.9999327413526337e-05, + "loss": 0.8947, + "step": 1426 + }, + { + "epoch": 0.03, + "grad_norm": 1.9289761334394946, + "learning_rate": 1.9999318534693904e-05, + "loss": 1.005, + "step": 1427 + }, + { + "epoch": 0.03, + "grad_norm": 2.0040437161363744, + "learning_rate": 1.9999309597642275e-05, + "loss": 0.9929, + "step": 1428 + }, + { + "epoch": 0.03, + "grad_norm": 2.1897627957015064, + "learning_rate": 1.9999300602371495e-05, + "loss": 1.1456, + "step": 1429 + }, + { + "epoch": 0.03, + "grad_norm": 2.3083163204042076, + "learning_rate": 1.9999291548881628e-05, + "loss": 0.9886, + "step": 1430 + }, + { + "epoch": 0.03, + "grad_norm": 2.2254264917960205, + "learning_rate": 1.9999282437172716e-05, + "loss": 1.0266, + "step": 1431 + }, + { + "epoch": 0.03, + "grad_norm": 2.138175309573074, + "learning_rate": 1.9999273267244813e-05, + "loss": 1.1087, + "step": 1432 + }, + { + "epoch": 0.03, + "grad_norm": 2.354741239171817, + "learning_rate": 1.9999264039097982e-05, + "loss": 0.97, + "step": 1433 + }, + { + "epoch": 0.03, + "grad_norm": 2.025609459936761, + "learning_rate": 1.9999254752732265e-05, + "loss": 1.1925, + "step": 1434 + }, + { + "epoch": 0.03, + "grad_norm": 2.796730716575983, + "learning_rate": 1.999924540814772e-05, + "loss": 1.0698, + "step": 1435 + }, + { + "epoch": 0.03, + "grad_norm": 2.170315429542712, + "learning_rate": 1.999923600534441e-05, + "loss": 1.2019, + "step": 1436 + }, + { + "epoch": 0.03, + "grad_norm": 2.4326129199855435, + "learning_rate": 1.9999226544322377e-05, + "loss": 1.0566, + "step": 1437 + }, + { + "epoch": 0.03, + "grad_norm": 2.982923670560315, + "learning_rate": 1.999921702508168e-05, + "loss": 1.0298, + "step": 1438 + }, + { + "epoch": 0.03, + "grad_norm": 2.2095500383472855, + "learning_rate": 1.9999207447622376e-05, + "loss": 1.2241, + "step": 1439 + }, + { + "epoch": 0.03, + "grad_norm": 2.492526869851429, + "learning_rate": 1.999919781194452e-05, + "loss": 1.1548, + "step": 1440 + }, + { + "epoch": 0.03, + "grad_norm": 2.7910956036938934, + "learning_rate": 1.9999188118048173e-05, + "loss": 1.1435, + "step": 1441 + }, + { + "epoch": 0.03, + "grad_norm": 3.4758178594569666, + "learning_rate": 1.9999178365933388e-05, + "loss": 1.0741, + "step": 1442 + }, + { + "epoch": 0.03, + "grad_norm": 2.4198444120641494, + "learning_rate": 1.9999168555600215e-05, + "loss": 1.1386, + "step": 1443 + }, + { + "epoch": 0.03, + "grad_norm": 2.0160859511008837, + "learning_rate": 1.999915868704872e-05, + "loss": 0.9946, + "step": 1444 + }, + { + "epoch": 0.03, + "grad_norm": 1.1778559829548125, + "learning_rate": 1.9999148760278956e-05, + "loss": 0.9267, + "step": 1445 + }, + { + "epoch": 0.03, + "grad_norm": 2.250631354541688, + "learning_rate": 1.9999138775290985e-05, + "loss": 1.0818, + "step": 1446 + }, + { + "epoch": 0.03, + "grad_norm": 2.1559017505289564, + "learning_rate": 1.999912873208486e-05, + "loss": 1.0815, + "step": 1447 + }, + { + "epoch": 0.03, + "grad_norm": 2.4063213643763683, + "learning_rate": 1.9999118630660643e-05, + "loss": 1.0703, + "step": 1448 + }, + { + "epoch": 0.03, + "grad_norm": 2.1662450751429145, + "learning_rate": 1.999910847101839e-05, + "loss": 1.1925, + "step": 1449 + }, + { + "epoch": 0.03, + "grad_norm": 1.1675742266064393, + "learning_rate": 1.9999098253158166e-05, + "loss": 1.009, + "step": 1450 + }, + { + "epoch": 0.03, + "grad_norm": 2.470420725236201, + "learning_rate": 1.999908797708002e-05, + "loss": 1.0903, + "step": 1451 + }, + { + "epoch": 0.03, + "grad_norm": 2.159639620524817, + "learning_rate": 1.9999077642784024e-05, + "loss": 1.1358, + "step": 1452 + }, + { + "epoch": 0.03, + "grad_norm": 2.222719815297264, + "learning_rate": 1.999906725027023e-05, + "loss": 1.0615, + "step": 1453 + }, + { + "epoch": 0.03, + "grad_norm": 2.08801588634168, + "learning_rate": 1.9999056799538698e-05, + "loss": 1.0605, + "step": 1454 + }, + { + "epoch": 0.03, + "grad_norm": 2.143805048231618, + "learning_rate": 1.9999046290589495e-05, + "loss": 1.1336, + "step": 1455 + }, + { + "epoch": 0.03, + "grad_norm": 2.4357919646153547, + "learning_rate": 1.9999035723422677e-05, + "loss": 1.1165, + "step": 1456 + }, + { + "epoch": 0.03, + "grad_norm": 2.179488084927508, + "learning_rate": 1.999902509803831e-05, + "loss": 1.2751, + "step": 1457 + }, + { + "epoch": 0.03, + "grad_norm": 2.055728789803291, + "learning_rate": 1.9999014414436454e-05, + "loss": 1.1117, + "step": 1458 + }, + { + "epoch": 0.03, + "grad_norm": 2.4520086097045803, + "learning_rate": 1.9999003672617165e-05, + "loss": 1.0238, + "step": 1459 + }, + { + "epoch": 0.03, + "grad_norm": 2.4508117853927947, + "learning_rate": 1.9998992872580516e-05, + "loss": 1.0959, + "step": 1460 + }, + { + "epoch": 0.03, + "grad_norm": 2.3330793306902335, + "learning_rate": 1.9998982014326563e-05, + "loss": 1.0123, + "step": 1461 + }, + { + "epoch": 0.03, + "grad_norm": 2.492407305883869, + "learning_rate": 1.9998971097855372e-05, + "loss": 0.9619, + "step": 1462 + }, + { + "epoch": 0.03, + "grad_norm": 2.0544897337160055, + "learning_rate": 1.9998960123167003e-05, + "loss": 0.9835, + "step": 1463 + }, + { + "epoch": 0.03, + "grad_norm": 2.1240094144402106, + "learning_rate": 1.9998949090261525e-05, + "loss": 1.0688, + "step": 1464 + }, + { + "epoch": 0.03, + "grad_norm": 2.1426302326267965, + "learning_rate": 1.9998937999138998e-05, + "loss": 1.0758, + "step": 1465 + }, + { + "epoch": 0.03, + "grad_norm": 2.1841749814963602, + "learning_rate": 1.999892684979949e-05, + "loss": 1.1565, + "step": 1466 + }, + { + "epoch": 0.03, + "grad_norm": 3.3105817759201734, + "learning_rate": 1.9998915642243064e-05, + "loss": 1.0952, + "step": 1467 + }, + { + "epoch": 0.03, + "grad_norm": 3.053342119350852, + "learning_rate": 1.9998904376469784e-05, + "loss": 1.0354, + "step": 1468 + }, + { + "epoch": 0.03, + "grad_norm": 2.350857499295771, + "learning_rate": 1.999889305247972e-05, + "loss": 1.0781, + "step": 1469 + }, + { + "epoch": 0.03, + "grad_norm": 2.423475282375007, + "learning_rate": 1.9998881670272933e-05, + "loss": 1.025, + "step": 1470 + }, + { + "epoch": 0.03, + "grad_norm": 1.229322847706223, + "learning_rate": 1.9998870229849492e-05, + "loss": 1.0192, + "step": 1471 + }, + { + "epoch": 0.03, + "grad_norm": 2.1839959003735925, + "learning_rate": 1.9998858731209463e-05, + "loss": 0.9574, + "step": 1472 + }, + { + "epoch": 0.03, + "grad_norm": 2.1700840947531916, + "learning_rate": 1.999884717435291e-05, + "loss": 1.035, + "step": 1473 + }, + { + "epoch": 0.03, + "grad_norm": 2.3285638160435123, + "learning_rate": 1.999883555927991e-05, + "loss": 1.0361, + "step": 1474 + }, + { + "epoch": 0.03, + "grad_norm": 2.3734448166962667, + "learning_rate": 1.999882388599052e-05, + "loss": 1.2013, + "step": 1475 + }, + { + "epoch": 0.03, + "grad_norm": 2.111835786917955, + "learning_rate": 1.999881215448481e-05, + "loss": 1.1769, + "step": 1476 + }, + { + "epoch": 0.03, + "grad_norm": 2.143674618248724, + "learning_rate": 1.999880036476285e-05, + "loss": 1.1595, + "step": 1477 + }, + { + "epoch": 0.03, + "grad_norm": 2.004844786408324, + "learning_rate": 1.999878851682471e-05, + "loss": 0.9939, + "step": 1478 + }, + { + "epoch": 0.03, + "grad_norm": 1.9525438994577362, + "learning_rate": 1.999877661067046e-05, + "loss": 1.1092, + "step": 1479 + }, + { + "epoch": 0.03, + "grad_norm": 2.7616763438360876, + "learning_rate": 1.9998764646300163e-05, + "loss": 1.1114, + "step": 1480 + }, + { + "epoch": 0.03, + "grad_norm": 2.692964874270461, + "learning_rate": 1.9998752623713894e-05, + "loss": 1.1098, + "step": 1481 + }, + { + "epoch": 0.03, + "grad_norm": 1.1278677421750705, + "learning_rate": 1.9998740542911722e-05, + "loss": 0.9452, + "step": 1482 + }, + { + "epoch": 0.03, + "grad_norm": 1.1807796041725598, + "learning_rate": 1.9998728403893717e-05, + "loss": 0.9621, + "step": 1483 + }, + { + "epoch": 0.03, + "grad_norm": 2.0776751564002565, + "learning_rate": 1.9998716206659948e-05, + "loss": 1.0686, + "step": 1484 + }, + { + "epoch": 0.03, + "grad_norm": 2.2680138798473513, + "learning_rate": 1.9998703951210492e-05, + "loss": 1.1431, + "step": 1485 + }, + { + "epoch": 0.04, + "grad_norm": 1.1066258396629698, + "learning_rate": 1.9998691637545417e-05, + "loss": 0.9716, + "step": 1486 + }, + { + "epoch": 0.04, + "grad_norm": 2.17555858069351, + "learning_rate": 1.9998679265664787e-05, + "loss": 1.1001, + "step": 1487 + }, + { + "epoch": 0.04, + "grad_norm": 2.072083920953625, + "learning_rate": 1.9998666835568684e-05, + "loss": 0.9215, + "step": 1488 + }, + { + "epoch": 0.04, + "grad_norm": 2.082072003074863, + "learning_rate": 1.999865434725718e-05, + "loss": 1.0381, + "step": 1489 + }, + { + "epoch": 0.04, + "grad_norm": 1.1427301703059685, + "learning_rate": 1.999864180073034e-05, + "loss": 0.9671, + "step": 1490 + }, + { + "epoch": 0.04, + "grad_norm": 1.141838417087071, + "learning_rate": 1.9998629195988243e-05, + "loss": 0.9121, + "step": 1491 + }, + { + "epoch": 0.04, + "grad_norm": 2.1969478640200504, + "learning_rate": 1.9998616533030965e-05, + "loss": 1.2058, + "step": 1492 + }, + { + "epoch": 0.04, + "grad_norm": 2.2902665431944964, + "learning_rate": 1.9998603811858573e-05, + "loss": 0.9411, + "step": 1493 + }, + { + "epoch": 0.04, + "grad_norm": 2.1015789318727482, + "learning_rate": 1.9998591032471144e-05, + "loss": 1.0076, + "step": 1494 + }, + { + "epoch": 0.04, + "grad_norm": 2.1313409450158796, + "learning_rate": 1.9998578194868754e-05, + "loss": 1.017, + "step": 1495 + }, + { + "epoch": 0.04, + "grad_norm": 2.4281174613738647, + "learning_rate": 1.9998565299051473e-05, + "loss": 0.9573, + "step": 1496 + }, + { + "epoch": 0.04, + "grad_norm": 2.755964792774656, + "learning_rate": 1.9998552345019383e-05, + "loss": 1.1241, + "step": 1497 + }, + { + "epoch": 0.04, + "grad_norm": 2.464069967906408, + "learning_rate": 1.9998539332772555e-05, + "loss": 1.1407, + "step": 1498 + }, + { + "epoch": 0.04, + "grad_norm": 2.3793502023991726, + "learning_rate": 1.9998526262311063e-05, + "loss": 1.1329, + "step": 1499 + }, + { + "epoch": 0.04, + "grad_norm": 2.2156751172379145, + "learning_rate": 1.999851313363499e-05, + "loss": 1.0213, + "step": 1500 + }, + { + "epoch": 0.04, + "grad_norm": 2.0837679277310595, + "learning_rate": 1.9998499946744404e-05, + "loss": 1.0496, + "step": 1501 + }, + { + "epoch": 0.04, + "grad_norm": 2.415304078312954, + "learning_rate": 1.999848670163939e-05, + "loss": 1.0883, + "step": 1502 + }, + { + "epoch": 0.04, + "grad_norm": 2.718457709829429, + "learning_rate": 1.9998473398320016e-05, + "loss": 1.1223, + "step": 1503 + }, + { + "epoch": 0.04, + "grad_norm": 2.4119486832472017, + "learning_rate": 1.9998460036786365e-05, + "loss": 1.0447, + "step": 1504 + }, + { + "epoch": 0.04, + "grad_norm": 1.9715521878910611, + "learning_rate": 1.999844661703852e-05, + "loss": 1.0417, + "step": 1505 + }, + { + "epoch": 0.04, + "grad_norm": 2.1288415881494904, + "learning_rate": 1.9998433139076547e-05, + "loss": 1.1027, + "step": 1506 + }, + { + "epoch": 0.04, + "grad_norm": 2.479147771087465, + "learning_rate": 1.9998419602900535e-05, + "loss": 1.0091, + "step": 1507 + }, + { + "epoch": 0.04, + "grad_norm": 2.252894313635414, + "learning_rate": 1.9998406008510555e-05, + "loss": 0.9394, + "step": 1508 + }, + { + "epoch": 0.04, + "grad_norm": 2.1573593969534692, + "learning_rate": 1.999839235590669e-05, + "loss": 1.051, + "step": 1509 + }, + { + "epoch": 0.04, + "grad_norm": 3.342647319005965, + "learning_rate": 1.999837864508902e-05, + "loss": 1.1735, + "step": 1510 + }, + { + "epoch": 0.04, + "grad_norm": 3.1668920005977736, + "learning_rate": 1.999836487605763e-05, + "loss": 1.1011, + "step": 1511 + }, + { + "epoch": 0.04, + "grad_norm": 2.5548427789602104, + "learning_rate": 1.9998351048812585e-05, + "loss": 1.0876, + "step": 1512 + }, + { + "epoch": 0.04, + "grad_norm": 2.41547296307455, + "learning_rate": 1.999833716335398e-05, + "loss": 1.0358, + "step": 1513 + }, + { + "epoch": 0.04, + "grad_norm": 2.600055369539211, + "learning_rate": 1.999832321968189e-05, + "loss": 1.0069, + "step": 1514 + }, + { + "epoch": 0.04, + "grad_norm": 2.3380849574442863, + "learning_rate": 1.9998309217796395e-05, + "loss": 0.9962, + "step": 1515 + }, + { + "epoch": 0.04, + "grad_norm": 2.1374612658756322, + "learning_rate": 1.9998295157697583e-05, + "loss": 0.9182, + "step": 1516 + }, + { + "epoch": 0.04, + "grad_norm": 2.4595717911301964, + "learning_rate": 1.9998281039385527e-05, + "loss": 1.0512, + "step": 1517 + }, + { + "epoch": 0.04, + "grad_norm": 2.8830249998488586, + "learning_rate": 1.9998266862860315e-05, + "loss": 1.0847, + "step": 1518 + }, + { + "epoch": 0.04, + "grad_norm": 3.1379634889201364, + "learning_rate": 1.9998252628122025e-05, + "loss": 1.0857, + "step": 1519 + }, + { + "epoch": 0.04, + "grad_norm": 2.2276669325077085, + "learning_rate": 1.9998238335170746e-05, + "loss": 1.0656, + "step": 1520 + }, + { + "epoch": 0.04, + "grad_norm": 2.1861264239271296, + "learning_rate": 1.9998223984006558e-05, + "loss": 1.0666, + "step": 1521 + }, + { + "epoch": 0.04, + "grad_norm": 2.5045282968165408, + "learning_rate": 1.9998209574629542e-05, + "loss": 1.0848, + "step": 1522 + }, + { + "epoch": 0.04, + "grad_norm": 2.1479291636919724, + "learning_rate": 1.9998195107039785e-05, + "loss": 1.069, + "step": 1523 + }, + { + "epoch": 0.04, + "grad_norm": 3.435755169535846, + "learning_rate": 1.9998180581237372e-05, + "loss": 1.1902, + "step": 1524 + }, + { + "epoch": 0.04, + "grad_norm": 2.2006646320117746, + "learning_rate": 1.9998165997222387e-05, + "loss": 0.9694, + "step": 1525 + }, + { + "epoch": 0.04, + "grad_norm": 2.3432979048367084, + "learning_rate": 1.9998151354994914e-05, + "loss": 1.0263, + "step": 1526 + }, + { + "epoch": 0.04, + "grad_norm": 2.333442915473936, + "learning_rate": 1.999813665455504e-05, + "loss": 1.0695, + "step": 1527 + }, + { + "epoch": 0.04, + "grad_norm": 2.0465141469815338, + "learning_rate": 1.9998121895902843e-05, + "loss": 1.0251, + "step": 1528 + }, + { + "epoch": 0.04, + "grad_norm": 2.276141306447814, + "learning_rate": 1.999810707903842e-05, + "loss": 1.0556, + "step": 1529 + }, + { + "epoch": 0.04, + "grad_norm": 2.0951662536912914, + "learning_rate": 1.999809220396185e-05, + "loss": 0.9943, + "step": 1530 + }, + { + "epoch": 0.04, + "grad_norm": 2.1440908012492943, + "learning_rate": 1.9998077270673222e-05, + "loss": 0.9928, + "step": 1531 + }, + { + "epoch": 0.04, + "grad_norm": 2.014665130429584, + "learning_rate": 1.9998062279172624e-05, + "loss": 0.9327, + "step": 1532 + }, + { + "epoch": 0.04, + "grad_norm": 3.457087874824339, + "learning_rate": 1.9998047229460144e-05, + "loss": 1.0978, + "step": 1533 + }, + { + "epoch": 0.04, + "grad_norm": 2.810581243515628, + "learning_rate": 1.9998032121535863e-05, + "loss": 1.076, + "step": 1534 + }, + { + "epoch": 0.04, + "grad_norm": 2.243847832034537, + "learning_rate": 1.9998016955399877e-05, + "loss": 1.0386, + "step": 1535 + }, + { + "epoch": 0.04, + "grad_norm": 2.2796771783150587, + "learning_rate": 1.999800173105227e-05, + "loss": 1.005, + "step": 1536 + }, + { + "epoch": 0.04, + "grad_norm": 2.298373196783109, + "learning_rate": 1.9997986448493133e-05, + "loss": 1.048, + "step": 1537 + }, + { + "epoch": 0.04, + "grad_norm": 2.1298526952434087, + "learning_rate": 1.999797110772255e-05, + "loss": 1.1157, + "step": 1538 + }, + { + "epoch": 0.04, + "grad_norm": 2.170690034214531, + "learning_rate": 1.9997955708740616e-05, + "loss": 1.0645, + "step": 1539 + }, + { + "epoch": 0.04, + "grad_norm": 2.1740202091631056, + "learning_rate": 1.999794025154742e-05, + "loss": 1.057, + "step": 1540 + }, + { + "epoch": 0.04, + "grad_norm": 2.286831081441198, + "learning_rate": 1.999792473614305e-05, + "loss": 0.999, + "step": 1541 + }, + { + "epoch": 0.04, + "grad_norm": 2.1215827697164396, + "learning_rate": 1.9997909162527596e-05, + "loss": 1.1445, + "step": 1542 + }, + { + "epoch": 0.04, + "grad_norm": 2.2758122350293686, + "learning_rate": 1.999789353070115e-05, + "loss": 1.0434, + "step": 1543 + }, + { + "epoch": 0.04, + "grad_norm": 2.2800171151567445, + "learning_rate": 1.99978778406638e-05, + "loss": 1.138, + "step": 1544 + }, + { + "epoch": 0.04, + "grad_norm": 2.476976690757785, + "learning_rate": 1.9997862092415645e-05, + "loss": 0.9515, + "step": 1545 + }, + { + "epoch": 0.04, + "grad_norm": 2.4207131491666356, + "learning_rate": 1.9997846285956767e-05, + "loss": 1.0972, + "step": 1546 + }, + { + "epoch": 0.04, + "grad_norm": 2.1043232256955107, + "learning_rate": 1.9997830421287265e-05, + "loss": 1.0786, + "step": 1547 + }, + { + "epoch": 0.04, + "grad_norm": 2.070865792101983, + "learning_rate": 1.999781449840723e-05, + "loss": 1.1861, + "step": 1548 + }, + { + "epoch": 0.04, + "grad_norm": 3.486660224593569, + "learning_rate": 1.999779851731675e-05, + "loss": 1.0395, + "step": 1549 + }, + { + "epoch": 0.04, + "grad_norm": 1.943869745626806, + "learning_rate": 1.9997782478015923e-05, + "loss": 1.1803, + "step": 1550 + }, + { + "epoch": 0.04, + "grad_norm": 2.091258460229292, + "learning_rate": 1.9997766380504842e-05, + "loss": 1.0222, + "step": 1551 + }, + { + "epoch": 0.04, + "grad_norm": 2.954448683143159, + "learning_rate": 1.9997750224783605e-05, + "loss": 1.1155, + "step": 1552 + }, + { + "epoch": 0.04, + "grad_norm": 2.438647901955143, + "learning_rate": 1.9997734010852294e-05, + "loss": 1.0233, + "step": 1553 + }, + { + "epoch": 0.04, + "grad_norm": 2.646897918293091, + "learning_rate": 1.9997717738711013e-05, + "loss": 0.9428, + "step": 1554 + }, + { + "epoch": 0.04, + "grad_norm": 2.1819534655893094, + "learning_rate": 1.9997701408359852e-05, + "loss": 1.0771, + "step": 1555 + }, + { + "epoch": 0.04, + "grad_norm": 2.136936624597192, + "learning_rate": 1.9997685019798913e-05, + "loss": 1.1122, + "step": 1556 + }, + { + "epoch": 0.04, + "grad_norm": 2.075778965833001, + "learning_rate": 1.999766857302828e-05, + "loss": 1.0189, + "step": 1557 + }, + { + "epoch": 0.04, + "grad_norm": 2.2878541491171673, + "learning_rate": 1.9997652068048057e-05, + "loss": 1.2621, + "step": 1558 + }, + { + "epoch": 0.04, + "grad_norm": 2.5698598701211863, + "learning_rate": 1.999763550485834e-05, + "loss": 1.0908, + "step": 1559 + }, + { + "epoch": 0.04, + "grad_norm": 2.2390808828600686, + "learning_rate": 1.9997618883459225e-05, + "loss": 1.1608, + "step": 1560 + }, + { + "epoch": 0.04, + "grad_norm": 1.99305074842535, + "learning_rate": 1.9997602203850807e-05, + "loss": 1.2597, + "step": 1561 + }, + { + "epoch": 0.04, + "grad_norm": 2.502929906848804, + "learning_rate": 1.9997585466033186e-05, + "loss": 1.1343, + "step": 1562 + }, + { + "epoch": 0.04, + "grad_norm": 2.207663115422151, + "learning_rate": 1.999756867000645e-05, + "loss": 1.0364, + "step": 1563 + }, + { + "epoch": 0.04, + "grad_norm": 2.355374459297828, + "learning_rate": 1.9997551815770712e-05, + "loss": 0.9813, + "step": 1564 + }, + { + "epoch": 0.04, + "grad_norm": 2.630606653327842, + "learning_rate": 1.9997534903326058e-05, + "loss": 1.0016, + "step": 1565 + }, + { + "epoch": 0.04, + "grad_norm": 1.1597110475268597, + "learning_rate": 1.9997517932672592e-05, + "loss": 0.9456, + "step": 1566 + }, + { + "epoch": 0.04, + "grad_norm": 2.361146816584798, + "learning_rate": 1.9997500903810412e-05, + "loss": 1.0124, + "step": 1567 + }, + { + "epoch": 0.04, + "grad_norm": 2.2517309436453212, + "learning_rate": 1.9997483816739613e-05, + "loss": 1.088, + "step": 1568 + }, + { + "epoch": 0.04, + "grad_norm": 2.3808187266322935, + "learning_rate": 1.99974666714603e-05, + "loss": 1.0139, + "step": 1569 + }, + { + "epoch": 0.04, + "grad_norm": 2.165328177504271, + "learning_rate": 1.9997449467972574e-05, + "loss": 1.2674, + "step": 1570 + }, + { + "epoch": 0.04, + "grad_norm": 2.18321204133328, + "learning_rate": 1.9997432206276528e-05, + "loss": 1.0813, + "step": 1571 + }, + { + "epoch": 0.04, + "grad_norm": 1.9221123375827365, + "learning_rate": 1.9997414886372272e-05, + "loss": 1.1464, + "step": 1572 + }, + { + "epoch": 0.04, + "grad_norm": 2.351963598319574, + "learning_rate": 1.9997397508259895e-05, + "loss": 1.024, + "step": 1573 + }, + { + "epoch": 0.04, + "grad_norm": 2.4560661871642773, + "learning_rate": 1.9997380071939506e-05, + "loss": 1.191, + "step": 1574 + }, + { + "epoch": 0.04, + "grad_norm": 2.3241915914607696, + "learning_rate": 1.999736257741121e-05, + "loss": 1.0453, + "step": 1575 + }, + { + "epoch": 0.04, + "grad_norm": 2.202090616059174, + "learning_rate": 1.9997345024675095e-05, + "loss": 1.1071, + "step": 1576 + }, + { + "epoch": 0.04, + "grad_norm": 1.2901943658003918, + "learning_rate": 1.999732741373128e-05, + "loss": 0.9604, + "step": 1577 + }, + { + "epoch": 0.04, + "grad_norm": 2.313533158996857, + "learning_rate": 1.9997309744579857e-05, + "loss": 1.1921, + "step": 1578 + }, + { + "epoch": 0.04, + "grad_norm": 2.6414625013136317, + "learning_rate": 1.9997292017220933e-05, + "loss": 1.2464, + "step": 1579 + }, + { + "epoch": 0.04, + "grad_norm": 2.256220868384402, + "learning_rate": 1.9997274231654605e-05, + "loss": 1.1709, + "step": 1580 + }, + { + "epoch": 0.04, + "grad_norm": 2.090995000096685, + "learning_rate": 1.9997256387880987e-05, + "loss": 1.0471, + "step": 1581 + }, + { + "epoch": 0.04, + "grad_norm": 1.1189327807389724, + "learning_rate": 1.9997238485900175e-05, + "loss": 0.893, + "step": 1582 + }, + { + "epoch": 0.04, + "grad_norm": 2.136763180387602, + "learning_rate": 1.9997220525712276e-05, + "loss": 0.9458, + "step": 1583 + }, + { + "epoch": 0.04, + "grad_norm": 2.08353577064907, + "learning_rate": 1.999720250731739e-05, + "loss": 1.1713, + "step": 1584 + }, + { + "epoch": 0.04, + "grad_norm": 1.1430233890663255, + "learning_rate": 1.9997184430715632e-05, + "loss": 0.974, + "step": 1585 + }, + { + "epoch": 0.04, + "grad_norm": 2.595531752446033, + "learning_rate": 1.9997166295907097e-05, + "loss": 1.0279, + "step": 1586 + }, + { + "epoch": 0.04, + "grad_norm": 1.98240837059683, + "learning_rate": 1.9997148102891897e-05, + "loss": 1.0144, + "step": 1587 + }, + { + "epoch": 0.04, + "grad_norm": 1.9758659207175602, + "learning_rate": 1.9997129851670136e-05, + "loss": 0.995, + "step": 1588 + }, + { + "epoch": 0.04, + "grad_norm": 1.985813963528918, + "learning_rate": 1.999711154224192e-05, + "loss": 1.156, + "step": 1589 + }, + { + "epoch": 0.04, + "grad_norm": 2.2292711085245345, + "learning_rate": 1.9997093174607354e-05, + "loss": 1.0268, + "step": 1590 + }, + { + "epoch": 0.04, + "grad_norm": 2.0374407810700714, + "learning_rate": 1.999707474876655e-05, + "loss": 0.9911, + "step": 1591 + }, + { + "epoch": 0.04, + "grad_norm": 1.8870257623059394, + "learning_rate": 1.9997056264719607e-05, + "loss": 1.233, + "step": 1592 + }, + { + "epoch": 0.04, + "grad_norm": 2.493739740391232, + "learning_rate": 1.999703772246664e-05, + "loss": 1.0532, + "step": 1593 + }, + { + "epoch": 0.04, + "grad_norm": 2.1671990644606054, + "learning_rate": 1.999701912200775e-05, + "loss": 1.1501, + "step": 1594 + }, + { + "epoch": 0.04, + "grad_norm": 3.0777883585357952, + "learning_rate": 1.9997000463343057e-05, + "loss": 1.0392, + "step": 1595 + }, + { + "epoch": 0.04, + "grad_norm": 2.026996505921947, + "learning_rate": 1.9996981746472657e-05, + "loss": 1.2338, + "step": 1596 + }, + { + "epoch": 0.04, + "grad_norm": 10.933298247983998, + "learning_rate": 1.9996962971396664e-05, + "loss": 1.0548, + "step": 1597 + }, + { + "epoch": 0.04, + "grad_norm": 1.9305477964764446, + "learning_rate": 1.9996944138115192e-05, + "loss": 1.0334, + "step": 1598 + }, + { + "epoch": 0.04, + "grad_norm": 2.1715857393360314, + "learning_rate": 1.9996925246628344e-05, + "loss": 0.9671, + "step": 1599 + }, + { + "epoch": 0.04, + "grad_norm": 2.2289453652331197, + "learning_rate": 1.999690629693623e-05, + "loss": 1.0527, + "step": 1600 + }, + { + "epoch": 0.04, + "grad_norm": 2.2261532736837992, + "learning_rate": 1.9996887289038968e-05, + "loss": 0.9464, + "step": 1601 + }, + { + "epoch": 0.04, + "grad_norm": 2.438329834413523, + "learning_rate": 1.999686822293666e-05, + "loss": 1.0094, + "step": 1602 + }, + { + "epoch": 0.04, + "grad_norm": 1.1210243184075386, + "learning_rate": 1.999684909862942e-05, + "loss": 0.9208, + "step": 1603 + }, + { + "epoch": 0.04, + "grad_norm": 2.288601467791561, + "learning_rate": 1.999682991611736e-05, + "loss": 1.1953, + "step": 1604 + }, + { + "epoch": 0.04, + "grad_norm": 2.051178181359232, + "learning_rate": 1.9996810675400588e-05, + "loss": 1.0025, + "step": 1605 + }, + { + "epoch": 0.04, + "grad_norm": 2.2082816987066556, + "learning_rate": 1.9996791376479223e-05, + "loss": 0.9704, + "step": 1606 + }, + { + "epoch": 0.04, + "grad_norm": 2.1472862557719017, + "learning_rate": 1.9996772019353373e-05, + "loss": 1.0284, + "step": 1607 + }, + { + "epoch": 0.04, + "grad_norm": 2.433645133253672, + "learning_rate": 1.9996752604023152e-05, + "loss": 1.101, + "step": 1608 + }, + { + "epoch": 0.04, + "grad_norm": 2.922733957361882, + "learning_rate": 1.999673313048867e-05, + "loss": 0.9814, + "step": 1609 + }, + { + "epoch": 0.04, + "grad_norm": 2.2006309228742635, + "learning_rate": 1.9996713598750047e-05, + "loss": 0.9767, + "step": 1610 + }, + { + "epoch": 0.04, + "grad_norm": 2.1944492334819703, + "learning_rate": 1.999669400880739e-05, + "loss": 1.1552, + "step": 1611 + }, + { + "epoch": 0.04, + "grad_norm": 2.007233755912263, + "learning_rate": 1.9996674360660815e-05, + "loss": 1.0056, + "step": 1612 + }, + { + "epoch": 0.04, + "grad_norm": 2.011425532110267, + "learning_rate": 1.999665465431044e-05, + "loss": 0.9578, + "step": 1613 + }, + { + "epoch": 0.04, + "grad_norm": 2.2773164602128246, + "learning_rate": 1.999663488975637e-05, + "loss": 1.1617, + "step": 1614 + }, + { + "epoch": 0.04, + "grad_norm": 2.4278351851246005, + "learning_rate": 1.9996615066998733e-05, + "loss": 0.9302, + "step": 1615 + }, + { + "epoch": 0.04, + "grad_norm": 2.415026014325919, + "learning_rate": 1.9996595186037635e-05, + "loss": 1.1175, + "step": 1616 + }, + { + "epoch": 0.04, + "grad_norm": 2.1082237834666593, + "learning_rate": 1.9996575246873196e-05, + "loss": 1.1269, + "step": 1617 + }, + { + "epoch": 0.04, + "grad_norm": 2.3983850506396958, + "learning_rate": 1.9996555249505533e-05, + "loss": 1.053, + "step": 1618 + }, + { + "epoch": 0.04, + "grad_norm": 2.1276003898238747, + "learning_rate": 1.999653519393476e-05, + "loss": 0.9794, + "step": 1619 + }, + { + "epoch": 0.04, + "grad_norm": 2.3052629312138326, + "learning_rate": 1.999651508016099e-05, + "loss": 1.1064, + "step": 1620 + }, + { + "epoch": 0.04, + "grad_norm": 2.090309029884562, + "learning_rate": 1.999649490818435e-05, + "loss": 1.1976, + "step": 1621 + }, + { + "epoch": 0.04, + "grad_norm": 2.0500294727153587, + "learning_rate": 1.999647467800495e-05, + "loss": 1.0703, + "step": 1622 + }, + { + "epoch": 0.04, + "grad_norm": 2.2372633306001855, + "learning_rate": 1.999645438962291e-05, + "loss": 1.1395, + "step": 1623 + }, + { + "epoch": 0.04, + "grad_norm": 2.3318594384241305, + "learning_rate": 1.9996434043038344e-05, + "loss": 1.131, + "step": 1624 + }, + { + "epoch": 0.04, + "grad_norm": 2.332952087055381, + "learning_rate": 1.9996413638251377e-05, + "loss": 1.2689, + "step": 1625 + }, + { + "epoch": 0.04, + "grad_norm": 2.15738302821931, + "learning_rate": 1.9996393175262123e-05, + "loss": 1.1632, + "step": 1626 + }, + { + "epoch": 0.04, + "grad_norm": 1.9346726697121668, + "learning_rate": 1.9996372654070708e-05, + "loss": 1.0941, + "step": 1627 + }, + { + "epoch": 0.04, + "grad_norm": 2.158675524850137, + "learning_rate": 1.999635207467724e-05, + "loss": 1.0209, + "step": 1628 + }, + { + "epoch": 0.04, + "grad_norm": 2.3544366526214966, + "learning_rate": 1.999633143708185e-05, + "loss": 1.081, + "step": 1629 + }, + { + "epoch": 0.04, + "grad_norm": 2.8974664383586277, + "learning_rate": 1.9996310741284652e-05, + "loss": 1.1032, + "step": 1630 + }, + { + "epoch": 0.04, + "grad_norm": 1.1897721412827047, + "learning_rate": 1.999628998728577e-05, + "loss": 0.9952, + "step": 1631 + }, + { + "epoch": 0.04, + "grad_norm": 2.275875776166386, + "learning_rate": 1.999626917508532e-05, + "loss": 1.0773, + "step": 1632 + }, + { + "epoch": 0.04, + "grad_norm": 1.159255770457733, + "learning_rate": 1.999624830468343e-05, + "loss": 0.9525, + "step": 1633 + }, + { + "epoch": 0.04, + "grad_norm": 2.2637339260304024, + "learning_rate": 1.9996227376080213e-05, + "loss": 1.1382, + "step": 1634 + }, + { + "epoch": 0.04, + "grad_norm": 1.1270180318658984, + "learning_rate": 1.99962063892758e-05, + "loss": 0.9376, + "step": 1635 + }, + { + "epoch": 0.04, + "grad_norm": 2.301850852534426, + "learning_rate": 1.9996185344270306e-05, + "loss": 1.3066, + "step": 1636 + }, + { + "epoch": 0.04, + "grad_norm": 2.081461697243956, + "learning_rate": 1.9996164241063857e-05, + "loss": 1.1903, + "step": 1637 + }, + { + "epoch": 0.04, + "grad_norm": 2.2071319071847757, + "learning_rate": 1.9996143079656575e-05, + "loss": 0.9432, + "step": 1638 + }, + { + "epoch": 0.04, + "grad_norm": 2.4259296157249466, + "learning_rate": 1.999612186004858e-05, + "loss": 1.0881, + "step": 1639 + }, + { + "epoch": 0.04, + "grad_norm": 3.168595090786728, + "learning_rate": 1.9996100582240002e-05, + "loss": 1.1632, + "step": 1640 + }, + { + "epoch": 0.04, + "grad_norm": 2.191851793732097, + "learning_rate": 1.9996079246230962e-05, + "loss": 1.0072, + "step": 1641 + }, + { + "epoch": 0.04, + "grad_norm": 2.1759469643986833, + "learning_rate": 1.9996057852021584e-05, + "loss": 1.1721, + "step": 1642 + }, + { + "epoch": 0.04, + "grad_norm": 2.120516711719957, + "learning_rate": 1.9996036399611992e-05, + "loss": 1.0291, + "step": 1643 + }, + { + "epoch": 0.04, + "grad_norm": 1.229914773638126, + "learning_rate": 1.999601488900231e-05, + "loss": 0.9518, + "step": 1644 + }, + { + "epoch": 0.04, + "grad_norm": 2.1153909063055516, + "learning_rate": 1.9995993320192668e-05, + "loss": 1.0131, + "step": 1645 + }, + { + "epoch": 0.04, + "grad_norm": 2.1736605978965504, + "learning_rate": 1.9995971693183185e-05, + "loss": 1.056, + "step": 1646 + }, + { + "epoch": 0.04, + "grad_norm": 2.5632077053406963, + "learning_rate": 1.999595000797399e-05, + "loss": 1.0034, + "step": 1647 + }, + { + "epoch": 0.04, + "grad_norm": 2.4848958591620125, + "learning_rate": 1.999592826456521e-05, + "loss": 1.1028, + "step": 1648 + }, + { + "epoch": 0.04, + "grad_norm": 2.386890009266545, + "learning_rate": 1.9995906462956973e-05, + "loss": 1.0123, + "step": 1649 + }, + { + "epoch": 0.04, + "grad_norm": 1.9630427134365331, + "learning_rate": 1.9995884603149403e-05, + "loss": 1.0883, + "step": 1650 + }, + { + "epoch": 0.04, + "grad_norm": 1.2279386209627425, + "learning_rate": 1.999586268514263e-05, + "loss": 0.9271, + "step": 1651 + }, + { + "epoch": 0.04, + "grad_norm": 1.2156974408671557, + "learning_rate": 1.9995840708936773e-05, + "loss": 0.9862, + "step": 1652 + }, + { + "epoch": 0.04, + "grad_norm": 2.1817129150049093, + "learning_rate": 1.9995818674531972e-05, + "loss": 1.0804, + "step": 1653 + }, + { + "epoch": 0.04, + "grad_norm": 2.077666072772572, + "learning_rate": 1.999579658192835e-05, + "loss": 1.1281, + "step": 1654 + }, + { + "epoch": 0.04, + "grad_norm": 2.046611689327961, + "learning_rate": 1.9995774431126036e-05, + "loss": 1.046, + "step": 1655 + }, + { + "epoch": 0.04, + "grad_norm": 2.2774107043547938, + "learning_rate": 1.999575222212516e-05, + "loss": 1.0517, + "step": 1656 + }, + { + "epoch": 0.04, + "grad_norm": 2.182466713200807, + "learning_rate": 1.9995729954925844e-05, + "loss": 1.0891, + "step": 1657 + }, + { + "epoch": 0.04, + "grad_norm": 2.7256649366204053, + "learning_rate": 1.9995707629528227e-05, + "loss": 1.1516, + "step": 1658 + }, + { + "epoch": 0.04, + "grad_norm": 2.054381024961396, + "learning_rate": 1.9995685245932436e-05, + "loss": 1.0623, + "step": 1659 + }, + { + "epoch": 0.04, + "grad_norm": 2.342417974247102, + "learning_rate": 1.99956628041386e-05, + "loss": 1.0099, + "step": 1660 + }, + { + "epoch": 0.04, + "grad_norm": 2.0452365841438866, + "learning_rate": 1.999564030414685e-05, + "loss": 1.0926, + "step": 1661 + }, + { + "epoch": 0.04, + "grad_norm": 2.190766633985501, + "learning_rate": 1.999561774595732e-05, + "loss": 1.1401, + "step": 1662 + }, + { + "epoch": 0.04, + "grad_norm": 2.2181988874009924, + "learning_rate": 1.999559512957014e-05, + "loss": 1.0327, + "step": 1663 + }, + { + "epoch": 0.04, + "grad_norm": 2.006026487623011, + "learning_rate": 1.9995572454985437e-05, + "loss": 1.0242, + "step": 1664 + }, + { + "epoch": 0.04, + "grad_norm": 1.8977323166617306, + "learning_rate": 1.999554972220335e-05, + "loss": 1.1856, + "step": 1665 + }, + { + "epoch": 0.04, + "grad_norm": 2.0585652643247005, + "learning_rate": 1.9995526931224004e-05, + "loss": 1.0717, + "step": 1666 + }, + { + "epoch": 0.04, + "grad_norm": 2.0964502825740547, + "learning_rate": 1.9995504082047538e-05, + "loss": 1.0854, + "step": 1667 + }, + { + "epoch": 0.04, + "grad_norm": 1.357621889678359, + "learning_rate": 1.9995481174674084e-05, + "loss": 1.0513, + "step": 1668 + }, + { + "epoch": 0.04, + "grad_norm": 2.127891338636574, + "learning_rate": 1.999545820910377e-05, + "loss": 1.0583, + "step": 1669 + }, + { + "epoch": 0.04, + "grad_norm": 2.046237386016571, + "learning_rate": 1.9995435185336735e-05, + "loss": 1.0822, + "step": 1670 + }, + { + "epoch": 0.04, + "grad_norm": 2.094830232113409, + "learning_rate": 1.9995412103373115e-05, + "loss": 1.0357, + "step": 1671 + }, + { + "epoch": 0.04, + "grad_norm": 2.3228398855052648, + "learning_rate": 1.999538896321304e-05, + "loss": 1.1573, + "step": 1672 + }, + { + "epoch": 0.04, + "grad_norm": 1.176622596071275, + "learning_rate": 1.9995365764856646e-05, + "loss": 1.0011, + "step": 1673 + }, + { + "epoch": 0.04, + "grad_norm": 2.1237371848132582, + "learning_rate": 1.9995342508304065e-05, + "loss": 1.1494, + "step": 1674 + }, + { + "epoch": 0.04, + "grad_norm": 1.8859662395012342, + "learning_rate": 1.999531919355544e-05, + "loss": 1.0674, + "step": 1675 + }, + { + "epoch": 0.04, + "grad_norm": 2.215602491448394, + "learning_rate": 1.9995295820610897e-05, + "loss": 1.0343, + "step": 1676 + }, + { + "epoch": 0.04, + "grad_norm": 3.885345126818207, + "learning_rate": 1.9995272389470582e-05, + "loss": 1.078, + "step": 1677 + }, + { + "epoch": 0.04, + "grad_norm": 1.1793672715361474, + "learning_rate": 1.9995248900134624e-05, + "loss": 0.9682, + "step": 1678 + }, + { + "epoch": 0.04, + "grad_norm": 1.850660162194594, + "learning_rate": 1.9995225352603166e-05, + "loss": 0.9134, + "step": 1679 + }, + { + "epoch": 0.04, + "grad_norm": 2.194759506856468, + "learning_rate": 1.9995201746876337e-05, + "loss": 1.0538, + "step": 1680 + }, + { + "epoch": 0.04, + "grad_norm": 1.9356288421011962, + "learning_rate": 1.999517808295428e-05, + "loss": 1.1508, + "step": 1681 + }, + { + "epoch": 0.04, + "grad_norm": 2.1043527228351646, + "learning_rate": 1.9995154360837135e-05, + "loss": 1.1346, + "step": 1682 + }, + { + "epoch": 0.04, + "grad_norm": 2.0626210202376245, + "learning_rate": 1.999513058052503e-05, + "loss": 1.0826, + "step": 1683 + }, + { + "epoch": 0.04, + "grad_norm": 2.2536399386884414, + "learning_rate": 1.999510674201812e-05, + "loss": 1.119, + "step": 1684 + }, + { + "epoch": 0.04, + "grad_norm": 2.4124616554868337, + "learning_rate": 1.9995082845316528e-05, + "loss": 1.118, + "step": 1685 + }, + { + "epoch": 0.04, + "grad_norm": 1.9063606194079394, + "learning_rate": 1.99950588904204e-05, + "loss": 1.0705, + "step": 1686 + }, + { + "epoch": 0.04, + "grad_norm": 2.153877538478253, + "learning_rate": 1.9995034877329876e-05, + "loss": 1.1286, + "step": 1687 + }, + { + "epoch": 0.04, + "grad_norm": 2.2880918440252076, + "learning_rate": 1.9995010806045096e-05, + "loss": 1.1597, + "step": 1688 + }, + { + "epoch": 0.04, + "grad_norm": 2.173669769666353, + "learning_rate": 1.9994986676566194e-05, + "loss": 1.1366, + "step": 1689 + }, + { + "epoch": 0.04, + "grad_norm": 2.188906707561779, + "learning_rate": 1.9994962488893316e-05, + "loss": 0.92, + "step": 1690 + }, + { + "epoch": 0.04, + "grad_norm": 2.1378487858197244, + "learning_rate": 1.9994938243026605e-05, + "loss": 1.1124, + "step": 1691 + }, + { + "epoch": 0.04, + "grad_norm": 1.2342835785564723, + "learning_rate": 1.99949139389662e-05, + "loss": 0.9957, + "step": 1692 + }, + { + "epoch": 0.04, + "grad_norm": 2.365458819988178, + "learning_rate": 1.999488957671224e-05, + "loss": 0.9708, + "step": 1693 + }, + { + "epoch": 0.04, + "grad_norm": 3.4726955405695517, + "learning_rate": 1.9994865156264868e-05, + "loss": 1.0304, + "step": 1694 + }, + { + "epoch": 0.04, + "grad_norm": 2.832666176703946, + "learning_rate": 1.9994840677624226e-05, + "loss": 0.9762, + "step": 1695 + }, + { + "epoch": 0.04, + "grad_norm": 2.2311921806145936, + "learning_rate": 1.999481614079046e-05, + "loss": 1.0726, + "step": 1696 + }, + { + "epoch": 0.04, + "grad_norm": 2.221426113245724, + "learning_rate": 1.9994791545763707e-05, + "loss": 1.1816, + "step": 1697 + }, + { + "epoch": 0.04, + "grad_norm": 2.3499221992858654, + "learning_rate": 1.9994766892544113e-05, + "loss": 1.1664, + "step": 1698 + }, + { + "epoch": 0.04, + "grad_norm": 2.4330791040959947, + "learning_rate": 1.9994742181131824e-05, + "loss": 0.9853, + "step": 1699 + }, + { + "epoch": 0.04, + "grad_norm": 2.098878270319984, + "learning_rate": 1.999471741152698e-05, + "loss": 1.0551, + "step": 1700 + }, + { + "epoch": 0.04, + "grad_norm": 2.06494011160683, + "learning_rate": 1.9994692583729728e-05, + "loss": 1.0834, + "step": 1701 + }, + { + "epoch": 0.04, + "grad_norm": 2.4997128176399563, + "learning_rate": 1.9994667697740212e-05, + "loss": 1.0985, + "step": 1702 + }, + { + "epoch": 0.04, + "grad_norm": 2.483634190876616, + "learning_rate": 1.9994642753558577e-05, + "loss": 1.0675, + "step": 1703 + }, + { + "epoch": 0.04, + "grad_norm": 2.5226504944699935, + "learning_rate": 1.999461775118497e-05, + "loss": 1.1015, + "step": 1704 + }, + { + "epoch": 0.04, + "grad_norm": 2.4183874566945263, + "learning_rate": 1.9994592690619526e-05, + "loss": 1.0886, + "step": 1705 + }, + { + "epoch": 0.04, + "grad_norm": 2.092881508527578, + "learning_rate": 1.9994567571862405e-05, + "loss": 1.1692, + "step": 1706 + }, + { + "epoch": 0.04, + "grad_norm": 2.176098339870025, + "learning_rate": 1.9994542394913748e-05, + "loss": 1.1659, + "step": 1707 + }, + { + "epoch": 0.04, + "grad_norm": 2.540463229559394, + "learning_rate": 1.99945171597737e-05, + "loss": 1.1503, + "step": 1708 + }, + { + "epoch": 0.04, + "grad_norm": 2.124421398436532, + "learning_rate": 1.9994491866442408e-05, + "loss": 1.159, + "step": 1709 + }, + { + "epoch": 0.04, + "grad_norm": 1.1631049144431462, + "learning_rate": 1.9994466514920023e-05, + "loss": 0.9637, + "step": 1710 + }, + { + "epoch": 0.04, + "grad_norm": 1.8958227567475405, + "learning_rate": 1.9994441105206684e-05, + "loss": 1.0529, + "step": 1711 + }, + { + "epoch": 0.04, + "grad_norm": 2.153959858621639, + "learning_rate": 1.9994415637302545e-05, + "loss": 1.0506, + "step": 1712 + }, + { + "epoch": 0.04, + "grad_norm": 2.3915689208364332, + "learning_rate": 1.999439011120776e-05, + "loss": 1.1275, + "step": 1713 + }, + { + "epoch": 0.04, + "grad_norm": 2.1062697242121375, + "learning_rate": 1.9994364526922467e-05, + "loss": 1.1706, + "step": 1714 + }, + { + "epoch": 0.04, + "grad_norm": 2.312134765897488, + "learning_rate": 1.9994338884446823e-05, + "loss": 1.0664, + "step": 1715 + }, + { + "epoch": 0.04, + "grad_norm": 1.1821908221600224, + "learning_rate": 1.999431318378097e-05, + "loss": 0.9725, + "step": 1716 + }, + { + "epoch": 0.04, + "grad_norm": 2.494152547487427, + "learning_rate": 1.9994287424925063e-05, + "loss": 0.9365, + "step": 1717 + }, + { + "epoch": 0.04, + "grad_norm": 2.1808730687029736, + "learning_rate": 1.9994261607879252e-05, + "loss": 1.1013, + "step": 1718 + }, + { + "epoch": 0.04, + "grad_norm": 2.4503165992724543, + "learning_rate": 1.9994235732643687e-05, + "loss": 1.0822, + "step": 1719 + }, + { + "epoch": 0.04, + "grad_norm": 2.0626323672828497, + "learning_rate": 1.9994209799218514e-05, + "loss": 1.0593, + "step": 1720 + }, + { + "epoch": 0.04, + "grad_norm": 2.4876425449614863, + "learning_rate": 1.9994183807603885e-05, + "loss": 1.1147, + "step": 1721 + }, + { + "epoch": 0.04, + "grad_norm": 2.265769686580129, + "learning_rate": 1.999415775779996e-05, + "loss": 1.0536, + "step": 1722 + }, + { + "epoch": 0.04, + "grad_norm": 1.1943248033199514, + "learning_rate": 1.999413164980688e-05, + "loss": 1.0405, + "step": 1723 + }, + { + "epoch": 0.04, + "grad_norm": 2.093035471817573, + "learning_rate": 1.9994105483624807e-05, + "loss": 1.0649, + "step": 1724 + }, + { + "epoch": 0.04, + "grad_norm": 2.231418736339215, + "learning_rate": 1.9994079259253883e-05, + "loss": 1.0331, + "step": 1725 + }, + { + "epoch": 0.04, + "grad_norm": 2.601842069603881, + "learning_rate": 1.9994052976694265e-05, + "loss": 0.9587, + "step": 1726 + }, + { + "epoch": 0.04, + "grad_norm": 2.1769864743914407, + "learning_rate": 1.999402663594611e-05, + "loss": 1.0517, + "step": 1727 + }, + { + "epoch": 0.04, + "grad_norm": 1.933224560130183, + "learning_rate": 1.9994000237009565e-05, + "loss": 1.0847, + "step": 1728 + }, + { + "epoch": 0.04, + "grad_norm": 2.237787025184904, + "learning_rate": 1.9993973779884785e-05, + "loss": 1.2092, + "step": 1729 + }, + { + "epoch": 0.04, + "grad_norm": 1.9617224466096468, + "learning_rate": 1.999394726457193e-05, + "loss": 1.0565, + "step": 1730 + }, + { + "epoch": 0.04, + "grad_norm": 2.0801144227366644, + "learning_rate": 1.9993920691071146e-05, + "loss": 1.023, + "step": 1731 + }, + { + "epoch": 0.04, + "grad_norm": 2.510184360891999, + "learning_rate": 1.9993894059382595e-05, + "loss": 0.9834, + "step": 1732 + }, + { + "epoch": 0.04, + "grad_norm": 2.676930745800279, + "learning_rate": 1.999386736950643e-05, + "loss": 0.9859, + "step": 1733 + }, + { + "epoch": 0.04, + "grad_norm": 2.173187070832274, + "learning_rate": 1.9993840621442802e-05, + "loss": 1.1474, + "step": 1734 + }, + { + "epoch": 0.04, + "grad_norm": 1.1291299851094467, + "learning_rate": 1.9993813815191868e-05, + "loss": 0.9403, + "step": 1735 + }, + { + "epoch": 0.04, + "grad_norm": 1.9502954451953205, + "learning_rate": 1.9993786950753792e-05, + "loss": 1.0323, + "step": 1736 + }, + { + "epoch": 0.04, + "grad_norm": 2.1527983386396192, + "learning_rate": 1.999376002812872e-05, + "loss": 1.0376, + "step": 1737 + }, + { + "epoch": 0.04, + "grad_norm": 2.1592734260755586, + "learning_rate": 1.9993733047316814e-05, + "loss": 1.1816, + "step": 1738 + }, + { + "epoch": 0.04, + "grad_norm": 1.1207022044454604, + "learning_rate": 1.999370600831823e-05, + "loss": 1.0045, + "step": 1739 + }, + { + "epoch": 0.04, + "grad_norm": 2.220393988867591, + "learning_rate": 1.9993678911133126e-05, + "loss": 1.0744, + "step": 1740 + }, + { + "epoch": 0.04, + "grad_norm": 1.9168598202344584, + "learning_rate": 1.9993651755761658e-05, + "loss": 1.1452, + "step": 1741 + }, + { + "epoch": 0.04, + "grad_norm": 1.956007031532968, + "learning_rate": 1.9993624542203987e-05, + "loss": 1.1333, + "step": 1742 + }, + { + "epoch": 0.04, + "grad_norm": 2.013541594225797, + "learning_rate": 1.9993597270460273e-05, + "loss": 1.1301, + "step": 1743 + }, + { + "epoch": 0.04, + "grad_norm": 2.1051815251513015, + "learning_rate": 1.999356994053067e-05, + "loss": 1.1981, + "step": 1744 + }, + { + "epoch": 0.04, + "grad_norm": 2.6251057224966448, + "learning_rate": 1.9993542552415334e-05, + "loss": 1.1305, + "step": 1745 + }, + { + "epoch": 0.04, + "grad_norm": 1.9594984791943661, + "learning_rate": 1.9993515106114433e-05, + "loss": 1.1395, + "step": 1746 + }, + { + "epoch": 0.04, + "grad_norm": 2.5191080578911027, + "learning_rate": 1.9993487601628124e-05, + "loss": 0.9447, + "step": 1747 + }, + { + "epoch": 0.04, + "grad_norm": 2.0200536176153716, + "learning_rate": 1.9993460038956565e-05, + "loss": 1.0768, + "step": 1748 + }, + { + "epoch": 0.04, + "grad_norm": 2.1862540139760362, + "learning_rate": 1.999343241809992e-05, + "loss": 1.0097, + "step": 1749 + }, + { + "epoch": 0.04, + "grad_norm": 2.2416801419627332, + "learning_rate": 1.9993404739058346e-05, + "loss": 1.0618, + "step": 1750 + }, + { + "epoch": 0.04, + "grad_norm": 1.141367261507592, + "learning_rate": 1.9993377001832005e-05, + "loss": 0.9437, + "step": 1751 + }, + { + "epoch": 0.04, + "grad_norm": 2.3039245724717516, + "learning_rate": 1.999334920642106e-05, + "loss": 0.9834, + "step": 1752 + }, + { + "epoch": 0.04, + "grad_norm": 2.1609510793198052, + "learning_rate": 1.999332135282567e-05, + "loss": 1.1244, + "step": 1753 + }, + { + "epoch": 0.04, + "grad_norm": 2.417954463475518, + "learning_rate": 1.9993293441046e-05, + "loss": 1.001, + "step": 1754 + }, + { + "epoch": 0.04, + "grad_norm": 2.330704619391159, + "learning_rate": 1.9993265471082213e-05, + "loss": 1.141, + "step": 1755 + }, + { + "epoch": 0.04, + "grad_norm": 2.1515132568319495, + "learning_rate": 1.999323744293447e-05, + "loss": 1.162, + "step": 1756 + }, + { + "epoch": 0.04, + "grad_norm": 2.291668076530568, + "learning_rate": 1.999320935660294e-05, + "loss": 1.1638, + "step": 1757 + }, + { + "epoch": 0.04, + "grad_norm": 2.343415798559413, + "learning_rate": 1.9993181212087775e-05, + "loss": 0.8974, + "step": 1758 + }, + { + "epoch": 0.04, + "grad_norm": 3.088982740032418, + "learning_rate": 1.9993153009389145e-05, + "loss": 1.0019, + "step": 1759 + }, + { + "epoch": 0.04, + "grad_norm": 2.3620710218921794, + "learning_rate": 1.999312474850722e-05, + "loss": 1.1145, + "step": 1760 + }, + { + "epoch": 0.04, + "grad_norm": 2.390862871715157, + "learning_rate": 1.9993096429442153e-05, + "loss": 1.1521, + "step": 1761 + }, + { + "epoch": 0.04, + "grad_norm": 2.212763309709292, + "learning_rate": 1.9993068052194115e-05, + "loss": 0.9144, + "step": 1762 + }, + { + "epoch": 0.04, + "grad_norm": 2.1171524387620164, + "learning_rate": 1.9993039616763276e-05, + "loss": 1.009, + "step": 1763 + }, + { + "epoch": 0.04, + "grad_norm": 1.1452467833188391, + "learning_rate": 1.999301112314979e-05, + "loss": 0.9416, + "step": 1764 + }, + { + "epoch": 0.04, + "grad_norm": 2.2966088090861794, + "learning_rate": 1.9992982571353837e-05, + "loss": 0.9072, + "step": 1765 + }, + { + "epoch": 0.04, + "grad_norm": 1.2011060019201387, + "learning_rate": 1.9992953961375573e-05, + "loss": 1.0026, + "step": 1766 + }, + { + "epoch": 0.04, + "grad_norm": 1.1977962430235605, + "learning_rate": 1.9992925293215164e-05, + "loss": 1.0343, + "step": 1767 + }, + { + "epoch": 0.04, + "grad_norm": 2.812887535284226, + "learning_rate": 1.999289656687278e-05, + "loss": 1.0446, + "step": 1768 + }, + { + "epoch": 0.04, + "grad_norm": 1.1209819639112821, + "learning_rate": 1.9992867782348595e-05, + "loss": 0.9221, + "step": 1769 + }, + { + "epoch": 0.04, + "grad_norm": 2.010495782259121, + "learning_rate": 1.9992838939642766e-05, + "loss": 1.1028, + "step": 1770 + }, + { + "epoch": 0.04, + "grad_norm": 2.2065777621825617, + "learning_rate": 1.9992810038755464e-05, + "loss": 1.072, + "step": 1771 + }, + { + "epoch": 0.04, + "grad_norm": 2.7227839109204757, + "learning_rate": 1.999278107968686e-05, + "loss": 0.9826, + "step": 1772 + }, + { + "epoch": 0.04, + "grad_norm": 2.739673751898487, + "learning_rate": 1.999275206243712e-05, + "loss": 0.9207, + "step": 1773 + }, + { + "epoch": 0.04, + "grad_norm": 2.406099760447923, + "learning_rate": 1.9992722987006412e-05, + "loss": 1.0697, + "step": 1774 + }, + { + "epoch": 0.04, + "grad_norm": 2.2651363579613477, + "learning_rate": 1.999269385339491e-05, + "loss": 0.9371, + "step": 1775 + }, + { + "epoch": 0.04, + "grad_norm": 2.3494672573738833, + "learning_rate": 1.9992664661602782e-05, + "loss": 1.0406, + "step": 1776 + }, + { + "epoch": 0.04, + "grad_norm": 1.313765472240525, + "learning_rate": 1.9992635411630194e-05, + "loss": 0.9238, + "step": 1777 + }, + { + "epoch": 0.04, + "grad_norm": 2.7624684801035664, + "learning_rate": 1.9992606103477318e-05, + "loss": 0.9588, + "step": 1778 + }, + { + "epoch": 0.04, + "grad_norm": 2.3087603285843863, + "learning_rate": 1.999257673714433e-05, + "loss": 1.0359, + "step": 1779 + }, + { + "epoch": 0.04, + "grad_norm": 2.091622674925229, + "learning_rate": 1.999254731263139e-05, + "loss": 1.0634, + "step": 1780 + }, + { + "epoch": 0.04, + "grad_norm": 2.434899070097137, + "learning_rate": 1.9992517829938683e-05, + "loss": 0.997, + "step": 1781 + }, + { + "epoch": 0.04, + "grad_norm": 2.486856144734205, + "learning_rate": 1.999248828906637e-05, + "loss": 1.145, + "step": 1782 + }, + { + "epoch": 0.04, + "grad_norm": 2.739428383485529, + "learning_rate": 1.9992458690014625e-05, + "loss": 1.0355, + "step": 1783 + }, + { + "epoch": 0.04, + "grad_norm": 1.1402058776448447, + "learning_rate": 1.9992429032783625e-05, + "loss": 0.9833, + "step": 1784 + }, + { + "epoch": 0.04, + "grad_norm": 2.0566439369891376, + "learning_rate": 1.9992399317373537e-05, + "loss": 1.0929, + "step": 1785 + }, + { + "epoch": 0.04, + "grad_norm": 1.902818557646527, + "learning_rate": 1.999236954378454e-05, + "loss": 1.0404, + "step": 1786 + }, + { + "epoch": 0.04, + "grad_norm": 2.4978959024048484, + "learning_rate": 1.9992339712016805e-05, + "loss": 1.0349, + "step": 1787 + }, + { + "epoch": 0.04, + "grad_norm": 2.473100028621427, + "learning_rate": 1.9992309822070503e-05, + "loss": 1.1681, + "step": 1788 + }, + { + "epoch": 0.04, + "grad_norm": 5.3016569354054495, + "learning_rate": 1.9992279873945807e-05, + "loss": 1.2474, + "step": 1789 + }, + { + "epoch": 0.04, + "grad_norm": 2.2105169491285066, + "learning_rate": 1.9992249867642894e-05, + "loss": 0.9809, + "step": 1790 + }, + { + "epoch": 0.04, + "grad_norm": 1.900014684537637, + "learning_rate": 1.999221980316194e-05, + "loss": 1.164, + "step": 1791 + }, + { + "epoch": 0.04, + "grad_norm": 2.1749954679526873, + "learning_rate": 1.9992189680503124e-05, + "loss": 1.0708, + "step": 1792 + }, + { + "epoch": 0.04, + "grad_norm": 1.9952157555681167, + "learning_rate": 1.999215949966661e-05, + "loss": 1.0565, + "step": 1793 + }, + { + "epoch": 0.04, + "grad_norm": 2.1139124660934874, + "learning_rate": 1.999212926065258e-05, + "loss": 1.0227, + "step": 1794 + }, + { + "epoch": 0.04, + "grad_norm": 2.809926238203372, + "learning_rate": 1.9992098963461213e-05, + "loss": 1.0065, + "step": 1795 + }, + { + "epoch": 0.04, + "grad_norm": 1.9251750010891218, + "learning_rate": 1.999206860809268e-05, + "loss": 1.075, + "step": 1796 + }, + { + "epoch": 0.04, + "grad_norm": 2.3610882681645147, + "learning_rate": 1.9992038194547163e-05, + "loss": 1.0929, + "step": 1797 + }, + { + "epoch": 0.04, + "grad_norm": 1.9955309031859558, + "learning_rate": 1.9992007722824832e-05, + "loss": 1.0185, + "step": 1798 + }, + { + "epoch": 0.04, + "grad_norm": 1.8545415179580198, + "learning_rate": 1.999197719292587e-05, + "loss": 1.0049, + "step": 1799 + }, + { + "epoch": 0.04, + "grad_norm": 2.6866791693714767, + "learning_rate": 1.9991946604850456e-05, + "loss": 1.0028, + "step": 1800 + }, + { + "epoch": 0.04, + "grad_norm": 2.442703671754254, + "learning_rate": 1.9991915958598766e-05, + "loss": 0.996, + "step": 1801 + }, + { + "epoch": 0.04, + "grad_norm": 2.4133510174805233, + "learning_rate": 1.9991885254170972e-05, + "loss": 1.172, + "step": 1802 + }, + { + "epoch": 0.04, + "grad_norm": 3.0424782341728642, + "learning_rate": 1.9991854491567265e-05, + "loss": 1.221, + "step": 1803 + }, + { + "epoch": 0.04, + "grad_norm": 2.2021297830438615, + "learning_rate": 1.9991823670787817e-05, + "loss": 1.1724, + "step": 1804 + }, + { + "epoch": 0.04, + "grad_norm": 1.2165384883149843, + "learning_rate": 1.9991792791832805e-05, + "loss": 1.029, + "step": 1805 + }, + { + "epoch": 0.04, + "grad_norm": 2.3563295401942703, + "learning_rate": 1.999176185470241e-05, + "loss": 0.9963, + "step": 1806 + }, + { + "epoch": 0.04, + "grad_norm": 1.9117928457627849, + "learning_rate": 1.9991730859396822e-05, + "loss": 1.0743, + "step": 1807 + }, + { + "epoch": 0.04, + "grad_norm": 1.234735649914742, + "learning_rate": 1.9991699805916207e-05, + "loss": 0.979, + "step": 1808 + }, + { + "epoch": 0.04, + "grad_norm": 2.626523127691918, + "learning_rate": 1.9991668694260756e-05, + "loss": 1.1703, + "step": 1809 + }, + { + "epoch": 0.04, + "grad_norm": 2.3045258190534086, + "learning_rate": 1.9991637524430642e-05, + "loss": 0.9305, + "step": 1810 + }, + { + "epoch": 0.04, + "grad_norm": 1.1379769823833816, + "learning_rate": 1.9991606296426054e-05, + "loss": 0.9707, + "step": 1811 + }, + { + "epoch": 0.04, + "grad_norm": 3.0936136655196798, + "learning_rate": 1.9991575010247173e-05, + "loss": 1.071, + "step": 1812 + }, + { + "epoch": 0.04, + "grad_norm": 2.14622028566336, + "learning_rate": 1.9991543665894177e-05, + "loss": 1.1188, + "step": 1813 + }, + { + "epoch": 0.04, + "grad_norm": 1.1717629433740615, + "learning_rate": 1.999151226336725e-05, + "loss": 1.0315, + "step": 1814 + }, + { + "epoch": 0.04, + "grad_norm": 1.208661355268937, + "learning_rate": 1.9991480802666573e-05, + "loss": 0.9068, + "step": 1815 + }, + { + "epoch": 0.04, + "grad_norm": 1.9435104991868095, + "learning_rate": 1.9991449283792338e-05, + "loss": 1.123, + "step": 1816 + }, + { + "epoch": 0.04, + "grad_norm": 2.079585943063883, + "learning_rate": 1.9991417706744718e-05, + "loss": 1.0636, + "step": 1817 + }, + { + "epoch": 0.04, + "grad_norm": 2.3145246207485553, + "learning_rate": 1.99913860715239e-05, + "loss": 1.0452, + "step": 1818 + }, + { + "epoch": 0.04, + "grad_norm": 2.077581300792039, + "learning_rate": 1.9991354378130074e-05, + "loss": 1.0269, + "step": 1819 + }, + { + "epoch": 0.04, + "grad_norm": 2.2198601987514888, + "learning_rate": 1.9991322626563413e-05, + "loss": 1.0535, + "step": 1820 + }, + { + "epoch": 0.04, + "grad_norm": 2.0614471485807506, + "learning_rate": 1.9991290816824112e-05, + "loss": 1.1645, + "step": 1821 + }, + { + "epoch": 0.04, + "grad_norm": 2.409771640677015, + "learning_rate": 1.999125894891235e-05, + "loss": 1.1641, + "step": 1822 + }, + { + "epoch": 0.04, + "grad_norm": 2.321165025455947, + "learning_rate": 1.999122702282832e-05, + "loss": 1.276, + "step": 1823 + }, + { + "epoch": 0.04, + "grad_norm": 2.0571100050535276, + "learning_rate": 1.99911950385722e-05, + "loss": 1.0592, + "step": 1824 + }, + { + "epoch": 0.04, + "grad_norm": 1.9981826325224583, + "learning_rate": 1.999116299614418e-05, + "loss": 1.0223, + "step": 1825 + }, + { + "epoch": 0.04, + "grad_norm": 2.072529532618809, + "learning_rate": 1.9991130895544443e-05, + "loss": 1.0356, + "step": 1826 + }, + { + "epoch": 0.04, + "grad_norm": 2.0455060833578815, + "learning_rate": 1.999109873677318e-05, + "loss": 1.0209, + "step": 1827 + }, + { + "epoch": 0.04, + "grad_norm": 2.2526153611606023, + "learning_rate": 1.9991066519830584e-05, + "loss": 1.0007, + "step": 1828 + }, + { + "epoch": 0.04, + "grad_norm": 2.3315707228042264, + "learning_rate": 1.9991034244716826e-05, + "loss": 1.0098, + "step": 1829 + }, + { + "epoch": 0.04, + "grad_norm": 1.2372336005916191, + "learning_rate": 1.9991001911432108e-05, + "loss": 1.0152, + "step": 1830 + }, + { + "epoch": 0.04, + "grad_norm": 2.330148624037533, + "learning_rate": 1.9990969519976614e-05, + "loss": 1.0021, + "step": 1831 + }, + { + "epoch": 0.04, + "grad_norm": 2.5076792003248514, + "learning_rate": 1.999093707035053e-05, + "loss": 1.057, + "step": 1832 + }, + { + "epoch": 0.04, + "grad_norm": 2.4992813355168857, + "learning_rate": 1.9990904562554048e-05, + "loss": 1.1304, + "step": 1833 + }, + { + "epoch": 0.04, + "grad_norm": 3.2217147584725128, + "learning_rate": 1.999087199658736e-05, + "loss": 1.1323, + "step": 1834 + }, + { + "epoch": 0.04, + "grad_norm": 2.326648586740779, + "learning_rate": 1.999083937245065e-05, + "loss": 1.1311, + "step": 1835 + }, + { + "epoch": 0.04, + "grad_norm": 2.3849611474590726, + "learning_rate": 1.999080669014411e-05, + "loss": 0.9969, + "step": 1836 + }, + { + "epoch": 0.04, + "grad_norm": 2.1306904205479147, + "learning_rate": 1.999077394966793e-05, + "loss": 1.1118, + "step": 1837 + }, + { + "epoch": 0.04, + "grad_norm": 3.009218588760998, + "learning_rate": 1.9990741151022302e-05, + "loss": 1.1976, + "step": 1838 + }, + { + "epoch": 0.04, + "grad_norm": 2.6584564062653806, + "learning_rate": 1.9990708294207414e-05, + "loss": 0.9088, + "step": 1839 + }, + { + "epoch": 0.04, + "grad_norm": 1.930320697273623, + "learning_rate": 1.999067537922346e-05, + "loss": 1.0624, + "step": 1840 + }, + { + "epoch": 0.04, + "grad_norm": 2.0458195015603753, + "learning_rate": 1.9990642406070634e-05, + "loss": 0.9529, + "step": 1841 + }, + { + "epoch": 0.04, + "grad_norm": 2.1887185183427484, + "learning_rate": 1.999060937474912e-05, + "loss": 1.0708, + "step": 1842 + }, + { + "epoch": 0.04, + "grad_norm": 2.006924116791548, + "learning_rate": 1.999057628525912e-05, + "loss": 1.1018, + "step": 1843 + }, + { + "epoch": 0.04, + "grad_norm": 2.2646944523323396, + "learning_rate": 1.9990543137600816e-05, + "loss": 1.0929, + "step": 1844 + }, + { + "epoch": 0.04, + "grad_norm": 2.0904434539078243, + "learning_rate": 1.9990509931774412e-05, + "loss": 1.0777, + "step": 1845 + }, + { + "epoch": 0.04, + "grad_norm": 2.1414443776797745, + "learning_rate": 1.9990476667780097e-05, + "loss": 1.119, + "step": 1846 + }, + { + "epoch": 0.04, + "grad_norm": 2.238864047739549, + "learning_rate": 1.999044334561806e-05, + "loss": 1.1032, + "step": 1847 + }, + { + "epoch": 0.04, + "grad_norm": 2.109386024512516, + "learning_rate": 1.9990409965288503e-05, + "loss": 1.0709, + "step": 1848 + }, + { + "epoch": 0.04, + "grad_norm": 2.207599581170092, + "learning_rate": 1.999037652679161e-05, + "loss": 1.1157, + "step": 1849 + }, + { + "epoch": 0.04, + "grad_norm": 2.349671624612379, + "learning_rate": 1.9990343030127588e-05, + "loss": 1.0778, + "step": 1850 + }, + { + "epoch": 0.04, + "grad_norm": 1.2626697690007176, + "learning_rate": 1.9990309475296623e-05, + "loss": 0.9333, + "step": 1851 + }, + { + "epoch": 0.04, + "grad_norm": 2.2207044970072833, + "learning_rate": 1.9990275862298916e-05, + "loss": 1.0294, + "step": 1852 + }, + { + "epoch": 0.04, + "grad_norm": 2.439108834297748, + "learning_rate": 1.9990242191134657e-05, + "loss": 0.993, + "step": 1853 + }, + { + "epoch": 0.04, + "grad_norm": 2.302787470953646, + "learning_rate": 1.999020846180405e-05, + "loss": 1.0886, + "step": 1854 + }, + { + "epoch": 0.04, + "grad_norm": 2.0848607284889065, + "learning_rate": 1.999017467430728e-05, + "loss": 1.0463, + "step": 1855 + }, + { + "epoch": 0.04, + "grad_norm": 1.9663174370641898, + "learning_rate": 1.9990140828644553e-05, + "loss": 1.1033, + "step": 1856 + }, + { + "epoch": 0.04, + "grad_norm": 2.1188758246272483, + "learning_rate": 1.999010692481606e-05, + "loss": 1.0202, + "step": 1857 + }, + { + "epoch": 0.04, + "grad_norm": 2.478679781746434, + "learning_rate": 1.999007296282201e-05, + "loss": 1.0594, + "step": 1858 + }, + { + "epoch": 0.04, + "grad_norm": 2.1874707920389485, + "learning_rate": 1.9990038942662585e-05, + "loss": 1.0517, + "step": 1859 + }, + { + "epoch": 0.04, + "grad_norm": 2.3940652958674726, + "learning_rate": 1.9990004864337994e-05, + "loss": 1.1228, + "step": 1860 + }, + { + "epoch": 0.04, + "grad_norm": 2.3463851151103863, + "learning_rate": 1.998997072784843e-05, + "loss": 0.9671, + "step": 1861 + }, + { + "epoch": 0.04, + "grad_norm": 2.1397911808299894, + "learning_rate": 1.9989936533194093e-05, + "loss": 0.9759, + "step": 1862 + }, + { + "epoch": 0.04, + "grad_norm": 4.127391198878549, + "learning_rate": 1.9989902280375185e-05, + "loss": 1.1217, + "step": 1863 + }, + { + "epoch": 0.04, + "grad_norm": 1.9994598526335248, + "learning_rate": 1.9989867969391904e-05, + "loss": 1.1566, + "step": 1864 + }, + { + "epoch": 0.04, + "grad_norm": 2.1601658768292773, + "learning_rate": 1.9989833600244446e-05, + "loss": 0.9903, + "step": 1865 + }, + { + "epoch": 0.04, + "grad_norm": 2.176199829850349, + "learning_rate": 1.9989799172933018e-05, + "loss": 1.1303, + "step": 1866 + }, + { + "epoch": 0.04, + "grad_norm": 2.273580436252365, + "learning_rate": 1.9989764687457813e-05, + "loss": 1.2324, + "step": 1867 + }, + { + "epoch": 0.04, + "grad_norm": 2.1286008146552473, + "learning_rate": 1.9989730143819033e-05, + "loss": 1.1215, + "step": 1868 + }, + { + "epoch": 0.04, + "grad_norm": 2.3283266980811876, + "learning_rate": 1.9989695542016887e-05, + "loss": 1.1587, + "step": 1869 + }, + { + "epoch": 0.04, + "grad_norm": 1.2331895486184339, + "learning_rate": 1.998966088205157e-05, + "loss": 1.0259, + "step": 1870 + }, + { + "epoch": 0.04, + "grad_norm": 3.429780686998126, + "learning_rate": 1.998962616392328e-05, + "loss": 1.0642, + "step": 1871 + }, + { + "epoch": 0.04, + "grad_norm": 2.0972112702239163, + "learning_rate": 1.998959138763223e-05, + "loss": 1.1012, + "step": 1872 + }, + { + "epoch": 0.04, + "grad_norm": 2.286274182809853, + "learning_rate": 1.998955655317861e-05, + "loss": 1.2201, + "step": 1873 + }, + { + "epoch": 0.04, + "grad_norm": 2.401251027430268, + "learning_rate": 1.998952166056263e-05, + "loss": 1.0839, + "step": 1874 + }, + { + "epoch": 0.04, + "grad_norm": 1.9752245466167477, + "learning_rate": 1.99894867097845e-05, + "loss": 1.1699, + "step": 1875 + }, + { + "epoch": 0.04, + "grad_norm": 2.1147654848920654, + "learning_rate": 1.998945170084441e-05, + "loss": 1.0545, + "step": 1876 + }, + { + "epoch": 0.04, + "grad_norm": 2.223605565078819, + "learning_rate": 1.998941663374257e-05, + "loss": 1.1295, + "step": 1877 + }, + { + "epoch": 0.04, + "grad_norm": 2.440687256417706, + "learning_rate": 1.998938150847918e-05, + "loss": 1.0852, + "step": 1878 + }, + { + "epoch": 0.04, + "grad_norm": 1.2263765892741323, + "learning_rate": 1.9989346325054454e-05, + "loss": 1.0741, + "step": 1879 + }, + { + "epoch": 0.04, + "grad_norm": 1.1498572156437492, + "learning_rate": 1.998931108346859e-05, + "loss": 0.9481, + "step": 1880 + }, + { + "epoch": 0.04, + "grad_norm": 2.252228524732887, + "learning_rate": 1.998927578372179e-05, + "loss": 1.1804, + "step": 1881 + }, + { + "epoch": 0.04, + "grad_norm": 2.400523708540354, + "learning_rate": 1.998924042581427e-05, + "loss": 1.1, + "step": 1882 + }, + { + "epoch": 0.04, + "grad_norm": 2.6064289874312023, + "learning_rate": 1.9989205009746227e-05, + "loss": 1.1035, + "step": 1883 + }, + { + "epoch": 0.04, + "grad_norm": 1.1536789487098256, + "learning_rate": 1.998916953551787e-05, + "loss": 0.9887, + "step": 1884 + }, + { + "epoch": 0.04, + "grad_norm": 2.1439958921019318, + "learning_rate": 1.9989134003129405e-05, + "loss": 0.9836, + "step": 1885 + }, + { + "epoch": 0.04, + "grad_norm": 1.89974346908375, + "learning_rate": 1.998909841258104e-05, + "loss": 1.1636, + "step": 1886 + }, + { + "epoch": 0.04, + "grad_norm": 2.28232182531226, + "learning_rate": 1.9989062763872978e-05, + "loss": 1.123, + "step": 1887 + }, + { + "epoch": 0.04, + "grad_norm": 2.3436130110125104, + "learning_rate": 1.9989027057005438e-05, + "loss": 0.8718, + "step": 1888 + }, + { + "epoch": 0.04, + "grad_norm": 1.216464621093371, + "learning_rate": 1.9988991291978613e-05, + "loss": 0.9911, + "step": 1889 + }, + { + "epoch": 0.04, + "grad_norm": 2.1124849709880493, + "learning_rate": 1.998895546879272e-05, + "loss": 1.0218, + "step": 1890 + }, + { + "epoch": 0.04, + "grad_norm": 2.1537752218089037, + "learning_rate": 1.9988919587447967e-05, + "loss": 1.0143, + "step": 1891 + }, + { + "epoch": 0.04, + "grad_norm": 2.118888546908689, + "learning_rate": 1.9988883647944563e-05, + "loss": 1.1793, + "step": 1892 + }, + { + "epoch": 0.04, + "grad_norm": 2.035049789945014, + "learning_rate": 1.998884765028271e-05, + "loss": 1.0511, + "step": 1893 + }, + { + "epoch": 0.04, + "grad_norm": 2.154249928616551, + "learning_rate": 1.9988811594462628e-05, + "loss": 1.1638, + "step": 1894 + }, + { + "epoch": 0.04, + "grad_norm": 2.2467914481414004, + "learning_rate": 1.9988775480484524e-05, + "loss": 1.0432, + "step": 1895 + }, + { + "epoch": 0.04, + "grad_norm": 2.275918465589491, + "learning_rate": 1.9988739308348603e-05, + "loss": 1.1818, + "step": 1896 + }, + { + "epoch": 0.04, + "grad_norm": 3.648712837415141, + "learning_rate": 1.998870307805508e-05, + "loss": 1.074, + "step": 1897 + }, + { + "epoch": 0.04, + "grad_norm": 2.055487028912065, + "learning_rate": 1.998866678960417e-05, + "loss": 1.168, + "step": 1898 + }, + { + "epoch": 0.04, + "grad_norm": 1.9906986757695586, + "learning_rate": 1.9988630442996072e-05, + "loss": 1.0738, + "step": 1899 + }, + { + "epoch": 0.04, + "grad_norm": 3.1601849155782165, + "learning_rate": 1.9988594038231012e-05, + "loss": 1.1526, + "step": 1900 + }, + { + "epoch": 0.04, + "grad_norm": 2.3518635545819766, + "learning_rate": 1.9988557575309192e-05, + "loss": 1.0251, + "step": 1901 + }, + { + "epoch": 0.04, + "grad_norm": 2.207145601344215, + "learning_rate": 1.9988521054230828e-05, + "loss": 1.0767, + "step": 1902 + }, + { + "epoch": 0.04, + "grad_norm": 1.2646144247452071, + "learning_rate": 1.9988484474996134e-05, + "loss": 1.0216, + "step": 1903 + }, + { + "epoch": 0.04, + "grad_norm": 2.3088965320269708, + "learning_rate": 1.9988447837605317e-05, + "loss": 1.218, + "step": 1904 + }, + { + "epoch": 0.04, + "grad_norm": 2.071844000534093, + "learning_rate": 1.9988411142058596e-05, + "loss": 0.9763, + "step": 1905 + }, + { + "epoch": 0.04, + "grad_norm": 2.1944102269199632, + "learning_rate": 1.9988374388356186e-05, + "loss": 1.0306, + "step": 1906 + }, + { + "epoch": 0.04, + "grad_norm": 2.2471061017345786, + "learning_rate": 1.9988337576498296e-05, + "loss": 1.0152, + "step": 1907 + }, + { + "epoch": 0.04, + "grad_norm": 2.0038592450283197, + "learning_rate": 1.998830070648514e-05, + "loss": 0.9996, + "step": 1908 + }, + { + "epoch": 0.04, + "grad_norm": 2.040957606211939, + "learning_rate": 1.9988263778316937e-05, + "loss": 1.1241, + "step": 1909 + }, + { + "epoch": 0.04, + "grad_norm": 2.207507391275861, + "learning_rate": 1.99882267919939e-05, + "loss": 0.9315, + "step": 1910 + }, + { + "epoch": 0.05, + "grad_norm": 2.4036178352391238, + "learning_rate": 1.9988189747516245e-05, + "loss": 1.1465, + "step": 1911 + }, + { + "epoch": 0.05, + "grad_norm": 2.003809062153739, + "learning_rate": 1.9988152644884186e-05, + "loss": 1.2115, + "step": 1912 + }, + { + "epoch": 0.05, + "grad_norm": 1.99323067922909, + "learning_rate": 1.998811548409794e-05, + "loss": 1.075, + "step": 1913 + }, + { + "epoch": 0.05, + "grad_norm": 2.423462508373546, + "learning_rate": 1.9988078265157727e-05, + "loss": 1.004, + "step": 1914 + }, + { + "epoch": 0.05, + "grad_norm": 2.139978623713932, + "learning_rate": 1.9988040988063755e-05, + "loss": 1.0951, + "step": 1915 + }, + { + "epoch": 0.05, + "grad_norm": 2.004723144837903, + "learning_rate": 1.998800365281625e-05, + "loss": 1.0977, + "step": 1916 + }, + { + "epoch": 0.05, + "grad_norm": 2.0496224474121987, + "learning_rate": 1.9987966259415423e-05, + "loss": 1.123, + "step": 1917 + }, + { + "epoch": 0.05, + "grad_norm": 1.1779454006224839, + "learning_rate": 1.9987928807861494e-05, + "loss": 0.99, + "step": 1918 + }, + { + "epoch": 0.05, + "grad_norm": 2.0451194709617173, + "learning_rate": 1.9987891298154684e-05, + "loss": 1.0986, + "step": 1919 + }, + { + "epoch": 0.05, + "grad_norm": 3.4653085545831326, + "learning_rate": 1.998785373029521e-05, + "loss": 1.0972, + "step": 1920 + }, + { + "epoch": 0.05, + "grad_norm": 2.3286713625710918, + "learning_rate": 1.9987816104283284e-05, + "loss": 1.0172, + "step": 1921 + }, + { + "epoch": 0.05, + "grad_norm": 2.1046626122675134, + "learning_rate": 1.998777842011913e-05, + "loss": 1.0542, + "step": 1922 + }, + { + "epoch": 0.05, + "grad_norm": 2.2169486649258636, + "learning_rate": 1.9987740677802976e-05, + "loss": 1.1431, + "step": 1923 + }, + { + "epoch": 0.05, + "grad_norm": 2.3240044749485214, + "learning_rate": 1.998770287733503e-05, + "loss": 1.213, + "step": 1924 + }, + { + "epoch": 0.05, + "grad_norm": 2.0957702272014442, + "learning_rate": 1.9987665018715514e-05, + "loss": 1.0113, + "step": 1925 + }, + { + "epoch": 0.05, + "grad_norm": 2.293290918954449, + "learning_rate": 1.9987627101944652e-05, + "loss": 1.0527, + "step": 1926 + }, + { + "epoch": 0.05, + "grad_norm": 2.284469016165225, + "learning_rate": 1.998758912702266e-05, + "loss": 0.9746, + "step": 1927 + }, + { + "epoch": 0.05, + "grad_norm": 2.0270033968247065, + "learning_rate": 1.9987551093949765e-05, + "loss": 1.1156, + "step": 1928 + }, + { + "epoch": 0.05, + "grad_norm": 2.325801188451062, + "learning_rate": 1.9987513002726186e-05, + "loss": 0.9654, + "step": 1929 + }, + { + "epoch": 0.05, + "grad_norm": 2.142108112067745, + "learning_rate": 1.9987474853352142e-05, + "loss": 1.0411, + "step": 1930 + }, + { + "epoch": 0.05, + "grad_norm": 2.1261802049249745, + "learning_rate": 1.998743664582786e-05, + "loss": 1.1295, + "step": 1931 + }, + { + "epoch": 0.05, + "grad_norm": 2.264949603779947, + "learning_rate": 1.9987398380153556e-05, + "loss": 0.9994, + "step": 1932 + }, + { + "epoch": 0.05, + "grad_norm": 2.3365250005352505, + "learning_rate": 1.9987360056329458e-05, + "loss": 1.2212, + "step": 1933 + }, + { + "epoch": 0.05, + "grad_norm": 2.36635827382364, + "learning_rate": 1.9987321674355792e-05, + "loss": 1.0402, + "step": 1934 + }, + { + "epoch": 0.05, + "grad_norm": 2.0611991654394486, + "learning_rate": 1.9987283234232777e-05, + "loss": 1.0627, + "step": 1935 + }, + { + "epoch": 0.05, + "grad_norm": 1.9243366412323586, + "learning_rate": 1.998724473596063e-05, + "loss": 1.0811, + "step": 1936 + }, + { + "epoch": 0.05, + "grad_norm": 2.658688331020738, + "learning_rate": 1.9987206179539587e-05, + "loss": 1.1531, + "step": 1937 + }, + { + "epoch": 0.05, + "grad_norm": 2.138351540656712, + "learning_rate": 1.9987167564969868e-05, + "loss": 1.1133, + "step": 1938 + }, + { + "epoch": 0.05, + "grad_norm": 2.4388204790563974, + "learning_rate": 1.9987128892251696e-05, + "loss": 0.9079, + "step": 1939 + }, + { + "epoch": 0.05, + "grad_norm": 2.3213557068080957, + "learning_rate": 1.99870901613853e-05, + "loss": 1.1511, + "step": 1940 + }, + { + "epoch": 0.05, + "grad_norm": 1.968223320149469, + "learning_rate": 1.99870513723709e-05, + "loss": 1.0541, + "step": 1941 + }, + { + "epoch": 0.05, + "grad_norm": 1.1000335888888813, + "learning_rate": 1.9987012525208728e-05, + "loss": 0.9557, + "step": 1942 + }, + { + "epoch": 0.05, + "grad_norm": 2.1602017503916566, + "learning_rate": 1.9986973619899006e-05, + "loss": 1.1752, + "step": 1943 + }, + { + "epoch": 0.05, + "grad_norm": 1.88036609803524, + "learning_rate": 1.9986934656441962e-05, + "loss": 1.0556, + "step": 1944 + }, + { + "epoch": 0.05, + "grad_norm": 2.265408062054676, + "learning_rate": 1.998689563483782e-05, + "loss": 1.1172, + "step": 1945 + }, + { + "epoch": 0.05, + "grad_norm": 2.7303304628695675, + "learning_rate": 1.9986856555086812e-05, + "loss": 1.1408, + "step": 1946 + }, + { + "epoch": 0.05, + "grad_norm": 2.0828136597504576, + "learning_rate": 1.9986817417189168e-05, + "loss": 1.0747, + "step": 1947 + }, + { + "epoch": 0.05, + "grad_norm": 2.274024988655702, + "learning_rate": 1.9986778221145103e-05, + "loss": 1.1111, + "step": 1948 + }, + { + "epoch": 0.05, + "grad_norm": 2.484906449254986, + "learning_rate": 1.9986738966954856e-05, + "loss": 1.0209, + "step": 1949 + }, + { + "epoch": 0.05, + "grad_norm": 1.9579069273303338, + "learning_rate": 1.9986699654618656e-05, + "loss": 1.016, + "step": 1950 + }, + { + "epoch": 0.05, + "grad_norm": 2.0312304853455596, + "learning_rate": 1.9986660284136728e-05, + "loss": 0.986, + "step": 1951 + }, + { + "epoch": 0.05, + "grad_norm": 2.060178791182491, + "learning_rate": 1.99866208555093e-05, + "loss": 1.0904, + "step": 1952 + }, + { + "epoch": 0.05, + "grad_norm": 2.7318053779650544, + "learning_rate": 1.9986581368736607e-05, + "loss": 0.9819, + "step": 1953 + }, + { + "epoch": 0.05, + "grad_norm": 1.1370268224820272, + "learning_rate": 1.9986541823818872e-05, + "loss": 0.9682, + "step": 1954 + }, + { + "epoch": 0.05, + "grad_norm": 3.5167947789311693, + "learning_rate": 1.9986502220756333e-05, + "loss": 1.0896, + "step": 1955 + }, + { + "epoch": 0.05, + "grad_norm": 1.8748324667073424, + "learning_rate": 1.998646255954921e-05, + "loss": 0.9848, + "step": 1956 + }, + { + "epoch": 0.05, + "grad_norm": 3.13957557630758, + "learning_rate": 1.9986422840197746e-05, + "loss": 1.0651, + "step": 1957 + }, + { + "epoch": 0.05, + "grad_norm": 2.104377236070159, + "learning_rate": 1.9986383062702165e-05, + "loss": 1.1132, + "step": 1958 + }, + { + "epoch": 0.05, + "grad_norm": 2.1481604542701156, + "learning_rate": 1.99863432270627e-05, + "loss": 1.1262, + "step": 1959 + }, + { + "epoch": 0.05, + "grad_norm": 3.420250287972538, + "learning_rate": 1.998630333327958e-05, + "loss": 1.0069, + "step": 1960 + }, + { + "epoch": 0.05, + "grad_norm": 2.1117738095436134, + "learning_rate": 1.9986263381353045e-05, + "loss": 0.9981, + "step": 1961 + }, + { + "epoch": 0.05, + "grad_norm": 1.1159362630228362, + "learning_rate": 1.998622337128332e-05, + "loss": 0.9631, + "step": 1962 + }, + { + "epoch": 0.05, + "grad_norm": 2.2662878452162767, + "learning_rate": 1.9986183303070644e-05, + "loss": 1.1681, + "step": 1963 + }, + { + "epoch": 0.05, + "grad_norm": 2.1323444996833043, + "learning_rate": 1.9986143176715246e-05, + "loss": 1.1101, + "step": 1964 + }, + { + "epoch": 0.05, + "grad_norm": 1.1383024872231833, + "learning_rate": 1.9986102992217356e-05, + "loss": 0.8544, + "step": 1965 + }, + { + "epoch": 0.05, + "grad_norm": 2.3963500650180847, + "learning_rate": 1.998606274957722e-05, + "loss": 1.1702, + "step": 1966 + }, + { + "epoch": 0.05, + "grad_norm": 2.1658940504712394, + "learning_rate": 1.9986022448795062e-05, + "loss": 1.2604, + "step": 1967 + }, + { + "epoch": 0.05, + "grad_norm": 2.3652417371219494, + "learning_rate": 1.998598208987112e-05, + "loss": 0.9965, + "step": 1968 + }, + { + "epoch": 0.05, + "grad_norm": 1.9572914630105802, + "learning_rate": 1.998594167280563e-05, + "loss": 0.9585, + "step": 1969 + }, + { + "epoch": 0.05, + "grad_norm": 3.5326557093181035, + "learning_rate": 1.998590119759882e-05, + "loss": 1.0999, + "step": 1970 + }, + { + "epoch": 0.05, + "grad_norm": 2.4760182935168453, + "learning_rate": 1.998586066425094e-05, + "loss": 1.0713, + "step": 1971 + }, + { + "epoch": 0.05, + "grad_norm": 2.130866061308534, + "learning_rate": 1.9985820072762213e-05, + "loss": 1.0557, + "step": 1972 + }, + { + "epoch": 0.05, + "grad_norm": 1.931754531240462, + "learning_rate": 1.9985779423132883e-05, + "loss": 1.1786, + "step": 1973 + }, + { + "epoch": 0.05, + "grad_norm": 2.3417495996381823, + "learning_rate": 1.9985738715363185e-05, + "loss": 1.0742, + "step": 1974 + }, + { + "epoch": 0.05, + "grad_norm": 1.1659278059800904, + "learning_rate": 1.9985697949453348e-05, + "loss": 0.9729, + "step": 1975 + }, + { + "epoch": 0.05, + "grad_norm": 2.147925527432872, + "learning_rate": 1.998565712540362e-05, + "loss": 0.9308, + "step": 1976 + }, + { + "epoch": 0.05, + "grad_norm": 2.2880327486782934, + "learning_rate": 1.9985616243214236e-05, + "loss": 1.064, + "step": 1977 + }, + { + "epoch": 0.05, + "grad_norm": 3.656499278026792, + "learning_rate": 1.9985575302885435e-05, + "loss": 0.9589, + "step": 1978 + }, + { + "epoch": 0.05, + "grad_norm": 1.1453943432140734, + "learning_rate": 1.9985534304417448e-05, + "loss": 1.0129, + "step": 1979 + }, + { + "epoch": 0.05, + "grad_norm": 2.744054386711551, + "learning_rate": 1.998549324781052e-05, + "loss": 1.0963, + "step": 1980 + }, + { + "epoch": 0.05, + "grad_norm": 2.3209799035883534, + "learning_rate": 1.998545213306489e-05, + "loss": 1.0138, + "step": 1981 + }, + { + "epoch": 0.05, + "grad_norm": 2.4162714103101304, + "learning_rate": 1.9985410960180795e-05, + "loss": 0.9652, + "step": 1982 + }, + { + "epoch": 0.05, + "grad_norm": 2.366690832913769, + "learning_rate": 1.9985369729158476e-05, + "loss": 1.0526, + "step": 1983 + }, + { + "epoch": 0.05, + "grad_norm": 2.0511498604176817, + "learning_rate": 1.9985328439998176e-05, + "loss": 1.0585, + "step": 1984 + }, + { + "epoch": 0.05, + "grad_norm": 2.077273510564627, + "learning_rate": 1.998528709270013e-05, + "loss": 1.1277, + "step": 1985 + }, + { + "epoch": 0.05, + "grad_norm": 2.386252839791671, + "learning_rate": 1.9985245687264583e-05, + "loss": 1.0182, + "step": 1986 + }, + { + "epoch": 0.05, + "grad_norm": 1.0877928524024123, + "learning_rate": 1.9985204223691773e-05, + "loss": 0.9797, + "step": 1987 + }, + { + "epoch": 0.05, + "grad_norm": 1.1325136135690443, + "learning_rate": 1.9985162701981942e-05, + "loss": 0.9465, + "step": 1988 + }, + { + "epoch": 0.05, + "grad_norm": 2.1284208567906684, + "learning_rate": 1.9985121122135334e-05, + "loss": 1.0838, + "step": 1989 + }, + { + "epoch": 0.05, + "grad_norm": 2.119192452609994, + "learning_rate": 1.998507948415219e-05, + "loss": 1.3012, + "step": 1990 + }, + { + "epoch": 0.05, + "grad_norm": 2.1523850117041117, + "learning_rate": 1.9985037788032753e-05, + "loss": 0.941, + "step": 1991 + }, + { + "epoch": 0.05, + "grad_norm": 2.153837694997592, + "learning_rate": 1.9984996033777263e-05, + "loss": 1.2061, + "step": 1992 + }, + { + "epoch": 0.05, + "grad_norm": 1.95351391413022, + "learning_rate": 1.9984954221385962e-05, + "loss": 1.2096, + "step": 1993 + }, + { + "epoch": 0.05, + "grad_norm": 2.027929622874626, + "learning_rate": 1.99849123508591e-05, + "loss": 1.1137, + "step": 1994 + }, + { + "epoch": 0.05, + "grad_norm": 2.078690717232401, + "learning_rate": 1.998487042219692e-05, + "loss": 0.985, + "step": 1995 + }, + { + "epoch": 0.05, + "grad_norm": 3.144720110678731, + "learning_rate": 1.9984828435399657e-05, + "loss": 1.0698, + "step": 1996 + }, + { + "epoch": 0.05, + "grad_norm": 2.3035770387857792, + "learning_rate": 1.9984786390467566e-05, + "loss": 0.9478, + "step": 1997 + }, + { + "epoch": 0.05, + "grad_norm": 2.1112363691268237, + "learning_rate": 1.9984744287400883e-05, + "loss": 1.201, + "step": 1998 + }, + { + "epoch": 0.05, + "grad_norm": 2.0231084781635875, + "learning_rate": 1.9984702126199863e-05, + "loss": 1.0771, + "step": 1999 + }, + { + "epoch": 0.05, + "grad_norm": 2.0090821723542063, + "learning_rate": 1.9984659906864742e-05, + "loss": 1.1846, + "step": 2000 + }, + { + "epoch": 0.05, + "grad_norm": 2.4230792876679663, + "learning_rate": 1.998461762939577e-05, + "loss": 1.0745, + "step": 2001 + }, + { + "epoch": 0.05, + "grad_norm": 2.3151500072243505, + "learning_rate": 1.9984575293793198e-05, + "loss": 1.0606, + "step": 2002 + }, + { + "epoch": 0.05, + "grad_norm": 2.672162463663433, + "learning_rate": 1.9984532900057263e-05, + "loss": 1.1056, + "step": 2003 + }, + { + "epoch": 0.05, + "grad_norm": 2.2143629185426135, + "learning_rate": 1.998449044818822e-05, + "loss": 1.0834, + "step": 2004 + }, + { + "epoch": 0.05, + "grad_norm": 2.613321836260453, + "learning_rate": 1.998444793818631e-05, + "loss": 1.0777, + "step": 2005 + }, + { + "epoch": 0.05, + "grad_norm": 1.1947640802177875, + "learning_rate": 1.9984405370051783e-05, + "loss": 0.9935, + "step": 2006 + }, + { + "epoch": 0.05, + "grad_norm": 2.2516318625153926, + "learning_rate": 1.9984362743784887e-05, + "loss": 1.0019, + "step": 2007 + }, + { + "epoch": 0.05, + "grad_norm": 2.320006807577154, + "learning_rate": 1.998432005938587e-05, + "loss": 1.0802, + "step": 2008 + }, + { + "epoch": 0.05, + "grad_norm": 2.3007386089074906, + "learning_rate": 1.998427731685498e-05, + "loss": 1.0888, + "step": 2009 + }, + { + "epoch": 0.05, + "grad_norm": 2.6314645301670923, + "learning_rate": 1.998423451619247e-05, + "loss": 1.0688, + "step": 2010 + }, + { + "epoch": 0.05, + "grad_norm": 2.2247290673262836, + "learning_rate": 1.998419165739858e-05, + "loss": 0.998, + "step": 2011 + }, + { + "epoch": 0.05, + "grad_norm": 2.7446183269088493, + "learning_rate": 1.998414874047357e-05, + "loss": 1.1011, + "step": 2012 + }, + { + "epoch": 0.05, + "grad_norm": 2.1384396905474437, + "learning_rate": 1.998410576541768e-05, + "loss": 1.0513, + "step": 2013 + }, + { + "epoch": 0.05, + "grad_norm": 2.074588303256057, + "learning_rate": 1.9984062732231174e-05, + "loss": 1.1634, + "step": 2014 + }, + { + "epoch": 0.05, + "grad_norm": 1.9047105068529824, + "learning_rate": 1.9984019640914285e-05, + "loss": 0.9548, + "step": 2015 + }, + { + "epoch": 0.05, + "grad_norm": 2.231132851892558, + "learning_rate": 1.9983976491467278e-05, + "loss": 1.0683, + "step": 2016 + }, + { + "epoch": 0.05, + "grad_norm": 2.062846223352965, + "learning_rate": 1.9983933283890398e-05, + "loss": 1.0537, + "step": 2017 + }, + { + "epoch": 0.05, + "grad_norm": 2.2509188408805123, + "learning_rate": 1.9983890018183896e-05, + "loss": 1.0233, + "step": 2018 + }, + { + "epoch": 0.05, + "grad_norm": 2.0285996905627117, + "learning_rate": 1.9983846694348027e-05, + "loss": 1.1557, + "step": 2019 + }, + { + "epoch": 0.05, + "grad_norm": 2.066394775566989, + "learning_rate": 1.9983803312383042e-05, + "loss": 1.1943, + "step": 2020 + }, + { + "epoch": 0.05, + "grad_norm": 1.1828963993963408, + "learning_rate": 1.998375987228919e-05, + "loss": 0.9654, + "step": 2021 + }, + { + "epoch": 0.05, + "grad_norm": 2.257303888356032, + "learning_rate": 1.998371637406673e-05, + "loss": 1.0588, + "step": 2022 + }, + { + "epoch": 0.05, + "grad_norm": 2.7002508471608766, + "learning_rate": 1.9983672817715914e-05, + "loss": 1.1011, + "step": 2023 + }, + { + "epoch": 0.05, + "grad_norm": 2.097568468558321, + "learning_rate": 1.9983629203236993e-05, + "loss": 1.1755, + "step": 2024 + }, + { + "epoch": 0.05, + "grad_norm": 1.9950883657182013, + "learning_rate": 1.998358553063022e-05, + "loss": 0.9917, + "step": 2025 + }, + { + "epoch": 0.05, + "grad_norm": 2.3943202183805927, + "learning_rate": 1.998354179989585e-05, + "loss": 1.0505, + "step": 2026 + }, + { + "epoch": 0.05, + "grad_norm": 2.367241720524429, + "learning_rate": 1.9983498011034143e-05, + "loss": 1.1005, + "step": 2027 + }, + { + "epoch": 0.05, + "grad_norm": 1.1498388781112578, + "learning_rate": 1.9983454164045352e-05, + "loss": 0.9826, + "step": 2028 + }, + { + "epoch": 0.05, + "grad_norm": 2.262783052746151, + "learning_rate": 1.998341025892972e-05, + "loss": 0.9668, + "step": 2029 + }, + { + "epoch": 0.05, + "grad_norm": 2.043316891466871, + "learning_rate": 1.9983366295687522e-05, + "loss": 1.0294, + "step": 2030 + }, + { + "epoch": 0.05, + "grad_norm": 2.081992339992695, + "learning_rate": 1.9983322274319005e-05, + "loss": 1.0624, + "step": 2031 + }, + { + "epoch": 0.05, + "grad_norm": 2.1351563668887836, + "learning_rate": 1.998327819482442e-05, + "loss": 1.0823, + "step": 2032 + }, + { + "epoch": 0.05, + "grad_norm": 2.2089876521463143, + "learning_rate": 1.998323405720403e-05, + "loss": 1.0354, + "step": 2033 + }, + { + "epoch": 0.05, + "grad_norm": 2.0020028290742378, + "learning_rate": 1.9983189861458093e-05, + "loss": 1.1776, + "step": 2034 + }, + { + "epoch": 0.05, + "grad_norm": 1.2039915894704412, + "learning_rate": 1.9983145607586862e-05, + "loss": 1.0253, + "step": 2035 + }, + { + "epoch": 0.05, + "grad_norm": 2.1093756658610423, + "learning_rate": 1.9983101295590597e-05, + "loss": 1.0994, + "step": 2036 + }, + { + "epoch": 0.05, + "grad_norm": 2.8806943180192826, + "learning_rate": 1.9983056925469558e-05, + "loss": 1.1257, + "step": 2037 + }, + { + "epoch": 0.05, + "grad_norm": 2.753195193915557, + "learning_rate": 1.9983012497223994e-05, + "loss": 1.2146, + "step": 2038 + }, + { + "epoch": 0.05, + "grad_norm": 2.2896288388435315, + "learning_rate": 1.998296801085418e-05, + "loss": 1.0911, + "step": 2039 + }, + { + "epoch": 0.05, + "grad_norm": 2.0845741524049357, + "learning_rate": 1.9982923466360357e-05, + "loss": 1.142, + "step": 2040 + }, + { + "epoch": 0.05, + "grad_norm": 2.3277495838791262, + "learning_rate": 1.9982878863742798e-05, + "loss": 0.9917, + "step": 2041 + }, + { + "epoch": 0.05, + "grad_norm": 2.269926099193964, + "learning_rate": 1.9982834203001757e-05, + "loss": 1.0061, + "step": 2042 + }, + { + "epoch": 0.05, + "grad_norm": 2.2721909146194053, + "learning_rate": 1.9982789484137495e-05, + "loss": 1.1233, + "step": 2043 + }, + { + "epoch": 0.05, + "grad_norm": 2.1844535200849853, + "learning_rate": 1.9982744707150272e-05, + "loss": 1.0017, + "step": 2044 + }, + { + "epoch": 0.05, + "grad_norm": 2.0162285287118844, + "learning_rate": 1.9982699872040346e-05, + "loss": 1.0988, + "step": 2045 + }, + { + "epoch": 0.05, + "grad_norm": 1.2720706499246588, + "learning_rate": 1.9982654978807986e-05, + "loss": 0.9891, + "step": 2046 + }, + { + "epoch": 0.05, + "grad_norm": 2.208510792894383, + "learning_rate": 1.9982610027453447e-05, + "loss": 0.9827, + "step": 2047 + }, + { + "epoch": 0.05, + "grad_norm": 2.0535841349211794, + "learning_rate": 1.9982565017976992e-05, + "loss": 1.0868, + "step": 2048 + }, + { + "epoch": 0.05, + "grad_norm": 2.1713419983197966, + "learning_rate": 1.998251995037888e-05, + "loss": 1.1488, + "step": 2049 + }, + { + "epoch": 0.05, + "grad_norm": 2.38606065735636, + "learning_rate": 1.9982474824659378e-05, + "loss": 0.9907, + "step": 2050 + }, + { + "epoch": 0.05, + "grad_norm": 4.859438515185576, + "learning_rate": 1.9982429640818747e-05, + "loss": 1.0822, + "step": 2051 + }, + { + "epoch": 0.05, + "grad_norm": 1.1511866227329186, + "learning_rate": 1.998238439885725e-05, + "loss": 1.0288, + "step": 2052 + }, + { + "epoch": 0.05, + "grad_norm": 2.1574125214083346, + "learning_rate": 1.998233909877515e-05, + "loss": 1.1292, + "step": 2053 + }, + { + "epoch": 0.05, + "grad_norm": 2.050928368343657, + "learning_rate": 1.9982293740572714e-05, + "loss": 1.025, + "step": 2054 + }, + { + "epoch": 0.05, + "grad_norm": 2.5292615582419637, + "learning_rate": 1.99822483242502e-05, + "loss": 1.1338, + "step": 2055 + }, + { + "epoch": 0.05, + "grad_norm": 2.222513220438843, + "learning_rate": 1.998220284980788e-05, + "loss": 1.1695, + "step": 2056 + }, + { + "epoch": 0.05, + "grad_norm": 3.683922830368823, + "learning_rate": 1.9982157317246015e-05, + "loss": 1.0385, + "step": 2057 + }, + { + "epoch": 0.05, + "grad_norm": 2.0523497684402856, + "learning_rate": 1.9982111726564865e-05, + "loss": 1.1062, + "step": 2058 + }, + { + "epoch": 0.05, + "grad_norm": 2.5710030432686266, + "learning_rate": 1.9982066077764702e-05, + "loss": 1.0814, + "step": 2059 + }, + { + "epoch": 0.05, + "grad_norm": 2.098462537560569, + "learning_rate": 1.998202037084579e-05, + "loss": 1.0944, + "step": 2060 + }, + { + "epoch": 0.05, + "grad_norm": 2.3686999938605537, + "learning_rate": 1.9981974605808398e-05, + "loss": 1.1048, + "step": 2061 + }, + { + "epoch": 0.05, + "grad_norm": 2.0379978582672744, + "learning_rate": 1.9981928782652788e-05, + "loss": 1.0598, + "step": 2062 + }, + { + "epoch": 0.05, + "grad_norm": 2.3085885789130725, + "learning_rate": 1.9981882901379228e-05, + "loss": 1.0864, + "step": 2063 + }, + { + "epoch": 0.05, + "grad_norm": 2.3454141601618543, + "learning_rate": 1.9981836961987987e-05, + "loss": 1.0733, + "step": 2064 + }, + { + "epoch": 0.05, + "grad_norm": 1.1104222376246726, + "learning_rate": 1.998179096447933e-05, + "loss": 1.0143, + "step": 2065 + }, + { + "epoch": 0.05, + "grad_norm": 2.1388568801420775, + "learning_rate": 1.9981744908853524e-05, + "loss": 1.139, + "step": 2066 + }, + { + "epoch": 0.05, + "grad_norm": 2.099805097550643, + "learning_rate": 1.998169879511084e-05, + "loss": 0.9883, + "step": 2067 + }, + { + "epoch": 0.05, + "grad_norm": 2.049339306235096, + "learning_rate": 1.9981652623251546e-05, + "loss": 1.1291, + "step": 2068 + }, + { + "epoch": 0.05, + "grad_norm": 2.7304821701361246, + "learning_rate": 1.998160639327591e-05, + "loss": 1.0546, + "step": 2069 + }, + { + "epoch": 0.05, + "grad_norm": 2.200118694038052, + "learning_rate": 1.9981560105184205e-05, + "loss": 1.1269, + "step": 2070 + }, + { + "epoch": 0.05, + "grad_norm": 2.4015900588734596, + "learning_rate": 1.998151375897669e-05, + "loss": 1.1256, + "step": 2071 + }, + { + "epoch": 0.05, + "grad_norm": 1.9076899634770426, + "learning_rate": 1.9981467354653645e-05, + "loss": 0.9248, + "step": 2072 + }, + { + "epoch": 0.05, + "grad_norm": 2.2904020106217096, + "learning_rate": 1.998142089221534e-05, + "loss": 1.1351, + "step": 2073 + }, + { + "epoch": 0.05, + "grad_norm": 2.269315438814509, + "learning_rate": 1.9981374371662035e-05, + "loss": 0.9911, + "step": 2074 + }, + { + "epoch": 0.05, + "grad_norm": 1.9363678797813184, + "learning_rate": 1.9981327792994016e-05, + "loss": 1.1238, + "step": 2075 + }, + { + "epoch": 0.05, + "grad_norm": 1.9207338714504922, + "learning_rate": 1.9981281156211543e-05, + "loss": 0.9678, + "step": 2076 + }, + { + "epoch": 0.05, + "grad_norm": 2.16832294878483, + "learning_rate": 1.9981234461314893e-05, + "loss": 1.0494, + "step": 2077 + }, + { + "epoch": 0.05, + "grad_norm": 2.2988090662218856, + "learning_rate": 1.9981187708304335e-05, + "loss": 1.0639, + "step": 2078 + }, + { + "epoch": 0.05, + "grad_norm": 2.2018681257910013, + "learning_rate": 1.9981140897180144e-05, + "loss": 0.9702, + "step": 2079 + }, + { + "epoch": 0.05, + "grad_norm": 1.9782549213496132, + "learning_rate": 1.9981094027942585e-05, + "loss": 1.1974, + "step": 2080 + }, + { + "epoch": 0.05, + "grad_norm": 2.081311159852487, + "learning_rate": 1.998104710059194e-05, + "loss": 1.1523, + "step": 2081 + }, + { + "epoch": 0.05, + "grad_norm": 2.498199127186432, + "learning_rate": 1.998100011512848e-05, + "loss": 1.1276, + "step": 2082 + }, + { + "epoch": 0.05, + "grad_norm": 2.2365959792081864, + "learning_rate": 1.9980953071552477e-05, + "loss": 0.9762, + "step": 2083 + }, + { + "epoch": 0.05, + "grad_norm": 1.9739794574183285, + "learning_rate": 1.9980905969864204e-05, + "loss": 1.0434, + "step": 2084 + }, + { + "epoch": 0.05, + "grad_norm": 2.3354063039087203, + "learning_rate": 1.998085881006394e-05, + "loss": 1.1559, + "step": 2085 + }, + { + "epoch": 0.05, + "grad_norm": 2.2778131239904655, + "learning_rate": 1.998081159215195e-05, + "loss": 0.9055, + "step": 2086 + }, + { + "epoch": 0.05, + "grad_norm": 2.907073621808223, + "learning_rate": 1.998076431612852e-05, + "loss": 1.0831, + "step": 2087 + }, + { + "epoch": 0.05, + "grad_norm": 2.1621991308400554, + "learning_rate": 1.998071698199392e-05, + "loss": 1.0548, + "step": 2088 + }, + { + "epoch": 0.05, + "grad_norm": 2.4300028066064883, + "learning_rate": 1.9980669589748426e-05, + "loss": 1.0757, + "step": 2089 + }, + { + "epoch": 0.05, + "grad_norm": 2.014109966115129, + "learning_rate": 1.998062213939231e-05, + "loss": 1.1529, + "step": 2090 + }, + { + "epoch": 0.05, + "grad_norm": 2.227962750659147, + "learning_rate": 1.9980574630925856e-05, + "loss": 1.0857, + "step": 2091 + }, + { + "epoch": 0.05, + "grad_norm": 2.1735390727384116, + "learning_rate": 1.9980527064349334e-05, + "loss": 1.15, + "step": 2092 + }, + { + "epoch": 0.05, + "grad_norm": 2.208245286462272, + "learning_rate": 1.9980479439663024e-05, + "loss": 1.1034, + "step": 2093 + }, + { + "epoch": 0.05, + "grad_norm": 1.908966666847473, + "learning_rate": 1.9980431756867206e-05, + "loss": 1.1893, + "step": 2094 + }, + { + "epoch": 0.05, + "grad_norm": 2.180417610246678, + "learning_rate": 1.9980384015962152e-05, + "loss": 1.1564, + "step": 2095 + }, + { + "epoch": 0.05, + "grad_norm": 2.0452669432300605, + "learning_rate": 1.9980336216948144e-05, + "loss": 1.119, + "step": 2096 + }, + { + "epoch": 0.05, + "grad_norm": 2.1660069603269414, + "learning_rate": 1.9980288359825457e-05, + "loss": 1.1709, + "step": 2097 + }, + { + "epoch": 0.05, + "grad_norm": 2.8799252917641978, + "learning_rate": 1.9980240444594372e-05, + "loss": 1.1111, + "step": 2098 + }, + { + "epoch": 0.05, + "grad_norm": 2.123701633213404, + "learning_rate": 1.998019247125517e-05, + "loss": 1.1635, + "step": 2099 + }, + { + "epoch": 0.05, + "grad_norm": 4.255025426555278, + "learning_rate": 1.998014443980812e-05, + "loss": 0.9998, + "step": 2100 + }, + { + "epoch": 0.05, + "grad_norm": 1.9876786925119647, + "learning_rate": 1.998009635025352e-05, + "loss": 1.2026, + "step": 2101 + }, + { + "epoch": 0.05, + "grad_norm": 2.178096639540559, + "learning_rate": 1.9980048202591632e-05, + "loss": 1.1316, + "step": 2102 + }, + { + "epoch": 0.05, + "grad_norm": 2.2082257999412938, + "learning_rate": 1.9979999996822744e-05, + "loss": 0.9657, + "step": 2103 + }, + { + "epoch": 0.05, + "grad_norm": 2.4374562916367295, + "learning_rate": 1.997995173294714e-05, + "loss": 1.1167, + "step": 2104 + }, + { + "epoch": 0.05, + "grad_norm": 2.4814108125006538, + "learning_rate": 1.9979903410965095e-05, + "loss": 1.0945, + "step": 2105 + }, + { + "epoch": 0.05, + "grad_norm": 2.1497472065529153, + "learning_rate": 1.9979855030876894e-05, + "loss": 1.1037, + "step": 2106 + }, + { + "epoch": 0.05, + "grad_norm": 2.262829740158922, + "learning_rate": 1.997980659268282e-05, + "loss": 0.9951, + "step": 2107 + }, + { + "epoch": 0.05, + "grad_norm": 2.100241243850292, + "learning_rate": 1.9979758096383147e-05, + "loss": 1.154, + "step": 2108 + }, + { + "epoch": 0.05, + "grad_norm": 2.0712402314163367, + "learning_rate": 1.9979709541978164e-05, + "loss": 0.8903, + "step": 2109 + }, + { + "epoch": 0.05, + "grad_norm": 2.1721820347339142, + "learning_rate": 1.9979660929468157e-05, + "loss": 1.1219, + "step": 2110 + }, + { + "epoch": 0.05, + "grad_norm": 2.7345487090689815, + "learning_rate": 1.99796122588534e-05, + "loss": 1.1632, + "step": 2111 + }, + { + "epoch": 0.05, + "grad_norm": 2.6340499749755932, + "learning_rate": 1.997956353013418e-05, + "loss": 1.0702, + "step": 2112 + }, + { + "epoch": 0.05, + "grad_norm": 1.162305393637044, + "learning_rate": 1.9979514743310787e-05, + "loss": 0.939, + "step": 2113 + }, + { + "epoch": 0.05, + "grad_norm": 2.1188398227923972, + "learning_rate": 1.9979465898383498e-05, + "loss": 1.0102, + "step": 2114 + }, + { + "epoch": 0.05, + "grad_norm": 2.760621838481602, + "learning_rate": 1.9979416995352595e-05, + "loss": 1.109, + "step": 2115 + }, + { + "epoch": 0.05, + "grad_norm": 2.056259143893802, + "learning_rate": 1.997936803421837e-05, + "loss": 1.0501, + "step": 2116 + }, + { + "epoch": 0.05, + "grad_norm": 2.8231130641866145, + "learning_rate": 1.9979319014981105e-05, + "loss": 1.0681, + "step": 2117 + }, + { + "epoch": 0.05, + "grad_norm": 2.3596197086852566, + "learning_rate": 1.9979269937641087e-05, + "loss": 1.0679, + "step": 2118 + }, + { + "epoch": 0.05, + "grad_norm": 2.515069873456969, + "learning_rate": 1.9979220802198597e-05, + "loss": 1.0091, + "step": 2119 + }, + { + "epoch": 0.05, + "grad_norm": 2.2258600617071593, + "learning_rate": 1.9979171608653926e-05, + "loss": 1.1991, + "step": 2120 + }, + { + "epoch": 0.05, + "grad_norm": 2.4902825238678, + "learning_rate": 1.9979122357007355e-05, + "loss": 1.0695, + "step": 2121 + }, + { + "epoch": 0.05, + "grad_norm": 2.2766591442905892, + "learning_rate": 1.9979073047259176e-05, + "loss": 1.1065, + "step": 2122 + }, + { + "epoch": 0.05, + "grad_norm": 2.01694715040635, + "learning_rate": 1.9979023679409676e-05, + "loss": 1.1699, + "step": 2123 + }, + { + "epoch": 0.05, + "grad_norm": 2.1081252082271287, + "learning_rate": 1.997897425345914e-05, + "loss": 0.9075, + "step": 2124 + }, + { + "epoch": 0.05, + "grad_norm": 2.4292251296075253, + "learning_rate": 1.9978924769407858e-05, + "loss": 1.017, + "step": 2125 + }, + { + "epoch": 0.05, + "grad_norm": 1.189574547068232, + "learning_rate": 1.9978875227256115e-05, + "loss": 0.9957, + "step": 2126 + }, + { + "epoch": 0.05, + "grad_norm": 1.1417281562563029, + "learning_rate": 1.99788256270042e-05, + "loss": 0.9723, + "step": 2127 + }, + { + "epoch": 0.05, + "grad_norm": 2.036380435330988, + "learning_rate": 1.9978775968652405e-05, + "loss": 1.033, + "step": 2128 + }, + { + "epoch": 0.05, + "grad_norm": 1.9921882632314867, + "learning_rate": 1.9978726252201018e-05, + "loss": 1.0107, + "step": 2129 + }, + { + "epoch": 0.05, + "grad_norm": 2.0621329089473184, + "learning_rate": 1.9978676477650326e-05, + "loss": 1.1792, + "step": 2130 + }, + { + "epoch": 0.05, + "grad_norm": 2.0408004858395725, + "learning_rate": 1.9978626645000617e-05, + "loss": 1.1204, + "step": 2131 + }, + { + "epoch": 0.05, + "grad_norm": 1.7849895404148088, + "learning_rate": 1.997857675425219e-05, + "loss": 1.0557, + "step": 2132 + }, + { + "epoch": 0.05, + "grad_norm": 2.15758471863793, + "learning_rate": 1.997852680540533e-05, + "loss": 1.2337, + "step": 2133 + }, + { + "epoch": 0.05, + "grad_norm": 1.1328558434629477, + "learning_rate": 1.9978476798460324e-05, + "loss": 0.957, + "step": 2134 + }, + { + "epoch": 0.05, + "grad_norm": 2.4047147390057244, + "learning_rate": 1.997842673341747e-05, + "loss": 0.9408, + "step": 2135 + }, + { + "epoch": 0.05, + "grad_norm": 1.134094096418273, + "learning_rate": 1.9978376610277053e-05, + "loss": 0.8516, + "step": 2136 + }, + { + "epoch": 0.05, + "grad_norm": 2.23924949081312, + "learning_rate": 1.997832642903937e-05, + "loss": 1.0982, + "step": 2137 + }, + { + "epoch": 0.05, + "grad_norm": 2.3076279222639884, + "learning_rate": 1.9978276189704713e-05, + "loss": 1.1547, + "step": 2138 + }, + { + "epoch": 0.05, + "grad_norm": 2.3328720838106904, + "learning_rate": 1.997822589227337e-05, + "loss": 1.0601, + "step": 2139 + }, + { + "epoch": 0.05, + "grad_norm": 1.1488225107637873, + "learning_rate": 1.9978175536745642e-05, + "loss": 1.0064, + "step": 2140 + }, + { + "epoch": 0.05, + "grad_norm": 2.2925037848004943, + "learning_rate": 1.9978125123121812e-05, + "loss": 1.1144, + "step": 2141 + }, + { + "epoch": 0.05, + "grad_norm": 2.06579825086662, + "learning_rate": 1.997807465140218e-05, + "loss": 0.9698, + "step": 2142 + }, + { + "epoch": 0.05, + "grad_norm": 1.1237141440924585, + "learning_rate": 1.997802412158704e-05, + "loss": 0.9787, + "step": 2143 + }, + { + "epoch": 0.05, + "grad_norm": 2.4036619547320175, + "learning_rate": 1.997797353367668e-05, + "loss": 1.1583, + "step": 2144 + }, + { + "epoch": 0.05, + "grad_norm": 2.2228605665884786, + "learning_rate": 1.9977922887671404e-05, + "loss": 0.9861, + "step": 2145 + }, + { + "epoch": 0.05, + "grad_norm": 2.3214626961468827, + "learning_rate": 1.9977872183571502e-05, + "loss": 1.1824, + "step": 2146 + }, + { + "epoch": 0.05, + "grad_norm": 2.22183939653168, + "learning_rate": 1.9977821421377266e-05, + "loss": 1.0203, + "step": 2147 + }, + { + "epoch": 0.05, + "grad_norm": 2.420486864540913, + "learning_rate": 1.9977770601088994e-05, + "loss": 1.1386, + "step": 2148 + }, + { + "epoch": 0.05, + "grad_norm": 2.8168645340738503, + "learning_rate": 1.9977719722706985e-05, + "loss": 0.9337, + "step": 2149 + }, + { + "epoch": 0.05, + "grad_norm": 2.1602185802538285, + "learning_rate": 1.9977668786231536e-05, + "loss": 1.1325, + "step": 2150 + }, + { + "epoch": 0.05, + "grad_norm": 3.281841328746511, + "learning_rate": 1.9977617791662935e-05, + "loss": 1.1528, + "step": 2151 + }, + { + "epoch": 0.05, + "grad_norm": 1.230127178572634, + "learning_rate": 1.9977566739001485e-05, + "loss": 0.9791, + "step": 2152 + }, + { + "epoch": 0.05, + "grad_norm": 1.9998256346687533, + "learning_rate": 1.9977515628247482e-05, + "loss": 1.0709, + "step": 2153 + }, + { + "epoch": 0.05, + "grad_norm": 2.16591314655123, + "learning_rate": 1.997746445940123e-05, + "loss": 1.1844, + "step": 2154 + }, + { + "epoch": 0.05, + "grad_norm": 2.920111856132944, + "learning_rate": 1.9977413232463017e-05, + "loss": 1.0783, + "step": 2155 + }, + { + "epoch": 0.05, + "grad_norm": 2.0568055094043127, + "learning_rate": 1.9977361947433144e-05, + "loss": 1.0529, + "step": 2156 + }, + { + "epoch": 0.05, + "grad_norm": 2.3748935571988223, + "learning_rate": 1.9977310604311912e-05, + "loss": 0.9653, + "step": 2157 + }, + { + "epoch": 0.05, + "grad_norm": 2.0325588478701406, + "learning_rate": 1.9977259203099623e-05, + "loss": 0.9941, + "step": 2158 + }, + { + "epoch": 0.05, + "grad_norm": 2.904710047134431, + "learning_rate": 1.9977207743796566e-05, + "loss": 1.094, + "step": 2159 + }, + { + "epoch": 0.05, + "grad_norm": 1.8472800733186177, + "learning_rate": 1.9977156226403052e-05, + "loss": 1.0513, + "step": 2160 + }, + { + "epoch": 0.05, + "grad_norm": 2.0805035949595596, + "learning_rate": 1.997710465091937e-05, + "loss": 0.9759, + "step": 2161 + }, + { + "epoch": 0.05, + "grad_norm": 1.9738332274105075, + "learning_rate": 1.9977053017345832e-05, + "loss": 1.0943, + "step": 2162 + }, + { + "epoch": 0.05, + "grad_norm": 2.570961601030872, + "learning_rate": 1.997700132568273e-05, + "loss": 1.0605, + "step": 2163 + }, + { + "epoch": 0.05, + "grad_norm": 2.4661456202745176, + "learning_rate": 1.997694957593037e-05, + "loss": 1.0813, + "step": 2164 + }, + { + "epoch": 0.05, + "grad_norm": 1.1969833382533448, + "learning_rate": 1.9976897768089047e-05, + "loss": 0.9629, + "step": 2165 + }, + { + "epoch": 0.05, + "grad_norm": 2.002991093001631, + "learning_rate": 1.9976845902159066e-05, + "loss": 1.1138, + "step": 2166 + }, + { + "epoch": 0.05, + "grad_norm": 1.2171388961703637, + "learning_rate": 1.997679397814073e-05, + "loss": 0.9567, + "step": 2167 + }, + { + "epoch": 0.05, + "grad_norm": 2.0553514287596655, + "learning_rate": 1.9976741996034344e-05, + "loss": 1.131, + "step": 2168 + }, + { + "epoch": 0.05, + "grad_norm": 2.2796112934035384, + "learning_rate": 1.9976689955840206e-05, + "loss": 1.0965, + "step": 2169 + }, + { + "epoch": 0.05, + "grad_norm": 1.147742755452406, + "learning_rate": 1.997663785755862e-05, + "loss": 0.9787, + "step": 2170 + }, + { + "epoch": 0.05, + "grad_norm": 2.1367092906020724, + "learning_rate": 1.997658570118989e-05, + "loss": 1.1921, + "step": 2171 + }, + { + "epoch": 0.05, + "grad_norm": 2.0125306135873475, + "learning_rate": 1.997653348673432e-05, + "loss": 1.1057, + "step": 2172 + }, + { + "epoch": 0.05, + "grad_norm": 2.0378642201518176, + "learning_rate": 1.9976481214192215e-05, + "loss": 1.1035, + "step": 2173 + }, + { + "epoch": 0.05, + "grad_norm": 2.3791744778012225, + "learning_rate": 1.9976428883563876e-05, + "loss": 1.0323, + "step": 2174 + }, + { + "epoch": 0.05, + "grad_norm": 2.2987815338508275, + "learning_rate": 1.997637649484961e-05, + "loss": 1.0826, + "step": 2175 + }, + { + "epoch": 0.05, + "grad_norm": 2.256115234612059, + "learning_rate": 1.997632404804972e-05, + "loss": 1.2668, + "step": 2176 + }, + { + "epoch": 0.05, + "grad_norm": 1.9748759046208109, + "learning_rate": 1.9976271543164514e-05, + "loss": 1.144, + "step": 2177 + }, + { + "epoch": 0.05, + "grad_norm": 1.907345831907123, + "learning_rate": 1.99762189801943e-05, + "loss": 1.1033, + "step": 2178 + }, + { + "epoch": 0.05, + "grad_norm": 2.245146139770438, + "learning_rate": 1.9976166359139375e-05, + "loss": 1.0559, + "step": 2179 + }, + { + "epoch": 0.05, + "grad_norm": 2.3884776916473163, + "learning_rate": 1.9976113680000055e-05, + "loss": 1.0962, + "step": 2180 + }, + { + "epoch": 0.05, + "grad_norm": 2.0064015521666705, + "learning_rate": 1.997606094277664e-05, + "loss": 1.0361, + "step": 2181 + }, + { + "epoch": 0.05, + "grad_norm": 2.7520576388504865, + "learning_rate": 1.997600814746944e-05, + "loss": 1.0071, + "step": 2182 + }, + { + "epoch": 0.05, + "grad_norm": 3.202393992241136, + "learning_rate": 1.9975955294078764e-05, + "loss": 0.9449, + "step": 2183 + }, + { + "epoch": 0.05, + "grad_norm": 2.051577458223619, + "learning_rate": 1.9975902382604916e-05, + "loss": 1.0183, + "step": 2184 + }, + { + "epoch": 0.05, + "grad_norm": 2.002267373821329, + "learning_rate": 1.997584941304821e-05, + "loss": 1.0562, + "step": 2185 + }, + { + "epoch": 0.05, + "grad_norm": 2.1372610233067846, + "learning_rate": 1.9975796385408948e-05, + "loss": 0.966, + "step": 2186 + }, + { + "epoch": 0.05, + "grad_norm": 2.3110850629207387, + "learning_rate": 1.997574329968744e-05, + "loss": 1.0821, + "step": 2187 + }, + { + "epoch": 0.05, + "grad_norm": 1.248060579888819, + "learning_rate": 1.9975690155883998e-05, + "loss": 0.9432, + "step": 2188 + }, + { + "epoch": 0.05, + "grad_norm": 3.3028238401133465, + "learning_rate": 1.997563695399893e-05, + "loss": 0.9712, + "step": 2189 + }, + { + "epoch": 0.05, + "grad_norm": 2.245481543865511, + "learning_rate": 1.997558369403254e-05, + "loss": 1.1433, + "step": 2190 + }, + { + "epoch": 0.05, + "grad_norm": 2.0874331187606403, + "learning_rate": 1.997553037598515e-05, + "loss": 1.0628, + "step": 2191 + }, + { + "epoch": 0.05, + "grad_norm": 1.181437723220865, + "learning_rate": 1.997547699985706e-05, + "loss": 1.013, + "step": 2192 + }, + { + "epoch": 0.05, + "grad_norm": 1.1569160559915561, + "learning_rate": 1.997542356564859e-05, + "loss": 1.0051, + "step": 2193 + }, + { + "epoch": 0.05, + "grad_norm": 2.1341522408568436, + "learning_rate": 1.997537007336004e-05, + "loss": 1.1276, + "step": 2194 + }, + { + "epoch": 0.05, + "grad_norm": 2.8022740149648238, + "learning_rate": 1.997531652299173e-05, + "loss": 1.1343, + "step": 2195 + }, + { + "epoch": 0.05, + "grad_norm": 1.8350631276585685, + "learning_rate": 1.9975262914543968e-05, + "loss": 1.0957, + "step": 2196 + }, + { + "epoch": 0.05, + "grad_norm": 1.9650328746122288, + "learning_rate": 1.997520924801707e-05, + "loss": 0.9667, + "step": 2197 + }, + { + "epoch": 0.05, + "grad_norm": 2.0189872436490512, + "learning_rate": 1.9975155523411343e-05, + "loss": 1.0514, + "step": 2198 + }, + { + "epoch": 0.05, + "grad_norm": 2.012116818834546, + "learning_rate": 1.9975101740727104e-05, + "loss": 1.0214, + "step": 2199 + }, + { + "epoch": 0.05, + "grad_norm": 2.053551435965606, + "learning_rate": 1.9975047899964663e-05, + "loss": 1.1049, + "step": 2200 + }, + { + "epoch": 0.05, + "grad_norm": 1.945261661181685, + "learning_rate": 1.9974994001124336e-05, + "loss": 1.1011, + "step": 2201 + }, + { + "epoch": 0.05, + "grad_norm": 1.1523280411393095, + "learning_rate": 1.9974940044206433e-05, + "loss": 0.9352, + "step": 2202 + }, + { + "epoch": 0.05, + "grad_norm": 2.636456197982921, + "learning_rate": 1.9974886029211275e-05, + "loss": 1.1038, + "step": 2203 + }, + { + "epoch": 0.05, + "grad_norm": 1.1198441581958527, + "learning_rate": 1.9974831956139172e-05, + "loss": 0.9756, + "step": 2204 + }, + { + "epoch": 0.05, + "grad_norm": 1.9726811702495013, + "learning_rate": 1.997477782499044e-05, + "loss": 1.1578, + "step": 2205 + }, + { + "epoch": 0.05, + "grad_norm": 2.289176161404848, + "learning_rate": 1.997472363576539e-05, + "loss": 1.1062, + "step": 2206 + }, + { + "epoch": 0.05, + "grad_norm": 2.020271507662879, + "learning_rate": 1.997466938846434e-05, + "loss": 1.0135, + "step": 2207 + }, + { + "epoch": 0.05, + "grad_norm": 2.0180305276556916, + "learning_rate": 1.9974615083087614e-05, + "loss": 0.9159, + "step": 2208 + }, + { + "epoch": 0.05, + "grad_norm": 2.340575753479826, + "learning_rate": 1.9974560719635517e-05, + "loss": 0.9927, + "step": 2209 + }, + { + "epoch": 0.05, + "grad_norm": 1.974387293763489, + "learning_rate": 1.997450629810837e-05, + "loss": 1.0805, + "step": 2210 + }, + { + "epoch": 0.05, + "grad_norm": 2.124652397821647, + "learning_rate": 1.997445181850649e-05, + "loss": 1.0789, + "step": 2211 + }, + { + "epoch": 0.05, + "grad_norm": 2.349058186480292, + "learning_rate": 1.997439728083019e-05, + "loss": 1.045, + "step": 2212 + }, + { + "epoch": 0.05, + "grad_norm": 2.174417999424716, + "learning_rate": 1.9974342685079795e-05, + "loss": 1.0294, + "step": 2213 + }, + { + "epoch": 0.05, + "grad_norm": 2.37390533331593, + "learning_rate": 1.997428803125562e-05, + "loss": 1.0097, + "step": 2214 + }, + { + "epoch": 0.05, + "grad_norm": 1.9603446057247416, + "learning_rate": 1.9974233319357977e-05, + "loss": 0.9924, + "step": 2215 + }, + { + "epoch": 0.05, + "grad_norm": 2.038465865105205, + "learning_rate": 1.9974178549387195e-05, + "loss": 1.2019, + "step": 2216 + }, + { + "epoch": 0.05, + "grad_norm": 2.1494843288636862, + "learning_rate": 1.9974123721343587e-05, + "loss": 1.1972, + "step": 2217 + }, + { + "epoch": 0.05, + "grad_norm": 3.4947171075855903, + "learning_rate": 1.9974068835227468e-05, + "loss": 1.0029, + "step": 2218 + }, + { + "epoch": 0.05, + "grad_norm": 2.4957052969180498, + "learning_rate": 1.997401389103917e-05, + "loss": 1.1615, + "step": 2219 + }, + { + "epoch": 0.05, + "grad_norm": 2.2822365688150317, + "learning_rate": 1.9973958888779e-05, + "loss": 1.2123, + "step": 2220 + }, + { + "epoch": 0.05, + "grad_norm": 2.811519128074669, + "learning_rate": 1.9973903828447286e-05, + "loss": 1.1779, + "step": 2221 + }, + { + "epoch": 0.05, + "grad_norm": 2.2559933268629093, + "learning_rate": 1.9973848710044345e-05, + "loss": 1.0457, + "step": 2222 + }, + { + "epoch": 0.05, + "grad_norm": 2.669018714752434, + "learning_rate": 1.99737935335705e-05, + "loss": 0.9859, + "step": 2223 + }, + { + "epoch": 0.05, + "grad_norm": 2.0279839188063336, + "learning_rate": 1.997373829902607e-05, + "loss": 1.1512, + "step": 2224 + }, + { + "epoch": 0.05, + "grad_norm": 1.963242620191225, + "learning_rate": 1.997368300641138e-05, + "loss": 1.0602, + "step": 2225 + }, + { + "epoch": 0.05, + "grad_norm": 2.255851930651416, + "learning_rate": 1.997362765572675e-05, + "loss": 1.1758, + "step": 2226 + }, + { + "epoch": 0.05, + "grad_norm": 2.509655251005601, + "learning_rate": 1.99735722469725e-05, + "loss": 0.9081, + "step": 2227 + }, + { + "epoch": 0.05, + "grad_norm": 2.6674566455519515, + "learning_rate": 1.9973516780148957e-05, + "loss": 1.1357, + "step": 2228 + }, + { + "epoch": 0.05, + "grad_norm": 2.2773356524441284, + "learning_rate": 1.9973461255256443e-05, + "loss": 1.1768, + "step": 2229 + }, + { + "epoch": 0.05, + "grad_norm": 1.960856997169203, + "learning_rate": 1.9973405672295276e-05, + "loss": 0.9854, + "step": 2230 + }, + { + "epoch": 0.05, + "grad_norm": 2.631649590018756, + "learning_rate": 1.9973350031265786e-05, + "loss": 1.1564, + "step": 2231 + }, + { + "epoch": 0.05, + "grad_norm": 2.076374133482105, + "learning_rate": 1.9973294332168294e-05, + "loss": 1.1756, + "step": 2232 + }, + { + "epoch": 0.05, + "grad_norm": 2.2156568530921876, + "learning_rate": 1.9973238575003123e-05, + "loss": 1.1879, + "step": 2233 + }, + { + "epoch": 0.05, + "grad_norm": 2.141652428983566, + "learning_rate": 1.9973182759770605e-05, + "loss": 1.1985, + "step": 2234 + }, + { + "epoch": 0.05, + "grad_norm": 1.977500463038365, + "learning_rate": 1.9973126886471053e-05, + "loss": 1.056, + "step": 2235 + }, + { + "epoch": 0.05, + "grad_norm": 2.4530416791692846, + "learning_rate": 1.9973070955104802e-05, + "loss": 1.1894, + "step": 2236 + }, + { + "epoch": 0.05, + "grad_norm": 3.0760843376541356, + "learning_rate": 1.9973014965672175e-05, + "loss": 0.9034, + "step": 2237 + }, + { + "epoch": 0.05, + "grad_norm": 1.2357284091731802, + "learning_rate": 1.9972958918173498e-05, + "loss": 0.9768, + "step": 2238 + }, + { + "epoch": 0.05, + "grad_norm": 2.321359854285932, + "learning_rate": 1.9972902812609093e-05, + "loss": 1.0056, + "step": 2239 + }, + { + "epoch": 0.05, + "grad_norm": 1.1824663412137557, + "learning_rate": 1.9972846648979292e-05, + "loss": 1.0129, + "step": 2240 + }, + { + "epoch": 0.05, + "grad_norm": 1.9759827244594563, + "learning_rate": 1.9972790427284424e-05, + "loss": 0.9957, + "step": 2241 + }, + { + "epoch": 0.05, + "grad_norm": 2.3768281507569835, + "learning_rate": 1.997273414752481e-05, + "loss": 0.9014, + "step": 2242 + }, + { + "epoch": 0.05, + "grad_norm": 2.0589059999161385, + "learning_rate": 1.997267780970078e-05, + "loss": 1.1017, + "step": 2243 + }, + { + "epoch": 0.05, + "grad_norm": 1.1931101430273425, + "learning_rate": 1.9972621413812664e-05, + "loss": 0.9967, + "step": 2244 + }, + { + "epoch": 0.05, + "grad_norm": 2.061544467808818, + "learning_rate": 1.997256495986079e-05, + "loss": 0.9491, + "step": 2245 + }, + { + "epoch": 0.05, + "grad_norm": 2.109524940530451, + "learning_rate": 1.9972508447845483e-05, + "loss": 1.1183, + "step": 2246 + }, + { + "epoch": 0.05, + "grad_norm": 2.0471112070856834, + "learning_rate": 1.9972451877767075e-05, + "loss": 1.0918, + "step": 2247 + }, + { + "epoch": 0.05, + "grad_norm": 2.7809247842389855, + "learning_rate": 1.9972395249625897e-05, + "loss": 1.0297, + "step": 2248 + }, + { + "epoch": 0.05, + "grad_norm": 2.448107737793077, + "learning_rate": 1.9972338563422273e-05, + "loss": 1.1386, + "step": 2249 + }, + { + "epoch": 0.05, + "grad_norm": 1.945929929415176, + "learning_rate": 1.997228181915654e-05, + "loss": 1.1699, + "step": 2250 + }, + { + "epoch": 0.05, + "grad_norm": 2.4300304334980742, + "learning_rate": 1.9972225016829023e-05, + "loss": 1.0446, + "step": 2251 + }, + { + "epoch": 0.05, + "grad_norm": 1.990506715316271, + "learning_rate": 1.9972168156440054e-05, + "loss": 1.1174, + "step": 2252 + }, + { + "epoch": 0.05, + "grad_norm": 3.8420555731749744, + "learning_rate": 1.9972111237989968e-05, + "loss": 1.1085, + "step": 2253 + }, + { + "epoch": 0.05, + "grad_norm": 3.362733283468021, + "learning_rate": 1.997205426147909e-05, + "loss": 1.051, + "step": 2254 + }, + { + "epoch": 0.05, + "grad_norm": 2.2705240773059367, + "learning_rate": 1.9971997226907756e-05, + "loss": 1.0029, + "step": 2255 + }, + { + "epoch": 0.05, + "grad_norm": 2.370623822735373, + "learning_rate": 1.9971940134276298e-05, + "loss": 1.0595, + "step": 2256 + }, + { + "epoch": 0.05, + "grad_norm": 2.20211032149322, + "learning_rate": 1.9971882983585048e-05, + "loss": 1.0879, + "step": 2257 + }, + { + "epoch": 0.05, + "grad_norm": 2.3253468572649107, + "learning_rate": 1.9971825774834338e-05, + "loss": 1.034, + "step": 2258 + }, + { + "epoch": 0.05, + "grad_norm": 2.1416621425756137, + "learning_rate": 1.99717685080245e-05, + "loss": 1.2445, + "step": 2259 + }, + { + "epoch": 0.05, + "grad_norm": 2.095424199918386, + "learning_rate": 1.997171118315587e-05, + "loss": 1.0927, + "step": 2260 + }, + { + "epoch": 0.05, + "grad_norm": 1.1783044241677052, + "learning_rate": 1.9971653800228782e-05, + "loss": 0.9958, + "step": 2261 + }, + { + "epoch": 0.05, + "grad_norm": 2.14672312556336, + "learning_rate": 1.9971596359243562e-05, + "loss": 1.018, + "step": 2262 + }, + { + "epoch": 0.05, + "grad_norm": 2.167083177250717, + "learning_rate": 1.9971538860200556e-05, + "loss": 0.9955, + "step": 2263 + }, + { + "epoch": 0.05, + "grad_norm": 2.2357630525399177, + "learning_rate": 1.9971481303100093e-05, + "loss": 1.0941, + "step": 2264 + }, + { + "epoch": 0.05, + "grad_norm": 2.535547838712617, + "learning_rate": 1.997142368794251e-05, + "loss": 1.0968, + "step": 2265 + }, + { + "epoch": 0.05, + "grad_norm": 1.9710153755536128, + "learning_rate": 1.997136601472814e-05, + "loss": 1.1086, + "step": 2266 + }, + { + "epoch": 0.05, + "grad_norm": 2.187610674084936, + "learning_rate": 1.997130828345732e-05, + "loss": 1.0752, + "step": 2267 + }, + { + "epoch": 0.05, + "grad_norm": 2.3249316756979166, + "learning_rate": 1.997125049413039e-05, + "loss": 1.0868, + "step": 2268 + }, + { + "epoch": 0.05, + "grad_norm": 2.407684207033599, + "learning_rate": 1.9971192646747678e-05, + "loss": 1.0151, + "step": 2269 + }, + { + "epoch": 0.05, + "grad_norm": 2.7589022456219627, + "learning_rate": 1.997113474130953e-05, + "loss": 1.0215, + "step": 2270 + }, + { + "epoch": 0.05, + "grad_norm": 2.0658423978070046, + "learning_rate": 1.9971076777816275e-05, + "loss": 0.9958, + "step": 2271 + }, + { + "epoch": 0.05, + "grad_norm": 2.600389833107808, + "learning_rate": 1.9971018756268256e-05, + "loss": 1.0096, + "step": 2272 + }, + { + "epoch": 0.05, + "grad_norm": 2.0517933237378934, + "learning_rate": 1.997096067666581e-05, + "loss": 1.1383, + "step": 2273 + }, + { + "epoch": 0.05, + "grad_norm": 3.0106639821743473, + "learning_rate": 1.997090253900927e-05, + "loss": 0.935, + "step": 2274 + }, + { + "epoch": 0.05, + "grad_norm": 1.209849709158252, + "learning_rate": 1.997084434329898e-05, + "loss": 0.9879, + "step": 2275 + }, + { + "epoch": 0.05, + "grad_norm": 2.357672272024379, + "learning_rate": 1.9970786089535283e-05, + "loss": 1.0181, + "step": 2276 + }, + { + "epoch": 0.05, + "grad_norm": 1.1414478400823727, + "learning_rate": 1.9970727777718505e-05, + "loss": 1.0252, + "step": 2277 + }, + { + "epoch": 0.05, + "grad_norm": 2.2321198783568774, + "learning_rate": 1.9970669407848997e-05, + "loss": 1.1609, + "step": 2278 + }, + { + "epoch": 0.05, + "grad_norm": 2.355976114541676, + "learning_rate": 1.9970610979927096e-05, + "loss": 1.1422, + "step": 2279 + }, + { + "epoch": 0.05, + "grad_norm": 2.334914923058556, + "learning_rate": 1.997055249395314e-05, + "loss": 0.9999, + "step": 2280 + }, + { + "epoch": 0.05, + "grad_norm": 2.26836472583003, + "learning_rate": 1.9970493949927473e-05, + "loss": 1.0811, + "step": 2281 + }, + { + "epoch": 0.05, + "grad_norm": 2.160470962665018, + "learning_rate": 1.9970435347850428e-05, + "loss": 0.8911, + "step": 2282 + }, + { + "epoch": 0.05, + "grad_norm": 2.2131333966048437, + "learning_rate": 1.9970376687722356e-05, + "loss": 0.9807, + "step": 2283 + }, + { + "epoch": 0.05, + "grad_norm": 2.2447868512123916, + "learning_rate": 1.9970317969543592e-05, + "loss": 1.1402, + "step": 2284 + }, + { + "epoch": 0.05, + "grad_norm": 2.0719453024093535, + "learning_rate": 1.9970259193314483e-05, + "loss": 1.2383, + "step": 2285 + }, + { + "epoch": 0.05, + "grad_norm": 2.021576048665855, + "learning_rate": 1.9970200359035366e-05, + "loss": 0.9832, + "step": 2286 + }, + { + "epoch": 0.05, + "grad_norm": 2.33296925912876, + "learning_rate": 1.9970141466706588e-05, + "loss": 1.1112, + "step": 2287 + }, + { + "epoch": 0.05, + "grad_norm": 2.0844985701611574, + "learning_rate": 1.997008251632849e-05, + "loss": 1.0543, + "step": 2288 + }, + { + "epoch": 0.05, + "grad_norm": 1.0362556073081435, + "learning_rate": 1.9970023507901416e-05, + "loss": 0.9836, + "step": 2289 + }, + { + "epoch": 0.05, + "grad_norm": 2.522035919145633, + "learning_rate": 1.9969964441425706e-05, + "loss": 1.1395, + "step": 2290 + }, + { + "epoch": 0.05, + "grad_norm": 1.1615179722854458, + "learning_rate": 1.9969905316901704e-05, + "loss": 0.9416, + "step": 2291 + }, + { + "epoch": 0.05, + "grad_norm": 2.0917059678716337, + "learning_rate": 1.9969846134329765e-05, + "loss": 1.1042, + "step": 2292 + }, + { + "epoch": 0.05, + "grad_norm": 2.0773884236021525, + "learning_rate": 1.996978689371022e-05, + "loss": 1.0725, + "step": 2293 + }, + { + "epoch": 0.05, + "grad_norm": 2.1676538579165303, + "learning_rate": 1.9969727595043417e-05, + "loss": 0.9991, + "step": 2294 + }, + { + "epoch": 0.05, + "grad_norm": 3.0652521500668204, + "learning_rate": 1.996966823832971e-05, + "loss": 1.116, + "step": 2295 + }, + { + "epoch": 0.05, + "grad_norm": 2.0850194726271645, + "learning_rate": 1.9969608823569435e-05, + "loss": 0.971, + "step": 2296 + }, + { + "epoch": 0.05, + "grad_norm": 2.033656918667178, + "learning_rate": 1.996954935076294e-05, + "loss": 1.031, + "step": 2297 + }, + { + "epoch": 0.05, + "grad_norm": 1.9847412772178346, + "learning_rate": 1.9969489819910572e-05, + "loss": 1.1057, + "step": 2298 + }, + { + "epoch": 0.05, + "grad_norm": 2.0980710496567965, + "learning_rate": 1.9969430231012682e-05, + "loss": 1.0299, + "step": 2299 + }, + { + "epoch": 0.05, + "grad_norm": 2.628852110889092, + "learning_rate": 1.9969370584069607e-05, + "loss": 0.9685, + "step": 2300 + }, + { + "epoch": 0.05, + "grad_norm": 2.136702050827228, + "learning_rate": 1.9969310879081705e-05, + "loss": 1.0608, + "step": 2301 + }, + { + "epoch": 0.05, + "grad_norm": 2.3283964065310463, + "learning_rate": 1.996925111604932e-05, + "loss": 1.0619, + "step": 2302 + }, + { + "epoch": 0.05, + "grad_norm": 2.244465747846587, + "learning_rate": 1.9969191294972794e-05, + "loss": 0.946, + "step": 2303 + }, + { + "epoch": 0.05, + "grad_norm": 2.317070804307833, + "learning_rate": 1.9969131415852482e-05, + "loss": 1.0828, + "step": 2304 + }, + { + "epoch": 0.05, + "grad_norm": 2.379505444040971, + "learning_rate": 1.996907147868873e-05, + "loss": 0.9816, + "step": 2305 + }, + { + "epoch": 0.05, + "grad_norm": 2.062700042339017, + "learning_rate": 1.9969011483481888e-05, + "loss": 1.1939, + "step": 2306 + }, + { + "epoch": 0.05, + "grad_norm": 2.3500403286876637, + "learning_rate": 1.9968951430232307e-05, + "loss": 1.0898, + "step": 2307 + }, + { + "epoch": 0.05, + "grad_norm": 1.9876806304249208, + "learning_rate": 1.9968891318940332e-05, + "loss": 0.9865, + "step": 2308 + }, + { + "epoch": 0.05, + "grad_norm": 2.5931473475403313, + "learning_rate": 1.996883114960632e-05, + "loss": 0.9217, + "step": 2309 + }, + { + "epoch": 0.05, + "grad_norm": 1.961439729251132, + "learning_rate": 1.996877092223061e-05, + "loss": 1.0984, + "step": 2310 + }, + { + "epoch": 0.05, + "grad_norm": 3.9749491794710567, + "learning_rate": 1.9968710636813564e-05, + "loss": 1.0087, + "step": 2311 + }, + { + "epoch": 0.05, + "grad_norm": 2.12887047838203, + "learning_rate": 1.996865029335553e-05, + "loss": 1.1325, + "step": 2312 + }, + { + "epoch": 0.05, + "grad_norm": 2.3152370371730027, + "learning_rate": 1.9968589891856854e-05, + "loss": 1.1069, + "step": 2313 + }, + { + "epoch": 0.05, + "grad_norm": 2.1430252818283537, + "learning_rate": 1.9968529432317894e-05, + "loss": 1.0347, + "step": 2314 + }, + { + "epoch": 0.05, + "grad_norm": 2.1318845165613642, + "learning_rate": 1.9968468914738997e-05, + "loss": 1.1215, + "step": 2315 + }, + { + "epoch": 0.05, + "grad_norm": 1.8852675577483053, + "learning_rate": 1.9968408339120522e-05, + "loss": 1.0694, + "step": 2316 + }, + { + "epoch": 0.05, + "grad_norm": 2.1656283343225464, + "learning_rate": 1.9968347705462817e-05, + "loss": 0.9788, + "step": 2317 + }, + { + "epoch": 0.05, + "grad_norm": 2.044190570870174, + "learning_rate": 1.9968287013766235e-05, + "loss": 1.2013, + "step": 2318 + }, + { + "epoch": 0.05, + "grad_norm": 1.9442526123129542, + "learning_rate": 1.996822626403113e-05, + "loss": 1.1448, + "step": 2319 + }, + { + "epoch": 0.05, + "grad_norm": 2.1213215814034414, + "learning_rate": 1.9968165456257855e-05, + "loss": 1.1245, + "step": 2320 + }, + { + "epoch": 0.05, + "grad_norm": 2.3687974132039566, + "learning_rate": 1.9968104590446766e-05, + "loss": 0.9927, + "step": 2321 + }, + { + "epoch": 0.05, + "grad_norm": 2.3372140324036295, + "learning_rate": 1.9968043666598216e-05, + "loss": 0.9837, + "step": 2322 + }, + { + "epoch": 0.05, + "grad_norm": 2.1593805771329904, + "learning_rate": 1.996798268471256e-05, + "loss": 0.9676, + "step": 2323 + }, + { + "epoch": 0.05, + "grad_norm": 2.5549319673256408, + "learning_rate": 1.9967921644790153e-05, + "loss": 1.1664, + "step": 2324 + }, + { + "epoch": 0.05, + "grad_norm": 2.6436333521302093, + "learning_rate": 1.9967860546831348e-05, + "loss": 0.9235, + "step": 2325 + }, + { + "epoch": 0.05, + "grad_norm": 2.676216527591101, + "learning_rate": 1.9967799390836505e-05, + "loss": 0.9718, + "step": 2326 + }, + { + "epoch": 0.05, + "grad_norm": 2.511362953852357, + "learning_rate": 1.996773817680598e-05, + "loss": 1.0235, + "step": 2327 + }, + { + "epoch": 0.05, + "grad_norm": 2.219347646324501, + "learning_rate": 1.9967676904740126e-05, + "loss": 0.9887, + "step": 2328 + }, + { + "epoch": 0.05, + "grad_norm": 3.9287502945624264, + "learning_rate": 1.99676155746393e-05, + "loss": 1.067, + "step": 2329 + }, + { + "epoch": 0.05, + "grad_norm": 1.1449384353566483, + "learning_rate": 1.9967554186503864e-05, + "loss": 0.9244, + "step": 2330 + }, + { + "epoch": 0.05, + "grad_norm": 1.9402285827062957, + "learning_rate": 1.9967492740334168e-05, + "loss": 1.0438, + "step": 2331 + }, + { + "epoch": 0.05, + "grad_norm": 3.071025450829972, + "learning_rate": 1.9967431236130576e-05, + "loss": 1.0819, + "step": 2332 + }, + { + "epoch": 0.05, + "grad_norm": 2.1351916550312047, + "learning_rate": 1.9967369673893443e-05, + "loss": 1.2085, + "step": 2333 + }, + { + "epoch": 0.05, + "grad_norm": 2.163067688768104, + "learning_rate": 1.9967308053623128e-05, + "loss": 0.9937, + "step": 2334 + }, + { + "epoch": 0.06, + "grad_norm": 1.9816265885769417, + "learning_rate": 1.996724637531999e-05, + "loss": 1.1775, + "step": 2335 + }, + { + "epoch": 0.06, + "grad_norm": 1.9795651625885737, + "learning_rate": 1.996718463898439e-05, + "loss": 1.0363, + "step": 2336 + }, + { + "epoch": 0.06, + "grad_norm": 2.1933426379731196, + "learning_rate": 1.996712284461668e-05, + "loss": 1.0926, + "step": 2337 + }, + { + "epoch": 0.06, + "grad_norm": 2.1889018119477632, + "learning_rate": 1.996706099221723e-05, + "loss": 1.0741, + "step": 2338 + }, + { + "epoch": 0.06, + "grad_norm": 2.358464132207462, + "learning_rate": 1.9966999081786396e-05, + "loss": 0.9571, + "step": 2339 + }, + { + "epoch": 0.06, + "grad_norm": 1.8274150304236811, + "learning_rate": 1.9966937113324536e-05, + "loss": 1.1131, + "step": 2340 + }, + { + "epoch": 0.06, + "grad_norm": 2.5219599369906347, + "learning_rate": 1.9966875086832012e-05, + "loss": 1.1093, + "step": 2341 + }, + { + "epoch": 0.06, + "grad_norm": 2.3544621662463183, + "learning_rate": 1.9966813002309187e-05, + "loss": 1.0894, + "step": 2342 + }, + { + "epoch": 0.06, + "grad_norm": 2.2402680028879187, + "learning_rate": 1.9966750859756422e-05, + "loss": 1.0719, + "step": 2343 + }, + { + "epoch": 0.06, + "grad_norm": 2.6392342719364414, + "learning_rate": 1.9966688659174074e-05, + "loss": 1.1143, + "step": 2344 + }, + { + "epoch": 0.06, + "grad_norm": 2.1447004524431037, + "learning_rate": 1.9966626400562515e-05, + "loss": 0.8848, + "step": 2345 + }, + { + "epoch": 0.06, + "grad_norm": 2.2719603382820326, + "learning_rate": 1.9966564083922098e-05, + "loss": 1.2028, + "step": 2346 + }, + { + "epoch": 0.06, + "grad_norm": 2.2283303717059253, + "learning_rate": 1.996650170925319e-05, + "loss": 1.08, + "step": 2347 + }, + { + "epoch": 0.06, + "grad_norm": 2.769937156132906, + "learning_rate": 1.9966439276556155e-05, + "loss": 1.0415, + "step": 2348 + }, + { + "epoch": 0.06, + "grad_norm": 2.12050347716433, + "learning_rate": 1.9966376785831354e-05, + "loss": 1.0759, + "step": 2349 + }, + { + "epoch": 0.06, + "grad_norm": 2.0992200024646848, + "learning_rate": 1.996631423707915e-05, + "loss": 1.003, + "step": 2350 + }, + { + "epoch": 0.06, + "grad_norm": 2.1075756601995903, + "learning_rate": 1.9966251630299915e-05, + "loss": 1.0716, + "step": 2351 + }, + { + "epoch": 0.06, + "grad_norm": 1.916009521924272, + "learning_rate": 1.9966188965494e-05, + "loss": 0.962, + "step": 2352 + }, + { + "epoch": 0.06, + "grad_norm": 1.8766040527642027, + "learning_rate": 1.9966126242661783e-05, + "loss": 1.0607, + "step": 2353 + }, + { + "epoch": 0.06, + "grad_norm": 2.476143744317625, + "learning_rate": 1.996606346180362e-05, + "loss": 0.8461, + "step": 2354 + }, + { + "epoch": 0.06, + "grad_norm": 1.200188956852994, + "learning_rate": 1.9966000622919886e-05, + "loss": 1.038, + "step": 2355 + }, + { + "epoch": 0.06, + "grad_norm": 1.9533161950704507, + "learning_rate": 1.996593772601094e-05, + "loss": 0.968, + "step": 2356 + }, + { + "epoch": 0.06, + "grad_norm": 2.2341664206911864, + "learning_rate": 1.996587477107715e-05, + "loss": 1.1104, + "step": 2357 + }, + { + "epoch": 0.06, + "grad_norm": 3.278034640221199, + "learning_rate": 1.9965811758118873e-05, + "loss": 0.8383, + "step": 2358 + }, + { + "epoch": 0.06, + "grad_norm": 2.295471007411432, + "learning_rate": 1.9965748687136494e-05, + "loss": 0.9553, + "step": 2359 + }, + { + "epoch": 0.06, + "grad_norm": 2.354370912979062, + "learning_rate": 1.9965685558130368e-05, + "loss": 1.0949, + "step": 2360 + }, + { + "epoch": 0.06, + "grad_norm": 1.222511650673943, + "learning_rate": 1.9965622371100866e-05, + "loss": 1.0381, + "step": 2361 + }, + { + "epoch": 0.06, + "grad_norm": 2.365703030456611, + "learning_rate": 1.9965559126048354e-05, + "loss": 1.0364, + "step": 2362 + }, + { + "epoch": 0.06, + "grad_norm": 2.2218980846374436, + "learning_rate": 1.9965495822973204e-05, + "loss": 1.0782, + "step": 2363 + }, + { + "epoch": 0.06, + "grad_norm": 2.188980397754815, + "learning_rate": 1.996543246187578e-05, + "loss": 1.0019, + "step": 2364 + }, + { + "epoch": 0.06, + "grad_norm": 2.06760073992132, + "learning_rate": 1.9965369042756453e-05, + "loss": 1.1151, + "step": 2365 + }, + { + "epoch": 0.06, + "grad_norm": 1.2312338467478718, + "learning_rate": 1.99653055656156e-05, + "loss": 1.0567, + "step": 2366 + }, + { + "epoch": 0.06, + "grad_norm": 2.1319630427045806, + "learning_rate": 1.9965242030453577e-05, + "loss": 1.0439, + "step": 2367 + }, + { + "epoch": 0.06, + "grad_norm": 2.0335417636344224, + "learning_rate": 1.9965178437270757e-05, + "loss": 0.9033, + "step": 2368 + }, + { + "epoch": 0.06, + "grad_norm": 2.1741934624185557, + "learning_rate": 1.9965114786067515e-05, + "loss": 1.0718, + "step": 2369 + }, + { + "epoch": 0.06, + "grad_norm": 2.433353208112653, + "learning_rate": 1.9965051076844224e-05, + "loss": 1.033, + "step": 2370 + }, + { + "epoch": 0.06, + "grad_norm": 2.5360803969603234, + "learning_rate": 1.9964987309601246e-05, + "loss": 1.099, + "step": 2371 + }, + { + "epoch": 0.06, + "grad_norm": 2.1428003597976106, + "learning_rate": 1.996492348433896e-05, + "loss": 0.9419, + "step": 2372 + }, + { + "epoch": 0.06, + "grad_norm": 1.1361113414314363, + "learning_rate": 1.9964859601057734e-05, + "loss": 0.9652, + "step": 2373 + }, + { + "epoch": 0.06, + "grad_norm": 1.2023208719743697, + "learning_rate": 1.996479565975794e-05, + "loss": 0.9933, + "step": 2374 + }, + { + "epoch": 0.06, + "grad_norm": 2.1243151286316038, + "learning_rate": 1.996473166043995e-05, + "loss": 1.0286, + "step": 2375 + }, + { + "epoch": 0.06, + "grad_norm": 1.1609567670375338, + "learning_rate": 1.996466760310414e-05, + "loss": 1.0017, + "step": 2376 + }, + { + "epoch": 0.06, + "grad_norm": 2.3950123971527577, + "learning_rate": 1.996460348775088e-05, + "loss": 0.9683, + "step": 2377 + }, + { + "epoch": 0.06, + "grad_norm": 2.2334579950638984, + "learning_rate": 1.9964539314380542e-05, + "loss": 1.0354, + "step": 2378 + }, + { + "epoch": 0.06, + "grad_norm": 2.3962889421033755, + "learning_rate": 1.99644750829935e-05, + "loss": 1.0207, + "step": 2379 + }, + { + "epoch": 0.06, + "grad_norm": 2.2193578178014044, + "learning_rate": 1.996441079359013e-05, + "loss": 0.9621, + "step": 2380 + }, + { + "epoch": 0.06, + "grad_norm": 2.2055616432570906, + "learning_rate": 1.996434644617081e-05, + "loss": 1.0684, + "step": 2381 + }, + { + "epoch": 0.06, + "grad_norm": 2.3043641885986745, + "learning_rate": 1.9964282040735907e-05, + "loss": 1.0632, + "step": 2382 + }, + { + "epoch": 0.06, + "grad_norm": 2.1273219416168923, + "learning_rate": 1.99642175772858e-05, + "loss": 1.1347, + "step": 2383 + }, + { + "epoch": 0.06, + "grad_norm": 2.111971990554125, + "learning_rate": 1.9964153055820863e-05, + "loss": 1.0697, + "step": 2384 + }, + { + "epoch": 0.06, + "grad_norm": 2.1157885408476593, + "learning_rate": 1.996408847634147e-05, + "loss": 1.0527, + "step": 2385 + }, + { + "epoch": 0.06, + "grad_norm": 2.1659247984086853, + "learning_rate": 1.9964023838848004e-05, + "loss": 1.1015, + "step": 2386 + }, + { + "epoch": 0.06, + "grad_norm": 2.2587172581968997, + "learning_rate": 1.9963959143340833e-05, + "loss": 1.1686, + "step": 2387 + }, + { + "epoch": 0.06, + "grad_norm": 2.1137265611652993, + "learning_rate": 1.996389438982034e-05, + "loss": 1.1036, + "step": 2388 + }, + { + "epoch": 0.06, + "grad_norm": 2.2466935179030254, + "learning_rate": 1.99638295782869e-05, + "loss": 1.0687, + "step": 2389 + }, + { + "epoch": 0.06, + "grad_norm": 2.0771583080841864, + "learning_rate": 1.996376470874089e-05, + "loss": 0.992, + "step": 2390 + }, + { + "epoch": 0.06, + "grad_norm": 2.206935093481372, + "learning_rate": 1.9963699781182685e-05, + "loss": 1.046, + "step": 2391 + }, + { + "epoch": 0.06, + "grad_norm": 1.902952250739689, + "learning_rate": 1.9963634795612666e-05, + "loss": 1.0967, + "step": 2392 + }, + { + "epoch": 0.06, + "grad_norm": 2.4092971189799917, + "learning_rate": 1.9963569752031207e-05, + "loss": 1.1171, + "step": 2393 + }, + { + "epoch": 0.06, + "grad_norm": 2.4501026405150887, + "learning_rate": 1.9963504650438698e-05, + "loss": 1.0649, + "step": 2394 + }, + { + "epoch": 0.06, + "grad_norm": 1.974042113642194, + "learning_rate": 1.9963439490835504e-05, + "loss": 0.9627, + "step": 2395 + }, + { + "epoch": 0.06, + "grad_norm": 2.1642704879443606, + "learning_rate": 1.9963374273222015e-05, + "loss": 1.072, + "step": 2396 + }, + { + "epoch": 0.06, + "grad_norm": 1.2732368860457686, + "learning_rate": 1.9963308997598604e-05, + "loss": 0.9774, + "step": 2397 + }, + { + "epoch": 0.06, + "grad_norm": 2.6125540600050448, + "learning_rate": 1.9963243663965658e-05, + "loss": 0.9132, + "step": 2398 + }, + { + "epoch": 0.06, + "grad_norm": 2.035251791714855, + "learning_rate": 1.9963178272323548e-05, + "loss": 1.0918, + "step": 2399 + }, + { + "epoch": 0.06, + "grad_norm": 1.9063683319445968, + "learning_rate": 1.996311282267266e-05, + "loss": 1.1777, + "step": 2400 + }, + { + "epoch": 0.06, + "grad_norm": 2.0975043131459086, + "learning_rate": 1.9963047315013376e-05, + "loss": 1.1547, + "step": 2401 + }, + { + "epoch": 0.06, + "grad_norm": 2.0166937006898484, + "learning_rate": 1.996298174934608e-05, + "loss": 0.9385, + "step": 2402 + }, + { + "epoch": 0.06, + "grad_norm": 2.5283398073107133, + "learning_rate": 1.9962916125671146e-05, + "loss": 1.0787, + "step": 2403 + }, + { + "epoch": 0.06, + "grad_norm": 2.1328367226005907, + "learning_rate": 1.996285044398896e-05, + "loss": 1.097, + "step": 2404 + }, + { + "epoch": 0.06, + "grad_norm": 2.2032441737276263, + "learning_rate": 1.9962784704299907e-05, + "loss": 1.0542, + "step": 2405 + }, + { + "epoch": 0.06, + "grad_norm": 1.1083100336169118, + "learning_rate": 1.9962718906604365e-05, + "loss": 1.0244, + "step": 2406 + }, + { + "epoch": 0.06, + "grad_norm": 2.6259220303114748, + "learning_rate": 1.9962653050902722e-05, + "loss": 1.0324, + "step": 2407 + }, + { + "epoch": 0.06, + "grad_norm": 2.091160232902421, + "learning_rate": 1.996258713719536e-05, + "loss": 1.0568, + "step": 2408 + }, + { + "epoch": 0.06, + "grad_norm": 2.231349721296818, + "learning_rate": 1.9962521165482657e-05, + "loss": 1.1325, + "step": 2409 + }, + { + "epoch": 0.06, + "grad_norm": 2.46076810108062, + "learning_rate": 1.9962455135765003e-05, + "loss": 1.008, + "step": 2410 + }, + { + "epoch": 0.06, + "grad_norm": 2.1754748904132923, + "learning_rate": 1.996238904804278e-05, + "loss": 0.9096, + "step": 2411 + }, + { + "epoch": 0.06, + "grad_norm": 2.1551419206324183, + "learning_rate": 1.9962322902316376e-05, + "loss": 0.9731, + "step": 2412 + }, + { + "epoch": 0.06, + "grad_norm": 1.1591329251814144, + "learning_rate": 1.9962256698586174e-05, + "loss": 0.9985, + "step": 2413 + }, + { + "epoch": 0.06, + "grad_norm": 2.871852675921005, + "learning_rate": 1.996219043685256e-05, + "loss": 1.169, + "step": 2414 + }, + { + "epoch": 0.06, + "grad_norm": 2.1971557722181694, + "learning_rate": 1.9962124117115914e-05, + "loss": 1.1518, + "step": 2415 + }, + { + "epoch": 0.06, + "grad_norm": 2.3616960289589675, + "learning_rate": 1.9962057739376634e-05, + "loss": 1.0595, + "step": 2416 + }, + { + "epoch": 0.06, + "grad_norm": 2.022664101944738, + "learning_rate": 1.9961991303635096e-05, + "loss": 1.0132, + "step": 2417 + }, + { + "epoch": 0.06, + "grad_norm": 2.150783417408257, + "learning_rate": 1.9961924809891693e-05, + "loss": 1.0955, + "step": 2418 + }, + { + "epoch": 0.06, + "grad_norm": 2.2646187802038784, + "learning_rate": 1.996185825814681e-05, + "loss": 0.9096, + "step": 2419 + }, + { + "epoch": 0.06, + "grad_norm": 2.2435378684684273, + "learning_rate": 1.9961791648400836e-05, + "loss": 1.058, + "step": 2420 + }, + { + "epoch": 0.06, + "grad_norm": 1.1174929041338513, + "learning_rate": 1.9961724980654154e-05, + "loss": 0.9214, + "step": 2421 + }, + { + "epoch": 0.06, + "grad_norm": 2.0401476228095827, + "learning_rate": 1.9961658254907153e-05, + "loss": 1.0114, + "step": 2422 + }, + { + "epoch": 0.06, + "grad_norm": 2.5388491192434546, + "learning_rate": 1.9961591471160228e-05, + "loss": 1.2018, + "step": 2423 + }, + { + "epoch": 0.06, + "grad_norm": 1.1249430015110298, + "learning_rate": 1.9961524629413765e-05, + "loss": 1.0209, + "step": 2424 + }, + { + "epoch": 0.06, + "grad_norm": 2.280607421235675, + "learning_rate": 1.9961457729668148e-05, + "loss": 1.1202, + "step": 2425 + }, + { + "epoch": 0.06, + "grad_norm": 2.130652164975364, + "learning_rate": 1.9961390771923772e-05, + "loss": 1.138, + "step": 2426 + }, + { + "epoch": 0.06, + "grad_norm": 1.9511721650878493, + "learning_rate": 1.9961323756181024e-05, + "loss": 1.0036, + "step": 2427 + }, + { + "epoch": 0.06, + "grad_norm": 4.117363499423357, + "learning_rate": 1.9961256682440298e-05, + "loss": 1.0636, + "step": 2428 + }, + { + "epoch": 0.06, + "grad_norm": 2.010721462808995, + "learning_rate": 1.996118955070198e-05, + "loss": 1.1131, + "step": 2429 + }, + { + "epoch": 0.06, + "grad_norm": 2.172850936182339, + "learning_rate": 1.9961122360966463e-05, + "loss": 1.0872, + "step": 2430 + }, + { + "epoch": 0.06, + "grad_norm": 2.09542140862558, + "learning_rate": 1.9961055113234137e-05, + "loss": 1.1897, + "step": 2431 + }, + { + "epoch": 0.06, + "grad_norm": 2.2149782111587895, + "learning_rate": 1.9960987807505398e-05, + "loss": 1.124, + "step": 2432 + }, + { + "epoch": 0.06, + "grad_norm": 2.3158657372504416, + "learning_rate": 1.996092044378063e-05, + "loss": 0.9427, + "step": 2433 + }, + { + "epoch": 0.06, + "grad_norm": 2.145760141427153, + "learning_rate": 1.996085302206023e-05, + "loss": 1.2129, + "step": 2434 + }, + { + "epoch": 0.06, + "grad_norm": 2.2223594199992447, + "learning_rate": 1.9960785542344594e-05, + "loss": 0.9993, + "step": 2435 + }, + { + "epoch": 0.06, + "grad_norm": 2.077217025817082, + "learning_rate": 1.996071800463411e-05, + "loss": 1.0516, + "step": 2436 + }, + { + "epoch": 0.06, + "grad_norm": 2.492644274584414, + "learning_rate": 1.996065040892917e-05, + "loss": 0.9486, + "step": 2437 + }, + { + "epoch": 0.06, + "grad_norm": 2.2969273067376745, + "learning_rate": 1.9960582755230172e-05, + "loss": 1.013, + "step": 2438 + }, + { + "epoch": 0.06, + "grad_norm": 2.0639834957033574, + "learning_rate": 1.9960515043537504e-05, + "loss": 1.1025, + "step": 2439 + }, + { + "epoch": 0.06, + "grad_norm": 2.187938021014696, + "learning_rate": 1.9960447273851567e-05, + "loss": 1.0346, + "step": 2440 + }, + { + "epoch": 0.06, + "grad_norm": 2.0408692552703536, + "learning_rate": 1.9960379446172754e-05, + "loss": 1.0811, + "step": 2441 + }, + { + "epoch": 0.06, + "grad_norm": 2.2057800603299826, + "learning_rate": 1.9960311560501457e-05, + "loss": 1.1166, + "step": 2442 + }, + { + "epoch": 0.06, + "grad_norm": 3.9597492662095, + "learning_rate": 1.996024361683807e-05, + "loss": 1.0397, + "step": 2443 + }, + { + "epoch": 0.06, + "grad_norm": 2.0866692068701953, + "learning_rate": 1.996017561518299e-05, + "loss": 1.0384, + "step": 2444 + }, + { + "epoch": 0.06, + "grad_norm": 2.329991333225188, + "learning_rate": 1.996010755553662e-05, + "loss": 1.1695, + "step": 2445 + }, + { + "epoch": 0.06, + "grad_norm": 1.908008153431948, + "learning_rate": 1.996003943789935e-05, + "loss": 0.9255, + "step": 2446 + }, + { + "epoch": 0.06, + "grad_norm": 3.4123212570993173, + "learning_rate": 1.9959971262271574e-05, + "loss": 1.0627, + "step": 2447 + }, + { + "epoch": 0.06, + "grad_norm": 2.544080051397715, + "learning_rate": 1.9959903028653695e-05, + "loss": 0.9829, + "step": 2448 + }, + { + "epoch": 0.06, + "grad_norm": 2.0630619499776977, + "learning_rate": 1.9959834737046107e-05, + "loss": 1.1072, + "step": 2449 + }, + { + "epoch": 0.06, + "grad_norm": 2.2151840097239277, + "learning_rate": 1.9959766387449203e-05, + "loss": 1.0732, + "step": 2450 + }, + { + "epoch": 0.06, + "grad_norm": 2.2340795392059514, + "learning_rate": 1.995969797986339e-05, + "loss": 1.2177, + "step": 2451 + }, + { + "epoch": 0.06, + "grad_norm": 2.1980855107237955, + "learning_rate": 1.9959629514289065e-05, + "loss": 0.9655, + "step": 2452 + }, + { + "epoch": 0.06, + "grad_norm": 2.528663371995504, + "learning_rate": 1.995956099072662e-05, + "loss": 1.0626, + "step": 2453 + }, + { + "epoch": 0.06, + "grad_norm": 2.1838117931633994, + "learning_rate": 1.995949240917646e-05, + "loss": 1.0239, + "step": 2454 + }, + { + "epoch": 0.06, + "grad_norm": 2.5437393327489053, + "learning_rate": 1.9959423769638978e-05, + "loss": 1.0024, + "step": 2455 + }, + { + "epoch": 0.06, + "grad_norm": 2.33825125498194, + "learning_rate": 1.995935507211458e-05, + "loss": 1.1949, + "step": 2456 + }, + { + "epoch": 0.06, + "grad_norm": 2.08047664970663, + "learning_rate": 1.9959286316603662e-05, + "loss": 1.1023, + "step": 2457 + }, + { + "epoch": 0.06, + "grad_norm": 2.2370953245456096, + "learning_rate": 1.995921750310663e-05, + "loss": 0.9682, + "step": 2458 + }, + { + "epoch": 0.06, + "grad_norm": 3.153106300824425, + "learning_rate": 1.9959148631623878e-05, + "loss": 1.1198, + "step": 2459 + }, + { + "epoch": 0.06, + "grad_norm": 2.1539481098812163, + "learning_rate": 1.995907970215581e-05, + "loss": 1.0807, + "step": 2460 + }, + { + "epoch": 0.06, + "grad_norm": 2.112302275205232, + "learning_rate": 1.9959010714702827e-05, + "loss": 1.1634, + "step": 2461 + }, + { + "epoch": 0.06, + "grad_norm": 2.268800110397379, + "learning_rate": 1.995894166926533e-05, + "loss": 0.9956, + "step": 2462 + }, + { + "epoch": 0.06, + "grad_norm": 1.841697783373261, + "learning_rate": 1.9958872565843724e-05, + "loss": 1.0364, + "step": 2463 + }, + { + "epoch": 0.06, + "grad_norm": 1.9187651438617337, + "learning_rate": 1.9958803404438406e-05, + "loss": 0.9962, + "step": 2464 + }, + { + "epoch": 0.06, + "grad_norm": 2.1865132844015363, + "learning_rate": 1.9958734185049782e-05, + "loss": 1.1698, + "step": 2465 + }, + { + "epoch": 0.06, + "grad_norm": 1.8742468745698226, + "learning_rate": 1.9958664907678257e-05, + "loss": 1.0519, + "step": 2466 + }, + { + "epoch": 0.06, + "grad_norm": 2.0828235873264753, + "learning_rate": 1.995859557232423e-05, + "loss": 0.9943, + "step": 2467 + }, + { + "epoch": 0.06, + "grad_norm": 3.141931694484915, + "learning_rate": 1.9958526178988107e-05, + "loss": 1.2504, + "step": 2468 + }, + { + "epoch": 0.06, + "grad_norm": 2.0432829121271463, + "learning_rate": 1.9958456727670296e-05, + "loss": 1.0988, + "step": 2469 + }, + { + "epoch": 0.06, + "grad_norm": 2.446956255341302, + "learning_rate": 1.9958387218371192e-05, + "loss": 1.1, + "step": 2470 + }, + { + "epoch": 0.06, + "grad_norm": 2.2194153245068327, + "learning_rate": 1.9958317651091206e-05, + "loss": 1.235, + "step": 2471 + }, + { + "epoch": 0.06, + "grad_norm": 1.95109362333151, + "learning_rate": 1.9958248025830743e-05, + "loss": 0.9162, + "step": 2472 + }, + { + "epoch": 0.06, + "grad_norm": 2.562973368076797, + "learning_rate": 1.9958178342590205e-05, + "loss": 1.1226, + "step": 2473 + }, + { + "epoch": 0.06, + "grad_norm": 2.2452029014393573, + "learning_rate": 1.995810860137e-05, + "loss": 1.1152, + "step": 2474 + }, + { + "epoch": 0.06, + "grad_norm": 2.035260412457373, + "learning_rate": 1.9958038802170537e-05, + "loss": 1.1267, + "step": 2475 + }, + { + "epoch": 0.06, + "grad_norm": 2.351375874907219, + "learning_rate": 1.995796894499222e-05, + "loss": 1.1522, + "step": 2476 + }, + { + "epoch": 0.06, + "grad_norm": 2.8708353052383893, + "learning_rate": 1.9957899029835453e-05, + "loss": 1.0068, + "step": 2477 + }, + { + "epoch": 0.06, + "grad_norm": 2.245145375419949, + "learning_rate": 1.9957829056700643e-05, + "loss": 1.1684, + "step": 2478 + }, + { + "epoch": 0.06, + "grad_norm": 2.0525216778211255, + "learning_rate": 1.9957759025588203e-05, + "loss": 1.0189, + "step": 2479 + }, + { + "epoch": 0.06, + "grad_norm": 2.4528720787592078, + "learning_rate": 1.9957688936498537e-05, + "loss": 1.0567, + "step": 2480 + }, + { + "epoch": 0.06, + "grad_norm": 2.68129822919947, + "learning_rate": 1.9957618789432053e-05, + "loss": 1.0008, + "step": 2481 + }, + { + "epoch": 0.06, + "grad_norm": 2.3101815094436047, + "learning_rate": 1.995754858438916e-05, + "loss": 1.1812, + "step": 2482 + }, + { + "epoch": 0.06, + "grad_norm": 2.0607349043294327, + "learning_rate": 1.9957478321370266e-05, + "loss": 1.1637, + "step": 2483 + }, + { + "epoch": 0.06, + "grad_norm": 2.046752378084067, + "learning_rate": 1.995740800037578e-05, + "loss": 0.9976, + "step": 2484 + }, + { + "epoch": 0.06, + "grad_norm": 5.475414708824379, + "learning_rate": 1.9957337621406112e-05, + "loss": 1.0419, + "step": 2485 + }, + { + "epoch": 0.06, + "grad_norm": 2.372981650629567, + "learning_rate": 1.9957267184461674e-05, + "loss": 1.0835, + "step": 2486 + }, + { + "epoch": 0.06, + "grad_norm": 2.390674608810954, + "learning_rate": 1.9957196689542873e-05, + "loss": 0.9989, + "step": 2487 + }, + { + "epoch": 0.06, + "grad_norm": 2.2225724104528886, + "learning_rate": 1.995712613665012e-05, + "loss": 1.1089, + "step": 2488 + }, + { + "epoch": 0.06, + "grad_norm": 2.116509724420983, + "learning_rate": 1.995705552578383e-05, + "loss": 1.1642, + "step": 2489 + }, + { + "epoch": 0.06, + "grad_norm": 3.1854636976970854, + "learning_rate": 1.9956984856944403e-05, + "loss": 1.1472, + "step": 2490 + }, + { + "epoch": 0.06, + "grad_norm": 2.30475696038785, + "learning_rate": 1.9956914130132264e-05, + "loss": 1.1023, + "step": 2491 + }, + { + "epoch": 0.06, + "grad_norm": 2.18238337072979, + "learning_rate": 1.9956843345347817e-05, + "loss": 1.1269, + "step": 2492 + }, + { + "epoch": 0.06, + "grad_norm": 2.2601231473116505, + "learning_rate": 1.9956772502591473e-05, + "loss": 0.9733, + "step": 2493 + }, + { + "epoch": 0.06, + "grad_norm": 2.6329071349106874, + "learning_rate": 1.9956701601863653e-05, + "loss": 1.1862, + "step": 2494 + }, + { + "epoch": 0.06, + "grad_norm": 2.0543134774071956, + "learning_rate": 1.9956630643164758e-05, + "loss": 1.1874, + "step": 2495 + }, + { + "epoch": 0.06, + "grad_norm": 2.446505505776486, + "learning_rate": 1.995655962649521e-05, + "loss": 1.0633, + "step": 2496 + }, + { + "epoch": 0.06, + "grad_norm": 2.2728755579443356, + "learning_rate": 1.995648855185542e-05, + "loss": 1.1891, + "step": 2497 + }, + { + "epoch": 0.06, + "grad_norm": 2.5620417213282662, + "learning_rate": 1.9956417419245802e-05, + "loss": 1.0342, + "step": 2498 + }, + { + "epoch": 0.06, + "grad_norm": 2.414454214731966, + "learning_rate": 1.995634622866677e-05, + "loss": 0.9672, + "step": 2499 + }, + { + "epoch": 0.06, + "grad_norm": 1.1788133040076418, + "learning_rate": 1.9956274980118736e-05, + "loss": 0.8975, + "step": 2500 + }, + { + "epoch": 0.06, + "grad_norm": 2.521108070323407, + "learning_rate": 1.9956203673602116e-05, + "loss": 0.8641, + "step": 2501 + }, + { + "epoch": 0.06, + "grad_norm": 2.1606458851438255, + "learning_rate": 1.995613230911733e-05, + "loss": 1.0043, + "step": 2502 + }, + { + "epoch": 0.06, + "grad_norm": 2.175942428151818, + "learning_rate": 1.9956060886664786e-05, + "loss": 1.0601, + "step": 2503 + }, + { + "epoch": 0.06, + "grad_norm": 2.284833188125143, + "learning_rate": 1.9955989406244906e-05, + "loss": 1.1103, + "step": 2504 + }, + { + "epoch": 0.06, + "grad_norm": 1.931754006461178, + "learning_rate": 1.9955917867858102e-05, + "loss": 1.0691, + "step": 2505 + }, + { + "epoch": 0.06, + "grad_norm": 2.7595514222709503, + "learning_rate": 1.9955846271504792e-05, + "loss": 0.9079, + "step": 2506 + }, + { + "epoch": 0.06, + "grad_norm": 2.995706096877254, + "learning_rate": 1.9955774617185395e-05, + "loss": 1.1, + "step": 2507 + }, + { + "epoch": 0.06, + "grad_norm": 2.601578396134312, + "learning_rate": 1.9955702904900325e-05, + "loss": 1.1214, + "step": 2508 + }, + { + "epoch": 0.06, + "grad_norm": 2.399611834350616, + "learning_rate": 1.995563113465e-05, + "loss": 1.0781, + "step": 2509 + }, + { + "epoch": 0.06, + "grad_norm": 1.9353076408529037, + "learning_rate": 1.9955559306434843e-05, + "loss": 0.9668, + "step": 2510 + }, + { + "epoch": 0.06, + "grad_norm": 2.6088913923389825, + "learning_rate": 1.995548742025526e-05, + "loss": 1.1618, + "step": 2511 + }, + { + "epoch": 0.06, + "grad_norm": 2.44457733405967, + "learning_rate": 1.9955415476111683e-05, + "loss": 0.9478, + "step": 2512 + }, + { + "epoch": 0.06, + "grad_norm": 2.6332500246151227, + "learning_rate": 1.9955343474004526e-05, + "loss": 0.9373, + "step": 2513 + }, + { + "epoch": 0.06, + "grad_norm": 2.1370554295684565, + "learning_rate": 1.9955271413934203e-05, + "loss": 1.128, + "step": 2514 + }, + { + "epoch": 0.06, + "grad_norm": 3.9834653789009176, + "learning_rate": 1.9955199295901144e-05, + "loss": 1.0368, + "step": 2515 + }, + { + "epoch": 0.06, + "grad_norm": 2.3562570699065866, + "learning_rate": 1.9955127119905756e-05, + "loss": 1.1927, + "step": 2516 + }, + { + "epoch": 0.06, + "grad_norm": 2.860106863835045, + "learning_rate": 1.995505488594847e-05, + "loss": 1.0525, + "step": 2517 + }, + { + "epoch": 0.06, + "grad_norm": 2.1877647595178384, + "learning_rate": 1.9954982594029703e-05, + "loss": 0.9611, + "step": 2518 + }, + { + "epoch": 0.06, + "grad_norm": 1.2445860462882974, + "learning_rate": 1.9954910244149873e-05, + "loss": 0.947, + "step": 2519 + }, + { + "epoch": 0.06, + "grad_norm": 3.2401810664522195, + "learning_rate": 1.9954837836309407e-05, + "loss": 1.0964, + "step": 2520 + }, + { + "epoch": 0.06, + "grad_norm": 2.1013913269909676, + "learning_rate": 1.9954765370508723e-05, + "loss": 1.0378, + "step": 2521 + }, + { + "epoch": 0.06, + "grad_norm": 1.935320501004146, + "learning_rate": 1.995469284674824e-05, + "loss": 0.9905, + "step": 2522 + }, + { + "epoch": 0.06, + "grad_norm": 1.9965730125468912, + "learning_rate": 1.995462026502839e-05, + "loss": 1.0883, + "step": 2523 + }, + { + "epoch": 0.06, + "grad_norm": 2.0910878448491848, + "learning_rate": 1.9954547625349584e-05, + "loss": 0.9653, + "step": 2524 + }, + { + "epoch": 0.06, + "grad_norm": 3.0363536913036615, + "learning_rate": 1.9954474927712252e-05, + "loss": 1.0275, + "step": 2525 + }, + { + "epoch": 0.06, + "grad_norm": 2.100875054221008, + "learning_rate": 1.9954402172116816e-05, + "loss": 1.0299, + "step": 2526 + }, + { + "epoch": 0.06, + "grad_norm": 2.7589783896407365, + "learning_rate": 1.99543293585637e-05, + "loss": 1.0615, + "step": 2527 + }, + { + "epoch": 0.06, + "grad_norm": 2.1842202638119694, + "learning_rate": 1.9954256487053326e-05, + "loss": 1.1554, + "step": 2528 + }, + { + "epoch": 0.06, + "grad_norm": 2.2747445462192197, + "learning_rate": 1.995418355758612e-05, + "loss": 1.1121, + "step": 2529 + }, + { + "epoch": 0.06, + "grad_norm": 2.1817243130652826, + "learning_rate": 1.99541105701625e-05, + "loss": 0.9588, + "step": 2530 + }, + { + "epoch": 0.06, + "grad_norm": 2.0018206882865544, + "learning_rate": 1.99540375247829e-05, + "loss": 1.0762, + "step": 2531 + }, + { + "epoch": 0.06, + "grad_norm": 2.1713764024441824, + "learning_rate": 1.9953964421447745e-05, + "loss": 1.0943, + "step": 2532 + }, + { + "epoch": 0.06, + "grad_norm": 2.418375038610901, + "learning_rate": 1.995389126015746e-05, + "loss": 1.1042, + "step": 2533 + }, + { + "epoch": 0.06, + "grad_norm": 2.1396288327868946, + "learning_rate": 1.995381804091246e-05, + "loss": 1.1614, + "step": 2534 + }, + { + "epoch": 0.06, + "grad_norm": 2.193296603139368, + "learning_rate": 1.9953744763713188e-05, + "loss": 1.102, + "step": 2535 + }, + { + "epoch": 0.06, + "grad_norm": 2.1736079152236067, + "learning_rate": 1.995367142856006e-05, + "loss": 1.0259, + "step": 2536 + }, + { + "epoch": 0.06, + "grad_norm": 2.614241703476862, + "learning_rate": 1.9953598035453505e-05, + "loss": 1.0123, + "step": 2537 + }, + { + "epoch": 0.06, + "grad_norm": 2.798390117460582, + "learning_rate": 1.995352458439395e-05, + "loss": 1.0656, + "step": 2538 + }, + { + "epoch": 0.06, + "grad_norm": 2.274967277215867, + "learning_rate": 1.9953451075381826e-05, + "loss": 1.0949, + "step": 2539 + }, + { + "epoch": 0.06, + "grad_norm": 2.285717073420907, + "learning_rate": 1.995337750841756e-05, + "loss": 1.2005, + "step": 2540 + }, + { + "epoch": 0.06, + "grad_norm": 2.6171873185171943, + "learning_rate": 1.9953303883501577e-05, + "loss": 1.1105, + "step": 2541 + }, + { + "epoch": 0.06, + "grad_norm": 1.9691699825127245, + "learning_rate": 1.9953230200634305e-05, + "loss": 1.0431, + "step": 2542 + }, + { + "epoch": 0.06, + "grad_norm": 2.2295767601613092, + "learning_rate": 1.995315645981618e-05, + "loss": 1.1275, + "step": 2543 + }, + { + "epoch": 0.06, + "grad_norm": 1.9921424386126705, + "learning_rate": 1.9953082661047624e-05, + "loss": 1.0505, + "step": 2544 + }, + { + "epoch": 0.06, + "grad_norm": 3.825361235815474, + "learning_rate": 1.995300880432907e-05, + "loss": 1.037, + "step": 2545 + }, + { + "epoch": 0.06, + "grad_norm": 2.3136474112325582, + "learning_rate": 1.9952934889660947e-05, + "loss": 0.9582, + "step": 2546 + }, + { + "epoch": 0.06, + "grad_norm": 2.2549698030395424, + "learning_rate": 1.9952860917043687e-05, + "loss": 1.2031, + "step": 2547 + }, + { + "epoch": 0.06, + "grad_norm": 2.266257039034935, + "learning_rate": 1.9952786886477718e-05, + "loss": 1.1442, + "step": 2548 + }, + { + "epoch": 0.06, + "grad_norm": 2.2915593148097786, + "learning_rate": 1.9952712797963477e-05, + "loss": 1.1864, + "step": 2549 + }, + { + "epoch": 0.06, + "grad_norm": 1.9679570450800028, + "learning_rate": 1.995263865150139e-05, + "loss": 1.1494, + "step": 2550 + }, + { + "epoch": 0.06, + "grad_norm": 2.306558920271625, + "learning_rate": 1.995256444709189e-05, + "loss": 1.148, + "step": 2551 + }, + { + "epoch": 0.06, + "grad_norm": 2.227995237528361, + "learning_rate": 1.9952490184735404e-05, + "loss": 1.1625, + "step": 2552 + }, + { + "epoch": 0.06, + "grad_norm": 2.1021256053602233, + "learning_rate": 1.9952415864432373e-05, + "loss": 1.1934, + "step": 2553 + }, + { + "epoch": 0.06, + "grad_norm": 1.9928515645801865, + "learning_rate": 1.9952341486183228e-05, + "loss": 1.1231, + "step": 2554 + }, + { + "epoch": 0.06, + "grad_norm": 2.2221698867849926, + "learning_rate": 1.9952267049988393e-05, + "loss": 1.0631, + "step": 2555 + }, + { + "epoch": 0.06, + "grad_norm": 1.1525035864492095, + "learning_rate": 1.9952192555848312e-05, + "loss": 0.929, + "step": 2556 + }, + { + "epoch": 0.06, + "grad_norm": 2.3018945046259005, + "learning_rate": 1.995211800376342e-05, + "loss": 1.1014, + "step": 2557 + }, + { + "epoch": 0.06, + "grad_norm": 2.2082195195251284, + "learning_rate": 1.995204339373414e-05, + "loss": 1.0842, + "step": 2558 + }, + { + "epoch": 0.06, + "grad_norm": 2.015722460710024, + "learning_rate": 1.9951968725760916e-05, + "loss": 1.0476, + "step": 2559 + }, + { + "epoch": 0.06, + "grad_norm": 2.1832342781728595, + "learning_rate": 1.9951893999844178e-05, + "loss": 1.0924, + "step": 2560 + }, + { + "epoch": 0.06, + "grad_norm": 2.1548242633494246, + "learning_rate": 1.995181921598436e-05, + "loss": 1.0879, + "step": 2561 + }, + { + "epoch": 0.06, + "grad_norm": 2.0719168760111892, + "learning_rate": 1.9951744374181905e-05, + "loss": 1.0946, + "step": 2562 + }, + { + "epoch": 0.06, + "grad_norm": 2.1230164816776824, + "learning_rate": 1.9951669474437244e-05, + "loss": 1.0885, + "step": 2563 + }, + { + "epoch": 0.06, + "grad_norm": 2.0627776427770343, + "learning_rate": 1.9951594516750807e-05, + "loss": 1.2366, + "step": 2564 + }, + { + "epoch": 0.06, + "grad_norm": 2.7311165645551503, + "learning_rate": 1.995151950112304e-05, + "loss": 1.0691, + "step": 2565 + }, + { + "epoch": 0.06, + "grad_norm": 2.026569867741104, + "learning_rate": 1.9951444427554375e-05, + "loss": 1.0426, + "step": 2566 + }, + { + "epoch": 0.06, + "grad_norm": 2.461794992663451, + "learning_rate": 1.9951369296045247e-05, + "loss": 1.0864, + "step": 2567 + }, + { + "epoch": 0.06, + "grad_norm": 2.2326539659253175, + "learning_rate": 1.9951294106596098e-05, + "loss": 1.1287, + "step": 2568 + }, + { + "epoch": 0.06, + "grad_norm": 2.1277867824393817, + "learning_rate": 1.9951218859207368e-05, + "loss": 1.0787, + "step": 2569 + }, + { + "epoch": 0.06, + "grad_norm": 2.444837615389925, + "learning_rate": 1.995114355387949e-05, + "loss": 1.0728, + "step": 2570 + }, + { + "epoch": 0.06, + "grad_norm": 2.521494327340109, + "learning_rate": 1.99510681906129e-05, + "loss": 1.0514, + "step": 2571 + }, + { + "epoch": 0.06, + "grad_norm": 1.8653810000958149, + "learning_rate": 1.9950992769408043e-05, + "loss": 1.1654, + "step": 2572 + }, + { + "epoch": 0.06, + "grad_norm": 2.5087715673504842, + "learning_rate": 1.9950917290265356e-05, + "loss": 1.1204, + "step": 2573 + }, + { + "epoch": 0.06, + "grad_norm": 2.4500242917069364, + "learning_rate": 1.9950841753185276e-05, + "loss": 1.0714, + "step": 2574 + }, + { + "epoch": 0.06, + "grad_norm": 1.9018799650453466, + "learning_rate": 1.995076615816825e-05, + "loss": 1.0105, + "step": 2575 + }, + { + "epoch": 0.06, + "grad_norm": 2.120778560216815, + "learning_rate": 1.9950690505214707e-05, + "loss": 0.9227, + "step": 2576 + }, + { + "epoch": 0.06, + "grad_norm": 2.2034221074825786, + "learning_rate": 1.9950614794325096e-05, + "loss": 0.9976, + "step": 2577 + }, + { + "epoch": 0.06, + "grad_norm": 3.0783561040230283, + "learning_rate": 1.9950539025499857e-05, + "loss": 0.9555, + "step": 2578 + }, + { + "epoch": 0.06, + "grad_norm": 2.0159867252850847, + "learning_rate": 1.995046319873943e-05, + "loss": 1.0395, + "step": 2579 + }, + { + "epoch": 0.06, + "grad_norm": 3.035510596802366, + "learning_rate": 1.9950387314044254e-05, + "loss": 1.2392, + "step": 2580 + }, + { + "epoch": 0.06, + "grad_norm": 2.0474535619444034, + "learning_rate": 1.995031137141477e-05, + "loss": 1.0338, + "step": 2581 + }, + { + "epoch": 0.06, + "grad_norm": 1.846236057707564, + "learning_rate": 1.9950235370851428e-05, + "loss": 1.0822, + "step": 2582 + }, + { + "epoch": 0.06, + "grad_norm": 4.1648888996911735, + "learning_rate": 1.9950159312354664e-05, + "loss": 1.2032, + "step": 2583 + }, + { + "epoch": 0.06, + "grad_norm": 2.0383199217366585, + "learning_rate": 1.9950083195924922e-05, + "loss": 1.0625, + "step": 2584 + }, + { + "epoch": 0.06, + "grad_norm": 2.154482851945378, + "learning_rate": 1.9950007021562643e-05, + "loss": 1.1812, + "step": 2585 + }, + { + "epoch": 0.06, + "grad_norm": 2.444021534045529, + "learning_rate": 1.9949930789268274e-05, + "loss": 1.1087, + "step": 2586 + }, + { + "epoch": 0.06, + "grad_norm": 2.221284882133503, + "learning_rate": 1.994985449904226e-05, + "loss": 1.0991, + "step": 2587 + }, + { + "epoch": 0.06, + "grad_norm": 2.018890216981058, + "learning_rate": 1.994977815088504e-05, + "loss": 1.0595, + "step": 2588 + }, + { + "epoch": 0.06, + "grad_norm": 2.058187645100828, + "learning_rate": 1.994970174479707e-05, + "loss": 1.1569, + "step": 2589 + }, + { + "epoch": 0.06, + "grad_norm": 1.9767851794796207, + "learning_rate": 1.994962528077878e-05, + "loss": 1.1603, + "step": 2590 + }, + { + "epoch": 0.06, + "grad_norm": 1.917824214337298, + "learning_rate": 1.9949548758830618e-05, + "loss": 1.1008, + "step": 2591 + }, + { + "epoch": 0.06, + "grad_norm": 2.349748842063109, + "learning_rate": 1.9949472178953036e-05, + "loss": 1.1408, + "step": 2592 + }, + { + "epoch": 0.06, + "grad_norm": 2.358015969363509, + "learning_rate": 1.9949395541146475e-05, + "loss": 1.095, + "step": 2593 + }, + { + "epoch": 0.06, + "grad_norm": 2.2885608206752126, + "learning_rate": 1.994931884541139e-05, + "loss": 1.1844, + "step": 2594 + }, + { + "epoch": 0.06, + "grad_norm": 2.075704855575552, + "learning_rate": 1.9949242091748217e-05, + "loss": 1.035, + "step": 2595 + }, + { + "epoch": 0.06, + "grad_norm": 1.1841014072956533, + "learning_rate": 1.9949165280157404e-05, + "loss": 1.0234, + "step": 2596 + }, + { + "epoch": 0.06, + "grad_norm": 2.165719571554104, + "learning_rate": 1.9949088410639407e-05, + "loss": 1.0151, + "step": 2597 + }, + { + "epoch": 0.06, + "grad_norm": 2.0761425690453184, + "learning_rate": 1.9949011483194663e-05, + "loss": 1.0871, + "step": 2598 + }, + { + "epoch": 0.06, + "grad_norm": 2.205693214475173, + "learning_rate": 1.9948934497823626e-05, + "loss": 1.1658, + "step": 2599 + }, + { + "epoch": 0.06, + "grad_norm": 2.0150986528741233, + "learning_rate": 1.994885745452674e-05, + "loss": 0.9423, + "step": 2600 + }, + { + "epoch": 0.06, + "grad_norm": 1.906749761207398, + "learning_rate": 1.994878035330446e-05, + "loss": 0.9589, + "step": 2601 + }, + { + "epoch": 0.06, + "grad_norm": 2.551351418435861, + "learning_rate": 1.9948703194157227e-05, + "loss": 1.055, + "step": 2602 + }, + { + "epoch": 0.06, + "grad_norm": 1.8773393285753253, + "learning_rate": 1.9948625977085497e-05, + "loss": 1.0391, + "step": 2603 + }, + { + "epoch": 0.06, + "grad_norm": 2.1544827378255715, + "learning_rate": 1.9948548702089715e-05, + "loss": 0.9537, + "step": 2604 + }, + { + "epoch": 0.06, + "grad_norm": 2.2752546325041525, + "learning_rate": 1.9948471369170336e-05, + "loss": 1.2113, + "step": 2605 + }, + { + "epoch": 0.06, + "grad_norm": 3.5895314168300057, + "learning_rate": 1.9948393978327802e-05, + "loss": 0.9531, + "step": 2606 + }, + { + "epoch": 0.06, + "grad_norm": 1.9560265429543597, + "learning_rate": 1.9948316529562573e-05, + "loss": 1.1985, + "step": 2607 + }, + { + "epoch": 0.06, + "grad_norm": 2.070102211218434, + "learning_rate": 1.9948239022875094e-05, + "loss": 1.0116, + "step": 2608 + }, + { + "epoch": 0.06, + "grad_norm": 2.00365518473696, + "learning_rate": 1.9948161458265817e-05, + "loss": 0.9644, + "step": 2609 + }, + { + "epoch": 0.06, + "grad_norm": 2.2288780634391805, + "learning_rate": 1.9948083835735198e-05, + "loss": 1.0481, + "step": 2610 + }, + { + "epoch": 0.06, + "grad_norm": 2.357888831589029, + "learning_rate": 1.994800615528368e-05, + "loss": 1.1212, + "step": 2611 + }, + { + "epoch": 0.06, + "grad_norm": 2.316597577129313, + "learning_rate": 1.9947928416911723e-05, + "loss": 1.0439, + "step": 2612 + }, + { + "epoch": 0.06, + "grad_norm": 2.0538449366139497, + "learning_rate": 1.994785062061978e-05, + "loss": 1.1071, + "step": 2613 + }, + { + "epoch": 0.06, + "grad_norm": 1.8924986888564115, + "learning_rate": 1.9947772766408298e-05, + "loss": 0.8313, + "step": 2614 + }, + { + "epoch": 0.06, + "grad_norm": 2.5181028984063394, + "learning_rate": 1.9947694854277734e-05, + "loss": 1.0863, + "step": 2615 + }, + { + "epoch": 0.06, + "grad_norm": 2.469729355644194, + "learning_rate": 1.9947616884228545e-05, + "loss": 0.8812, + "step": 2616 + }, + { + "epoch": 0.06, + "grad_norm": 2.078459020777473, + "learning_rate": 1.9947538856261177e-05, + "loss": 1.106, + "step": 2617 + }, + { + "epoch": 0.06, + "grad_norm": 2.055595728900525, + "learning_rate": 1.994746077037609e-05, + "loss": 1.1096, + "step": 2618 + }, + { + "epoch": 0.06, + "grad_norm": 2.05036054360396, + "learning_rate": 1.9947382626573736e-05, + "loss": 1.1239, + "step": 2619 + }, + { + "epoch": 0.06, + "grad_norm": 2.0977359168101457, + "learning_rate": 1.994730442485457e-05, + "loss": 1.0082, + "step": 2620 + }, + { + "epoch": 0.06, + "grad_norm": 2.206833342523566, + "learning_rate": 1.994722616521905e-05, + "loss": 1.1503, + "step": 2621 + }, + { + "epoch": 0.06, + "grad_norm": 1.1373152337411558, + "learning_rate": 1.994714784766763e-05, + "loss": 0.9424, + "step": 2622 + }, + { + "epoch": 0.06, + "grad_norm": 2.98662833530934, + "learning_rate": 1.9947069472200766e-05, + "loss": 1.0183, + "step": 2623 + }, + { + "epoch": 0.06, + "grad_norm": 1.1300847720788987, + "learning_rate": 1.9946991038818916e-05, + "loss": 0.9848, + "step": 2624 + }, + { + "epoch": 0.06, + "grad_norm": 2.14278944885048, + "learning_rate": 1.994691254752253e-05, + "loss": 1.2193, + "step": 2625 + }, + { + "epoch": 0.06, + "grad_norm": 2.1495264063848025, + "learning_rate": 1.9946833998312077e-05, + "loss": 1.1224, + "step": 2626 + }, + { + "epoch": 0.06, + "grad_norm": 2.342993410102491, + "learning_rate": 1.9946755391188e-05, + "loss": 1.122, + "step": 2627 + }, + { + "epoch": 0.06, + "grad_norm": 2.1384808389702985, + "learning_rate": 1.994667672615077e-05, + "loss": 0.9915, + "step": 2628 + }, + { + "epoch": 0.06, + "grad_norm": 2.2051346182872096, + "learning_rate": 1.9946598003200837e-05, + "loss": 1.1101, + "step": 2629 + }, + { + "epoch": 0.06, + "grad_norm": 2.1098429740249833, + "learning_rate": 1.994651922233866e-05, + "loss": 1.2462, + "step": 2630 + }, + { + "epoch": 0.06, + "grad_norm": 2.0843044636707373, + "learning_rate": 1.9946440383564703e-05, + "loss": 1.1728, + "step": 2631 + }, + { + "epoch": 0.06, + "grad_norm": 2.2047562257582185, + "learning_rate": 1.9946361486879416e-05, + "loss": 1.13, + "step": 2632 + }, + { + "epoch": 0.06, + "grad_norm": 2.117133531815156, + "learning_rate": 1.9946282532283266e-05, + "loss": 1.171, + "step": 2633 + }, + { + "epoch": 0.06, + "grad_norm": 2.0351827675447276, + "learning_rate": 1.994620351977671e-05, + "loss": 1.0499, + "step": 2634 + }, + { + "epoch": 0.06, + "grad_norm": 2.6423740754887812, + "learning_rate": 1.9946124449360205e-05, + "loss": 1.1096, + "step": 2635 + }, + { + "epoch": 0.06, + "grad_norm": 2.3284387670088136, + "learning_rate": 1.9946045321034223e-05, + "loss": 1.0864, + "step": 2636 + }, + { + "epoch": 0.06, + "grad_norm": 2.133890121432458, + "learning_rate": 1.9945966134799207e-05, + "loss": 1.0506, + "step": 2637 + }, + { + "epoch": 0.06, + "grad_norm": 2.136824694793011, + "learning_rate": 1.9945886890655632e-05, + "loss": 0.8785, + "step": 2638 + }, + { + "epoch": 0.06, + "grad_norm": 2.292740312806252, + "learning_rate": 1.9945807588603954e-05, + "loss": 1.1418, + "step": 2639 + }, + { + "epoch": 0.06, + "grad_norm": 2.26442185181756, + "learning_rate": 1.9945728228644636e-05, + "loss": 1.1149, + "step": 2640 + }, + { + "epoch": 0.06, + "grad_norm": 2.387761926404365, + "learning_rate": 1.994564881077814e-05, + "loss": 1.0082, + "step": 2641 + }, + { + "epoch": 0.06, + "grad_norm": 2.3797081897741634, + "learning_rate": 1.9945569335004925e-05, + "loss": 1.0255, + "step": 2642 + }, + { + "epoch": 0.06, + "grad_norm": 2.2568113995451555, + "learning_rate": 1.994548980132546e-05, + "loss": 1.1218, + "step": 2643 + }, + { + "epoch": 0.06, + "grad_norm": 2.114717391613573, + "learning_rate": 1.9945410209740202e-05, + "loss": 1.0565, + "step": 2644 + }, + { + "epoch": 0.06, + "grad_norm": 1.884172613980419, + "learning_rate": 1.994533056024962e-05, + "loss": 0.9895, + "step": 2645 + }, + { + "epoch": 0.06, + "grad_norm": 1.9490455172289936, + "learning_rate": 1.994525085285417e-05, + "loss": 1.0991, + "step": 2646 + }, + { + "epoch": 0.06, + "grad_norm": 2.2786385170501235, + "learning_rate": 1.9945171087554324e-05, + "loss": 1.0894, + "step": 2647 + }, + { + "epoch": 0.06, + "grad_norm": 2.7257199100725003, + "learning_rate": 1.9945091264350545e-05, + "loss": 0.9561, + "step": 2648 + }, + { + "epoch": 0.06, + "grad_norm": 2.203417533596689, + "learning_rate": 1.9945011383243296e-05, + "loss": 1.1141, + "step": 2649 + }, + { + "epoch": 0.06, + "grad_norm": 2.1373703331859013, + "learning_rate": 1.994493144423304e-05, + "loss": 1.046, + "step": 2650 + }, + { + "epoch": 0.06, + "grad_norm": 4.782549193932529, + "learning_rate": 1.9944851447320243e-05, + "loss": 1.1914, + "step": 2651 + }, + { + "epoch": 0.06, + "grad_norm": 2.1667664527648585, + "learning_rate": 1.9944771392505374e-05, + "loss": 1.0771, + "step": 2652 + }, + { + "epoch": 0.06, + "grad_norm": 2.647526315534747, + "learning_rate": 1.99446912797889e-05, + "loss": 1.295, + "step": 2653 + }, + { + "epoch": 0.06, + "grad_norm": 2.039395038027411, + "learning_rate": 1.9944611109171284e-05, + "loss": 1.0291, + "step": 2654 + }, + { + "epoch": 0.06, + "grad_norm": 2.363461797493314, + "learning_rate": 1.9944530880652993e-05, + "loss": 1.1265, + "step": 2655 + }, + { + "epoch": 0.06, + "grad_norm": 2.1400307714620754, + "learning_rate": 1.9944450594234492e-05, + "loss": 1.0169, + "step": 2656 + }, + { + "epoch": 0.06, + "grad_norm": 2.1775346260237747, + "learning_rate": 1.9944370249916254e-05, + "loss": 1.1452, + "step": 2657 + }, + { + "epoch": 0.06, + "grad_norm": 2.0708144782705555, + "learning_rate": 1.9944289847698744e-05, + "loss": 1.0621, + "step": 2658 + }, + { + "epoch": 0.06, + "grad_norm": 2.1956414426704343, + "learning_rate": 1.994420938758243e-05, + "loss": 0.9287, + "step": 2659 + }, + { + "epoch": 0.06, + "grad_norm": 3.624953504546289, + "learning_rate": 1.9944128869567782e-05, + "loss": 1.0766, + "step": 2660 + }, + { + "epoch": 0.06, + "grad_norm": 2.0731938807276755, + "learning_rate": 1.9944048293655266e-05, + "loss": 1.1505, + "step": 2661 + }, + { + "epoch": 0.06, + "grad_norm": 2.1130883763229846, + "learning_rate": 1.994396765984535e-05, + "loss": 1.1786, + "step": 2662 + }, + { + "epoch": 0.06, + "grad_norm": 2.210413052141052, + "learning_rate": 1.9943886968138512e-05, + "loss": 1.0341, + "step": 2663 + }, + { + "epoch": 0.06, + "grad_norm": 2.180342898099178, + "learning_rate": 1.9943806218535212e-05, + "loss": 0.981, + "step": 2664 + }, + { + "epoch": 0.06, + "grad_norm": 2.2067397985399557, + "learning_rate": 1.9943725411035926e-05, + "loss": 1.1564, + "step": 2665 + }, + { + "epoch": 0.06, + "grad_norm": 2.0269546224923447, + "learning_rate": 1.994364454564112e-05, + "loss": 1.0911, + "step": 2666 + }, + { + "epoch": 0.06, + "grad_norm": 2.432761278980679, + "learning_rate": 1.9943563622351267e-05, + "loss": 0.9831, + "step": 2667 + }, + { + "epoch": 0.06, + "grad_norm": 2.0918738353687254, + "learning_rate": 1.994348264116684e-05, + "loss": 0.9653, + "step": 2668 + }, + { + "epoch": 0.06, + "grad_norm": 3.0637702664299735, + "learning_rate": 1.994340160208831e-05, + "loss": 1.1406, + "step": 2669 + }, + { + "epoch": 0.06, + "grad_norm": 2.0497425184867795, + "learning_rate": 1.994332050511615e-05, + "loss": 1.137, + "step": 2670 + }, + { + "epoch": 0.06, + "grad_norm": 1.195483981852251, + "learning_rate": 1.9943239350250823e-05, + "loss": 0.9243, + "step": 2671 + }, + { + "epoch": 0.06, + "grad_norm": 2.3116342843725963, + "learning_rate": 1.9943158137492815e-05, + "loss": 1.1768, + "step": 2672 + }, + { + "epoch": 0.06, + "grad_norm": 3.7630496156531135, + "learning_rate": 1.994307686684259e-05, + "loss": 1.0847, + "step": 2673 + }, + { + "epoch": 0.06, + "grad_norm": 2.1489819741654315, + "learning_rate": 1.9942995538300624e-05, + "loss": 0.991, + "step": 2674 + }, + { + "epoch": 0.06, + "grad_norm": 2.180764047546616, + "learning_rate": 1.994291415186739e-05, + "loss": 1.0402, + "step": 2675 + }, + { + "epoch": 0.06, + "grad_norm": 2.2673693787962734, + "learning_rate": 1.994283270754336e-05, + "loss": 1.1695, + "step": 2676 + }, + { + "epoch": 0.06, + "grad_norm": 1.126694092491984, + "learning_rate": 1.994275120532901e-05, + "loss": 0.9791, + "step": 2677 + }, + { + "epoch": 0.06, + "grad_norm": 2.37446332864954, + "learning_rate": 1.994266964522482e-05, + "loss": 0.9665, + "step": 2678 + }, + { + "epoch": 0.06, + "grad_norm": 1.1641088046874633, + "learning_rate": 1.9942588027231254e-05, + "loss": 0.941, + "step": 2679 + }, + { + "epoch": 0.06, + "grad_norm": 2.5403334499450927, + "learning_rate": 1.9942506351348796e-05, + "loss": 1.1454, + "step": 2680 + }, + { + "epoch": 0.06, + "grad_norm": 1.975417514760013, + "learning_rate": 1.9942424617577915e-05, + "loss": 1.0666, + "step": 2681 + }, + { + "epoch": 0.06, + "grad_norm": 1.9675365729031173, + "learning_rate": 1.994234282591909e-05, + "loss": 1.0843, + "step": 2682 + }, + { + "epoch": 0.06, + "grad_norm": 2.9804376992946144, + "learning_rate": 1.9942260976372802e-05, + "loss": 0.9945, + "step": 2683 + }, + { + "epoch": 0.06, + "grad_norm": 2.1166294572591413, + "learning_rate": 1.994217906893952e-05, + "loss": 1.0549, + "step": 2684 + }, + { + "epoch": 0.06, + "grad_norm": 2.17254069526228, + "learning_rate": 1.9942097103619728e-05, + "loss": 0.896, + "step": 2685 + }, + { + "epoch": 0.06, + "grad_norm": 2.0221305080602128, + "learning_rate": 1.9942015080413894e-05, + "loss": 0.9515, + "step": 2686 + }, + { + "epoch": 0.06, + "grad_norm": 2.172676223688277, + "learning_rate": 1.99419329993225e-05, + "loss": 1.0307, + "step": 2687 + }, + { + "epoch": 0.06, + "grad_norm": 2.5661553944955413, + "learning_rate": 1.9941850860346024e-05, + "loss": 0.9803, + "step": 2688 + }, + { + "epoch": 0.06, + "grad_norm": 7.666308632888027, + "learning_rate": 1.9941768663484948e-05, + "loss": 1.0843, + "step": 2689 + }, + { + "epoch": 0.06, + "grad_norm": 2.2061045373170547, + "learning_rate": 1.9941686408739748e-05, + "loss": 1.1685, + "step": 2690 + }, + { + "epoch": 0.06, + "grad_norm": 1.8969219674760138, + "learning_rate": 1.99416040961109e-05, + "loss": 1.0685, + "step": 2691 + }, + { + "epoch": 0.06, + "grad_norm": 2.4626665565517567, + "learning_rate": 1.9941521725598888e-05, + "loss": 0.9918, + "step": 2692 + }, + { + "epoch": 0.06, + "grad_norm": 2.0623544152582527, + "learning_rate": 1.9941439297204185e-05, + "loss": 1.0614, + "step": 2693 + }, + { + "epoch": 0.06, + "grad_norm": 2.1434712556664812, + "learning_rate": 1.994135681092728e-05, + "loss": 0.9693, + "step": 2694 + }, + { + "epoch": 0.06, + "grad_norm": 1.8323347159584475, + "learning_rate": 1.9941274266768643e-05, + "loss": 1.0129, + "step": 2695 + }, + { + "epoch": 0.06, + "grad_norm": 1.9561341716992946, + "learning_rate": 1.9941191664728763e-05, + "loss": 1.0581, + "step": 2696 + }, + { + "epoch": 0.06, + "grad_norm": 2.4924523944683816, + "learning_rate": 1.9941109004808117e-05, + "loss": 1.059, + "step": 2697 + }, + { + "epoch": 0.06, + "grad_norm": 2.4239104923205512, + "learning_rate": 1.9941026287007186e-05, + "loss": 1.0604, + "step": 2698 + }, + { + "epoch": 0.06, + "grad_norm": 2.1142964401416635, + "learning_rate": 1.9940943511326453e-05, + "loss": 1.0489, + "step": 2699 + }, + { + "epoch": 0.06, + "grad_norm": 2.138021348950958, + "learning_rate": 1.9940860677766404e-05, + "loss": 1.1746, + "step": 2700 + }, + { + "epoch": 0.06, + "grad_norm": 2.2812096033087674, + "learning_rate": 1.9940777786327513e-05, + "loss": 1.1737, + "step": 2701 + }, + { + "epoch": 0.06, + "grad_norm": 2.3679630506386977, + "learning_rate": 1.9940694837010265e-05, + "loss": 1.0547, + "step": 2702 + }, + { + "epoch": 0.06, + "grad_norm": 2.1582292384729773, + "learning_rate": 1.994061182981515e-05, + "loss": 1.0376, + "step": 2703 + }, + { + "epoch": 0.06, + "grad_norm": 3.1660767992188137, + "learning_rate": 1.994052876474264e-05, + "loss": 1.1339, + "step": 2704 + }, + { + "epoch": 0.06, + "grad_norm": 1.2391682422228916, + "learning_rate": 1.9940445641793228e-05, + "loss": 1.0014, + "step": 2705 + }, + { + "epoch": 0.06, + "grad_norm": 1.976468053005894, + "learning_rate": 1.994036246096739e-05, + "loss": 1.0378, + "step": 2706 + }, + { + "epoch": 0.06, + "grad_norm": 2.017513330998783, + "learning_rate": 1.9940279222265618e-05, + "loss": 1.0093, + "step": 2707 + }, + { + "epoch": 0.06, + "grad_norm": 2.515170959701697, + "learning_rate": 1.9940195925688395e-05, + "loss": 0.9359, + "step": 2708 + }, + { + "epoch": 0.06, + "grad_norm": 2.078841161748314, + "learning_rate": 1.9940112571236203e-05, + "loss": 0.9976, + "step": 2709 + }, + { + "epoch": 0.06, + "grad_norm": 2.48009424019978, + "learning_rate": 1.9940029158909525e-05, + "loss": 1.0526, + "step": 2710 + }, + { + "epoch": 0.06, + "grad_norm": 2.1985756179749, + "learning_rate": 1.9939945688708854e-05, + "loss": 1.0212, + "step": 2711 + }, + { + "epoch": 0.06, + "grad_norm": 2.2128697003564626, + "learning_rate": 1.9939862160634674e-05, + "loss": 1.0877, + "step": 2712 + }, + { + "epoch": 0.06, + "grad_norm": 2.042144986086268, + "learning_rate": 1.9939778574687465e-05, + "loss": 1.0465, + "step": 2713 + }, + { + "epoch": 0.06, + "grad_norm": 1.1260729245608565, + "learning_rate": 1.993969493086772e-05, + "loss": 1.0309, + "step": 2714 + }, + { + "epoch": 0.06, + "grad_norm": 1.2381709187675205, + "learning_rate": 1.9939611229175926e-05, + "loss": 1.0321, + "step": 2715 + }, + { + "epoch": 0.06, + "grad_norm": 1.2432002828674968, + "learning_rate": 1.9939527469612567e-05, + "loss": 1.0448, + "step": 2716 + }, + { + "epoch": 0.06, + "grad_norm": 2.356905932389591, + "learning_rate": 1.9939443652178136e-05, + "loss": 1.0898, + "step": 2717 + }, + { + "epoch": 0.06, + "grad_norm": 2.4364311141524104, + "learning_rate": 1.9939359776873112e-05, + "loss": 1.1414, + "step": 2718 + }, + { + "epoch": 0.06, + "grad_norm": 2.150909671173729, + "learning_rate": 1.9939275843697988e-05, + "loss": 1.0377, + "step": 2719 + }, + { + "epoch": 0.06, + "grad_norm": 2.0916105553827933, + "learning_rate": 1.9939191852653257e-05, + "loss": 1.1981, + "step": 2720 + }, + { + "epoch": 0.06, + "grad_norm": 2.051950500587544, + "learning_rate": 1.9939107803739402e-05, + "loss": 1.0997, + "step": 2721 + }, + { + "epoch": 0.06, + "grad_norm": 2.1524787364410587, + "learning_rate": 1.9939023696956916e-05, + "loss": 1.1399, + "step": 2722 + }, + { + "epoch": 0.06, + "grad_norm": 1.8063207738208984, + "learning_rate": 1.993893953230629e-05, + "loss": 1.0658, + "step": 2723 + }, + { + "epoch": 0.06, + "grad_norm": 2.1504062425512287, + "learning_rate": 1.993885530978801e-05, + "loss": 1.0733, + "step": 2724 + }, + { + "epoch": 0.06, + "grad_norm": 2.242959571102598, + "learning_rate": 1.9938771029402564e-05, + "loss": 1.0937, + "step": 2725 + }, + { + "epoch": 0.06, + "grad_norm": 2.2588490851517338, + "learning_rate": 1.993868669115045e-05, + "loss": 1.1289, + "step": 2726 + }, + { + "epoch": 0.06, + "grad_norm": 2.1735836931202948, + "learning_rate": 1.993860229503215e-05, + "loss": 0.9627, + "step": 2727 + }, + { + "epoch": 0.06, + "grad_norm": 2.046730421780567, + "learning_rate": 1.9938517841048167e-05, + "loss": 1.1279, + "step": 2728 + }, + { + "epoch": 0.06, + "grad_norm": 1.1567879908078793, + "learning_rate": 1.9938433329198985e-05, + "loss": 0.8966, + "step": 2729 + }, + { + "epoch": 0.06, + "grad_norm": 2.1615778167094253, + "learning_rate": 1.99383487594851e-05, + "loss": 1.0786, + "step": 2730 + }, + { + "epoch": 0.06, + "grad_norm": 1.9781241182559581, + "learning_rate": 1.9938264131906998e-05, + "loss": 1.1033, + "step": 2731 + }, + { + "epoch": 0.06, + "grad_norm": 6.588948365858162, + "learning_rate": 1.993817944646518e-05, + "loss": 1.0807, + "step": 2732 + }, + { + "epoch": 0.06, + "grad_norm": 1.2073869031738222, + "learning_rate": 1.993809470316013e-05, + "loss": 1.028, + "step": 2733 + }, + { + "epoch": 0.06, + "grad_norm": 2.6140031133865853, + "learning_rate": 1.993800990199235e-05, + "loss": 1.1238, + "step": 2734 + }, + { + "epoch": 0.06, + "grad_norm": 2.1151519553606475, + "learning_rate": 1.993792504296233e-05, + "loss": 0.961, + "step": 2735 + }, + { + "epoch": 0.06, + "grad_norm": 1.9809178699207797, + "learning_rate": 1.993784012607056e-05, + "loss": 1.0584, + "step": 2736 + }, + { + "epoch": 0.06, + "grad_norm": 2.49168006426883, + "learning_rate": 1.993775515131754e-05, + "loss": 0.9303, + "step": 2737 + }, + { + "epoch": 0.06, + "grad_norm": 2.2384086712215057, + "learning_rate": 1.9937670118703768e-05, + "loss": 0.9953, + "step": 2738 + }, + { + "epoch": 0.06, + "grad_norm": 1.14847829561371, + "learning_rate": 1.993758502822973e-05, + "loss": 0.9909, + "step": 2739 + }, + { + "epoch": 0.06, + "grad_norm": 2.0106468794116643, + "learning_rate": 1.993749987989593e-05, + "loss": 1.1712, + "step": 2740 + }, + { + "epoch": 0.06, + "grad_norm": 1.9743084144034293, + "learning_rate": 1.9937414673702852e-05, + "loss": 1.0725, + "step": 2741 + }, + { + "epoch": 0.06, + "grad_norm": 1.9568556812834075, + "learning_rate": 1.9937329409651007e-05, + "loss": 1.0684, + "step": 2742 + }, + { + "epoch": 0.06, + "grad_norm": 2.0095536692061096, + "learning_rate": 1.993724408774088e-05, + "loss": 1.1334, + "step": 2743 + }, + { + "epoch": 0.06, + "grad_norm": 2.234436530215517, + "learning_rate": 1.9937158707972975e-05, + "loss": 0.9367, + "step": 2744 + }, + { + "epoch": 0.06, + "grad_norm": 2.1972007851040654, + "learning_rate": 1.9937073270347782e-05, + "loss": 0.9581, + "step": 2745 + }, + { + "epoch": 0.06, + "grad_norm": 2.5288766715348707, + "learning_rate": 1.9936987774865805e-05, + "loss": 1.0915, + "step": 2746 + }, + { + "epoch": 0.06, + "grad_norm": 1.1457445773154102, + "learning_rate": 1.9936902221527536e-05, + "loss": 0.8962, + "step": 2747 + }, + { + "epoch": 0.06, + "grad_norm": 1.9450503472721983, + "learning_rate": 1.993681661033348e-05, + "loss": 0.9429, + "step": 2748 + }, + { + "epoch": 0.06, + "grad_norm": 1.9941214541626202, + "learning_rate": 1.993673094128413e-05, + "loss": 1.0059, + "step": 2749 + }, + { + "epoch": 0.06, + "grad_norm": 2.3593316615976403, + "learning_rate": 1.9936645214379988e-05, + "loss": 1.0749, + "step": 2750 + }, + { + "epoch": 0.06, + "grad_norm": 1.1186678666586736, + "learning_rate": 1.993655942962155e-05, + "loss": 0.9403, + "step": 2751 + }, + { + "epoch": 0.06, + "grad_norm": 2.850309669754671, + "learning_rate": 1.993647358700932e-05, + "loss": 1.0422, + "step": 2752 + }, + { + "epoch": 0.06, + "grad_norm": 3.1624227358816106, + "learning_rate": 1.993638768654379e-05, + "loss": 1.0922, + "step": 2753 + }, + { + "epoch": 0.06, + "grad_norm": 2.330685846351926, + "learning_rate": 1.993630172822547e-05, + "loss": 1.1562, + "step": 2754 + }, + { + "epoch": 0.06, + "grad_norm": 2.510137082416157, + "learning_rate": 1.9936215712054854e-05, + "loss": 1.0385, + "step": 2755 + }, + { + "epoch": 0.06, + "grad_norm": 2.021567132022931, + "learning_rate": 1.9936129638032443e-05, + "loss": 1.0369, + "step": 2756 + }, + { + "epoch": 0.06, + "grad_norm": 2.6435913756107245, + "learning_rate": 1.9936043506158742e-05, + "loss": 1.064, + "step": 2757 + }, + { + "epoch": 0.06, + "grad_norm": 2.3785161639496595, + "learning_rate": 1.9935957316434246e-05, + "loss": 0.9301, + "step": 2758 + }, + { + "epoch": 0.06, + "grad_norm": 2.1980705155537166, + "learning_rate": 1.9935871068859463e-05, + "loss": 1.1542, + "step": 2759 + }, + { + "epoch": 0.07, + "grad_norm": 2.2302260928678876, + "learning_rate": 1.9935784763434894e-05, + "loss": 1.193, + "step": 2760 + }, + { + "epoch": 0.07, + "grad_norm": 2.3516891348760103, + "learning_rate": 1.993569840016104e-05, + "loss": 0.8678, + "step": 2761 + }, + { + "epoch": 0.07, + "grad_norm": 2.234365425827115, + "learning_rate": 1.9935611979038403e-05, + "loss": 1.0082, + "step": 2762 + }, + { + "epoch": 0.07, + "grad_norm": 1.8690001063457573, + "learning_rate": 1.993552550006749e-05, + "loss": 0.9422, + "step": 2763 + }, + { + "epoch": 0.07, + "grad_norm": 2.180804932153207, + "learning_rate": 1.99354389632488e-05, + "loss": 1.0826, + "step": 2764 + }, + { + "epoch": 0.07, + "grad_norm": 2.118257390930115, + "learning_rate": 1.9935352368582837e-05, + "loss": 1.0494, + "step": 2765 + }, + { + "epoch": 0.07, + "grad_norm": 2.191486472983558, + "learning_rate": 1.9935265716070113e-05, + "loss": 0.9909, + "step": 2766 + }, + { + "epoch": 0.07, + "grad_norm": 2.0811775441783285, + "learning_rate": 1.9935179005711124e-05, + "loss": 1.1719, + "step": 2767 + }, + { + "epoch": 0.07, + "grad_norm": 1.9388685119655509, + "learning_rate": 1.993509223750637e-05, + "loss": 1.0775, + "step": 2768 + }, + { + "epoch": 0.07, + "grad_norm": 3.128229790957105, + "learning_rate": 1.9935005411456373e-05, + "loss": 1.0283, + "step": 2769 + }, + { + "epoch": 0.07, + "grad_norm": 1.9886900737005058, + "learning_rate": 1.9934918527561627e-05, + "loss": 1.1124, + "step": 2770 + }, + { + "epoch": 0.07, + "grad_norm": 1.1536021635158722, + "learning_rate": 1.993483158582264e-05, + "loss": 0.9901, + "step": 2771 + }, + { + "epoch": 0.07, + "grad_norm": 2.461129101416753, + "learning_rate": 1.9934744586239917e-05, + "loss": 1.1706, + "step": 2772 + }, + { + "epoch": 0.07, + "grad_norm": 2.148435943017742, + "learning_rate": 1.9934657528813965e-05, + "loss": 1.0214, + "step": 2773 + }, + { + "epoch": 0.07, + "grad_norm": 3.119404135073133, + "learning_rate": 1.993457041354529e-05, + "loss": 1.176, + "step": 2774 + }, + { + "epoch": 0.07, + "grad_norm": 3.9857558925969903, + "learning_rate": 1.9934483240434406e-05, + "loss": 1.0961, + "step": 2775 + }, + { + "epoch": 0.07, + "grad_norm": 2.4454610393563443, + "learning_rate": 1.9934396009481815e-05, + "loss": 1.1115, + "step": 2776 + }, + { + "epoch": 0.07, + "grad_norm": 2.33132307029222, + "learning_rate": 1.9934308720688024e-05, + "loss": 1.0664, + "step": 2777 + }, + { + "epoch": 0.07, + "grad_norm": 2.2354540148338597, + "learning_rate": 1.9934221374053538e-05, + "loss": 1.0114, + "step": 2778 + }, + { + "epoch": 0.07, + "grad_norm": 1.1671568341404632, + "learning_rate": 1.9934133969578875e-05, + "loss": 0.9908, + "step": 2779 + }, + { + "epoch": 0.07, + "grad_norm": 2.2652065799183387, + "learning_rate": 1.9934046507264537e-05, + "loss": 1.0488, + "step": 2780 + }, + { + "epoch": 0.07, + "grad_norm": 2.073220304983682, + "learning_rate": 1.9933958987111035e-05, + "loss": 0.9394, + "step": 2781 + }, + { + "epoch": 0.07, + "grad_norm": 1.9467762244321554, + "learning_rate": 1.993387140911888e-05, + "loss": 1.0451, + "step": 2782 + }, + { + "epoch": 0.07, + "grad_norm": 2.446761858626013, + "learning_rate": 1.993378377328858e-05, + "loss": 1.0059, + "step": 2783 + }, + { + "epoch": 0.07, + "grad_norm": 2.1071536998993516, + "learning_rate": 1.9933696079620647e-05, + "loss": 1.1366, + "step": 2784 + }, + { + "epoch": 0.07, + "grad_norm": 2.3534394801453242, + "learning_rate": 1.993360832811559e-05, + "loss": 1.0601, + "step": 2785 + }, + { + "epoch": 0.07, + "grad_norm": 2.455954483104016, + "learning_rate": 1.993352051877392e-05, + "loss": 1.1275, + "step": 2786 + }, + { + "epoch": 0.07, + "grad_norm": 2.044185663254151, + "learning_rate": 1.9933432651596146e-05, + "loss": 1.0447, + "step": 2787 + }, + { + "epoch": 0.07, + "grad_norm": 2.1707997507718146, + "learning_rate": 1.9933344726582787e-05, + "loss": 1.0879, + "step": 2788 + }, + { + "epoch": 0.07, + "grad_norm": 2.233654729647683, + "learning_rate": 1.9933256743734346e-05, + "loss": 1.0346, + "step": 2789 + }, + { + "epoch": 0.07, + "grad_norm": 2.534614511718709, + "learning_rate": 1.9933168703051343e-05, + "loss": 1.092, + "step": 2790 + }, + { + "epoch": 0.07, + "grad_norm": 3.283767121941717, + "learning_rate": 1.9933080604534285e-05, + "loss": 1.0067, + "step": 2791 + }, + { + "epoch": 0.07, + "grad_norm": 2.221509812326623, + "learning_rate": 1.9932992448183688e-05, + "loss": 1.0448, + "step": 2792 + }, + { + "epoch": 0.07, + "grad_norm": 2.529024131072562, + "learning_rate": 1.9932904234000063e-05, + "loss": 1.1322, + "step": 2793 + }, + { + "epoch": 0.07, + "grad_norm": 2.4144980341913396, + "learning_rate": 1.9932815961983925e-05, + "loss": 1.0089, + "step": 2794 + }, + { + "epoch": 0.07, + "grad_norm": 1.1680148913565196, + "learning_rate": 1.993272763213579e-05, + "loss": 0.9693, + "step": 2795 + }, + { + "epoch": 0.07, + "grad_norm": 2.656927543367129, + "learning_rate": 1.9932639244456165e-05, + "loss": 0.9529, + "step": 2796 + }, + { + "epoch": 0.07, + "grad_norm": 2.047317589550877, + "learning_rate": 1.9932550798945575e-05, + "loss": 1.0747, + "step": 2797 + }, + { + "epoch": 0.07, + "grad_norm": 2.052112019620463, + "learning_rate": 1.9932462295604526e-05, + "loss": 0.9551, + "step": 2798 + }, + { + "epoch": 0.07, + "grad_norm": 2.2646207038724206, + "learning_rate": 1.993237373443354e-05, + "loss": 1.026, + "step": 2799 + }, + { + "epoch": 0.07, + "grad_norm": 2.28962532618901, + "learning_rate": 1.9932285115433126e-05, + "loss": 1.136, + "step": 2800 + }, + { + "epoch": 0.07, + "grad_norm": 1.98525098601256, + "learning_rate": 1.9932196438603805e-05, + "loss": 1.0452, + "step": 2801 + }, + { + "epoch": 0.07, + "grad_norm": 2.5810030523570076, + "learning_rate": 1.993210770394609e-05, + "loss": 1.2151, + "step": 2802 + }, + { + "epoch": 0.07, + "grad_norm": 2.2691602314389128, + "learning_rate": 1.9932018911460503e-05, + "loss": 1.0067, + "step": 2803 + }, + { + "epoch": 0.07, + "grad_norm": 2.095447725254439, + "learning_rate": 1.9931930061147556e-05, + "loss": 1.0894, + "step": 2804 + }, + { + "epoch": 0.07, + "grad_norm": 2.1800658553926, + "learning_rate": 1.993184115300777e-05, + "loss": 1.2217, + "step": 2805 + }, + { + "epoch": 0.07, + "grad_norm": 2.6707526025935726, + "learning_rate": 1.993175218704166e-05, + "loss": 1.1243, + "step": 2806 + }, + { + "epoch": 0.07, + "grad_norm": 2.206505388090541, + "learning_rate": 1.9931663163249744e-05, + "loss": 0.9998, + "step": 2807 + }, + { + "epoch": 0.07, + "grad_norm": 1.1068576404711576, + "learning_rate": 1.993157408163254e-05, + "loss": 0.9638, + "step": 2808 + }, + { + "epoch": 0.07, + "grad_norm": 2.0682241001145543, + "learning_rate": 1.9931484942190567e-05, + "loss": 0.9879, + "step": 2809 + }, + { + "epoch": 0.07, + "grad_norm": 2.5050919970601644, + "learning_rate": 1.9931395744924345e-05, + "loss": 1.1303, + "step": 2810 + }, + { + "epoch": 0.07, + "grad_norm": 1.9524666051388553, + "learning_rate": 1.993130648983439e-05, + "loss": 1.1031, + "step": 2811 + }, + { + "epoch": 0.07, + "grad_norm": 2.150097150356838, + "learning_rate": 1.9931217176921228e-05, + "loss": 1.1777, + "step": 2812 + }, + { + "epoch": 0.07, + "grad_norm": 1.1685694844850278, + "learning_rate": 1.9931127806185376e-05, + "loss": 0.9252, + "step": 2813 + }, + { + "epoch": 0.07, + "grad_norm": 1.1002505101420055, + "learning_rate": 1.993103837762735e-05, + "loss": 0.9735, + "step": 2814 + }, + { + "epoch": 0.07, + "grad_norm": 2.1255873325231684, + "learning_rate": 1.9930948891247678e-05, + "loss": 0.8902, + "step": 2815 + }, + { + "epoch": 0.07, + "grad_norm": 2.0717404828870816, + "learning_rate": 1.9930859347046875e-05, + "loss": 1.1291, + "step": 2816 + }, + { + "epoch": 0.07, + "grad_norm": 2.2116171567058736, + "learning_rate": 1.9930769745025468e-05, + "loss": 1.0466, + "step": 2817 + }, + { + "epoch": 0.07, + "grad_norm": 2.535097926964853, + "learning_rate": 1.993068008518397e-05, + "loss": 1.2003, + "step": 2818 + }, + { + "epoch": 0.07, + "grad_norm": 2.0755528829720387, + "learning_rate": 1.993059036752291e-05, + "loss": 1.0327, + "step": 2819 + }, + { + "epoch": 0.07, + "grad_norm": 2.1780721088213917, + "learning_rate": 1.9930500592042815e-05, + "loss": 1.1221, + "step": 2820 + }, + { + "epoch": 0.07, + "grad_norm": 2.2946781174072632, + "learning_rate": 1.9930410758744195e-05, + "loss": 1.0289, + "step": 2821 + }, + { + "epoch": 0.07, + "grad_norm": 1.9748174173840665, + "learning_rate": 1.9930320867627584e-05, + "loss": 1.0303, + "step": 2822 + }, + { + "epoch": 0.07, + "grad_norm": 2.1869582668763945, + "learning_rate": 1.99302309186935e-05, + "loss": 1.1139, + "step": 2823 + }, + { + "epoch": 0.07, + "grad_norm": 2.3514824131622514, + "learning_rate": 1.9930140911942465e-05, + "loss": 1.075, + "step": 2824 + }, + { + "epoch": 0.07, + "grad_norm": 1.8620405948185674, + "learning_rate": 1.993005084737501e-05, + "loss": 0.9707, + "step": 2825 + }, + { + "epoch": 0.07, + "grad_norm": 2.058625766945818, + "learning_rate": 1.9929960724991652e-05, + "loss": 1.0526, + "step": 2826 + }, + { + "epoch": 0.07, + "grad_norm": 1.149091783164174, + "learning_rate": 1.9929870544792918e-05, + "loss": 1.0024, + "step": 2827 + }, + { + "epoch": 0.07, + "grad_norm": 1.2214554316708974, + "learning_rate": 1.9929780306779337e-05, + "loss": 0.9709, + "step": 2828 + }, + { + "epoch": 0.07, + "grad_norm": 2.021827598588041, + "learning_rate": 1.9929690010951427e-05, + "loss": 0.9667, + "step": 2829 + }, + { + "epoch": 0.07, + "grad_norm": 2.4874930730732543, + "learning_rate": 1.992959965730972e-05, + "loss": 1.0713, + "step": 2830 + }, + { + "epoch": 0.07, + "grad_norm": 1.8569148388962962, + "learning_rate": 1.9929509245854743e-05, + "loss": 1.022, + "step": 2831 + }, + { + "epoch": 0.07, + "grad_norm": 2.028803618283833, + "learning_rate": 1.9929418776587018e-05, + "loss": 1.1004, + "step": 2832 + }, + { + "epoch": 0.07, + "grad_norm": 2.2215681572052084, + "learning_rate": 1.9929328249507072e-05, + "loss": 1.1376, + "step": 2833 + }, + { + "epoch": 0.07, + "grad_norm": 2.0540285350124816, + "learning_rate": 1.9929237664615434e-05, + "loss": 0.9662, + "step": 2834 + }, + { + "epoch": 0.07, + "grad_norm": 1.9500357036529554, + "learning_rate": 1.992914702191263e-05, + "loss": 1.0334, + "step": 2835 + }, + { + "epoch": 0.07, + "grad_norm": 2.5310233527574617, + "learning_rate": 1.9929056321399193e-05, + "loss": 1.0788, + "step": 2836 + }, + { + "epoch": 0.07, + "grad_norm": 2.438157190258675, + "learning_rate": 1.992896556307564e-05, + "loss": 1.0401, + "step": 2837 + }, + { + "epoch": 0.07, + "grad_norm": 2.167507887411403, + "learning_rate": 1.992887474694251e-05, + "loss": 1.0007, + "step": 2838 + }, + { + "epoch": 0.07, + "grad_norm": 1.9394320577349287, + "learning_rate": 1.992878387300033e-05, + "loss": 1.0198, + "step": 2839 + }, + { + "epoch": 0.07, + "grad_norm": 2.308945052713642, + "learning_rate": 1.992869294124962e-05, + "loss": 1.1611, + "step": 2840 + }, + { + "epoch": 0.07, + "grad_norm": 2.230871441251919, + "learning_rate": 1.9928601951690922e-05, + "loss": 1.1879, + "step": 2841 + }, + { + "epoch": 0.07, + "grad_norm": 2.3354717296821046, + "learning_rate": 1.992851090432476e-05, + "loss": 1.0503, + "step": 2842 + }, + { + "epoch": 0.07, + "grad_norm": 2.8792052780899895, + "learning_rate": 1.9928419799151658e-05, + "loss": 1.0074, + "step": 2843 + }, + { + "epoch": 0.07, + "grad_norm": 1.1482901813208664, + "learning_rate": 1.992832863617216e-05, + "loss": 0.962, + "step": 2844 + }, + { + "epoch": 0.07, + "grad_norm": 1.9971044115452457, + "learning_rate": 1.9928237415386785e-05, + "loss": 1.002, + "step": 2845 + }, + { + "epoch": 0.07, + "grad_norm": 2.025369473115137, + "learning_rate": 1.992814613679607e-05, + "loss": 1.1937, + "step": 2846 + }, + { + "epoch": 0.07, + "grad_norm": 2.172991527164177, + "learning_rate": 1.992805480040055e-05, + "loss": 1.041, + "step": 2847 + }, + { + "epoch": 0.07, + "grad_norm": 2.3122037419110755, + "learning_rate": 1.9927963406200743e-05, + "loss": 1.2195, + "step": 2848 + }, + { + "epoch": 0.07, + "grad_norm": 2.0521405764928695, + "learning_rate": 1.9927871954197194e-05, + "loss": 1.064, + "step": 2849 + }, + { + "epoch": 0.07, + "grad_norm": 3.523367083280926, + "learning_rate": 1.9927780444390433e-05, + "loss": 1.0478, + "step": 2850 + }, + { + "epoch": 0.07, + "grad_norm": 2.1818517421209407, + "learning_rate": 1.9927688876780988e-05, + "loss": 1.0654, + "step": 2851 + }, + { + "epoch": 0.07, + "grad_norm": 2.4841287611968905, + "learning_rate": 1.9927597251369398e-05, + "loss": 1.0471, + "step": 2852 + }, + { + "epoch": 0.07, + "grad_norm": 2.089763221125044, + "learning_rate": 1.992750556815619e-05, + "loss": 0.9923, + "step": 2853 + }, + { + "epoch": 0.07, + "grad_norm": 2.2796220991747354, + "learning_rate": 1.9927413827141904e-05, + "loss": 1.0095, + "step": 2854 + }, + { + "epoch": 0.07, + "grad_norm": 1.9569665585583578, + "learning_rate": 1.9927322028327074e-05, + "loss": 1.0647, + "step": 2855 + }, + { + "epoch": 0.07, + "grad_norm": 2.2287497273157535, + "learning_rate": 1.992723017171223e-05, + "loss": 0.9931, + "step": 2856 + }, + { + "epoch": 0.07, + "grad_norm": 2.5351765210086254, + "learning_rate": 1.9927138257297908e-05, + "loss": 1.1397, + "step": 2857 + }, + { + "epoch": 0.07, + "grad_norm": 2.147642263383063, + "learning_rate": 1.9927046285084642e-05, + "loss": 1.0988, + "step": 2858 + }, + { + "epoch": 0.07, + "grad_norm": 1.9873951338353761, + "learning_rate": 1.9926954255072976e-05, + "loss": 1.0429, + "step": 2859 + }, + { + "epoch": 0.07, + "grad_norm": 2.4330583325144968, + "learning_rate": 1.9926862167263437e-05, + "loss": 1.0362, + "step": 2860 + }, + { + "epoch": 0.07, + "grad_norm": 2.257697008715818, + "learning_rate": 1.9926770021656564e-05, + "loss": 1.073, + "step": 2861 + }, + { + "epoch": 0.07, + "grad_norm": 2.409796199555926, + "learning_rate": 1.992667781825289e-05, + "loss": 1.0413, + "step": 2862 + }, + { + "epoch": 0.07, + "grad_norm": 2.148657288471305, + "learning_rate": 1.9926585557052958e-05, + "loss": 0.9873, + "step": 2863 + }, + { + "epoch": 0.07, + "grad_norm": 2.001693216757955, + "learning_rate": 1.99264932380573e-05, + "loss": 1.1725, + "step": 2864 + }, + { + "epoch": 0.07, + "grad_norm": 2.1575838213442164, + "learning_rate": 1.9926400861266458e-05, + "loss": 1.1107, + "step": 2865 + }, + { + "epoch": 0.07, + "grad_norm": 2.0248163128871712, + "learning_rate": 1.9926308426680966e-05, + "loss": 1.1052, + "step": 2866 + }, + { + "epoch": 0.07, + "grad_norm": 2.085800258581078, + "learning_rate": 1.9926215934301366e-05, + "loss": 1.0337, + "step": 2867 + }, + { + "epoch": 0.07, + "grad_norm": 2.1930778794948904, + "learning_rate": 1.992612338412819e-05, + "loss": 1.169, + "step": 2868 + }, + { + "epoch": 0.07, + "grad_norm": 1.9477282717508049, + "learning_rate": 1.9926030776161986e-05, + "loss": 1.0778, + "step": 2869 + }, + { + "epoch": 0.07, + "grad_norm": 2.4098051743254065, + "learning_rate": 1.9925938110403286e-05, + "loss": 1.0707, + "step": 2870 + }, + { + "epoch": 0.07, + "grad_norm": 2.0146465638966062, + "learning_rate": 1.992584538685263e-05, + "loss": 1.1046, + "step": 2871 + }, + { + "epoch": 0.07, + "grad_norm": 1.1349531802905397, + "learning_rate": 1.9925752605510566e-05, + "loss": 1.0683, + "step": 2872 + }, + { + "epoch": 0.07, + "grad_norm": 2.550827605334162, + "learning_rate": 1.9925659766377622e-05, + "loss": 1.0365, + "step": 2873 + }, + { + "epoch": 0.07, + "grad_norm": 2.5392851580138465, + "learning_rate": 1.9925566869454345e-05, + "loss": 0.9811, + "step": 2874 + }, + { + "epoch": 0.07, + "grad_norm": 2.21000408441496, + "learning_rate": 1.9925473914741275e-05, + "loss": 1.2094, + "step": 2875 + }, + { + "epoch": 0.07, + "grad_norm": 1.9168875610339107, + "learning_rate": 1.9925380902238956e-05, + "loss": 1.0013, + "step": 2876 + }, + { + "epoch": 0.07, + "grad_norm": 2.22589474607425, + "learning_rate": 1.9925287831947922e-05, + "loss": 1.0917, + "step": 2877 + }, + { + "epoch": 0.07, + "grad_norm": 2.2873192655434136, + "learning_rate": 1.9925194703868727e-05, + "loss": 1.0582, + "step": 2878 + }, + { + "epoch": 0.07, + "grad_norm": 2.2575573981270534, + "learning_rate": 1.99251015180019e-05, + "loss": 0.8546, + "step": 2879 + }, + { + "epoch": 0.07, + "grad_norm": 2.1475541241282596, + "learning_rate": 1.9925008274347993e-05, + "loss": 0.9161, + "step": 2880 + }, + { + "epoch": 0.07, + "grad_norm": 2.536427318911218, + "learning_rate": 1.9924914972907547e-05, + "loss": 1.0484, + "step": 2881 + }, + { + "epoch": 0.07, + "grad_norm": 2.0155529295341963, + "learning_rate": 1.9924821613681103e-05, + "loss": 1.13, + "step": 2882 + }, + { + "epoch": 0.07, + "grad_norm": 2.107106517975659, + "learning_rate": 1.9924728196669205e-05, + "loss": 1.0174, + "step": 2883 + }, + { + "epoch": 0.07, + "grad_norm": 1.7952480547086118, + "learning_rate": 1.9924634721872396e-05, + "loss": 1.062, + "step": 2884 + }, + { + "epoch": 0.07, + "grad_norm": 2.195383667428661, + "learning_rate": 1.992454118929122e-05, + "loss": 1.0442, + "step": 2885 + }, + { + "epoch": 0.07, + "grad_norm": 2.1848468076906467, + "learning_rate": 1.9924447598926227e-05, + "loss": 1.0378, + "step": 2886 + }, + { + "epoch": 0.07, + "grad_norm": 2.1626844418859767, + "learning_rate": 1.9924353950777956e-05, + "loss": 1.2622, + "step": 2887 + }, + { + "epoch": 0.07, + "grad_norm": 2.069818959923516, + "learning_rate": 1.9924260244846954e-05, + "loss": 1.1693, + "step": 2888 + }, + { + "epoch": 0.07, + "grad_norm": 2.178480823954348, + "learning_rate": 1.9924166481133767e-05, + "loss": 1.1162, + "step": 2889 + }, + { + "epoch": 0.07, + "grad_norm": 2.278187817307125, + "learning_rate": 1.9924072659638944e-05, + "loss": 1.0737, + "step": 2890 + }, + { + "epoch": 0.07, + "grad_norm": 2.106800054891054, + "learning_rate": 1.9923978780363025e-05, + "loss": 0.8771, + "step": 2891 + }, + { + "epoch": 0.07, + "grad_norm": 2.362436292072532, + "learning_rate": 1.992388484330656e-05, + "loss": 1.0786, + "step": 2892 + }, + { + "epoch": 0.07, + "grad_norm": 2.0946234185271733, + "learning_rate": 1.9923790848470098e-05, + "loss": 1.074, + "step": 2893 + }, + { + "epoch": 0.07, + "grad_norm": 2.9651842261063686, + "learning_rate": 1.992369679585418e-05, + "loss": 1.0034, + "step": 2894 + }, + { + "epoch": 0.07, + "grad_norm": 2.1656424470306375, + "learning_rate": 1.992360268545936e-05, + "loss": 1.1126, + "step": 2895 + }, + { + "epoch": 0.07, + "grad_norm": 2.1726700621247845, + "learning_rate": 1.992350851728618e-05, + "loss": 1.0834, + "step": 2896 + }, + { + "epoch": 0.07, + "grad_norm": 2.3877621402437654, + "learning_rate": 1.9923414291335198e-05, + "loss": 1.0997, + "step": 2897 + }, + { + "epoch": 0.07, + "grad_norm": 2.4818155712612344, + "learning_rate": 1.992332000760695e-05, + "loss": 0.9769, + "step": 2898 + }, + { + "epoch": 0.07, + "grad_norm": 1.9216291203210478, + "learning_rate": 1.9923225666101994e-05, + "loss": 1.1639, + "step": 2899 + }, + { + "epoch": 0.07, + "grad_norm": 2.06927449176558, + "learning_rate": 1.9923131266820878e-05, + "loss": 1.0296, + "step": 2900 + }, + { + "epoch": 0.07, + "grad_norm": 2.0670826978573924, + "learning_rate": 1.9923036809764145e-05, + "loss": 1.0739, + "step": 2901 + }, + { + "epoch": 0.07, + "grad_norm": 2.0124893311544843, + "learning_rate": 1.9922942294932356e-05, + "loss": 1.1117, + "step": 2902 + }, + { + "epoch": 0.07, + "grad_norm": 2.5491456576668603, + "learning_rate": 1.9922847722326052e-05, + "loss": 1.2453, + "step": 2903 + }, + { + "epoch": 0.07, + "grad_norm": 2.190932767655254, + "learning_rate": 1.992275309194579e-05, + "loss": 1.178, + "step": 2904 + }, + { + "epoch": 0.07, + "grad_norm": 2.038747418863809, + "learning_rate": 1.9922658403792115e-05, + "loss": 1.0725, + "step": 2905 + }, + { + "epoch": 0.07, + "grad_norm": 2.004816978393798, + "learning_rate": 1.9922563657865584e-05, + "loss": 1.1174, + "step": 2906 + }, + { + "epoch": 0.07, + "grad_norm": 2.778432012920504, + "learning_rate": 1.9922468854166744e-05, + "loss": 1.0328, + "step": 2907 + }, + { + "epoch": 0.07, + "grad_norm": 1.9260946067464422, + "learning_rate": 1.9922373992696146e-05, + "loss": 1.0932, + "step": 2908 + }, + { + "epoch": 0.07, + "grad_norm": 2.0592410029836667, + "learning_rate": 1.992227907345435e-05, + "loss": 1.1144, + "step": 2909 + }, + { + "epoch": 0.07, + "grad_norm": 2.037409418000821, + "learning_rate": 1.9922184096441904e-05, + "loss": 1.0415, + "step": 2910 + }, + { + "epoch": 0.07, + "grad_norm": 1.1620158668772311, + "learning_rate": 1.992208906165936e-05, + "loss": 0.9636, + "step": 2911 + }, + { + "epoch": 0.07, + "grad_norm": 1.961924750779347, + "learning_rate": 1.992199396910727e-05, + "loss": 1.1029, + "step": 2912 + }, + { + "epoch": 0.07, + "grad_norm": 2.3993063091471014, + "learning_rate": 1.992189881878619e-05, + "loss": 1.0356, + "step": 2913 + }, + { + "epoch": 0.07, + "grad_norm": 2.464381330440199, + "learning_rate": 1.9921803610696675e-05, + "loss": 1.2032, + "step": 2914 + }, + { + "epoch": 0.07, + "grad_norm": 6.227304393872376, + "learning_rate": 1.992170834483928e-05, + "loss": 0.9493, + "step": 2915 + }, + { + "epoch": 0.07, + "grad_norm": 2.432354312553881, + "learning_rate": 1.9921613021214555e-05, + "loss": 1.0979, + "step": 2916 + }, + { + "epoch": 0.07, + "grad_norm": 2.365597638939805, + "learning_rate": 1.992151763982306e-05, + "loss": 1.0171, + "step": 2917 + }, + { + "epoch": 0.07, + "grad_norm": 1.968720281301905, + "learning_rate": 1.9921422200665347e-05, + "loss": 1.1078, + "step": 2918 + }, + { + "epoch": 0.07, + "grad_norm": 2.0099336654220368, + "learning_rate": 1.9921326703741976e-05, + "loss": 1.0907, + "step": 2919 + }, + { + "epoch": 0.07, + "grad_norm": 2.1864327046764656, + "learning_rate": 1.9921231149053495e-05, + "loss": 0.9813, + "step": 2920 + }, + { + "epoch": 0.07, + "grad_norm": 2.00609389570097, + "learning_rate": 1.9921135536600465e-05, + "loss": 1.1028, + "step": 2921 + }, + { + "epoch": 0.07, + "grad_norm": 1.9760431275915755, + "learning_rate": 1.9921039866383447e-05, + "loss": 0.9459, + "step": 2922 + }, + { + "epoch": 0.07, + "grad_norm": 1.9621892468354611, + "learning_rate": 1.9920944138402993e-05, + "loss": 0.8573, + "step": 2923 + }, + { + "epoch": 0.07, + "grad_norm": 2.3824297465646103, + "learning_rate": 1.9920848352659662e-05, + "loss": 1.0584, + "step": 2924 + }, + { + "epoch": 0.07, + "grad_norm": 1.9759931276695453, + "learning_rate": 1.9920752509154008e-05, + "loss": 1.0524, + "step": 2925 + }, + { + "epoch": 0.07, + "grad_norm": 2.2535945358204246, + "learning_rate": 1.992065660788659e-05, + "loss": 0.9617, + "step": 2926 + }, + { + "epoch": 0.07, + "grad_norm": 2.2875127742005477, + "learning_rate": 1.9920560648857976e-05, + "loss": 1.0211, + "step": 2927 + }, + { + "epoch": 0.07, + "grad_norm": 1.9834524707956438, + "learning_rate": 1.9920464632068713e-05, + "loss": 1.0361, + "step": 2928 + }, + { + "epoch": 0.07, + "grad_norm": 2.523749534643557, + "learning_rate": 1.9920368557519362e-05, + "loss": 1.032, + "step": 2929 + }, + { + "epoch": 0.07, + "grad_norm": 2.107600219134359, + "learning_rate": 1.992027242521049e-05, + "loss": 0.951, + "step": 2930 + }, + { + "epoch": 0.07, + "grad_norm": 2.098141872412311, + "learning_rate": 1.9920176235142645e-05, + "loss": 1.0641, + "step": 2931 + }, + { + "epoch": 0.07, + "grad_norm": 4.970174462706873, + "learning_rate": 1.9920079987316398e-05, + "loss": 1.0192, + "step": 2932 + }, + { + "epoch": 0.07, + "grad_norm": 2.1319606718739648, + "learning_rate": 1.9919983681732302e-05, + "loss": 1.2713, + "step": 2933 + }, + { + "epoch": 0.07, + "grad_norm": 2.2288297567410384, + "learning_rate": 1.9919887318390923e-05, + "loss": 1.1747, + "step": 2934 + }, + { + "epoch": 0.07, + "grad_norm": 2.622156434136596, + "learning_rate": 1.9919790897292817e-05, + "loss": 1.0603, + "step": 2935 + }, + { + "epoch": 0.07, + "grad_norm": 1.1303905118753301, + "learning_rate": 1.9919694418438547e-05, + "loss": 0.8995, + "step": 2936 + }, + { + "epoch": 0.07, + "grad_norm": 1.983015999415005, + "learning_rate": 1.9919597881828678e-05, + "loss": 1.1068, + "step": 2937 + }, + { + "epoch": 0.07, + "grad_norm": 2.0757457956700422, + "learning_rate": 1.991950128746377e-05, + "loss": 1.0594, + "step": 2938 + }, + { + "epoch": 0.07, + "grad_norm": 2.077858778968792, + "learning_rate": 1.9919404635344383e-05, + "loss": 1.0035, + "step": 2939 + }, + { + "epoch": 0.07, + "grad_norm": 3.266015532090608, + "learning_rate": 1.991930792547108e-05, + "loss": 1.1167, + "step": 2940 + }, + { + "epoch": 0.07, + "grad_norm": 2.061027034412749, + "learning_rate": 1.991921115784443e-05, + "loss": 1.0117, + "step": 2941 + }, + { + "epoch": 0.07, + "grad_norm": 1.060908317994129, + "learning_rate": 1.9919114332464994e-05, + "loss": 0.9922, + "step": 2942 + }, + { + "epoch": 0.07, + "grad_norm": 2.1786761791922533, + "learning_rate": 1.991901744933333e-05, + "loss": 1.0373, + "step": 2943 + }, + { + "epoch": 0.07, + "grad_norm": 1.9448879637580356, + "learning_rate": 1.9918920508450003e-05, + "loss": 1.0676, + "step": 2944 + }, + { + "epoch": 0.07, + "grad_norm": 2.0306457746463202, + "learning_rate": 1.9918823509815585e-05, + "loss": 1.1211, + "step": 2945 + }, + { + "epoch": 0.07, + "grad_norm": 2.196103581869107, + "learning_rate": 1.9918726453430635e-05, + "loss": 1.0277, + "step": 2946 + }, + { + "epoch": 0.07, + "grad_norm": 1.812799312509761, + "learning_rate": 1.9918629339295722e-05, + "loss": 1.1071, + "step": 2947 + }, + { + "epoch": 0.07, + "grad_norm": 2.0138057029257403, + "learning_rate": 1.9918532167411407e-05, + "loss": 0.9514, + "step": 2948 + }, + { + "epoch": 0.07, + "grad_norm": 2.2962804877619223, + "learning_rate": 1.991843493777826e-05, + "loss": 1.1293, + "step": 2949 + }, + { + "epoch": 0.07, + "grad_norm": 2.0834874286386493, + "learning_rate": 1.991833765039684e-05, + "loss": 1.089, + "step": 2950 + }, + { + "epoch": 0.07, + "grad_norm": 2.1467505312566955, + "learning_rate": 1.991824030526772e-05, + "loss": 0.9986, + "step": 2951 + }, + { + "epoch": 0.07, + "grad_norm": 2.339197268020623, + "learning_rate": 1.9918142902391463e-05, + "loss": 1.1693, + "step": 2952 + }, + { + "epoch": 0.07, + "grad_norm": 3.140825809412629, + "learning_rate": 1.9918045441768642e-05, + "loss": 1.0261, + "step": 2953 + }, + { + "epoch": 0.07, + "grad_norm": 2.1974002203513656, + "learning_rate": 1.9917947923399818e-05, + "loss": 1.1462, + "step": 2954 + }, + { + "epoch": 0.07, + "grad_norm": 2.4757485688039615, + "learning_rate": 1.9917850347285563e-05, + "loss": 1.1298, + "step": 2955 + }, + { + "epoch": 0.07, + "grad_norm": 2.2880130670029564, + "learning_rate": 1.9917752713426442e-05, + "loss": 1.0337, + "step": 2956 + }, + { + "epoch": 0.07, + "grad_norm": 1.1985145520065033, + "learning_rate": 1.9917655021823022e-05, + "loss": 1.0345, + "step": 2957 + }, + { + "epoch": 0.07, + "grad_norm": 2.1719676752587764, + "learning_rate": 1.991755727247588e-05, + "loss": 1.0724, + "step": 2958 + }, + { + "epoch": 0.07, + "grad_norm": 2.4512594851380713, + "learning_rate": 1.9917459465385578e-05, + "loss": 0.9099, + "step": 2959 + }, + { + "epoch": 0.07, + "grad_norm": 1.1745173605522263, + "learning_rate": 1.9917361600552685e-05, + "loss": 0.9453, + "step": 2960 + }, + { + "epoch": 0.07, + "grad_norm": 2.2092375415242858, + "learning_rate": 1.9917263677977773e-05, + "loss": 1.1841, + "step": 2961 + }, + { + "epoch": 0.07, + "grad_norm": 1.9492359903582237, + "learning_rate": 1.9917165697661415e-05, + "loss": 1.0483, + "step": 2962 + }, + { + "epoch": 0.07, + "grad_norm": 2.259287763859694, + "learning_rate": 1.9917067659604176e-05, + "loss": 1.2208, + "step": 2963 + }, + { + "epoch": 0.07, + "grad_norm": 1.9260530755104386, + "learning_rate": 1.991696956380663e-05, + "loss": 1.0717, + "step": 2964 + }, + { + "epoch": 0.07, + "grad_norm": 1.861717457596766, + "learning_rate": 1.991687141026935e-05, + "loss": 1.207, + "step": 2965 + }, + { + "epoch": 0.07, + "grad_norm": 2.2872261682101254, + "learning_rate": 1.99167731989929e-05, + "loss": 1.1045, + "step": 2966 + }, + { + "epoch": 0.07, + "grad_norm": 2.214681135095699, + "learning_rate": 1.991667492997786e-05, + "loss": 1.0869, + "step": 2967 + }, + { + "epoch": 0.07, + "grad_norm": 2.2563081621066425, + "learning_rate": 1.99165766032248e-05, + "loss": 1.127, + "step": 2968 + }, + { + "epoch": 0.07, + "grad_norm": 2.1615905621692306, + "learning_rate": 1.991647821873429e-05, + "loss": 1.0597, + "step": 2969 + }, + { + "epoch": 0.07, + "grad_norm": 2.200531379250738, + "learning_rate": 1.9916379776506908e-05, + "loss": 1.0819, + "step": 2970 + }, + { + "epoch": 0.07, + "grad_norm": 2.0175500823488126, + "learning_rate": 1.9916281276543215e-05, + "loss": 1.0361, + "step": 2971 + }, + { + "epoch": 0.07, + "grad_norm": 2.1986288690612135, + "learning_rate": 1.99161827188438e-05, + "loss": 1.1143, + "step": 2972 + }, + { + "epoch": 0.07, + "grad_norm": 2.119702833613661, + "learning_rate": 1.9916084103409228e-05, + "loss": 1.0096, + "step": 2973 + }, + { + "epoch": 0.07, + "grad_norm": 2.0249088034056504, + "learning_rate": 1.9915985430240075e-05, + "loss": 1.0053, + "step": 2974 + }, + { + "epoch": 0.07, + "grad_norm": 1.9311917390192965, + "learning_rate": 1.9915886699336915e-05, + "loss": 1.0272, + "step": 2975 + }, + { + "epoch": 0.07, + "grad_norm": 2.2277962814365218, + "learning_rate": 1.9915787910700322e-05, + "loss": 1.131, + "step": 2976 + }, + { + "epoch": 0.07, + "grad_norm": 1.1500756702069095, + "learning_rate": 1.9915689064330875e-05, + "loss": 1.0212, + "step": 2977 + }, + { + "epoch": 0.07, + "grad_norm": 1.9490765963389511, + "learning_rate": 1.991559016022915e-05, + "loss": 1.0842, + "step": 2978 + }, + { + "epoch": 0.07, + "grad_norm": 2.0032916642840983, + "learning_rate": 1.9915491198395714e-05, + "loss": 1.0457, + "step": 2979 + }, + { + "epoch": 0.07, + "grad_norm": 2.3280618660242314, + "learning_rate": 1.991539217883115e-05, + "loss": 1.1465, + "step": 2980 + }, + { + "epoch": 0.07, + "grad_norm": 2.1284831575937226, + "learning_rate": 1.991529310153603e-05, + "loss": 1.1361, + "step": 2981 + }, + { + "epoch": 0.07, + "grad_norm": 2.200858866481066, + "learning_rate": 1.9915193966510945e-05, + "loss": 1.0707, + "step": 2982 + }, + { + "epoch": 0.07, + "grad_norm": 3.424154910961848, + "learning_rate": 1.9915094773756454e-05, + "loss": 1.0859, + "step": 2983 + }, + { + "epoch": 0.07, + "grad_norm": 2.1325656497809278, + "learning_rate": 1.9914995523273144e-05, + "loss": 1.1365, + "step": 2984 + }, + { + "epoch": 0.07, + "grad_norm": 2.110224301284423, + "learning_rate": 1.991489621506159e-05, + "loss": 1.1313, + "step": 2985 + }, + { + "epoch": 0.07, + "grad_norm": 2.170072613183587, + "learning_rate": 1.9914796849122374e-05, + "loss": 1.1018, + "step": 2986 + }, + { + "epoch": 0.07, + "grad_norm": 1.9826250557061749, + "learning_rate": 1.991469742545607e-05, + "loss": 1.1463, + "step": 2987 + }, + { + "epoch": 0.07, + "grad_norm": 2.199698457902296, + "learning_rate": 1.9914597944063257e-05, + "loss": 0.9859, + "step": 2988 + }, + { + "epoch": 0.07, + "grad_norm": 2.1930971410825215, + "learning_rate": 1.991449840494452e-05, + "loss": 1.0095, + "step": 2989 + }, + { + "epoch": 0.07, + "grad_norm": 2.5282811002132766, + "learning_rate": 1.991439880810043e-05, + "loss": 1.2793, + "step": 2990 + }, + { + "epoch": 0.07, + "grad_norm": 2.235107388123992, + "learning_rate": 1.991429915353157e-05, + "loss": 1.0072, + "step": 2991 + }, + { + "epoch": 0.07, + "grad_norm": 2.176280017576744, + "learning_rate": 1.9914199441238528e-05, + "loss": 1.0396, + "step": 2992 + }, + { + "epoch": 0.07, + "grad_norm": 1.9455711335724106, + "learning_rate": 1.9914099671221874e-05, + "loss": 1.081, + "step": 2993 + }, + { + "epoch": 0.07, + "grad_norm": 2.1015267401217823, + "learning_rate": 1.9913999843482195e-05, + "loss": 1.0634, + "step": 2994 + }, + { + "epoch": 0.07, + "grad_norm": 2.264247125893083, + "learning_rate": 1.9913899958020067e-05, + "loss": 1.1225, + "step": 2995 + }, + { + "epoch": 0.07, + "grad_norm": 2.456355587916114, + "learning_rate": 1.9913800014836076e-05, + "loss": 1.0054, + "step": 2996 + }, + { + "epoch": 0.07, + "grad_norm": 2.1719664783389065, + "learning_rate": 1.9913700013930804e-05, + "loss": 1.0932, + "step": 2997 + }, + { + "epoch": 0.07, + "grad_norm": 2.2259013896873223, + "learning_rate": 1.991359995530483e-05, + "loss": 1.1939, + "step": 2998 + }, + { + "epoch": 0.07, + "grad_norm": 2.455608503716985, + "learning_rate": 1.991349983895874e-05, + "loss": 1.1355, + "step": 2999 + }, + { + "epoch": 0.07, + "grad_norm": 2.322960045123313, + "learning_rate": 1.9913399664893113e-05, + "loss": 1.0207, + "step": 3000 + }, + { + "epoch": 0.07, + "grad_norm": 2.510822732646857, + "learning_rate": 1.991329943310854e-05, + "loss": 1.0395, + "step": 3001 + }, + { + "epoch": 0.07, + "grad_norm": 2.2040577496664437, + "learning_rate": 1.9913199143605595e-05, + "loss": 1.0919, + "step": 3002 + }, + { + "epoch": 0.07, + "grad_norm": 2.0767423159206704, + "learning_rate": 1.9913098796384865e-05, + "loss": 1.127, + "step": 3003 + }, + { + "epoch": 0.07, + "grad_norm": 2.1143805554112873, + "learning_rate": 1.991299839144694e-05, + "loss": 1.1354, + "step": 3004 + }, + { + "epoch": 0.07, + "grad_norm": 1.1279551026341983, + "learning_rate": 1.9912897928792395e-05, + "loss": 0.9972, + "step": 3005 + }, + { + "epoch": 0.07, + "grad_norm": 2.097056231447447, + "learning_rate": 1.9912797408421823e-05, + "loss": 1.1402, + "step": 3006 + }, + { + "epoch": 0.07, + "grad_norm": 2.0818902270646285, + "learning_rate": 1.9912696830335806e-05, + "loss": 1.0732, + "step": 3007 + }, + { + "epoch": 0.07, + "grad_norm": 2.243459097091603, + "learning_rate": 1.991259619453493e-05, + "loss": 1.1732, + "step": 3008 + }, + { + "epoch": 0.07, + "grad_norm": 2.03105318773447, + "learning_rate": 1.9912495501019777e-05, + "loss": 1.131, + "step": 3009 + }, + { + "epoch": 0.07, + "grad_norm": 2.2981909883400884, + "learning_rate": 1.9912394749790942e-05, + "loss": 1.0502, + "step": 3010 + }, + { + "epoch": 0.07, + "grad_norm": 2.1869537631576867, + "learning_rate": 1.9912293940849002e-05, + "loss": 1.1347, + "step": 3011 + }, + { + "epoch": 0.07, + "grad_norm": 2.5612662452028507, + "learning_rate": 1.991219307419455e-05, + "loss": 1.0748, + "step": 3012 + }, + { + "epoch": 0.07, + "grad_norm": 2.8371651328271152, + "learning_rate": 1.9912092149828175e-05, + "loss": 1.0168, + "step": 3013 + }, + { + "epoch": 0.07, + "grad_norm": 2.9641995135273254, + "learning_rate": 1.9911991167750458e-05, + "loss": 0.9664, + "step": 3014 + }, + { + "epoch": 0.07, + "grad_norm": 1.124629077820768, + "learning_rate": 1.9911890127961993e-05, + "loss": 0.909, + "step": 3015 + }, + { + "epoch": 0.07, + "grad_norm": 2.0095723797680654, + "learning_rate": 1.991178903046336e-05, + "loss": 1.1156, + "step": 3016 + }, + { + "epoch": 0.07, + "grad_norm": 2.137617019589321, + "learning_rate": 1.991168787525516e-05, + "loss": 0.9752, + "step": 3017 + }, + { + "epoch": 0.07, + "grad_norm": 1.958217824223083, + "learning_rate": 1.991158666233797e-05, + "loss": 0.9983, + "step": 3018 + }, + { + "epoch": 0.07, + "grad_norm": 2.028583778911836, + "learning_rate": 1.991148539171239e-05, + "loss": 1.1426, + "step": 3019 + }, + { + "epoch": 0.07, + "grad_norm": 1.1469631330680203, + "learning_rate": 1.9911384063378996e-05, + "loss": 1.0526, + "step": 3020 + }, + { + "epoch": 0.07, + "grad_norm": 2.16367544613418, + "learning_rate": 1.9911282677338395e-05, + "loss": 1.0467, + "step": 3021 + }, + { + "epoch": 0.07, + "grad_norm": 2.069557970132992, + "learning_rate": 1.9911181233591163e-05, + "loss": 0.9501, + "step": 3022 + }, + { + "epoch": 0.07, + "grad_norm": 2.071749368402756, + "learning_rate": 1.9911079732137897e-05, + "loss": 1.0898, + "step": 3023 + }, + { + "epoch": 0.07, + "grad_norm": 2.1726228874402826, + "learning_rate": 1.9910978172979187e-05, + "loss": 1.2189, + "step": 3024 + }, + { + "epoch": 0.07, + "grad_norm": 2.6429636841659048, + "learning_rate": 1.9910876556115626e-05, + "loss": 1.0644, + "step": 3025 + }, + { + "epoch": 0.07, + "grad_norm": 2.195210912880746, + "learning_rate": 1.9910774881547803e-05, + "loss": 1.0847, + "step": 3026 + }, + { + "epoch": 0.07, + "grad_norm": 2.5394660434620224, + "learning_rate": 1.9910673149276308e-05, + "loss": 1.1145, + "step": 3027 + }, + { + "epoch": 0.07, + "grad_norm": 2.1012984614262837, + "learning_rate": 1.991057135930174e-05, + "loss": 1.0377, + "step": 3028 + }, + { + "epoch": 0.07, + "grad_norm": 1.1607264438478067, + "learning_rate": 1.991046951162469e-05, + "loss": 0.975, + "step": 3029 + }, + { + "epoch": 0.07, + "grad_norm": 2.2135770579856002, + "learning_rate": 1.9910367606245744e-05, + "loss": 1.016, + "step": 3030 + }, + { + "epoch": 0.07, + "grad_norm": 2.7656718681325674, + "learning_rate": 1.9910265643165502e-05, + "loss": 1.186, + "step": 3031 + }, + { + "epoch": 0.07, + "grad_norm": 3.3018613764935196, + "learning_rate": 1.9910163622384556e-05, + "loss": 1.0079, + "step": 3032 + }, + { + "epoch": 0.07, + "grad_norm": 2.333195117754363, + "learning_rate": 1.9910061543903502e-05, + "loss": 1.0514, + "step": 3033 + }, + { + "epoch": 0.07, + "grad_norm": 2.6179136779377714, + "learning_rate": 1.9909959407722926e-05, + "loss": 1.0807, + "step": 3034 + }, + { + "epoch": 0.07, + "grad_norm": 2.3675232987344903, + "learning_rate": 1.990985721384343e-05, + "loss": 1.0298, + "step": 3035 + }, + { + "epoch": 0.07, + "grad_norm": 2.093034751200465, + "learning_rate": 1.9909754962265615e-05, + "loss": 1.1406, + "step": 3036 + }, + { + "epoch": 0.07, + "grad_norm": 2.0813148108621, + "learning_rate": 1.9909652652990065e-05, + "loss": 1.153, + "step": 3037 + }, + { + "epoch": 0.07, + "grad_norm": 1.1622964862645684, + "learning_rate": 1.990955028601738e-05, + "loss": 1.0706, + "step": 3038 + }, + { + "epoch": 0.07, + "grad_norm": 2.267040905947008, + "learning_rate": 1.9909447861348154e-05, + "loss": 1.1872, + "step": 3039 + }, + { + "epoch": 0.07, + "grad_norm": 2.147888717571085, + "learning_rate": 1.9909345378982987e-05, + "loss": 1.1706, + "step": 3040 + }, + { + "epoch": 0.07, + "grad_norm": 2.3016532479497767, + "learning_rate": 1.9909242838922472e-05, + "loss": 1.028, + "step": 3041 + }, + { + "epoch": 0.07, + "grad_norm": 1.9563243442510445, + "learning_rate": 1.9909140241167212e-05, + "loss": 1.1537, + "step": 3042 + }, + { + "epoch": 0.07, + "grad_norm": 2.3338126436603774, + "learning_rate": 1.99090375857178e-05, + "loss": 1.0024, + "step": 3043 + }, + { + "epoch": 0.07, + "grad_norm": 1.9813897250572121, + "learning_rate": 1.990893487257483e-05, + "loss": 1.0569, + "step": 3044 + }, + { + "epoch": 0.07, + "grad_norm": 2.553456835021127, + "learning_rate": 1.990883210173891e-05, + "loss": 0.9512, + "step": 3045 + }, + { + "epoch": 0.07, + "grad_norm": 2.228771441288357, + "learning_rate": 1.9908729273210627e-05, + "loss": 1.1298, + "step": 3046 + }, + { + "epoch": 0.07, + "grad_norm": 2.0844317473755485, + "learning_rate": 1.9908626386990586e-05, + "loss": 0.9637, + "step": 3047 + }, + { + "epoch": 0.07, + "grad_norm": 2.019626733691953, + "learning_rate": 1.9908523443079384e-05, + "loss": 1.1788, + "step": 3048 + }, + { + "epoch": 0.07, + "grad_norm": 1.0568530483734175, + "learning_rate": 1.9908420441477626e-05, + "loss": 0.9445, + "step": 3049 + }, + { + "epoch": 0.07, + "grad_norm": 2.0804606265089594, + "learning_rate": 1.9908317382185904e-05, + "loss": 1.0686, + "step": 3050 + }, + { + "epoch": 0.07, + "grad_norm": 2.1917865166336923, + "learning_rate": 1.9908214265204825e-05, + "loss": 0.949, + "step": 3051 + }, + { + "epoch": 0.07, + "grad_norm": 2.175056809044507, + "learning_rate": 1.990811109053498e-05, + "loss": 1.0653, + "step": 3052 + }, + { + "epoch": 0.07, + "grad_norm": 2.1110655970305916, + "learning_rate": 1.990800785817698e-05, + "loss": 1.1953, + "step": 3053 + }, + { + "epoch": 0.07, + "grad_norm": 2.482240264118477, + "learning_rate": 1.990790456813142e-05, + "loss": 0.9386, + "step": 3054 + }, + { + "epoch": 0.07, + "grad_norm": 2.0180089975507975, + "learning_rate": 1.99078012203989e-05, + "loss": 1.0761, + "step": 3055 + }, + { + "epoch": 0.07, + "grad_norm": 2.6450700017414106, + "learning_rate": 1.9907697814980025e-05, + "loss": 1.0683, + "step": 3056 + }, + { + "epoch": 0.07, + "grad_norm": 2.217695864543195, + "learning_rate": 1.9907594351875402e-05, + "loss": 1.0634, + "step": 3057 + }, + { + "epoch": 0.07, + "grad_norm": 2.114784352787273, + "learning_rate": 1.9907490831085626e-05, + "loss": 1.1837, + "step": 3058 + }, + { + "epoch": 0.07, + "grad_norm": 2.271849008180131, + "learning_rate": 1.99073872526113e-05, + "loss": 1.0912, + "step": 3059 + }, + { + "epoch": 0.07, + "grad_norm": 2.2216200910686354, + "learning_rate": 1.9907283616453027e-05, + "loss": 1.0557, + "step": 3060 + }, + { + "epoch": 0.07, + "grad_norm": 2.113414367107525, + "learning_rate": 1.9907179922611414e-05, + "loss": 1.024, + "step": 3061 + }, + { + "epoch": 0.07, + "grad_norm": 2.1241822219559854, + "learning_rate": 1.9907076171087064e-05, + "loss": 1.0096, + "step": 3062 + }, + { + "epoch": 0.07, + "grad_norm": 2.4486707146676254, + "learning_rate": 1.990697236188058e-05, + "loss": 1.1324, + "step": 3063 + }, + { + "epoch": 0.07, + "grad_norm": 2.400249345321626, + "learning_rate": 1.9906868494992566e-05, + "loss": 1.0694, + "step": 3064 + }, + { + "epoch": 0.07, + "grad_norm": 1.1787038386493809, + "learning_rate": 1.9906764570423624e-05, + "loss": 0.9979, + "step": 3065 + }, + { + "epoch": 0.07, + "grad_norm": 2.0810922548133464, + "learning_rate": 1.9906660588174367e-05, + "loss": 0.9973, + "step": 3066 + }, + { + "epoch": 0.07, + "grad_norm": 2.200010401227223, + "learning_rate": 1.9906556548245395e-05, + "loss": 1.0372, + "step": 3067 + }, + { + "epoch": 0.07, + "grad_norm": 2.1739234525329545, + "learning_rate": 1.9906452450637312e-05, + "loss": 1.0573, + "step": 3068 + }, + { + "epoch": 0.07, + "grad_norm": 1.884912587400392, + "learning_rate": 1.9906348295350727e-05, + "loss": 1.0376, + "step": 3069 + }, + { + "epoch": 0.07, + "grad_norm": 2.062755777981078, + "learning_rate": 1.990624408238625e-05, + "loss": 1.1565, + "step": 3070 + }, + { + "epoch": 0.07, + "grad_norm": 2.0953878424855943, + "learning_rate": 1.990613981174448e-05, + "loss": 1.2021, + "step": 3071 + }, + { + "epoch": 0.07, + "grad_norm": 2.1421458839371095, + "learning_rate": 1.9906035483426024e-05, + "loss": 1.0931, + "step": 3072 + }, + { + "epoch": 0.07, + "grad_norm": 2.5854410859395207, + "learning_rate": 1.99059310974315e-05, + "loss": 0.9663, + "step": 3073 + }, + { + "epoch": 0.07, + "grad_norm": 1.2044167608815401, + "learning_rate": 1.9905826653761506e-05, + "loss": 0.927, + "step": 3074 + }, + { + "epoch": 0.07, + "grad_norm": 2.1145513308778106, + "learning_rate": 1.9905722152416652e-05, + "loss": 1.2121, + "step": 3075 + }, + { + "epoch": 0.07, + "grad_norm": 2.1290731428616256, + "learning_rate": 1.990561759339755e-05, + "loss": 1.017, + "step": 3076 + }, + { + "epoch": 0.07, + "grad_norm": 3.1685845203473098, + "learning_rate": 1.9905512976704805e-05, + "loss": 1.0172, + "step": 3077 + }, + { + "epoch": 0.07, + "grad_norm": 2.3652805003748876, + "learning_rate": 1.9905408302339025e-05, + "loss": 1.0811, + "step": 3078 + }, + { + "epoch": 0.07, + "grad_norm": 2.1205321665336214, + "learning_rate": 1.9905303570300828e-05, + "loss": 1.1213, + "step": 3079 + }, + { + "epoch": 0.07, + "grad_norm": 2.3690965255059244, + "learning_rate": 1.990519878059081e-05, + "loss": 1.0817, + "step": 3080 + }, + { + "epoch": 0.07, + "grad_norm": 2.0000365775924687, + "learning_rate": 1.9905093933209593e-05, + "loss": 1.0593, + "step": 3081 + }, + { + "epoch": 0.07, + "grad_norm": 2.9608477647911267, + "learning_rate": 1.9904989028157784e-05, + "loss": 1.1236, + "step": 3082 + }, + { + "epoch": 0.07, + "grad_norm": 2.3554671471979374, + "learning_rate": 1.990488406543599e-05, + "loss": 1.1397, + "step": 3083 + }, + { + "epoch": 0.07, + "grad_norm": 2.125753900084944, + "learning_rate": 1.9904779045044828e-05, + "loss": 1.0684, + "step": 3084 + }, + { + "epoch": 0.07, + "grad_norm": 2.3459106045058795, + "learning_rate": 1.9904673966984907e-05, + "loss": 1.0934, + "step": 3085 + }, + { + "epoch": 0.07, + "grad_norm": 2.053070909265341, + "learning_rate": 1.9904568831256838e-05, + "loss": 1.0073, + "step": 3086 + }, + { + "epoch": 0.07, + "grad_norm": 2.1358674752413447, + "learning_rate": 1.990446363786123e-05, + "loss": 1.0936, + "step": 3087 + }, + { + "epoch": 0.07, + "grad_norm": 2.2996272796188495, + "learning_rate": 1.9904358386798704e-05, + "loss": 1.1176, + "step": 3088 + }, + { + "epoch": 0.07, + "grad_norm": 2.3997679611643874, + "learning_rate": 1.9904253078069865e-05, + "loss": 1.035, + "step": 3089 + }, + { + "epoch": 0.07, + "grad_norm": 2.1973379203770866, + "learning_rate": 1.990414771167533e-05, + "loss": 1.0728, + "step": 3090 + }, + { + "epoch": 0.07, + "grad_norm": 2.3515545674202913, + "learning_rate": 1.9904042287615712e-05, + "loss": 0.9968, + "step": 3091 + }, + { + "epoch": 0.07, + "grad_norm": 2.0444880508575314, + "learning_rate": 1.990393680589162e-05, + "loss": 0.961, + "step": 3092 + }, + { + "epoch": 0.07, + "grad_norm": 2.2345853337452457, + "learning_rate": 1.9903831266503676e-05, + "loss": 0.933, + "step": 3093 + }, + { + "epoch": 0.07, + "grad_norm": 1.971198291379117, + "learning_rate": 1.9903725669452492e-05, + "loss": 1.164, + "step": 3094 + }, + { + "epoch": 0.07, + "grad_norm": 2.035993581580944, + "learning_rate": 1.990362001473868e-05, + "loss": 0.9821, + "step": 3095 + }, + { + "epoch": 0.07, + "grad_norm": 2.5374817809821475, + "learning_rate": 1.9903514302362856e-05, + "loss": 1.0354, + "step": 3096 + }, + { + "epoch": 0.07, + "grad_norm": 2.1584586825574212, + "learning_rate": 1.9903408532325637e-05, + "loss": 1.1308, + "step": 3097 + }, + { + "epoch": 0.07, + "grad_norm": 2.1380566602646702, + "learning_rate": 1.990330270462764e-05, + "loss": 1.2537, + "step": 3098 + }, + { + "epoch": 0.07, + "grad_norm": 1.8357291341797055, + "learning_rate": 1.9903196819269476e-05, + "loss": 0.9246, + "step": 3099 + }, + { + "epoch": 0.07, + "grad_norm": 2.0147455777275827, + "learning_rate": 1.9903090876251767e-05, + "loss": 1.0654, + "step": 3100 + }, + { + "epoch": 0.07, + "grad_norm": 2.424862221263038, + "learning_rate": 1.9902984875575127e-05, + "loss": 1.0739, + "step": 3101 + }, + { + "epoch": 0.07, + "grad_norm": 1.1572936175431079, + "learning_rate": 1.9902878817240175e-05, + "loss": 0.9813, + "step": 3102 + }, + { + "epoch": 0.07, + "grad_norm": 3.133032716409911, + "learning_rate": 1.9902772701247527e-05, + "loss": 1.1967, + "step": 3103 + }, + { + "epoch": 0.07, + "grad_norm": 1.1063501659283697, + "learning_rate": 1.9902666527597797e-05, + "loss": 0.9673, + "step": 3104 + }, + { + "epoch": 0.07, + "grad_norm": 2.2452441696472483, + "learning_rate": 1.9902560296291612e-05, + "loss": 1.0153, + "step": 3105 + }, + { + "epoch": 0.07, + "grad_norm": 2.7478079490324046, + "learning_rate": 1.9902454007329585e-05, + "loss": 1.0301, + "step": 3106 + }, + { + "epoch": 0.07, + "grad_norm": 2.043917913203123, + "learning_rate": 1.9902347660712336e-05, + "loss": 1.0486, + "step": 3107 + }, + { + "epoch": 0.07, + "grad_norm": 2.6331180183355776, + "learning_rate": 1.990224125644048e-05, + "loss": 1.1018, + "step": 3108 + }, + { + "epoch": 0.07, + "grad_norm": 1.9612394385153369, + "learning_rate": 1.9902134794514648e-05, + "loss": 1.0288, + "step": 3109 + }, + { + "epoch": 0.07, + "grad_norm": 2.159007924290347, + "learning_rate": 1.9902028274935446e-05, + "loss": 1.2298, + "step": 3110 + }, + { + "epoch": 0.07, + "grad_norm": 2.1340053447645886, + "learning_rate": 1.9901921697703503e-05, + "loss": 0.9481, + "step": 3111 + }, + { + "epoch": 0.07, + "grad_norm": 2.141048208681084, + "learning_rate": 1.9901815062819437e-05, + "loss": 0.9812, + "step": 3112 + }, + { + "epoch": 0.07, + "grad_norm": 2.031399916999051, + "learning_rate": 1.9901708370283866e-05, + "loss": 0.9911, + "step": 3113 + }, + { + "epoch": 0.07, + "grad_norm": 2.281624036655819, + "learning_rate": 1.9901601620097417e-05, + "loss": 1.153, + "step": 3114 + }, + { + "epoch": 0.07, + "grad_norm": 2.095708238439019, + "learning_rate": 1.990149481226071e-05, + "loss": 1.1821, + "step": 3115 + }, + { + "epoch": 0.07, + "grad_norm": 1.192247334820174, + "learning_rate": 1.9901387946774363e-05, + "loss": 1.0148, + "step": 3116 + }, + { + "epoch": 0.07, + "grad_norm": 2.2869524919249162, + "learning_rate": 1.9901281023639002e-05, + "loss": 0.9746, + "step": 3117 + }, + { + "epoch": 0.07, + "grad_norm": 2.0104230262804723, + "learning_rate": 1.9901174042855243e-05, + "loss": 1.0739, + "step": 3118 + }, + { + "epoch": 0.07, + "grad_norm": 2.2440428007001203, + "learning_rate": 1.9901067004423722e-05, + "loss": 1.0389, + "step": 3119 + }, + { + "epoch": 0.07, + "grad_norm": 1.1664496627161978, + "learning_rate": 1.990095990834505e-05, + "loss": 1.0063, + "step": 3120 + }, + { + "epoch": 0.07, + "grad_norm": 2.3857763174171094, + "learning_rate": 1.9900852754619853e-05, + "loss": 1.1273, + "step": 3121 + }, + { + "epoch": 0.07, + "grad_norm": 2.138845508420462, + "learning_rate": 1.9900745543248762e-05, + "loss": 1.1341, + "step": 3122 + }, + { + "epoch": 0.07, + "grad_norm": 2.7521749734292973, + "learning_rate": 1.990063827423239e-05, + "loss": 1.0587, + "step": 3123 + }, + { + "epoch": 0.07, + "grad_norm": 2.313815308963567, + "learning_rate": 1.990053094757137e-05, + "loss": 1.1357, + "step": 3124 + }, + { + "epoch": 0.07, + "grad_norm": 2.1003331546479274, + "learning_rate": 1.9900423563266325e-05, + "loss": 1.0667, + "step": 3125 + }, + { + "epoch": 0.07, + "grad_norm": 2.078624880110693, + "learning_rate": 1.9900316121317883e-05, + "loss": 1.0656, + "step": 3126 + }, + { + "epoch": 0.07, + "grad_norm": 2.300244059373704, + "learning_rate": 1.990020862172666e-05, + "loss": 1.2015, + "step": 3127 + }, + { + "epoch": 0.07, + "grad_norm": 2.2251276145958316, + "learning_rate": 1.990010106449329e-05, + "loss": 1.2715, + "step": 3128 + }, + { + "epoch": 0.07, + "grad_norm": 2.057257719667742, + "learning_rate": 1.9899993449618397e-05, + "loss": 1.0559, + "step": 3129 + }, + { + "epoch": 0.07, + "grad_norm": 1.946527763238528, + "learning_rate": 1.989988577710261e-05, + "loss": 1.056, + "step": 3130 + }, + { + "epoch": 0.07, + "grad_norm": 2.239517425249533, + "learning_rate": 1.9899778046946552e-05, + "loss": 1.0031, + "step": 3131 + }, + { + "epoch": 0.07, + "grad_norm": 2.0713439706613452, + "learning_rate": 1.9899670259150855e-05, + "loss": 1.0388, + "step": 3132 + }, + { + "epoch": 0.07, + "grad_norm": 2.0266092147834516, + "learning_rate": 1.9899562413716137e-05, + "loss": 0.9223, + "step": 3133 + }, + { + "epoch": 0.07, + "grad_norm": 2.4454714656574006, + "learning_rate": 1.989945451064304e-05, + "loss": 1.0983, + "step": 3134 + }, + { + "epoch": 0.07, + "grad_norm": 1.9366575711388851, + "learning_rate": 1.989934654993218e-05, + "loss": 1.0822, + "step": 3135 + }, + { + "epoch": 0.07, + "grad_norm": 2.026986142042329, + "learning_rate": 1.989923853158419e-05, + "loss": 1.1273, + "step": 3136 + }, + { + "epoch": 0.07, + "grad_norm": 1.233290947566744, + "learning_rate": 1.9899130455599703e-05, + "loss": 0.987, + "step": 3137 + }, + { + "epoch": 0.07, + "grad_norm": 2.212023533392144, + "learning_rate": 1.989902232197934e-05, + "loss": 1.1238, + "step": 3138 + }, + { + "epoch": 0.07, + "grad_norm": 2.668846653365077, + "learning_rate": 1.989891413072374e-05, + "loss": 1.1161, + "step": 3139 + }, + { + "epoch": 0.07, + "grad_norm": 1.9151483600746602, + "learning_rate": 1.9898805881833523e-05, + "loss": 1.0244, + "step": 3140 + }, + { + "epoch": 0.07, + "grad_norm": 3.4949330237142466, + "learning_rate": 1.9898697575309326e-05, + "loss": 0.9562, + "step": 3141 + }, + { + "epoch": 0.07, + "grad_norm": 3.7890944698609714, + "learning_rate": 1.989858921115178e-05, + "loss": 1.0394, + "step": 3142 + }, + { + "epoch": 0.07, + "grad_norm": 2.253534567555457, + "learning_rate": 1.989848078936151e-05, + "loss": 1.1113, + "step": 3143 + }, + { + "epoch": 0.07, + "grad_norm": 2.0925151432588134, + "learning_rate": 1.9898372309939153e-05, + "loss": 1.0014, + "step": 3144 + }, + { + "epoch": 0.07, + "grad_norm": 2.4977502316797895, + "learning_rate": 1.989826377288534e-05, + "loss": 1.1182, + "step": 3145 + }, + { + "epoch": 0.07, + "grad_norm": 1.1124720916334863, + "learning_rate": 1.98981551782007e-05, + "loss": 1.0059, + "step": 3146 + }, + { + "epoch": 0.07, + "grad_norm": 2.2884883040426227, + "learning_rate": 1.9898046525885868e-05, + "loss": 1.1635, + "step": 3147 + }, + { + "epoch": 0.07, + "grad_norm": 2.290050928190842, + "learning_rate": 1.9897937815941474e-05, + "loss": 1.0615, + "step": 3148 + }, + { + "epoch": 0.07, + "grad_norm": 2.138594855267277, + "learning_rate": 1.9897829048368153e-05, + "loss": 1.1676, + "step": 3149 + }, + { + "epoch": 0.07, + "grad_norm": 1.1488533256565845, + "learning_rate": 1.989772022316654e-05, + "loss": 0.9761, + "step": 3150 + }, + { + "epoch": 0.07, + "grad_norm": 2.174858482877, + "learning_rate": 1.9897611340337263e-05, + "loss": 1.0842, + "step": 3151 + }, + { + "epoch": 0.07, + "grad_norm": 2.282227272756692, + "learning_rate": 1.989750239988096e-05, + "loss": 0.9971, + "step": 3152 + }, + { + "epoch": 0.07, + "grad_norm": 2.2203564580868522, + "learning_rate": 1.9897393401798265e-05, + "loss": 1.1246, + "step": 3153 + }, + { + "epoch": 0.07, + "grad_norm": 2.1515493949980677, + "learning_rate": 1.989728434608981e-05, + "loss": 1.1228, + "step": 3154 + }, + { + "epoch": 0.07, + "grad_norm": 2.221159341471258, + "learning_rate": 1.9897175232756236e-05, + "loss": 1.1095, + "step": 3155 + }, + { + "epoch": 0.07, + "grad_norm": 2.3520785267329636, + "learning_rate": 1.9897066061798172e-05, + "loss": 1.2332, + "step": 3156 + }, + { + "epoch": 0.07, + "grad_norm": 1.2619065897436486, + "learning_rate": 1.9896956833216257e-05, + "loss": 1.0499, + "step": 3157 + }, + { + "epoch": 0.07, + "grad_norm": 2.2088256389128795, + "learning_rate": 1.9896847547011124e-05, + "loss": 1.0567, + "step": 3158 + }, + { + "epoch": 0.07, + "grad_norm": 1.9559061520253815, + "learning_rate": 1.9896738203183413e-05, + "loss": 1.0208, + "step": 3159 + }, + { + "epoch": 0.07, + "grad_norm": 2.1822147846501574, + "learning_rate": 1.9896628801733763e-05, + "loss": 1.1784, + "step": 3160 + }, + { + "epoch": 0.07, + "grad_norm": 2.093765990997747, + "learning_rate": 1.9896519342662798e-05, + "loss": 1.2138, + "step": 3161 + }, + { + "epoch": 0.07, + "grad_norm": 1.2490734395300167, + "learning_rate": 1.989640982597117e-05, + "loss": 1.0224, + "step": 3162 + }, + { + "epoch": 0.07, + "grad_norm": 2.762773961616038, + "learning_rate": 1.989630025165951e-05, + "loss": 1.0118, + "step": 3163 + }, + { + "epoch": 0.07, + "grad_norm": 1.1637459943106505, + "learning_rate": 1.9896190619728458e-05, + "loss": 0.943, + "step": 3164 + }, + { + "epoch": 0.07, + "grad_norm": 2.5616807567542605, + "learning_rate": 1.9896080930178646e-05, + "loss": 1.0514, + "step": 3165 + }, + { + "epoch": 0.07, + "grad_norm": 2.000770537529449, + "learning_rate": 1.9895971183010726e-05, + "loss": 1.0764, + "step": 3166 + }, + { + "epoch": 0.07, + "grad_norm": 1.995325757727942, + "learning_rate": 1.9895861378225324e-05, + "loss": 1.0951, + "step": 3167 + }, + { + "epoch": 0.07, + "grad_norm": 2.651495677642897, + "learning_rate": 1.989575151582308e-05, + "loss": 0.9957, + "step": 3168 + }, + { + "epoch": 0.07, + "grad_norm": 2.0887754692363485, + "learning_rate": 1.9895641595804644e-05, + "loss": 1.166, + "step": 3169 + }, + { + "epoch": 0.07, + "grad_norm": 2.1660296727319164, + "learning_rate": 1.9895531618170646e-05, + "loss": 1.1017, + "step": 3170 + }, + { + "epoch": 0.07, + "grad_norm": 1.9546635406822561, + "learning_rate": 1.9895421582921735e-05, + "loss": 1.011, + "step": 3171 + }, + { + "epoch": 0.07, + "grad_norm": 2.8568906359490067, + "learning_rate": 1.989531149005854e-05, + "loss": 0.9892, + "step": 3172 + }, + { + "epoch": 0.07, + "grad_norm": 1.279617724150931, + "learning_rate": 1.9895201339581717e-05, + "loss": 1.0241, + "step": 3173 + }, + { + "epoch": 0.07, + "grad_norm": 1.9460127639649931, + "learning_rate": 1.989509113149189e-05, + "loss": 1.2039, + "step": 3174 + }, + { + "epoch": 0.07, + "grad_norm": 1.952887686498006, + "learning_rate": 1.9894980865789716e-05, + "loss": 0.9665, + "step": 3175 + }, + { + "epoch": 0.07, + "grad_norm": 2.101802609355941, + "learning_rate": 1.9894870542475827e-05, + "loss": 1.0875, + "step": 3176 + }, + { + "epoch": 0.07, + "grad_norm": 1.8658042145867897, + "learning_rate": 1.989476016155087e-05, + "loss": 1.1433, + "step": 3177 + }, + { + "epoch": 0.07, + "grad_norm": 1.110828778348033, + "learning_rate": 1.9894649723015487e-05, + "loss": 0.9564, + "step": 3178 + }, + { + "epoch": 0.07, + "grad_norm": 2.087215533681179, + "learning_rate": 1.9894539226870325e-05, + "loss": 1.0774, + "step": 3179 + }, + { + "epoch": 0.07, + "grad_norm": 2.1801626346016616, + "learning_rate": 1.9894428673116016e-05, + "loss": 1.0734, + "step": 3180 + }, + { + "epoch": 0.07, + "grad_norm": 1.9792574122004871, + "learning_rate": 1.9894318061753214e-05, + "loss": 1.1282, + "step": 3181 + }, + { + "epoch": 0.07, + "grad_norm": 1.9906607869246071, + "learning_rate": 1.989420739278256e-05, + "loss": 0.986, + "step": 3182 + }, + { + "epoch": 0.07, + "grad_norm": 1.9454036723836419, + "learning_rate": 1.98940966662047e-05, + "loss": 1.0618, + "step": 3183 + }, + { + "epoch": 0.08, + "grad_norm": 2.0530180402760054, + "learning_rate": 1.9893985882020272e-05, + "loss": 1.1593, + "step": 3184 + }, + { + "epoch": 0.08, + "grad_norm": 2.6557647135556888, + "learning_rate": 1.989387504022993e-05, + "loss": 1.1583, + "step": 3185 + }, + { + "epoch": 0.08, + "grad_norm": 2.028113157271009, + "learning_rate": 1.9893764140834312e-05, + "loss": 1.0454, + "step": 3186 + }, + { + "epoch": 0.08, + "grad_norm": 1.280946610419917, + "learning_rate": 1.9893653183834068e-05, + "loss": 1.0169, + "step": 3187 + }, + { + "epoch": 0.08, + "grad_norm": 2.2960583984219305, + "learning_rate": 1.9893542169229844e-05, + "loss": 1.0821, + "step": 3188 + }, + { + "epoch": 0.08, + "grad_norm": 2.2588122452635973, + "learning_rate": 1.9893431097022284e-05, + "loss": 0.9227, + "step": 3189 + }, + { + "epoch": 0.08, + "grad_norm": 1.1441653792025583, + "learning_rate": 1.989331996721204e-05, + "loss": 0.9807, + "step": 3190 + }, + { + "epoch": 0.08, + "grad_norm": 2.0606993658354593, + "learning_rate": 1.9893208779799747e-05, + "loss": 1.1173, + "step": 3191 + }, + { + "epoch": 0.08, + "grad_norm": 2.049479280624887, + "learning_rate": 1.989309753478607e-05, + "loss": 1.0696, + "step": 3192 + }, + { + "epoch": 0.08, + "grad_norm": 1.198880219477418, + "learning_rate": 1.989298623217164e-05, + "loss": 0.9515, + "step": 3193 + }, + { + "epoch": 0.08, + "grad_norm": 1.856192634569911, + "learning_rate": 1.989287487195711e-05, + "loss": 1.0236, + "step": 3194 + }, + { + "epoch": 0.08, + "grad_norm": 2.261662725469567, + "learning_rate": 1.9892763454143134e-05, + "loss": 1.146, + "step": 3195 + }, + { + "epoch": 0.08, + "grad_norm": 2.3074778879217943, + "learning_rate": 1.989265197873036e-05, + "loss": 1.0894, + "step": 3196 + }, + { + "epoch": 0.08, + "grad_norm": 2.2655312977109325, + "learning_rate": 1.989254044571943e-05, + "loss": 1.0131, + "step": 3197 + }, + { + "epoch": 0.08, + "grad_norm": 2.8792357143826504, + "learning_rate": 1.9892428855111e-05, + "loss": 1.0756, + "step": 3198 + }, + { + "epoch": 0.08, + "grad_norm": 2.1025956437939834, + "learning_rate": 1.989231720690572e-05, + "loss": 1.0548, + "step": 3199 + }, + { + "epoch": 0.08, + "grad_norm": 2.2364485512504575, + "learning_rate": 1.989220550110423e-05, + "loss": 1.0631, + "step": 3200 + }, + { + "epoch": 0.08, + "grad_norm": 2.4412671805827157, + "learning_rate": 1.989209373770719e-05, + "loss": 1.1752, + "step": 3201 + }, + { + "epoch": 0.08, + "grad_norm": 2.157300693970296, + "learning_rate": 1.9891981916715254e-05, + "loss": 1.0772, + "step": 3202 + }, + { + "epoch": 0.08, + "grad_norm": 3.5378279328017013, + "learning_rate": 1.9891870038129066e-05, + "loss": 1.0947, + "step": 3203 + }, + { + "epoch": 0.08, + "grad_norm": 2.177017864786462, + "learning_rate": 1.9891758101949275e-05, + "loss": 1.0985, + "step": 3204 + }, + { + "epoch": 0.08, + "grad_norm": 2.8805034593780574, + "learning_rate": 1.9891646108176536e-05, + "loss": 1.0878, + "step": 3205 + }, + { + "epoch": 0.08, + "grad_norm": 1.9820033219987665, + "learning_rate": 1.9891534056811507e-05, + "loss": 1.1004, + "step": 3206 + }, + { + "epoch": 0.08, + "grad_norm": 2.2324560700740443, + "learning_rate": 1.9891421947854833e-05, + "loss": 1.1682, + "step": 3207 + }, + { + "epoch": 0.08, + "grad_norm": 2.4377740249750444, + "learning_rate": 1.9891309781307167e-05, + "loss": 1.0885, + "step": 3208 + }, + { + "epoch": 0.08, + "grad_norm": 2.0971742596604095, + "learning_rate": 1.9891197557169164e-05, + "loss": 1.0692, + "step": 3209 + }, + { + "epoch": 0.08, + "grad_norm": 1.98050116701983, + "learning_rate": 1.989108527544148e-05, + "loss": 0.9989, + "step": 3210 + }, + { + "epoch": 0.08, + "grad_norm": 1.2373862777821842, + "learning_rate": 1.9890972936124764e-05, + "loss": 1.0298, + "step": 3211 + }, + { + "epoch": 0.08, + "grad_norm": 2.2657830031152537, + "learning_rate": 1.989086053921967e-05, + "loss": 1.0711, + "step": 3212 + }, + { + "epoch": 0.08, + "grad_norm": 1.2107761717819179, + "learning_rate": 1.989074808472686e-05, + "loss": 1.011, + "step": 3213 + }, + { + "epoch": 0.08, + "grad_norm": 1.1086402465453051, + "learning_rate": 1.989063557264698e-05, + "loss": 0.9539, + "step": 3214 + }, + { + "epoch": 0.08, + "grad_norm": 1.097187715190572, + "learning_rate": 1.9890523002980687e-05, + "loss": 0.9204, + "step": 3215 + }, + { + "epoch": 0.08, + "grad_norm": 2.2755753669316454, + "learning_rate": 1.989041037572864e-05, + "loss": 1.0663, + "step": 3216 + }, + { + "epoch": 0.08, + "grad_norm": 2.0798679388888046, + "learning_rate": 1.989029769089149e-05, + "loss": 1.0365, + "step": 3217 + }, + { + "epoch": 0.08, + "grad_norm": 2.14671657237463, + "learning_rate": 1.98901849484699e-05, + "loss": 1.1109, + "step": 3218 + }, + { + "epoch": 0.08, + "grad_norm": 2.034010235224786, + "learning_rate": 1.9890072148464518e-05, + "loss": 1.0487, + "step": 3219 + }, + { + "epoch": 0.08, + "grad_norm": 1.9947767511062067, + "learning_rate": 1.988995929087601e-05, + "loss": 1.1463, + "step": 3220 + }, + { + "epoch": 0.08, + "grad_norm": 2.1199461383315854, + "learning_rate": 1.988984637570502e-05, + "loss": 1.0503, + "step": 3221 + }, + { + "epoch": 0.08, + "grad_norm": 1.1069376633699284, + "learning_rate": 1.9889733402952222e-05, + "loss": 1.0093, + "step": 3222 + }, + { + "epoch": 0.08, + "grad_norm": 2.291645590886643, + "learning_rate": 1.988962037261826e-05, + "loss": 1.1514, + "step": 3223 + }, + { + "epoch": 0.08, + "grad_norm": 2.6034438658316685, + "learning_rate": 1.98895072847038e-05, + "loss": 0.9264, + "step": 3224 + }, + { + "epoch": 0.08, + "grad_norm": 1.127889057455627, + "learning_rate": 1.9889394139209492e-05, + "loss": 0.9062, + "step": 3225 + }, + { + "epoch": 0.08, + "grad_norm": 2.46201732792298, + "learning_rate": 1.9889280936136007e-05, + "loss": 1.141, + "step": 3226 + }, + { + "epoch": 0.08, + "grad_norm": 2.6694968066764946, + "learning_rate": 1.9889167675483994e-05, + "loss": 1.0627, + "step": 3227 + }, + { + "epoch": 0.08, + "grad_norm": 1.9890668957182984, + "learning_rate": 1.9889054357254113e-05, + "loss": 1.0307, + "step": 3228 + }, + { + "epoch": 0.08, + "grad_norm": 2.155581360051634, + "learning_rate": 1.988894098144703e-05, + "loss": 1.0561, + "step": 3229 + }, + { + "epoch": 0.08, + "grad_norm": 2.0518010565899933, + "learning_rate": 1.9888827548063404e-05, + "loss": 1.0852, + "step": 3230 + }, + { + "epoch": 0.08, + "grad_norm": 2.0599312327564254, + "learning_rate": 1.988871405710389e-05, + "loss": 1.0916, + "step": 3231 + }, + { + "epoch": 0.08, + "grad_norm": 2.0613345076999083, + "learning_rate": 1.9888600508569157e-05, + "loss": 0.9818, + "step": 3232 + }, + { + "epoch": 0.08, + "grad_norm": 2.2014642005580694, + "learning_rate": 1.9888486902459852e-05, + "loss": 1.0702, + "step": 3233 + }, + { + "epoch": 0.08, + "grad_norm": 1.9306745431207457, + "learning_rate": 1.9888373238776654e-05, + "loss": 1.1384, + "step": 3234 + }, + { + "epoch": 0.08, + "grad_norm": 2.6493082715246046, + "learning_rate": 1.9888259517520212e-05, + "loss": 0.9879, + "step": 3235 + }, + { + "epoch": 0.08, + "grad_norm": 3.1711013005867037, + "learning_rate": 1.9888145738691196e-05, + "loss": 0.9428, + "step": 3236 + }, + { + "epoch": 0.08, + "grad_norm": 2.3559280940240432, + "learning_rate": 1.9888031902290265e-05, + "loss": 0.9941, + "step": 3237 + }, + { + "epoch": 0.08, + "grad_norm": 2.3665531710341083, + "learning_rate": 1.9887918008318076e-05, + "loss": 1.1094, + "step": 3238 + }, + { + "epoch": 0.08, + "grad_norm": 2.0323704504083495, + "learning_rate": 1.98878040567753e-05, + "loss": 1.0113, + "step": 3239 + }, + { + "epoch": 0.08, + "grad_norm": 3.0556566445415134, + "learning_rate": 1.9887690047662604e-05, + "loss": 1.0664, + "step": 3240 + }, + { + "epoch": 0.08, + "grad_norm": 2.147765349127514, + "learning_rate": 1.988757598098064e-05, + "loss": 1.2042, + "step": 3241 + }, + { + "epoch": 0.08, + "grad_norm": 1.247825086464708, + "learning_rate": 1.988746185673008e-05, + "loss": 1.0077, + "step": 3242 + }, + { + "epoch": 0.08, + "grad_norm": 1.9519604452382195, + "learning_rate": 1.9887347674911587e-05, + "loss": 0.9763, + "step": 3243 + }, + { + "epoch": 0.08, + "grad_norm": 3.025349990058982, + "learning_rate": 1.9887233435525827e-05, + "loss": 1.0633, + "step": 3244 + }, + { + "epoch": 0.08, + "grad_norm": 2.411420182947785, + "learning_rate": 1.9887119138573462e-05, + "loss": 0.9131, + "step": 3245 + }, + { + "epoch": 0.08, + "grad_norm": 2.157702796213529, + "learning_rate": 1.988700478405516e-05, + "loss": 0.978, + "step": 3246 + }, + { + "epoch": 0.08, + "grad_norm": 2.1878343066993726, + "learning_rate": 1.9886890371971585e-05, + "loss": 1.1328, + "step": 3247 + }, + { + "epoch": 0.08, + "grad_norm": 2.4447577150136306, + "learning_rate": 1.9886775902323405e-05, + "loss": 1.1774, + "step": 3248 + }, + { + "epoch": 0.08, + "grad_norm": 2.200458975480693, + "learning_rate": 1.9886661375111286e-05, + "loss": 1.1109, + "step": 3249 + }, + { + "epoch": 0.08, + "grad_norm": 2.474334856391671, + "learning_rate": 1.9886546790335895e-05, + "loss": 1.0648, + "step": 3250 + }, + { + "epoch": 0.08, + "grad_norm": 2.204027122145583, + "learning_rate": 1.9886432147997897e-05, + "loss": 1.0958, + "step": 3251 + }, + { + "epoch": 0.08, + "grad_norm": 2.117497310908794, + "learning_rate": 1.988631744809796e-05, + "loss": 1.1039, + "step": 3252 + }, + { + "epoch": 0.08, + "grad_norm": 2.982514029383163, + "learning_rate": 1.9886202690636755e-05, + "loss": 1.1238, + "step": 3253 + }, + { + "epoch": 0.08, + "grad_norm": 2.0708745075522668, + "learning_rate": 1.9886087875614946e-05, + "loss": 1.0642, + "step": 3254 + }, + { + "epoch": 0.08, + "grad_norm": 2.08970613271609, + "learning_rate": 1.9885973003033207e-05, + "loss": 1.2274, + "step": 3255 + }, + { + "epoch": 0.08, + "grad_norm": 2.186487960655401, + "learning_rate": 1.98858580728922e-05, + "loss": 1.1308, + "step": 3256 + }, + { + "epoch": 0.08, + "grad_norm": 2.152360033256671, + "learning_rate": 1.9885743085192598e-05, + "loss": 1.1491, + "step": 3257 + }, + { + "epoch": 0.08, + "grad_norm": 1.0934124387373625, + "learning_rate": 1.988562803993507e-05, + "loss": 0.9935, + "step": 3258 + }, + { + "epoch": 0.08, + "grad_norm": 1.1080879696672623, + "learning_rate": 1.988551293712029e-05, + "loss": 0.9638, + "step": 3259 + }, + { + "epoch": 0.08, + "grad_norm": 2.140577141068412, + "learning_rate": 1.9885397776748918e-05, + "loss": 0.9717, + "step": 3260 + }, + { + "epoch": 0.08, + "grad_norm": 1.9417227068825673, + "learning_rate": 1.9885282558821633e-05, + "loss": 1.1366, + "step": 3261 + }, + { + "epoch": 0.08, + "grad_norm": 1.1332559562325424, + "learning_rate": 1.98851672833391e-05, + "loss": 0.965, + "step": 3262 + }, + { + "epoch": 0.08, + "grad_norm": 3.2593842386725584, + "learning_rate": 1.9885051950301998e-05, + "loss": 1.098, + "step": 3263 + }, + { + "epoch": 0.08, + "grad_norm": 2.157531158128622, + "learning_rate": 1.9884936559710995e-05, + "loss": 0.8701, + "step": 3264 + }, + { + "epoch": 0.08, + "grad_norm": 1.897952966647107, + "learning_rate": 1.988482111156676e-05, + "loss": 1.0861, + "step": 3265 + }, + { + "epoch": 0.08, + "grad_norm": 2.036432254965549, + "learning_rate": 1.9884705605869965e-05, + "loss": 1.1009, + "step": 3266 + }, + { + "epoch": 0.08, + "grad_norm": 2.047715220167109, + "learning_rate": 1.9884590042621284e-05, + "loss": 1.1542, + "step": 3267 + }, + { + "epoch": 0.08, + "grad_norm": 2.4181697086751734, + "learning_rate": 1.9884474421821393e-05, + "loss": 1.2012, + "step": 3268 + }, + { + "epoch": 0.08, + "grad_norm": 2.2551446239584005, + "learning_rate": 1.9884358743470963e-05, + "loss": 1.0453, + "step": 3269 + }, + { + "epoch": 0.08, + "grad_norm": 2.1214801759490403, + "learning_rate": 1.9884243007570665e-05, + "loss": 1.116, + "step": 3270 + }, + { + "epoch": 0.08, + "grad_norm": 2.3146398027399164, + "learning_rate": 1.9884127214121176e-05, + "loss": 1.041, + "step": 3271 + }, + { + "epoch": 0.08, + "grad_norm": 2.030444527789238, + "learning_rate": 1.9884011363123168e-05, + "loss": 1.1275, + "step": 3272 + }, + { + "epoch": 0.08, + "grad_norm": 2.128998890226023, + "learning_rate": 1.988389545457732e-05, + "loss": 1.1009, + "step": 3273 + }, + { + "epoch": 0.08, + "grad_norm": 2.07571073257268, + "learning_rate": 1.9883779488484297e-05, + "loss": 1.1027, + "step": 3274 + }, + { + "epoch": 0.08, + "grad_norm": 2.863459861090787, + "learning_rate": 1.9883663464844786e-05, + "loss": 1.1333, + "step": 3275 + }, + { + "epoch": 0.08, + "grad_norm": 3.624462911090153, + "learning_rate": 1.9883547383659456e-05, + "loss": 1.0143, + "step": 3276 + }, + { + "epoch": 0.08, + "grad_norm": 1.8488017473927352, + "learning_rate": 1.9883431244928984e-05, + "loss": 0.9622, + "step": 3277 + }, + { + "epoch": 0.08, + "grad_norm": 2.15886631065947, + "learning_rate": 1.9883315048654048e-05, + "loss": 0.962, + "step": 3278 + }, + { + "epoch": 0.08, + "grad_norm": 1.1166974752802012, + "learning_rate": 1.988319879483532e-05, + "loss": 0.9286, + "step": 3279 + }, + { + "epoch": 0.08, + "grad_norm": 2.3313947953080585, + "learning_rate": 1.988308248347348e-05, + "loss": 1.0744, + "step": 3280 + }, + { + "epoch": 0.08, + "grad_norm": 2.0400470060802856, + "learning_rate": 1.9882966114569205e-05, + "loss": 1.0067, + "step": 3281 + }, + { + "epoch": 0.08, + "grad_norm": 2.1689498889295997, + "learning_rate": 1.9882849688123172e-05, + "loss": 1.1651, + "step": 3282 + }, + { + "epoch": 0.08, + "grad_norm": 2.252331132502813, + "learning_rate": 1.988273320413606e-05, + "loss": 1.0758, + "step": 3283 + }, + { + "epoch": 0.08, + "grad_norm": 2.112671671207256, + "learning_rate": 1.9882616662608547e-05, + "loss": 1.0294, + "step": 3284 + }, + { + "epoch": 0.08, + "grad_norm": 1.986213991075337, + "learning_rate": 1.988250006354131e-05, + "loss": 1.0499, + "step": 3285 + }, + { + "epoch": 0.08, + "grad_norm": 1.9476129754497251, + "learning_rate": 1.988238340693503e-05, + "loss": 1.0168, + "step": 3286 + }, + { + "epoch": 0.08, + "grad_norm": 2.1543357013210755, + "learning_rate": 1.9882266692790387e-05, + "loss": 1.1687, + "step": 3287 + }, + { + "epoch": 0.08, + "grad_norm": 2.254921258278134, + "learning_rate": 1.9882149921108053e-05, + "loss": 1.0939, + "step": 3288 + }, + { + "epoch": 0.08, + "grad_norm": 2.1787286455912955, + "learning_rate": 1.988203309188872e-05, + "loss": 1.036, + "step": 3289 + }, + { + "epoch": 0.08, + "grad_norm": 2.2837634430045908, + "learning_rate": 1.9881916205133057e-05, + "loss": 1.0191, + "step": 3290 + }, + { + "epoch": 0.08, + "grad_norm": 2.7841634814286804, + "learning_rate": 1.988179926084175e-05, + "loss": 1.0645, + "step": 3291 + }, + { + "epoch": 0.08, + "grad_norm": 1.9491813842186803, + "learning_rate": 1.988168225901548e-05, + "loss": 1.2428, + "step": 3292 + }, + { + "epoch": 0.08, + "grad_norm": 2.412030065638081, + "learning_rate": 1.9881565199654926e-05, + "loss": 0.9932, + "step": 3293 + }, + { + "epoch": 0.08, + "grad_norm": 2.4873878327878303, + "learning_rate": 1.9881448082760774e-05, + "loss": 0.9653, + "step": 3294 + }, + { + "epoch": 0.08, + "grad_norm": 2.331489368441746, + "learning_rate": 1.9881330908333702e-05, + "loss": 0.9943, + "step": 3295 + }, + { + "epoch": 0.08, + "grad_norm": 2.4212769328950774, + "learning_rate": 1.988121367637439e-05, + "loss": 1.1093, + "step": 3296 + }, + { + "epoch": 0.08, + "grad_norm": 2.1687149050422807, + "learning_rate": 1.9881096386883526e-05, + "loss": 1.0814, + "step": 3297 + }, + { + "epoch": 0.08, + "grad_norm": 2.0219145160561856, + "learning_rate": 1.988097903986179e-05, + "loss": 1.2059, + "step": 3298 + }, + { + "epoch": 0.08, + "grad_norm": 2.5172160465707325, + "learning_rate": 1.9880861635309867e-05, + "loss": 1.069, + "step": 3299 + }, + { + "epoch": 0.08, + "grad_norm": 1.9455467443589072, + "learning_rate": 1.9880744173228437e-05, + "loss": 1.0429, + "step": 3300 + }, + { + "epoch": 0.08, + "grad_norm": 2.2673691476465496, + "learning_rate": 1.988062665361819e-05, + "loss": 1.029, + "step": 3301 + }, + { + "epoch": 0.08, + "grad_norm": 2.265498721574474, + "learning_rate": 1.9880509076479803e-05, + "loss": 1.0808, + "step": 3302 + }, + { + "epoch": 0.08, + "grad_norm": 1.174089554786104, + "learning_rate": 1.9880391441813963e-05, + "loss": 0.963, + "step": 3303 + }, + { + "epoch": 0.08, + "grad_norm": 2.2593393983021177, + "learning_rate": 1.988027374962136e-05, + "loss": 1.045, + "step": 3304 + }, + { + "epoch": 0.08, + "grad_norm": 2.066710497482838, + "learning_rate": 1.9880155999902672e-05, + "loss": 1.0785, + "step": 3305 + }, + { + "epoch": 0.08, + "grad_norm": 2.343672608150805, + "learning_rate": 1.988003819265859e-05, + "loss": 1.0731, + "step": 3306 + }, + { + "epoch": 0.08, + "grad_norm": 2.0984421998942073, + "learning_rate": 1.98799203278898e-05, + "loss": 1.1626, + "step": 3307 + }, + { + "epoch": 0.08, + "grad_norm": 2.7511782722880813, + "learning_rate": 1.987980240559698e-05, + "loss": 1.0537, + "step": 3308 + }, + { + "epoch": 0.08, + "grad_norm": 1.1002629979562295, + "learning_rate": 1.9879684425780825e-05, + "loss": 0.9588, + "step": 3309 + }, + { + "epoch": 0.08, + "grad_norm": 1.9347022538490202, + "learning_rate": 1.987956638844202e-05, + "loss": 1.1104, + "step": 3310 + }, + { + "epoch": 0.08, + "grad_norm": 2.2006380895569926, + "learning_rate": 1.9879448293581253e-05, + "loss": 1.1228, + "step": 3311 + }, + { + "epoch": 0.08, + "grad_norm": 2.603679355375858, + "learning_rate": 1.9879330141199206e-05, + "loss": 1.0487, + "step": 3312 + }, + { + "epoch": 0.08, + "grad_norm": 2.075319996061352, + "learning_rate": 1.9879211931296572e-05, + "loss": 1.1207, + "step": 3313 + }, + { + "epoch": 0.08, + "grad_norm": 2.0199772933175257, + "learning_rate": 1.9879093663874044e-05, + "loss": 1.1672, + "step": 3314 + }, + { + "epoch": 0.08, + "grad_norm": 2.2711947609897605, + "learning_rate": 1.98789753389323e-05, + "loss": 1.2133, + "step": 3315 + }, + { + "epoch": 0.08, + "grad_norm": 2.026142904324733, + "learning_rate": 1.9878856956472033e-05, + "loss": 1.082, + "step": 3316 + }, + { + "epoch": 0.08, + "grad_norm": 2.1456021184861687, + "learning_rate": 1.987873851649394e-05, + "loss": 1.0903, + "step": 3317 + }, + { + "epoch": 0.08, + "grad_norm": 1.8343427236130465, + "learning_rate": 1.9878620018998696e-05, + "loss": 1.0494, + "step": 3318 + }, + { + "epoch": 0.08, + "grad_norm": 2.016884816668641, + "learning_rate": 1.9878501463987002e-05, + "loss": 1.0813, + "step": 3319 + }, + { + "epoch": 0.08, + "grad_norm": 2.981511716918446, + "learning_rate": 1.9878382851459545e-05, + "loss": 0.9885, + "step": 3320 + }, + { + "epoch": 0.08, + "grad_norm": 2.3130928802675936, + "learning_rate": 1.9878264181417016e-05, + "loss": 1.2195, + "step": 3321 + }, + { + "epoch": 0.08, + "grad_norm": 2.2525878124134926, + "learning_rate": 1.9878145453860106e-05, + "loss": 1.0648, + "step": 3322 + }, + { + "epoch": 0.08, + "grad_norm": 2.2624026843179528, + "learning_rate": 1.9878026668789503e-05, + "loss": 1.2437, + "step": 3323 + }, + { + "epoch": 0.08, + "grad_norm": 1.142107141860364, + "learning_rate": 1.9877907826205902e-05, + "loss": 1.044, + "step": 3324 + }, + { + "epoch": 0.08, + "grad_norm": 1.7817494634317648, + "learning_rate": 1.9877788926109997e-05, + "loss": 1.1302, + "step": 3325 + }, + { + "epoch": 0.08, + "grad_norm": 2.2936057388120776, + "learning_rate": 1.9877669968502476e-05, + "loss": 1.0343, + "step": 3326 + }, + { + "epoch": 0.08, + "grad_norm": 2.01323391956478, + "learning_rate": 1.9877550953384037e-05, + "loss": 1.119, + "step": 3327 + }, + { + "epoch": 0.08, + "grad_norm": 1.9414668582828365, + "learning_rate": 1.9877431880755364e-05, + "loss": 1.1285, + "step": 3328 + }, + { + "epoch": 0.08, + "grad_norm": 2.2913303678362684, + "learning_rate": 1.9877312750617155e-05, + "loss": 0.9854, + "step": 3329 + }, + { + "epoch": 0.08, + "grad_norm": 2.2632534762686634, + "learning_rate": 1.9877193562970105e-05, + "loss": 1.029, + "step": 3330 + }, + { + "epoch": 0.08, + "grad_norm": 2.490468395868134, + "learning_rate": 1.987707431781491e-05, + "loss": 1.0764, + "step": 3331 + }, + { + "epoch": 0.08, + "grad_norm": 2.144682576728334, + "learning_rate": 1.987695501515226e-05, + "loss": 1.0877, + "step": 3332 + }, + { + "epoch": 0.08, + "grad_norm": 2.0414165351764058, + "learning_rate": 1.9876835654982848e-05, + "loss": 1.0207, + "step": 3333 + }, + { + "epoch": 0.08, + "grad_norm": 2.2631883085689384, + "learning_rate": 1.9876716237307376e-05, + "loss": 1.0271, + "step": 3334 + }, + { + "epoch": 0.08, + "grad_norm": 2.417016976450948, + "learning_rate": 1.9876596762126532e-05, + "loss": 1.0855, + "step": 3335 + }, + { + "epoch": 0.08, + "grad_norm": 2.4421127902010147, + "learning_rate": 1.9876477229441014e-05, + "loss": 1.0481, + "step": 3336 + }, + { + "epoch": 0.08, + "grad_norm": 2.2143039789911247, + "learning_rate": 1.987635763925152e-05, + "loss": 1.2653, + "step": 3337 + }, + { + "epoch": 0.08, + "grad_norm": 2.1894472957536886, + "learning_rate": 1.9876237991558746e-05, + "loss": 1.2088, + "step": 3338 + }, + { + "epoch": 0.08, + "grad_norm": 2.027798902795328, + "learning_rate": 1.987611828636339e-05, + "loss": 1.1485, + "step": 3339 + }, + { + "epoch": 0.08, + "grad_norm": 2.0165761516531036, + "learning_rate": 1.987599852366614e-05, + "loss": 1.1247, + "step": 3340 + }, + { + "epoch": 0.08, + "grad_norm": 2.442543193490568, + "learning_rate": 1.9875878703467707e-05, + "loss": 1.0793, + "step": 3341 + }, + { + "epoch": 0.08, + "grad_norm": 2.0366111487957714, + "learning_rate": 1.987575882576878e-05, + "loss": 0.992, + "step": 3342 + }, + { + "epoch": 0.08, + "grad_norm": 2.563703503962869, + "learning_rate": 1.987563889057006e-05, + "loss": 1.022, + "step": 3343 + }, + { + "epoch": 0.08, + "grad_norm": 1.0820101328581173, + "learning_rate": 1.9875518897872236e-05, + "loss": 0.9785, + "step": 3344 + }, + { + "epoch": 0.08, + "grad_norm": 2.297757471766736, + "learning_rate": 1.987539884767602e-05, + "loss": 1.1597, + "step": 3345 + }, + { + "epoch": 0.08, + "grad_norm": 1.9552284740492398, + "learning_rate": 1.9875278739982104e-05, + "loss": 1.0484, + "step": 3346 + }, + { + "epoch": 0.08, + "grad_norm": 2.0629543159442623, + "learning_rate": 1.9875158574791192e-05, + "loss": 1.1173, + "step": 3347 + }, + { + "epoch": 0.08, + "grad_norm": 2.951375298207756, + "learning_rate": 1.9875038352103975e-05, + "loss": 1.1965, + "step": 3348 + }, + { + "epoch": 0.08, + "grad_norm": 2.1772397769050134, + "learning_rate": 1.9874918071921162e-05, + "loss": 0.8961, + "step": 3349 + }, + { + "epoch": 0.08, + "grad_norm": 2.442936589119608, + "learning_rate": 1.9874797734243446e-05, + "loss": 1.148, + "step": 3350 + }, + { + "epoch": 0.08, + "grad_norm": 2.24075476568761, + "learning_rate": 1.9874677339071536e-05, + "loss": 0.9603, + "step": 3351 + }, + { + "epoch": 0.08, + "grad_norm": 1.9146497134658063, + "learning_rate": 1.9874556886406126e-05, + "loss": 0.9897, + "step": 3352 + }, + { + "epoch": 0.08, + "grad_norm": 2.931517751083248, + "learning_rate": 1.987443637624792e-05, + "loss": 0.9893, + "step": 3353 + }, + { + "epoch": 0.08, + "grad_norm": 2.3401732161064106, + "learning_rate": 1.987431580859762e-05, + "loss": 1.0251, + "step": 3354 + }, + { + "epoch": 0.08, + "grad_norm": 1.1852547801852864, + "learning_rate": 1.9874195183455925e-05, + "loss": 0.9763, + "step": 3355 + }, + { + "epoch": 0.08, + "grad_norm": 1.990615472488195, + "learning_rate": 1.987407450082354e-05, + "loss": 1.0452, + "step": 3356 + }, + { + "epoch": 0.08, + "grad_norm": 2.146491716066881, + "learning_rate": 1.987395376070117e-05, + "loss": 1.0569, + "step": 3357 + }, + { + "epoch": 0.08, + "grad_norm": 2.012447411897982, + "learning_rate": 1.9873832963089512e-05, + "loss": 1.2342, + "step": 3358 + }, + { + "epoch": 0.08, + "grad_norm": 2.498152997347891, + "learning_rate": 1.9873712107989273e-05, + "loss": 1.1147, + "step": 3359 + }, + { + "epoch": 0.08, + "grad_norm": 2.24776540901117, + "learning_rate": 1.9873591195401157e-05, + "loss": 1.096, + "step": 3360 + }, + { + "epoch": 0.08, + "grad_norm": 2.2083041932467613, + "learning_rate": 1.9873470225325868e-05, + "loss": 1.1147, + "step": 3361 + }, + { + "epoch": 0.08, + "grad_norm": 2.3953412924746518, + "learning_rate": 1.987334919776411e-05, + "loss": 0.9985, + "step": 3362 + }, + { + "epoch": 0.08, + "grad_norm": 2.1302831838421032, + "learning_rate": 1.9873228112716583e-05, + "loss": 1.0245, + "step": 3363 + }, + { + "epoch": 0.08, + "grad_norm": 2.568059138782304, + "learning_rate": 1.9873106970184e-05, + "loss": 1.1016, + "step": 3364 + }, + { + "epoch": 0.08, + "grad_norm": 2.4489561537917215, + "learning_rate": 1.987298577016706e-05, + "loss": 1.0789, + "step": 3365 + }, + { + "epoch": 0.08, + "grad_norm": 2.5004273239390122, + "learning_rate": 1.9872864512666473e-05, + "loss": 1.0611, + "step": 3366 + }, + { + "epoch": 0.08, + "grad_norm": 2.1529423066168842, + "learning_rate": 1.987274319768294e-05, + "loss": 0.9624, + "step": 3367 + }, + { + "epoch": 0.08, + "grad_norm": 1.8408748562343804, + "learning_rate": 1.9872621825217174e-05, + "loss": 1.0281, + "step": 3368 + }, + { + "epoch": 0.08, + "grad_norm": 1.9875775992200984, + "learning_rate": 1.9872500395269876e-05, + "loss": 1.1372, + "step": 3369 + }, + { + "epoch": 0.08, + "grad_norm": 2.457225832508101, + "learning_rate": 1.9872378907841757e-05, + "loss": 0.9752, + "step": 3370 + }, + { + "epoch": 0.08, + "grad_norm": 1.9112030196342122, + "learning_rate": 1.987225736293352e-05, + "loss": 1.0794, + "step": 3371 + }, + { + "epoch": 0.08, + "grad_norm": 2.051340579891193, + "learning_rate": 1.9872135760545877e-05, + "loss": 1.0614, + "step": 3372 + }, + { + "epoch": 0.08, + "grad_norm": 2.1657049156806107, + "learning_rate": 1.9872014100679533e-05, + "loss": 1.064, + "step": 3373 + }, + { + "epoch": 0.08, + "grad_norm": 2.111969927020074, + "learning_rate": 1.98718923833352e-05, + "loss": 1.0732, + "step": 3374 + }, + { + "epoch": 0.08, + "grad_norm": 2.047896533980715, + "learning_rate": 1.987177060851358e-05, + "loss": 1.0345, + "step": 3375 + }, + { + "epoch": 0.08, + "grad_norm": 2.226870331482299, + "learning_rate": 1.987164877621539e-05, + "loss": 0.9826, + "step": 3376 + }, + { + "epoch": 0.08, + "grad_norm": 1.2320472173118726, + "learning_rate": 1.9871526886441334e-05, + "loss": 0.999, + "step": 3377 + }, + { + "epoch": 0.08, + "grad_norm": 2.4483684738129727, + "learning_rate": 1.9871404939192123e-05, + "loss": 0.9567, + "step": 3378 + }, + { + "epoch": 0.08, + "grad_norm": 2.1203076717086127, + "learning_rate": 1.9871282934468467e-05, + "loss": 1.0299, + "step": 3379 + }, + { + "epoch": 0.08, + "grad_norm": 2.1343367713853008, + "learning_rate": 1.9871160872271078e-05, + "loss": 1.1472, + "step": 3380 + }, + { + "epoch": 0.08, + "grad_norm": 1.1567265329613772, + "learning_rate": 1.9871038752600663e-05, + "loss": 0.9913, + "step": 3381 + }, + { + "epoch": 0.08, + "grad_norm": 1.1794266380746643, + "learning_rate": 1.987091657545794e-05, + "loss": 1.039, + "step": 3382 + }, + { + "epoch": 0.08, + "grad_norm": 2.413754230170961, + "learning_rate": 1.9870794340843608e-05, + "loss": 1.2293, + "step": 3383 + }, + { + "epoch": 0.08, + "grad_norm": 1.1830973833510783, + "learning_rate": 1.987067204875839e-05, + "loss": 1.033, + "step": 3384 + }, + { + "epoch": 0.08, + "grad_norm": 2.26329192691348, + "learning_rate": 1.9870549699202994e-05, + "loss": 1.0892, + "step": 3385 + }, + { + "epoch": 0.08, + "grad_norm": 2.272552034624227, + "learning_rate": 1.9870427292178136e-05, + "loss": 1.2025, + "step": 3386 + }, + { + "epoch": 0.08, + "grad_norm": 2.660347944321877, + "learning_rate": 1.9870304827684522e-05, + "loss": 1.0444, + "step": 3387 + }, + { + "epoch": 0.08, + "grad_norm": 1.9572518843212006, + "learning_rate": 1.9870182305722867e-05, + "loss": 1.0124, + "step": 3388 + }, + { + "epoch": 0.08, + "grad_norm": 2.1000959996809168, + "learning_rate": 1.9870059726293892e-05, + "loss": 1.1572, + "step": 3389 + }, + { + "epoch": 0.08, + "grad_norm": 2.2081790536684975, + "learning_rate": 1.9869937089398296e-05, + "loss": 1.0765, + "step": 3390 + }, + { + "epoch": 0.08, + "grad_norm": 2.0667758669224416, + "learning_rate": 1.986981439503681e-05, + "loss": 1.0766, + "step": 3391 + }, + { + "epoch": 0.08, + "grad_norm": 2.064975531508741, + "learning_rate": 1.9869691643210134e-05, + "loss": 0.9399, + "step": 3392 + }, + { + "epoch": 0.08, + "grad_norm": 2.2721295079201798, + "learning_rate": 1.986956883391899e-05, + "loss": 1.0476, + "step": 3393 + }, + { + "epoch": 0.08, + "grad_norm": 2.1590023281722024, + "learning_rate": 1.9869445967164093e-05, + "loss": 1.2178, + "step": 3394 + }, + { + "epoch": 0.08, + "grad_norm": 1.9826340262344602, + "learning_rate": 1.9869323042946155e-05, + "loss": 1.1296, + "step": 3395 + }, + { + "epoch": 0.08, + "grad_norm": 1.8361083531396496, + "learning_rate": 1.9869200061265893e-05, + "loss": 0.9109, + "step": 3396 + }, + { + "epoch": 0.08, + "grad_norm": 2.117029615045502, + "learning_rate": 1.9869077022124026e-05, + "loss": 1.1255, + "step": 3397 + }, + { + "epoch": 0.08, + "grad_norm": 2.226378027380644, + "learning_rate": 1.9868953925521267e-05, + "loss": 0.9888, + "step": 3398 + }, + { + "epoch": 0.08, + "grad_norm": 2.5048121110519594, + "learning_rate": 1.9868830771458337e-05, + "loss": 0.899, + "step": 3399 + }, + { + "epoch": 0.08, + "grad_norm": 2.150915975229599, + "learning_rate": 1.9868707559935945e-05, + "loss": 1.1222, + "step": 3400 + }, + { + "epoch": 0.08, + "grad_norm": 2.2875684919503296, + "learning_rate": 1.9868584290954812e-05, + "loss": 0.9683, + "step": 3401 + }, + { + "epoch": 0.08, + "grad_norm": 2.430561042704472, + "learning_rate": 1.9868460964515663e-05, + "loss": 1.1057, + "step": 3402 + }, + { + "epoch": 0.08, + "grad_norm": 2.3175421721578173, + "learning_rate": 1.9868337580619207e-05, + "loss": 0.986, + "step": 3403 + }, + { + "epoch": 0.08, + "grad_norm": 2.1769423768089458, + "learning_rate": 1.9868214139266168e-05, + "loss": 1.0543, + "step": 3404 + }, + { + "epoch": 0.08, + "grad_norm": 2.3288094720284116, + "learning_rate": 1.9868090640457258e-05, + "loss": 0.9934, + "step": 3405 + }, + { + "epoch": 0.08, + "grad_norm": 2.1664545113970477, + "learning_rate": 1.98679670841932e-05, + "loss": 1.195, + "step": 3406 + }, + { + "epoch": 0.08, + "grad_norm": 2.214836093260298, + "learning_rate": 1.9867843470474714e-05, + "loss": 1.0761, + "step": 3407 + }, + { + "epoch": 0.08, + "grad_norm": 2.192536379059764, + "learning_rate": 1.9867719799302524e-05, + "loss": 1.1116, + "step": 3408 + }, + { + "epoch": 0.08, + "grad_norm": 1.91744629487295, + "learning_rate": 1.986759607067734e-05, + "loss": 1.0127, + "step": 3409 + }, + { + "epoch": 0.08, + "grad_norm": 2.2265022028111874, + "learning_rate": 1.9867472284599888e-05, + "loss": 1.1465, + "step": 3410 + }, + { + "epoch": 0.08, + "grad_norm": 1.203494649151465, + "learning_rate": 1.986734844107089e-05, + "loss": 0.9687, + "step": 3411 + }, + { + "epoch": 0.08, + "grad_norm": 2.31662551596376, + "learning_rate": 1.9867224540091064e-05, + "loss": 1.1242, + "step": 3412 + }, + { + "epoch": 0.08, + "grad_norm": 1.1500968090367474, + "learning_rate": 1.9867100581661135e-05, + "loss": 1.0284, + "step": 3413 + }, + { + "epoch": 0.08, + "grad_norm": 3.1256537899432195, + "learning_rate": 1.9866976565781823e-05, + "loss": 1.0787, + "step": 3414 + }, + { + "epoch": 0.08, + "grad_norm": 2.2923761059667087, + "learning_rate": 1.9866852492453846e-05, + "loss": 1.0526, + "step": 3415 + }, + { + "epoch": 0.08, + "grad_norm": 1.155056302616258, + "learning_rate": 1.986672836167793e-05, + "loss": 1.0373, + "step": 3416 + }, + { + "epoch": 0.08, + "grad_norm": 2.045273673252379, + "learning_rate": 1.9866604173454802e-05, + "loss": 0.9546, + "step": 3417 + }, + { + "epoch": 0.08, + "grad_norm": 2.122208085768638, + "learning_rate": 1.986647992778518e-05, + "loss": 1.1516, + "step": 3418 + }, + { + "epoch": 0.08, + "grad_norm": 2.3748317209109544, + "learning_rate": 1.986635562466978e-05, + "loss": 1.1822, + "step": 3419 + }, + { + "epoch": 0.08, + "grad_norm": 2.1091679255075184, + "learning_rate": 1.9866231264109343e-05, + "loss": 1.009, + "step": 3420 + }, + { + "epoch": 0.08, + "grad_norm": 2.174883190672934, + "learning_rate": 1.986610684610458e-05, + "loss": 1.1549, + "step": 3421 + }, + { + "epoch": 0.08, + "grad_norm": 2.3942791311520675, + "learning_rate": 1.9865982370656224e-05, + "loss": 1.0752, + "step": 3422 + }, + { + "epoch": 0.08, + "grad_norm": 1.9560614478393912, + "learning_rate": 1.9865857837764992e-05, + "loss": 1.1485, + "step": 3423 + }, + { + "epoch": 0.08, + "grad_norm": 1.9777014518514338, + "learning_rate": 1.986573324743161e-05, + "loss": 1.0521, + "step": 3424 + }, + { + "epoch": 0.08, + "grad_norm": 1.8688144351257794, + "learning_rate": 1.986560859965681e-05, + "loss": 0.987, + "step": 3425 + }, + { + "epoch": 0.08, + "grad_norm": 2.09907226540094, + "learning_rate": 1.9865483894441312e-05, + "loss": 1.0635, + "step": 3426 + }, + { + "epoch": 0.08, + "grad_norm": 1.9886014843590667, + "learning_rate": 1.986535913178584e-05, + "loss": 1.1576, + "step": 3427 + }, + { + "epoch": 0.08, + "grad_norm": 2.020821051449154, + "learning_rate": 1.9865234311691127e-05, + "loss": 0.9943, + "step": 3428 + }, + { + "epoch": 0.08, + "grad_norm": 2.2425996465672737, + "learning_rate": 1.9865109434157896e-05, + "loss": 1.0083, + "step": 3429 + }, + { + "epoch": 0.08, + "grad_norm": 2.1548152514015104, + "learning_rate": 1.9864984499186873e-05, + "loss": 0.9823, + "step": 3430 + }, + { + "epoch": 0.08, + "grad_norm": 2.2748148775040336, + "learning_rate": 1.9864859506778787e-05, + "loss": 1.0011, + "step": 3431 + }, + { + "epoch": 0.08, + "grad_norm": 2.435243676600058, + "learning_rate": 1.9864734456934372e-05, + "loss": 0.9777, + "step": 3432 + }, + { + "epoch": 0.08, + "grad_norm": 2.10836546623861, + "learning_rate": 1.986460934965434e-05, + "loss": 1.0428, + "step": 3433 + }, + { + "epoch": 0.08, + "grad_norm": 2.144155116048775, + "learning_rate": 1.9864484184939438e-05, + "loss": 1.0249, + "step": 3434 + }, + { + "epoch": 0.08, + "grad_norm": 2.2702965642730955, + "learning_rate": 1.986435896279038e-05, + "loss": 1.0222, + "step": 3435 + }, + { + "epoch": 0.08, + "grad_norm": 2.591608609025285, + "learning_rate": 1.9864233683207907e-05, + "loss": 1.1287, + "step": 3436 + }, + { + "epoch": 0.08, + "grad_norm": 2.025073288577184, + "learning_rate": 1.986410834619274e-05, + "loss": 1.0497, + "step": 3437 + }, + { + "epoch": 0.08, + "grad_norm": 2.23047651424827, + "learning_rate": 1.986398295174561e-05, + "loss": 1.0681, + "step": 3438 + }, + { + "epoch": 0.08, + "grad_norm": 2.5166034784520663, + "learning_rate": 1.9863857499867248e-05, + "loss": 1.0216, + "step": 3439 + }, + { + "epoch": 0.08, + "grad_norm": 3.4250924393622815, + "learning_rate": 1.9863731990558386e-05, + "loss": 1.0725, + "step": 3440 + }, + { + "epoch": 0.08, + "grad_norm": 1.97123144788046, + "learning_rate": 1.9863606423819753e-05, + "loss": 1.0803, + "step": 3441 + }, + { + "epoch": 0.08, + "grad_norm": 2.378098900314342, + "learning_rate": 1.9863480799652078e-05, + "loss": 1.0346, + "step": 3442 + }, + { + "epoch": 0.08, + "grad_norm": 1.8408562668089428, + "learning_rate": 1.9863355118056102e-05, + "loss": 1.1899, + "step": 3443 + }, + { + "epoch": 0.08, + "grad_norm": 2.0059899696907633, + "learning_rate": 1.9863229379032542e-05, + "loss": 1.0548, + "step": 3444 + }, + { + "epoch": 0.08, + "grad_norm": 2.1507906037757527, + "learning_rate": 1.9863103582582144e-05, + "loss": 0.9725, + "step": 3445 + }, + { + "epoch": 0.08, + "grad_norm": 2.154109303067687, + "learning_rate": 1.9862977728705633e-05, + "loss": 1.1071, + "step": 3446 + }, + { + "epoch": 0.08, + "grad_norm": 2.333587833859352, + "learning_rate": 1.986285181740374e-05, + "loss": 0.937, + "step": 3447 + }, + { + "epoch": 0.08, + "grad_norm": 2.0861670274261748, + "learning_rate": 1.98627258486772e-05, + "loss": 0.9669, + "step": 3448 + }, + { + "epoch": 0.08, + "grad_norm": 2.267720884873625, + "learning_rate": 1.9862599822526755e-05, + "loss": 1.0425, + "step": 3449 + }, + { + "epoch": 0.08, + "grad_norm": 1.1053679381526955, + "learning_rate": 1.9862473738953127e-05, + "loss": 1.015, + "step": 3450 + }, + { + "epoch": 0.08, + "grad_norm": 2.020214715116678, + "learning_rate": 1.9862347597957054e-05, + "loss": 1.156, + "step": 3451 + }, + { + "epoch": 0.08, + "grad_norm": 2.2432865046920005, + "learning_rate": 1.9862221399539273e-05, + "loss": 1.0412, + "step": 3452 + }, + { + "epoch": 0.08, + "grad_norm": 2.064629102440814, + "learning_rate": 1.9862095143700512e-05, + "loss": 1.195, + "step": 3453 + }, + { + "epoch": 0.08, + "grad_norm": 2.4662695067651614, + "learning_rate": 1.986196883044152e-05, + "loss": 1.1278, + "step": 3454 + }, + { + "epoch": 0.08, + "grad_norm": 2.08726419304005, + "learning_rate": 1.9861842459763014e-05, + "loss": 1.0988, + "step": 3455 + }, + { + "epoch": 0.08, + "grad_norm": 2.435983215809776, + "learning_rate": 1.9861716031665744e-05, + "loss": 1.1224, + "step": 3456 + }, + { + "epoch": 0.08, + "grad_norm": 1.208354920801698, + "learning_rate": 1.986158954615044e-05, + "loss": 1.0094, + "step": 3457 + }, + { + "epoch": 0.08, + "grad_norm": 2.8142111012052147, + "learning_rate": 1.986146300321784e-05, + "loss": 1.102, + "step": 3458 + }, + { + "epoch": 0.08, + "grad_norm": 2.0115629246924476, + "learning_rate": 1.9861336402868678e-05, + "loss": 1.0987, + "step": 3459 + }, + { + "epoch": 0.08, + "grad_norm": 2.058567413174274, + "learning_rate": 1.9861209745103693e-05, + "loss": 1.2193, + "step": 3460 + }, + { + "epoch": 0.08, + "grad_norm": 3.41879836061929, + "learning_rate": 1.9861083029923623e-05, + "loss": 1.1132, + "step": 3461 + }, + { + "epoch": 0.08, + "grad_norm": 2.2662135714165745, + "learning_rate": 1.986095625732921e-05, + "loss": 0.9085, + "step": 3462 + }, + { + "epoch": 0.08, + "grad_norm": 2.0590586616867075, + "learning_rate": 1.9860829427321184e-05, + "loss": 1.0949, + "step": 3463 + }, + { + "epoch": 0.08, + "grad_norm": 2.380946342113664, + "learning_rate": 1.9860702539900288e-05, + "loss": 1.0295, + "step": 3464 + }, + { + "epoch": 0.08, + "grad_norm": 2.0617241750409447, + "learning_rate": 1.986057559506726e-05, + "loss": 1.1245, + "step": 3465 + }, + { + "epoch": 0.08, + "grad_norm": 2.089383747264275, + "learning_rate": 1.9860448592822837e-05, + "loss": 1.05, + "step": 3466 + }, + { + "epoch": 0.08, + "grad_norm": 2.000908894346506, + "learning_rate": 1.9860321533167763e-05, + "loss": 0.8974, + "step": 3467 + }, + { + "epoch": 0.08, + "grad_norm": 2.2590579687352568, + "learning_rate": 1.9860194416102777e-05, + "loss": 1.0879, + "step": 3468 + }, + { + "epoch": 0.08, + "grad_norm": 2.3979641061987556, + "learning_rate": 1.9860067241628616e-05, + "loss": 0.9513, + "step": 3469 + }, + { + "epoch": 0.08, + "grad_norm": 1.2062573104412047, + "learning_rate": 1.985994000974602e-05, + "loss": 1.0901, + "step": 3470 + }, + { + "epoch": 0.08, + "grad_norm": 2.5945691269838136, + "learning_rate": 1.985981272045573e-05, + "loss": 1.1495, + "step": 3471 + }, + { + "epoch": 0.08, + "grad_norm": 2.426758223136391, + "learning_rate": 1.9859685373758493e-05, + "loss": 1.1045, + "step": 3472 + }, + { + "epoch": 0.08, + "grad_norm": 2.2345277735693303, + "learning_rate": 1.9859557969655045e-05, + "loss": 1.1433, + "step": 3473 + }, + { + "epoch": 0.08, + "grad_norm": 1.1842818278431235, + "learning_rate": 1.9859430508146128e-05, + "loss": 0.9686, + "step": 3474 + }, + { + "epoch": 0.08, + "grad_norm": 2.9304576810326717, + "learning_rate": 1.9859302989232486e-05, + "loss": 1.1446, + "step": 3475 + }, + { + "epoch": 0.08, + "grad_norm": 2.022481242003808, + "learning_rate": 1.9859175412914863e-05, + "loss": 1.1399, + "step": 3476 + }, + { + "epoch": 0.08, + "grad_norm": 2.252903395419985, + "learning_rate": 1.9859047779194e-05, + "loss": 1.1051, + "step": 3477 + }, + { + "epoch": 0.08, + "grad_norm": 2.162298641884654, + "learning_rate": 1.9858920088070636e-05, + "loss": 0.9519, + "step": 3478 + }, + { + "epoch": 0.08, + "grad_norm": 2.2602795776405884, + "learning_rate": 1.9858792339545515e-05, + "loss": 1.1008, + "step": 3479 + }, + { + "epoch": 0.08, + "grad_norm": 1.1446454571761717, + "learning_rate": 1.9858664533619388e-05, + "loss": 0.9236, + "step": 3480 + }, + { + "epoch": 0.08, + "grad_norm": 2.3740686440174925, + "learning_rate": 1.9858536670292995e-05, + "loss": 1.1132, + "step": 3481 + }, + { + "epoch": 0.08, + "grad_norm": 2.4353781318943737, + "learning_rate": 1.985840874956708e-05, + "loss": 1.086, + "step": 3482 + }, + { + "epoch": 0.08, + "grad_norm": 2.340315576977413, + "learning_rate": 1.9858280771442387e-05, + "loss": 1.1864, + "step": 3483 + }, + { + "epoch": 0.08, + "grad_norm": 1.9539062257243935, + "learning_rate": 1.985815273591966e-05, + "loss": 1.0417, + "step": 3484 + }, + { + "epoch": 0.08, + "grad_norm": 1.1780745855499326, + "learning_rate": 1.9858024642999655e-05, + "loss": 0.9305, + "step": 3485 + }, + { + "epoch": 0.08, + "grad_norm": 2.3281999681350634, + "learning_rate": 1.98578964926831e-05, + "loss": 1.0144, + "step": 3486 + }, + { + "epoch": 0.08, + "grad_norm": 2.5508208813042113, + "learning_rate": 1.9857768284970754e-05, + "loss": 1.1433, + "step": 3487 + }, + { + "epoch": 0.08, + "grad_norm": 2.121355654312518, + "learning_rate": 1.9857640019863364e-05, + "loss": 1.0147, + "step": 3488 + }, + { + "epoch": 0.08, + "grad_norm": 2.0652857633366, + "learning_rate": 1.985751169736167e-05, + "loss": 1.1837, + "step": 3489 + }, + { + "epoch": 0.08, + "grad_norm": 2.1004639296499246, + "learning_rate": 1.985738331746642e-05, + "loss": 1.0017, + "step": 3490 + }, + { + "epoch": 0.08, + "grad_norm": 4.759296450953759, + "learning_rate": 1.9857254880178368e-05, + "loss": 1.0377, + "step": 3491 + }, + { + "epoch": 0.08, + "grad_norm": 3.163001986915013, + "learning_rate": 1.9857126385498255e-05, + "loss": 1.102, + "step": 3492 + }, + { + "epoch": 0.08, + "grad_norm": 1.9276312456377471, + "learning_rate": 1.9856997833426828e-05, + "loss": 1.2147, + "step": 3493 + }, + { + "epoch": 0.08, + "grad_norm": 2.0932116281145694, + "learning_rate": 1.9856869223964844e-05, + "loss": 1.114, + "step": 3494 + }, + { + "epoch": 0.08, + "grad_norm": 2.2993253889270915, + "learning_rate": 1.9856740557113046e-05, + "loss": 1.179, + "step": 3495 + }, + { + "epoch": 0.08, + "grad_norm": 1.2077899046702196, + "learning_rate": 1.9856611832872185e-05, + "loss": 1.0035, + "step": 3496 + }, + { + "epoch": 0.08, + "grad_norm": 2.18532430940525, + "learning_rate": 1.9856483051243005e-05, + "loss": 1.1467, + "step": 3497 + }, + { + "epoch": 0.08, + "grad_norm": 1.896269244785816, + "learning_rate": 1.9856354212226262e-05, + "loss": 1.0545, + "step": 3498 + }, + { + "epoch": 0.08, + "grad_norm": 2.0862198026502567, + "learning_rate": 1.9856225315822705e-05, + "loss": 1.1825, + "step": 3499 + }, + { + "epoch": 0.08, + "grad_norm": 1.1683844331775866, + "learning_rate": 1.9856096362033083e-05, + "loss": 1.0102, + "step": 3500 + }, + { + "epoch": 0.08, + "grad_norm": 2.2512922340780306, + "learning_rate": 1.9855967350858147e-05, + "loss": 1.0689, + "step": 3501 + }, + { + "epoch": 0.08, + "grad_norm": 2.368743598509511, + "learning_rate": 1.985583828229865e-05, + "loss": 1.0947, + "step": 3502 + }, + { + "epoch": 0.08, + "grad_norm": 2.402945230170654, + "learning_rate": 1.985570915635534e-05, + "loss": 1.1469, + "step": 3503 + }, + { + "epoch": 0.08, + "grad_norm": 2.6002187611131418, + "learning_rate": 1.985557997302898e-05, + "loss": 1.0786, + "step": 3504 + }, + { + "epoch": 0.08, + "grad_norm": 2.5321072357410035, + "learning_rate": 1.98554507323203e-05, + "loss": 1.12, + "step": 3505 + }, + { + "epoch": 0.08, + "grad_norm": 2.0819734764179816, + "learning_rate": 1.9855321434230075e-05, + "loss": 0.9937, + "step": 3506 + }, + { + "epoch": 0.08, + "grad_norm": 2.04275689789694, + "learning_rate": 1.9855192078759045e-05, + "loss": 0.9545, + "step": 3507 + }, + { + "epoch": 0.08, + "grad_norm": 2.960933379806405, + "learning_rate": 1.9855062665907967e-05, + "loss": 1.1126, + "step": 3508 + }, + { + "epoch": 0.08, + "grad_norm": 2.15027389287489, + "learning_rate": 1.9854933195677596e-05, + "loss": 1.0679, + "step": 3509 + }, + { + "epoch": 0.08, + "grad_norm": 2.0926711672262472, + "learning_rate": 1.9854803668068682e-05, + "loss": 1.1775, + "step": 3510 + }, + { + "epoch": 0.08, + "grad_norm": 2.3548832969620737, + "learning_rate": 1.985467408308198e-05, + "loss": 0.922, + "step": 3511 + }, + { + "epoch": 0.08, + "grad_norm": 2.199017655061525, + "learning_rate": 1.9854544440718247e-05, + "loss": 0.9575, + "step": 3512 + }, + { + "epoch": 0.08, + "grad_norm": 2.5561187841221753, + "learning_rate": 1.9854414740978235e-05, + "loss": 1.0872, + "step": 3513 + }, + { + "epoch": 0.08, + "grad_norm": 2.3273841272927185, + "learning_rate": 1.98542849838627e-05, + "loss": 1.2436, + "step": 3514 + }, + { + "epoch": 0.08, + "grad_norm": 2.1553776419407336, + "learning_rate": 1.98541551693724e-05, + "loss": 0.9657, + "step": 3515 + }, + { + "epoch": 0.08, + "grad_norm": 2.00142727795907, + "learning_rate": 1.985402529750809e-05, + "loss": 1.14, + "step": 3516 + }, + { + "epoch": 0.08, + "grad_norm": 2.334158136172389, + "learning_rate": 1.9853895368270524e-05, + "loss": 1.1564, + "step": 3517 + }, + { + "epoch": 0.08, + "grad_norm": 2.060394190604849, + "learning_rate": 1.9853765381660456e-05, + "loss": 1.1344, + "step": 3518 + }, + { + "epoch": 0.08, + "grad_norm": 2.2714949416540478, + "learning_rate": 1.985363533767865e-05, + "loss": 1.0594, + "step": 3519 + }, + { + "epoch": 0.08, + "grad_norm": 2.222586701081016, + "learning_rate": 1.9853505236325857e-05, + "loss": 1.1378, + "step": 3520 + }, + { + "epoch": 0.08, + "grad_norm": 2.0616656246013583, + "learning_rate": 1.985337507760284e-05, + "loss": 1.1432, + "step": 3521 + }, + { + "epoch": 0.08, + "grad_norm": 2.00534543749436, + "learning_rate": 1.985324486151035e-05, + "loss": 1.1591, + "step": 3522 + }, + { + "epoch": 0.08, + "grad_norm": 2.980377576622149, + "learning_rate": 1.9853114588049154e-05, + "loss": 0.9975, + "step": 3523 + }, + { + "epoch": 0.08, + "grad_norm": 2.4658522109999423, + "learning_rate": 1.9852984257220004e-05, + "loss": 0.9357, + "step": 3524 + }, + { + "epoch": 0.08, + "grad_norm": 2.2710838107161244, + "learning_rate": 1.9852853869023656e-05, + "loss": 1.1167, + "step": 3525 + }, + { + "epoch": 0.08, + "grad_norm": 2.1229029506216315, + "learning_rate": 1.9852723423460878e-05, + "loss": 1.1483, + "step": 3526 + }, + { + "epoch": 0.08, + "grad_norm": 2.035013691575524, + "learning_rate": 1.9852592920532424e-05, + "loss": 1.098, + "step": 3527 + }, + { + "epoch": 0.08, + "grad_norm": 1.113588039055176, + "learning_rate": 1.9852462360239053e-05, + "loss": 0.9943, + "step": 3528 + }, + { + "epoch": 0.08, + "grad_norm": 2.0349914633708206, + "learning_rate": 1.9852331742581528e-05, + "loss": 1.135, + "step": 3529 + }, + { + "epoch": 0.08, + "grad_norm": 2.212713405456198, + "learning_rate": 1.9852201067560607e-05, + "loss": 1.1814, + "step": 3530 + }, + { + "epoch": 0.08, + "grad_norm": 2.262806659777414, + "learning_rate": 1.9852070335177057e-05, + "loss": 1.1029, + "step": 3531 + }, + { + "epoch": 0.08, + "grad_norm": 1.918924169569528, + "learning_rate": 1.985193954543163e-05, + "loss": 0.9484, + "step": 3532 + }, + { + "epoch": 0.08, + "grad_norm": 2.155401931021166, + "learning_rate": 1.985180869832509e-05, + "loss": 1.0907, + "step": 3533 + }, + { + "epoch": 0.08, + "grad_norm": 2.3503160274175077, + "learning_rate": 1.9851677793858206e-05, + "loss": 1.0501, + "step": 3534 + }, + { + "epoch": 0.08, + "grad_norm": 2.1909936019794127, + "learning_rate": 1.985154683203173e-05, + "loss": 1.1752, + "step": 3535 + }, + { + "epoch": 0.08, + "grad_norm": 2.1549183881912812, + "learning_rate": 1.9851415812846432e-05, + "loss": 1.0985, + "step": 3536 + }, + { + "epoch": 0.08, + "grad_norm": 2.0841498099233835, + "learning_rate": 1.985128473630307e-05, + "loss": 1.0538, + "step": 3537 + }, + { + "epoch": 0.08, + "grad_norm": 2.428247559573331, + "learning_rate": 1.9851153602402414e-05, + "loss": 1.1137, + "step": 3538 + }, + { + "epoch": 0.08, + "grad_norm": 2.1497232043886947, + "learning_rate": 1.9851022411145222e-05, + "loss": 1.0918, + "step": 3539 + }, + { + "epoch": 0.08, + "grad_norm": 2.1327299704780103, + "learning_rate": 1.9850891162532255e-05, + "loss": 1.1211, + "step": 3540 + }, + { + "epoch": 0.08, + "grad_norm": 1.1964920194863, + "learning_rate": 1.9850759856564285e-05, + "loss": 1.0639, + "step": 3541 + }, + { + "epoch": 0.08, + "grad_norm": 2.7873993727236472, + "learning_rate": 1.9850628493242067e-05, + "loss": 0.9266, + "step": 3542 + }, + { + "epoch": 0.08, + "grad_norm": 1.895570807697469, + "learning_rate": 1.9850497072566377e-05, + "loss": 1.1426, + "step": 3543 + }, + { + "epoch": 0.08, + "grad_norm": 1.9211893549838226, + "learning_rate": 1.9850365594537972e-05, + "loss": 1.0882, + "step": 3544 + }, + { + "epoch": 0.08, + "grad_norm": 2.802001160570594, + "learning_rate": 1.985023405915762e-05, + "loss": 0.8736, + "step": 3545 + }, + { + "epoch": 0.08, + "grad_norm": 1.1597188705516381, + "learning_rate": 1.9850102466426086e-05, + "loss": 1.0222, + "step": 3546 + }, + { + "epoch": 0.08, + "grad_norm": 1.9903481122111224, + "learning_rate": 1.984997081634414e-05, + "loss": 1.0004, + "step": 3547 + }, + { + "epoch": 0.08, + "grad_norm": 2.0574213875342133, + "learning_rate": 1.9849839108912544e-05, + "loss": 0.994, + "step": 3548 + }, + { + "epoch": 0.08, + "grad_norm": 1.820240344845099, + "learning_rate": 1.9849707344132065e-05, + "loss": 1.0251, + "step": 3549 + }, + { + "epoch": 0.08, + "grad_norm": 2.239462415474472, + "learning_rate": 1.984957552200347e-05, + "loss": 0.966, + "step": 3550 + }, + { + "epoch": 0.08, + "grad_norm": 1.9144336096174266, + "learning_rate": 1.984944364252753e-05, + "loss": 1.0319, + "step": 3551 + }, + { + "epoch": 0.08, + "grad_norm": 2.1956274822237636, + "learning_rate": 1.984931170570501e-05, + "loss": 0.9957, + "step": 3552 + }, + { + "epoch": 0.08, + "grad_norm": 1.9946817925677969, + "learning_rate": 1.984917971153668e-05, + "loss": 1.0687, + "step": 3553 + }, + { + "epoch": 0.08, + "grad_norm": 2.186342429948675, + "learning_rate": 1.984904766002331e-05, + "loss": 1.0025, + "step": 3554 + }, + { + "epoch": 0.08, + "grad_norm": 2.207546343236624, + "learning_rate": 1.984891555116566e-05, + "loss": 1.0819, + "step": 3555 + }, + { + "epoch": 0.08, + "grad_norm": 1.1180132112402357, + "learning_rate": 1.9848783384964513e-05, + "loss": 1.0329, + "step": 3556 + }, + { + "epoch": 0.08, + "grad_norm": 2.3533460262138637, + "learning_rate": 1.9848651161420625e-05, + "loss": 1.0077, + "step": 3557 + }, + { + "epoch": 0.08, + "grad_norm": 2.244218548653567, + "learning_rate": 1.9848518880534775e-05, + "loss": 1.1177, + "step": 3558 + }, + { + "epoch": 0.08, + "grad_norm": 2.232708738365647, + "learning_rate": 1.9848386542307728e-05, + "loss": 0.9156, + "step": 3559 + }, + { + "epoch": 0.08, + "grad_norm": 2.2702236310457575, + "learning_rate": 1.9848254146740257e-05, + "loss": 1.1731, + "step": 3560 + }, + { + "epoch": 0.08, + "grad_norm": 2.1189185331577343, + "learning_rate": 1.9848121693833135e-05, + "loss": 1.118, + "step": 3561 + }, + { + "epoch": 0.08, + "grad_norm": 2.134888257189043, + "learning_rate": 1.9847989183587127e-05, + "loss": 1.0791, + "step": 3562 + }, + { + "epoch": 0.08, + "grad_norm": 1.163059148431895, + "learning_rate": 1.9847856616003013e-05, + "loss": 1.0423, + "step": 3563 + }, + { + "epoch": 0.08, + "grad_norm": 2.5096822812940554, + "learning_rate": 1.9847723991081555e-05, + "loss": 1.075, + "step": 3564 + }, + { + "epoch": 0.08, + "grad_norm": 2.3475187862098434, + "learning_rate": 1.9847591308823535e-05, + "loss": 1.1115, + "step": 3565 + }, + { + "epoch": 0.08, + "grad_norm": 1.9963733072347751, + "learning_rate": 1.9847458569229716e-05, + "loss": 1.0053, + "step": 3566 + }, + { + "epoch": 0.08, + "grad_norm": 1.9916188158241153, + "learning_rate": 1.9847325772300878e-05, + "loss": 1.1545, + "step": 3567 + }, + { + "epoch": 0.08, + "grad_norm": 2.1674772672034446, + "learning_rate": 1.984719291803779e-05, + "loss": 1.1434, + "step": 3568 + }, + { + "epoch": 0.08, + "grad_norm": 2.1090580779432835, + "learning_rate": 1.984706000644123e-05, + "loss": 1.0284, + "step": 3569 + }, + { + "epoch": 0.08, + "grad_norm": 2.8134441122284333, + "learning_rate": 1.984692703751197e-05, + "loss": 1.0293, + "step": 3570 + }, + { + "epoch": 0.08, + "grad_norm": 1.1724185370531632, + "learning_rate": 1.984679401125078e-05, + "loss": 0.987, + "step": 3571 + }, + { + "epoch": 0.08, + "grad_norm": 2.038829292740374, + "learning_rate": 1.9846660927658437e-05, + "loss": 1.0775, + "step": 3572 + }, + { + "epoch": 0.08, + "grad_norm": 1.9654166583683488, + "learning_rate": 1.9846527786735722e-05, + "loss": 1.1435, + "step": 3573 + }, + { + "epoch": 0.08, + "grad_norm": 2.0392524185014334, + "learning_rate": 1.98463945884834e-05, + "loss": 1.2144, + "step": 3574 + }, + { + "epoch": 0.08, + "grad_norm": 2.099528653012532, + "learning_rate": 1.984626133290225e-05, + "loss": 1.1046, + "step": 3575 + }, + { + "epoch": 0.08, + "grad_norm": 2.43562348894577, + "learning_rate": 1.9846128019993054e-05, + "loss": 1.1046, + "step": 3576 + }, + { + "epoch": 0.08, + "grad_norm": 2.6076865630692563, + "learning_rate": 1.984599464975658e-05, + "loss": 1.0529, + "step": 3577 + }, + { + "epoch": 0.08, + "grad_norm": 2.146197185327623, + "learning_rate": 1.984586122219361e-05, + "loss": 0.904, + "step": 3578 + }, + { + "epoch": 0.08, + "grad_norm": 2.4988634247214905, + "learning_rate": 1.9845727737304917e-05, + "loss": 0.9388, + "step": 3579 + }, + { + "epoch": 0.08, + "grad_norm": 2.2721856299451852, + "learning_rate": 1.9845594195091283e-05, + "loss": 0.9489, + "step": 3580 + }, + { + "epoch": 0.08, + "grad_norm": 1.949104833714264, + "learning_rate": 1.9845460595553482e-05, + "loss": 0.9706, + "step": 3581 + }, + { + "epoch": 0.08, + "grad_norm": 2.194686953991002, + "learning_rate": 1.9845326938692292e-05, + "loss": 0.9859, + "step": 3582 + }, + { + "epoch": 0.08, + "grad_norm": 2.106298288326299, + "learning_rate": 1.984519322450849e-05, + "loss": 1.1298, + "step": 3583 + }, + { + "epoch": 0.08, + "grad_norm": 2.617914961572096, + "learning_rate": 1.9845059453002858e-05, + "loss": 1.0308, + "step": 3584 + }, + { + "epoch": 0.08, + "grad_norm": 1.1085909890455687, + "learning_rate": 1.984492562417617e-05, + "loss": 1.0195, + "step": 3585 + }, + { + "epoch": 0.08, + "grad_norm": 2.512579872075526, + "learning_rate": 1.984479173802921e-05, + "loss": 1.0933, + "step": 3586 + }, + { + "epoch": 0.08, + "grad_norm": 1.920009931779274, + "learning_rate": 1.9844657794562755e-05, + "loss": 1.0841, + "step": 3587 + }, + { + "epoch": 0.08, + "grad_norm": 3.899245824021934, + "learning_rate": 1.984452379377759e-05, + "loss": 1.1648, + "step": 3588 + }, + { + "epoch": 0.08, + "grad_norm": 2.229358676826112, + "learning_rate": 1.9844389735674484e-05, + "loss": 1.0693, + "step": 3589 + }, + { + "epoch": 0.08, + "grad_norm": 2.042375537941074, + "learning_rate": 1.9844255620254227e-05, + "loss": 1.023, + "step": 3590 + }, + { + "epoch": 0.08, + "grad_norm": 2.3594162148703153, + "learning_rate": 1.98441214475176e-05, + "loss": 1.0107, + "step": 3591 + }, + { + "epoch": 0.08, + "grad_norm": 1.2023962504916579, + "learning_rate": 1.9843987217465378e-05, + "loss": 1.0672, + "step": 3592 + }, + { + "epoch": 0.08, + "grad_norm": 2.093761321293413, + "learning_rate": 1.9843852930098345e-05, + "loss": 1.1366, + "step": 3593 + }, + { + "epoch": 0.08, + "grad_norm": 2.654340971807784, + "learning_rate": 1.9843718585417284e-05, + "loss": 1.0191, + "step": 3594 + }, + { + "epoch": 0.08, + "grad_norm": 2.407588674998933, + "learning_rate": 1.9843584183422983e-05, + "loss": 1.0906, + "step": 3595 + }, + { + "epoch": 0.08, + "grad_norm": 2.4742003020320618, + "learning_rate": 1.984344972411621e-05, + "loss": 1.0722, + "step": 3596 + }, + { + "epoch": 0.08, + "grad_norm": 2.0738311318292912, + "learning_rate": 1.984331520749776e-05, + "loss": 1.1127, + "step": 3597 + }, + { + "epoch": 0.08, + "grad_norm": 2.263643289892428, + "learning_rate": 1.9843180633568412e-05, + "loss": 1.143, + "step": 3598 + }, + { + "epoch": 0.08, + "grad_norm": 2.2604964562499164, + "learning_rate": 1.9843046002328953e-05, + "loss": 1.1046, + "step": 3599 + }, + { + "epoch": 0.08, + "grad_norm": 2.1024599700505786, + "learning_rate": 1.984291131378016e-05, + "loss": 1.073, + "step": 3600 + }, + { + "epoch": 0.08, + "grad_norm": 2.260311190570866, + "learning_rate": 1.9842776567922825e-05, + "loss": 1.0466, + "step": 3601 + }, + { + "epoch": 0.08, + "grad_norm": 2.863339516078095, + "learning_rate": 1.9842641764757724e-05, + "loss": 1.0487, + "step": 3602 + }, + { + "epoch": 0.08, + "grad_norm": 2.4760336266080714, + "learning_rate": 1.984250690428565e-05, + "loss": 1.0403, + "step": 3603 + }, + { + "epoch": 0.08, + "grad_norm": 1.0824715150276611, + "learning_rate": 1.9842371986507384e-05, + "loss": 0.9513, + "step": 3604 + }, + { + "epoch": 0.08, + "grad_norm": 2.0289518138218545, + "learning_rate": 1.9842237011423712e-05, + "loss": 1.0355, + "step": 3605 + }, + { + "epoch": 0.08, + "grad_norm": 2.32105657649422, + "learning_rate": 1.9842101979035416e-05, + "loss": 0.947, + "step": 3606 + }, + { + "epoch": 0.08, + "grad_norm": 2.1421959639645682, + "learning_rate": 1.984196688934329e-05, + "loss": 1.0423, + "step": 3607 + }, + { + "epoch": 0.09, + "grad_norm": 2.2108972287975517, + "learning_rate": 1.984183174234812e-05, + "loss": 1.1614, + "step": 3608 + }, + { + "epoch": 0.09, + "grad_norm": 2.0090008739545495, + "learning_rate": 1.9841696538050686e-05, + "loss": 1.065, + "step": 3609 + }, + { + "epoch": 0.09, + "grad_norm": 2.416434344448416, + "learning_rate": 1.984156127645178e-05, + "loss": 1.0034, + "step": 3610 + }, + { + "epoch": 0.09, + "grad_norm": 2.122388857835392, + "learning_rate": 1.9841425957552188e-05, + "loss": 1.1786, + "step": 3611 + }, + { + "epoch": 0.09, + "grad_norm": 2.0254305556949244, + "learning_rate": 1.98412905813527e-05, + "loss": 1.0599, + "step": 3612 + }, + { + "epoch": 0.09, + "grad_norm": 2.150199367331563, + "learning_rate": 1.98411551478541e-05, + "loss": 1.0299, + "step": 3613 + }, + { + "epoch": 0.09, + "grad_norm": 2.2979323026023044, + "learning_rate": 1.984101965705718e-05, + "loss": 1.1809, + "step": 3614 + }, + { + "epoch": 0.09, + "grad_norm": 2.125519151548887, + "learning_rate": 1.984088410896273e-05, + "loss": 0.9919, + "step": 3615 + }, + { + "epoch": 0.09, + "grad_norm": 2.216874078317911, + "learning_rate": 1.9840748503571535e-05, + "loss": 1.1001, + "step": 3616 + }, + { + "epoch": 0.09, + "grad_norm": 1.1783993511842583, + "learning_rate": 1.9840612840884386e-05, + "loss": 1.0119, + "step": 3617 + }, + { + "epoch": 0.09, + "grad_norm": 2.187078076995596, + "learning_rate": 1.9840477120902076e-05, + "loss": 1.0629, + "step": 3618 + }, + { + "epoch": 0.09, + "grad_norm": 2.0133401138863753, + "learning_rate": 1.984034134362539e-05, + "loss": 0.9925, + "step": 3619 + }, + { + "epoch": 0.09, + "grad_norm": 2.4549902893224154, + "learning_rate": 1.9840205509055123e-05, + "loss": 0.9244, + "step": 3620 + }, + { + "epoch": 0.09, + "grad_norm": 1.0973039117586096, + "learning_rate": 1.9840069617192063e-05, + "loss": 0.9763, + "step": 3621 + }, + { + "epoch": 0.09, + "grad_norm": 2.085037290573189, + "learning_rate": 1.9839933668037003e-05, + "loss": 1.0009, + "step": 3622 + }, + { + "epoch": 0.09, + "grad_norm": 2.0681440729430136, + "learning_rate": 1.9839797661590734e-05, + "loss": 1.0012, + "step": 3623 + }, + { + "epoch": 0.09, + "grad_norm": 1.117499116619312, + "learning_rate": 1.9839661597854048e-05, + "loss": 1.0017, + "step": 3624 + }, + { + "epoch": 0.09, + "grad_norm": 2.2948161678115353, + "learning_rate": 1.983952547682774e-05, + "loss": 1.0102, + "step": 3625 + }, + { + "epoch": 0.09, + "grad_norm": 2.089715329351157, + "learning_rate": 1.9839389298512593e-05, + "loss": 1.0681, + "step": 3626 + }, + { + "epoch": 0.09, + "grad_norm": 1.9472978984058011, + "learning_rate": 1.9839253062909408e-05, + "loss": 1.1418, + "step": 3627 + }, + { + "epoch": 0.09, + "grad_norm": 2.5892984660753204, + "learning_rate": 1.983911677001898e-05, + "loss": 1.1638, + "step": 3628 + }, + { + "epoch": 0.09, + "grad_norm": 2.3425650126252604, + "learning_rate": 1.98389804198421e-05, + "loss": 1.062, + "step": 3629 + }, + { + "epoch": 0.09, + "grad_norm": 2.3015268495788814, + "learning_rate": 1.9838844012379557e-05, + "loss": 1.1169, + "step": 3630 + }, + { + "epoch": 0.09, + "grad_norm": 1.1135485497617077, + "learning_rate": 1.983870754763215e-05, + "loss": 0.9659, + "step": 3631 + }, + { + "epoch": 0.09, + "grad_norm": 1.1052137285442267, + "learning_rate": 1.9838571025600674e-05, + "loss": 0.9828, + "step": 3632 + }, + { + "epoch": 0.09, + "grad_norm": 2.544115051804753, + "learning_rate": 1.983843444628592e-05, + "loss": 1.2207, + "step": 3633 + }, + { + "epoch": 0.09, + "grad_norm": 2.230646863175, + "learning_rate": 1.983829780968869e-05, + "loss": 1.0958, + "step": 3634 + }, + { + "epoch": 0.09, + "grad_norm": 1.950204569076381, + "learning_rate": 1.9838161115809773e-05, + "loss": 1.0215, + "step": 3635 + }, + { + "epoch": 0.09, + "grad_norm": 2.171974916205254, + "learning_rate": 1.983802436464997e-05, + "loss": 0.9949, + "step": 3636 + }, + { + "epoch": 0.09, + "grad_norm": 1.1365918177919707, + "learning_rate": 1.983788755621007e-05, + "loss": 0.9874, + "step": 3637 + }, + { + "epoch": 0.09, + "grad_norm": 2.1714238122251266, + "learning_rate": 1.9837750690490876e-05, + "loss": 1.1047, + "step": 3638 + }, + { + "epoch": 0.09, + "grad_norm": 2.0472417313612215, + "learning_rate": 1.9837613767493186e-05, + "loss": 1.1152, + "step": 3639 + }, + { + "epoch": 0.09, + "grad_norm": 2.0160035737871542, + "learning_rate": 1.983747678721779e-05, + "loss": 1.0291, + "step": 3640 + }, + { + "epoch": 0.09, + "grad_norm": 2.184108180702499, + "learning_rate": 1.983733974966549e-05, + "loss": 1.1719, + "step": 3641 + }, + { + "epoch": 0.09, + "grad_norm": 2.285076574114372, + "learning_rate": 1.9837202654837085e-05, + "loss": 1.143, + "step": 3642 + }, + { + "epoch": 0.09, + "grad_norm": 1.7711431133582956, + "learning_rate": 1.983706550273337e-05, + "loss": 1.1328, + "step": 3643 + }, + { + "epoch": 0.09, + "grad_norm": 2.3774436467666082, + "learning_rate": 1.9836928293355146e-05, + "loss": 1.0566, + "step": 3644 + }, + { + "epoch": 0.09, + "grad_norm": 1.1514557291967, + "learning_rate": 1.9836791026703213e-05, + "loss": 1.0622, + "step": 3645 + }, + { + "epoch": 0.09, + "grad_norm": 2.1622434539409765, + "learning_rate": 1.983665370277837e-05, + "loss": 1.0282, + "step": 3646 + }, + { + "epoch": 0.09, + "grad_norm": 2.6143519534580095, + "learning_rate": 1.9836516321581408e-05, + "loss": 1.0846, + "step": 3647 + }, + { + "epoch": 0.09, + "grad_norm": 1.1440180797368735, + "learning_rate": 1.9836378883113137e-05, + "loss": 0.9405, + "step": 3648 + }, + { + "epoch": 0.09, + "grad_norm": 2.3425408236332816, + "learning_rate": 1.9836241387374354e-05, + "loss": 1.1253, + "step": 3649 + }, + { + "epoch": 0.09, + "grad_norm": 2.5568498729086815, + "learning_rate": 1.9836103834365864e-05, + "loss": 1.053, + "step": 3650 + }, + { + "epoch": 0.09, + "grad_norm": 2.1693235657492784, + "learning_rate": 1.983596622408846e-05, + "loss": 1.1114, + "step": 3651 + }, + { + "epoch": 0.09, + "grad_norm": 1.92021946162726, + "learning_rate": 1.9835828556542944e-05, + "loss": 0.9731, + "step": 3652 + }, + { + "epoch": 0.09, + "grad_norm": 2.1673843816232616, + "learning_rate": 1.9835690831730125e-05, + "loss": 1.1372, + "step": 3653 + }, + { + "epoch": 0.09, + "grad_norm": 2.196189144290599, + "learning_rate": 1.9835553049650796e-05, + "loss": 1.0412, + "step": 3654 + }, + { + "epoch": 0.09, + "grad_norm": 2.170764757889105, + "learning_rate": 1.9835415210305763e-05, + "loss": 1.0975, + "step": 3655 + }, + { + "epoch": 0.09, + "grad_norm": 1.1049607579002567, + "learning_rate": 1.983527731369583e-05, + "loss": 1.0043, + "step": 3656 + }, + { + "epoch": 0.09, + "grad_norm": 2.0717528225096937, + "learning_rate": 1.98351393598218e-05, + "loss": 1.0311, + "step": 3657 + }, + { + "epoch": 0.09, + "grad_norm": 2.077079557488248, + "learning_rate": 1.9835001348684476e-05, + "loss": 1.0466, + "step": 3658 + }, + { + "epoch": 0.09, + "grad_norm": 2.256706149876867, + "learning_rate": 1.9834863280284657e-05, + "loss": 1.031, + "step": 3659 + }, + { + "epoch": 0.09, + "grad_norm": 2.440644381094504, + "learning_rate": 1.983472515462315e-05, + "loss": 1.025, + "step": 3660 + }, + { + "epoch": 0.09, + "grad_norm": 2.361395256032443, + "learning_rate": 1.983458697170076e-05, + "loss": 0.9792, + "step": 3661 + }, + { + "epoch": 0.09, + "grad_norm": 2.107978229536912, + "learning_rate": 1.9834448731518296e-05, + "loss": 1.1085, + "step": 3662 + }, + { + "epoch": 0.09, + "grad_norm": 1.1184559588925533, + "learning_rate": 1.9834310434076553e-05, + "loss": 0.979, + "step": 3663 + }, + { + "epoch": 0.09, + "grad_norm": 2.2907562974956743, + "learning_rate": 1.9834172079376342e-05, + "loss": 1.0588, + "step": 3664 + }, + { + "epoch": 0.09, + "grad_norm": 2.006151386092334, + "learning_rate": 1.983403366741847e-05, + "loss": 1.0791, + "step": 3665 + }, + { + "epoch": 0.09, + "grad_norm": 2.2661643521842603, + "learning_rate": 1.9833895198203737e-05, + "loss": 1.0088, + "step": 3666 + }, + { + "epoch": 0.09, + "grad_norm": 2.075129907587106, + "learning_rate": 1.9833756671732956e-05, + "loss": 1.0079, + "step": 3667 + }, + { + "epoch": 0.09, + "grad_norm": 2.179210018128636, + "learning_rate": 1.983361808800693e-05, + "loss": 1.0299, + "step": 3668 + }, + { + "epoch": 0.09, + "grad_norm": 1.940133025914272, + "learning_rate": 1.9833479447026464e-05, + "loss": 1.081, + "step": 3669 + }, + { + "epoch": 0.09, + "grad_norm": 1.9866999498044813, + "learning_rate": 1.983334074879237e-05, + "loss": 1.1221, + "step": 3670 + }, + { + "epoch": 0.09, + "grad_norm": 1.098787055314286, + "learning_rate": 1.9833201993305452e-05, + "loss": 0.8946, + "step": 3671 + }, + { + "epoch": 0.09, + "grad_norm": 2.184408832566538, + "learning_rate": 1.9833063180566517e-05, + "loss": 1.2668, + "step": 3672 + }, + { + "epoch": 0.09, + "grad_norm": 1.9143051208997974, + "learning_rate": 1.9832924310576377e-05, + "loss": 1.1515, + "step": 3673 + }, + { + "epoch": 0.09, + "grad_norm": 1.1744404178100187, + "learning_rate": 1.983278538333584e-05, + "loss": 1.0487, + "step": 3674 + }, + { + "epoch": 0.09, + "grad_norm": 2.668359234737554, + "learning_rate": 1.983264639884571e-05, + "loss": 1.1488, + "step": 3675 + }, + { + "epoch": 0.09, + "grad_norm": 2.0533049294797117, + "learning_rate": 1.98325073571068e-05, + "loss": 1.0614, + "step": 3676 + }, + { + "epoch": 0.09, + "grad_norm": 1.9341341923303628, + "learning_rate": 1.983236825811992e-05, + "loss": 0.9832, + "step": 3677 + }, + { + "epoch": 0.09, + "grad_norm": 2.095401253450003, + "learning_rate": 1.9832229101885882e-05, + "loss": 1.0243, + "step": 3678 + }, + { + "epoch": 0.09, + "grad_norm": 1.0733590293478414, + "learning_rate": 1.983208988840549e-05, + "loss": 0.8971, + "step": 3679 + }, + { + "epoch": 0.09, + "grad_norm": 1.892661726141482, + "learning_rate": 1.9831950617679562e-05, + "loss": 1.0996, + "step": 3680 + }, + { + "epoch": 0.09, + "grad_norm": 2.4974026281707236, + "learning_rate": 1.98318112897089e-05, + "loss": 1.0277, + "step": 3681 + }, + { + "epoch": 0.09, + "grad_norm": 1.953048182526572, + "learning_rate": 1.983167190449432e-05, + "loss": 1.0926, + "step": 3682 + }, + { + "epoch": 0.09, + "grad_norm": 2.233388940455699, + "learning_rate": 1.9831532462036634e-05, + "loss": 1.0474, + "step": 3683 + }, + { + "epoch": 0.09, + "grad_norm": 1.9602575459016767, + "learning_rate": 1.983139296233666e-05, + "loss": 1.2112, + "step": 3684 + }, + { + "epoch": 0.09, + "grad_norm": 2.5252739299621627, + "learning_rate": 1.9831253405395196e-05, + "loss": 1.1521, + "step": 3685 + }, + { + "epoch": 0.09, + "grad_norm": 1.974759630529257, + "learning_rate": 1.9831113791213062e-05, + "loss": 0.9153, + "step": 3686 + }, + { + "epoch": 0.09, + "grad_norm": 2.205598037839999, + "learning_rate": 1.9830974119791075e-05, + "loss": 1.0094, + "step": 3687 + }, + { + "epoch": 0.09, + "grad_norm": 2.12453218467541, + "learning_rate": 1.9830834391130043e-05, + "loss": 1.0016, + "step": 3688 + }, + { + "epoch": 0.09, + "grad_norm": 2.0893343501841235, + "learning_rate": 1.983069460523078e-05, + "loss": 1.0704, + "step": 3689 + }, + { + "epoch": 0.09, + "grad_norm": 2.1608260301555386, + "learning_rate": 1.98305547620941e-05, + "loss": 1.1017, + "step": 3690 + }, + { + "epoch": 0.09, + "grad_norm": 2.0118719138707712, + "learning_rate": 1.9830414861720818e-05, + "loss": 1.061, + "step": 3691 + }, + { + "epoch": 0.09, + "grad_norm": 1.1214270582237835, + "learning_rate": 1.9830274904111754e-05, + "loss": 1.038, + "step": 3692 + }, + { + "epoch": 0.09, + "grad_norm": 2.05783477170488, + "learning_rate": 1.983013488926771e-05, + "loss": 0.9961, + "step": 3693 + }, + { + "epoch": 0.09, + "grad_norm": 2.0344483533035844, + "learning_rate": 1.9829994817189512e-05, + "loss": 0.9407, + "step": 3694 + }, + { + "epoch": 0.09, + "grad_norm": 2.0241486303598952, + "learning_rate": 1.982985468787797e-05, + "loss": 1.0628, + "step": 3695 + }, + { + "epoch": 0.09, + "grad_norm": 2.0795303936557854, + "learning_rate": 1.9829714501333905e-05, + "loss": 1.1156, + "step": 3696 + }, + { + "epoch": 0.09, + "grad_norm": 1.9632226457958917, + "learning_rate": 1.9829574257558128e-05, + "loss": 1.0118, + "step": 3697 + }, + { + "epoch": 0.09, + "grad_norm": 2.071835847460585, + "learning_rate": 1.982943395655146e-05, + "loss": 1.1984, + "step": 3698 + }, + { + "epoch": 0.09, + "grad_norm": 2.2221440824433167, + "learning_rate": 1.982929359831471e-05, + "loss": 0.9971, + "step": 3699 + }, + { + "epoch": 0.09, + "grad_norm": 2.645514579829773, + "learning_rate": 1.9829153182848705e-05, + "loss": 1.1423, + "step": 3700 + }, + { + "epoch": 0.09, + "grad_norm": 5.225150534554131, + "learning_rate": 1.9829012710154258e-05, + "loss": 1.1606, + "step": 3701 + }, + { + "epoch": 0.09, + "grad_norm": 2.518607994519784, + "learning_rate": 1.9828872180232185e-05, + "loss": 0.9856, + "step": 3702 + }, + { + "epoch": 0.09, + "grad_norm": 2.3586243945724763, + "learning_rate": 1.982873159308331e-05, + "loss": 1.1549, + "step": 3703 + }, + { + "epoch": 0.09, + "grad_norm": 2.255059919262335, + "learning_rate": 1.9828590948708446e-05, + "loss": 1.1812, + "step": 3704 + }, + { + "epoch": 0.09, + "grad_norm": 2.9581740851471072, + "learning_rate": 1.9828450247108413e-05, + "loss": 0.9313, + "step": 3705 + }, + { + "epoch": 0.09, + "grad_norm": 2.0343258797640074, + "learning_rate": 1.982830948828403e-05, + "loss": 0.9795, + "step": 3706 + }, + { + "epoch": 0.09, + "grad_norm": 1.9244533609624652, + "learning_rate": 1.9828168672236125e-05, + "loss": 1.0723, + "step": 3707 + }, + { + "epoch": 0.09, + "grad_norm": 3.371995502369009, + "learning_rate": 1.9828027798965502e-05, + "loss": 1.1379, + "step": 3708 + }, + { + "epoch": 0.09, + "grad_norm": 2.1027583642633254, + "learning_rate": 1.9827886868472993e-05, + "loss": 1.0963, + "step": 3709 + }, + { + "epoch": 0.09, + "grad_norm": 2.2622240103148865, + "learning_rate": 1.9827745880759414e-05, + "loss": 1.0003, + "step": 3710 + }, + { + "epoch": 0.09, + "grad_norm": 1.195825611863724, + "learning_rate": 1.982760483582559e-05, + "loss": 0.9976, + "step": 3711 + }, + { + "epoch": 0.09, + "grad_norm": 1.1752306062249567, + "learning_rate": 1.9827463733672334e-05, + "loss": 1.0355, + "step": 3712 + }, + { + "epoch": 0.09, + "grad_norm": 2.1194941381842964, + "learning_rate": 1.9827322574300477e-05, + "loss": 1.0969, + "step": 3713 + }, + { + "epoch": 0.09, + "grad_norm": 2.3548896363909253, + "learning_rate": 1.9827181357710837e-05, + "loss": 1.0818, + "step": 3714 + }, + { + "epoch": 0.09, + "grad_norm": 1.1661199648018432, + "learning_rate": 1.9827040083904236e-05, + "loss": 0.9596, + "step": 3715 + }, + { + "epoch": 0.09, + "grad_norm": 1.9867296760099635, + "learning_rate": 1.9826898752881494e-05, + "loss": 1.1895, + "step": 3716 + }, + { + "epoch": 0.09, + "grad_norm": 2.536495638049759, + "learning_rate": 1.9826757364643435e-05, + "loss": 1.086, + "step": 3717 + }, + { + "epoch": 0.09, + "grad_norm": 1.1853735608941367, + "learning_rate": 1.9826615919190886e-05, + "loss": 1.0401, + "step": 3718 + }, + { + "epoch": 0.09, + "grad_norm": 2.2982122306009227, + "learning_rate": 1.9826474416524667e-05, + "loss": 1.1091, + "step": 3719 + }, + { + "epoch": 0.09, + "grad_norm": 1.1403873510182074, + "learning_rate": 1.9826332856645603e-05, + "loss": 0.922, + "step": 3720 + }, + { + "epoch": 0.09, + "grad_norm": 2.1522407875989154, + "learning_rate": 1.9826191239554523e-05, + "loss": 1.0707, + "step": 3721 + }, + { + "epoch": 0.09, + "grad_norm": 1.9750522716763523, + "learning_rate": 1.9826049565252242e-05, + "loss": 0.9276, + "step": 3722 + }, + { + "epoch": 0.09, + "grad_norm": 2.294603979839267, + "learning_rate": 1.982590783373959e-05, + "loss": 1.0003, + "step": 3723 + }, + { + "epoch": 0.09, + "grad_norm": 2.21289266081596, + "learning_rate": 1.9825766045017395e-05, + "loss": 1.0787, + "step": 3724 + }, + { + "epoch": 0.09, + "grad_norm": 1.8580128351879692, + "learning_rate": 1.9825624199086473e-05, + "loss": 0.9837, + "step": 3725 + }, + { + "epoch": 0.09, + "grad_norm": 1.1494345156746668, + "learning_rate": 1.982548229594766e-05, + "loss": 0.944, + "step": 3726 + }, + { + "epoch": 0.09, + "grad_norm": 2.2088307402928256, + "learning_rate": 1.9825340335601776e-05, + "loss": 1.0361, + "step": 3727 + }, + { + "epoch": 0.09, + "grad_norm": 1.9479912644089052, + "learning_rate": 1.9825198318049656e-05, + "loss": 1.0387, + "step": 3728 + }, + { + "epoch": 0.09, + "grad_norm": 2.005960497679737, + "learning_rate": 1.9825056243292112e-05, + "loss": 1.0092, + "step": 3729 + }, + { + "epoch": 0.09, + "grad_norm": 2.1481377863987268, + "learning_rate": 1.9824914111329988e-05, + "loss": 0.9675, + "step": 3730 + }, + { + "epoch": 0.09, + "grad_norm": 2.065011089035002, + "learning_rate": 1.98247719221641e-05, + "loss": 1.0746, + "step": 3731 + }, + { + "epoch": 0.09, + "grad_norm": 2.574500570659089, + "learning_rate": 1.982462967579528e-05, + "loss": 1.2162, + "step": 3732 + }, + { + "epoch": 0.09, + "grad_norm": 1.1266571677744386, + "learning_rate": 1.9824487372224354e-05, + "loss": 0.9388, + "step": 3733 + }, + { + "epoch": 0.09, + "grad_norm": 2.0453455828966516, + "learning_rate": 1.9824345011452152e-05, + "loss": 1.0795, + "step": 3734 + }, + { + "epoch": 0.09, + "grad_norm": 2.0822841623081043, + "learning_rate": 1.9824202593479507e-05, + "loss": 1.0463, + "step": 3735 + }, + { + "epoch": 0.09, + "grad_norm": 2.0977896378461307, + "learning_rate": 1.982406011830724e-05, + "loss": 1.1634, + "step": 3736 + }, + { + "epoch": 0.09, + "grad_norm": 1.1930162108878795, + "learning_rate": 1.9823917585936185e-05, + "loss": 0.965, + "step": 3737 + }, + { + "epoch": 0.09, + "grad_norm": 2.016822767941156, + "learning_rate": 1.9823774996367177e-05, + "loss": 1.035, + "step": 3738 + }, + { + "epoch": 0.09, + "grad_norm": 1.9355807573276456, + "learning_rate": 1.9823632349601033e-05, + "loss": 1.0643, + "step": 3739 + }, + { + "epoch": 0.09, + "grad_norm": 1.086658180540199, + "learning_rate": 1.9823489645638596e-05, + "loss": 0.9498, + "step": 3740 + }, + { + "epoch": 0.09, + "grad_norm": 2.251045991608819, + "learning_rate": 1.9823346884480694e-05, + "loss": 1.1122, + "step": 3741 + }, + { + "epoch": 0.09, + "grad_norm": 4.724840149062001, + "learning_rate": 1.982320406612815e-05, + "loss": 0.9446, + "step": 3742 + }, + { + "epoch": 0.09, + "grad_norm": 3.0082464522044114, + "learning_rate": 1.9823061190581807e-05, + "loss": 1.002, + "step": 3743 + }, + { + "epoch": 0.09, + "grad_norm": 1.181043237225652, + "learning_rate": 1.982291825784249e-05, + "loss": 0.8815, + "step": 3744 + }, + { + "epoch": 0.09, + "grad_norm": 1.9545625360255174, + "learning_rate": 1.9822775267911033e-05, + "loss": 1.0237, + "step": 3745 + }, + { + "epoch": 0.09, + "grad_norm": 2.0492645075999834, + "learning_rate": 1.982263222078827e-05, + "loss": 0.9731, + "step": 3746 + }, + { + "epoch": 0.09, + "grad_norm": 2.7306091203613114, + "learning_rate": 1.982248911647503e-05, + "loss": 1.0713, + "step": 3747 + }, + { + "epoch": 0.09, + "grad_norm": 2.2345685399870923, + "learning_rate": 1.982234595497215e-05, + "loss": 1.1281, + "step": 3748 + }, + { + "epoch": 0.09, + "grad_norm": 2.9465166564459104, + "learning_rate": 1.9822202736280462e-05, + "loss": 1.003, + "step": 3749 + }, + { + "epoch": 0.09, + "grad_norm": 2.0812153150074644, + "learning_rate": 1.9822059460400804e-05, + "loss": 1.1221, + "step": 3750 + }, + { + "epoch": 0.09, + "grad_norm": 2.0463626759353937, + "learning_rate": 1.9821916127334e-05, + "loss": 0.9955, + "step": 3751 + }, + { + "epoch": 0.09, + "grad_norm": 1.9942107142769825, + "learning_rate": 1.9821772737080892e-05, + "loss": 1.0785, + "step": 3752 + }, + { + "epoch": 0.09, + "grad_norm": 2.1461824248339867, + "learning_rate": 1.9821629289642316e-05, + "loss": 0.9634, + "step": 3753 + }, + { + "epoch": 0.09, + "grad_norm": 2.0078367553490826, + "learning_rate": 1.9821485785019103e-05, + "loss": 1.0379, + "step": 3754 + }, + { + "epoch": 0.09, + "grad_norm": 1.9902935400658108, + "learning_rate": 1.982134222321209e-05, + "loss": 0.9285, + "step": 3755 + }, + { + "epoch": 0.09, + "grad_norm": 2.176918911395523, + "learning_rate": 1.9821198604222114e-05, + "loss": 0.9575, + "step": 3756 + }, + { + "epoch": 0.09, + "grad_norm": 4.071780729927131, + "learning_rate": 1.9821054928050012e-05, + "loss": 1.03, + "step": 3757 + }, + { + "epoch": 0.09, + "grad_norm": 2.027519657610626, + "learning_rate": 1.9820911194696613e-05, + "loss": 1.1128, + "step": 3758 + }, + { + "epoch": 0.09, + "grad_norm": 2.26395053255114, + "learning_rate": 1.9820767404162765e-05, + "loss": 1.0613, + "step": 3759 + }, + { + "epoch": 0.09, + "grad_norm": 2.166203586460821, + "learning_rate": 1.98206235564493e-05, + "loss": 0.9981, + "step": 3760 + }, + { + "epoch": 0.09, + "grad_norm": 1.9072892084248878, + "learning_rate": 1.9820479651557056e-05, + "loss": 0.9696, + "step": 3761 + }, + { + "epoch": 0.09, + "grad_norm": 1.9810361023660787, + "learning_rate": 1.982033568948687e-05, + "loss": 1.1428, + "step": 3762 + }, + { + "epoch": 0.09, + "grad_norm": 2.250065001548004, + "learning_rate": 1.982019167023958e-05, + "loss": 1.1023, + "step": 3763 + }, + { + "epoch": 0.09, + "grad_norm": 2.3180549427928128, + "learning_rate": 1.9820047593816024e-05, + "loss": 1.0082, + "step": 3764 + }, + { + "epoch": 0.09, + "grad_norm": 1.911484255588411, + "learning_rate": 1.981990346021704e-05, + "loss": 1.0866, + "step": 3765 + }, + { + "epoch": 0.09, + "grad_norm": 2.2105465040292547, + "learning_rate": 1.9819759269443475e-05, + "loss": 1.1579, + "step": 3766 + }, + { + "epoch": 0.09, + "grad_norm": 1.9636128312807621, + "learning_rate": 1.9819615021496158e-05, + "loss": 1.1398, + "step": 3767 + }, + { + "epoch": 0.09, + "grad_norm": 2.4189737034911465, + "learning_rate": 1.9819470716375933e-05, + "loss": 0.999, + "step": 3768 + }, + { + "epoch": 0.09, + "grad_norm": 2.547739856606222, + "learning_rate": 1.9819326354083644e-05, + "loss": 0.9377, + "step": 3769 + }, + { + "epoch": 0.09, + "grad_norm": 2.5216656037274823, + "learning_rate": 1.9819181934620128e-05, + "loss": 1.0437, + "step": 3770 + }, + { + "epoch": 0.09, + "grad_norm": 2.2424481352364016, + "learning_rate": 1.9819037457986224e-05, + "loss": 1.1491, + "step": 3771 + }, + { + "epoch": 0.09, + "grad_norm": 2.035811779959456, + "learning_rate": 1.9818892924182774e-05, + "loss": 1.026, + "step": 3772 + }, + { + "epoch": 0.09, + "grad_norm": 1.1582670184116906, + "learning_rate": 1.9818748333210624e-05, + "loss": 0.9907, + "step": 3773 + }, + { + "epoch": 0.09, + "grad_norm": 2.357741103567031, + "learning_rate": 1.981860368507061e-05, + "loss": 1.0258, + "step": 3774 + }, + { + "epoch": 0.09, + "grad_norm": 2.0876740272131737, + "learning_rate": 1.981845897976358e-05, + "loss": 1.2125, + "step": 3775 + }, + { + "epoch": 0.09, + "grad_norm": 1.9543391285391971, + "learning_rate": 1.9818314217290373e-05, + "loss": 1.1312, + "step": 3776 + }, + { + "epoch": 0.09, + "grad_norm": 2.0388175074516512, + "learning_rate": 1.981816939765183e-05, + "loss": 1.0662, + "step": 3777 + }, + { + "epoch": 0.09, + "grad_norm": 2.41644231918578, + "learning_rate": 1.9818024520848797e-05, + "loss": 1.0965, + "step": 3778 + }, + { + "epoch": 0.09, + "grad_norm": 2.0179802524660286, + "learning_rate": 1.9817879586882117e-05, + "loss": 1.0972, + "step": 3779 + }, + { + "epoch": 0.09, + "grad_norm": 3.073818284311175, + "learning_rate": 1.9817734595752632e-05, + "loss": 0.9827, + "step": 3780 + }, + { + "epoch": 0.09, + "grad_norm": 1.191673696509262, + "learning_rate": 1.9817589547461193e-05, + "loss": 0.9926, + "step": 3781 + }, + { + "epoch": 0.09, + "grad_norm": 2.618113546709386, + "learning_rate": 1.9817444442008634e-05, + "loss": 1.0607, + "step": 3782 + }, + { + "epoch": 0.09, + "grad_norm": 2.63224455973479, + "learning_rate": 1.981729927939581e-05, + "loss": 0.9382, + "step": 3783 + }, + { + "epoch": 0.09, + "grad_norm": 1.955096408237366, + "learning_rate": 1.981715405962356e-05, + "loss": 1.1058, + "step": 3784 + }, + { + "epoch": 0.09, + "grad_norm": 2.29290677688315, + "learning_rate": 1.981700878269273e-05, + "loss": 1.0328, + "step": 3785 + }, + { + "epoch": 0.09, + "grad_norm": 2.054286199112817, + "learning_rate": 1.9816863448604165e-05, + "loss": 1.1642, + "step": 3786 + }, + { + "epoch": 0.09, + "grad_norm": 2.197075233370031, + "learning_rate": 1.9816718057358715e-05, + "loss": 1.0624, + "step": 3787 + }, + { + "epoch": 0.09, + "grad_norm": 2.1172740314191905, + "learning_rate": 1.9816572608957227e-05, + "loss": 1.0921, + "step": 3788 + }, + { + "epoch": 0.09, + "grad_norm": 2.124373564253476, + "learning_rate": 1.981642710340054e-05, + "loss": 0.9915, + "step": 3789 + }, + { + "epoch": 0.09, + "grad_norm": 2.61218617384345, + "learning_rate": 1.981628154068951e-05, + "loss": 1.2848, + "step": 3790 + }, + { + "epoch": 0.09, + "grad_norm": 2.1489696240346157, + "learning_rate": 1.981613592082498e-05, + "loss": 1.0452, + "step": 3791 + }, + { + "epoch": 0.09, + "grad_norm": 2.1678436009052073, + "learning_rate": 1.98159902438078e-05, + "loss": 0.9586, + "step": 3792 + }, + { + "epoch": 0.09, + "grad_norm": 2.9691829592339496, + "learning_rate": 1.9815844509638817e-05, + "loss": 0.9989, + "step": 3793 + }, + { + "epoch": 0.09, + "grad_norm": 2.154068299757947, + "learning_rate": 1.981569871831888e-05, + "loss": 1.1002, + "step": 3794 + }, + { + "epoch": 0.09, + "grad_norm": 2.2137284724448825, + "learning_rate": 1.9815552869848836e-05, + "loss": 1.116, + "step": 3795 + }, + { + "epoch": 0.09, + "grad_norm": 1.198934755476608, + "learning_rate": 1.9815406964229537e-05, + "loss": 1.0016, + "step": 3796 + }, + { + "epoch": 0.09, + "grad_norm": 2.1446488129766386, + "learning_rate": 1.981526100146183e-05, + "loss": 1.122, + "step": 3797 + }, + { + "epoch": 0.09, + "grad_norm": 2.2405366301858503, + "learning_rate": 1.9815114981546564e-05, + "loss": 1.0965, + "step": 3798 + }, + { + "epoch": 0.09, + "grad_norm": 2.146068169812311, + "learning_rate": 1.9814968904484595e-05, + "loss": 1.0788, + "step": 3799 + }, + { + "epoch": 0.09, + "grad_norm": 2.0807927703903735, + "learning_rate": 1.981482277027677e-05, + "loss": 1.0133, + "step": 3800 + }, + { + "epoch": 0.09, + "grad_norm": 2.064891318798234, + "learning_rate": 1.9814676578923936e-05, + "loss": 1.1095, + "step": 3801 + }, + { + "epoch": 0.09, + "grad_norm": 1.8663320830182741, + "learning_rate": 1.981453033042695e-05, + "loss": 1.1478, + "step": 3802 + }, + { + "epoch": 0.09, + "grad_norm": 2.1240393552512553, + "learning_rate": 1.981438402478666e-05, + "loss": 1.0848, + "step": 3803 + }, + { + "epoch": 0.09, + "grad_norm": 2.1431402458573454, + "learning_rate": 1.981423766200392e-05, + "loss": 1.0151, + "step": 3804 + }, + { + "epoch": 0.09, + "grad_norm": 2.0267338455494714, + "learning_rate": 1.9814091242079583e-05, + "loss": 1.1382, + "step": 3805 + }, + { + "epoch": 0.09, + "grad_norm": 2.7457351408025454, + "learning_rate": 1.98139447650145e-05, + "loss": 1.176, + "step": 3806 + }, + { + "epoch": 0.09, + "grad_norm": 2.0366757046482373, + "learning_rate": 1.981379823080952e-05, + "loss": 1.07, + "step": 3807 + }, + { + "epoch": 0.09, + "grad_norm": 1.951118427099209, + "learning_rate": 1.98136516394655e-05, + "loss": 1.1237, + "step": 3808 + }, + { + "epoch": 0.09, + "grad_norm": 2.044015469465662, + "learning_rate": 1.9813504990983293e-05, + "loss": 1.0894, + "step": 3809 + }, + { + "epoch": 0.09, + "grad_norm": 2.4404443181901816, + "learning_rate": 1.9813358285363754e-05, + "loss": 1.1173, + "step": 3810 + }, + { + "epoch": 0.09, + "grad_norm": 2.105141282381784, + "learning_rate": 1.981321152260774e-05, + "loss": 0.9944, + "step": 3811 + }, + { + "epoch": 0.09, + "grad_norm": 2.109350650810179, + "learning_rate": 1.9813064702716094e-05, + "loss": 1.0322, + "step": 3812 + }, + { + "epoch": 0.09, + "grad_norm": 1.196705790746021, + "learning_rate": 1.9812917825689683e-05, + "loss": 0.9454, + "step": 3813 + }, + { + "epoch": 0.09, + "grad_norm": 1.9532567445906288, + "learning_rate": 1.981277089152936e-05, + "loss": 1.1023, + "step": 3814 + }, + { + "epoch": 0.09, + "grad_norm": 2.2511658123503318, + "learning_rate": 1.981262390023597e-05, + "loss": 1.1282, + "step": 3815 + }, + { + "epoch": 0.09, + "grad_norm": 1.1249774497639753, + "learning_rate": 1.9812476851810384e-05, + "loss": 0.91, + "step": 3816 + }, + { + "epoch": 0.09, + "grad_norm": 1.1279533833743587, + "learning_rate": 1.981232974625345e-05, + "loss": 1.0492, + "step": 3817 + }, + { + "epoch": 0.09, + "grad_norm": 1.9076522592804641, + "learning_rate": 1.9812182583566026e-05, + "loss": 1.0976, + "step": 3818 + }, + { + "epoch": 0.09, + "grad_norm": 3.034847346441786, + "learning_rate": 1.981203536374897e-05, + "loss": 1.1045, + "step": 3819 + }, + { + "epoch": 0.09, + "grad_norm": 2.722423765027377, + "learning_rate": 1.9811888086803137e-05, + "loss": 1.0491, + "step": 3820 + }, + { + "epoch": 0.09, + "grad_norm": 2.0056211259607952, + "learning_rate": 1.981174075272938e-05, + "loss": 1.0733, + "step": 3821 + }, + { + "epoch": 0.09, + "grad_norm": 2.6373319459426185, + "learning_rate": 1.9811593361528566e-05, + "loss": 1.1418, + "step": 3822 + }, + { + "epoch": 0.09, + "grad_norm": 3.115017969064956, + "learning_rate": 1.981144591320155e-05, + "loss": 0.9094, + "step": 3823 + }, + { + "epoch": 0.09, + "grad_norm": 1.8658817052998384, + "learning_rate": 1.9811298407749188e-05, + "loss": 0.9912, + "step": 3824 + }, + { + "epoch": 0.09, + "grad_norm": 2.2546654616511073, + "learning_rate": 1.9811150845172342e-05, + "loss": 1.0425, + "step": 3825 + }, + { + "epoch": 0.09, + "grad_norm": 2.200524780826677, + "learning_rate": 1.981100322547187e-05, + "loss": 1.0767, + "step": 3826 + }, + { + "epoch": 0.09, + "grad_norm": 1.8485215703237357, + "learning_rate": 1.9810855548648628e-05, + "loss": 1.0106, + "step": 3827 + }, + { + "epoch": 0.09, + "grad_norm": 2.1596420927134097, + "learning_rate": 1.981070781470348e-05, + "loss": 1.2415, + "step": 3828 + }, + { + "epoch": 0.09, + "grad_norm": 2.191543823720256, + "learning_rate": 1.9810560023637285e-05, + "loss": 1.0744, + "step": 3829 + }, + { + "epoch": 0.09, + "grad_norm": 2.2167403889278035, + "learning_rate": 1.9810412175450907e-05, + "loss": 1.1312, + "step": 3830 + }, + { + "epoch": 0.09, + "grad_norm": 1.9547686491484293, + "learning_rate": 1.9810264270145198e-05, + "loss": 1.1515, + "step": 3831 + }, + { + "epoch": 0.09, + "grad_norm": 2.126453550179555, + "learning_rate": 1.9810116307721028e-05, + "loss": 1.1014, + "step": 3832 + }, + { + "epoch": 0.09, + "grad_norm": 2.4129883423401166, + "learning_rate": 1.9809968288179254e-05, + "loss": 1.1539, + "step": 3833 + }, + { + "epoch": 0.09, + "grad_norm": 2.072130855807915, + "learning_rate": 1.980982021152074e-05, + "loss": 0.9574, + "step": 3834 + }, + { + "epoch": 0.09, + "grad_norm": 1.8325366320651968, + "learning_rate": 1.9809672077746347e-05, + "loss": 1.0042, + "step": 3835 + }, + { + "epoch": 0.09, + "grad_norm": 2.0756998684042522, + "learning_rate": 1.9809523886856937e-05, + "loss": 1.2231, + "step": 3836 + }, + { + "epoch": 0.09, + "grad_norm": 2.337878763595805, + "learning_rate": 1.9809375638853373e-05, + "loss": 1.0418, + "step": 3837 + }, + { + "epoch": 0.09, + "grad_norm": 2.665828252415526, + "learning_rate": 1.9809227333736518e-05, + "loss": 1.0246, + "step": 3838 + }, + { + "epoch": 0.09, + "grad_norm": 2.0055149172272375, + "learning_rate": 1.9809078971507234e-05, + "loss": 1.0409, + "step": 3839 + }, + { + "epoch": 0.09, + "grad_norm": 2.2915581429137193, + "learning_rate": 1.980893055216639e-05, + "loss": 1.0331, + "step": 3840 + }, + { + "epoch": 0.09, + "grad_norm": 2.2158249395388583, + "learning_rate": 1.980878207571485e-05, + "loss": 1.1329, + "step": 3841 + }, + { + "epoch": 0.09, + "grad_norm": 1.9776537302115424, + "learning_rate": 1.980863354215347e-05, + "loss": 1.11, + "step": 3842 + }, + { + "epoch": 0.09, + "grad_norm": 3.288395186268319, + "learning_rate": 1.980848495148312e-05, + "loss": 1.0521, + "step": 3843 + }, + { + "epoch": 0.09, + "grad_norm": 2.001535470241759, + "learning_rate": 1.9808336303704666e-05, + "loss": 1.0529, + "step": 3844 + }, + { + "epoch": 0.09, + "grad_norm": 2.3590797020344465, + "learning_rate": 1.9808187598818974e-05, + "loss": 1.0795, + "step": 3845 + }, + { + "epoch": 0.09, + "grad_norm": 2.125322744941544, + "learning_rate": 1.9808038836826906e-05, + "loss": 0.9521, + "step": 3846 + }, + { + "epoch": 0.09, + "grad_norm": 2.2502484833724177, + "learning_rate": 1.980789001772933e-05, + "loss": 1.036, + "step": 3847 + }, + { + "epoch": 0.09, + "grad_norm": 2.7257539885441817, + "learning_rate": 1.980774114152712e-05, + "loss": 1.0642, + "step": 3848 + }, + { + "epoch": 0.09, + "grad_norm": 1.1876898427746974, + "learning_rate": 1.980759220822113e-05, + "loss": 0.9865, + "step": 3849 + }, + { + "epoch": 0.09, + "grad_norm": 1.9534579557098182, + "learning_rate": 1.9807443217812236e-05, + "loss": 1.0325, + "step": 3850 + }, + { + "epoch": 0.09, + "grad_norm": 2.0016249754085, + "learning_rate": 1.9807294170301303e-05, + "loss": 1.0505, + "step": 3851 + }, + { + "epoch": 0.09, + "grad_norm": 2.2434919565632527, + "learning_rate": 1.9807145065689196e-05, + "loss": 1.1416, + "step": 3852 + }, + { + "epoch": 0.09, + "grad_norm": 2.322823824841589, + "learning_rate": 1.9806995903976785e-05, + "loss": 0.9518, + "step": 3853 + }, + { + "epoch": 0.09, + "grad_norm": 2.2064014099442293, + "learning_rate": 1.980684668516494e-05, + "loss": 1.0056, + "step": 3854 + }, + { + "epoch": 0.09, + "grad_norm": 2.0565975458437076, + "learning_rate": 1.9806697409254532e-05, + "loss": 0.9593, + "step": 3855 + }, + { + "epoch": 0.09, + "grad_norm": 2.610150674496005, + "learning_rate": 1.9806548076246425e-05, + "loss": 1.0637, + "step": 3856 + }, + { + "epoch": 0.09, + "grad_norm": 2.2699922782376296, + "learning_rate": 1.9806398686141487e-05, + "loss": 1.0739, + "step": 3857 + }, + { + "epoch": 0.09, + "grad_norm": 2.2000720262166933, + "learning_rate": 1.9806249238940596e-05, + "loss": 1.1063, + "step": 3858 + }, + { + "epoch": 0.09, + "grad_norm": 2.033539321680666, + "learning_rate": 1.9806099734644615e-05, + "loss": 1.0938, + "step": 3859 + }, + { + "epoch": 0.09, + "grad_norm": 2.330268116976808, + "learning_rate": 1.980595017325442e-05, + "loss": 1.0549, + "step": 3860 + }, + { + "epoch": 0.09, + "grad_norm": 2.2597734769323723, + "learning_rate": 1.9805800554770872e-05, + "loss": 1.027, + "step": 3861 + }, + { + "epoch": 0.09, + "grad_norm": 2.257628557626385, + "learning_rate": 1.980565087919485e-05, + "loss": 1.0405, + "step": 3862 + }, + { + "epoch": 0.09, + "grad_norm": 2.1154187462007514, + "learning_rate": 1.9805501146527228e-05, + "loss": 1.0973, + "step": 3863 + }, + { + "epoch": 0.09, + "grad_norm": 2.1735429147197176, + "learning_rate": 1.9805351356768872e-05, + "loss": 1.1131, + "step": 3864 + }, + { + "epoch": 0.09, + "grad_norm": 2.44191187705775, + "learning_rate": 1.980520150992066e-05, + "loss": 0.9891, + "step": 3865 + }, + { + "epoch": 0.09, + "grad_norm": 2.2105956043850647, + "learning_rate": 1.9805051605983454e-05, + "loss": 1.0247, + "step": 3866 + }, + { + "epoch": 0.09, + "grad_norm": 1.1890168356836708, + "learning_rate": 1.9804901644958137e-05, + "loss": 0.9386, + "step": 3867 + }, + { + "epoch": 0.09, + "grad_norm": 2.1272709643388956, + "learning_rate": 1.9804751626845577e-05, + "loss": 0.9708, + "step": 3868 + }, + { + "epoch": 0.09, + "grad_norm": 2.2206223466032884, + "learning_rate": 1.9804601551646652e-05, + "loss": 1.1469, + "step": 3869 + }, + { + "epoch": 0.09, + "grad_norm": 2.3505871393987534, + "learning_rate": 1.980445141936223e-05, + "loss": 1.1361, + "step": 3870 + }, + { + "epoch": 0.09, + "grad_norm": 2.292758983232116, + "learning_rate": 1.9804301229993185e-05, + "loss": 1.2096, + "step": 3871 + }, + { + "epoch": 0.09, + "grad_norm": 2.2667784045597745, + "learning_rate": 1.98041509835404e-05, + "loss": 1.1209, + "step": 3872 + }, + { + "epoch": 0.09, + "grad_norm": 2.1559203879004474, + "learning_rate": 1.9804000680004744e-05, + "loss": 1.1331, + "step": 3873 + }, + { + "epoch": 0.09, + "grad_norm": 2.2056592693080854, + "learning_rate": 1.9803850319387087e-05, + "loss": 1.1166, + "step": 3874 + }, + { + "epoch": 0.09, + "grad_norm": 2.1762456461456092, + "learning_rate": 1.9803699901688313e-05, + "loss": 1.0753, + "step": 3875 + }, + { + "epoch": 0.09, + "grad_norm": 1.7982577150572923, + "learning_rate": 1.9803549426909293e-05, + "loss": 1.0762, + "step": 3876 + }, + { + "epoch": 0.09, + "grad_norm": 2.0298410204471593, + "learning_rate": 1.9803398895050905e-05, + "loss": 1.0117, + "step": 3877 + }, + { + "epoch": 0.09, + "grad_norm": 2.5968671864971298, + "learning_rate": 1.9803248306114026e-05, + "loss": 1.2492, + "step": 3878 + }, + { + "epoch": 0.09, + "grad_norm": 2.378478882619842, + "learning_rate": 1.980309766009953e-05, + "loss": 1.1015, + "step": 3879 + }, + { + "epoch": 0.09, + "grad_norm": 2.327176115701427, + "learning_rate": 1.9802946957008296e-05, + "loss": 0.9774, + "step": 3880 + }, + { + "epoch": 0.09, + "grad_norm": 1.153536403153507, + "learning_rate": 1.98027961968412e-05, + "loss": 1.0218, + "step": 3881 + }, + { + "epoch": 0.09, + "grad_norm": 2.010497752376644, + "learning_rate": 1.9802645379599122e-05, + "loss": 1.1232, + "step": 3882 + }, + { + "epoch": 0.09, + "grad_norm": 2.1873129204810478, + "learning_rate": 1.980249450528294e-05, + "loss": 1.0835, + "step": 3883 + }, + { + "epoch": 0.09, + "grad_norm": 2.523335292203365, + "learning_rate": 1.9802343573893536e-05, + "loss": 1.0914, + "step": 3884 + }, + { + "epoch": 0.09, + "grad_norm": 2.3352655242419074, + "learning_rate": 1.9802192585431778e-05, + "loss": 1.0791, + "step": 3885 + }, + { + "epoch": 0.09, + "grad_norm": 2.201077465035685, + "learning_rate": 1.9802041539898553e-05, + "loss": 0.9524, + "step": 3886 + }, + { + "epoch": 0.09, + "grad_norm": 2.2309513549081244, + "learning_rate": 1.980189043729474e-05, + "loss": 0.8722, + "step": 3887 + }, + { + "epoch": 0.09, + "grad_norm": 2.160190680084598, + "learning_rate": 1.9801739277621215e-05, + "loss": 1.093, + "step": 3888 + }, + { + "epoch": 0.09, + "grad_norm": 3.985138860168479, + "learning_rate": 1.980158806087886e-05, + "loss": 0.9475, + "step": 3889 + }, + { + "epoch": 0.09, + "grad_norm": 3.0952268908793235, + "learning_rate": 1.9801436787068563e-05, + "loss": 1.0668, + "step": 3890 + }, + { + "epoch": 0.09, + "grad_norm": 2.3181512255332897, + "learning_rate": 1.980128545619119e-05, + "loss": 1.1267, + "step": 3891 + }, + { + "epoch": 0.09, + "grad_norm": 1.967432477556697, + "learning_rate": 1.9801134068247636e-05, + "loss": 1.0871, + "step": 3892 + }, + { + "epoch": 0.09, + "grad_norm": 1.3567351693425425, + "learning_rate": 1.9800982623238775e-05, + "loss": 1.0044, + "step": 3893 + }, + { + "epoch": 0.09, + "grad_norm": 2.311036668366325, + "learning_rate": 1.9800831121165488e-05, + "loss": 1.0032, + "step": 3894 + }, + { + "epoch": 0.09, + "grad_norm": 1.9547229745807202, + "learning_rate": 1.9800679562028662e-05, + "loss": 0.9986, + "step": 3895 + }, + { + "epoch": 0.09, + "grad_norm": 1.1138725701365415, + "learning_rate": 1.9800527945829175e-05, + "loss": 0.9136, + "step": 3896 + }, + { + "epoch": 0.09, + "grad_norm": 1.131488503573892, + "learning_rate": 1.9800376272567913e-05, + "loss": 1.0452, + "step": 3897 + }, + { + "epoch": 0.09, + "grad_norm": 2.71869964679726, + "learning_rate": 1.9800224542245756e-05, + "loss": 1.0738, + "step": 3898 + }, + { + "epoch": 0.09, + "grad_norm": 2.274612807487952, + "learning_rate": 1.980007275486359e-05, + "loss": 1.0559, + "step": 3899 + }, + { + "epoch": 0.09, + "grad_norm": 3.338954258587049, + "learning_rate": 1.9799920910422296e-05, + "loss": 1.1357, + "step": 3900 + }, + { + "epoch": 0.09, + "grad_norm": 1.2496038195552737, + "learning_rate": 1.9799769008922762e-05, + "loss": 1.02, + "step": 3901 + }, + { + "epoch": 0.09, + "grad_norm": 1.991895383099752, + "learning_rate": 1.979961705036587e-05, + "loss": 0.9936, + "step": 3902 + }, + { + "epoch": 0.09, + "grad_norm": 2.094626427637914, + "learning_rate": 1.9799465034752505e-05, + "loss": 1.0501, + "step": 3903 + }, + { + "epoch": 0.09, + "grad_norm": 1.9961646458414757, + "learning_rate": 1.979931296208355e-05, + "loss": 1.021, + "step": 3904 + }, + { + "epoch": 0.09, + "grad_norm": 2.1954247422219724, + "learning_rate": 1.9799160832359896e-05, + "loss": 1.1574, + "step": 3905 + }, + { + "epoch": 0.09, + "grad_norm": 2.1754754515463635, + "learning_rate": 1.9799008645582424e-05, + "loss": 1.0721, + "step": 3906 + }, + { + "epoch": 0.09, + "grad_norm": 2.151267413450391, + "learning_rate": 1.9798856401752017e-05, + "loss": 0.9618, + "step": 3907 + }, + { + "epoch": 0.09, + "grad_norm": 2.2919054079449896, + "learning_rate": 1.979870410086957e-05, + "loss": 1.0487, + "step": 3908 + }, + { + "epoch": 0.09, + "grad_norm": 2.3725771524465387, + "learning_rate": 1.979855174293597e-05, + "loss": 1.1549, + "step": 3909 + }, + { + "epoch": 0.09, + "grad_norm": 1.2553825687624385, + "learning_rate": 1.9798399327952093e-05, + "loss": 0.9659, + "step": 3910 + }, + { + "epoch": 0.09, + "grad_norm": 2.240748913520293, + "learning_rate": 1.9798246855918833e-05, + "loss": 0.9833, + "step": 3911 + }, + { + "epoch": 0.09, + "grad_norm": 2.2433888486109703, + "learning_rate": 1.979809432683708e-05, + "loss": 0.999, + "step": 3912 + }, + { + "epoch": 0.09, + "grad_norm": 2.07923057329691, + "learning_rate": 1.9797941740707718e-05, + "loss": 1.2397, + "step": 3913 + }, + { + "epoch": 0.09, + "grad_norm": 2.147357029805787, + "learning_rate": 1.9797789097531638e-05, + "loss": 1.1336, + "step": 3914 + }, + { + "epoch": 0.09, + "grad_norm": 1.2037503437206882, + "learning_rate": 1.979763639730973e-05, + "loss": 0.9098, + "step": 3915 + }, + { + "epoch": 0.09, + "grad_norm": 2.0046606992961546, + "learning_rate": 1.979748364004288e-05, + "loss": 0.9031, + "step": 3916 + }, + { + "epoch": 0.09, + "grad_norm": 2.1507480491685165, + "learning_rate": 1.9797330825731976e-05, + "loss": 1.0328, + "step": 3917 + }, + { + "epoch": 0.09, + "grad_norm": 2.309058890768232, + "learning_rate": 1.979717795437791e-05, + "loss": 1.0904, + "step": 3918 + }, + { + "epoch": 0.09, + "grad_norm": 2.015255415239652, + "learning_rate": 1.9797025025981576e-05, + "loss": 1.1724, + "step": 3919 + }, + { + "epoch": 0.09, + "grad_norm": 1.9847431168188385, + "learning_rate": 1.979687204054386e-05, + "loss": 1.0019, + "step": 3920 + }, + { + "epoch": 0.09, + "grad_norm": 1.9753227956685642, + "learning_rate": 1.9796718998065653e-05, + "loss": 1.0374, + "step": 3921 + }, + { + "epoch": 0.09, + "grad_norm": 2.111435817535685, + "learning_rate": 1.9796565898547845e-05, + "loss": 1.0379, + "step": 3922 + }, + { + "epoch": 0.09, + "grad_norm": 2.454916795579142, + "learning_rate": 1.979641274199133e-05, + "loss": 1.1922, + "step": 3923 + }, + { + "epoch": 0.09, + "grad_norm": 2.3403014609435226, + "learning_rate": 1.9796259528396998e-05, + "loss": 1.1549, + "step": 3924 + }, + { + "epoch": 0.09, + "grad_norm": 2.17848564669055, + "learning_rate": 1.979610625776574e-05, + "loss": 1.2196, + "step": 3925 + }, + { + "epoch": 0.09, + "grad_norm": 2.1995681537858482, + "learning_rate": 1.9795952930098453e-05, + "loss": 1.1121, + "step": 3926 + }, + { + "epoch": 0.09, + "grad_norm": 2.714801050697697, + "learning_rate": 1.9795799545396022e-05, + "loss": 1.0098, + "step": 3927 + }, + { + "epoch": 0.09, + "grad_norm": 2.0042910750705163, + "learning_rate": 1.9795646103659346e-05, + "loss": 1.0242, + "step": 3928 + }, + { + "epoch": 0.09, + "grad_norm": 2.1071970342857864, + "learning_rate": 1.9795492604889323e-05, + "loss": 0.9934, + "step": 3929 + }, + { + "epoch": 0.09, + "grad_norm": 1.9255321297545511, + "learning_rate": 1.9795339049086837e-05, + "loss": 1.0686, + "step": 3930 + }, + { + "epoch": 0.09, + "grad_norm": 2.2108417450654745, + "learning_rate": 1.9795185436252785e-05, + "loss": 1.154, + "step": 3931 + }, + { + "epoch": 0.09, + "grad_norm": 2.033602220748086, + "learning_rate": 1.9795031766388063e-05, + "loss": 0.9254, + "step": 3932 + }, + { + "epoch": 0.09, + "grad_norm": 2.0880893915090915, + "learning_rate": 1.9794878039493564e-05, + "loss": 1.1931, + "step": 3933 + }, + { + "epoch": 0.09, + "grad_norm": 2.7317696261354034, + "learning_rate": 1.9794724255570187e-05, + "loss": 1.1001, + "step": 3934 + }, + { + "epoch": 0.09, + "grad_norm": 3.413619037192627, + "learning_rate": 1.9794570414618823e-05, + "loss": 1.003, + "step": 3935 + }, + { + "epoch": 0.09, + "grad_norm": 2.27332756440929, + "learning_rate": 1.9794416516640365e-05, + "loss": 1.0474, + "step": 3936 + }, + { + "epoch": 0.09, + "grad_norm": 1.9673226630582978, + "learning_rate": 1.9794262561635716e-05, + "loss": 1.1498, + "step": 3937 + }, + { + "epoch": 0.09, + "grad_norm": 3.113014861332449, + "learning_rate": 1.9794108549605774e-05, + "loss": 1.1317, + "step": 3938 + }, + { + "epoch": 0.09, + "grad_norm": 2.0049989272309316, + "learning_rate": 1.979395448055143e-05, + "loss": 1.0591, + "step": 3939 + }, + { + "epoch": 0.09, + "grad_norm": 1.8975335319304631, + "learning_rate": 1.9793800354473575e-05, + "loss": 1.0559, + "step": 3940 + }, + { + "epoch": 0.09, + "grad_norm": 2.066224515333744, + "learning_rate": 1.979364617137312e-05, + "loss": 1.1502, + "step": 3941 + }, + { + "epoch": 0.09, + "grad_norm": 2.039692721040945, + "learning_rate": 1.9793491931250955e-05, + "loss": 1.1935, + "step": 3942 + }, + { + "epoch": 0.09, + "grad_norm": 2.082475094695674, + "learning_rate": 1.9793337634107977e-05, + "loss": 1.0545, + "step": 3943 + }, + { + "epoch": 0.09, + "grad_norm": 2.073539475950439, + "learning_rate": 1.9793183279945093e-05, + "loss": 1.0757, + "step": 3944 + }, + { + "epoch": 0.09, + "grad_norm": 1.1296692475747774, + "learning_rate": 1.979302886876319e-05, + "loss": 0.9838, + "step": 3945 + }, + { + "epoch": 0.09, + "grad_norm": 2.3101704258358047, + "learning_rate": 1.9792874400563172e-05, + "loss": 1.1176, + "step": 3946 + }, + { + "epoch": 0.09, + "grad_norm": 2.3779812347026223, + "learning_rate": 1.979271987534594e-05, + "loss": 1.0855, + "step": 3947 + }, + { + "epoch": 0.09, + "grad_norm": 2.1207545248053834, + "learning_rate": 1.9792565293112396e-05, + "loss": 1.1264, + "step": 3948 + }, + { + "epoch": 0.09, + "grad_norm": 2.151625515241651, + "learning_rate": 1.9792410653863436e-05, + "loss": 1.0987, + "step": 3949 + }, + { + "epoch": 0.09, + "grad_norm": 2.308220088901729, + "learning_rate": 1.9792255957599956e-05, + "loss": 1.0359, + "step": 3950 + }, + { + "epoch": 0.09, + "grad_norm": 2.165094924101997, + "learning_rate": 1.9792101204322866e-05, + "loss": 1.0472, + "step": 3951 + }, + { + "epoch": 0.09, + "grad_norm": 2.065010283095618, + "learning_rate": 1.979194639403306e-05, + "loss": 1.1444, + "step": 3952 + }, + { + "epoch": 0.09, + "grad_norm": 2.337110342191011, + "learning_rate": 1.9791791526731444e-05, + "loss": 0.9873, + "step": 3953 + }, + { + "epoch": 0.09, + "grad_norm": 2.081654019964984, + "learning_rate": 1.9791636602418916e-05, + "loss": 1.2181, + "step": 3954 + }, + { + "epoch": 0.09, + "grad_norm": 1.83550077707426, + "learning_rate": 1.979148162109638e-05, + "loss": 0.8835, + "step": 3955 + }, + { + "epoch": 0.09, + "grad_norm": 7.923753124783135, + "learning_rate": 1.9791326582764742e-05, + "loss": 0.9951, + "step": 3956 + }, + { + "epoch": 0.09, + "grad_norm": 2.8133241877990827, + "learning_rate": 1.9791171487424895e-05, + "loss": 1.094, + "step": 3957 + }, + { + "epoch": 0.09, + "grad_norm": 2.3020233467249644, + "learning_rate": 1.979101633507775e-05, + "loss": 1.1691, + "step": 3958 + }, + { + "epoch": 0.09, + "grad_norm": 1.7995789734219552, + "learning_rate": 1.979086112572421e-05, + "loss": 1.124, + "step": 3959 + }, + { + "epoch": 0.09, + "grad_norm": 2.1984275094085217, + "learning_rate": 1.9790705859365178e-05, + "loss": 1.2428, + "step": 3960 + }, + { + "epoch": 0.09, + "grad_norm": 1.2421509983664711, + "learning_rate": 1.9790550536001554e-05, + "loss": 0.9804, + "step": 3961 + }, + { + "epoch": 0.09, + "grad_norm": 2.1289348869410025, + "learning_rate": 1.9790395155634243e-05, + "loss": 1.1807, + "step": 3962 + }, + { + "epoch": 0.09, + "grad_norm": 2.2708819675824374, + "learning_rate": 1.9790239718264155e-05, + "loss": 0.9233, + "step": 3963 + }, + { + "epoch": 0.09, + "grad_norm": 2.2776878323556713, + "learning_rate": 1.9790084223892192e-05, + "loss": 1.0396, + "step": 3964 + }, + { + "epoch": 0.09, + "grad_norm": 2.25190355861735, + "learning_rate": 1.978992867251926e-05, + "loss": 1.157, + "step": 3965 + }, + { + "epoch": 0.09, + "grad_norm": 2.14984835861883, + "learning_rate": 1.9789773064146257e-05, + "loss": 1.0919, + "step": 3966 + }, + { + "epoch": 0.09, + "grad_norm": 2.2890965147754185, + "learning_rate": 1.97896173987741e-05, + "loss": 1.0574, + "step": 3967 + }, + { + "epoch": 0.09, + "grad_norm": 1.9795029879380281, + "learning_rate": 1.9789461676403694e-05, + "loss": 1.0565, + "step": 3968 + }, + { + "epoch": 0.09, + "grad_norm": 1.9764840628539915, + "learning_rate": 1.9789305897035942e-05, + "loss": 1.0406, + "step": 3969 + }, + { + "epoch": 0.09, + "grad_norm": 2.5032396865019564, + "learning_rate": 1.978915006067175e-05, + "loss": 1.0503, + "step": 3970 + }, + { + "epoch": 0.09, + "grad_norm": 2.3050733419086535, + "learning_rate": 1.978899416731203e-05, + "loss": 0.9949, + "step": 3971 + }, + { + "epoch": 0.09, + "grad_norm": 1.8895315549986136, + "learning_rate": 1.978883821695768e-05, + "loss": 0.9395, + "step": 3972 + }, + { + "epoch": 0.09, + "grad_norm": 2.298849673522647, + "learning_rate": 1.978868220960962e-05, + "loss": 1.1157, + "step": 3973 + }, + { + "epoch": 0.09, + "grad_norm": 2.295315278503932, + "learning_rate": 1.9788526145268752e-05, + "loss": 1.0031, + "step": 3974 + }, + { + "epoch": 0.09, + "grad_norm": 2.5734879368222505, + "learning_rate": 1.978837002393599e-05, + "loss": 1.1162, + "step": 3975 + }, + { + "epoch": 0.09, + "grad_norm": 2.1683354045143957, + "learning_rate": 1.9788213845612233e-05, + "loss": 1.0143, + "step": 3976 + }, + { + "epoch": 0.09, + "grad_norm": 2.1612728591080863, + "learning_rate": 1.9788057610298396e-05, + "loss": 1.116, + "step": 3977 + }, + { + "epoch": 0.09, + "grad_norm": 2.223690645948888, + "learning_rate": 1.9787901317995393e-05, + "loss": 1.0197, + "step": 3978 + }, + { + "epoch": 0.09, + "grad_norm": 2.3783607284151844, + "learning_rate": 1.9787744968704125e-05, + "loss": 1.1453, + "step": 3979 + }, + { + "epoch": 0.09, + "grad_norm": 2.02787417308999, + "learning_rate": 1.9787588562425514e-05, + "loss": 1.164, + "step": 3980 + }, + { + "epoch": 0.09, + "grad_norm": 1.9942036039895017, + "learning_rate": 1.9787432099160457e-05, + "loss": 1.1548, + "step": 3981 + }, + { + "epoch": 0.09, + "grad_norm": 1.975270088482155, + "learning_rate": 1.9787275578909873e-05, + "loss": 0.9223, + "step": 3982 + }, + { + "epoch": 0.09, + "grad_norm": 2.1612303440755336, + "learning_rate": 1.9787119001674677e-05, + "loss": 1.081, + "step": 3983 + }, + { + "epoch": 0.09, + "grad_norm": 1.964723690490547, + "learning_rate": 1.978696236745577e-05, + "loss": 1.2106, + "step": 3984 + }, + { + "epoch": 0.09, + "grad_norm": 2.427158393265414, + "learning_rate": 1.978680567625407e-05, + "loss": 1.0038, + "step": 3985 + }, + { + "epoch": 0.09, + "grad_norm": 1.223566357462695, + "learning_rate": 1.9786648928070493e-05, + "loss": 0.9586, + "step": 3986 + }, + { + "epoch": 0.09, + "grad_norm": 1.9389006460682008, + "learning_rate": 1.978649212290594e-05, + "loss": 0.9716, + "step": 3987 + }, + { + "epoch": 0.09, + "grad_norm": 2.3083535711293792, + "learning_rate": 1.978633526076134e-05, + "loss": 1.1013, + "step": 3988 + }, + { + "epoch": 0.09, + "grad_norm": 1.1448294240809933, + "learning_rate": 1.9786178341637594e-05, + "loss": 0.9672, + "step": 3989 + }, + { + "epoch": 0.09, + "grad_norm": 2.3246858486066984, + "learning_rate": 1.9786021365535616e-05, + "loss": 0.957, + "step": 3990 + }, + { + "epoch": 0.09, + "grad_norm": 2.248375064615984, + "learning_rate": 1.9785864332456327e-05, + "loss": 0.9761, + "step": 3991 + }, + { + "epoch": 0.09, + "grad_norm": 2.23879712700346, + "learning_rate": 1.9785707242400638e-05, + "loss": 1.0069, + "step": 3992 + }, + { + "epoch": 0.09, + "grad_norm": 2.2345776361311764, + "learning_rate": 1.9785550095369465e-05, + "loss": 1.0683, + "step": 3993 + }, + { + "epoch": 0.09, + "grad_norm": 1.85072429965082, + "learning_rate": 1.9785392891363718e-05, + "loss": 1.0087, + "step": 3994 + }, + { + "epoch": 0.09, + "grad_norm": 2.468515270088072, + "learning_rate": 1.9785235630384316e-05, + "loss": 0.9788, + "step": 3995 + }, + { + "epoch": 0.09, + "grad_norm": 2.2703763299219086, + "learning_rate": 1.9785078312432176e-05, + "loss": 0.9087, + "step": 3996 + }, + { + "epoch": 0.09, + "grad_norm": 2.295962422232596, + "learning_rate": 1.9784920937508208e-05, + "loss": 1.0749, + "step": 3997 + }, + { + "epoch": 0.09, + "grad_norm": 2.818251174507308, + "learning_rate": 1.978476350561334e-05, + "loss": 1.1861, + "step": 3998 + }, + { + "epoch": 0.09, + "grad_norm": 2.337880174992725, + "learning_rate": 1.9784606016748474e-05, + "loss": 1.1398, + "step": 3999 + }, + { + "epoch": 0.09, + "grad_norm": 2.0029775411756394, + "learning_rate": 1.9784448470914538e-05, + "loss": 0.9692, + "step": 4000 + }, + { + "epoch": 0.09, + "grad_norm": 2.293788639195551, + "learning_rate": 1.9784290868112442e-05, + "loss": 1.0508, + "step": 4001 + }, + { + "epoch": 0.09, + "grad_norm": 2.136782528033947, + "learning_rate": 1.978413320834311e-05, + "loss": 1.1096, + "step": 4002 + }, + { + "epoch": 0.09, + "grad_norm": 2.0331814650949327, + "learning_rate": 1.9783975491607457e-05, + "loss": 1.0727, + "step": 4003 + }, + { + "epoch": 0.09, + "grad_norm": 2.329531838101273, + "learning_rate": 1.9783817717906398e-05, + "loss": 0.9655, + "step": 4004 + }, + { + "epoch": 0.09, + "grad_norm": 2.0681048705289746, + "learning_rate": 1.9783659887240855e-05, + "loss": 1.054, + "step": 4005 + }, + { + "epoch": 0.09, + "grad_norm": 2.229501778109521, + "learning_rate": 1.978350199961175e-05, + "loss": 1.1729, + "step": 4006 + }, + { + "epoch": 0.09, + "grad_norm": 2.1344312190412498, + "learning_rate": 1.978334405502e-05, + "loss": 1.0886, + "step": 4007 + }, + { + "epoch": 0.09, + "grad_norm": 2.0945394457982567, + "learning_rate": 1.9783186053466518e-05, + "loss": 1.0836, + "step": 4008 + }, + { + "epoch": 0.09, + "grad_norm": 2.245778418257639, + "learning_rate": 1.9783027994952233e-05, + "loss": 1.1247, + "step": 4009 + }, + { + "epoch": 0.09, + "grad_norm": 2.2558076559471134, + "learning_rate": 1.9782869879478058e-05, + "loss": 1.0879, + "step": 4010 + }, + { + "epoch": 0.09, + "grad_norm": 2.1238427544070313, + "learning_rate": 1.9782711707044923e-05, + "loss": 0.9045, + "step": 4011 + }, + { + "epoch": 0.09, + "grad_norm": 2.4001588115854786, + "learning_rate": 1.9782553477653738e-05, + "loss": 1.0441, + "step": 4012 + }, + { + "epoch": 0.09, + "grad_norm": 1.12134906764059, + "learning_rate": 1.9782395191305436e-05, + "loss": 0.9881, + "step": 4013 + }, + { + "epoch": 0.09, + "grad_norm": 2.1615062665459814, + "learning_rate": 1.9782236848000927e-05, + "loss": 1.0758, + "step": 4014 + }, + { + "epoch": 0.09, + "grad_norm": 2.174229288187028, + "learning_rate": 1.9782078447741136e-05, + "loss": 1.3046, + "step": 4015 + }, + { + "epoch": 0.09, + "grad_norm": 2.0648283239671352, + "learning_rate": 1.9781919990526993e-05, + "loss": 1.1262, + "step": 4016 + }, + { + "epoch": 0.09, + "grad_norm": 1.120430413246485, + "learning_rate": 1.9781761476359414e-05, + "loss": 0.9957, + "step": 4017 + }, + { + "epoch": 0.09, + "grad_norm": 2.384502972261199, + "learning_rate": 1.9781602905239323e-05, + "loss": 1.1614, + "step": 4018 + }, + { + "epoch": 0.09, + "grad_norm": 1.9824511495317274, + "learning_rate": 1.978144427716764e-05, + "loss": 1.0481, + "step": 4019 + }, + { + "epoch": 0.09, + "grad_norm": 2.345442432155754, + "learning_rate": 1.9781285592145292e-05, + "loss": 1.11, + "step": 4020 + }, + { + "epoch": 0.09, + "grad_norm": 2.3574011579739436, + "learning_rate": 1.9781126850173204e-05, + "loss": 1.1923, + "step": 4021 + }, + { + "epoch": 0.09, + "grad_norm": 1.1118505826144431, + "learning_rate": 1.97809680512523e-05, + "loss": 1.0158, + "step": 4022 + }, + { + "epoch": 0.09, + "grad_norm": 1.9875618676836888, + "learning_rate": 1.9780809195383498e-05, + "loss": 0.9924, + "step": 4023 + }, + { + "epoch": 0.09, + "grad_norm": 2.0395503896316067, + "learning_rate": 1.9780650282567737e-05, + "loss": 1.1372, + "step": 4024 + }, + { + "epoch": 0.09, + "grad_norm": 1.9784017899161137, + "learning_rate": 1.9780491312805927e-05, + "loss": 1.2382, + "step": 4025 + }, + { + "epoch": 0.09, + "grad_norm": 2.1270378681996025, + "learning_rate": 1.9780332286099004e-05, + "loss": 1.1455, + "step": 4026 + }, + { + "epoch": 0.09, + "grad_norm": 2.1688705612442396, + "learning_rate": 1.9780173202447885e-05, + "loss": 1.1303, + "step": 4027 + }, + { + "epoch": 0.09, + "grad_norm": 2.308637085498196, + "learning_rate": 1.9780014061853506e-05, + "loss": 1.0963, + "step": 4028 + }, + { + "epoch": 0.09, + "grad_norm": 2.1026677109126823, + "learning_rate": 1.9779854864316784e-05, + "loss": 1.1543, + "step": 4029 + }, + { + "epoch": 0.09, + "grad_norm": 2.3708161182884053, + "learning_rate": 1.9779695609838656e-05, + "loss": 1.065, + "step": 4030 + }, + { + "epoch": 0.09, + "grad_norm": 2.512503534823542, + "learning_rate": 1.9779536298420042e-05, + "loss": 1.1367, + "step": 4031 + }, + { + "epoch": 0.09, + "grad_norm": 1.9979694579620955, + "learning_rate": 1.977937693006187e-05, + "loss": 1.088, + "step": 4032 + }, + { + "epoch": 0.1, + "grad_norm": 2.0479493007566214, + "learning_rate": 1.9779217504765072e-05, + "loss": 1.2298, + "step": 4033 + }, + { + "epoch": 0.1, + "grad_norm": 2.2718400151984173, + "learning_rate": 1.9779058022530572e-05, + "loss": 1.0786, + "step": 4034 + }, + { + "epoch": 0.1, + "grad_norm": 2.5597413328657685, + "learning_rate": 1.97788984833593e-05, + "loss": 1.0686, + "step": 4035 + }, + { + "epoch": 0.1, + "grad_norm": 2.003361996257663, + "learning_rate": 1.9778738887252187e-05, + "loss": 0.9625, + "step": 4036 + }, + { + "epoch": 0.1, + "grad_norm": 1.1792247789191412, + "learning_rate": 1.977857923421016e-05, + "loss": 1.0303, + "step": 4037 + }, + { + "epoch": 0.1, + "grad_norm": 2.110471898061161, + "learning_rate": 1.9778419524234148e-05, + "loss": 1.0394, + "step": 4038 + }, + { + "epoch": 0.1, + "grad_norm": 1.1341868925711525, + "learning_rate": 1.977825975732508e-05, + "loss": 0.9094, + "step": 4039 + }, + { + "epoch": 0.1, + "grad_norm": 2.2005295447759288, + "learning_rate": 1.977809993348389e-05, + "loss": 0.9892, + "step": 4040 + }, + { + "epoch": 0.1, + "grad_norm": 2.117923192068428, + "learning_rate": 1.9777940052711506e-05, + "loss": 1.118, + "step": 4041 + }, + { + "epoch": 0.1, + "grad_norm": 2.754423481309457, + "learning_rate": 1.977778011500886e-05, + "loss": 1.142, + "step": 4042 + }, + { + "epoch": 0.1, + "grad_norm": 1.3084551896213608, + "learning_rate": 1.977762012037688e-05, + "loss": 1.0344, + "step": 4043 + }, + { + "epoch": 0.1, + "grad_norm": 2.37565859647668, + "learning_rate": 1.9777460068816503e-05, + "loss": 0.9645, + "step": 4044 + }, + { + "epoch": 0.1, + "grad_norm": 1.9596325611598522, + "learning_rate": 1.9777299960328657e-05, + "loss": 1.0866, + "step": 4045 + }, + { + "epoch": 0.1, + "grad_norm": 2.0449858798227636, + "learning_rate": 1.9777139794914276e-05, + "loss": 1.0591, + "step": 4046 + }, + { + "epoch": 0.1, + "grad_norm": 1.1416523557446758, + "learning_rate": 1.9776979572574292e-05, + "loss": 1.0032, + "step": 4047 + }, + { + "epoch": 0.1, + "grad_norm": 2.0237472130420024, + "learning_rate": 1.9776819293309636e-05, + "loss": 1.1133, + "step": 4048 + }, + { + "epoch": 0.1, + "grad_norm": 2.8095550685034114, + "learning_rate": 1.9776658957121242e-05, + "loss": 1.0935, + "step": 4049 + }, + { + "epoch": 0.1, + "grad_norm": 2.195827721058748, + "learning_rate": 1.9776498564010045e-05, + "loss": 1.1278, + "step": 4050 + }, + { + "epoch": 0.1, + "grad_norm": 2.0074323062751898, + "learning_rate": 1.977633811397698e-05, + "loss": 0.935, + "step": 4051 + }, + { + "epoch": 0.1, + "grad_norm": 1.9595404333269848, + "learning_rate": 1.9776177607022976e-05, + "loss": 1.0418, + "step": 4052 + }, + { + "epoch": 0.1, + "grad_norm": 1.199184067555447, + "learning_rate": 1.9776017043148975e-05, + "loss": 0.9768, + "step": 4053 + }, + { + "epoch": 0.1, + "grad_norm": 2.0617046214463777, + "learning_rate": 1.9775856422355904e-05, + "loss": 1.1455, + "step": 4054 + }, + { + "epoch": 0.1, + "grad_norm": 2.1595221780103135, + "learning_rate": 1.9775695744644706e-05, + "loss": 0.917, + "step": 4055 + }, + { + "epoch": 0.1, + "grad_norm": 2.137884014164897, + "learning_rate": 1.977553501001631e-05, + "loss": 1.1708, + "step": 4056 + }, + { + "epoch": 0.1, + "grad_norm": 1.925051570677512, + "learning_rate": 1.9775374218471653e-05, + "loss": 1.0389, + "step": 4057 + }, + { + "epoch": 0.1, + "grad_norm": 2.2174699434776595, + "learning_rate": 1.9775213370011675e-05, + "loss": 1.1947, + "step": 4058 + }, + { + "epoch": 0.1, + "grad_norm": 3.416424777590846, + "learning_rate": 1.9775052464637306e-05, + "loss": 0.9897, + "step": 4059 + }, + { + "epoch": 0.1, + "grad_norm": 2.582524538170542, + "learning_rate": 1.9774891502349493e-05, + "loss": 0.9992, + "step": 4060 + }, + { + "epoch": 0.1, + "grad_norm": 1.9901494523040724, + "learning_rate": 1.9774730483149166e-05, + "loss": 0.9618, + "step": 4061 + }, + { + "epoch": 0.1, + "grad_norm": 2.1085451410988374, + "learning_rate": 1.977456940703726e-05, + "loss": 1.1503, + "step": 4062 + }, + { + "epoch": 0.1, + "grad_norm": 2.411956367774995, + "learning_rate": 1.9774408274014715e-05, + "loss": 1.0672, + "step": 4063 + }, + { + "epoch": 0.1, + "grad_norm": 2.0346034593954445, + "learning_rate": 1.9774247084082474e-05, + "loss": 0.961, + "step": 4064 + }, + { + "epoch": 0.1, + "grad_norm": 2.1075991218164876, + "learning_rate": 1.977408583724147e-05, + "loss": 1.0267, + "step": 4065 + }, + { + "epoch": 0.1, + "grad_norm": 1.1794348225672524, + "learning_rate": 1.9773924533492646e-05, + "loss": 0.9601, + "step": 4066 + }, + { + "epoch": 0.1, + "grad_norm": 1.1168283826507035, + "learning_rate": 1.9773763172836936e-05, + "loss": 1.0138, + "step": 4067 + }, + { + "epoch": 0.1, + "grad_norm": 2.3930770464763547, + "learning_rate": 1.9773601755275287e-05, + "loss": 0.9452, + "step": 4068 + }, + { + "epoch": 0.1, + "grad_norm": 1.9865219821967732, + "learning_rate": 1.9773440280808633e-05, + "loss": 1.1825, + "step": 4069 + }, + { + "epoch": 0.1, + "grad_norm": 1.9974076248509516, + "learning_rate": 1.9773278749437912e-05, + "loss": 0.9164, + "step": 4070 + }, + { + "epoch": 0.1, + "grad_norm": 1.9554567688486515, + "learning_rate": 1.977311716116407e-05, + "loss": 1.0787, + "step": 4071 + }, + { + "epoch": 0.1, + "grad_norm": 2.1221698307773305, + "learning_rate": 1.9772955515988045e-05, + "loss": 0.9648, + "step": 4072 + }, + { + "epoch": 0.1, + "grad_norm": 1.0775965744495588, + "learning_rate": 1.9772793813910783e-05, + "loss": 0.9922, + "step": 4073 + }, + { + "epoch": 0.1, + "grad_norm": 1.9185208045202795, + "learning_rate": 1.9772632054933216e-05, + "loss": 1.0997, + "step": 4074 + }, + { + "epoch": 0.1, + "grad_norm": 1.8330965152118623, + "learning_rate": 1.9772470239056295e-05, + "loss": 1.0664, + "step": 4075 + }, + { + "epoch": 0.1, + "grad_norm": 1.1291728242243677, + "learning_rate": 1.9772308366280955e-05, + "loss": 0.9798, + "step": 4076 + }, + { + "epoch": 0.1, + "grad_norm": 2.0070474466186883, + "learning_rate": 1.9772146436608143e-05, + "loss": 1.0581, + "step": 4077 + }, + { + "epoch": 0.1, + "grad_norm": 2.04091462166025, + "learning_rate": 1.9771984450038802e-05, + "loss": 1.0358, + "step": 4078 + }, + { + "epoch": 0.1, + "grad_norm": 2.3262754686687455, + "learning_rate": 1.9771822406573876e-05, + "loss": 0.9708, + "step": 4079 + }, + { + "epoch": 0.1, + "grad_norm": 1.8395800290144653, + "learning_rate": 1.97716603062143e-05, + "loss": 0.9198, + "step": 4080 + }, + { + "epoch": 0.1, + "grad_norm": 2.166766636513496, + "learning_rate": 1.9771498148961027e-05, + "loss": 1.1288, + "step": 4081 + }, + { + "epoch": 0.1, + "grad_norm": 2.077622568461086, + "learning_rate": 1.9771335934815e-05, + "loss": 1.1723, + "step": 4082 + }, + { + "epoch": 0.1, + "grad_norm": 2.216949290127813, + "learning_rate": 1.977117366377716e-05, + "loss": 1.0632, + "step": 4083 + }, + { + "epoch": 0.1, + "grad_norm": 2.118088534456479, + "learning_rate": 1.9771011335848453e-05, + "loss": 1.1106, + "step": 4084 + }, + { + "epoch": 0.1, + "grad_norm": 1.9892722219657841, + "learning_rate": 1.9770848951029825e-05, + "loss": 1.0756, + "step": 4085 + }, + { + "epoch": 0.1, + "grad_norm": 2.4871215845120345, + "learning_rate": 1.977068650932222e-05, + "loss": 1.0616, + "step": 4086 + }, + { + "epoch": 0.1, + "grad_norm": 2.236933011910918, + "learning_rate": 1.9770524010726587e-05, + "loss": 1.0592, + "step": 4087 + }, + { + "epoch": 0.1, + "grad_norm": 2.4237217108423295, + "learning_rate": 1.977036145524387e-05, + "loss": 1.0335, + "step": 4088 + }, + { + "epoch": 0.1, + "grad_norm": 1.0406556491161614, + "learning_rate": 1.9770198842875015e-05, + "loss": 0.992, + "step": 4089 + }, + { + "epoch": 0.1, + "grad_norm": 1.9267186504305132, + "learning_rate": 1.9770036173620965e-05, + "loss": 1.0417, + "step": 4090 + }, + { + "epoch": 0.1, + "grad_norm": 2.1966860440709923, + "learning_rate": 1.9769873447482676e-05, + "loss": 1.0676, + "step": 4091 + }, + { + "epoch": 0.1, + "grad_norm": 1.1261026967499965, + "learning_rate": 1.976971066446109e-05, + "loss": 0.9517, + "step": 4092 + }, + { + "epoch": 0.1, + "grad_norm": 2.5000650335957766, + "learning_rate": 1.9769547824557155e-05, + "loss": 1.1715, + "step": 4093 + }, + { + "epoch": 0.1, + "grad_norm": 2.069004429010706, + "learning_rate": 1.976938492777182e-05, + "loss": 1.1054, + "step": 4094 + }, + { + "epoch": 0.1, + "grad_norm": 2.224185319909058, + "learning_rate": 1.976922197410603e-05, + "loss": 1.1326, + "step": 4095 + }, + { + "epoch": 0.1, + "grad_norm": 1.9726707380655064, + "learning_rate": 1.976905896356074e-05, + "loss": 1.11, + "step": 4096 + }, + { + "epoch": 0.1, + "grad_norm": 1.949210278477349, + "learning_rate": 1.9768895896136893e-05, + "loss": 1.1499, + "step": 4097 + }, + { + "epoch": 0.1, + "grad_norm": 2.3837239809579414, + "learning_rate": 1.9768732771835446e-05, + "loss": 1.0248, + "step": 4098 + }, + { + "epoch": 0.1, + "grad_norm": 2.612166401242183, + "learning_rate": 1.976856959065734e-05, + "loss": 1.038, + "step": 4099 + }, + { + "epoch": 0.1, + "grad_norm": 2.0114492734722598, + "learning_rate": 1.976840635260353e-05, + "loss": 1.0668, + "step": 4100 + }, + { + "epoch": 0.1, + "grad_norm": 2.04179912773732, + "learning_rate": 1.9768243057674967e-05, + "loss": 0.987, + "step": 4101 + }, + { + "epoch": 0.1, + "grad_norm": 3.508532185578172, + "learning_rate": 1.97680797058726e-05, + "loss": 0.9653, + "step": 4102 + }, + { + "epoch": 0.1, + "grad_norm": 1.113940722597386, + "learning_rate": 1.9767916297197378e-05, + "loss": 0.9695, + "step": 4103 + }, + { + "epoch": 0.1, + "grad_norm": 2.396155982885345, + "learning_rate": 1.976775283165026e-05, + "loss": 0.9931, + "step": 4104 + }, + { + "epoch": 0.1, + "grad_norm": 2.1035189287660914, + "learning_rate": 1.976758930923219e-05, + "loss": 1.0528, + "step": 4105 + }, + { + "epoch": 0.1, + "grad_norm": 2.2253831935930575, + "learning_rate": 1.976742572994412e-05, + "loss": 1.0886, + "step": 4106 + }, + { + "epoch": 0.1, + "grad_norm": 2.4815486625030205, + "learning_rate": 1.9767262093787007e-05, + "loss": 1.113, + "step": 4107 + }, + { + "epoch": 0.1, + "grad_norm": 2.219615415332147, + "learning_rate": 1.97670984007618e-05, + "loss": 1.1068, + "step": 4108 + }, + { + "epoch": 0.1, + "grad_norm": 2.088942450063223, + "learning_rate": 1.976693465086946e-05, + "loss": 1.0709, + "step": 4109 + }, + { + "epoch": 0.1, + "grad_norm": 2.1502746286341616, + "learning_rate": 1.976677084411093e-05, + "loss": 1.0883, + "step": 4110 + }, + { + "epoch": 0.1, + "grad_norm": 2.4096082962130065, + "learning_rate": 1.9766606980487167e-05, + "loss": 1.0826, + "step": 4111 + }, + { + "epoch": 0.1, + "grad_norm": 2.000572587763075, + "learning_rate": 1.9766443059999127e-05, + "loss": 1.079, + "step": 4112 + }, + { + "epoch": 0.1, + "grad_norm": 2.09114705232714, + "learning_rate": 1.9766279082647763e-05, + "loss": 1.0787, + "step": 4113 + }, + { + "epoch": 0.1, + "grad_norm": 2.3237084457626755, + "learning_rate": 1.976611504843403e-05, + "loss": 1.152, + "step": 4114 + }, + { + "epoch": 0.1, + "grad_norm": 1.9331482013903925, + "learning_rate": 1.9765950957358887e-05, + "loss": 0.9968, + "step": 4115 + }, + { + "epoch": 0.1, + "grad_norm": 1.9863127487003827, + "learning_rate": 1.9765786809423283e-05, + "loss": 1.1283, + "step": 4116 + }, + { + "epoch": 0.1, + "grad_norm": 2.1091488410102817, + "learning_rate": 1.9765622604628177e-05, + "loss": 1.0232, + "step": 4117 + }, + { + "epoch": 0.1, + "grad_norm": 2.326966859710692, + "learning_rate": 1.976545834297452e-05, + "loss": 1.078, + "step": 4118 + }, + { + "epoch": 0.1, + "grad_norm": 2.0891657419085017, + "learning_rate": 1.9765294024463276e-05, + "loss": 1.1522, + "step": 4119 + }, + { + "epoch": 0.1, + "grad_norm": 1.9431060630715145, + "learning_rate": 1.97651296490954e-05, + "loss": 1.0268, + "step": 4120 + }, + { + "epoch": 0.1, + "grad_norm": 2.0581673013261312, + "learning_rate": 1.9764965216871848e-05, + "loss": 1.2093, + "step": 4121 + }, + { + "epoch": 0.1, + "grad_norm": 2.2854563113611377, + "learning_rate": 1.9764800727793576e-05, + "loss": 1.0613, + "step": 4122 + }, + { + "epoch": 0.1, + "grad_norm": 2.252198554971631, + "learning_rate": 1.9764636181861544e-05, + "loss": 1.1005, + "step": 4123 + }, + { + "epoch": 0.1, + "grad_norm": 2.0633471840378834, + "learning_rate": 1.9764471579076705e-05, + "loss": 0.9764, + "step": 4124 + }, + { + "epoch": 0.1, + "grad_norm": 1.8865799389420075, + "learning_rate": 1.9764306919440023e-05, + "loss": 1.107, + "step": 4125 + }, + { + "epoch": 0.1, + "grad_norm": 1.0789581530416208, + "learning_rate": 1.9764142202952456e-05, + "loss": 0.9549, + "step": 4126 + }, + { + "epoch": 0.1, + "grad_norm": 3.230054537182663, + "learning_rate": 1.976397742961496e-05, + "loss": 1.1709, + "step": 4127 + }, + { + "epoch": 0.1, + "grad_norm": 2.0424112763746534, + "learning_rate": 1.9763812599428498e-05, + "loss": 1.117, + "step": 4128 + }, + { + "epoch": 0.1, + "grad_norm": 2.339768956518217, + "learning_rate": 1.9763647712394028e-05, + "loss": 0.9365, + "step": 4129 + }, + { + "epoch": 0.1, + "grad_norm": 1.8617398613702447, + "learning_rate": 1.9763482768512507e-05, + "loss": 1.1462, + "step": 4130 + }, + { + "epoch": 0.1, + "grad_norm": 1.9992668483409977, + "learning_rate": 1.97633177677849e-05, + "loss": 0.8974, + "step": 4131 + }, + { + "epoch": 0.1, + "grad_norm": 1.8755842714765079, + "learning_rate": 1.976315271021217e-05, + "loss": 1.0438, + "step": 4132 + }, + { + "epoch": 0.1, + "grad_norm": 1.9716007850771957, + "learning_rate": 1.976298759579527e-05, + "loss": 0.9892, + "step": 4133 + }, + { + "epoch": 0.1, + "grad_norm": 2.2394174011739034, + "learning_rate": 1.9762822424535163e-05, + "loss": 1.0136, + "step": 4134 + }, + { + "epoch": 0.1, + "grad_norm": 1.9586006602417219, + "learning_rate": 1.9762657196432814e-05, + "loss": 0.9801, + "step": 4135 + }, + { + "epoch": 0.1, + "grad_norm": 2.0888612603544416, + "learning_rate": 1.9762491911489184e-05, + "loss": 0.985, + "step": 4136 + }, + { + "epoch": 0.1, + "grad_norm": 1.1545861154323622, + "learning_rate": 1.976232656970524e-05, + "loss": 1.0326, + "step": 4137 + }, + { + "epoch": 0.1, + "grad_norm": 2.1631125874739805, + "learning_rate": 1.9762161171081934e-05, + "loss": 0.9583, + "step": 4138 + }, + { + "epoch": 0.1, + "grad_norm": 2.7190780068399136, + "learning_rate": 1.976199571562024e-05, + "loss": 0.9587, + "step": 4139 + }, + { + "epoch": 0.1, + "grad_norm": 1.0794608300233888, + "learning_rate": 1.9761830203321114e-05, + "loss": 0.9413, + "step": 4140 + }, + { + "epoch": 0.1, + "grad_norm": 2.299081377819711, + "learning_rate": 1.9761664634185523e-05, + "loss": 1.1768, + "step": 4141 + }, + { + "epoch": 0.1, + "grad_norm": 2.296242899125293, + "learning_rate": 1.976149900821443e-05, + "loss": 1.2573, + "step": 4142 + }, + { + "epoch": 0.1, + "grad_norm": 2.2199826761994936, + "learning_rate": 1.9761333325408795e-05, + "loss": 1.0695, + "step": 4143 + }, + { + "epoch": 0.1, + "grad_norm": 2.8030219799440035, + "learning_rate": 1.9761167585769588e-05, + "loss": 1.0163, + "step": 4144 + }, + { + "epoch": 0.1, + "grad_norm": 2.071724581265732, + "learning_rate": 1.9761001789297778e-05, + "loss": 1.1803, + "step": 4145 + }, + { + "epoch": 0.1, + "grad_norm": 2.1380512500835196, + "learning_rate": 1.9760835935994322e-05, + "loss": 1.0279, + "step": 4146 + }, + { + "epoch": 0.1, + "grad_norm": 2.240824689597336, + "learning_rate": 1.976067002586019e-05, + "loss": 1.004, + "step": 4147 + }, + { + "epoch": 0.1, + "grad_norm": 1.9141505676075299, + "learning_rate": 1.9760504058896348e-05, + "loss": 1.091, + "step": 4148 + }, + { + "epoch": 0.1, + "grad_norm": 2.23971921190688, + "learning_rate": 1.9760338035103758e-05, + "loss": 1.0035, + "step": 4149 + }, + { + "epoch": 0.1, + "grad_norm": 6.641354451085762, + "learning_rate": 1.976017195448339e-05, + "loss": 1.0338, + "step": 4150 + }, + { + "epoch": 0.1, + "grad_norm": 2.2816798384088206, + "learning_rate": 1.9760005817036215e-05, + "loss": 1.1118, + "step": 4151 + }, + { + "epoch": 0.1, + "grad_norm": 2.23856753932544, + "learning_rate": 1.9759839622763194e-05, + "loss": 1.0795, + "step": 4152 + }, + { + "epoch": 0.1, + "grad_norm": 2.1394541811979626, + "learning_rate": 1.9759673371665297e-05, + "loss": 1.0826, + "step": 4153 + }, + { + "epoch": 0.1, + "grad_norm": 3.3567588361854837, + "learning_rate": 1.975950706374349e-05, + "loss": 1.095, + "step": 4154 + }, + { + "epoch": 0.1, + "grad_norm": 2.0063466436235475, + "learning_rate": 1.9759340698998746e-05, + "loss": 1.1455, + "step": 4155 + }, + { + "epoch": 0.1, + "grad_norm": 2.2459702570282656, + "learning_rate": 1.9759174277432026e-05, + "loss": 1.0697, + "step": 4156 + }, + { + "epoch": 0.1, + "grad_norm": 2.098574222285572, + "learning_rate": 1.9759007799044307e-05, + "loss": 0.9941, + "step": 4157 + }, + { + "epoch": 0.1, + "grad_norm": 2.0539681601870066, + "learning_rate": 1.9758841263836553e-05, + "loss": 1.0473, + "step": 4158 + }, + { + "epoch": 0.1, + "grad_norm": 2.5849748570723508, + "learning_rate": 1.9758674671809742e-05, + "loss": 1.1708, + "step": 4159 + }, + { + "epoch": 0.1, + "grad_norm": 2.0740694446447097, + "learning_rate": 1.975850802296483e-05, + "loss": 0.9402, + "step": 4160 + }, + { + "epoch": 0.1, + "grad_norm": 1.0835985345449477, + "learning_rate": 1.97583413173028e-05, + "loss": 0.947, + "step": 4161 + }, + { + "epoch": 0.1, + "grad_norm": 1.749871002811577, + "learning_rate": 1.9758174554824614e-05, + "loss": 0.9873, + "step": 4162 + }, + { + "epoch": 0.1, + "grad_norm": 2.496121802269394, + "learning_rate": 1.9758007735531247e-05, + "loss": 1.0913, + "step": 4163 + }, + { + "epoch": 0.1, + "grad_norm": 2.3291713819095476, + "learning_rate": 1.975784085942367e-05, + "loss": 1.0787, + "step": 4164 + }, + { + "epoch": 0.1, + "grad_norm": 2.0112808515966805, + "learning_rate": 1.9757673926502854e-05, + "loss": 0.9863, + "step": 4165 + }, + { + "epoch": 0.1, + "grad_norm": 2.235191891447561, + "learning_rate": 1.975750693676977e-05, + "loss": 1.0441, + "step": 4166 + }, + { + "epoch": 0.1, + "grad_norm": 2.6757857656788833, + "learning_rate": 1.9757339890225395e-05, + "loss": 1.1251, + "step": 4167 + }, + { + "epoch": 0.1, + "grad_norm": 2.188894141368145, + "learning_rate": 1.9757172786870695e-05, + "loss": 1.0912, + "step": 4168 + }, + { + "epoch": 0.1, + "grad_norm": 2.205099127961292, + "learning_rate": 1.9757005626706648e-05, + "loss": 1.1763, + "step": 4169 + }, + { + "epoch": 0.1, + "grad_norm": 2.026100951709879, + "learning_rate": 1.9756838409734222e-05, + "loss": 1.1915, + "step": 4170 + }, + { + "epoch": 0.1, + "grad_norm": 2.2577538956020655, + "learning_rate": 1.9756671135954396e-05, + "loss": 1.0982, + "step": 4171 + }, + { + "epoch": 0.1, + "grad_norm": 2.1022386257628902, + "learning_rate": 1.975650380536814e-05, + "loss": 1.0682, + "step": 4172 + }, + { + "epoch": 0.1, + "grad_norm": 1.7878723141275557, + "learning_rate": 1.9756336417976435e-05, + "loss": 1.0914, + "step": 4173 + }, + { + "epoch": 0.1, + "grad_norm": 2.334060276265757, + "learning_rate": 1.9756168973780245e-05, + "loss": 1.1229, + "step": 4174 + }, + { + "epoch": 0.1, + "grad_norm": 2.294756198898746, + "learning_rate": 1.975600147278055e-05, + "loss": 1.0276, + "step": 4175 + }, + { + "epoch": 0.1, + "grad_norm": 1.132098199231644, + "learning_rate": 1.975583391497833e-05, + "loss": 0.9414, + "step": 4176 + }, + { + "epoch": 0.1, + "grad_norm": 2.295780976684798, + "learning_rate": 1.9755666300374554e-05, + "loss": 1.0784, + "step": 4177 + }, + { + "epoch": 0.1, + "grad_norm": 2.087155354259482, + "learning_rate": 1.9755498628970202e-05, + "loss": 1.0018, + "step": 4178 + }, + { + "epoch": 0.1, + "grad_norm": 2.2334705925133176, + "learning_rate": 1.9755330900766247e-05, + "loss": 1.2298, + "step": 4179 + }, + { + "epoch": 0.1, + "grad_norm": 2.2478465849426614, + "learning_rate": 1.975516311576367e-05, + "loss": 1.1148, + "step": 4180 + }, + { + "epoch": 0.1, + "grad_norm": 1.240012384432021, + "learning_rate": 1.975499527396344e-05, + "loss": 1.0082, + "step": 4181 + }, + { + "epoch": 0.1, + "grad_norm": 2.0097427007722732, + "learning_rate": 1.975482737536654e-05, + "loss": 1.0954, + "step": 4182 + }, + { + "epoch": 0.1, + "grad_norm": 1.4525696650824822, + "learning_rate": 1.975465941997395e-05, + "loss": 0.9586, + "step": 4183 + }, + { + "epoch": 0.1, + "grad_norm": 2.06672191845341, + "learning_rate": 1.9754491407786642e-05, + "loss": 1.0777, + "step": 4184 + }, + { + "epoch": 0.1, + "grad_norm": 1.9674981843505144, + "learning_rate": 1.9754323338805597e-05, + "loss": 1.0249, + "step": 4185 + }, + { + "epoch": 0.1, + "grad_norm": 1.8685987948676237, + "learning_rate": 1.9754155213031793e-05, + "loss": 1.0257, + "step": 4186 + }, + { + "epoch": 0.1, + "grad_norm": 2.3431728643878604, + "learning_rate": 1.9753987030466212e-05, + "loss": 1.1133, + "step": 4187 + }, + { + "epoch": 0.1, + "grad_norm": 2.6652781769587266, + "learning_rate": 1.975381879110983e-05, + "loss": 0.9744, + "step": 4188 + }, + { + "epoch": 0.1, + "grad_norm": 3.3625595979871905, + "learning_rate": 1.975365049496362e-05, + "loss": 1.0101, + "step": 4189 + }, + { + "epoch": 0.1, + "grad_norm": 2.4030074619669546, + "learning_rate": 1.9753482142028578e-05, + "loss": 1.1063, + "step": 4190 + }, + { + "epoch": 0.1, + "grad_norm": 2.273440047841103, + "learning_rate": 1.975331373230567e-05, + "loss": 1.2098, + "step": 4191 + }, + { + "epoch": 0.1, + "grad_norm": 1.2650652570428709, + "learning_rate": 1.9753145265795885e-05, + "loss": 1.0043, + "step": 4192 + }, + { + "epoch": 0.1, + "grad_norm": 2.0086883009432626, + "learning_rate": 1.9752976742500198e-05, + "loss": 0.9641, + "step": 4193 + }, + { + "epoch": 0.1, + "grad_norm": 1.1559805127097216, + "learning_rate": 1.975280816241959e-05, + "loss": 0.9847, + "step": 4194 + }, + { + "epoch": 0.1, + "grad_norm": 2.410060098737949, + "learning_rate": 1.9752639525555048e-05, + "loss": 1.0186, + "step": 4195 + }, + { + "epoch": 0.1, + "grad_norm": 2.5090069643701125, + "learning_rate": 1.9752470831907553e-05, + "loss": 1.091, + "step": 4196 + }, + { + "epoch": 0.1, + "grad_norm": 1.124070548521089, + "learning_rate": 1.9752302081478084e-05, + "loss": 0.9614, + "step": 4197 + }, + { + "epoch": 0.1, + "grad_norm": 2.2670564265954547, + "learning_rate": 1.9752133274267626e-05, + "loss": 1.0928, + "step": 4198 + }, + { + "epoch": 0.1, + "grad_norm": 2.018931408375094, + "learning_rate": 1.9751964410277157e-05, + "loss": 1.1493, + "step": 4199 + }, + { + "epoch": 0.1, + "grad_norm": 1.0687354982222514, + "learning_rate": 1.9751795489507665e-05, + "loss": 1.0029, + "step": 4200 + }, + { + "epoch": 0.1, + "grad_norm": 2.8556668831529746, + "learning_rate": 1.9751626511960134e-05, + "loss": 1.023, + "step": 4201 + }, + { + "epoch": 0.1, + "grad_norm": 2.0345645195771507, + "learning_rate": 1.9751457477635544e-05, + "loss": 1.0609, + "step": 4202 + }, + { + "epoch": 0.1, + "grad_norm": 2.3547602819882556, + "learning_rate": 1.9751288386534885e-05, + "loss": 1.1088, + "step": 4203 + }, + { + "epoch": 0.1, + "grad_norm": 2.2358978945099612, + "learning_rate": 1.9751119238659132e-05, + "loss": 1.1069, + "step": 4204 + }, + { + "epoch": 0.1, + "grad_norm": 2.0638629584956365, + "learning_rate": 1.9750950034009282e-05, + "loss": 1.0456, + "step": 4205 + }, + { + "epoch": 0.1, + "grad_norm": 1.1558702894915713, + "learning_rate": 1.975078077258631e-05, + "loss": 0.9517, + "step": 4206 + }, + { + "epoch": 0.1, + "grad_norm": 2.172088425170332, + "learning_rate": 1.9750611454391208e-05, + "loss": 1.2469, + "step": 4207 + }, + { + "epoch": 0.1, + "grad_norm": 1.8773955118797871, + "learning_rate": 1.9750442079424953e-05, + "loss": 0.9096, + "step": 4208 + }, + { + "epoch": 0.1, + "grad_norm": 1.9536486053753284, + "learning_rate": 1.9750272647688544e-05, + "loss": 1.1418, + "step": 4209 + }, + { + "epoch": 0.1, + "grad_norm": 2.7092795272243317, + "learning_rate": 1.9750103159182957e-05, + "loss": 1.1624, + "step": 4210 + }, + { + "epoch": 0.1, + "grad_norm": 1.9577254807647604, + "learning_rate": 1.974993361390918e-05, + "loss": 1.0244, + "step": 4211 + }, + { + "epoch": 0.1, + "grad_norm": 2.003881082274499, + "learning_rate": 1.9749764011868212e-05, + "loss": 1.1283, + "step": 4212 + }, + { + "epoch": 0.1, + "grad_norm": 2.2275889883170166, + "learning_rate": 1.9749594353061024e-05, + "loss": 1.011, + "step": 4213 + }, + { + "epoch": 0.1, + "grad_norm": 2.1377601815188636, + "learning_rate": 1.974942463748861e-05, + "loss": 0.9983, + "step": 4214 + }, + { + "epoch": 0.1, + "grad_norm": 2.2344423897457566, + "learning_rate": 1.9749254865151966e-05, + "loss": 1.0588, + "step": 4215 + }, + { + "epoch": 0.1, + "grad_norm": 1.950490761298227, + "learning_rate": 1.9749085036052066e-05, + "loss": 1.0319, + "step": 4216 + }, + { + "epoch": 0.1, + "grad_norm": 2.2858191364194758, + "learning_rate": 1.974891515018991e-05, + "loss": 1.0334, + "step": 4217 + }, + { + "epoch": 0.1, + "grad_norm": 1.253172488000994, + "learning_rate": 1.9748745207566483e-05, + "loss": 1.0993, + "step": 4218 + }, + { + "epoch": 0.1, + "grad_norm": 2.238772099201223, + "learning_rate": 1.9748575208182776e-05, + "loss": 1.1772, + "step": 4219 + }, + { + "epoch": 0.1, + "grad_norm": 2.302301951688483, + "learning_rate": 1.9748405152039777e-05, + "loss": 1.0073, + "step": 4220 + }, + { + "epoch": 0.1, + "grad_norm": 2.1496169992339764, + "learning_rate": 1.974823503913848e-05, + "loss": 1.238, + "step": 4221 + }, + { + "epoch": 0.1, + "grad_norm": 1.1424712527454677, + "learning_rate": 1.9748064869479866e-05, + "loss": 1.0414, + "step": 4222 + }, + { + "epoch": 0.1, + "grad_norm": 2.2502121353480775, + "learning_rate": 1.9747894643064938e-05, + "loss": 1.0742, + "step": 4223 + }, + { + "epoch": 0.1, + "grad_norm": 2.1593323083851654, + "learning_rate": 1.9747724359894677e-05, + "loss": 1.1059, + "step": 4224 + }, + { + "epoch": 0.1, + "grad_norm": 1.8645264388051668, + "learning_rate": 1.974755401997008e-05, + "loss": 1.0554, + "step": 4225 + }, + { + "epoch": 0.1, + "grad_norm": 2.9754171060653536, + "learning_rate": 1.974738362329214e-05, + "loss": 0.9902, + "step": 4226 + }, + { + "epoch": 0.1, + "grad_norm": 2.2167333225139427, + "learning_rate": 1.9747213169861844e-05, + "loss": 1.0134, + "step": 4227 + }, + { + "epoch": 0.1, + "grad_norm": 2.2710719218950395, + "learning_rate": 1.974704265968019e-05, + "loss": 1.1712, + "step": 4228 + }, + { + "epoch": 0.1, + "grad_norm": 1.0928362009332535, + "learning_rate": 1.9746872092748162e-05, + "loss": 1.026, + "step": 4229 + }, + { + "epoch": 0.1, + "grad_norm": 1.9564600724180194, + "learning_rate": 1.9746701469066766e-05, + "loss": 1.0269, + "step": 4230 + }, + { + "epoch": 0.1, + "grad_norm": 2.4610763198436207, + "learning_rate": 1.974653078863698e-05, + "loss": 1.1402, + "step": 4231 + }, + { + "epoch": 0.1, + "grad_norm": 2.1215785670718588, + "learning_rate": 1.974636005145981e-05, + "loss": 1.0653, + "step": 4232 + }, + { + "epoch": 0.1, + "grad_norm": 1.9984487209180275, + "learning_rate": 1.9746189257536252e-05, + "loss": 1.1621, + "step": 4233 + }, + { + "epoch": 0.1, + "grad_norm": 2.3175195877858465, + "learning_rate": 1.9746018406867285e-05, + "loss": 1.0269, + "step": 4234 + }, + { + "epoch": 0.1, + "grad_norm": 2.425242708997548, + "learning_rate": 1.9745847499453918e-05, + "loss": 1.089, + "step": 4235 + }, + { + "epoch": 0.1, + "grad_norm": 1.9962714222556994, + "learning_rate": 1.974567653529714e-05, + "loss": 0.9766, + "step": 4236 + }, + { + "epoch": 0.1, + "grad_norm": 2.2733790099880213, + "learning_rate": 1.9745505514397952e-05, + "loss": 1.054, + "step": 4237 + }, + { + "epoch": 0.1, + "grad_norm": 5.341248637733926, + "learning_rate": 1.9745334436757342e-05, + "loss": 1.1827, + "step": 4238 + }, + { + "epoch": 0.1, + "grad_norm": 1.8678375321824072, + "learning_rate": 1.9745163302376306e-05, + "loss": 1.0194, + "step": 4239 + }, + { + "epoch": 0.1, + "grad_norm": 1.1303453623765782, + "learning_rate": 1.974499211125585e-05, + "loss": 0.9106, + "step": 4240 + }, + { + "epoch": 0.1, + "grad_norm": 2.12923347414708, + "learning_rate": 1.9744820863396962e-05, + "loss": 1.0234, + "step": 4241 + }, + { + "epoch": 0.1, + "grad_norm": 2.092438576235731, + "learning_rate": 1.9744649558800644e-05, + "loss": 1.1457, + "step": 4242 + }, + { + "epoch": 0.1, + "grad_norm": 2.620510838720737, + "learning_rate": 1.974447819746789e-05, + "loss": 1.0964, + "step": 4243 + }, + { + "epoch": 0.1, + "grad_norm": 2.3982025529991438, + "learning_rate": 1.9744306779399697e-05, + "loss": 0.9932, + "step": 4244 + }, + { + "epoch": 0.1, + "grad_norm": 1.9960050049344962, + "learning_rate": 1.9744135304597068e-05, + "loss": 0.9754, + "step": 4245 + }, + { + "epoch": 0.1, + "grad_norm": 2.012156662452895, + "learning_rate": 1.9743963773060996e-05, + "loss": 1.0984, + "step": 4246 + }, + { + "epoch": 0.1, + "grad_norm": 2.2169299988434057, + "learning_rate": 1.974379218479248e-05, + "loss": 1.1406, + "step": 4247 + }, + { + "epoch": 0.1, + "grad_norm": 2.020107962724568, + "learning_rate": 1.9743620539792525e-05, + "loss": 1.038, + "step": 4248 + }, + { + "epoch": 0.1, + "grad_norm": 1.952849977561241, + "learning_rate": 1.9743448838062126e-05, + "loss": 1.0233, + "step": 4249 + }, + { + "epoch": 0.1, + "grad_norm": 2.0192697794897825, + "learning_rate": 1.974327707960228e-05, + "loss": 0.9977, + "step": 4250 + }, + { + "epoch": 0.1, + "grad_norm": 2.2274976372786557, + "learning_rate": 1.9743105264413994e-05, + "loss": 0.9543, + "step": 4251 + }, + { + "epoch": 0.1, + "grad_norm": 1.9504257648840977, + "learning_rate": 1.9742933392498263e-05, + "loss": 1.003, + "step": 4252 + }, + { + "epoch": 0.1, + "grad_norm": 2.2687898269527307, + "learning_rate": 1.974276146385609e-05, + "loss": 1.0056, + "step": 4253 + }, + { + "epoch": 0.1, + "grad_norm": 1.8757974322163988, + "learning_rate": 1.9742589478488474e-05, + "loss": 1.0397, + "step": 4254 + }, + { + "epoch": 0.1, + "grad_norm": 2.051970156094778, + "learning_rate": 1.974241743639642e-05, + "loss": 1.1094, + "step": 4255 + }, + { + "epoch": 0.1, + "grad_norm": 2.2933627515909962, + "learning_rate": 1.9742245337580924e-05, + "loss": 1.0461, + "step": 4256 + }, + { + "epoch": 0.1, + "grad_norm": 2.6025523258751573, + "learning_rate": 1.974207318204299e-05, + "loss": 1.042, + "step": 4257 + }, + { + "epoch": 0.1, + "grad_norm": 1.9763776611233819, + "learning_rate": 1.9741900969783625e-05, + "loss": 0.9381, + "step": 4258 + }, + { + "epoch": 0.1, + "grad_norm": 1.1370545557761709, + "learning_rate": 1.974172870080383e-05, + "loss": 0.9697, + "step": 4259 + }, + { + "epoch": 0.1, + "grad_norm": 2.220258775517195, + "learning_rate": 1.97415563751046e-05, + "loss": 1.0787, + "step": 4260 + }, + { + "epoch": 0.1, + "grad_norm": 3.3458447701956233, + "learning_rate": 1.974138399268695e-05, + "loss": 1.0695, + "step": 4261 + }, + { + "epoch": 0.1, + "grad_norm": 1.1309921746895306, + "learning_rate": 1.9741211553551877e-05, + "loss": 0.9613, + "step": 4262 + }, + { + "epoch": 0.1, + "grad_norm": 2.237778899044327, + "learning_rate": 1.9741039057700385e-05, + "loss": 1.066, + "step": 4263 + }, + { + "epoch": 0.1, + "grad_norm": 2.0714835098591378, + "learning_rate": 1.974086650513348e-05, + "loss": 1.0694, + "step": 4264 + }, + { + "epoch": 0.1, + "grad_norm": 2.2507304718020715, + "learning_rate": 1.974069389585217e-05, + "loss": 1.0997, + "step": 4265 + }, + { + "epoch": 0.1, + "grad_norm": 2.090654968635306, + "learning_rate": 1.9740521229857453e-05, + "loss": 1.0993, + "step": 4266 + }, + { + "epoch": 0.1, + "grad_norm": 2.8427710389802354, + "learning_rate": 1.9740348507150338e-05, + "loss": 0.9975, + "step": 4267 + }, + { + "epoch": 0.1, + "grad_norm": 1.1205165072473293, + "learning_rate": 1.9740175727731832e-05, + "loss": 0.9564, + "step": 4268 + }, + { + "epoch": 0.1, + "grad_norm": 2.3262860554794886, + "learning_rate": 1.9740002891602936e-05, + "loss": 1.1041, + "step": 4269 + }, + { + "epoch": 0.1, + "grad_norm": 1.9841211490527269, + "learning_rate": 1.9739829998764662e-05, + "loss": 1.1625, + "step": 4270 + }, + { + "epoch": 0.1, + "grad_norm": 1.137319701838549, + "learning_rate": 1.9739657049218014e-05, + "loss": 0.9591, + "step": 4271 + }, + { + "epoch": 0.1, + "grad_norm": 2.0372052222715142, + "learning_rate": 1.9739484042963997e-05, + "loss": 1.2162, + "step": 4272 + }, + { + "epoch": 0.1, + "grad_norm": 2.207342915556418, + "learning_rate": 1.973931098000362e-05, + "loss": 1.0716, + "step": 4273 + }, + { + "epoch": 0.1, + "grad_norm": 2.010066522751188, + "learning_rate": 1.9739137860337893e-05, + "loss": 0.9874, + "step": 4274 + }, + { + "epoch": 0.1, + "grad_norm": 3.8862917604440783, + "learning_rate": 1.9738964683967823e-05, + "loss": 0.9959, + "step": 4275 + }, + { + "epoch": 0.1, + "grad_norm": 2.1849960800903707, + "learning_rate": 1.9738791450894415e-05, + "loss": 1.1578, + "step": 4276 + }, + { + "epoch": 0.1, + "grad_norm": 1.0750833027709916, + "learning_rate": 1.9738618161118684e-05, + "loss": 1.0651, + "step": 4277 + }, + { + "epoch": 0.1, + "grad_norm": 2.2615665721634275, + "learning_rate": 1.973844481464163e-05, + "loss": 1.0008, + "step": 4278 + }, + { + "epoch": 0.1, + "grad_norm": 2.0298572989273254, + "learning_rate": 1.973827141146427e-05, + "loss": 1.0506, + "step": 4279 + }, + { + "epoch": 0.1, + "grad_norm": 1.923108892043254, + "learning_rate": 1.973809795158761e-05, + "loss": 0.9975, + "step": 4280 + }, + { + "epoch": 0.1, + "grad_norm": 2.2245046932794956, + "learning_rate": 1.973792443501266e-05, + "loss": 1.0349, + "step": 4281 + }, + { + "epoch": 0.1, + "grad_norm": 2.0308907704715704, + "learning_rate": 1.9737750861740434e-05, + "loss": 0.9599, + "step": 4282 + }, + { + "epoch": 0.1, + "grad_norm": 1.2247913161177328, + "learning_rate": 1.9737577231771934e-05, + "loss": 1.0162, + "step": 4283 + }, + { + "epoch": 0.1, + "grad_norm": 1.1272794294083777, + "learning_rate": 1.973740354510818e-05, + "loss": 0.9594, + "step": 4284 + }, + { + "epoch": 0.1, + "grad_norm": 2.301846493399294, + "learning_rate": 1.973722980175018e-05, + "loss": 1.1788, + "step": 4285 + }, + { + "epoch": 0.1, + "grad_norm": 2.0017282457352015, + "learning_rate": 1.9737056001698946e-05, + "loss": 1.2094, + "step": 4286 + }, + { + "epoch": 0.1, + "grad_norm": 2.1151722626359315, + "learning_rate": 1.9736882144955487e-05, + "loss": 0.9955, + "step": 4287 + }, + { + "epoch": 0.1, + "grad_norm": 2.130566250117899, + "learning_rate": 1.973670823152082e-05, + "loss": 0.9844, + "step": 4288 + }, + { + "epoch": 0.1, + "grad_norm": 1.1606792533825838, + "learning_rate": 1.9736534261395953e-05, + "loss": 1.0689, + "step": 4289 + }, + { + "epoch": 0.1, + "grad_norm": 2.311729701061519, + "learning_rate": 1.9736360234581902e-05, + "loss": 0.9379, + "step": 4290 + }, + { + "epoch": 0.1, + "grad_norm": 1.945044204582999, + "learning_rate": 1.9736186151079677e-05, + "loss": 1.0555, + "step": 4291 + }, + { + "epoch": 0.1, + "grad_norm": 2.852269252794488, + "learning_rate": 1.9736012010890294e-05, + "loss": 1.1951, + "step": 4292 + }, + { + "epoch": 0.1, + "grad_norm": 2.0358766553601693, + "learning_rate": 1.9735837814014772e-05, + "loss": 1.0312, + "step": 4293 + }, + { + "epoch": 0.1, + "grad_norm": 1.991390320867394, + "learning_rate": 1.9735663560454115e-05, + "loss": 1.0347, + "step": 4294 + }, + { + "epoch": 0.1, + "grad_norm": 2.286519567132478, + "learning_rate": 1.9735489250209342e-05, + "loss": 1.1172, + "step": 4295 + }, + { + "epoch": 0.1, + "grad_norm": 2.0081127979809135, + "learning_rate": 1.973531488328147e-05, + "loss": 1.0007, + "step": 4296 + }, + { + "epoch": 0.1, + "grad_norm": 2.171569644843591, + "learning_rate": 1.9735140459671512e-05, + "loss": 1.1003, + "step": 4297 + }, + { + "epoch": 0.1, + "grad_norm": 2.888340993205429, + "learning_rate": 1.9734965979380487e-05, + "loss": 1.0417, + "step": 4298 + }, + { + "epoch": 0.1, + "grad_norm": 2.247893298050273, + "learning_rate": 1.9734791442409403e-05, + "loss": 1.1224, + "step": 4299 + }, + { + "epoch": 0.1, + "grad_norm": 2.1565827464096614, + "learning_rate": 1.9734616848759288e-05, + "loss": 0.9549, + "step": 4300 + }, + { + "epoch": 0.1, + "grad_norm": 2.0464032451696847, + "learning_rate": 1.9734442198431146e-05, + "loss": 1.0536, + "step": 4301 + }, + { + "epoch": 0.1, + "grad_norm": 2.3688328577275977, + "learning_rate": 1.9734267491426003e-05, + "loss": 1.0925, + "step": 4302 + }, + { + "epoch": 0.1, + "grad_norm": 2.095010999561638, + "learning_rate": 1.973409272774487e-05, + "loss": 1.3531, + "step": 4303 + }, + { + "epoch": 0.1, + "grad_norm": 2.0290123981748502, + "learning_rate": 1.9733917907388772e-05, + "loss": 1.2078, + "step": 4304 + }, + { + "epoch": 0.1, + "grad_norm": 1.141233229178686, + "learning_rate": 1.9733743030358716e-05, + "loss": 1.0091, + "step": 4305 + }, + { + "epoch": 0.1, + "grad_norm": 2.2628591816265238, + "learning_rate": 1.973356809665573e-05, + "loss": 0.9278, + "step": 4306 + }, + { + "epoch": 0.1, + "grad_norm": 2.14373710894497, + "learning_rate": 1.9733393106280826e-05, + "loss": 0.8908, + "step": 4307 + }, + { + "epoch": 0.1, + "grad_norm": 2.2119402460733766, + "learning_rate": 1.9733218059235032e-05, + "loss": 1.0088, + "step": 4308 + }, + { + "epoch": 0.1, + "grad_norm": 2.1972719424474816, + "learning_rate": 1.9733042955519356e-05, + "loss": 1.0278, + "step": 4309 + }, + { + "epoch": 0.1, + "grad_norm": 2.213623130283373, + "learning_rate": 1.9732867795134824e-05, + "loss": 1.0263, + "step": 4310 + }, + { + "epoch": 0.1, + "grad_norm": 1.9621869243884917, + "learning_rate": 1.9732692578082453e-05, + "loss": 1.043, + "step": 4311 + }, + { + "epoch": 0.1, + "grad_norm": 1.971116510860168, + "learning_rate": 1.9732517304363266e-05, + "loss": 1.1712, + "step": 4312 + }, + { + "epoch": 0.1, + "grad_norm": 2.0582330039184664, + "learning_rate": 1.9732341973978283e-05, + "loss": 1.099, + "step": 4313 + }, + { + "epoch": 0.1, + "grad_norm": 2.3642228186765974, + "learning_rate": 1.973216658692852e-05, + "loss": 1.1631, + "step": 4314 + }, + { + "epoch": 0.1, + "grad_norm": 1.8659842623969198, + "learning_rate": 1.9731991143215005e-05, + "loss": 1.1269, + "step": 4315 + }, + { + "epoch": 0.1, + "grad_norm": 2.4505926785323515, + "learning_rate": 1.9731815642838756e-05, + "loss": 1.0783, + "step": 4316 + }, + { + "epoch": 0.1, + "grad_norm": 2.326686739494512, + "learning_rate": 1.9731640085800794e-05, + "loss": 1.1976, + "step": 4317 + }, + { + "epoch": 0.1, + "grad_norm": 2.103805797863942, + "learning_rate": 1.9731464472102145e-05, + "loss": 1.1402, + "step": 4318 + }, + { + "epoch": 0.1, + "grad_norm": 2.6112724458219434, + "learning_rate": 1.9731288801743826e-05, + "loss": 1.0291, + "step": 4319 + }, + { + "epoch": 0.1, + "grad_norm": 2.2341457644075753, + "learning_rate": 1.9731113074726863e-05, + "loss": 1.159, + "step": 4320 + }, + { + "epoch": 0.1, + "grad_norm": 2.9779809152887173, + "learning_rate": 1.973093729105228e-05, + "loss": 0.9898, + "step": 4321 + }, + { + "epoch": 0.1, + "grad_norm": 1.202388952801656, + "learning_rate": 1.9730761450721103e-05, + "loss": 0.9825, + "step": 4322 + }, + { + "epoch": 0.1, + "grad_norm": 2.3048046229796153, + "learning_rate": 1.9730585553734345e-05, + "loss": 1.0146, + "step": 4323 + }, + { + "epoch": 0.1, + "grad_norm": 1.9515509368460007, + "learning_rate": 1.9730409600093042e-05, + "loss": 1.1017, + "step": 4324 + }, + { + "epoch": 0.1, + "grad_norm": 2.0111338799907146, + "learning_rate": 1.9730233589798213e-05, + "loss": 1.0524, + "step": 4325 + }, + { + "epoch": 0.1, + "grad_norm": 2.228341243509174, + "learning_rate": 1.9730057522850882e-05, + "loss": 1.0178, + "step": 4326 + }, + { + "epoch": 0.1, + "grad_norm": 2.1100409272289817, + "learning_rate": 1.9729881399252077e-05, + "loss": 1.0148, + "step": 4327 + }, + { + "epoch": 0.1, + "grad_norm": 2.3956129685308976, + "learning_rate": 1.972970521900282e-05, + "loss": 1.0815, + "step": 4328 + }, + { + "epoch": 0.1, + "grad_norm": 2.09940834469167, + "learning_rate": 1.972952898210414e-05, + "loss": 0.9047, + "step": 4329 + }, + { + "epoch": 0.1, + "grad_norm": 1.9524794251918842, + "learning_rate": 1.9729352688557064e-05, + "loss": 1.0036, + "step": 4330 + }, + { + "epoch": 0.1, + "grad_norm": 2.5630245107160268, + "learning_rate": 1.9729176338362613e-05, + "loss": 1.1258, + "step": 4331 + }, + { + "epoch": 0.1, + "grad_norm": 2.2931817235486687, + "learning_rate": 1.9728999931521818e-05, + "loss": 1.1473, + "step": 4332 + }, + { + "epoch": 0.1, + "grad_norm": 1.993426222056928, + "learning_rate": 1.9728823468035708e-05, + "loss": 1.0983, + "step": 4333 + }, + { + "epoch": 0.1, + "grad_norm": 2.044057760731423, + "learning_rate": 1.9728646947905307e-05, + "loss": 1.1141, + "step": 4334 + }, + { + "epoch": 0.1, + "grad_norm": 3.4153820298008393, + "learning_rate": 1.972847037113164e-05, + "loss": 0.9709, + "step": 4335 + }, + { + "epoch": 0.1, + "grad_norm": 1.904568343497653, + "learning_rate": 1.9728293737715742e-05, + "loss": 1.0442, + "step": 4336 + }, + { + "epoch": 0.1, + "grad_norm": 2.0284119452967, + "learning_rate": 1.9728117047658636e-05, + "loss": 0.9109, + "step": 4337 + }, + { + "epoch": 0.1, + "grad_norm": 2.16514798790694, + "learning_rate": 1.9727940300961354e-05, + "loss": 1.0159, + "step": 4338 + }, + { + "epoch": 0.1, + "grad_norm": 1.914629510115968, + "learning_rate": 1.9727763497624924e-05, + "loss": 1.0206, + "step": 4339 + }, + { + "epoch": 0.1, + "grad_norm": 2.0175053570333654, + "learning_rate": 1.9727586637650373e-05, + "loss": 1.0359, + "step": 4340 + }, + { + "epoch": 0.1, + "grad_norm": 1.1984591912278275, + "learning_rate": 1.9727409721038736e-05, + "loss": 1.0238, + "step": 4341 + }, + { + "epoch": 0.1, + "grad_norm": 1.8409863346525017, + "learning_rate": 1.9727232747791037e-05, + "loss": 0.8729, + "step": 4342 + }, + { + "epoch": 0.1, + "grad_norm": 2.004970100484346, + "learning_rate": 1.972705571790831e-05, + "loss": 1.0723, + "step": 4343 + }, + { + "epoch": 0.1, + "grad_norm": 2.0456385024012214, + "learning_rate": 1.9726878631391585e-05, + "loss": 1.2116, + "step": 4344 + }, + { + "epoch": 0.1, + "grad_norm": 2.3241940368813077, + "learning_rate": 1.9726701488241895e-05, + "loss": 1.019, + "step": 4345 + }, + { + "epoch": 0.1, + "grad_norm": 2.0029770523768606, + "learning_rate": 1.9726524288460266e-05, + "loss": 1.1376, + "step": 4346 + }, + { + "epoch": 0.1, + "grad_norm": 4.241848600358327, + "learning_rate": 1.9726347032047735e-05, + "loss": 1.0339, + "step": 4347 + }, + { + "epoch": 0.1, + "grad_norm": 2.204631509313227, + "learning_rate": 1.9726169719005336e-05, + "loss": 1.3973, + "step": 4348 + }, + { + "epoch": 0.1, + "grad_norm": 2.2644128018859555, + "learning_rate": 1.9725992349334092e-05, + "loss": 0.9424, + "step": 4349 + }, + { + "epoch": 0.1, + "grad_norm": 2.286169531925324, + "learning_rate": 1.9725814923035044e-05, + "loss": 1.0064, + "step": 4350 + }, + { + "epoch": 0.1, + "grad_norm": 2.1644298369947044, + "learning_rate": 1.9725637440109223e-05, + "loss": 1.0808, + "step": 4351 + }, + { + "epoch": 0.1, + "grad_norm": 2.105701438540954, + "learning_rate": 1.972545990055766e-05, + "loss": 1.0849, + "step": 4352 + }, + { + "epoch": 0.1, + "grad_norm": 2.15662159559826, + "learning_rate": 1.972528230438139e-05, + "loss": 1.0674, + "step": 4353 + }, + { + "epoch": 0.1, + "grad_norm": 2.4734175869305206, + "learning_rate": 1.972510465158145e-05, + "loss": 1.2493, + "step": 4354 + }, + { + "epoch": 0.1, + "grad_norm": 2.45476703524052, + "learning_rate": 1.972492694215887e-05, + "loss": 1.0045, + "step": 4355 + }, + { + "epoch": 0.1, + "grad_norm": 2.299007658421818, + "learning_rate": 1.972474917611469e-05, + "loss": 1.022, + "step": 4356 + }, + { + "epoch": 0.1, + "grad_norm": 2.202871293650991, + "learning_rate": 1.9724571353449933e-05, + "loss": 1.0495, + "step": 4357 + }, + { + "epoch": 0.1, + "grad_norm": 2.012625398594589, + "learning_rate": 1.972439347416565e-05, + "loss": 1.0314, + "step": 4358 + }, + { + "epoch": 0.1, + "grad_norm": 1.2213519830462232, + "learning_rate": 1.972421553826287e-05, + "loss": 0.934, + "step": 4359 + }, + { + "epoch": 0.1, + "grad_norm": 2.2321468276239744, + "learning_rate": 1.9724037545742624e-05, + "loss": 1.0265, + "step": 4360 + }, + { + "epoch": 0.1, + "grad_norm": 2.5546838768735034, + "learning_rate": 1.9723859496605958e-05, + "loss": 1.1815, + "step": 4361 + }, + { + "epoch": 0.1, + "grad_norm": 2.476480085429892, + "learning_rate": 1.97236813908539e-05, + "loss": 0.9836, + "step": 4362 + }, + { + "epoch": 0.1, + "grad_norm": 2.137589840609138, + "learning_rate": 1.972350322848749e-05, + "loss": 1.0733, + "step": 4363 + }, + { + "epoch": 0.1, + "grad_norm": 2.1628030716886637, + "learning_rate": 1.972332500950777e-05, + "loss": 1.1505, + "step": 4364 + }, + { + "epoch": 0.1, + "grad_norm": 2.0452401400609475, + "learning_rate": 1.9723146733915766e-05, + "loss": 1.1002, + "step": 4365 + }, + { + "epoch": 0.1, + "grad_norm": 2.3650885495869463, + "learning_rate": 1.9722968401712527e-05, + "loss": 1.239, + "step": 4366 + }, + { + "epoch": 0.1, + "grad_norm": 1.1305123609587966, + "learning_rate": 1.972279001289909e-05, + "loss": 0.9477, + "step": 4367 + }, + { + "epoch": 0.1, + "grad_norm": 2.2510180980466354, + "learning_rate": 1.972261156747649e-05, + "loss": 1.047, + "step": 4368 + }, + { + "epoch": 0.1, + "grad_norm": 2.0985271008752435, + "learning_rate": 1.9722433065445767e-05, + "loss": 1.0059, + "step": 4369 + }, + { + "epoch": 0.1, + "grad_norm": 2.6696908750764456, + "learning_rate": 1.972225450680796e-05, + "loss": 1.0002, + "step": 4370 + }, + { + "epoch": 0.1, + "grad_norm": 1.1107320651493162, + "learning_rate": 1.972207589156411e-05, + "loss": 0.9799, + "step": 4371 + }, + { + "epoch": 0.1, + "grad_norm": 2.1977676288622967, + "learning_rate": 1.9721897219715254e-05, + "loss": 1.1549, + "step": 4372 + }, + { + "epoch": 0.1, + "grad_norm": 1.9337443849982225, + "learning_rate": 1.972171849126244e-05, + "loss": 0.9977, + "step": 4373 + }, + { + "epoch": 0.1, + "grad_norm": 2.0389919243671377, + "learning_rate": 1.9721539706206697e-05, + "loss": 1.1729, + "step": 4374 + }, + { + "epoch": 0.1, + "grad_norm": 2.2082859893883398, + "learning_rate": 1.9721360864549078e-05, + "loss": 0.9628, + "step": 4375 + }, + { + "epoch": 0.1, + "grad_norm": 2.3896835237467537, + "learning_rate": 1.9721181966290614e-05, + "loss": 1.0923, + "step": 4376 + }, + { + "epoch": 0.1, + "grad_norm": 2.081697667744795, + "learning_rate": 1.972100301143235e-05, + "loss": 1.0948, + "step": 4377 + }, + { + "epoch": 0.1, + "grad_norm": 2.7770897763256337, + "learning_rate": 1.9720823999975333e-05, + "loss": 1.0486, + "step": 4378 + }, + { + "epoch": 0.1, + "grad_norm": 1.997021777815382, + "learning_rate": 1.97206449319206e-05, + "loss": 0.9997, + "step": 4379 + }, + { + "epoch": 0.1, + "grad_norm": 2.0560660080277833, + "learning_rate": 1.9720465807269192e-05, + "loss": 1.0487, + "step": 4380 + }, + { + "epoch": 0.1, + "grad_norm": 2.0083793917601205, + "learning_rate": 1.972028662602216e-05, + "loss": 1.0279, + "step": 4381 + }, + { + "epoch": 0.1, + "grad_norm": 1.2058293068294026, + "learning_rate": 1.9720107388180536e-05, + "loss": 0.9956, + "step": 4382 + }, + { + "epoch": 0.1, + "grad_norm": 2.044491256750717, + "learning_rate": 1.9719928093745372e-05, + "loss": 1.0448, + "step": 4383 + }, + { + "epoch": 0.1, + "grad_norm": 2.8924733215141853, + "learning_rate": 1.971974874271771e-05, + "loss": 1.0375, + "step": 4384 + }, + { + "epoch": 0.1, + "grad_norm": 1.9868843986200286, + "learning_rate": 1.9719569335098592e-05, + "loss": 1.1289, + "step": 4385 + }, + { + "epoch": 0.1, + "grad_norm": 1.8008512990286687, + "learning_rate": 1.9719389870889064e-05, + "loss": 1.0416, + "step": 4386 + }, + { + "epoch": 0.1, + "grad_norm": 2.2028606963056654, + "learning_rate": 1.9719210350090172e-05, + "loss": 0.9982, + "step": 4387 + }, + { + "epoch": 0.1, + "grad_norm": 2.092501697329341, + "learning_rate": 1.971903077270296e-05, + "loss": 1.0002, + "step": 4388 + }, + { + "epoch": 0.1, + "grad_norm": 2.522348534896394, + "learning_rate": 1.9718851138728474e-05, + "loss": 1.0611, + "step": 4389 + }, + { + "epoch": 0.1, + "grad_norm": 2.052687749119902, + "learning_rate": 1.971867144816776e-05, + "loss": 0.9933, + "step": 4390 + }, + { + "epoch": 0.1, + "grad_norm": 5.081349213935528, + "learning_rate": 1.9718491701021865e-05, + "loss": 1.1247, + "step": 4391 + }, + { + "epoch": 0.1, + "grad_norm": 1.9878981575792245, + "learning_rate": 1.971831189729183e-05, + "loss": 1.0435, + "step": 4392 + }, + { + "epoch": 0.1, + "grad_norm": 2.1082127821981897, + "learning_rate": 1.9718132036978712e-05, + "loss": 1.079, + "step": 4393 + }, + { + "epoch": 0.1, + "grad_norm": 1.1673992426674813, + "learning_rate": 1.971795212008355e-05, + "loss": 0.9344, + "step": 4394 + }, + { + "epoch": 0.1, + "grad_norm": 2.2248585276158472, + "learning_rate": 1.9717772146607396e-05, + "loss": 1.2026, + "step": 4395 + }, + { + "epoch": 0.1, + "grad_norm": 2.321455815512025, + "learning_rate": 1.971759211655129e-05, + "loss": 1.0574, + "step": 4396 + }, + { + "epoch": 0.1, + "grad_norm": 2.195609380206751, + "learning_rate": 1.971741202991629e-05, + "loss": 1.1914, + "step": 4397 + }, + { + "epoch": 0.1, + "grad_norm": 1.160031859275041, + "learning_rate": 1.9717231886703444e-05, + "loss": 0.9981, + "step": 4398 + }, + { + "epoch": 0.1, + "grad_norm": 2.40226303581393, + "learning_rate": 1.9717051686913793e-05, + "loss": 1.069, + "step": 4399 + }, + { + "epoch": 0.1, + "grad_norm": 2.1566602043103624, + "learning_rate": 1.9716871430548394e-05, + "loss": 1.1533, + "step": 4400 + }, + { + "epoch": 0.1, + "grad_norm": 2.150716418148447, + "learning_rate": 1.971669111760829e-05, + "loss": 1.0266, + "step": 4401 + }, + { + "epoch": 0.1, + "grad_norm": 2.1032980436219, + "learning_rate": 1.9716510748094536e-05, + "loss": 1.141, + "step": 4402 + }, + { + "epoch": 0.1, + "grad_norm": 2.088809554948837, + "learning_rate": 1.971633032200818e-05, + "loss": 1.1476, + "step": 4403 + }, + { + "epoch": 0.1, + "grad_norm": 1.830806794806854, + "learning_rate": 1.971614983935027e-05, + "loss": 1.1303, + "step": 4404 + }, + { + "epoch": 0.1, + "grad_norm": 1.9954489880406492, + "learning_rate": 1.9715969300121863e-05, + "loss": 0.9987, + "step": 4405 + }, + { + "epoch": 0.1, + "grad_norm": 2.0007315368603837, + "learning_rate": 1.971578870432401e-05, + "loss": 1.0823, + "step": 4406 + }, + { + "epoch": 0.1, + "grad_norm": 2.389833167052055, + "learning_rate": 1.9715608051957753e-05, + "loss": 1.0408, + "step": 4407 + }, + { + "epoch": 0.1, + "grad_norm": 2.2273820284889583, + "learning_rate": 1.971542734302415e-05, + "loss": 1.1644, + "step": 4408 + }, + { + "epoch": 0.1, + "grad_norm": 2.554065267386367, + "learning_rate": 1.9715246577524258e-05, + "loss": 1.0391, + "step": 4409 + }, + { + "epoch": 0.1, + "grad_norm": 2.0138606906560765, + "learning_rate": 1.9715065755459122e-05, + "loss": 1.0447, + "step": 4410 + }, + { + "epoch": 0.1, + "grad_norm": 2.231965313157988, + "learning_rate": 1.9714884876829797e-05, + "loss": 1.1189, + "step": 4411 + }, + { + "epoch": 0.1, + "grad_norm": 2.1600471447687317, + "learning_rate": 1.9714703941637333e-05, + "loss": 0.9718, + "step": 4412 + }, + { + "epoch": 0.1, + "grad_norm": 2.106845240657732, + "learning_rate": 1.9714522949882794e-05, + "loss": 1.1318, + "step": 4413 + }, + { + "epoch": 0.1, + "grad_norm": 2.1656483556186763, + "learning_rate": 1.9714341901567223e-05, + "loss": 1.1597, + "step": 4414 + }, + { + "epoch": 0.1, + "grad_norm": 2.089014918643915, + "learning_rate": 1.971416079669168e-05, + "loss": 0.8672, + "step": 4415 + }, + { + "epoch": 0.1, + "grad_norm": 1.8298764129743175, + "learning_rate": 1.9713979635257215e-05, + "loss": 1.0488, + "step": 4416 + }, + { + "epoch": 0.1, + "grad_norm": 2.0584771874985313, + "learning_rate": 1.9713798417264885e-05, + "loss": 1.1555, + "step": 4417 + }, + { + "epoch": 0.1, + "grad_norm": 1.9286371884551237, + "learning_rate": 1.9713617142715748e-05, + "loss": 1.1113, + "step": 4418 + }, + { + "epoch": 0.1, + "grad_norm": 1.9643209580242433, + "learning_rate": 1.9713435811610855e-05, + "loss": 1.1218, + "step": 4419 + }, + { + "epoch": 0.1, + "grad_norm": 2.3453429666825354, + "learning_rate": 1.9713254423951263e-05, + "loss": 1.0608, + "step": 4420 + }, + { + "epoch": 0.1, + "grad_norm": 1.2211134593809003, + "learning_rate": 1.971307297973803e-05, + "loss": 0.9374, + "step": 4421 + }, + { + "epoch": 0.1, + "grad_norm": 2.2992328335258265, + "learning_rate": 1.971289147897221e-05, + "loss": 1.0108, + "step": 4422 + }, + { + "epoch": 0.1, + "grad_norm": 2.1331679957042944, + "learning_rate": 1.971270992165486e-05, + "loss": 1.0846, + "step": 4423 + }, + { + "epoch": 0.1, + "grad_norm": 5.914214893294745, + "learning_rate": 1.971252830778704e-05, + "loss": 1.0386, + "step": 4424 + }, + { + "epoch": 0.1, + "grad_norm": 2.1692558273052667, + "learning_rate": 1.9712346637369803e-05, + "loss": 1.0482, + "step": 4425 + }, + { + "epoch": 0.1, + "grad_norm": 2.346351068921043, + "learning_rate": 1.971216491040421e-05, + "loss": 1.1385, + "step": 4426 + }, + { + "epoch": 0.1, + "grad_norm": 2.149909545137819, + "learning_rate": 1.971198312689132e-05, + "loss": 1.0611, + "step": 4427 + }, + { + "epoch": 0.1, + "grad_norm": 2.3301709697001134, + "learning_rate": 1.9711801286832186e-05, + "loss": 1.1741, + "step": 4428 + }, + { + "epoch": 0.1, + "grad_norm": 2.1804615053519867, + "learning_rate": 1.971161939022787e-05, + "loss": 1.1084, + "step": 4429 + }, + { + "epoch": 0.1, + "grad_norm": 2.0451145550438854, + "learning_rate": 1.9711437437079436e-05, + "loss": 0.9859, + "step": 4430 + }, + { + "epoch": 0.1, + "grad_norm": 1.7759713457480442, + "learning_rate": 1.9711255427387935e-05, + "loss": 1.0334, + "step": 4431 + }, + { + "epoch": 0.1, + "grad_norm": 2.2329017620024816, + "learning_rate": 1.9711073361154433e-05, + "loss": 1.1331, + "step": 4432 + }, + { + "epoch": 0.1, + "grad_norm": 2.3276056410033377, + "learning_rate": 1.9710891238379984e-05, + "loss": 1.0641, + "step": 4433 + }, + { + "epoch": 0.1, + "grad_norm": 2.321107331180514, + "learning_rate": 1.9710709059065653e-05, + "loss": 1.0089, + "step": 4434 + }, + { + "epoch": 0.1, + "grad_norm": 2.3522149862278745, + "learning_rate": 1.97105268232125e-05, + "loss": 1.0367, + "step": 4435 + }, + { + "epoch": 0.1, + "grad_norm": 1.1404508815428633, + "learning_rate": 1.9710344530821586e-05, + "loss": 0.9461, + "step": 4436 + }, + { + "epoch": 0.1, + "grad_norm": 2.1745947884064765, + "learning_rate": 1.9710162181893972e-05, + "loss": 1.2614, + "step": 4437 + }, + { + "epoch": 0.1, + "grad_norm": 1.156323804102749, + "learning_rate": 1.9709979776430717e-05, + "loss": 0.9608, + "step": 4438 + }, + { + "epoch": 0.1, + "grad_norm": 2.2129977996575683, + "learning_rate": 1.9709797314432885e-05, + "loss": 1.1079, + "step": 4439 + }, + { + "epoch": 0.1, + "grad_norm": 1.8651626432067099, + "learning_rate": 1.9709614795901543e-05, + "loss": 1.0764, + "step": 4440 + }, + { + "epoch": 0.1, + "grad_norm": 2.4384242768500806, + "learning_rate": 1.9709432220837748e-05, + "loss": 1.1234, + "step": 4441 + }, + { + "epoch": 0.1, + "grad_norm": 2.1879742646827194, + "learning_rate": 1.9709249589242564e-05, + "loss": 1.2298, + "step": 4442 + }, + { + "epoch": 0.1, + "grad_norm": 2.962690095065754, + "learning_rate": 1.970906690111706e-05, + "loss": 0.9308, + "step": 4443 + }, + { + "epoch": 0.1, + "grad_norm": 2.152169936486692, + "learning_rate": 1.9708884156462286e-05, + "loss": 1.2339, + "step": 4444 + }, + { + "epoch": 0.1, + "grad_norm": 2.0736555753333232, + "learning_rate": 1.970870135527932e-05, + "loss": 1.0468, + "step": 4445 + }, + { + "epoch": 0.1, + "grad_norm": 2.1695601734193284, + "learning_rate": 1.970851849756922e-05, + "loss": 1.1051, + "step": 4446 + }, + { + "epoch": 0.1, + "grad_norm": 2.630964627528772, + "learning_rate": 1.9708335583333052e-05, + "loss": 1.1614, + "step": 4447 + }, + { + "epoch": 0.1, + "grad_norm": 2.0544354706656396, + "learning_rate": 1.970815261257188e-05, + "loss": 1.1474, + "step": 4448 + }, + { + "epoch": 0.1, + "grad_norm": 1.1383254912135128, + "learning_rate": 1.970796958528677e-05, + "loss": 1.0255, + "step": 4449 + }, + { + "epoch": 0.1, + "grad_norm": 1.1022869504807224, + "learning_rate": 1.970778650147879e-05, + "loss": 0.948, + "step": 4450 + }, + { + "epoch": 0.1, + "grad_norm": 2.2069381797628433, + "learning_rate": 1.9707603361149e-05, + "loss": 1.1118, + "step": 4451 + }, + { + "epoch": 0.1, + "grad_norm": 2.1938536388341965, + "learning_rate": 1.970742016429847e-05, + "loss": 1.0131, + "step": 4452 + }, + { + "epoch": 0.1, + "grad_norm": 1.03993076803874, + "learning_rate": 1.970723691092827e-05, + "loss": 0.9916, + "step": 4453 + }, + { + "epoch": 0.1, + "grad_norm": 2.488596128994443, + "learning_rate": 1.970705360103946e-05, + "loss": 1.0462, + "step": 4454 + }, + { + "epoch": 0.1, + "grad_norm": 1.1313346752768787, + "learning_rate": 1.9706870234633114e-05, + "loss": 1.0014, + "step": 4455 + }, + { + "epoch": 0.1, + "grad_norm": 2.266549058854213, + "learning_rate": 1.970668681171029e-05, + "loss": 1.1147, + "step": 4456 + }, + { + "epoch": 0.11, + "grad_norm": 2.290957716907559, + "learning_rate": 1.970650333227207e-05, + "loss": 1.102, + "step": 4457 + }, + { + "epoch": 0.11, + "grad_norm": 2.6807044518664287, + "learning_rate": 1.970631979631951e-05, + "loss": 1.0964, + "step": 4458 + }, + { + "epoch": 0.11, + "grad_norm": 1.926855024649277, + "learning_rate": 1.9706136203853684e-05, + "loss": 1.1112, + "step": 4459 + }, + { + "epoch": 0.11, + "grad_norm": 1.894108088232678, + "learning_rate": 1.970595255487566e-05, + "loss": 1.0018, + "step": 4460 + }, + { + "epoch": 0.11, + "grad_norm": 1.0969789119935338, + "learning_rate": 1.970576884938651e-05, + "loss": 0.9666, + "step": 4461 + }, + { + "epoch": 0.11, + "grad_norm": 2.195628508506437, + "learning_rate": 1.9705585087387297e-05, + "loss": 1.0354, + "step": 4462 + }, + { + "epoch": 0.11, + "grad_norm": 2.1260567639787196, + "learning_rate": 1.9705401268879095e-05, + "loss": 1.0906, + "step": 4463 + }, + { + "epoch": 0.11, + "grad_norm": 2.000564596667878, + "learning_rate": 1.9705217393862977e-05, + "loss": 1.1592, + "step": 4464 + }, + { + "epoch": 0.11, + "grad_norm": 1.1816876317360872, + "learning_rate": 1.970503346234001e-05, + "loss": 1.0154, + "step": 4465 + }, + { + "epoch": 0.11, + "grad_norm": 2.040451938758497, + "learning_rate": 1.9704849474311266e-05, + "loss": 1.1101, + "step": 4466 + }, + { + "epoch": 0.11, + "grad_norm": 2.1827638446683286, + "learning_rate": 1.9704665429777815e-05, + "loss": 1.1984, + "step": 4467 + }, + { + "epoch": 0.11, + "grad_norm": 2.163555908491533, + "learning_rate": 1.9704481328740726e-05, + "loss": 1.1324, + "step": 4468 + }, + { + "epoch": 0.11, + "grad_norm": 1.1250455171595881, + "learning_rate": 1.9704297171201082e-05, + "loss": 0.9644, + "step": 4469 + }, + { + "epoch": 0.11, + "grad_norm": 2.5001974793385937, + "learning_rate": 1.970411295715994e-05, + "loss": 1.0693, + "step": 4470 + }, + { + "epoch": 0.11, + "grad_norm": 2.5026367745779234, + "learning_rate": 1.9703928686618384e-05, + "loss": 1.1213, + "step": 4471 + }, + { + "epoch": 0.11, + "grad_norm": 1.9643127530161764, + "learning_rate": 1.9703744359577483e-05, + "loss": 1.1877, + "step": 4472 + }, + { + "epoch": 0.11, + "grad_norm": 2.0102510085095435, + "learning_rate": 1.9703559976038308e-05, + "loss": 0.9146, + "step": 4473 + }, + { + "epoch": 0.11, + "grad_norm": 1.1614520105017347, + "learning_rate": 1.9703375536001937e-05, + "loss": 0.9964, + "step": 4474 + }, + { + "epoch": 0.11, + "grad_norm": 1.1374444813970637, + "learning_rate": 1.970319103946944e-05, + "loss": 0.9958, + "step": 4475 + }, + { + "epoch": 0.11, + "grad_norm": 2.9618609111090684, + "learning_rate": 1.9703006486441895e-05, + "loss": 0.9635, + "step": 4476 + }, + { + "epoch": 0.11, + "grad_norm": 2.610357144949553, + "learning_rate": 1.9702821876920374e-05, + "loss": 1.0497, + "step": 4477 + }, + { + "epoch": 0.11, + "grad_norm": 1.960379562767535, + "learning_rate": 1.970263721090595e-05, + "loss": 0.9906, + "step": 4478 + }, + { + "epoch": 0.11, + "grad_norm": 2.1106558292003403, + "learning_rate": 1.97024524883997e-05, + "loss": 1.0782, + "step": 4479 + }, + { + "epoch": 0.11, + "grad_norm": 1.925251296867285, + "learning_rate": 1.9702267709402703e-05, + "loss": 1.1055, + "step": 4480 + }, + { + "epoch": 0.11, + "grad_norm": 2.1230325461485773, + "learning_rate": 1.970208287391603e-05, + "loss": 1.0311, + "step": 4481 + }, + { + "epoch": 0.11, + "grad_norm": 1.2162216638613081, + "learning_rate": 1.970189798194076e-05, + "loss": 1.057, + "step": 4482 + }, + { + "epoch": 0.11, + "grad_norm": 2.1340194607767784, + "learning_rate": 1.9701713033477965e-05, + "loss": 1.1678, + "step": 4483 + }, + { + "epoch": 0.11, + "grad_norm": 2.049789518388339, + "learning_rate": 1.9701528028528728e-05, + "loss": 1.0671, + "step": 4484 + }, + { + "epoch": 0.11, + "grad_norm": 1.8982745507714, + "learning_rate": 1.970134296709412e-05, + "loss": 0.9791, + "step": 4485 + }, + { + "epoch": 0.11, + "grad_norm": 2.0810067796943432, + "learning_rate": 1.970115784917523e-05, + "loss": 1.1622, + "step": 4486 + }, + { + "epoch": 0.11, + "grad_norm": 2.1444015831779386, + "learning_rate": 1.970097267477312e-05, + "loss": 1.0673, + "step": 4487 + }, + { + "epoch": 0.11, + "grad_norm": 2.3737041100923246, + "learning_rate": 1.9700787443888877e-05, + "loss": 1.0415, + "step": 4488 + }, + { + "epoch": 0.11, + "grad_norm": 2.4501630633479428, + "learning_rate": 1.9700602156523582e-05, + "loss": 1.0492, + "step": 4489 + }, + { + "epoch": 0.11, + "grad_norm": 2.250268247958851, + "learning_rate": 1.9700416812678303e-05, + "loss": 1.2066, + "step": 4490 + }, + { + "epoch": 0.11, + "grad_norm": 2.397414510689171, + "learning_rate": 1.9700231412354134e-05, + "loss": 1.0908, + "step": 4491 + }, + { + "epoch": 0.11, + "grad_norm": 2.1886235003846113, + "learning_rate": 1.9700045955552143e-05, + "loss": 1.1074, + "step": 4492 + }, + { + "epoch": 0.11, + "grad_norm": 2.1083851375884772, + "learning_rate": 1.9699860442273417e-05, + "loss": 1.0717, + "step": 4493 + }, + { + "epoch": 0.11, + "grad_norm": 1.138788239506377, + "learning_rate": 1.9699674872519026e-05, + "loss": 1.0152, + "step": 4494 + }, + { + "epoch": 0.11, + "grad_norm": 2.1805513906348413, + "learning_rate": 1.969948924629006e-05, + "loss": 1.1158, + "step": 4495 + }, + { + "epoch": 0.11, + "grad_norm": 2.06924411786901, + "learning_rate": 1.9699303563587596e-05, + "loss": 1.1114, + "step": 4496 + }, + { + "epoch": 0.11, + "grad_norm": 2.2992330132196597, + "learning_rate": 1.9699117824412718e-05, + "loss": 1.1087, + "step": 4497 + }, + { + "epoch": 0.11, + "grad_norm": 2.359207718765789, + "learning_rate": 1.9698932028766506e-05, + "loss": 1.0435, + "step": 4498 + }, + { + "epoch": 0.11, + "grad_norm": 2.2229678894218807, + "learning_rate": 1.969874617665004e-05, + "loss": 1.1289, + "step": 4499 + }, + { + "epoch": 0.11, + "grad_norm": 2.1700571860933953, + "learning_rate": 1.96985602680644e-05, + "loss": 1.1283, + "step": 4500 + }, + { + "epoch": 0.11, + "grad_norm": 2.069214444646754, + "learning_rate": 1.969837430301068e-05, + "loss": 1.0562, + "step": 4501 + }, + { + "epoch": 0.11, + "grad_norm": 2.0399884120182756, + "learning_rate": 1.9698188281489944e-05, + "loss": 0.9742, + "step": 4502 + }, + { + "epoch": 0.11, + "grad_norm": 2.3684976336931496, + "learning_rate": 1.969800220350329e-05, + "loss": 1.2149, + "step": 4503 + }, + { + "epoch": 0.11, + "grad_norm": 2.3111999928662104, + "learning_rate": 1.96978160690518e-05, + "loss": 1.0809, + "step": 4504 + }, + { + "epoch": 0.11, + "grad_norm": 1.168701604982913, + "learning_rate": 1.969762987813655e-05, + "loss": 1.0249, + "step": 4505 + }, + { + "epoch": 0.11, + "grad_norm": 2.2291251237338177, + "learning_rate": 1.969744363075863e-05, + "loss": 1.0236, + "step": 4506 + }, + { + "epoch": 0.11, + "grad_norm": 1.8995447138395214, + "learning_rate": 1.9697257326919123e-05, + "loss": 1.0262, + "step": 4507 + }, + { + "epoch": 0.11, + "grad_norm": 1.1365926150672396, + "learning_rate": 1.9697070966619115e-05, + "loss": 1.0131, + "step": 4508 + }, + { + "epoch": 0.11, + "grad_norm": 1.0420772033820629, + "learning_rate": 1.969688454985969e-05, + "loss": 0.9847, + "step": 4509 + }, + { + "epoch": 0.11, + "grad_norm": 2.032723255790662, + "learning_rate": 1.969669807664193e-05, + "loss": 1.1108, + "step": 4510 + }, + { + "epoch": 0.11, + "grad_norm": 2.6873570280455685, + "learning_rate": 1.969651154696693e-05, + "loss": 1.0552, + "step": 4511 + }, + { + "epoch": 0.11, + "grad_norm": 2.3986091072336726, + "learning_rate": 1.9696324960835763e-05, + "loss": 1.0092, + "step": 4512 + }, + { + "epoch": 0.11, + "grad_norm": 2.0876241869466763, + "learning_rate": 1.9696138318249526e-05, + "loss": 1.0035, + "step": 4513 + }, + { + "epoch": 0.11, + "grad_norm": 1.9412092356789985, + "learning_rate": 1.96959516192093e-05, + "loss": 1.0465, + "step": 4514 + }, + { + "epoch": 0.11, + "grad_norm": 2.0347279238169484, + "learning_rate": 1.9695764863716173e-05, + "loss": 1.1051, + "step": 4515 + }, + { + "epoch": 0.11, + "grad_norm": 2.4417186810853275, + "learning_rate": 1.9695578051771238e-05, + "loss": 1.0193, + "step": 4516 + }, + { + "epoch": 0.11, + "grad_norm": 1.1546430040575777, + "learning_rate": 1.9695391183375575e-05, + "loss": 0.972, + "step": 4517 + }, + { + "epoch": 0.11, + "grad_norm": 2.010944518559719, + "learning_rate": 1.9695204258530273e-05, + "loss": 1.1028, + "step": 4518 + }, + { + "epoch": 0.11, + "grad_norm": 2.263763719317331, + "learning_rate": 1.9695017277236425e-05, + "loss": 0.9989, + "step": 4519 + }, + { + "epoch": 0.11, + "grad_norm": 2.422359050554035, + "learning_rate": 1.969483023949512e-05, + "loss": 0.9799, + "step": 4520 + }, + { + "epoch": 0.11, + "grad_norm": 2.142732930763258, + "learning_rate": 1.969464314530744e-05, + "loss": 0.9922, + "step": 4521 + }, + { + "epoch": 0.11, + "grad_norm": 2.159644341733071, + "learning_rate": 1.9694455994674476e-05, + "loss": 1.1225, + "step": 4522 + }, + { + "epoch": 0.11, + "grad_norm": 2.1643868080262365, + "learning_rate": 1.9694268787597328e-05, + "loss": 1.1156, + "step": 4523 + }, + { + "epoch": 0.11, + "grad_norm": 2.032296750010564, + "learning_rate": 1.9694081524077072e-05, + "loss": 1.0379, + "step": 4524 + }, + { + "epoch": 0.11, + "grad_norm": 2.2285595409143344, + "learning_rate": 1.9693894204114803e-05, + "loss": 1.1786, + "step": 4525 + }, + { + "epoch": 0.11, + "grad_norm": 2.199511130201226, + "learning_rate": 1.9693706827711615e-05, + "loss": 1.0648, + "step": 4526 + }, + { + "epoch": 0.11, + "grad_norm": 2.293136256186399, + "learning_rate": 1.9693519394868595e-05, + "loss": 0.9961, + "step": 4527 + }, + { + "epoch": 0.11, + "grad_norm": 1.9103961993284226, + "learning_rate": 1.9693331905586842e-05, + "loss": 1.1098, + "step": 4528 + }, + { + "epoch": 0.11, + "grad_norm": 2.1132557141582176, + "learning_rate": 1.9693144359867436e-05, + "loss": 1.2034, + "step": 4529 + }, + { + "epoch": 0.11, + "grad_norm": 2.5339414321018072, + "learning_rate": 1.969295675771148e-05, + "loss": 1.0152, + "step": 4530 + }, + { + "epoch": 0.11, + "grad_norm": 2.177762140484076, + "learning_rate": 1.9692769099120054e-05, + "loss": 0.887, + "step": 4531 + }, + { + "epoch": 0.11, + "grad_norm": 2.075167841673392, + "learning_rate": 1.9692581384094262e-05, + "loss": 0.9544, + "step": 4532 + }, + { + "epoch": 0.11, + "grad_norm": 2.491872054965651, + "learning_rate": 1.9692393612635192e-05, + "loss": 0.9265, + "step": 4533 + }, + { + "epoch": 0.11, + "grad_norm": 1.9957787617232463, + "learning_rate": 1.9692205784743938e-05, + "loss": 1.0229, + "step": 4534 + }, + { + "epoch": 0.11, + "grad_norm": 2.1111565790612903, + "learning_rate": 1.9692017900421594e-05, + "loss": 1.1299, + "step": 4535 + }, + { + "epoch": 0.11, + "grad_norm": 2.33950781216679, + "learning_rate": 1.969182995966925e-05, + "loss": 1.0267, + "step": 4536 + }, + { + "epoch": 0.11, + "grad_norm": 2.658057911155671, + "learning_rate": 1.969164196248801e-05, + "loss": 1.0149, + "step": 4537 + }, + { + "epoch": 0.11, + "grad_norm": 1.1638854671437677, + "learning_rate": 1.9691453908878957e-05, + "loss": 0.9408, + "step": 4538 + }, + { + "epoch": 0.11, + "grad_norm": 2.2593185852018047, + "learning_rate": 1.969126579884319e-05, + "loss": 0.9817, + "step": 4539 + }, + { + "epoch": 0.11, + "grad_norm": 2.249215917429808, + "learning_rate": 1.969107763238181e-05, + "loss": 1.1182, + "step": 4540 + }, + { + "epoch": 0.11, + "grad_norm": 2.278307024789872, + "learning_rate": 1.9690889409495905e-05, + "loss": 1.0768, + "step": 4541 + }, + { + "epoch": 0.11, + "grad_norm": 2.1895967041063384, + "learning_rate": 1.9690701130186576e-05, + "loss": 1.1491, + "step": 4542 + }, + { + "epoch": 0.11, + "grad_norm": 2.0114449667734453, + "learning_rate": 1.9690512794454915e-05, + "loss": 1.0844, + "step": 4543 + }, + { + "epoch": 0.11, + "grad_norm": 2.1150039755850303, + "learning_rate": 1.969032440230202e-05, + "loss": 1.1489, + "step": 4544 + }, + { + "epoch": 0.11, + "grad_norm": 1.9692293685249807, + "learning_rate": 1.969013595372899e-05, + "loss": 0.9933, + "step": 4545 + }, + { + "epoch": 0.11, + "grad_norm": 2.2963288719861414, + "learning_rate": 1.9689947448736918e-05, + "loss": 1.1485, + "step": 4546 + }, + { + "epoch": 0.11, + "grad_norm": 2.0668093695106986, + "learning_rate": 1.9689758887326905e-05, + "loss": 1.0606, + "step": 4547 + }, + { + "epoch": 0.11, + "grad_norm": 2.1536727327103, + "learning_rate": 1.9689570269500053e-05, + "loss": 1.0673, + "step": 4548 + }, + { + "epoch": 0.11, + "grad_norm": 2.02174051523304, + "learning_rate": 1.968938159525745e-05, + "loss": 1.0402, + "step": 4549 + }, + { + "epoch": 0.11, + "grad_norm": 3.233198340658294, + "learning_rate": 1.9689192864600203e-05, + "loss": 1.1751, + "step": 4550 + }, + { + "epoch": 0.11, + "grad_norm": 2.05749787845613, + "learning_rate": 1.9689004077529405e-05, + "loss": 0.9408, + "step": 4551 + }, + { + "epoch": 0.11, + "grad_norm": 1.7894539309857695, + "learning_rate": 1.968881523404616e-05, + "loss": 1.0329, + "step": 4552 + }, + { + "epoch": 0.11, + "grad_norm": 2.0044080851116357, + "learning_rate": 1.968862633415156e-05, + "loss": 1.1505, + "step": 4553 + }, + { + "epoch": 0.11, + "grad_norm": 1.993584949275497, + "learning_rate": 1.9688437377846714e-05, + "loss": 0.9687, + "step": 4554 + }, + { + "epoch": 0.11, + "grad_norm": 2.4577547397945723, + "learning_rate": 1.968824836513272e-05, + "loss": 0.9768, + "step": 4555 + }, + { + "epoch": 0.11, + "grad_norm": 2.3420020037276656, + "learning_rate": 1.9688059296010676e-05, + "loss": 0.9961, + "step": 4556 + }, + { + "epoch": 0.11, + "grad_norm": 1.96310971292356, + "learning_rate": 1.968787017048168e-05, + "loss": 0.9636, + "step": 4557 + }, + { + "epoch": 0.11, + "grad_norm": 3.3716268294063205, + "learning_rate": 1.968768098854684e-05, + "loss": 1.0004, + "step": 4558 + }, + { + "epoch": 0.11, + "grad_norm": 2.0255757758084574, + "learning_rate": 1.9687491750207255e-05, + "loss": 1.0145, + "step": 4559 + }, + { + "epoch": 0.11, + "grad_norm": 1.9451917379594512, + "learning_rate": 1.9687302455464025e-05, + "loss": 1.2573, + "step": 4560 + }, + { + "epoch": 0.11, + "grad_norm": 1.1577533287560315, + "learning_rate": 1.9687113104318252e-05, + "loss": 0.9291, + "step": 4561 + }, + { + "epoch": 0.11, + "grad_norm": 1.9911061969021868, + "learning_rate": 1.968692369677104e-05, + "loss": 1.0363, + "step": 4562 + }, + { + "epoch": 0.11, + "grad_norm": 2.125247498969197, + "learning_rate": 1.968673423282349e-05, + "loss": 1.021, + "step": 4563 + }, + { + "epoch": 0.11, + "grad_norm": 1.827359647970165, + "learning_rate": 1.968654471247671e-05, + "loss": 1.0279, + "step": 4564 + }, + { + "epoch": 0.11, + "grad_norm": 2.7184760365693865, + "learning_rate": 1.9686355135731797e-05, + "loss": 1.1692, + "step": 4565 + }, + { + "epoch": 0.11, + "grad_norm": 2.2219730166787572, + "learning_rate": 1.968616550258986e-05, + "loss": 1.0617, + "step": 4566 + }, + { + "epoch": 0.11, + "grad_norm": 2.4170545460141626, + "learning_rate": 1.9685975813052e-05, + "loss": 1.0601, + "step": 4567 + }, + { + "epoch": 0.11, + "grad_norm": 1.6948595261749517, + "learning_rate": 1.968578606711932e-05, + "loss": 1.0782, + "step": 4568 + }, + { + "epoch": 0.11, + "grad_norm": 2.1545559261767857, + "learning_rate": 1.968559626479293e-05, + "loss": 1.0839, + "step": 4569 + }, + { + "epoch": 0.11, + "grad_norm": 2.393316055084493, + "learning_rate": 1.9685406406073933e-05, + "loss": 0.996, + "step": 4570 + }, + { + "epoch": 0.11, + "grad_norm": 2.187117990704024, + "learning_rate": 1.968521649096343e-05, + "loss": 1.1481, + "step": 4571 + }, + { + "epoch": 0.11, + "grad_norm": 1.9137402716977812, + "learning_rate": 1.9685026519462535e-05, + "loss": 0.9787, + "step": 4572 + }, + { + "epoch": 0.11, + "grad_norm": 1.1163703985193747, + "learning_rate": 1.9684836491572342e-05, + "loss": 1.0405, + "step": 4573 + }, + { + "epoch": 0.11, + "grad_norm": 2.1395976266438477, + "learning_rate": 1.9684646407293974e-05, + "loss": 1.0364, + "step": 4574 + }, + { + "epoch": 0.11, + "grad_norm": 1.1242491089984414, + "learning_rate": 1.9684456266628524e-05, + "loss": 0.9952, + "step": 4575 + }, + { + "epoch": 0.11, + "grad_norm": 2.0609536413536014, + "learning_rate": 1.96842660695771e-05, + "loss": 1.0626, + "step": 4576 + }, + { + "epoch": 0.11, + "grad_norm": 2.3493792299473952, + "learning_rate": 1.9684075816140817e-05, + "loss": 1.0661, + "step": 4577 + }, + { + "epoch": 0.11, + "grad_norm": 2.0379047989032464, + "learning_rate": 1.968388550632078e-05, + "loss": 0.998, + "step": 4578 + }, + { + "epoch": 0.11, + "grad_norm": 2.2508878730753543, + "learning_rate": 1.9683695140118093e-05, + "loss": 0.9324, + "step": 4579 + }, + { + "epoch": 0.11, + "grad_norm": 1.9251792354353636, + "learning_rate": 1.9683504717533868e-05, + "loss": 1.0292, + "step": 4580 + }, + { + "epoch": 0.11, + "grad_norm": 1.8303096147995317, + "learning_rate": 1.9683314238569215e-05, + "loss": 0.9591, + "step": 4581 + }, + { + "epoch": 0.11, + "grad_norm": 2.1021609667344645, + "learning_rate": 1.9683123703225236e-05, + "loss": 1.0467, + "step": 4582 + }, + { + "epoch": 0.11, + "grad_norm": 2.2994100208752424, + "learning_rate": 1.968293311150305e-05, + "loss": 1.1697, + "step": 4583 + }, + { + "epoch": 0.11, + "grad_norm": 1.9659424297116028, + "learning_rate": 1.9682742463403755e-05, + "loss": 0.9526, + "step": 4584 + }, + { + "epoch": 0.11, + "grad_norm": 2.135277448114838, + "learning_rate": 1.9682551758928473e-05, + "loss": 1.0898, + "step": 4585 + }, + { + "epoch": 0.11, + "grad_norm": 2.323023125116727, + "learning_rate": 1.968236099807831e-05, + "loss": 1.0509, + "step": 4586 + }, + { + "epoch": 0.11, + "grad_norm": 2.0658427588572486, + "learning_rate": 1.9682170180854373e-05, + "loss": 1.0844, + "step": 4587 + }, + { + "epoch": 0.11, + "grad_norm": 1.1623283498483932, + "learning_rate": 1.9681979307257778e-05, + "loss": 1.0145, + "step": 4588 + }, + { + "epoch": 0.11, + "grad_norm": 1.9012809975362728, + "learning_rate": 1.9681788377289633e-05, + "loss": 1.1904, + "step": 4589 + }, + { + "epoch": 0.11, + "grad_norm": 2.1666843323272054, + "learning_rate": 1.9681597390951052e-05, + "loss": 1.1264, + "step": 4590 + }, + { + "epoch": 0.11, + "grad_norm": 2.138149575901282, + "learning_rate": 1.9681406348243147e-05, + "loss": 0.9626, + "step": 4591 + }, + { + "epoch": 0.11, + "grad_norm": 3.047356744272378, + "learning_rate": 1.968121524916703e-05, + "loss": 1.0103, + "step": 4592 + }, + { + "epoch": 0.11, + "grad_norm": 2.19107912698095, + "learning_rate": 1.9681024093723805e-05, + "loss": 1.0634, + "step": 4593 + }, + { + "epoch": 0.11, + "grad_norm": 2.4031293514916716, + "learning_rate": 1.9680832881914598e-05, + "loss": 1.069, + "step": 4594 + }, + { + "epoch": 0.11, + "grad_norm": 2.6013438583799804, + "learning_rate": 1.9680641613740515e-05, + "loss": 0.9703, + "step": 4595 + }, + { + "epoch": 0.11, + "grad_norm": 2.3866473218855795, + "learning_rate": 1.968045028920268e-05, + "loss": 1.0911, + "step": 4596 + }, + { + "epoch": 0.11, + "grad_norm": 2.6046744417123238, + "learning_rate": 1.968025890830219e-05, + "loss": 1.0813, + "step": 4597 + }, + { + "epoch": 0.11, + "grad_norm": 1.9855468832646772, + "learning_rate": 1.968006747104017e-05, + "loss": 1.0518, + "step": 4598 + }, + { + "epoch": 0.11, + "grad_norm": 2.3346162324799957, + "learning_rate": 1.9679875977417734e-05, + "loss": 1.0874, + "step": 4599 + }, + { + "epoch": 0.11, + "grad_norm": 2.401746086755994, + "learning_rate": 1.9679684427435993e-05, + "loss": 0.9787, + "step": 4600 + }, + { + "epoch": 0.11, + "grad_norm": 2.412840620004131, + "learning_rate": 1.9679492821096065e-05, + "loss": 1.1229, + "step": 4601 + }, + { + "epoch": 0.11, + "grad_norm": 2.189583593287454, + "learning_rate": 1.9679301158399067e-05, + "loss": 1.1599, + "step": 4602 + }, + { + "epoch": 0.11, + "grad_norm": 2.0182584043148797, + "learning_rate": 1.9679109439346114e-05, + "loss": 1.1185, + "step": 4603 + }, + { + "epoch": 0.11, + "grad_norm": 2.332238495226514, + "learning_rate": 1.967891766393832e-05, + "loss": 1.0514, + "step": 4604 + }, + { + "epoch": 0.11, + "grad_norm": 2.7724690152507785, + "learning_rate": 1.9678725832176803e-05, + "loss": 1.0866, + "step": 4605 + }, + { + "epoch": 0.11, + "grad_norm": 1.9275873832548327, + "learning_rate": 1.967853394406268e-05, + "loss": 1.0339, + "step": 4606 + }, + { + "epoch": 0.11, + "grad_norm": 2.2351864342966437, + "learning_rate": 1.9678341999597066e-05, + "loss": 0.8173, + "step": 4607 + }, + { + "epoch": 0.11, + "grad_norm": 1.0703369668831912, + "learning_rate": 1.9678149998781083e-05, + "loss": 1.0274, + "step": 4608 + }, + { + "epoch": 0.11, + "grad_norm": 2.4041285874281684, + "learning_rate": 1.9677957941615847e-05, + "loss": 1.0582, + "step": 4609 + }, + { + "epoch": 0.11, + "grad_norm": 2.0826036315519, + "learning_rate": 1.9677765828102477e-05, + "loss": 1.13, + "step": 4610 + }, + { + "epoch": 0.11, + "grad_norm": 2.322154848654534, + "learning_rate": 1.967757365824209e-05, + "loss": 1.0099, + "step": 4611 + }, + { + "epoch": 0.11, + "grad_norm": 2.132665478438839, + "learning_rate": 1.9677381432035803e-05, + "loss": 1.1199, + "step": 4612 + }, + { + "epoch": 0.11, + "grad_norm": 2.0105532003410307, + "learning_rate": 1.967718914948474e-05, + "loss": 1.1293, + "step": 4613 + }, + { + "epoch": 0.11, + "grad_norm": 2.194632846373141, + "learning_rate": 1.9676996810590016e-05, + "loss": 1.0298, + "step": 4614 + }, + { + "epoch": 0.11, + "grad_norm": 1.9796954348442013, + "learning_rate": 1.9676804415352752e-05, + "loss": 1.0436, + "step": 4615 + }, + { + "epoch": 0.11, + "grad_norm": 2.084426157909238, + "learning_rate": 1.967661196377407e-05, + "loss": 1.1004, + "step": 4616 + }, + { + "epoch": 0.11, + "grad_norm": 2.328343438475327, + "learning_rate": 1.9676419455855093e-05, + "loss": 1.0614, + "step": 4617 + }, + { + "epoch": 0.11, + "grad_norm": 2.121695884726867, + "learning_rate": 1.9676226891596934e-05, + "loss": 0.9092, + "step": 4618 + }, + { + "epoch": 0.11, + "grad_norm": 2.9590632001682327, + "learning_rate": 1.967603427100072e-05, + "loss": 1.1694, + "step": 4619 + }, + { + "epoch": 0.11, + "grad_norm": 2.272211612270462, + "learning_rate": 1.967584159406757e-05, + "loss": 1.0279, + "step": 4620 + }, + { + "epoch": 0.11, + "grad_norm": 2.259252167287934, + "learning_rate": 1.967564886079861e-05, + "loss": 1.1458, + "step": 4621 + }, + { + "epoch": 0.11, + "grad_norm": 2.2152435722498165, + "learning_rate": 1.9675456071194957e-05, + "loss": 1.1041, + "step": 4622 + }, + { + "epoch": 0.11, + "grad_norm": 1.9204044130997433, + "learning_rate": 1.9675263225257736e-05, + "loss": 1.0178, + "step": 4623 + }, + { + "epoch": 0.11, + "grad_norm": 2.30232259586853, + "learning_rate": 1.967507032298807e-05, + "loss": 1.1654, + "step": 4624 + }, + { + "epoch": 0.11, + "grad_norm": 2.0630654916478774, + "learning_rate": 1.9674877364387083e-05, + "loss": 1.1518, + "step": 4625 + }, + { + "epoch": 0.11, + "grad_norm": 2.2270359601483403, + "learning_rate": 1.9674684349455893e-05, + "loss": 0.9958, + "step": 4626 + }, + { + "epoch": 0.11, + "grad_norm": 1.9632154002694495, + "learning_rate": 1.967449127819563e-05, + "loss": 1.1139, + "step": 4627 + }, + { + "epoch": 0.11, + "grad_norm": 1.8686100696081058, + "learning_rate": 1.9674298150607415e-05, + "loss": 1.0501, + "step": 4628 + }, + { + "epoch": 0.11, + "grad_norm": 1.9432564246881334, + "learning_rate": 1.9674104966692376e-05, + "loss": 1.0627, + "step": 4629 + }, + { + "epoch": 0.11, + "grad_norm": 2.1593206599871033, + "learning_rate": 1.9673911726451632e-05, + "loss": 0.997, + "step": 4630 + }, + { + "epoch": 0.11, + "grad_norm": 2.211623821521316, + "learning_rate": 1.967371842988631e-05, + "loss": 1.0589, + "step": 4631 + }, + { + "epoch": 0.11, + "grad_norm": 2.328069245487692, + "learning_rate": 1.967352507699754e-05, + "loss": 1.1118, + "step": 4632 + }, + { + "epoch": 0.11, + "grad_norm": 2.424968385903193, + "learning_rate": 1.9673331667786445e-05, + "loss": 1.1231, + "step": 4633 + }, + { + "epoch": 0.11, + "grad_norm": 1.1343870498893072, + "learning_rate": 1.9673138202254148e-05, + "loss": 0.9616, + "step": 4634 + }, + { + "epoch": 0.11, + "grad_norm": 1.8598450808537148, + "learning_rate": 1.9672944680401776e-05, + "loss": 1.0149, + "step": 4635 + }, + { + "epoch": 0.11, + "grad_norm": 2.2233459189051286, + "learning_rate": 1.967275110223046e-05, + "loss": 1.128, + "step": 4636 + }, + { + "epoch": 0.11, + "grad_norm": 1.9932952758936333, + "learning_rate": 1.9672557467741325e-05, + "loss": 0.9868, + "step": 4637 + }, + { + "epoch": 0.11, + "grad_norm": 2.36344489602908, + "learning_rate": 1.9672363776935503e-05, + "loss": 1.0765, + "step": 4638 + }, + { + "epoch": 0.11, + "grad_norm": 2.124003203723163, + "learning_rate": 1.9672170029814112e-05, + "loss": 1.2825, + "step": 4639 + }, + { + "epoch": 0.11, + "grad_norm": 2.541528170778532, + "learning_rate": 1.967197622637828e-05, + "loss": 0.9093, + "step": 4640 + }, + { + "epoch": 0.11, + "grad_norm": 2.443399043260507, + "learning_rate": 1.967178236662915e-05, + "loss": 1.1266, + "step": 4641 + }, + { + "epoch": 0.11, + "grad_norm": 2.005471090870633, + "learning_rate": 1.9671588450567833e-05, + "loss": 0.9399, + "step": 4642 + }, + { + "epoch": 0.11, + "grad_norm": 1.9927552224779945, + "learning_rate": 1.9671394478195472e-05, + "loss": 0.9957, + "step": 4643 + }, + { + "epoch": 0.11, + "grad_norm": 1.8535223319235792, + "learning_rate": 1.9671200449513188e-05, + "loss": 1.0033, + "step": 4644 + }, + { + "epoch": 0.11, + "grad_norm": 4.209319370860821, + "learning_rate": 1.9671006364522115e-05, + "loss": 1.0634, + "step": 4645 + }, + { + "epoch": 0.11, + "grad_norm": 2.2158157288481646, + "learning_rate": 1.9670812223223378e-05, + "loss": 1.1638, + "step": 4646 + }, + { + "epoch": 0.11, + "grad_norm": 2.72274658998712, + "learning_rate": 1.9670618025618112e-05, + "loss": 1.035, + "step": 4647 + }, + { + "epoch": 0.11, + "grad_norm": 2.093329208250178, + "learning_rate": 1.9670423771707442e-05, + "loss": 0.9638, + "step": 4648 + }, + { + "epoch": 0.11, + "grad_norm": 2.216476777047631, + "learning_rate": 1.967022946149251e-05, + "loss": 0.9694, + "step": 4649 + }, + { + "epoch": 0.11, + "grad_norm": 2.0034218026193544, + "learning_rate": 1.9670035094974435e-05, + "loss": 1.1255, + "step": 4650 + }, + { + "epoch": 0.11, + "grad_norm": 4.8148633633176505, + "learning_rate": 1.9669840672154358e-05, + "loss": 1.1147, + "step": 4651 + }, + { + "epoch": 0.11, + "grad_norm": 2.5635327140471134, + "learning_rate": 1.9669646193033406e-05, + "loss": 1.0731, + "step": 4652 + }, + { + "epoch": 0.11, + "grad_norm": 2.131523898027105, + "learning_rate": 1.9669451657612712e-05, + "loss": 1.0777, + "step": 4653 + }, + { + "epoch": 0.11, + "grad_norm": 1.9471808804166637, + "learning_rate": 1.966925706589341e-05, + "loss": 0.8894, + "step": 4654 + }, + { + "epoch": 0.11, + "grad_norm": 2.628280287737846, + "learning_rate": 1.966906241787663e-05, + "loss": 0.953, + "step": 4655 + }, + { + "epoch": 0.11, + "grad_norm": 1.1615731310393205, + "learning_rate": 1.966886771356351e-05, + "loss": 0.986, + "step": 4656 + }, + { + "epoch": 0.11, + "grad_norm": 2.1152757657554937, + "learning_rate": 1.966867295295518e-05, + "loss": 1.0493, + "step": 4657 + }, + { + "epoch": 0.11, + "grad_norm": 2.254237730337407, + "learning_rate": 1.9668478136052776e-05, + "loss": 1.1605, + "step": 4658 + }, + { + "epoch": 0.11, + "grad_norm": 2.1502331681253857, + "learning_rate": 1.966828326285743e-05, + "loss": 1.0307, + "step": 4659 + }, + { + "epoch": 0.11, + "grad_norm": 2.2841598702329846, + "learning_rate": 1.966808833337028e-05, + "loss": 1.0878, + "step": 4660 + }, + { + "epoch": 0.11, + "grad_norm": 2.040187820625668, + "learning_rate": 1.9667893347592458e-05, + "loss": 1.0948, + "step": 4661 + }, + { + "epoch": 0.11, + "grad_norm": 2.404679183367736, + "learning_rate": 1.96676983055251e-05, + "loss": 1.0327, + "step": 4662 + }, + { + "epoch": 0.11, + "grad_norm": 2.496321990468469, + "learning_rate": 1.9667503207169337e-05, + "loss": 1.0321, + "step": 4663 + }, + { + "epoch": 0.11, + "grad_norm": 2.4187963337427876, + "learning_rate": 1.9667308052526318e-05, + "loss": 1.265, + "step": 4664 + }, + { + "epoch": 0.11, + "grad_norm": 2.317114241164473, + "learning_rate": 1.9667112841597164e-05, + "loss": 0.971, + "step": 4665 + }, + { + "epoch": 0.11, + "grad_norm": 1.9921403070544705, + "learning_rate": 1.9666917574383026e-05, + "loss": 1.1183, + "step": 4666 + }, + { + "epoch": 0.11, + "grad_norm": 4.544096402881972, + "learning_rate": 1.9666722250885026e-05, + "loss": 1.081, + "step": 4667 + }, + { + "epoch": 0.11, + "grad_norm": 1.9600156963864301, + "learning_rate": 1.9666526871104312e-05, + "loss": 0.9872, + "step": 4668 + }, + { + "epoch": 0.11, + "grad_norm": 2.094467228950973, + "learning_rate": 1.9666331435042017e-05, + "loss": 0.9849, + "step": 4669 + }, + { + "epoch": 0.11, + "grad_norm": 1.884725242513202, + "learning_rate": 1.9666135942699278e-05, + "loss": 1.0191, + "step": 4670 + }, + { + "epoch": 0.11, + "grad_norm": 2.0361402529756645, + "learning_rate": 1.966594039407724e-05, + "loss": 1.2028, + "step": 4671 + }, + { + "epoch": 0.11, + "grad_norm": 2.240212996941244, + "learning_rate": 1.9665744789177033e-05, + "loss": 0.9781, + "step": 4672 + }, + { + "epoch": 0.11, + "grad_norm": 2.164556759615851, + "learning_rate": 1.96655491279998e-05, + "loss": 1.0586, + "step": 4673 + }, + { + "epoch": 0.11, + "grad_norm": 2.093093952773157, + "learning_rate": 1.9665353410546683e-05, + "loss": 1.0632, + "step": 4674 + }, + { + "epoch": 0.11, + "grad_norm": 3.4306824235997593, + "learning_rate": 1.9665157636818816e-05, + "loss": 0.9436, + "step": 4675 + }, + { + "epoch": 0.11, + "grad_norm": 1.8309770125126348, + "learning_rate": 1.9664961806817343e-05, + "loss": 1.0568, + "step": 4676 + }, + { + "epoch": 0.11, + "grad_norm": 2.837918607704445, + "learning_rate": 1.96647659205434e-05, + "loss": 1.0044, + "step": 4677 + }, + { + "epoch": 0.11, + "grad_norm": 3.2463481088451527, + "learning_rate": 1.9664569977998132e-05, + "loss": 1.1793, + "step": 4678 + }, + { + "epoch": 0.11, + "grad_norm": 2.118101679673423, + "learning_rate": 1.9664373979182678e-05, + "loss": 1.1966, + "step": 4679 + }, + { + "epoch": 0.11, + "grad_norm": 1.9206899860499804, + "learning_rate": 1.9664177924098178e-05, + "loss": 1.0846, + "step": 4680 + }, + { + "epoch": 0.11, + "grad_norm": 1.9700746028413663, + "learning_rate": 1.9663981812745777e-05, + "loss": 1.0707, + "step": 4681 + }, + { + "epoch": 0.11, + "grad_norm": 2.6422698784081367, + "learning_rate": 1.966378564512661e-05, + "loss": 1.1109, + "step": 4682 + }, + { + "epoch": 0.11, + "grad_norm": 2.360406074991344, + "learning_rate": 1.9663589421241827e-05, + "loss": 1.1342, + "step": 4683 + }, + { + "epoch": 0.11, + "grad_norm": 2.6112644111818044, + "learning_rate": 1.9663393141092565e-05, + "loss": 1.0171, + "step": 4684 + }, + { + "epoch": 0.11, + "grad_norm": 2.495519457823718, + "learning_rate": 1.966319680467997e-05, + "loss": 1.0354, + "step": 4685 + }, + { + "epoch": 0.11, + "grad_norm": 2.060888785203329, + "learning_rate": 1.9663000412005182e-05, + "loss": 1.1376, + "step": 4686 + }, + { + "epoch": 0.11, + "grad_norm": 2.3192676724878094, + "learning_rate": 1.966280396306935e-05, + "loss": 1.1468, + "step": 4687 + }, + { + "epoch": 0.11, + "grad_norm": 2.175296933719605, + "learning_rate": 1.966260745787361e-05, + "loss": 0.9674, + "step": 4688 + }, + { + "epoch": 0.11, + "grad_norm": 2.1519582395809604, + "learning_rate": 1.966241089641911e-05, + "loss": 1.0823, + "step": 4689 + }, + { + "epoch": 0.11, + "grad_norm": 1.1494137051759226, + "learning_rate": 1.9662214278707e-05, + "loss": 0.967, + "step": 4690 + }, + { + "epoch": 0.11, + "grad_norm": 2.147707161799924, + "learning_rate": 1.9662017604738416e-05, + "loss": 1.0703, + "step": 4691 + }, + { + "epoch": 0.11, + "grad_norm": 2.0983016237929895, + "learning_rate": 1.9661820874514504e-05, + "loss": 1.0392, + "step": 4692 + }, + { + "epoch": 0.11, + "grad_norm": 2.0358671407822473, + "learning_rate": 1.9661624088036414e-05, + "loss": 0.9854, + "step": 4693 + }, + { + "epoch": 0.11, + "grad_norm": 2.7347008987763726, + "learning_rate": 1.966142724530529e-05, + "loss": 1.0873, + "step": 4694 + }, + { + "epoch": 0.11, + "grad_norm": 2.044864405750827, + "learning_rate": 1.966123034632228e-05, + "loss": 1.0518, + "step": 4695 + }, + { + "epoch": 0.11, + "grad_norm": 1.1124226338672079, + "learning_rate": 1.9661033391088526e-05, + "loss": 0.9849, + "step": 4696 + }, + { + "epoch": 0.11, + "grad_norm": 1.1581335987689496, + "learning_rate": 1.9660836379605176e-05, + "loss": 0.9704, + "step": 4697 + }, + { + "epoch": 0.11, + "grad_norm": 2.1904515806575984, + "learning_rate": 1.966063931187338e-05, + "loss": 0.976, + "step": 4698 + }, + { + "epoch": 0.11, + "grad_norm": 1.9701357141366578, + "learning_rate": 1.9660442187894283e-05, + "loss": 1.1321, + "step": 4699 + }, + { + "epoch": 0.11, + "grad_norm": 2.0811684936255537, + "learning_rate": 1.9660245007669035e-05, + "loss": 0.9893, + "step": 4700 + }, + { + "epoch": 0.11, + "grad_norm": 2.543652981529459, + "learning_rate": 1.966004777119878e-05, + "loss": 1.0609, + "step": 4701 + }, + { + "epoch": 0.11, + "grad_norm": 1.1990546769913017, + "learning_rate": 1.9659850478484668e-05, + "loss": 1.062, + "step": 4702 + }, + { + "epoch": 0.11, + "grad_norm": 2.322560442641587, + "learning_rate": 1.965965312952785e-05, + "loss": 0.9713, + "step": 4703 + }, + { + "epoch": 0.11, + "grad_norm": 2.2208312625872373, + "learning_rate": 1.965945572432947e-05, + "loss": 0.9972, + "step": 4704 + }, + { + "epoch": 0.11, + "grad_norm": 2.4468116150745187, + "learning_rate": 1.9659258262890683e-05, + "loss": 1.1748, + "step": 4705 + }, + { + "epoch": 0.11, + "grad_norm": 2.1362591712658285, + "learning_rate": 1.9659060745212638e-05, + "loss": 1.0284, + "step": 4706 + }, + { + "epoch": 0.11, + "grad_norm": 1.967031159979324, + "learning_rate": 1.9658863171296484e-05, + "loss": 0.9978, + "step": 4707 + }, + { + "epoch": 0.11, + "grad_norm": 1.095443988184754, + "learning_rate": 1.9658665541143366e-05, + "loss": 0.9662, + "step": 4708 + }, + { + "epoch": 0.11, + "grad_norm": 1.98007865043798, + "learning_rate": 1.965846785475444e-05, + "loss": 1.0613, + "step": 4709 + }, + { + "epoch": 0.11, + "grad_norm": 1.1098470581788946, + "learning_rate": 1.965827011213086e-05, + "loss": 1.0172, + "step": 4710 + }, + { + "epoch": 0.11, + "grad_norm": 2.378495555674139, + "learning_rate": 1.965807231327377e-05, + "loss": 1.1069, + "step": 4711 + }, + { + "epoch": 0.11, + "grad_norm": 2.1951941153018026, + "learning_rate": 1.965787445818433e-05, + "loss": 0.9698, + "step": 4712 + }, + { + "epoch": 0.11, + "grad_norm": 1.969356970753703, + "learning_rate": 1.9657676546863682e-05, + "loss": 1.0734, + "step": 4713 + }, + { + "epoch": 0.11, + "grad_norm": 2.068843262523481, + "learning_rate": 1.9657478579312984e-05, + "loss": 1.1106, + "step": 4714 + }, + { + "epoch": 0.11, + "grad_norm": 2.4684466251234385, + "learning_rate": 1.965728055553339e-05, + "loss": 1.0106, + "step": 4715 + }, + { + "epoch": 0.11, + "grad_norm": 2.143331267111912, + "learning_rate": 1.965708247552605e-05, + "loss": 1.1128, + "step": 4716 + }, + { + "epoch": 0.11, + "grad_norm": 2.082807105468602, + "learning_rate": 1.9656884339292123e-05, + "loss": 1.0555, + "step": 4717 + }, + { + "epoch": 0.11, + "grad_norm": 2.6542266008075615, + "learning_rate": 1.9656686146832754e-05, + "loss": 0.9442, + "step": 4718 + }, + { + "epoch": 0.11, + "grad_norm": 2.2086477324595335, + "learning_rate": 1.96564878981491e-05, + "loss": 1.1434, + "step": 4719 + }, + { + "epoch": 0.11, + "grad_norm": 2.0897470412873216, + "learning_rate": 1.965628959324232e-05, + "loss": 1.1463, + "step": 4720 + }, + { + "epoch": 0.11, + "grad_norm": 2.3584541225375135, + "learning_rate": 1.9656091232113562e-05, + "loss": 1.1227, + "step": 4721 + }, + { + "epoch": 0.11, + "grad_norm": 2.3922865781944505, + "learning_rate": 1.9655892814763984e-05, + "loss": 1.0692, + "step": 4722 + }, + { + "epoch": 0.11, + "grad_norm": 1.9177250639333983, + "learning_rate": 1.9655694341194743e-05, + "loss": 1.1174, + "step": 4723 + }, + { + "epoch": 0.11, + "grad_norm": 2.4213780342210933, + "learning_rate": 1.965549581140699e-05, + "loss": 1.224, + "step": 4724 + }, + { + "epoch": 0.11, + "grad_norm": 2.2613817418486994, + "learning_rate": 1.9655297225401884e-05, + "loss": 1.0964, + "step": 4725 + }, + { + "epoch": 0.11, + "grad_norm": 2.0207455952862223, + "learning_rate": 1.965509858318058e-05, + "loss": 1.1277, + "step": 4726 + }, + { + "epoch": 0.11, + "grad_norm": 2.500803555721971, + "learning_rate": 1.9654899884744238e-05, + "loss": 0.9933, + "step": 4727 + }, + { + "epoch": 0.11, + "grad_norm": 2.0075888008734046, + "learning_rate": 1.965470113009401e-05, + "loss": 0.9673, + "step": 4728 + }, + { + "epoch": 0.11, + "grad_norm": 2.1412208008167117, + "learning_rate": 1.9654502319231056e-05, + "loss": 1.0688, + "step": 4729 + }, + { + "epoch": 0.11, + "grad_norm": 2.0197134362338245, + "learning_rate": 1.9654303452156535e-05, + "loss": 1.0622, + "step": 4730 + }, + { + "epoch": 0.11, + "grad_norm": 1.972825933230843, + "learning_rate": 1.96541045288716e-05, + "loss": 1.0421, + "step": 4731 + }, + { + "epoch": 0.11, + "grad_norm": 2.1381515586336604, + "learning_rate": 1.9653905549377413e-05, + "loss": 1.2039, + "step": 4732 + }, + { + "epoch": 0.11, + "grad_norm": 2.0298517291159266, + "learning_rate": 1.965370651367513e-05, + "loss": 1.0872, + "step": 4733 + }, + { + "epoch": 0.11, + "grad_norm": 2.4400578462086417, + "learning_rate": 1.9653507421765914e-05, + "loss": 1.0172, + "step": 4734 + }, + { + "epoch": 0.11, + "grad_norm": 2.0123422489772302, + "learning_rate": 1.965330827365092e-05, + "loss": 1.1414, + "step": 4735 + }, + { + "epoch": 0.11, + "grad_norm": 1.826055979876045, + "learning_rate": 1.965310906933131e-05, + "loss": 1.0389, + "step": 4736 + }, + { + "epoch": 0.11, + "grad_norm": 2.599017380001938, + "learning_rate": 1.9652909808808242e-05, + "loss": 0.9639, + "step": 4737 + }, + { + "epoch": 0.11, + "grad_norm": 1.9572489104300297, + "learning_rate": 1.965271049208288e-05, + "loss": 1.0297, + "step": 4738 + }, + { + "epoch": 0.11, + "grad_norm": 2.0684983943628645, + "learning_rate": 1.9652511119156375e-05, + "loss": 0.9498, + "step": 4739 + }, + { + "epoch": 0.11, + "grad_norm": 2.3905034645008674, + "learning_rate": 1.96523116900299e-05, + "loss": 1.0705, + "step": 4740 + }, + { + "epoch": 0.11, + "grad_norm": 2.084730170609441, + "learning_rate": 1.9652112204704608e-05, + "loss": 1.1339, + "step": 4741 + }, + { + "epoch": 0.11, + "grad_norm": 2.055208120105233, + "learning_rate": 1.9651912663181665e-05, + "loss": 1.2135, + "step": 4742 + }, + { + "epoch": 0.11, + "grad_norm": 2.270490588666334, + "learning_rate": 1.965171306546223e-05, + "loss": 0.9797, + "step": 4743 + }, + { + "epoch": 0.11, + "grad_norm": 2.9160992402948818, + "learning_rate": 1.9651513411547466e-05, + "loss": 1.1559, + "step": 4744 + }, + { + "epoch": 0.11, + "grad_norm": 2.158655824370163, + "learning_rate": 1.9651313701438533e-05, + "loss": 1.0203, + "step": 4745 + }, + { + "epoch": 0.11, + "grad_norm": 1.169613831672882, + "learning_rate": 1.9651113935136596e-05, + "loss": 1.064, + "step": 4746 + }, + { + "epoch": 0.11, + "grad_norm": 1.167184417645854, + "learning_rate": 1.965091411264282e-05, + "loss": 1.0109, + "step": 4747 + }, + { + "epoch": 0.11, + "grad_norm": 2.0676450796140156, + "learning_rate": 1.9650714233958367e-05, + "loss": 1.0772, + "step": 4748 + }, + { + "epoch": 0.11, + "grad_norm": 2.4422100672210028, + "learning_rate": 1.96505142990844e-05, + "loss": 0.95, + "step": 4749 + }, + { + "epoch": 0.11, + "grad_norm": 2.19459410172893, + "learning_rate": 1.9650314308022083e-05, + "loss": 1.1011, + "step": 4750 + }, + { + "epoch": 0.11, + "grad_norm": 2.4164384504307046, + "learning_rate": 1.9650114260772583e-05, + "loss": 1.066, + "step": 4751 + }, + { + "epoch": 0.11, + "grad_norm": 2.126086646061175, + "learning_rate": 1.964991415733706e-05, + "loss": 1.1393, + "step": 4752 + }, + { + "epoch": 0.11, + "grad_norm": 2.708851931616864, + "learning_rate": 1.964971399771668e-05, + "loss": 1.053, + "step": 4753 + }, + { + "epoch": 0.11, + "grad_norm": 2.546704884582266, + "learning_rate": 1.964951378191261e-05, + "loss": 1.1612, + "step": 4754 + }, + { + "epoch": 0.11, + "grad_norm": 1.9893350913908732, + "learning_rate": 1.9649313509926022e-05, + "loss": 1.0413, + "step": 4755 + }, + { + "epoch": 0.11, + "grad_norm": 2.20628206240983, + "learning_rate": 1.964911318175807e-05, + "loss": 1.0696, + "step": 4756 + }, + { + "epoch": 0.11, + "grad_norm": 2.049480254331229, + "learning_rate": 1.964891279740993e-05, + "loss": 1.1297, + "step": 4757 + }, + { + "epoch": 0.11, + "grad_norm": 2.366026935751893, + "learning_rate": 1.9648712356882763e-05, + "loss": 1.022, + "step": 4758 + }, + { + "epoch": 0.11, + "grad_norm": 1.8862036203927053, + "learning_rate": 1.9648511860177737e-05, + "loss": 1.1507, + "step": 4759 + }, + { + "epoch": 0.11, + "grad_norm": 2.4853903187538533, + "learning_rate": 1.964831130729602e-05, + "loss": 0.9925, + "step": 4760 + }, + { + "epoch": 0.11, + "grad_norm": 1.9747876898964336, + "learning_rate": 1.9648110698238782e-05, + "loss": 1.0981, + "step": 4761 + }, + { + "epoch": 0.11, + "grad_norm": 1.9609580722215105, + "learning_rate": 1.964791003300719e-05, + "loss": 0.9462, + "step": 4762 + }, + { + "epoch": 0.11, + "grad_norm": 1.986793825087307, + "learning_rate": 1.9647709311602408e-05, + "loss": 1.0334, + "step": 4763 + }, + { + "epoch": 0.11, + "grad_norm": 1.886761313221415, + "learning_rate": 1.964750853402561e-05, + "loss": 0.9073, + "step": 4764 + }, + { + "epoch": 0.11, + "grad_norm": 2.033263182692086, + "learning_rate": 1.9647307700277963e-05, + "loss": 1.1511, + "step": 4765 + }, + { + "epoch": 0.11, + "grad_norm": 2.0861461723441996, + "learning_rate": 1.9647106810360636e-05, + "loss": 1.0547, + "step": 4766 + }, + { + "epoch": 0.11, + "grad_norm": 2.1673652426445895, + "learning_rate": 1.9646905864274798e-05, + "loss": 1.1076, + "step": 4767 + }, + { + "epoch": 0.11, + "grad_norm": 2.0831532712574026, + "learning_rate": 1.964670486202162e-05, + "loss": 1.0937, + "step": 4768 + }, + { + "epoch": 0.11, + "grad_norm": 2.0295092276661886, + "learning_rate": 1.9646503803602272e-05, + "loss": 1.2469, + "step": 4769 + }, + { + "epoch": 0.11, + "grad_norm": 2.601643309755768, + "learning_rate": 1.9646302689017925e-05, + "loss": 1.084, + "step": 4770 + }, + { + "epoch": 0.11, + "grad_norm": 1.9793926329483413, + "learning_rate": 1.964610151826975e-05, + "loss": 1.0939, + "step": 4771 + }, + { + "epoch": 0.11, + "grad_norm": 2.2327982874295165, + "learning_rate": 1.9645900291358918e-05, + "loss": 1.1757, + "step": 4772 + }, + { + "epoch": 0.11, + "grad_norm": 2.337612115891201, + "learning_rate": 1.9645699008286602e-05, + "loss": 1.1144, + "step": 4773 + }, + { + "epoch": 0.11, + "grad_norm": 1.998645613870823, + "learning_rate": 1.9645497669053973e-05, + "loss": 0.9619, + "step": 4774 + }, + { + "epoch": 0.11, + "grad_norm": 1.8547362034707118, + "learning_rate": 1.9645296273662202e-05, + "loss": 1.1113, + "step": 4775 + }, + { + "epoch": 0.11, + "grad_norm": 2.0484661301499623, + "learning_rate": 1.964509482211246e-05, + "loss": 0.9105, + "step": 4776 + }, + { + "epoch": 0.11, + "grad_norm": 2.232817366586725, + "learning_rate": 1.9644893314405924e-05, + "loss": 1.0845, + "step": 4777 + }, + { + "epoch": 0.11, + "grad_norm": 2.350157873978056, + "learning_rate": 1.964469175054377e-05, + "loss": 1.0269, + "step": 4778 + }, + { + "epoch": 0.11, + "grad_norm": 2.3475679107336567, + "learning_rate": 1.964449013052716e-05, + "loss": 0.9145, + "step": 4779 + }, + { + "epoch": 0.11, + "grad_norm": 2.513769241539131, + "learning_rate": 1.964428845435728e-05, + "loss": 1.2188, + "step": 4780 + }, + { + "epoch": 0.11, + "grad_norm": 1.9908488999431255, + "learning_rate": 1.9644086722035295e-05, + "loss": 1.0149, + "step": 4781 + }, + { + "epoch": 0.11, + "grad_norm": 2.3185578054851073, + "learning_rate": 1.9643884933562387e-05, + "loss": 1.0103, + "step": 4782 + }, + { + "epoch": 0.11, + "grad_norm": 1.1453152721981024, + "learning_rate": 1.9643683088939726e-05, + "loss": 0.9709, + "step": 4783 + }, + { + "epoch": 0.11, + "grad_norm": 2.040928375966527, + "learning_rate": 1.9643481188168492e-05, + "loss": 0.9599, + "step": 4784 + }, + { + "epoch": 0.11, + "grad_norm": 1.1734102343519914, + "learning_rate": 1.9643279231249855e-05, + "loss": 0.9876, + "step": 4785 + }, + { + "epoch": 0.11, + "grad_norm": 2.1862536794672818, + "learning_rate": 1.9643077218184993e-05, + "loss": 1.0563, + "step": 4786 + }, + { + "epoch": 0.11, + "grad_norm": 2.4292413561247836, + "learning_rate": 1.9642875148975084e-05, + "loss": 1.0384, + "step": 4787 + }, + { + "epoch": 0.11, + "grad_norm": 2.1636777959337294, + "learning_rate": 1.9642673023621303e-05, + "loss": 1.2125, + "step": 4788 + }, + { + "epoch": 0.11, + "grad_norm": 2.038134608466705, + "learning_rate": 1.9642470842124823e-05, + "loss": 1.1195, + "step": 4789 + }, + { + "epoch": 0.11, + "grad_norm": 2.4264879217642052, + "learning_rate": 1.964226860448683e-05, + "loss": 0.9628, + "step": 4790 + }, + { + "epoch": 0.11, + "grad_norm": 2.0762352238461736, + "learning_rate": 1.9642066310708493e-05, + "loss": 1.077, + "step": 4791 + }, + { + "epoch": 0.11, + "grad_norm": 2.42038773970144, + "learning_rate": 1.9641863960790998e-05, + "loss": 1.0452, + "step": 4792 + }, + { + "epoch": 0.11, + "grad_norm": 2.1673784256667536, + "learning_rate": 1.9641661554735513e-05, + "loss": 1.0676, + "step": 4793 + }, + { + "epoch": 0.11, + "grad_norm": 2.628457985207885, + "learning_rate": 1.9641459092543224e-05, + "loss": 1.1827, + "step": 4794 + }, + { + "epoch": 0.11, + "grad_norm": 2.447417585059214, + "learning_rate": 1.9641256574215308e-05, + "loss": 1.0644, + "step": 4795 + }, + { + "epoch": 0.11, + "grad_norm": 2.1099232068343476, + "learning_rate": 1.9641053999752942e-05, + "loss": 0.9845, + "step": 4796 + }, + { + "epoch": 0.11, + "grad_norm": 2.076092257368169, + "learning_rate": 1.964085136915731e-05, + "loss": 1.1112, + "step": 4797 + }, + { + "epoch": 0.11, + "grad_norm": 2.7950800759115717, + "learning_rate": 1.964064868242959e-05, + "loss": 0.9377, + "step": 4798 + }, + { + "epoch": 0.11, + "grad_norm": 2.2118794246934104, + "learning_rate": 1.964044593957096e-05, + "loss": 0.9865, + "step": 4799 + }, + { + "epoch": 0.11, + "grad_norm": 1.9981364641911843, + "learning_rate": 1.96402431405826e-05, + "loss": 1.0632, + "step": 4800 + }, + { + "epoch": 0.11, + "grad_norm": 2.649983848262347, + "learning_rate": 1.9640040285465693e-05, + "loss": 1.0984, + "step": 4801 + }, + { + "epoch": 0.11, + "grad_norm": 2.2877795773029277, + "learning_rate": 1.963983737422142e-05, + "loss": 0.9587, + "step": 4802 + }, + { + "epoch": 0.11, + "grad_norm": 2.1338567651007434, + "learning_rate": 1.9639634406850962e-05, + "loss": 1.0597, + "step": 4803 + }, + { + "epoch": 0.11, + "grad_norm": 1.8932622452064838, + "learning_rate": 1.96394313833555e-05, + "loss": 1.0204, + "step": 4804 + }, + { + "epoch": 0.11, + "grad_norm": 2.1247144301884817, + "learning_rate": 1.9639228303736216e-05, + "loss": 1.0483, + "step": 4805 + }, + { + "epoch": 0.11, + "grad_norm": 2.484902878001852, + "learning_rate": 1.9639025167994295e-05, + "loss": 0.938, + "step": 4806 + }, + { + "epoch": 0.11, + "grad_norm": 1.9646782273283738, + "learning_rate": 1.9638821976130916e-05, + "loss": 1.0678, + "step": 4807 + }, + { + "epoch": 0.11, + "grad_norm": 2.181979829421008, + "learning_rate": 1.9638618728147266e-05, + "loss": 0.8823, + "step": 4808 + }, + { + "epoch": 0.11, + "grad_norm": 2.1680038769342036, + "learning_rate": 1.9638415424044527e-05, + "loss": 0.9827, + "step": 4809 + }, + { + "epoch": 0.11, + "grad_norm": 2.0840014896744727, + "learning_rate": 1.963821206382388e-05, + "loss": 1.0892, + "step": 4810 + }, + { + "epoch": 0.11, + "grad_norm": 2.2417187913186853, + "learning_rate": 1.963800864748651e-05, + "loss": 0.9525, + "step": 4811 + }, + { + "epoch": 0.11, + "grad_norm": 2.4035519027576036, + "learning_rate": 1.9637805175033603e-05, + "loss": 1.031, + "step": 4812 + }, + { + "epoch": 0.11, + "grad_norm": 2.2094798965756985, + "learning_rate": 1.9637601646466343e-05, + "loss": 0.9489, + "step": 4813 + }, + { + "epoch": 0.11, + "grad_norm": 2.545495386012103, + "learning_rate": 1.9637398061785914e-05, + "loss": 1.275, + "step": 4814 + }, + { + "epoch": 0.11, + "grad_norm": 2.596109564744786, + "learning_rate": 1.9637194420993504e-05, + "loss": 1.1481, + "step": 4815 + }, + { + "epoch": 0.11, + "grad_norm": 3.5784605759180073, + "learning_rate": 1.9636990724090297e-05, + "loss": 1.0191, + "step": 4816 + }, + { + "epoch": 0.11, + "grad_norm": 1.8723135409848495, + "learning_rate": 1.963678697107748e-05, + "loss": 1.1568, + "step": 4817 + }, + { + "epoch": 0.11, + "grad_norm": 1.2013248710578333, + "learning_rate": 1.9636583161956235e-05, + "loss": 0.97, + "step": 4818 + }, + { + "epoch": 0.11, + "grad_norm": 2.488971525567875, + "learning_rate": 1.9636379296727754e-05, + "loss": 0.9654, + "step": 4819 + }, + { + "epoch": 0.11, + "grad_norm": 1.9600840907852024, + "learning_rate": 1.9636175375393223e-05, + "loss": 1.0807, + "step": 4820 + }, + { + "epoch": 0.11, + "grad_norm": 1.997175810692513, + "learning_rate": 1.9635971397953826e-05, + "loss": 0.9958, + "step": 4821 + }, + { + "epoch": 0.11, + "grad_norm": 2.862780930355369, + "learning_rate": 1.963576736441075e-05, + "loss": 1.0356, + "step": 4822 + }, + { + "epoch": 0.11, + "grad_norm": 1.1924427842969536, + "learning_rate": 1.9635563274765194e-05, + "loss": 0.9168, + "step": 4823 + }, + { + "epoch": 0.11, + "grad_norm": 2.5095827681726433, + "learning_rate": 1.9635359129018334e-05, + "loss": 1.0581, + "step": 4824 + }, + { + "epoch": 0.11, + "grad_norm": 1.98088493608541, + "learning_rate": 1.963515492717136e-05, + "loss": 0.9446, + "step": 4825 + }, + { + "epoch": 0.11, + "grad_norm": 2.3005894030862177, + "learning_rate": 1.9634950669225465e-05, + "loss": 1.0818, + "step": 4826 + }, + { + "epoch": 0.11, + "grad_norm": 5.602192275081561, + "learning_rate": 1.9634746355181836e-05, + "loss": 1.0368, + "step": 4827 + }, + { + "epoch": 0.11, + "grad_norm": 2.679035440211452, + "learning_rate": 1.9634541985041667e-05, + "loss": 1.0612, + "step": 4828 + }, + { + "epoch": 0.11, + "grad_norm": 2.0850317937951806, + "learning_rate": 1.9634337558806142e-05, + "loss": 0.9108, + "step": 4829 + }, + { + "epoch": 0.11, + "grad_norm": 2.079877432579708, + "learning_rate": 1.9634133076476453e-05, + "loss": 1.1925, + "step": 4830 + }, + { + "epoch": 0.11, + "grad_norm": 2.324957000016496, + "learning_rate": 1.9633928538053792e-05, + "loss": 0.8153, + "step": 4831 + }, + { + "epoch": 0.11, + "grad_norm": 2.226302176553815, + "learning_rate": 1.9633723943539352e-05, + "loss": 1.0718, + "step": 4832 + }, + { + "epoch": 0.11, + "grad_norm": 2.336462426570883, + "learning_rate": 1.9633519292934317e-05, + "loss": 1.0956, + "step": 4833 + }, + { + "epoch": 0.11, + "grad_norm": 2.050282274919218, + "learning_rate": 1.9633314586239884e-05, + "loss": 1.0919, + "step": 4834 + }, + { + "epoch": 0.11, + "grad_norm": 2.170728676834178, + "learning_rate": 1.9633109823457243e-05, + "loss": 1.0826, + "step": 4835 + }, + { + "epoch": 0.11, + "grad_norm": 2.5603854340719288, + "learning_rate": 1.9632905004587588e-05, + "loss": 0.9381, + "step": 4836 + }, + { + "epoch": 0.11, + "grad_norm": 2.2176513209920135, + "learning_rate": 1.963270012963211e-05, + "loss": 1.1396, + "step": 4837 + }, + { + "epoch": 0.11, + "grad_norm": 1.9273005548634392, + "learning_rate": 1.9632495198592003e-05, + "loss": 1.1875, + "step": 4838 + }, + { + "epoch": 0.11, + "grad_norm": 1.879489747978109, + "learning_rate": 1.9632290211468458e-05, + "loss": 1.1747, + "step": 4839 + }, + { + "epoch": 0.11, + "grad_norm": 2.2030730940997816, + "learning_rate": 1.963208516826267e-05, + "loss": 1.0858, + "step": 4840 + }, + { + "epoch": 0.11, + "grad_norm": 2.0034419393488387, + "learning_rate": 1.9631880068975834e-05, + "loss": 1.0322, + "step": 4841 + }, + { + "epoch": 0.11, + "grad_norm": 2.1481183725373576, + "learning_rate": 1.9631674913609137e-05, + "loss": 1.0428, + "step": 4842 + }, + { + "epoch": 0.11, + "grad_norm": 2.897535560509677, + "learning_rate": 1.9631469702163786e-05, + "loss": 1.043, + "step": 4843 + }, + { + "epoch": 0.11, + "grad_norm": 2.617274905905232, + "learning_rate": 1.9631264434640965e-05, + "loss": 0.9888, + "step": 4844 + }, + { + "epoch": 0.11, + "grad_norm": 2.616769286447459, + "learning_rate": 1.9631059111041876e-05, + "loss": 1.2103, + "step": 4845 + }, + { + "epoch": 0.11, + "grad_norm": 1.2368299942459933, + "learning_rate": 1.9630853731367715e-05, + "loss": 1.0306, + "step": 4846 + }, + { + "epoch": 0.11, + "grad_norm": 3.0433641882697886, + "learning_rate": 1.9630648295619668e-05, + "loss": 0.8478, + "step": 4847 + }, + { + "epoch": 0.11, + "grad_norm": 1.9997471555497837, + "learning_rate": 1.963044280379894e-05, + "loss": 1.1384, + "step": 4848 + }, + { + "epoch": 0.11, + "grad_norm": 1.1247457491811752, + "learning_rate": 1.9630237255906724e-05, + "loss": 1.0219, + "step": 4849 + }, + { + "epoch": 0.11, + "grad_norm": 2.597040786917438, + "learning_rate": 1.963003165194422e-05, + "loss": 1.11, + "step": 4850 + }, + { + "epoch": 0.11, + "grad_norm": 2.276042859793579, + "learning_rate": 1.9629825991912622e-05, + "loss": 1.0161, + "step": 4851 + }, + { + "epoch": 0.11, + "grad_norm": 2.0632314484521395, + "learning_rate": 1.962962027581313e-05, + "loss": 1.1099, + "step": 4852 + }, + { + "epoch": 0.11, + "grad_norm": 2.0869845336918513, + "learning_rate": 1.9629414503646936e-05, + "loss": 1.153, + "step": 4853 + }, + { + "epoch": 0.11, + "grad_norm": 2.0509684990365726, + "learning_rate": 1.9629208675415246e-05, + "loss": 1.1822, + "step": 4854 + }, + { + "epoch": 0.11, + "grad_norm": 1.0335975507735164, + "learning_rate": 1.9629002791119252e-05, + "loss": 0.9251, + "step": 4855 + }, + { + "epoch": 0.11, + "grad_norm": 2.2110085882428265, + "learning_rate": 1.9628796850760156e-05, + "loss": 1.1425, + "step": 4856 + }, + { + "epoch": 0.11, + "grad_norm": 2.0957702515869068, + "learning_rate": 1.962859085433916e-05, + "loss": 0.9436, + "step": 4857 + }, + { + "epoch": 0.11, + "grad_norm": 2.0345523793156266, + "learning_rate": 1.9628384801857454e-05, + "loss": 1.0607, + "step": 4858 + }, + { + "epoch": 0.11, + "grad_norm": 1.9835088066184616, + "learning_rate": 1.9628178693316243e-05, + "loss": 1.1741, + "step": 4859 + }, + { + "epoch": 0.11, + "grad_norm": 2.0747012326596503, + "learning_rate": 1.962797252871673e-05, + "loss": 1.1494, + "step": 4860 + }, + { + "epoch": 0.11, + "grad_norm": 2.23726926577404, + "learning_rate": 1.9627766308060115e-05, + "loss": 1.0285, + "step": 4861 + }, + { + "epoch": 0.11, + "grad_norm": 2.219611947649693, + "learning_rate": 1.9627560031347594e-05, + "loss": 1.138, + "step": 4862 + }, + { + "epoch": 0.11, + "grad_norm": 2.016174932095269, + "learning_rate": 1.962735369858037e-05, + "loss": 1.0585, + "step": 4863 + }, + { + "epoch": 0.11, + "grad_norm": 2.1156121264602907, + "learning_rate": 1.9627147309759646e-05, + "loss": 1.1265, + "step": 4864 + }, + { + "epoch": 0.11, + "grad_norm": 2.435164702860958, + "learning_rate": 1.9626940864886622e-05, + "loss": 0.9568, + "step": 4865 + }, + { + "epoch": 0.11, + "grad_norm": 2.106156205656507, + "learning_rate": 1.96267343639625e-05, + "loss": 1.0312, + "step": 4866 + }, + { + "epoch": 0.11, + "grad_norm": 2.227665685715612, + "learning_rate": 1.9626527806988485e-05, + "loss": 1.2097, + "step": 4867 + }, + { + "epoch": 0.11, + "grad_norm": 2.239357228969005, + "learning_rate": 1.9626321193965774e-05, + "loss": 1.0519, + "step": 4868 + }, + { + "epoch": 0.11, + "grad_norm": 2.4996060214628977, + "learning_rate": 1.9626114524895574e-05, + "loss": 0.9982, + "step": 4869 + }, + { + "epoch": 0.11, + "grad_norm": 1.9050238938739583, + "learning_rate": 1.962590779977909e-05, + "loss": 1.1653, + "step": 4870 + }, + { + "epoch": 0.11, + "grad_norm": 1.9038096843969758, + "learning_rate": 1.9625701018617524e-05, + "loss": 1.068, + "step": 4871 + }, + { + "epoch": 0.11, + "grad_norm": 2.581806410633924, + "learning_rate": 1.9625494181412076e-05, + "loss": 1.1228, + "step": 4872 + }, + { + "epoch": 0.11, + "grad_norm": 2.3876791942770095, + "learning_rate": 1.9625287288163954e-05, + "loss": 1.0576, + "step": 4873 + }, + { + "epoch": 0.11, + "grad_norm": 2.821379995564173, + "learning_rate": 1.9625080338874363e-05, + "loss": 1.0735, + "step": 4874 + }, + { + "epoch": 0.11, + "grad_norm": 2.2916190731448745, + "learning_rate": 1.9624873333544507e-05, + "loss": 1.0899, + "step": 4875 + }, + { + "epoch": 0.11, + "grad_norm": 2.047721876252476, + "learning_rate": 1.9624666272175588e-05, + "loss": 1.0578, + "step": 4876 + }, + { + "epoch": 0.11, + "grad_norm": 2.1846403275713624, + "learning_rate": 1.962445915476882e-05, + "loss": 0.9919, + "step": 4877 + }, + { + "epoch": 0.11, + "grad_norm": 2.6726069740914613, + "learning_rate": 1.9624251981325403e-05, + "loss": 1.1761, + "step": 4878 + }, + { + "epoch": 0.11, + "grad_norm": 2.804698214454029, + "learning_rate": 1.9624044751846543e-05, + "loss": 0.937, + "step": 4879 + }, + { + "epoch": 0.11, + "grad_norm": 2.2349004666337557, + "learning_rate": 1.9623837466333445e-05, + "loss": 0.9194, + "step": 4880 + }, + { + "epoch": 0.11, + "grad_norm": 2.32503914548958, + "learning_rate": 1.962363012478732e-05, + "loss": 1.0752, + "step": 4881 + }, + { + "epoch": 0.12, + "grad_norm": 2.025069930070798, + "learning_rate": 1.9623422727209376e-05, + "loss": 1.0587, + "step": 4882 + }, + { + "epoch": 0.12, + "grad_norm": 2.0994411341218604, + "learning_rate": 1.9623215273600817e-05, + "loss": 1.2099, + "step": 4883 + }, + { + "epoch": 0.12, + "grad_norm": 2.8364247647797662, + "learning_rate": 1.962300776396285e-05, + "loss": 1.0047, + "step": 4884 + }, + { + "epoch": 0.12, + "grad_norm": 2.0700391130060645, + "learning_rate": 1.9622800198296686e-05, + "loss": 1.0488, + "step": 4885 + }, + { + "epoch": 0.12, + "grad_norm": 2.511059570635173, + "learning_rate": 1.9622592576603536e-05, + "loss": 1.0117, + "step": 4886 + }, + { + "epoch": 0.12, + "grad_norm": 2.0678377944850337, + "learning_rate": 1.96223848988846e-05, + "loss": 1.1007, + "step": 4887 + }, + { + "epoch": 0.12, + "grad_norm": 1.095011045994938, + "learning_rate": 1.9622177165141098e-05, + "loss": 0.9485, + "step": 4888 + }, + { + "epoch": 0.12, + "grad_norm": 2.013492690840566, + "learning_rate": 1.962196937537423e-05, + "loss": 1.1417, + "step": 4889 + }, + { + "epoch": 0.12, + "grad_norm": 2.2416223518742537, + "learning_rate": 1.962176152958521e-05, + "loss": 1.0016, + "step": 4890 + }, + { + "epoch": 0.12, + "grad_norm": 2.02857194374144, + "learning_rate": 1.962155362777525e-05, + "loss": 0.9928, + "step": 4891 + }, + { + "epoch": 0.12, + "grad_norm": 1.9811845214623565, + "learning_rate": 1.9621345669945562e-05, + "loss": 1.0917, + "step": 4892 + }, + { + "epoch": 0.12, + "grad_norm": 2.5263355574417563, + "learning_rate": 1.9621137656097347e-05, + "loss": 1.1407, + "step": 4893 + }, + { + "epoch": 0.12, + "grad_norm": 2.0509929033459797, + "learning_rate": 1.9620929586231828e-05, + "loss": 1.1439, + "step": 4894 + }, + { + "epoch": 0.12, + "grad_norm": 2.0617657215929905, + "learning_rate": 1.9620721460350207e-05, + "loss": 1.0729, + "step": 4895 + }, + { + "epoch": 0.12, + "grad_norm": 2.085931306800328, + "learning_rate": 1.9620513278453704e-05, + "loss": 1.1664, + "step": 4896 + }, + { + "epoch": 0.12, + "grad_norm": 2.2225852271550535, + "learning_rate": 1.9620305040543523e-05, + "loss": 1.1061, + "step": 4897 + }, + { + "epoch": 0.12, + "grad_norm": 1.8966013502037926, + "learning_rate": 1.9620096746620882e-05, + "loss": 0.9651, + "step": 4898 + }, + { + "epoch": 0.12, + "grad_norm": 1.9381032849773032, + "learning_rate": 1.9619888396686992e-05, + "loss": 1.1, + "step": 4899 + }, + { + "epoch": 0.12, + "grad_norm": 2.0765256137393737, + "learning_rate": 1.9619679990743065e-05, + "loss": 1.0317, + "step": 4900 + }, + { + "epoch": 0.12, + "grad_norm": 2.331792796066806, + "learning_rate": 1.9619471528790317e-05, + "loss": 1.0093, + "step": 4901 + }, + { + "epoch": 0.12, + "grad_norm": 2.523334280082681, + "learning_rate": 1.961926301082996e-05, + "loss": 0.9985, + "step": 4902 + }, + { + "epoch": 0.12, + "grad_norm": 2.3477068802773937, + "learning_rate": 1.9619054436863206e-05, + "loss": 0.9783, + "step": 4903 + }, + { + "epoch": 0.12, + "grad_norm": 1.1420687489642891, + "learning_rate": 1.9618845806891276e-05, + "loss": 1.0105, + "step": 4904 + }, + { + "epoch": 0.12, + "grad_norm": 3.4462060122483216, + "learning_rate": 1.9618637120915376e-05, + "loss": 1.1557, + "step": 4905 + }, + { + "epoch": 0.12, + "grad_norm": 2.028722227463984, + "learning_rate": 1.9618428378936728e-05, + "loss": 1.1209, + "step": 4906 + }, + { + "epoch": 0.12, + "grad_norm": 2.026505722687675, + "learning_rate": 1.9618219580956544e-05, + "loss": 1.0145, + "step": 4907 + }, + { + "epoch": 0.12, + "grad_norm": 1.994368620573779, + "learning_rate": 1.9618010726976044e-05, + "loss": 1.1009, + "step": 4908 + }, + { + "epoch": 0.12, + "grad_norm": 2.195395573939676, + "learning_rate": 1.9617801816996437e-05, + "loss": 1.082, + "step": 4909 + }, + { + "epoch": 0.12, + "grad_norm": 2.2747861082689496, + "learning_rate": 1.961759285101894e-05, + "loss": 0.9991, + "step": 4910 + }, + { + "epoch": 0.12, + "grad_norm": 2.160002275798823, + "learning_rate": 1.961738382904478e-05, + "loss": 1.0225, + "step": 4911 + }, + { + "epoch": 0.12, + "grad_norm": 1.9975843075906443, + "learning_rate": 1.961717475107516e-05, + "loss": 1.1082, + "step": 4912 + }, + { + "epoch": 0.12, + "grad_norm": 2.0447047037651895, + "learning_rate": 1.9616965617111305e-05, + "loss": 1.0707, + "step": 4913 + }, + { + "epoch": 0.12, + "grad_norm": 2.2663439006272634, + "learning_rate": 1.9616756427154434e-05, + "loss": 1.18, + "step": 4914 + }, + { + "epoch": 0.12, + "grad_norm": 2.292411443627897, + "learning_rate": 1.961654718120576e-05, + "loss": 1.064, + "step": 4915 + }, + { + "epoch": 0.12, + "grad_norm": 1.952085330815211, + "learning_rate": 1.9616337879266507e-05, + "loss": 0.9882, + "step": 4916 + }, + { + "epoch": 0.12, + "grad_norm": 1.137756899570634, + "learning_rate": 1.9616128521337886e-05, + "loss": 0.9673, + "step": 4917 + }, + { + "epoch": 0.12, + "grad_norm": 2.080978003690739, + "learning_rate": 1.9615919107421123e-05, + "loss": 0.9772, + "step": 4918 + }, + { + "epoch": 0.12, + "grad_norm": 1.2315527583961192, + "learning_rate": 1.9615709637517437e-05, + "loss": 1.1104, + "step": 4919 + }, + { + "epoch": 0.12, + "grad_norm": 1.8929233949614486, + "learning_rate": 1.961550011162804e-05, + "loss": 1.0407, + "step": 4920 + }, + { + "epoch": 0.12, + "grad_norm": 2.0934737305914477, + "learning_rate": 1.9615290529754162e-05, + "loss": 1.139, + "step": 4921 + }, + { + "epoch": 0.12, + "grad_norm": 2.099474218919838, + "learning_rate": 1.961508089189702e-05, + "loss": 1.1858, + "step": 4922 + }, + { + "epoch": 0.12, + "grad_norm": 2.0745325287817264, + "learning_rate": 1.9614871198057826e-05, + "loss": 1.1738, + "step": 4923 + }, + { + "epoch": 0.12, + "grad_norm": 2.2529059107277734, + "learning_rate": 1.961466144823781e-05, + "loss": 1.0672, + "step": 4924 + }, + { + "epoch": 0.12, + "grad_norm": 1.82778149481588, + "learning_rate": 1.9614451642438193e-05, + "loss": 0.8917, + "step": 4925 + }, + { + "epoch": 0.12, + "grad_norm": 2.0529302287879054, + "learning_rate": 1.9614241780660196e-05, + "loss": 1.1894, + "step": 4926 + }, + { + "epoch": 0.12, + "grad_norm": 2.040585907389859, + "learning_rate": 1.9614031862905034e-05, + "loss": 1.1043, + "step": 4927 + }, + { + "epoch": 0.12, + "grad_norm": 2.9087168815058155, + "learning_rate": 1.961382188917394e-05, + "loss": 1.1382, + "step": 4928 + }, + { + "epoch": 0.12, + "grad_norm": 2.1632972240590096, + "learning_rate": 1.9613611859468128e-05, + "loss": 1.1877, + "step": 4929 + }, + { + "epoch": 0.12, + "grad_norm": 1.1906655336520908, + "learning_rate": 1.9613401773788826e-05, + "loss": 0.9861, + "step": 4930 + }, + { + "epoch": 0.12, + "grad_norm": 1.9670781522305973, + "learning_rate": 1.9613191632137256e-05, + "loss": 1.194, + "step": 4931 + }, + { + "epoch": 0.12, + "grad_norm": 2.235422222025597, + "learning_rate": 1.961298143451464e-05, + "loss": 0.9739, + "step": 4932 + }, + { + "epoch": 0.12, + "grad_norm": 2.0701924008046784, + "learning_rate": 1.9612771180922203e-05, + "loss": 1.1494, + "step": 4933 + }, + { + "epoch": 0.12, + "grad_norm": 2.07077002969102, + "learning_rate": 1.9612560871361163e-05, + "loss": 1.1301, + "step": 4934 + }, + { + "epoch": 0.12, + "grad_norm": 1.1634733886348871, + "learning_rate": 1.9612350505832757e-05, + "loss": 0.9752, + "step": 4935 + }, + { + "epoch": 0.12, + "grad_norm": 2.3007991668435865, + "learning_rate": 1.9612140084338204e-05, + "loss": 1.031, + "step": 4936 + }, + { + "epoch": 0.12, + "grad_norm": 1.1251585156635517, + "learning_rate": 1.9611929606878724e-05, + "loss": 0.8812, + "step": 4937 + }, + { + "epoch": 0.12, + "grad_norm": 2.0115342931762865, + "learning_rate": 1.9611719073455547e-05, + "loss": 1.0771, + "step": 4938 + }, + { + "epoch": 0.12, + "grad_norm": 2.0275306137786453, + "learning_rate": 1.9611508484069902e-05, + "loss": 1.0876, + "step": 4939 + }, + { + "epoch": 0.12, + "grad_norm": 4.4119365169302975, + "learning_rate": 1.961129783872301e-05, + "loss": 1.1297, + "step": 4940 + }, + { + "epoch": 0.12, + "grad_norm": 1.882027521202472, + "learning_rate": 1.9611087137416097e-05, + "loss": 1.0279, + "step": 4941 + }, + { + "epoch": 0.12, + "grad_norm": 1.9245666257981728, + "learning_rate": 1.9610876380150396e-05, + "loss": 0.935, + "step": 4942 + }, + { + "epoch": 0.12, + "grad_norm": 1.9554057476756777, + "learning_rate": 1.9610665566927127e-05, + "loss": 1.228, + "step": 4943 + }, + { + "epoch": 0.12, + "grad_norm": 2.1073248221514405, + "learning_rate": 1.961045469774752e-05, + "loss": 1.0141, + "step": 4944 + }, + { + "epoch": 0.12, + "grad_norm": 2.2162302531364393, + "learning_rate": 1.9610243772612804e-05, + "loss": 1.0059, + "step": 4945 + }, + { + "epoch": 0.12, + "grad_norm": 2.521216649502672, + "learning_rate": 1.9610032791524206e-05, + "loss": 1.0133, + "step": 4946 + }, + { + "epoch": 0.12, + "grad_norm": 2.009277990348035, + "learning_rate": 1.9609821754482957e-05, + "loss": 1.1033, + "step": 4947 + }, + { + "epoch": 0.12, + "grad_norm": 2.3159094203499255, + "learning_rate": 1.960961066149028e-05, + "loss": 1.1462, + "step": 4948 + }, + { + "epoch": 0.12, + "grad_norm": 2.9310225110560704, + "learning_rate": 1.960939951254741e-05, + "loss": 1.0131, + "step": 4949 + }, + { + "epoch": 0.12, + "grad_norm": 1.9396628047265818, + "learning_rate": 1.9609188307655572e-05, + "loss": 1.0428, + "step": 4950 + }, + { + "epoch": 0.12, + "grad_norm": 1.2500016815949095, + "learning_rate": 1.9608977046815997e-05, + "loss": 1.0288, + "step": 4951 + }, + { + "epoch": 0.12, + "grad_norm": 1.9251925870135562, + "learning_rate": 1.9608765730029917e-05, + "loss": 1.1371, + "step": 4952 + }, + { + "epoch": 0.12, + "grad_norm": 1.9807773144914893, + "learning_rate": 1.960855435729856e-05, + "loss": 1.0892, + "step": 4953 + }, + { + "epoch": 0.12, + "grad_norm": 2.7553417659539234, + "learning_rate": 1.9608342928623157e-05, + "loss": 0.9566, + "step": 4954 + }, + { + "epoch": 0.12, + "grad_norm": 2.2906053743669865, + "learning_rate": 1.960813144400494e-05, + "loss": 1.0448, + "step": 4955 + }, + { + "epoch": 0.12, + "grad_norm": 2.054618465556575, + "learning_rate": 1.960791990344514e-05, + "loss": 1.1563, + "step": 4956 + }, + { + "epoch": 0.12, + "grad_norm": 2.5901107545264748, + "learning_rate": 1.960770830694499e-05, + "loss": 1.0832, + "step": 4957 + }, + { + "epoch": 0.12, + "grad_norm": 2.5910391326920825, + "learning_rate": 1.960749665450572e-05, + "loss": 1.0682, + "step": 4958 + }, + { + "epoch": 0.12, + "grad_norm": 2.0065451936964283, + "learning_rate": 1.9607284946128562e-05, + "loss": 1.1312, + "step": 4959 + }, + { + "epoch": 0.12, + "grad_norm": 2.0894982148110444, + "learning_rate": 1.9607073181814748e-05, + "loss": 1.151, + "step": 4960 + }, + { + "epoch": 0.12, + "grad_norm": 2.9256405363636055, + "learning_rate": 1.9606861361565514e-05, + "loss": 1.025, + "step": 4961 + }, + { + "epoch": 0.12, + "grad_norm": 2.3208219008364623, + "learning_rate": 1.960664948538209e-05, + "loss": 1.1136, + "step": 4962 + }, + { + "epoch": 0.12, + "grad_norm": 2.0304595822712224, + "learning_rate": 1.9606437553265713e-05, + "loss": 1.1023, + "step": 4963 + }, + { + "epoch": 0.12, + "grad_norm": 2.1754272688135576, + "learning_rate": 1.9606225565217615e-05, + "loss": 0.9305, + "step": 4964 + }, + { + "epoch": 0.12, + "grad_norm": 2.2164321382995333, + "learning_rate": 1.960601352123903e-05, + "loss": 1.0445, + "step": 4965 + }, + { + "epoch": 0.12, + "grad_norm": 2.662569708356558, + "learning_rate": 1.9605801421331194e-05, + "loss": 1.0464, + "step": 4966 + }, + { + "epoch": 0.12, + "grad_norm": 2.210589275511278, + "learning_rate": 1.960558926549534e-05, + "loss": 0.9914, + "step": 4967 + }, + { + "epoch": 0.12, + "grad_norm": 1.9867554076574665, + "learning_rate": 1.9605377053732705e-05, + "loss": 1.1693, + "step": 4968 + }, + { + "epoch": 0.12, + "grad_norm": 4.725455512410713, + "learning_rate": 1.9605164786044523e-05, + "loss": 1.024, + "step": 4969 + }, + { + "epoch": 0.12, + "grad_norm": 2.159343234510339, + "learning_rate": 1.9604952462432032e-05, + "loss": 0.9729, + "step": 4970 + }, + { + "epoch": 0.12, + "grad_norm": 2.372992221571827, + "learning_rate": 1.9604740082896467e-05, + "loss": 0.9966, + "step": 4971 + }, + { + "epoch": 0.12, + "grad_norm": 1.1764916334603401, + "learning_rate": 1.960452764743906e-05, + "loss": 1.0069, + "step": 4972 + }, + { + "epoch": 0.12, + "grad_norm": 2.479595220403794, + "learning_rate": 1.960431515606106e-05, + "loss": 1.0984, + "step": 4973 + }, + { + "epoch": 0.12, + "grad_norm": 2.017560658974763, + "learning_rate": 1.960410260876369e-05, + "loss": 1.1348, + "step": 4974 + }, + { + "epoch": 0.12, + "grad_norm": 1.949442907671665, + "learning_rate": 1.9603890005548198e-05, + "loss": 1.1205, + "step": 4975 + }, + { + "epoch": 0.12, + "grad_norm": 2.5155326972748075, + "learning_rate": 1.9603677346415815e-05, + "loss": 1.0426, + "step": 4976 + }, + { + "epoch": 0.12, + "grad_norm": 2.6297388743883294, + "learning_rate": 1.9603464631367782e-05, + "loss": 1.027, + "step": 4977 + }, + { + "epoch": 0.12, + "grad_norm": 2.1866599155782533, + "learning_rate": 1.960325186040534e-05, + "loss": 1.1644, + "step": 4978 + }, + { + "epoch": 0.12, + "grad_norm": 2.4299515028020773, + "learning_rate": 1.9603039033529723e-05, + "loss": 1.1326, + "step": 4979 + }, + { + "epoch": 0.12, + "grad_norm": 2.03842211881161, + "learning_rate": 1.9602826150742174e-05, + "loss": 0.9744, + "step": 4980 + }, + { + "epoch": 0.12, + "grad_norm": 2.2328103042330465, + "learning_rate": 1.960261321204393e-05, + "loss": 1.1711, + "step": 4981 + }, + { + "epoch": 0.12, + "grad_norm": 2.2409201015682196, + "learning_rate": 1.960240021743623e-05, + "loss": 1.1724, + "step": 4982 + }, + { + "epoch": 0.12, + "grad_norm": 2.2308617910919106, + "learning_rate": 1.960218716692032e-05, + "loss": 1.0268, + "step": 4983 + }, + { + "epoch": 0.12, + "grad_norm": 2.0640912392916295, + "learning_rate": 1.9601974060497433e-05, + "loss": 1.0904, + "step": 4984 + }, + { + "epoch": 0.12, + "grad_norm": 2.385653932053897, + "learning_rate": 1.9601760898168814e-05, + "loss": 1.033, + "step": 4985 + }, + { + "epoch": 0.12, + "grad_norm": 2.036534862252122, + "learning_rate": 1.9601547679935706e-05, + "loss": 1.042, + "step": 4986 + }, + { + "epoch": 0.12, + "grad_norm": 2.5897848101348933, + "learning_rate": 1.9601334405799344e-05, + "loss": 1.1621, + "step": 4987 + }, + { + "epoch": 0.12, + "grad_norm": 8.157099971213679, + "learning_rate": 1.9601121075760977e-05, + "loss": 1.0814, + "step": 4988 + }, + { + "epoch": 0.12, + "grad_norm": 2.5234319748463685, + "learning_rate": 1.960090768982184e-05, + "loss": 1.0851, + "step": 4989 + }, + { + "epoch": 0.12, + "grad_norm": 2.382128552280535, + "learning_rate": 1.960069424798318e-05, + "loss": 0.9564, + "step": 4990 + }, + { + "epoch": 0.12, + "grad_norm": 2.104404757646251, + "learning_rate": 1.9600480750246235e-05, + "loss": 1.0473, + "step": 4991 + }, + { + "epoch": 0.12, + "grad_norm": 2.3517327860867723, + "learning_rate": 1.9600267196612253e-05, + "loss": 1.0021, + "step": 4992 + }, + { + "epoch": 0.12, + "grad_norm": 1.1870802232282838, + "learning_rate": 1.960005358708248e-05, + "loss": 0.9945, + "step": 4993 + }, + { + "epoch": 0.12, + "grad_norm": 2.2059188902719504, + "learning_rate": 1.9599839921658154e-05, + "loss": 1.0824, + "step": 4994 + }, + { + "epoch": 0.12, + "grad_norm": 2.1903009909401128, + "learning_rate": 1.9599626200340523e-05, + "loss": 1.0489, + "step": 4995 + }, + { + "epoch": 0.12, + "grad_norm": 2.1978565804203796, + "learning_rate": 1.9599412423130823e-05, + "loss": 1.1171, + "step": 4996 + }, + { + "epoch": 0.12, + "grad_norm": 1.806024098748329, + "learning_rate": 1.9599198590030308e-05, + "loss": 1.0392, + "step": 4997 + }, + { + "epoch": 0.12, + "grad_norm": 1.1704218909208879, + "learning_rate": 1.959898470104022e-05, + "loss": 0.9626, + "step": 4998 + }, + { + "epoch": 0.12, + "grad_norm": 2.152522626490796, + "learning_rate": 1.9598770756161807e-05, + "loss": 1.2366, + "step": 4999 + }, + { + "epoch": 0.12, + "grad_norm": 2.166127453863376, + "learning_rate": 1.9598556755396307e-05, + "loss": 1.0763, + "step": 5000 + }, + { + "epoch": 0.12, + "grad_norm": 2.5959008759911995, + "learning_rate": 1.9598342698744973e-05, + "loss": 1.0421, + "step": 5001 + }, + { + "epoch": 0.12, + "grad_norm": 2.752692595766774, + "learning_rate": 1.9598128586209047e-05, + "loss": 1.0573, + "step": 5002 + }, + { + "epoch": 0.12, + "grad_norm": 2.8566156942138266, + "learning_rate": 1.959791441778978e-05, + "loss": 1.1741, + "step": 5003 + }, + { + "epoch": 0.12, + "grad_norm": 2.189078033115277, + "learning_rate": 1.9597700193488418e-05, + "loss": 0.968, + "step": 5004 + }, + { + "epoch": 0.12, + "grad_norm": 1.9831929242416682, + "learning_rate": 1.9597485913306203e-05, + "loss": 1.0499, + "step": 5005 + }, + { + "epoch": 0.12, + "grad_norm": 2.2392772402546535, + "learning_rate": 1.959727157724439e-05, + "loss": 1.116, + "step": 5006 + }, + { + "epoch": 0.12, + "grad_norm": 2.0188437212474493, + "learning_rate": 1.959705718530422e-05, + "loss": 1.0624, + "step": 5007 + }, + { + "epoch": 0.12, + "grad_norm": 1.9186814189806167, + "learning_rate": 1.9596842737486945e-05, + "loss": 1.1727, + "step": 5008 + }, + { + "epoch": 0.12, + "grad_norm": 1.910459437752704, + "learning_rate": 1.9596628233793814e-05, + "loss": 1.1186, + "step": 5009 + }, + { + "epoch": 0.12, + "grad_norm": 1.9962495894168892, + "learning_rate": 1.9596413674226076e-05, + "loss": 1.0535, + "step": 5010 + }, + { + "epoch": 0.12, + "grad_norm": 2.0677822590496815, + "learning_rate": 1.959619905878498e-05, + "loss": 1.065, + "step": 5011 + }, + { + "epoch": 0.12, + "grad_norm": 3.382567696655605, + "learning_rate": 1.9595984387471772e-05, + "loss": 0.94, + "step": 5012 + }, + { + "epoch": 0.12, + "grad_norm": 2.5054597896532274, + "learning_rate": 1.959576966028771e-05, + "loss": 1.0474, + "step": 5013 + }, + { + "epoch": 0.12, + "grad_norm": 1.1899469893601982, + "learning_rate": 1.9595554877234036e-05, + "loss": 1.0603, + "step": 5014 + }, + { + "epoch": 0.12, + "grad_norm": 2.269735382111525, + "learning_rate": 1.9595340038312e-05, + "loss": 1.0782, + "step": 5015 + }, + { + "epoch": 0.12, + "grad_norm": 1.976900671123798, + "learning_rate": 1.959512514352286e-05, + "loss": 1.1273, + "step": 5016 + }, + { + "epoch": 0.12, + "grad_norm": 2.007116108849235, + "learning_rate": 1.9594910192867866e-05, + "loss": 0.9946, + "step": 5017 + }, + { + "epoch": 0.12, + "grad_norm": 2.1999954961702186, + "learning_rate": 1.9594695186348265e-05, + "loss": 1.0571, + "step": 5018 + }, + { + "epoch": 0.12, + "grad_norm": 2.3551679726781343, + "learning_rate": 1.959448012396531e-05, + "loss": 1.0493, + "step": 5019 + }, + { + "epoch": 0.12, + "grad_norm": 2.147190088081042, + "learning_rate": 1.9594265005720255e-05, + "loss": 1.0964, + "step": 5020 + }, + { + "epoch": 0.12, + "grad_norm": 1.9533299155624304, + "learning_rate": 1.9594049831614352e-05, + "loss": 1.1212, + "step": 5021 + }, + { + "epoch": 0.12, + "grad_norm": 1.2126357810759814, + "learning_rate": 1.9593834601648853e-05, + "loss": 0.924, + "step": 5022 + }, + { + "epoch": 0.12, + "grad_norm": 2.365440194320243, + "learning_rate": 1.9593619315825013e-05, + "loss": 1.153, + "step": 5023 + }, + { + "epoch": 0.12, + "grad_norm": 2.243542729492662, + "learning_rate": 1.9593403974144084e-05, + "loss": 1.0733, + "step": 5024 + }, + { + "epoch": 0.12, + "grad_norm": 1.1313770948505781, + "learning_rate": 1.9593188576607316e-05, + "loss": 0.991, + "step": 5025 + }, + { + "epoch": 0.12, + "grad_norm": 2.6168646188031346, + "learning_rate": 1.959297312321597e-05, + "loss": 1.0771, + "step": 5026 + }, + { + "epoch": 0.12, + "grad_norm": 2.0573301323280444, + "learning_rate": 1.95927576139713e-05, + "loss": 0.939, + "step": 5027 + }, + { + "epoch": 0.12, + "grad_norm": 2.222270926519324, + "learning_rate": 1.959254204887455e-05, + "loss": 1.0213, + "step": 5028 + }, + { + "epoch": 0.12, + "grad_norm": 2.4029891983620137, + "learning_rate": 1.959232642792699e-05, + "loss": 1.0764, + "step": 5029 + }, + { + "epoch": 0.12, + "grad_norm": 2.255437104121576, + "learning_rate": 1.959211075112987e-05, + "loss": 1.0529, + "step": 5030 + }, + { + "epoch": 0.12, + "grad_norm": 2.386505961004895, + "learning_rate": 1.959189501848444e-05, + "loss": 1.0943, + "step": 5031 + }, + { + "epoch": 0.12, + "grad_norm": 1.8808114886501248, + "learning_rate": 1.9591679229991963e-05, + "loss": 1.0656, + "step": 5032 + }, + { + "epoch": 0.12, + "grad_norm": 2.194419475785188, + "learning_rate": 1.959146338565369e-05, + "loss": 1.125, + "step": 5033 + }, + { + "epoch": 0.12, + "grad_norm": 3.8309462940616275, + "learning_rate": 1.959124748547088e-05, + "loss": 1.0718, + "step": 5034 + }, + { + "epoch": 0.12, + "grad_norm": 2.1415082659623286, + "learning_rate": 1.9591031529444795e-05, + "loss": 1.0877, + "step": 5035 + }, + { + "epoch": 0.12, + "grad_norm": 1.9594062686513514, + "learning_rate": 1.9590815517576684e-05, + "loss": 1.1523, + "step": 5036 + }, + { + "epoch": 0.12, + "grad_norm": 2.066433358273784, + "learning_rate": 1.9590599449867813e-05, + "loss": 1.095, + "step": 5037 + }, + { + "epoch": 0.12, + "grad_norm": 1.163488883780453, + "learning_rate": 1.959038332631943e-05, + "loss": 1.0169, + "step": 5038 + }, + { + "epoch": 0.12, + "grad_norm": 2.165110049312111, + "learning_rate": 1.9590167146932804e-05, + "loss": 1.0973, + "step": 5039 + }, + { + "epoch": 0.12, + "grad_norm": 1.9285366800513746, + "learning_rate": 1.958995091170919e-05, + "loss": 0.9619, + "step": 5040 + }, + { + "epoch": 0.12, + "grad_norm": 2.0438167348294565, + "learning_rate": 1.958973462064984e-05, + "loss": 0.9978, + "step": 5041 + }, + { + "epoch": 0.12, + "grad_norm": 2.3198214079081683, + "learning_rate": 1.9589518273756023e-05, + "loss": 1.1337, + "step": 5042 + }, + { + "epoch": 0.12, + "grad_norm": 2.742353774978881, + "learning_rate": 1.9589301871028993e-05, + "loss": 0.9458, + "step": 5043 + }, + { + "epoch": 0.12, + "grad_norm": 1.9245536956026392, + "learning_rate": 1.958908541247001e-05, + "loss": 1.0241, + "step": 5044 + }, + { + "epoch": 0.12, + "grad_norm": 2.0456817388145083, + "learning_rate": 1.9588868898080337e-05, + "loss": 1.115, + "step": 5045 + }, + { + "epoch": 0.12, + "grad_norm": 1.9423578059579956, + "learning_rate": 1.9588652327861236e-05, + "loss": 1.094, + "step": 5046 + }, + { + "epoch": 0.12, + "grad_norm": 2.2108747773321604, + "learning_rate": 1.9588435701813962e-05, + "loss": 1.0395, + "step": 5047 + }, + { + "epoch": 0.12, + "grad_norm": 2.2884707899550687, + "learning_rate": 1.958821901993978e-05, + "loss": 1.0043, + "step": 5048 + }, + { + "epoch": 0.12, + "grad_norm": 1.9531988942949714, + "learning_rate": 1.9588002282239955e-05, + "loss": 0.9611, + "step": 5049 + }, + { + "epoch": 0.12, + "grad_norm": 1.2752889147915398, + "learning_rate": 1.9587785488715742e-05, + "loss": 1.0633, + "step": 5050 + }, + { + "epoch": 0.12, + "grad_norm": 2.113144461727227, + "learning_rate": 1.9587568639368405e-05, + "loss": 1.1003, + "step": 5051 + }, + { + "epoch": 0.12, + "grad_norm": 2.159917408421517, + "learning_rate": 1.958735173419921e-05, + "loss": 1.1544, + "step": 5052 + }, + { + "epoch": 0.12, + "grad_norm": 2.3996201569524387, + "learning_rate": 1.958713477320942e-05, + "loss": 1.0953, + "step": 5053 + }, + { + "epoch": 0.12, + "grad_norm": 2.3600845037870584, + "learning_rate": 1.9586917756400293e-05, + "loss": 1.0174, + "step": 5054 + }, + { + "epoch": 0.12, + "grad_norm": 2.197841818071868, + "learning_rate": 1.95867006837731e-05, + "loss": 1.0142, + "step": 5055 + }, + { + "epoch": 0.12, + "grad_norm": 2.743261323079185, + "learning_rate": 1.9586483555329098e-05, + "loss": 1.0424, + "step": 5056 + }, + { + "epoch": 0.12, + "grad_norm": 1.8814880051180058, + "learning_rate": 1.958626637106955e-05, + "loss": 1.0336, + "step": 5057 + }, + { + "epoch": 0.12, + "grad_norm": 1.1715171029027514, + "learning_rate": 1.9586049130995728e-05, + "loss": 1.0271, + "step": 5058 + }, + { + "epoch": 0.12, + "grad_norm": 1.878934738573699, + "learning_rate": 1.958583183510889e-05, + "loss": 1.1849, + "step": 5059 + }, + { + "epoch": 0.12, + "grad_norm": 2.117942719240646, + "learning_rate": 1.9585614483410308e-05, + "loss": 1.018, + "step": 5060 + }, + { + "epoch": 0.12, + "grad_norm": 2.267624505974972, + "learning_rate": 1.9585397075901242e-05, + "loss": 1.1024, + "step": 5061 + }, + { + "epoch": 0.12, + "grad_norm": 2.2765756751533752, + "learning_rate": 1.958517961258296e-05, + "loss": 1.1826, + "step": 5062 + }, + { + "epoch": 0.12, + "grad_norm": 1.9307582952341598, + "learning_rate": 1.958496209345673e-05, + "loss": 1.0894, + "step": 5063 + }, + { + "epoch": 0.12, + "grad_norm": 2.118485037524843, + "learning_rate": 1.9584744518523812e-05, + "loss": 1.0331, + "step": 5064 + }, + { + "epoch": 0.12, + "grad_norm": 2.3579419933827137, + "learning_rate": 1.958452688778548e-05, + "loss": 0.9911, + "step": 5065 + }, + { + "epoch": 0.12, + "grad_norm": 1.176646011826955, + "learning_rate": 1.9584309201242993e-05, + "loss": 0.9596, + "step": 5066 + }, + { + "epoch": 0.12, + "grad_norm": 2.054666290835509, + "learning_rate": 1.958409145889763e-05, + "loss": 1.1701, + "step": 5067 + }, + { + "epoch": 0.12, + "grad_norm": 2.184248007875264, + "learning_rate": 1.9583873660750646e-05, + "loss": 1.032, + "step": 5068 + }, + { + "epoch": 0.12, + "grad_norm": 2.2772750075288752, + "learning_rate": 1.958365580680332e-05, + "loss": 1.2261, + "step": 5069 + }, + { + "epoch": 0.12, + "grad_norm": 2.1896996028648217, + "learning_rate": 1.9583437897056915e-05, + "loss": 1.0977, + "step": 5070 + }, + { + "epoch": 0.12, + "grad_norm": 4.4788304489434605, + "learning_rate": 1.95832199315127e-05, + "loss": 1.082, + "step": 5071 + }, + { + "epoch": 0.12, + "grad_norm": 2.7400412710685185, + "learning_rate": 1.9583001910171947e-05, + "loss": 1.0318, + "step": 5072 + }, + { + "epoch": 0.12, + "grad_norm": 2.2499496232666574, + "learning_rate": 1.958278383303592e-05, + "loss": 1.0507, + "step": 5073 + }, + { + "epoch": 0.12, + "grad_norm": 3.607844408267138, + "learning_rate": 1.9582565700105893e-05, + "loss": 1.051, + "step": 5074 + }, + { + "epoch": 0.12, + "grad_norm": 2.220157111306372, + "learning_rate": 1.958234751138313e-05, + "loss": 1.132, + "step": 5075 + }, + { + "epoch": 0.12, + "grad_norm": 2.3733414711409226, + "learning_rate": 1.9582129266868914e-05, + "loss": 1.0895, + "step": 5076 + }, + { + "epoch": 0.12, + "grad_norm": 2.2833864568870457, + "learning_rate": 1.9581910966564502e-05, + "loss": 1.0589, + "step": 5077 + }, + { + "epoch": 0.12, + "grad_norm": 2.222678515208774, + "learning_rate": 1.9581692610471174e-05, + "loss": 1.1456, + "step": 5078 + }, + { + "epoch": 0.12, + "grad_norm": 2.1744798853248657, + "learning_rate": 1.9581474198590196e-05, + "loss": 1.0982, + "step": 5079 + }, + { + "epoch": 0.12, + "grad_norm": 1.9585908352332229, + "learning_rate": 1.9581255730922842e-05, + "loss": 1.1324, + "step": 5080 + }, + { + "epoch": 0.12, + "grad_norm": 2.3063380189161977, + "learning_rate": 1.9581037207470382e-05, + "loss": 1.0839, + "step": 5081 + }, + { + "epoch": 0.12, + "grad_norm": 1.9368610354722844, + "learning_rate": 1.9580818628234095e-05, + "loss": 1.0921, + "step": 5082 + }, + { + "epoch": 0.12, + "grad_norm": 2.2157012893616295, + "learning_rate": 1.9580599993215243e-05, + "loss": 0.9415, + "step": 5083 + }, + { + "epoch": 0.12, + "grad_norm": 2.1109269325156315, + "learning_rate": 1.9580381302415107e-05, + "loss": 1.0091, + "step": 5084 + }, + { + "epoch": 0.12, + "grad_norm": 1.252353116691743, + "learning_rate": 1.9580162555834962e-05, + "loss": 1.0419, + "step": 5085 + }, + { + "epoch": 0.12, + "grad_norm": 1.1029220522779581, + "learning_rate": 1.957994375347607e-05, + "loss": 0.8905, + "step": 5086 + }, + { + "epoch": 0.12, + "grad_norm": 2.210733388794275, + "learning_rate": 1.957972489533972e-05, + "loss": 1.0418, + "step": 5087 + }, + { + "epoch": 0.12, + "grad_norm": 2.145435454550957, + "learning_rate": 1.9579505981427177e-05, + "loss": 1.1715, + "step": 5088 + }, + { + "epoch": 0.12, + "grad_norm": 2.268006528160461, + "learning_rate": 1.9579287011739712e-05, + "loss": 1.0984, + "step": 5089 + }, + { + "epoch": 0.12, + "grad_norm": 1.1579521841725782, + "learning_rate": 1.957906798627861e-05, + "loss": 1.0035, + "step": 5090 + }, + { + "epoch": 0.12, + "grad_norm": 2.2505644674476284, + "learning_rate": 1.957884890504514e-05, + "loss": 0.9462, + "step": 5091 + }, + { + "epoch": 0.12, + "grad_norm": 2.224673825788075, + "learning_rate": 1.957862976804058e-05, + "loss": 1.009, + "step": 5092 + }, + { + "epoch": 0.12, + "grad_norm": 3.1375346401612205, + "learning_rate": 1.9578410575266203e-05, + "loss": 1.069, + "step": 5093 + }, + { + "epoch": 0.12, + "grad_norm": 2.2214268957670376, + "learning_rate": 1.9578191326723292e-05, + "loss": 1.0737, + "step": 5094 + }, + { + "epoch": 0.12, + "grad_norm": 2.0413836534096603, + "learning_rate": 1.9577972022413113e-05, + "loss": 0.9523, + "step": 5095 + }, + { + "epoch": 0.12, + "grad_norm": 1.8629586476708004, + "learning_rate": 1.9577752662336952e-05, + "loss": 0.9623, + "step": 5096 + }, + { + "epoch": 0.12, + "grad_norm": 2.158690805268997, + "learning_rate": 1.957753324649608e-05, + "loss": 1.0196, + "step": 5097 + }, + { + "epoch": 0.12, + "grad_norm": 2.399373692179494, + "learning_rate": 1.957731377489178e-05, + "loss": 1.0556, + "step": 5098 + }, + { + "epoch": 0.12, + "grad_norm": 2.134154601943324, + "learning_rate": 1.9577094247525324e-05, + "loss": 1.1287, + "step": 5099 + }, + { + "epoch": 0.12, + "grad_norm": 2.1752162049617367, + "learning_rate": 1.9576874664397998e-05, + "loss": 0.9959, + "step": 5100 + }, + { + "epoch": 0.12, + "grad_norm": 1.8046203049622236, + "learning_rate": 1.9576655025511072e-05, + "loss": 1.0456, + "step": 5101 + }, + { + "epoch": 0.12, + "grad_norm": 1.1116460096637815, + "learning_rate": 1.957643533086583e-05, + "loss": 0.9821, + "step": 5102 + }, + { + "epoch": 0.12, + "grad_norm": 2.132822621084834, + "learning_rate": 1.957621558046355e-05, + "loss": 1.0934, + "step": 5103 + }, + { + "epoch": 0.12, + "grad_norm": 1.082981121329404, + "learning_rate": 1.9575995774305506e-05, + "loss": 0.9607, + "step": 5104 + }, + { + "epoch": 0.12, + "grad_norm": 1.1455350494154863, + "learning_rate": 1.9575775912392992e-05, + "loss": 0.9705, + "step": 5105 + }, + { + "epoch": 0.12, + "grad_norm": 2.322389393328373, + "learning_rate": 1.9575555994727273e-05, + "loss": 1.0518, + "step": 5106 + }, + { + "epoch": 0.12, + "grad_norm": 2.2210601814838222, + "learning_rate": 1.9575336021309635e-05, + "loss": 1.0587, + "step": 5107 + }, + { + "epoch": 0.12, + "grad_norm": 2.2466116736711004, + "learning_rate": 1.957511599214136e-05, + "loss": 1.0955, + "step": 5108 + }, + { + "epoch": 0.12, + "grad_norm": 1.0688444911270365, + "learning_rate": 1.957489590722373e-05, + "loss": 0.9862, + "step": 5109 + }, + { + "epoch": 0.12, + "grad_norm": 1.9689743095387844, + "learning_rate": 1.9574675766558023e-05, + "loss": 1.0335, + "step": 5110 + }, + { + "epoch": 0.12, + "grad_norm": 2.2674463931631283, + "learning_rate": 1.9574455570145524e-05, + "loss": 1.0876, + "step": 5111 + }, + { + "epoch": 0.12, + "grad_norm": 2.179040492010567, + "learning_rate": 1.9574235317987514e-05, + "loss": 1.0185, + "step": 5112 + }, + { + "epoch": 0.12, + "grad_norm": 6.017187964181019, + "learning_rate": 1.9574015010085273e-05, + "loss": 0.9273, + "step": 5113 + }, + { + "epoch": 0.12, + "grad_norm": 2.3189911615198633, + "learning_rate": 1.9573794646440084e-05, + "loss": 1.0795, + "step": 5114 + }, + { + "epoch": 0.12, + "grad_norm": 2.1746854496759966, + "learning_rate": 1.9573574227053233e-05, + "loss": 0.9728, + "step": 5115 + }, + { + "epoch": 0.12, + "grad_norm": 2.09962319436072, + "learning_rate": 1.9573353751926004e-05, + "loss": 1.0721, + "step": 5116 + }, + { + "epoch": 0.12, + "grad_norm": 2.111809056991402, + "learning_rate": 1.9573133221059676e-05, + "loss": 1.0089, + "step": 5117 + }, + { + "epoch": 0.12, + "grad_norm": 2.7565294632364776, + "learning_rate": 1.9572912634455538e-05, + "loss": 1.1475, + "step": 5118 + }, + { + "epoch": 0.12, + "grad_norm": 2.3967890423847984, + "learning_rate": 1.9572691992114867e-05, + "loss": 1.0857, + "step": 5119 + }, + { + "epoch": 0.12, + "grad_norm": 2.1628980415059313, + "learning_rate": 1.9572471294038956e-05, + "loss": 1.1178, + "step": 5120 + }, + { + "epoch": 0.12, + "grad_norm": 4.074587177024697, + "learning_rate": 1.9572250540229088e-05, + "loss": 1.0566, + "step": 5121 + }, + { + "epoch": 0.12, + "grad_norm": 2.2243000112439244, + "learning_rate": 1.9572029730686546e-05, + "loss": 1.027, + "step": 5122 + }, + { + "epoch": 0.12, + "grad_norm": 2.3899505293336913, + "learning_rate": 1.9571808865412617e-05, + "loss": 0.9731, + "step": 5123 + }, + { + "epoch": 0.12, + "grad_norm": 2.360944960087408, + "learning_rate": 1.9571587944408582e-05, + "loss": 0.9662, + "step": 5124 + }, + { + "epoch": 0.12, + "grad_norm": 2.045262009338604, + "learning_rate": 1.957136696767574e-05, + "loss": 1.0601, + "step": 5125 + }, + { + "epoch": 0.12, + "grad_norm": 2.400916784684769, + "learning_rate": 1.9571145935215358e-05, + "loss": 1.1916, + "step": 5126 + }, + { + "epoch": 0.12, + "grad_norm": 1.9880649654175324, + "learning_rate": 1.9570924847028744e-05, + "loss": 1.1365, + "step": 5127 + }, + { + "epoch": 0.12, + "grad_norm": 1.905199204655351, + "learning_rate": 1.957070370311717e-05, + "loss": 1.0527, + "step": 5128 + }, + { + "epoch": 0.12, + "grad_norm": 2.006201783338745, + "learning_rate": 1.9570482503481933e-05, + "loss": 0.905, + "step": 5129 + }, + { + "epoch": 0.12, + "grad_norm": 2.267834277947158, + "learning_rate": 1.957026124812431e-05, + "loss": 1.0701, + "step": 5130 + }, + { + "epoch": 0.12, + "grad_norm": 2.298393086099759, + "learning_rate": 1.9570039937045604e-05, + "loss": 1.0835, + "step": 5131 + }, + { + "epoch": 0.12, + "grad_norm": 2.1700014035212067, + "learning_rate": 1.9569818570247093e-05, + "loss": 0.9563, + "step": 5132 + }, + { + "epoch": 0.12, + "grad_norm": 2.048051349229198, + "learning_rate": 1.9569597147730066e-05, + "loss": 1.1706, + "step": 5133 + }, + { + "epoch": 0.12, + "grad_norm": 2.367483689456995, + "learning_rate": 1.9569375669495815e-05, + "loss": 1.1798, + "step": 5134 + }, + { + "epoch": 0.12, + "grad_norm": 2.7518983148317484, + "learning_rate": 1.9569154135545634e-05, + "loss": 0.9647, + "step": 5135 + }, + { + "epoch": 0.12, + "grad_norm": 2.0506432394864547, + "learning_rate": 1.9568932545880804e-05, + "loss": 1.1523, + "step": 5136 + }, + { + "epoch": 0.12, + "grad_norm": 1.879103963755929, + "learning_rate": 1.9568710900502622e-05, + "loss": 1.042, + "step": 5137 + }, + { + "epoch": 0.12, + "grad_norm": 2.4001436789353345, + "learning_rate": 1.9568489199412373e-05, + "loss": 1.0576, + "step": 5138 + }, + { + "epoch": 0.12, + "grad_norm": 1.9213323569090255, + "learning_rate": 1.956826744261135e-05, + "loss": 1.0883, + "step": 5139 + }, + { + "epoch": 0.12, + "grad_norm": 1.285938575215197, + "learning_rate": 1.9568045630100846e-05, + "loss": 1.074, + "step": 5140 + }, + { + "epoch": 0.12, + "grad_norm": 6.466413466324627, + "learning_rate": 1.9567823761882154e-05, + "loss": 0.897, + "step": 5141 + }, + { + "epoch": 0.12, + "grad_norm": 1.159930986793984, + "learning_rate": 1.9567601837956557e-05, + "loss": 0.9907, + "step": 5142 + }, + { + "epoch": 0.12, + "grad_norm": 2.0490558132282435, + "learning_rate": 1.9567379858325358e-05, + "loss": 1.0171, + "step": 5143 + }, + { + "epoch": 0.12, + "grad_norm": 2.1519500390503437, + "learning_rate": 1.9567157822989843e-05, + "loss": 1.0556, + "step": 5144 + }, + { + "epoch": 0.12, + "grad_norm": 1.819456242058533, + "learning_rate": 1.9566935731951307e-05, + "loss": 1.0605, + "step": 5145 + }, + { + "epoch": 0.12, + "grad_norm": 2.2807737658237017, + "learning_rate": 1.956671358521104e-05, + "loss": 1.0732, + "step": 5146 + }, + { + "epoch": 0.12, + "grad_norm": 2.052571380166427, + "learning_rate": 1.9566491382770342e-05, + "loss": 1.2471, + "step": 5147 + }, + { + "epoch": 0.12, + "grad_norm": 2.1218530275512713, + "learning_rate": 1.9566269124630498e-05, + "loss": 1.0294, + "step": 5148 + }, + { + "epoch": 0.12, + "grad_norm": 2.043661155668926, + "learning_rate": 1.9566046810792806e-05, + "loss": 1.0355, + "step": 5149 + }, + { + "epoch": 0.12, + "grad_norm": 2.1590899645513684, + "learning_rate": 1.9565824441258562e-05, + "loss": 1.0861, + "step": 5150 + }, + { + "epoch": 0.12, + "grad_norm": 1.9319049850296655, + "learning_rate": 1.9565602016029062e-05, + "loss": 1.0986, + "step": 5151 + }, + { + "epoch": 0.12, + "grad_norm": 1.1637970634546235, + "learning_rate": 1.95653795351056e-05, + "loss": 0.9774, + "step": 5152 + }, + { + "epoch": 0.12, + "grad_norm": 2.0228327713731047, + "learning_rate": 1.9565156998489465e-05, + "loss": 1.1845, + "step": 5153 + }, + { + "epoch": 0.12, + "grad_norm": 1.9546270755863997, + "learning_rate": 1.956493440618196e-05, + "loss": 1.152, + "step": 5154 + }, + { + "epoch": 0.12, + "grad_norm": 2.27771625163222, + "learning_rate": 1.9564711758184378e-05, + "loss": 0.9761, + "step": 5155 + }, + { + "epoch": 0.12, + "grad_norm": 1.9881676522462806, + "learning_rate": 1.956448905449802e-05, + "loss": 1.1447, + "step": 5156 + }, + { + "epoch": 0.12, + "grad_norm": 1.0780525092446214, + "learning_rate": 1.9564266295124174e-05, + "loss": 1.0333, + "step": 5157 + }, + { + "epoch": 0.12, + "grad_norm": 1.9531470749322803, + "learning_rate": 1.956404348006414e-05, + "loss": 0.9962, + "step": 5158 + }, + { + "epoch": 0.12, + "grad_norm": 1.9622190942318969, + "learning_rate": 1.956382060931922e-05, + "loss": 1.1182, + "step": 5159 + }, + { + "epoch": 0.12, + "grad_norm": 1.1083117112919814, + "learning_rate": 1.956359768289071e-05, + "loss": 0.9918, + "step": 5160 + }, + { + "epoch": 0.12, + "grad_norm": 2.214350620672435, + "learning_rate": 1.9563374700779904e-05, + "loss": 1.1047, + "step": 5161 + }, + { + "epoch": 0.12, + "grad_norm": 3.240502139976582, + "learning_rate": 1.9563151662988106e-05, + "loss": 1.0421, + "step": 5162 + }, + { + "epoch": 0.12, + "grad_norm": 2.2373963963291903, + "learning_rate": 1.9562928569516607e-05, + "loss": 1.0047, + "step": 5163 + }, + { + "epoch": 0.12, + "grad_norm": 2.018648285903458, + "learning_rate": 1.9562705420366712e-05, + "loss": 1.1138, + "step": 5164 + }, + { + "epoch": 0.12, + "grad_norm": 2.073259630390413, + "learning_rate": 1.9562482215539722e-05, + "loss": 1.2189, + "step": 5165 + }, + { + "epoch": 0.12, + "grad_norm": 2.4894437513761565, + "learning_rate": 1.9562258955036928e-05, + "loss": 0.9998, + "step": 5166 + }, + { + "epoch": 0.12, + "grad_norm": 2.4101917443529035, + "learning_rate": 1.9562035638859638e-05, + "loss": 1.1779, + "step": 5167 + }, + { + "epoch": 0.12, + "grad_norm": 2.233991325930904, + "learning_rate": 1.9561812267009148e-05, + "loss": 0.9746, + "step": 5168 + }, + { + "epoch": 0.12, + "grad_norm": 2.3519818278241638, + "learning_rate": 1.956158883948676e-05, + "loss": 1.1904, + "step": 5169 + }, + { + "epoch": 0.12, + "grad_norm": 2.15166534893611, + "learning_rate": 1.9561365356293774e-05, + "loss": 0.9143, + "step": 5170 + }, + { + "epoch": 0.12, + "grad_norm": 1.906428625529915, + "learning_rate": 1.9561141817431494e-05, + "loss": 1.1422, + "step": 5171 + }, + { + "epoch": 0.12, + "grad_norm": 2.2268075222586194, + "learning_rate": 1.9560918222901216e-05, + "loss": 0.9928, + "step": 5172 + }, + { + "epoch": 0.12, + "grad_norm": 2.6492589350799367, + "learning_rate": 1.9560694572704245e-05, + "loss": 1.2152, + "step": 5173 + }, + { + "epoch": 0.12, + "grad_norm": 2.1163058815259665, + "learning_rate": 1.9560470866841888e-05, + "loss": 1.0928, + "step": 5174 + }, + { + "epoch": 0.12, + "grad_norm": 2.292741044990644, + "learning_rate": 1.9560247105315443e-05, + "loss": 1.1062, + "step": 5175 + }, + { + "epoch": 0.12, + "grad_norm": 1.925816984164281, + "learning_rate": 1.9560023288126208e-05, + "loss": 0.9457, + "step": 5176 + }, + { + "epoch": 0.12, + "grad_norm": 2.0812039772637516, + "learning_rate": 1.955979941527549e-05, + "loss": 1.0859, + "step": 5177 + }, + { + "epoch": 0.12, + "grad_norm": 2.112652949204429, + "learning_rate": 1.9559575486764596e-05, + "loss": 1.1169, + "step": 5178 + }, + { + "epoch": 0.12, + "grad_norm": 2.256813927881103, + "learning_rate": 1.9559351502594826e-05, + "loss": 1.0368, + "step": 5179 + }, + { + "epoch": 0.12, + "grad_norm": 1.9654477193202924, + "learning_rate": 1.9559127462767487e-05, + "loss": 0.8934, + "step": 5180 + }, + { + "epoch": 0.12, + "grad_norm": 1.1838034012943648, + "learning_rate": 1.9558903367283882e-05, + "loss": 1.0404, + "step": 5181 + }, + { + "epoch": 0.12, + "grad_norm": 1.9962949175534657, + "learning_rate": 1.9558679216145314e-05, + "loss": 1.1553, + "step": 5182 + }, + { + "epoch": 0.12, + "grad_norm": 2.0332445962406855, + "learning_rate": 1.9558455009353086e-05, + "loss": 1.0872, + "step": 5183 + }, + { + "epoch": 0.12, + "grad_norm": 2.2908415276996164, + "learning_rate": 1.955823074690851e-05, + "loss": 1.0707, + "step": 5184 + }, + { + "epoch": 0.12, + "grad_norm": 2.9138326465957993, + "learning_rate": 1.955800642881289e-05, + "loss": 1.0583, + "step": 5185 + }, + { + "epoch": 0.12, + "grad_norm": 1.14178735642278, + "learning_rate": 1.9557782055067528e-05, + "loss": 0.9533, + "step": 5186 + }, + { + "epoch": 0.12, + "grad_norm": 1.972834535959963, + "learning_rate": 1.9557557625673734e-05, + "loss": 1.0487, + "step": 5187 + }, + { + "epoch": 0.12, + "grad_norm": 2.6933841162168317, + "learning_rate": 1.9557333140632816e-05, + "loss": 1.0964, + "step": 5188 + }, + { + "epoch": 0.12, + "grad_norm": 1.2128775892693597, + "learning_rate": 1.9557108599946076e-05, + "loss": 1.0402, + "step": 5189 + }, + { + "epoch": 0.12, + "grad_norm": 2.162913167331194, + "learning_rate": 1.9556884003614824e-05, + "loss": 1.0774, + "step": 5190 + }, + { + "epoch": 0.12, + "grad_norm": 2.1525972599112486, + "learning_rate": 1.9556659351640368e-05, + "loss": 1.1817, + "step": 5191 + }, + { + "epoch": 0.12, + "grad_norm": 2.161050009100025, + "learning_rate": 1.9556434644024013e-05, + "loss": 0.9787, + "step": 5192 + }, + { + "epoch": 0.12, + "grad_norm": 1.8940219973873116, + "learning_rate": 1.9556209880767073e-05, + "loss": 1.0676, + "step": 5193 + }, + { + "epoch": 0.12, + "grad_norm": 3.4320695540959, + "learning_rate": 1.9555985061870854e-05, + "loss": 1.0816, + "step": 5194 + }, + { + "epoch": 0.12, + "grad_norm": 2.2176072424999087, + "learning_rate": 1.9555760187336665e-05, + "loss": 1.18, + "step": 5195 + }, + { + "epoch": 0.12, + "grad_norm": 2.17253638671254, + "learning_rate": 1.955553525716581e-05, + "loss": 0.9652, + "step": 5196 + }, + { + "epoch": 0.12, + "grad_norm": 2.0635866886347234, + "learning_rate": 1.9555310271359608e-05, + "loss": 1.0817, + "step": 5197 + }, + { + "epoch": 0.12, + "grad_norm": 2.317564880482577, + "learning_rate": 1.9555085229919366e-05, + "loss": 1.0948, + "step": 5198 + }, + { + "epoch": 0.12, + "grad_norm": 2.170814583487086, + "learning_rate": 1.955486013284639e-05, + "loss": 1.0915, + "step": 5199 + }, + { + "epoch": 0.12, + "grad_norm": 2.276036152934824, + "learning_rate": 1.9554634980141993e-05, + "loss": 0.9961, + "step": 5200 + }, + { + "epoch": 0.12, + "grad_norm": 2.199541975787207, + "learning_rate": 1.955440977180749e-05, + "loss": 1.1785, + "step": 5201 + }, + { + "epoch": 0.12, + "grad_norm": 2.1698833316440864, + "learning_rate": 1.9554184507844183e-05, + "loss": 0.9614, + "step": 5202 + }, + { + "epoch": 0.12, + "grad_norm": 2.1635553420071334, + "learning_rate": 1.9553959188253388e-05, + "loss": 1.1587, + "step": 5203 + }, + { + "epoch": 0.12, + "grad_norm": 1.1760841560450752, + "learning_rate": 1.9553733813036425e-05, + "loss": 0.9906, + "step": 5204 + }, + { + "epoch": 0.12, + "grad_norm": 1.0966304230218857, + "learning_rate": 1.955350838219459e-05, + "loss": 0.938, + "step": 5205 + }, + { + "epoch": 0.12, + "grad_norm": 2.0222143934024435, + "learning_rate": 1.9553282895729212e-05, + "loss": 0.9935, + "step": 5206 + }, + { + "epoch": 0.12, + "grad_norm": 1.0997465687111083, + "learning_rate": 1.9553057353641595e-05, + "loss": 0.9856, + "step": 5207 + }, + { + "epoch": 0.12, + "grad_norm": 2.196829175623789, + "learning_rate": 1.955283175593305e-05, + "loss": 1.0709, + "step": 5208 + }, + { + "epoch": 0.12, + "grad_norm": 2.279523030565869, + "learning_rate": 1.95526061026049e-05, + "loss": 1.0749, + "step": 5209 + }, + { + "epoch": 0.12, + "grad_norm": 2.0848705369862484, + "learning_rate": 1.955238039365845e-05, + "loss": 1.1006, + "step": 5210 + }, + { + "epoch": 0.12, + "grad_norm": 2.2414926540946745, + "learning_rate": 1.9552154629095015e-05, + "loss": 1.1599, + "step": 5211 + }, + { + "epoch": 0.12, + "grad_norm": 2.487588973661599, + "learning_rate": 1.9551928808915914e-05, + "loss": 1.0947, + "step": 5212 + }, + { + "epoch": 0.12, + "grad_norm": 2.323468541234268, + "learning_rate": 1.9551702933122462e-05, + "loss": 1.0393, + "step": 5213 + }, + { + "epoch": 0.12, + "grad_norm": 2.685347559815138, + "learning_rate": 1.9551477001715965e-05, + "loss": 0.9586, + "step": 5214 + }, + { + "epoch": 0.12, + "grad_norm": 2.0504866558743804, + "learning_rate": 1.955125101469775e-05, + "loss": 1.1302, + "step": 5215 + }, + { + "epoch": 0.12, + "grad_norm": 1.0785146299452235, + "learning_rate": 1.9551024972069127e-05, + "loss": 0.9404, + "step": 5216 + }, + { + "epoch": 0.12, + "grad_norm": 2.007194385862707, + "learning_rate": 1.955079887383141e-05, + "loss": 0.9822, + "step": 5217 + }, + { + "epoch": 0.12, + "grad_norm": 2.2048214928006304, + "learning_rate": 1.9550572719985922e-05, + "loss": 1.1862, + "step": 5218 + }, + { + "epoch": 0.12, + "grad_norm": 1.1335003210614845, + "learning_rate": 1.955034651053397e-05, + "loss": 0.9999, + "step": 5219 + }, + { + "epoch": 0.12, + "grad_norm": 2.0907498256039294, + "learning_rate": 1.9550120245476886e-05, + "loss": 1.1061, + "step": 5220 + }, + { + "epoch": 0.12, + "grad_norm": 1.760684606421985, + "learning_rate": 1.9549893924815974e-05, + "loss": 1.0376, + "step": 5221 + }, + { + "epoch": 0.12, + "grad_norm": 2.5444890487307017, + "learning_rate": 1.9549667548552557e-05, + "loss": 1.0452, + "step": 5222 + }, + { + "epoch": 0.12, + "grad_norm": 2.0531564182478776, + "learning_rate": 1.9549441116687945e-05, + "loss": 1.0893, + "step": 5223 + }, + { + "epoch": 0.12, + "grad_norm": 2.1393942654775104, + "learning_rate": 1.954921462922347e-05, + "loss": 1.0635, + "step": 5224 + }, + { + "epoch": 0.12, + "grad_norm": 1.9594725523985952, + "learning_rate": 1.9548988086160445e-05, + "loss": 1.0184, + "step": 5225 + }, + { + "epoch": 0.12, + "grad_norm": 1.2458628474927678, + "learning_rate": 1.954876148750019e-05, + "loss": 1.0279, + "step": 5226 + }, + { + "epoch": 0.12, + "grad_norm": 1.9489190584580875, + "learning_rate": 1.9548534833244017e-05, + "loss": 0.9059, + "step": 5227 + }, + { + "epoch": 0.12, + "grad_norm": 2.8779712687913217, + "learning_rate": 1.9548308123393255e-05, + "loss": 1.0404, + "step": 5228 + }, + { + "epoch": 0.12, + "grad_norm": 2.188177304487187, + "learning_rate": 1.9548081357949217e-05, + "loss": 1.1229, + "step": 5229 + }, + { + "epoch": 0.12, + "grad_norm": 2.4173578499169883, + "learning_rate": 1.954785453691323e-05, + "loss": 1.0419, + "step": 5230 + }, + { + "epoch": 0.12, + "grad_norm": 2.148786585797583, + "learning_rate": 1.9547627660286607e-05, + "loss": 1.1383, + "step": 5231 + }, + { + "epoch": 0.12, + "grad_norm": 1.9074798324016908, + "learning_rate": 1.9547400728070675e-05, + "loss": 0.9698, + "step": 5232 + }, + { + "epoch": 0.12, + "grad_norm": 2.429829063998507, + "learning_rate": 1.9547173740266753e-05, + "loss": 1.0491, + "step": 5233 + }, + { + "epoch": 0.12, + "grad_norm": 2.2629101133281115, + "learning_rate": 1.9546946696876164e-05, + "loss": 1.0949, + "step": 5234 + }, + { + "epoch": 0.12, + "grad_norm": 22.972129181353523, + "learning_rate": 1.954671959790023e-05, + "loss": 0.9967, + "step": 5235 + }, + { + "epoch": 0.12, + "grad_norm": 2.137144605196825, + "learning_rate": 1.9546492443340268e-05, + "loss": 1.1557, + "step": 5236 + }, + { + "epoch": 0.12, + "grad_norm": 1.898985893782141, + "learning_rate": 1.9546265233197607e-05, + "loss": 1.1677, + "step": 5237 + }, + { + "epoch": 0.12, + "grad_norm": 2.2056709652823674, + "learning_rate": 1.954603796747357e-05, + "loss": 1.1363, + "step": 5238 + }, + { + "epoch": 0.12, + "grad_norm": 2.1734922619580677, + "learning_rate": 1.9545810646169474e-05, + "loss": 0.9963, + "step": 5239 + }, + { + "epoch": 0.12, + "grad_norm": 2.3860281069497673, + "learning_rate": 1.9545583269286646e-05, + "loss": 1.0791, + "step": 5240 + }, + { + "epoch": 0.12, + "grad_norm": 2.106913350672186, + "learning_rate": 1.9545355836826415e-05, + "loss": 1.1174, + "step": 5241 + }, + { + "epoch": 0.12, + "grad_norm": 2.262135638142345, + "learning_rate": 1.9545128348790096e-05, + "loss": 1.1514, + "step": 5242 + }, + { + "epoch": 0.12, + "grad_norm": 2.3855816428552536, + "learning_rate": 1.9544900805179016e-05, + "loss": 1.0928, + "step": 5243 + }, + { + "epoch": 0.12, + "grad_norm": 2.055869684559542, + "learning_rate": 1.9544673205994505e-05, + "loss": 1.149, + "step": 5244 + }, + { + "epoch": 0.12, + "grad_norm": 2.1874831160677144, + "learning_rate": 1.9544445551237883e-05, + "loss": 1.0924, + "step": 5245 + }, + { + "epoch": 0.12, + "grad_norm": 2.195097392631319, + "learning_rate": 1.9544217840910483e-05, + "loss": 1.1639, + "step": 5246 + }, + { + "epoch": 0.12, + "grad_norm": 2.138823241272734, + "learning_rate": 1.9543990075013618e-05, + "loss": 1.033, + "step": 5247 + }, + { + "epoch": 0.12, + "grad_norm": 2.2228238376966214, + "learning_rate": 1.9543762253548623e-05, + "loss": 0.9741, + "step": 5248 + }, + { + "epoch": 0.12, + "grad_norm": 5.170315189626023, + "learning_rate": 1.9543534376516824e-05, + "loss": 1.0541, + "step": 5249 + }, + { + "epoch": 0.12, + "grad_norm": 2.2857778034012117, + "learning_rate": 1.954330644391955e-05, + "loss": 1.1606, + "step": 5250 + }, + { + "epoch": 0.12, + "grad_norm": 4.942224876158154, + "learning_rate": 1.9543078455758118e-05, + "loss": 1.0894, + "step": 5251 + }, + { + "epoch": 0.12, + "grad_norm": 2.1492705227898665, + "learning_rate": 1.9542850412033863e-05, + "loss": 1.079, + "step": 5252 + }, + { + "epoch": 0.12, + "grad_norm": 1.9984639885806343, + "learning_rate": 1.9542622312748113e-05, + "loss": 1.1465, + "step": 5253 + }, + { + "epoch": 0.12, + "grad_norm": 2.133332532540318, + "learning_rate": 1.9542394157902196e-05, + "loss": 1.0627, + "step": 5254 + }, + { + "epoch": 0.12, + "grad_norm": 2.2763196567104647, + "learning_rate": 1.9542165947497436e-05, + "loss": 0.9856, + "step": 5255 + }, + { + "epoch": 0.12, + "grad_norm": 2.193509746048168, + "learning_rate": 1.954193768153517e-05, + "loss": 1.1035, + "step": 5256 + }, + { + "epoch": 0.12, + "grad_norm": 2.107176513746657, + "learning_rate": 1.9541709360016715e-05, + "loss": 1.0866, + "step": 5257 + }, + { + "epoch": 0.12, + "grad_norm": 2.238285351898436, + "learning_rate": 1.9541480982943413e-05, + "loss": 1.1158, + "step": 5258 + }, + { + "epoch": 0.12, + "grad_norm": 2.24354764094242, + "learning_rate": 1.9541252550316585e-05, + "loss": 1.0318, + "step": 5259 + }, + { + "epoch": 0.12, + "grad_norm": 2.7536648094075398, + "learning_rate": 1.9541024062137565e-05, + "loss": 1.1159, + "step": 5260 + }, + { + "epoch": 0.12, + "grad_norm": 2.1690884466273985, + "learning_rate": 1.954079551840768e-05, + "loss": 1.1123, + "step": 5261 + }, + { + "epoch": 0.12, + "grad_norm": 2.038384105988896, + "learning_rate": 1.954056691912827e-05, + "loss": 1.0667, + "step": 5262 + }, + { + "epoch": 0.12, + "grad_norm": 1.189981607228439, + "learning_rate": 1.954033826430065e-05, + "loss": 1.0359, + "step": 5263 + }, + { + "epoch": 0.12, + "grad_norm": 1.1644272106915612, + "learning_rate": 1.9540109553926165e-05, + "loss": 1.0596, + "step": 5264 + }, + { + "epoch": 0.12, + "grad_norm": 2.101916072153253, + "learning_rate": 1.953988078800614e-05, + "loss": 1.1372, + "step": 5265 + }, + { + "epoch": 0.12, + "grad_norm": 2.8843339864756032, + "learning_rate": 1.953965196654191e-05, + "loss": 1.0295, + "step": 5266 + }, + { + "epoch": 0.12, + "grad_norm": 2.463642636293927, + "learning_rate": 1.9539423089534805e-05, + "loss": 1.0604, + "step": 5267 + }, + { + "epoch": 0.12, + "grad_norm": 2.5743001991369, + "learning_rate": 1.953919415698616e-05, + "loss": 1.1179, + "step": 5268 + }, + { + "epoch": 0.12, + "grad_norm": 3.131616311272146, + "learning_rate": 1.9538965168897307e-05, + "loss": 1.026, + "step": 5269 + }, + { + "epoch": 0.12, + "grad_norm": 2.766389240125337, + "learning_rate": 1.9538736125269578e-05, + "loss": 1.0508, + "step": 5270 + }, + { + "epoch": 0.12, + "grad_norm": 2.084793543740354, + "learning_rate": 1.9538507026104304e-05, + "loss": 1.0811, + "step": 5271 + }, + { + "epoch": 0.12, + "grad_norm": 1.1473274258796906, + "learning_rate": 1.953827787140283e-05, + "loss": 0.9315, + "step": 5272 + }, + { + "epoch": 0.12, + "grad_norm": 2.5090864271825404, + "learning_rate": 1.9538048661166474e-05, + "loss": 1.0092, + "step": 5273 + }, + { + "epoch": 0.12, + "grad_norm": 3.3045010370379186, + "learning_rate": 1.9537819395396586e-05, + "loss": 1.0683, + "step": 5274 + }, + { + "epoch": 0.12, + "grad_norm": 1.079659295835071, + "learning_rate": 1.9537590074094492e-05, + "loss": 0.9601, + "step": 5275 + }, + { + "epoch": 0.12, + "grad_norm": 2.035375536945006, + "learning_rate": 1.9537360697261524e-05, + "loss": 1.1165, + "step": 5276 + }, + { + "epoch": 0.12, + "grad_norm": 1.9672813926768713, + "learning_rate": 1.953713126489903e-05, + "loss": 1.0405, + "step": 5277 + }, + { + "epoch": 0.12, + "grad_norm": 2.334643162841981, + "learning_rate": 1.9536901777008332e-05, + "loss": 1.1448, + "step": 5278 + }, + { + "epoch": 0.12, + "grad_norm": 2.5387473088188544, + "learning_rate": 1.9536672233590774e-05, + "loss": 1.2401, + "step": 5279 + }, + { + "epoch": 0.12, + "grad_norm": 2.114994875812424, + "learning_rate": 1.9536442634647696e-05, + "loss": 1.0479, + "step": 5280 + }, + { + "epoch": 0.12, + "grad_norm": 2.2437517279031205, + "learning_rate": 1.9536212980180422e-05, + "loss": 1.0014, + "step": 5281 + }, + { + "epoch": 0.12, + "grad_norm": 2.069290379359825, + "learning_rate": 1.95359832701903e-05, + "loss": 1.2716, + "step": 5282 + }, + { + "epoch": 0.12, + "grad_norm": 2.2141870019771983, + "learning_rate": 1.9535753504678667e-05, + "loss": 1.0958, + "step": 5283 + }, + { + "epoch": 0.12, + "grad_norm": 2.34958106683467, + "learning_rate": 1.9535523683646857e-05, + "loss": 1.084, + "step": 5284 + }, + { + "epoch": 0.12, + "grad_norm": 2.129847840547184, + "learning_rate": 1.9535293807096208e-05, + "loss": 0.9408, + "step": 5285 + }, + { + "epoch": 0.12, + "grad_norm": 2.4847835176408686, + "learning_rate": 1.953506387502806e-05, + "loss": 1.0099, + "step": 5286 + }, + { + "epoch": 0.12, + "grad_norm": 1.9972468987112446, + "learning_rate": 1.9534833887443752e-05, + "loss": 0.9596, + "step": 5287 + }, + { + "epoch": 0.12, + "grad_norm": 2.446933136458087, + "learning_rate": 1.9534603844344623e-05, + "loss": 1.1441, + "step": 5288 + }, + { + "epoch": 0.12, + "grad_norm": 2.9579588227829356, + "learning_rate": 1.9534373745732008e-05, + "loss": 1.0625, + "step": 5289 + }, + { + "epoch": 0.12, + "grad_norm": 2.166682131664607, + "learning_rate": 1.9534143591607256e-05, + "loss": 1.1874, + "step": 5290 + }, + { + "epoch": 0.12, + "grad_norm": 1.2121063459507881, + "learning_rate": 1.9533913381971698e-05, + "loss": 0.9596, + "step": 5291 + }, + { + "epoch": 0.12, + "grad_norm": 2.0740184654686713, + "learning_rate": 1.953368311682668e-05, + "loss": 0.9832, + "step": 5292 + }, + { + "epoch": 0.12, + "grad_norm": 1.1644452347099412, + "learning_rate": 1.953345279617354e-05, + "loss": 0.9988, + "step": 5293 + }, + { + "epoch": 0.12, + "grad_norm": 2.1589042299374652, + "learning_rate": 1.9533222420013618e-05, + "loss": 1.0241, + "step": 5294 + }, + { + "epoch": 0.12, + "grad_norm": 2.289701217261712, + "learning_rate": 1.9532991988348254e-05, + "loss": 1.0858, + "step": 5295 + }, + { + "epoch": 0.12, + "grad_norm": 2.2822998739643894, + "learning_rate": 1.9532761501178797e-05, + "loss": 1.0689, + "step": 5296 + }, + { + "epoch": 0.12, + "grad_norm": 2.7999710441732866, + "learning_rate": 1.9532530958506585e-05, + "loss": 1.1671, + "step": 5297 + }, + { + "epoch": 0.12, + "grad_norm": 2.235663517224108, + "learning_rate": 1.953230036033296e-05, + "loss": 1.0364, + "step": 5298 + }, + { + "epoch": 0.12, + "grad_norm": 1.977945968067091, + "learning_rate": 1.9532069706659263e-05, + "loss": 1.0617, + "step": 5299 + }, + { + "epoch": 0.12, + "grad_norm": 2.025610101953586, + "learning_rate": 1.9531838997486838e-05, + "loss": 1.0731, + "step": 5300 + }, + { + "epoch": 0.12, + "grad_norm": 2.1928911485309186, + "learning_rate": 1.953160823281703e-05, + "loss": 1.2579, + "step": 5301 + }, + { + "epoch": 0.12, + "grad_norm": 2.0130246549795614, + "learning_rate": 1.953137741265118e-05, + "loss": 1.1155, + "step": 5302 + }, + { + "epoch": 0.12, + "grad_norm": 1.9017157737149806, + "learning_rate": 1.9531146536990634e-05, + "loss": 1.0019, + "step": 5303 + }, + { + "epoch": 0.12, + "grad_norm": 2.0476120983646653, + "learning_rate": 1.9530915605836735e-05, + "loss": 1.0238, + "step": 5304 + }, + { + "epoch": 0.12, + "grad_norm": 2.1822772102637717, + "learning_rate": 1.9530684619190827e-05, + "loss": 0.9897, + "step": 5305 + }, + { + "epoch": 0.13, + "grad_norm": 2.2890005856007694, + "learning_rate": 1.9530453577054257e-05, + "loss": 1.1468, + "step": 5306 + }, + { + "epoch": 0.13, + "grad_norm": 2.319320113077891, + "learning_rate": 1.9530222479428368e-05, + "loss": 0.9888, + "step": 5307 + }, + { + "epoch": 0.13, + "grad_norm": 2.3746360207564203, + "learning_rate": 1.9529991326314506e-05, + "loss": 1.0883, + "step": 5308 + }, + { + "epoch": 0.13, + "grad_norm": 2.5103649477689753, + "learning_rate": 1.952976011771402e-05, + "loss": 0.9857, + "step": 5309 + }, + { + "epoch": 0.13, + "grad_norm": 2.270160427590819, + "learning_rate": 1.9529528853628254e-05, + "loss": 1.104, + "step": 5310 + }, + { + "epoch": 0.13, + "grad_norm": 2.6273560829146794, + "learning_rate": 1.952929753405855e-05, + "loss": 1.1148, + "step": 5311 + }, + { + "epoch": 0.13, + "grad_norm": 3.4988577051808596, + "learning_rate": 1.952906615900626e-05, + "loss": 0.8636, + "step": 5312 + }, + { + "epoch": 0.13, + "grad_norm": 1.2208530189037219, + "learning_rate": 1.952883472847273e-05, + "loss": 0.938, + "step": 5313 + }, + { + "epoch": 0.13, + "grad_norm": 2.2789330298315296, + "learning_rate": 1.9528603242459308e-05, + "loss": 1.071, + "step": 5314 + }, + { + "epoch": 0.13, + "grad_norm": 2.614364397459332, + "learning_rate": 1.952837170096734e-05, + "loss": 1.0871, + "step": 5315 + }, + { + "epoch": 0.13, + "grad_norm": 2.720114243201002, + "learning_rate": 1.952814010399818e-05, + "loss": 1.0486, + "step": 5316 + }, + { + "epoch": 0.13, + "grad_norm": 2.365626887399615, + "learning_rate": 1.9527908451553168e-05, + "loss": 1.0218, + "step": 5317 + }, + { + "epoch": 0.13, + "grad_norm": 2.215565533678141, + "learning_rate": 1.9527676743633654e-05, + "loss": 1.1098, + "step": 5318 + }, + { + "epoch": 0.13, + "grad_norm": 2.1065742300028063, + "learning_rate": 1.9527444980240993e-05, + "loss": 1.0925, + "step": 5319 + }, + { + "epoch": 0.13, + "grad_norm": 2.100420555838312, + "learning_rate": 1.952721316137653e-05, + "loss": 1.0686, + "step": 5320 + }, + { + "epoch": 0.13, + "grad_norm": 2.286209299689284, + "learning_rate": 1.952698128704162e-05, + "loss": 1.2193, + "step": 5321 + }, + { + "epoch": 0.13, + "grad_norm": 2.7388633636122024, + "learning_rate": 1.9526749357237604e-05, + "loss": 0.9721, + "step": 5322 + }, + { + "epoch": 0.13, + "grad_norm": 2.4001851914440997, + "learning_rate": 1.952651737196584e-05, + "loss": 1.0774, + "step": 5323 + }, + { + "epoch": 0.13, + "grad_norm": 2.6136679757451455, + "learning_rate": 1.9526285331227672e-05, + "loss": 1.0483, + "step": 5324 + }, + { + "epoch": 0.13, + "grad_norm": 1.9005306086730291, + "learning_rate": 1.9526053235024457e-05, + "loss": 1.1085, + "step": 5325 + }, + { + "epoch": 0.13, + "grad_norm": 2.5341293885819858, + "learning_rate": 1.9525821083357545e-05, + "loss": 1.104, + "step": 5326 + }, + { + "epoch": 0.13, + "grad_norm": 2.5053045585167313, + "learning_rate": 1.9525588876228284e-05, + "loss": 1.1197, + "step": 5327 + }, + { + "epoch": 0.13, + "grad_norm": 2.040565261365291, + "learning_rate": 1.952535661363803e-05, + "loss": 1.0198, + "step": 5328 + }, + { + "epoch": 0.13, + "grad_norm": 2.0966991857626813, + "learning_rate": 1.9525124295588135e-05, + "loss": 1.1675, + "step": 5329 + }, + { + "epoch": 0.13, + "grad_norm": 2.062899605803362, + "learning_rate": 1.952489192207995e-05, + "loss": 1.1036, + "step": 5330 + }, + { + "epoch": 0.13, + "grad_norm": 1.9402322734737285, + "learning_rate": 1.952465949311483e-05, + "loss": 1.0493, + "step": 5331 + }, + { + "epoch": 0.13, + "grad_norm": 2.126468201644471, + "learning_rate": 1.9524427008694125e-05, + "loss": 0.8193, + "step": 5332 + }, + { + "epoch": 0.13, + "grad_norm": 2.0382480865283052, + "learning_rate": 1.952419446881919e-05, + "loss": 1.0957, + "step": 5333 + }, + { + "epoch": 0.13, + "grad_norm": 2.071832609934184, + "learning_rate": 1.9523961873491385e-05, + "loss": 1.1717, + "step": 5334 + }, + { + "epoch": 0.13, + "grad_norm": 1.8082452253113972, + "learning_rate": 1.9523729222712053e-05, + "loss": 0.984, + "step": 5335 + }, + { + "epoch": 0.13, + "grad_norm": 1.849431143111857, + "learning_rate": 1.9523496516482557e-05, + "loss": 1.1006, + "step": 5336 + }, + { + "epoch": 0.13, + "grad_norm": 2.6867485643656033, + "learning_rate": 1.9523263754804248e-05, + "loss": 1.0203, + "step": 5337 + }, + { + "epoch": 0.13, + "grad_norm": 2.2062940925681556, + "learning_rate": 1.9523030937678484e-05, + "loss": 1.0328, + "step": 5338 + }, + { + "epoch": 0.13, + "grad_norm": 2.293971545904544, + "learning_rate": 1.9522798065106615e-05, + "loss": 1.0769, + "step": 5339 + }, + { + "epoch": 0.13, + "grad_norm": 2.162618278147567, + "learning_rate": 1.9522565137090005e-05, + "loss": 1.1913, + "step": 5340 + }, + { + "epoch": 0.13, + "grad_norm": 2.2737798128916547, + "learning_rate": 1.9522332153630006e-05, + "loss": 1.098, + "step": 5341 + }, + { + "epoch": 0.13, + "grad_norm": 2.3109648821994946, + "learning_rate": 1.952209911472797e-05, + "loss": 1.0586, + "step": 5342 + }, + { + "epoch": 0.13, + "grad_norm": 2.141193684924511, + "learning_rate": 1.952186602038526e-05, + "loss": 1.044, + "step": 5343 + }, + { + "epoch": 0.13, + "grad_norm": 1.1012649597180655, + "learning_rate": 1.9521632870603235e-05, + "loss": 0.9399, + "step": 5344 + }, + { + "epoch": 0.13, + "grad_norm": 2.2034990366426466, + "learning_rate": 1.952139966538325e-05, + "loss": 1.0931, + "step": 5345 + }, + { + "epoch": 0.13, + "grad_norm": 2.23943465210841, + "learning_rate": 1.9521166404726656e-05, + "loss": 0.8957, + "step": 5346 + }, + { + "epoch": 0.13, + "grad_norm": 1.1839902932721729, + "learning_rate": 1.952093308863482e-05, + "loss": 1.0259, + "step": 5347 + }, + { + "epoch": 0.13, + "grad_norm": 2.1716514136128096, + "learning_rate": 1.9520699717109096e-05, + "loss": 1.093, + "step": 5348 + }, + { + "epoch": 0.13, + "grad_norm": 2.392036172444193, + "learning_rate": 1.9520466290150842e-05, + "loss": 1.0867, + "step": 5349 + }, + { + "epoch": 0.13, + "grad_norm": 1.9384514792618726, + "learning_rate": 1.9520232807761424e-05, + "loss": 1.0987, + "step": 5350 + }, + { + "epoch": 0.13, + "grad_norm": 1.9756112407029376, + "learning_rate": 1.9519999269942194e-05, + "loss": 1.0556, + "step": 5351 + }, + { + "epoch": 0.13, + "grad_norm": 2.0310179987105674, + "learning_rate": 1.9519765676694513e-05, + "loss": 1.1002, + "step": 5352 + }, + { + "epoch": 0.13, + "grad_norm": 1.9937417401291042, + "learning_rate": 1.9519532028019744e-05, + "loss": 1.2524, + "step": 5353 + }, + { + "epoch": 0.13, + "grad_norm": 2.0485297168485443, + "learning_rate": 1.9519298323919245e-05, + "loss": 0.9246, + "step": 5354 + }, + { + "epoch": 0.13, + "grad_norm": 1.1126746231493858, + "learning_rate": 1.9519064564394383e-05, + "loss": 0.9776, + "step": 5355 + }, + { + "epoch": 0.13, + "grad_norm": 1.9863910989756286, + "learning_rate": 1.9518830749446507e-05, + "loss": 1.0158, + "step": 5356 + }, + { + "epoch": 0.13, + "grad_norm": 2.070348282664493, + "learning_rate": 1.9518596879076982e-05, + "loss": 1.1607, + "step": 5357 + }, + { + "epoch": 0.13, + "grad_norm": 2.2495865253590304, + "learning_rate": 1.9518362953287178e-05, + "loss": 0.9987, + "step": 5358 + }, + { + "epoch": 0.13, + "grad_norm": 2.221657498427032, + "learning_rate": 1.9518128972078448e-05, + "loss": 1.0674, + "step": 5359 + }, + { + "epoch": 0.13, + "grad_norm": 2.5304692800635538, + "learning_rate": 1.951789493545216e-05, + "loss": 0.9328, + "step": 5360 + }, + { + "epoch": 0.13, + "grad_norm": 2.3422873421750854, + "learning_rate": 1.9517660843409678e-05, + "loss": 1.1317, + "step": 5361 + }, + { + "epoch": 0.13, + "grad_norm": 2.06101338599392, + "learning_rate": 1.9517426695952358e-05, + "loss": 1.1217, + "step": 5362 + }, + { + "epoch": 0.13, + "grad_norm": 1.1229498410805812, + "learning_rate": 1.9517192493081563e-05, + "loss": 0.9898, + "step": 5363 + }, + { + "epoch": 0.13, + "grad_norm": 2.1558045271773683, + "learning_rate": 1.9516958234798663e-05, + "loss": 1.1665, + "step": 5364 + }, + { + "epoch": 0.13, + "grad_norm": 1.9855562776286255, + "learning_rate": 1.951672392110502e-05, + "loss": 1.1426, + "step": 5365 + }, + { + "epoch": 0.13, + "grad_norm": 2.0377673761286537, + "learning_rate": 1.9516489552002e-05, + "loss": 1.094, + "step": 5366 + }, + { + "epoch": 0.13, + "grad_norm": 2.1849385096409195, + "learning_rate": 1.951625512749096e-05, + "loss": 1.0921, + "step": 5367 + }, + { + "epoch": 0.13, + "grad_norm": 1.977056016794943, + "learning_rate": 1.9516020647573272e-05, + "loss": 1.1843, + "step": 5368 + }, + { + "epoch": 0.13, + "grad_norm": 1.943617394345904, + "learning_rate": 1.95157861122503e-05, + "loss": 0.908, + "step": 5369 + }, + { + "epoch": 0.13, + "grad_norm": 2.203464055645551, + "learning_rate": 1.9515551521523406e-05, + "loss": 1.0019, + "step": 5370 + }, + { + "epoch": 0.13, + "grad_norm": 1.7813209051760968, + "learning_rate": 1.9515316875393958e-05, + "loss": 1.0804, + "step": 5371 + }, + { + "epoch": 0.13, + "grad_norm": 2.1171530106240897, + "learning_rate": 1.9515082173863327e-05, + "loss": 1.206, + "step": 5372 + }, + { + "epoch": 0.13, + "grad_norm": 1.080673667955343, + "learning_rate": 1.9514847416932873e-05, + "loss": 0.9756, + "step": 5373 + }, + { + "epoch": 0.13, + "grad_norm": 2.502711547040719, + "learning_rate": 1.9514612604603963e-05, + "loss": 1.1249, + "step": 5374 + }, + { + "epoch": 0.13, + "grad_norm": 2.1359148022257375, + "learning_rate": 1.9514377736877967e-05, + "loss": 1.049, + "step": 5375 + }, + { + "epoch": 0.13, + "grad_norm": 2.262779521260114, + "learning_rate": 1.9514142813756253e-05, + "loss": 1.0963, + "step": 5376 + }, + { + "epoch": 0.13, + "grad_norm": 2.4032249057794437, + "learning_rate": 1.9513907835240187e-05, + "loss": 1.0059, + "step": 5377 + }, + { + "epoch": 0.13, + "grad_norm": 2.2494177108291886, + "learning_rate": 1.9513672801331135e-05, + "loss": 0.9187, + "step": 5378 + }, + { + "epoch": 0.13, + "grad_norm": 2.6946492644636475, + "learning_rate": 1.951343771203047e-05, + "loss": 0.8916, + "step": 5379 + }, + { + "epoch": 0.13, + "grad_norm": 2.302560727938671, + "learning_rate": 1.951320256733956e-05, + "loss": 1.0463, + "step": 5380 + }, + { + "epoch": 0.13, + "grad_norm": 3.120331460164975, + "learning_rate": 1.951296736725977e-05, + "loss": 0.9564, + "step": 5381 + }, + { + "epoch": 0.13, + "grad_norm": 2.1491113882469257, + "learning_rate": 1.9512732111792477e-05, + "loss": 1.1055, + "step": 5382 + }, + { + "epoch": 0.13, + "grad_norm": 2.3403860173882163, + "learning_rate": 1.9512496800939043e-05, + "loss": 1.072, + "step": 5383 + }, + { + "epoch": 0.13, + "grad_norm": 1.217127864421004, + "learning_rate": 1.9512261434700842e-05, + "loss": 0.9915, + "step": 5384 + }, + { + "epoch": 0.13, + "grad_norm": 1.9791533821715226, + "learning_rate": 1.951202601307924e-05, + "loss": 1.1441, + "step": 5385 + }, + { + "epoch": 0.13, + "grad_norm": 1.266807741576718, + "learning_rate": 1.9511790536075613e-05, + "loss": 1.1214, + "step": 5386 + }, + { + "epoch": 0.13, + "grad_norm": 2.001080143790232, + "learning_rate": 1.951155500369133e-05, + "loss": 1.0855, + "step": 5387 + }, + { + "epoch": 0.13, + "grad_norm": 2.2649494461917214, + "learning_rate": 1.9511319415927766e-05, + "loss": 1.0241, + "step": 5388 + }, + { + "epoch": 0.13, + "grad_norm": 2.1386205186948466, + "learning_rate": 1.9511083772786285e-05, + "loss": 1.0554, + "step": 5389 + }, + { + "epoch": 0.13, + "grad_norm": 2.1668559644446543, + "learning_rate": 1.9510848074268263e-05, + "loss": 0.8996, + "step": 5390 + }, + { + "epoch": 0.13, + "grad_norm": 2.1137065233685846, + "learning_rate": 1.9510612320375073e-05, + "loss": 1.0345, + "step": 5391 + }, + { + "epoch": 0.13, + "grad_norm": 6.738489668566833, + "learning_rate": 1.951037651110809e-05, + "loss": 1.0255, + "step": 5392 + }, + { + "epoch": 0.13, + "grad_norm": 2.5540532336760653, + "learning_rate": 1.9510140646468682e-05, + "loss": 1.0628, + "step": 5393 + }, + { + "epoch": 0.13, + "grad_norm": 2.2185383445147195, + "learning_rate": 1.9509904726458223e-05, + "loss": 1.1366, + "step": 5394 + }, + { + "epoch": 0.13, + "grad_norm": 3.0087801023279903, + "learning_rate": 1.9509668751078087e-05, + "loss": 0.929, + "step": 5395 + }, + { + "epoch": 0.13, + "grad_norm": 1.9908097726212999, + "learning_rate": 1.9509432720329653e-05, + "loss": 1.1152, + "step": 5396 + }, + { + "epoch": 0.13, + "grad_norm": 2.0410199789863213, + "learning_rate": 1.9509196634214288e-05, + "loss": 1.0839, + "step": 5397 + }, + { + "epoch": 0.13, + "grad_norm": 2.742874988379705, + "learning_rate": 1.950896049273337e-05, + "loss": 0.9925, + "step": 5398 + }, + { + "epoch": 0.13, + "grad_norm": 1.8662675808899387, + "learning_rate": 1.9508724295888277e-05, + "loss": 1.0182, + "step": 5399 + }, + { + "epoch": 0.13, + "grad_norm": 2.0885418898991794, + "learning_rate": 1.9508488043680376e-05, + "loss": 1.0597, + "step": 5400 + }, + { + "epoch": 0.13, + "grad_norm": 2.2558107107660277, + "learning_rate": 1.950825173611105e-05, + "loss": 1.1333, + "step": 5401 + }, + { + "epoch": 0.13, + "grad_norm": 1.8940422351778698, + "learning_rate": 1.950801537318167e-05, + "loss": 0.8891, + "step": 5402 + }, + { + "epoch": 0.13, + "grad_norm": 2.558464484552982, + "learning_rate": 1.9507778954893614e-05, + "loss": 0.9359, + "step": 5403 + }, + { + "epoch": 0.13, + "grad_norm": 2.117762441359436, + "learning_rate": 1.950754248124826e-05, + "loss": 1.113, + "step": 5404 + }, + { + "epoch": 0.13, + "grad_norm": 2.126745492457385, + "learning_rate": 1.9507305952246988e-05, + "loss": 1.1807, + "step": 5405 + }, + { + "epoch": 0.13, + "grad_norm": 1.9827470025398433, + "learning_rate": 1.9507069367891166e-05, + "loss": 1.0393, + "step": 5406 + }, + { + "epoch": 0.13, + "grad_norm": 2.115651709108536, + "learning_rate": 1.950683272818218e-05, + "loss": 1.085, + "step": 5407 + }, + { + "epoch": 0.13, + "grad_norm": 1.9069859284044053, + "learning_rate": 1.9506596033121397e-05, + "loss": 1.0483, + "step": 5408 + }, + { + "epoch": 0.13, + "grad_norm": 2.6043217657413735, + "learning_rate": 1.9506359282710207e-05, + "loss": 1.122, + "step": 5409 + }, + { + "epoch": 0.13, + "grad_norm": 2.067585907899286, + "learning_rate": 1.950612247694998e-05, + "loss": 1.1136, + "step": 5410 + }, + { + "epoch": 0.13, + "grad_norm": 2.351012915607266, + "learning_rate": 1.9505885615842102e-05, + "loss": 1.1889, + "step": 5411 + }, + { + "epoch": 0.13, + "grad_norm": 2.228301915541613, + "learning_rate": 1.950564869938795e-05, + "loss": 1.0797, + "step": 5412 + }, + { + "epoch": 0.13, + "grad_norm": 2.023036580325187, + "learning_rate": 1.95054117275889e-05, + "loss": 1.0074, + "step": 5413 + }, + { + "epoch": 0.13, + "grad_norm": 1.845950843003864, + "learning_rate": 1.9505174700446333e-05, + "loss": 1.1368, + "step": 5414 + }, + { + "epoch": 0.13, + "grad_norm": 2.266529264944706, + "learning_rate": 1.950493761796163e-05, + "loss": 1.092, + "step": 5415 + }, + { + "epoch": 0.13, + "grad_norm": 2.078150684737542, + "learning_rate": 1.950470048013617e-05, + "loss": 1.1455, + "step": 5416 + }, + { + "epoch": 0.13, + "grad_norm": 1.2314607133741788, + "learning_rate": 1.9504463286971332e-05, + "loss": 1.0287, + "step": 5417 + }, + { + "epoch": 0.13, + "grad_norm": 2.316007470249908, + "learning_rate": 1.9504226038468506e-05, + "loss": 0.9549, + "step": 5418 + }, + { + "epoch": 0.13, + "grad_norm": 2.0043834500116846, + "learning_rate": 1.950398873462906e-05, + "loss": 1.2211, + "step": 5419 + }, + { + "epoch": 0.13, + "grad_norm": 1.1489661777461424, + "learning_rate": 1.9503751375454388e-05, + "loss": 0.9954, + "step": 5420 + }, + { + "epoch": 0.13, + "grad_norm": 2.4404596716720617, + "learning_rate": 1.9503513960945864e-05, + "loss": 1.0688, + "step": 5421 + }, + { + "epoch": 0.13, + "grad_norm": 1.0971445266585886, + "learning_rate": 1.9503276491104876e-05, + "loss": 0.9361, + "step": 5422 + }, + { + "epoch": 0.13, + "grad_norm": 2.3670362949078356, + "learning_rate": 1.9503038965932802e-05, + "loss": 0.9577, + "step": 5423 + }, + { + "epoch": 0.13, + "grad_norm": 2.5666829013425527, + "learning_rate": 1.9502801385431028e-05, + "loss": 1.1138, + "step": 5424 + }, + { + "epoch": 0.13, + "grad_norm": 2.3617235009290436, + "learning_rate": 1.9502563749600933e-05, + "loss": 1.1437, + "step": 5425 + }, + { + "epoch": 0.13, + "grad_norm": 2.1709014803109614, + "learning_rate": 1.9502326058443904e-05, + "loss": 1.0858, + "step": 5426 + }, + { + "epoch": 0.13, + "grad_norm": 2.561318010705401, + "learning_rate": 1.9502088311961324e-05, + "loss": 1.1272, + "step": 5427 + }, + { + "epoch": 0.13, + "grad_norm": 1.1400431090161056, + "learning_rate": 1.9501850510154578e-05, + "loss": 1.03, + "step": 5428 + }, + { + "epoch": 0.13, + "grad_norm": 2.299579212247833, + "learning_rate": 1.950161265302505e-05, + "loss": 1.0855, + "step": 5429 + }, + { + "epoch": 0.13, + "grad_norm": 2.3621004172220417, + "learning_rate": 1.9501374740574128e-05, + "loss": 0.9841, + "step": 5430 + }, + { + "epoch": 0.13, + "grad_norm": 1.9995537400633416, + "learning_rate": 1.9501136772803192e-05, + "loss": 1.008, + "step": 5431 + }, + { + "epoch": 0.13, + "grad_norm": 1.078721964641792, + "learning_rate": 1.9500898749713628e-05, + "loss": 0.8818, + "step": 5432 + }, + { + "epoch": 0.13, + "grad_norm": 2.211601079603554, + "learning_rate": 1.9500660671306826e-05, + "loss": 1.0985, + "step": 5433 + }, + { + "epoch": 0.13, + "grad_norm": 1.9459877447243144, + "learning_rate": 1.950042253758417e-05, + "loss": 1.0606, + "step": 5434 + }, + { + "epoch": 0.13, + "grad_norm": 1.9132016026736443, + "learning_rate": 1.950018434854704e-05, + "loss": 0.9937, + "step": 5435 + }, + { + "epoch": 0.13, + "grad_norm": 1.9990297705489912, + "learning_rate": 1.9499946104196835e-05, + "loss": 1.1396, + "step": 5436 + }, + { + "epoch": 0.13, + "grad_norm": 2.3761067965249922, + "learning_rate": 1.9499707804534934e-05, + "loss": 1.0491, + "step": 5437 + }, + { + "epoch": 0.13, + "grad_norm": 2.3051275312080413, + "learning_rate": 1.949946944956273e-05, + "loss": 1.0855, + "step": 5438 + }, + { + "epoch": 0.13, + "grad_norm": 2.0568564737338155, + "learning_rate": 1.9499231039281603e-05, + "loss": 1.0914, + "step": 5439 + }, + { + "epoch": 0.13, + "grad_norm": 5.139119000372019, + "learning_rate": 1.9498992573692948e-05, + "loss": 1.1463, + "step": 5440 + }, + { + "epoch": 0.13, + "grad_norm": 2.1595638793633176, + "learning_rate": 1.949875405279815e-05, + "loss": 1.1947, + "step": 5441 + }, + { + "epoch": 0.13, + "grad_norm": 2.2782224565870783, + "learning_rate": 1.94985154765986e-05, + "loss": 0.95, + "step": 5442 + }, + { + "epoch": 0.13, + "grad_norm": 2.131888718657185, + "learning_rate": 1.949827684509568e-05, + "loss": 1.1542, + "step": 5443 + }, + { + "epoch": 0.13, + "grad_norm": 1.075672179900509, + "learning_rate": 1.9498038158290792e-05, + "loss": 0.9807, + "step": 5444 + }, + { + "epoch": 0.13, + "grad_norm": 2.465602708931142, + "learning_rate": 1.9497799416185318e-05, + "loss": 1.1125, + "step": 5445 + }, + { + "epoch": 0.13, + "grad_norm": 2.0911468150035404, + "learning_rate": 1.949756061878065e-05, + "loss": 1.1719, + "step": 5446 + }, + { + "epoch": 0.13, + "grad_norm": 2.7838083265073346, + "learning_rate": 1.949732176607817e-05, + "loss": 1.1514, + "step": 5447 + }, + { + "epoch": 0.13, + "grad_norm": 2.104303751273575, + "learning_rate": 1.9497082858079284e-05, + "loss": 1.1371, + "step": 5448 + }, + { + "epoch": 0.13, + "grad_norm": 1.1766070144633771, + "learning_rate": 1.9496843894785372e-05, + "loss": 0.9742, + "step": 5449 + }, + { + "epoch": 0.13, + "grad_norm": 2.0804496439511464, + "learning_rate": 1.9496604876197826e-05, + "loss": 1.1243, + "step": 5450 + }, + { + "epoch": 0.13, + "grad_norm": 2.2009579499339784, + "learning_rate": 1.949636580231804e-05, + "loss": 1.0761, + "step": 5451 + }, + { + "epoch": 0.13, + "grad_norm": 2.282471805934984, + "learning_rate": 1.949612667314741e-05, + "loss": 1.0858, + "step": 5452 + }, + { + "epoch": 0.13, + "grad_norm": 2.309607323739121, + "learning_rate": 1.949588748868732e-05, + "loss": 1.0208, + "step": 5453 + }, + { + "epoch": 0.13, + "grad_norm": 2.373368788105025, + "learning_rate": 1.9495648248939166e-05, + "loss": 1.0985, + "step": 5454 + }, + { + "epoch": 0.13, + "grad_norm": 1.9151413336800314, + "learning_rate": 1.9495408953904346e-05, + "loss": 0.9663, + "step": 5455 + }, + { + "epoch": 0.13, + "grad_norm": 1.1640485764317217, + "learning_rate": 1.9495169603584244e-05, + "loss": 0.9646, + "step": 5456 + }, + { + "epoch": 0.13, + "grad_norm": 2.285343278257454, + "learning_rate": 1.949493019798026e-05, + "loss": 1.2066, + "step": 5457 + }, + { + "epoch": 0.13, + "grad_norm": 2.5586478324080977, + "learning_rate": 1.9494690737093787e-05, + "loss": 1.0369, + "step": 5458 + }, + { + "epoch": 0.13, + "grad_norm": 2.012084624203593, + "learning_rate": 1.9494451220926224e-05, + "loss": 0.9831, + "step": 5459 + }, + { + "epoch": 0.13, + "grad_norm": 2.196109788794355, + "learning_rate": 1.9494211649478952e-05, + "loss": 0.9893, + "step": 5460 + }, + { + "epoch": 0.13, + "grad_norm": 2.2333546754296725, + "learning_rate": 1.949397202275338e-05, + "loss": 0.9993, + "step": 5461 + }, + { + "epoch": 0.13, + "grad_norm": 2.013604586880244, + "learning_rate": 1.9493732340750892e-05, + "loss": 1.1533, + "step": 5462 + }, + { + "epoch": 0.13, + "grad_norm": 2.4929829890139796, + "learning_rate": 1.949349260347289e-05, + "loss": 1.0991, + "step": 5463 + }, + { + "epoch": 0.13, + "grad_norm": 2.4969510681017, + "learning_rate": 1.949325281092077e-05, + "loss": 1.0326, + "step": 5464 + }, + { + "epoch": 0.13, + "grad_norm": 2.2707169999899293, + "learning_rate": 1.9493012963095923e-05, + "loss": 1.0321, + "step": 5465 + }, + { + "epoch": 0.13, + "grad_norm": 2.4616480198992168, + "learning_rate": 1.9492773059999752e-05, + "loss": 1.1317, + "step": 5466 + }, + { + "epoch": 0.13, + "grad_norm": 2.047356709688768, + "learning_rate": 1.949253310163365e-05, + "loss": 1.0791, + "step": 5467 + }, + { + "epoch": 0.13, + "grad_norm": 2.164086402083683, + "learning_rate": 1.9492293087999016e-05, + "loss": 1.0931, + "step": 5468 + }, + { + "epoch": 0.13, + "grad_norm": 1.154820876950165, + "learning_rate": 1.9492053019097246e-05, + "loss": 0.9731, + "step": 5469 + }, + { + "epoch": 0.13, + "grad_norm": 1.9676696176094923, + "learning_rate": 1.949181289492974e-05, + "loss": 0.9843, + "step": 5470 + }, + { + "epoch": 0.13, + "grad_norm": 2.409014280792976, + "learning_rate": 1.9491572715497894e-05, + "loss": 1.0986, + "step": 5471 + }, + { + "epoch": 0.13, + "grad_norm": 2.176577401953118, + "learning_rate": 1.9491332480803106e-05, + "loss": 1.0607, + "step": 5472 + }, + { + "epoch": 0.13, + "grad_norm": 6.511667931031736, + "learning_rate": 1.949109219084677e-05, + "loss": 1.0317, + "step": 5473 + }, + { + "epoch": 0.13, + "grad_norm": 2.2106104591409155, + "learning_rate": 1.9490851845630297e-05, + "loss": 1.1514, + "step": 5474 + }, + { + "epoch": 0.13, + "grad_norm": 2.5241654679472827, + "learning_rate": 1.9490611445155078e-05, + "loss": 0.9626, + "step": 5475 + }, + { + "epoch": 0.13, + "grad_norm": 2.159356355340786, + "learning_rate": 1.9490370989422516e-05, + "loss": 1.1619, + "step": 5476 + }, + { + "epoch": 0.13, + "grad_norm": 2.00222677988906, + "learning_rate": 1.9490130478434008e-05, + "loss": 1.1211, + "step": 5477 + }, + { + "epoch": 0.13, + "grad_norm": 2.137380189223518, + "learning_rate": 1.9489889912190954e-05, + "loss": 1.1209, + "step": 5478 + }, + { + "epoch": 0.13, + "grad_norm": 2.1805978681483174, + "learning_rate": 1.948964929069476e-05, + "loss": 0.9556, + "step": 5479 + }, + { + "epoch": 0.13, + "grad_norm": 2.2498265306029377, + "learning_rate": 1.948940861394682e-05, + "loss": 0.9972, + "step": 5480 + }, + { + "epoch": 0.13, + "grad_norm": 1.1973369817876627, + "learning_rate": 1.948916788194854e-05, + "loss": 0.9875, + "step": 5481 + }, + { + "epoch": 0.13, + "grad_norm": 2.3019131637168453, + "learning_rate": 1.9488927094701322e-05, + "loss": 1.1018, + "step": 5482 + }, + { + "epoch": 0.13, + "grad_norm": 2.518125261802848, + "learning_rate": 1.9488686252206564e-05, + "loss": 1.0008, + "step": 5483 + }, + { + "epoch": 0.13, + "grad_norm": 2.7822225675423655, + "learning_rate": 1.948844535446567e-05, + "loss": 1.0317, + "step": 5484 + }, + { + "epoch": 0.13, + "grad_norm": 2.3991045152485238, + "learning_rate": 1.948820440148005e-05, + "loss": 1.0499, + "step": 5485 + }, + { + "epoch": 0.13, + "grad_norm": 2.19282947006848, + "learning_rate": 1.948796339325109e-05, + "loss": 1.0014, + "step": 5486 + }, + { + "epoch": 0.13, + "grad_norm": 2.5516425192291776, + "learning_rate": 1.948772232978021e-05, + "loss": 1.0635, + "step": 5487 + }, + { + "epoch": 0.13, + "grad_norm": 2.2148345271344048, + "learning_rate": 1.9487481211068807e-05, + "loss": 1.2466, + "step": 5488 + }, + { + "epoch": 0.13, + "grad_norm": 2.0809951269790967, + "learning_rate": 1.9487240037118283e-05, + "loss": 0.9177, + "step": 5489 + }, + { + "epoch": 0.13, + "grad_norm": 2.000287432873148, + "learning_rate": 1.9486998807930043e-05, + "loss": 1.156, + "step": 5490 + }, + { + "epoch": 0.13, + "grad_norm": 3.5957686038455456, + "learning_rate": 1.9486757523505495e-05, + "loss": 0.9641, + "step": 5491 + }, + { + "epoch": 0.13, + "grad_norm": 2.278474681103141, + "learning_rate": 1.9486516183846036e-05, + "loss": 1.0413, + "step": 5492 + }, + { + "epoch": 0.13, + "grad_norm": 2.659582804199996, + "learning_rate": 1.9486274788953083e-05, + "loss": 1.2247, + "step": 5493 + }, + { + "epoch": 0.13, + "grad_norm": 1.1452524910004733, + "learning_rate": 1.948603333882803e-05, + "loss": 0.9821, + "step": 5494 + }, + { + "epoch": 0.13, + "grad_norm": 1.1123740816783403, + "learning_rate": 1.9485791833472286e-05, + "loss": 1.041, + "step": 5495 + }, + { + "epoch": 0.13, + "grad_norm": 2.6303267465467686, + "learning_rate": 1.9485550272887265e-05, + "loss": 1.1039, + "step": 5496 + }, + { + "epoch": 0.13, + "grad_norm": 2.202690199864643, + "learning_rate": 1.9485308657074364e-05, + "loss": 1.0209, + "step": 5497 + }, + { + "epoch": 0.13, + "grad_norm": 1.1519554528571507, + "learning_rate": 1.9485066986034992e-05, + "loss": 0.9904, + "step": 5498 + }, + { + "epoch": 0.13, + "grad_norm": 2.189429172628161, + "learning_rate": 1.948482525977056e-05, + "loss": 1.0678, + "step": 5499 + }, + { + "epoch": 0.13, + "grad_norm": 2.375752900058565, + "learning_rate": 1.948458347828247e-05, + "loss": 1.055, + "step": 5500 + }, + { + "epoch": 0.13, + "grad_norm": 2.002662090976573, + "learning_rate": 1.9484341641572134e-05, + "loss": 1.0058, + "step": 5501 + }, + { + "epoch": 0.13, + "grad_norm": 1.0742069607868046, + "learning_rate": 1.9484099749640956e-05, + "loss": 0.9626, + "step": 5502 + }, + { + "epoch": 0.13, + "grad_norm": 2.809100306328975, + "learning_rate": 1.9483857802490348e-05, + "loss": 1.1179, + "step": 5503 + }, + { + "epoch": 0.13, + "grad_norm": 2.153036581861293, + "learning_rate": 1.9483615800121717e-05, + "loss": 0.9771, + "step": 5504 + }, + { + "epoch": 0.13, + "grad_norm": 1.1241205390595423, + "learning_rate": 1.9483373742536475e-05, + "loss": 0.9823, + "step": 5505 + }, + { + "epoch": 0.13, + "grad_norm": 2.632019803378113, + "learning_rate": 1.9483131629736024e-05, + "loss": 1.0382, + "step": 5506 + }, + { + "epoch": 0.13, + "grad_norm": 2.150721376866039, + "learning_rate": 1.9482889461721782e-05, + "loss": 1.0, + "step": 5507 + }, + { + "epoch": 0.13, + "grad_norm": 2.09782377553955, + "learning_rate": 1.948264723849515e-05, + "loss": 0.9761, + "step": 5508 + }, + { + "epoch": 0.13, + "grad_norm": 2.293811574555748, + "learning_rate": 1.948240496005755e-05, + "loss": 1.1528, + "step": 5509 + }, + { + "epoch": 0.13, + "grad_norm": 2.0646591489168666, + "learning_rate": 1.9482162626410383e-05, + "loss": 1.0744, + "step": 5510 + }, + { + "epoch": 0.13, + "grad_norm": 1.1269967828521859, + "learning_rate": 1.9481920237555064e-05, + "loss": 1.0578, + "step": 5511 + }, + { + "epoch": 0.13, + "grad_norm": 1.828413716110437, + "learning_rate": 1.9481677793493005e-05, + "loss": 1.1247, + "step": 5512 + }, + { + "epoch": 0.13, + "grad_norm": 2.1493424190685713, + "learning_rate": 1.9481435294225614e-05, + "loss": 1.0829, + "step": 5513 + }, + { + "epoch": 0.13, + "grad_norm": 2.272061123243631, + "learning_rate": 1.9481192739754307e-05, + "loss": 1.0383, + "step": 5514 + }, + { + "epoch": 0.13, + "grad_norm": 3.794254933921255, + "learning_rate": 1.948095013008049e-05, + "loss": 1.0054, + "step": 5515 + }, + { + "epoch": 0.13, + "grad_norm": 1.8732979717297007, + "learning_rate": 1.9480707465205582e-05, + "loss": 1.0377, + "step": 5516 + }, + { + "epoch": 0.13, + "grad_norm": 2.2454740090159615, + "learning_rate": 1.9480464745130997e-05, + "loss": 1.0546, + "step": 5517 + }, + { + "epoch": 0.13, + "grad_norm": 2.123520461851966, + "learning_rate": 1.948022196985814e-05, + "loss": 1.1051, + "step": 5518 + }, + { + "epoch": 0.13, + "grad_norm": 2.4750063388386114, + "learning_rate": 1.9479979139388432e-05, + "loss": 1.0483, + "step": 5519 + }, + { + "epoch": 0.13, + "grad_norm": 2.3227742311319357, + "learning_rate": 1.9479736253723282e-05, + "loss": 1.0996, + "step": 5520 + }, + { + "epoch": 0.13, + "grad_norm": 2.3160158657830485, + "learning_rate": 1.9479493312864107e-05, + "loss": 1.1922, + "step": 5521 + }, + { + "epoch": 0.13, + "grad_norm": 2.0707881939925277, + "learning_rate": 1.9479250316812323e-05, + "loss": 1.0169, + "step": 5522 + }, + { + "epoch": 0.13, + "grad_norm": 2.213006646161037, + "learning_rate": 1.9479007265569342e-05, + "loss": 1.0871, + "step": 5523 + }, + { + "epoch": 0.13, + "grad_norm": 2.2314596711996866, + "learning_rate": 1.9478764159136578e-05, + "loss": 1.1269, + "step": 5524 + }, + { + "epoch": 0.13, + "grad_norm": 2.1782716623944833, + "learning_rate": 1.9478520997515446e-05, + "loss": 1.0757, + "step": 5525 + }, + { + "epoch": 0.13, + "grad_norm": 2.047052206720587, + "learning_rate": 1.9478277780707367e-05, + "loss": 1.0297, + "step": 5526 + }, + { + "epoch": 0.13, + "grad_norm": 2.1487714773381645, + "learning_rate": 1.9478034508713753e-05, + "loss": 1.0409, + "step": 5527 + }, + { + "epoch": 0.13, + "grad_norm": 2.2526581136104475, + "learning_rate": 1.9477791181536024e-05, + "loss": 1.1434, + "step": 5528 + }, + { + "epoch": 0.13, + "grad_norm": 2.0442256035883397, + "learning_rate": 1.9477547799175592e-05, + "loss": 1.0114, + "step": 5529 + }, + { + "epoch": 0.13, + "grad_norm": 2.4620324356608916, + "learning_rate": 1.9477304361633878e-05, + "loss": 1.0021, + "step": 5530 + }, + { + "epoch": 0.13, + "grad_norm": 1.328832183519726, + "learning_rate": 1.9477060868912294e-05, + "loss": 1.0416, + "step": 5531 + }, + { + "epoch": 0.13, + "grad_norm": 2.198571330793663, + "learning_rate": 1.9476817321012263e-05, + "loss": 1.1484, + "step": 5532 + }, + { + "epoch": 0.13, + "grad_norm": 2.126706704248076, + "learning_rate": 1.94765737179352e-05, + "loss": 1.0326, + "step": 5533 + }, + { + "epoch": 0.13, + "grad_norm": 2.0274679409146694, + "learning_rate": 1.9476330059682528e-05, + "loss": 0.996, + "step": 5534 + }, + { + "epoch": 0.13, + "grad_norm": 2.0649238967581622, + "learning_rate": 1.9476086346255657e-05, + "loss": 1.126, + "step": 5535 + }, + { + "epoch": 0.13, + "grad_norm": 1.076517002164408, + "learning_rate": 1.9475842577656018e-05, + "loss": 1.0288, + "step": 5536 + }, + { + "epoch": 0.13, + "grad_norm": 1.9743899298687506, + "learning_rate": 1.947559875388502e-05, + "loss": 1.063, + "step": 5537 + }, + { + "epoch": 0.13, + "grad_norm": 2.4635520703120086, + "learning_rate": 1.9475354874944083e-05, + "loss": 1.0206, + "step": 5538 + }, + { + "epoch": 0.13, + "grad_norm": 2.083200974680579, + "learning_rate": 1.9475110940834633e-05, + "loss": 1.1407, + "step": 5539 + }, + { + "epoch": 0.13, + "grad_norm": 2.1662862410228527, + "learning_rate": 1.947486695155809e-05, + "loss": 1.0995, + "step": 5540 + }, + { + "epoch": 0.13, + "grad_norm": 2.314237366450072, + "learning_rate": 1.947462290711587e-05, + "loss": 1.1179, + "step": 5541 + }, + { + "epoch": 0.13, + "grad_norm": 2.1407618916119135, + "learning_rate": 1.9474378807509395e-05, + "loss": 0.9386, + "step": 5542 + }, + { + "epoch": 0.13, + "grad_norm": 2.131223454958576, + "learning_rate": 1.9474134652740083e-05, + "loss": 1.0682, + "step": 5543 + }, + { + "epoch": 0.13, + "grad_norm": 1.9240066927126118, + "learning_rate": 1.9473890442809366e-05, + "loss": 1.2169, + "step": 5544 + }, + { + "epoch": 0.13, + "grad_norm": 1.8967922375281887, + "learning_rate": 1.9473646177718658e-05, + "loss": 1.1422, + "step": 5545 + }, + { + "epoch": 0.13, + "grad_norm": 2.356727771492962, + "learning_rate": 1.947340185746938e-05, + "loss": 1.0486, + "step": 5546 + }, + { + "epoch": 0.13, + "grad_norm": 2.0734690233499102, + "learning_rate": 1.9473157482062958e-05, + "loss": 1.2358, + "step": 5547 + }, + { + "epoch": 0.13, + "grad_norm": 2.093044587190984, + "learning_rate": 1.9472913051500816e-05, + "loss": 1.0998, + "step": 5548 + }, + { + "epoch": 0.13, + "grad_norm": 2.0285400421178754, + "learning_rate": 1.9472668565784375e-05, + "loss": 1.0169, + "step": 5549 + }, + { + "epoch": 0.13, + "grad_norm": 2.466340623197113, + "learning_rate": 1.9472424024915056e-05, + "loss": 1.0609, + "step": 5550 + }, + { + "epoch": 0.13, + "grad_norm": 2.426395265987443, + "learning_rate": 1.9472179428894288e-05, + "loss": 1.1114, + "step": 5551 + }, + { + "epoch": 0.13, + "grad_norm": 2.634337346151071, + "learning_rate": 1.9471934777723492e-05, + "loss": 1.0459, + "step": 5552 + }, + { + "epoch": 0.13, + "grad_norm": 2.267886135032713, + "learning_rate": 1.9471690071404096e-05, + "loss": 1.0976, + "step": 5553 + }, + { + "epoch": 0.13, + "grad_norm": 2.5817816788816943, + "learning_rate": 1.9471445309937515e-05, + "loss": 1.1744, + "step": 5554 + }, + { + "epoch": 0.13, + "grad_norm": 2.0950646930055896, + "learning_rate": 1.9471200493325186e-05, + "loss": 1.0014, + "step": 5555 + }, + { + "epoch": 0.13, + "grad_norm": 2.518922747552857, + "learning_rate": 1.9470955621568525e-05, + "loss": 1.1236, + "step": 5556 + }, + { + "epoch": 0.13, + "grad_norm": 2.242498384789629, + "learning_rate": 1.9470710694668966e-05, + "loss": 1.0129, + "step": 5557 + }, + { + "epoch": 0.13, + "grad_norm": 2.1403095664909535, + "learning_rate": 1.9470465712627928e-05, + "loss": 1.0714, + "step": 5558 + }, + { + "epoch": 0.13, + "grad_norm": 2.490909996383375, + "learning_rate": 1.9470220675446844e-05, + "loss": 1.0309, + "step": 5559 + }, + { + "epoch": 0.13, + "grad_norm": 2.6430214194129085, + "learning_rate": 1.946997558312713e-05, + "loss": 1.0645, + "step": 5560 + }, + { + "epoch": 0.13, + "grad_norm": 2.318361240167679, + "learning_rate": 1.9469730435670227e-05, + "loss": 1.1006, + "step": 5561 + }, + { + "epoch": 0.13, + "grad_norm": 2.465704086321714, + "learning_rate": 1.9469485233077554e-05, + "loss": 1.1457, + "step": 5562 + }, + { + "epoch": 0.13, + "grad_norm": 2.0349182792437315, + "learning_rate": 1.9469239975350538e-05, + "loss": 1.1356, + "step": 5563 + }, + { + "epoch": 0.13, + "grad_norm": 1.956469930499822, + "learning_rate": 1.946899466249061e-05, + "loss": 0.9907, + "step": 5564 + }, + { + "epoch": 0.13, + "grad_norm": 1.1868266755233172, + "learning_rate": 1.9468749294499196e-05, + "loss": 0.8787, + "step": 5565 + }, + { + "epoch": 0.13, + "grad_norm": 1.0997649149713908, + "learning_rate": 1.9468503871377728e-05, + "loss": 0.9461, + "step": 5566 + }, + { + "epoch": 0.13, + "grad_norm": 2.256883110046935, + "learning_rate": 1.946825839312763e-05, + "loss": 1.0262, + "step": 5567 + }, + { + "epoch": 0.13, + "grad_norm": 2.690703669113356, + "learning_rate": 1.9468012859750337e-05, + "loss": 1.1399, + "step": 5568 + }, + { + "epoch": 0.13, + "grad_norm": 2.029880859925913, + "learning_rate": 1.9467767271247275e-05, + "loss": 1.0713, + "step": 5569 + }, + { + "epoch": 0.13, + "grad_norm": 2.073740702815589, + "learning_rate": 1.9467521627619874e-05, + "loss": 1.0243, + "step": 5570 + }, + { + "epoch": 0.13, + "grad_norm": 2.534359052039932, + "learning_rate": 1.9467275928869565e-05, + "loss": 1.1196, + "step": 5571 + }, + { + "epoch": 0.13, + "grad_norm": 1.9123600654436315, + "learning_rate": 1.9467030174997777e-05, + "loss": 1.1817, + "step": 5572 + }, + { + "epoch": 0.13, + "grad_norm": 1.9392528977405303, + "learning_rate": 1.9466784366005943e-05, + "loss": 1.2794, + "step": 5573 + }, + { + "epoch": 0.13, + "grad_norm": 2.2846929910988982, + "learning_rate": 1.9466538501895493e-05, + "loss": 1.0088, + "step": 5574 + }, + { + "epoch": 0.13, + "grad_norm": 1.9780654014866141, + "learning_rate": 1.946629258266786e-05, + "loss": 1.2052, + "step": 5575 + }, + { + "epoch": 0.13, + "grad_norm": 2.2189880480762927, + "learning_rate": 1.9466046608324472e-05, + "loss": 1.1231, + "step": 5576 + }, + { + "epoch": 0.13, + "grad_norm": 2.1169427450023206, + "learning_rate": 1.9465800578866764e-05, + "loss": 1.0816, + "step": 5577 + }, + { + "epoch": 0.13, + "grad_norm": 2.1260959276216185, + "learning_rate": 1.9465554494296173e-05, + "loss": 1.0653, + "step": 5578 + }, + { + "epoch": 0.13, + "grad_norm": 2.368861873007841, + "learning_rate": 1.9465308354614125e-05, + "loss": 0.9944, + "step": 5579 + }, + { + "epoch": 0.13, + "grad_norm": 2.2702107156168014, + "learning_rate": 1.946506215982205e-05, + "loss": 0.8269, + "step": 5580 + }, + { + "epoch": 0.13, + "grad_norm": 2.252686272027046, + "learning_rate": 1.946481590992139e-05, + "loss": 1.1327, + "step": 5581 + }, + { + "epoch": 0.13, + "grad_norm": 4.449863599151083, + "learning_rate": 1.9464569604913575e-05, + "loss": 0.96, + "step": 5582 + }, + { + "epoch": 0.13, + "grad_norm": 1.8695042008304488, + "learning_rate": 1.9464323244800042e-05, + "loss": 1.0007, + "step": 5583 + }, + { + "epoch": 0.13, + "grad_norm": 2.0860966454742194, + "learning_rate": 1.946407682958222e-05, + "loss": 0.9582, + "step": 5584 + }, + { + "epoch": 0.13, + "grad_norm": 2.1298801548034794, + "learning_rate": 1.9463830359261547e-05, + "loss": 1.0414, + "step": 5585 + }, + { + "epoch": 0.13, + "grad_norm": 1.1234214106347378, + "learning_rate": 1.946358383383946e-05, + "loss": 1.0225, + "step": 5586 + }, + { + "epoch": 0.13, + "grad_norm": 2.16533870595778, + "learning_rate": 1.9463337253317385e-05, + "loss": 1.0123, + "step": 5587 + }, + { + "epoch": 0.13, + "grad_norm": 1.151070033093567, + "learning_rate": 1.946309061769677e-05, + "loss": 0.9696, + "step": 5588 + }, + { + "epoch": 0.13, + "grad_norm": 1.0923754386612057, + "learning_rate": 1.9462843926979044e-05, + "loss": 1.038, + "step": 5589 + }, + { + "epoch": 0.13, + "grad_norm": 2.012326925436781, + "learning_rate": 1.9462597181165647e-05, + "loss": 1.1272, + "step": 5590 + }, + { + "epoch": 0.13, + "grad_norm": 2.054372425747401, + "learning_rate": 1.946235038025801e-05, + "loss": 1.0562, + "step": 5591 + }, + { + "epoch": 0.13, + "grad_norm": 2.224311125076984, + "learning_rate": 1.9462103524257574e-05, + "loss": 1.1161, + "step": 5592 + }, + { + "epoch": 0.13, + "grad_norm": 2.429368530008861, + "learning_rate": 1.9461856613165775e-05, + "loss": 1.109, + "step": 5593 + }, + { + "epoch": 0.13, + "grad_norm": 2.041918079879984, + "learning_rate": 1.946160964698405e-05, + "loss": 1.0409, + "step": 5594 + }, + { + "epoch": 0.13, + "grad_norm": 2.183657242418282, + "learning_rate": 1.9461362625713845e-05, + "loss": 1.0438, + "step": 5595 + }, + { + "epoch": 0.13, + "grad_norm": 1.099675311323085, + "learning_rate": 1.9461115549356588e-05, + "loss": 0.971, + "step": 5596 + }, + { + "epoch": 0.13, + "grad_norm": 2.189098033941666, + "learning_rate": 1.9460868417913718e-05, + "loss": 1.1405, + "step": 5597 + }, + { + "epoch": 0.13, + "grad_norm": 2.2002360122808873, + "learning_rate": 1.946062123138668e-05, + "loss": 1.0311, + "step": 5598 + }, + { + "epoch": 0.13, + "grad_norm": 2.110798465835888, + "learning_rate": 1.9460373989776905e-05, + "loss": 1.1275, + "step": 5599 + }, + { + "epoch": 0.13, + "grad_norm": 2.062335725032407, + "learning_rate": 1.9460126693085842e-05, + "loss": 1.0622, + "step": 5600 + }, + { + "epoch": 0.13, + "grad_norm": 2.1429218257372593, + "learning_rate": 1.9459879341314927e-05, + "loss": 1.0267, + "step": 5601 + }, + { + "epoch": 0.13, + "grad_norm": 2.387354006390867, + "learning_rate": 1.9459631934465596e-05, + "loss": 1.0391, + "step": 5602 + }, + { + "epoch": 0.13, + "grad_norm": 1.17770112502354, + "learning_rate": 1.9459384472539297e-05, + "loss": 1.0153, + "step": 5603 + }, + { + "epoch": 0.13, + "grad_norm": 2.019679647662008, + "learning_rate": 1.9459136955537465e-05, + "loss": 1.1109, + "step": 5604 + }, + { + "epoch": 0.13, + "grad_norm": 2.201460832044663, + "learning_rate": 1.9458889383461543e-05, + "loss": 1.1161, + "step": 5605 + }, + { + "epoch": 0.13, + "grad_norm": 2.5816675937666465, + "learning_rate": 1.9458641756312973e-05, + "loss": 1.1164, + "step": 5606 + }, + { + "epoch": 0.13, + "grad_norm": 2.2370907211017235, + "learning_rate": 1.9458394074093198e-05, + "loss": 1.0003, + "step": 5607 + }, + { + "epoch": 0.13, + "grad_norm": 2.222231889994378, + "learning_rate": 1.9458146336803653e-05, + "loss": 1.0212, + "step": 5608 + }, + { + "epoch": 0.13, + "grad_norm": 2.1837634567019855, + "learning_rate": 1.945789854444579e-05, + "loss": 1.0642, + "step": 5609 + }, + { + "epoch": 0.13, + "grad_norm": 2.5461166330951146, + "learning_rate": 1.9457650697021046e-05, + "loss": 0.983, + "step": 5610 + }, + { + "epoch": 0.13, + "grad_norm": 2.3794384688491728, + "learning_rate": 1.9457402794530864e-05, + "loss": 1.1348, + "step": 5611 + }, + { + "epoch": 0.13, + "grad_norm": 2.2980440864553366, + "learning_rate": 1.945715483697669e-05, + "loss": 1.2075, + "step": 5612 + }, + { + "epoch": 0.13, + "grad_norm": 2.8253425077852765, + "learning_rate": 1.9456906824359965e-05, + "loss": 1.1046, + "step": 5613 + }, + { + "epoch": 0.13, + "grad_norm": 2.370199598096588, + "learning_rate": 1.9456658756682136e-05, + "loss": 1.1541, + "step": 5614 + }, + { + "epoch": 0.13, + "grad_norm": 2.13644797247828, + "learning_rate": 1.9456410633944645e-05, + "loss": 1.0848, + "step": 5615 + }, + { + "epoch": 0.13, + "grad_norm": 1.11701128914641, + "learning_rate": 1.9456162456148938e-05, + "loss": 0.958, + "step": 5616 + }, + { + "epoch": 0.13, + "grad_norm": 2.387104884673864, + "learning_rate": 1.9455914223296458e-05, + "loss": 1.0882, + "step": 5617 + }, + { + "epoch": 0.13, + "grad_norm": 2.3433067141487625, + "learning_rate": 1.9455665935388653e-05, + "loss": 0.8945, + "step": 5618 + }, + { + "epoch": 0.13, + "grad_norm": 2.0271018963313536, + "learning_rate": 1.945541759242697e-05, + "loss": 1.0022, + "step": 5619 + }, + { + "epoch": 0.13, + "grad_norm": 1.101098688576892, + "learning_rate": 1.945516919441285e-05, + "loss": 1.0237, + "step": 5620 + }, + { + "epoch": 0.13, + "grad_norm": 2.274732256260502, + "learning_rate": 1.945492074134774e-05, + "loss": 0.955, + "step": 5621 + }, + { + "epoch": 0.13, + "grad_norm": 2.320846840272797, + "learning_rate": 1.945467223323309e-05, + "loss": 1.0485, + "step": 5622 + }, + { + "epoch": 0.13, + "grad_norm": 1.928708020582287, + "learning_rate": 1.945442367007034e-05, + "loss": 1.0373, + "step": 5623 + }, + { + "epoch": 0.13, + "grad_norm": 2.0324712191400645, + "learning_rate": 1.945417505186095e-05, + "loss": 1.0174, + "step": 5624 + }, + { + "epoch": 0.13, + "grad_norm": 2.2060899873626356, + "learning_rate": 1.9453926378606354e-05, + "loss": 1.0282, + "step": 5625 + }, + { + "epoch": 0.13, + "grad_norm": 2.2588008766096253, + "learning_rate": 1.9453677650308007e-05, + "loss": 1.101, + "step": 5626 + }, + { + "epoch": 0.13, + "grad_norm": 2.012038958976612, + "learning_rate": 1.945342886696736e-05, + "loss": 1.1016, + "step": 5627 + }, + { + "epoch": 0.13, + "grad_norm": 2.6200135223306686, + "learning_rate": 1.945318002858585e-05, + "loss": 1.1083, + "step": 5628 + }, + { + "epoch": 0.13, + "grad_norm": 2.271291749015936, + "learning_rate": 1.9452931135164936e-05, + "loss": 0.9944, + "step": 5629 + }, + { + "epoch": 0.13, + "grad_norm": 2.1908147561008877, + "learning_rate": 1.9452682186706067e-05, + "loss": 1.1083, + "step": 5630 + }, + { + "epoch": 0.13, + "grad_norm": 2.5224387188766206, + "learning_rate": 1.945243318321069e-05, + "loss": 1.155, + "step": 5631 + }, + { + "epoch": 0.13, + "grad_norm": 2.01101887042025, + "learning_rate": 1.9452184124680247e-05, + "loss": 1.1139, + "step": 5632 + }, + { + "epoch": 0.13, + "grad_norm": 2.338171116694816, + "learning_rate": 1.94519350111162e-05, + "loss": 0.9743, + "step": 5633 + }, + { + "epoch": 0.13, + "grad_norm": 2.1719462179138485, + "learning_rate": 1.9451685842519997e-05, + "loss": 0.9594, + "step": 5634 + }, + { + "epoch": 0.13, + "grad_norm": 2.4520086986774774, + "learning_rate": 1.9451436618893086e-05, + "loss": 1.0383, + "step": 5635 + }, + { + "epoch": 0.13, + "grad_norm": 7.11286927125518, + "learning_rate": 1.9451187340236918e-05, + "loss": 0.9855, + "step": 5636 + }, + { + "epoch": 0.13, + "grad_norm": 2.305616646807856, + "learning_rate": 1.9450938006552943e-05, + "loss": 0.9957, + "step": 5637 + }, + { + "epoch": 0.13, + "grad_norm": 2.422828899175301, + "learning_rate": 1.945068861784262e-05, + "loss": 1.1316, + "step": 5638 + }, + { + "epoch": 0.13, + "grad_norm": 2.113839967483324, + "learning_rate": 1.9450439174107388e-05, + "loss": 1.0204, + "step": 5639 + }, + { + "epoch": 0.13, + "grad_norm": 2.2990718167893855, + "learning_rate": 1.9450189675348712e-05, + "loss": 1.1815, + "step": 5640 + }, + { + "epoch": 0.13, + "grad_norm": 2.1007761519990154, + "learning_rate": 1.944994012156804e-05, + "loss": 1.0227, + "step": 5641 + }, + { + "epoch": 0.13, + "grad_norm": 1.1446341150606456, + "learning_rate": 1.9449690512766825e-05, + "loss": 1.0463, + "step": 5642 + }, + { + "epoch": 0.13, + "grad_norm": 2.057775830447761, + "learning_rate": 1.944944084894652e-05, + "loss": 1.1328, + "step": 5643 + }, + { + "epoch": 0.13, + "grad_norm": 2.4383000694570445, + "learning_rate": 1.944919113010858e-05, + "loss": 0.9206, + "step": 5644 + }, + { + "epoch": 0.13, + "grad_norm": 1.934859880100627, + "learning_rate": 1.9448941356254453e-05, + "loss": 1.0655, + "step": 5645 + }, + { + "epoch": 0.13, + "grad_norm": 2.4734244750216723, + "learning_rate": 1.94486915273856e-05, + "loss": 1.1469, + "step": 5646 + }, + { + "epoch": 0.13, + "grad_norm": 1.1191036929158178, + "learning_rate": 1.9448441643503473e-05, + "loss": 1.0272, + "step": 5647 + }, + { + "epoch": 0.13, + "grad_norm": 2.4706937540954677, + "learning_rate": 1.944819170460953e-05, + "loss": 1.2293, + "step": 5648 + }, + { + "epoch": 0.13, + "grad_norm": 2.305898723310186, + "learning_rate": 1.9447941710705223e-05, + "loss": 1.1083, + "step": 5649 + }, + { + "epoch": 0.13, + "grad_norm": 2.2353798706992016, + "learning_rate": 1.9447691661792007e-05, + "loss": 0.9587, + "step": 5650 + }, + { + "epoch": 0.13, + "grad_norm": 2.5193473880054245, + "learning_rate": 1.944744155787134e-05, + "loss": 1.2548, + "step": 5651 + }, + { + "epoch": 0.13, + "grad_norm": 2.51473689738483, + "learning_rate": 1.944719139894468e-05, + "loss": 1.0096, + "step": 5652 + }, + { + "epoch": 0.13, + "grad_norm": 2.028409527470473, + "learning_rate": 1.9446941185013474e-05, + "loss": 1.0599, + "step": 5653 + }, + { + "epoch": 0.13, + "grad_norm": 1.1194060250804185, + "learning_rate": 1.944669091607919e-05, + "loss": 0.949, + "step": 5654 + }, + { + "epoch": 0.13, + "grad_norm": 2.0364370091391777, + "learning_rate": 1.944644059214328e-05, + "loss": 1.1409, + "step": 5655 + }, + { + "epoch": 0.13, + "grad_norm": 2.609468984303286, + "learning_rate": 1.94461902132072e-05, + "loss": 0.9481, + "step": 5656 + }, + { + "epoch": 0.13, + "grad_norm": 1.8820061203358458, + "learning_rate": 1.9445939779272415e-05, + "loss": 1.1457, + "step": 5657 + }, + { + "epoch": 0.13, + "grad_norm": 1.0772424046900477, + "learning_rate": 1.9445689290340372e-05, + "loss": 1.0481, + "step": 5658 + }, + { + "epoch": 0.13, + "grad_norm": 2.093867220150486, + "learning_rate": 1.944543874641254e-05, + "loss": 1.0312, + "step": 5659 + }, + { + "epoch": 0.13, + "grad_norm": 2.124361051824991, + "learning_rate": 1.9445188147490373e-05, + "loss": 1.0929, + "step": 5660 + }, + { + "epoch": 0.13, + "grad_norm": 2.564033499935242, + "learning_rate": 1.9444937493575326e-05, + "loss": 1.1569, + "step": 5661 + }, + { + "epoch": 0.13, + "grad_norm": 2.3885159447446145, + "learning_rate": 1.9444686784668868e-05, + "loss": 1.0585, + "step": 5662 + }, + { + "epoch": 0.13, + "grad_norm": 1.13094377490991, + "learning_rate": 1.944443602077245e-05, + "loss": 0.9642, + "step": 5663 + }, + { + "epoch": 0.13, + "grad_norm": 2.2620748405730384, + "learning_rate": 1.9444185201887536e-05, + "loss": 0.9994, + "step": 5664 + }, + { + "epoch": 0.13, + "grad_norm": 1.1214394217167716, + "learning_rate": 1.9443934328015587e-05, + "loss": 0.9915, + "step": 5665 + }, + { + "epoch": 0.13, + "grad_norm": 2.2577813981233876, + "learning_rate": 1.944368339915806e-05, + "loss": 0.9584, + "step": 5666 + }, + { + "epoch": 0.13, + "grad_norm": 1.8608581593492222, + "learning_rate": 1.9443432415316424e-05, + "loss": 1.0092, + "step": 5667 + }, + { + "epoch": 0.13, + "grad_norm": 2.1036860018235046, + "learning_rate": 1.944318137649213e-05, + "loss": 1.1966, + "step": 5668 + }, + { + "epoch": 0.13, + "grad_norm": 2.3870408306953648, + "learning_rate": 1.9442930282686647e-05, + "loss": 1.1341, + "step": 5669 + }, + { + "epoch": 0.13, + "grad_norm": 2.27783458603269, + "learning_rate": 1.9442679133901433e-05, + "loss": 1.1376, + "step": 5670 + }, + { + "epoch": 0.13, + "grad_norm": 2.259998996249894, + "learning_rate": 1.9442427930137948e-05, + "loss": 1.0574, + "step": 5671 + }, + { + "epoch": 0.13, + "grad_norm": 1.9456467647695335, + "learning_rate": 1.9442176671397663e-05, + "loss": 1.1193, + "step": 5672 + }, + { + "epoch": 0.13, + "grad_norm": 4.300222330215989, + "learning_rate": 1.9441925357682033e-05, + "loss": 1.0423, + "step": 5673 + }, + { + "epoch": 0.13, + "grad_norm": 2.192724963615063, + "learning_rate": 1.9441673988992525e-05, + "loss": 1.0166, + "step": 5674 + }, + { + "epoch": 0.13, + "grad_norm": 2.0051394287672504, + "learning_rate": 1.9441422565330606e-05, + "loss": 1.0795, + "step": 5675 + }, + { + "epoch": 0.13, + "grad_norm": 1.9802120122742848, + "learning_rate": 1.9441171086697732e-05, + "loss": 0.9792, + "step": 5676 + }, + { + "epoch": 0.13, + "grad_norm": 2.4606889163027406, + "learning_rate": 1.944091955309537e-05, + "loss": 1.1238, + "step": 5677 + }, + { + "epoch": 0.13, + "grad_norm": 2.0623184720489185, + "learning_rate": 1.9440667964524987e-05, + "loss": 0.9575, + "step": 5678 + }, + { + "epoch": 0.13, + "grad_norm": 3.2081820725878094, + "learning_rate": 1.9440416320988046e-05, + "loss": 1.0418, + "step": 5679 + }, + { + "epoch": 0.13, + "grad_norm": 2.0570526606353345, + "learning_rate": 1.9440164622486013e-05, + "loss": 1.0881, + "step": 5680 + }, + { + "epoch": 0.13, + "grad_norm": 1.9595366585623937, + "learning_rate": 1.9439912869020353e-05, + "loss": 1.2196, + "step": 5681 + }, + { + "epoch": 0.13, + "grad_norm": 2.0887590871235284, + "learning_rate": 1.943966106059253e-05, + "loss": 1.1083, + "step": 5682 + }, + { + "epoch": 0.13, + "grad_norm": 1.115723541580784, + "learning_rate": 1.9439409197204012e-05, + "loss": 1.0134, + "step": 5683 + }, + { + "epoch": 0.13, + "grad_norm": 1.991540754864275, + "learning_rate": 1.9439157278856268e-05, + "loss": 1.1518, + "step": 5684 + }, + { + "epoch": 0.13, + "grad_norm": 2.156514827905873, + "learning_rate": 1.9438905305550756e-05, + "loss": 0.9438, + "step": 5685 + }, + { + "epoch": 0.13, + "grad_norm": 1.8897205711120892, + "learning_rate": 1.9438653277288953e-05, + "loss": 1.0981, + "step": 5686 + }, + { + "epoch": 0.13, + "grad_norm": 2.337075846970485, + "learning_rate": 1.9438401194072324e-05, + "loss": 1.0372, + "step": 5687 + }, + { + "epoch": 0.13, + "grad_norm": 2.2726785843878834, + "learning_rate": 1.9438149055902335e-05, + "loss": 1.0566, + "step": 5688 + }, + { + "epoch": 0.13, + "grad_norm": 2.9892080696831242, + "learning_rate": 1.943789686278045e-05, + "loss": 1.0167, + "step": 5689 + }, + { + "epoch": 0.13, + "grad_norm": 1.9270490586595914, + "learning_rate": 1.9437644614708143e-05, + "loss": 1.1634, + "step": 5690 + }, + { + "epoch": 0.13, + "grad_norm": 2.2090835439411145, + "learning_rate": 1.9437392311686884e-05, + "loss": 1.0048, + "step": 5691 + }, + { + "epoch": 0.13, + "grad_norm": 2.077892402488676, + "learning_rate": 1.9437139953718137e-05, + "loss": 1.1196, + "step": 5692 + }, + { + "epoch": 0.13, + "grad_norm": 2.2611858820282857, + "learning_rate": 1.9436887540803372e-05, + "loss": 0.9952, + "step": 5693 + }, + { + "epoch": 0.13, + "grad_norm": 2.120694844793097, + "learning_rate": 1.943663507294406e-05, + "loss": 1.0548, + "step": 5694 + }, + { + "epoch": 0.13, + "grad_norm": 2.6093300974181806, + "learning_rate": 1.9436382550141675e-05, + "loss": 0.9948, + "step": 5695 + }, + { + "epoch": 0.13, + "grad_norm": 2.2272638945763648, + "learning_rate": 1.943612997239768e-05, + "loss": 1.0953, + "step": 5696 + }, + { + "epoch": 0.13, + "grad_norm": 2.096057637829445, + "learning_rate": 1.943587733971355e-05, + "loss": 1.0841, + "step": 5697 + }, + { + "epoch": 0.13, + "grad_norm": 2.0887811495025197, + "learning_rate": 1.9435624652090754e-05, + "loss": 1.0125, + "step": 5698 + }, + { + "epoch": 0.13, + "grad_norm": 2.025112744717702, + "learning_rate": 1.9435371909530766e-05, + "loss": 1.0736, + "step": 5699 + }, + { + "epoch": 0.13, + "grad_norm": 2.6819484439212724, + "learning_rate": 1.9435119112035054e-05, + "loss": 1.1016, + "step": 5700 + }, + { + "epoch": 0.13, + "grad_norm": 1.1240747662324175, + "learning_rate": 1.943486625960509e-05, + "loss": 0.9454, + "step": 5701 + }, + { + "epoch": 0.13, + "grad_norm": 2.3697226181774034, + "learning_rate": 1.943461335224235e-05, + "loss": 1.0504, + "step": 5702 + }, + { + "epoch": 0.13, + "grad_norm": 1.186020536556418, + "learning_rate": 1.9434360389948303e-05, + "loss": 1.0505, + "step": 5703 + }, + { + "epoch": 0.13, + "grad_norm": 2.023355446099007, + "learning_rate": 1.9434107372724424e-05, + "loss": 1.1425, + "step": 5704 + }, + { + "epoch": 0.13, + "grad_norm": 2.029640036335231, + "learning_rate": 1.9433854300572184e-05, + "loss": 0.9342, + "step": 5705 + }, + { + "epoch": 0.13, + "grad_norm": 2.114648959977284, + "learning_rate": 1.943360117349306e-05, + "loss": 1.0233, + "step": 5706 + }, + { + "epoch": 0.13, + "grad_norm": 2.0546090106286763, + "learning_rate": 1.943334799148852e-05, + "loss": 1.1066, + "step": 5707 + }, + { + "epoch": 0.13, + "grad_norm": 2.133706257520652, + "learning_rate": 1.9433094754560042e-05, + "loss": 1.0861, + "step": 5708 + }, + { + "epoch": 0.13, + "grad_norm": 2.0810336267872604, + "learning_rate": 1.94328414627091e-05, + "loss": 1.1069, + "step": 5709 + }, + { + "epoch": 0.13, + "grad_norm": 1.961438848705033, + "learning_rate": 1.943258811593717e-05, + "loss": 1.1106, + "step": 5710 + }, + { + "epoch": 0.13, + "grad_norm": 2.204311976817514, + "learning_rate": 1.9432334714245723e-05, + "loss": 1.1337, + "step": 5711 + }, + { + "epoch": 0.13, + "grad_norm": 1.1393043565373682, + "learning_rate": 1.943208125763624e-05, + "loss": 1.0019, + "step": 5712 + }, + { + "epoch": 0.13, + "grad_norm": 2.2257249844372864, + "learning_rate": 1.943182774611019e-05, + "loss": 1.187, + "step": 5713 + }, + { + "epoch": 0.13, + "grad_norm": 2.6115524014504463, + "learning_rate": 1.943157417966906e-05, + "loss": 0.9286, + "step": 5714 + }, + { + "epoch": 0.13, + "grad_norm": 2.0921254064762733, + "learning_rate": 1.9431320558314313e-05, + "loss": 1.0277, + "step": 5715 + }, + { + "epoch": 0.13, + "grad_norm": 2.109935936916072, + "learning_rate": 1.9431066882047434e-05, + "loss": 1.1048, + "step": 5716 + }, + { + "epoch": 0.13, + "grad_norm": 2.0634291905653663, + "learning_rate": 1.9430813150869896e-05, + "loss": 1.1708, + "step": 5717 + }, + { + "epoch": 0.13, + "grad_norm": 2.4577918532299896, + "learning_rate": 1.943055936478318e-05, + "loss": 0.9883, + "step": 5718 + }, + { + "epoch": 0.13, + "grad_norm": 2.1379012132417654, + "learning_rate": 1.943030552378876e-05, + "loss": 1.1004, + "step": 5719 + }, + { + "epoch": 0.13, + "grad_norm": 2.3151695091410525, + "learning_rate": 1.9430051627888117e-05, + "loss": 0.9174, + "step": 5720 + }, + { + "epoch": 0.13, + "grad_norm": 2.616175359047599, + "learning_rate": 1.942979767708273e-05, + "loss": 1.0593, + "step": 5721 + }, + { + "epoch": 0.13, + "grad_norm": 2.9261323289423062, + "learning_rate": 1.942954367137407e-05, + "loss": 0.9867, + "step": 5722 + }, + { + "epoch": 0.13, + "grad_norm": 2.3338906752161246, + "learning_rate": 1.9429289610763624e-05, + "loss": 1.1014, + "step": 5723 + }, + { + "epoch": 0.13, + "grad_norm": 2.0767723579465995, + "learning_rate": 1.9429035495252872e-05, + "loss": 1.0929, + "step": 5724 + }, + { + "epoch": 0.13, + "grad_norm": 1.1113747594707897, + "learning_rate": 1.9428781324843285e-05, + "loss": 0.9367, + "step": 5725 + }, + { + "epoch": 0.13, + "grad_norm": 2.225029306134192, + "learning_rate": 1.942852709953635e-05, + "loss": 1.0786, + "step": 5726 + }, + { + "epoch": 0.13, + "grad_norm": 2.083374755442618, + "learning_rate": 1.9428272819333545e-05, + "loss": 1.1366, + "step": 5727 + }, + { + "epoch": 0.13, + "grad_norm": 4.747858892597006, + "learning_rate": 1.9428018484236353e-05, + "loss": 1.0856, + "step": 5728 + }, + { + "epoch": 0.13, + "grad_norm": 2.2643077689929703, + "learning_rate": 1.9427764094246248e-05, + "loss": 1.1723, + "step": 5729 + }, + { + "epoch": 0.13, + "grad_norm": 1.1216717513091106, + "learning_rate": 1.942750964936472e-05, + "loss": 1.01, + "step": 5730 + }, + { + "epoch": 0.14, + "grad_norm": 1.1782553401876616, + "learning_rate": 1.9427255149593245e-05, + "loss": 1.0046, + "step": 5731 + }, + { + "epoch": 0.14, + "grad_norm": 1.9967495590155018, + "learning_rate": 1.9427000594933305e-05, + "loss": 0.9966, + "step": 5732 + }, + { + "epoch": 0.14, + "grad_norm": 2.6643248831558664, + "learning_rate": 1.9426745985386385e-05, + "loss": 1.092, + "step": 5733 + }, + { + "epoch": 0.14, + "grad_norm": 2.015033351621663, + "learning_rate": 1.9426491320953963e-05, + "loss": 1.1368, + "step": 5734 + }, + { + "epoch": 0.14, + "grad_norm": 1.153738323813017, + "learning_rate": 1.9426236601637523e-05, + "loss": 1.0011, + "step": 5735 + }, + { + "epoch": 0.14, + "grad_norm": 2.095888606779736, + "learning_rate": 1.9425981827438552e-05, + "loss": 0.9835, + "step": 5736 + }, + { + "epoch": 0.14, + "grad_norm": 2.3300630967869838, + "learning_rate": 1.942572699835853e-05, + "loss": 0.9871, + "step": 5737 + }, + { + "epoch": 0.14, + "grad_norm": 1.9041109805583902, + "learning_rate": 1.942547211439894e-05, + "loss": 1.1482, + "step": 5738 + }, + { + "epoch": 0.14, + "grad_norm": 1.9102114076953916, + "learning_rate": 1.9425217175561268e-05, + "loss": 1.0273, + "step": 5739 + }, + { + "epoch": 0.14, + "grad_norm": 2.2948838031938537, + "learning_rate": 1.9424962181847e-05, + "loss": 1.0344, + "step": 5740 + }, + { + "epoch": 0.14, + "grad_norm": 2.332064677083559, + "learning_rate": 1.9424707133257615e-05, + "loss": 0.9864, + "step": 5741 + }, + { + "epoch": 0.14, + "grad_norm": 2.182780058008886, + "learning_rate": 1.9424452029794603e-05, + "loss": 1.0663, + "step": 5742 + }, + { + "epoch": 0.14, + "grad_norm": 1.9445682218157863, + "learning_rate": 1.9424196871459448e-05, + "loss": 0.9178, + "step": 5743 + }, + { + "epoch": 0.14, + "grad_norm": 1.1508515255179146, + "learning_rate": 1.9423941658253634e-05, + "loss": 0.9786, + "step": 5744 + }, + { + "epoch": 0.14, + "grad_norm": 2.1148824804132476, + "learning_rate": 1.942368639017865e-05, + "loss": 1.0666, + "step": 5745 + }, + { + "epoch": 0.14, + "grad_norm": 2.4492536202751656, + "learning_rate": 1.9423431067235977e-05, + "loss": 0.8827, + "step": 5746 + }, + { + "epoch": 0.14, + "grad_norm": 2.1161560968854825, + "learning_rate": 1.9423175689427105e-05, + "loss": 1.133, + "step": 5747 + }, + { + "epoch": 0.14, + "grad_norm": 2.1485617940884807, + "learning_rate": 1.9422920256753523e-05, + "loss": 0.9501, + "step": 5748 + }, + { + "epoch": 0.14, + "grad_norm": 1.9337357712135412, + "learning_rate": 1.9422664769216716e-05, + "loss": 1.2393, + "step": 5749 + }, + { + "epoch": 0.14, + "grad_norm": 1.9177388991388662, + "learning_rate": 1.9422409226818175e-05, + "loss": 1.2088, + "step": 5750 + }, + { + "epoch": 0.14, + "grad_norm": 1.1413454697743024, + "learning_rate": 1.9422153629559376e-05, + "loss": 0.9976, + "step": 5751 + }, + { + "epoch": 0.14, + "grad_norm": 2.0215273037097834, + "learning_rate": 1.942189797744182e-05, + "loss": 1.0728, + "step": 5752 + }, + { + "epoch": 0.14, + "grad_norm": 1.079714361805834, + "learning_rate": 1.942164227046699e-05, + "loss": 0.9987, + "step": 5753 + }, + { + "epoch": 0.14, + "grad_norm": 1.1375031689649497, + "learning_rate": 1.9421386508636378e-05, + "loss": 0.9784, + "step": 5754 + }, + { + "epoch": 0.14, + "grad_norm": 2.040610035382119, + "learning_rate": 1.9421130691951472e-05, + "loss": 1.1496, + "step": 5755 + }, + { + "epoch": 0.14, + "grad_norm": 2.512754491037985, + "learning_rate": 1.942087482041376e-05, + "loss": 1.0912, + "step": 5756 + }, + { + "epoch": 0.14, + "grad_norm": 1.8832496555026241, + "learning_rate": 1.9420618894024727e-05, + "loss": 1.0294, + "step": 5757 + }, + { + "epoch": 0.14, + "grad_norm": 2.1656574601538425, + "learning_rate": 1.942036291278587e-05, + "loss": 1.1208, + "step": 5758 + }, + { + "epoch": 0.14, + "grad_norm": 2.194599713818943, + "learning_rate": 1.942010687669868e-05, + "loss": 1.0054, + "step": 5759 + }, + { + "epoch": 0.14, + "grad_norm": 2.399800606954885, + "learning_rate": 1.9419850785764645e-05, + "loss": 0.9883, + "step": 5760 + }, + { + "epoch": 0.14, + "grad_norm": 2.2758012514671018, + "learning_rate": 1.9419594639985257e-05, + "loss": 1.1382, + "step": 5761 + }, + { + "epoch": 0.14, + "grad_norm": 2.2230123356592806, + "learning_rate": 1.9419338439362007e-05, + "loss": 1.1298, + "step": 5762 + }, + { + "epoch": 0.14, + "grad_norm": 2.0153453085552058, + "learning_rate": 1.9419082183896385e-05, + "loss": 1.031, + "step": 5763 + }, + { + "epoch": 0.14, + "grad_norm": 1.1183618535095288, + "learning_rate": 1.9418825873589883e-05, + "loss": 0.9388, + "step": 5764 + }, + { + "epoch": 0.14, + "grad_norm": 2.1639658794670016, + "learning_rate": 1.9418569508443998e-05, + "loss": 1.0665, + "step": 5765 + }, + { + "epoch": 0.14, + "grad_norm": 1.9426580275336687, + "learning_rate": 1.941831308846022e-05, + "loss": 1.0745, + "step": 5766 + }, + { + "epoch": 0.14, + "grad_norm": 2.0719854208562074, + "learning_rate": 1.9418056613640038e-05, + "loss": 1.0826, + "step": 5767 + }, + { + "epoch": 0.14, + "grad_norm": 2.7773020521721046, + "learning_rate": 1.941780008398495e-05, + "loss": 1.082, + "step": 5768 + }, + { + "epoch": 0.14, + "grad_norm": 2.4196082578523885, + "learning_rate": 1.941754349949645e-05, + "loss": 1.1064, + "step": 5769 + }, + { + "epoch": 0.14, + "grad_norm": 2.34819810436942, + "learning_rate": 1.941728686017603e-05, + "loss": 1.0461, + "step": 5770 + }, + { + "epoch": 0.14, + "grad_norm": 2.058583047715042, + "learning_rate": 1.9417030166025182e-05, + "loss": 1.0238, + "step": 5771 + }, + { + "epoch": 0.14, + "grad_norm": 1.2198592019174381, + "learning_rate": 1.9416773417045406e-05, + "loss": 1.0207, + "step": 5772 + }, + { + "epoch": 0.14, + "grad_norm": 2.092516390131231, + "learning_rate": 1.9416516613238193e-05, + "loss": 1.1777, + "step": 5773 + }, + { + "epoch": 0.14, + "grad_norm": 2.34962150518278, + "learning_rate": 1.9416259754605034e-05, + "loss": 1.0428, + "step": 5774 + }, + { + "epoch": 0.14, + "grad_norm": 2.0108482050123233, + "learning_rate": 1.9416002841147438e-05, + "loss": 1.0538, + "step": 5775 + }, + { + "epoch": 0.14, + "grad_norm": 2.0978147621752044, + "learning_rate": 1.9415745872866887e-05, + "loss": 1.0534, + "step": 5776 + }, + { + "epoch": 0.14, + "grad_norm": 2.142344207195877, + "learning_rate": 1.9415488849764884e-05, + "loss": 1.0332, + "step": 5777 + }, + { + "epoch": 0.14, + "grad_norm": 2.5494095270742947, + "learning_rate": 1.9415231771842925e-05, + "loss": 1.1464, + "step": 5778 + }, + { + "epoch": 0.14, + "grad_norm": 2.395013572090505, + "learning_rate": 1.9414974639102505e-05, + "loss": 1.0707, + "step": 5779 + }, + { + "epoch": 0.14, + "grad_norm": 2.008306654349508, + "learning_rate": 1.941471745154512e-05, + "loss": 1.1642, + "step": 5780 + }, + { + "epoch": 0.14, + "grad_norm": 2.035583309996999, + "learning_rate": 1.941446020917227e-05, + "loss": 1.0879, + "step": 5781 + }, + { + "epoch": 0.14, + "grad_norm": 1.8826081022288126, + "learning_rate": 1.9414202911985453e-05, + "loss": 1.0001, + "step": 5782 + }, + { + "epoch": 0.14, + "grad_norm": 2.9692740994949554, + "learning_rate": 1.9413945559986164e-05, + "loss": 0.9865, + "step": 5783 + }, + { + "epoch": 0.14, + "grad_norm": 2.15892926974119, + "learning_rate": 1.9413688153175907e-05, + "loss": 1.0623, + "step": 5784 + }, + { + "epoch": 0.14, + "grad_norm": 1.9798200370618495, + "learning_rate": 1.9413430691556175e-05, + "loss": 1.1275, + "step": 5785 + }, + { + "epoch": 0.14, + "grad_norm": 1.9996856061899184, + "learning_rate": 1.9413173175128472e-05, + "loss": 0.9736, + "step": 5786 + }, + { + "epoch": 0.14, + "grad_norm": 2.348956144326403, + "learning_rate": 1.9412915603894294e-05, + "loss": 1.0212, + "step": 5787 + }, + { + "epoch": 0.14, + "grad_norm": 2.134276921785619, + "learning_rate": 1.941265797785514e-05, + "loss": 1.0367, + "step": 5788 + }, + { + "epoch": 0.14, + "grad_norm": 2.2646693153249697, + "learning_rate": 1.941240029701251e-05, + "loss": 1.0248, + "step": 5789 + }, + { + "epoch": 0.14, + "grad_norm": 1.9654881242404385, + "learning_rate": 1.9412142561367906e-05, + "loss": 1.0842, + "step": 5790 + }, + { + "epoch": 0.14, + "grad_norm": 2.1117998503989712, + "learning_rate": 1.941188477092283e-05, + "loss": 1.0632, + "step": 5791 + }, + { + "epoch": 0.14, + "grad_norm": 2.3893482432227127, + "learning_rate": 1.941162692567878e-05, + "loss": 0.9181, + "step": 5792 + }, + { + "epoch": 0.14, + "grad_norm": 3.2062398142181117, + "learning_rate": 1.941136902563726e-05, + "loss": 1.0706, + "step": 5793 + }, + { + "epoch": 0.14, + "grad_norm": 2.005958185691904, + "learning_rate": 1.941111107079977e-05, + "loss": 1.0876, + "step": 5794 + }, + { + "epoch": 0.14, + "grad_norm": 2.1927794821915727, + "learning_rate": 1.9410853061167806e-05, + "loss": 1.1434, + "step": 5795 + }, + { + "epoch": 0.14, + "grad_norm": 2.249647951565368, + "learning_rate": 1.9410594996742884e-05, + "loss": 1.2291, + "step": 5796 + }, + { + "epoch": 0.14, + "grad_norm": 2.1827099933991083, + "learning_rate": 1.9410336877526494e-05, + "loss": 1.1131, + "step": 5797 + }, + { + "epoch": 0.14, + "grad_norm": 2.0441435792013376, + "learning_rate": 1.9410078703520147e-05, + "loss": 1.176, + "step": 5798 + }, + { + "epoch": 0.14, + "grad_norm": 1.9855643732752808, + "learning_rate": 1.9409820474725338e-05, + "loss": 1.0503, + "step": 5799 + }, + { + "epoch": 0.14, + "grad_norm": 2.132389351177835, + "learning_rate": 1.9409562191143577e-05, + "loss": 1.035, + "step": 5800 + }, + { + "epoch": 0.14, + "grad_norm": 2.0545017570722783, + "learning_rate": 1.9409303852776365e-05, + "loss": 1.0354, + "step": 5801 + }, + { + "epoch": 0.14, + "grad_norm": 3.087326218621124, + "learning_rate": 1.940904545962521e-05, + "loss": 1.0535, + "step": 5802 + }, + { + "epoch": 0.14, + "grad_norm": 2.1670481650741085, + "learning_rate": 1.940878701169161e-05, + "loss": 0.8996, + "step": 5803 + }, + { + "epoch": 0.14, + "grad_norm": 1.1624203348053443, + "learning_rate": 1.9408528508977076e-05, + "loss": 0.9439, + "step": 5804 + }, + { + "epoch": 0.14, + "grad_norm": 2.2383167509938495, + "learning_rate": 1.940826995148311e-05, + "loss": 1.0157, + "step": 5805 + }, + { + "epoch": 0.14, + "grad_norm": 1.9042314546914179, + "learning_rate": 1.9408011339211215e-05, + "loss": 1.1446, + "step": 5806 + }, + { + "epoch": 0.14, + "grad_norm": 2.412989134817116, + "learning_rate": 1.94077526721629e-05, + "loss": 0.9802, + "step": 5807 + }, + { + "epoch": 0.14, + "grad_norm": 1.1247131500183691, + "learning_rate": 1.940749395033967e-05, + "loss": 0.946, + "step": 5808 + }, + { + "epoch": 0.14, + "grad_norm": 2.042812719367694, + "learning_rate": 1.9407235173743038e-05, + "loss": 1.1149, + "step": 5809 + }, + { + "epoch": 0.14, + "grad_norm": 1.9031759994437294, + "learning_rate": 1.94069763423745e-05, + "loss": 1.2174, + "step": 5810 + }, + { + "epoch": 0.14, + "grad_norm": 2.3431312921588185, + "learning_rate": 1.9406717456235566e-05, + "loss": 1.0814, + "step": 5811 + }, + { + "epoch": 0.14, + "grad_norm": 1.9292041351991425, + "learning_rate": 1.9406458515327747e-05, + "loss": 1.1625, + "step": 5812 + }, + { + "epoch": 0.14, + "grad_norm": 2.2214524716735458, + "learning_rate": 1.940619951965255e-05, + "loss": 1.0804, + "step": 5813 + }, + { + "epoch": 0.14, + "grad_norm": 2.0906636204104956, + "learning_rate": 1.9405940469211478e-05, + "loss": 0.9939, + "step": 5814 + }, + { + "epoch": 0.14, + "grad_norm": 1.803928901259412, + "learning_rate": 1.9405681364006047e-05, + "loss": 1.0402, + "step": 5815 + }, + { + "epoch": 0.14, + "grad_norm": 1.150980885000898, + "learning_rate": 1.9405422204037758e-05, + "loss": 1.0114, + "step": 5816 + }, + { + "epoch": 0.14, + "grad_norm": 2.2565996222941758, + "learning_rate": 1.940516298930812e-05, + "loss": 1.0793, + "step": 5817 + }, + { + "epoch": 0.14, + "grad_norm": 2.0591378259822513, + "learning_rate": 1.940490371981865e-05, + "loss": 1.0856, + "step": 5818 + }, + { + "epoch": 0.14, + "grad_norm": 1.9959948341660168, + "learning_rate": 1.9404644395570852e-05, + "loss": 1.1248, + "step": 5819 + }, + { + "epoch": 0.14, + "grad_norm": 1.1835439745505265, + "learning_rate": 1.940438501656624e-05, + "loss": 0.9774, + "step": 5820 + }, + { + "epoch": 0.14, + "grad_norm": 2.1932736227803855, + "learning_rate": 1.9404125582806316e-05, + "loss": 1.146, + "step": 5821 + }, + { + "epoch": 0.14, + "grad_norm": 2.204114148472052, + "learning_rate": 1.9403866094292598e-05, + "loss": 1.0102, + "step": 5822 + }, + { + "epoch": 0.14, + "grad_norm": 2.0844008448337483, + "learning_rate": 1.9403606551026596e-05, + "loss": 1.0537, + "step": 5823 + }, + { + "epoch": 0.14, + "grad_norm": 2.5210231748704195, + "learning_rate": 1.9403346953009817e-05, + "loss": 1.0473, + "step": 5824 + }, + { + "epoch": 0.14, + "grad_norm": 2.1044795912168714, + "learning_rate": 1.9403087300243776e-05, + "loss": 1.0297, + "step": 5825 + }, + { + "epoch": 0.14, + "grad_norm": 1.9501237780928382, + "learning_rate": 1.9402827592729984e-05, + "loss": 0.991, + "step": 5826 + }, + { + "epoch": 0.14, + "grad_norm": 2.1090349434336337, + "learning_rate": 1.9402567830469952e-05, + "loss": 1.1037, + "step": 5827 + }, + { + "epoch": 0.14, + "grad_norm": 2.2154939535349403, + "learning_rate": 1.9402308013465192e-05, + "loss": 1.0825, + "step": 5828 + }, + { + "epoch": 0.14, + "grad_norm": 2.0784416960163217, + "learning_rate": 1.9402048141717217e-05, + "loss": 1.0914, + "step": 5829 + }, + { + "epoch": 0.14, + "grad_norm": 2.0753646272139914, + "learning_rate": 1.9401788215227547e-05, + "loss": 1.1537, + "step": 5830 + }, + { + "epoch": 0.14, + "grad_norm": 2.241706543604891, + "learning_rate": 1.9401528233997684e-05, + "loss": 0.9993, + "step": 5831 + }, + { + "epoch": 0.14, + "grad_norm": 2.147979811987957, + "learning_rate": 1.940126819802915e-05, + "loss": 1.0685, + "step": 5832 + }, + { + "epoch": 0.14, + "grad_norm": 2.060507775676987, + "learning_rate": 1.9401008107323457e-05, + "loss": 1.1428, + "step": 5833 + }, + { + "epoch": 0.14, + "grad_norm": 1.8984526313023704, + "learning_rate": 1.9400747961882115e-05, + "loss": 0.9226, + "step": 5834 + }, + { + "epoch": 0.14, + "grad_norm": 2.034311854603247, + "learning_rate": 1.9400487761706646e-05, + "loss": 1.1126, + "step": 5835 + }, + { + "epoch": 0.14, + "grad_norm": 2.073609708890132, + "learning_rate": 1.9400227506798557e-05, + "loss": 1.0726, + "step": 5836 + }, + { + "epoch": 0.14, + "grad_norm": 2.2359550106514523, + "learning_rate": 1.939996719715937e-05, + "loss": 1.0083, + "step": 5837 + }, + { + "epoch": 0.14, + "grad_norm": 2.1457325760550985, + "learning_rate": 1.93997068327906e-05, + "loss": 1.1268, + "step": 5838 + }, + { + "epoch": 0.14, + "grad_norm": 1.9020208405005015, + "learning_rate": 1.9399446413693756e-05, + "loss": 1.0157, + "step": 5839 + }, + { + "epoch": 0.14, + "grad_norm": 1.9827247125236929, + "learning_rate": 1.939918593987036e-05, + "loss": 0.9873, + "step": 5840 + }, + { + "epoch": 0.14, + "grad_norm": 2.2118291198555062, + "learning_rate": 1.939892541132193e-05, + "loss": 1.0933, + "step": 5841 + }, + { + "epoch": 0.14, + "grad_norm": 2.4332857023098704, + "learning_rate": 1.9398664828049983e-05, + "loss": 1.0105, + "step": 5842 + }, + { + "epoch": 0.14, + "grad_norm": 2.2859168243956165, + "learning_rate": 1.939840419005603e-05, + "loss": 1.1318, + "step": 5843 + }, + { + "epoch": 0.14, + "grad_norm": 2.250727666843008, + "learning_rate": 1.9398143497341593e-05, + "loss": 1.1567, + "step": 5844 + }, + { + "epoch": 0.14, + "grad_norm": 2.1005428340600374, + "learning_rate": 1.939788274990819e-05, + "loss": 1.0225, + "step": 5845 + }, + { + "epoch": 0.14, + "grad_norm": 2.1306123385448497, + "learning_rate": 1.9397621947757338e-05, + "loss": 1.0557, + "step": 5846 + }, + { + "epoch": 0.14, + "grad_norm": 2.297279463403461, + "learning_rate": 1.9397361090890554e-05, + "loss": 1.1215, + "step": 5847 + }, + { + "epoch": 0.14, + "grad_norm": 2.391006660781547, + "learning_rate": 1.939710017930936e-05, + "loss": 1.0802, + "step": 5848 + }, + { + "epoch": 0.14, + "grad_norm": 2.4479791063751746, + "learning_rate": 1.939683921301527e-05, + "loss": 0.9775, + "step": 5849 + }, + { + "epoch": 0.14, + "grad_norm": 2.1087679025760173, + "learning_rate": 1.9396578192009812e-05, + "loss": 1.1066, + "step": 5850 + }, + { + "epoch": 0.14, + "grad_norm": 2.1923284381589077, + "learning_rate": 1.9396317116294502e-05, + "loss": 1.1667, + "step": 5851 + }, + { + "epoch": 0.14, + "grad_norm": 2.4421609723835656, + "learning_rate": 1.9396055985870855e-05, + "loss": 0.9844, + "step": 5852 + }, + { + "epoch": 0.14, + "grad_norm": 1.9813081446858536, + "learning_rate": 1.93957948007404e-05, + "loss": 1.0721, + "step": 5853 + }, + { + "epoch": 0.14, + "grad_norm": 2.1478064971564095, + "learning_rate": 1.939553356090465e-05, + "loss": 1.2458, + "step": 5854 + }, + { + "epoch": 0.14, + "grad_norm": 2.0754935086243305, + "learning_rate": 1.9395272266365127e-05, + "loss": 1.1681, + "step": 5855 + }, + { + "epoch": 0.14, + "grad_norm": 2.2604239151200454, + "learning_rate": 1.9395010917123355e-05, + "loss": 1.0058, + "step": 5856 + }, + { + "epoch": 0.14, + "grad_norm": 2.4829111012969944, + "learning_rate": 1.9394749513180857e-05, + "loss": 1.1198, + "step": 5857 + }, + { + "epoch": 0.14, + "grad_norm": 2.4160499097628034, + "learning_rate": 1.9394488054539153e-05, + "loss": 1.042, + "step": 5858 + }, + { + "epoch": 0.14, + "grad_norm": 1.8320392423426957, + "learning_rate": 1.9394226541199764e-05, + "loss": 1.0748, + "step": 5859 + }, + { + "epoch": 0.14, + "grad_norm": 2.415924862478601, + "learning_rate": 1.9393964973164216e-05, + "loss": 0.9994, + "step": 5860 + }, + { + "epoch": 0.14, + "grad_norm": 1.9050887378579264, + "learning_rate": 1.939370335043403e-05, + "loss": 1.0909, + "step": 5861 + }, + { + "epoch": 0.14, + "grad_norm": 2.565503837510725, + "learning_rate": 1.9393441673010725e-05, + "loss": 1.212, + "step": 5862 + }, + { + "epoch": 0.14, + "grad_norm": 3.611354930711407, + "learning_rate": 1.939317994089583e-05, + "loss": 1.0632, + "step": 5863 + }, + { + "epoch": 0.14, + "grad_norm": 2.077222561152602, + "learning_rate": 1.939291815409087e-05, + "loss": 1.0317, + "step": 5864 + }, + { + "epoch": 0.14, + "grad_norm": 2.01822200157319, + "learning_rate": 1.9392656312597365e-05, + "loss": 0.915, + "step": 5865 + }, + { + "epoch": 0.14, + "grad_norm": 2.1874296889117355, + "learning_rate": 1.939239441641684e-05, + "loss": 1.018, + "step": 5866 + }, + { + "epoch": 0.14, + "grad_norm": 2.0760596428687, + "learning_rate": 1.9392132465550824e-05, + "loss": 0.9991, + "step": 5867 + }, + { + "epoch": 0.14, + "grad_norm": 2.004452573862813, + "learning_rate": 1.939187046000084e-05, + "loss": 0.9323, + "step": 5868 + }, + { + "epoch": 0.14, + "grad_norm": 2.2201016864753957, + "learning_rate": 1.939160839976841e-05, + "loss": 1.0772, + "step": 5869 + }, + { + "epoch": 0.14, + "grad_norm": 2.155738533647049, + "learning_rate": 1.939134628485506e-05, + "loss": 1.0341, + "step": 5870 + }, + { + "epoch": 0.14, + "grad_norm": 1.153195614255025, + "learning_rate": 1.9391084115262323e-05, + "loss": 0.9543, + "step": 5871 + }, + { + "epoch": 0.14, + "grad_norm": 3.022529177057142, + "learning_rate": 1.9390821890991723e-05, + "loss": 1.0853, + "step": 5872 + }, + { + "epoch": 0.14, + "grad_norm": 2.218146414576147, + "learning_rate": 1.9390559612044783e-05, + "loss": 1.0865, + "step": 5873 + }, + { + "epoch": 0.14, + "grad_norm": 2.0917264493962056, + "learning_rate": 1.9390297278423028e-05, + "loss": 1.2092, + "step": 5874 + }, + { + "epoch": 0.14, + "grad_norm": 2.0782078156426027, + "learning_rate": 1.9390034890127993e-05, + "loss": 1.0894, + "step": 5875 + }, + { + "epoch": 0.14, + "grad_norm": 2.631625635595045, + "learning_rate": 1.93897724471612e-05, + "loss": 1.0452, + "step": 5876 + }, + { + "epoch": 0.14, + "grad_norm": 2.3315109358079966, + "learning_rate": 1.938950994952418e-05, + "loss": 1.0633, + "step": 5877 + }, + { + "epoch": 0.14, + "grad_norm": 2.076363584068386, + "learning_rate": 1.938924739721846e-05, + "loss": 1.0878, + "step": 5878 + }, + { + "epoch": 0.14, + "grad_norm": 2.081621050516686, + "learning_rate": 1.9388984790245575e-05, + "loss": 1.1478, + "step": 5879 + }, + { + "epoch": 0.14, + "grad_norm": 2.217671268218665, + "learning_rate": 1.938872212860704e-05, + "loss": 1.1515, + "step": 5880 + }, + { + "epoch": 0.14, + "grad_norm": 2.757736398593141, + "learning_rate": 1.9388459412304397e-05, + "loss": 1.0255, + "step": 5881 + }, + { + "epoch": 0.14, + "grad_norm": 1.164715535527508, + "learning_rate": 1.938819664133917e-05, + "loss": 1.0035, + "step": 5882 + }, + { + "epoch": 0.14, + "grad_norm": 2.1639232265761046, + "learning_rate": 1.938793381571289e-05, + "loss": 0.9112, + "step": 5883 + }, + { + "epoch": 0.14, + "grad_norm": 2.1611551888243037, + "learning_rate": 1.938767093542708e-05, + "loss": 1.1008, + "step": 5884 + }, + { + "epoch": 0.14, + "grad_norm": 2.31022238281758, + "learning_rate": 1.938740800048329e-05, + "loss": 1.1045, + "step": 5885 + }, + { + "epoch": 0.14, + "grad_norm": 2.18384092660163, + "learning_rate": 1.938714501088303e-05, + "loss": 0.9539, + "step": 5886 + }, + { + "epoch": 0.14, + "grad_norm": 2.665841837518076, + "learning_rate": 1.9386881966627844e-05, + "loss": 1.1046, + "step": 5887 + }, + { + "epoch": 0.14, + "grad_norm": 2.3227347495865036, + "learning_rate": 1.9386618867719258e-05, + "loss": 1.3264, + "step": 5888 + }, + { + "epoch": 0.14, + "grad_norm": 1.9894162028603644, + "learning_rate": 1.9386355714158806e-05, + "loss": 1.1187, + "step": 5889 + }, + { + "epoch": 0.14, + "grad_norm": 2.3453128238465863, + "learning_rate": 1.9386092505948018e-05, + "loss": 0.978, + "step": 5890 + }, + { + "epoch": 0.14, + "grad_norm": 2.1774650274054443, + "learning_rate": 1.9385829243088426e-05, + "loss": 0.996, + "step": 5891 + }, + { + "epoch": 0.14, + "grad_norm": 1.9746121902709872, + "learning_rate": 1.938556592558157e-05, + "loss": 0.9464, + "step": 5892 + }, + { + "epoch": 0.14, + "grad_norm": 2.2210810329488813, + "learning_rate": 1.9385302553428974e-05, + "loss": 1.0286, + "step": 5893 + }, + { + "epoch": 0.14, + "grad_norm": 3.524047813176357, + "learning_rate": 1.9385039126632176e-05, + "loss": 1.1165, + "step": 5894 + }, + { + "epoch": 0.14, + "grad_norm": 2.1199108163468914, + "learning_rate": 1.9384775645192707e-05, + "loss": 0.9843, + "step": 5895 + }, + { + "epoch": 0.14, + "grad_norm": 2.167963693694499, + "learning_rate": 1.9384512109112104e-05, + "loss": 0.9633, + "step": 5896 + }, + { + "epoch": 0.14, + "grad_norm": 1.133258092323606, + "learning_rate": 1.9384248518391903e-05, + "loss": 0.9668, + "step": 5897 + }, + { + "epoch": 0.14, + "grad_norm": 1.9929180890623632, + "learning_rate": 1.9383984873033636e-05, + "loss": 0.9861, + "step": 5898 + }, + { + "epoch": 0.14, + "grad_norm": 2.1961586073541017, + "learning_rate": 1.9383721173038836e-05, + "loss": 1.1223, + "step": 5899 + }, + { + "epoch": 0.14, + "grad_norm": 2.2561876988007126, + "learning_rate": 1.938345741840904e-05, + "loss": 1.0397, + "step": 5900 + }, + { + "epoch": 0.14, + "grad_norm": 2.5370733945547648, + "learning_rate": 1.9383193609145786e-05, + "loss": 1.2338, + "step": 5901 + }, + { + "epoch": 0.14, + "grad_norm": 1.1985792681360017, + "learning_rate": 1.9382929745250605e-05, + "loss": 0.908, + "step": 5902 + }, + { + "epoch": 0.14, + "grad_norm": 2.5885531586845105, + "learning_rate": 1.938266582672504e-05, + "loss": 1.164, + "step": 5903 + }, + { + "epoch": 0.14, + "grad_norm": 2.0481624363899744, + "learning_rate": 1.9382401853570624e-05, + "loss": 1.0202, + "step": 5904 + }, + { + "epoch": 0.14, + "grad_norm": 1.906616621617618, + "learning_rate": 1.938213782578889e-05, + "loss": 1.1645, + "step": 5905 + }, + { + "epoch": 0.14, + "grad_norm": 2.008993674911678, + "learning_rate": 1.9381873743381384e-05, + "loss": 1.1925, + "step": 5906 + }, + { + "epoch": 0.14, + "grad_norm": 2.056857845571698, + "learning_rate": 1.9381609606349635e-05, + "loss": 1.1603, + "step": 5907 + }, + { + "epoch": 0.14, + "grad_norm": 2.122751268680996, + "learning_rate": 1.9381345414695188e-05, + "loss": 1.0948, + "step": 5908 + }, + { + "epoch": 0.14, + "grad_norm": 2.5389423277733334, + "learning_rate": 1.9381081168419577e-05, + "loss": 0.8593, + "step": 5909 + }, + { + "epoch": 0.14, + "grad_norm": 2.1400142864644556, + "learning_rate": 1.938081686752434e-05, + "loss": 1.1991, + "step": 5910 + }, + { + "epoch": 0.14, + "grad_norm": 2.336769319109554, + "learning_rate": 1.938055251201102e-05, + "loss": 1.0602, + "step": 5911 + }, + { + "epoch": 0.14, + "grad_norm": 1.1898974996362695, + "learning_rate": 1.9380288101881152e-05, + "loss": 1.0286, + "step": 5912 + }, + { + "epoch": 0.14, + "grad_norm": 1.0854629923495303, + "learning_rate": 1.9380023637136272e-05, + "loss": 0.9894, + "step": 5913 + }, + { + "epoch": 0.14, + "grad_norm": 1.1301266669608623, + "learning_rate": 1.937975911777793e-05, + "loss": 0.9388, + "step": 5914 + }, + { + "epoch": 0.14, + "grad_norm": 2.028671284675806, + "learning_rate": 1.937949454380766e-05, + "loss": 1.0403, + "step": 5915 + }, + { + "epoch": 0.14, + "grad_norm": 2.2868624614178708, + "learning_rate": 1.9379229915227006e-05, + "loss": 1.0634, + "step": 5916 + }, + { + "epoch": 0.14, + "grad_norm": 2.164419046700642, + "learning_rate": 1.9378965232037503e-05, + "loss": 1.0021, + "step": 5917 + }, + { + "epoch": 0.14, + "grad_norm": 3.4933791590259107, + "learning_rate": 1.9378700494240697e-05, + "loss": 1.0925, + "step": 5918 + }, + { + "epoch": 0.14, + "grad_norm": 2.105688782042554, + "learning_rate": 1.9378435701838127e-05, + "loss": 1.0467, + "step": 5919 + }, + { + "epoch": 0.14, + "grad_norm": 1.9899449478432778, + "learning_rate": 1.937817085483133e-05, + "loss": 1.0357, + "step": 5920 + }, + { + "epoch": 0.14, + "grad_norm": 2.176965593413195, + "learning_rate": 1.937790595322186e-05, + "loss": 0.994, + "step": 5921 + }, + { + "epoch": 0.14, + "grad_norm": 1.1694136214691087, + "learning_rate": 1.937764099701125e-05, + "loss": 1.0406, + "step": 5922 + }, + { + "epoch": 0.14, + "grad_norm": 2.2804900754661355, + "learning_rate": 1.9377375986201047e-05, + "loss": 1.0639, + "step": 5923 + }, + { + "epoch": 0.14, + "grad_norm": 1.826101636237056, + "learning_rate": 1.9377110920792788e-05, + "loss": 0.9018, + "step": 5924 + }, + { + "epoch": 0.14, + "grad_norm": 2.022595141064371, + "learning_rate": 1.937684580078802e-05, + "loss": 0.9462, + "step": 5925 + }, + { + "epoch": 0.14, + "grad_norm": 2.216034216704466, + "learning_rate": 1.9376580626188292e-05, + "loss": 1.0173, + "step": 5926 + }, + { + "epoch": 0.14, + "grad_norm": 2.097388523370154, + "learning_rate": 1.937631539699514e-05, + "loss": 0.9273, + "step": 5927 + }, + { + "epoch": 0.14, + "grad_norm": 2.173866693415043, + "learning_rate": 1.9376050113210113e-05, + "loss": 0.9263, + "step": 5928 + }, + { + "epoch": 0.14, + "grad_norm": 1.1618427770215345, + "learning_rate": 1.9375784774834754e-05, + "loss": 1.0318, + "step": 5929 + }, + { + "epoch": 0.14, + "grad_norm": 2.195548356697118, + "learning_rate": 1.9375519381870608e-05, + "loss": 1.0956, + "step": 5930 + }, + { + "epoch": 0.14, + "grad_norm": 2.0807986341361686, + "learning_rate": 1.9375253934319217e-05, + "loss": 0.9574, + "step": 5931 + }, + { + "epoch": 0.14, + "grad_norm": 2.4039238827132774, + "learning_rate": 1.9374988432182133e-05, + "loss": 1.0353, + "step": 5932 + }, + { + "epoch": 0.14, + "grad_norm": 2.1780737304571196, + "learning_rate": 1.9374722875460894e-05, + "loss": 0.9437, + "step": 5933 + }, + { + "epoch": 0.14, + "grad_norm": 2.042789485439622, + "learning_rate": 1.9374457264157055e-05, + "loss": 1.0945, + "step": 5934 + }, + { + "epoch": 0.14, + "grad_norm": 1.1573594451614893, + "learning_rate": 1.9374191598272154e-05, + "loss": 1.0016, + "step": 5935 + }, + { + "epoch": 0.14, + "grad_norm": 2.8244265053537454, + "learning_rate": 1.9373925877807744e-05, + "loss": 1.0043, + "step": 5936 + }, + { + "epoch": 0.14, + "grad_norm": 1.1851228125641868, + "learning_rate": 1.937366010276537e-05, + "loss": 0.9914, + "step": 5937 + }, + { + "epoch": 0.14, + "grad_norm": 1.1337005204243407, + "learning_rate": 1.9373394273146577e-05, + "loss": 0.9884, + "step": 5938 + }, + { + "epoch": 0.14, + "grad_norm": 1.9527887236069033, + "learning_rate": 1.9373128388952914e-05, + "loss": 1.0841, + "step": 5939 + }, + { + "epoch": 0.14, + "grad_norm": 2.4422197485769663, + "learning_rate": 1.9372862450185934e-05, + "loss": 1.0536, + "step": 5940 + }, + { + "epoch": 0.14, + "grad_norm": 2.292466108949515, + "learning_rate": 1.937259645684718e-05, + "loss": 1.1043, + "step": 5941 + }, + { + "epoch": 0.14, + "grad_norm": 2.7734574381017616, + "learning_rate": 1.9372330408938203e-05, + "loss": 1.048, + "step": 5942 + }, + { + "epoch": 0.14, + "grad_norm": 2.4727314210489126, + "learning_rate": 1.937206430646055e-05, + "loss": 1.0287, + "step": 5943 + }, + { + "epoch": 0.14, + "grad_norm": 1.1507242816671608, + "learning_rate": 1.937179814941577e-05, + "loss": 1.0717, + "step": 5944 + }, + { + "epoch": 0.14, + "grad_norm": 2.032529978875807, + "learning_rate": 1.9371531937805417e-05, + "loss": 1.0696, + "step": 5945 + }, + { + "epoch": 0.14, + "grad_norm": 1.2297312172379655, + "learning_rate": 1.9371265671631038e-05, + "loss": 1.0132, + "step": 5946 + }, + { + "epoch": 0.14, + "grad_norm": 2.0396307998505754, + "learning_rate": 1.9370999350894178e-05, + "loss": 1.1447, + "step": 5947 + }, + { + "epoch": 0.14, + "grad_norm": 2.967824949273763, + "learning_rate": 1.93707329755964e-05, + "loss": 1.0938, + "step": 5948 + }, + { + "epoch": 0.14, + "grad_norm": 1.8932518960413913, + "learning_rate": 1.9370466545739246e-05, + "loss": 1.003, + "step": 5949 + }, + { + "epoch": 0.14, + "grad_norm": 2.038363885593442, + "learning_rate": 1.9370200061324267e-05, + "loss": 1.17, + "step": 5950 + }, + { + "epoch": 0.14, + "grad_norm": 1.9863544884810438, + "learning_rate": 1.9369933522353017e-05, + "loss": 1.0214, + "step": 5951 + }, + { + "epoch": 0.14, + "grad_norm": 2.353325791142428, + "learning_rate": 1.936966692882705e-05, + "loss": 1.2475, + "step": 5952 + }, + { + "epoch": 0.14, + "grad_norm": 2.087187273088791, + "learning_rate": 1.9369400280747913e-05, + "loss": 1.0406, + "step": 5953 + }, + { + "epoch": 0.14, + "grad_norm": 2.3641250567958734, + "learning_rate": 1.9369133578117162e-05, + "loss": 0.9665, + "step": 5954 + }, + { + "epoch": 0.14, + "grad_norm": 2.199229506865225, + "learning_rate": 1.936886682093635e-05, + "loss": 1.0906, + "step": 5955 + }, + { + "epoch": 0.14, + "grad_norm": 1.101623819319351, + "learning_rate": 1.9368600009207027e-05, + "loss": 0.9597, + "step": 5956 + }, + { + "epoch": 0.14, + "grad_norm": 2.1599498267732673, + "learning_rate": 1.9368333142930754e-05, + "loss": 1.0589, + "step": 5957 + }, + { + "epoch": 0.14, + "grad_norm": 2.020879578301927, + "learning_rate": 1.9368066222109073e-05, + "loss": 1.0229, + "step": 5958 + }, + { + "epoch": 0.14, + "grad_norm": 3.963731866229171, + "learning_rate": 1.936779924674355e-05, + "loss": 0.8483, + "step": 5959 + }, + { + "epoch": 0.14, + "grad_norm": 2.83692782360948, + "learning_rate": 1.936753221683573e-05, + "loss": 1.0221, + "step": 5960 + }, + { + "epoch": 0.14, + "grad_norm": 2.0406448211088826, + "learning_rate": 1.936726513238717e-05, + "loss": 1.0158, + "step": 5961 + }, + { + "epoch": 0.14, + "grad_norm": 1.150134235131421, + "learning_rate": 1.9366997993399432e-05, + "loss": 1.0391, + "step": 5962 + }, + { + "epoch": 0.14, + "grad_norm": 1.1220192600900363, + "learning_rate": 1.9366730799874063e-05, + "loss": 0.9908, + "step": 5963 + }, + { + "epoch": 0.14, + "grad_norm": 3.112761397856137, + "learning_rate": 1.9366463551812624e-05, + "loss": 1.1913, + "step": 5964 + }, + { + "epoch": 0.14, + "grad_norm": 2.12784355482941, + "learning_rate": 1.936619624921667e-05, + "loss": 1.0358, + "step": 5965 + }, + { + "epoch": 0.14, + "grad_norm": 2.3343027940120855, + "learning_rate": 1.936592889208775e-05, + "loss": 0.9254, + "step": 5966 + }, + { + "epoch": 0.14, + "grad_norm": 1.7651275106904805, + "learning_rate": 1.936566148042743e-05, + "loss": 1.0563, + "step": 5967 + }, + { + "epoch": 0.14, + "grad_norm": 1.9841772085554152, + "learning_rate": 1.9365394014237268e-05, + "loss": 0.9912, + "step": 5968 + }, + { + "epoch": 0.14, + "grad_norm": 2.4766958688525906, + "learning_rate": 1.9365126493518813e-05, + "loss": 0.9265, + "step": 5969 + }, + { + "epoch": 0.14, + "grad_norm": 2.4108816155439037, + "learning_rate": 1.9364858918273626e-05, + "loss": 1.1576, + "step": 5970 + }, + { + "epoch": 0.14, + "grad_norm": 2.3339042395437635, + "learning_rate": 1.9364591288503267e-05, + "loss": 1.0581, + "step": 5971 + }, + { + "epoch": 0.14, + "grad_norm": 2.029202323882559, + "learning_rate": 1.9364323604209292e-05, + "loss": 0.976, + "step": 5972 + }, + { + "epoch": 0.14, + "grad_norm": 2.34949743238268, + "learning_rate": 1.936405586539326e-05, + "loss": 1.159, + "step": 5973 + }, + { + "epoch": 0.14, + "grad_norm": 2.6257265472169045, + "learning_rate": 1.936378807205673e-05, + "loss": 1.0814, + "step": 5974 + }, + { + "epoch": 0.14, + "grad_norm": 2.247845221215615, + "learning_rate": 1.9363520224201263e-05, + "loss": 1.0959, + "step": 5975 + }, + { + "epoch": 0.14, + "grad_norm": 1.1462549785522926, + "learning_rate": 1.9363252321828417e-05, + "loss": 0.9979, + "step": 5976 + }, + { + "epoch": 0.14, + "grad_norm": 2.067021077569738, + "learning_rate": 1.936298436493975e-05, + "loss": 1.2193, + "step": 5977 + }, + { + "epoch": 0.14, + "grad_norm": 2.018742278102007, + "learning_rate": 1.9362716353536825e-05, + "loss": 1.0944, + "step": 5978 + }, + { + "epoch": 0.14, + "grad_norm": 2.2708210280965213, + "learning_rate": 1.93624482876212e-05, + "loss": 0.961, + "step": 5979 + }, + { + "epoch": 0.14, + "grad_norm": 2.0914368241671997, + "learning_rate": 1.9362180167194435e-05, + "loss": 1.0959, + "step": 5980 + }, + { + "epoch": 0.14, + "grad_norm": 1.8613419394665947, + "learning_rate": 1.9361911992258096e-05, + "loss": 1.147, + "step": 5981 + }, + { + "epoch": 0.14, + "grad_norm": 2.101558879734842, + "learning_rate": 1.936164376281374e-05, + "loss": 1.1326, + "step": 5982 + }, + { + "epoch": 0.14, + "grad_norm": 2.2365240223635383, + "learning_rate": 1.9361375478862933e-05, + "loss": 1.0937, + "step": 5983 + }, + { + "epoch": 0.14, + "grad_norm": 1.9065184368163746, + "learning_rate": 1.9361107140407233e-05, + "loss": 1.0053, + "step": 5984 + }, + { + "epoch": 0.14, + "grad_norm": 2.9814897667300526, + "learning_rate": 1.9360838747448202e-05, + "loss": 1.0488, + "step": 5985 + }, + { + "epoch": 0.14, + "grad_norm": 2.44350467594041, + "learning_rate": 1.9360570299987405e-05, + "loss": 1.0827, + "step": 5986 + }, + { + "epoch": 0.14, + "grad_norm": 2.386043999986969, + "learning_rate": 1.9360301798026404e-05, + "loss": 1.0607, + "step": 5987 + }, + { + "epoch": 0.14, + "grad_norm": 2.439828735912139, + "learning_rate": 1.9360033241566764e-05, + "loss": 1.1339, + "step": 5988 + }, + { + "epoch": 0.14, + "grad_norm": 1.143959263264176, + "learning_rate": 1.9359764630610044e-05, + "loss": 1.0028, + "step": 5989 + }, + { + "epoch": 0.14, + "grad_norm": 2.5746382966028287, + "learning_rate": 1.9359495965157814e-05, + "loss": 1.0355, + "step": 5990 + }, + { + "epoch": 0.14, + "grad_norm": 2.2473119135836583, + "learning_rate": 1.9359227245211633e-05, + "loss": 1.1241, + "step": 5991 + }, + { + "epoch": 0.14, + "grad_norm": 2.156955903197366, + "learning_rate": 1.935895847077307e-05, + "loss": 0.9671, + "step": 5992 + }, + { + "epoch": 0.14, + "grad_norm": 2.2339376760622183, + "learning_rate": 1.9358689641843686e-05, + "loss": 1.032, + "step": 5993 + }, + { + "epoch": 0.14, + "grad_norm": 2.4750840104291907, + "learning_rate": 1.935842075842505e-05, + "loss": 1.0999, + "step": 5994 + }, + { + "epoch": 0.14, + "grad_norm": 1.9060645108478471, + "learning_rate": 1.9358151820518725e-05, + "loss": 0.9894, + "step": 5995 + }, + { + "epoch": 0.14, + "grad_norm": 2.0651565895390913, + "learning_rate": 1.9357882828126278e-05, + "loss": 1.1041, + "step": 5996 + }, + { + "epoch": 0.14, + "grad_norm": 2.2603015637106827, + "learning_rate": 1.935761378124927e-05, + "loss": 1.0581, + "step": 5997 + }, + { + "epoch": 0.14, + "grad_norm": 1.837134922741876, + "learning_rate": 1.935734467988928e-05, + "loss": 0.9203, + "step": 5998 + }, + { + "epoch": 0.14, + "grad_norm": 2.2081270712929952, + "learning_rate": 1.935707552404786e-05, + "loss": 0.9865, + "step": 5999 + }, + { + "epoch": 0.14, + "grad_norm": 1.9835472632763533, + "learning_rate": 1.9356806313726584e-05, + "loss": 1.071, + "step": 6000 + }, + { + "epoch": 0.14, + "grad_norm": 2.55040532455672, + "learning_rate": 1.9356537048927023e-05, + "loss": 1.012, + "step": 6001 + }, + { + "epoch": 0.14, + "grad_norm": 1.916088117701874, + "learning_rate": 1.9356267729650738e-05, + "loss": 0.9797, + "step": 6002 + }, + { + "epoch": 0.14, + "grad_norm": 1.1823420919304246, + "learning_rate": 1.93559983558993e-05, + "loss": 1.0001, + "step": 6003 + }, + { + "epoch": 0.14, + "grad_norm": 2.3308690177549565, + "learning_rate": 1.9355728927674284e-05, + "loss": 0.9576, + "step": 6004 + }, + { + "epoch": 0.14, + "grad_norm": 1.9456094771027352, + "learning_rate": 1.9355459444977245e-05, + "loss": 1.1403, + "step": 6005 + }, + { + "epoch": 0.14, + "grad_norm": 1.0718893142552801, + "learning_rate": 1.935518990780976e-05, + "loss": 1.0069, + "step": 6006 + }, + { + "epoch": 0.14, + "grad_norm": 2.136118203946188, + "learning_rate": 1.93549203161734e-05, + "loss": 0.9513, + "step": 6007 + }, + { + "epoch": 0.14, + "grad_norm": 2.9025114604311852, + "learning_rate": 1.9354650670069732e-05, + "loss": 1.0396, + "step": 6008 + }, + { + "epoch": 0.14, + "grad_norm": 1.2156897880780415, + "learning_rate": 1.9354380969500326e-05, + "loss": 1.0491, + "step": 6009 + }, + { + "epoch": 0.14, + "grad_norm": 1.8980442882768693, + "learning_rate": 1.9354111214466752e-05, + "loss": 0.9132, + "step": 6010 + }, + { + "epoch": 0.14, + "grad_norm": 2.391892530343651, + "learning_rate": 1.9353841404970582e-05, + "loss": 1.0023, + "step": 6011 + }, + { + "epoch": 0.14, + "grad_norm": 2.296381932680876, + "learning_rate": 1.9353571541013382e-05, + "loss": 1.0408, + "step": 6012 + }, + { + "epoch": 0.14, + "grad_norm": 2.429217063033157, + "learning_rate": 1.935330162259673e-05, + "loss": 0.9958, + "step": 6013 + }, + { + "epoch": 0.14, + "grad_norm": 2.2283539123059954, + "learning_rate": 1.9353031649722196e-05, + "loss": 1.1172, + "step": 6014 + }, + { + "epoch": 0.14, + "grad_norm": 2.046137798937783, + "learning_rate": 1.9352761622391348e-05, + "loss": 0.9991, + "step": 6015 + }, + { + "epoch": 0.14, + "grad_norm": 2.2206972941168575, + "learning_rate": 1.935249154060576e-05, + "loss": 1.1341, + "step": 6016 + }, + { + "epoch": 0.14, + "grad_norm": 2.1293707563948887, + "learning_rate": 1.9352221404367012e-05, + "loss": 0.9702, + "step": 6017 + }, + { + "epoch": 0.14, + "grad_norm": 2.081028809301495, + "learning_rate": 1.9351951213676664e-05, + "loss": 1.2573, + "step": 6018 + }, + { + "epoch": 0.14, + "grad_norm": 2.0587429730196476, + "learning_rate": 1.9351680968536298e-05, + "loss": 0.9811, + "step": 6019 + }, + { + "epoch": 0.14, + "grad_norm": 2.0047206721354316, + "learning_rate": 1.9351410668947482e-05, + "loss": 1.1121, + "step": 6020 + }, + { + "epoch": 0.14, + "grad_norm": 1.981272525759895, + "learning_rate": 1.9351140314911798e-05, + "loss": 1.0941, + "step": 6021 + }, + { + "epoch": 0.14, + "grad_norm": 2.18756151879182, + "learning_rate": 1.9350869906430807e-05, + "loss": 1.0025, + "step": 6022 + }, + { + "epoch": 0.14, + "grad_norm": 1.7323093899245354, + "learning_rate": 1.9350599443506097e-05, + "loss": 1.1031, + "step": 6023 + }, + { + "epoch": 0.14, + "grad_norm": 3.1704396358560594, + "learning_rate": 1.935032892613923e-05, + "loss": 1.0159, + "step": 6024 + }, + { + "epoch": 0.14, + "grad_norm": 2.405816663757632, + "learning_rate": 1.9350058354331796e-05, + "loss": 1.0698, + "step": 6025 + }, + { + "epoch": 0.14, + "grad_norm": 2.3717242514975414, + "learning_rate": 1.9349787728085357e-05, + "loss": 1.0628, + "step": 6026 + }, + { + "epoch": 0.14, + "grad_norm": 1.1538151923663718, + "learning_rate": 1.9349517047401493e-05, + "loss": 0.9895, + "step": 6027 + }, + { + "epoch": 0.14, + "grad_norm": 2.018734607766105, + "learning_rate": 1.9349246312281783e-05, + "loss": 1.0236, + "step": 6028 + }, + { + "epoch": 0.14, + "grad_norm": 2.0186705421093243, + "learning_rate": 1.9348975522727802e-05, + "loss": 1.0846, + "step": 6029 + }, + { + "epoch": 0.14, + "grad_norm": 2.7067719307975637, + "learning_rate": 1.9348704678741125e-05, + "loss": 1.1345, + "step": 6030 + }, + { + "epoch": 0.14, + "grad_norm": 2.0151247048903755, + "learning_rate": 1.934843378032333e-05, + "loss": 1.1627, + "step": 6031 + }, + { + "epoch": 0.14, + "grad_norm": 2.3452612634822643, + "learning_rate": 1.934816282747599e-05, + "loss": 0.9723, + "step": 6032 + }, + { + "epoch": 0.14, + "grad_norm": 2.246148079235797, + "learning_rate": 1.9347891820200692e-05, + "loss": 1.0261, + "step": 6033 + }, + { + "epoch": 0.14, + "grad_norm": 2.271870029926439, + "learning_rate": 1.9347620758499003e-05, + "loss": 1.085, + "step": 6034 + }, + { + "epoch": 0.14, + "grad_norm": 2.5849853877635915, + "learning_rate": 1.9347349642372512e-05, + "loss": 1.1854, + "step": 6035 + }, + { + "epoch": 0.14, + "grad_norm": 2.146149388112136, + "learning_rate": 1.9347078471822786e-05, + "loss": 0.9759, + "step": 6036 + }, + { + "epoch": 0.14, + "grad_norm": 1.9343247225254172, + "learning_rate": 1.9346807246851417e-05, + "loss": 1.0345, + "step": 6037 + }, + { + "epoch": 0.14, + "grad_norm": 2.1536150556508673, + "learning_rate": 1.9346535967459973e-05, + "loss": 1.0711, + "step": 6038 + }, + { + "epoch": 0.14, + "grad_norm": 2.269075537202193, + "learning_rate": 1.9346264633650036e-05, + "loss": 1.0861, + "step": 6039 + }, + { + "epoch": 0.14, + "grad_norm": 2.497459185627187, + "learning_rate": 1.934599324542319e-05, + "loss": 1.1876, + "step": 6040 + }, + { + "epoch": 0.14, + "grad_norm": 2.1460205951199, + "learning_rate": 1.934572180278101e-05, + "loss": 1.0373, + "step": 6041 + }, + { + "epoch": 0.14, + "grad_norm": 2.3974304159400885, + "learning_rate": 1.934545030572508e-05, + "loss": 1.0411, + "step": 6042 + }, + { + "epoch": 0.14, + "grad_norm": 2.4056957015929945, + "learning_rate": 1.934517875425698e-05, + "loss": 1.0483, + "step": 6043 + }, + { + "epoch": 0.14, + "grad_norm": 1.9729410589100662, + "learning_rate": 1.934490714837829e-05, + "loss": 1.1867, + "step": 6044 + }, + { + "epoch": 0.14, + "grad_norm": 2.1506290153151992, + "learning_rate": 1.9344635488090593e-05, + "loss": 1.1039, + "step": 6045 + }, + { + "epoch": 0.14, + "grad_norm": 1.1518077723228175, + "learning_rate": 1.934436377339547e-05, + "loss": 0.9597, + "step": 6046 + }, + { + "epoch": 0.14, + "grad_norm": 1.9925120300904788, + "learning_rate": 1.9344092004294503e-05, + "loss": 1.0476, + "step": 6047 + }, + { + "epoch": 0.14, + "grad_norm": 2.2918230909864037, + "learning_rate": 1.9343820180789273e-05, + "loss": 1.1273, + "step": 6048 + }, + { + "epoch": 0.14, + "grad_norm": 2.2743304851690307, + "learning_rate": 1.9343548302881364e-05, + "loss": 1.1843, + "step": 6049 + }, + { + "epoch": 0.14, + "grad_norm": 1.8983584283768447, + "learning_rate": 1.9343276370572357e-05, + "loss": 1.1086, + "step": 6050 + }, + { + "epoch": 0.14, + "grad_norm": 1.1496114946145555, + "learning_rate": 1.9343004383863838e-05, + "loss": 0.9644, + "step": 6051 + }, + { + "epoch": 0.14, + "grad_norm": 1.9937685064462882, + "learning_rate": 1.9342732342757392e-05, + "loss": 0.961, + "step": 6052 + }, + { + "epoch": 0.14, + "grad_norm": 2.01321290635783, + "learning_rate": 1.9342460247254596e-05, + "loss": 1.0132, + "step": 6053 + }, + { + "epoch": 0.14, + "grad_norm": 3.000496342541859, + "learning_rate": 1.9342188097357042e-05, + "loss": 1.0562, + "step": 6054 + }, + { + "epoch": 0.14, + "grad_norm": 2.6037254149705698, + "learning_rate": 1.934191589306631e-05, + "loss": 1.1325, + "step": 6055 + }, + { + "epoch": 0.14, + "grad_norm": 2.135751822336073, + "learning_rate": 1.9341643634383986e-05, + "loss": 1.1907, + "step": 6056 + }, + { + "epoch": 0.14, + "grad_norm": 2.118565010197051, + "learning_rate": 1.9341371321311654e-05, + "loss": 1.175, + "step": 6057 + }, + { + "epoch": 0.14, + "grad_norm": 2.112881116546821, + "learning_rate": 1.9341098953850905e-05, + "loss": 1.0595, + "step": 6058 + }, + { + "epoch": 0.14, + "grad_norm": 1.1944386946062662, + "learning_rate": 1.934082653200332e-05, + "loss": 1.0793, + "step": 6059 + }, + { + "epoch": 0.14, + "grad_norm": 2.1544607927229893, + "learning_rate": 1.934055405577048e-05, + "loss": 0.9943, + "step": 6060 + }, + { + "epoch": 0.14, + "grad_norm": 2.361886997402205, + "learning_rate": 1.9340281525153983e-05, + "loss": 0.9651, + "step": 6061 + }, + { + "epoch": 0.14, + "grad_norm": 2.2796235964826077, + "learning_rate": 1.9340008940155407e-05, + "loss": 1.0395, + "step": 6062 + }, + { + "epoch": 0.14, + "grad_norm": 2.0490100462657463, + "learning_rate": 1.933973630077634e-05, + "loss": 1.0797, + "step": 6063 + }, + { + "epoch": 0.14, + "grad_norm": 1.1281759746640885, + "learning_rate": 1.9339463607018375e-05, + "loss": 0.98, + "step": 6064 + }, + { + "epoch": 0.14, + "grad_norm": 2.0411720163487135, + "learning_rate": 1.9339190858883096e-05, + "loss": 1.0842, + "step": 6065 + }, + { + "epoch": 0.14, + "grad_norm": 2.1660210232714707, + "learning_rate": 1.933891805637209e-05, + "loss": 1.0789, + "step": 6066 + }, + { + "epoch": 0.14, + "grad_norm": 2.1537797006404316, + "learning_rate": 1.9338645199486947e-05, + "loss": 1.1455, + "step": 6067 + }, + { + "epoch": 0.14, + "grad_norm": 2.2143056718806133, + "learning_rate": 1.9338372288229253e-05, + "loss": 1.2554, + "step": 6068 + }, + { + "epoch": 0.14, + "grad_norm": 2.4702665955445404, + "learning_rate": 1.93380993226006e-05, + "loss": 1.0765, + "step": 6069 + }, + { + "epoch": 0.14, + "grad_norm": 3.3567436958007186, + "learning_rate": 1.933782630260258e-05, + "loss": 0.9544, + "step": 6070 + }, + { + "epoch": 0.14, + "grad_norm": 1.1149857640829364, + "learning_rate": 1.9337553228236772e-05, + "loss": 0.9707, + "step": 6071 + }, + { + "epoch": 0.14, + "grad_norm": 4.959692244078139, + "learning_rate": 1.933728009950478e-05, + "loss": 1.0286, + "step": 6072 + }, + { + "epoch": 0.14, + "grad_norm": 2.09994999085479, + "learning_rate": 1.9337006916408182e-05, + "loss": 1.1443, + "step": 6073 + }, + { + "epoch": 0.14, + "grad_norm": 2.2145412065401864, + "learning_rate": 1.9336733678948577e-05, + "loss": 1.0215, + "step": 6074 + }, + { + "epoch": 0.14, + "grad_norm": 2.4555845413554787, + "learning_rate": 1.933646038712755e-05, + "loss": 1.1227, + "step": 6075 + }, + { + "epoch": 0.14, + "grad_norm": 2.4042816953776405, + "learning_rate": 1.9336187040946695e-05, + "loss": 0.9572, + "step": 6076 + }, + { + "epoch": 0.14, + "grad_norm": 2.471209196235074, + "learning_rate": 1.9335913640407606e-05, + "loss": 1.173, + "step": 6077 + }, + { + "epoch": 0.14, + "grad_norm": 2.3007187356677186, + "learning_rate": 1.9335640185511872e-05, + "loss": 1.0629, + "step": 6078 + }, + { + "epoch": 0.14, + "grad_norm": 2.097603390447659, + "learning_rate": 1.933536667626108e-05, + "loss": 1.0788, + "step": 6079 + }, + { + "epoch": 0.14, + "grad_norm": 1.9217806416531749, + "learning_rate": 1.933509311265683e-05, + "loss": 0.9955, + "step": 6080 + }, + { + "epoch": 0.14, + "grad_norm": 2.0403181299924555, + "learning_rate": 1.9334819494700714e-05, + "loss": 0.9694, + "step": 6081 + }, + { + "epoch": 0.14, + "grad_norm": 1.180047502123527, + "learning_rate": 1.9334545822394327e-05, + "loss": 0.9294, + "step": 6082 + }, + { + "epoch": 0.14, + "grad_norm": 2.190819664019223, + "learning_rate": 1.9334272095739252e-05, + "loss": 1.142, + "step": 6083 + }, + { + "epoch": 0.14, + "grad_norm": 2.382750505330767, + "learning_rate": 1.9333998314737093e-05, + "loss": 0.9647, + "step": 6084 + }, + { + "epoch": 0.14, + "grad_norm": 1.9440187092932308, + "learning_rate": 1.9333724479389443e-05, + "loss": 1.1761, + "step": 6085 + }, + { + "epoch": 0.14, + "grad_norm": 2.115764165216237, + "learning_rate": 1.9333450589697887e-05, + "loss": 0.972, + "step": 6086 + }, + { + "epoch": 0.14, + "grad_norm": 2.1121896321358173, + "learning_rate": 1.9333176645664035e-05, + "loss": 1.1906, + "step": 6087 + }, + { + "epoch": 0.14, + "grad_norm": 2.112040272517175, + "learning_rate": 1.9332902647289468e-05, + "loss": 0.9854, + "step": 6088 + }, + { + "epoch": 0.14, + "grad_norm": 1.776881553477306, + "learning_rate": 1.9332628594575792e-05, + "loss": 0.9756, + "step": 6089 + }, + { + "epoch": 0.14, + "grad_norm": 2.162796167507451, + "learning_rate": 1.933235448752459e-05, + "loss": 0.963, + "step": 6090 + }, + { + "epoch": 0.14, + "grad_norm": 2.114757155220135, + "learning_rate": 1.9332080326137476e-05, + "loss": 0.9762, + "step": 6091 + }, + { + "epoch": 0.14, + "grad_norm": 1.9348970958737473, + "learning_rate": 1.9331806110416027e-05, + "loss": 1.1544, + "step": 6092 + }, + { + "epoch": 0.14, + "grad_norm": 1.8595895106281488, + "learning_rate": 1.9331531840361852e-05, + "loss": 1.0907, + "step": 6093 + }, + { + "epoch": 0.14, + "grad_norm": 2.1562009180398336, + "learning_rate": 1.9331257515976545e-05, + "loss": 1.0128, + "step": 6094 + }, + { + "epoch": 0.14, + "grad_norm": 2.4661972245058457, + "learning_rate": 1.93309831372617e-05, + "loss": 1.0547, + "step": 6095 + }, + { + "epoch": 0.14, + "grad_norm": 2.55463057045908, + "learning_rate": 1.933070870421892e-05, + "loss": 1.0067, + "step": 6096 + }, + { + "epoch": 0.14, + "grad_norm": 1.213898853600961, + "learning_rate": 1.9330434216849796e-05, + "loss": 1.0317, + "step": 6097 + }, + { + "epoch": 0.14, + "grad_norm": 2.1986035828825967, + "learning_rate": 1.9330159675155933e-05, + "loss": 0.9477, + "step": 6098 + }, + { + "epoch": 0.14, + "grad_norm": 1.1721428596815677, + "learning_rate": 1.9329885079138927e-05, + "loss": 1.0736, + "step": 6099 + }, + { + "epoch": 0.14, + "grad_norm": 1.9882777647060959, + "learning_rate": 1.9329610428800376e-05, + "loss": 1.1376, + "step": 6100 + }, + { + "epoch": 0.14, + "grad_norm": 1.9546884687682022, + "learning_rate": 1.9329335724141877e-05, + "loss": 0.9047, + "step": 6101 + }, + { + "epoch": 0.14, + "grad_norm": 1.8304802981688857, + "learning_rate": 1.9329060965165032e-05, + "loss": 1.0861, + "step": 6102 + }, + { + "epoch": 0.14, + "grad_norm": 2.25857412467619, + "learning_rate": 1.9328786151871444e-05, + "loss": 1.1469, + "step": 6103 + }, + { + "epoch": 0.14, + "grad_norm": 2.23491514494552, + "learning_rate": 1.9328511284262705e-05, + "loss": 1.0816, + "step": 6104 + }, + { + "epoch": 0.14, + "grad_norm": 2.1060314314025264, + "learning_rate": 1.9328236362340425e-05, + "loss": 1.0395, + "step": 6105 + }, + { + "epoch": 0.14, + "grad_norm": 1.9531779456114882, + "learning_rate": 1.9327961386106198e-05, + "loss": 1.073, + "step": 6106 + }, + { + "epoch": 0.14, + "grad_norm": 2.0773819214680564, + "learning_rate": 1.9327686355561622e-05, + "loss": 1.0942, + "step": 6107 + }, + { + "epoch": 0.14, + "grad_norm": 1.9820315506451476, + "learning_rate": 1.932741127070831e-05, + "loss": 1.0592, + "step": 6108 + }, + { + "epoch": 0.14, + "grad_norm": 1.9430289559082359, + "learning_rate": 1.932713613154785e-05, + "loss": 1.0815, + "step": 6109 + }, + { + "epoch": 0.14, + "grad_norm": 1.9580240779305755, + "learning_rate": 1.9326860938081854e-05, + "loss": 0.9984, + "step": 6110 + }, + { + "epoch": 0.14, + "grad_norm": 2.0322863141752467, + "learning_rate": 1.932658569031192e-05, + "loss": 1.0815, + "step": 6111 + }, + { + "epoch": 0.14, + "grad_norm": 2.3609561066072606, + "learning_rate": 1.9326310388239654e-05, + "loss": 1.0475, + "step": 6112 + }, + { + "epoch": 0.14, + "grad_norm": 2.219611477711065, + "learning_rate": 1.9326035031866653e-05, + "loss": 1.0424, + "step": 6113 + }, + { + "epoch": 0.14, + "grad_norm": 2.0927529300640817, + "learning_rate": 1.9325759621194528e-05, + "loss": 1.0257, + "step": 6114 + }, + { + "epoch": 0.14, + "grad_norm": 1.9707224789879842, + "learning_rate": 1.9325484156224872e-05, + "loss": 1.0419, + "step": 6115 + }, + { + "epoch": 0.14, + "grad_norm": 2.037542112534067, + "learning_rate": 1.9325208636959297e-05, + "loss": 1.0957, + "step": 6116 + }, + { + "epoch": 0.14, + "grad_norm": 2.1545074833105735, + "learning_rate": 1.932493306339941e-05, + "loss": 1.0977, + "step": 6117 + }, + { + "epoch": 0.14, + "grad_norm": 2.156825360623104, + "learning_rate": 1.9324657435546802e-05, + "loss": 1.0392, + "step": 6118 + }, + { + "epoch": 0.14, + "grad_norm": 1.7559773593917372, + "learning_rate": 1.932438175340309e-05, + "loss": 1.0236, + "step": 6119 + }, + { + "epoch": 0.14, + "grad_norm": 1.9650569491060046, + "learning_rate": 1.9324106016969875e-05, + "loss": 1.0749, + "step": 6120 + }, + { + "epoch": 0.14, + "grad_norm": 2.3143139539797404, + "learning_rate": 1.9323830226248763e-05, + "loss": 1.0214, + "step": 6121 + }, + { + "epoch": 0.14, + "grad_norm": 4.298899974725176, + "learning_rate": 1.932355438124136e-05, + "loss": 1.0367, + "step": 6122 + }, + { + "epoch": 0.14, + "grad_norm": 2.003962879305611, + "learning_rate": 1.9323278481949274e-05, + "loss": 0.9777, + "step": 6123 + }, + { + "epoch": 0.14, + "grad_norm": 2.3925820068892305, + "learning_rate": 1.9323002528374104e-05, + "loss": 1.1116, + "step": 6124 + }, + { + "epoch": 0.14, + "grad_norm": 2.6526482177173087, + "learning_rate": 1.9322726520517465e-05, + "loss": 1.0589, + "step": 6125 + }, + { + "epoch": 0.14, + "grad_norm": 2.196524248298894, + "learning_rate": 1.932245045838096e-05, + "loss": 1.0592, + "step": 6126 + }, + { + "epoch": 0.14, + "grad_norm": 2.055589615205437, + "learning_rate": 1.9322174341966196e-05, + "loss": 1.1566, + "step": 6127 + }, + { + "epoch": 0.14, + "grad_norm": 2.311735298305228, + "learning_rate": 1.9321898171274785e-05, + "loss": 0.9427, + "step": 6128 + }, + { + "epoch": 0.14, + "grad_norm": 2.005034429021838, + "learning_rate": 1.9321621946308328e-05, + "loss": 1.1289, + "step": 6129 + }, + { + "epoch": 0.14, + "grad_norm": 2.800219688604105, + "learning_rate": 1.9321345667068435e-05, + "loss": 1.091, + "step": 6130 + }, + { + "epoch": 0.14, + "grad_norm": 2.4328171479117824, + "learning_rate": 1.9321069333556722e-05, + "loss": 1.131, + "step": 6131 + }, + { + "epoch": 0.14, + "grad_norm": 2.126402511515258, + "learning_rate": 1.9320792945774788e-05, + "loss": 1.0461, + "step": 6132 + }, + { + "epoch": 0.14, + "grad_norm": 2.0200782115666795, + "learning_rate": 1.932051650372425e-05, + "loss": 1.0827, + "step": 6133 + }, + { + "epoch": 0.14, + "grad_norm": 2.5792912963566814, + "learning_rate": 1.932024000740671e-05, + "loss": 1.1409, + "step": 6134 + }, + { + "epoch": 0.14, + "grad_norm": 1.154534437693626, + "learning_rate": 1.9319963456823788e-05, + "loss": 0.9857, + "step": 6135 + }, + { + "epoch": 0.14, + "grad_norm": 2.1164142069718936, + "learning_rate": 1.931968685197708e-05, + "loss": 1.0129, + "step": 6136 + }, + { + "epoch": 0.14, + "grad_norm": 2.3881178232439946, + "learning_rate": 1.931941019286821e-05, + "loss": 0.9235, + "step": 6137 + }, + { + "epoch": 0.14, + "grad_norm": 2.1359965451364946, + "learning_rate": 1.931913347949878e-05, + "loss": 1.0782, + "step": 6138 + }, + { + "epoch": 0.14, + "grad_norm": 2.2149802246634502, + "learning_rate": 1.931885671187041e-05, + "loss": 1.0668, + "step": 6139 + }, + { + "epoch": 0.14, + "grad_norm": 2.316382296950568, + "learning_rate": 1.9318579889984705e-05, + "loss": 0.9728, + "step": 6140 + }, + { + "epoch": 0.14, + "grad_norm": 2.010633357415937, + "learning_rate": 1.9318303013843272e-05, + "loss": 1.122, + "step": 6141 + }, + { + "epoch": 0.14, + "grad_norm": 1.0785448657917678, + "learning_rate": 1.9318026083447733e-05, + "loss": 1.0095, + "step": 6142 + }, + { + "epoch": 0.14, + "grad_norm": 1.953069075260589, + "learning_rate": 1.9317749098799695e-05, + "loss": 1.0259, + "step": 6143 + }, + { + "epoch": 0.14, + "grad_norm": 2.0430390674550116, + "learning_rate": 1.9317472059900773e-05, + "loss": 0.9988, + "step": 6144 + }, + { + "epoch": 0.14, + "grad_norm": 2.1810645833452376, + "learning_rate": 1.9317194966752576e-05, + "loss": 1.1496, + "step": 6145 + }, + { + "epoch": 0.14, + "grad_norm": 2.288437503815251, + "learning_rate": 1.9316917819356723e-05, + "loss": 0.9998, + "step": 6146 + }, + { + "epoch": 0.14, + "grad_norm": 1.1167923974334018, + "learning_rate": 1.9316640617714823e-05, + "loss": 0.9634, + "step": 6147 + }, + { + "epoch": 0.14, + "grad_norm": 2.146950825087234, + "learning_rate": 1.9316363361828493e-05, + "loss": 1.0314, + "step": 6148 + }, + { + "epoch": 0.14, + "grad_norm": 2.3907315890385785, + "learning_rate": 1.9316086051699345e-05, + "loss": 1.0629, + "step": 6149 + }, + { + "epoch": 0.14, + "grad_norm": 1.9582518481387834, + "learning_rate": 1.9315808687328996e-05, + "loss": 1.0158, + "step": 6150 + }, + { + "epoch": 0.14, + "grad_norm": 1.9452301957345108, + "learning_rate": 1.9315531268719058e-05, + "loss": 1.0848, + "step": 6151 + }, + { + "epoch": 0.14, + "grad_norm": 2.9860496484714374, + "learning_rate": 1.931525379587115e-05, + "loss": 1.1925, + "step": 6152 + }, + { + "epoch": 0.14, + "grad_norm": 2.1951280270098494, + "learning_rate": 1.9314976268786883e-05, + "loss": 1.0462, + "step": 6153 + }, + { + "epoch": 0.14, + "grad_norm": 1.1017555140004884, + "learning_rate": 1.9314698687467876e-05, + "loss": 0.9774, + "step": 6154 + }, + { + "epoch": 0.15, + "grad_norm": 1.1140743828874753, + "learning_rate": 1.9314421051915744e-05, + "loss": 0.9895, + "step": 6155 + }, + { + "epoch": 0.15, + "grad_norm": 2.096527448323284, + "learning_rate": 1.93141433621321e-05, + "loss": 1.0529, + "step": 6156 + }, + { + "epoch": 0.15, + "grad_norm": 2.3904228518450856, + "learning_rate": 1.931386561811857e-05, + "loss": 1.0018, + "step": 6157 + }, + { + "epoch": 0.15, + "grad_norm": 2.0996451964760268, + "learning_rate": 1.9313587819876765e-05, + "loss": 1.178, + "step": 6158 + }, + { + "epoch": 0.15, + "grad_norm": 2.1691471844399146, + "learning_rate": 1.93133099674083e-05, + "loss": 1.0678, + "step": 6159 + }, + { + "epoch": 0.15, + "grad_norm": 2.143726711007237, + "learning_rate": 1.9313032060714797e-05, + "loss": 1.1286, + "step": 6160 + }, + { + "epoch": 0.15, + "grad_norm": 1.9336142761653377, + "learning_rate": 1.9312754099797876e-05, + "loss": 1.1805, + "step": 6161 + }, + { + "epoch": 0.15, + "grad_norm": 1.9874505426508815, + "learning_rate": 1.931247608465915e-05, + "loss": 1.166, + "step": 6162 + }, + { + "epoch": 0.15, + "grad_norm": 2.0202439500243696, + "learning_rate": 1.931219801530024e-05, + "loss": 0.9082, + "step": 6163 + }, + { + "epoch": 0.15, + "grad_norm": 2.0228255321801707, + "learning_rate": 1.9311919891722764e-05, + "loss": 1.0813, + "step": 6164 + }, + { + "epoch": 0.15, + "grad_norm": 2.4656260496043316, + "learning_rate": 1.9311641713928342e-05, + "loss": 1.0753, + "step": 6165 + }, + { + "epoch": 0.15, + "grad_norm": 2.1293911369124356, + "learning_rate": 1.9311363481918594e-05, + "loss": 1.091, + "step": 6166 + }, + { + "epoch": 0.15, + "grad_norm": 1.1873471340493162, + "learning_rate": 1.9311085195695142e-05, + "loss": 0.9759, + "step": 6167 + }, + { + "epoch": 0.15, + "grad_norm": 2.1084524663890756, + "learning_rate": 1.9310806855259604e-05, + "loss": 1.0349, + "step": 6168 + }, + { + "epoch": 0.15, + "grad_norm": 1.1276275219702903, + "learning_rate": 1.93105284606136e-05, + "loss": 0.9568, + "step": 6169 + }, + { + "epoch": 0.15, + "grad_norm": 1.8621087883911744, + "learning_rate": 1.9310250011758752e-05, + "loss": 1.1428, + "step": 6170 + }, + { + "epoch": 0.15, + "grad_norm": 2.2584811323165748, + "learning_rate": 1.9309971508696675e-05, + "loss": 0.9845, + "step": 6171 + }, + { + "epoch": 0.15, + "grad_norm": 3.1700672008668054, + "learning_rate": 1.9309692951429003e-05, + "loss": 1.1528, + "step": 6172 + }, + { + "epoch": 0.15, + "grad_norm": 2.1161499201678957, + "learning_rate": 1.9309414339957355e-05, + "loss": 1.0346, + "step": 6173 + }, + { + "epoch": 0.15, + "grad_norm": 2.1395052970135584, + "learning_rate": 1.9309135674283343e-05, + "loss": 1.0065, + "step": 6174 + }, + { + "epoch": 0.15, + "grad_norm": 2.200019906472777, + "learning_rate": 1.9308856954408594e-05, + "loss": 1.1127, + "step": 6175 + }, + { + "epoch": 0.15, + "grad_norm": 2.407185569437358, + "learning_rate": 1.9308578180334737e-05, + "loss": 1.1182, + "step": 6176 + }, + { + "epoch": 0.15, + "grad_norm": 1.097131042640014, + "learning_rate": 1.930829935206339e-05, + "loss": 0.9415, + "step": 6177 + }, + { + "epoch": 0.15, + "grad_norm": 2.117897800912682, + "learning_rate": 1.9308020469596177e-05, + "loss": 1.0508, + "step": 6178 + }, + { + "epoch": 0.15, + "grad_norm": 2.114417807282587, + "learning_rate": 1.930774153293472e-05, + "loss": 1.0387, + "step": 6179 + }, + { + "epoch": 0.15, + "grad_norm": 2.1535224660714243, + "learning_rate": 1.9307462542080646e-05, + "loss": 1.0162, + "step": 6180 + }, + { + "epoch": 0.15, + "grad_norm": 2.2187289318617385, + "learning_rate": 1.930718349703558e-05, + "loss": 1.0171, + "step": 6181 + }, + { + "epoch": 0.15, + "grad_norm": 1.9999497786145242, + "learning_rate": 1.9306904397801144e-05, + "loss": 1.1099, + "step": 6182 + }, + { + "epoch": 0.15, + "grad_norm": 2.2017202985640814, + "learning_rate": 1.9306625244378964e-05, + "loss": 0.9294, + "step": 6183 + }, + { + "epoch": 0.15, + "grad_norm": 2.0415893721328326, + "learning_rate": 1.9306346036770667e-05, + "loss": 1.0823, + "step": 6184 + }, + { + "epoch": 0.15, + "grad_norm": 2.028351393343509, + "learning_rate": 1.9306066774977875e-05, + "loss": 1.075, + "step": 6185 + }, + { + "epoch": 0.15, + "grad_norm": 2.0351212422092866, + "learning_rate": 1.930578745900222e-05, + "loss": 1.0462, + "step": 6186 + }, + { + "epoch": 0.15, + "grad_norm": 2.1745885663247244, + "learning_rate": 1.930550808884532e-05, + "loss": 0.843, + "step": 6187 + }, + { + "epoch": 0.15, + "grad_norm": 2.1782846517875263, + "learning_rate": 1.9305228664508805e-05, + "loss": 1.0348, + "step": 6188 + }, + { + "epoch": 0.15, + "grad_norm": 2.3182546723318675, + "learning_rate": 1.9304949185994303e-05, + "loss": 0.9543, + "step": 6189 + }, + { + "epoch": 0.15, + "grad_norm": 1.974928311914452, + "learning_rate": 1.9304669653303444e-05, + "loss": 0.9844, + "step": 6190 + }, + { + "epoch": 0.15, + "grad_norm": 2.3061464219963295, + "learning_rate": 1.930439006643785e-05, + "loss": 1.0906, + "step": 6191 + }, + { + "epoch": 0.15, + "grad_norm": 2.320891079990905, + "learning_rate": 1.930411042539915e-05, + "loss": 0.9495, + "step": 6192 + }, + { + "epoch": 0.15, + "grad_norm": 1.1393524909722357, + "learning_rate": 1.930383073018898e-05, + "loss": 1.0661, + "step": 6193 + }, + { + "epoch": 0.15, + "grad_norm": 2.6933606364224407, + "learning_rate": 1.9303550980808956e-05, + "loss": 0.9852, + "step": 6194 + }, + { + "epoch": 0.15, + "grad_norm": 2.0484965063299647, + "learning_rate": 1.9303271177260715e-05, + "loss": 0.9972, + "step": 6195 + }, + { + "epoch": 0.15, + "grad_norm": 2.224346920738157, + "learning_rate": 1.9302991319545883e-05, + "loss": 1.234, + "step": 6196 + }, + { + "epoch": 0.15, + "grad_norm": 2.114430920978323, + "learning_rate": 1.930271140766609e-05, + "loss": 1.081, + "step": 6197 + }, + { + "epoch": 0.15, + "grad_norm": 2.202461442144581, + "learning_rate": 1.930243144162297e-05, + "loss": 1.0269, + "step": 6198 + }, + { + "epoch": 0.15, + "grad_norm": 2.1256409885330387, + "learning_rate": 1.930215142141814e-05, + "loss": 1.1944, + "step": 6199 + }, + { + "epoch": 0.15, + "grad_norm": 1.9233033301942135, + "learning_rate": 1.930187134705325e-05, + "loss": 1.0378, + "step": 6200 + }, + { + "epoch": 0.15, + "grad_norm": 2.1234227183826078, + "learning_rate": 1.930159121852991e-05, + "loss": 1.0858, + "step": 6201 + }, + { + "epoch": 0.15, + "grad_norm": 1.995388594996153, + "learning_rate": 1.930131103584977e-05, + "loss": 0.9331, + "step": 6202 + }, + { + "epoch": 0.15, + "grad_norm": 1.9384864922180296, + "learning_rate": 1.9301030799014445e-05, + "loss": 1.0129, + "step": 6203 + }, + { + "epoch": 0.15, + "grad_norm": 2.176168453119368, + "learning_rate": 1.9300750508025577e-05, + "loss": 1.0803, + "step": 6204 + }, + { + "epoch": 0.15, + "grad_norm": 2.1059684384950494, + "learning_rate": 1.93004701628848e-05, + "loss": 1.0178, + "step": 6205 + }, + { + "epoch": 0.15, + "grad_norm": 1.110576168877194, + "learning_rate": 1.9300189763593736e-05, + "loss": 0.975, + "step": 6206 + }, + { + "epoch": 0.15, + "grad_norm": 2.326303866316666, + "learning_rate": 1.9299909310154025e-05, + "loss": 1.0854, + "step": 6207 + }, + { + "epoch": 0.15, + "grad_norm": 2.1915028981283444, + "learning_rate": 1.9299628802567293e-05, + "loss": 1.1173, + "step": 6208 + }, + { + "epoch": 0.15, + "grad_norm": 2.01794638866403, + "learning_rate": 1.9299348240835182e-05, + "loss": 0.9656, + "step": 6209 + }, + { + "epoch": 0.15, + "grad_norm": 1.9255286512555492, + "learning_rate": 1.9299067624959322e-05, + "loss": 0.9513, + "step": 6210 + }, + { + "epoch": 0.15, + "grad_norm": 2.0400887455192858, + "learning_rate": 1.9298786954941344e-05, + "loss": 1.0357, + "step": 6211 + }, + { + "epoch": 0.15, + "grad_norm": 2.216625059397209, + "learning_rate": 1.9298506230782886e-05, + "loss": 1.0485, + "step": 6212 + }, + { + "epoch": 0.15, + "grad_norm": 2.1311482019379384, + "learning_rate": 1.929822545248558e-05, + "loss": 1.0862, + "step": 6213 + }, + { + "epoch": 0.15, + "grad_norm": 2.166983756340785, + "learning_rate": 1.9297944620051063e-05, + "loss": 1.0676, + "step": 6214 + }, + { + "epoch": 0.15, + "grad_norm": 2.3128429210233357, + "learning_rate": 1.9297663733480966e-05, + "loss": 1.0802, + "step": 6215 + }, + { + "epoch": 0.15, + "grad_norm": 2.0832713594631387, + "learning_rate": 1.929738279277693e-05, + "loss": 0.9669, + "step": 6216 + }, + { + "epoch": 0.15, + "grad_norm": 2.0710929845691965, + "learning_rate": 1.9297101797940586e-05, + "loss": 1.0311, + "step": 6217 + }, + { + "epoch": 0.15, + "grad_norm": 2.313727829759069, + "learning_rate": 1.9296820748973572e-05, + "loss": 1.0699, + "step": 6218 + }, + { + "epoch": 0.15, + "grad_norm": 2.08625935007432, + "learning_rate": 1.9296539645877524e-05, + "loss": 1.0906, + "step": 6219 + }, + { + "epoch": 0.15, + "grad_norm": 2.0122947254895123, + "learning_rate": 1.9296258488654082e-05, + "loss": 1.0982, + "step": 6220 + }, + { + "epoch": 0.15, + "grad_norm": 2.0858311171719497, + "learning_rate": 1.9295977277304875e-05, + "loss": 1.0432, + "step": 6221 + }, + { + "epoch": 0.15, + "grad_norm": 2.1662216838807087, + "learning_rate": 1.9295696011831547e-05, + "loss": 1.0447, + "step": 6222 + }, + { + "epoch": 0.15, + "grad_norm": 1.9698885758598244, + "learning_rate": 1.9295414692235738e-05, + "loss": 1.1804, + "step": 6223 + }, + { + "epoch": 0.15, + "grad_norm": 1.14400070428217, + "learning_rate": 1.9295133318519077e-05, + "loss": 0.9575, + "step": 6224 + }, + { + "epoch": 0.15, + "grad_norm": 2.119211677379713, + "learning_rate": 1.9294851890683208e-05, + "loss": 1.0983, + "step": 6225 + }, + { + "epoch": 0.15, + "grad_norm": 1.1932309043693115, + "learning_rate": 1.929457040872977e-05, + "loss": 0.9624, + "step": 6226 + }, + { + "epoch": 0.15, + "grad_norm": 1.1209689735048436, + "learning_rate": 1.92942888726604e-05, + "loss": 1.0017, + "step": 6227 + }, + { + "epoch": 0.15, + "grad_norm": 2.0481343350397436, + "learning_rate": 1.9294007282476737e-05, + "loss": 0.9711, + "step": 6228 + }, + { + "epoch": 0.15, + "grad_norm": 2.5724543605962498, + "learning_rate": 1.929372563818042e-05, + "loss": 1.1085, + "step": 6229 + }, + { + "epoch": 0.15, + "grad_norm": 1.9777017683040103, + "learning_rate": 1.9293443939773092e-05, + "loss": 1.1169, + "step": 6230 + }, + { + "epoch": 0.15, + "grad_norm": 2.0310481762114496, + "learning_rate": 1.9293162187256392e-05, + "loss": 1.0478, + "step": 6231 + }, + { + "epoch": 0.15, + "grad_norm": 2.1224872340397547, + "learning_rate": 1.9292880380631956e-05, + "loss": 1.1289, + "step": 6232 + }, + { + "epoch": 0.15, + "grad_norm": 2.152588444528317, + "learning_rate": 1.929259851990143e-05, + "loss": 0.9811, + "step": 6233 + }, + { + "epoch": 0.15, + "grad_norm": 2.060186390734041, + "learning_rate": 1.9292316605066455e-05, + "loss": 1.0452, + "step": 6234 + }, + { + "epoch": 0.15, + "grad_norm": 1.9064092166552549, + "learning_rate": 1.929203463612867e-05, + "loss": 0.9764, + "step": 6235 + }, + { + "epoch": 0.15, + "grad_norm": 2.4072841330795636, + "learning_rate": 1.9291752613089717e-05, + "loss": 1.1825, + "step": 6236 + }, + { + "epoch": 0.15, + "grad_norm": 2.594990951555271, + "learning_rate": 1.929147053595124e-05, + "loss": 1.0272, + "step": 6237 + }, + { + "epoch": 0.15, + "grad_norm": 2.340778547995798, + "learning_rate": 1.9291188404714876e-05, + "loss": 0.9631, + "step": 6238 + }, + { + "epoch": 0.15, + "grad_norm": 2.1491584418936083, + "learning_rate": 1.9290906219382277e-05, + "loss": 1.0153, + "step": 6239 + }, + { + "epoch": 0.15, + "grad_norm": 2.2093773030944526, + "learning_rate": 1.9290623979955077e-05, + "loss": 1.0503, + "step": 6240 + }, + { + "epoch": 0.15, + "grad_norm": 2.2302554292851275, + "learning_rate": 1.9290341686434924e-05, + "loss": 1.0357, + "step": 6241 + }, + { + "epoch": 0.15, + "grad_norm": 2.182169997297219, + "learning_rate": 1.929005933882346e-05, + "loss": 1.0847, + "step": 6242 + }, + { + "epoch": 0.15, + "grad_norm": 2.3790143213854074, + "learning_rate": 1.928977693712233e-05, + "loss": 1.2081, + "step": 6243 + }, + { + "epoch": 0.15, + "grad_norm": 2.2165382954950403, + "learning_rate": 1.9289494481333177e-05, + "loss": 1.0875, + "step": 6244 + }, + { + "epoch": 0.15, + "grad_norm": 2.0799453482538977, + "learning_rate": 1.9289211971457644e-05, + "loss": 1.0643, + "step": 6245 + }, + { + "epoch": 0.15, + "grad_norm": 1.966460224606452, + "learning_rate": 1.928892940749738e-05, + "loss": 1.1017, + "step": 6246 + }, + { + "epoch": 0.15, + "grad_norm": 2.2851707502866887, + "learning_rate": 1.9288646789454032e-05, + "loss": 1.129, + "step": 6247 + }, + { + "epoch": 0.15, + "grad_norm": 2.4524230535502, + "learning_rate": 1.9288364117329233e-05, + "loss": 1.0703, + "step": 6248 + }, + { + "epoch": 0.15, + "grad_norm": 1.9246067390961943, + "learning_rate": 1.9288081391124643e-05, + "loss": 1.107, + "step": 6249 + }, + { + "epoch": 0.15, + "grad_norm": 1.9315020571905013, + "learning_rate": 1.9287798610841903e-05, + "loss": 1.1191, + "step": 6250 + }, + { + "epoch": 0.15, + "grad_norm": 1.2458596669261506, + "learning_rate": 1.928751577648266e-05, + "loss": 0.9734, + "step": 6251 + }, + { + "epoch": 0.15, + "grad_norm": 1.903550317972817, + "learning_rate": 1.9287232888048554e-05, + "loss": 0.9561, + "step": 6252 + }, + { + "epoch": 0.15, + "grad_norm": 2.0663330023678905, + "learning_rate": 1.9286949945541243e-05, + "loss": 0.9906, + "step": 6253 + }, + { + "epoch": 0.15, + "grad_norm": 2.4632457519475457, + "learning_rate": 1.9286666948962368e-05, + "loss": 1.146, + "step": 6254 + }, + { + "epoch": 0.15, + "grad_norm": 2.0482404951320663, + "learning_rate": 1.9286383898313578e-05, + "loss": 1.0826, + "step": 6255 + }, + { + "epoch": 0.15, + "grad_norm": 2.1591306458264783, + "learning_rate": 1.9286100793596523e-05, + "loss": 1.0977, + "step": 6256 + }, + { + "epoch": 0.15, + "grad_norm": 2.0111746931943557, + "learning_rate": 1.9285817634812844e-05, + "loss": 1.0547, + "step": 6257 + }, + { + "epoch": 0.15, + "grad_norm": 1.9452001770704146, + "learning_rate": 1.92855344219642e-05, + "loss": 1.092, + "step": 6258 + }, + { + "epoch": 0.15, + "grad_norm": 1.9302122296965023, + "learning_rate": 1.9285251155052233e-05, + "loss": 1.04, + "step": 6259 + }, + { + "epoch": 0.15, + "grad_norm": 2.0159459443052974, + "learning_rate": 1.928496783407859e-05, + "loss": 1.1909, + "step": 6260 + }, + { + "epoch": 0.15, + "grad_norm": 1.1077827790818902, + "learning_rate": 1.928468445904493e-05, + "loss": 0.9378, + "step": 6261 + }, + { + "epoch": 0.15, + "grad_norm": 2.3086525627977728, + "learning_rate": 1.9284401029952898e-05, + "loss": 1.0925, + "step": 6262 + }, + { + "epoch": 0.15, + "grad_norm": 2.103049547172551, + "learning_rate": 1.928411754680414e-05, + "loss": 1.1589, + "step": 6263 + }, + { + "epoch": 0.15, + "grad_norm": 2.181021462380686, + "learning_rate": 1.928383400960031e-05, + "loss": 1.1149, + "step": 6264 + }, + { + "epoch": 0.15, + "grad_norm": 2.4969977317760024, + "learning_rate": 1.9283550418343063e-05, + "loss": 0.9998, + "step": 6265 + }, + { + "epoch": 0.15, + "grad_norm": 2.5666962997494815, + "learning_rate": 1.9283266773034047e-05, + "loss": 1.1116, + "step": 6266 + }, + { + "epoch": 0.15, + "grad_norm": 1.1584173007703387, + "learning_rate": 1.928298307367491e-05, + "loss": 1.0399, + "step": 6267 + }, + { + "epoch": 0.15, + "grad_norm": 2.080479546550974, + "learning_rate": 1.928269932026731e-05, + "loss": 1.0152, + "step": 6268 + }, + { + "epoch": 0.15, + "grad_norm": 2.4727480556445487, + "learning_rate": 1.928241551281289e-05, + "loss": 0.9591, + "step": 6269 + }, + { + "epoch": 0.15, + "grad_norm": 2.0602037929505315, + "learning_rate": 1.928213165131331e-05, + "loss": 0.96, + "step": 6270 + }, + { + "epoch": 0.15, + "grad_norm": 2.5602738651711614, + "learning_rate": 1.9281847735770224e-05, + "loss": 1.0439, + "step": 6271 + }, + { + "epoch": 0.15, + "grad_norm": 1.152826979107836, + "learning_rate": 1.928156376618528e-05, + "loss": 0.99, + "step": 6272 + }, + { + "epoch": 0.15, + "grad_norm": 2.2257241037870825, + "learning_rate": 1.9281279742560134e-05, + "loss": 1.0304, + "step": 6273 + }, + { + "epoch": 0.15, + "grad_norm": 2.006135199483353, + "learning_rate": 1.9280995664896435e-05, + "loss": 1.124, + "step": 6274 + }, + { + "epoch": 0.15, + "grad_norm": 2.139106165869182, + "learning_rate": 1.9280711533195843e-05, + "loss": 1.2828, + "step": 6275 + }, + { + "epoch": 0.15, + "grad_norm": 1.9701147676834807, + "learning_rate": 1.9280427347460013e-05, + "loss": 1.0326, + "step": 6276 + }, + { + "epoch": 0.15, + "grad_norm": 2.152912140132529, + "learning_rate": 1.9280143107690593e-05, + "loss": 1.0873, + "step": 6277 + }, + { + "epoch": 0.15, + "grad_norm": 2.1242863183290424, + "learning_rate": 1.9279858813889246e-05, + "loss": 0.9977, + "step": 6278 + }, + { + "epoch": 0.15, + "grad_norm": 2.2235411142164456, + "learning_rate": 1.927957446605762e-05, + "loss": 1.1314, + "step": 6279 + }, + { + "epoch": 0.15, + "grad_norm": 2.5335563369854945, + "learning_rate": 1.9279290064197375e-05, + "loss": 0.9581, + "step": 6280 + }, + { + "epoch": 0.15, + "grad_norm": 1.9864367172401893, + "learning_rate": 1.9279005608310165e-05, + "loss": 1.025, + "step": 6281 + }, + { + "epoch": 0.15, + "grad_norm": 1.9128559819214672, + "learning_rate": 1.9278721098397645e-05, + "loss": 1.1968, + "step": 6282 + }, + { + "epoch": 0.15, + "grad_norm": 2.3092654279000637, + "learning_rate": 1.9278436534461476e-05, + "loss": 0.9282, + "step": 6283 + }, + { + "epoch": 0.15, + "grad_norm": 2.1225499248832205, + "learning_rate": 1.927815191650331e-05, + "loss": 1.18, + "step": 6284 + }, + { + "epoch": 0.15, + "grad_norm": 1.1587791397599607, + "learning_rate": 1.9277867244524805e-05, + "loss": 1.0785, + "step": 6285 + }, + { + "epoch": 0.15, + "grad_norm": 2.1607186088147907, + "learning_rate": 1.927758251852762e-05, + "loss": 1.0081, + "step": 6286 + }, + { + "epoch": 0.15, + "grad_norm": 2.161744024140962, + "learning_rate": 1.9277297738513415e-05, + "loss": 1.0512, + "step": 6287 + }, + { + "epoch": 0.15, + "grad_norm": 2.569624358767445, + "learning_rate": 1.9277012904483846e-05, + "loss": 1.1051, + "step": 6288 + }, + { + "epoch": 0.15, + "grad_norm": 2.22370819295741, + "learning_rate": 1.927672801644057e-05, + "loss": 1.0812, + "step": 6289 + }, + { + "epoch": 0.15, + "grad_norm": 2.4064102850011024, + "learning_rate": 1.9276443074385246e-05, + "loss": 1.2938, + "step": 6290 + }, + { + "epoch": 0.15, + "grad_norm": 2.3943992909176934, + "learning_rate": 1.927615807831953e-05, + "loss": 1.0047, + "step": 6291 + }, + { + "epoch": 0.15, + "grad_norm": 1.0818114937625622, + "learning_rate": 1.927587302824509e-05, + "loss": 1.0237, + "step": 6292 + }, + { + "epoch": 0.15, + "grad_norm": 1.946709873710889, + "learning_rate": 1.9275587924163578e-05, + "loss": 1.0352, + "step": 6293 + }, + { + "epoch": 0.15, + "grad_norm": 1.103942821108497, + "learning_rate": 1.927530276607666e-05, + "loss": 1.0876, + "step": 6294 + }, + { + "epoch": 0.15, + "grad_norm": 2.0448557235908362, + "learning_rate": 1.9275017553985986e-05, + "loss": 1.1712, + "step": 6295 + }, + { + "epoch": 0.15, + "grad_norm": 1.999216663566155, + "learning_rate": 1.927473228789323e-05, + "loss": 0.9385, + "step": 6296 + }, + { + "epoch": 0.15, + "grad_norm": 1.2335549645610968, + "learning_rate": 1.927444696780004e-05, + "loss": 0.9829, + "step": 6297 + }, + { + "epoch": 0.15, + "grad_norm": 2.62899245975273, + "learning_rate": 1.9274161593708086e-05, + "loss": 1.2217, + "step": 6298 + }, + { + "epoch": 0.15, + "grad_norm": 1.0846458014085152, + "learning_rate": 1.927387616561903e-05, + "loss": 0.982, + "step": 6299 + }, + { + "epoch": 0.15, + "grad_norm": 1.1918865423938854, + "learning_rate": 1.927359068353453e-05, + "loss": 0.9795, + "step": 6300 + }, + { + "epoch": 0.15, + "grad_norm": 2.077498114094376, + "learning_rate": 1.9273305147456247e-05, + "loss": 0.9422, + "step": 6301 + }, + { + "epoch": 0.15, + "grad_norm": 2.062048350764629, + "learning_rate": 1.9273019557385845e-05, + "loss": 1.1434, + "step": 6302 + }, + { + "epoch": 0.15, + "grad_norm": 2.6305112530543386, + "learning_rate": 1.927273391332499e-05, + "loss": 1.092, + "step": 6303 + }, + { + "epoch": 0.15, + "grad_norm": 2.1776485591389303, + "learning_rate": 1.927244821527534e-05, + "loss": 1.0248, + "step": 6304 + }, + { + "epoch": 0.15, + "grad_norm": 2.4305108897813397, + "learning_rate": 1.9272162463238562e-05, + "loss": 1.0189, + "step": 6305 + }, + { + "epoch": 0.15, + "grad_norm": 2.128054022001411, + "learning_rate": 1.927187665721632e-05, + "loss": 1.1272, + "step": 6306 + }, + { + "epoch": 0.15, + "grad_norm": 2.208542721669602, + "learning_rate": 1.9271590797210276e-05, + "loss": 0.901, + "step": 6307 + }, + { + "epoch": 0.15, + "grad_norm": 2.112327690083037, + "learning_rate": 1.9271304883222095e-05, + "loss": 1.1656, + "step": 6308 + }, + { + "epoch": 0.15, + "grad_norm": 2.321300237353512, + "learning_rate": 1.927101891525344e-05, + "loss": 0.9934, + "step": 6309 + }, + { + "epoch": 0.15, + "grad_norm": 1.9679202749116984, + "learning_rate": 1.927073289330598e-05, + "loss": 1.272, + "step": 6310 + }, + { + "epoch": 0.15, + "grad_norm": 2.071542582718268, + "learning_rate": 1.9270446817381377e-05, + "loss": 0.9021, + "step": 6311 + }, + { + "epoch": 0.15, + "grad_norm": 1.1463679089194434, + "learning_rate": 1.9270160687481296e-05, + "loss": 0.9881, + "step": 6312 + }, + { + "epoch": 0.15, + "grad_norm": 1.8559135636656674, + "learning_rate": 1.926987450360741e-05, + "loss": 1.1511, + "step": 6313 + }, + { + "epoch": 0.15, + "grad_norm": 1.1420551973048132, + "learning_rate": 1.9269588265761374e-05, + "loss": 0.9998, + "step": 6314 + }, + { + "epoch": 0.15, + "grad_norm": 2.5671678131488367, + "learning_rate": 1.9269301973944864e-05, + "loss": 1.0889, + "step": 6315 + }, + { + "epoch": 0.15, + "grad_norm": 2.4587586404760895, + "learning_rate": 1.926901562815954e-05, + "loss": 0.8602, + "step": 6316 + }, + { + "epoch": 0.15, + "grad_norm": 2.0439407104352894, + "learning_rate": 1.9268729228407076e-05, + "loss": 1.0465, + "step": 6317 + }, + { + "epoch": 0.15, + "grad_norm": 1.065244885819613, + "learning_rate": 1.9268442774689132e-05, + "loss": 0.908, + "step": 6318 + }, + { + "epoch": 0.15, + "grad_norm": 1.921759232057178, + "learning_rate": 1.9268156267007383e-05, + "loss": 1.1072, + "step": 6319 + }, + { + "epoch": 0.15, + "grad_norm": 1.1179627694388836, + "learning_rate": 1.9267869705363493e-05, + "loss": 0.9241, + "step": 6320 + }, + { + "epoch": 0.15, + "grad_norm": 2.0293223310146127, + "learning_rate": 1.926758308975913e-05, + "loss": 1.1177, + "step": 6321 + }, + { + "epoch": 0.15, + "grad_norm": 2.4661511955531066, + "learning_rate": 1.9267296420195966e-05, + "loss": 0.9609, + "step": 6322 + }, + { + "epoch": 0.15, + "grad_norm": 1.1605383600149235, + "learning_rate": 1.926700969667567e-05, + "loss": 1.003, + "step": 6323 + }, + { + "epoch": 0.15, + "grad_norm": 2.5162047177635922, + "learning_rate": 1.9266722919199907e-05, + "loss": 0.9758, + "step": 6324 + }, + { + "epoch": 0.15, + "grad_norm": 2.1713171524877493, + "learning_rate": 1.926643608777035e-05, + "loss": 0.9872, + "step": 6325 + }, + { + "epoch": 0.15, + "grad_norm": 1.9228978188410986, + "learning_rate": 1.926614920238867e-05, + "loss": 1.1255, + "step": 6326 + }, + { + "epoch": 0.15, + "grad_norm": 2.612577000622844, + "learning_rate": 1.926586226305653e-05, + "loss": 1.1175, + "step": 6327 + }, + { + "epoch": 0.15, + "grad_norm": 2.037623756318318, + "learning_rate": 1.9265575269775612e-05, + "loss": 1.0533, + "step": 6328 + }, + { + "epoch": 0.15, + "grad_norm": 1.9992475593073333, + "learning_rate": 1.926528822254758e-05, + "loss": 0.997, + "step": 6329 + }, + { + "epoch": 0.15, + "grad_norm": 1.9855962609300026, + "learning_rate": 1.9265001121374108e-05, + "loss": 1.1044, + "step": 6330 + }, + { + "epoch": 0.15, + "grad_norm": 2.0231823802947932, + "learning_rate": 1.9264713966256864e-05, + "loss": 1.0817, + "step": 6331 + }, + { + "epoch": 0.15, + "grad_norm": 2.4397587731688524, + "learning_rate": 1.9264426757197523e-05, + "loss": 1.0258, + "step": 6332 + }, + { + "epoch": 0.15, + "grad_norm": 1.9228897265303255, + "learning_rate": 1.9264139494197755e-05, + "loss": 0.9053, + "step": 6333 + }, + { + "epoch": 0.15, + "grad_norm": 2.1814511181821303, + "learning_rate": 1.9263852177259234e-05, + "loss": 0.9851, + "step": 6334 + }, + { + "epoch": 0.15, + "grad_norm": 1.9570260367876084, + "learning_rate": 1.9263564806383637e-05, + "loss": 0.9891, + "step": 6335 + }, + { + "epoch": 0.15, + "grad_norm": 2.839199281629642, + "learning_rate": 1.9263277381572628e-05, + "loss": 0.9201, + "step": 6336 + }, + { + "epoch": 0.15, + "grad_norm": 2.269220942701763, + "learning_rate": 1.926298990282789e-05, + "loss": 1.0624, + "step": 6337 + }, + { + "epoch": 0.15, + "grad_norm": 2.008947602280761, + "learning_rate": 1.9262702370151087e-05, + "loss": 0.9408, + "step": 6338 + }, + { + "epoch": 0.15, + "grad_norm": 1.2545665360954614, + "learning_rate": 1.92624147835439e-05, + "loss": 0.9976, + "step": 6339 + }, + { + "epoch": 0.15, + "grad_norm": 2.8459814879856773, + "learning_rate": 1.9262127143008e-05, + "loss": 1.1263, + "step": 6340 + }, + { + "epoch": 0.15, + "grad_norm": 1.1181218476634849, + "learning_rate": 1.9261839448545066e-05, + "loss": 1.0363, + "step": 6341 + }, + { + "epoch": 0.15, + "grad_norm": 1.188374498215562, + "learning_rate": 1.926155170015677e-05, + "loss": 0.9334, + "step": 6342 + }, + { + "epoch": 0.15, + "grad_norm": 2.3104390351353086, + "learning_rate": 1.926126389784479e-05, + "loss": 1.14, + "step": 6343 + }, + { + "epoch": 0.15, + "grad_norm": 2.7276030582298465, + "learning_rate": 1.9260976041610797e-05, + "loss": 1.1238, + "step": 6344 + }, + { + "epoch": 0.15, + "grad_norm": 1.0899987571987013, + "learning_rate": 1.926068813145647e-05, + "loss": 0.9689, + "step": 6345 + }, + { + "epoch": 0.15, + "grad_norm": 1.115004050664329, + "learning_rate": 1.9260400167383485e-05, + "loss": 0.9073, + "step": 6346 + }, + { + "epoch": 0.15, + "grad_norm": 2.067903008377982, + "learning_rate": 1.9260112149393517e-05, + "loss": 1.0778, + "step": 6347 + }, + { + "epoch": 0.15, + "grad_norm": 1.1581996108334702, + "learning_rate": 1.9259824077488248e-05, + "loss": 0.97, + "step": 6348 + }, + { + "epoch": 0.15, + "grad_norm": 1.1152660355775408, + "learning_rate": 1.925953595166935e-05, + "loss": 0.9821, + "step": 6349 + }, + { + "epoch": 0.15, + "grad_norm": 2.180851316185181, + "learning_rate": 1.92592477719385e-05, + "loss": 0.9897, + "step": 6350 + }, + { + "epoch": 0.15, + "grad_norm": 1.9958482889293538, + "learning_rate": 1.925895953829738e-05, + "loss": 1.0202, + "step": 6351 + }, + { + "epoch": 0.15, + "grad_norm": 2.2286999359399218, + "learning_rate": 1.9258671250747665e-05, + "loss": 1.0635, + "step": 6352 + }, + { + "epoch": 0.15, + "grad_norm": 2.5360682130950916, + "learning_rate": 1.9258382909291037e-05, + "loss": 1.015, + "step": 6353 + }, + { + "epoch": 0.15, + "grad_norm": 2.078564406042805, + "learning_rate": 1.925809451392917e-05, + "loss": 1.0721, + "step": 6354 + }, + { + "epoch": 0.15, + "grad_norm": 2.035580636469951, + "learning_rate": 1.9257806064663746e-05, + "loss": 1.0516, + "step": 6355 + }, + { + "epoch": 0.15, + "grad_norm": 2.1257351262141215, + "learning_rate": 1.9257517561496443e-05, + "loss": 1.1337, + "step": 6356 + }, + { + "epoch": 0.15, + "grad_norm": 1.1549992260095605, + "learning_rate": 1.9257229004428944e-05, + "loss": 0.9907, + "step": 6357 + }, + { + "epoch": 0.15, + "grad_norm": 2.289166741267907, + "learning_rate": 1.925694039346293e-05, + "loss": 1.0968, + "step": 6358 + }, + { + "epoch": 0.15, + "grad_norm": 2.2942179513387027, + "learning_rate": 1.9256651728600073e-05, + "loss": 1.1587, + "step": 6359 + }, + { + "epoch": 0.15, + "grad_norm": 2.2206907232151294, + "learning_rate": 1.925636300984206e-05, + "loss": 1.0292, + "step": 6360 + }, + { + "epoch": 0.15, + "grad_norm": 1.9618941895758693, + "learning_rate": 1.9256074237190573e-05, + "loss": 1.0705, + "step": 6361 + }, + { + "epoch": 0.15, + "grad_norm": 2.413609137582511, + "learning_rate": 1.925578541064729e-05, + "loss": 1.0419, + "step": 6362 + }, + { + "epoch": 0.15, + "grad_norm": 2.058576517286322, + "learning_rate": 1.9255496530213895e-05, + "loss": 1.0643, + "step": 6363 + }, + { + "epoch": 0.15, + "grad_norm": 2.004127098252055, + "learning_rate": 1.925520759589207e-05, + "loss": 0.9449, + "step": 6364 + }, + { + "epoch": 0.15, + "grad_norm": 1.9606985832613113, + "learning_rate": 1.9254918607683493e-05, + "loss": 1.0232, + "step": 6365 + }, + { + "epoch": 0.15, + "grad_norm": 2.4164775444977713, + "learning_rate": 1.925462956558985e-05, + "loss": 1.1337, + "step": 6366 + }, + { + "epoch": 0.15, + "grad_norm": 2.069821689228007, + "learning_rate": 1.9254340469612826e-05, + "loss": 0.9, + "step": 6367 + }, + { + "epoch": 0.15, + "grad_norm": 2.3547127501522014, + "learning_rate": 1.9254051319754103e-05, + "loss": 1.0893, + "step": 6368 + }, + { + "epoch": 0.15, + "grad_norm": 1.9877505439629486, + "learning_rate": 1.925376211601536e-05, + "loss": 1.2046, + "step": 6369 + }, + { + "epoch": 0.15, + "grad_norm": 1.8317221737181029, + "learning_rate": 1.9253472858398284e-05, + "loss": 1.0037, + "step": 6370 + }, + { + "epoch": 0.15, + "grad_norm": 2.813440711184601, + "learning_rate": 1.925318354690456e-05, + "loss": 1.104, + "step": 6371 + }, + { + "epoch": 0.15, + "grad_norm": 1.1745681151909617, + "learning_rate": 1.9252894181535874e-05, + "loss": 0.9599, + "step": 6372 + }, + { + "epoch": 0.15, + "grad_norm": 2.3467116476748764, + "learning_rate": 1.9252604762293908e-05, + "loss": 0.8803, + "step": 6373 + }, + { + "epoch": 0.15, + "grad_norm": 2.0324910203531097, + "learning_rate": 1.9252315289180348e-05, + "loss": 0.9584, + "step": 6374 + }, + { + "epoch": 0.15, + "grad_norm": 2.146354842876747, + "learning_rate": 1.9252025762196878e-05, + "loss": 1.0624, + "step": 6375 + }, + { + "epoch": 0.15, + "grad_norm": 1.9861676387249714, + "learning_rate": 1.9251736181345186e-05, + "loss": 1.055, + "step": 6376 + }, + { + "epoch": 0.15, + "grad_norm": 2.770134371060247, + "learning_rate": 1.925144654662696e-05, + "loss": 1.0392, + "step": 6377 + }, + { + "epoch": 0.15, + "grad_norm": 2.202480952180547, + "learning_rate": 1.9251156858043877e-05, + "loss": 1.1319, + "step": 6378 + }, + { + "epoch": 0.15, + "grad_norm": 2.0108162333428954, + "learning_rate": 1.925086711559763e-05, + "loss": 1.2352, + "step": 6379 + }, + { + "epoch": 0.15, + "grad_norm": 2.308058860902183, + "learning_rate": 1.9250577319289912e-05, + "loss": 1.0891, + "step": 6380 + }, + { + "epoch": 0.15, + "grad_norm": 2.036256578054935, + "learning_rate": 1.92502874691224e-05, + "loss": 0.9976, + "step": 6381 + }, + { + "epoch": 0.15, + "grad_norm": 2.2342313245324683, + "learning_rate": 1.9249997565096783e-05, + "loss": 1.1157, + "step": 6382 + }, + { + "epoch": 0.15, + "grad_norm": 2.180030442847454, + "learning_rate": 1.9249707607214757e-05, + "loss": 1.0848, + "step": 6383 + }, + { + "epoch": 0.15, + "grad_norm": 2.2780805049516117, + "learning_rate": 1.9249417595478005e-05, + "loss": 0.9978, + "step": 6384 + }, + { + "epoch": 0.15, + "grad_norm": 2.6053626672110966, + "learning_rate": 1.9249127529888212e-05, + "loss": 1.0516, + "step": 6385 + }, + { + "epoch": 0.15, + "grad_norm": 2.0815992307475417, + "learning_rate": 1.924883741044707e-05, + "loss": 1.1043, + "step": 6386 + }, + { + "epoch": 0.15, + "grad_norm": 2.3634312210201514, + "learning_rate": 1.924854723715627e-05, + "loss": 0.9578, + "step": 6387 + }, + { + "epoch": 0.15, + "grad_norm": 2.1118236774253254, + "learning_rate": 1.92482570100175e-05, + "loss": 1.0992, + "step": 6388 + }, + { + "epoch": 0.15, + "grad_norm": 2.336680217668822, + "learning_rate": 1.924796672903245e-05, + "loss": 1.051, + "step": 6389 + }, + { + "epoch": 0.15, + "grad_norm": 1.1196266850974754, + "learning_rate": 1.924767639420281e-05, + "loss": 0.9914, + "step": 6390 + }, + { + "epoch": 0.15, + "grad_norm": 2.203898239116957, + "learning_rate": 1.924738600553027e-05, + "loss": 1.1234, + "step": 6391 + }, + { + "epoch": 0.15, + "grad_norm": 2.2772218224191327, + "learning_rate": 1.924709556301652e-05, + "loss": 1.099, + "step": 6392 + }, + { + "epoch": 0.15, + "grad_norm": 3.766468035591343, + "learning_rate": 1.9246805066663255e-05, + "loss": 0.9999, + "step": 6393 + }, + { + "epoch": 0.15, + "grad_norm": 2.180157936467426, + "learning_rate": 1.9246514516472157e-05, + "loss": 1.1893, + "step": 6394 + }, + { + "epoch": 0.15, + "grad_norm": 2.539812850164654, + "learning_rate": 1.9246223912444933e-05, + "loss": 1.0399, + "step": 6395 + }, + { + "epoch": 0.15, + "grad_norm": 2.213072002655031, + "learning_rate": 1.924593325458326e-05, + "loss": 0.9637, + "step": 6396 + }, + { + "epoch": 0.15, + "grad_norm": 2.125623986547644, + "learning_rate": 1.9245642542888836e-05, + "loss": 1.0444, + "step": 6397 + }, + { + "epoch": 0.15, + "grad_norm": 1.9401370274424516, + "learning_rate": 1.9245351777363355e-05, + "loss": 1.1876, + "step": 6398 + }, + { + "epoch": 0.15, + "grad_norm": 2.124468915416524, + "learning_rate": 1.924506095800851e-05, + "loss": 1.0121, + "step": 6399 + }, + { + "epoch": 0.15, + "grad_norm": 2.0704807263754414, + "learning_rate": 1.9244770084825993e-05, + "loss": 1.0102, + "step": 6400 + }, + { + "epoch": 0.15, + "grad_norm": 3.1934991841669107, + "learning_rate": 1.92444791578175e-05, + "loss": 0.9963, + "step": 6401 + }, + { + "epoch": 0.15, + "grad_norm": 2.174492001590806, + "learning_rate": 1.924418817698472e-05, + "loss": 1.1079, + "step": 6402 + }, + { + "epoch": 0.15, + "grad_norm": 2.131472304283468, + "learning_rate": 1.9243897142329347e-05, + "loss": 1.0134, + "step": 6403 + }, + { + "epoch": 0.15, + "grad_norm": 2.15956348336344, + "learning_rate": 1.9243606053853084e-05, + "loss": 1.1575, + "step": 6404 + }, + { + "epoch": 0.15, + "grad_norm": 2.1101914459380646, + "learning_rate": 1.9243314911557614e-05, + "loss": 0.9548, + "step": 6405 + }, + { + "epoch": 0.15, + "grad_norm": 1.9553996562705385, + "learning_rate": 1.9243023715444646e-05, + "loss": 1.0249, + "step": 6406 + }, + { + "epoch": 0.15, + "grad_norm": 1.876085040189484, + "learning_rate": 1.9242732465515862e-05, + "loss": 1.0672, + "step": 6407 + }, + { + "epoch": 0.15, + "grad_norm": 2.16870241778148, + "learning_rate": 1.9242441161772965e-05, + "loss": 1.0304, + "step": 6408 + }, + { + "epoch": 0.15, + "grad_norm": 3.0640447787006306, + "learning_rate": 1.9242149804217648e-05, + "loss": 1.0924, + "step": 6409 + }, + { + "epoch": 0.15, + "grad_norm": 2.3368937395441987, + "learning_rate": 1.9241858392851612e-05, + "loss": 0.9702, + "step": 6410 + }, + { + "epoch": 0.15, + "grad_norm": 2.0617718595479504, + "learning_rate": 1.9241566927676548e-05, + "loss": 1.0355, + "step": 6411 + }, + { + "epoch": 0.15, + "grad_norm": 2.5206904463387816, + "learning_rate": 1.9241275408694156e-05, + "loss": 1.0572, + "step": 6412 + }, + { + "epoch": 0.15, + "grad_norm": 2.0244907124495737, + "learning_rate": 1.924098383590613e-05, + "loss": 1.0223, + "step": 6413 + }, + { + "epoch": 0.15, + "grad_norm": 2.0133344867585774, + "learning_rate": 1.9240692209314176e-05, + "loss": 1.1591, + "step": 6414 + }, + { + "epoch": 0.15, + "grad_norm": 2.2931086630148627, + "learning_rate": 1.9240400528919983e-05, + "loss": 1.0678, + "step": 6415 + }, + { + "epoch": 0.15, + "grad_norm": 2.5200615148367658, + "learning_rate": 1.9240108794725252e-05, + "loss": 1.1753, + "step": 6416 + }, + { + "epoch": 0.15, + "grad_norm": 2.0769593231653234, + "learning_rate": 1.9239817006731684e-05, + "loss": 0.9258, + "step": 6417 + }, + { + "epoch": 0.15, + "grad_norm": 2.934194634711879, + "learning_rate": 1.9239525164940977e-05, + "loss": 0.9623, + "step": 6418 + }, + { + "epoch": 0.15, + "grad_norm": 1.1612688055921097, + "learning_rate": 1.9239233269354828e-05, + "loss": 1.106, + "step": 6419 + }, + { + "epoch": 0.15, + "grad_norm": 2.4496604011314553, + "learning_rate": 1.9238941319974938e-05, + "loss": 1.0362, + "step": 6420 + }, + { + "epoch": 0.15, + "grad_norm": 1.0829392795016164, + "learning_rate": 1.9238649316803004e-05, + "loss": 1.0084, + "step": 6421 + }, + { + "epoch": 0.15, + "grad_norm": 2.2859947782337855, + "learning_rate": 1.923835725984073e-05, + "loss": 1.1711, + "step": 6422 + }, + { + "epoch": 0.15, + "grad_norm": 2.780626787976259, + "learning_rate": 1.923806514908982e-05, + "loss": 1.0385, + "step": 6423 + }, + { + "epoch": 0.15, + "grad_norm": 2.00229381918746, + "learning_rate": 1.9237772984551966e-05, + "loss": 0.9962, + "step": 6424 + }, + { + "epoch": 0.15, + "grad_norm": 2.1727603890010547, + "learning_rate": 1.9237480766228872e-05, + "loss": 1.0076, + "step": 6425 + }, + { + "epoch": 0.15, + "grad_norm": 2.0175504929514303, + "learning_rate": 1.923718849412224e-05, + "loss": 1.016, + "step": 6426 + }, + { + "epoch": 0.15, + "grad_norm": 2.055902053439668, + "learning_rate": 1.9236896168233775e-05, + "loss": 1.1933, + "step": 6427 + }, + { + "epoch": 0.15, + "grad_norm": 2.1639373936580464, + "learning_rate": 1.9236603788565176e-05, + "loss": 1.0767, + "step": 6428 + }, + { + "epoch": 0.15, + "grad_norm": 2.0907422504502904, + "learning_rate": 1.9236311355118144e-05, + "loss": 1.0252, + "step": 6429 + }, + { + "epoch": 0.15, + "grad_norm": 1.1313439450637344, + "learning_rate": 1.9236018867894382e-05, + "loss": 0.9744, + "step": 6430 + }, + { + "epoch": 0.15, + "grad_norm": 2.839032609743652, + "learning_rate": 1.9235726326895595e-05, + "loss": 1.1117, + "step": 6431 + }, + { + "epoch": 0.15, + "grad_norm": 2.043624029775946, + "learning_rate": 1.9235433732123488e-05, + "loss": 1.0177, + "step": 6432 + }, + { + "epoch": 0.15, + "grad_norm": 1.9535609876865156, + "learning_rate": 1.923514108357976e-05, + "loss": 1.0315, + "step": 6433 + }, + { + "epoch": 0.15, + "grad_norm": 2.2883108437976287, + "learning_rate": 1.9234848381266117e-05, + "loss": 1.058, + "step": 6434 + }, + { + "epoch": 0.15, + "grad_norm": 4.4590031319130565, + "learning_rate": 1.9234555625184265e-05, + "loss": 1.027, + "step": 6435 + }, + { + "epoch": 0.15, + "grad_norm": 2.0335372209202767, + "learning_rate": 1.9234262815335903e-05, + "loss": 1.0727, + "step": 6436 + }, + { + "epoch": 0.15, + "grad_norm": 1.919805813851093, + "learning_rate": 1.923396995172274e-05, + "loss": 1.022, + "step": 6437 + }, + { + "epoch": 0.15, + "grad_norm": 2.007783764209581, + "learning_rate": 1.9233677034346484e-05, + "loss": 1.149, + "step": 6438 + }, + { + "epoch": 0.15, + "grad_norm": 1.1760313181602036, + "learning_rate": 1.923338406320883e-05, + "loss": 0.9951, + "step": 6439 + }, + { + "epoch": 0.15, + "grad_norm": 2.0618275763532004, + "learning_rate": 1.92330910383115e-05, + "loss": 1.1202, + "step": 6440 + }, + { + "epoch": 0.15, + "grad_norm": 2.216472285512754, + "learning_rate": 1.9232797959656184e-05, + "loss": 1.2591, + "step": 6441 + }, + { + "epoch": 0.15, + "grad_norm": 2.008346730529023, + "learning_rate": 1.9232504827244594e-05, + "loss": 1.057, + "step": 6442 + }, + { + "epoch": 0.15, + "grad_norm": 2.513992228394224, + "learning_rate": 1.9232211641078443e-05, + "loss": 1.2777, + "step": 6443 + }, + { + "epoch": 0.15, + "grad_norm": 2.19645762019354, + "learning_rate": 1.9231918401159433e-05, + "loss": 0.9075, + "step": 6444 + }, + { + "epoch": 0.15, + "grad_norm": 2.01094503581455, + "learning_rate": 1.923162510748927e-05, + "loss": 1.1113, + "step": 6445 + }, + { + "epoch": 0.15, + "grad_norm": 2.1388209252914367, + "learning_rate": 1.9231331760069662e-05, + "loss": 1.1222, + "step": 6446 + }, + { + "epoch": 0.15, + "grad_norm": 2.0970706257638256, + "learning_rate": 1.9231038358902317e-05, + "loss": 1.0403, + "step": 6447 + }, + { + "epoch": 0.15, + "grad_norm": 1.9628986805660387, + "learning_rate": 1.9230744903988943e-05, + "loss": 0.9176, + "step": 6448 + }, + { + "epoch": 0.15, + "grad_norm": 3.527563242060876, + "learning_rate": 1.9230451395331253e-05, + "loss": 0.9848, + "step": 6449 + }, + { + "epoch": 0.15, + "grad_norm": 2.4482930467806043, + "learning_rate": 1.9230157832930953e-05, + "loss": 0.9012, + "step": 6450 + }, + { + "epoch": 0.15, + "grad_norm": 3.3819044827320353, + "learning_rate": 1.922986421678975e-05, + "loss": 1.1057, + "step": 6451 + }, + { + "epoch": 0.15, + "grad_norm": 2.5231844054322994, + "learning_rate": 1.9229570546909354e-05, + "loss": 1.0817, + "step": 6452 + }, + { + "epoch": 0.15, + "grad_norm": 2.577482681563974, + "learning_rate": 1.922927682329148e-05, + "loss": 1.0261, + "step": 6453 + }, + { + "epoch": 0.15, + "grad_norm": 2.316600657116274, + "learning_rate": 1.922898304593783e-05, + "loss": 0.9515, + "step": 6454 + }, + { + "epoch": 0.15, + "grad_norm": 2.1662532885190324, + "learning_rate": 1.922868921485012e-05, + "loss": 1.0677, + "step": 6455 + }, + { + "epoch": 0.15, + "grad_norm": 2.2654700226230626, + "learning_rate": 1.9228395330030064e-05, + "loss": 1.0436, + "step": 6456 + }, + { + "epoch": 0.15, + "grad_norm": 2.0008977421390584, + "learning_rate": 1.9228101391479366e-05, + "loss": 1.1272, + "step": 6457 + }, + { + "epoch": 0.15, + "grad_norm": 1.0854468567625595, + "learning_rate": 1.922780739919974e-05, + "loss": 0.999, + "step": 6458 + }, + { + "epoch": 0.15, + "grad_norm": 1.9539667386360695, + "learning_rate": 1.9227513353192895e-05, + "loss": 0.9391, + "step": 6459 + }, + { + "epoch": 0.15, + "grad_norm": 2.383864092945085, + "learning_rate": 1.9227219253460552e-05, + "loss": 1.1248, + "step": 6460 + }, + { + "epoch": 0.15, + "grad_norm": 2.2543240468617403, + "learning_rate": 1.9226925100004415e-05, + "loss": 1.0175, + "step": 6461 + }, + { + "epoch": 0.15, + "grad_norm": 1.097266841901673, + "learning_rate": 1.9226630892826196e-05, + "loss": 1.0151, + "step": 6462 + }, + { + "epoch": 0.15, + "grad_norm": 1.9761217764100787, + "learning_rate": 1.9226336631927613e-05, + "loss": 1.1656, + "step": 6463 + }, + { + "epoch": 0.15, + "grad_norm": 1.9289105914390734, + "learning_rate": 1.9226042317310377e-05, + "loss": 1.0763, + "step": 6464 + }, + { + "epoch": 0.15, + "grad_norm": 1.1749095786684167, + "learning_rate": 1.9225747948976205e-05, + "loss": 1.023, + "step": 6465 + }, + { + "epoch": 0.15, + "grad_norm": 1.9789291422303328, + "learning_rate": 1.92254535269268e-05, + "loss": 0.9829, + "step": 6466 + }, + { + "epoch": 0.15, + "grad_norm": 2.369288302624237, + "learning_rate": 1.9225159051163888e-05, + "loss": 1.1554, + "step": 6467 + }, + { + "epoch": 0.15, + "grad_norm": 2.0385041911266835, + "learning_rate": 1.9224864521689182e-05, + "loss": 1.0409, + "step": 6468 + }, + { + "epoch": 0.15, + "grad_norm": 1.8600190048142524, + "learning_rate": 1.922456993850439e-05, + "loss": 1.1224, + "step": 6469 + }, + { + "epoch": 0.15, + "grad_norm": 2.2269134172799876, + "learning_rate": 1.9224275301611234e-05, + "loss": 0.9977, + "step": 6470 + }, + { + "epoch": 0.15, + "grad_norm": 2.044765258623859, + "learning_rate": 1.9223980611011424e-05, + "loss": 1.085, + "step": 6471 + }, + { + "epoch": 0.15, + "grad_norm": 2.2905928200983596, + "learning_rate": 1.922368586670668e-05, + "loss": 1.1281, + "step": 6472 + }, + { + "epoch": 0.15, + "grad_norm": 2.2486235670587016, + "learning_rate": 1.9223391068698717e-05, + "loss": 1.2275, + "step": 6473 + }, + { + "epoch": 0.15, + "grad_norm": 2.101385760032446, + "learning_rate": 1.9223096216989253e-05, + "loss": 1.1691, + "step": 6474 + }, + { + "epoch": 0.15, + "grad_norm": 1.959298018414065, + "learning_rate": 1.922280131158e-05, + "loss": 1.1828, + "step": 6475 + }, + { + "epoch": 0.15, + "grad_norm": 1.9619542953710407, + "learning_rate": 1.9222506352472677e-05, + "loss": 0.9998, + "step": 6476 + }, + { + "epoch": 0.15, + "grad_norm": 2.1279477435549965, + "learning_rate": 1.9222211339669005e-05, + "loss": 1.0247, + "step": 6477 + }, + { + "epoch": 0.15, + "grad_norm": 2.1806584470038537, + "learning_rate": 1.9221916273170697e-05, + "loss": 1.0845, + "step": 6478 + }, + { + "epoch": 0.15, + "grad_norm": 2.1671648771130414, + "learning_rate": 1.9221621152979476e-05, + "loss": 1.1003, + "step": 6479 + }, + { + "epoch": 0.15, + "grad_norm": 2.6047494271024365, + "learning_rate": 1.9221325979097053e-05, + "loss": 1.0143, + "step": 6480 + }, + { + "epoch": 0.15, + "grad_norm": 1.1116780114098421, + "learning_rate": 1.9221030751525154e-05, + "loss": 1.0468, + "step": 6481 + }, + { + "epoch": 0.15, + "grad_norm": 2.444690592412731, + "learning_rate": 1.9220735470265492e-05, + "loss": 1.0543, + "step": 6482 + }, + { + "epoch": 0.15, + "grad_norm": 1.085432234591325, + "learning_rate": 1.9220440135319794e-05, + "loss": 0.948, + "step": 6483 + }, + { + "epoch": 0.15, + "grad_norm": 2.147050152547517, + "learning_rate": 1.9220144746689768e-05, + "loss": 1.1215, + "step": 6484 + }, + { + "epoch": 0.15, + "grad_norm": 3.1244443547712577, + "learning_rate": 1.9219849304377142e-05, + "loss": 1.0716, + "step": 6485 + }, + { + "epoch": 0.15, + "grad_norm": 2.0191808350207348, + "learning_rate": 1.9219553808383636e-05, + "loss": 1.0438, + "step": 6486 + }, + { + "epoch": 0.15, + "grad_norm": 2.0949411316773827, + "learning_rate": 1.921925825871097e-05, + "loss": 0.9344, + "step": 6487 + }, + { + "epoch": 0.15, + "grad_norm": 2.0512184730573417, + "learning_rate": 1.921896265536086e-05, + "loss": 1.2052, + "step": 6488 + }, + { + "epoch": 0.15, + "grad_norm": 2.3986976244444866, + "learning_rate": 1.9218666998335034e-05, + "loss": 1.1338, + "step": 6489 + }, + { + "epoch": 0.15, + "grad_norm": 2.2700450064578823, + "learning_rate": 1.921837128763521e-05, + "loss": 1.0861, + "step": 6490 + }, + { + "epoch": 0.15, + "grad_norm": 2.1177575529009927, + "learning_rate": 1.9218075523263104e-05, + "loss": 1.0687, + "step": 6491 + }, + { + "epoch": 0.15, + "grad_norm": 2.4377009107627674, + "learning_rate": 1.9217779705220452e-05, + "loss": 0.938, + "step": 6492 + }, + { + "epoch": 0.15, + "grad_norm": 2.234016695358254, + "learning_rate": 1.9217483833508962e-05, + "loss": 1.0538, + "step": 6493 + }, + { + "epoch": 0.15, + "grad_norm": 2.2835783030176415, + "learning_rate": 1.921718790813037e-05, + "loss": 1.1081, + "step": 6494 + }, + { + "epoch": 0.15, + "grad_norm": 2.069831050254993, + "learning_rate": 1.9216891929086382e-05, + "loss": 0.9589, + "step": 6495 + }, + { + "epoch": 0.15, + "grad_norm": 1.8610178849263126, + "learning_rate": 1.9216595896378737e-05, + "loss": 1.0681, + "step": 6496 + }, + { + "epoch": 0.15, + "grad_norm": 2.0579973917602534, + "learning_rate": 1.9216299810009156e-05, + "loss": 1.0022, + "step": 6497 + }, + { + "epoch": 0.15, + "grad_norm": 2.1342082960453097, + "learning_rate": 1.9216003669979356e-05, + "loss": 0.9449, + "step": 6498 + }, + { + "epoch": 0.15, + "grad_norm": 2.1618064400654755, + "learning_rate": 1.9215707476291064e-05, + "loss": 1.0361, + "step": 6499 + }, + { + "epoch": 0.15, + "grad_norm": 1.8874970764688561, + "learning_rate": 1.9215411228946005e-05, + "loss": 1.0506, + "step": 6500 + }, + { + "epoch": 0.15, + "grad_norm": 2.1508467036422663, + "learning_rate": 1.9215114927945905e-05, + "loss": 1.1526, + "step": 6501 + }, + { + "epoch": 0.15, + "grad_norm": 2.2982483311335415, + "learning_rate": 1.9214818573292492e-05, + "loss": 1.1165, + "step": 6502 + }, + { + "epoch": 0.15, + "grad_norm": 1.8913884293927206, + "learning_rate": 1.9214522164987484e-05, + "loss": 1.1302, + "step": 6503 + }, + { + "epoch": 0.15, + "grad_norm": 2.146325008934166, + "learning_rate": 1.921422570303261e-05, + "loss": 1.0857, + "step": 6504 + }, + { + "epoch": 0.15, + "grad_norm": 3.23463627842882, + "learning_rate": 1.9213929187429602e-05, + "loss": 0.988, + "step": 6505 + }, + { + "epoch": 0.15, + "grad_norm": 2.7205753803900046, + "learning_rate": 1.9213632618180175e-05, + "loss": 1.07, + "step": 6506 + }, + { + "epoch": 0.15, + "grad_norm": 2.1385915012803136, + "learning_rate": 1.9213335995286065e-05, + "loss": 1.046, + "step": 6507 + }, + { + "epoch": 0.15, + "grad_norm": 2.304281361403674, + "learning_rate": 1.9213039318748995e-05, + "loss": 1.0334, + "step": 6508 + }, + { + "epoch": 0.15, + "grad_norm": 1.8848748107523143, + "learning_rate": 1.9212742588570694e-05, + "loss": 1.0877, + "step": 6509 + }, + { + "epoch": 0.15, + "grad_norm": 2.308475341882726, + "learning_rate": 1.921244580475289e-05, + "loss": 1.037, + "step": 6510 + }, + { + "epoch": 0.15, + "grad_norm": 2.346005646460255, + "learning_rate": 1.9212148967297307e-05, + "loss": 1.1484, + "step": 6511 + }, + { + "epoch": 0.15, + "grad_norm": 2.1562235373189673, + "learning_rate": 1.921185207620568e-05, + "loss": 0.9005, + "step": 6512 + }, + { + "epoch": 0.15, + "grad_norm": 1.09593203243615, + "learning_rate": 1.9211555131479728e-05, + "loss": 0.9541, + "step": 6513 + }, + { + "epoch": 0.15, + "grad_norm": 2.513405090459764, + "learning_rate": 1.921125813312119e-05, + "loss": 0.9407, + "step": 6514 + }, + { + "epoch": 0.15, + "grad_norm": 2.7090709174131633, + "learning_rate": 1.921096108113179e-05, + "loss": 0.9497, + "step": 6515 + }, + { + "epoch": 0.15, + "grad_norm": 2.171100441354246, + "learning_rate": 1.921066397551326e-05, + "loss": 0.9524, + "step": 6516 + }, + { + "epoch": 0.15, + "grad_norm": 2.2579966583125763, + "learning_rate": 1.9210366816267327e-05, + "loss": 0.9738, + "step": 6517 + }, + { + "epoch": 0.15, + "grad_norm": 1.9742551152938355, + "learning_rate": 1.9210069603395724e-05, + "loss": 1.129, + "step": 6518 + }, + { + "epoch": 0.15, + "grad_norm": 3.5003912604050114, + "learning_rate": 1.920977233690018e-05, + "loss": 1.0913, + "step": 6519 + }, + { + "epoch": 0.15, + "grad_norm": 1.9746953910007086, + "learning_rate": 1.920947501678242e-05, + "loss": 1.1254, + "step": 6520 + }, + { + "epoch": 0.15, + "grad_norm": 1.9265017614841236, + "learning_rate": 1.9209177643044187e-05, + "loss": 1.2279, + "step": 6521 + }, + { + "epoch": 0.15, + "grad_norm": 2.492808330516292, + "learning_rate": 1.9208880215687203e-05, + "loss": 1.0138, + "step": 6522 + }, + { + "epoch": 0.15, + "grad_norm": 1.1711702267991635, + "learning_rate": 1.9208582734713202e-05, + "loss": 0.9377, + "step": 6523 + }, + { + "epoch": 0.15, + "grad_norm": 2.0995105326474537, + "learning_rate": 1.920828520012392e-05, + "loss": 1.01, + "step": 6524 + }, + { + "epoch": 0.15, + "grad_norm": 2.1366118631224986, + "learning_rate": 1.9207987611921085e-05, + "loss": 1.12, + "step": 6525 + }, + { + "epoch": 0.15, + "grad_norm": 2.3658949325711665, + "learning_rate": 1.9207689970106434e-05, + "loss": 1.0375, + "step": 6526 + }, + { + "epoch": 0.15, + "grad_norm": 2.1340451835363243, + "learning_rate": 1.9207392274681692e-05, + "loss": 1.1094, + "step": 6527 + }, + { + "epoch": 0.15, + "grad_norm": 2.423649643294602, + "learning_rate": 1.92070945256486e-05, + "loss": 0.9955, + "step": 6528 + }, + { + "epoch": 0.15, + "grad_norm": 2.103866334176895, + "learning_rate": 1.920679672300889e-05, + "loss": 1.1601, + "step": 6529 + }, + { + "epoch": 0.15, + "grad_norm": 2.161452399910209, + "learning_rate": 1.920649886676429e-05, + "loss": 1.1233, + "step": 6530 + }, + { + "epoch": 0.15, + "grad_norm": 2.064710947767024, + "learning_rate": 1.9206200956916542e-05, + "loss": 0.9747, + "step": 6531 + }, + { + "epoch": 0.15, + "grad_norm": 2.101498409294612, + "learning_rate": 1.9205902993467377e-05, + "loss": 1.1416, + "step": 6532 + }, + { + "epoch": 0.15, + "grad_norm": 2.7913734321469406, + "learning_rate": 1.920560497641853e-05, + "loss": 1.0898, + "step": 6533 + }, + { + "epoch": 0.15, + "grad_norm": 1.990837177908102, + "learning_rate": 1.9205306905771735e-05, + "loss": 0.993, + "step": 6534 + }, + { + "epoch": 0.15, + "grad_norm": 2.061048819436318, + "learning_rate": 1.920500878152873e-05, + "loss": 1.1225, + "step": 6535 + }, + { + "epoch": 0.15, + "grad_norm": 2.2070994407744786, + "learning_rate": 1.920471060369125e-05, + "loss": 1.1095, + "step": 6536 + }, + { + "epoch": 0.15, + "grad_norm": 2.4342956299071448, + "learning_rate": 1.920441237226103e-05, + "loss": 1.0613, + "step": 6537 + }, + { + "epoch": 0.15, + "grad_norm": 1.8589377815842563, + "learning_rate": 1.9204114087239806e-05, + "loss": 1.0934, + "step": 6538 + }, + { + "epoch": 0.15, + "grad_norm": 2.26222425578176, + "learning_rate": 1.920381574862932e-05, + "loss": 1.0151, + "step": 6539 + }, + { + "epoch": 0.15, + "grad_norm": 2.0570456681296774, + "learning_rate": 1.92035173564313e-05, + "loss": 1.1836, + "step": 6540 + }, + { + "epoch": 0.15, + "grad_norm": 2.4861476059907366, + "learning_rate": 1.9203218910647492e-05, + "loss": 0.9796, + "step": 6541 + }, + { + "epoch": 0.15, + "grad_norm": 2.198673542572597, + "learning_rate": 1.920292041127963e-05, + "loss": 0.9718, + "step": 6542 + }, + { + "epoch": 0.15, + "grad_norm": 2.5738129367364317, + "learning_rate": 1.920262185832945e-05, + "loss": 0.993, + "step": 6543 + }, + { + "epoch": 0.15, + "grad_norm": 2.0902309661729435, + "learning_rate": 1.920232325179869e-05, + "loss": 1.0023, + "step": 6544 + }, + { + "epoch": 0.15, + "grad_norm": 1.9215078045499776, + "learning_rate": 1.9202024591689094e-05, + "loss": 1.0234, + "step": 6545 + }, + { + "epoch": 0.15, + "grad_norm": 1.9603524095686995, + "learning_rate": 1.9201725878002394e-05, + "loss": 1.0554, + "step": 6546 + }, + { + "epoch": 0.15, + "grad_norm": 2.227532485231408, + "learning_rate": 1.9201427110740335e-05, + "loss": 1.1509, + "step": 6547 + }, + { + "epoch": 0.15, + "grad_norm": 1.889761716936663, + "learning_rate": 1.9201128289904654e-05, + "loss": 1.1129, + "step": 6548 + }, + { + "epoch": 0.15, + "grad_norm": 2.4309370420001346, + "learning_rate": 1.9200829415497095e-05, + "loss": 1.1128, + "step": 6549 + }, + { + "epoch": 0.15, + "grad_norm": 2.086555674210245, + "learning_rate": 1.920053048751939e-05, + "loss": 1.0654, + "step": 6550 + }, + { + "epoch": 0.15, + "grad_norm": 2.026470234273043, + "learning_rate": 1.9200231505973282e-05, + "loss": 0.975, + "step": 6551 + }, + { + "epoch": 0.15, + "grad_norm": 1.9524937704603158, + "learning_rate": 1.9199932470860516e-05, + "loss": 0.9643, + "step": 6552 + }, + { + "epoch": 0.15, + "grad_norm": 2.1602637932877378, + "learning_rate": 1.919963338218283e-05, + "loss": 0.959, + "step": 6553 + }, + { + "epoch": 0.15, + "grad_norm": 2.1256545167454486, + "learning_rate": 1.9199334239941968e-05, + "loss": 0.9925, + "step": 6554 + }, + { + "epoch": 0.15, + "grad_norm": 4.74981835396976, + "learning_rate": 1.9199035044139666e-05, + "loss": 1.0417, + "step": 6555 + }, + { + "epoch": 0.15, + "grad_norm": 2.4051220625909306, + "learning_rate": 1.919873579477767e-05, + "loss": 1.0532, + "step": 6556 + }, + { + "epoch": 0.15, + "grad_norm": 1.9042279323911462, + "learning_rate": 1.9198436491857725e-05, + "loss": 1.1178, + "step": 6557 + }, + { + "epoch": 0.15, + "grad_norm": 2.113301775822935, + "learning_rate": 1.9198137135381567e-05, + "loss": 0.9403, + "step": 6558 + }, + { + "epoch": 0.15, + "grad_norm": 2.0484069063600305, + "learning_rate": 1.9197837725350944e-05, + "loss": 1.1082, + "step": 6559 + }, + { + "epoch": 0.15, + "grad_norm": 3.296225291963936, + "learning_rate": 1.91975382617676e-05, + "loss": 1.0781, + "step": 6560 + }, + { + "epoch": 0.15, + "grad_norm": 2.047257238244236, + "learning_rate": 1.9197238744633275e-05, + "loss": 1.0569, + "step": 6561 + }, + { + "epoch": 0.15, + "grad_norm": 1.9834615310210788, + "learning_rate": 1.919693917394971e-05, + "loss": 1.1149, + "step": 6562 + }, + { + "epoch": 0.15, + "grad_norm": 1.89427845143295, + "learning_rate": 1.9196639549718655e-05, + "loss": 1.2215, + "step": 6563 + }, + { + "epoch": 0.15, + "grad_norm": 3.5578047714444727, + "learning_rate": 1.919633987194186e-05, + "loss": 1.0433, + "step": 6564 + }, + { + "epoch": 0.15, + "grad_norm": 1.9394574524244375, + "learning_rate": 1.9196040140621056e-05, + "loss": 1.0345, + "step": 6565 + }, + { + "epoch": 0.15, + "grad_norm": 2.528991880401903, + "learning_rate": 1.9195740355757998e-05, + "loss": 1.0827, + "step": 6566 + }, + { + "epoch": 0.15, + "grad_norm": 1.9504933470956887, + "learning_rate": 1.9195440517354424e-05, + "loss": 1.0349, + "step": 6567 + }, + { + "epoch": 0.15, + "grad_norm": 1.9737186801328646, + "learning_rate": 1.9195140625412087e-05, + "loss": 1.1575, + "step": 6568 + }, + { + "epoch": 0.15, + "grad_norm": 1.89001513663886, + "learning_rate": 1.919484067993273e-05, + "loss": 1.0464, + "step": 6569 + }, + { + "epoch": 0.15, + "grad_norm": 2.3440623564995686, + "learning_rate": 1.91945406809181e-05, + "loss": 1.0842, + "step": 6570 + }, + { + "epoch": 0.15, + "grad_norm": 1.9337004765764576, + "learning_rate": 1.919424062836994e-05, + "loss": 0.8967, + "step": 6571 + }, + { + "epoch": 0.15, + "grad_norm": 1.97905062279856, + "learning_rate": 1.9193940522290008e-05, + "loss": 1.1554, + "step": 6572 + }, + { + "epoch": 0.15, + "grad_norm": 2.466146060086931, + "learning_rate": 1.9193640362680038e-05, + "loss": 1.146, + "step": 6573 + }, + { + "epoch": 0.15, + "grad_norm": 3.0034164083889507, + "learning_rate": 1.919334014954178e-05, + "loss": 1.1358, + "step": 6574 + }, + { + "epoch": 0.15, + "grad_norm": 1.9337991891924649, + "learning_rate": 1.919303988287699e-05, + "loss": 1.1163, + "step": 6575 + }, + { + "epoch": 0.15, + "grad_norm": 2.0752369855174564, + "learning_rate": 1.9192739562687412e-05, + "loss": 1.0134, + "step": 6576 + }, + { + "epoch": 0.15, + "grad_norm": 2.0964916399884426, + "learning_rate": 1.919243918897479e-05, + "loss": 1.0084, + "step": 6577 + }, + { + "epoch": 0.15, + "grad_norm": 2.0403511254705227, + "learning_rate": 1.9192138761740883e-05, + "loss": 1.0732, + "step": 6578 + }, + { + "epoch": 0.15, + "grad_norm": 2.096811367154359, + "learning_rate": 1.9191838280987428e-05, + "loss": 1.1211, + "step": 6579 + }, + { + "epoch": 0.16, + "grad_norm": 2.0994821639668695, + "learning_rate": 1.919153774671618e-05, + "loss": 1.095, + "step": 6580 + }, + { + "epoch": 0.16, + "grad_norm": 2.008084280641247, + "learning_rate": 1.919123715892889e-05, + "loss": 1.0237, + "step": 6581 + }, + { + "epoch": 0.16, + "grad_norm": 2.093920027855771, + "learning_rate": 1.919093651762731e-05, + "loss": 1.1792, + "step": 6582 + }, + { + "epoch": 0.16, + "grad_norm": 1.8483016723964716, + "learning_rate": 1.919063582281319e-05, + "loss": 1.0602, + "step": 6583 + }, + { + "epoch": 0.16, + "grad_norm": 2.3700732319209488, + "learning_rate": 1.9190335074488275e-05, + "loss": 1.0075, + "step": 6584 + }, + { + "epoch": 0.16, + "grad_norm": 4.296303901975365, + "learning_rate": 1.9190034272654316e-05, + "loss": 1.031, + "step": 6585 + }, + { + "epoch": 0.16, + "grad_norm": 2.233405057541177, + "learning_rate": 1.9189733417313073e-05, + "loss": 1.0154, + "step": 6586 + }, + { + "epoch": 0.16, + "grad_norm": 2.0427271241238336, + "learning_rate": 1.918943250846629e-05, + "loss": 1.1146, + "step": 6587 + }, + { + "epoch": 0.16, + "grad_norm": 1.8673367090292374, + "learning_rate": 1.9189131546115725e-05, + "loss": 1.127, + "step": 6588 + }, + { + "epoch": 0.16, + "grad_norm": 2.0047558641168224, + "learning_rate": 1.9188830530263124e-05, + "loss": 1.0169, + "step": 6589 + }, + { + "epoch": 0.16, + "grad_norm": 2.167611507030509, + "learning_rate": 1.9188529460910244e-05, + "loss": 0.9604, + "step": 6590 + }, + { + "epoch": 0.16, + "grad_norm": 1.1961463558948517, + "learning_rate": 1.9188228338058835e-05, + "loss": 1.083, + "step": 6591 + }, + { + "epoch": 0.16, + "grad_norm": 1.880106282575667, + "learning_rate": 1.918792716171065e-05, + "loss": 1.0955, + "step": 6592 + }, + { + "epoch": 0.16, + "grad_norm": 1.8456976998970998, + "learning_rate": 1.9187625931867448e-05, + "loss": 1.0523, + "step": 6593 + }, + { + "epoch": 0.16, + "grad_norm": 2.5921157159826618, + "learning_rate": 1.9187324648530977e-05, + "loss": 0.9291, + "step": 6594 + }, + { + "epoch": 0.16, + "grad_norm": 5.175643405698181, + "learning_rate": 1.9187023311702996e-05, + "loss": 0.9128, + "step": 6595 + }, + { + "epoch": 0.16, + "grad_norm": 2.195302235462248, + "learning_rate": 1.9186721921385257e-05, + "loss": 0.9511, + "step": 6596 + }, + { + "epoch": 0.16, + "grad_norm": 3.093695180550552, + "learning_rate": 1.9186420477579508e-05, + "loss": 1.0569, + "step": 6597 + }, + { + "epoch": 0.16, + "grad_norm": 2.0911469932606774, + "learning_rate": 1.918611898028752e-05, + "loss": 1.071, + "step": 6598 + }, + { + "epoch": 0.16, + "grad_norm": 2.6442409474722584, + "learning_rate": 1.9185817429511033e-05, + "loss": 0.9868, + "step": 6599 + }, + { + "epoch": 0.16, + "grad_norm": 2.4737661514383227, + "learning_rate": 1.9185515825251806e-05, + "loss": 1.0496, + "step": 6600 + }, + { + "epoch": 0.16, + "grad_norm": 2.1417113093154514, + "learning_rate": 1.91852141675116e-05, + "loss": 1.0614, + "step": 6601 + }, + { + "epoch": 0.16, + "grad_norm": 2.2419008867655843, + "learning_rate": 1.9184912456292172e-05, + "loss": 0.9646, + "step": 6602 + }, + { + "epoch": 0.16, + "grad_norm": 2.4941219998092174, + "learning_rate": 1.9184610691595273e-05, + "loss": 1.0468, + "step": 6603 + }, + { + "epoch": 0.16, + "grad_norm": 1.099242386296075, + "learning_rate": 1.9184308873422666e-05, + "loss": 0.9238, + "step": 6604 + }, + { + "epoch": 0.16, + "grad_norm": 2.397404014412754, + "learning_rate": 1.9184007001776103e-05, + "loss": 1.0262, + "step": 6605 + }, + { + "epoch": 0.16, + "grad_norm": 2.098253788144121, + "learning_rate": 1.9183705076657343e-05, + "loss": 1.035, + "step": 6606 + }, + { + "epoch": 0.16, + "grad_norm": 2.049273728368686, + "learning_rate": 1.9183403098068145e-05, + "loss": 1.1799, + "step": 6607 + }, + { + "epoch": 0.16, + "grad_norm": 2.496020198836815, + "learning_rate": 1.918310106601027e-05, + "loss": 1.0095, + "step": 6608 + }, + { + "epoch": 0.16, + "grad_norm": 1.1508962616160119, + "learning_rate": 1.9182798980485468e-05, + "loss": 0.9925, + "step": 6609 + }, + { + "epoch": 0.16, + "grad_norm": 2.0518804667386505, + "learning_rate": 1.9182496841495504e-05, + "loss": 1.0307, + "step": 6610 + }, + { + "epoch": 0.16, + "grad_norm": 4.679801781306551, + "learning_rate": 1.9182194649042138e-05, + "loss": 1.1276, + "step": 6611 + }, + { + "epoch": 0.16, + "grad_norm": 2.23620212222864, + "learning_rate": 1.918189240312713e-05, + "loss": 1.1025, + "step": 6612 + }, + { + "epoch": 0.16, + "grad_norm": 2.6752871558385065, + "learning_rate": 1.918159010375223e-05, + "loss": 1.0484, + "step": 6613 + }, + { + "epoch": 0.16, + "grad_norm": 1.9816186049127305, + "learning_rate": 1.9181287750919214e-05, + "loss": 1.0562, + "step": 6614 + }, + { + "epoch": 0.16, + "grad_norm": 3.252139579236138, + "learning_rate": 1.918098534462983e-05, + "loss": 0.9985, + "step": 6615 + }, + { + "epoch": 0.16, + "grad_norm": 2.5809779781816693, + "learning_rate": 1.918068288488584e-05, + "loss": 0.9099, + "step": 6616 + }, + { + "epoch": 0.16, + "grad_norm": 1.0799298111314142, + "learning_rate": 1.918038037168901e-05, + "loss": 0.9561, + "step": 6617 + }, + { + "epoch": 0.16, + "grad_norm": 1.119110890833155, + "learning_rate": 1.9180077805041103e-05, + "loss": 1.0044, + "step": 6618 + }, + { + "epoch": 0.16, + "grad_norm": 2.2410128474737796, + "learning_rate": 1.917977518494387e-05, + "loss": 1.1371, + "step": 6619 + }, + { + "epoch": 0.16, + "grad_norm": 2.117431712648137, + "learning_rate": 1.9179472511399084e-05, + "loss": 1.0295, + "step": 6620 + }, + { + "epoch": 0.16, + "grad_norm": 2.078291171778123, + "learning_rate": 1.91791697844085e-05, + "loss": 1.0541, + "step": 6621 + }, + { + "epoch": 0.16, + "grad_norm": 1.183082184973754, + "learning_rate": 1.917886700397389e-05, + "loss": 0.8854, + "step": 6622 + }, + { + "epoch": 0.16, + "grad_norm": 2.1983012970366436, + "learning_rate": 1.9178564170097004e-05, + "loss": 1.2332, + "step": 6623 + }, + { + "epoch": 0.16, + "grad_norm": 2.5980798544430175, + "learning_rate": 1.9178261282779614e-05, + "loss": 1.0949, + "step": 6624 + }, + { + "epoch": 0.16, + "grad_norm": 1.840055620290869, + "learning_rate": 1.9177958342023478e-05, + "loss": 0.9041, + "step": 6625 + }, + { + "epoch": 0.16, + "grad_norm": 2.2007899250210614, + "learning_rate": 1.9177655347830368e-05, + "loss": 1.0894, + "step": 6626 + }, + { + "epoch": 0.16, + "grad_norm": 1.940406331697847, + "learning_rate": 1.9177352300202038e-05, + "loss": 0.9985, + "step": 6627 + }, + { + "epoch": 0.16, + "grad_norm": 2.3998679239612466, + "learning_rate": 1.917704919914026e-05, + "loss": 1.1886, + "step": 6628 + }, + { + "epoch": 0.16, + "grad_norm": 1.801607852667973, + "learning_rate": 1.9176746044646796e-05, + "loss": 1.1796, + "step": 6629 + }, + { + "epoch": 0.16, + "grad_norm": 2.1184167689209263, + "learning_rate": 1.917644283672341e-05, + "loss": 1.0091, + "step": 6630 + }, + { + "epoch": 0.16, + "grad_norm": 1.1152481320012009, + "learning_rate": 1.9176139575371867e-05, + "loss": 1.0016, + "step": 6631 + }, + { + "epoch": 0.16, + "grad_norm": 1.9979345495406229, + "learning_rate": 1.9175836260593937e-05, + "loss": 0.9822, + "step": 6632 + }, + { + "epoch": 0.16, + "grad_norm": 1.1019018752486929, + "learning_rate": 1.9175532892391383e-05, + "loss": 1.0214, + "step": 6633 + }, + { + "epoch": 0.16, + "grad_norm": 1.1181737507895988, + "learning_rate": 1.9175229470765972e-05, + "loss": 1.0133, + "step": 6634 + }, + { + "epoch": 0.16, + "grad_norm": 2.2381702095136538, + "learning_rate": 1.917492599571947e-05, + "loss": 1.1363, + "step": 6635 + }, + { + "epoch": 0.16, + "grad_norm": 1.9773487967336691, + "learning_rate": 1.9174622467253644e-05, + "loss": 1.0818, + "step": 6636 + }, + { + "epoch": 0.16, + "grad_norm": 2.0056181592449795, + "learning_rate": 1.917431888537026e-05, + "loss": 1.1371, + "step": 6637 + }, + { + "epoch": 0.16, + "grad_norm": 2.1749113694358644, + "learning_rate": 1.917401525007109e-05, + "loss": 1.0073, + "step": 6638 + }, + { + "epoch": 0.16, + "grad_norm": 2.0892708619280698, + "learning_rate": 1.9173711561357895e-05, + "loss": 1.0402, + "step": 6639 + }, + { + "epoch": 0.16, + "grad_norm": 1.9721038551279384, + "learning_rate": 1.917340781923245e-05, + "loss": 0.9908, + "step": 6640 + }, + { + "epoch": 0.16, + "grad_norm": 1.957449286963971, + "learning_rate": 1.917310402369652e-05, + "loss": 1.0485, + "step": 6641 + }, + { + "epoch": 0.16, + "grad_norm": 1.0974905408215916, + "learning_rate": 1.9172800174751873e-05, + "loss": 1.0315, + "step": 6642 + }, + { + "epoch": 0.16, + "grad_norm": 2.103442739155501, + "learning_rate": 1.9172496272400282e-05, + "loss": 1.0182, + "step": 6643 + }, + { + "epoch": 0.16, + "grad_norm": 2.36000956903387, + "learning_rate": 1.917219231664351e-05, + "loss": 1.0899, + "step": 6644 + }, + { + "epoch": 0.16, + "grad_norm": 2.408659536053029, + "learning_rate": 1.917188830748333e-05, + "loss": 1.096, + "step": 6645 + }, + { + "epoch": 0.16, + "grad_norm": 2.2002530009973285, + "learning_rate": 1.9171584244921516e-05, + "loss": 1.0705, + "step": 6646 + }, + { + "epoch": 0.16, + "grad_norm": 3.080584951826532, + "learning_rate": 1.9171280128959834e-05, + "loss": 0.9889, + "step": 6647 + }, + { + "epoch": 0.16, + "grad_norm": 2.0628051568753767, + "learning_rate": 1.9170975959600053e-05, + "loss": 1.0211, + "step": 6648 + }, + { + "epoch": 0.16, + "grad_norm": 2.281402145171994, + "learning_rate": 1.917067173684395e-05, + "loss": 0.9918, + "step": 6649 + }, + { + "epoch": 0.16, + "grad_norm": 1.9071642341283426, + "learning_rate": 1.917036746069329e-05, + "loss": 1.0581, + "step": 6650 + }, + { + "epoch": 0.16, + "grad_norm": 2.0456248967705073, + "learning_rate": 1.917006313114985e-05, + "loss": 1.0466, + "step": 6651 + }, + { + "epoch": 0.16, + "grad_norm": 2.203402984443756, + "learning_rate": 1.9169758748215396e-05, + "loss": 0.9848, + "step": 6652 + }, + { + "epoch": 0.16, + "grad_norm": 1.8740588395913038, + "learning_rate": 1.9169454311891705e-05, + "loss": 0.9973, + "step": 6653 + }, + { + "epoch": 0.16, + "grad_norm": 2.0596311416885875, + "learning_rate": 1.916914982218055e-05, + "loss": 0.9417, + "step": 6654 + }, + { + "epoch": 0.16, + "grad_norm": 2.2210915599758363, + "learning_rate": 1.9168845279083697e-05, + "loss": 0.9909, + "step": 6655 + }, + { + "epoch": 0.16, + "grad_norm": 3.0193036716611807, + "learning_rate": 1.9168540682602928e-05, + "loss": 1.1003, + "step": 6656 + }, + { + "epoch": 0.16, + "grad_norm": 2.15435805173353, + "learning_rate": 1.916823603274001e-05, + "loss": 1.1211, + "step": 6657 + }, + { + "epoch": 0.16, + "grad_norm": 2.293716401785027, + "learning_rate": 1.916793132949672e-05, + "loss": 1.0654, + "step": 6658 + }, + { + "epoch": 0.16, + "grad_norm": 2.6414483133451507, + "learning_rate": 1.916762657287483e-05, + "loss": 1.0814, + "step": 6659 + }, + { + "epoch": 0.16, + "grad_norm": 1.9111088470794515, + "learning_rate": 1.916732176287612e-05, + "loss": 1.1387, + "step": 6660 + }, + { + "epoch": 0.16, + "grad_norm": 1.9760184584145857, + "learning_rate": 1.9167016899502357e-05, + "loss": 1.0308, + "step": 6661 + }, + { + "epoch": 0.16, + "grad_norm": 2.4750327174957483, + "learning_rate": 1.916671198275532e-05, + "loss": 1.043, + "step": 6662 + }, + { + "epoch": 0.16, + "grad_norm": 1.8640598458152693, + "learning_rate": 1.9166407012636785e-05, + "loss": 1.0042, + "step": 6663 + }, + { + "epoch": 0.16, + "grad_norm": 2.302129788257265, + "learning_rate": 1.9166101989148527e-05, + "loss": 1.043, + "step": 6664 + }, + { + "epoch": 0.16, + "grad_norm": 2.196460419089518, + "learning_rate": 1.9165796912292317e-05, + "loss": 1.1231, + "step": 6665 + }, + { + "epoch": 0.16, + "grad_norm": 1.9938432660222432, + "learning_rate": 1.9165491782069942e-05, + "loss": 1.0068, + "step": 6666 + }, + { + "epoch": 0.16, + "grad_norm": 2.263330819609591, + "learning_rate": 1.9165186598483166e-05, + "loss": 1.1556, + "step": 6667 + }, + { + "epoch": 0.16, + "grad_norm": 2.1905078297884826, + "learning_rate": 1.9164881361533775e-05, + "loss": 0.9201, + "step": 6668 + }, + { + "epoch": 0.16, + "grad_norm": 2.2839501769363384, + "learning_rate": 1.9164576071223542e-05, + "loss": 1.2651, + "step": 6669 + }, + { + "epoch": 0.16, + "grad_norm": 1.979481358303264, + "learning_rate": 1.9164270727554247e-05, + "loss": 0.9939, + "step": 6670 + }, + { + "epoch": 0.16, + "grad_norm": 1.89852121226471, + "learning_rate": 1.9163965330527664e-05, + "loss": 1.1855, + "step": 6671 + }, + { + "epoch": 0.16, + "grad_norm": 2.081320117271032, + "learning_rate": 1.9163659880145575e-05, + "loss": 1.1413, + "step": 6672 + }, + { + "epoch": 0.16, + "grad_norm": 3.903400303946994, + "learning_rate": 1.9163354376409757e-05, + "loss": 1.1707, + "step": 6673 + }, + { + "epoch": 0.16, + "grad_norm": 2.260906000361125, + "learning_rate": 1.9163048819321987e-05, + "loss": 1.0536, + "step": 6674 + }, + { + "epoch": 0.16, + "grad_norm": 1.8776643992666429, + "learning_rate": 1.916274320888405e-05, + "loss": 0.9383, + "step": 6675 + }, + { + "epoch": 0.16, + "grad_norm": 2.0523354450527194, + "learning_rate": 1.916243754509772e-05, + "loss": 1.0872, + "step": 6676 + }, + { + "epoch": 0.16, + "grad_norm": 2.0017041847086783, + "learning_rate": 1.9162131827964774e-05, + "loss": 1.1127, + "step": 6677 + }, + { + "epoch": 0.16, + "grad_norm": 2.1746612917153985, + "learning_rate": 1.9161826057487e-05, + "loss": 1.0221, + "step": 6678 + }, + { + "epoch": 0.16, + "grad_norm": 2.189420623372166, + "learning_rate": 1.9161520233666172e-05, + "loss": 0.9962, + "step": 6679 + }, + { + "epoch": 0.16, + "grad_norm": 2.337140944076565, + "learning_rate": 1.9161214356504072e-05, + "loss": 1.0948, + "step": 6680 + }, + { + "epoch": 0.16, + "grad_norm": 2.073295914200434, + "learning_rate": 1.916090842600248e-05, + "loss": 1.1411, + "step": 6681 + }, + { + "epoch": 0.16, + "grad_norm": 1.9550906658127665, + "learning_rate": 1.9160602442163182e-05, + "loss": 1.0543, + "step": 6682 + }, + { + "epoch": 0.16, + "grad_norm": 2.15967171142792, + "learning_rate": 1.916029640498796e-05, + "loss": 0.9842, + "step": 6683 + }, + { + "epoch": 0.16, + "grad_norm": 2.303896095387545, + "learning_rate": 1.915999031447859e-05, + "loss": 1.0821, + "step": 6684 + }, + { + "epoch": 0.16, + "grad_norm": 1.1391208524786953, + "learning_rate": 1.915968417063685e-05, + "loss": 0.9775, + "step": 6685 + }, + { + "epoch": 0.16, + "grad_norm": 2.068831670945592, + "learning_rate": 1.9159377973464536e-05, + "loss": 1.1964, + "step": 6686 + }, + { + "epoch": 0.16, + "grad_norm": 2.1244921263622283, + "learning_rate": 1.9159071722963422e-05, + "loss": 1.0717, + "step": 6687 + }, + { + "epoch": 0.16, + "grad_norm": 2.3108730217981663, + "learning_rate": 1.915876541913529e-05, + "loss": 0.9452, + "step": 6688 + }, + { + "epoch": 0.16, + "grad_norm": 2.480473046130117, + "learning_rate": 1.9158459061981928e-05, + "loss": 1.0392, + "step": 6689 + }, + { + "epoch": 0.16, + "grad_norm": 2.1966012921050857, + "learning_rate": 1.9158152651505116e-05, + "loss": 1.1675, + "step": 6690 + }, + { + "epoch": 0.16, + "grad_norm": 2.9510210514344406, + "learning_rate": 1.9157846187706642e-05, + "loss": 0.9934, + "step": 6691 + }, + { + "epoch": 0.16, + "grad_norm": 1.9126989256494469, + "learning_rate": 1.9157539670588287e-05, + "loss": 0.8422, + "step": 6692 + }, + { + "epoch": 0.16, + "grad_norm": 2.160114577328604, + "learning_rate": 1.915723310015184e-05, + "loss": 0.9835, + "step": 6693 + }, + { + "epoch": 0.16, + "grad_norm": 2.193465880129063, + "learning_rate": 1.915692647639908e-05, + "loss": 1.085, + "step": 6694 + }, + { + "epoch": 0.16, + "grad_norm": 1.973083516154335, + "learning_rate": 1.9156619799331796e-05, + "loss": 1.0903, + "step": 6695 + }, + { + "epoch": 0.16, + "grad_norm": 2.192725822531133, + "learning_rate": 1.9156313068951772e-05, + "loss": 1.1563, + "step": 6696 + }, + { + "epoch": 0.16, + "grad_norm": 2.0919354408024717, + "learning_rate": 1.9156006285260792e-05, + "loss": 1.029, + "step": 6697 + }, + { + "epoch": 0.16, + "grad_norm": 1.938502642314365, + "learning_rate": 1.9155699448260648e-05, + "loss": 1.0078, + "step": 6698 + }, + { + "epoch": 0.16, + "grad_norm": 2.200347377450967, + "learning_rate": 1.9155392557953126e-05, + "loss": 0.9341, + "step": 6699 + }, + { + "epoch": 0.16, + "grad_norm": 1.0990796714749111, + "learning_rate": 1.9155085614340004e-05, + "loss": 0.955, + "step": 6700 + }, + { + "epoch": 0.16, + "grad_norm": 1.8876501085115391, + "learning_rate": 1.915477861742308e-05, + "loss": 0.9248, + "step": 6701 + }, + { + "epoch": 0.16, + "grad_norm": 2.2788283590600247, + "learning_rate": 1.9154471567204132e-05, + "loss": 0.8735, + "step": 6702 + }, + { + "epoch": 0.16, + "grad_norm": 2.2559286506345244, + "learning_rate": 1.9154164463684953e-05, + "loss": 1.0624, + "step": 6703 + }, + { + "epoch": 0.16, + "grad_norm": 2.1068600401963273, + "learning_rate": 1.915385730686733e-05, + "loss": 1.1199, + "step": 6704 + }, + { + "epoch": 0.16, + "grad_norm": 2.161737582241363, + "learning_rate": 1.9153550096753056e-05, + "loss": 1.1082, + "step": 6705 + }, + { + "epoch": 0.16, + "grad_norm": 2.341962547582338, + "learning_rate": 1.9153242833343912e-05, + "loss": 1.1106, + "step": 6706 + }, + { + "epoch": 0.16, + "grad_norm": 2.2069919837986407, + "learning_rate": 1.915293551664169e-05, + "loss": 1.155, + "step": 6707 + }, + { + "epoch": 0.16, + "grad_norm": 2.5258702827002586, + "learning_rate": 1.9152628146648183e-05, + "loss": 0.9983, + "step": 6708 + }, + { + "epoch": 0.16, + "grad_norm": 3.120768267498001, + "learning_rate": 1.9152320723365176e-05, + "loss": 1.0645, + "step": 6709 + }, + { + "epoch": 0.16, + "grad_norm": 2.0110036251347263, + "learning_rate": 1.9152013246794455e-05, + "loss": 1.0978, + "step": 6710 + }, + { + "epoch": 0.16, + "grad_norm": 2.284869884007463, + "learning_rate": 1.915170571693782e-05, + "loss": 1.2173, + "step": 6711 + }, + { + "epoch": 0.16, + "grad_norm": 2.0439675039890157, + "learning_rate": 1.915139813379706e-05, + "loss": 1.0085, + "step": 6712 + }, + { + "epoch": 0.16, + "grad_norm": 2.237182173003249, + "learning_rate": 1.9151090497373955e-05, + "loss": 1.0016, + "step": 6713 + }, + { + "epoch": 0.16, + "grad_norm": 2.275627804948, + "learning_rate": 1.915078280767031e-05, + "loss": 0.9835, + "step": 6714 + }, + { + "epoch": 0.16, + "grad_norm": 2.45795581462923, + "learning_rate": 1.9150475064687907e-05, + "loss": 0.9983, + "step": 6715 + }, + { + "epoch": 0.16, + "grad_norm": 1.1226736912371351, + "learning_rate": 1.9150167268428544e-05, + "loss": 0.9967, + "step": 6716 + }, + { + "epoch": 0.16, + "grad_norm": 1.2372314612857966, + "learning_rate": 1.9149859418894007e-05, + "loss": 1.0452, + "step": 6717 + }, + { + "epoch": 0.16, + "grad_norm": 1.943072023305478, + "learning_rate": 1.9149551516086095e-05, + "loss": 0.9853, + "step": 6718 + }, + { + "epoch": 0.16, + "grad_norm": 1.9515535502299188, + "learning_rate": 1.9149243560006598e-05, + "loss": 1.2235, + "step": 6719 + }, + { + "epoch": 0.16, + "grad_norm": 2.268463826956561, + "learning_rate": 1.9148935550657306e-05, + "loss": 0.9458, + "step": 6720 + }, + { + "epoch": 0.16, + "grad_norm": 1.8426740419421155, + "learning_rate": 1.9148627488040014e-05, + "loss": 1.2116, + "step": 6721 + }, + { + "epoch": 0.16, + "grad_norm": 2.26060489130529, + "learning_rate": 1.914831937215652e-05, + "loss": 1.0336, + "step": 6722 + }, + { + "epoch": 0.16, + "grad_norm": 2.3754009395432507, + "learning_rate": 1.9148011203008607e-05, + "loss": 1.1186, + "step": 6723 + }, + { + "epoch": 0.16, + "grad_norm": 2.200969333646966, + "learning_rate": 1.9147702980598085e-05, + "loss": 1.0869, + "step": 6724 + }, + { + "epoch": 0.16, + "grad_norm": 2.106044144490394, + "learning_rate": 1.9147394704926734e-05, + "loss": 0.9353, + "step": 6725 + }, + { + "epoch": 0.16, + "grad_norm": 3.469022038422973, + "learning_rate": 1.914708637599636e-05, + "loss": 1.0538, + "step": 6726 + }, + { + "epoch": 0.16, + "grad_norm": 3.3954242464111286, + "learning_rate": 1.914677799380875e-05, + "loss": 1.0259, + "step": 6727 + }, + { + "epoch": 0.16, + "grad_norm": 2.0179784311648636, + "learning_rate": 1.9146469558365703e-05, + "loss": 1.0094, + "step": 6728 + }, + { + "epoch": 0.16, + "grad_norm": 2.6622522963671487, + "learning_rate": 1.9146161069669015e-05, + "loss": 0.9639, + "step": 6729 + }, + { + "epoch": 0.16, + "grad_norm": 1.1076692934683514, + "learning_rate": 1.914585252772048e-05, + "loss": 0.9247, + "step": 6730 + }, + { + "epoch": 0.16, + "grad_norm": 1.1203773782718864, + "learning_rate": 1.9145543932521898e-05, + "loss": 0.9259, + "step": 6731 + }, + { + "epoch": 0.16, + "grad_norm": 1.767856697002195, + "learning_rate": 1.9145235284075064e-05, + "loss": 1.0712, + "step": 6732 + }, + { + "epoch": 0.16, + "grad_norm": 2.107941957966845, + "learning_rate": 1.9144926582381777e-05, + "loss": 1.0029, + "step": 6733 + }, + { + "epoch": 0.16, + "grad_norm": 1.8155009644707294, + "learning_rate": 1.9144617827443826e-05, + "loss": 0.9507, + "step": 6734 + }, + { + "epoch": 0.16, + "grad_norm": 1.942129686216119, + "learning_rate": 1.914430901926302e-05, + "loss": 1.0131, + "step": 6735 + }, + { + "epoch": 0.16, + "grad_norm": 2.005928141795071, + "learning_rate": 1.9144000157841146e-05, + "loss": 1.0305, + "step": 6736 + }, + { + "epoch": 0.16, + "grad_norm": 2.023835533232899, + "learning_rate": 1.9143691243180015e-05, + "loss": 1.0327, + "step": 6737 + }, + { + "epoch": 0.16, + "grad_norm": 2.063897517351362, + "learning_rate": 1.9143382275281416e-05, + "loss": 1.1452, + "step": 6738 + }, + { + "epoch": 0.16, + "grad_norm": 1.9973205290155118, + "learning_rate": 1.914307325414715e-05, + "loss": 0.8821, + "step": 6739 + }, + { + "epoch": 0.16, + "grad_norm": 2.0380059840833513, + "learning_rate": 1.914276417977902e-05, + "loss": 0.9655, + "step": 6740 + }, + { + "epoch": 0.16, + "grad_norm": 1.9806177957399405, + "learning_rate": 1.9142455052178817e-05, + "loss": 1.0749, + "step": 6741 + }, + { + "epoch": 0.16, + "grad_norm": 2.044268832185499, + "learning_rate": 1.914214587134835e-05, + "loss": 1.0942, + "step": 6742 + }, + { + "epoch": 0.16, + "grad_norm": 2.0787995924552938, + "learning_rate": 1.9141836637289414e-05, + "loss": 1.1443, + "step": 6743 + }, + { + "epoch": 0.16, + "grad_norm": 1.1432222663785925, + "learning_rate": 1.9141527350003812e-05, + "loss": 0.9152, + "step": 6744 + }, + { + "epoch": 0.16, + "grad_norm": 1.2116504552715925, + "learning_rate": 1.9141218009493344e-05, + "loss": 0.9699, + "step": 6745 + }, + { + "epoch": 0.16, + "grad_norm": 2.383884511140024, + "learning_rate": 1.9140908615759807e-05, + "loss": 1.0644, + "step": 6746 + }, + { + "epoch": 0.16, + "grad_norm": 3.045303093225695, + "learning_rate": 1.914059916880501e-05, + "loss": 1.1273, + "step": 6747 + }, + { + "epoch": 0.16, + "grad_norm": 1.1254786766122367, + "learning_rate": 1.9140289668630753e-05, + "loss": 0.9945, + "step": 6748 + }, + { + "epoch": 0.16, + "grad_norm": 2.113136389332193, + "learning_rate": 1.9139980115238827e-05, + "loss": 0.9941, + "step": 6749 + }, + { + "epoch": 0.16, + "grad_norm": 2.1820198287043517, + "learning_rate": 1.913967050863105e-05, + "loss": 1.136, + "step": 6750 + }, + { + "epoch": 0.16, + "grad_norm": 1.8868790263140391, + "learning_rate": 1.9139360848809215e-05, + "loss": 1.1145, + "step": 6751 + }, + { + "epoch": 0.16, + "grad_norm": 2.1041292315676623, + "learning_rate": 1.913905113577513e-05, + "loss": 1.1118, + "step": 6752 + }, + { + "epoch": 0.16, + "grad_norm": 2.5634199119847727, + "learning_rate": 1.9138741369530597e-05, + "loss": 1.1135, + "step": 6753 + }, + { + "epoch": 0.16, + "grad_norm": 1.9751945013297139, + "learning_rate": 1.9138431550077414e-05, + "loss": 1.0508, + "step": 6754 + }, + { + "epoch": 0.16, + "grad_norm": 2.1212876039862985, + "learning_rate": 1.913812167741739e-05, + "loss": 1.0571, + "step": 6755 + }, + { + "epoch": 0.16, + "grad_norm": 2.3852904824543932, + "learning_rate": 1.9137811751552332e-05, + "loss": 1.2238, + "step": 6756 + }, + { + "epoch": 0.16, + "grad_norm": 2.810324727206735, + "learning_rate": 1.913750177248404e-05, + "loss": 0.9917, + "step": 6757 + }, + { + "epoch": 0.16, + "grad_norm": 3.0804981824848405, + "learning_rate": 1.913719174021432e-05, + "loss": 1.082, + "step": 6758 + }, + { + "epoch": 0.16, + "grad_norm": 2.1724046755512596, + "learning_rate": 1.9136881654744973e-05, + "loss": 1.0018, + "step": 6759 + }, + { + "epoch": 0.16, + "grad_norm": 1.9694405591825184, + "learning_rate": 1.913657151607781e-05, + "loss": 1.118, + "step": 6760 + }, + { + "epoch": 0.16, + "grad_norm": 2.01511204889077, + "learning_rate": 1.913626132421464e-05, + "loss": 1.0918, + "step": 6761 + }, + { + "epoch": 0.16, + "grad_norm": 2.095416875058571, + "learning_rate": 1.913595107915726e-05, + "loss": 1.1096, + "step": 6762 + }, + { + "epoch": 0.16, + "grad_norm": 2.485011007644337, + "learning_rate": 1.913564078090748e-05, + "loss": 1.082, + "step": 6763 + }, + { + "epoch": 0.16, + "grad_norm": 2.044131797215262, + "learning_rate": 1.9135330429467105e-05, + "loss": 1.1067, + "step": 6764 + }, + { + "epoch": 0.16, + "grad_norm": 1.1549849295052244, + "learning_rate": 1.913502002483795e-05, + "loss": 1.0124, + "step": 6765 + }, + { + "epoch": 0.16, + "grad_norm": 2.0241830814105426, + "learning_rate": 1.9134709567021808e-05, + "loss": 1.1422, + "step": 6766 + }, + { + "epoch": 0.16, + "grad_norm": 2.125682314937945, + "learning_rate": 1.9134399056020504e-05, + "loss": 0.9924, + "step": 6767 + }, + { + "epoch": 0.16, + "grad_norm": 1.960522676661723, + "learning_rate": 1.913408849183583e-05, + "loss": 1.1187, + "step": 6768 + }, + { + "epoch": 0.16, + "grad_norm": 2.0974693974328673, + "learning_rate": 1.9133777874469605e-05, + "loss": 1.1118, + "step": 6769 + }, + { + "epoch": 0.16, + "grad_norm": 1.8318991513046718, + "learning_rate": 1.913346720392363e-05, + "loss": 1.1181, + "step": 6770 + }, + { + "epoch": 0.16, + "grad_norm": 2.3429915603698825, + "learning_rate": 1.913315648019972e-05, + "loss": 1.0466, + "step": 6771 + }, + { + "epoch": 0.16, + "grad_norm": 2.13941737380706, + "learning_rate": 1.9132845703299683e-05, + "loss": 1.1622, + "step": 6772 + }, + { + "epoch": 0.16, + "grad_norm": 1.7833686178752848, + "learning_rate": 1.9132534873225323e-05, + "loss": 1.0388, + "step": 6773 + }, + { + "epoch": 0.16, + "grad_norm": 2.183526554698772, + "learning_rate": 1.9132223989978457e-05, + "loss": 0.9928, + "step": 6774 + }, + { + "epoch": 0.16, + "grad_norm": 2.2013268168187743, + "learning_rate": 1.913191305356089e-05, + "loss": 1.1524, + "step": 6775 + }, + { + "epoch": 0.16, + "grad_norm": 2.0607716374502454, + "learning_rate": 1.9131602063974434e-05, + "loss": 1.0031, + "step": 6776 + }, + { + "epoch": 0.16, + "grad_norm": 2.1303383271457954, + "learning_rate": 1.9131291021220902e-05, + "loss": 1.0356, + "step": 6777 + }, + { + "epoch": 0.16, + "grad_norm": 2.039955596485498, + "learning_rate": 1.91309799253021e-05, + "loss": 1.1581, + "step": 6778 + }, + { + "epoch": 0.16, + "grad_norm": 2.213575328776643, + "learning_rate": 1.9130668776219843e-05, + "loss": 0.9588, + "step": 6779 + }, + { + "epoch": 0.16, + "grad_norm": 2.5205769060146803, + "learning_rate": 1.913035757397594e-05, + "loss": 0.9173, + "step": 6780 + }, + { + "epoch": 0.16, + "grad_norm": 2.407446529617934, + "learning_rate": 1.913004631857221e-05, + "loss": 0.9921, + "step": 6781 + }, + { + "epoch": 0.16, + "grad_norm": 2.2727222782098417, + "learning_rate": 1.9129735010010452e-05, + "loss": 1.18, + "step": 6782 + }, + { + "epoch": 0.16, + "grad_norm": 2.1441234339177604, + "learning_rate": 1.9129423648292492e-05, + "loss": 1.1999, + "step": 6783 + }, + { + "epoch": 0.16, + "grad_norm": 1.9499529354015364, + "learning_rate": 1.9129112233420138e-05, + "loss": 1.0156, + "step": 6784 + }, + { + "epoch": 0.16, + "grad_norm": 1.1089421333600125, + "learning_rate": 1.9128800765395196e-05, + "loss": 0.9868, + "step": 6785 + }, + { + "epoch": 0.16, + "grad_norm": 3.367422860006278, + "learning_rate": 1.9128489244219492e-05, + "loss": 1.2138, + "step": 6786 + }, + { + "epoch": 0.16, + "grad_norm": 2.2683604456334243, + "learning_rate": 1.9128177669894832e-05, + "loss": 1.0427, + "step": 6787 + }, + { + "epoch": 0.16, + "grad_norm": 2.3049596133866634, + "learning_rate": 1.9127866042423033e-05, + "loss": 0.9543, + "step": 6788 + }, + { + "epoch": 0.16, + "grad_norm": 1.9082692679271729, + "learning_rate": 1.9127554361805902e-05, + "loss": 1.1303, + "step": 6789 + }, + { + "epoch": 0.16, + "grad_norm": 2.3341508555592676, + "learning_rate": 1.9127242628045268e-05, + "loss": 1.074, + "step": 6790 + }, + { + "epoch": 0.16, + "grad_norm": 2.599500769127761, + "learning_rate": 1.912693084114293e-05, + "loss": 1.0424, + "step": 6791 + }, + { + "epoch": 0.16, + "grad_norm": 2.177493713962937, + "learning_rate": 1.9126619001100715e-05, + "loss": 1.0721, + "step": 6792 + }, + { + "epoch": 0.16, + "grad_norm": 2.130774919936761, + "learning_rate": 1.9126307107920433e-05, + "loss": 1.0527, + "step": 6793 + }, + { + "epoch": 0.16, + "grad_norm": 1.8327041840925495, + "learning_rate": 1.9125995161603904e-05, + "loss": 1.1147, + "step": 6794 + }, + { + "epoch": 0.16, + "grad_norm": 1.952934596079096, + "learning_rate": 1.912568316215294e-05, + "loss": 0.9815, + "step": 6795 + }, + { + "epoch": 0.16, + "grad_norm": 1.9782703224996925, + "learning_rate": 1.9125371109569362e-05, + "loss": 1.1245, + "step": 6796 + }, + { + "epoch": 0.16, + "grad_norm": 2.2068166455439218, + "learning_rate": 1.912505900385498e-05, + "loss": 1.1789, + "step": 6797 + }, + { + "epoch": 0.16, + "grad_norm": 2.2051958156386995, + "learning_rate": 1.912474684501162e-05, + "loss": 1.1408, + "step": 6798 + }, + { + "epoch": 0.16, + "grad_norm": 2.252023821615992, + "learning_rate": 1.9124434633041097e-05, + "loss": 1.1841, + "step": 6799 + }, + { + "epoch": 0.16, + "grad_norm": 2.1134907300023187, + "learning_rate": 1.9124122367945223e-05, + "loss": 0.996, + "step": 6800 + }, + { + "epoch": 0.16, + "grad_norm": 2.10789896672884, + "learning_rate": 1.912381004972582e-05, + "loss": 1.1031, + "step": 6801 + }, + { + "epoch": 0.16, + "grad_norm": 2.349266982955085, + "learning_rate": 1.9123497678384706e-05, + "loss": 1.0705, + "step": 6802 + }, + { + "epoch": 0.16, + "grad_norm": 2.0024014759431132, + "learning_rate": 1.91231852539237e-05, + "loss": 0.9769, + "step": 6803 + }, + { + "epoch": 0.16, + "grad_norm": 2.532904481875946, + "learning_rate": 1.9122872776344622e-05, + "loss": 0.9849, + "step": 6804 + }, + { + "epoch": 0.16, + "grad_norm": 2.1584484511105626, + "learning_rate": 1.912256024564929e-05, + "loss": 1.1449, + "step": 6805 + }, + { + "epoch": 0.16, + "grad_norm": 2.1371537847511815, + "learning_rate": 1.9122247661839524e-05, + "loss": 1.074, + "step": 6806 + }, + { + "epoch": 0.16, + "grad_norm": 2.086072446689737, + "learning_rate": 1.9121935024917144e-05, + "loss": 1.0748, + "step": 6807 + }, + { + "epoch": 0.16, + "grad_norm": 1.1243631455873395, + "learning_rate": 1.9121622334883973e-05, + "loss": 0.9827, + "step": 6808 + }, + { + "epoch": 0.16, + "grad_norm": 2.0955289129913406, + "learning_rate": 1.9121309591741822e-05, + "loss": 1.0628, + "step": 6809 + }, + { + "epoch": 0.16, + "grad_norm": 2.240676698402502, + "learning_rate": 1.9120996795492525e-05, + "loss": 1.1376, + "step": 6810 + }, + { + "epoch": 0.16, + "grad_norm": 2.797049435794744, + "learning_rate": 1.9120683946137894e-05, + "loss": 1.0688, + "step": 6811 + }, + { + "epoch": 0.16, + "grad_norm": 2.177804373356026, + "learning_rate": 1.9120371043679756e-05, + "loss": 1.0879, + "step": 6812 + }, + { + "epoch": 0.16, + "grad_norm": 2.360386144383151, + "learning_rate": 1.912005808811993e-05, + "loss": 1.0279, + "step": 6813 + }, + { + "epoch": 0.16, + "grad_norm": 2.2821896623102105, + "learning_rate": 1.9119745079460238e-05, + "loss": 1.0736, + "step": 6814 + }, + { + "epoch": 0.16, + "grad_norm": 2.356426988483803, + "learning_rate": 1.91194320177025e-05, + "loss": 1.0818, + "step": 6815 + }, + { + "epoch": 0.16, + "grad_norm": 2.05056617131232, + "learning_rate": 1.9119118902848546e-05, + "loss": 1.105, + "step": 6816 + }, + { + "epoch": 0.16, + "grad_norm": 1.1514809671954982, + "learning_rate": 1.9118805734900192e-05, + "loss": 0.9673, + "step": 6817 + }, + { + "epoch": 0.16, + "grad_norm": 2.327755983600929, + "learning_rate": 1.9118492513859264e-05, + "loss": 1.1477, + "step": 6818 + }, + { + "epoch": 0.16, + "grad_norm": 1.9580361763761915, + "learning_rate": 1.911817923972759e-05, + "loss": 1.1645, + "step": 6819 + }, + { + "epoch": 0.16, + "grad_norm": 2.140862005945968, + "learning_rate": 1.9117865912506984e-05, + "loss": 1.1187, + "step": 6820 + }, + { + "epoch": 0.16, + "grad_norm": 2.320552692981876, + "learning_rate": 1.9117552532199278e-05, + "loss": 0.9339, + "step": 6821 + }, + { + "epoch": 0.16, + "grad_norm": 1.8115736627942212, + "learning_rate": 1.9117239098806296e-05, + "loss": 1.0267, + "step": 6822 + }, + { + "epoch": 0.16, + "grad_norm": 2.2899648291892953, + "learning_rate": 1.911692561232986e-05, + "loss": 1.169, + "step": 6823 + }, + { + "epoch": 0.16, + "grad_norm": 1.1906808495485621, + "learning_rate": 1.9116612072771798e-05, + "loss": 1.0197, + "step": 6824 + }, + { + "epoch": 0.16, + "grad_norm": 2.472485758077291, + "learning_rate": 1.911629848013393e-05, + "loss": 0.9957, + "step": 6825 + }, + { + "epoch": 0.16, + "grad_norm": 2.1940775790485554, + "learning_rate": 1.911598483441809e-05, + "loss": 1.0616, + "step": 6826 + }, + { + "epoch": 0.16, + "grad_norm": 2.149021255016562, + "learning_rate": 1.91156711356261e-05, + "loss": 1.0366, + "step": 6827 + }, + { + "epoch": 0.16, + "grad_norm": 2.0696347328305253, + "learning_rate": 1.911535738375979e-05, + "loss": 1.0299, + "step": 6828 + }, + { + "epoch": 0.16, + "grad_norm": 2.2145373871655565, + "learning_rate": 1.9115043578820978e-05, + "loss": 1.0182, + "step": 6829 + }, + { + "epoch": 0.16, + "grad_norm": 2.0169293751415256, + "learning_rate": 1.9114729720811497e-05, + "loss": 1.0292, + "step": 6830 + }, + { + "epoch": 0.16, + "grad_norm": 1.8320563128191611, + "learning_rate": 1.9114415809733176e-05, + "loss": 1.0601, + "step": 6831 + }, + { + "epoch": 0.16, + "grad_norm": 2.1252896537924935, + "learning_rate": 1.911410184558784e-05, + "loss": 1.1003, + "step": 6832 + }, + { + "epoch": 0.16, + "grad_norm": 1.213683146437298, + "learning_rate": 1.911378782837732e-05, + "loss": 1.0288, + "step": 6833 + }, + { + "epoch": 0.16, + "grad_norm": 1.985225282247199, + "learning_rate": 1.9113473758103435e-05, + "loss": 1.0076, + "step": 6834 + }, + { + "epoch": 0.16, + "grad_norm": 1.1845367893513539, + "learning_rate": 1.9113159634768025e-05, + "loss": 0.9484, + "step": 6835 + }, + { + "epoch": 0.16, + "grad_norm": 2.056772970789399, + "learning_rate": 1.9112845458372914e-05, + "loss": 1.2408, + "step": 6836 + }, + { + "epoch": 0.16, + "grad_norm": 1.80800633026059, + "learning_rate": 1.9112531228919933e-05, + "loss": 1.0492, + "step": 6837 + }, + { + "epoch": 0.16, + "grad_norm": 1.8766713910923594, + "learning_rate": 1.911221694641091e-05, + "loss": 1.2522, + "step": 6838 + }, + { + "epoch": 0.16, + "grad_norm": 2.311815596120184, + "learning_rate": 1.9111902610847674e-05, + "loss": 1.0664, + "step": 6839 + }, + { + "epoch": 0.16, + "grad_norm": 1.8616542255932436, + "learning_rate": 1.9111588222232056e-05, + "loss": 1.1652, + "step": 6840 + }, + { + "epoch": 0.16, + "grad_norm": 2.0820326845368102, + "learning_rate": 1.9111273780565886e-05, + "loss": 0.9935, + "step": 6841 + }, + { + "epoch": 0.16, + "grad_norm": 1.8433679210307512, + "learning_rate": 1.9110959285850995e-05, + "loss": 1.0687, + "step": 6842 + }, + { + "epoch": 0.16, + "grad_norm": 2.030893579949997, + "learning_rate": 1.9110644738089216e-05, + "loss": 1.0356, + "step": 6843 + }, + { + "epoch": 0.16, + "grad_norm": 2.186623320517276, + "learning_rate": 1.911033013728238e-05, + "loss": 1.048, + "step": 6844 + }, + { + "epoch": 0.16, + "grad_norm": 2.448607865760487, + "learning_rate": 1.9110015483432318e-05, + "loss": 1.0597, + "step": 6845 + }, + { + "epoch": 0.16, + "grad_norm": 2.100365422767892, + "learning_rate": 1.9109700776540857e-05, + "loss": 1.088, + "step": 6846 + }, + { + "epoch": 0.16, + "grad_norm": 1.8766585940388119, + "learning_rate": 1.910938601660984e-05, + "loss": 1.1238, + "step": 6847 + }, + { + "epoch": 0.16, + "grad_norm": 2.291741521862003, + "learning_rate": 1.910907120364109e-05, + "loss": 1.0435, + "step": 6848 + }, + { + "epoch": 0.16, + "grad_norm": 1.9752086772059172, + "learning_rate": 1.9108756337636446e-05, + "loss": 1.0183, + "step": 6849 + }, + { + "epoch": 0.16, + "grad_norm": 2.123644290495901, + "learning_rate": 1.9108441418597737e-05, + "loss": 1.1362, + "step": 6850 + }, + { + "epoch": 0.16, + "grad_norm": 1.1674252582372233, + "learning_rate": 1.9108126446526797e-05, + "loss": 0.9344, + "step": 6851 + }, + { + "epoch": 0.16, + "grad_norm": 1.1526735918730187, + "learning_rate": 1.9107811421425465e-05, + "loss": 0.9653, + "step": 6852 + }, + { + "epoch": 0.16, + "grad_norm": 1.7955822191742021, + "learning_rate": 1.9107496343295567e-05, + "loss": 1.0852, + "step": 6853 + }, + { + "epoch": 0.16, + "grad_norm": 2.0384124266785513, + "learning_rate": 1.9107181212138945e-05, + "loss": 0.8677, + "step": 6854 + }, + { + "epoch": 0.16, + "grad_norm": 2.2345819469129355, + "learning_rate": 1.910686602795743e-05, + "loss": 1.0971, + "step": 6855 + }, + { + "epoch": 0.16, + "grad_norm": 3.7252087941518552, + "learning_rate": 1.9106550790752857e-05, + "loss": 1.0373, + "step": 6856 + }, + { + "epoch": 0.16, + "grad_norm": 2.19824371333601, + "learning_rate": 1.9106235500527065e-05, + "loss": 1.1101, + "step": 6857 + }, + { + "epoch": 0.16, + "grad_norm": 1.9865126627231446, + "learning_rate": 1.9105920157281883e-05, + "loss": 1.0415, + "step": 6858 + }, + { + "epoch": 0.16, + "grad_norm": 5.99409105563806, + "learning_rate": 1.9105604761019153e-05, + "loss": 1.0051, + "step": 6859 + }, + { + "epoch": 0.16, + "grad_norm": 1.8810002315320102, + "learning_rate": 1.910528931174071e-05, + "loss": 1.1287, + "step": 6860 + }, + { + "epoch": 0.16, + "grad_norm": 1.9995329221129887, + "learning_rate": 1.9104973809448388e-05, + "loss": 1.1433, + "step": 6861 + }, + { + "epoch": 0.16, + "grad_norm": 1.9529494256735958, + "learning_rate": 1.9104658254144025e-05, + "loss": 1.0778, + "step": 6862 + }, + { + "epoch": 0.16, + "grad_norm": 2.3196471544570256, + "learning_rate": 1.9104342645829464e-05, + "loss": 0.9762, + "step": 6863 + }, + { + "epoch": 0.16, + "grad_norm": 2.2608768181024312, + "learning_rate": 1.9104026984506533e-05, + "loss": 1.1359, + "step": 6864 + }, + { + "epoch": 0.16, + "grad_norm": 2.911384050932415, + "learning_rate": 1.9103711270177077e-05, + "loss": 0.988, + "step": 6865 + }, + { + "epoch": 0.16, + "grad_norm": 1.1897572952960587, + "learning_rate": 1.9103395502842932e-05, + "loss": 1.0141, + "step": 6866 + }, + { + "epoch": 0.16, + "grad_norm": 2.2311551108814602, + "learning_rate": 1.9103079682505938e-05, + "loss": 1.0646, + "step": 6867 + }, + { + "epoch": 0.16, + "grad_norm": 2.0161683673728725, + "learning_rate": 1.9102763809167927e-05, + "loss": 1.0937, + "step": 6868 + }, + { + "epoch": 0.16, + "grad_norm": 2.230383940320699, + "learning_rate": 1.9102447882830745e-05, + "loss": 1.0373, + "step": 6869 + }, + { + "epoch": 0.16, + "grad_norm": 2.0719421395573754, + "learning_rate": 1.910213190349623e-05, + "loss": 1.2025, + "step": 6870 + }, + { + "epoch": 0.16, + "grad_norm": 2.2077097804452332, + "learning_rate": 1.9101815871166222e-05, + "loss": 0.9787, + "step": 6871 + }, + { + "epoch": 0.16, + "grad_norm": 1.9205457559243717, + "learning_rate": 1.9101499785842563e-05, + "loss": 1.0544, + "step": 6872 + }, + { + "epoch": 0.16, + "grad_norm": 1.9808815428577233, + "learning_rate": 1.9101183647527086e-05, + "loss": 1.0914, + "step": 6873 + }, + { + "epoch": 0.16, + "grad_norm": 2.3133302495798422, + "learning_rate": 1.910086745622164e-05, + "loss": 0.9179, + "step": 6874 + }, + { + "epoch": 0.16, + "grad_norm": 2.0435215486347733, + "learning_rate": 1.910055121192806e-05, + "loss": 1.1296, + "step": 6875 + }, + { + "epoch": 0.16, + "grad_norm": 2.234657104313076, + "learning_rate": 1.910023491464819e-05, + "loss": 1.2285, + "step": 6876 + }, + { + "epoch": 0.16, + "grad_norm": 1.948914947609634, + "learning_rate": 1.9099918564383873e-05, + "loss": 1.1472, + "step": 6877 + }, + { + "epoch": 0.16, + "grad_norm": 2.0608301807510174, + "learning_rate": 1.9099602161136947e-05, + "loss": 1.1351, + "step": 6878 + }, + { + "epoch": 0.16, + "grad_norm": 1.8981655936954944, + "learning_rate": 1.9099285704909256e-05, + "loss": 1.0629, + "step": 6879 + }, + { + "epoch": 0.16, + "grad_norm": 2.029521537998421, + "learning_rate": 1.9098969195702646e-05, + "loss": 1.0248, + "step": 6880 + }, + { + "epoch": 0.16, + "grad_norm": 1.1447017183880066, + "learning_rate": 1.9098652633518956e-05, + "loss": 0.9895, + "step": 6881 + }, + { + "epoch": 0.16, + "grad_norm": 1.8761381873858758, + "learning_rate": 1.9098336018360027e-05, + "loss": 1.0389, + "step": 6882 + }, + { + "epoch": 0.16, + "grad_norm": 2.0263130557631697, + "learning_rate": 1.9098019350227705e-05, + "loss": 1.1138, + "step": 6883 + }, + { + "epoch": 0.16, + "grad_norm": 2.5609335035509075, + "learning_rate": 1.9097702629123836e-05, + "loss": 1.0658, + "step": 6884 + }, + { + "epoch": 0.16, + "grad_norm": 2.1939277356252855, + "learning_rate": 1.9097385855050263e-05, + "loss": 1.0788, + "step": 6885 + }, + { + "epoch": 0.16, + "grad_norm": 2.7487821821215754, + "learning_rate": 1.9097069028008827e-05, + "loss": 1.1119, + "step": 6886 + }, + { + "epoch": 0.16, + "grad_norm": 2.2059298554302336, + "learning_rate": 1.9096752148001378e-05, + "loss": 1.0427, + "step": 6887 + }, + { + "epoch": 0.16, + "grad_norm": 2.0114804596647446, + "learning_rate": 1.9096435215029757e-05, + "loss": 0.9727, + "step": 6888 + }, + { + "epoch": 0.16, + "grad_norm": 3.256126233795546, + "learning_rate": 1.9096118229095806e-05, + "loss": 0.9552, + "step": 6889 + }, + { + "epoch": 0.16, + "grad_norm": 2.145724066243745, + "learning_rate": 1.909580119020138e-05, + "loss": 1.1447, + "step": 6890 + }, + { + "epoch": 0.16, + "grad_norm": 1.2278481836241835, + "learning_rate": 1.9095484098348314e-05, + "loss": 0.9364, + "step": 6891 + }, + { + "epoch": 0.16, + "grad_norm": 1.9225760522130035, + "learning_rate": 1.9095166953538467e-05, + "loss": 1.0115, + "step": 6892 + }, + { + "epoch": 0.16, + "grad_norm": 2.0643673040313404, + "learning_rate": 1.9094849755773675e-05, + "loss": 1.093, + "step": 6893 + }, + { + "epoch": 0.16, + "grad_norm": 2.2134484312346205, + "learning_rate": 1.909453250505579e-05, + "loss": 1.0222, + "step": 6894 + }, + { + "epoch": 0.16, + "grad_norm": 2.3401329176485506, + "learning_rate": 1.9094215201386656e-05, + "loss": 1.1882, + "step": 6895 + }, + { + "epoch": 0.16, + "grad_norm": 2.1793865532115175, + "learning_rate": 1.9093897844768122e-05, + "loss": 1.0773, + "step": 6896 + }, + { + "epoch": 0.16, + "grad_norm": 2.103985455149469, + "learning_rate": 1.9093580435202037e-05, + "loss": 0.9597, + "step": 6897 + }, + { + "epoch": 0.16, + "grad_norm": 1.8982453693611874, + "learning_rate": 1.9093262972690246e-05, + "loss": 1.0534, + "step": 6898 + }, + { + "epoch": 0.16, + "grad_norm": 2.0559313779824184, + "learning_rate": 1.90929454572346e-05, + "loss": 1.0082, + "step": 6899 + }, + { + "epoch": 0.16, + "grad_norm": 1.099431919029806, + "learning_rate": 1.9092627888836945e-05, + "loss": 1.0112, + "step": 6900 + }, + { + "epoch": 0.16, + "grad_norm": 2.303957577868043, + "learning_rate": 1.9092310267499135e-05, + "loss": 1.0308, + "step": 6901 + }, + { + "epoch": 0.16, + "grad_norm": 1.1015706039825577, + "learning_rate": 1.9091992593223015e-05, + "loss": 0.9889, + "step": 6902 + }, + { + "epoch": 0.16, + "grad_norm": 1.9954774882382553, + "learning_rate": 1.9091674866010434e-05, + "loss": 1.0822, + "step": 6903 + }, + { + "epoch": 0.16, + "grad_norm": 1.977088798078296, + "learning_rate": 1.9091357085863246e-05, + "loss": 1.0768, + "step": 6904 + }, + { + "epoch": 0.16, + "grad_norm": 1.0933989395705206, + "learning_rate": 1.9091039252783296e-05, + "loss": 0.9865, + "step": 6905 + }, + { + "epoch": 0.16, + "grad_norm": 2.0517206765127254, + "learning_rate": 1.909072136677244e-05, + "loss": 1.0287, + "step": 6906 + }, + { + "epoch": 0.16, + "grad_norm": 2.033077752583919, + "learning_rate": 1.909040342783252e-05, + "loss": 1.0083, + "step": 6907 + }, + { + "epoch": 0.16, + "grad_norm": 2.1479004635212804, + "learning_rate": 1.90900854359654e-05, + "loss": 1.0876, + "step": 6908 + }, + { + "epoch": 0.16, + "grad_norm": 1.9834736135267304, + "learning_rate": 1.9089767391172925e-05, + "loss": 1.0082, + "step": 6909 + }, + { + "epoch": 0.16, + "grad_norm": 2.083554812060529, + "learning_rate": 1.9089449293456944e-05, + "loss": 0.9709, + "step": 6910 + }, + { + "epoch": 0.16, + "grad_norm": 2.2465216013544076, + "learning_rate": 1.908913114281931e-05, + "loss": 1.0777, + "step": 6911 + }, + { + "epoch": 0.16, + "grad_norm": 2.240089975158146, + "learning_rate": 1.908881293926188e-05, + "loss": 1.2018, + "step": 6912 + }, + { + "epoch": 0.16, + "grad_norm": 1.1245544641216039, + "learning_rate": 1.9088494682786502e-05, + "loss": 0.9838, + "step": 6913 + }, + { + "epoch": 0.16, + "grad_norm": 1.9949595979736772, + "learning_rate": 1.908817637339503e-05, + "loss": 1.1039, + "step": 6914 + }, + { + "epoch": 0.16, + "grad_norm": 2.229757555918306, + "learning_rate": 1.908785801108932e-05, + "loss": 1.038, + "step": 6915 + }, + { + "epoch": 0.16, + "grad_norm": 2.56302382185995, + "learning_rate": 1.9087539595871222e-05, + "loss": 1.0204, + "step": 6916 + }, + { + "epoch": 0.16, + "grad_norm": 2.190851550714002, + "learning_rate": 1.908722112774259e-05, + "loss": 1.1794, + "step": 6917 + }, + { + "epoch": 0.16, + "grad_norm": 2.3638764117693745, + "learning_rate": 1.9086902606705286e-05, + "loss": 1.0157, + "step": 6918 + }, + { + "epoch": 0.16, + "grad_norm": 2.0487153141094554, + "learning_rate": 1.9086584032761148e-05, + "loss": 0.9481, + "step": 6919 + }, + { + "epoch": 0.16, + "grad_norm": 2.422259765390144, + "learning_rate": 1.9086265405912047e-05, + "loss": 1.0835, + "step": 6920 + }, + { + "epoch": 0.16, + "grad_norm": 2.071269633149418, + "learning_rate": 1.9085946726159833e-05, + "loss": 0.9887, + "step": 6921 + }, + { + "epoch": 0.16, + "grad_norm": 1.1324825422926132, + "learning_rate": 1.908562799350636e-05, + "loss": 1.0643, + "step": 6922 + }, + { + "epoch": 0.16, + "grad_norm": 2.1138540155451, + "learning_rate": 1.9085309207953483e-05, + "loss": 1.0935, + "step": 6923 + }, + { + "epoch": 0.16, + "grad_norm": 2.317219298958268, + "learning_rate": 1.908499036950306e-05, + "loss": 0.9702, + "step": 6924 + }, + { + "epoch": 0.16, + "grad_norm": 2.0436162020676165, + "learning_rate": 1.9084671478156947e-05, + "loss": 0.9458, + "step": 6925 + }, + { + "epoch": 0.16, + "grad_norm": 1.962161904260691, + "learning_rate": 1.9084352533917e-05, + "loss": 1.0273, + "step": 6926 + }, + { + "epoch": 0.16, + "grad_norm": 2.034134759351858, + "learning_rate": 1.908403353678508e-05, + "loss": 1.0438, + "step": 6927 + }, + { + "epoch": 0.16, + "grad_norm": 2.873975147084006, + "learning_rate": 1.9083714486763035e-05, + "loss": 1.0029, + "step": 6928 + }, + { + "epoch": 0.16, + "grad_norm": 2.271427964115215, + "learning_rate": 1.9083395383852733e-05, + "loss": 1.1518, + "step": 6929 + }, + { + "epoch": 0.16, + "grad_norm": 2.1623744657788735, + "learning_rate": 1.9083076228056023e-05, + "loss": 1.1353, + "step": 6930 + }, + { + "epoch": 0.16, + "grad_norm": 2.048708154554188, + "learning_rate": 1.908275701937477e-05, + "loss": 1.0449, + "step": 6931 + }, + { + "epoch": 0.16, + "grad_norm": 2.124348747071421, + "learning_rate": 1.9082437757810827e-05, + "loss": 1.1082, + "step": 6932 + }, + { + "epoch": 0.16, + "grad_norm": 2.100523209754742, + "learning_rate": 1.908211844336606e-05, + "loss": 1.0038, + "step": 6933 + }, + { + "epoch": 0.16, + "grad_norm": 2.084008258729196, + "learning_rate": 1.9081799076042323e-05, + "loss": 1.0954, + "step": 6934 + }, + { + "epoch": 0.16, + "grad_norm": 2.2864359891309483, + "learning_rate": 1.9081479655841476e-05, + "loss": 0.9238, + "step": 6935 + }, + { + "epoch": 0.16, + "grad_norm": 2.214222989435628, + "learning_rate": 1.908116018276538e-05, + "loss": 1.2105, + "step": 6936 + }, + { + "epoch": 0.16, + "grad_norm": 1.8536149892462002, + "learning_rate": 1.9080840656815895e-05, + "loss": 1.0608, + "step": 6937 + }, + { + "epoch": 0.16, + "grad_norm": 2.073478040416993, + "learning_rate": 1.908052107799488e-05, + "loss": 1.1412, + "step": 6938 + }, + { + "epoch": 0.16, + "grad_norm": 1.9237982869859975, + "learning_rate": 1.9080201446304198e-05, + "loss": 1.0461, + "step": 6939 + }, + { + "epoch": 0.16, + "grad_norm": 1.9354198815443449, + "learning_rate": 1.9079881761745706e-05, + "loss": 1.0157, + "step": 6940 + }, + { + "epoch": 0.16, + "grad_norm": 1.9196382783174957, + "learning_rate": 1.907956202432127e-05, + "loss": 0.9799, + "step": 6941 + }, + { + "epoch": 0.16, + "grad_norm": 2.090054308119434, + "learning_rate": 1.9079242234032747e-05, + "loss": 1.1596, + "step": 6942 + }, + { + "epoch": 0.16, + "grad_norm": 2.6589376804916625, + "learning_rate": 1.9078922390882e-05, + "loss": 1.0671, + "step": 6943 + }, + { + "epoch": 0.16, + "grad_norm": 2.0103323434477884, + "learning_rate": 1.9078602494870894e-05, + "loss": 1.0102, + "step": 6944 + }, + { + "epoch": 0.16, + "grad_norm": 2.0637903729742333, + "learning_rate": 1.907828254600129e-05, + "loss": 1.0901, + "step": 6945 + }, + { + "epoch": 0.16, + "grad_norm": 2.2358872042478244, + "learning_rate": 1.9077962544275053e-05, + "loss": 1.0725, + "step": 6946 + }, + { + "epoch": 0.16, + "grad_norm": 2.122663462864931, + "learning_rate": 1.9077642489694043e-05, + "loss": 1.1225, + "step": 6947 + }, + { + "epoch": 0.16, + "grad_norm": 2.023169287097267, + "learning_rate": 1.9077322382260123e-05, + "loss": 1.0277, + "step": 6948 + }, + { + "epoch": 0.16, + "grad_norm": 1.9040373378345892, + "learning_rate": 1.907700222197516e-05, + "loss": 1.0668, + "step": 6949 + }, + { + "epoch": 0.16, + "grad_norm": 2.5870660875885076, + "learning_rate": 1.9076682008841015e-05, + "loss": 0.9854, + "step": 6950 + }, + { + "epoch": 0.16, + "grad_norm": 2.226851956057286, + "learning_rate": 1.9076361742859556e-05, + "loss": 1.0877, + "step": 6951 + }, + { + "epoch": 0.16, + "grad_norm": 1.1854822252736266, + "learning_rate": 1.9076041424032642e-05, + "loss": 1.0863, + "step": 6952 + }, + { + "epoch": 0.16, + "grad_norm": 2.024668909594486, + "learning_rate": 1.9075721052362145e-05, + "loss": 1.056, + "step": 6953 + }, + { + "epoch": 0.16, + "grad_norm": 1.1279647819624552, + "learning_rate": 1.9075400627849925e-05, + "loss": 1.0295, + "step": 6954 + }, + { + "epoch": 0.16, + "grad_norm": 1.8507462532736954, + "learning_rate": 1.9075080150497848e-05, + "loss": 1.0369, + "step": 6955 + }, + { + "epoch": 0.16, + "grad_norm": 2.46236855535531, + "learning_rate": 1.907475962030778e-05, + "loss": 1.1342, + "step": 6956 + }, + { + "epoch": 0.16, + "grad_norm": 1.9748430607849443, + "learning_rate": 1.9074439037281594e-05, + "loss": 0.9575, + "step": 6957 + }, + { + "epoch": 0.16, + "grad_norm": 1.2344636998418623, + "learning_rate": 1.907411840142115e-05, + "loss": 1.0609, + "step": 6958 + }, + { + "epoch": 0.16, + "grad_norm": 2.2794557426561943, + "learning_rate": 1.907379771272831e-05, + "loss": 1.0482, + "step": 6959 + }, + { + "epoch": 0.16, + "grad_norm": 2.133213271814756, + "learning_rate": 1.9073476971204952e-05, + "loss": 1.0052, + "step": 6960 + }, + { + "epoch": 0.16, + "grad_norm": 2.051137390718155, + "learning_rate": 1.9073156176852936e-05, + "loss": 1.0522, + "step": 6961 + }, + { + "epoch": 0.16, + "grad_norm": 2.234354912310782, + "learning_rate": 1.9072835329674134e-05, + "loss": 1.0455, + "step": 6962 + }, + { + "epoch": 0.16, + "grad_norm": 2.079043188323808, + "learning_rate": 1.907251442967041e-05, + "loss": 1.1565, + "step": 6963 + }, + { + "epoch": 0.16, + "grad_norm": 1.9272343101460168, + "learning_rate": 1.9072193476843636e-05, + "loss": 0.9436, + "step": 6964 + }, + { + "epoch": 0.16, + "grad_norm": 1.8970770155431744, + "learning_rate": 1.907187247119568e-05, + "loss": 1.0362, + "step": 6965 + }, + { + "epoch": 0.16, + "grad_norm": 2.47003029031906, + "learning_rate": 1.907155141272841e-05, + "loss": 0.9852, + "step": 6966 + }, + { + "epoch": 0.16, + "grad_norm": 1.9897764233459945, + "learning_rate": 1.9071230301443698e-05, + "loss": 0.9875, + "step": 6967 + }, + { + "epoch": 0.16, + "grad_norm": 2.0802291461583597, + "learning_rate": 1.907090913734341e-05, + "loss": 0.9815, + "step": 6968 + }, + { + "epoch": 0.16, + "grad_norm": 2.8295137535048034, + "learning_rate": 1.9070587920429414e-05, + "loss": 1.0895, + "step": 6969 + }, + { + "epoch": 0.16, + "grad_norm": 2.148373292266966, + "learning_rate": 1.9070266650703585e-05, + "loss": 1.0917, + "step": 6970 + }, + { + "epoch": 0.16, + "grad_norm": 1.2312000056421435, + "learning_rate": 1.9069945328167792e-05, + "loss": 0.9901, + "step": 6971 + }, + { + "epoch": 0.16, + "grad_norm": 1.9138266206415337, + "learning_rate": 1.906962395282391e-05, + "loss": 0.9786, + "step": 6972 + }, + { + "epoch": 0.16, + "grad_norm": 2.391375082602138, + "learning_rate": 1.90693025246738e-05, + "loss": 0.991, + "step": 6973 + }, + { + "epoch": 0.16, + "grad_norm": 1.913244135621458, + "learning_rate": 1.9068981043719343e-05, + "loss": 1.0777, + "step": 6974 + }, + { + "epoch": 0.16, + "grad_norm": 2.0421798351657685, + "learning_rate": 1.9068659509962407e-05, + "loss": 1.1439, + "step": 6975 + }, + { + "epoch": 0.16, + "grad_norm": 1.9633556367908307, + "learning_rate": 1.9068337923404863e-05, + "loss": 1.0524, + "step": 6976 + }, + { + "epoch": 0.16, + "grad_norm": 2.989127890945401, + "learning_rate": 1.9068016284048584e-05, + "loss": 1.052, + "step": 6977 + }, + { + "epoch": 0.16, + "grad_norm": 2.3804622296224296, + "learning_rate": 1.9067694591895443e-05, + "loss": 1.032, + "step": 6978 + }, + { + "epoch": 0.16, + "grad_norm": 1.131602489087598, + "learning_rate": 1.9067372846947315e-05, + "loss": 0.9791, + "step": 6979 + }, + { + "epoch": 0.16, + "grad_norm": 1.8833216007576317, + "learning_rate": 1.906705104920607e-05, + "loss": 0.9263, + "step": 6980 + }, + { + "epoch": 0.16, + "grad_norm": 2.2807339868409207, + "learning_rate": 1.9066729198673588e-05, + "loss": 1.005, + "step": 6981 + }, + { + "epoch": 0.16, + "grad_norm": 1.9974566766777933, + "learning_rate": 1.9066407295351732e-05, + "loss": 1.1381, + "step": 6982 + }, + { + "epoch": 0.16, + "grad_norm": 2.0159569109390345, + "learning_rate": 1.9066085339242387e-05, + "loss": 1.1845, + "step": 6983 + }, + { + "epoch": 0.16, + "grad_norm": 2.003876433396993, + "learning_rate": 1.906576333034742e-05, + "loss": 1.2298, + "step": 6984 + }, + { + "epoch": 0.16, + "grad_norm": 1.9998912355624963, + "learning_rate": 1.906544126866871e-05, + "loss": 1.0824, + "step": 6985 + }, + { + "epoch": 0.16, + "grad_norm": 1.9861484766818953, + "learning_rate": 1.906511915420813e-05, + "loss": 0.9613, + "step": 6986 + }, + { + "epoch": 0.16, + "grad_norm": 1.9411481590642912, + "learning_rate": 1.9064796986967557e-05, + "loss": 1.0935, + "step": 6987 + }, + { + "epoch": 0.16, + "grad_norm": 2.243661661097593, + "learning_rate": 1.9064474766948864e-05, + "loss": 1.0355, + "step": 6988 + }, + { + "epoch": 0.16, + "grad_norm": 4.775814155020322, + "learning_rate": 1.906415249415393e-05, + "loss": 1.0454, + "step": 6989 + }, + { + "epoch": 0.16, + "grad_norm": 1.9228527726343472, + "learning_rate": 1.9063830168584633e-05, + "loss": 1.0323, + "step": 6990 + }, + { + "epoch": 0.16, + "grad_norm": 2.4087944936664027, + "learning_rate": 1.9063507790242845e-05, + "loss": 1.0868, + "step": 6991 + }, + { + "epoch": 0.16, + "grad_norm": 2.5392168573997216, + "learning_rate": 1.9063185359130446e-05, + "loss": 1.0646, + "step": 6992 + }, + { + "epoch": 0.16, + "grad_norm": 1.2963196792797556, + "learning_rate": 1.906286287524931e-05, + "loss": 0.943, + "step": 6993 + }, + { + "epoch": 0.16, + "grad_norm": 7.100707482421858, + "learning_rate": 1.9062540338601324e-05, + "loss": 1.1133, + "step": 6994 + }, + { + "epoch": 0.16, + "grad_norm": 1.9542069960722455, + "learning_rate": 1.906221774918835e-05, + "loss": 1.0419, + "step": 6995 + }, + { + "epoch": 0.16, + "grad_norm": 2.1667684538511565, + "learning_rate": 1.906189510701228e-05, + "loss": 1.0894, + "step": 6996 + }, + { + "epoch": 0.16, + "grad_norm": 2.072074278402403, + "learning_rate": 1.906157241207499e-05, + "loss": 1.0894, + "step": 6997 + }, + { + "epoch": 0.16, + "grad_norm": 1.2177514234463482, + "learning_rate": 1.9061249664378353e-05, + "loss": 0.9581, + "step": 6998 + }, + { + "epoch": 0.16, + "grad_norm": 2.2856478047281255, + "learning_rate": 1.906092686392425e-05, + "loss": 1.1057, + "step": 6999 + }, + { + "epoch": 0.16, + "grad_norm": 2.1191293209627027, + "learning_rate": 1.9060604010714566e-05, + "loss": 0.95, + "step": 7000 + }, + { + "epoch": 0.16, + "grad_norm": 2.085597024099596, + "learning_rate": 1.9060281104751178e-05, + "loss": 1.0416, + "step": 7001 + }, + { + "epoch": 0.16, + "grad_norm": 2.320442125845049, + "learning_rate": 1.9059958146035963e-05, + "loss": 0.9476, + "step": 7002 + }, + { + "epoch": 0.16, + "grad_norm": 2.0541871859155214, + "learning_rate": 1.90596351345708e-05, + "loss": 1.1078, + "step": 7003 + }, + { + "epoch": 0.17, + "grad_norm": 1.8371347642588334, + "learning_rate": 1.9059312070357577e-05, + "loss": 1.0528, + "step": 7004 + }, + { + "epoch": 0.17, + "grad_norm": 2.008056613667033, + "learning_rate": 1.9058988953398166e-05, + "loss": 1.0845, + "step": 7005 + }, + { + "epoch": 0.17, + "grad_norm": 3.592658773636702, + "learning_rate": 1.9058665783694458e-05, + "loss": 1.0598, + "step": 7006 + }, + { + "epoch": 0.17, + "grad_norm": 2.3943338674524983, + "learning_rate": 1.9058342561248332e-05, + "loss": 1.0943, + "step": 7007 + }, + { + "epoch": 0.17, + "grad_norm": 1.13793963256553, + "learning_rate": 1.9058019286061662e-05, + "loss": 0.9467, + "step": 7008 + }, + { + "epoch": 0.17, + "grad_norm": 2.1726335854438683, + "learning_rate": 1.905769595813634e-05, + "loss": 1.0108, + "step": 7009 + }, + { + "epoch": 0.17, + "grad_norm": 1.9990633687772486, + "learning_rate": 1.9057372577474244e-05, + "loss": 1.1157, + "step": 7010 + }, + { + "epoch": 0.17, + "grad_norm": 1.1481826726167512, + "learning_rate": 1.9057049144077256e-05, + "loss": 0.9999, + "step": 7011 + }, + { + "epoch": 0.17, + "grad_norm": 2.2900658487977927, + "learning_rate": 1.905672565794726e-05, + "loss": 1.1078, + "step": 7012 + }, + { + "epoch": 0.17, + "grad_norm": 1.9186656699006872, + "learning_rate": 1.9056402119086142e-05, + "loss": 0.9941, + "step": 7013 + }, + { + "epoch": 0.17, + "grad_norm": 2.0448371105993743, + "learning_rate": 1.9056078527495784e-05, + "loss": 0.9522, + "step": 7014 + }, + { + "epoch": 0.17, + "grad_norm": 1.906516374577314, + "learning_rate": 1.9055754883178068e-05, + "loss": 1.0877, + "step": 7015 + }, + { + "epoch": 0.17, + "grad_norm": 2.0701969989461166, + "learning_rate": 1.905543118613488e-05, + "loss": 1.0199, + "step": 7016 + }, + { + "epoch": 0.17, + "grad_norm": 2.5355506933233776, + "learning_rate": 1.9055107436368104e-05, + "loss": 1.0169, + "step": 7017 + }, + { + "epoch": 0.17, + "grad_norm": 2.148340217815377, + "learning_rate": 1.905478363387963e-05, + "loss": 1.0812, + "step": 7018 + }, + { + "epoch": 0.17, + "grad_norm": 2.1949558307704864, + "learning_rate": 1.9054459778671334e-05, + "loss": 0.9578, + "step": 7019 + }, + { + "epoch": 0.17, + "grad_norm": 2.0199132229851706, + "learning_rate": 1.9054135870745106e-05, + "loss": 0.9502, + "step": 7020 + }, + { + "epoch": 0.17, + "grad_norm": 1.16445553668021, + "learning_rate": 1.9053811910102838e-05, + "loss": 1.0161, + "step": 7021 + }, + { + "epoch": 0.17, + "grad_norm": 1.8925157701329802, + "learning_rate": 1.9053487896746406e-05, + "loss": 1.0695, + "step": 7022 + }, + { + "epoch": 0.17, + "grad_norm": 1.159448225383586, + "learning_rate": 1.9053163830677705e-05, + "loss": 0.9978, + "step": 7023 + }, + { + "epoch": 0.17, + "grad_norm": 2.241565207300163, + "learning_rate": 1.9052839711898614e-05, + "loss": 1.0681, + "step": 7024 + }, + { + "epoch": 0.17, + "grad_norm": 1.1848303595119145, + "learning_rate": 1.9052515540411027e-05, + "loss": 1.066, + "step": 7025 + }, + { + "epoch": 0.17, + "grad_norm": 4.511564403557719, + "learning_rate": 1.905219131621683e-05, + "loss": 1.0098, + "step": 7026 + }, + { + "epoch": 0.17, + "grad_norm": 2.206158067447515, + "learning_rate": 1.9051867039317903e-05, + "loss": 1.035, + "step": 7027 + }, + { + "epoch": 0.17, + "grad_norm": 2.0398738398372593, + "learning_rate": 1.9051542709716147e-05, + "loss": 1.1118, + "step": 7028 + }, + { + "epoch": 0.17, + "grad_norm": 2.036975367578486, + "learning_rate": 1.9051218327413438e-05, + "loss": 1.0874, + "step": 7029 + }, + { + "epoch": 0.17, + "grad_norm": 2.236189663444471, + "learning_rate": 1.9050893892411675e-05, + "loss": 1.0409, + "step": 7030 + }, + { + "epoch": 0.17, + "grad_norm": 2.052914285352502, + "learning_rate": 1.905056940471274e-05, + "loss": 1.1019, + "step": 7031 + }, + { + "epoch": 0.17, + "grad_norm": 1.1581377506300463, + "learning_rate": 1.9050244864318524e-05, + "loss": 0.8995, + "step": 7032 + }, + { + "epoch": 0.17, + "grad_norm": 1.0979178664564262, + "learning_rate": 1.904992027123092e-05, + "loss": 0.9928, + "step": 7033 + }, + { + "epoch": 0.17, + "grad_norm": 2.2441128995907453, + "learning_rate": 1.9049595625451813e-05, + "loss": 1.2115, + "step": 7034 + }, + { + "epoch": 0.17, + "grad_norm": 2.2342455575743525, + "learning_rate": 1.9049270926983095e-05, + "loss": 0.9994, + "step": 7035 + }, + { + "epoch": 0.17, + "grad_norm": 2.2152118010178796, + "learning_rate": 1.9048946175826658e-05, + "loss": 1.0934, + "step": 7036 + }, + { + "epoch": 0.17, + "grad_norm": 2.3355835784842394, + "learning_rate": 1.904862137198439e-05, + "loss": 0.8665, + "step": 7037 + }, + { + "epoch": 0.17, + "grad_norm": 2.2554731536909123, + "learning_rate": 1.904829651545818e-05, + "loss": 1.1479, + "step": 7038 + }, + { + "epoch": 0.17, + "grad_norm": 2.3161266571883012, + "learning_rate": 1.904797160624993e-05, + "loss": 0.949, + "step": 7039 + }, + { + "epoch": 0.17, + "grad_norm": 2.2471319575704856, + "learning_rate": 1.904764664436152e-05, + "loss": 1.1551, + "step": 7040 + }, + { + "epoch": 0.17, + "grad_norm": 1.9555021244489925, + "learning_rate": 1.904732162979485e-05, + "loss": 1.0921, + "step": 7041 + }, + { + "epoch": 0.17, + "grad_norm": 2.2819754224028923, + "learning_rate": 1.904699656255181e-05, + "loss": 1.0289, + "step": 7042 + }, + { + "epoch": 0.17, + "grad_norm": 1.8443537754964394, + "learning_rate": 1.9046671442634288e-05, + "loss": 1.1728, + "step": 7043 + }, + { + "epoch": 0.17, + "grad_norm": 2.424172324348587, + "learning_rate": 1.9046346270044183e-05, + "loss": 0.9474, + "step": 7044 + }, + { + "epoch": 0.17, + "grad_norm": 2.1295205900439726, + "learning_rate": 1.904602104478339e-05, + "loss": 1.2433, + "step": 7045 + }, + { + "epoch": 0.17, + "grad_norm": 1.9491446024422459, + "learning_rate": 1.904569576685379e-05, + "loss": 1.0527, + "step": 7046 + }, + { + "epoch": 0.17, + "grad_norm": 3.5189434583024615, + "learning_rate": 1.904537043625729e-05, + "loss": 1.035, + "step": 7047 + }, + { + "epoch": 0.17, + "grad_norm": 1.7527343720881796, + "learning_rate": 1.9045045052995776e-05, + "loss": 1.0023, + "step": 7048 + }, + { + "epoch": 0.17, + "grad_norm": 2.914694688733327, + "learning_rate": 1.904471961707115e-05, + "loss": 0.9696, + "step": 7049 + }, + { + "epoch": 0.17, + "grad_norm": 1.9591983826717598, + "learning_rate": 1.9044394128485303e-05, + "loss": 1.0106, + "step": 7050 + }, + { + "epoch": 0.17, + "grad_norm": 2.1117638160243213, + "learning_rate": 1.9044068587240128e-05, + "loss": 1.0835, + "step": 7051 + }, + { + "epoch": 0.17, + "grad_norm": 2.897226395000022, + "learning_rate": 1.9043742993337523e-05, + "loss": 1.0738, + "step": 7052 + }, + { + "epoch": 0.17, + "grad_norm": 2.0448579926245087, + "learning_rate": 1.9043417346779386e-05, + "loss": 0.9735, + "step": 7053 + }, + { + "epoch": 0.17, + "grad_norm": 2.9289897023315317, + "learning_rate": 1.9043091647567608e-05, + "loss": 0.8759, + "step": 7054 + }, + { + "epoch": 0.17, + "grad_norm": 2.240394450104205, + "learning_rate": 1.9042765895704085e-05, + "loss": 1.0458, + "step": 7055 + }, + { + "epoch": 0.17, + "grad_norm": 2.1257392863584146, + "learning_rate": 1.904244009119072e-05, + "loss": 1.1236, + "step": 7056 + }, + { + "epoch": 0.17, + "grad_norm": 2.0201464185081597, + "learning_rate": 1.90421142340294e-05, + "loss": 1.072, + "step": 7057 + }, + { + "epoch": 0.17, + "grad_norm": 1.2665412303170185, + "learning_rate": 1.9041788324222035e-05, + "loss": 0.995, + "step": 7058 + }, + { + "epoch": 0.17, + "grad_norm": 2.1785059623656164, + "learning_rate": 1.9041462361770513e-05, + "loss": 0.9814, + "step": 7059 + }, + { + "epoch": 0.17, + "grad_norm": 1.986074620173816, + "learning_rate": 1.9041136346676733e-05, + "loss": 0.9888, + "step": 7060 + }, + { + "epoch": 0.17, + "grad_norm": 2.0880277270166565, + "learning_rate": 1.9040810278942596e-05, + "loss": 1.0958, + "step": 7061 + }, + { + "epoch": 0.17, + "grad_norm": 2.3038768184047393, + "learning_rate": 1.904048415857e-05, + "loss": 1.1326, + "step": 7062 + }, + { + "epoch": 0.17, + "grad_norm": 2.4049271281203928, + "learning_rate": 1.9040157985560842e-05, + "loss": 0.8873, + "step": 7063 + }, + { + "epoch": 0.17, + "grad_norm": 2.444461880227211, + "learning_rate": 1.9039831759917023e-05, + "loss": 1.0455, + "step": 7064 + }, + { + "epoch": 0.17, + "grad_norm": 2.2709917667382964, + "learning_rate": 1.9039505481640447e-05, + "loss": 1.0762, + "step": 7065 + }, + { + "epoch": 0.17, + "grad_norm": 1.8140556504610443, + "learning_rate": 1.9039179150733e-05, + "loss": 1.0035, + "step": 7066 + }, + { + "epoch": 0.17, + "grad_norm": 1.0859726704589854, + "learning_rate": 1.9038852767196592e-05, + "loss": 0.9721, + "step": 7067 + }, + { + "epoch": 0.17, + "grad_norm": 2.424050114903408, + "learning_rate": 1.9038526331033122e-05, + "loss": 1.0853, + "step": 7068 + }, + { + "epoch": 0.17, + "grad_norm": 2.6201948896553016, + "learning_rate": 1.903819984224449e-05, + "loss": 1.0975, + "step": 7069 + }, + { + "epoch": 0.17, + "grad_norm": 2.0663532695802007, + "learning_rate": 1.90378733008326e-05, + "loss": 1.0425, + "step": 7070 + }, + { + "epoch": 0.17, + "grad_norm": 1.9017284809389083, + "learning_rate": 1.903754670679935e-05, + "loss": 0.9506, + "step": 7071 + }, + { + "epoch": 0.17, + "grad_norm": 2.476850225836892, + "learning_rate": 1.9037220060146637e-05, + "loss": 1.1286, + "step": 7072 + }, + { + "epoch": 0.17, + "grad_norm": 2.1725333526216053, + "learning_rate": 1.9036893360876373e-05, + "loss": 1.1457, + "step": 7073 + }, + { + "epoch": 0.17, + "grad_norm": 2.133611023281929, + "learning_rate": 1.9036566608990452e-05, + "loss": 1.0994, + "step": 7074 + }, + { + "epoch": 0.17, + "grad_norm": 2.190337557455952, + "learning_rate": 1.9036239804490778e-05, + "loss": 1.0327, + "step": 7075 + }, + { + "epoch": 0.17, + "grad_norm": 2.373532330637468, + "learning_rate": 1.9035912947379258e-05, + "loss": 1.0176, + "step": 7076 + }, + { + "epoch": 0.17, + "grad_norm": 1.962986539518539, + "learning_rate": 1.903558603765779e-05, + "loss": 1.0363, + "step": 7077 + }, + { + "epoch": 0.17, + "grad_norm": 2.520462216589665, + "learning_rate": 1.9035259075328278e-05, + "loss": 1.0158, + "step": 7078 + }, + { + "epoch": 0.17, + "grad_norm": 2.04792203040664, + "learning_rate": 1.9034932060392633e-05, + "loss": 1.0571, + "step": 7079 + }, + { + "epoch": 0.17, + "grad_norm": 2.3154751373175952, + "learning_rate": 1.903460499285275e-05, + "loss": 0.9146, + "step": 7080 + }, + { + "epoch": 0.17, + "grad_norm": 2.0597984681385793, + "learning_rate": 1.9034277872710533e-05, + "loss": 1.1806, + "step": 7081 + }, + { + "epoch": 0.17, + "grad_norm": 1.1250638378677305, + "learning_rate": 1.9033950699967896e-05, + "loss": 0.9694, + "step": 7082 + }, + { + "epoch": 0.17, + "grad_norm": 2.1001977309095374, + "learning_rate": 1.9033623474626736e-05, + "loss": 0.9576, + "step": 7083 + }, + { + "epoch": 0.17, + "grad_norm": 1.9629498517889323, + "learning_rate": 1.9033296196688957e-05, + "loss": 0.9981, + "step": 7084 + }, + { + "epoch": 0.17, + "grad_norm": 2.57040782689372, + "learning_rate": 1.903296886615647e-05, + "loss": 1.0823, + "step": 7085 + }, + { + "epoch": 0.17, + "grad_norm": 2.284087035080569, + "learning_rate": 1.9032641483031177e-05, + "loss": 1.0548, + "step": 7086 + }, + { + "epoch": 0.17, + "grad_norm": 2.3435984168108575, + "learning_rate": 1.9032314047314985e-05, + "loss": 1.0122, + "step": 7087 + }, + { + "epoch": 0.17, + "grad_norm": 2.0791261930426725, + "learning_rate": 1.9031986559009804e-05, + "loss": 1.0168, + "step": 7088 + }, + { + "epoch": 0.17, + "grad_norm": 3.5312498420213143, + "learning_rate": 1.9031659018117537e-05, + "loss": 1.1367, + "step": 7089 + }, + { + "epoch": 0.17, + "grad_norm": 2.0424448749341733, + "learning_rate": 1.903133142464009e-05, + "loss": 0.9883, + "step": 7090 + }, + { + "epoch": 0.17, + "grad_norm": 2.182386022418861, + "learning_rate": 1.9031003778579376e-05, + "loss": 1.1597, + "step": 7091 + }, + { + "epoch": 0.17, + "grad_norm": 1.9285095992205699, + "learning_rate": 1.9030676079937296e-05, + "loss": 1.1955, + "step": 7092 + }, + { + "epoch": 0.17, + "grad_norm": 2.0356369532537655, + "learning_rate": 1.903034832871576e-05, + "loss": 1.1105, + "step": 7093 + }, + { + "epoch": 0.17, + "grad_norm": 2.712180091016883, + "learning_rate": 1.903002052491668e-05, + "loss": 1.1619, + "step": 7094 + }, + { + "epoch": 0.17, + "grad_norm": 2.0037083027254634, + "learning_rate": 1.902969266854196e-05, + "loss": 1.0973, + "step": 7095 + }, + { + "epoch": 0.17, + "grad_norm": 2.25479303362368, + "learning_rate": 1.902936475959351e-05, + "loss": 0.9845, + "step": 7096 + }, + { + "epoch": 0.17, + "grad_norm": 2.4788055444812946, + "learning_rate": 1.902903679807324e-05, + "loss": 1.0659, + "step": 7097 + }, + { + "epoch": 0.17, + "grad_norm": 2.4470545102276846, + "learning_rate": 1.902870878398306e-05, + "loss": 1.067, + "step": 7098 + }, + { + "epoch": 0.17, + "grad_norm": 2.320814196056053, + "learning_rate": 1.9028380717324875e-05, + "loss": 1.0989, + "step": 7099 + }, + { + "epoch": 0.17, + "grad_norm": 2.587145812168815, + "learning_rate": 1.9028052598100607e-05, + "loss": 1.0689, + "step": 7100 + }, + { + "epoch": 0.17, + "grad_norm": 2.0083832147093323, + "learning_rate": 1.902772442631215e-05, + "loss": 1.0853, + "step": 7101 + }, + { + "epoch": 0.17, + "grad_norm": 1.976647445395101, + "learning_rate": 1.902739620196143e-05, + "loss": 1.0557, + "step": 7102 + }, + { + "epoch": 0.17, + "grad_norm": 2.2346015716098804, + "learning_rate": 1.902706792505035e-05, + "loss": 1.0414, + "step": 7103 + }, + { + "epoch": 0.17, + "grad_norm": 2.2955206591054274, + "learning_rate": 1.902673959558082e-05, + "loss": 1.1567, + "step": 7104 + }, + { + "epoch": 0.17, + "grad_norm": 2.4241386115002084, + "learning_rate": 1.9026411213554758e-05, + "loss": 1.0435, + "step": 7105 + }, + { + "epoch": 0.17, + "grad_norm": 2.480105015882594, + "learning_rate": 1.9026082778974065e-05, + "loss": 1.0172, + "step": 7106 + }, + { + "epoch": 0.17, + "grad_norm": 2.0970996486961737, + "learning_rate": 1.902575429184067e-05, + "loss": 1.1062, + "step": 7107 + }, + { + "epoch": 0.17, + "grad_norm": 1.99444911046061, + "learning_rate": 1.902542575215647e-05, + "loss": 0.9138, + "step": 7108 + }, + { + "epoch": 0.17, + "grad_norm": 2.0352329887325413, + "learning_rate": 1.902509715992338e-05, + "loss": 0.9734, + "step": 7109 + }, + { + "epoch": 0.17, + "grad_norm": 2.0754201251158597, + "learning_rate": 1.9024768515143325e-05, + "loss": 1.153, + "step": 7110 + }, + { + "epoch": 0.17, + "grad_norm": 2.086259560444014, + "learning_rate": 1.902443981781821e-05, + "loss": 1.0753, + "step": 7111 + }, + { + "epoch": 0.17, + "grad_norm": 2.440463409232824, + "learning_rate": 1.9024111067949945e-05, + "loss": 0.7887, + "step": 7112 + }, + { + "epoch": 0.17, + "grad_norm": 2.2269343309854834, + "learning_rate": 1.902378226554045e-05, + "loss": 1.0068, + "step": 7113 + }, + { + "epoch": 0.17, + "grad_norm": 2.092654743680292, + "learning_rate": 1.902345341059164e-05, + "loss": 1.1297, + "step": 7114 + }, + { + "epoch": 0.17, + "grad_norm": 1.8428240013816861, + "learning_rate": 1.9023124503105424e-05, + "loss": 0.9841, + "step": 7115 + }, + { + "epoch": 0.17, + "grad_norm": 1.9728765872142002, + "learning_rate": 1.9022795543083722e-05, + "loss": 1.1835, + "step": 7116 + }, + { + "epoch": 0.17, + "grad_norm": 2.268384743186186, + "learning_rate": 1.9022466530528448e-05, + "loss": 1.0557, + "step": 7117 + }, + { + "epoch": 0.17, + "grad_norm": 2.1759144033441333, + "learning_rate": 1.9022137465441518e-05, + "loss": 0.9549, + "step": 7118 + }, + { + "epoch": 0.17, + "grad_norm": 1.907466599257256, + "learning_rate": 1.9021808347824847e-05, + "loss": 1.079, + "step": 7119 + }, + { + "epoch": 0.17, + "grad_norm": 2.38225091751849, + "learning_rate": 1.902147917768035e-05, + "loss": 0.9328, + "step": 7120 + }, + { + "epoch": 0.17, + "grad_norm": 2.024044420207077, + "learning_rate": 1.9021149955009943e-05, + "loss": 0.998, + "step": 7121 + }, + { + "epoch": 0.17, + "grad_norm": 2.0641960653468137, + "learning_rate": 1.902082067981555e-05, + "loss": 1.0371, + "step": 7122 + }, + { + "epoch": 0.17, + "grad_norm": 2.12199017206182, + "learning_rate": 1.902049135209908e-05, + "loss": 1.0988, + "step": 7123 + }, + { + "epoch": 0.17, + "grad_norm": 1.8940992632173428, + "learning_rate": 1.9020161971862454e-05, + "loss": 1.0041, + "step": 7124 + }, + { + "epoch": 0.17, + "grad_norm": 2.0612493011540365, + "learning_rate": 1.9019832539107592e-05, + "loss": 1.0595, + "step": 7125 + }, + { + "epoch": 0.17, + "grad_norm": 2.1915802022797157, + "learning_rate": 1.9019503053836406e-05, + "loss": 1.0492, + "step": 7126 + }, + { + "epoch": 0.17, + "grad_norm": 1.961805546884457, + "learning_rate": 1.901917351605082e-05, + "loss": 1.0906, + "step": 7127 + }, + { + "epoch": 0.17, + "grad_norm": 2.1540456746009577, + "learning_rate": 1.901884392575275e-05, + "loss": 1.0289, + "step": 7128 + }, + { + "epoch": 0.17, + "grad_norm": 1.6719711959957932, + "learning_rate": 1.9018514282944115e-05, + "loss": 0.9142, + "step": 7129 + }, + { + "epoch": 0.17, + "grad_norm": 2.0260096480447225, + "learning_rate": 1.901818458762683e-05, + "loss": 1.0946, + "step": 7130 + }, + { + "epoch": 0.17, + "grad_norm": 2.049001913693499, + "learning_rate": 1.9017854839802824e-05, + "loss": 0.9945, + "step": 7131 + }, + { + "epoch": 0.17, + "grad_norm": 2.542480404829827, + "learning_rate": 1.9017525039474013e-05, + "loss": 0.8847, + "step": 7132 + }, + { + "epoch": 0.17, + "grad_norm": 1.2429503392364185, + "learning_rate": 1.9017195186642315e-05, + "loss": 0.9911, + "step": 7133 + }, + { + "epoch": 0.17, + "grad_norm": 2.9676133557006295, + "learning_rate": 1.9016865281309652e-05, + "loss": 0.9897, + "step": 7134 + }, + { + "epoch": 0.17, + "grad_norm": 2.67116159846001, + "learning_rate": 1.9016535323477945e-05, + "loss": 0.9805, + "step": 7135 + }, + { + "epoch": 0.17, + "grad_norm": 2.1575016285207016, + "learning_rate": 1.9016205313149114e-05, + "loss": 1.0368, + "step": 7136 + }, + { + "epoch": 0.17, + "grad_norm": 2.095450447289131, + "learning_rate": 1.901587525032508e-05, + "loss": 1.0962, + "step": 7137 + }, + { + "epoch": 0.17, + "grad_norm": 2.3315834101819513, + "learning_rate": 1.9015545135007766e-05, + "loss": 1.0411, + "step": 7138 + }, + { + "epoch": 0.17, + "grad_norm": 1.9787564598258265, + "learning_rate": 1.9015214967199095e-05, + "loss": 1.1066, + "step": 7139 + }, + { + "epoch": 0.17, + "grad_norm": 2.0552982462410054, + "learning_rate": 1.901488474690099e-05, + "loss": 1.0337, + "step": 7140 + }, + { + "epoch": 0.17, + "grad_norm": 1.0930428965322116, + "learning_rate": 1.901455447411537e-05, + "loss": 0.9549, + "step": 7141 + }, + { + "epoch": 0.17, + "grad_norm": 1.740395640704928, + "learning_rate": 1.901422414884416e-05, + "loss": 1.1354, + "step": 7142 + }, + { + "epoch": 0.17, + "grad_norm": 2.279261968986075, + "learning_rate": 1.9013893771089283e-05, + "loss": 1.0463, + "step": 7143 + }, + { + "epoch": 0.17, + "grad_norm": 2.1251809221973708, + "learning_rate": 1.9013563340852664e-05, + "loss": 1.08, + "step": 7144 + }, + { + "epoch": 0.17, + "grad_norm": 2.13088609169943, + "learning_rate": 1.9013232858136225e-05, + "loss": 1.094, + "step": 7145 + }, + { + "epoch": 0.17, + "grad_norm": 1.9224936239684773, + "learning_rate": 1.901290232294189e-05, + "loss": 1.0722, + "step": 7146 + }, + { + "epoch": 0.17, + "grad_norm": 1.9933865753327409, + "learning_rate": 1.9012571735271584e-05, + "loss": 0.8533, + "step": 7147 + }, + { + "epoch": 0.17, + "grad_norm": 2.121455386763956, + "learning_rate": 1.9012241095127234e-05, + "loss": 1.1361, + "step": 7148 + }, + { + "epoch": 0.17, + "grad_norm": 2.232159675248471, + "learning_rate": 1.9011910402510762e-05, + "loss": 1.1526, + "step": 7149 + }, + { + "epoch": 0.17, + "grad_norm": 2.4468229377501247, + "learning_rate": 1.9011579657424092e-05, + "loss": 1.1436, + "step": 7150 + }, + { + "epoch": 0.17, + "grad_norm": 1.1677538967587162, + "learning_rate": 1.9011248859869155e-05, + "loss": 0.953, + "step": 7151 + }, + { + "epoch": 0.17, + "grad_norm": 2.0222502948252865, + "learning_rate": 1.9010918009847877e-05, + "loss": 1.1621, + "step": 7152 + }, + { + "epoch": 0.17, + "grad_norm": 2.0953084325368674, + "learning_rate": 1.9010587107362177e-05, + "loss": 0.9691, + "step": 7153 + }, + { + "epoch": 0.17, + "grad_norm": 2.6233910947572614, + "learning_rate": 1.9010256152413987e-05, + "loss": 1.0845, + "step": 7154 + }, + { + "epoch": 0.17, + "grad_norm": 2.009531754161707, + "learning_rate": 1.900992514500523e-05, + "loss": 0.8956, + "step": 7155 + }, + { + "epoch": 0.17, + "grad_norm": 1.0527803133653342, + "learning_rate": 1.900959408513784e-05, + "loss": 0.9359, + "step": 7156 + }, + { + "epoch": 0.17, + "grad_norm": 1.9765731922315621, + "learning_rate": 1.9009262972813746e-05, + "loss": 1.1188, + "step": 7157 + }, + { + "epoch": 0.17, + "grad_norm": 2.428143306105327, + "learning_rate": 1.9008931808034864e-05, + "loss": 1.0762, + "step": 7158 + }, + { + "epoch": 0.17, + "grad_norm": 2.0327196175567077, + "learning_rate": 1.900860059080313e-05, + "loss": 1.0917, + "step": 7159 + }, + { + "epoch": 0.17, + "grad_norm": 2.067330075709898, + "learning_rate": 1.900826932112047e-05, + "loss": 0.9011, + "step": 7160 + }, + { + "epoch": 0.17, + "grad_norm": 1.9508739318894992, + "learning_rate": 1.9007937998988818e-05, + "loss": 1.0899, + "step": 7161 + }, + { + "epoch": 0.17, + "grad_norm": 2.524707206255668, + "learning_rate": 1.9007606624410094e-05, + "loss": 1.049, + "step": 7162 + }, + { + "epoch": 0.17, + "grad_norm": 2.6866035841201774, + "learning_rate": 1.9007275197386238e-05, + "loss": 1.0549, + "step": 7163 + }, + { + "epoch": 0.17, + "grad_norm": 1.8578348225974144, + "learning_rate": 1.900694371791917e-05, + "loss": 1.2822, + "step": 7164 + }, + { + "epoch": 0.17, + "grad_norm": 2.1080587373158086, + "learning_rate": 1.9006612186010825e-05, + "loss": 1.2087, + "step": 7165 + }, + { + "epoch": 0.17, + "grad_norm": 1.9808752419397255, + "learning_rate": 1.9006280601663132e-05, + "loss": 0.9986, + "step": 7166 + }, + { + "epoch": 0.17, + "grad_norm": 1.969432066707196, + "learning_rate": 1.9005948964878024e-05, + "loss": 1.1316, + "step": 7167 + }, + { + "epoch": 0.17, + "grad_norm": 2.0291672236997376, + "learning_rate": 1.900561727565743e-05, + "loss": 1.0285, + "step": 7168 + }, + { + "epoch": 0.17, + "grad_norm": 2.1037996790122047, + "learning_rate": 1.900528553400328e-05, + "loss": 1.0393, + "step": 7169 + }, + { + "epoch": 0.17, + "grad_norm": 2.3213771422008382, + "learning_rate": 1.9004953739917503e-05, + "loss": 0.9609, + "step": 7170 + }, + { + "epoch": 0.17, + "grad_norm": 1.1828369189284063, + "learning_rate": 1.900462189340204e-05, + "loss": 0.9454, + "step": 7171 + }, + { + "epoch": 0.17, + "grad_norm": 2.3846047142698947, + "learning_rate": 1.9004289994458815e-05, + "loss": 1.1149, + "step": 7172 + }, + { + "epoch": 0.17, + "grad_norm": 2.1179808755178042, + "learning_rate": 1.9003958043089763e-05, + "loss": 1.0953, + "step": 7173 + }, + { + "epoch": 0.17, + "grad_norm": 2.030364890467749, + "learning_rate": 1.900362603929682e-05, + "loss": 0.9483, + "step": 7174 + }, + { + "epoch": 0.17, + "grad_norm": 2.2235197391165684, + "learning_rate": 1.900329398308191e-05, + "loss": 1.0583, + "step": 7175 + }, + { + "epoch": 0.17, + "grad_norm": 2.1598533796154262, + "learning_rate": 1.9002961874446973e-05, + "loss": 1.1425, + "step": 7176 + }, + { + "epoch": 0.17, + "grad_norm": 1.9399840833663353, + "learning_rate": 1.9002629713393946e-05, + "loss": 1.0169, + "step": 7177 + }, + { + "epoch": 0.17, + "grad_norm": 2.145245772200579, + "learning_rate": 1.9002297499924752e-05, + "loss": 1.0706, + "step": 7178 + }, + { + "epoch": 0.17, + "grad_norm": 2.1995246326809434, + "learning_rate": 1.900196523404134e-05, + "loss": 1.0516, + "step": 7179 + }, + { + "epoch": 0.17, + "grad_norm": 2.1219638194508867, + "learning_rate": 1.900163291574563e-05, + "loss": 1.0363, + "step": 7180 + }, + { + "epoch": 0.17, + "grad_norm": 2.18096939670653, + "learning_rate": 1.9001300545039566e-05, + "loss": 1.0435, + "step": 7181 + }, + { + "epoch": 0.17, + "grad_norm": 2.0947339345673344, + "learning_rate": 1.900096812192508e-05, + "loss": 1.0694, + "step": 7182 + }, + { + "epoch": 0.17, + "grad_norm": 2.023803195991269, + "learning_rate": 1.9000635646404108e-05, + "loss": 1.0737, + "step": 7183 + }, + { + "epoch": 0.17, + "grad_norm": 2.0239463844216963, + "learning_rate": 1.9000303118478584e-05, + "loss": 0.9964, + "step": 7184 + }, + { + "epoch": 0.17, + "grad_norm": 2.0466461397788507, + "learning_rate": 1.899997053815045e-05, + "loss": 1.0008, + "step": 7185 + }, + { + "epoch": 0.17, + "grad_norm": 1.9167185544125984, + "learning_rate": 1.8999637905421633e-05, + "loss": 1.0988, + "step": 7186 + }, + { + "epoch": 0.17, + "grad_norm": 1.9535251922693515, + "learning_rate": 1.899930522029408e-05, + "loss": 1.1025, + "step": 7187 + }, + { + "epoch": 0.17, + "grad_norm": 1.2031882398336775, + "learning_rate": 1.899897248276972e-05, + "loss": 0.9686, + "step": 7188 + }, + { + "epoch": 0.17, + "grad_norm": 2.201156595848961, + "learning_rate": 1.8998639692850492e-05, + "loss": 1.1514, + "step": 7189 + }, + { + "epoch": 0.17, + "grad_norm": 2.0874285061630014, + "learning_rate": 1.899830685053834e-05, + "loss": 1.097, + "step": 7190 + }, + { + "epoch": 0.17, + "grad_norm": 2.09694008382878, + "learning_rate": 1.8997973955835193e-05, + "loss": 1.2092, + "step": 7191 + }, + { + "epoch": 0.17, + "grad_norm": 2.0374892945525795, + "learning_rate": 1.899764100874299e-05, + "loss": 1.0586, + "step": 7192 + }, + { + "epoch": 0.17, + "grad_norm": 2.233112244936261, + "learning_rate": 1.8997308009263675e-05, + "loss": 1.0731, + "step": 7193 + }, + { + "epoch": 0.17, + "grad_norm": 2.14338693429423, + "learning_rate": 1.8996974957399186e-05, + "loss": 1.0004, + "step": 7194 + }, + { + "epoch": 0.17, + "grad_norm": 2.154846892649168, + "learning_rate": 1.8996641853151462e-05, + "loss": 1.0616, + "step": 7195 + }, + { + "epoch": 0.17, + "grad_norm": 2.221567844572955, + "learning_rate": 1.8996308696522435e-05, + "loss": 1.1146, + "step": 7196 + }, + { + "epoch": 0.17, + "grad_norm": 2.0839357749175345, + "learning_rate": 1.8995975487514052e-05, + "loss": 1.2341, + "step": 7197 + }, + { + "epoch": 0.17, + "grad_norm": 2.149744507780891, + "learning_rate": 1.8995642226128256e-05, + "loss": 0.9429, + "step": 7198 + }, + { + "epoch": 0.17, + "grad_norm": 1.9294547266856379, + "learning_rate": 1.899530891236698e-05, + "loss": 1.0717, + "step": 7199 + }, + { + "epoch": 0.17, + "grad_norm": 1.9512654818790711, + "learning_rate": 1.8994975546232166e-05, + "loss": 0.9198, + "step": 7200 + }, + { + "epoch": 0.17, + "grad_norm": 1.9602432373167789, + "learning_rate": 1.899464212772576e-05, + "loss": 0.951, + "step": 7201 + }, + { + "epoch": 0.17, + "grad_norm": 1.1397572208466387, + "learning_rate": 1.8994308656849697e-05, + "loss": 1.0012, + "step": 7202 + }, + { + "epoch": 0.17, + "grad_norm": 2.0081806591544598, + "learning_rate": 1.8993975133605925e-05, + "loss": 0.9171, + "step": 7203 + }, + { + "epoch": 0.17, + "grad_norm": 1.9212641504835926, + "learning_rate": 1.8993641557996377e-05, + "loss": 1.2019, + "step": 7204 + }, + { + "epoch": 0.17, + "grad_norm": 2.2516166104938122, + "learning_rate": 1.8993307930023005e-05, + "loss": 1.1215, + "step": 7205 + }, + { + "epoch": 0.17, + "grad_norm": 2.419253911372209, + "learning_rate": 1.8992974249687747e-05, + "loss": 1.1536, + "step": 7206 + }, + { + "epoch": 0.17, + "grad_norm": 2.0993871717641017, + "learning_rate": 1.8992640516992544e-05, + "loss": 1.0231, + "step": 7207 + }, + { + "epoch": 0.17, + "grad_norm": 2.2171760854417166, + "learning_rate": 1.899230673193934e-05, + "loss": 1.1085, + "step": 7208 + }, + { + "epoch": 0.17, + "grad_norm": 2.001867349229199, + "learning_rate": 1.899197289453008e-05, + "loss": 0.9695, + "step": 7209 + }, + { + "epoch": 0.17, + "grad_norm": 1.159940434621682, + "learning_rate": 1.8991639004766708e-05, + "loss": 0.999, + "step": 7210 + }, + { + "epoch": 0.17, + "grad_norm": 1.826951390166292, + "learning_rate": 1.8991305062651167e-05, + "loss": 0.9808, + "step": 7211 + }, + { + "epoch": 0.17, + "grad_norm": 1.0643661625362797, + "learning_rate": 1.8990971068185402e-05, + "loss": 0.9863, + "step": 7212 + }, + { + "epoch": 0.17, + "grad_norm": 2.067573785070113, + "learning_rate": 1.8990637021371355e-05, + "loss": 1.0301, + "step": 7213 + }, + { + "epoch": 0.17, + "grad_norm": 1.937334787387039, + "learning_rate": 1.8990302922210976e-05, + "loss": 1.1074, + "step": 7214 + }, + { + "epoch": 0.17, + "grad_norm": 2.1187446533027705, + "learning_rate": 1.89899687707062e-05, + "loss": 0.9712, + "step": 7215 + }, + { + "epoch": 0.17, + "grad_norm": 2.206144985501457, + "learning_rate": 1.8989634566858987e-05, + "loss": 1.0756, + "step": 7216 + }, + { + "epoch": 0.17, + "grad_norm": 2.243029343965123, + "learning_rate": 1.8989300310671273e-05, + "loss": 1.1693, + "step": 7217 + }, + { + "epoch": 0.17, + "grad_norm": 2.3195489080928717, + "learning_rate": 1.8988966002145007e-05, + "loss": 1.0788, + "step": 7218 + }, + { + "epoch": 0.17, + "grad_norm": 1.982831711287126, + "learning_rate": 1.8988631641282134e-05, + "loss": 1.1362, + "step": 7219 + }, + { + "epoch": 0.17, + "grad_norm": 2.2123846447172912, + "learning_rate": 1.89882972280846e-05, + "loss": 0.9576, + "step": 7220 + }, + { + "epoch": 0.17, + "grad_norm": 2.2170961257844897, + "learning_rate": 1.8987962762554358e-05, + "loss": 1.2581, + "step": 7221 + }, + { + "epoch": 0.17, + "grad_norm": 2.4303480057204667, + "learning_rate": 1.898762824469335e-05, + "loss": 1.135, + "step": 7222 + }, + { + "epoch": 0.17, + "grad_norm": 1.8428743209222365, + "learning_rate": 1.8987293674503525e-05, + "loss": 1.1269, + "step": 7223 + }, + { + "epoch": 0.17, + "grad_norm": 1.1730206043764553, + "learning_rate": 1.898695905198683e-05, + "loss": 0.967, + "step": 7224 + }, + { + "epoch": 0.17, + "grad_norm": 1.9897524939475069, + "learning_rate": 1.8986624377145216e-05, + "loss": 0.901, + "step": 7225 + }, + { + "epoch": 0.17, + "grad_norm": 2.7139313517957495, + "learning_rate": 1.8986289649980627e-05, + "loss": 1.0396, + "step": 7226 + }, + { + "epoch": 0.17, + "grad_norm": 2.172737169147703, + "learning_rate": 1.8985954870495016e-05, + "loss": 1.0234, + "step": 7227 + }, + { + "epoch": 0.17, + "grad_norm": 2.3024699973999923, + "learning_rate": 1.8985620038690332e-05, + "loss": 1.0191, + "step": 7228 + }, + { + "epoch": 0.17, + "grad_norm": 2.165111367937224, + "learning_rate": 1.8985285154568522e-05, + "loss": 1.0388, + "step": 7229 + }, + { + "epoch": 0.17, + "grad_norm": 2.762939640876728, + "learning_rate": 1.8984950218131538e-05, + "loss": 1.087, + "step": 7230 + }, + { + "epoch": 0.17, + "grad_norm": 2.3258617401170283, + "learning_rate": 1.8984615229381325e-05, + "loss": 1.0124, + "step": 7231 + }, + { + "epoch": 0.17, + "grad_norm": 2.377492505525473, + "learning_rate": 1.8984280188319845e-05, + "loss": 0.9668, + "step": 7232 + }, + { + "epoch": 0.17, + "grad_norm": 2.360130019961199, + "learning_rate": 1.8983945094949035e-05, + "loss": 1.0587, + "step": 7233 + }, + { + "epoch": 0.17, + "grad_norm": 2.064863915769826, + "learning_rate": 1.8983609949270854e-05, + "loss": 1.1088, + "step": 7234 + }, + { + "epoch": 0.17, + "grad_norm": 1.8868764271905276, + "learning_rate": 1.8983274751287254e-05, + "loss": 1.0837, + "step": 7235 + }, + { + "epoch": 0.17, + "grad_norm": 2.0802375449174253, + "learning_rate": 1.8982939501000182e-05, + "loss": 0.945, + "step": 7236 + }, + { + "epoch": 0.17, + "grad_norm": 2.069815577506754, + "learning_rate": 1.8982604198411594e-05, + "loss": 1.0892, + "step": 7237 + }, + { + "epoch": 0.17, + "grad_norm": 2.1708812766570182, + "learning_rate": 1.898226884352344e-05, + "loss": 1.0549, + "step": 7238 + }, + { + "epoch": 0.17, + "grad_norm": 2.5933873845176705, + "learning_rate": 1.8981933436337675e-05, + "loss": 0.8839, + "step": 7239 + }, + { + "epoch": 0.17, + "grad_norm": 2.1252861596149017, + "learning_rate": 1.8981597976856247e-05, + "loss": 1.0253, + "step": 7240 + }, + { + "epoch": 0.17, + "grad_norm": 1.1153935387334128, + "learning_rate": 1.8981262465081112e-05, + "loss": 1.0138, + "step": 7241 + }, + { + "epoch": 0.17, + "grad_norm": 1.885462725524725, + "learning_rate": 1.8980926901014226e-05, + "loss": 0.9953, + "step": 7242 + }, + { + "epoch": 0.17, + "grad_norm": 1.922762566134816, + "learning_rate": 1.8980591284657536e-05, + "loss": 1.0968, + "step": 7243 + }, + { + "epoch": 0.17, + "grad_norm": 2.5771448733249835, + "learning_rate": 1.8980255616013004e-05, + "loss": 1.1794, + "step": 7244 + }, + { + "epoch": 0.17, + "grad_norm": 2.0183470269239456, + "learning_rate": 1.8979919895082578e-05, + "loss": 1.0301, + "step": 7245 + }, + { + "epoch": 0.17, + "grad_norm": 2.000215850065701, + "learning_rate": 1.8979584121868217e-05, + "loss": 1.0578, + "step": 7246 + }, + { + "epoch": 0.17, + "grad_norm": 3.0295622882746343, + "learning_rate": 1.8979248296371876e-05, + "loss": 1.0699, + "step": 7247 + }, + { + "epoch": 0.17, + "grad_norm": 2.115161141693514, + "learning_rate": 1.897891241859551e-05, + "loss": 0.943, + "step": 7248 + }, + { + "epoch": 0.17, + "grad_norm": 2.2444327936559865, + "learning_rate": 1.8978576488541068e-05, + "loss": 1.0805, + "step": 7249 + }, + { + "epoch": 0.17, + "grad_norm": 2.029331250526032, + "learning_rate": 1.897824050621051e-05, + "loss": 1.0782, + "step": 7250 + }, + { + "epoch": 0.17, + "grad_norm": 1.9981982259115727, + "learning_rate": 1.8977904471605798e-05, + "loss": 1.0259, + "step": 7251 + }, + { + "epoch": 0.17, + "grad_norm": 2.2689435445363735, + "learning_rate": 1.8977568384728885e-05, + "loss": 1.1543, + "step": 7252 + }, + { + "epoch": 0.17, + "grad_norm": 1.9800506609928976, + "learning_rate": 1.8977232245581722e-05, + "loss": 0.9916, + "step": 7253 + }, + { + "epoch": 0.17, + "grad_norm": 2.081699218831782, + "learning_rate": 1.8976896054166273e-05, + "loss": 1.1401, + "step": 7254 + }, + { + "epoch": 0.17, + "grad_norm": 2.2424748038102553, + "learning_rate": 1.8976559810484494e-05, + "loss": 1.1384, + "step": 7255 + }, + { + "epoch": 0.17, + "grad_norm": 3.003942169334653, + "learning_rate": 1.8976223514538342e-05, + "loss": 0.8784, + "step": 7256 + }, + { + "epoch": 0.17, + "grad_norm": 2.3128326086606386, + "learning_rate": 1.8975887166329773e-05, + "loss": 1.0053, + "step": 7257 + }, + { + "epoch": 0.17, + "grad_norm": 2.1585446556240946, + "learning_rate": 1.8975550765860745e-05, + "loss": 0.9404, + "step": 7258 + }, + { + "epoch": 0.17, + "grad_norm": 2.024363262143625, + "learning_rate": 1.8975214313133223e-05, + "loss": 1.1708, + "step": 7259 + }, + { + "epoch": 0.17, + "grad_norm": 1.9613273548190426, + "learning_rate": 1.8974877808149162e-05, + "loss": 1.0025, + "step": 7260 + }, + { + "epoch": 0.17, + "grad_norm": 2.4878599185002472, + "learning_rate": 1.897454125091052e-05, + "loss": 1.1321, + "step": 7261 + }, + { + "epoch": 0.17, + "grad_norm": 2.9079674998550953, + "learning_rate": 1.8974204641419256e-05, + "loss": 1.2124, + "step": 7262 + }, + { + "epoch": 0.17, + "grad_norm": 2.2035903642081074, + "learning_rate": 1.8973867979677332e-05, + "loss": 0.9424, + "step": 7263 + }, + { + "epoch": 0.17, + "grad_norm": 2.1920101308857243, + "learning_rate": 1.8973531265686706e-05, + "loss": 0.9407, + "step": 7264 + }, + { + "epoch": 0.17, + "grad_norm": 2.0880343531132386, + "learning_rate": 1.8973194499449342e-05, + "loss": 0.9318, + "step": 7265 + }, + { + "epoch": 0.17, + "grad_norm": 1.1658691752779222, + "learning_rate": 1.89728576809672e-05, + "loss": 1.043, + "step": 7266 + }, + { + "epoch": 0.17, + "grad_norm": 2.279559894559991, + "learning_rate": 1.897252081024224e-05, + "loss": 1.0519, + "step": 7267 + }, + { + "epoch": 0.17, + "grad_norm": 2.2892615464480053, + "learning_rate": 1.8972183887276422e-05, + "loss": 0.9952, + "step": 7268 + }, + { + "epoch": 0.17, + "grad_norm": 2.3129786587225687, + "learning_rate": 1.8971846912071707e-05, + "loss": 1.234, + "step": 7269 + }, + { + "epoch": 0.17, + "grad_norm": 2.381842128554803, + "learning_rate": 1.897150988463006e-05, + "loss": 1.0275, + "step": 7270 + }, + { + "epoch": 0.17, + "grad_norm": 1.1545098890019763, + "learning_rate": 1.8971172804953444e-05, + "loss": 1.0252, + "step": 7271 + }, + { + "epoch": 0.17, + "grad_norm": 1.8399381176777583, + "learning_rate": 1.897083567304382e-05, + "loss": 1.0671, + "step": 7272 + }, + { + "epoch": 0.17, + "grad_norm": 1.9741577353321533, + "learning_rate": 1.8970498488903147e-05, + "loss": 0.9926, + "step": 7273 + }, + { + "epoch": 0.17, + "grad_norm": 3.0023570047316, + "learning_rate": 1.8970161252533393e-05, + "loss": 1.0948, + "step": 7274 + }, + { + "epoch": 0.17, + "grad_norm": 1.888453920475029, + "learning_rate": 1.8969823963936524e-05, + "loss": 1.0276, + "step": 7275 + }, + { + "epoch": 0.17, + "grad_norm": 1.9747671733184329, + "learning_rate": 1.8969486623114496e-05, + "loss": 0.9711, + "step": 7276 + }, + { + "epoch": 0.17, + "grad_norm": 2.259600100496921, + "learning_rate": 1.8969149230069278e-05, + "loss": 0.9532, + "step": 7277 + }, + { + "epoch": 0.17, + "grad_norm": 2.4906507225358125, + "learning_rate": 1.8968811784802833e-05, + "loss": 1.0986, + "step": 7278 + }, + { + "epoch": 0.17, + "grad_norm": 1.9683397689746798, + "learning_rate": 1.896847428731713e-05, + "loss": 0.9713, + "step": 7279 + }, + { + "epoch": 0.17, + "grad_norm": 2.1007500115120523, + "learning_rate": 1.8968136737614123e-05, + "loss": 0.9601, + "step": 7280 + }, + { + "epoch": 0.17, + "grad_norm": 1.818442540976307, + "learning_rate": 1.8967799135695788e-05, + "loss": 1.0041, + "step": 7281 + }, + { + "epoch": 0.17, + "grad_norm": 2.222405230086059, + "learning_rate": 1.896746148156409e-05, + "loss": 1.13, + "step": 7282 + }, + { + "epoch": 0.17, + "grad_norm": 2.7044135801739335, + "learning_rate": 1.896712377522099e-05, + "loss": 1.2034, + "step": 7283 + }, + { + "epoch": 0.17, + "grad_norm": 2.643063404984306, + "learning_rate": 1.8966786016668458e-05, + "loss": 1.1685, + "step": 7284 + }, + { + "epoch": 0.17, + "grad_norm": 2.262105423748964, + "learning_rate": 1.8966448205908458e-05, + "loss": 1.0226, + "step": 7285 + }, + { + "epoch": 0.17, + "grad_norm": 2.122214818643327, + "learning_rate": 1.8966110342942954e-05, + "loss": 1.0626, + "step": 7286 + }, + { + "epoch": 0.17, + "grad_norm": 2.1026550152804515, + "learning_rate": 1.8965772427773923e-05, + "loss": 0.9651, + "step": 7287 + }, + { + "epoch": 0.17, + "grad_norm": 2.619312425198114, + "learning_rate": 1.8965434460403325e-05, + "loss": 1.0774, + "step": 7288 + }, + { + "epoch": 0.17, + "grad_norm": 2.567770316740899, + "learning_rate": 1.896509644083313e-05, + "loss": 1.1376, + "step": 7289 + }, + { + "epoch": 0.17, + "grad_norm": 2.4497080950804073, + "learning_rate": 1.8964758369065303e-05, + "loss": 1.1994, + "step": 7290 + }, + { + "epoch": 0.17, + "grad_norm": 1.9689928865399184, + "learning_rate": 1.8964420245101816e-05, + "loss": 0.9857, + "step": 7291 + }, + { + "epoch": 0.17, + "grad_norm": 2.2609300434425235, + "learning_rate": 1.8964082068944634e-05, + "loss": 1.1525, + "step": 7292 + }, + { + "epoch": 0.17, + "grad_norm": 2.2498443531260692, + "learning_rate": 1.896374384059573e-05, + "loss": 1.055, + "step": 7293 + }, + { + "epoch": 0.17, + "grad_norm": 1.1513349981152012, + "learning_rate": 1.896340556005707e-05, + "loss": 1.0223, + "step": 7294 + }, + { + "epoch": 0.17, + "grad_norm": 2.114722475810913, + "learning_rate": 1.8963067227330624e-05, + "loss": 1.1191, + "step": 7295 + }, + { + "epoch": 0.17, + "grad_norm": 2.485790001918956, + "learning_rate": 1.8962728842418366e-05, + "loss": 1.0023, + "step": 7296 + }, + { + "epoch": 0.17, + "grad_norm": 2.310328842107481, + "learning_rate": 1.8962390405322263e-05, + "loss": 1.022, + "step": 7297 + }, + { + "epoch": 0.17, + "grad_norm": 1.8968523162243742, + "learning_rate": 1.8962051916044288e-05, + "loss": 1.1316, + "step": 7298 + }, + { + "epoch": 0.17, + "grad_norm": 2.251365405132713, + "learning_rate": 1.8961713374586403e-05, + "loss": 1.0948, + "step": 7299 + }, + { + "epoch": 0.17, + "grad_norm": 2.130967732614994, + "learning_rate": 1.8961374780950588e-05, + "loss": 0.9993, + "step": 7300 + }, + { + "epoch": 0.17, + "grad_norm": 1.7926760001581672, + "learning_rate": 1.8961036135138813e-05, + "loss": 1.0287, + "step": 7301 + }, + { + "epoch": 0.17, + "grad_norm": 2.3665317144000935, + "learning_rate": 1.8960697437153048e-05, + "loss": 1.0933, + "step": 7302 + }, + { + "epoch": 0.17, + "grad_norm": 2.271939252398364, + "learning_rate": 1.8960358686995265e-05, + "loss": 1.0752, + "step": 7303 + }, + { + "epoch": 0.17, + "grad_norm": 2.210221930274947, + "learning_rate": 1.896001988466744e-05, + "loss": 1.177, + "step": 7304 + }, + { + "epoch": 0.17, + "grad_norm": 1.984037269411808, + "learning_rate": 1.895968103017154e-05, + "loss": 1.0232, + "step": 7305 + }, + { + "epoch": 0.17, + "grad_norm": 2.283425498131847, + "learning_rate": 1.8959342123509538e-05, + "loss": 0.9953, + "step": 7306 + }, + { + "epoch": 0.17, + "grad_norm": 2.8866773097144196, + "learning_rate": 1.8959003164683416e-05, + "loss": 0.9905, + "step": 7307 + }, + { + "epoch": 0.17, + "grad_norm": 2.9098367901327946, + "learning_rate": 1.8958664153695138e-05, + "loss": 1.1523, + "step": 7308 + }, + { + "epoch": 0.17, + "grad_norm": 2.1370062303413775, + "learning_rate": 1.895832509054668e-05, + "loss": 0.9908, + "step": 7309 + }, + { + "epoch": 0.17, + "grad_norm": 2.323855186625609, + "learning_rate": 1.8957985975240016e-05, + "loss": 0.9646, + "step": 7310 + }, + { + "epoch": 0.17, + "grad_norm": 2.6437832378069035, + "learning_rate": 1.8957646807777123e-05, + "loss": 1.1206, + "step": 7311 + }, + { + "epoch": 0.17, + "grad_norm": 1.795362308498506, + "learning_rate": 1.8957307588159978e-05, + "loss": 1.1664, + "step": 7312 + }, + { + "epoch": 0.17, + "grad_norm": 1.9963374016000643, + "learning_rate": 1.8956968316390547e-05, + "loss": 1.2329, + "step": 7313 + }, + { + "epoch": 0.17, + "grad_norm": 2.370675116005996, + "learning_rate": 1.8956628992470812e-05, + "loss": 1.1392, + "step": 7314 + }, + { + "epoch": 0.17, + "grad_norm": 1.2076610494983577, + "learning_rate": 1.895628961640275e-05, + "loss": 0.9675, + "step": 7315 + }, + { + "epoch": 0.17, + "grad_norm": 2.129417168156437, + "learning_rate": 1.895595018818833e-05, + "loss": 1.1189, + "step": 7316 + }, + { + "epoch": 0.17, + "grad_norm": 2.294367348858692, + "learning_rate": 1.8955610707829536e-05, + "loss": 0.9024, + "step": 7317 + }, + { + "epoch": 0.17, + "grad_norm": 1.9296035599333277, + "learning_rate": 1.895527117532834e-05, + "loss": 1.0006, + "step": 7318 + }, + { + "epoch": 0.17, + "grad_norm": 2.144864545216464, + "learning_rate": 1.8954931590686722e-05, + "loss": 0.9238, + "step": 7319 + }, + { + "epoch": 0.17, + "grad_norm": 2.007565080138308, + "learning_rate": 1.8954591953906657e-05, + "loss": 0.9521, + "step": 7320 + }, + { + "epoch": 0.17, + "grad_norm": 2.6074458716749156, + "learning_rate": 1.8954252264990122e-05, + "loss": 1.1108, + "step": 7321 + }, + { + "epoch": 0.17, + "grad_norm": 2.6357480509651436, + "learning_rate": 1.895391252393909e-05, + "loss": 1.0996, + "step": 7322 + }, + { + "epoch": 0.17, + "grad_norm": 2.182306112292924, + "learning_rate": 1.8953572730755552e-05, + "loss": 1.0139, + "step": 7323 + }, + { + "epoch": 0.17, + "grad_norm": 2.3414032732979964, + "learning_rate": 1.8953232885441474e-05, + "loss": 1.0659, + "step": 7324 + }, + { + "epoch": 0.17, + "grad_norm": 2.1051940095118242, + "learning_rate": 1.8952892987998844e-05, + "loss": 1.2257, + "step": 7325 + }, + { + "epoch": 0.17, + "grad_norm": 2.025050806794159, + "learning_rate": 1.8952553038429634e-05, + "loss": 1.0702, + "step": 7326 + }, + { + "epoch": 0.17, + "grad_norm": 2.0444846895099684, + "learning_rate": 1.8952213036735826e-05, + "loss": 1.1804, + "step": 7327 + }, + { + "epoch": 0.17, + "grad_norm": 2.211768347106762, + "learning_rate": 1.8951872982919398e-05, + "loss": 0.9936, + "step": 7328 + }, + { + "epoch": 0.17, + "grad_norm": 2.1635397139096244, + "learning_rate": 1.8951532876982335e-05, + "loss": 1.0615, + "step": 7329 + }, + { + "epoch": 0.17, + "grad_norm": 1.972171642069124, + "learning_rate": 1.895119271892661e-05, + "loss": 1.0945, + "step": 7330 + }, + { + "epoch": 0.17, + "grad_norm": 1.9189226269897717, + "learning_rate": 1.895085250875421e-05, + "loss": 1.1299, + "step": 7331 + }, + { + "epoch": 0.17, + "grad_norm": 2.7472260751715543, + "learning_rate": 1.8950512246467113e-05, + "loss": 1.0791, + "step": 7332 + }, + { + "epoch": 0.17, + "grad_norm": 2.130738116853986, + "learning_rate": 1.8950171932067297e-05, + "loss": 1.0747, + "step": 7333 + }, + { + "epoch": 0.17, + "grad_norm": 1.9708029404627376, + "learning_rate": 1.8949831565556747e-05, + "loss": 1.106, + "step": 7334 + }, + { + "epoch": 0.17, + "grad_norm": 2.222489262519466, + "learning_rate": 1.8949491146937446e-05, + "loss": 1.0181, + "step": 7335 + }, + { + "epoch": 0.17, + "grad_norm": 2.1696775641794868, + "learning_rate": 1.8949150676211374e-05, + "loss": 1.1778, + "step": 7336 + }, + { + "epoch": 0.17, + "grad_norm": 2.038040044891145, + "learning_rate": 1.8948810153380514e-05, + "loss": 1.0189, + "step": 7337 + }, + { + "epoch": 0.17, + "grad_norm": 2.1072228358037166, + "learning_rate": 1.8948469578446845e-05, + "loss": 1.0177, + "step": 7338 + }, + { + "epoch": 0.17, + "grad_norm": 1.2166325502587576, + "learning_rate": 1.8948128951412356e-05, + "loss": 1.0239, + "step": 7339 + }, + { + "epoch": 0.17, + "grad_norm": 1.909477722583824, + "learning_rate": 1.8947788272279028e-05, + "loss": 1.0958, + "step": 7340 + }, + { + "epoch": 0.17, + "grad_norm": 1.0787115411698025, + "learning_rate": 1.8947447541048844e-05, + "loss": 0.957, + "step": 7341 + }, + { + "epoch": 0.17, + "grad_norm": 2.204647712856267, + "learning_rate": 1.8947106757723782e-05, + "loss": 1.0858, + "step": 7342 + }, + { + "epoch": 0.17, + "grad_norm": 2.1381119661427337, + "learning_rate": 1.8946765922305835e-05, + "loss": 0.9329, + "step": 7343 + }, + { + "epoch": 0.17, + "grad_norm": 2.180907389517777, + "learning_rate": 1.8946425034796988e-05, + "loss": 1.1009, + "step": 7344 + }, + { + "epoch": 0.17, + "grad_norm": 2.20948143109007, + "learning_rate": 1.8946084095199216e-05, + "loss": 1.0817, + "step": 7345 + }, + { + "epoch": 0.17, + "grad_norm": 2.0984544025842493, + "learning_rate": 1.8945743103514516e-05, + "loss": 0.9938, + "step": 7346 + }, + { + "epoch": 0.17, + "grad_norm": 2.3316521917594026, + "learning_rate": 1.8945402059744865e-05, + "loss": 1.0855, + "step": 7347 + }, + { + "epoch": 0.17, + "grad_norm": 1.129882886477759, + "learning_rate": 1.894506096389225e-05, + "loss": 0.9409, + "step": 7348 + }, + { + "epoch": 0.17, + "grad_norm": 2.2631291939494544, + "learning_rate": 1.894471981595866e-05, + "loss": 1.0063, + "step": 7349 + }, + { + "epoch": 0.17, + "grad_norm": 2.107095253688883, + "learning_rate": 1.8944378615946077e-05, + "loss": 1.1886, + "step": 7350 + }, + { + "epoch": 0.17, + "grad_norm": 5.951203114011631, + "learning_rate": 1.8944037363856492e-05, + "loss": 1.0161, + "step": 7351 + }, + { + "epoch": 0.17, + "grad_norm": 2.179737950582027, + "learning_rate": 1.894369605969189e-05, + "loss": 1.0321, + "step": 7352 + }, + { + "epoch": 0.17, + "grad_norm": 2.2549770071420974, + "learning_rate": 1.894335470345426e-05, + "loss": 0.98, + "step": 7353 + }, + { + "epoch": 0.17, + "grad_norm": 1.0803427581136629, + "learning_rate": 1.894301329514558e-05, + "loss": 0.9639, + "step": 7354 + }, + { + "epoch": 0.17, + "grad_norm": 2.9042278145619753, + "learning_rate": 1.894267183476785e-05, + "loss": 1.1409, + "step": 7355 + }, + { + "epoch": 0.17, + "grad_norm": 1.2476386192021973, + "learning_rate": 1.8942330322323055e-05, + "loss": 0.911, + "step": 7356 + }, + { + "epoch": 0.17, + "grad_norm": 2.600297239102879, + "learning_rate": 1.894198875781318e-05, + "loss": 1.1277, + "step": 7357 + }, + { + "epoch": 0.17, + "grad_norm": 2.18915352251765, + "learning_rate": 1.8941647141240213e-05, + "loss": 1.1603, + "step": 7358 + }, + { + "epoch": 0.17, + "grad_norm": 2.2185147235853218, + "learning_rate": 1.894130547260615e-05, + "loss": 1.0351, + "step": 7359 + }, + { + "epoch": 0.17, + "grad_norm": 2.664379452878225, + "learning_rate": 1.8940963751912976e-05, + "loss": 0.9944, + "step": 7360 + }, + { + "epoch": 0.17, + "grad_norm": 1.8557221244679643, + "learning_rate": 1.8940621979162678e-05, + "loss": 1.0624, + "step": 7361 + }, + { + "epoch": 0.17, + "grad_norm": 1.9813189607251644, + "learning_rate": 1.8940280154357248e-05, + "loss": 1.0041, + "step": 7362 + }, + { + "epoch": 0.17, + "grad_norm": 2.10402604027484, + "learning_rate": 1.8939938277498675e-05, + "loss": 1.0226, + "step": 7363 + }, + { + "epoch": 0.17, + "grad_norm": 2.14255123427998, + "learning_rate": 1.8939596348588955e-05, + "loss": 1.0149, + "step": 7364 + }, + { + "epoch": 0.17, + "grad_norm": 1.1249224175335244, + "learning_rate": 1.893925436763007e-05, + "loss": 0.9481, + "step": 7365 + }, + { + "epoch": 0.17, + "grad_norm": 2.195721589332145, + "learning_rate": 1.893891233462402e-05, + "loss": 1.112, + "step": 7366 + }, + { + "epoch": 0.17, + "grad_norm": 1.1285219936877764, + "learning_rate": 1.8938570249572793e-05, + "loss": 0.9887, + "step": 7367 + }, + { + "epoch": 0.17, + "grad_norm": 2.206932104941046, + "learning_rate": 1.8938228112478377e-05, + "loss": 1.2576, + "step": 7368 + }, + { + "epoch": 0.17, + "grad_norm": 2.0668483943462523, + "learning_rate": 1.8937885923342768e-05, + "loss": 0.9884, + "step": 7369 + }, + { + "epoch": 0.17, + "grad_norm": 1.1197887876084562, + "learning_rate": 1.893754368216796e-05, + "loss": 0.9115, + "step": 7370 + }, + { + "epoch": 0.17, + "grad_norm": 2.0551423380281615, + "learning_rate": 1.893720138895594e-05, + "loss": 1.0077, + "step": 7371 + }, + { + "epoch": 0.17, + "grad_norm": 2.2713339673600155, + "learning_rate": 1.8936859043708705e-05, + "loss": 1.0266, + "step": 7372 + }, + { + "epoch": 0.17, + "grad_norm": 2.086921382506532, + "learning_rate": 1.893651664642825e-05, + "loss": 1.0074, + "step": 7373 + }, + { + "epoch": 0.17, + "grad_norm": 2.089687252925407, + "learning_rate": 1.893617419711656e-05, + "loss": 1.1167, + "step": 7374 + }, + { + "epoch": 0.17, + "grad_norm": 2.1031756921120985, + "learning_rate": 1.893583169577564e-05, + "loss": 0.9993, + "step": 7375 + }, + { + "epoch": 0.17, + "grad_norm": 2.131113029949946, + "learning_rate": 1.8935489142407474e-05, + "loss": 1.0868, + "step": 7376 + }, + { + "epoch": 0.17, + "grad_norm": 2.240337695164532, + "learning_rate": 1.8935146537014064e-05, + "loss": 1.1857, + "step": 7377 + }, + { + "epoch": 0.17, + "grad_norm": 2.50603370699344, + "learning_rate": 1.8934803879597402e-05, + "loss": 1.0306, + "step": 7378 + }, + { + "epoch": 0.17, + "grad_norm": 2.0717821830290215, + "learning_rate": 1.8934461170159483e-05, + "loss": 1.1332, + "step": 7379 + }, + { + "epoch": 0.17, + "grad_norm": 1.944670184670545, + "learning_rate": 1.8934118408702306e-05, + "loss": 1.0693, + "step": 7380 + }, + { + "epoch": 0.17, + "grad_norm": 2.406122888673289, + "learning_rate": 1.8933775595227856e-05, + "loss": 1.0289, + "step": 7381 + }, + { + "epoch": 0.17, + "grad_norm": 2.0232336007626572, + "learning_rate": 1.893343272973814e-05, + "loss": 1.0817, + "step": 7382 + }, + { + "epoch": 0.17, + "grad_norm": 2.0705863584822097, + "learning_rate": 1.8933089812235153e-05, + "loss": 0.9824, + "step": 7383 + }, + { + "epoch": 0.17, + "grad_norm": 2.0165370269267466, + "learning_rate": 1.8932746842720884e-05, + "loss": 1.1102, + "step": 7384 + }, + { + "epoch": 0.17, + "grad_norm": 2.382073667376283, + "learning_rate": 1.8932403821197338e-05, + "loss": 1.0093, + "step": 7385 + }, + { + "epoch": 0.17, + "grad_norm": 2.159360944976313, + "learning_rate": 1.8932060747666504e-05, + "loss": 1.1642, + "step": 7386 + }, + { + "epoch": 0.17, + "grad_norm": 2.392560517187917, + "learning_rate": 1.8931717622130392e-05, + "loss": 1.0958, + "step": 7387 + }, + { + "epoch": 0.17, + "grad_norm": 2.0773092294197624, + "learning_rate": 1.8931374444590987e-05, + "loss": 1.0057, + "step": 7388 + }, + { + "epoch": 0.17, + "grad_norm": 1.987381632531391, + "learning_rate": 1.8931031215050295e-05, + "loss": 1.055, + "step": 7389 + }, + { + "epoch": 0.17, + "grad_norm": 2.3583364521987185, + "learning_rate": 1.893068793351031e-05, + "loss": 1.027, + "step": 7390 + }, + { + "epoch": 0.17, + "grad_norm": 2.2480182142526948, + "learning_rate": 1.8930344599973035e-05, + "loss": 1.1273, + "step": 7391 + }, + { + "epoch": 0.17, + "grad_norm": 2.0613817365494045, + "learning_rate": 1.8930001214440464e-05, + "loss": 1.0355, + "step": 7392 + }, + { + "epoch": 0.17, + "grad_norm": 2.79393962536526, + "learning_rate": 1.89296577769146e-05, + "loss": 1.0077, + "step": 7393 + }, + { + "epoch": 0.17, + "grad_norm": 2.110494311706494, + "learning_rate": 1.892931428739744e-05, + "loss": 0.9468, + "step": 7394 + }, + { + "epoch": 0.17, + "grad_norm": 2.5656913649097355, + "learning_rate": 1.8928970745890985e-05, + "loss": 0.9861, + "step": 7395 + }, + { + "epoch": 0.17, + "grad_norm": 2.056439435049051, + "learning_rate": 1.8928627152397237e-05, + "loss": 1.0951, + "step": 7396 + }, + { + "epoch": 0.17, + "grad_norm": 1.8981470536597302, + "learning_rate": 1.8928283506918194e-05, + "loss": 1.3121, + "step": 7397 + }, + { + "epoch": 0.17, + "grad_norm": 2.7901363242356925, + "learning_rate": 1.892793980945586e-05, + "loss": 1.1559, + "step": 7398 + }, + { + "epoch": 0.17, + "grad_norm": 2.1087190431005323, + "learning_rate": 1.892759606001223e-05, + "loss": 0.9672, + "step": 7399 + }, + { + "epoch": 0.17, + "grad_norm": 1.2236603989121146, + "learning_rate": 1.8927252258589312e-05, + "loss": 1.0304, + "step": 7400 + }, + { + "epoch": 0.17, + "grad_norm": 2.076233043828208, + "learning_rate": 1.8926908405189105e-05, + "loss": 1.0472, + "step": 7401 + }, + { + "epoch": 0.17, + "grad_norm": 1.1204508567769957, + "learning_rate": 1.8926564499813608e-05, + "loss": 0.9703, + "step": 7402 + }, + { + "epoch": 0.17, + "grad_norm": 2.052855973607447, + "learning_rate": 1.8926220542464833e-05, + "loss": 1.0326, + "step": 7403 + }, + { + "epoch": 0.17, + "grad_norm": 1.149911996472901, + "learning_rate": 1.8925876533144772e-05, + "loss": 0.9481, + "step": 7404 + }, + { + "epoch": 0.17, + "grad_norm": 2.0269452699816712, + "learning_rate": 1.892553247185543e-05, + "loss": 1.0799, + "step": 7405 + }, + { + "epoch": 0.17, + "grad_norm": 2.2825704855330313, + "learning_rate": 1.8925188358598815e-05, + "loss": 1.0391, + "step": 7406 + }, + { + "epoch": 0.17, + "grad_norm": 1.952646213020963, + "learning_rate": 1.8924844193376926e-05, + "loss": 0.946, + "step": 7407 + }, + { + "epoch": 0.17, + "grad_norm": 2.0240847155906088, + "learning_rate": 1.892449997619177e-05, + "loss": 1.1612, + "step": 7408 + }, + { + "epoch": 0.17, + "grad_norm": 2.073375443368587, + "learning_rate": 1.8924155707045348e-05, + "loss": 1.1072, + "step": 7409 + }, + { + "epoch": 0.17, + "grad_norm": 2.2797884910438184, + "learning_rate": 1.8923811385939666e-05, + "loss": 1.1187, + "step": 7410 + }, + { + "epoch": 0.17, + "grad_norm": 2.189060996388294, + "learning_rate": 1.892346701287673e-05, + "loss": 1.1035, + "step": 7411 + }, + { + "epoch": 0.17, + "grad_norm": 2.1963090367763733, + "learning_rate": 1.8923122587858543e-05, + "loss": 1.1498, + "step": 7412 + }, + { + "epoch": 0.17, + "grad_norm": 2.233946836684038, + "learning_rate": 1.8922778110887114e-05, + "loss": 1.0927, + "step": 7413 + }, + { + "epoch": 0.17, + "grad_norm": 1.896056098636855, + "learning_rate": 1.8922433581964443e-05, + "loss": 1.1264, + "step": 7414 + }, + { + "epoch": 0.17, + "grad_norm": 2.0893655273264917, + "learning_rate": 1.892208900109254e-05, + "loss": 1.1663, + "step": 7415 + }, + { + "epoch": 0.17, + "grad_norm": 1.9405884497804844, + "learning_rate": 1.892174436827341e-05, + "loss": 1.1004, + "step": 7416 + }, + { + "epoch": 0.17, + "grad_norm": 1.2641218736951252, + "learning_rate": 1.892139968350906e-05, + "loss": 0.9485, + "step": 7417 + }, + { + "epoch": 0.17, + "grad_norm": 1.9119928194658287, + "learning_rate": 1.8921054946801497e-05, + "loss": 1.2112, + "step": 7418 + }, + { + "epoch": 0.17, + "grad_norm": 2.350431046099496, + "learning_rate": 1.8920710158152727e-05, + "loss": 0.8845, + "step": 7419 + }, + { + "epoch": 0.17, + "grad_norm": 1.9704949895938941, + "learning_rate": 1.8920365317564758e-05, + "loss": 0.962, + "step": 7420 + }, + { + "epoch": 0.17, + "grad_norm": 1.1241072110626282, + "learning_rate": 1.8920020425039598e-05, + "loss": 0.9331, + "step": 7421 + }, + { + "epoch": 0.17, + "grad_norm": 2.0726421019950707, + "learning_rate": 1.8919675480579256e-05, + "loss": 1.1094, + "step": 7422 + }, + { + "epoch": 0.17, + "grad_norm": 2.2643969941856894, + "learning_rate": 1.891933048418574e-05, + "loss": 1.208, + "step": 7423 + }, + { + "epoch": 0.17, + "grad_norm": 1.1262086098164525, + "learning_rate": 1.8918985435861056e-05, + "loss": 1.0144, + "step": 7424 + }, + { + "epoch": 0.17, + "grad_norm": 1.1172127009138701, + "learning_rate": 1.8918640335607215e-05, + "loss": 0.9165, + "step": 7425 + }, + { + "epoch": 0.17, + "grad_norm": 1.9436715748666447, + "learning_rate": 1.8918295183426226e-05, + "loss": 1.0558, + "step": 7426 + }, + { + "epoch": 0.17, + "grad_norm": 2.147230107936837, + "learning_rate": 1.89179499793201e-05, + "loss": 1.0795, + "step": 7427 + }, + { + "epoch": 0.17, + "grad_norm": 5.961363223979705, + "learning_rate": 1.891760472329085e-05, + "loss": 0.9406, + "step": 7428 + }, + { + "epoch": 0.18, + "grad_norm": 2.525424264291406, + "learning_rate": 1.891725941534048e-05, + "loss": 1.0117, + "step": 7429 + }, + { + "epoch": 0.18, + "grad_norm": 2.0949926001775068, + "learning_rate": 1.8916914055471e-05, + "loss": 0.9507, + "step": 7430 + }, + { + "epoch": 0.18, + "grad_norm": 2.321861052332848, + "learning_rate": 1.891656864368442e-05, + "loss": 1.0602, + "step": 7431 + }, + { + "epoch": 0.18, + "grad_norm": 2.2948147137965873, + "learning_rate": 1.891622317998276e-05, + "loss": 1.1061, + "step": 7432 + }, + { + "epoch": 0.18, + "grad_norm": 2.625542807105962, + "learning_rate": 1.8915877664368026e-05, + "loss": 1.0505, + "step": 7433 + }, + { + "epoch": 0.18, + "grad_norm": 2.2833581104776584, + "learning_rate": 1.8915532096842225e-05, + "loss": 0.9899, + "step": 7434 + }, + { + "epoch": 0.18, + "grad_norm": 2.150847293770832, + "learning_rate": 1.8915186477407376e-05, + "loss": 1.2068, + "step": 7435 + }, + { + "epoch": 0.18, + "grad_norm": 1.2683550337625218, + "learning_rate": 1.891484080606549e-05, + "loss": 0.9977, + "step": 7436 + }, + { + "epoch": 0.18, + "grad_norm": 1.9614390801778256, + "learning_rate": 1.8914495082818578e-05, + "loss": 1.0006, + "step": 7437 + }, + { + "epoch": 0.18, + "grad_norm": 2.399619440690716, + "learning_rate": 1.891414930766865e-05, + "loss": 1.0648, + "step": 7438 + }, + { + "epoch": 0.18, + "grad_norm": 2.1181264429238396, + "learning_rate": 1.891380348061773e-05, + "loss": 1.0992, + "step": 7439 + }, + { + "epoch": 0.18, + "grad_norm": 2.1306334266703324, + "learning_rate": 1.891345760166782e-05, + "loss": 1.1248, + "step": 7440 + }, + { + "epoch": 0.18, + "grad_norm": 1.9360839462386097, + "learning_rate": 1.8913111670820936e-05, + "loss": 1.1625, + "step": 7441 + }, + { + "epoch": 0.18, + "grad_norm": 1.8221641051127107, + "learning_rate": 1.8912765688079097e-05, + "loss": 0.9483, + "step": 7442 + }, + { + "epoch": 0.18, + "grad_norm": 2.18370339953417, + "learning_rate": 1.8912419653444313e-05, + "loss": 0.9431, + "step": 7443 + }, + { + "epoch": 0.18, + "grad_norm": 2.2367402074984106, + "learning_rate": 1.8912073566918597e-05, + "loss": 0.9603, + "step": 7444 + }, + { + "epoch": 0.18, + "grad_norm": 3.5440863408533003, + "learning_rate": 1.891172742850397e-05, + "loss": 0.9064, + "step": 7445 + }, + { + "epoch": 0.18, + "grad_norm": 2.1339122042705903, + "learning_rate": 1.8911381238202447e-05, + "loss": 1.0236, + "step": 7446 + }, + { + "epoch": 0.18, + "grad_norm": 1.1998521659187442, + "learning_rate": 1.891103499601604e-05, + "loss": 1.0077, + "step": 7447 + }, + { + "epoch": 0.18, + "grad_norm": 2.0635971020814443, + "learning_rate": 1.8910688701946763e-05, + "loss": 1.1025, + "step": 7448 + }, + { + "epoch": 0.18, + "grad_norm": 2.093195023885668, + "learning_rate": 1.8910342355996637e-05, + "loss": 0.9869, + "step": 7449 + }, + { + "epoch": 0.18, + "grad_norm": 2.587644544093778, + "learning_rate": 1.8909995958167678e-05, + "loss": 1.0608, + "step": 7450 + }, + { + "epoch": 0.18, + "grad_norm": 2.0649561323361074, + "learning_rate": 1.89096495084619e-05, + "loss": 1.0748, + "step": 7451 + }, + { + "epoch": 0.18, + "grad_norm": 2.0648079152683616, + "learning_rate": 1.8909303006881323e-05, + "loss": 1.112, + "step": 7452 + }, + { + "epoch": 0.18, + "grad_norm": 2.441576412469772, + "learning_rate": 1.8908956453427963e-05, + "loss": 1.149, + "step": 7453 + }, + { + "epoch": 0.18, + "grad_norm": 1.964377558762014, + "learning_rate": 1.890860984810384e-05, + "loss": 1.0449, + "step": 7454 + }, + { + "epoch": 0.18, + "grad_norm": 1.9860298118870163, + "learning_rate": 1.8908263190910966e-05, + "loss": 0.9706, + "step": 7455 + }, + { + "epoch": 0.18, + "grad_norm": 1.9953358994094326, + "learning_rate": 1.8907916481851364e-05, + "loss": 1.0713, + "step": 7456 + }, + { + "epoch": 0.18, + "grad_norm": 2.087765375547649, + "learning_rate": 1.8907569720927055e-05, + "loss": 1.1547, + "step": 7457 + }, + { + "epoch": 0.18, + "grad_norm": 2.026552546537562, + "learning_rate": 1.8907222908140053e-05, + "loss": 1.0357, + "step": 7458 + }, + { + "epoch": 0.18, + "grad_norm": 1.9678507408689971, + "learning_rate": 1.8906876043492382e-05, + "loss": 1.1034, + "step": 7459 + }, + { + "epoch": 0.18, + "grad_norm": 1.961881969423442, + "learning_rate": 1.8906529126986057e-05, + "loss": 1.053, + "step": 7460 + }, + { + "epoch": 0.18, + "grad_norm": 2.1729141332425312, + "learning_rate": 1.8906182158623097e-05, + "loss": 1.0839, + "step": 7461 + }, + { + "epoch": 0.18, + "grad_norm": 2.0537954571310935, + "learning_rate": 1.8905835138405526e-05, + "loss": 1.0294, + "step": 7462 + }, + { + "epoch": 0.18, + "grad_norm": 2.296761845904137, + "learning_rate": 1.8905488066335366e-05, + "loss": 0.9999, + "step": 7463 + }, + { + "epoch": 0.18, + "grad_norm": 2.2227614052951346, + "learning_rate": 1.8905140942414633e-05, + "loss": 1.1151, + "step": 7464 + }, + { + "epoch": 0.18, + "grad_norm": 1.8967206157873835, + "learning_rate": 1.890479376664535e-05, + "loss": 1.1209, + "step": 7465 + }, + { + "epoch": 0.18, + "grad_norm": 1.1701630972943844, + "learning_rate": 1.890444653902954e-05, + "loss": 0.9412, + "step": 7466 + }, + { + "epoch": 0.18, + "grad_norm": 1.9106144387253863, + "learning_rate": 1.890409925956922e-05, + "loss": 0.993, + "step": 7467 + }, + { + "epoch": 0.18, + "grad_norm": 1.1528831651957077, + "learning_rate": 1.8903751928266415e-05, + "loss": 0.9824, + "step": 7468 + }, + { + "epoch": 0.18, + "grad_norm": 1.9563541142706466, + "learning_rate": 1.890340454512315e-05, + "loss": 1.109, + "step": 7469 + }, + { + "epoch": 0.18, + "grad_norm": 2.3138776405744284, + "learning_rate": 1.8903057110141443e-05, + "loss": 1.1273, + "step": 7470 + }, + { + "epoch": 0.18, + "grad_norm": 1.9223614946540668, + "learning_rate": 1.8902709623323322e-05, + "loss": 1.0413, + "step": 7471 + }, + { + "epoch": 0.18, + "grad_norm": 2.073953154454614, + "learning_rate": 1.8902362084670806e-05, + "loss": 1.1061, + "step": 7472 + }, + { + "epoch": 0.18, + "grad_norm": 2.765057670009657, + "learning_rate": 1.8902014494185918e-05, + "loss": 1.1202, + "step": 7473 + }, + { + "epoch": 0.18, + "grad_norm": 1.1861099889562539, + "learning_rate": 1.8901666851870684e-05, + "loss": 1.0074, + "step": 7474 + }, + { + "epoch": 0.18, + "grad_norm": 2.246844967563797, + "learning_rate": 1.8901319157727127e-05, + "loss": 1.0843, + "step": 7475 + }, + { + "epoch": 0.18, + "grad_norm": 2.2250464514909045, + "learning_rate": 1.8900971411757273e-05, + "loss": 1.1475, + "step": 7476 + }, + { + "epoch": 0.18, + "grad_norm": 2.29468682921785, + "learning_rate": 1.8900623613963144e-05, + "loss": 1.092, + "step": 7477 + }, + { + "epoch": 0.18, + "grad_norm": 2.250951681806374, + "learning_rate": 1.890027576434677e-05, + "loss": 0.9746, + "step": 7478 + }, + { + "epoch": 0.18, + "grad_norm": 1.0534488667612105, + "learning_rate": 1.889992786291017e-05, + "loss": 1.0559, + "step": 7479 + }, + { + "epoch": 0.18, + "grad_norm": 2.1293586140984115, + "learning_rate": 1.889957990965537e-05, + "loss": 1.1589, + "step": 7480 + }, + { + "epoch": 0.18, + "grad_norm": 2.1642683234483244, + "learning_rate": 1.88992319045844e-05, + "loss": 0.923, + "step": 7481 + }, + { + "epoch": 0.18, + "grad_norm": 1.9983830504758138, + "learning_rate": 1.8898883847699284e-05, + "loss": 1.2194, + "step": 7482 + }, + { + "epoch": 0.18, + "grad_norm": 2.1963905993288204, + "learning_rate": 1.889853573900205e-05, + "loss": 1.0385, + "step": 7483 + }, + { + "epoch": 0.18, + "grad_norm": 2.203034238750829, + "learning_rate": 1.8898187578494723e-05, + "loss": 1.0654, + "step": 7484 + }, + { + "epoch": 0.18, + "grad_norm": 2.280330854556836, + "learning_rate": 1.889783936617933e-05, + "loss": 0.9081, + "step": 7485 + }, + { + "epoch": 0.18, + "grad_norm": 1.9659748045111116, + "learning_rate": 1.88974911020579e-05, + "loss": 1.1392, + "step": 7486 + }, + { + "epoch": 0.18, + "grad_norm": 2.8028709072658513, + "learning_rate": 1.889714278613246e-05, + "loss": 0.9859, + "step": 7487 + }, + { + "epoch": 0.18, + "grad_norm": 1.999307362566701, + "learning_rate": 1.889679441840504e-05, + "loss": 1.1236, + "step": 7488 + }, + { + "epoch": 0.18, + "grad_norm": 1.1274714042568417, + "learning_rate": 1.8896445998877664e-05, + "loss": 1.0435, + "step": 7489 + }, + { + "epoch": 0.18, + "grad_norm": 2.6086858752936792, + "learning_rate": 1.8896097527552362e-05, + "loss": 1.2159, + "step": 7490 + }, + { + "epoch": 0.18, + "grad_norm": 2.042736550272621, + "learning_rate": 1.8895749004431168e-05, + "loss": 1.1512, + "step": 7491 + }, + { + "epoch": 0.18, + "grad_norm": 1.9814859166833254, + "learning_rate": 1.88954004295161e-05, + "loss": 1.134, + "step": 7492 + }, + { + "epoch": 0.18, + "grad_norm": 2.120034869004694, + "learning_rate": 1.8895051802809202e-05, + "loss": 1.0517, + "step": 7493 + }, + { + "epoch": 0.18, + "grad_norm": 2.093348022160725, + "learning_rate": 1.8894703124312492e-05, + "loss": 1.1005, + "step": 7494 + }, + { + "epoch": 0.18, + "grad_norm": 1.9826321234341049, + "learning_rate": 1.8894354394028005e-05, + "loss": 0.9257, + "step": 7495 + }, + { + "epoch": 0.18, + "grad_norm": 3.685015151461668, + "learning_rate": 1.889400561195777e-05, + "loss": 1.1511, + "step": 7496 + }, + { + "epoch": 0.18, + "grad_norm": 2.757382957842075, + "learning_rate": 1.889365677810382e-05, + "loss": 1.1007, + "step": 7497 + }, + { + "epoch": 0.18, + "grad_norm": 1.727187264528098, + "learning_rate": 1.8893307892468184e-05, + "loss": 1.1203, + "step": 7498 + }, + { + "epoch": 0.18, + "grad_norm": 1.998131557480463, + "learning_rate": 1.8892958955052893e-05, + "loss": 1.1299, + "step": 7499 + }, + { + "epoch": 0.18, + "grad_norm": 2.0532873043718975, + "learning_rate": 1.889260996585998e-05, + "loss": 1.0694, + "step": 7500 + }, + { + "epoch": 0.18, + "grad_norm": 2.5597369796715195, + "learning_rate": 1.889226092489148e-05, + "loss": 1.1111, + "step": 7501 + }, + { + "epoch": 0.18, + "grad_norm": 1.9814126710943354, + "learning_rate": 1.8891911832149418e-05, + "loss": 1.0555, + "step": 7502 + }, + { + "epoch": 0.18, + "grad_norm": 2.268946015484778, + "learning_rate": 1.889156268763583e-05, + "loss": 1.0246, + "step": 7503 + }, + { + "epoch": 0.18, + "grad_norm": 2.647719781112826, + "learning_rate": 1.889121349135275e-05, + "loss": 1.0482, + "step": 7504 + }, + { + "epoch": 0.18, + "grad_norm": 2.098299167603817, + "learning_rate": 1.889086424330221e-05, + "loss": 1.1372, + "step": 7505 + }, + { + "epoch": 0.18, + "grad_norm": 2.147617401012871, + "learning_rate": 1.8890514943486242e-05, + "loss": 1.0555, + "step": 7506 + }, + { + "epoch": 0.18, + "grad_norm": 2.146149461015817, + "learning_rate": 1.8890165591906883e-05, + "loss": 1.0397, + "step": 7507 + }, + { + "epoch": 0.18, + "grad_norm": 2.4022108528741883, + "learning_rate": 1.8889816188566163e-05, + "loss": 1.0377, + "step": 7508 + }, + { + "epoch": 0.18, + "grad_norm": 2.191142858032834, + "learning_rate": 1.8889466733466124e-05, + "loss": 1.1115, + "step": 7509 + }, + { + "epoch": 0.18, + "grad_norm": 2.2074884571958866, + "learning_rate": 1.888911722660879e-05, + "loss": 1.0244, + "step": 7510 + }, + { + "epoch": 0.18, + "grad_norm": 2.870011728782293, + "learning_rate": 1.88887676679962e-05, + "loss": 0.9992, + "step": 7511 + }, + { + "epoch": 0.18, + "grad_norm": 2.5327524949439946, + "learning_rate": 1.8888418057630397e-05, + "loss": 1.07, + "step": 7512 + }, + { + "epoch": 0.18, + "grad_norm": 2.2487028674476313, + "learning_rate": 1.8888068395513407e-05, + "loss": 1.0877, + "step": 7513 + }, + { + "epoch": 0.18, + "grad_norm": 2.1309623909707116, + "learning_rate": 1.8887718681647265e-05, + "loss": 0.9955, + "step": 7514 + }, + { + "epoch": 0.18, + "grad_norm": 2.1985579629905563, + "learning_rate": 1.8887368916034017e-05, + "loss": 1.0647, + "step": 7515 + }, + { + "epoch": 0.18, + "grad_norm": 1.1261300450107443, + "learning_rate": 1.8887019098675692e-05, + "loss": 0.9612, + "step": 7516 + }, + { + "epoch": 0.18, + "grad_norm": 2.1856810249052336, + "learning_rate": 1.8886669229574326e-05, + "loss": 1.0945, + "step": 7517 + }, + { + "epoch": 0.18, + "grad_norm": 2.387096487810926, + "learning_rate": 1.888631930873196e-05, + "loss": 1.0638, + "step": 7518 + }, + { + "epoch": 0.18, + "grad_norm": 2.6021810152917433, + "learning_rate": 1.888596933615063e-05, + "loss": 1.0433, + "step": 7519 + }, + { + "epoch": 0.18, + "grad_norm": 2.623463075284107, + "learning_rate": 1.888561931183237e-05, + "loss": 1.1103, + "step": 7520 + }, + { + "epoch": 0.18, + "grad_norm": 2.7844335255669312, + "learning_rate": 1.8885269235779226e-05, + "loss": 1.1093, + "step": 7521 + }, + { + "epoch": 0.18, + "grad_norm": 2.4044688000245387, + "learning_rate": 1.888491910799323e-05, + "loss": 1.1372, + "step": 7522 + }, + { + "epoch": 0.18, + "grad_norm": 2.0729413215146044, + "learning_rate": 1.888456892847642e-05, + "loss": 1.1355, + "step": 7523 + }, + { + "epoch": 0.18, + "grad_norm": 2.4540096964796843, + "learning_rate": 1.8884218697230838e-05, + "loss": 0.8735, + "step": 7524 + }, + { + "epoch": 0.18, + "grad_norm": 2.1720045512268937, + "learning_rate": 1.8883868414258525e-05, + "loss": 1.1155, + "step": 7525 + }, + { + "epoch": 0.18, + "grad_norm": 2.2501475319740827, + "learning_rate": 1.8883518079561512e-05, + "loss": 1.1883, + "step": 7526 + }, + { + "epoch": 0.18, + "grad_norm": 2.03855942840693, + "learning_rate": 1.8883167693141845e-05, + "loss": 1.1814, + "step": 7527 + }, + { + "epoch": 0.18, + "grad_norm": 1.1727555157817717, + "learning_rate": 1.8882817255001567e-05, + "loss": 0.9575, + "step": 7528 + }, + { + "epoch": 0.18, + "grad_norm": 2.274052611891799, + "learning_rate": 1.8882466765142715e-05, + "loss": 1.0228, + "step": 7529 + }, + { + "epoch": 0.18, + "grad_norm": 2.0810411734524834, + "learning_rate": 1.8882116223567327e-05, + "loss": 1.0116, + "step": 7530 + }, + { + "epoch": 0.18, + "grad_norm": 2.2891255295183726, + "learning_rate": 1.8881765630277447e-05, + "loss": 1.0315, + "step": 7531 + }, + { + "epoch": 0.18, + "grad_norm": 2.2922495589594094, + "learning_rate": 1.8881414985275114e-05, + "loss": 1.0311, + "step": 7532 + }, + { + "epoch": 0.18, + "grad_norm": 2.0246511523856414, + "learning_rate": 1.8881064288562373e-05, + "loss": 1.1676, + "step": 7533 + }, + { + "epoch": 0.18, + "grad_norm": 1.9967315189423966, + "learning_rate": 1.888071354014126e-05, + "loss": 0.9422, + "step": 7534 + }, + { + "epoch": 0.18, + "grad_norm": 2.058267473030595, + "learning_rate": 1.8880362740013827e-05, + "loss": 1.0602, + "step": 7535 + }, + { + "epoch": 0.18, + "grad_norm": 1.0590670133012203, + "learning_rate": 1.8880011888182108e-05, + "loss": 1.0303, + "step": 7536 + }, + { + "epoch": 0.18, + "grad_norm": 2.2984170690516796, + "learning_rate": 1.887966098464815e-05, + "loss": 1.0236, + "step": 7537 + }, + { + "epoch": 0.18, + "grad_norm": 2.2167939845667073, + "learning_rate": 1.8879310029413993e-05, + "loss": 0.9764, + "step": 7538 + }, + { + "epoch": 0.18, + "grad_norm": 1.886271447909949, + "learning_rate": 1.8878959022481682e-05, + "loss": 1.0994, + "step": 7539 + }, + { + "epoch": 0.18, + "grad_norm": 2.035819933543201, + "learning_rate": 1.887860796385326e-05, + "loss": 0.946, + "step": 7540 + }, + { + "epoch": 0.18, + "grad_norm": 2.8857931734476776, + "learning_rate": 1.8878256853530772e-05, + "loss": 1.0917, + "step": 7541 + }, + { + "epoch": 0.18, + "grad_norm": 2.2332080663811857, + "learning_rate": 1.8877905691516263e-05, + "loss": 1.1209, + "step": 7542 + }, + { + "epoch": 0.18, + "grad_norm": 2.024783710962769, + "learning_rate": 1.8877554477811775e-05, + "loss": 1.1605, + "step": 7543 + }, + { + "epoch": 0.18, + "grad_norm": 2.0602304241610425, + "learning_rate": 1.8877203212419354e-05, + "loss": 0.8628, + "step": 7544 + }, + { + "epoch": 0.18, + "grad_norm": 2.4239125638588903, + "learning_rate": 1.8876851895341044e-05, + "loss": 1.0539, + "step": 7545 + }, + { + "epoch": 0.18, + "grad_norm": 2.2586637241381324, + "learning_rate": 1.8876500526578897e-05, + "loss": 0.9924, + "step": 7546 + }, + { + "epoch": 0.18, + "grad_norm": 2.118371891969198, + "learning_rate": 1.8876149106134947e-05, + "loss": 1.0246, + "step": 7547 + }, + { + "epoch": 0.18, + "grad_norm": 1.9493729015832395, + "learning_rate": 1.8875797634011254e-05, + "loss": 0.9931, + "step": 7548 + }, + { + "epoch": 0.18, + "grad_norm": 2.040669643700937, + "learning_rate": 1.887544611020985e-05, + "loss": 1.0832, + "step": 7549 + }, + { + "epoch": 0.18, + "grad_norm": 3.1886882238542573, + "learning_rate": 1.8875094534732793e-05, + "loss": 1.089, + "step": 7550 + }, + { + "epoch": 0.18, + "grad_norm": 1.8518074822722541, + "learning_rate": 1.8874742907582124e-05, + "loss": 1.0946, + "step": 7551 + }, + { + "epoch": 0.18, + "grad_norm": 2.343851074185106, + "learning_rate": 1.8874391228759894e-05, + "loss": 1.0283, + "step": 7552 + }, + { + "epoch": 0.18, + "grad_norm": 1.103748288988281, + "learning_rate": 1.8874039498268147e-05, + "loss": 1.0039, + "step": 7553 + }, + { + "epoch": 0.18, + "grad_norm": 1.989763245303881, + "learning_rate": 1.8873687716108935e-05, + "loss": 0.9846, + "step": 7554 + }, + { + "epoch": 0.18, + "grad_norm": 2.2655021692761146, + "learning_rate": 1.88733358822843e-05, + "loss": 1.1027, + "step": 7555 + }, + { + "epoch": 0.18, + "grad_norm": 2.046841829518913, + "learning_rate": 1.8872983996796298e-05, + "loss": 1.2245, + "step": 7556 + }, + { + "epoch": 0.18, + "grad_norm": 1.0640490649801124, + "learning_rate": 1.8872632059646973e-05, + "loss": 1.0495, + "step": 7557 + }, + { + "epoch": 0.18, + "grad_norm": 2.2697588280000365, + "learning_rate": 1.8872280070838373e-05, + "loss": 1.0097, + "step": 7558 + }, + { + "epoch": 0.18, + "grad_norm": 2.04203844975809, + "learning_rate": 1.8871928030372553e-05, + "loss": 0.9704, + "step": 7559 + }, + { + "epoch": 0.18, + "grad_norm": 1.9948684133033034, + "learning_rate": 1.8871575938251557e-05, + "loss": 0.8812, + "step": 7560 + }, + { + "epoch": 0.18, + "grad_norm": 2.1191511784018267, + "learning_rate": 1.887122379447744e-05, + "loss": 1.0625, + "step": 7561 + }, + { + "epoch": 0.18, + "grad_norm": 2.305436247897389, + "learning_rate": 1.8870871599052244e-05, + "loss": 0.998, + "step": 7562 + }, + { + "epoch": 0.18, + "grad_norm": 1.0875672740756686, + "learning_rate": 1.8870519351978026e-05, + "loss": 0.9484, + "step": 7563 + }, + { + "epoch": 0.18, + "grad_norm": 2.0343633908498693, + "learning_rate": 1.8870167053256838e-05, + "loss": 1.0821, + "step": 7564 + }, + { + "epoch": 0.18, + "grad_norm": 1.12922584425928, + "learning_rate": 1.8869814702890733e-05, + "loss": 0.983, + "step": 7565 + }, + { + "epoch": 0.18, + "grad_norm": 2.306353441735019, + "learning_rate": 1.8869462300881755e-05, + "loss": 0.9854, + "step": 7566 + }, + { + "epoch": 0.18, + "grad_norm": 2.1321972774695235, + "learning_rate": 1.8869109847231956e-05, + "loss": 0.9948, + "step": 7567 + }, + { + "epoch": 0.18, + "grad_norm": 2.3183242709628322, + "learning_rate": 1.8868757341943396e-05, + "loss": 1.1375, + "step": 7568 + }, + { + "epoch": 0.18, + "grad_norm": 2.0125711706764795, + "learning_rate": 1.886840478501812e-05, + "loss": 1.0379, + "step": 7569 + }, + { + "epoch": 0.18, + "grad_norm": 1.8454488233439716, + "learning_rate": 1.8868052176458188e-05, + "loss": 1.1552, + "step": 7570 + }, + { + "epoch": 0.18, + "grad_norm": 2.046198959174613, + "learning_rate": 1.8867699516265646e-05, + "loss": 1.0318, + "step": 7571 + }, + { + "epoch": 0.18, + "grad_norm": 2.4418229565717278, + "learning_rate": 1.886734680444255e-05, + "loss": 1.0542, + "step": 7572 + }, + { + "epoch": 0.18, + "grad_norm": 1.8471181550444697, + "learning_rate": 1.8866994040990954e-05, + "loss": 1.0358, + "step": 7573 + }, + { + "epoch": 0.18, + "grad_norm": 2.0847558418724836, + "learning_rate": 1.8866641225912908e-05, + "loss": 1.1108, + "step": 7574 + }, + { + "epoch": 0.18, + "grad_norm": 2.314376977078301, + "learning_rate": 1.8866288359210474e-05, + "loss": 1.2493, + "step": 7575 + }, + { + "epoch": 0.18, + "grad_norm": 2.271108050632692, + "learning_rate": 1.88659354408857e-05, + "loss": 1.028, + "step": 7576 + }, + { + "epoch": 0.18, + "grad_norm": 2.65557827116928, + "learning_rate": 1.8865582470940644e-05, + "loss": 1.1057, + "step": 7577 + }, + { + "epoch": 0.18, + "grad_norm": 2.128544829032257, + "learning_rate": 1.886522944937736e-05, + "loss": 1.1882, + "step": 7578 + }, + { + "epoch": 0.18, + "grad_norm": 2.1347591548596516, + "learning_rate": 1.8864876376197904e-05, + "loss": 0.9066, + "step": 7579 + }, + { + "epoch": 0.18, + "grad_norm": 4.974307134183679, + "learning_rate": 1.886452325140433e-05, + "loss": 0.9311, + "step": 7580 + }, + { + "epoch": 0.18, + "grad_norm": 2.7469171941769215, + "learning_rate": 1.8864170074998695e-05, + "loss": 0.979, + "step": 7581 + }, + { + "epoch": 0.18, + "grad_norm": 2.070429470698543, + "learning_rate": 1.8863816846983055e-05, + "loss": 0.9445, + "step": 7582 + }, + { + "epoch": 0.18, + "grad_norm": 2.0534637361802335, + "learning_rate": 1.886346356735947e-05, + "loss": 1.0848, + "step": 7583 + }, + { + "epoch": 0.18, + "grad_norm": 2.346615874390795, + "learning_rate": 1.8863110236129993e-05, + "loss": 1.0545, + "step": 7584 + }, + { + "epoch": 0.18, + "grad_norm": 2.072558011389818, + "learning_rate": 1.886275685329668e-05, + "loss": 0.9571, + "step": 7585 + }, + { + "epoch": 0.18, + "grad_norm": 2.206822630213083, + "learning_rate": 1.8862403418861593e-05, + "loss": 1.0106, + "step": 7586 + }, + { + "epoch": 0.18, + "grad_norm": 2.114597642999692, + "learning_rate": 1.8862049932826787e-05, + "loss": 1.0776, + "step": 7587 + }, + { + "epoch": 0.18, + "grad_norm": 2.0346601600259113, + "learning_rate": 1.886169639519432e-05, + "loss": 1.0322, + "step": 7588 + }, + { + "epoch": 0.18, + "grad_norm": 3.759981240617039, + "learning_rate": 1.886134280596625e-05, + "loss": 0.9404, + "step": 7589 + }, + { + "epoch": 0.18, + "grad_norm": 1.9263130042871999, + "learning_rate": 1.886098916514464e-05, + "loss": 0.9835, + "step": 7590 + }, + { + "epoch": 0.18, + "grad_norm": 2.0971674003200627, + "learning_rate": 1.886063547273154e-05, + "loss": 0.9984, + "step": 7591 + }, + { + "epoch": 0.18, + "grad_norm": 1.871324774975486, + "learning_rate": 1.886028172872902e-05, + "loss": 0.9278, + "step": 7592 + }, + { + "epoch": 0.18, + "grad_norm": 2.391362290525908, + "learning_rate": 1.885992793313913e-05, + "loss": 1.0258, + "step": 7593 + }, + { + "epoch": 0.18, + "grad_norm": 2.1307091839764203, + "learning_rate": 1.885957408596394e-05, + "loss": 1.1013, + "step": 7594 + }, + { + "epoch": 0.18, + "grad_norm": 2.2913348879665243, + "learning_rate": 1.8859220187205505e-05, + "loss": 1.1388, + "step": 7595 + }, + { + "epoch": 0.18, + "grad_norm": 2.161110928353985, + "learning_rate": 1.8858866236865883e-05, + "loss": 1.1322, + "step": 7596 + }, + { + "epoch": 0.18, + "grad_norm": 2.0001872962068945, + "learning_rate": 1.8858512234947138e-05, + "loss": 1.082, + "step": 7597 + }, + { + "epoch": 0.18, + "grad_norm": 2.1263577540484384, + "learning_rate": 1.8858158181451327e-05, + "loss": 1.0649, + "step": 7598 + }, + { + "epoch": 0.18, + "grad_norm": 1.9395640449316254, + "learning_rate": 1.885780407638052e-05, + "loss": 1.2245, + "step": 7599 + }, + { + "epoch": 0.18, + "grad_norm": 2.566391731737171, + "learning_rate": 1.885744991973677e-05, + "loss": 1.059, + "step": 7600 + }, + { + "epoch": 0.18, + "grad_norm": 2.096659309330819, + "learning_rate": 1.885709571152214e-05, + "loss": 1.0174, + "step": 7601 + }, + { + "epoch": 0.18, + "grad_norm": 2.2092797269417535, + "learning_rate": 1.8856741451738697e-05, + "loss": 0.997, + "step": 7602 + }, + { + "epoch": 0.18, + "grad_norm": 2.219095846941275, + "learning_rate": 1.88563871403885e-05, + "loss": 1.0919, + "step": 7603 + }, + { + "epoch": 0.18, + "grad_norm": 2.302184569304338, + "learning_rate": 1.8856032777473617e-05, + "loss": 1.0677, + "step": 7604 + }, + { + "epoch": 0.18, + "grad_norm": 2.3118635504252767, + "learning_rate": 1.8855678362996105e-05, + "loss": 0.92, + "step": 7605 + }, + { + "epoch": 0.18, + "grad_norm": 2.26682499520206, + "learning_rate": 1.885532389695803e-05, + "loss": 1.1031, + "step": 7606 + }, + { + "epoch": 0.18, + "grad_norm": 2.538023011461194, + "learning_rate": 1.8854969379361456e-05, + "loss": 1.0314, + "step": 7607 + }, + { + "epoch": 0.18, + "grad_norm": 2.4240067198719917, + "learning_rate": 1.8854614810208446e-05, + "loss": 0.991, + "step": 7608 + }, + { + "epoch": 0.18, + "grad_norm": 2.1651678467510047, + "learning_rate": 1.8854260189501068e-05, + "loss": 1.0322, + "step": 7609 + }, + { + "epoch": 0.18, + "grad_norm": 2.295991747423155, + "learning_rate": 1.8853905517241384e-05, + "loss": 1.0945, + "step": 7610 + }, + { + "epoch": 0.18, + "grad_norm": 2.17214295051682, + "learning_rate": 1.885355079343145e-05, + "loss": 1.1498, + "step": 7611 + }, + { + "epoch": 0.18, + "grad_norm": 1.9969942445654685, + "learning_rate": 1.8853196018073352e-05, + "loss": 1.0632, + "step": 7612 + }, + { + "epoch": 0.18, + "grad_norm": 2.1310015179945814, + "learning_rate": 1.8852841191169138e-05, + "loss": 1.0354, + "step": 7613 + }, + { + "epoch": 0.18, + "grad_norm": 2.556408495579251, + "learning_rate": 1.885248631272088e-05, + "loss": 1.0404, + "step": 7614 + }, + { + "epoch": 0.18, + "grad_norm": 3.2913957381261705, + "learning_rate": 1.8852131382730645e-05, + "loss": 1.0666, + "step": 7615 + }, + { + "epoch": 0.18, + "grad_norm": 2.0900020276499283, + "learning_rate": 1.88517764012005e-05, + "loss": 1.1105, + "step": 7616 + }, + { + "epoch": 0.18, + "grad_norm": 1.1206422897389716, + "learning_rate": 1.8851421368132507e-05, + "loss": 0.9566, + "step": 7617 + }, + { + "epoch": 0.18, + "grad_norm": 2.1150300834509905, + "learning_rate": 1.8851066283528738e-05, + "loss": 0.939, + "step": 7618 + }, + { + "epoch": 0.18, + "grad_norm": 2.18688723521162, + "learning_rate": 1.8850711147391257e-05, + "loss": 1.0605, + "step": 7619 + }, + { + "epoch": 0.18, + "grad_norm": 1.1557204067298974, + "learning_rate": 1.8850355959722138e-05, + "loss": 1.0255, + "step": 7620 + }, + { + "epoch": 0.18, + "grad_norm": 2.3097792691191157, + "learning_rate": 1.885000072052344e-05, + "loss": 0.9233, + "step": 7621 + }, + { + "epoch": 0.18, + "grad_norm": 2.2030950193994605, + "learning_rate": 1.884964542979724e-05, + "loss": 0.9808, + "step": 7622 + }, + { + "epoch": 0.18, + "grad_norm": 1.1956846132858119, + "learning_rate": 1.88492900875456e-05, + "loss": 0.9821, + "step": 7623 + }, + { + "epoch": 0.18, + "grad_norm": 2.0124165216026926, + "learning_rate": 1.8848934693770594e-05, + "loss": 1.0476, + "step": 7624 + }, + { + "epoch": 0.18, + "grad_norm": 2.2832488237144455, + "learning_rate": 1.8848579248474286e-05, + "loss": 1.0894, + "step": 7625 + }, + { + "epoch": 0.18, + "grad_norm": 2.414345355113564, + "learning_rate": 1.884822375165875e-05, + "loss": 1.1378, + "step": 7626 + }, + { + "epoch": 0.18, + "grad_norm": 2.3356123559621405, + "learning_rate": 1.8847868203326053e-05, + "loss": 1.001, + "step": 7627 + }, + { + "epoch": 0.18, + "grad_norm": 2.73077268334632, + "learning_rate": 1.8847512603478267e-05, + "loss": 1.2118, + "step": 7628 + }, + { + "epoch": 0.18, + "grad_norm": 2.0548232415818153, + "learning_rate": 1.8847156952117462e-05, + "loss": 1.0519, + "step": 7629 + }, + { + "epoch": 0.18, + "grad_norm": 2.0121772000684874, + "learning_rate": 1.8846801249245706e-05, + "loss": 1.0633, + "step": 7630 + }, + { + "epoch": 0.18, + "grad_norm": 1.8556087675720456, + "learning_rate": 1.8846445494865074e-05, + "loss": 1.0791, + "step": 7631 + }, + { + "epoch": 0.18, + "grad_norm": 1.8898743578189356, + "learning_rate": 1.8846089688977635e-05, + "loss": 1.1257, + "step": 7632 + }, + { + "epoch": 0.18, + "grad_norm": 2.4155309338578625, + "learning_rate": 1.884573383158546e-05, + "loss": 1.0755, + "step": 7633 + }, + { + "epoch": 0.18, + "grad_norm": 1.8426030646402978, + "learning_rate": 1.8845377922690623e-05, + "loss": 1.0219, + "step": 7634 + }, + { + "epoch": 0.18, + "grad_norm": 2.1388760462721943, + "learning_rate": 1.88450219622952e-05, + "loss": 1.0378, + "step": 7635 + }, + { + "epoch": 0.18, + "grad_norm": 1.1416785185274063, + "learning_rate": 1.884466595040125e-05, + "loss": 0.9716, + "step": 7636 + }, + { + "epoch": 0.18, + "grad_norm": 2.3486155337810377, + "learning_rate": 1.8844309887010862e-05, + "loss": 1.0159, + "step": 7637 + }, + { + "epoch": 0.18, + "grad_norm": 2.1337568407222816, + "learning_rate": 1.88439537721261e-05, + "loss": 1.07, + "step": 7638 + }, + { + "epoch": 0.18, + "grad_norm": 2.120589913568752, + "learning_rate": 1.8843597605749035e-05, + "loss": 1.0937, + "step": 7639 + }, + { + "epoch": 0.18, + "grad_norm": 2.2715686660889354, + "learning_rate": 1.884324138788175e-05, + "loss": 1.0929, + "step": 7640 + }, + { + "epoch": 0.18, + "grad_norm": 3.7512395133959857, + "learning_rate": 1.8842885118526314e-05, + "loss": 1.0719, + "step": 7641 + }, + { + "epoch": 0.18, + "grad_norm": 1.950303479066469, + "learning_rate": 1.8842528797684795e-05, + "loss": 0.9931, + "step": 7642 + }, + { + "epoch": 0.18, + "grad_norm": 1.8546057914149452, + "learning_rate": 1.884217242535928e-05, + "loss": 1.0499, + "step": 7643 + }, + { + "epoch": 0.18, + "grad_norm": 2.2083194443999563, + "learning_rate": 1.8841816001551836e-05, + "loss": 0.9621, + "step": 7644 + }, + { + "epoch": 0.18, + "grad_norm": 2.094613230802803, + "learning_rate": 1.8841459526264542e-05, + "loss": 1.0523, + "step": 7645 + }, + { + "epoch": 0.18, + "grad_norm": 1.9403009398358393, + "learning_rate": 1.8841102999499468e-05, + "loss": 0.99, + "step": 7646 + }, + { + "epoch": 0.18, + "grad_norm": 2.0021153243579684, + "learning_rate": 1.8840746421258694e-05, + "loss": 0.9689, + "step": 7647 + }, + { + "epoch": 0.18, + "grad_norm": 1.9107620873078353, + "learning_rate": 1.8840389791544297e-05, + "loss": 1.1568, + "step": 7648 + }, + { + "epoch": 0.18, + "grad_norm": 1.980015261093927, + "learning_rate": 1.884003311035835e-05, + "loss": 1.0722, + "step": 7649 + }, + { + "epoch": 0.18, + "grad_norm": 2.0494311952470117, + "learning_rate": 1.8839676377702934e-05, + "loss": 1.0286, + "step": 7650 + }, + { + "epoch": 0.18, + "grad_norm": 2.1241200699145413, + "learning_rate": 1.883931959358012e-05, + "loss": 1.1282, + "step": 7651 + }, + { + "epoch": 0.18, + "grad_norm": 2.011374928087682, + "learning_rate": 1.8838962757991993e-05, + "loss": 1.0898, + "step": 7652 + }, + { + "epoch": 0.18, + "grad_norm": 2.0676325206656863, + "learning_rate": 1.883860587094062e-05, + "loss": 1.0555, + "step": 7653 + }, + { + "epoch": 0.18, + "grad_norm": 2.1398405143815, + "learning_rate": 1.883824893242809e-05, + "loss": 0.9305, + "step": 7654 + }, + { + "epoch": 0.18, + "grad_norm": 2.0344610301243913, + "learning_rate": 1.8837891942456477e-05, + "loss": 1.0054, + "step": 7655 + }, + { + "epoch": 0.18, + "grad_norm": 1.2574420818158116, + "learning_rate": 1.883753490102786e-05, + "loss": 0.9786, + "step": 7656 + }, + { + "epoch": 0.18, + "grad_norm": 2.3012181069294018, + "learning_rate": 1.8837177808144313e-05, + "loss": 1.0853, + "step": 7657 + }, + { + "epoch": 0.18, + "grad_norm": 1.1782975346943336, + "learning_rate": 1.8836820663807924e-05, + "loss": 1.0324, + "step": 7658 + }, + { + "epoch": 0.18, + "grad_norm": 2.0159721805553197, + "learning_rate": 1.8836463468020764e-05, + "loss": 0.9617, + "step": 7659 + }, + { + "epoch": 0.18, + "grad_norm": 2.284443493891234, + "learning_rate": 1.8836106220784915e-05, + "loss": 0.9899, + "step": 7660 + }, + { + "epoch": 0.18, + "grad_norm": 1.9240244801520123, + "learning_rate": 1.883574892210246e-05, + "loss": 0.9248, + "step": 7661 + }, + { + "epoch": 0.18, + "grad_norm": 2.202309781442142, + "learning_rate": 1.8835391571975475e-05, + "loss": 1.0682, + "step": 7662 + }, + { + "epoch": 0.18, + "grad_norm": 2.9027971893638638, + "learning_rate": 1.8835034170406046e-05, + "loss": 1.0871, + "step": 7663 + }, + { + "epoch": 0.18, + "grad_norm": 1.9528644728351625, + "learning_rate": 1.883467671739625e-05, + "loss": 1.0098, + "step": 7664 + }, + { + "epoch": 0.18, + "grad_norm": 2.408868185423534, + "learning_rate": 1.8834319212948168e-05, + "loss": 1.2731, + "step": 7665 + }, + { + "epoch": 0.18, + "grad_norm": 2.286153903993842, + "learning_rate": 1.8833961657063887e-05, + "loss": 0.936, + "step": 7666 + }, + { + "epoch": 0.18, + "grad_norm": 1.186725824661777, + "learning_rate": 1.883360404974548e-05, + "loss": 0.9792, + "step": 7667 + }, + { + "epoch": 0.18, + "grad_norm": 1.9854925298392514, + "learning_rate": 1.8833246390995034e-05, + "loss": 1.0648, + "step": 7668 + }, + { + "epoch": 0.18, + "grad_norm": 1.9330642792884845, + "learning_rate": 1.883288868081463e-05, + "loss": 1.1034, + "step": 7669 + }, + { + "epoch": 0.18, + "grad_norm": 1.0473048666245361, + "learning_rate": 1.8832530919206356e-05, + "loss": 0.9595, + "step": 7670 + }, + { + "epoch": 0.18, + "grad_norm": 2.058920371679783, + "learning_rate": 1.883217310617229e-05, + "loss": 1.1929, + "step": 7671 + }, + { + "epoch": 0.18, + "grad_norm": 2.029573231848303, + "learning_rate": 1.883181524171451e-05, + "loss": 1.0833, + "step": 7672 + }, + { + "epoch": 0.18, + "grad_norm": 1.9095114020987298, + "learning_rate": 1.8831457325835113e-05, + "loss": 1.0429, + "step": 7673 + }, + { + "epoch": 0.18, + "grad_norm": 2.0381315922615886, + "learning_rate": 1.8831099358536172e-05, + "loss": 1.0169, + "step": 7674 + }, + { + "epoch": 0.18, + "grad_norm": 2.2801515771296064, + "learning_rate": 1.8830741339819774e-05, + "loss": 1.0112, + "step": 7675 + }, + { + "epoch": 0.18, + "grad_norm": 1.966658773233185, + "learning_rate": 1.8830383269688005e-05, + "loss": 0.9519, + "step": 7676 + }, + { + "epoch": 0.18, + "grad_norm": 1.0935520296299557, + "learning_rate": 1.8830025148142948e-05, + "loss": 0.9385, + "step": 7677 + }, + { + "epoch": 0.18, + "grad_norm": 2.062071978320622, + "learning_rate": 1.882966697518669e-05, + "loss": 1.1369, + "step": 7678 + }, + { + "epoch": 0.18, + "grad_norm": 2.1349988378743494, + "learning_rate": 1.882930875082132e-05, + "loss": 1.1118, + "step": 7679 + }, + { + "epoch": 0.18, + "grad_norm": 1.177827427815306, + "learning_rate": 1.8828950475048913e-05, + "loss": 1.0153, + "step": 7680 + }, + { + "epoch": 0.18, + "grad_norm": 2.305294347027657, + "learning_rate": 1.8828592147871562e-05, + "loss": 1.1056, + "step": 7681 + }, + { + "epoch": 0.18, + "grad_norm": 2.0093641364354924, + "learning_rate": 1.8828233769291353e-05, + "loss": 1.0377, + "step": 7682 + }, + { + "epoch": 0.18, + "grad_norm": 2.5685604247028473, + "learning_rate": 1.8827875339310372e-05, + "loss": 1.0284, + "step": 7683 + }, + { + "epoch": 0.18, + "grad_norm": 2.142542352243009, + "learning_rate": 1.8827516857930706e-05, + "loss": 1.0218, + "step": 7684 + }, + { + "epoch": 0.18, + "grad_norm": 2.285856352547772, + "learning_rate": 1.882715832515444e-05, + "loss": 1.0638, + "step": 7685 + }, + { + "epoch": 0.18, + "grad_norm": 2.3360821478211578, + "learning_rate": 1.8826799740983665e-05, + "loss": 1.0874, + "step": 7686 + }, + { + "epoch": 0.18, + "grad_norm": 1.956199811659726, + "learning_rate": 1.882644110542047e-05, + "loss": 0.9713, + "step": 7687 + }, + { + "epoch": 0.18, + "grad_norm": 2.2179224474655106, + "learning_rate": 1.8826082418466938e-05, + "loss": 1.0013, + "step": 7688 + }, + { + "epoch": 0.18, + "grad_norm": 2.515540055560835, + "learning_rate": 1.882572368012516e-05, + "loss": 1.0921, + "step": 7689 + }, + { + "epoch": 0.18, + "grad_norm": 2.201177223862493, + "learning_rate": 1.8825364890397225e-05, + "loss": 1.0756, + "step": 7690 + }, + { + "epoch": 0.18, + "grad_norm": 1.1775751522242186, + "learning_rate": 1.8825006049285225e-05, + "loss": 1.0526, + "step": 7691 + }, + { + "epoch": 0.18, + "grad_norm": 1.8886597212224843, + "learning_rate": 1.882464715679124e-05, + "loss": 0.9907, + "step": 7692 + }, + { + "epoch": 0.18, + "grad_norm": 1.805291530436335, + "learning_rate": 1.8824288212917367e-05, + "loss": 0.9939, + "step": 7693 + }, + { + "epoch": 0.18, + "grad_norm": 2.5588986968539094, + "learning_rate": 1.8823929217665696e-05, + "loss": 1.1548, + "step": 7694 + }, + { + "epoch": 0.18, + "grad_norm": 2.072600509305261, + "learning_rate": 1.8823570171038316e-05, + "loss": 0.9731, + "step": 7695 + }, + { + "epoch": 0.18, + "grad_norm": 2.1066388783809282, + "learning_rate": 1.8823211073037316e-05, + "loss": 1.0775, + "step": 7696 + }, + { + "epoch": 0.18, + "grad_norm": 2.0848873527459837, + "learning_rate": 1.8822851923664785e-05, + "loss": 1.1357, + "step": 7697 + }, + { + "epoch": 0.18, + "grad_norm": 1.14118185080123, + "learning_rate": 1.882249272292282e-05, + "loss": 0.92, + "step": 7698 + }, + { + "epoch": 0.18, + "grad_norm": 2.106726210765647, + "learning_rate": 1.8822133470813507e-05, + "loss": 0.9624, + "step": 7699 + }, + { + "epoch": 0.18, + "grad_norm": 2.1174002499597346, + "learning_rate": 1.8821774167338944e-05, + "loss": 1.0601, + "step": 7700 + }, + { + "epoch": 0.18, + "grad_norm": 1.999714460213352, + "learning_rate": 1.8821414812501214e-05, + "loss": 1.0733, + "step": 7701 + }, + { + "epoch": 0.18, + "grad_norm": 2.3499191856125545, + "learning_rate": 1.8821055406302418e-05, + "loss": 1.0858, + "step": 7702 + }, + { + "epoch": 0.18, + "grad_norm": 2.123789586863293, + "learning_rate": 1.882069594874464e-05, + "loss": 1.1072, + "step": 7703 + }, + { + "epoch": 0.18, + "grad_norm": 2.322897779512849, + "learning_rate": 1.882033643982998e-05, + "loss": 1.0851, + "step": 7704 + }, + { + "epoch": 0.18, + "grad_norm": 2.082076590412923, + "learning_rate": 1.8819976879560534e-05, + "loss": 1.2218, + "step": 7705 + }, + { + "epoch": 0.18, + "grad_norm": 2.1520461079594813, + "learning_rate": 1.8819617267938383e-05, + "loss": 1.1277, + "step": 7706 + }, + { + "epoch": 0.18, + "grad_norm": 2.162741105512857, + "learning_rate": 1.881925760496563e-05, + "loss": 1.0095, + "step": 7707 + }, + { + "epoch": 0.18, + "grad_norm": 3.468789740095303, + "learning_rate": 1.881889789064437e-05, + "loss": 0.9615, + "step": 7708 + }, + { + "epoch": 0.18, + "grad_norm": 2.0790988628473848, + "learning_rate": 1.881853812497669e-05, + "loss": 1.1332, + "step": 7709 + }, + { + "epoch": 0.18, + "grad_norm": 2.1940217044603156, + "learning_rate": 1.881817830796469e-05, + "loss": 0.9878, + "step": 7710 + }, + { + "epoch": 0.18, + "grad_norm": 2.4071085152785145, + "learning_rate": 1.8817818439610466e-05, + "loss": 0.9747, + "step": 7711 + }, + { + "epoch": 0.18, + "grad_norm": 2.033367740753719, + "learning_rate": 1.8817458519916113e-05, + "loss": 1.059, + "step": 7712 + }, + { + "epoch": 0.18, + "grad_norm": 2.0580420974170175, + "learning_rate": 1.881709854888372e-05, + "loss": 0.9772, + "step": 7713 + }, + { + "epoch": 0.18, + "grad_norm": 2.0194482305523795, + "learning_rate": 1.881673852651539e-05, + "loss": 1.0629, + "step": 7714 + }, + { + "epoch": 0.18, + "grad_norm": 2.206448689308159, + "learning_rate": 1.881637845281322e-05, + "loss": 1.0903, + "step": 7715 + }, + { + "epoch": 0.18, + "grad_norm": 2.005278568720121, + "learning_rate": 1.88160183277793e-05, + "loss": 1.0311, + "step": 7716 + }, + { + "epoch": 0.18, + "grad_norm": 1.7684181213048111, + "learning_rate": 1.8815658151415734e-05, + "loss": 1.0665, + "step": 7717 + }, + { + "epoch": 0.18, + "grad_norm": 1.966791935917355, + "learning_rate": 1.8815297923724612e-05, + "loss": 1.1148, + "step": 7718 + }, + { + "epoch": 0.18, + "grad_norm": 2.0233168690879153, + "learning_rate": 1.8814937644708035e-05, + "loss": 1.0378, + "step": 7719 + }, + { + "epoch": 0.18, + "grad_norm": 2.339490533280605, + "learning_rate": 1.8814577314368098e-05, + "loss": 0.9645, + "step": 7720 + }, + { + "epoch": 0.18, + "grad_norm": 1.9865211464065409, + "learning_rate": 1.8814216932706905e-05, + "loss": 1.1066, + "step": 7721 + }, + { + "epoch": 0.18, + "grad_norm": 2.0590736184548746, + "learning_rate": 1.8813856499726548e-05, + "loss": 0.993, + "step": 7722 + }, + { + "epoch": 0.18, + "grad_norm": 1.8887538201164982, + "learning_rate": 1.881349601542913e-05, + "loss": 1.0821, + "step": 7723 + }, + { + "epoch": 0.18, + "grad_norm": 1.786517875918586, + "learning_rate": 1.8813135479816748e-05, + "loss": 1.0669, + "step": 7724 + }, + { + "epoch": 0.18, + "grad_norm": 2.601309072333826, + "learning_rate": 1.88127748928915e-05, + "loss": 1.0648, + "step": 7725 + }, + { + "epoch": 0.18, + "grad_norm": 2.63067493857763, + "learning_rate": 1.8812414254655483e-05, + "loss": 0.9613, + "step": 7726 + }, + { + "epoch": 0.18, + "grad_norm": 2.2359418246980813, + "learning_rate": 1.8812053565110804e-05, + "loss": 0.8946, + "step": 7727 + }, + { + "epoch": 0.18, + "grad_norm": 1.2237071292911217, + "learning_rate": 1.881169282425956e-05, + "loss": 0.9468, + "step": 7728 + }, + { + "epoch": 0.18, + "grad_norm": 2.016149502359781, + "learning_rate": 1.881133203210385e-05, + "loss": 0.9554, + "step": 7729 + }, + { + "epoch": 0.18, + "grad_norm": 2.026189638885105, + "learning_rate": 1.8810971188645775e-05, + "loss": 1.1058, + "step": 7730 + }, + { + "epoch": 0.18, + "grad_norm": 1.9226774724884164, + "learning_rate": 1.8810610293887435e-05, + "loss": 1.0611, + "step": 7731 + }, + { + "epoch": 0.18, + "grad_norm": 2.161371966069379, + "learning_rate": 1.8810249347830935e-05, + "loss": 0.9868, + "step": 7732 + }, + { + "epoch": 0.18, + "grad_norm": 2.1111375359221127, + "learning_rate": 1.8809888350478372e-05, + "loss": 0.9757, + "step": 7733 + }, + { + "epoch": 0.18, + "grad_norm": 2.2944366165241887, + "learning_rate": 1.880952730183185e-05, + "loss": 0.9701, + "step": 7734 + }, + { + "epoch": 0.18, + "grad_norm": 2.3487100312412528, + "learning_rate": 1.880916620189347e-05, + "loss": 1.1257, + "step": 7735 + }, + { + "epoch": 0.18, + "grad_norm": 2.0061157182552387, + "learning_rate": 1.880880505066534e-05, + "loss": 1.115, + "step": 7736 + }, + { + "epoch": 0.18, + "grad_norm": 1.9566156692982974, + "learning_rate": 1.8808443848149556e-05, + "loss": 1.1433, + "step": 7737 + }, + { + "epoch": 0.18, + "grad_norm": 1.979376273861702, + "learning_rate": 1.880808259434822e-05, + "loss": 1.1979, + "step": 7738 + }, + { + "epoch": 0.18, + "grad_norm": 1.9843304366384114, + "learning_rate": 1.8807721289263442e-05, + "loss": 0.9796, + "step": 7739 + }, + { + "epoch": 0.18, + "grad_norm": 1.9120301075812947, + "learning_rate": 1.880735993289732e-05, + "loss": 1.0216, + "step": 7740 + }, + { + "epoch": 0.18, + "grad_norm": 2.0280453215333227, + "learning_rate": 1.8806998525251964e-05, + "loss": 1.1697, + "step": 7741 + }, + { + "epoch": 0.18, + "grad_norm": 2.7080771930513285, + "learning_rate": 1.880663706632947e-05, + "loss": 1.0127, + "step": 7742 + }, + { + "epoch": 0.18, + "grad_norm": 2.113520644721182, + "learning_rate": 1.8806275556131948e-05, + "loss": 1.2388, + "step": 7743 + }, + { + "epoch": 0.18, + "grad_norm": 2.186983807043506, + "learning_rate": 1.88059139946615e-05, + "loss": 0.8635, + "step": 7744 + }, + { + "epoch": 0.18, + "grad_norm": 2.237219482195569, + "learning_rate": 1.8805552381920238e-05, + "loss": 0.9965, + "step": 7745 + }, + { + "epoch": 0.18, + "grad_norm": 1.958834518080546, + "learning_rate": 1.880519071791026e-05, + "loss": 1.0136, + "step": 7746 + }, + { + "epoch": 0.18, + "grad_norm": 2.0498397798246786, + "learning_rate": 1.8804829002633672e-05, + "loss": 1.1281, + "step": 7747 + }, + { + "epoch": 0.18, + "grad_norm": 2.228315370620081, + "learning_rate": 1.8804467236092584e-05, + "loss": 0.9647, + "step": 7748 + }, + { + "epoch": 0.18, + "grad_norm": 2.2632014356195875, + "learning_rate": 1.8804105418289097e-05, + "loss": 1.0673, + "step": 7749 + }, + { + "epoch": 0.18, + "grad_norm": 2.6956258635697274, + "learning_rate": 1.8803743549225323e-05, + "loss": 1.0249, + "step": 7750 + }, + { + "epoch": 0.18, + "grad_norm": 2.0633081456792666, + "learning_rate": 1.880338162890337e-05, + "loss": 1.0579, + "step": 7751 + }, + { + "epoch": 0.18, + "grad_norm": 1.1056231515787343, + "learning_rate": 1.8803019657325338e-05, + "loss": 0.9895, + "step": 7752 + }, + { + "epoch": 0.18, + "grad_norm": 1.8456108631557997, + "learning_rate": 1.880265763449334e-05, + "loss": 1.1877, + "step": 7753 + }, + { + "epoch": 0.18, + "grad_norm": 1.9847415135981248, + "learning_rate": 1.880229556040948e-05, + "loss": 1.1029, + "step": 7754 + }, + { + "epoch": 0.18, + "grad_norm": 1.1714522754655665, + "learning_rate": 1.8801933435075872e-05, + "loss": 0.9897, + "step": 7755 + }, + { + "epoch": 0.18, + "grad_norm": 2.141994321800665, + "learning_rate": 1.880157125849462e-05, + "loss": 1.1562, + "step": 7756 + }, + { + "epoch": 0.18, + "grad_norm": 1.0896708308983998, + "learning_rate": 1.8801209030667834e-05, + "loss": 0.9806, + "step": 7757 + }, + { + "epoch": 0.18, + "grad_norm": 1.9871921917522821, + "learning_rate": 1.880084675159762e-05, + "loss": 1.1607, + "step": 7758 + }, + { + "epoch": 0.18, + "grad_norm": 2.01779010900547, + "learning_rate": 1.8800484421286094e-05, + "loss": 1.1723, + "step": 7759 + }, + { + "epoch": 0.18, + "grad_norm": 2.511055447095848, + "learning_rate": 1.880012203973536e-05, + "loss": 1.1295, + "step": 7760 + }, + { + "epoch": 0.18, + "grad_norm": 2.033095410541662, + "learning_rate": 1.8799759606947527e-05, + "loss": 1.0769, + "step": 7761 + }, + { + "epoch": 0.18, + "grad_norm": 2.2598841249937114, + "learning_rate": 1.8799397122924708e-05, + "loss": 1.0583, + "step": 7762 + }, + { + "epoch": 0.18, + "grad_norm": 2.0515842534554913, + "learning_rate": 1.8799034587669016e-05, + "loss": 1.169, + "step": 7763 + }, + { + "epoch": 0.18, + "grad_norm": 1.1815908354523537, + "learning_rate": 1.8798672001182553e-05, + "loss": 1.0198, + "step": 7764 + }, + { + "epoch": 0.18, + "grad_norm": 2.3173210423761486, + "learning_rate": 1.8798309363467446e-05, + "loss": 1.0438, + "step": 7765 + }, + { + "epoch": 0.18, + "grad_norm": 1.9271117544239729, + "learning_rate": 1.8797946674525788e-05, + "loss": 1.0472, + "step": 7766 + }, + { + "epoch": 0.18, + "grad_norm": 2.115872091440888, + "learning_rate": 1.87975839343597e-05, + "loss": 1.0494, + "step": 7767 + }, + { + "epoch": 0.18, + "grad_norm": 2.023568581096614, + "learning_rate": 1.8797221142971295e-05, + "loss": 1.1338, + "step": 7768 + }, + { + "epoch": 0.18, + "grad_norm": 2.018064184508357, + "learning_rate": 1.8796858300362682e-05, + "loss": 0.9967, + "step": 7769 + }, + { + "epoch": 0.18, + "grad_norm": 3.217658398081215, + "learning_rate": 1.879649540653598e-05, + "loss": 0.9738, + "step": 7770 + }, + { + "epoch": 0.18, + "grad_norm": 2.1073336254650448, + "learning_rate": 1.879613246149329e-05, + "loss": 1.1643, + "step": 7771 + }, + { + "epoch": 0.18, + "grad_norm": 2.4102240213637507, + "learning_rate": 1.8795769465236735e-05, + "loss": 1.0454, + "step": 7772 + }, + { + "epoch": 0.18, + "grad_norm": 2.158825739317997, + "learning_rate": 1.8795406417768426e-05, + "loss": 1.2107, + "step": 7773 + }, + { + "epoch": 0.18, + "grad_norm": 1.1197598377036446, + "learning_rate": 1.8795043319090477e-05, + "loss": 1.0417, + "step": 7774 + }, + { + "epoch": 0.18, + "grad_norm": 1.7790914733242373, + "learning_rate": 1.8794680169205e-05, + "loss": 1.0081, + "step": 7775 + }, + { + "epoch": 0.18, + "grad_norm": 2.257004759601036, + "learning_rate": 1.879431696811411e-05, + "loss": 1.0073, + "step": 7776 + }, + { + "epoch": 0.18, + "grad_norm": 2.0861702402516533, + "learning_rate": 1.8793953715819922e-05, + "loss": 0.985, + "step": 7777 + }, + { + "epoch": 0.18, + "grad_norm": 2.279292110913748, + "learning_rate": 1.8793590412324555e-05, + "loss": 1.0262, + "step": 7778 + }, + { + "epoch": 0.18, + "grad_norm": 1.9086995142942347, + "learning_rate": 1.8793227057630115e-05, + "loss": 0.9253, + "step": 7779 + }, + { + "epoch": 0.18, + "grad_norm": 2.1472576798910876, + "learning_rate": 1.8792863651738727e-05, + "loss": 1.1964, + "step": 7780 + }, + { + "epoch": 0.18, + "grad_norm": 2.0102264679469375, + "learning_rate": 1.8792500194652505e-05, + "loss": 1.013, + "step": 7781 + }, + { + "epoch": 0.18, + "grad_norm": 2.1979782483248265, + "learning_rate": 1.879213668637356e-05, + "loss": 1.1084, + "step": 7782 + }, + { + "epoch": 0.18, + "grad_norm": 2.420664650591447, + "learning_rate": 1.879177312690401e-05, + "loss": 0.9543, + "step": 7783 + }, + { + "epoch": 0.18, + "grad_norm": 2.2793363734862333, + "learning_rate": 1.8791409516245974e-05, + "loss": 1.1902, + "step": 7784 + }, + { + "epoch": 0.18, + "grad_norm": 4.004027667007307, + "learning_rate": 1.879104585440157e-05, + "loss": 0.9599, + "step": 7785 + }, + { + "epoch": 0.18, + "grad_norm": 2.0700512225223098, + "learning_rate": 1.8790682141372913e-05, + "loss": 1.1119, + "step": 7786 + }, + { + "epoch": 0.18, + "grad_norm": 1.9011753302145575, + "learning_rate": 1.8790318377162126e-05, + "loss": 0.9727, + "step": 7787 + }, + { + "epoch": 0.18, + "grad_norm": 2.066487461479117, + "learning_rate": 1.8789954561771314e-05, + "loss": 0.9661, + "step": 7788 + }, + { + "epoch": 0.18, + "grad_norm": 1.8550669057402984, + "learning_rate": 1.878959069520261e-05, + "loss": 0.9937, + "step": 7789 + }, + { + "epoch": 0.18, + "grad_norm": 2.076488920588085, + "learning_rate": 1.8789226777458126e-05, + "loss": 1.0364, + "step": 7790 + }, + { + "epoch": 0.18, + "grad_norm": 2.438944441000722, + "learning_rate": 1.878886280853998e-05, + "loss": 0.965, + "step": 7791 + }, + { + "epoch": 0.18, + "grad_norm": 2.5936768930649996, + "learning_rate": 1.878849878845029e-05, + "loss": 1.0599, + "step": 7792 + }, + { + "epoch": 0.18, + "grad_norm": 1.9421701120628634, + "learning_rate": 1.8788134717191182e-05, + "loss": 1.0436, + "step": 7793 + }, + { + "epoch": 0.18, + "grad_norm": 2.267075614588769, + "learning_rate": 1.8787770594764768e-05, + "loss": 0.9614, + "step": 7794 + }, + { + "epoch": 0.18, + "grad_norm": 2.279893417978338, + "learning_rate": 1.8787406421173173e-05, + "loss": 1.0586, + "step": 7795 + }, + { + "epoch": 0.18, + "grad_norm": 2.034183756979444, + "learning_rate": 1.8787042196418514e-05, + "loss": 1.0745, + "step": 7796 + }, + { + "epoch": 0.18, + "grad_norm": 2.032734441971988, + "learning_rate": 1.8786677920502913e-05, + "loss": 1.1348, + "step": 7797 + }, + { + "epoch": 0.18, + "grad_norm": 1.8547557627935238, + "learning_rate": 1.8786313593428493e-05, + "loss": 1.0489, + "step": 7798 + }, + { + "epoch": 0.18, + "grad_norm": 2.434743275353152, + "learning_rate": 1.8785949215197377e-05, + "loss": 1.158, + "step": 7799 + }, + { + "epoch": 0.18, + "grad_norm": 2.014658846530106, + "learning_rate": 1.8785584785811677e-05, + "loss": 1.0214, + "step": 7800 + }, + { + "epoch": 0.18, + "grad_norm": 1.9349619123903643, + "learning_rate": 1.8785220305273526e-05, + "loss": 1.0712, + "step": 7801 + }, + { + "epoch": 0.18, + "grad_norm": 2.0844865270382322, + "learning_rate": 1.8784855773585037e-05, + "loss": 1.0913, + "step": 7802 + }, + { + "epoch": 0.18, + "grad_norm": 2.54546394950807, + "learning_rate": 1.8784491190748338e-05, + "loss": 0.9536, + "step": 7803 + }, + { + "epoch": 0.18, + "grad_norm": 2.4028875372254093, + "learning_rate": 1.878412655676555e-05, + "loss": 0.9856, + "step": 7804 + }, + { + "epoch": 0.18, + "grad_norm": 2.877742211254941, + "learning_rate": 1.8783761871638798e-05, + "loss": 1.0988, + "step": 7805 + }, + { + "epoch": 0.18, + "grad_norm": 2.1750429033255885, + "learning_rate": 1.87833971353702e-05, + "loss": 1.0996, + "step": 7806 + }, + { + "epoch": 0.18, + "grad_norm": 1.1410378050419567, + "learning_rate": 1.8783032347961884e-05, + "loss": 0.9056, + "step": 7807 + }, + { + "epoch": 0.18, + "grad_norm": 2.0794157392856847, + "learning_rate": 1.8782667509415974e-05, + "loss": 1.0932, + "step": 7808 + }, + { + "epoch": 0.18, + "grad_norm": 2.1304113718466158, + "learning_rate": 1.8782302619734594e-05, + "loss": 1.0306, + "step": 7809 + }, + { + "epoch": 0.18, + "grad_norm": 1.9215681665337834, + "learning_rate": 1.8781937678919867e-05, + "loss": 1.0616, + "step": 7810 + }, + { + "epoch": 0.18, + "grad_norm": 2.7976044159558566, + "learning_rate": 1.878157268697392e-05, + "loss": 1.0458, + "step": 7811 + }, + { + "epoch": 0.18, + "grad_norm": 1.059093570030813, + "learning_rate": 1.8781207643898873e-05, + "loss": 1.0096, + "step": 7812 + }, + { + "epoch": 0.18, + "grad_norm": 2.0632601282355942, + "learning_rate": 1.8780842549696857e-05, + "loss": 0.9847, + "step": 7813 + }, + { + "epoch": 0.18, + "grad_norm": 2.1115610715274085, + "learning_rate": 1.8780477404369996e-05, + "loss": 0.9576, + "step": 7814 + }, + { + "epoch": 0.18, + "grad_norm": 2.651077206834746, + "learning_rate": 1.8780112207920415e-05, + "loss": 1.1052, + "step": 7815 + }, + { + "epoch": 0.18, + "grad_norm": 2.014207067290121, + "learning_rate": 1.8779746960350243e-05, + "loss": 0.891, + "step": 7816 + }, + { + "epoch": 0.18, + "grad_norm": 1.9168291960269623, + "learning_rate": 1.8779381661661603e-05, + "loss": 1.1277, + "step": 7817 + }, + { + "epoch": 0.18, + "grad_norm": 2.5179609238147576, + "learning_rate": 1.877901631185662e-05, + "loss": 1.005, + "step": 7818 + }, + { + "epoch": 0.18, + "grad_norm": 2.2393967876071788, + "learning_rate": 1.877865091093743e-05, + "loss": 1.2302, + "step": 7819 + }, + { + "epoch": 0.18, + "grad_norm": 1.9805529241948021, + "learning_rate": 1.8778285458906152e-05, + "loss": 1.0967, + "step": 7820 + }, + { + "epoch": 0.18, + "grad_norm": 2.221801256479777, + "learning_rate": 1.8777919955764917e-05, + "loss": 1.0864, + "step": 7821 + }, + { + "epoch": 0.18, + "grad_norm": 2.5811754393557296, + "learning_rate": 1.8777554401515855e-05, + "loss": 1.0474, + "step": 7822 + }, + { + "epoch": 0.18, + "grad_norm": 1.9145600347385292, + "learning_rate": 1.877718879616109e-05, + "loss": 0.9759, + "step": 7823 + }, + { + "epoch": 0.18, + "grad_norm": 2.1748513428497405, + "learning_rate": 1.877682313970276e-05, + "loss": 1.0982, + "step": 7824 + }, + { + "epoch": 0.18, + "grad_norm": 1.8259655007114928, + "learning_rate": 1.877645743214298e-05, + "loss": 1.0815, + "step": 7825 + }, + { + "epoch": 0.18, + "grad_norm": 2.324164808973706, + "learning_rate": 1.8776091673483886e-05, + "loss": 1.1429, + "step": 7826 + }, + { + "epoch": 0.18, + "grad_norm": 2.4179582397235593, + "learning_rate": 1.877572586372761e-05, + "loss": 1.0046, + "step": 7827 + }, + { + "epoch": 0.18, + "grad_norm": 2.155951594180722, + "learning_rate": 1.877536000287628e-05, + "loss": 1.0086, + "step": 7828 + }, + { + "epoch": 0.18, + "grad_norm": 2.2087024687811168, + "learning_rate": 1.8774994090932025e-05, + "loss": 1.1034, + "step": 7829 + }, + { + "epoch": 0.18, + "grad_norm": 2.1189586666211864, + "learning_rate": 1.8774628127896976e-05, + "loss": 1.0347, + "step": 7830 + }, + { + "epoch": 0.18, + "grad_norm": 2.30607129283082, + "learning_rate": 1.8774262113773262e-05, + "loss": 0.9505, + "step": 7831 + }, + { + "epoch": 0.18, + "grad_norm": 2.1288229068576094, + "learning_rate": 1.877389604856302e-05, + "loss": 1.0607, + "step": 7832 + }, + { + "epoch": 0.18, + "grad_norm": 2.0396508781806073, + "learning_rate": 1.8773529932268374e-05, + "loss": 1.0064, + "step": 7833 + }, + { + "epoch": 0.18, + "grad_norm": 2.3547680777015567, + "learning_rate": 1.8773163764891464e-05, + "loss": 1.1276, + "step": 7834 + }, + { + "epoch": 0.18, + "grad_norm": 2.1992553332904246, + "learning_rate": 1.8772797546434415e-05, + "loss": 1.0661, + "step": 7835 + }, + { + "epoch": 0.18, + "grad_norm": 2.870909001689559, + "learning_rate": 1.8772431276899358e-05, + "loss": 1.0104, + "step": 7836 + }, + { + "epoch": 0.18, + "grad_norm": 2.2497745179899598, + "learning_rate": 1.877206495628843e-05, + "loss": 1.134, + "step": 7837 + }, + { + "epoch": 0.18, + "grad_norm": 2.1257567646914963, + "learning_rate": 1.8771698584603766e-05, + "loss": 1.103, + "step": 7838 + }, + { + "epoch": 0.18, + "grad_norm": 1.8695485453198193, + "learning_rate": 1.8771332161847493e-05, + "loss": 0.9942, + "step": 7839 + }, + { + "epoch": 0.18, + "grad_norm": 1.1318637774123457, + "learning_rate": 1.8770965688021747e-05, + "loss": 0.9618, + "step": 7840 + }, + { + "epoch": 0.18, + "grad_norm": 2.518125825358059, + "learning_rate": 1.8770599163128663e-05, + "loss": 1.0691, + "step": 7841 + }, + { + "epoch": 0.18, + "grad_norm": 2.555581710466691, + "learning_rate": 1.8770232587170374e-05, + "loss": 1.1072, + "step": 7842 + }, + { + "epoch": 0.18, + "grad_norm": 2.0481429133815574, + "learning_rate": 1.8769865960149012e-05, + "loss": 1.0154, + "step": 7843 + }, + { + "epoch": 0.18, + "grad_norm": 2.190506145717745, + "learning_rate": 1.8769499282066716e-05, + "loss": 1.0692, + "step": 7844 + }, + { + "epoch": 0.18, + "grad_norm": 2.231170513315185, + "learning_rate": 1.876913255292562e-05, + "loss": 1.0059, + "step": 7845 + }, + { + "epoch": 0.18, + "grad_norm": 1.9537018734638647, + "learning_rate": 1.8768765772727856e-05, + "loss": 1.0793, + "step": 7846 + }, + { + "epoch": 0.18, + "grad_norm": 1.8706989006103114, + "learning_rate": 1.876839894147556e-05, + "loss": 1.009, + "step": 7847 + }, + { + "epoch": 0.18, + "grad_norm": 1.9019883843302414, + "learning_rate": 1.8768032059170874e-05, + "loss": 1.1163, + "step": 7848 + }, + { + "epoch": 0.18, + "grad_norm": 1.9529450690120307, + "learning_rate": 1.8767665125815927e-05, + "loss": 1.1475, + "step": 7849 + }, + { + "epoch": 0.18, + "grad_norm": 2.2030351988054067, + "learning_rate": 1.876729814141286e-05, + "loss": 0.9972, + "step": 7850 + }, + { + "epoch": 0.18, + "grad_norm": 2.02431072317766, + "learning_rate": 1.87669311059638e-05, + "loss": 1.0442, + "step": 7851 + }, + { + "epoch": 0.18, + "grad_norm": 1.9111157399722454, + "learning_rate": 1.8766564019470898e-05, + "loss": 1.1882, + "step": 7852 + }, + { + "epoch": 0.19, + "grad_norm": 1.9429298785295517, + "learning_rate": 1.8766196881936284e-05, + "loss": 0.9305, + "step": 7853 + }, + { + "epoch": 0.19, + "grad_norm": 2.045779431871879, + "learning_rate": 1.8765829693362097e-05, + "loss": 0.8822, + "step": 7854 + }, + { + "epoch": 0.19, + "grad_norm": 1.929101375825268, + "learning_rate": 1.8765462453750472e-05, + "loss": 1.15, + "step": 7855 + }, + { + "epoch": 0.19, + "grad_norm": 2.165799889836744, + "learning_rate": 1.876509516310355e-05, + "loss": 1.0601, + "step": 7856 + }, + { + "epoch": 0.19, + "grad_norm": 1.1178157487731872, + "learning_rate": 1.8764727821423473e-05, + "loss": 0.9755, + "step": 7857 + }, + { + "epoch": 0.19, + "grad_norm": 2.9496078692399412, + "learning_rate": 1.876436042871237e-05, + "loss": 1.0432, + "step": 7858 + }, + { + "epoch": 0.19, + "grad_norm": 1.9656550570208897, + "learning_rate": 1.8763992984972388e-05, + "loss": 0.9468, + "step": 7859 + }, + { + "epoch": 0.19, + "grad_norm": 1.9672501867565948, + "learning_rate": 1.8763625490205664e-05, + "loss": 0.928, + "step": 7860 + }, + { + "epoch": 0.19, + "grad_norm": 2.258748144017472, + "learning_rate": 1.876325794441434e-05, + "loss": 1.0531, + "step": 7861 + }, + { + "epoch": 0.19, + "grad_norm": 2.0591037885123846, + "learning_rate": 1.876289034760055e-05, + "loss": 1.0221, + "step": 7862 + }, + { + "epoch": 0.19, + "grad_norm": 2.342400209151075, + "learning_rate": 1.8762522699766443e-05, + "loss": 0.9831, + "step": 7863 + }, + { + "epoch": 0.19, + "grad_norm": 2.036766884208102, + "learning_rate": 1.876215500091415e-05, + "loss": 1.001, + "step": 7864 + }, + { + "epoch": 0.19, + "grad_norm": 1.0756343145420189, + "learning_rate": 1.876178725104582e-05, + "loss": 0.9915, + "step": 7865 + }, + { + "epoch": 0.19, + "grad_norm": 1.1592970867370158, + "learning_rate": 1.8761419450163586e-05, + "loss": 1.0494, + "step": 7866 + }, + { + "epoch": 0.19, + "grad_norm": 2.0894469392477752, + "learning_rate": 1.87610515982696e-05, + "loss": 0.9952, + "step": 7867 + }, + { + "epoch": 0.19, + "grad_norm": 2.0563056576564023, + "learning_rate": 1.8760683695365994e-05, + "loss": 0.9543, + "step": 7868 + }, + { + "epoch": 0.19, + "grad_norm": 2.072108862899697, + "learning_rate": 1.8760315741454917e-05, + "loss": 1.0759, + "step": 7869 + }, + { + "epoch": 0.19, + "grad_norm": 2.0214698426442994, + "learning_rate": 1.8759947736538507e-05, + "loss": 0.9813, + "step": 7870 + }, + { + "epoch": 0.19, + "grad_norm": 2.148675587964925, + "learning_rate": 1.8759579680618907e-05, + "loss": 1.1875, + "step": 7871 + }, + { + "epoch": 0.19, + "grad_norm": 1.920516578129515, + "learning_rate": 1.8759211573698263e-05, + "loss": 1.1055, + "step": 7872 + }, + { + "epoch": 0.19, + "grad_norm": 1.9854618205842833, + "learning_rate": 1.875884341577871e-05, + "loss": 1.1214, + "step": 7873 + }, + { + "epoch": 0.19, + "grad_norm": 2.9079952739744006, + "learning_rate": 1.87584752068624e-05, + "loss": 1.0667, + "step": 7874 + }, + { + "epoch": 0.19, + "grad_norm": 2.0330733813023922, + "learning_rate": 1.8758106946951484e-05, + "loss": 1.1084, + "step": 7875 + }, + { + "epoch": 0.19, + "grad_norm": 2.170462072950841, + "learning_rate": 1.8757738636048088e-05, + "loss": 1.1178, + "step": 7876 + }, + { + "epoch": 0.19, + "grad_norm": 2.0687098925416136, + "learning_rate": 1.8757370274154362e-05, + "loss": 0.9954, + "step": 7877 + }, + { + "epoch": 0.19, + "grad_norm": 2.014541754322593, + "learning_rate": 1.875700186127246e-05, + "loss": 1.1268, + "step": 7878 + }, + { + "epoch": 0.19, + "grad_norm": 2.076063078140692, + "learning_rate": 1.8756633397404518e-05, + "loss": 1.0002, + "step": 7879 + }, + { + "epoch": 0.19, + "grad_norm": 2.0876331637852767, + "learning_rate": 1.8756264882552685e-05, + "loss": 1.0103, + "step": 7880 + }, + { + "epoch": 0.19, + "grad_norm": 2.029181649697313, + "learning_rate": 1.8755896316719104e-05, + "loss": 1.0433, + "step": 7881 + }, + { + "epoch": 0.19, + "grad_norm": 2.0947785587793226, + "learning_rate": 1.8755527699905927e-05, + "loss": 1.0617, + "step": 7882 + }, + { + "epoch": 0.19, + "grad_norm": 1.2666052496186577, + "learning_rate": 1.8755159032115292e-05, + "loss": 0.9939, + "step": 7883 + }, + { + "epoch": 0.19, + "grad_norm": 1.8502013530298362, + "learning_rate": 1.8754790313349347e-05, + "loss": 0.9603, + "step": 7884 + }, + { + "epoch": 0.19, + "grad_norm": 1.8893913182213216, + "learning_rate": 1.8754421543610247e-05, + "loss": 1.0084, + "step": 7885 + }, + { + "epoch": 0.19, + "grad_norm": 2.4131729640188593, + "learning_rate": 1.875405272290013e-05, + "loss": 0.963, + "step": 7886 + }, + { + "epoch": 0.19, + "grad_norm": 2.927630959393194, + "learning_rate": 1.8753683851221146e-05, + "loss": 1.0061, + "step": 7887 + }, + { + "epoch": 0.19, + "grad_norm": 1.9120250228866316, + "learning_rate": 1.8753314928575444e-05, + "loss": 0.9615, + "step": 7888 + }, + { + "epoch": 0.19, + "grad_norm": 2.1181105627397416, + "learning_rate": 1.875294595496517e-05, + "loss": 1.0726, + "step": 7889 + }, + { + "epoch": 0.19, + "grad_norm": 1.202459949019557, + "learning_rate": 1.8752576930392473e-05, + "loss": 0.926, + "step": 7890 + }, + { + "epoch": 0.19, + "grad_norm": 2.1550353021949666, + "learning_rate": 1.8752207854859507e-05, + "loss": 1.1315, + "step": 7891 + }, + { + "epoch": 0.19, + "grad_norm": 2.1015408060131167, + "learning_rate": 1.875183872836841e-05, + "loss": 1.0649, + "step": 7892 + }, + { + "epoch": 0.19, + "grad_norm": 3.0287339174245447, + "learning_rate": 1.8751469550921342e-05, + "loss": 1.0714, + "step": 7893 + }, + { + "epoch": 0.19, + "grad_norm": 2.007367054995582, + "learning_rate": 1.8751100322520447e-05, + "loss": 1.0265, + "step": 7894 + }, + { + "epoch": 0.19, + "grad_norm": 1.1657028329817047, + "learning_rate": 1.8750731043167873e-05, + "loss": 0.9961, + "step": 7895 + }, + { + "epoch": 0.19, + "grad_norm": 2.0966771018727908, + "learning_rate": 1.8750361712865774e-05, + "loss": 0.9818, + "step": 7896 + }, + { + "epoch": 0.19, + "grad_norm": 2.1641724878938153, + "learning_rate": 1.8749992331616296e-05, + "loss": 1.0546, + "step": 7897 + }, + { + "epoch": 0.19, + "grad_norm": 2.364026524430919, + "learning_rate": 1.8749622899421597e-05, + "loss": 0.965, + "step": 7898 + }, + { + "epoch": 0.19, + "grad_norm": 2.4691682437972604, + "learning_rate": 1.874925341628382e-05, + "loss": 1.1086, + "step": 7899 + }, + { + "epoch": 0.19, + "grad_norm": 1.9716385227268796, + "learning_rate": 1.874888388220512e-05, + "loss": 1.0282, + "step": 7900 + }, + { + "epoch": 0.19, + "grad_norm": 2.1221104300788833, + "learning_rate": 1.874851429718765e-05, + "loss": 1.1324, + "step": 7901 + }, + { + "epoch": 0.19, + "grad_norm": 2.1563468288107175, + "learning_rate": 1.8748144661233558e-05, + "loss": 0.8779, + "step": 7902 + }, + { + "epoch": 0.19, + "grad_norm": 1.8468608966501099, + "learning_rate": 1.8747774974345e-05, + "loss": 0.9993, + "step": 7903 + }, + { + "epoch": 0.19, + "grad_norm": 1.844935940882818, + "learning_rate": 1.8747405236524125e-05, + "loss": 1.1992, + "step": 7904 + }, + { + "epoch": 0.19, + "grad_norm": 2.0461935169673517, + "learning_rate": 1.8747035447773086e-05, + "loss": 0.9731, + "step": 7905 + }, + { + "epoch": 0.19, + "grad_norm": 1.9207823532951855, + "learning_rate": 1.874666560809404e-05, + "loss": 1.1119, + "step": 7906 + }, + { + "epoch": 0.19, + "grad_norm": 2.1613073590944496, + "learning_rate": 1.8746295717489138e-05, + "loss": 0.9706, + "step": 7907 + }, + { + "epoch": 0.19, + "grad_norm": 1.9532925340801646, + "learning_rate": 1.874592577596053e-05, + "loss": 1.0124, + "step": 7908 + }, + { + "epoch": 0.19, + "grad_norm": 2.1813198149773063, + "learning_rate": 1.8745555783510378e-05, + "loss": 1.071, + "step": 7909 + }, + { + "epoch": 0.19, + "grad_norm": 2.253723268140868, + "learning_rate": 1.874518574014083e-05, + "loss": 1.1203, + "step": 7910 + }, + { + "epoch": 0.19, + "grad_norm": 1.9850851439367274, + "learning_rate": 1.874481564585404e-05, + "loss": 0.9642, + "step": 7911 + }, + { + "epoch": 0.19, + "grad_norm": 1.9936717541861766, + "learning_rate": 1.8744445500652165e-05, + "loss": 1.0861, + "step": 7912 + }, + { + "epoch": 0.19, + "grad_norm": 2.1163781040199274, + "learning_rate": 1.8744075304537358e-05, + "loss": 1.0792, + "step": 7913 + }, + { + "epoch": 0.19, + "grad_norm": 2.3334306995913594, + "learning_rate": 1.8743705057511777e-05, + "loss": 1.0921, + "step": 7914 + }, + { + "epoch": 0.19, + "grad_norm": 2.3612000808912406, + "learning_rate": 1.8743334759577582e-05, + "loss": 1.0323, + "step": 7915 + }, + { + "epoch": 0.19, + "grad_norm": 2.211712122955756, + "learning_rate": 1.874296441073692e-05, + "loss": 0.9912, + "step": 7916 + }, + { + "epoch": 0.19, + "grad_norm": 1.88848189163472, + "learning_rate": 1.8742594010991953e-05, + "loss": 1.0786, + "step": 7917 + }, + { + "epoch": 0.19, + "grad_norm": 2.1785518365011542, + "learning_rate": 1.8742223560344835e-05, + "loss": 1.11, + "step": 7918 + }, + { + "epoch": 0.19, + "grad_norm": 2.2152803794198612, + "learning_rate": 1.874185305879772e-05, + "loss": 1.108, + "step": 7919 + }, + { + "epoch": 0.19, + "grad_norm": 1.152995753305818, + "learning_rate": 1.8741482506352776e-05, + "loss": 1.0259, + "step": 7920 + }, + { + "epoch": 0.19, + "grad_norm": 2.362470072156671, + "learning_rate": 1.8741111903012148e-05, + "loss": 1.0567, + "step": 7921 + }, + { + "epoch": 0.19, + "grad_norm": 3.9041487259926084, + "learning_rate": 1.8740741248778002e-05, + "loss": 0.9711, + "step": 7922 + }, + { + "epoch": 0.19, + "grad_norm": 3.733382000208449, + "learning_rate": 1.8740370543652487e-05, + "loss": 0.9189, + "step": 7923 + }, + { + "epoch": 0.19, + "grad_norm": 1.122079095254129, + "learning_rate": 1.8739999787637775e-05, + "loss": 0.9548, + "step": 7924 + }, + { + "epoch": 0.19, + "grad_norm": 1.8769197052278723, + "learning_rate": 1.8739628980736016e-05, + "loss": 1.0687, + "step": 7925 + }, + { + "epoch": 0.19, + "grad_norm": 2.3409253414353524, + "learning_rate": 1.8739258122949367e-05, + "loss": 1.0601, + "step": 7926 + }, + { + "epoch": 0.19, + "grad_norm": 2.104044124573439, + "learning_rate": 1.873888721427999e-05, + "loss": 1.0781, + "step": 7927 + }, + { + "epoch": 0.19, + "grad_norm": 2.2415037779342035, + "learning_rate": 1.8738516254730052e-05, + "loss": 0.9066, + "step": 7928 + }, + { + "epoch": 0.19, + "grad_norm": 2.4472266546687265, + "learning_rate": 1.87381452443017e-05, + "loss": 1.0664, + "step": 7929 + }, + { + "epoch": 0.19, + "grad_norm": 1.881964623951236, + "learning_rate": 1.87377741829971e-05, + "loss": 1.1842, + "step": 7930 + }, + { + "epoch": 0.19, + "grad_norm": 2.807646241576752, + "learning_rate": 1.873740307081841e-05, + "loss": 1.0549, + "step": 7931 + }, + { + "epoch": 0.19, + "grad_norm": 1.696508375091298, + "learning_rate": 1.8737031907767797e-05, + "loss": 1.1131, + "step": 7932 + }, + { + "epoch": 0.19, + "grad_norm": 2.268670868885196, + "learning_rate": 1.873666069384742e-05, + "loss": 1.033, + "step": 7933 + }, + { + "epoch": 0.19, + "grad_norm": 2.2316526640082026, + "learning_rate": 1.8736289429059436e-05, + "loss": 1.1508, + "step": 7934 + }, + { + "epoch": 0.19, + "grad_norm": 1.1655086410757447, + "learning_rate": 1.873591811340601e-05, + "loss": 0.991, + "step": 7935 + }, + { + "epoch": 0.19, + "grad_norm": 2.155138721067046, + "learning_rate": 1.87355467468893e-05, + "loss": 1.0495, + "step": 7936 + }, + { + "epoch": 0.19, + "grad_norm": 2.7535835242724174, + "learning_rate": 1.873517532951147e-05, + "loss": 0.9876, + "step": 7937 + }, + { + "epoch": 0.19, + "grad_norm": 2.342122756404818, + "learning_rate": 1.8734803861274692e-05, + "loss": 1.076, + "step": 7938 + }, + { + "epoch": 0.19, + "grad_norm": 2.0047393562622378, + "learning_rate": 1.8734432342181113e-05, + "loss": 1.0231, + "step": 7939 + }, + { + "epoch": 0.19, + "grad_norm": 2.4188833427037557, + "learning_rate": 1.8734060772232906e-05, + "loss": 1.0307, + "step": 7940 + }, + { + "epoch": 0.19, + "grad_norm": 1.9657549692077643, + "learning_rate": 1.873368915143223e-05, + "loss": 1.138, + "step": 7941 + }, + { + "epoch": 0.19, + "grad_norm": 1.8870356973003972, + "learning_rate": 1.8733317479781256e-05, + "loss": 1.0817, + "step": 7942 + }, + { + "epoch": 0.19, + "grad_norm": 1.810425157581043, + "learning_rate": 1.8732945757282137e-05, + "loss": 1.0113, + "step": 7943 + }, + { + "epoch": 0.19, + "grad_norm": 2.23580248852107, + "learning_rate": 1.8732573983937046e-05, + "loss": 1.1023, + "step": 7944 + }, + { + "epoch": 0.19, + "grad_norm": 1.9267615441761334, + "learning_rate": 1.8732202159748146e-05, + "loss": 1.0501, + "step": 7945 + }, + { + "epoch": 0.19, + "grad_norm": 2.1901768196373466, + "learning_rate": 1.8731830284717597e-05, + "loss": 1.0397, + "step": 7946 + }, + { + "epoch": 0.19, + "grad_norm": 2.0377601168660724, + "learning_rate": 1.873145835884757e-05, + "loss": 1.1854, + "step": 7947 + }, + { + "epoch": 0.19, + "grad_norm": 1.1326113758687972, + "learning_rate": 1.8731086382140228e-05, + "loss": 1.0279, + "step": 7948 + }, + { + "epoch": 0.19, + "grad_norm": 1.9652160710427518, + "learning_rate": 1.8730714354597738e-05, + "loss": 1.189, + "step": 7949 + }, + { + "epoch": 0.19, + "grad_norm": 2.102344874500795, + "learning_rate": 1.873034227622226e-05, + "loss": 1.0411, + "step": 7950 + }, + { + "epoch": 0.19, + "grad_norm": 2.279002044923147, + "learning_rate": 1.8729970147015968e-05, + "loss": 1.0194, + "step": 7951 + }, + { + "epoch": 0.19, + "grad_norm": 3.6691308787538395, + "learning_rate": 1.8729597966981025e-05, + "loss": 0.9224, + "step": 7952 + }, + { + "epoch": 0.19, + "grad_norm": 1.993166248391825, + "learning_rate": 1.8729225736119602e-05, + "loss": 1.0664, + "step": 7953 + }, + { + "epoch": 0.19, + "grad_norm": 2.141392567872087, + "learning_rate": 1.872885345443386e-05, + "loss": 1.0074, + "step": 7954 + }, + { + "epoch": 0.19, + "grad_norm": 1.115822878849992, + "learning_rate": 1.872848112192597e-05, + "loss": 1.0103, + "step": 7955 + }, + { + "epoch": 0.19, + "grad_norm": 1.9975001374495929, + "learning_rate": 1.8728108738598103e-05, + "loss": 1.0051, + "step": 7956 + }, + { + "epoch": 0.19, + "grad_norm": 2.0192719854342633, + "learning_rate": 1.872773630445242e-05, + "loss": 1.0792, + "step": 7957 + }, + { + "epoch": 0.19, + "grad_norm": 1.9914836181764555, + "learning_rate": 1.8727363819491097e-05, + "loss": 1.055, + "step": 7958 + }, + { + "epoch": 0.19, + "grad_norm": 2.107014329187812, + "learning_rate": 1.8726991283716295e-05, + "loss": 0.994, + "step": 7959 + }, + { + "epoch": 0.19, + "grad_norm": 1.0931244645163751, + "learning_rate": 1.8726618697130186e-05, + "loss": 0.8803, + "step": 7960 + }, + { + "epoch": 0.19, + "grad_norm": 2.6222123482308435, + "learning_rate": 1.872624605973494e-05, + "loss": 1.0216, + "step": 7961 + }, + { + "epoch": 0.19, + "grad_norm": 1.9777310253503346, + "learning_rate": 1.8725873371532733e-05, + "loss": 1.0373, + "step": 7962 + }, + { + "epoch": 0.19, + "grad_norm": 1.8940358907966113, + "learning_rate": 1.872550063252572e-05, + "loss": 0.9932, + "step": 7963 + }, + { + "epoch": 0.19, + "grad_norm": 3.0799936058903388, + "learning_rate": 1.8725127842716086e-05, + "loss": 1.1142, + "step": 7964 + }, + { + "epoch": 0.19, + "grad_norm": 2.2556485486287747, + "learning_rate": 1.8724755002105996e-05, + "loss": 1.065, + "step": 7965 + }, + { + "epoch": 0.19, + "grad_norm": 2.1462503017392938, + "learning_rate": 1.8724382110697615e-05, + "loss": 1.1774, + "step": 7966 + }, + { + "epoch": 0.19, + "grad_norm": 2.1191796679350072, + "learning_rate": 1.8724009168493126e-05, + "loss": 1.0062, + "step": 7967 + }, + { + "epoch": 0.19, + "grad_norm": 2.160410157385394, + "learning_rate": 1.872363617549469e-05, + "loss": 1.0276, + "step": 7968 + }, + { + "epoch": 0.19, + "grad_norm": 1.1105439154524788, + "learning_rate": 1.872326313170448e-05, + "loss": 0.9987, + "step": 7969 + }, + { + "epoch": 0.19, + "grad_norm": 2.274440600308764, + "learning_rate": 1.8722890037124674e-05, + "loss": 1.1608, + "step": 7970 + }, + { + "epoch": 0.19, + "grad_norm": 2.0810277530747676, + "learning_rate": 1.8722516891757435e-05, + "loss": 1.0596, + "step": 7971 + }, + { + "epoch": 0.19, + "grad_norm": 2.184761955861592, + "learning_rate": 1.872214369560495e-05, + "loss": 1.0464, + "step": 7972 + }, + { + "epoch": 0.19, + "grad_norm": 2.5444232768982644, + "learning_rate": 1.8721770448669377e-05, + "loss": 1.0818, + "step": 7973 + }, + { + "epoch": 0.19, + "grad_norm": 2.1737305218899543, + "learning_rate": 1.8721397150952894e-05, + "loss": 1.0434, + "step": 7974 + }, + { + "epoch": 0.19, + "grad_norm": 2.1070538403066843, + "learning_rate": 1.8721023802457678e-05, + "loss": 1.0849, + "step": 7975 + }, + { + "epoch": 0.19, + "grad_norm": 1.9061667681135124, + "learning_rate": 1.87206504031859e-05, + "loss": 1.0637, + "step": 7976 + }, + { + "epoch": 0.19, + "grad_norm": 2.0099373602501354, + "learning_rate": 1.8720276953139737e-05, + "loss": 0.997, + "step": 7977 + }, + { + "epoch": 0.19, + "grad_norm": 2.3340588649530436, + "learning_rate": 1.871990345232136e-05, + "loss": 1.0687, + "step": 7978 + }, + { + "epoch": 0.19, + "grad_norm": 2.310073682356384, + "learning_rate": 1.871952990073294e-05, + "loss": 1.0934, + "step": 7979 + }, + { + "epoch": 0.19, + "grad_norm": 2.0150039411660003, + "learning_rate": 1.871915629837666e-05, + "loss": 1.0163, + "step": 7980 + }, + { + "epoch": 0.19, + "grad_norm": 2.340199002437315, + "learning_rate": 1.8718782645254688e-05, + "loss": 0.9814, + "step": 7981 + }, + { + "epoch": 0.19, + "grad_norm": 1.9620239360911784, + "learning_rate": 1.8718408941369208e-05, + "loss": 1.1064, + "step": 7982 + }, + { + "epoch": 0.19, + "grad_norm": 2.252810758652069, + "learning_rate": 1.871803518672239e-05, + "loss": 1.0626, + "step": 7983 + }, + { + "epoch": 0.19, + "grad_norm": 2.2531844050086516, + "learning_rate": 1.871766138131641e-05, + "loss": 0.9072, + "step": 7984 + }, + { + "epoch": 0.19, + "grad_norm": 2.018560422341072, + "learning_rate": 1.8717287525153443e-05, + "loss": 1.164, + "step": 7985 + }, + { + "epoch": 0.19, + "grad_norm": 1.793239265336225, + "learning_rate": 1.8716913618235676e-05, + "loss": 0.9507, + "step": 7986 + }, + { + "epoch": 0.19, + "grad_norm": 1.9965058064014856, + "learning_rate": 1.871653966056527e-05, + "loss": 1.1637, + "step": 7987 + }, + { + "epoch": 0.19, + "grad_norm": 2.122794855551771, + "learning_rate": 1.8716165652144415e-05, + "loss": 0.9208, + "step": 7988 + }, + { + "epoch": 0.19, + "grad_norm": 2.0249412247498846, + "learning_rate": 1.871579159297528e-05, + "loss": 1.0553, + "step": 7989 + }, + { + "epoch": 0.19, + "grad_norm": 2.082329688078216, + "learning_rate": 1.871541748306005e-05, + "loss": 1.1134, + "step": 7990 + }, + { + "epoch": 0.19, + "grad_norm": 2.074262656943277, + "learning_rate": 1.87150433224009e-05, + "loss": 1.0199, + "step": 7991 + }, + { + "epoch": 0.19, + "grad_norm": 2.129355978709704, + "learning_rate": 1.871466911100001e-05, + "loss": 1.0341, + "step": 7992 + }, + { + "epoch": 0.19, + "grad_norm": 1.9201815291091482, + "learning_rate": 1.8714294848859555e-05, + "loss": 1.0307, + "step": 7993 + }, + { + "epoch": 0.19, + "grad_norm": 1.1093470557733054, + "learning_rate": 1.8713920535981716e-05, + "loss": 0.986, + "step": 7994 + }, + { + "epoch": 0.19, + "grad_norm": 1.1843918593604528, + "learning_rate": 1.8713546172368676e-05, + "loss": 0.9557, + "step": 7995 + }, + { + "epoch": 0.19, + "grad_norm": 2.0632255784520344, + "learning_rate": 1.8713171758022608e-05, + "loss": 1.0448, + "step": 7996 + }, + { + "epoch": 0.19, + "grad_norm": 3.230267928069936, + "learning_rate": 1.8712797292945698e-05, + "loss": 1.1439, + "step": 7997 + }, + { + "epoch": 0.19, + "grad_norm": 1.7714176771326249, + "learning_rate": 1.8712422777140123e-05, + "loss": 0.9979, + "step": 7998 + }, + { + "epoch": 0.19, + "grad_norm": 1.9352218971486959, + "learning_rate": 1.8712048210608062e-05, + "loss": 1.0429, + "step": 7999 + }, + { + "epoch": 0.19, + "grad_norm": 2.1521598481842443, + "learning_rate": 1.87116735933517e-05, + "loss": 0.9688, + "step": 8000 + }, + { + "epoch": 0.19, + "grad_norm": 1.8920409999818764, + "learning_rate": 1.8711298925373218e-05, + "loss": 0.858, + "step": 8001 + }, + { + "epoch": 0.19, + "grad_norm": 2.2278081614520238, + "learning_rate": 1.8710924206674795e-05, + "loss": 0.9776, + "step": 8002 + }, + { + "epoch": 0.19, + "grad_norm": 2.1488517975699457, + "learning_rate": 1.871054943725861e-05, + "loss": 1.1012, + "step": 8003 + }, + { + "epoch": 0.19, + "grad_norm": 2.1169155823697006, + "learning_rate": 1.871017461712685e-05, + "loss": 1.1417, + "step": 8004 + }, + { + "epoch": 0.19, + "grad_norm": 2.002581230928964, + "learning_rate": 1.87097997462817e-05, + "loss": 1.0112, + "step": 8005 + }, + { + "epoch": 0.19, + "grad_norm": 1.9401169471138417, + "learning_rate": 1.8709424824725335e-05, + "loss": 1.0782, + "step": 8006 + }, + { + "epoch": 0.19, + "grad_norm": 1.9848654232037048, + "learning_rate": 1.8709049852459942e-05, + "loss": 1.0624, + "step": 8007 + }, + { + "epoch": 0.19, + "grad_norm": 1.9442640669207465, + "learning_rate": 1.87086748294877e-05, + "loss": 0.9733, + "step": 8008 + }, + { + "epoch": 0.19, + "grad_norm": 1.9332342926329433, + "learning_rate": 1.87082997558108e-05, + "loss": 1.0681, + "step": 8009 + }, + { + "epoch": 0.19, + "grad_norm": 1.9030689703050832, + "learning_rate": 1.8707924631431424e-05, + "loss": 1.1023, + "step": 8010 + }, + { + "epoch": 0.19, + "grad_norm": 1.9384734485226467, + "learning_rate": 1.870754945635175e-05, + "loss": 1.0471, + "step": 8011 + }, + { + "epoch": 0.19, + "grad_norm": 2.4987643196242257, + "learning_rate": 1.8707174230573967e-05, + "loss": 1.1261, + "step": 8012 + }, + { + "epoch": 0.19, + "grad_norm": 2.1667198162120247, + "learning_rate": 1.870679895410026e-05, + "loss": 1.053, + "step": 8013 + }, + { + "epoch": 0.19, + "grad_norm": 2.038392355110637, + "learning_rate": 1.870642362693281e-05, + "loss": 1.0518, + "step": 8014 + }, + { + "epoch": 0.19, + "grad_norm": 1.955509302946216, + "learning_rate": 1.870604824907381e-05, + "loss": 1.1067, + "step": 8015 + }, + { + "epoch": 0.19, + "grad_norm": 2.0865569058974307, + "learning_rate": 1.8705672820525437e-05, + "loss": 1.0319, + "step": 8016 + }, + { + "epoch": 0.19, + "grad_norm": 1.920506415414627, + "learning_rate": 1.870529734128988e-05, + "loss": 1.119, + "step": 8017 + }, + { + "epoch": 0.19, + "grad_norm": 2.4655173818127345, + "learning_rate": 1.8704921811369328e-05, + "loss": 1.12, + "step": 8018 + }, + { + "epoch": 0.19, + "grad_norm": 1.1492833157949782, + "learning_rate": 1.8704546230765964e-05, + "loss": 1.0448, + "step": 8019 + }, + { + "epoch": 0.19, + "grad_norm": 1.8465564159369985, + "learning_rate": 1.870417059948198e-05, + "loss": 1.0275, + "step": 8020 + }, + { + "epoch": 0.19, + "grad_norm": 2.132085243226704, + "learning_rate": 1.8703794917519555e-05, + "loss": 1.0309, + "step": 8021 + }, + { + "epoch": 0.19, + "grad_norm": 2.0090014887707954, + "learning_rate": 1.8703419184880882e-05, + "loss": 1.0372, + "step": 8022 + }, + { + "epoch": 0.19, + "grad_norm": 2.2346488479086624, + "learning_rate": 1.8703043401568144e-05, + "loss": 1.2358, + "step": 8023 + }, + { + "epoch": 0.19, + "grad_norm": 2.059114471638633, + "learning_rate": 1.8702667567583533e-05, + "loss": 1.1745, + "step": 8024 + }, + { + "epoch": 0.19, + "grad_norm": 2.225636028951763, + "learning_rate": 1.870229168292924e-05, + "loss": 0.9876, + "step": 8025 + }, + { + "epoch": 0.19, + "grad_norm": 2.1024982295166357, + "learning_rate": 1.8701915747607446e-05, + "loss": 0.9953, + "step": 8026 + }, + { + "epoch": 0.19, + "grad_norm": 2.216784387187822, + "learning_rate": 1.8701539761620348e-05, + "loss": 1.0254, + "step": 8027 + }, + { + "epoch": 0.19, + "grad_norm": 1.9451752138096292, + "learning_rate": 1.8701163724970125e-05, + "loss": 0.9802, + "step": 8028 + }, + { + "epoch": 0.19, + "grad_norm": 2.463420997237347, + "learning_rate": 1.8700787637658976e-05, + "loss": 1.0023, + "step": 8029 + }, + { + "epoch": 0.19, + "grad_norm": 1.8784985154416454, + "learning_rate": 1.8700411499689088e-05, + "loss": 1.1497, + "step": 8030 + }, + { + "epoch": 0.19, + "grad_norm": 1.9439707979112484, + "learning_rate": 1.8700035311062644e-05, + "loss": 1.0496, + "step": 8031 + }, + { + "epoch": 0.19, + "grad_norm": 1.9745911343615896, + "learning_rate": 1.8699659071781845e-05, + "loss": 1.1228, + "step": 8032 + }, + { + "epoch": 0.19, + "grad_norm": 2.471393001404608, + "learning_rate": 1.8699282781848873e-05, + "loss": 1.0072, + "step": 8033 + }, + { + "epoch": 0.19, + "grad_norm": 1.1229549876230016, + "learning_rate": 1.8698906441265925e-05, + "loss": 1.012, + "step": 8034 + }, + { + "epoch": 0.19, + "grad_norm": 2.1383519203325356, + "learning_rate": 1.8698530050035188e-05, + "loss": 1.1197, + "step": 8035 + }, + { + "epoch": 0.19, + "grad_norm": 2.3806958521827415, + "learning_rate": 1.8698153608158856e-05, + "loss": 1.0619, + "step": 8036 + }, + { + "epoch": 0.19, + "grad_norm": 2.0739300694434553, + "learning_rate": 1.869777711563912e-05, + "loss": 1.0823, + "step": 8037 + }, + { + "epoch": 0.19, + "grad_norm": 2.152842666238917, + "learning_rate": 1.8697400572478173e-05, + "loss": 1.1478, + "step": 8038 + }, + { + "epoch": 0.19, + "grad_norm": 1.9717042464681827, + "learning_rate": 1.8697023978678205e-05, + "loss": 1.1271, + "step": 8039 + }, + { + "epoch": 0.19, + "grad_norm": 2.083329880036035, + "learning_rate": 1.869664733424141e-05, + "loss": 1.1524, + "step": 8040 + }, + { + "epoch": 0.19, + "grad_norm": 1.905598022796085, + "learning_rate": 1.869627063916998e-05, + "loss": 1.1331, + "step": 8041 + }, + { + "epoch": 0.19, + "grad_norm": 1.188898034735103, + "learning_rate": 1.869589389346611e-05, + "loss": 0.9986, + "step": 8042 + }, + { + "epoch": 0.19, + "grad_norm": 2.1905686823971524, + "learning_rate": 1.8695517097131993e-05, + "loss": 1.1428, + "step": 8043 + }, + { + "epoch": 0.19, + "grad_norm": 2.1318038901105867, + "learning_rate": 1.8695140250169825e-05, + "loss": 1.0974, + "step": 8044 + }, + { + "epoch": 0.19, + "grad_norm": 1.9695547748363316, + "learning_rate": 1.8694763352581792e-05, + "loss": 1.1984, + "step": 8045 + }, + { + "epoch": 0.19, + "grad_norm": 2.1707047954396383, + "learning_rate": 1.86943864043701e-05, + "loss": 1.1571, + "step": 8046 + }, + { + "epoch": 0.19, + "grad_norm": 2.2442902681318393, + "learning_rate": 1.8694009405536934e-05, + "loss": 1.019, + "step": 8047 + }, + { + "epoch": 0.19, + "grad_norm": 1.975737921082055, + "learning_rate": 1.8693632356084493e-05, + "loss": 1.1473, + "step": 8048 + }, + { + "epoch": 0.19, + "grad_norm": 2.5770590862621865, + "learning_rate": 1.8693255256014976e-05, + "loss": 1.1294, + "step": 8049 + }, + { + "epoch": 0.19, + "grad_norm": 1.9611718107979395, + "learning_rate": 1.8692878105330568e-05, + "loss": 1.184, + "step": 8050 + }, + { + "epoch": 0.19, + "grad_norm": 2.217532493146344, + "learning_rate": 1.8692500904033478e-05, + "loss": 1.0253, + "step": 8051 + }, + { + "epoch": 0.19, + "grad_norm": 1.0903199981455955, + "learning_rate": 1.869212365212589e-05, + "loss": 1.0035, + "step": 8052 + }, + { + "epoch": 0.19, + "grad_norm": 2.059404894234838, + "learning_rate": 1.8691746349610008e-05, + "loss": 1.1393, + "step": 8053 + }, + { + "epoch": 0.19, + "grad_norm": 2.3108315998791684, + "learning_rate": 1.869136899648803e-05, + "loss": 0.9955, + "step": 8054 + }, + { + "epoch": 0.19, + "grad_norm": 2.1601924379078574, + "learning_rate": 1.8690991592762147e-05, + "loss": 1.0562, + "step": 8055 + }, + { + "epoch": 0.19, + "grad_norm": 2.0243830696876564, + "learning_rate": 1.869061413843456e-05, + "loss": 1.1249, + "step": 8056 + }, + { + "epoch": 0.19, + "grad_norm": 2.492779069977987, + "learning_rate": 1.869023663350747e-05, + "loss": 1.0087, + "step": 8057 + }, + { + "epoch": 0.19, + "grad_norm": 2.038759175550346, + "learning_rate": 1.8689859077983064e-05, + "loss": 1.0181, + "step": 8058 + }, + { + "epoch": 0.19, + "grad_norm": 2.15473181725308, + "learning_rate": 1.868948147186355e-05, + "loss": 1.0637, + "step": 8059 + }, + { + "epoch": 0.19, + "grad_norm": 1.8799588627728203, + "learning_rate": 1.8689103815151127e-05, + "loss": 1.037, + "step": 8060 + }, + { + "epoch": 0.19, + "grad_norm": 1.1225494025154046, + "learning_rate": 1.8688726107847986e-05, + "loss": 0.9929, + "step": 8061 + }, + { + "epoch": 0.19, + "grad_norm": 2.204130740008503, + "learning_rate": 1.8688348349956334e-05, + "loss": 1.0201, + "step": 8062 + }, + { + "epoch": 0.19, + "grad_norm": 2.034678076024832, + "learning_rate": 1.8687970541478367e-05, + "loss": 1.2344, + "step": 8063 + }, + { + "epoch": 0.19, + "grad_norm": 2.084892490925128, + "learning_rate": 1.8687592682416284e-05, + "loss": 1.1625, + "step": 8064 + }, + { + "epoch": 0.19, + "grad_norm": 2.2203440868730144, + "learning_rate": 1.8687214772772283e-05, + "loss": 0.9712, + "step": 8065 + }, + { + "epoch": 0.19, + "grad_norm": 2.4830748960068365, + "learning_rate": 1.8686836812548573e-05, + "loss": 0.9379, + "step": 8066 + }, + { + "epoch": 0.19, + "grad_norm": 1.1630830998950619, + "learning_rate": 1.8686458801747347e-05, + "loss": 0.944, + "step": 8067 + }, + { + "epoch": 0.19, + "grad_norm": 2.1141968300946945, + "learning_rate": 1.8686080740370807e-05, + "loss": 1.128, + "step": 8068 + }, + { + "epoch": 0.19, + "grad_norm": 1.073418140848472, + "learning_rate": 1.8685702628421155e-05, + "loss": 0.9975, + "step": 8069 + }, + { + "epoch": 0.19, + "grad_norm": 1.9749003175011015, + "learning_rate": 1.8685324465900594e-05, + "loss": 1.1933, + "step": 8070 + }, + { + "epoch": 0.19, + "grad_norm": 2.1416098173332894, + "learning_rate": 1.868494625281132e-05, + "loss": 1.0255, + "step": 8071 + }, + { + "epoch": 0.19, + "grad_norm": 1.0524011997692218, + "learning_rate": 1.8684567989155545e-05, + "loss": 0.9192, + "step": 8072 + }, + { + "epoch": 0.19, + "grad_norm": 2.3954898382182033, + "learning_rate": 1.868418967493546e-05, + "loss": 0.9165, + "step": 8073 + }, + { + "epoch": 0.19, + "grad_norm": 1.8878982314499804, + "learning_rate": 1.8683811310153276e-05, + "loss": 1.0506, + "step": 8074 + }, + { + "epoch": 0.19, + "grad_norm": 2.2638788554460514, + "learning_rate": 1.8683432894811195e-05, + "loss": 1.0591, + "step": 8075 + }, + { + "epoch": 0.19, + "grad_norm": 2.3285788085379173, + "learning_rate": 1.8683054428911414e-05, + "loss": 1.1182, + "step": 8076 + }, + { + "epoch": 0.19, + "grad_norm": 1.971852357694899, + "learning_rate": 1.8682675912456147e-05, + "loss": 0.9673, + "step": 8077 + }, + { + "epoch": 0.19, + "grad_norm": 2.0962815356432665, + "learning_rate": 1.868229734544759e-05, + "loss": 1.0088, + "step": 8078 + }, + { + "epoch": 0.19, + "grad_norm": 2.348523514716506, + "learning_rate": 1.8681918727887945e-05, + "loss": 1.1597, + "step": 8079 + }, + { + "epoch": 0.19, + "grad_norm": 1.8253778378622225, + "learning_rate": 1.8681540059779425e-05, + "loss": 1.0225, + "step": 8080 + }, + { + "epoch": 0.19, + "grad_norm": 2.0204155616479595, + "learning_rate": 1.8681161341124227e-05, + "loss": 1.1096, + "step": 8081 + }, + { + "epoch": 0.19, + "grad_norm": 2.1825049773116785, + "learning_rate": 1.8680782571924564e-05, + "loss": 0.9517, + "step": 8082 + }, + { + "epoch": 0.19, + "grad_norm": 1.8827950734750665, + "learning_rate": 1.868040375218263e-05, + "loss": 0.991, + "step": 8083 + }, + { + "epoch": 0.19, + "grad_norm": 2.0756974974718254, + "learning_rate": 1.868002488190064e-05, + "loss": 1.0826, + "step": 8084 + }, + { + "epoch": 0.19, + "grad_norm": 2.24710010975491, + "learning_rate": 1.86796459610808e-05, + "loss": 1.058, + "step": 8085 + }, + { + "epoch": 0.19, + "grad_norm": 1.9436795399860953, + "learning_rate": 1.8679266989725313e-05, + "loss": 1.0162, + "step": 8086 + }, + { + "epoch": 0.19, + "grad_norm": 2.9619651823666486, + "learning_rate": 1.8678887967836382e-05, + "loss": 1.1641, + "step": 8087 + }, + { + "epoch": 0.19, + "grad_norm": 1.723809702579431, + "learning_rate": 1.867850889541622e-05, + "loss": 1.0483, + "step": 8088 + }, + { + "epoch": 0.19, + "grad_norm": 2.1644743655902103, + "learning_rate": 1.867812977246703e-05, + "loss": 1.1034, + "step": 8089 + }, + { + "epoch": 0.19, + "grad_norm": 2.4560521233850525, + "learning_rate": 1.8677750598991023e-05, + "loss": 1.0812, + "step": 8090 + }, + { + "epoch": 0.19, + "grad_norm": 1.169449620445667, + "learning_rate": 1.8677371374990406e-05, + "loss": 0.9637, + "step": 8091 + }, + { + "epoch": 0.19, + "grad_norm": 2.0072195314734453, + "learning_rate": 1.8676992100467383e-05, + "loss": 1.1468, + "step": 8092 + }, + { + "epoch": 0.19, + "grad_norm": 2.360023827836763, + "learning_rate": 1.8676612775424165e-05, + "loss": 1.0653, + "step": 8093 + }, + { + "epoch": 0.19, + "grad_norm": 2.643597271780243, + "learning_rate": 1.8676233399862963e-05, + "loss": 1.1303, + "step": 8094 + }, + { + "epoch": 0.19, + "grad_norm": 1.099745419849922, + "learning_rate": 1.8675853973785976e-05, + "loss": 0.9381, + "step": 8095 + }, + { + "epoch": 0.19, + "grad_norm": 2.574126115660402, + "learning_rate": 1.8675474497195428e-05, + "loss": 1.0188, + "step": 8096 + }, + { + "epoch": 0.19, + "grad_norm": 2.7190414087900536, + "learning_rate": 1.867509497009352e-05, + "loss": 0.9469, + "step": 8097 + }, + { + "epoch": 0.19, + "grad_norm": 1.0931435281971165, + "learning_rate": 1.867471539248246e-05, + "loss": 0.9611, + "step": 8098 + }, + { + "epoch": 0.19, + "grad_norm": 2.140492392743284, + "learning_rate": 1.8674335764364464e-05, + "loss": 1.1149, + "step": 8099 + }, + { + "epoch": 0.19, + "grad_norm": 2.3828742135568888, + "learning_rate": 1.8673956085741734e-05, + "loss": 1.1609, + "step": 8100 + }, + { + "epoch": 0.19, + "grad_norm": 1.9972357142816466, + "learning_rate": 1.8673576356616488e-05, + "loss": 1.0441, + "step": 8101 + }, + { + "epoch": 0.19, + "grad_norm": 2.0729771903811387, + "learning_rate": 1.8673196576990938e-05, + "loss": 1.0378, + "step": 8102 + }, + { + "epoch": 0.19, + "grad_norm": 2.164889071208938, + "learning_rate": 1.8672816746867285e-05, + "loss": 1.0102, + "step": 8103 + }, + { + "epoch": 0.19, + "grad_norm": 2.1376076082250384, + "learning_rate": 1.867243686624775e-05, + "loss": 1.0273, + "step": 8104 + }, + { + "epoch": 0.19, + "grad_norm": 2.5691182684284524, + "learning_rate": 1.8672056935134545e-05, + "loss": 1.06, + "step": 8105 + }, + { + "epoch": 0.19, + "grad_norm": 2.0333881703928154, + "learning_rate": 1.8671676953529875e-05, + "loss": 1.0769, + "step": 8106 + }, + { + "epoch": 0.19, + "grad_norm": 2.8292217804362463, + "learning_rate": 1.8671296921435956e-05, + "loss": 1.0675, + "step": 8107 + }, + { + "epoch": 0.19, + "grad_norm": 2.435601935399046, + "learning_rate": 1.8670916838855005e-05, + "loss": 0.8781, + "step": 8108 + }, + { + "epoch": 0.19, + "grad_norm": 1.1307991246288518, + "learning_rate": 1.867053670578923e-05, + "loss": 1.029, + "step": 8109 + }, + { + "epoch": 0.19, + "grad_norm": 1.984333485986656, + "learning_rate": 1.8670156522240843e-05, + "loss": 1.1722, + "step": 8110 + }, + { + "epoch": 0.19, + "grad_norm": 2.2095261920323175, + "learning_rate": 1.866977628821206e-05, + "loss": 1.008, + "step": 8111 + }, + { + "epoch": 0.19, + "grad_norm": 2.352956217450554, + "learning_rate": 1.8669396003705095e-05, + "loss": 1.0628, + "step": 8112 + }, + { + "epoch": 0.19, + "grad_norm": 2.2431699299559127, + "learning_rate": 1.8669015668722162e-05, + "loss": 1.0098, + "step": 8113 + }, + { + "epoch": 0.19, + "grad_norm": 2.36298185941709, + "learning_rate": 1.8668635283265475e-05, + "loss": 0.9933, + "step": 8114 + }, + { + "epoch": 0.19, + "grad_norm": 2.305889157630052, + "learning_rate": 1.8668254847337252e-05, + "loss": 1.1349, + "step": 8115 + }, + { + "epoch": 0.19, + "grad_norm": 1.1862852324474478, + "learning_rate": 1.86678743609397e-05, + "loss": 0.9685, + "step": 8116 + }, + { + "epoch": 0.19, + "grad_norm": 2.362225795357335, + "learning_rate": 1.8667493824075042e-05, + "loss": 0.9502, + "step": 8117 + }, + { + "epoch": 0.19, + "grad_norm": 1.1173456753615632, + "learning_rate": 1.866711323674549e-05, + "loss": 1.0159, + "step": 8118 + }, + { + "epoch": 0.19, + "grad_norm": 2.3181692507829843, + "learning_rate": 1.8666732598953263e-05, + "loss": 1.0305, + "step": 8119 + }, + { + "epoch": 0.19, + "grad_norm": 2.4804424874912883, + "learning_rate": 1.8666351910700572e-05, + "loss": 1.0997, + "step": 8120 + }, + { + "epoch": 0.19, + "grad_norm": 2.0973661581100425, + "learning_rate": 1.8665971171989638e-05, + "loss": 1.003, + "step": 8121 + }, + { + "epoch": 0.19, + "grad_norm": 1.961307166052522, + "learning_rate": 1.8665590382822675e-05, + "loss": 0.9334, + "step": 8122 + }, + { + "epoch": 0.19, + "grad_norm": 2.0237332075854573, + "learning_rate": 1.8665209543201904e-05, + "loss": 1.1324, + "step": 8123 + }, + { + "epoch": 0.19, + "grad_norm": 3.755683276000192, + "learning_rate": 1.866482865312954e-05, + "loss": 0.9969, + "step": 8124 + }, + { + "epoch": 0.19, + "grad_norm": 2.5152370010391296, + "learning_rate": 1.8664447712607798e-05, + "loss": 1.1063, + "step": 8125 + }, + { + "epoch": 0.19, + "grad_norm": 1.9405959151245378, + "learning_rate": 1.8664066721638895e-05, + "loss": 1.0055, + "step": 8126 + }, + { + "epoch": 0.19, + "grad_norm": 2.053398438003987, + "learning_rate": 1.8663685680225056e-05, + "loss": 1.1176, + "step": 8127 + }, + { + "epoch": 0.19, + "grad_norm": 1.9952663923899923, + "learning_rate": 1.8663304588368495e-05, + "loss": 1.2009, + "step": 8128 + }, + { + "epoch": 0.19, + "grad_norm": 2.097756343678712, + "learning_rate": 1.8662923446071436e-05, + "loss": 1.1102, + "step": 8129 + }, + { + "epoch": 0.19, + "grad_norm": 2.228705888695147, + "learning_rate": 1.866254225333609e-05, + "loss": 0.9308, + "step": 8130 + }, + { + "epoch": 0.19, + "grad_norm": 2.0619163649216348, + "learning_rate": 1.8662161010164682e-05, + "loss": 1.1397, + "step": 8131 + }, + { + "epoch": 0.19, + "grad_norm": 1.96265362914904, + "learning_rate": 1.8661779716559428e-05, + "loss": 1.077, + "step": 8132 + }, + { + "epoch": 0.19, + "grad_norm": 1.948088324119047, + "learning_rate": 1.866139837252255e-05, + "loss": 1.1253, + "step": 8133 + }, + { + "epoch": 0.19, + "grad_norm": 2.1009532358176606, + "learning_rate": 1.8661016978056273e-05, + "loss": 1.0391, + "step": 8134 + }, + { + "epoch": 0.19, + "grad_norm": 1.7939801469721555, + "learning_rate": 1.8660635533162813e-05, + "loss": 1.1166, + "step": 8135 + }, + { + "epoch": 0.19, + "grad_norm": 2.2682988830930024, + "learning_rate": 1.866025403784439e-05, + "loss": 1.0665, + "step": 8136 + }, + { + "epoch": 0.19, + "grad_norm": 1.9562698099318565, + "learning_rate": 1.8659872492103222e-05, + "loss": 1.0161, + "step": 8137 + }, + { + "epoch": 0.19, + "grad_norm": 1.9568472132551638, + "learning_rate": 1.865949089594154e-05, + "loss": 0.969, + "step": 8138 + }, + { + "epoch": 0.19, + "grad_norm": 2.1946426399234773, + "learning_rate": 1.8659109249361557e-05, + "loss": 1.1067, + "step": 8139 + }, + { + "epoch": 0.19, + "grad_norm": 2.333387362566515, + "learning_rate": 1.8658727552365498e-05, + "loss": 1.2434, + "step": 8140 + }, + { + "epoch": 0.19, + "grad_norm": 2.3369437424854884, + "learning_rate": 1.865834580495559e-05, + "loss": 0.9746, + "step": 8141 + }, + { + "epoch": 0.19, + "grad_norm": 2.0378758299789386, + "learning_rate": 1.865796400713405e-05, + "loss": 0.9856, + "step": 8142 + }, + { + "epoch": 0.19, + "grad_norm": 2.130743180034386, + "learning_rate": 1.86575821589031e-05, + "loss": 1.1304, + "step": 8143 + }, + { + "epoch": 0.19, + "grad_norm": 2.042457785629761, + "learning_rate": 1.8657200260264966e-05, + "loss": 1.0947, + "step": 8144 + }, + { + "epoch": 0.19, + "grad_norm": 2.047736474720268, + "learning_rate": 1.865681831122187e-05, + "loss": 0.97, + "step": 8145 + }, + { + "epoch": 0.19, + "grad_norm": 1.9874294459761341, + "learning_rate": 1.865643631177604e-05, + "loss": 0.9263, + "step": 8146 + }, + { + "epoch": 0.19, + "grad_norm": 3.0481809197041443, + "learning_rate": 1.86560542619297e-05, + "loss": 1.0879, + "step": 8147 + }, + { + "epoch": 0.19, + "grad_norm": 2.0136740952284344, + "learning_rate": 1.865567216168507e-05, + "loss": 1.1086, + "step": 8148 + }, + { + "epoch": 0.19, + "grad_norm": 2.439669167524681, + "learning_rate": 1.865529001104437e-05, + "loss": 1.0064, + "step": 8149 + }, + { + "epoch": 0.19, + "grad_norm": 2.3123796503583893, + "learning_rate": 1.8654907810009838e-05, + "loss": 1.1272, + "step": 8150 + }, + { + "epoch": 0.19, + "grad_norm": 1.9323394195320034, + "learning_rate": 1.865452555858369e-05, + "loss": 1.1072, + "step": 8151 + }, + { + "epoch": 0.19, + "grad_norm": 1.849164152386948, + "learning_rate": 1.8654143256768157e-05, + "loss": 0.9571, + "step": 8152 + }, + { + "epoch": 0.19, + "grad_norm": 2.22704244405921, + "learning_rate": 1.865376090456546e-05, + "loss": 1.0057, + "step": 8153 + }, + { + "epoch": 0.19, + "grad_norm": 1.069863974826253, + "learning_rate": 1.8653378501977827e-05, + "loss": 1.0057, + "step": 8154 + }, + { + "epoch": 0.19, + "grad_norm": 1.962460047791081, + "learning_rate": 1.8652996049007485e-05, + "loss": 1.0758, + "step": 8155 + }, + { + "epoch": 0.19, + "grad_norm": 2.2869587172876518, + "learning_rate": 1.865261354565666e-05, + "loss": 0.909, + "step": 8156 + }, + { + "epoch": 0.19, + "grad_norm": 2.1648132076719393, + "learning_rate": 1.8652230991927582e-05, + "loss": 1.1664, + "step": 8157 + }, + { + "epoch": 0.19, + "grad_norm": 2.454902126043711, + "learning_rate": 1.8651848387822473e-05, + "loss": 1.0358, + "step": 8158 + }, + { + "epoch": 0.19, + "grad_norm": 2.3488009080391627, + "learning_rate": 1.8651465733343562e-05, + "loss": 1.0579, + "step": 8159 + }, + { + "epoch": 0.19, + "grad_norm": 2.2177005194800636, + "learning_rate": 1.8651083028493082e-05, + "loss": 0.9814, + "step": 8160 + }, + { + "epoch": 0.19, + "grad_norm": 2.038013681338183, + "learning_rate": 1.8650700273273257e-05, + "loss": 1.0322, + "step": 8161 + }, + { + "epoch": 0.19, + "grad_norm": 1.856108180592584, + "learning_rate": 1.8650317467686318e-05, + "loss": 1.126, + "step": 8162 + }, + { + "epoch": 0.19, + "grad_norm": 2.3581911150906256, + "learning_rate": 1.864993461173449e-05, + "loss": 0.972, + "step": 8163 + }, + { + "epoch": 0.19, + "grad_norm": 2.375910978846741, + "learning_rate": 1.8649551705420004e-05, + "loss": 1.1177, + "step": 8164 + }, + { + "epoch": 0.19, + "grad_norm": 2.114006464952845, + "learning_rate": 1.864916874874509e-05, + "loss": 1.1768, + "step": 8165 + }, + { + "epoch": 0.19, + "grad_norm": 2.1223163598145405, + "learning_rate": 1.8648785741711975e-05, + "loss": 1.041, + "step": 8166 + }, + { + "epoch": 0.19, + "grad_norm": 2.388461421222859, + "learning_rate": 1.8648402684322893e-05, + "loss": 1.0396, + "step": 8167 + }, + { + "epoch": 0.19, + "grad_norm": 2.088819005993685, + "learning_rate": 1.8648019576580074e-05, + "loss": 1.0876, + "step": 8168 + }, + { + "epoch": 0.19, + "grad_norm": 2.5738680353528154, + "learning_rate": 1.8647636418485748e-05, + "loss": 1.1017, + "step": 8169 + }, + { + "epoch": 0.19, + "grad_norm": 2.5216483267654892, + "learning_rate": 1.8647253210042143e-05, + "loss": 0.9762, + "step": 8170 + }, + { + "epoch": 0.19, + "grad_norm": 2.5407908239038215, + "learning_rate": 1.8646869951251493e-05, + "loss": 0.9938, + "step": 8171 + }, + { + "epoch": 0.19, + "grad_norm": 2.073587566946365, + "learning_rate": 1.8646486642116026e-05, + "loss": 1.1035, + "step": 8172 + }, + { + "epoch": 0.19, + "grad_norm": 2.1660689458312232, + "learning_rate": 1.864610328263798e-05, + "loss": 1.0584, + "step": 8173 + }, + { + "epoch": 0.19, + "grad_norm": 2.096533275557337, + "learning_rate": 1.864571987281958e-05, + "loss": 0.9827, + "step": 8174 + }, + { + "epoch": 0.19, + "grad_norm": 2.235252531550695, + "learning_rate": 1.8645336412663062e-05, + "loss": 1.05, + "step": 8175 + }, + { + "epoch": 0.19, + "grad_norm": 2.0993597243727247, + "learning_rate": 1.864495290217066e-05, + "loss": 1.0941, + "step": 8176 + }, + { + "epoch": 0.19, + "grad_norm": 2.5770310015228133, + "learning_rate": 1.8644569341344604e-05, + "loss": 1.0901, + "step": 8177 + }, + { + "epoch": 0.19, + "grad_norm": 2.643768933625911, + "learning_rate": 1.864418573018713e-05, + "loss": 0.9946, + "step": 8178 + }, + { + "epoch": 0.19, + "grad_norm": 2.8535585125246614, + "learning_rate": 1.864380206870047e-05, + "loss": 0.989, + "step": 8179 + }, + { + "epoch": 0.19, + "grad_norm": 2.0963153166594317, + "learning_rate": 1.8643418356886854e-05, + "loss": 1.03, + "step": 8180 + }, + { + "epoch": 0.19, + "grad_norm": 1.9891549949615557, + "learning_rate": 1.8643034594748525e-05, + "loss": 1.1071, + "step": 8181 + }, + { + "epoch": 0.19, + "grad_norm": 2.084286463743943, + "learning_rate": 1.864265078228771e-05, + "loss": 1.007, + "step": 8182 + }, + { + "epoch": 0.19, + "grad_norm": 1.9497111136787564, + "learning_rate": 1.8642266919506644e-05, + "loss": 1.0006, + "step": 8183 + }, + { + "epoch": 0.19, + "grad_norm": 1.891626557021076, + "learning_rate": 1.864188300640757e-05, + "loss": 0.9672, + "step": 8184 + }, + { + "epoch": 0.19, + "grad_norm": 2.501763614137361, + "learning_rate": 1.864149904299271e-05, + "loss": 1.0344, + "step": 8185 + }, + { + "epoch": 0.19, + "grad_norm": 2.1074785411859804, + "learning_rate": 1.8641115029264312e-05, + "loss": 1.1222, + "step": 8186 + }, + { + "epoch": 0.19, + "grad_norm": 2.086790541397041, + "learning_rate": 1.8640730965224602e-05, + "loss": 1.0758, + "step": 8187 + }, + { + "epoch": 0.19, + "grad_norm": 1.9914947039099251, + "learning_rate": 1.8640346850875825e-05, + "loss": 1.021, + "step": 8188 + }, + { + "epoch": 0.19, + "grad_norm": 2.10077430622252, + "learning_rate": 1.863996268622021e-05, + "loss": 1.0512, + "step": 8189 + }, + { + "epoch": 0.19, + "grad_norm": 2.1236831944998285, + "learning_rate": 1.8639578471259997e-05, + "loss": 1.0279, + "step": 8190 + }, + { + "epoch": 0.19, + "grad_norm": 2.228709680916102, + "learning_rate": 1.8639194205997423e-05, + "loss": 1.0795, + "step": 8191 + }, + { + "epoch": 0.19, + "grad_norm": 1.9880296010640726, + "learning_rate": 1.863880989043473e-05, + "loss": 1.0007, + "step": 8192 + }, + { + "epoch": 0.19, + "grad_norm": 2.4621048637365752, + "learning_rate": 1.8638425524574148e-05, + "loss": 1.0807, + "step": 8193 + }, + { + "epoch": 0.19, + "grad_norm": 2.0410590780916604, + "learning_rate": 1.8638041108417916e-05, + "loss": 1.0687, + "step": 8194 + }, + { + "epoch": 0.19, + "grad_norm": 1.8470524840791844, + "learning_rate": 1.8637656641968272e-05, + "loss": 1.0605, + "step": 8195 + }, + { + "epoch": 0.19, + "grad_norm": 2.2134299593208238, + "learning_rate": 1.863727212522746e-05, + "loss": 1.0726, + "step": 8196 + }, + { + "epoch": 0.19, + "grad_norm": 1.8099052526579944, + "learning_rate": 1.8636887558197713e-05, + "loss": 0.972, + "step": 8197 + }, + { + "epoch": 0.19, + "grad_norm": 2.0610586269879567, + "learning_rate": 1.8636502940881276e-05, + "loss": 1.1444, + "step": 8198 + }, + { + "epoch": 0.19, + "grad_norm": 1.962085889974387, + "learning_rate": 1.863611827328038e-05, + "loss": 1.1286, + "step": 8199 + }, + { + "epoch": 0.19, + "grad_norm": 2.1924599283500057, + "learning_rate": 1.8635733555397273e-05, + "loss": 1.1374, + "step": 8200 + }, + { + "epoch": 0.19, + "grad_norm": 2.603882768911672, + "learning_rate": 1.863534878723419e-05, + "loss": 1.1015, + "step": 8201 + }, + { + "epoch": 0.19, + "grad_norm": 2.3559794765846087, + "learning_rate": 1.863496396879337e-05, + "loss": 1.1048, + "step": 8202 + }, + { + "epoch": 0.19, + "grad_norm": 3.5191168568924565, + "learning_rate": 1.863457910007706e-05, + "loss": 1.027, + "step": 8203 + }, + { + "epoch": 0.19, + "grad_norm": 2.4339787391471233, + "learning_rate": 1.8634194181087494e-05, + "loss": 1.1164, + "step": 8204 + }, + { + "epoch": 0.19, + "grad_norm": 3.536389230714079, + "learning_rate": 1.8633809211826917e-05, + "loss": 0.9905, + "step": 8205 + }, + { + "epoch": 0.19, + "grad_norm": 2.0699492018811263, + "learning_rate": 1.8633424192297566e-05, + "loss": 1.1198, + "step": 8206 + }, + { + "epoch": 0.19, + "grad_norm": 2.066516699976013, + "learning_rate": 1.8633039122501692e-05, + "loss": 1.0875, + "step": 8207 + }, + { + "epoch": 0.19, + "grad_norm": 1.2221919143897453, + "learning_rate": 1.8632654002441528e-05, + "loss": 0.944, + "step": 8208 + }, + { + "epoch": 0.19, + "grad_norm": 2.1864572260927693, + "learning_rate": 1.863226883211932e-05, + "loss": 1.1439, + "step": 8209 + }, + { + "epoch": 0.19, + "grad_norm": 2.111012035496844, + "learning_rate": 1.863188361153731e-05, + "loss": 0.9515, + "step": 8210 + }, + { + "epoch": 0.19, + "grad_norm": 1.2113899673863153, + "learning_rate": 1.8631498340697736e-05, + "loss": 0.968, + "step": 8211 + }, + { + "epoch": 0.19, + "grad_norm": 2.7024713245945873, + "learning_rate": 1.8631113019602854e-05, + "loss": 1.1148, + "step": 8212 + }, + { + "epoch": 0.19, + "grad_norm": 2.035002821288896, + "learning_rate": 1.8630727648254896e-05, + "loss": 1.0401, + "step": 8213 + }, + { + "epoch": 0.19, + "grad_norm": 2.33521230091308, + "learning_rate": 1.8630342226656108e-05, + "loss": 0.9663, + "step": 8214 + }, + { + "epoch": 0.19, + "grad_norm": 1.1007401861559907, + "learning_rate": 1.8629956754808734e-05, + "loss": 0.9416, + "step": 8215 + }, + { + "epoch": 0.19, + "grad_norm": 2.612002540254135, + "learning_rate": 1.8629571232715024e-05, + "loss": 1.1402, + "step": 8216 + }, + { + "epoch": 0.19, + "grad_norm": 2.6973342723625513, + "learning_rate": 1.8629185660377216e-05, + "loss": 1.0683, + "step": 8217 + }, + { + "epoch": 0.19, + "grad_norm": 2.346579320259819, + "learning_rate": 1.8628800037797558e-05, + "loss": 1.0488, + "step": 8218 + }, + { + "epoch": 0.19, + "grad_norm": 1.8038094152387045, + "learning_rate": 1.8628414364978292e-05, + "loss": 1.1681, + "step": 8219 + }, + { + "epoch": 0.19, + "grad_norm": 2.001573764864619, + "learning_rate": 1.8628028641921668e-05, + "loss": 0.9873, + "step": 8220 + }, + { + "epoch": 0.19, + "grad_norm": 2.1672198868764205, + "learning_rate": 1.862764286862993e-05, + "loss": 1.0665, + "step": 8221 + }, + { + "epoch": 0.19, + "grad_norm": 2.1130508099557517, + "learning_rate": 1.862725704510532e-05, + "loss": 1.0226, + "step": 8222 + }, + { + "epoch": 0.19, + "grad_norm": 2.0691853512231524, + "learning_rate": 1.8626871171350093e-05, + "loss": 0.9324, + "step": 8223 + }, + { + "epoch": 0.19, + "grad_norm": 1.906112953674429, + "learning_rate": 1.8626485247366486e-05, + "loss": 1.2349, + "step": 8224 + }, + { + "epoch": 0.19, + "grad_norm": 2.1013494707886715, + "learning_rate": 1.8626099273156753e-05, + "loss": 1.1292, + "step": 8225 + }, + { + "epoch": 0.19, + "grad_norm": 1.933704954258527, + "learning_rate": 1.862571324872314e-05, + "loss": 0.9628, + "step": 8226 + }, + { + "epoch": 0.19, + "grad_norm": 1.0639559994671794, + "learning_rate": 1.8625327174067896e-05, + "loss": 0.955, + "step": 8227 + }, + { + "epoch": 0.19, + "grad_norm": 1.863643286304242, + "learning_rate": 1.8624941049193262e-05, + "loss": 1.1025, + "step": 8228 + }, + { + "epoch": 0.19, + "grad_norm": 2.276446753375151, + "learning_rate": 1.862455487410149e-05, + "loss": 1.1773, + "step": 8229 + }, + { + "epoch": 0.19, + "grad_norm": 3.245969951302524, + "learning_rate": 1.8624168648794833e-05, + "loss": 1.0034, + "step": 8230 + }, + { + "epoch": 0.19, + "grad_norm": 2.041414523728382, + "learning_rate": 1.8623782373275535e-05, + "loss": 0.9454, + "step": 8231 + }, + { + "epoch": 0.19, + "grad_norm": 1.8194281693666856, + "learning_rate": 1.8623396047545844e-05, + "loss": 1.0057, + "step": 8232 + }, + { + "epoch": 0.19, + "grad_norm": 2.9281888084481342, + "learning_rate": 1.862300967160801e-05, + "loss": 0.8758, + "step": 8233 + }, + { + "epoch": 0.19, + "grad_norm": 2.389904311144786, + "learning_rate": 1.8622623245464286e-05, + "loss": 1.0811, + "step": 8234 + }, + { + "epoch": 0.19, + "grad_norm": 2.1406917941199493, + "learning_rate": 1.862223676911692e-05, + "loss": 1.0499, + "step": 8235 + }, + { + "epoch": 0.19, + "grad_norm": 2.2898196684662038, + "learning_rate": 1.862185024256816e-05, + "loss": 1.1062, + "step": 8236 + }, + { + "epoch": 0.19, + "grad_norm": 2.2717701737459994, + "learning_rate": 1.862146366582026e-05, + "loss": 0.9866, + "step": 8237 + }, + { + "epoch": 0.19, + "grad_norm": 2.046992683940535, + "learning_rate": 1.862107703887547e-05, + "loss": 1.0504, + "step": 8238 + }, + { + "epoch": 0.19, + "grad_norm": 1.993643683583596, + "learning_rate": 1.8620690361736037e-05, + "loss": 1.1556, + "step": 8239 + }, + { + "epoch": 0.19, + "grad_norm": 2.560542002963531, + "learning_rate": 1.8620303634404218e-05, + "loss": 1.0459, + "step": 8240 + }, + { + "epoch": 0.19, + "grad_norm": 2.1103784280538265, + "learning_rate": 1.8619916856882258e-05, + "loss": 0.9724, + "step": 8241 + }, + { + "epoch": 0.19, + "grad_norm": 1.899180524193038, + "learning_rate": 1.8619530029172417e-05, + "loss": 1.0686, + "step": 8242 + }, + { + "epoch": 0.19, + "grad_norm": 4.151914009786051, + "learning_rate": 1.861914315127694e-05, + "loss": 1.0539, + "step": 8243 + }, + { + "epoch": 0.19, + "grad_norm": 2.039682387141865, + "learning_rate": 1.8618756223198086e-05, + "loss": 1.0563, + "step": 8244 + }, + { + "epoch": 0.19, + "grad_norm": 2.1943335173431824, + "learning_rate": 1.8618369244938103e-05, + "loss": 0.8886, + "step": 8245 + }, + { + "epoch": 0.19, + "grad_norm": 2.2050869785685228, + "learning_rate": 1.8617982216499247e-05, + "loss": 1.0907, + "step": 8246 + }, + { + "epoch": 0.19, + "grad_norm": 3.1604173896305863, + "learning_rate": 1.861759513788377e-05, + "loss": 1.0955, + "step": 8247 + }, + { + "epoch": 0.19, + "grad_norm": 2.046345490409179, + "learning_rate": 1.8617208009093922e-05, + "loss": 1.1737, + "step": 8248 + }, + { + "epoch": 0.19, + "grad_norm": 2.084208437427562, + "learning_rate": 1.8616820830131966e-05, + "loss": 1.0827, + "step": 8249 + }, + { + "epoch": 0.19, + "grad_norm": 2.198330352202311, + "learning_rate": 1.861643360100015e-05, + "loss": 0.947, + "step": 8250 + }, + { + "epoch": 0.19, + "grad_norm": 2.0656492724152185, + "learning_rate": 1.861604632170073e-05, + "loss": 1.2245, + "step": 8251 + }, + { + "epoch": 0.19, + "grad_norm": 2.231896654264402, + "learning_rate": 1.861565899223596e-05, + "loss": 1.1471, + "step": 8252 + }, + { + "epoch": 0.19, + "grad_norm": 1.8858382403927076, + "learning_rate": 1.8615271612608094e-05, + "loss": 0.9359, + "step": 8253 + }, + { + "epoch": 0.19, + "grad_norm": 1.148687377757259, + "learning_rate": 1.861488418281939e-05, + "loss": 0.9529, + "step": 8254 + }, + { + "epoch": 0.19, + "grad_norm": 1.9113224700640012, + "learning_rate": 1.8614496702872102e-05, + "loss": 0.9494, + "step": 8255 + }, + { + "epoch": 0.19, + "grad_norm": 1.157179207174931, + "learning_rate": 1.8614109172768487e-05, + "loss": 0.9588, + "step": 8256 + }, + { + "epoch": 0.19, + "grad_norm": 1.1576390581493023, + "learning_rate": 1.8613721592510804e-05, + "loss": 1.014, + "step": 8257 + }, + { + "epoch": 0.19, + "grad_norm": 2.2852212387201143, + "learning_rate": 1.8613333962101303e-05, + "loss": 0.9934, + "step": 8258 + }, + { + "epoch": 0.19, + "grad_norm": 1.9812744916791993, + "learning_rate": 1.8612946281542247e-05, + "loss": 1.0045, + "step": 8259 + }, + { + "epoch": 0.19, + "grad_norm": 2.12974595037254, + "learning_rate": 1.861255855083589e-05, + "loss": 1.1325, + "step": 8260 + }, + { + "epoch": 0.19, + "grad_norm": 1.927386628012426, + "learning_rate": 1.861217076998449e-05, + "loss": 1.1329, + "step": 8261 + }, + { + "epoch": 0.19, + "grad_norm": 2.1926922615875037, + "learning_rate": 1.8611782938990308e-05, + "loss": 1.0877, + "step": 8262 + }, + { + "epoch": 0.19, + "grad_norm": 2.1559074909870053, + "learning_rate": 1.86113950578556e-05, + "loss": 1.141, + "step": 8263 + }, + { + "epoch": 0.19, + "grad_norm": 1.7611851331413952, + "learning_rate": 1.8611007126582617e-05, + "loss": 1.1236, + "step": 8264 + }, + { + "epoch": 0.19, + "grad_norm": 2.2445431274460956, + "learning_rate": 1.861061914517363e-05, + "loss": 0.9878, + "step": 8265 + }, + { + "epoch": 0.19, + "grad_norm": 2.4048530631958234, + "learning_rate": 1.861023111363089e-05, + "loss": 1.0448, + "step": 8266 + }, + { + "epoch": 0.19, + "grad_norm": 2.076886162298107, + "learning_rate": 1.860984303195666e-05, + "loss": 1.1478, + "step": 8267 + }, + { + "epoch": 0.19, + "grad_norm": 2.0889188744257092, + "learning_rate": 1.8609454900153195e-05, + "loss": 1.1471, + "step": 8268 + }, + { + "epoch": 0.19, + "grad_norm": 2.00067554990038, + "learning_rate": 1.860906671822276e-05, + "loss": 1.0971, + "step": 8269 + }, + { + "epoch": 0.19, + "grad_norm": 1.959632413355954, + "learning_rate": 1.8608678486167615e-05, + "loss": 1.1605, + "step": 8270 + }, + { + "epoch": 0.19, + "grad_norm": 2.241727433082144, + "learning_rate": 1.8608290203990015e-05, + "loss": 1.0141, + "step": 8271 + }, + { + "epoch": 0.19, + "grad_norm": 2.1189897230801837, + "learning_rate": 1.8607901871692226e-05, + "loss": 1.0702, + "step": 8272 + }, + { + "epoch": 0.19, + "grad_norm": 2.2580409474775647, + "learning_rate": 1.8607513489276507e-05, + "loss": 1.1144, + "step": 8273 + }, + { + "epoch": 0.19, + "grad_norm": 2.0874218001211813, + "learning_rate": 1.860712505674512e-05, + "loss": 1.0219, + "step": 8274 + }, + { + "epoch": 0.19, + "grad_norm": 2.061077710674619, + "learning_rate": 1.8606736574100325e-05, + "loss": 1.0487, + "step": 8275 + }, + { + "epoch": 0.19, + "grad_norm": 2.0096003304099903, + "learning_rate": 1.8606348041344383e-05, + "loss": 0.9996, + "step": 8276 + }, + { + "epoch": 0.19, + "grad_norm": 2.0518992881122817, + "learning_rate": 1.860595945847956e-05, + "loss": 1.082, + "step": 8277 + }, + { + "epoch": 0.2, + "grad_norm": 1.99980707176354, + "learning_rate": 1.8605570825508117e-05, + "loss": 1.1001, + "step": 8278 + }, + { + "epoch": 0.2, + "grad_norm": 2.2458375598091993, + "learning_rate": 1.8605182142432317e-05, + "loss": 1.0736, + "step": 8279 + }, + { + "epoch": 0.2, + "grad_norm": 2.968546305667251, + "learning_rate": 1.860479340925442e-05, + "loss": 1.0436, + "step": 8280 + }, + { + "epoch": 0.2, + "grad_norm": 1.9303167682720246, + "learning_rate": 1.860440462597669e-05, + "loss": 1.0847, + "step": 8281 + }, + { + "epoch": 0.2, + "grad_norm": 1.8432267179387085, + "learning_rate": 1.8604015792601395e-05, + "loss": 1.073, + "step": 8282 + }, + { + "epoch": 0.2, + "grad_norm": 2.137482307562418, + "learning_rate": 1.8603626909130796e-05, + "loss": 0.9752, + "step": 8283 + }, + { + "epoch": 0.2, + "grad_norm": 1.825160652946987, + "learning_rate": 1.8603237975567157e-05, + "loss": 1.1098, + "step": 8284 + }, + { + "epoch": 0.2, + "grad_norm": 1.1980038614172779, + "learning_rate": 1.860284899191274e-05, + "loss": 1.012, + "step": 8285 + }, + { + "epoch": 0.2, + "grad_norm": 2.1297870999929556, + "learning_rate": 1.8602459958169817e-05, + "loss": 1.152, + "step": 8286 + }, + { + "epoch": 0.2, + "grad_norm": 2.6998500755893464, + "learning_rate": 1.8602070874340646e-05, + "loss": 1.0263, + "step": 8287 + }, + { + "epoch": 0.2, + "grad_norm": 2.098863783126464, + "learning_rate": 1.8601681740427495e-05, + "loss": 1.1256, + "step": 8288 + }, + { + "epoch": 0.2, + "grad_norm": 1.996500448447899, + "learning_rate": 1.8601292556432632e-05, + "loss": 1.0606, + "step": 8289 + }, + { + "epoch": 0.2, + "grad_norm": 1.980504376915814, + "learning_rate": 1.860090332235832e-05, + "loss": 0.9925, + "step": 8290 + }, + { + "epoch": 0.2, + "grad_norm": 2.1992406073275146, + "learning_rate": 1.860051403820682e-05, + "loss": 0.9981, + "step": 8291 + }, + { + "epoch": 0.2, + "grad_norm": 1.1115429980368885, + "learning_rate": 1.860012470398041e-05, + "loss": 1.0272, + "step": 8292 + }, + { + "epoch": 0.2, + "grad_norm": 2.065747733350086, + "learning_rate": 1.859973531968135e-05, + "loss": 1.1049, + "step": 8293 + }, + { + "epoch": 0.2, + "grad_norm": 1.181895181235296, + "learning_rate": 1.859934588531191e-05, + "loss": 0.926, + "step": 8294 + }, + { + "epoch": 0.2, + "grad_norm": 2.2970106614363495, + "learning_rate": 1.8598956400874354e-05, + "loss": 1.0044, + "step": 8295 + }, + { + "epoch": 0.2, + "grad_norm": 1.1796837148029344, + "learning_rate": 1.8598566866370952e-05, + "loss": 0.9563, + "step": 8296 + }, + { + "epoch": 0.2, + "grad_norm": 1.0893420456264897, + "learning_rate": 1.859817728180397e-05, + "loss": 0.9619, + "step": 8297 + }, + { + "epoch": 0.2, + "grad_norm": 1.1106850137856954, + "learning_rate": 1.859778764717568e-05, + "loss": 0.9945, + "step": 8298 + }, + { + "epoch": 0.2, + "grad_norm": 2.047648330137469, + "learning_rate": 1.8597397962488347e-05, + "loss": 1.0006, + "step": 8299 + }, + { + "epoch": 0.2, + "grad_norm": 2.228736352478932, + "learning_rate": 1.8597008227744242e-05, + "loss": 0.9419, + "step": 8300 + }, + { + "epoch": 0.2, + "grad_norm": 2.0786855806927633, + "learning_rate": 1.8596618442945634e-05, + "loss": 1.0314, + "step": 8301 + }, + { + "epoch": 0.2, + "grad_norm": 2.1699708405484324, + "learning_rate": 1.8596228608094792e-05, + "loss": 1.0199, + "step": 8302 + }, + { + "epoch": 0.2, + "grad_norm": 2.0738866659404964, + "learning_rate": 1.8595838723193986e-05, + "loss": 1.019, + "step": 8303 + }, + { + "epoch": 0.2, + "grad_norm": 2.2884282593994394, + "learning_rate": 1.859544878824548e-05, + "loss": 1.0272, + "step": 8304 + }, + { + "epoch": 0.2, + "grad_norm": 2.3001992078598983, + "learning_rate": 1.859505880325155e-05, + "loss": 1.0796, + "step": 8305 + }, + { + "epoch": 0.2, + "grad_norm": 2.1669569470349557, + "learning_rate": 1.8594668768214474e-05, + "loss": 1.1658, + "step": 8306 + }, + { + "epoch": 0.2, + "grad_norm": 2.31787937213857, + "learning_rate": 1.859427868313651e-05, + "loss": 0.9819, + "step": 8307 + }, + { + "epoch": 0.2, + "grad_norm": 2.937728189897282, + "learning_rate": 1.8593888548019938e-05, + "loss": 0.9455, + "step": 8308 + }, + { + "epoch": 0.2, + "grad_norm": 1.8660468828363446, + "learning_rate": 1.8593498362867023e-05, + "loss": 1.237, + "step": 8309 + }, + { + "epoch": 0.2, + "grad_norm": 1.2593861072913626, + "learning_rate": 1.859310812768004e-05, + "loss": 1.0082, + "step": 8310 + }, + { + "epoch": 0.2, + "grad_norm": 2.0179032793530345, + "learning_rate": 1.8592717842461262e-05, + "loss": 1.036, + "step": 8311 + }, + { + "epoch": 0.2, + "grad_norm": 2.8936728557310123, + "learning_rate": 1.859232750721296e-05, + "loss": 1.0674, + "step": 8312 + }, + { + "epoch": 0.2, + "grad_norm": 1.9079259557177697, + "learning_rate": 1.8591937121937408e-05, + "loss": 1.1268, + "step": 8313 + }, + { + "epoch": 0.2, + "grad_norm": 2.0656602717257853, + "learning_rate": 1.859154668663688e-05, + "loss": 1.0731, + "step": 8314 + }, + { + "epoch": 0.2, + "grad_norm": 1.9760376573615623, + "learning_rate": 1.8591156201313642e-05, + "loss": 1.1954, + "step": 8315 + }, + { + "epoch": 0.2, + "grad_norm": 2.1300541149900982, + "learning_rate": 1.8590765665969974e-05, + "loss": 1.1405, + "step": 8316 + }, + { + "epoch": 0.2, + "grad_norm": 1.1846613338947543, + "learning_rate": 1.859037508060815e-05, + "loss": 0.95, + "step": 8317 + }, + { + "epoch": 0.2, + "grad_norm": 1.0759900515189118, + "learning_rate": 1.8589984445230444e-05, + "loss": 1.0219, + "step": 8318 + }, + { + "epoch": 0.2, + "grad_norm": 1.9660367287538287, + "learning_rate": 1.8589593759839125e-05, + "loss": 1.1593, + "step": 8319 + }, + { + "epoch": 0.2, + "grad_norm": 2.4572026498283073, + "learning_rate": 1.8589203024436477e-05, + "loss": 1.0331, + "step": 8320 + }, + { + "epoch": 0.2, + "grad_norm": 1.9321847573297037, + "learning_rate": 1.8588812239024766e-05, + "loss": 1.0565, + "step": 8321 + }, + { + "epoch": 0.2, + "grad_norm": 2.6238452353444197, + "learning_rate": 1.858842140360627e-05, + "loss": 1.0877, + "step": 8322 + }, + { + "epoch": 0.2, + "grad_norm": 2.447396921044077, + "learning_rate": 1.8588030518183265e-05, + "loss": 1.0316, + "step": 8323 + }, + { + "epoch": 0.2, + "grad_norm": 2.7748946707137985, + "learning_rate": 1.8587639582758032e-05, + "loss": 1.0761, + "step": 8324 + }, + { + "epoch": 0.2, + "grad_norm": 2.2474472723588783, + "learning_rate": 1.8587248597332837e-05, + "loss": 1.0875, + "step": 8325 + }, + { + "epoch": 0.2, + "grad_norm": 2.0845745681232097, + "learning_rate": 1.8586857561909967e-05, + "loss": 1.011, + "step": 8326 + }, + { + "epoch": 0.2, + "grad_norm": 2.0207893109575603, + "learning_rate": 1.8586466476491692e-05, + "loss": 0.9794, + "step": 8327 + }, + { + "epoch": 0.2, + "grad_norm": 2.1895658433613128, + "learning_rate": 1.8586075341080293e-05, + "loss": 1.1659, + "step": 8328 + }, + { + "epoch": 0.2, + "grad_norm": 2.1033517067256224, + "learning_rate": 1.858568415567804e-05, + "loss": 1.1136, + "step": 8329 + }, + { + "epoch": 0.2, + "grad_norm": 1.1401146588447926, + "learning_rate": 1.8585292920287217e-05, + "loss": 0.9813, + "step": 8330 + }, + { + "epoch": 0.2, + "grad_norm": 2.426452285378917, + "learning_rate": 1.8584901634910105e-05, + "loss": 0.9551, + "step": 8331 + }, + { + "epoch": 0.2, + "grad_norm": 1.9865062826196207, + "learning_rate": 1.8584510299548975e-05, + "loss": 1.0483, + "step": 8332 + }, + { + "epoch": 0.2, + "grad_norm": 2.0937856790808445, + "learning_rate": 1.8584118914206105e-05, + "loss": 1.0839, + "step": 8333 + }, + { + "epoch": 0.2, + "grad_norm": 1.200610035826194, + "learning_rate": 1.8583727478883782e-05, + "loss": 0.9355, + "step": 8334 + }, + { + "epoch": 0.2, + "grad_norm": 2.0186059886552465, + "learning_rate": 1.8583335993584275e-05, + "loss": 0.9608, + "step": 8335 + }, + { + "epoch": 0.2, + "grad_norm": 2.4776629639022203, + "learning_rate": 1.858294445830987e-05, + "loss": 1.1628, + "step": 8336 + }, + { + "epoch": 0.2, + "grad_norm": 2.3708677374651796, + "learning_rate": 1.858255287306285e-05, + "loss": 0.9321, + "step": 8337 + }, + { + "epoch": 0.2, + "grad_norm": 2.121089029043434, + "learning_rate": 1.8582161237845486e-05, + "loss": 1.1297, + "step": 8338 + }, + { + "epoch": 0.2, + "grad_norm": 2.5794701651365814, + "learning_rate": 1.8581769552660064e-05, + "loss": 0.9757, + "step": 8339 + }, + { + "epoch": 0.2, + "grad_norm": 2.413023453076768, + "learning_rate": 1.858137781750886e-05, + "loss": 1.1432, + "step": 8340 + }, + { + "epoch": 0.2, + "grad_norm": 2.2085020420298824, + "learning_rate": 1.8580986032394157e-05, + "loss": 1.0999, + "step": 8341 + }, + { + "epoch": 0.2, + "grad_norm": 1.229048280289969, + "learning_rate": 1.858059419731824e-05, + "loss": 1.0231, + "step": 8342 + }, + { + "epoch": 0.2, + "grad_norm": 2.049145199675218, + "learning_rate": 1.8580202312283383e-05, + "loss": 1.0606, + "step": 8343 + }, + { + "epoch": 0.2, + "grad_norm": 2.1400048571375616, + "learning_rate": 1.8579810377291875e-05, + "loss": 1.0421, + "step": 8344 + }, + { + "epoch": 0.2, + "grad_norm": 2.3780020787971576, + "learning_rate": 1.8579418392345993e-05, + "loss": 1.0837, + "step": 8345 + }, + { + "epoch": 0.2, + "grad_norm": 2.010724979371465, + "learning_rate": 1.8579026357448023e-05, + "loss": 1.1157, + "step": 8346 + }, + { + "epoch": 0.2, + "grad_norm": 2.542527208801562, + "learning_rate": 1.8578634272600246e-05, + "loss": 1.0975, + "step": 8347 + }, + { + "epoch": 0.2, + "grad_norm": 2.641246104695652, + "learning_rate": 1.8578242137804938e-05, + "loss": 1.22, + "step": 8348 + }, + { + "epoch": 0.2, + "grad_norm": 1.9854462757448093, + "learning_rate": 1.8577849953064394e-05, + "loss": 1.0699, + "step": 8349 + }, + { + "epoch": 0.2, + "grad_norm": 2.2620762725057184, + "learning_rate": 1.8577457718380893e-05, + "loss": 0.9846, + "step": 8350 + }, + { + "epoch": 0.2, + "grad_norm": 2.4160368571701514, + "learning_rate": 1.8577065433756712e-05, + "loss": 1.1796, + "step": 8351 + }, + { + "epoch": 0.2, + "grad_norm": 2.3794087392522543, + "learning_rate": 1.8576673099194145e-05, + "loss": 1.2083, + "step": 8352 + }, + { + "epoch": 0.2, + "grad_norm": 2.222754019789238, + "learning_rate": 1.857628071469547e-05, + "loss": 1.1075, + "step": 8353 + }, + { + "epoch": 0.2, + "grad_norm": 2.2892652082523117, + "learning_rate": 1.8575888280262975e-05, + "loss": 1.1254, + "step": 8354 + }, + { + "epoch": 0.2, + "grad_norm": 2.193821079926077, + "learning_rate": 1.8575495795898942e-05, + "loss": 1.0493, + "step": 8355 + }, + { + "epoch": 0.2, + "grad_norm": 2.507987008250985, + "learning_rate": 1.8575103261605657e-05, + "loss": 1.1867, + "step": 8356 + }, + { + "epoch": 0.2, + "grad_norm": 2.134700322832011, + "learning_rate": 1.8574710677385412e-05, + "loss": 1.0283, + "step": 8357 + }, + { + "epoch": 0.2, + "grad_norm": 2.327231255116998, + "learning_rate": 1.857431804324048e-05, + "loss": 1.1649, + "step": 8358 + }, + { + "epoch": 0.2, + "grad_norm": 1.9746807923947642, + "learning_rate": 1.8573925359173154e-05, + "loss": 0.9986, + "step": 8359 + }, + { + "epoch": 0.2, + "grad_norm": 2.5709551829154456, + "learning_rate": 1.8573532625185723e-05, + "loss": 1.0646, + "step": 8360 + }, + { + "epoch": 0.2, + "grad_norm": 1.9984201516549698, + "learning_rate": 1.857313984128047e-05, + "loss": 1.2159, + "step": 8361 + }, + { + "epoch": 0.2, + "grad_norm": 2.013552479252191, + "learning_rate": 1.8572747007459678e-05, + "loss": 1.0339, + "step": 8362 + }, + { + "epoch": 0.2, + "grad_norm": 2.0342046632666486, + "learning_rate": 1.8572354123725645e-05, + "loss": 1.2184, + "step": 8363 + }, + { + "epoch": 0.2, + "grad_norm": 2.147937698424417, + "learning_rate": 1.857196119008065e-05, + "loss": 0.9051, + "step": 8364 + }, + { + "epoch": 0.2, + "grad_norm": 2.0553636631559677, + "learning_rate": 1.857156820652698e-05, + "loss": 1.1546, + "step": 8365 + }, + { + "epoch": 0.2, + "grad_norm": 2.904960455144104, + "learning_rate": 1.8571175173066935e-05, + "loss": 1.0164, + "step": 8366 + }, + { + "epoch": 0.2, + "grad_norm": 1.9648100892031402, + "learning_rate": 1.8570782089702787e-05, + "loss": 1.0199, + "step": 8367 + }, + { + "epoch": 0.2, + "grad_norm": 1.1397824188227788, + "learning_rate": 1.8570388956436836e-05, + "loss": 0.9889, + "step": 8368 + }, + { + "epoch": 0.2, + "grad_norm": 2.026543212506931, + "learning_rate": 1.8569995773271366e-05, + "loss": 1.1778, + "step": 8369 + }, + { + "epoch": 0.2, + "grad_norm": 1.9676005179814309, + "learning_rate": 1.8569602540208664e-05, + "loss": 1.125, + "step": 8370 + }, + { + "epoch": 0.2, + "grad_norm": 2.0952729999629467, + "learning_rate": 1.8569209257251028e-05, + "loss": 1.1265, + "step": 8371 + }, + { + "epoch": 0.2, + "grad_norm": 2.0388759425156455, + "learning_rate": 1.8568815924400742e-05, + "loss": 1.0267, + "step": 8372 + }, + { + "epoch": 0.2, + "grad_norm": 2.598893247690359, + "learning_rate": 1.856842254166009e-05, + "loss": 0.956, + "step": 8373 + }, + { + "epoch": 0.2, + "grad_norm": 2.005359587393706, + "learning_rate": 1.8568029109031377e-05, + "loss": 1.0442, + "step": 8374 + }, + { + "epoch": 0.2, + "grad_norm": 1.8634980672119836, + "learning_rate": 1.8567635626516885e-05, + "loss": 1.0395, + "step": 8375 + }, + { + "epoch": 0.2, + "grad_norm": 2.212541112271141, + "learning_rate": 1.8567242094118902e-05, + "loss": 1.2175, + "step": 8376 + }, + { + "epoch": 0.2, + "grad_norm": 2.073354178767149, + "learning_rate": 1.8566848511839724e-05, + "loss": 1.0484, + "step": 8377 + }, + { + "epoch": 0.2, + "grad_norm": 1.8797521295578705, + "learning_rate": 1.8566454879681646e-05, + "loss": 0.9837, + "step": 8378 + }, + { + "epoch": 0.2, + "grad_norm": 2.2244452268577946, + "learning_rate": 1.856606119764695e-05, + "loss": 1.0451, + "step": 8379 + }, + { + "epoch": 0.2, + "grad_norm": 2.0747421113098934, + "learning_rate": 1.856566746573794e-05, + "loss": 1.0412, + "step": 8380 + }, + { + "epoch": 0.2, + "grad_norm": 1.975729588015701, + "learning_rate": 1.85652736839569e-05, + "loss": 1.1386, + "step": 8381 + }, + { + "epoch": 0.2, + "grad_norm": 4.610747225420086, + "learning_rate": 1.8564879852306122e-05, + "loss": 1.2773, + "step": 8382 + }, + { + "epoch": 0.2, + "grad_norm": 2.048576271770908, + "learning_rate": 1.8564485970787904e-05, + "loss": 0.9536, + "step": 8383 + }, + { + "epoch": 0.2, + "grad_norm": 2.232520183891694, + "learning_rate": 1.8564092039404537e-05, + "loss": 1.0803, + "step": 8384 + }, + { + "epoch": 0.2, + "grad_norm": 1.175307881752367, + "learning_rate": 1.8563698058158313e-05, + "loss": 0.986, + "step": 8385 + }, + { + "epoch": 0.2, + "grad_norm": 2.665665319854466, + "learning_rate": 1.8563304027051532e-05, + "loss": 1.013, + "step": 8386 + }, + { + "epoch": 0.2, + "grad_norm": 2.7165684113316275, + "learning_rate": 1.856290994608648e-05, + "loss": 1.0458, + "step": 8387 + }, + { + "epoch": 0.2, + "grad_norm": 2.207180529056921, + "learning_rate": 1.8562515815265456e-05, + "loss": 1.0906, + "step": 8388 + }, + { + "epoch": 0.2, + "grad_norm": 2.3727120310409675, + "learning_rate": 1.8562121634590756e-05, + "loss": 0.9775, + "step": 8389 + }, + { + "epoch": 0.2, + "grad_norm": 2.1645575730763453, + "learning_rate": 1.856172740406467e-05, + "loss": 1.0588, + "step": 8390 + }, + { + "epoch": 0.2, + "grad_norm": 1.9694618106627473, + "learning_rate": 1.85613331236895e-05, + "loss": 1.0594, + "step": 8391 + }, + { + "epoch": 0.2, + "grad_norm": 2.050763341848203, + "learning_rate": 1.8560938793467537e-05, + "loss": 0.9239, + "step": 8392 + }, + { + "epoch": 0.2, + "grad_norm": 1.9756766376002388, + "learning_rate": 1.8560544413401077e-05, + "loss": 1.1016, + "step": 8393 + }, + { + "epoch": 0.2, + "grad_norm": 2.1183651558930636, + "learning_rate": 1.8560149983492418e-05, + "loss": 1.0526, + "step": 8394 + }, + { + "epoch": 0.2, + "grad_norm": 2.119837422012809, + "learning_rate": 1.855975550374385e-05, + "loss": 1.1048, + "step": 8395 + }, + { + "epoch": 0.2, + "grad_norm": 1.123397454398934, + "learning_rate": 1.8559360974157683e-05, + "loss": 1.0152, + "step": 8396 + }, + { + "epoch": 0.2, + "grad_norm": 2.090580306482865, + "learning_rate": 1.8558966394736205e-05, + "loss": 1.0484, + "step": 8397 + }, + { + "epoch": 0.2, + "grad_norm": 2.587275481393979, + "learning_rate": 1.8558571765481712e-05, + "loss": 0.9851, + "step": 8398 + }, + { + "epoch": 0.2, + "grad_norm": 1.765642422008855, + "learning_rate": 1.855817708639651e-05, + "loss": 1.0382, + "step": 8399 + }, + { + "epoch": 0.2, + "grad_norm": 2.02511764480539, + "learning_rate": 1.8557782357482884e-05, + "loss": 1.0688, + "step": 8400 + }, + { + "epoch": 0.2, + "grad_norm": 2.3783796501557917, + "learning_rate": 1.8557387578743146e-05, + "loss": 1.1451, + "step": 8401 + }, + { + "epoch": 0.2, + "grad_norm": 1.9750735175328094, + "learning_rate": 1.855699275017959e-05, + "loss": 0.9822, + "step": 8402 + }, + { + "epoch": 0.2, + "grad_norm": 1.9259592332138686, + "learning_rate": 1.855659787179451e-05, + "loss": 0.8778, + "step": 8403 + }, + { + "epoch": 0.2, + "grad_norm": 2.523625596483187, + "learning_rate": 1.8556202943590206e-05, + "loss": 1.085, + "step": 8404 + }, + { + "epoch": 0.2, + "grad_norm": 3.4578471406683597, + "learning_rate": 1.855580796556898e-05, + "loss": 1.05, + "step": 8405 + }, + { + "epoch": 0.2, + "grad_norm": 2.645851314803618, + "learning_rate": 1.855541293773313e-05, + "loss": 1.0089, + "step": 8406 + }, + { + "epoch": 0.2, + "grad_norm": 2.2650730746398806, + "learning_rate": 1.855501786008496e-05, + "loss": 1.1208, + "step": 8407 + }, + { + "epoch": 0.2, + "grad_norm": 2.205242994470151, + "learning_rate": 1.855462273262677e-05, + "loss": 0.9924, + "step": 8408 + }, + { + "epoch": 0.2, + "grad_norm": 2.236498952623222, + "learning_rate": 1.855422755536085e-05, + "loss": 0.9887, + "step": 8409 + }, + { + "epoch": 0.2, + "grad_norm": 2.3512545037196357, + "learning_rate": 1.855383232828952e-05, + "loss": 1.0528, + "step": 8410 + }, + { + "epoch": 0.2, + "grad_norm": 2.4716140398481037, + "learning_rate": 1.855343705141506e-05, + "loss": 1.0851, + "step": 8411 + }, + { + "epoch": 0.2, + "grad_norm": 2.2248708485070803, + "learning_rate": 1.855304172473979e-05, + "loss": 1.1272, + "step": 8412 + }, + { + "epoch": 0.2, + "grad_norm": 2.000877776111405, + "learning_rate": 1.8552646348266e-05, + "loss": 1.0151, + "step": 8413 + }, + { + "epoch": 0.2, + "grad_norm": 2.305843843295976, + "learning_rate": 1.8552250921995995e-05, + "loss": 0.9755, + "step": 8414 + }, + { + "epoch": 0.2, + "grad_norm": 1.9248542648619371, + "learning_rate": 1.8551855445932077e-05, + "loss": 1.1472, + "step": 8415 + }, + { + "epoch": 0.2, + "grad_norm": 2.533699051898206, + "learning_rate": 1.8551459920076547e-05, + "loss": 1.0684, + "step": 8416 + }, + { + "epoch": 0.2, + "grad_norm": 2.246713123961481, + "learning_rate": 1.8551064344431712e-05, + "loss": 0.9846, + "step": 8417 + }, + { + "epoch": 0.2, + "grad_norm": 2.292493851229574, + "learning_rate": 1.8550668718999873e-05, + "loss": 1.0647, + "step": 8418 + }, + { + "epoch": 0.2, + "grad_norm": 2.320641110150578, + "learning_rate": 1.8550273043783337e-05, + "loss": 1.0953, + "step": 8419 + }, + { + "epoch": 0.2, + "grad_norm": 2.1166078254082206, + "learning_rate": 1.85498773187844e-05, + "loss": 1.0751, + "step": 8420 + }, + { + "epoch": 0.2, + "grad_norm": 2.094825957014225, + "learning_rate": 1.8549481544005373e-05, + "loss": 1.0577, + "step": 8421 + }, + { + "epoch": 0.2, + "grad_norm": 1.9607983423198694, + "learning_rate": 1.8549085719448553e-05, + "loss": 1.0746, + "step": 8422 + }, + { + "epoch": 0.2, + "grad_norm": 2.0908299036754787, + "learning_rate": 1.8548689845116255e-05, + "loss": 0.9752, + "step": 8423 + }, + { + "epoch": 0.2, + "grad_norm": 1.1027329450112238, + "learning_rate": 1.8548293921010776e-05, + "loss": 0.9745, + "step": 8424 + }, + { + "epoch": 0.2, + "grad_norm": 1.9818328813976853, + "learning_rate": 1.854789794713442e-05, + "loss": 1.0677, + "step": 8425 + }, + { + "epoch": 0.2, + "grad_norm": 2.098029203842992, + "learning_rate": 1.8547501923489503e-05, + "loss": 1.0923, + "step": 8426 + }, + { + "epoch": 0.2, + "grad_norm": 2.1020488957209444, + "learning_rate": 1.854710585007832e-05, + "loss": 1.0585, + "step": 8427 + }, + { + "epoch": 0.2, + "grad_norm": 2.101685179507401, + "learning_rate": 1.8546709726903178e-05, + "loss": 1.0504, + "step": 8428 + }, + { + "epoch": 0.2, + "grad_norm": 2.1318665563199266, + "learning_rate": 1.854631355396639e-05, + "loss": 1.1802, + "step": 8429 + }, + { + "epoch": 0.2, + "grad_norm": 1.075470778975699, + "learning_rate": 1.8545917331270257e-05, + "loss": 0.968, + "step": 8430 + }, + { + "epoch": 0.2, + "grad_norm": 1.9449341762895538, + "learning_rate": 1.8545521058817088e-05, + "loss": 1.0254, + "step": 8431 + }, + { + "epoch": 0.2, + "grad_norm": 2.007913781159975, + "learning_rate": 1.854512473660919e-05, + "loss": 0.8743, + "step": 8432 + }, + { + "epoch": 0.2, + "grad_norm": 1.0853714107445966, + "learning_rate": 1.8544728364648875e-05, + "loss": 0.9936, + "step": 8433 + }, + { + "epoch": 0.2, + "grad_norm": 2.057667452130808, + "learning_rate": 1.8544331942938442e-05, + "loss": 1.1183, + "step": 8434 + }, + { + "epoch": 0.2, + "grad_norm": 1.9363917672951514, + "learning_rate": 1.85439354714802e-05, + "loss": 1.0735, + "step": 8435 + }, + { + "epoch": 0.2, + "grad_norm": 2.178977991726285, + "learning_rate": 1.8543538950276465e-05, + "loss": 1.0224, + "step": 8436 + }, + { + "epoch": 0.2, + "grad_norm": 2.0877674889897753, + "learning_rate": 1.8543142379329543e-05, + "loss": 1.1558, + "step": 8437 + }, + { + "epoch": 0.2, + "grad_norm": 2.3193143461125896, + "learning_rate": 1.8542745758641736e-05, + "loss": 1.1235, + "step": 8438 + }, + { + "epoch": 0.2, + "grad_norm": 2.073349118563473, + "learning_rate": 1.8542349088215362e-05, + "loss": 1.2035, + "step": 8439 + }, + { + "epoch": 0.2, + "grad_norm": 2.519592766509806, + "learning_rate": 1.8541952368052727e-05, + "loss": 0.9753, + "step": 8440 + }, + { + "epoch": 0.2, + "grad_norm": 2.1166566326384864, + "learning_rate": 1.8541555598156143e-05, + "loss": 1.1007, + "step": 8441 + }, + { + "epoch": 0.2, + "grad_norm": 2.2477537968946244, + "learning_rate": 1.8541158778527914e-05, + "loss": 1.0888, + "step": 8442 + }, + { + "epoch": 0.2, + "grad_norm": 2.025773486350266, + "learning_rate": 1.8540761909170356e-05, + "loss": 1.0057, + "step": 8443 + }, + { + "epoch": 0.2, + "grad_norm": 2.021096660686015, + "learning_rate": 1.8540364990085778e-05, + "loss": 0.9963, + "step": 8444 + }, + { + "epoch": 0.2, + "grad_norm": 2.1126131003656474, + "learning_rate": 1.8539968021276493e-05, + "loss": 1.0183, + "step": 8445 + }, + { + "epoch": 0.2, + "grad_norm": 1.9304980914334144, + "learning_rate": 1.853957100274481e-05, + "loss": 1.0413, + "step": 8446 + }, + { + "epoch": 0.2, + "grad_norm": 1.9005484030328503, + "learning_rate": 1.853917393449304e-05, + "loss": 1.0497, + "step": 8447 + }, + { + "epoch": 0.2, + "grad_norm": 2.1514628057274114, + "learning_rate": 1.8538776816523495e-05, + "loss": 1.1233, + "step": 8448 + }, + { + "epoch": 0.2, + "grad_norm": 1.9483930082430412, + "learning_rate": 1.853837964883849e-05, + "loss": 0.9546, + "step": 8449 + }, + { + "epoch": 0.2, + "grad_norm": 2.4248113950897, + "learning_rate": 1.8537982431440333e-05, + "loss": 1.1019, + "step": 8450 + }, + { + "epoch": 0.2, + "grad_norm": 2.1934354494295083, + "learning_rate": 1.8537585164331343e-05, + "loss": 1.0013, + "step": 8451 + }, + { + "epoch": 0.2, + "grad_norm": 2.429540268137574, + "learning_rate": 1.8537187847513825e-05, + "loss": 1.0904, + "step": 8452 + }, + { + "epoch": 0.2, + "grad_norm": 2.7493074133232724, + "learning_rate": 1.85367904809901e-05, + "loss": 1.0197, + "step": 8453 + }, + { + "epoch": 0.2, + "grad_norm": 2.687256680746049, + "learning_rate": 1.8536393064762474e-05, + "loss": 1.1167, + "step": 8454 + }, + { + "epoch": 0.2, + "grad_norm": 2.0537857004303337, + "learning_rate": 1.853599559883327e-05, + "loss": 1.0957, + "step": 8455 + }, + { + "epoch": 0.2, + "grad_norm": 1.9639595091012625, + "learning_rate": 1.8535598083204792e-05, + "loss": 1.0884, + "step": 8456 + }, + { + "epoch": 0.2, + "grad_norm": 2.0622835373085215, + "learning_rate": 1.8535200517879364e-05, + "loss": 1.0676, + "step": 8457 + }, + { + "epoch": 0.2, + "grad_norm": 2.0525209451353703, + "learning_rate": 1.8534802902859296e-05, + "loss": 1.0258, + "step": 8458 + }, + { + "epoch": 0.2, + "grad_norm": 2.046504728163624, + "learning_rate": 1.85344052381469e-05, + "loss": 1.1007, + "step": 8459 + }, + { + "epoch": 0.2, + "grad_norm": 1.9120536614407082, + "learning_rate": 1.8534007523744498e-05, + "loss": 0.9385, + "step": 8460 + }, + { + "epoch": 0.2, + "grad_norm": 1.8542998489366482, + "learning_rate": 1.85336097596544e-05, + "loss": 1.1061, + "step": 8461 + }, + { + "epoch": 0.2, + "grad_norm": 2.3010905359289713, + "learning_rate": 1.8533211945878928e-05, + "loss": 1.0163, + "step": 8462 + }, + { + "epoch": 0.2, + "grad_norm": 2.0974849630661425, + "learning_rate": 1.8532814082420393e-05, + "loss": 0.8868, + "step": 8463 + }, + { + "epoch": 0.2, + "grad_norm": 2.2650942607776825, + "learning_rate": 1.8532416169281113e-05, + "loss": 1.0604, + "step": 8464 + }, + { + "epoch": 0.2, + "grad_norm": 2.0480299678879863, + "learning_rate": 1.8532018206463404e-05, + "loss": 1.111, + "step": 8465 + }, + { + "epoch": 0.2, + "grad_norm": 2.439999817864138, + "learning_rate": 1.853162019396958e-05, + "loss": 0.9836, + "step": 8466 + }, + { + "epoch": 0.2, + "grad_norm": 1.1716702661884264, + "learning_rate": 1.8531222131801965e-05, + "loss": 0.9355, + "step": 8467 + }, + { + "epoch": 0.2, + "grad_norm": 2.186197239141499, + "learning_rate": 1.8530824019962877e-05, + "loss": 1.0282, + "step": 8468 + }, + { + "epoch": 0.2, + "grad_norm": 2.22823810406105, + "learning_rate": 1.8530425858454627e-05, + "loss": 1.0312, + "step": 8469 + }, + { + "epoch": 0.2, + "grad_norm": 2.1014880571333454, + "learning_rate": 1.853002764727954e-05, + "loss": 0.9592, + "step": 8470 + }, + { + "epoch": 0.2, + "grad_norm": 2.1648140808889025, + "learning_rate": 1.8529629386439933e-05, + "loss": 1.0681, + "step": 8471 + }, + { + "epoch": 0.2, + "grad_norm": 2.593132136473693, + "learning_rate": 1.8529231075938117e-05, + "loss": 1.0083, + "step": 8472 + }, + { + "epoch": 0.2, + "grad_norm": 2.3427034993125884, + "learning_rate": 1.8528832715776422e-05, + "loss": 1.0432, + "step": 8473 + }, + { + "epoch": 0.2, + "grad_norm": 1.1492780015026003, + "learning_rate": 1.852843430595716e-05, + "loss": 1.0001, + "step": 8474 + }, + { + "epoch": 0.2, + "grad_norm": 2.357049117883546, + "learning_rate": 1.8528035846482656e-05, + "loss": 1.0283, + "step": 8475 + }, + { + "epoch": 0.2, + "grad_norm": 1.1351671722053038, + "learning_rate": 1.8527637337355225e-05, + "loss": 0.983, + "step": 8476 + }, + { + "epoch": 0.2, + "grad_norm": 2.88036420001343, + "learning_rate": 1.852723877857719e-05, + "loss": 1.1499, + "step": 8477 + }, + { + "epoch": 0.2, + "grad_norm": 2.258385368412614, + "learning_rate": 1.8526840170150874e-05, + "loss": 1.1037, + "step": 8478 + }, + { + "epoch": 0.2, + "grad_norm": 1.2274820537196662, + "learning_rate": 1.852644151207859e-05, + "loss": 1.0743, + "step": 8479 + }, + { + "epoch": 0.2, + "grad_norm": 2.1557444791547318, + "learning_rate": 1.8526042804362666e-05, + "loss": 1.0964, + "step": 8480 + }, + { + "epoch": 0.2, + "grad_norm": 2.340653080212369, + "learning_rate": 1.852564404700542e-05, + "loss": 1.009, + "step": 8481 + }, + { + "epoch": 0.2, + "grad_norm": 2.014496329949923, + "learning_rate": 1.852524524000918e-05, + "loss": 0.9464, + "step": 8482 + }, + { + "epoch": 0.2, + "grad_norm": 1.8986610801685235, + "learning_rate": 1.852484638337626e-05, + "loss": 0.9913, + "step": 8483 + }, + { + "epoch": 0.2, + "grad_norm": 2.0851538837992214, + "learning_rate": 1.852444747710899e-05, + "loss": 1.0185, + "step": 8484 + }, + { + "epoch": 0.2, + "grad_norm": 2.3799028493595786, + "learning_rate": 1.852404852120968e-05, + "loss": 1.0852, + "step": 8485 + }, + { + "epoch": 0.2, + "grad_norm": 1.6815273490994254, + "learning_rate": 1.852364951568067e-05, + "loss": 1.0488, + "step": 8486 + }, + { + "epoch": 0.2, + "grad_norm": 1.997203112945821, + "learning_rate": 1.8523250460524264e-05, + "loss": 1.0335, + "step": 8487 + }, + { + "epoch": 0.2, + "grad_norm": 1.8962418253049516, + "learning_rate": 1.85228513557428e-05, + "loss": 1.1153, + "step": 8488 + }, + { + "epoch": 0.2, + "grad_norm": 2.0370734010219214, + "learning_rate": 1.85224522013386e-05, + "loss": 1.1064, + "step": 8489 + }, + { + "epoch": 0.2, + "grad_norm": 1.9047649043145076, + "learning_rate": 1.8522052997313982e-05, + "loss": 1.005, + "step": 8490 + }, + { + "epoch": 0.2, + "grad_norm": 1.8198793817168186, + "learning_rate": 1.8521653743671276e-05, + "loss": 1.0098, + "step": 8491 + }, + { + "epoch": 0.2, + "grad_norm": 2.4932314828698794, + "learning_rate": 1.8521254440412798e-05, + "loss": 1.1179, + "step": 8492 + }, + { + "epoch": 0.2, + "grad_norm": 1.9782138146025234, + "learning_rate": 1.8520855087540887e-05, + "loss": 1.0534, + "step": 8493 + }, + { + "epoch": 0.2, + "grad_norm": 1.9406848467572708, + "learning_rate": 1.8520455685057856e-05, + "loss": 0.9545, + "step": 8494 + }, + { + "epoch": 0.2, + "grad_norm": 2.113295686228709, + "learning_rate": 1.8520056232966033e-05, + "loss": 1.0162, + "step": 8495 + }, + { + "epoch": 0.2, + "grad_norm": 2.0013574244273267, + "learning_rate": 1.8519656731267746e-05, + "loss": 1.1148, + "step": 8496 + }, + { + "epoch": 0.2, + "grad_norm": 2.0551325173105424, + "learning_rate": 1.8519257179965322e-05, + "loss": 0.9945, + "step": 8497 + }, + { + "epoch": 0.2, + "grad_norm": 1.116131636413314, + "learning_rate": 1.8518857579061083e-05, + "loss": 0.9646, + "step": 8498 + }, + { + "epoch": 0.2, + "grad_norm": 2.6173705525399695, + "learning_rate": 1.851845792855736e-05, + "loss": 1.0532, + "step": 8499 + }, + { + "epoch": 0.2, + "grad_norm": 1.8652569966667045, + "learning_rate": 1.8518058228456478e-05, + "loss": 1.1236, + "step": 8500 + }, + { + "epoch": 0.2, + "grad_norm": 1.8471002646997916, + "learning_rate": 1.851765847876076e-05, + "loss": 0.9475, + "step": 8501 + }, + { + "epoch": 0.2, + "grad_norm": 2.546632935954035, + "learning_rate": 1.8517258679472546e-05, + "loss": 1.0951, + "step": 8502 + }, + { + "epoch": 0.2, + "grad_norm": 2.0107712980935695, + "learning_rate": 1.851685883059415e-05, + "loss": 1.0674, + "step": 8503 + }, + { + "epoch": 0.2, + "grad_norm": 2.14705241464869, + "learning_rate": 1.8516458932127906e-05, + "loss": 1.1207, + "step": 8504 + }, + { + "epoch": 0.2, + "grad_norm": 2.838722353376258, + "learning_rate": 1.8516058984076145e-05, + "loss": 0.99, + "step": 8505 + }, + { + "epoch": 0.2, + "grad_norm": 2.2228951991951247, + "learning_rate": 1.851565898644119e-05, + "loss": 1.0002, + "step": 8506 + }, + { + "epoch": 0.2, + "grad_norm": 2.2168515434960367, + "learning_rate": 1.851525893922537e-05, + "loss": 1.1795, + "step": 8507 + }, + { + "epoch": 0.2, + "grad_norm": 2.048884523286419, + "learning_rate": 1.8514858842431017e-05, + "loss": 0.9268, + "step": 8508 + }, + { + "epoch": 0.2, + "grad_norm": 2.3120294747984973, + "learning_rate": 1.8514458696060463e-05, + "loss": 1.1064, + "step": 8509 + }, + { + "epoch": 0.2, + "grad_norm": 2.851699554402686, + "learning_rate": 1.8514058500116033e-05, + "loss": 0.8976, + "step": 8510 + }, + { + "epoch": 0.2, + "grad_norm": 2.077274836526577, + "learning_rate": 1.851365825460006e-05, + "loss": 1.0405, + "step": 8511 + }, + { + "epoch": 0.2, + "grad_norm": 2.0872639483589746, + "learning_rate": 1.851325795951487e-05, + "loss": 0.9335, + "step": 8512 + }, + { + "epoch": 0.2, + "grad_norm": 2.1750059506806614, + "learning_rate": 1.85128576148628e-05, + "loss": 0.9587, + "step": 8513 + }, + { + "epoch": 0.2, + "grad_norm": 2.382505562235475, + "learning_rate": 1.851245722064618e-05, + "loss": 0.9334, + "step": 8514 + }, + { + "epoch": 0.2, + "grad_norm": 2.0845114652282932, + "learning_rate": 1.8512056776867335e-05, + "loss": 1.0169, + "step": 8515 + }, + { + "epoch": 0.2, + "grad_norm": 1.9935730742153857, + "learning_rate": 1.8511656283528603e-05, + "loss": 0.8922, + "step": 8516 + }, + { + "epoch": 0.2, + "grad_norm": 2.1436332266806963, + "learning_rate": 1.851125574063231e-05, + "loss": 1.0087, + "step": 8517 + }, + { + "epoch": 0.2, + "grad_norm": 2.32824439054263, + "learning_rate": 1.8510855148180794e-05, + "loss": 1.1248, + "step": 8518 + }, + { + "epoch": 0.2, + "grad_norm": 2.109118722760845, + "learning_rate": 1.851045450617639e-05, + "loss": 1.0238, + "step": 8519 + }, + { + "epoch": 0.2, + "grad_norm": 1.9250234881603734, + "learning_rate": 1.8510053814621417e-05, + "loss": 1.0221, + "step": 8520 + }, + { + "epoch": 0.2, + "grad_norm": 1.9403837037265426, + "learning_rate": 1.850965307351822e-05, + "loss": 1.1277, + "step": 8521 + }, + { + "epoch": 0.2, + "grad_norm": 2.002297354737239, + "learning_rate": 1.8509252282869127e-05, + "loss": 1.1305, + "step": 8522 + }, + { + "epoch": 0.2, + "grad_norm": 1.8859563157246002, + "learning_rate": 1.8508851442676475e-05, + "loss": 1.0771, + "step": 8523 + }, + { + "epoch": 0.2, + "grad_norm": 1.9288797977021555, + "learning_rate": 1.8508450552942597e-05, + "loss": 1.0384, + "step": 8524 + }, + { + "epoch": 0.2, + "grad_norm": 2.0813982170739056, + "learning_rate": 1.8508049613669824e-05, + "loss": 1.12, + "step": 8525 + }, + { + "epoch": 0.2, + "grad_norm": 2.3590749143583203, + "learning_rate": 1.850764862486049e-05, + "loss": 1.0531, + "step": 8526 + }, + { + "epoch": 0.2, + "grad_norm": 2.017521431473501, + "learning_rate": 1.850724758651694e-05, + "loss": 1.1581, + "step": 8527 + }, + { + "epoch": 0.2, + "grad_norm": 1.1779422398649344, + "learning_rate": 1.8506846498641496e-05, + "loss": 0.968, + "step": 8528 + }, + { + "epoch": 0.2, + "grad_norm": 2.74743367499766, + "learning_rate": 1.8506445361236502e-05, + "loss": 1.1123, + "step": 8529 + }, + { + "epoch": 0.2, + "grad_norm": 2.3450600150841914, + "learning_rate": 1.8506044174304282e-05, + "loss": 1.0012, + "step": 8530 + }, + { + "epoch": 0.2, + "grad_norm": 1.8894934835898949, + "learning_rate": 1.8505642937847188e-05, + "loss": 1.1362, + "step": 8531 + }, + { + "epoch": 0.2, + "grad_norm": 2.101782295922645, + "learning_rate": 1.8505241651867543e-05, + "loss": 0.9313, + "step": 8532 + }, + { + "epoch": 0.2, + "grad_norm": 2.3023488829836727, + "learning_rate": 1.8504840316367692e-05, + "loss": 1.0931, + "step": 8533 + }, + { + "epoch": 0.2, + "grad_norm": 2.0321073982940736, + "learning_rate": 1.8504438931349967e-05, + "loss": 1.0447, + "step": 8534 + }, + { + "epoch": 0.2, + "grad_norm": 3.2594347385735176, + "learning_rate": 1.8504037496816708e-05, + "loss": 1.0745, + "step": 8535 + }, + { + "epoch": 0.2, + "grad_norm": 1.8969324502894072, + "learning_rate": 1.8503636012770247e-05, + "loss": 0.9852, + "step": 8536 + }, + { + "epoch": 0.2, + "grad_norm": 2.1031311095690657, + "learning_rate": 1.8503234479212924e-05, + "loss": 1.0437, + "step": 8537 + }, + { + "epoch": 0.2, + "grad_norm": 1.9051754356560728, + "learning_rate": 1.850283289614708e-05, + "loss": 1.0522, + "step": 8538 + }, + { + "epoch": 0.2, + "grad_norm": 2.429527649266826, + "learning_rate": 1.850243126357505e-05, + "loss": 1.0191, + "step": 8539 + }, + { + "epoch": 0.2, + "grad_norm": 2.3085321598300923, + "learning_rate": 1.8502029581499173e-05, + "loss": 1.0685, + "step": 8540 + }, + { + "epoch": 0.2, + "grad_norm": 1.8123632324962995, + "learning_rate": 1.850162784992179e-05, + "loss": 1.0685, + "step": 8541 + }, + { + "epoch": 0.2, + "grad_norm": 1.8466549359301376, + "learning_rate": 1.8501226068845238e-05, + "loss": 1.1471, + "step": 8542 + }, + { + "epoch": 0.2, + "grad_norm": 1.8439238623397862, + "learning_rate": 1.8500824238271855e-05, + "loss": 1.0738, + "step": 8543 + }, + { + "epoch": 0.2, + "grad_norm": 2.392322527311725, + "learning_rate": 1.8500422358203982e-05, + "loss": 0.9487, + "step": 8544 + }, + { + "epoch": 0.2, + "grad_norm": 2.1700319848935634, + "learning_rate": 1.850002042864396e-05, + "loss": 1.0198, + "step": 8545 + }, + { + "epoch": 0.2, + "grad_norm": 2.1605325090064857, + "learning_rate": 1.849961844959413e-05, + "loss": 1.1045, + "step": 8546 + }, + { + "epoch": 0.2, + "grad_norm": 1.8677334779042805, + "learning_rate": 1.8499216421056823e-05, + "loss": 1.0222, + "step": 8547 + }, + { + "epoch": 0.2, + "grad_norm": 2.217093380191071, + "learning_rate": 1.8498814343034396e-05, + "loss": 1.1005, + "step": 8548 + }, + { + "epoch": 0.2, + "grad_norm": 2.6000176711159657, + "learning_rate": 1.8498412215529175e-05, + "loss": 1.0318, + "step": 8549 + }, + { + "epoch": 0.2, + "grad_norm": 1.1635756062392504, + "learning_rate": 1.849801003854351e-05, + "loss": 0.9892, + "step": 8550 + }, + { + "epoch": 0.2, + "grad_norm": 2.2874037753074097, + "learning_rate": 1.849760781207974e-05, + "loss": 1.0377, + "step": 8551 + }, + { + "epoch": 0.2, + "grad_norm": 1.939030780526076, + "learning_rate": 1.8497205536140207e-05, + "loss": 1.085, + "step": 8552 + }, + { + "epoch": 0.2, + "grad_norm": 1.9048954931238191, + "learning_rate": 1.8496803210727252e-05, + "loss": 1.0225, + "step": 8553 + }, + { + "epoch": 0.2, + "grad_norm": 1.9805649187013679, + "learning_rate": 1.8496400835843223e-05, + "loss": 1.0402, + "step": 8554 + }, + { + "epoch": 0.2, + "grad_norm": 2.0494298516113085, + "learning_rate": 1.849599841149045e-05, + "loss": 1.0217, + "step": 8555 + }, + { + "epoch": 0.2, + "grad_norm": 1.8487550374657864, + "learning_rate": 1.8495595937671293e-05, + "loss": 1.1552, + "step": 8556 + }, + { + "epoch": 0.2, + "grad_norm": 1.9944579017296107, + "learning_rate": 1.8495193414388084e-05, + "loss": 1.0368, + "step": 8557 + }, + { + "epoch": 0.2, + "grad_norm": 2.982616733432086, + "learning_rate": 1.849479084164317e-05, + "loss": 1.0403, + "step": 8558 + }, + { + "epoch": 0.2, + "grad_norm": 2.457380812819132, + "learning_rate": 1.8494388219438892e-05, + "loss": 0.9981, + "step": 8559 + }, + { + "epoch": 0.2, + "grad_norm": 2.0938672745255595, + "learning_rate": 1.8493985547777598e-05, + "loss": 1.1086, + "step": 8560 + }, + { + "epoch": 0.2, + "grad_norm": 1.911316229454623, + "learning_rate": 1.8493582826661628e-05, + "loss": 1.1289, + "step": 8561 + }, + { + "epoch": 0.2, + "grad_norm": 1.098880805954009, + "learning_rate": 1.8493180056093334e-05, + "loss": 0.9885, + "step": 8562 + }, + { + "epoch": 0.2, + "grad_norm": 2.097850873944988, + "learning_rate": 1.8492777236075056e-05, + "loss": 0.8991, + "step": 8563 + }, + { + "epoch": 0.2, + "grad_norm": 2.0079087363503807, + "learning_rate": 1.849237436660914e-05, + "loss": 0.9669, + "step": 8564 + }, + { + "epoch": 0.2, + "grad_norm": 1.9592750420794762, + "learning_rate": 1.849197144769793e-05, + "loss": 1.0594, + "step": 8565 + }, + { + "epoch": 0.2, + "grad_norm": 2.309349751475599, + "learning_rate": 1.8491568479343775e-05, + "loss": 1.0174, + "step": 8566 + }, + { + "epoch": 0.2, + "grad_norm": 2.0446777073305045, + "learning_rate": 1.849116546154902e-05, + "loss": 1.1995, + "step": 8567 + }, + { + "epoch": 0.2, + "grad_norm": 1.7732419095699337, + "learning_rate": 1.849076239431601e-05, + "loss": 0.9947, + "step": 8568 + }, + { + "epoch": 0.2, + "grad_norm": 1.8708850106055095, + "learning_rate": 1.8490359277647096e-05, + "loss": 1.0271, + "step": 8569 + }, + { + "epoch": 0.2, + "grad_norm": 2.044758526131099, + "learning_rate": 1.8489956111544624e-05, + "loss": 1.01, + "step": 8570 + }, + { + "epoch": 0.2, + "grad_norm": 1.854991936215177, + "learning_rate": 1.8489552896010932e-05, + "loss": 0.931, + "step": 8571 + }, + { + "epoch": 0.2, + "grad_norm": 2.16370944429388, + "learning_rate": 1.8489149631048378e-05, + "loss": 1.0877, + "step": 8572 + }, + { + "epoch": 0.2, + "grad_norm": 2.3015322636437805, + "learning_rate": 1.848874631665931e-05, + "loss": 1.1201, + "step": 8573 + }, + { + "epoch": 0.2, + "grad_norm": 2.1196595725561123, + "learning_rate": 1.8488342952846074e-05, + "loss": 1.017, + "step": 8574 + }, + { + "epoch": 0.2, + "grad_norm": 2.4751567144943136, + "learning_rate": 1.8487939539611016e-05, + "loss": 1.0941, + "step": 8575 + }, + { + "epoch": 0.2, + "grad_norm": 2.111011985203064, + "learning_rate": 1.8487536076956488e-05, + "loss": 1.0573, + "step": 8576 + }, + { + "epoch": 0.2, + "grad_norm": 1.9097755181652627, + "learning_rate": 1.8487132564884835e-05, + "loss": 0.9787, + "step": 8577 + }, + { + "epoch": 0.2, + "grad_norm": 2.1692054400179606, + "learning_rate": 1.848672900339841e-05, + "loss": 1.1681, + "step": 8578 + }, + { + "epoch": 0.2, + "grad_norm": 1.8918012696750157, + "learning_rate": 1.848632539249957e-05, + "loss": 1.0219, + "step": 8579 + }, + { + "epoch": 0.2, + "grad_norm": 1.1816551979994843, + "learning_rate": 1.8485921732190644e-05, + "loss": 0.9762, + "step": 8580 + }, + { + "epoch": 0.2, + "grad_norm": 1.9182798635621563, + "learning_rate": 1.8485518022474004e-05, + "loss": 1.0139, + "step": 8581 + }, + { + "epoch": 0.2, + "grad_norm": 2.7204160653543905, + "learning_rate": 1.8485114263351988e-05, + "loss": 1.134, + "step": 8582 + }, + { + "epoch": 0.2, + "grad_norm": 2.4815447490233593, + "learning_rate": 1.8484710454826952e-05, + "loss": 1.0762, + "step": 8583 + }, + { + "epoch": 0.2, + "grad_norm": 2.3466184297941006, + "learning_rate": 1.8484306596901243e-05, + "loss": 1.1404, + "step": 8584 + }, + { + "epoch": 0.2, + "grad_norm": 2.0806041400543185, + "learning_rate": 1.8483902689577217e-05, + "loss": 1.0181, + "step": 8585 + }, + { + "epoch": 0.2, + "grad_norm": 5.061165618480096, + "learning_rate": 1.8483498732857224e-05, + "loss": 0.873, + "step": 8586 + }, + { + "epoch": 0.2, + "grad_norm": 1.1213297551908745, + "learning_rate": 1.8483094726743613e-05, + "loss": 0.9666, + "step": 8587 + }, + { + "epoch": 0.2, + "grad_norm": 2.5821122738941757, + "learning_rate": 1.8482690671238738e-05, + "loss": 0.8856, + "step": 8588 + }, + { + "epoch": 0.2, + "grad_norm": 2.1591707414470216, + "learning_rate": 1.8482286566344957e-05, + "loss": 1.0494, + "step": 8589 + }, + { + "epoch": 0.2, + "grad_norm": 1.7772693416718917, + "learning_rate": 1.8481882412064616e-05, + "loss": 1.072, + "step": 8590 + }, + { + "epoch": 0.2, + "grad_norm": 2.344901152973894, + "learning_rate": 1.848147820840007e-05, + "loss": 1.0605, + "step": 8591 + }, + { + "epoch": 0.2, + "grad_norm": 2.8028082183254823, + "learning_rate": 1.8481073955353673e-05, + "loss": 0.9485, + "step": 8592 + }, + { + "epoch": 0.2, + "grad_norm": 2.297097975018855, + "learning_rate": 1.8480669652927778e-05, + "loss": 1.1227, + "step": 8593 + }, + { + "epoch": 0.2, + "grad_norm": 2.3787093942369246, + "learning_rate": 1.848026530112474e-05, + "loss": 0.9419, + "step": 8594 + }, + { + "epoch": 0.2, + "grad_norm": 1.911670115391637, + "learning_rate": 1.8479860899946912e-05, + "loss": 0.947, + "step": 8595 + }, + { + "epoch": 0.2, + "grad_norm": 1.9543578870061145, + "learning_rate": 1.847945644939665e-05, + "loss": 0.879, + "step": 8596 + }, + { + "epoch": 0.2, + "grad_norm": 2.2324829697071182, + "learning_rate": 1.8479051949476307e-05, + "loss": 1.0427, + "step": 8597 + }, + { + "epoch": 0.2, + "grad_norm": 1.7532517274156718, + "learning_rate": 1.847864740018824e-05, + "loss": 1.116, + "step": 8598 + }, + { + "epoch": 0.2, + "grad_norm": 1.0851697883832516, + "learning_rate": 1.8478242801534803e-05, + "loss": 1.0576, + "step": 8599 + }, + { + "epoch": 0.2, + "grad_norm": 1.1459041693901528, + "learning_rate": 1.8477838153518354e-05, + "loss": 1.0012, + "step": 8600 + }, + { + "epoch": 0.2, + "grad_norm": 2.071149403631231, + "learning_rate": 1.8477433456141242e-05, + "loss": 1.1156, + "step": 8601 + }, + { + "epoch": 0.2, + "grad_norm": 2.356744115871879, + "learning_rate": 1.8477028709405834e-05, + "loss": 1.0426, + "step": 8602 + }, + { + "epoch": 0.2, + "grad_norm": 1.9719178533941757, + "learning_rate": 1.8476623913314476e-05, + "loss": 1.0074, + "step": 8603 + }, + { + "epoch": 0.2, + "grad_norm": 2.4277079383019875, + "learning_rate": 1.8476219067869534e-05, + "loss": 1.0942, + "step": 8604 + }, + { + "epoch": 0.2, + "grad_norm": 1.9174753759232497, + "learning_rate": 1.847581417307336e-05, + "loss": 0.9383, + "step": 8605 + }, + { + "epoch": 0.2, + "grad_norm": 2.027969665114006, + "learning_rate": 1.8475409228928314e-05, + "loss": 1.0083, + "step": 8606 + }, + { + "epoch": 0.2, + "grad_norm": 3.605310794941699, + "learning_rate": 1.847500423543675e-05, + "loss": 1.0913, + "step": 8607 + }, + { + "epoch": 0.2, + "grad_norm": 1.1109977806915035, + "learning_rate": 1.847459919260103e-05, + "loss": 0.9984, + "step": 8608 + }, + { + "epoch": 0.2, + "grad_norm": 1.958776241163155, + "learning_rate": 1.8474194100423508e-05, + "loss": 1.0202, + "step": 8609 + }, + { + "epoch": 0.2, + "grad_norm": 1.94812429652222, + "learning_rate": 1.8473788958906548e-05, + "loss": 0.9961, + "step": 8610 + }, + { + "epoch": 0.2, + "grad_norm": 2.2202096978457844, + "learning_rate": 1.8473383768052504e-05, + "loss": 0.8623, + "step": 8611 + }, + { + "epoch": 0.2, + "grad_norm": 4.210004135561478, + "learning_rate": 1.8472978527863737e-05, + "loss": 0.991, + "step": 8612 + }, + { + "epoch": 0.2, + "grad_norm": 2.44008456637128, + "learning_rate": 1.8472573238342606e-05, + "loss": 1.1004, + "step": 8613 + }, + { + "epoch": 0.2, + "grad_norm": 2.175482746584197, + "learning_rate": 1.8472167899491472e-05, + "loss": 1.1638, + "step": 8614 + }, + { + "epoch": 0.2, + "grad_norm": 1.977370820129584, + "learning_rate": 1.8471762511312693e-05, + "loss": 1.1341, + "step": 8615 + }, + { + "epoch": 0.2, + "grad_norm": 2.255967608658602, + "learning_rate": 1.8471357073808635e-05, + "loss": 1.1027, + "step": 8616 + }, + { + "epoch": 0.2, + "grad_norm": 2.007010596684786, + "learning_rate": 1.8470951586981648e-05, + "loss": 1.2259, + "step": 8617 + }, + { + "epoch": 0.2, + "grad_norm": 1.9216478379984592, + "learning_rate": 1.8470546050834104e-05, + "loss": 1.0641, + "step": 8618 + }, + { + "epoch": 0.2, + "grad_norm": 1.797069698973046, + "learning_rate": 1.8470140465368355e-05, + "loss": 1.1286, + "step": 8619 + }, + { + "epoch": 0.2, + "grad_norm": 1.9366571548068408, + "learning_rate": 1.8469734830586773e-05, + "loss": 1.1191, + "step": 8620 + }, + { + "epoch": 0.2, + "grad_norm": 2.388952101277511, + "learning_rate": 1.8469329146491708e-05, + "loss": 0.9466, + "step": 8621 + }, + { + "epoch": 0.2, + "grad_norm": 3.4333043289074014, + "learning_rate": 1.846892341308553e-05, + "loss": 0.9692, + "step": 8622 + }, + { + "epoch": 0.2, + "grad_norm": 2.468549325451089, + "learning_rate": 1.8468517630370594e-05, + "loss": 0.9896, + "step": 8623 + }, + { + "epoch": 0.2, + "grad_norm": 2.178574581102606, + "learning_rate": 1.846811179834927e-05, + "loss": 1.2275, + "step": 8624 + }, + { + "epoch": 0.2, + "grad_norm": 2.1935391536052573, + "learning_rate": 1.846770591702392e-05, + "loss": 1.0915, + "step": 8625 + }, + { + "epoch": 0.2, + "grad_norm": 2.143700162539214, + "learning_rate": 1.8467299986396903e-05, + "loss": 1.1356, + "step": 8626 + }, + { + "epoch": 0.2, + "grad_norm": 2.1926474902103696, + "learning_rate": 1.8466894006470588e-05, + "loss": 0.9293, + "step": 8627 + }, + { + "epoch": 0.2, + "grad_norm": 1.988195805901146, + "learning_rate": 1.846648797724733e-05, + "loss": 0.9721, + "step": 8628 + }, + { + "epoch": 0.2, + "grad_norm": 1.1997130067402475, + "learning_rate": 1.8466081898729507e-05, + "loss": 1.0192, + "step": 8629 + }, + { + "epoch": 0.2, + "grad_norm": 2.3735731246813443, + "learning_rate": 1.8465675770919468e-05, + "loss": 1.1244, + "step": 8630 + }, + { + "epoch": 0.2, + "grad_norm": 2.150326826728253, + "learning_rate": 1.8465269593819586e-05, + "loss": 0.9523, + "step": 8631 + }, + { + "epoch": 0.2, + "grad_norm": 2.079931726810989, + "learning_rate": 1.8464863367432225e-05, + "loss": 0.9734, + "step": 8632 + }, + { + "epoch": 0.2, + "grad_norm": 2.292176987463658, + "learning_rate": 1.8464457091759748e-05, + "loss": 1.0195, + "step": 8633 + }, + { + "epoch": 0.2, + "grad_norm": 1.9908074531548725, + "learning_rate": 1.8464050766804526e-05, + "loss": 1.0731, + "step": 8634 + }, + { + "epoch": 0.2, + "grad_norm": 2.0945031891724915, + "learning_rate": 1.8463644392568916e-05, + "loss": 0.9106, + "step": 8635 + }, + { + "epoch": 0.2, + "grad_norm": 2.2507696610024617, + "learning_rate": 1.846323796905529e-05, + "loss": 1.1011, + "step": 8636 + }, + { + "epoch": 0.2, + "grad_norm": 2.6097639513695556, + "learning_rate": 1.8462831496266013e-05, + "loss": 1.1018, + "step": 8637 + }, + { + "epoch": 0.2, + "grad_norm": 2.7179904651308924, + "learning_rate": 1.8462424974203452e-05, + "loss": 0.9969, + "step": 8638 + }, + { + "epoch": 0.2, + "grad_norm": 2.1202935791612116, + "learning_rate": 1.8462018402869977e-05, + "loss": 1.0288, + "step": 8639 + }, + { + "epoch": 0.2, + "grad_norm": 2.0248575189512197, + "learning_rate": 1.8461611782267947e-05, + "loss": 0.9864, + "step": 8640 + }, + { + "epoch": 0.2, + "grad_norm": 2.132492474918313, + "learning_rate": 1.8461205112399736e-05, + "loss": 1.0334, + "step": 8641 + }, + { + "epoch": 0.2, + "grad_norm": 2.4244511564939324, + "learning_rate": 1.8460798393267713e-05, + "loss": 0.9258, + "step": 8642 + }, + { + "epoch": 0.2, + "grad_norm": 2.4018029523033744, + "learning_rate": 1.846039162487424e-05, + "loss": 0.9895, + "step": 8643 + }, + { + "epoch": 0.2, + "grad_norm": 1.9643029163963175, + "learning_rate": 1.8459984807221686e-05, + "loss": 1.1444, + "step": 8644 + }, + { + "epoch": 0.2, + "grad_norm": 1.976530333608748, + "learning_rate": 1.845957794031243e-05, + "loss": 0.9477, + "step": 8645 + }, + { + "epoch": 0.2, + "grad_norm": 2.299273952263129, + "learning_rate": 1.8459171024148826e-05, + "loss": 0.944, + "step": 8646 + }, + { + "epoch": 0.2, + "grad_norm": 2.2911122656177247, + "learning_rate": 1.845876405873325e-05, + "loss": 1.113, + "step": 8647 + }, + { + "epoch": 0.2, + "grad_norm": 2.713515682473203, + "learning_rate": 1.8458357044068073e-05, + "loss": 1.0427, + "step": 8648 + }, + { + "epoch": 0.2, + "grad_norm": 2.2570498955494696, + "learning_rate": 1.8457949980155666e-05, + "loss": 1.1257, + "step": 8649 + }, + { + "epoch": 0.2, + "grad_norm": 1.8863534972417324, + "learning_rate": 1.8457542866998394e-05, + "loss": 1.0541, + "step": 8650 + }, + { + "epoch": 0.2, + "grad_norm": 1.8767564113009079, + "learning_rate": 1.8457135704598634e-05, + "loss": 1.0627, + "step": 8651 + }, + { + "epoch": 0.2, + "grad_norm": 2.350497019060226, + "learning_rate": 1.8456728492958748e-05, + "loss": 1.018, + "step": 8652 + }, + { + "epoch": 0.2, + "grad_norm": 2.122256713519703, + "learning_rate": 1.845632123208111e-05, + "loss": 1.062, + "step": 8653 + }, + { + "epoch": 0.2, + "grad_norm": 2.0505388601427113, + "learning_rate": 1.8455913921968095e-05, + "loss": 1.0691, + "step": 8654 + }, + { + "epoch": 0.2, + "grad_norm": 2.8923705977666816, + "learning_rate": 1.845550656262207e-05, + "loss": 1.0763, + "step": 8655 + }, + { + "epoch": 0.2, + "grad_norm": 1.8858200871553406, + "learning_rate": 1.8455099154045416e-05, + "loss": 1.1391, + "step": 8656 + }, + { + "epoch": 0.2, + "grad_norm": 1.1366961077929054, + "learning_rate": 1.8454691696240493e-05, + "loss": 0.9655, + "step": 8657 + }, + { + "epoch": 0.2, + "grad_norm": 2.014844981668709, + "learning_rate": 1.8454284189209674e-05, + "loss": 1.0136, + "step": 8658 + }, + { + "epoch": 0.2, + "grad_norm": 2.3598393548154664, + "learning_rate": 1.8453876632955342e-05, + "loss": 1.0059, + "step": 8659 + }, + { + "epoch": 0.2, + "grad_norm": 2.399920631719317, + "learning_rate": 1.845346902747986e-05, + "loss": 0.9833, + "step": 8660 + }, + { + "epoch": 0.2, + "grad_norm": 1.1081155117598782, + "learning_rate": 1.8453061372785605e-05, + "loss": 1.0069, + "step": 8661 + }, + { + "epoch": 0.2, + "grad_norm": 2.3488519456311496, + "learning_rate": 1.8452653668874954e-05, + "loss": 1.0557, + "step": 8662 + }, + { + "epoch": 0.2, + "grad_norm": 1.8550806677431975, + "learning_rate": 1.8452245915750276e-05, + "loss": 1.0339, + "step": 8663 + }, + { + "epoch": 0.2, + "grad_norm": 2.0806152051096802, + "learning_rate": 1.8451838113413946e-05, + "loss": 1.1018, + "step": 8664 + }, + { + "epoch": 0.2, + "grad_norm": 2.4117270618427713, + "learning_rate": 1.8451430261868336e-05, + "loss": 1.0825, + "step": 8665 + }, + { + "epoch": 0.2, + "grad_norm": 1.9453708255021152, + "learning_rate": 1.8451022361115828e-05, + "loss": 0.9883, + "step": 8666 + }, + { + "epoch": 0.2, + "grad_norm": 2.017304141984847, + "learning_rate": 1.845061441115879e-05, + "loss": 1.0449, + "step": 8667 + }, + { + "epoch": 0.2, + "grad_norm": 2.4936749893039387, + "learning_rate": 1.84502064119996e-05, + "loss": 1.1288, + "step": 8668 + }, + { + "epoch": 0.2, + "grad_norm": 1.9424836332527349, + "learning_rate": 1.844979836364063e-05, + "loss": 1.1514, + "step": 8669 + }, + { + "epoch": 0.2, + "grad_norm": 4.601946084124233, + "learning_rate": 1.8449390266084263e-05, + "loss": 0.9945, + "step": 8670 + }, + { + "epoch": 0.2, + "grad_norm": 1.165345381511075, + "learning_rate": 1.8448982119332868e-05, + "loss": 0.9698, + "step": 8671 + }, + { + "epoch": 0.2, + "grad_norm": 2.4294833227057655, + "learning_rate": 1.8448573923388825e-05, + "loss": 1.1712, + "step": 8672 + }, + { + "epoch": 0.2, + "grad_norm": 2.0027693620837725, + "learning_rate": 1.844816567825451e-05, + "loss": 1.1147, + "step": 8673 + }, + { + "epoch": 0.2, + "grad_norm": 3.5005141287349684, + "learning_rate": 1.8447757383932302e-05, + "loss": 1.023, + "step": 8674 + }, + { + "epoch": 0.2, + "grad_norm": 1.998937736588043, + "learning_rate": 1.8447349040424572e-05, + "loss": 1.0508, + "step": 8675 + }, + { + "epoch": 0.2, + "grad_norm": 1.953133268696448, + "learning_rate": 1.8446940647733705e-05, + "loss": 1.147, + "step": 8676 + }, + { + "epoch": 0.2, + "grad_norm": 2.1046687860252513, + "learning_rate": 1.8446532205862072e-05, + "loss": 0.9714, + "step": 8677 + }, + { + "epoch": 0.2, + "grad_norm": 1.1384599343862503, + "learning_rate": 1.8446123714812055e-05, + "loss": 0.9876, + "step": 8678 + }, + { + "epoch": 0.2, + "grad_norm": 1.8913251414411978, + "learning_rate": 1.8445715174586033e-05, + "loss": 0.9994, + "step": 8679 + }, + { + "epoch": 0.2, + "grad_norm": 1.0771418776187027, + "learning_rate": 1.8445306585186384e-05, + "loss": 0.9571, + "step": 8680 + }, + { + "epoch": 0.2, + "grad_norm": 1.085441348977846, + "learning_rate": 1.8444897946615486e-05, + "loss": 0.9127, + "step": 8681 + }, + { + "epoch": 0.2, + "grad_norm": 2.2631499614088955, + "learning_rate": 1.8444489258875718e-05, + "loss": 1.008, + "step": 8682 + }, + { + "epoch": 0.2, + "grad_norm": 1.0555164304946651, + "learning_rate": 1.844408052196946e-05, + "loss": 0.98, + "step": 8683 + }, + { + "epoch": 0.2, + "grad_norm": 2.3045315543246705, + "learning_rate": 1.8443671735899092e-05, + "loss": 1.0391, + "step": 8684 + }, + { + "epoch": 0.2, + "grad_norm": 2.0127134256366896, + "learning_rate": 1.8443262900666995e-05, + "loss": 1.0582, + "step": 8685 + }, + { + "epoch": 0.2, + "grad_norm": 2.17538068007982, + "learning_rate": 1.8442854016275545e-05, + "loss": 1.0671, + "step": 8686 + }, + { + "epoch": 0.2, + "grad_norm": 2.8441690407867144, + "learning_rate": 1.8442445082727127e-05, + "loss": 1.0757, + "step": 8687 + }, + { + "epoch": 0.2, + "grad_norm": 1.179416827940018, + "learning_rate": 1.8442036100024124e-05, + "loss": 1.0138, + "step": 8688 + }, + { + "epoch": 0.2, + "grad_norm": 1.9571113280942174, + "learning_rate": 1.8441627068168914e-05, + "loss": 1.0348, + "step": 8689 + }, + { + "epoch": 0.2, + "grad_norm": 2.099208360234754, + "learning_rate": 1.8441217987163874e-05, + "loss": 0.9473, + "step": 8690 + }, + { + "epoch": 0.2, + "grad_norm": 2.3325012027434817, + "learning_rate": 1.8440808857011392e-05, + "loss": 1.2059, + "step": 8691 + }, + { + "epoch": 0.2, + "grad_norm": 2.1217730403248494, + "learning_rate": 1.844039967771385e-05, + "loss": 1.0419, + "step": 8692 + }, + { + "epoch": 0.2, + "grad_norm": 2.131770786634533, + "learning_rate": 1.843999044927363e-05, + "loss": 1.0693, + "step": 8693 + }, + { + "epoch": 0.2, + "grad_norm": 1.9511865153732368, + "learning_rate": 1.843958117169311e-05, + "loss": 1.0284, + "step": 8694 + }, + { + "epoch": 0.2, + "grad_norm": 1.9057489031737809, + "learning_rate": 1.843917184497468e-05, + "loss": 1.0996, + "step": 8695 + }, + { + "epoch": 0.2, + "grad_norm": 2.0658876664336048, + "learning_rate": 1.843876246912072e-05, + "loss": 1.0994, + "step": 8696 + }, + { + "epoch": 0.2, + "grad_norm": 1.9763573280436877, + "learning_rate": 1.843835304413361e-05, + "loss": 1.0827, + "step": 8697 + }, + { + "epoch": 0.2, + "grad_norm": 1.1977042420374358, + "learning_rate": 1.8437943570015737e-05, + "loss": 0.9954, + "step": 8698 + }, + { + "epoch": 0.2, + "grad_norm": 1.9289309488430595, + "learning_rate": 1.843753404676949e-05, + "loss": 1.068, + "step": 8699 + }, + { + "epoch": 0.2, + "grad_norm": 2.2367503731654304, + "learning_rate": 1.8437124474397247e-05, + "loss": 1.0312, + "step": 8700 + }, + { + "epoch": 0.2, + "grad_norm": 1.862342224295397, + "learning_rate": 1.843671485290139e-05, + "loss": 1.1374, + "step": 8701 + }, + { + "epoch": 0.21, + "grad_norm": 3.9710762910611077, + "learning_rate": 1.843630518228431e-05, + "loss": 1.0551, + "step": 8702 + }, + { + "epoch": 0.21, + "grad_norm": 1.9019131605190482, + "learning_rate": 1.843589546254839e-05, + "loss": 1.0392, + "step": 8703 + }, + { + "epoch": 0.21, + "grad_norm": 2.1509592368185766, + "learning_rate": 1.8435485693696015e-05, + "loss": 1.072, + "step": 8704 + }, + { + "epoch": 0.21, + "grad_norm": 2.077195117024192, + "learning_rate": 1.8435075875729577e-05, + "loss": 1.1127, + "step": 8705 + }, + { + "epoch": 0.21, + "grad_norm": 1.1612855930652508, + "learning_rate": 1.843466600865145e-05, + "loss": 0.9453, + "step": 8706 + }, + { + "epoch": 0.21, + "grad_norm": 1.8910242307682268, + "learning_rate": 1.843425609246403e-05, + "loss": 1.0478, + "step": 8707 + }, + { + "epoch": 0.21, + "grad_norm": 1.941505812928527, + "learning_rate": 1.8433846127169704e-05, + "loss": 0.9395, + "step": 8708 + }, + { + "epoch": 0.21, + "grad_norm": 1.1299205386738715, + "learning_rate": 1.843343611277085e-05, + "loss": 0.9978, + "step": 8709 + }, + { + "epoch": 0.21, + "grad_norm": 2.249234222153186, + "learning_rate": 1.8433026049269866e-05, + "loss": 1.0291, + "step": 8710 + }, + { + "epoch": 0.21, + "grad_norm": 2.1626124643131908, + "learning_rate": 1.8432615936669133e-05, + "loss": 1.0251, + "step": 8711 + }, + { + "epoch": 0.21, + "grad_norm": 1.9439429994239947, + "learning_rate": 1.8432205774971036e-05, + "loss": 0.9596, + "step": 8712 + }, + { + "epoch": 0.21, + "grad_norm": 1.1701028494616263, + "learning_rate": 1.8431795564177975e-05, + "loss": 0.9835, + "step": 8713 + }, + { + "epoch": 0.21, + "grad_norm": 2.2430171136838783, + "learning_rate": 1.8431385304292324e-05, + "loss": 1.0418, + "step": 8714 + }, + { + "epoch": 0.21, + "grad_norm": 2.360389476865394, + "learning_rate": 1.8430974995316482e-05, + "loss": 1.1522, + "step": 8715 + }, + { + "epoch": 0.21, + "grad_norm": 2.0026195310161174, + "learning_rate": 1.8430564637252835e-05, + "loss": 1.0533, + "step": 8716 + }, + { + "epoch": 0.21, + "grad_norm": 2.11812434790187, + "learning_rate": 1.843015423010377e-05, + "loss": 1.0998, + "step": 8717 + }, + { + "epoch": 0.21, + "grad_norm": 2.170855886659676, + "learning_rate": 1.8429743773871678e-05, + "loss": 1.0628, + "step": 8718 + }, + { + "epoch": 0.21, + "grad_norm": 1.955287611717749, + "learning_rate": 1.842933326855895e-05, + "loss": 1.0083, + "step": 8719 + }, + { + "epoch": 0.21, + "grad_norm": 2.722292596760992, + "learning_rate": 1.842892271416797e-05, + "loss": 1.1189, + "step": 8720 + }, + { + "epoch": 0.21, + "grad_norm": 1.8404175776939737, + "learning_rate": 1.842851211070114e-05, + "loss": 1.1524, + "step": 8721 + }, + { + "epoch": 0.21, + "grad_norm": 1.9898269022581065, + "learning_rate": 1.8428101458160843e-05, + "loss": 1.1527, + "step": 8722 + }, + { + "epoch": 0.21, + "grad_norm": 2.109273825091077, + "learning_rate": 1.842769075654947e-05, + "loss": 1.0769, + "step": 8723 + }, + { + "epoch": 0.21, + "grad_norm": 2.481145236176716, + "learning_rate": 1.842728000586941e-05, + "loss": 0.9952, + "step": 8724 + }, + { + "epoch": 0.21, + "grad_norm": 2.094299124076038, + "learning_rate": 1.8426869206123062e-05, + "loss": 0.9729, + "step": 8725 + }, + { + "epoch": 0.21, + "grad_norm": 2.3428670572089727, + "learning_rate": 1.8426458357312814e-05, + "loss": 1.0134, + "step": 8726 + }, + { + "epoch": 0.21, + "grad_norm": 2.3284042139162646, + "learning_rate": 1.8426047459441053e-05, + "loss": 0.9397, + "step": 8727 + }, + { + "epoch": 0.21, + "grad_norm": 2.1061841253888836, + "learning_rate": 1.842563651251018e-05, + "loss": 1.0941, + "step": 8728 + }, + { + "epoch": 0.21, + "grad_norm": 2.159587535723061, + "learning_rate": 1.842522551652258e-05, + "loss": 1.0623, + "step": 8729 + }, + { + "epoch": 0.21, + "grad_norm": 2.3932853324026544, + "learning_rate": 1.8424814471480658e-05, + "loss": 1.0625, + "step": 8730 + }, + { + "epoch": 0.21, + "grad_norm": 3.499398279234311, + "learning_rate": 1.842440337738679e-05, + "loss": 1.0496, + "step": 8731 + }, + { + "epoch": 0.21, + "grad_norm": 2.099283399338845, + "learning_rate": 1.842399223424338e-05, + "loss": 0.9882, + "step": 8732 + }, + { + "epoch": 0.21, + "grad_norm": 1.937348275963882, + "learning_rate": 1.8423581042052823e-05, + "loss": 0.9965, + "step": 8733 + }, + { + "epoch": 0.21, + "grad_norm": 1.9949430154171774, + "learning_rate": 1.842316980081751e-05, + "loss": 1.038, + "step": 8734 + }, + { + "epoch": 0.21, + "grad_norm": 2.0336765859482897, + "learning_rate": 1.8422758510539837e-05, + "loss": 0.9881, + "step": 8735 + }, + { + "epoch": 0.21, + "grad_norm": 1.9947743546066263, + "learning_rate": 1.8422347171222194e-05, + "loss": 0.8932, + "step": 8736 + }, + { + "epoch": 0.21, + "grad_norm": 2.1747528692220772, + "learning_rate": 1.8421935782866982e-05, + "loss": 1.1444, + "step": 8737 + }, + { + "epoch": 0.21, + "grad_norm": 2.510639668617271, + "learning_rate": 1.842152434547659e-05, + "loss": 1.0922, + "step": 8738 + }, + { + "epoch": 0.21, + "grad_norm": 3.2911888147735366, + "learning_rate": 1.8421112859053417e-05, + "loss": 0.9666, + "step": 8739 + }, + { + "epoch": 0.21, + "grad_norm": 1.9058652366380726, + "learning_rate": 1.8420701323599865e-05, + "loss": 1.0891, + "step": 8740 + }, + { + "epoch": 0.21, + "grad_norm": 1.1310170885986193, + "learning_rate": 1.8420289739118318e-05, + "loss": 0.9325, + "step": 8741 + }, + { + "epoch": 0.21, + "grad_norm": 1.8466614236135277, + "learning_rate": 1.841987810561118e-05, + "loss": 1.0036, + "step": 8742 + }, + { + "epoch": 0.21, + "grad_norm": 1.9177742703962708, + "learning_rate": 1.8419466423080847e-05, + "loss": 1.0364, + "step": 8743 + }, + { + "epoch": 0.21, + "grad_norm": 2.2661457763751196, + "learning_rate": 1.8419054691529715e-05, + "loss": 1.1136, + "step": 8744 + }, + { + "epoch": 0.21, + "grad_norm": 2.1419968385942445, + "learning_rate": 1.841864291096018e-05, + "loss": 1.1184, + "step": 8745 + }, + { + "epoch": 0.21, + "grad_norm": 2.0575027065086204, + "learning_rate": 1.841823108137464e-05, + "loss": 1.1185, + "step": 8746 + }, + { + "epoch": 0.21, + "grad_norm": 1.170750313223251, + "learning_rate": 1.8417819202775497e-05, + "loss": 0.9819, + "step": 8747 + }, + { + "epoch": 0.21, + "grad_norm": 1.9502937255530512, + "learning_rate": 1.841740727516514e-05, + "loss": 1.1687, + "step": 8748 + }, + { + "epoch": 0.21, + "grad_norm": 2.2020064927767065, + "learning_rate": 1.8416995298545978e-05, + "loss": 1.1281, + "step": 8749 + }, + { + "epoch": 0.21, + "grad_norm": 2.220829050474543, + "learning_rate": 1.84165832729204e-05, + "loss": 1.0389, + "step": 8750 + }, + { + "epoch": 0.21, + "grad_norm": 2.5728230004653496, + "learning_rate": 1.8416171198290812e-05, + "loss": 0.987, + "step": 8751 + }, + { + "epoch": 0.21, + "grad_norm": 2.0651239960270855, + "learning_rate": 1.8415759074659614e-05, + "loss": 1.1876, + "step": 8752 + }, + { + "epoch": 0.21, + "grad_norm": 2.050513003948278, + "learning_rate": 1.8415346902029196e-05, + "loss": 0.9097, + "step": 8753 + }, + { + "epoch": 0.21, + "grad_norm": 2.306267010594065, + "learning_rate": 1.8414934680401968e-05, + "loss": 0.9956, + "step": 8754 + }, + { + "epoch": 0.21, + "grad_norm": 2.229410863350565, + "learning_rate": 1.8414522409780328e-05, + "loss": 1.0238, + "step": 8755 + }, + { + "epoch": 0.21, + "grad_norm": 2.1308658461642933, + "learning_rate": 1.8414110090166672e-05, + "loss": 1.1493, + "step": 8756 + }, + { + "epoch": 0.21, + "grad_norm": 2.3555110767263945, + "learning_rate": 1.84136977215634e-05, + "loss": 1.1294, + "step": 8757 + }, + { + "epoch": 0.21, + "grad_norm": 2.020350834213785, + "learning_rate": 1.8413285303972922e-05, + "loss": 1.0185, + "step": 8758 + }, + { + "epoch": 0.21, + "grad_norm": 2.210241124992578, + "learning_rate": 1.8412872837397627e-05, + "loss": 1.0714, + "step": 8759 + }, + { + "epoch": 0.21, + "grad_norm": 2.110000994963363, + "learning_rate": 1.841246032183993e-05, + "loss": 0.9836, + "step": 8760 + }, + { + "epoch": 0.21, + "grad_norm": 2.1611547376328097, + "learning_rate": 1.8412047757302222e-05, + "loss": 1.0394, + "step": 8761 + }, + { + "epoch": 0.21, + "grad_norm": 2.1454621279970043, + "learning_rate": 1.8411635143786906e-05, + "loss": 0.9874, + "step": 8762 + }, + { + "epoch": 0.21, + "grad_norm": 2.056738386746907, + "learning_rate": 1.841122248129639e-05, + "loss": 1.0475, + "step": 8763 + }, + { + "epoch": 0.21, + "grad_norm": 2.140195902746497, + "learning_rate": 1.8410809769833075e-05, + "loss": 1.0448, + "step": 8764 + }, + { + "epoch": 0.21, + "grad_norm": 2.0664169824137626, + "learning_rate": 1.8410397009399363e-05, + "loss": 1.0566, + "step": 8765 + }, + { + "epoch": 0.21, + "grad_norm": 2.007675619703355, + "learning_rate": 1.8409984199997653e-05, + "loss": 0.9804, + "step": 8766 + }, + { + "epoch": 0.21, + "grad_norm": 2.070958323108862, + "learning_rate": 1.8409571341630358e-05, + "loss": 1.0812, + "step": 8767 + }, + { + "epoch": 0.21, + "grad_norm": 1.9444850171615493, + "learning_rate": 1.8409158434299872e-05, + "loss": 0.9759, + "step": 8768 + }, + { + "epoch": 0.21, + "grad_norm": 2.35661516415307, + "learning_rate": 1.8408745478008602e-05, + "loss": 1.0468, + "step": 8769 + }, + { + "epoch": 0.21, + "grad_norm": 1.1778734286777253, + "learning_rate": 1.840833247275896e-05, + "loss": 0.9802, + "step": 8770 + }, + { + "epoch": 0.21, + "grad_norm": 1.9023050659474388, + "learning_rate": 1.8407919418553337e-05, + "loss": 1.1042, + "step": 8771 + }, + { + "epoch": 0.21, + "grad_norm": 1.876183291231635, + "learning_rate": 1.840750631539415e-05, + "loss": 1.0337, + "step": 8772 + }, + { + "epoch": 0.21, + "grad_norm": 2.0849022948618954, + "learning_rate": 1.84070931632838e-05, + "loss": 1.04, + "step": 8773 + }, + { + "epoch": 0.21, + "grad_norm": 2.0901991005064935, + "learning_rate": 1.840667996222469e-05, + "loss": 0.9574, + "step": 8774 + }, + { + "epoch": 0.21, + "grad_norm": 1.9533225230235083, + "learning_rate": 1.8406266712219225e-05, + "loss": 0.9927, + "step": 8775 + }, + { + "epoch": 0.21, + "grad_norm": 2.0398152215108496, + "learning_rate": 1.840585341326982e-05, + "loss": 1.1315, + "step": 8776 + }, + { + "epoch": 0.21, + "grad_norm": 2.089371979352452, + "learning_rate": 1.840544006537887e-05, + "loss": 1.045, + "step": 8777 + }, + { + "epoch": 0.21, + "grad_norm": 1.9370719473831388, + "learning_rate": 1.840502666854879e-05, + "loss": 1.0627, + "step": 8778 + }, + { + "epoch": 0.21, + "grad_norm": 2.146051915628594, + "learning_rate": 1.840461322278198e-05, + "loss": 0.9275, + "step": 8779 + }, + { + "epoch": 0.21, + "grad_norm": 2.3221680944133016, + "learning_rate": 1.8404199728080854e-05, + "loss": 1.2166, + "step": 8780 + }, + { + "epoch": 0.21, + "grad_norm": 1.9803124763268825, + "learning_rate": 1.8403786184447818e-05, + "loss": 1.0787, + "step": 8781 + }, + { + "epoch": 0.21, + "grad_norm": 1.1209108374990395, + "learning_rate": 1.8403372591885272e-05, + "loss": 0.9725, + "step": 8782 + }, + { + "epoch": 0.21, + "grad_norm": 1.1332764107262787, + "learning_rate": 1.8402958950395637e-05, + "loss": 1.0544, + "step": 8783 + }, + { + "epoch": 0.21, + "grad_norm": 2.088174988595515, + "learning_rate": 1.8402545259981312e-05, + "loss": 1.0922, + "step": 8784 + }, + { + "epoch": 0.21, + "grad_norm": 2.0354246401408966, + "learning_rate": 1.8402131520644708e-05, + "loss": 1.103, + "step": 8785 + }, + { + "epoch": 0.21, + "grad_norm": 1.983511849893398, + "learning_rate": 1.8401717732388234e-05, + "loss": 1.0649, + "step": 8786 + }, + { + "epoch": 0.21, + "grad_norm": 1.2404430076879205, + "learning_rate": 1.8401303895214303e-05, + "loss": 1.0067, + "step": 8787 + }, + { + "epoch": 0.21, + "grad_norm": 1.9068794693369135, + "learning_rate": 1.8400890009125314e-05, + "loss": 1.0145, + "step": 8788 + }, + { + "epoch": 0.21, + "grad_norm": 2.312958793862877, + "learning_rate": 1.8400476074123686e-05, + "loss": 1.059, + "step": 8789 + }, + { + "epoch": 0.21, + "grad_norm": 1.8135661557292404, + "learning_rate": 1.840006209021183e-05, + "loss": 0.904, + "step": 8790 + }, + { + "epoch": 0.21, + "grad_norm": 2.2470337191465637, + "learning_rate": 1.839964805739215e-05, + "loss": 1.0289, + "step": 8791 + }, + { + "epoch": 0.21, + "grad_norm": 2.3031221681832035, + "learning_rate": 1.8399233975667062e-05, + "loss": 1.0603, + "step": 8792 + }, + { + "epoch": 0.21, + "grad_norm": 2.223659993413088, + "learning_rate": 1.8398819845038973e-05, + "loss": 1.1552, + "step": 8793 + }, + { + "epoch": 0.21, + "grad_norm": 1.9931648827718853, + "learning_rate": 1.83984056655103e-05, + "loss": 1.0945, + "step": 8794 + }, + { + "epoch": 0.21, + "grad_norm": 2.6261856583521954, + "learning_rate": 1.8397991437083445e-05, + "loss": 1.0842, + "step": 8795 + }, + { + "epoch": 0.21, + "grad_norm": 2.2919558868410417, + "learning_rate": 1.8397577159760828e-05, + "loss": 1.0058, + "step": 8796 + }, + { + "epoch": 0.21, + "grad_norm": 2.055698998283488, + "learning_rate": 1.8397162833544856e-05, + "loss": 0.955, + "step": 8797 + }, + { + "epoch": 0.21, + "grad_norm": 2.08565996632931, + "learning_rate": 1.8396748458437945e-05, + "loss": 0.9411, + "step": 8798 + }, + { + "epoch": 0.21, + "grad_norm": 2.0856521201162, + "learning_rate": 1.8396334034442506e-05, + "loss": 0.9869, + "step": 8799 + }, + { + "epoch": 0.21, + "grad_norm": 2.1371319423498343, + "learning_rate": 1.839591956156095e-05, + "loss": 1.079, + "step": 8800 + }, + { + "epoch": 0.21, + "grad_norm": 2.5439243449410536, + "learning_rate": 1.8395505039795696e-05, + "loss": 1.0071, + "step": 8801 + }, + { + "epoch": 0.21, + "grad_norm": 2.203779464197342, + "learning_rate": 1.8395090469149153e-05, + "loss": 1.0087, + "step": 8802 + }, + { + "epoch": 0.21, + "grad_norm": 2.313239593394091, + "learning_rate": 1.8394675849623733e-05, + "loss": 1.0406, + "step": 8803 + }, + { + "epoch": 0.21, + "grad_norm": 2.365917618023618, + "learning_rate": 1.8394261181221853e-05, + "loss": 1.1183, + "step": 8804 + }, + { + "epoch": 0.21, + "grad_norm": 1.9521037646940977, + "learning_rate": 1.8393846463945925e-05, + "loss": 1.0911, + "step": 8805 + }, + { + "epoch": 0.21, + "grad_norm": 1.9105152180930292, + "learning_rate": 1.839343169779837e-05, + "loss": 1.0495, + "step": 8806 + }, + { + "epoch": 0.21, + "grad_norm": 2.023439831598883, + "learning_rate": 1.83930168827816e-05, + "loss": 1.1071, + "step": 8807 + }, + { + "epoch": 0.21, + "grad_norm": 1.9951426218038886, + "learning_rate": 1.8392602018898023e-05, + "loss": 1.0867, + "step": 8808 + }, + { + "epoch": 0.21, + "grad_norm": 2.0488638090639024, + "learning_rate": 1.8392187106150058e-05, + "loss": 1.0198, + "step": 8809 + }, + { + "epoch": 0.21, + "grad_norm": 2.574125307395824, + "learning_rate": 1.8391772144540127e-05, + "loss": 0.8996, + "step": 8810 + }, + { + "epoch": 0.21, + "grad_norm": 1.8056418395338274, + "learning_rate": 1.8391357134070642e-05, + "loss": 1.0801, + "step": 8811 + }, + { + "epoch": 0.21, + "grad_norm": 1.2292382672705955, + "learning_rate": 1.839094207474402e-05, + "loss": 1.0254, + "step": 8812 + }, + { + "epoch": 0.21, + "grad_norm": 1.1376896518353612, + "learning_rate": 1.8390526966562673e-05, + "loss": 0.989, + "step": 8813 + }, + { + "epoch": 0.21, + "grad_norm": 2.536320075071965, + "learning_rate": 1.8390111809529024e-05, + "loss": 0.9852, + "step": 8814 + }, + { + "epoch": 0.21, + "grad_norm": 2.1779164627312713, + "learning_rate": 1.8389696603645487e-05, + "loss": 1.0999, + "step": 8815 + }, + { + "epoch": 0.21, + "grad_norm": 1.8182378325216675, + "learning_rate": 1.8389281348914482e-05, + "loss": 1.0788, + "step": 8816 + }, + { + "epoch": 0.21, + "grad_norm": 1.8528304553318011, + "learning_rate": 1.8388866045338424e-05, + "loss": 0.9712, + "step": 8817 + }, + { + "epoch": 0.21, + "grad_norm": 2.2771216342679264, + "learning_rate": 1.8388450692919733e-05, + "loss": 1.0196, + "step": 8818 + }, + { + "epoch": 0.21, + "grad_norm": 2.5150497769011055, + "learning_rate": 1.8388035291660823e-05, + "loss": 0.9558, + "step": 8819 + }, + { + "epoch": 0.21, + "grad_norm": 1.9048643912229692, + "learning_rate": 1.8387619841564118e-05, + "loss": 1.101, + "step": 8820 + }, + { + "epoch": 0.21, + "grad_norm": 2.3136767978654484, + "learning_rate": 1.8387204342632034e-05, + "loss": 1.1143, + "step": 8821 + }, + { + "epoch": 0.21, + "grad_norm": 1.968249910400105, + "learning_rate": 1.8386788794866995e-05, + "loss": 1.1917, + "step": 8822 + }, + { + "epoch": 0.21, + "grad_norm": 2.159410407961067, + "learning_rate": 1.8386373198271412e-05, + "loss": 1.0773, + "step": 8823 + }, + { + "epoch": 0.21, + "grad_norm": 2.228608435805267, + "learning_rate": 1.8385957552847707e-05, + "loss": 1.0643, + "step": 8824 + }, + { + "epoch": 0.21, + "grad_norm": 2.2061649262670313, + "learning_rate": 1.8385541858598306e-05, + "loss": 1.0192, + "step": 8825 + }, + { + "epoch": 0.21, + "grad_norm": 2.03517721631202, + "learning_rate": 1.8385126115525627e-05, + "loss": 1.1314, + "step": 8826 + }, + { + "epoch": 0.21, + "grad_norm": 2.085091037175343, + "learning_rate": 1.8384710323632083e-05, + "loss": 1.0525, + "step": 8827 + }, + { + "epoch": 0.21, + "grad_norm": 2.1148801730036553, + "learning_rate": 1.838429448292011e-05, + "loss": 0.9294, + "step": 8828 + }, + { + "epoch": 0.21, + "grad_norm": 2.2810212338637927, + "learning_rate": 1.838387859339211e-05, + "loss": 1.1902, + "step": 8829 + }, + { + "epoch": 0.21, + "grad_norm": 1.1645818835129789, + "learning_rate": 1.8383462655050525e-05, + "loss": 0.9315, + "step": 8830 + }, + { + "epoch": 0.21, + "grad_norm": 2.000687639585955, + "learning_rate": 1.838304666789776e-05, + "loss": 1.0272, + "step": 8831 + }, + { + "epoch": 0.21, + "grad_norm": 3.6909881594711207, + "learning_rate": 1.8382630631936243e-05, + "loss": 1.1402, + "step": 8832 + }, + { + "epoch": 0.21, + "grad_norm": 2.075585512838036, + "learning_rate": 1.8382214547168396e-05, + "loss": 1.1366, + "step": 8833 + }, + { + "epoch": 0.21, + "grad_norm": 2.299623952939757, + "learning_rate": 1.8381798413596646e-05, + "loss": 1.0234, + "step": 8834 + }, + { + "epoch": 0.21, + "grad_norm": 2.1176185327505124, + "learning_rate": 1.838138223122341e-05, + "loss": 1.0239, + "step": 8835 + }, + { + "epoch": 0.21, + "grad_norm": 2.3802279868512692, + "learning_rate": 1.8380966000051113e-05, + "loss": 1.0512, + "step": 8836 + }, + { + "epoch": 0.21, + "grad_norm": 2.282308221697868, + "learning_rate": 1.8380549720082175e-05, + "loss": 1.1199, + "step": 8837 + }, + { + "epoch": 0.21, + "grad_norm": 2.0003494835352815, + "learning_rate": 1.838013339131903e-05, + "loss": 1.007, + "step": 8838 + }, + { + "epoch": 0.21, + "grad_norm": 2.266913774009598, + "learning_rate": 1.837971701376409e-05, + "loss": 0.9514, + "step": 8839 + }, + { + "epoch": 0.21, + "grad_norm": 2.1186906821436593, + "learning_rate": 1.837930058741979e-05, + "loss": 1.025, + "step": 8840 + }, + { + "epoch": 0.21, + "grad_norm": 1.9645864435411091, + "learning_rate": 1.8378884112288543e-05, + "loss": 1.089, + "step": 8841 + }, + { + "epoch": 0.21, + "grad_norm": 1.1993378118423894, + "learning_rate": 1.8378467588372785e-05, + "loss": 0.9506, + "step": 8842 + }, + { + "epoch": 0.21, + "grad_norm": 1.8549090333023008, + "learning_rate": 1.8378051015674933e-05, + "loss": 1.0699, + "step": 8843 + }, + { + "epoch": 0.21, + "grad_norm": 2.2330196645933156, + "learning_rate": 1.837763439419742e-05, + "loss": 1.0598, + "step": 8844 + }, + { + "epoch": 0.21, + "grad_norm": 2.0684189105155775, + "learning_rate": 1.8377217723942662e-05, + "loss": 1.0426, + "step": 8845 + }, + { + "epoch": 0.21, + "grad_norm": 2.0983673607241697, + "learning_rate": 1.837680100491309e-05, + "loss": 1.0863, + "step": 8846 + }, + { + "epoch": 0.21, + "grad_norm": 2.9553517216849623, + "learning_rate": 1.8376384237111134e-05, + "loss": 0.9963, + "step": 8847 + }, + { + "epoch": 0.21, + "grad_norm": 2.0158116542764017, + "learning_rate": 1.8375967420539215e-05, + "loss": 1.0238, + "step": 8848 + }, + { + "epoch": 0.21, + "grad_norm": 2.604700923406964, + "learning_rate": 1.8375550555199764e-05, + "loss": 1.0268, + "step": 8849 + }, + { + "epoch": 0.21, + "grad_norm": 2.381223533094838, + "learning_rate": 1.8375133641095206e-05, + "loss": 1.0147, + "step": 8850 + }, + { + "epoch": 0.21, + "grad_norm": 1.9100226891786931, + "learning_rate": 1.8374716678227964e-05, + "loss": 1.0635, + "step": 8851 + }, + { + "epoch": 0.21, + "grad_norm": 2.105644678365508, + "learning_rate": 1.8374299666600474e-05, + "loss": 1.0044, + "step": 8852 + }, + { + "epoch": 0.21, + "grad_norm": 2.094941087221595, + "learning_rate": 1.837388260621516e-05, + "loss": 0.9778, + "step": 8853 + }, + { + "epoch": 0.21, + "grad_norm": 1.9394567657419244, + "learning_rate": 1.8373465497074448e-05, + "loss": 1.0974, + "step": 8854 + }, + { + "epoch": 0.21, + "grad_norm": 2.429553487670066, + "learning_rate": 1.837304833918077e-05, + "loss": 0.9294, + "step": 8855 + }, + { + "epoch": 0.21, + "grad_norm": 1.1034186202429765, + "learning_rate": 1.8372631132536552e-05, + "loss": 0.9756, + "step": 8856 + }, + { + "epoch": 0.21, + "grad_norm": 2.1078417885999223, + "learning_rate": 1.837221387714423e-05, + "loss": 1.0784, + "step": 8857 + }, + { + "epoch": 0.21, + "grad_norm": 1.1353138335623338, + "learning_rate": 1.8371796573006223e-05, + "loss": 1.0066, + "step": 8858 + }, + { + "epoch": 0.21, + "grad_norm": 2.426424354922872, + "learning_rate": 1.8371379220124968e-05, + "loss": 1.0661, + "step": 8859 + }, + { + "epoch": 0.21, + "grad_norm": 2.090502587083312, + "learning_rate": 1.8370961818502893e-05, + "loss": 1.1184, + "step": 8860 + }, + { + "epoch": 0.21, + "grad_norm": 2.056064821263428, + "learning_rate": 1.837054436814243e-05, + "loss": 0.9914, + "step": 8861 + }, + { + "epoch": 0.21, + "grad_norm": 2.221651475905411, + "learning_rate": 1.8370126869046002e-05, + "loss": 1.1298, + "step": 8862 + }, + { + "epoch": 0.21, + "grad_norm": 1.7665899154042966, + "learning_rate": 1.836970932121605e-05, + "loss": 1.0064, + "step": 8863 + }, + { + "epoch": 0.21, + "grad_norm": 1.9342807538512892, + "learning_rate": 1.8369291724655e-05, + "loss": 1.0842, + "step": 8864 + }, + { + "epoch": 0.21, + "grad_norm": 3.029263045428288, + "learning_rate": 1.836887407936528e-05, + "loss": 1.1011, + "step": 8865 + }, + { + "epoch": 0.21, + "grad_norm": 1.8022109950411676, + "learning_rate": 1.8368456385349333e-05, + "loss": 1.1071, + "step": 8866 + }, + { + "epoch": 0.21, + "grad_norm": 2.210851614567869, + "learning_rate": 1.8368038642609577e-05, + "loss": 0.9808, + "step": 8867 + }, + { + "epoch": 0.21, + "grad_norm": 2.123829593911399, + "learning_rate": 1.8367620851148455e-05, + "loss": 1.0483, + "step": 8868 + }, + { + "epoch": 0.21, + "grad_norm": 1.9503951657966458, + "learning_rate": 1.8367203010968393e-05, + "loss": 1.1858, + "step": 8869 + }, + { + "epoch": 0.21, + "grad_norm": 2.424709124032055, + "learning_rate": 1.8366785122071828e-05, + "loss": 0.9399, + "step": 8870 + }, + { + "epoch": 0.21, + "grad_norm": 2.3392694908725504, + "learning_rate": 1.8366367184461188e-05, + "loss": 1.0063, + "step": 8871 + }, + { + "epoch": 0.21, + "grad_norm": 2.1948427911339676, + "learning_rate": 1.8365949198138913e-05, + "loss": 1.2098, + "step": 8872 + }, + { + "epoch": 0.21, + "grad_norm": 1.9999513688871209, + "learning_rate": 1.836553116310743e-05, + "loss": 1.1568, + "step": 8873 + }, + { + "epoch": 0.21, + "grad_norm": 2.0374143750795843, + "learning_rate": 1.8365113079369176e-05, + "loss": 1.1256, + "step": 8874 + }, + { + "epoch": 0.21, + "grad_norm": 2.245865627397156, + "learning_rate": 1.8364694946926586e-05, + "loss": 0.9465, + "step": 8875 + }, + { + "epoch": 0.21, + "grad_norm": 2.003262784959778, + "learning_rate": 1.8364276765782095e-05, + "loss": 0.9935, + "step": 8876 + }, + { + "epoch": 0.21, + "grad_norm": 2.3612018895384894, + "learning_rate": 1.8363858535938136e-05, + "loss": 1.1069, + "step": 8877 + }, + { + "epoch": 0.21, + "grad_norm": 2.2237317103480922, + "learning_rate": 1.8363440257397145e-05, + "loss": 0.99, + "step": 8878 + }, + { + "epoch": 0.21, + "grad_norm": 2.1594512052509534, + "learning_rate": 1.8363021930161555e-05, + "loss": 0.9882, + "step": 8879 + }, + { + "epoch": 0.21, + "grad_norm": 2.1928549138230125, + "learning_rate": 1.8362603554233807e-05, + "loss": 1.1021, + "step": 8880 + }, + { + "epoch": 0.21, + "grad_norm": 1.8075890300593829, + "learning_rate": 1.8362185129616333e-05, + "loss": 1.0177, + "step": 8881 + }, + { + "epoch": 0.21, + "grad_norm": 2.218841778236294, + "learning_rate": 1.8361766656311568e-05, + "loss": 1.1337, + "step": 8882 + }, + { + "epoch": 0.21, + "grad_norm": 2.1505618032138503, + "learning_rate": 1.836134813432195e-05, + "loss": 1.0772, + "step": 8883 + }, + { + "epoch": 0.21, + "grad_norm": 2.0049789865990055, + "learning_rate": 1.8360929563649918e-05, + "loss": 1.1833, + "step": 8884 + }, + { + "epoch": 0.21, + "grad_norm": 1.1248178438065377, + "learning_rate": 1.8360510944297905e-05, + "loss": 1.0163, + "step": 8885 + }, + { + "epoch": 0.21, + "grad_norm": 1.9883599453099743, + "learning_rate": 1.836009227626835e-05, + "loss": 1.1319, + "step": 8886 + }, + { + "epoch": 0.21, + "grad_norm": 2.255659716923424, + "learning_rate": 1.8359673559563696e-05, + "loss": 1.061, + "step": 8887 + }, + { + "epoch": 0.21, + "grad_norm": 2.2096434646254575, + "learning_rate": 1.8359254794186368e-05, + "loss": 1.1005, + "step": 8888 + }, + { + "epoch": 0.21, + "grad_norm": 1.9551924715891638, + "learning_rate": 1.8358835980138822e-05, + "loss": 1.0487, + "step": 8889 + }, + { + "epoch": 0.21, + "grad_norm": 2.096036180066587, + "learning_rate": 1.8358417117423476e-05, + "loss": 0.974, + "step": 8890 + }, + { + "epoch": 0.21, + "grad_norm": 2.153611128711762, + "learning_rate": 1.8357998206042788e-05, + "loss": 1.1663, + "step": 8891 + }, + { + "epoch": 0.21, + "grad_norm": 2.1461414858591983, + "learning_rate": 1.8357579245999182e-05, + "loss": 1.101, + "step": 8892 + }, + { + "epoch": 0.21, + "grad_norm": 1.9863406652619224, + "learning_rate": 1.8357160237295106e-05, + "loss": 1.0683, + "step": 8893 + }, + { + "epoch": 0.21, + "grad_norm": 2.1895289002806217, + "learning_rate": 1.8356741179933e-05, + "loss": 1.0925, + "step": 8894 + }, + { + "epoch": 0.21, + "grad_norm": 2.066516092807185, + "learning_rate": 1.83563220739153e-05, + "loss": 1.0935, + "step": 8895 + }, + { + "epoch": 0.21, + "grad_norm": 2.1747518732358184, + "learning_rate": 1.835590291924444e-05, + "loss": 1.1024, + "step": 8896 + }, + { + "epoch": 0.21, + "grad_norm": 2.2966671235506237, + "learning_rate": 1.8355483715922876e-05, + "loss": 1.0928, + "step": 8897 + }, + { + "epoch": 0.21, + "grad_norm": 1.8890711268430336, + "learning_rate": 1.835506446395304e-05, + "loss": 0.9865, + "step": 8898 + }, + { + "epoch": 0.21, + "grad_norm": 1.9636651500241855, + "learning_rate": 1.835464516333737e-05, + "loss": 0.987, + "step": 8899 + }, + { + "epoch": 0.21, + "grad_norm": 2.232535670872765, + "learning_rate": 1.8354225814078313e-05, + "loss": 0.9953, + "step": 8900 + }, + { + "epoch": 0.21, + "grad_norm": 2.191185623935347, + "learning_rate": 1.835380641617831e-05, + "loss": 1.1423, + "step": 8901 + }, + { + "epoch": 0.21, + "grad_norm": 1.9747622386660593, + "learning_rate": 1.83533869696398e-05, + "loss": 1.123, + "step": 8902 + }, + { + "epoch": 0.21, + "grad_norm": 2.1115025806429015, + "learning_rate": 1.8352967474465223e-05, + "loss": 1.0788, + "step": 8903 + }, + { + "epoch": 0.21, + "grad_norm": 2.011301161655147, + "learning_rate": 1.8352547930657027e-05, + "loss": 1.1863, + "step": 8904 + }, + { + "epoch": 0.21, + "grad_norm": 1.1972435732175084, + "learning_rate": 1.8352128338217654e-05, + "loss": 0.9287, + "step": 8905 + }, + { + "epoch": 0.21, + "grad_norm": 2.563050895136118, + "learning_rate": 1.8351708697149543e-05, + "loss": 0.9671, + "step": 8906 + }, + { + "epoch": 0.21, + "grad_norm": 2.0872970459686537, + "learning_rate": 1.8351289007455145e-05, + "loss": 1.0761, + "step": 8907 + }, + { + "epoch": 0.21, + "grad_norm": 2.0776764683221303, + "learning_rate": 1.835086926913689e-05, + "loss": 1.0894, + "step": 8908 + }, + { + "epoch": 0.21, + "grad_norm": 2.8315637912205, + "learning_rate": 1.8350449482197236e-05, + "loss": 1.0673, + "step": 8909 + }, + { + "epoch": 0.21, + "grad_norm": 2.0913115355982854, + "learning_rate": 1.8350029646638625e-05, + "loss": 1.0468, + "step": 8910 + }, + { + "epoch": 0.21, + "grad_norm": 1.9173524992724496, + "learning_rate": 1.834960976246349e-05, + "loss": 1.0192, + "step": 8911 + }, + { + "epoch": 0.21, + "grad_norm": 2.3044051677312725, + "learning_rate": 1.8349189829674288e-05, + "loss": 0.9589, + "step": 8912 + }, + { + "epoch": 0.21, + "grad_norm": 2.239307131954061, + "learning_rate": 1.834876984827346e-05, + "loss": 1.0607, + "step": 8913 + }, + { + "epoch": 0.21, + "grad_norm": 2.016461876818182, + "learning_rate": 1.8348349818263445e-05, + "loss": 1.0132, + "step": 8914 + }, + { + "epoch": 0.21, + "grad_norm": 2.002935374199333, + "learning_rate": 1.83479297396467e-05, + "loss": 0.9952, + "step": 8915 + }, + { + "epoch": 0.21, + "grad_norm": 1.900595844227985, + "learning_rate": 1.8347509612425666e-05, + "loss": 1.0635, + "step": 8916 + }, + { + "epoch": 0.21, + "grad_norm": 1.973654032406274, + "learning_rate": 1.8347089436602786e-05, + "loss": 1.087, + "step": 8917 + }, + { + "epoch": 0.21, + "grad_norm": 1.1221817881908671, + "learning_rate": 1.8346669212180507e-05, + "loss": 0.9702, + "step": 8918 + }, + { + "epoch": 0.21, + "grad_norm": 1.9774303621864144, + "learning_rate": 1.834624893916128e-05, + "loss": 1.0912, + "step": 8919 + }, + { + "epoch": 0.21, + "grad_norm": 1.1182488703945284, + "learning_rate": 1.834582861754755e-05, + "loss": 0.9772, + "step": 8920 + }, + { + "epoch": 0.21, + "grad_norm": 2.099128669779226, + "learning_rate": 1.834540824734176e-05, + "loss": 1.1444, + "step": 8921 + }, + { + "epoch": 0.21, + "grad_norm": 2.175094255796604, + "learning_rate": 1.8344987828546364e-05, + "loss": 1.1085, + "step": 8922 + }, + { + "epoch": 0.21, + "grad_norm": 2.2133669441584916, + "learning_rate": 1.8344567361163804e-05, + "loss": 1.0503, + "step": 8923 + }, + { + "epoch": 0.21, + "grad_norm": 2.2959736617407627, + "learning_rate": 1.8344146845196534e-05, + "loss": 1.0957, + "step": 8924 + }, + { + "epoch": 0.21, + "grad_norm": 2.0629216507303334, + "learning_rate": 1.8343726280646996e-05, + "loss": 1.1475, + "step": 8925 + }, + { + "epoch": 0.21, + "grad_norm": 2.4956626936841864, + "learning_rate": 1.8343305667517644e-05, + "loss": 0.9745, + "step": 8926 + }, + { + "epoch": 0.21, + "grad_norm": 1.21125536253867, + "learning_rate": 1.8342885005810926e-05, + "loss": 0.9211, + "step": 8927 + }, + { + "epoch": 0.21, + "grad_norm": 2.195767362737719, + "learning_rate": 1.834246429552929e-05, + "loss": 0.9773, + "step": 8928 + }, + { + "epoch": 0.21, + "grad_norm": 2.057305131513055, + "learning_rate": 1.8342043536675185e-05, + "loss": 1.076, + "step": 8929 + }, + { + "epoch": 0.21, + "grad_norm": 2.114633966774991, + "learning_rate": 1.8341622729251062e-05, + "loss": 1.0805, + "step": 8930 + }, + { + "epoch": 0.21, + "grad_norm": 2.0752021562800036, + "learning_rate": 1.834120187325937e-05, + "loss": 1.1231, + "step": 8931 + }, + { + "epoch": 0.21, + "grad_norm": 2.103847512567503, + "learning_rate": 1.834078096870256e-05, + "loss": 0.9558, + "step": 8932 + }, + { + "epoch": 0.21, + "grad_norm": 1.0647941314362666, + "learning_rate": 1.834036001558308e-05, + "loss": 1.018, + "step": 8933 + }, + { + "epoch": 0.21, + "grad_norm": 1.9408352703494312, + "learning_rate": 1.833993901390339e-05, + "loss": 0.9971, + "step": 8934 + }, + { + "epoch": 0.21, + "grad_norm": 2.249184905060394, + "learning_rate": 1.8339517963665928e-05, + "loss": 0.9903, + "step": 8935 + }, + { + "epoch": 0.21, + "grad_norm": 2.1528532884136213, + "learning_rate": 1.8339096864873153e-05, + "loss": 1.0427, + "step": 8936 + }, + { + "epoch": 0.21, + "grad_norm": 2.118597368404945, + "learning_rate": 1.833867571752752e-05, + "loss": 1.0464, + "step": 8937 + }, + { + "epoch": 0.21, + "grad_norm": 2.8578817603054416, + "learning_rate": 1.8338254521631476e-05, + "loss": 0.9125, + "step": 8938 + }, + { + "epoch": 0.21, + "grad_norm": 1.1450549260506802, + "learning_rate": 1.8337833277187472e-05, + "loss": 1.0005, + "step": 8939 + }, + { + "epoch": 0.21, + "grad_norm": 1.166573736352938, + "learning_rate": 1.8337411984197965e-05, + "loss": 0.9901, + "step": 8940 + }, + { + "epoch": 0.21, + "grad_norm": 2.1951186252108674, + "learning_rate": 1.8336990642665405e-05, + "loss": 1.1831, + "step": 8941 + }, + { + "epoch": 0.21, + "grad_norm": 1.9120194800718293, + "learning_rate": 1.8336569252592243e-05, + "loss": 1.0386, + "step": 8942 + }, + { + "epoch": 0.21, + "grad_norm": 1.964894248823479, + "learning_rate": 1.8336147813980936e-05, + "loss": 1.1528, + "step": 8943 + }, + { + "epoch": 0.21, + "grad_norm": 3.414277321270561, + "learning_rate": 1.8335726326833943e-05, + "loss": 0.8003, + "step": 8944 + }, + { + "epoch": 0.21, + "grad_norm": 2.3120178356512695, + "learning_rate": 1.8335304791153707e-05, + "loss": 1.0862, + "step": 8945 + }, + { + "epoch": 0.21, + "grad_norm": 1.9704114832946669, + "learning_rate": 1.833488320694269e-05, + "loss": 1.2322, + "step": 8946 + }, + { + "epoch": 0.21, + "grad_norm": 2.106128778603064, + "learning_rate": 1.8334461574203346e-05, + "loss": 1.1246, + "step": 8947 + }, + { + "epoch": 0.21, + "grad_norm": 1.9389442710774527, + "learning_rate": 1.8334039892938126e-05, + "loss": 1.0545, + "step": 8948 + }, + { + "epoch": 0.21, + "grad_norm": 2.1282421461909165, + "learning_rate": 1.8333618163149485e-05, + "loss": 1.026, + "step": 8949 + }, + { + "epoch": 0.21, + "grad_norm": 2.0677056205820397, + "learning_rate": 1.8333196384839884e-05, + "loss": 1.0149, + "step": 8950 + }, + { + "epoch": 0.21, + "grad_norm": 1.9255430578897113, + "learning_rate": 1.833277455801177e-05, + "loss": 1.1503, + "step": 8951 + }, + { + "epoch": 0.21, + "grad_norm": 1.9869927529378408, + "learning_rate": 1.8332352682667613e-05, + "loss": 1.1335, + "step": 8952 + }, + { + "epoch": 0.21, + "grad_norm": 1.9573154200319158, + "learning_rate": 1.8331930758809856e-05, + "loss": 0.9781, + "step": 8953 + }, + { + "epoch": 0.21, + "grad_norm": 1.9276909891544172, + "learning_rate": 1.833150878644096e-05, + "loss": 1.1368, + "step": 8954 + }, + { + "epoch": 0.21, + "grad_norm": 1.9736478282125092, + "learning_rate": 1.8331086765563384e-05, + "loss": 0.9867, + "step": 8955 + }, + { + "epoch": 0.21, + "grad_norm": 2.182289137605034, + "learning_rate": 1.833066469617958e-05, + "loss": 1.0253, + "step": 8956 + }, + { + "epoch": 0.21, + "grad_norm": 2.0929293002618063, + "learning_rate": 1.8330242578292013e-05, + "loss": 0.8297, + "step": 8957 + }, + { + "epoch": 0.21, + "grad_norm": 2.6395334669641404, + "learning_rate": 1.832982041190313e-05, + "loss": 1.038, + "step": 8958 + }, + { + "epoch": 0.21, + "grad_norm": 2.9758192913909283, + "learning_rate": 1.83293981970154e-05, + "loss": 0.9807, + "step": 8959 + }, + { + "epoch": 0.21, + "grad_norm": 1.9362611880409781, + "learning_rate": 1.8328975933631277e-05, + "loss": 1.0972, + "step": 8960 + }, + { + "epoch": 0.21, + "grad_norm": 1.9102482584625122, + "learning_rate": 1.832855362175322e-05, + "loss": 1.043, + "step": 8961 + }, + { + "epoch": 0.21, + "grad_norm": 2.221737756703829, + "learning_rate": 1.832813126138368e-05, + "loss": 1.027, + "step": 8962 + }, + { + "epoch": 0.21, + "grad_norm": 1.9106470286418562, + "learning_rate": 1.8327708852525132e-05, + "loss": 0.9793, + "step": 8963 + }, + { + "epoch": 0.21, + "grad_norm": 2.142623017639717, + "learning_rate": 1.832728639518002e-05, + "loss": 0.8922, + "step": 8964 + }, + { + "epoch": 0.21, + "grad_norm": 1.7615020626226048, + "learning_rate": 1.8326863889350816e-05, + "loss": 1.0384, + "step": 8965 + }, + { + "epoch": 0.21, + "grad_norm": 2.0469061189617364, + "learning_rate": 1.832644133503997e-05, + "loss": 1.0781, + "step": 8966 + }, + { + "epoch": 0.21, + "grad_norm": 1.9110485526395147, + "learning_rate": 1.832601873224995e-05, + "loss": 1.0539, + "step": 8967 + }, + { + "epoch": 0.21, + "grad_norm": 2.0237754563482393, + "learning_rate": 1.832559608098321e-05, + "loss": 1.0554, + "step": 8968 + }, + { + "epoch": 0.21, + "grad_norm": 1.8920455297387544, + "learning_rate": 1.8325173381242213e-05, + "loss": 1.0132, + "step": 8969 + }, + { + "epoch": 0.21, + "grad_norm": 3.272239479060897, + "learning_rate": 1.8324750633029424e-05, + "loss": 1.0564, + "step": 8970 + }, + { + "epoch": 0.21, + "grad_norm": 1.939514810384537, + "learning_rate": 1.83243278363473e-05, + "loss": 1.0893, + "step": 8971 + }, + { + "epoch": 0.21, + "grad_norm": 2.1738173415129913, + "learning_rate": 1.8323904991198304e-05, + "loss": 1.1028, + "step": 8972 + }, + { + "epoch": 0.21, + "grad_norm": 2.100135527045147, + "learning_rate": 1.8323482097584896e-05, + "loss": 1.1837, + "step": 8973 + }, + { + "epoch": 0.21, + "grad_norm": 2.246095197303969, + "learning_rate": 1.8323059155509542e-05, + "loss": 1.1184, + "step": 8974 + }, + { + "epoch": 0.21, + "grad_norm": 2.087872706283743, + "learning_rate": 1.8322636164974705e-05, + "loss": 0.9034, + "step": 8975 + }, + { + "epoch": 0.21, + "grad_norm": 2.1414333980670235, + "learning_rate": 1.832221312598284e-05, + "loss": 1.0561, + "step": 8976 + }, + { + "epoch": 0.21, + "grad_norm": 1.8825587438249503, + "learning_rate": 1.8321790038536415e-05, + "loss": 1.1524, + "step": 8977 + }, + { + "epoch": 0.21, + "grad_norm": 2.017385813536523, + "learning_rate": 1.83213669026379e-05, + "loss": 1.0855, + "step": 8978 + }, + { + "epoch": 0.21, + "grad_norm": 2.006754074492133, + "learning_rate": 1.8320943718289747e-05, + "loss": 1.0429, + "step": 8979 + }, + { + "epoch": 0.21, + "grad_norm": 2.0639092678960704, + "learning_rate": 1.832052048549443e-05, + "loss": 1.0583, + "step": 8980 + }, + { + "epoch": 0.21, + "grad_norm": 2.25993752526356, + "learning_rate": 1.8320097204254402e-05, + "loss": 1.0619, + "step": 8981 + }, + { + "epoch": 0.21, + "grad_norm": 2.2856181582644974, + "learning_rate": 1.831967387457214e-05, + "loss": 1.1454, + "step": 8982 + }, + { + "epoch": 0.21, + "grad_norm": 1.8001943120987787, + "learning_rate": 1.83192504964501e-05, + "loss": 1.0213, + "step": 8983 + }, + { + "epoch": 0.21, + "grad_norm": 1.0993437533024013, + "learning_rate": 1.831882706989075e-05, + "loss": 0.9387, + "step": 8984 + }, + { + "epoch": 0.21, + "grad_norm": 2.4726321164041183, + "learning_rate": 1.8318403594896557e-05, + "loss": 1.0371, + "step": 8985 + }, + { + "epoch": 0.21, + "grad_norm": 2.272345892961672, + "learning_rate": 1.8317980071469982e-05, + "loss": 1.033, + "step": 8986 + }, + { + "epoch": 0.21, + "grad_norm": 3.9054683584252454, + "learning_rate": 1.8317556499613492e-05, + "loss": 1.1261, + "step": 8987 + }, + { + "epoch": 0.21, + "grad_norm": 2.605390970675532, + "learning_rate": 1.831713287932956e-05, + "loss": 1.108, + "step": 8988 + }, + { + "epoch": 0.21, + "grad_norm": 2.482574259281285, + "learning_rate": 1.8316709210620646e-05, + "loss": 1.0687, + "step": 8989 + }, + { + "epoch": 0.21, + "grad_norm": 2.4055697047227578, + "learning_rate": 1.8316285493489213e-05, + "loss": 1.1339, + "step": 8990 + }, + { + "epoch": 0.21, + "grad_norm": 2.0136297499739526, + "learning_rate": 1.831586172793774e-05, + "loss": 1.043, + "step": 8991 + }, + { + "epoch": 0.21, + "grad_norm": 2.614569503455833, + "learning_rate": 1.831543791396868e-05, + "loss": 1.0095, + "step": 8992 + }, + { + "epoch": 0.21, + "grad_norm": 1.1334841262285524, + "learning_rate": 1.831501405158451e-05, + "loss": 0.9867, + "step": 8993 + }, + { + "epoch": 0.21, + "grad_norm": 2.6853711375910927, + "learning_rate": 1.83145901407877e-05, + "loss": 1.0798, + "step": 8994 + }, + { + "epoch": 0.21, + "grad_norm": 3.1284779050511666, + "learning_rate": 1.831416618158071e-05, + "loss": 1.068, + "step": 8995 + }, + { + "epoch": 0.21, + "grad_norm": 2.095260423930366, + "learning_rate": 1.8313742173966017e-05, + "loss": 0.9745, + "step": 8996 + }, + { + "epoch": 0.21, + "grad_norm": 1.957081337723717, + "learning_rate": 1.831331811794608e-05, + "loss": 1.025, + "step": 8997 + }, + { + "epoch": 0.21, + "grad_norm": 2.0210216676241877, + "learning_rate": 1.831289401352337e-05, + "loss": 0.9974, + "step": 8998 + }, + { + "epoch": 0.21, + "grad_norm": 1.9246043924081737, + "learning_rate": 1.8312469860700366e-05, + "loss": 1.1562, + "step": 8999 + }, + { + "epoch": 0.21, + "grad_norm": 2.324555338501697, + "learning_rate": 1.831204565947953e-05, + "loss": 0.9796, + "step": 9000 + }, + { + "epoch": 0.21, + "grad_norm": 2.184268245864827, + "learning_rate": 1.831162140986333e-05, + "loss": 1.0388, + "step": 9001 + }, + { + "epoch": 0.21, + "grad_norm": 2.10561860113921, + "learning_rate": 1.831119711185424e-05, + "loss": 0.9811, + "step": 9002 + }, + { + "epoch": 0.21, + "grad_norm": 2.098689707011012, + "learning_rate": 1.831077276545473e-05, + "loss": 1.0564, + "step": 9003 + }, + { + "epoch": 0.21, + "grad_norm": 2.103731119120881, + "learning_rate": 1.831034837066727e-05, + "loss": 1.0037, + "step": 9004 + }, + { + "epoch": 0.21, + "grad_norm": 1.954091057637168, + "learning_rate": 1.830992392749433e-05, + "loss": 1.0754, + "step": 9005 + }, + { + "epoch": 0.21, + "grad_norm": 1.1457765906997524, + "learning_rate": 1.8309499435938377e-05, + "loss": 1.0633, + "step": 9006 + }, + { + "epoch": 0.21, + "grad_norm": 1.9715360124575887, + "learning_rate": 1.8309074896001893e-05, + "loss": 1.0098, + "step": 9007 + }, + { + "epoch": 0.21, + "grad_norm": 2.0353941838967247, + "learning_rate": 1.8308650307687347e-05, + "loss": 0.9079, + "step": 9008 + }, + { + "epoch": 0.21, + "grad_norm": 2.2021832894741338, + "learning_rate": 1.83082256709972e-05, + "loss": 0.9238, + "step": 9009 + }, + { + "epoch": 0.21, + "grad_norm": 1.8348675300446018, + "learning_rate": 1.8307800985933938e-05, + "loss": 1.1206, + "step": 9010 + }, + { + "epoch": 0.21, + "grad_norm": 1.1553752206189074, + "learning_rate": 1.830737625250003e-05, + "loss": 1.0102, + "step": 9011 + }, + { + "epoch": 0.21, + "grad_norm": 2.096328757731342, + "learning_rate": 1.8306951470697946e-05, + "loss": 1.0612, + "step": 9012 + }, + { + "epoch": 0.21, + "grad_norm": 2.030129730288678, + "learning_rate": 1.830652664053016e-05, + "loss": 1.0149, + "step": 9013 + }, + { + "epoch": 0.21, + "grad_norm": 2.1525288080299676, + "learning_rate": 1.8306101761999147e-05, + "loss": 1.1083, + "step": 9014 + }, + { + "epoch": 0.21, + "grad_norm": 1.7276378483262125, + "learning_rate": 1.8305676835107378e-05, + "loss": 1.0052, + "step": 9015 + }, + { + "epoch": 0.21, + "grad_norm": 2.281930903596705, + "learning_rate": 1.830525185985733e-05, + "loss": 0.9876, + "step": 9016 + }, + { + "epoch": 0.21, + "grad_norm": 2.086431145856421, + "learning_rate": 1.8304826836251477e-05, + "loss": 0.9782, + "step": 9017 + }, + { + "epoch": 0.21, + "grad_norm": 2.083752033868421, + "learning_rate": 1.8304401764292292e-05, + "loss": 1.1198, + "step": 9018 + }, + { + "epoch": 0.21, + "grad_norm": 2.205111071392852, + "learning_rate": 1.830397664398225e-05, + "loss": 0.9632, + "step": 9019 + }, + { + "epoch": 0.21, + "grad_norm": 2.383601373880973, + "learning_rate": 1.830355147532383e-05, + "loss": 1.1038, + "step": 9020 + }, + { + "epoch": 0.21, + "grad_norm": 1.8518857050131492, + "learning_rate": 1.83031262583195e-05, + "loss": 0.9451, + "step": 9021 + }, + { + "epoch": 0.21, + "grad_norm": 1.940345526256226, + "learning_rate": 1.8302700992971744e-05, + "loss": 1.0449, + "step": 9022 + }, + { + "epoch": 0.21, + "grad_norm": 2.097938346530536, + "learning_rate": 1.8302275679283033e-05, + "loss": 1.0249, + "step": 9023 + }, + { + "epoch": 0.21, + "grad_norm": 2.058193434049283, + "learning_rate": 1.8301850317255843e-05, + "loss": 1.0403, + "step": 9024 + }, + { + "epoch": 0.21, + "grad_norm": 1.8817384569143358, + "learning_rate": 1.8301424906892653e-05, + "loss": 1.0613, + "step": 9025 + }, + { + "epoch": 0.21, + "grad_norm": 1.9262971314627388, + "learning_rate": 1.830099944819594e-05, + "loss": 0.9567, + "step": 9026 + }, + { + "epoch": 0.21, + "grad_norm": 2.0596885429034084, + "learning_rate": 1.830057394116818e-05, + "loss": 1.0413, + "step": 9027 + }, + { + "epoch": 0.21, + "grad_norm": 1.174828751476912, + "learning_rate": 1.8300148385811848e-05, + "loss": 0.9508, + "step": 9028 + }, + { + "epoch": 0.21, + "grad_norm": 1.965503921100094, + "learning_rate": 1.8299722782129428e-05, + "loss": 1.1933, + "step": 9029 + }, + { + "epoch": 0.21, + "grad_norm": 1.9321938593688222, + "learning_rate": 1.8299297130123395e-05, + "loss": 1.1435, + "step": 9030 + }, + { + "epoch": 0.21, + "grad_norm": 2.074241233513453, + "learning_rate": 1.8298871429796223e-05, + "loss": 1.103, + "step": 9031 + }, + { + "epoch": 0.21, + "grad_norm": 2.075030826837729, + "learning_rate": 1.8298445681150395e-05, + "loss": 1.0878, + "step": 9032 + }, + { + "epoch": 0.21, + "grad_norm": 2.2301165118691437, + "learning_rate": 1.829801988418839e-05, + "loss": 1.0262, + "step": 9033 + }, + { + "epoch": 0.21, + "grad_norm": 1.9041415115704268, + "learning_rate": 1.8297594038912686e-05, + "loss": 1.1078, + "step": 9034 + }, + { + "epoch": 0.21, + "grad_norm": 2.1625597762177167, + "learning_rate": 1.8297168145325762e-05, + "loss": 1.0761, + "step": 9035 + }, + { + "epoch": 0.21, + "grad_norm": 2.058293642263443, + "learning_rate": 1.82967422034301e-05, + "loss": 1.0811, + "step": 9036 + }, + { + "epoch": 0.21, + "grad_norm": 1.8683466899320775, + "learning_rate": 1.8296316213228174e-05, + "loss": 0.9981, + "step": 9037 + }, + { + "epoch": 0.21, + "grad_norm": 2.003168903242493, + "learning_rate": 1.829589017472247e-05, + "loss": 0.9926, + "step": 9038 + }, + { + "epoch": 0.21, + "grad_norm": 2.0552047591274802, + "learning_rate": 1.829546408791547e-05, + "loss": 1.0968, + "step": 9039 + }, + { + "epoch": 0.21, + "grad_norm": 2.3698696813562425, + "learning_rate": 1.829503795280965e-05, + "loss": 1.1582, + "step": 9040 + }, + { + "epoch": 0.21, + "grad_norm": 1.9319975763505035, + "learning_rate": 1.8294611769407493e-05, + "loss": 1.0264, + "step": 9041 + }, + { + "epoch": 0.21, + "grad_norm": 4.821405286606961, + "learning_rate": 1.829418553771148e-05, + "loss": 0.9513, + "step": 9042 + }, + { + "epoch": 0.21, + "grad_norm": 1.9057882648826072, + "learning_rate": 1.8293759257724095e-05, + "loss": 1.1383, + "step": 9043 + }, + { + "epoch": 0.21, + "grad_norm": 1.9443225247621252, + "learning_rate": 1.8293332929447817e-05, + "loss": 1.1083, + "step": 9044 + }, + { + "epoch": 0.21, + "grad_norm": 2.1576821516158295, + "learning_rate": 1.8292906552885122e-05, + "loss": 1.0919, + "step": 9045 + }, + { + "epoch": 0.21, + "grad_norm": 2.0408439872066526, + "learning_rate": 1.8292480128038507e-05, + "loss": 1.0926, + "step": 9046 + }, + { + "epoch": 0.21, + "grad_norm": 1.927351330178612, + "learning_rate": 1.8292053654910447e-05, + "loss": 1.05, + "step": 9047 + }, + { + "epoch": 0.21, + "grad_norm": 1.9368279596868045, + "learning_rate": 1.8291627133503424e-05, + "loss": 0.988, + "step": 9048 + }, + { + "epoch": 0.21, + "grad_norm": 1.0913995804076018, + "learning_rate": 1.829120056381992e-05, + "loss": 1.0107, + "step": 9049 + }, + { + "epoch": 0.21, + "grad_norm": 1.846296781732412, + "learning_rate": 1.8290773945862428e-05, + "loss": 1.1756, + "step": 9050 + }, + { + "epoch": 0.21, + "grad_norm": 2.2062515006983885, + "learning_rate": 1.829034727963342e-05, + "loss": 1.081, + "step": 9051 + }, + { + "epoch": 0.21, + "grad_norm": 1.1692284802359232, + "learning_rate": 1.828992056513539e-05, + "loss": 0.9886, + "step": 9052 + }, + { + "epoch": 0.21, + "grad_norm": 2.0868209949540963, + "learning_rate": 1.8289493802370813e-05, + "loss": 1.0326, + "step": 9053 + }, + { + "epoch": 0.21, + "grad_norm": 2.414907283423103, + "learning_rate": 1.8289066991342182e-05, + "loss": 1.0349, + "step": 9054 + }, + { + "epoch": 0.21, + "grad_norm": 2.0128259541221865, + "learning_rate": 1.8288640132051975e-05, + "loss": 1.0552, + "step": 9055 + }, + { + "epoch": 0.21, + "grad_norm": 2.104051523362527, + "learning_rate": 1.8288213224502685e-05, + "loss": 1.0944, + "step": 9056 + }, + { + "epoch": 0.21, + "grad_norm": 1.8823851053555676, + "learning_rate": 1.828778626869679e-05, + "loss": 0.9875, + "step": 9057 + }, + { + "epoch": 0.21, + "grad_norm": 2.001873207241657, + "learning_rate": 1.828735926463678e-05, + "loss": 1.047, + "step": 9058 + }, + { + "epoch": 0.21, + "grad_norm": 1.9806124379487393, + "learning_rate": 1.8286932212325143e-05, + "loss": 1.0565, + "step": 9059 + }, + { + "epoch": 0.21, + "grad_norm": 2.0169787734171014, + "learning_rate": 1.828650511176436e-05, + "loss": 1.1056, + "step": 9060 + }, + { + "epoch": 0.21, + "grad_norm": 2.1641312415690495, + "learning_rate": 1.8286077962956926e-05, + "loss": 0.9838, + "step": 9061 + }, + { + "epoch": 0.21, + "grad_norm": 3.086430199364516, + "learning_rate": 1.828565076590532e-05, + "loss": 0.9871, + "step": 9062 + }, + { + "epoch": 0.21, + "grad_norm": 2.0168144013308096, + "learning_rate": 1.828522352061203e-05, + "loss": 1.0903, + "step": 9063 + }, + { + "epoch": 0.21, + "grad_norm": 2.2242900230454854, + "learning_rate": 1.828479622707955e-05, + "loss": 1.1172, + "step": 9064 + }, + { + "epoch": 0.21, + "grad_norm": 2.3633018735111855, + "learning_rate": 1.828436888531036e-05, + "loss": 0.9635, + "step": 9065 + }, + { + "epoch": 0.21, + "grad_norm": 1.9684438343204695, + "learning_rate": 1.828394149530695e-05, + "loss": 1.0035, + "step": 9066 + }, + { + "epoch": 0.21, + "grad_norm": 2.369428545848556, + "learning_rate": 1.8283514057071813e-05, + "loss": 1.0441, + "step": 9067 + }, + { + "epoch": 0.21, + "grad_norm": 2.1191276348832155, + "learning_rate": 1.8283086570607434e-05, + "loss": 1.0766, + "step": 9068 + }, + { + "epoch": 0.21, + "grad_norm": 1.955222407575235, + "learning_rate": 1.8282659035916302e-05, + "loss": 0.9884, + "step": 9069 + }, + { + "epoch": 0.21, + "grad_norm": 2.219920900955218, + "learning_rate": 1.828223145300091e-05, + "loss": 1.0609, + "step": 9070 + }, + { + "epoch": 0.21, + "grad_norm": 2.0608714974546585, + "learning_rate": 1.828180382186374e-05, + "loss": 1.1685, + "step": 9071 + }, + { + "epoch": 0.21, + "grad_norm": 2.1112481103409966, + "learning_rate": 1.8281376142507288e-05, + "loss": 1.07, + "step": 9072 + }, + { + "epoch": 0.21, + "grad_norm": 2.0770608707984457, + "learning_rate": 1.8280948414934042e-05, + "loss": 0.9677, + "step": 9073 + }, + { + "epoch": 0.21, + "grad_norm": 2.3164712907206058, + "learning_rate": 1.828052063914649e-05, + "loss": 0.9902, + "step": 9074 + }, + { + "epoch": 0.21, + "grad_norm": 2.099199763308177, + "learning_rate": 1.828009281514713e-05, + "loss": 0.924, + "step": 9075 + }, + { + "epoch": 0.21, + "grad_norm": 2.0234923628002788, + "learning_rate": 1.8279664942938448e-05, + "loss": 1.0474, + "step": 9076 + }, + { + "epoch": 0.21, + "grad_norm": 1.8948974400491878, + "learning_rate": 1.8279237022522932e-05, + "loss": 0.9914, + "step": 9077 + }, + { + "epoch": 0.21, + "grad_norm": 2.3998382748019726, + "learning_rate": 1.8278809053903076e-05, + "loss": 1.0968, + "step": 9078 + }, + { + "epoch": 0.21, + "grad_norm": 2.251578257919718, + "learning_rate": 1.8278381037081377e-05, + "loss": 1.0277, + "step": 9079 + }, + { + "epoch": 0.21, + "grad_norm": 2.0177516992995184, + "learning_rate": 1.827795297206032e-05, + "loss": 0.9624, + "step": 9080 + }, + { + "epoch": 0.21, + "grad_norm": 2.2159516502400507, + "learning_rate": 1.8277524858842404e-05, + "loss": 1.0568, + "step": 9081 + }, + { + "epoch": 0.21, + "grad_norm": 1.8552718862450182, + "learning_rate": 1.827709669743011e-05, + "loss": 1.0976, + "step": 9082 + }, + { + "epoch": 0.21, + "grad_norm": 1.8951894768827464, + "learning_rate": 1.8276668487825946e-05, + "loss": 1.0612, + "step": 9083 + }, + { + "epoch": 0.21, + "grad_norm": 2.1404613186261834, + "learning_rate": 1.8276240230032396e-05, + "loss": 0.9644, + "step": 9084 + }, + { + "epoch": 0.21, + "grad_norm": 1.9732248740204579, + "learning_rate": 1.8275811924051955e-05, + "loss": 0.937, + "step": 9085 + }, + { + "epoch": 0.21, + "grad_norm": 2.1506022641825258, + "learning_rate": 1.8275383569887112e-05, + "loss": 1.0258, + "step": 9086 + }, + { + "epoch": 0.21, + "grad_norm": 1.2254972942561038, + "learning_rate": 1.827495516754037e-05, + "loss": 1.0194, + "step": 9087 + }, + { + "epoch": 0.21, + "grad_norm": 1.1718490516253433, + "learning_rate": 1.827452671701422e-05, + "loss": 0.988, + "step": 9088 + }, + { + "epoch": 0.21, + "grad_norm": 2.663844730598729, + "learning_rate": 1.8274098218311154e-05, + "loss": 1.2075, + "step": 9089 + }, + { + "epoch": 0.21, + "grad_norm": 2.040202394829533, + "learning_rate": 1.827366967143367e-05, + "loss": 1.013, + "step": 9090 + }, + { + "epoch": 0.21, + "grad_norm": 2.348587059875998, + "learning_rate": 1.8273241076384264e-05, + "loss": 1.0105, + "step": 9091 + }, + { + "epoch": 0.21, + "grad_norm": 2.1922048751396708, + "learning_rate": 1.8272812433165424e-05, + "loss": 0.9778, + "step": 9092 + }, + { + "epoch": 0.21, + "grad_norm": 2.1765332309359686, + "learning_rate": 1.8272383741779652e-05, + "loss": 1.0426, + "step": 9093 + }, + { + "epoch": 0.21, + "grad_norm": 2.1578825490297078, + "learning_rate": 1.8271955002229446e-05, + "loss": 0.9797, + "step": 9094 + }, + { + "epoch": 0.21, + "grad_norm": 2.082688423253345, + "learning_rate": 1.8271526214517298e-05, + "loss": 1.1072, + "step": 9095 + }, + { + "epoch": 0.21, + "grad_norm": 2.21402379635941, + "learning_rate": 1.8271097378645703e-05, + "loss": 1.1134, + "step": 9096 + }, + { + "epoch": 0.21, + "grad_norm": 2.011651881736207, + "learning_rate": 1.827066849461716e-05, + "loss": 0.8984, + "step": 9097 + }, + { + "epoch": 0.21, + "grad_norm": 2.4129184560960084, + "learning_rate": 1.827023956243417e-05, + "loss": 1.0042, + "step": 9098 + }, + { + "epoch": 0.21, + "grad_norm": 2.0010612215291426, + "learning_rate": 1.826981058209923e-05, + "loss": 1.037, + "step": 9099 + }, + { + "epoch": 0.21, + "grad_norm": 2.0376556485587427, + "learning_rate": 1.826938155361483e-05, + "loss": 1.0536, + "step": 9100 + }, + { + "epoch": 0.21, + "grad_norm": 2.0566804142532793, + "learning_rate": 1.8268952476983473e-05, + "loss": 0.9967, + "step": 9101 + }, + { + "epoch": 0.21, + "grad_norm": 2.3177339600858966, + "learning_rate": 1.8268523352207656e-05, + "loss": 1.0794, + "step": 9102 + }, + { + "epoch": 0.21, + "grad_norm": 2.068766586830201, + "learning_rate": 1.826809417928988e-05, + "loss": 0.988, + "step": 9103 + }, + { + "epoch": 0.21, + "grad_norm": 2.251946083679837, + "learning_rate": 1.826766495823264e-05, + "loss": 1.122, + "step": 9104 + }, + { + "epoch": 0.21, + "grad_norm": 1.9864530113836465, + "learning_rate": 1.826723568903844e-05, + "loss": 1.0196, + "step": 9105 + }, + { + "epoch": 0.21, + "grad_norm": 2.10428887268005, + "learning_rate": 1.8266806371709773e-05, + "loss": 1.2008, + "step": 9106 + }, + { + "epoch": 0.21, + "grad_norm": 2.0108653737610838, + "learning_rate": 1.8266377006249144e-05, + "loss": 1.0729, + "step": 9107 + }, + { + "epoch": 0.21, + "grad_norm": 1.9715153195566113, + "learning_rate": 1.826594759265905e-05, + "loss": 1.1303, + "step": 9108 + }, + { + "epoch": 0.21, + "grad_norm": 2.0574552348256803, + "learning_rate": 1.8265518130941992e-05, + "loss": 1.0988, + "step": 9109 + }, + { + "epoch": 0.21, + "grad_norm": 2.0587417838401545, + "learning_rate": 1.8265088621100474e-05, + "loss": 1.0502, + "step": 9110 + }, + { + "epoch": 0.21, + "grad_norm": 2.0616581078620397, + "learning_rate": 1.8264659063136988e-05, + "loss": 1.0992, + "step": 9111 + }, + { + "epoch": 0.21, + "grad_norm": 1.2090263551994385, + "learning_rate": 1.8264229457054045e-05, + "loss": 0.977, + "step": 9112 + }, + { + "epoch": 0.21, + "grad_norm": 1.9455783400951232, + "learning_rate": 1.826379980285414e-05, + "loss": 1.0843, + "step": 9113 + }, + { + "epoch": 0.21, + "grad_norm": 2.1906892401194455, + "learning_rate": 1.826337010053978e-05, + "loss": 1.0578, + "step": 9114 + }, + { + "epoch": 0.21, + "grad_norm": 2.112124723683724, + "learning_rate": 1.8262940350113456e-05, + "loss": 1.0585, + "step": 9115 + }, + { + "epoch": 0.21, + "grad_norm": 2.7893966767952914, + "learning_rate": 1.8262510551577678e-05, + "loss": 0.9427, + "step": 9116 + }, + { + "epoch": 0.21, + "grad_norm": 2.1280887275381253, + "learning_rate": 1.8262080704934955e-05, + "loss": 0.9927, + "step": 9117 + }, + { + "epoch": 0.21, + "grad_norm": 1.1353849700417329, + "learning_rate": 1.8261650810187778e-05, + "loss": 0.9743, + "step": 9118 + }, + { + "epoch": 0.21, + "grad_norm": 1.9743935897968379, + "learning_rate": 1.8261220867338653e-05, + "loss": 0.9268, + "step": 9119 + }, + { + "epoch": 0.21, + "grad_norm": 2.296379196917937, + "learning_rate": 1.8260790876390085e-05, + "loss": 1.1025, + "step": 9120 + }, + { + "epoch": 0.21, + "grad_norm": 1.9127867646325365, + "learning_rate": 1.826036083734458e-05, + "loss": 0.9979, + "step": 9121 + }, + { + "epoch": 0.21, + "grad_norm": 2.020381810401869, + "learning_rate": 1.8259930750204637e-05, + "loss": 1.1359, + "step": 9122 + }, + { + "epoch": 0.21, + "grad_norm": 2.1300286738723995, + "learning_rate": 1.825950061497276e-05, + "loss": 1.01, + "step": 9123 + }, + { + "epoch": 0.21, + "grad_norm": 1.7521188303026018, + "learning_rate": 1.8259070431651457e-05, + "loss": 1.209, + "step": 9124 + }, + { + "epoch": 0.21, + "grad_norm": 2.0104019268292124, + "learning_rate": 1.8258640200243235e-05, + "loss": 0.9069, + "step": 9125 + }, + { + "epoch": 0.22, + "grad_norm": 2.0807513443750025, + "learning_rate": 1.825820992075059e-05, + "loss": 1.1379, + "step": 9126 + }, + { + "epoch": 0.22, + "grad_norm": 2.283650365872183, + "learning_rate": 1.8257779593176034e-05, + "loss": 1.0982, + "step": 9127 + }, + { + "epoch": 0.22, + "grad_norm": 1.6883122759018738, + "learning_rate": 1.8257349217522072e-05, + "loss": 1.004, + "step": 9128 + }, + { + "epoch": 0.22, + "grad_norm": 2.337454045096266, + "learning_rate": 1.8256918793791208e-05, + "loss": 1.0714, + "step": 9129 + }, + { + "epoch": 0.22, + "grad_norm": 1.9847272850968463, + "learning_rate": 1.825648832198595e-05, + "loss": 1.0695, + "step": 9130 + }, + { + "epoch": 0.22, + "grad_norm": 1.927319283904184, + "learning_rate": 1.82560578021088e-05, + "loss": 1.0871, + "step": 9131 + }, + { + "epoch": 0.22, + "grad_norm": 1.9388625309612422, + "learning_rate": 1.8255627234162267e-05, + "loss": 0.9913, + "step": 9132 + }, + { + "epoch": 0.22, + "grad_norm": 2.040219203921601, + "learning_rate": 1.8255196618148863e-05, + "loss": 1.1188, + "step": 9133 + }, + { + "epoch": 0.22, + "grad_norm": 1.8947239434918488, + "learning_rate": 1.825476595407109e-05, + "loss": 1.0175, + "step": 9134 + }, + { + "epoch": 0.22, + "grad_norm": 1.8895156476088688, + "learning_rate": 1.8254335241931453e-05, + "loss": 1.1148, + "step": 9135 + }, + { + "epoch": 0.22, + "grad_norm": 2.479590355603866, + "learning_rate": 1.8253904481732464e-05, + "loss": 0.9248, + "step": 9136 + }, + { + "epoch": 0.22, + "grad_norm": 1.7646758897278942, + "learning_rate": 1.825347367347663e-05, + "loss": 1.0845, + "step": 9137 + }, + { + "epoch": 0.22, + "grad_norm": 2.1243743111203024, + "learning_rate": 1.8253042817166458e-05, + "loss": 0.8973, + "step": 9138 + }, + { + "epoch": 0.22, + "grad_norm": 2.0566431596392585, + "learning_rate": 1.825261191280446e-05, + "loss": 1.0453, + "step": 9139 + }, + { + "epoch": 0.22, + "grad_norm": 2.0721240526160853, + "learning_rate": 1.8252180960393145e-05, + "loss": 1.0628, + "step": 9140 + }, + { + "epoch": 0.22, + "grad_norm": 1.8365400554406572, + "learning_rate": 1.8251749959935017e-05, + "loss": 1.0751, + "step": 9141 + }, + { + "epoch": 0.22, + "grad_norm": 1.1467260850422152, + "learning_rate": 1.8251318911432586e-05, + "loss": 1.0269, + "step": 9142 + }, + { + "epoch": 0.22, + "grad_norm": 2.4172874635412547, + "learning_rate": 1.8250887814888368e-05, + "loss": 1.0241, + "step": 9143 + }, + { + "epoch": 0.22, + "grad_norm": 2.6809473731376565, + "learning_rate": 1.8250456670304866e-05, + "loss": 1.0958, + "step": 9144 + }, + { + "epoch": 0.22, + "grad_norm": 1.9516066349890018, + "learning_rate": 1.8250025477684598e-05, + "loss": 1.068, + "step": 9145 + }, + { + "epoch": 0.22, + "grad_norm": 2.0074672541302125, + "learning_rate": 1.8249594237030064e-05, + "loss": 0.9726, + "step": 9146 + }, + { + "epoch": 0.22, + "grad_norm": 2.1814332493574793, + "learning_rate": 1.8249162948343784e-05, + "loss": 1.0398, + "step": 9147 + }, + { + "epoch": 0.22, + "grad_norm": 2.370250297953284, + "learning_rate": 1.8248731611628266e-05, + "loss": 1.0377, + "step": 9148 + }, + { + "epoch": 0.22, + "grad_norm": 2.0100575320590335, + "learning_rate": 1.8248300226886018e-05, + "loss": 1.0813, + "step": 9149 + }, + { + "epoch": 0.22, + "grad_norm": 2.4223574120416416, + "learning_rate": 1.8247868794119556e-05, + "loss": 1.0528, + "step": 9150 + }, + { + "epoch": 0.22, + "grad_norm": 1.940372850174104, + "learning_rate": 1.8247437313331394e-05, + "loss": 1.1422, + "step": 9151 + }, + { + "epoch": 0.22, + "grad_norm": 1.1295137649627491, + "learning_rate": 1.8247005784524034e-05, + "loss": 0.9469, + "step": 9152 + }, + { + "epoch": 0.22, + "grad_norm": 1.1233717360425957, + "learning_rate": 1.82465742077e-05, + "loss": 0.9093, + "step": 9153 + }, + { + "epoch": 0.22, + "grad_norm": 1.7641901888166431, + "learning_rate": 1.8246142582861802e-05, + "loss": 1.0388, + "step": 9154 + }, + { + "epoch": 0.22, + "grad_norm": 2.0635472785180813, + "learning_rate": 1.8245710910011945e-05, + "loss": 1.081, + "step": 9155 + }, + { + "epoch": 0.22, + "grad_norm": 2.2653904460478436, + "learning_rate": 1.8245279189152952e-05, + "loss": 1.0293, + "step": 9156 + }, + { + "epoch": 0.22, + "grad_norm": 1.8539234732409757, + "learning_rate": 1.8244847420287334e-05, + "loss": 1.083, + "step": 9157 + }, + { + "epoch": 0.22, + "grad_norm": 2.0393163160635517, + "learning_rate": 1.8244415603417603e-05, + "loss": 1.11, + "step": 9158 + }, + { + "epoch": 0.22, + "grad_norm": 2.336224348280032, + "learning_rate": 1.8243983738546276e-05, + "loss": 1.0198, + "step": 9159 + }, + { + "epoch": 0.22, + "grad_norm": 1.951204273644134, + "learning_rate": 1.8243551825675864e-05, + "loss": 1.0335, + "step": 9160 + }, + { + "epoch": 0.22, + "grad_norm": 2.0957012722557353, + "learning_rate": 1.8243119864808886e-05, + "loss": 0.9068, + "step": 9161 + }, + { + "epoch": 0.22, + "grad_norm": 1.9749076174209264, + "learning_rate": 1.824268785594785e-05, + "loss": 0.9797, + "step": 9162 + }, + { + "epoch": 0.22, + "grad_norm": 2.3119311090050227, + "learning_rate": 1.824225579909528e-05, + "loss": 1.0236, + "step": 9163 + }, + { + "epoch": 0.22, + "grad_norm": 3.3160840805137006, + "learning_rate": 1.824182369425368e-05, + "loss": 1.0802, + "step": 9164 + }, + { + "epoch": 0.22, + "grad_norm": 2.063273344829131, + "learning_rate": 1.824139154142558e-05, + "loss": 0.9674, + "step": 9165 + }, + { + "epoch": 0.22, + "grad_norm": 2.1803702217531207, + "learning_rate": 1.8240959340613488e-05, + "loss": 1.0465, + "step": 9166 + }, + { + "epoch": 0.22, + "grad_norm": 2.1093797648340735, + "learning_rate": 1.8240527091819923e-05, + "loss": 1.0268, + "step": 9167 + }, + { + "epoch": 0.22, + "grad_norm": 1.9192880131563144, + "learning_rate": 1.82400947950474e-05, + "loss": 1.1596, + "step": 9168 + }, + { + "epoch": 0.22, + "grad_norm": 1.8902769534993804, + "learning_rate": 1.8239662450298433e-05, + "loss": 1.1211, + "step": 9169 + }, + { + "epoch": 0.22, + "grad_norm": 2.075547163137991, + "learning_rate": 1.8239230057575542e-05, + "loss": 1.0135, + "step": 9170 + }, + { + "epoch": 0.22, + "grad_norm": 2.201812320737686, + "learning_rate": 1.823879761688125e-05, + "loss": 1.2715, + "step": 9171 + }, + { + "epoch": 0.22, + "grad_norm": 4.62551670927487, + "learning_rate": 1.8238365128218066e-05, + "loss": 0.8995, + "step": 9172 + }, + { + "epoch": 0.22, + "grad_norm": 2.0375231829347142, + "learning_rate": 1.8237932591588513e-05, + "loss": 1.0279, + "step": 9173 + }, + { + "epoch": 0.22, + "grad_norm": 2.2150543470606276, + "learning_rate": 1.823750000699511e-05, + "loss": 1.1397, + "step": 9174 + }, + { + "epoch": 0.22, + "grad_norm": 1.9450873809967426, + "learning_rate": 1.8237067374440372e-05, + "loss": 1.0131, + "step": 9175 + }, + { + "epoch": 0.22, + "grad_norm": 1.8574805770010518, + "learning_rate": 1.823663469392682e-05, + "loss": 1.0603, + "step": 9176 + }, + { + "epoch": 0.22, + "grad_norm": 2.293307885198835, + "learning_rate": 1.8236201965456975e-05, + "loss": 1.0082, + "step": 9177 + }, + { + "epoch": 0.22, + "grad_norm": 2.205333781690037, + "learning_rate": 1.8235769189033354e-05, + "loss": 1.0825, + "step": 9178 + }, + { + "epoch": 0.22, + "grad_norm": 2.1890622359186485, + "learning_rate": 1.8235336364658475e-05, + "loss": 1.0874, + "step": 9179 + }, + { + "epoch": 0.22, + "grad_norm": 2.2090671644582827, + "learning_rate": 1.8234903492334864e-05, + "loss": 1.0805, + "step": 9180 + }, + { + "epoch": 0.22, + "grad_norm": 2.1681158117406043, + "learning_rate": 1.8234470572065034e-05, + "loss": 1.0358, + "step": 9181 + }, + { + "epoch": 0.22, + "grad_norm": 1.8105341265053658, + "learning_rate": 1.8234037603851512e-05, + "loss": 1.0177, + "step": 9182 + }, + { + "epoch": 0.22, + "grad_norm": 1.8912179992023521, + "learning_rate": 1.8233604587696818e-05, + "loss": 1.0811, + "step": 9183 + }, + { + "epoch": 0.22, + "grad_norm": 1.875584391765897, + "learning_rate": 1.823317152360347e-05, + "loss": 1.1512, + "step": 9184 + }, + { + "epoch": 0.22, + "grad_norm": 1.9513600606399695, + "learning_rate": 1.823273841157399e-05, + "loss": 0.9696, + "step": 9185 + }, + { + "epoch": 0.22, + "grad_norm": 2.1716568199608766, + "learning_rate": 1.82323052516109e-05, + "loss": 1.0452, + "step": 9186 + }, + { + "epoch": 0.22, + "grad_norm": 1.9537739167463528, + "learning_rate": 1.823187204371672e-05, + "loss": 1.1445, + "step": 9187 + }, + { + "epoch": 0.22, + "grad_norm": 1.831538380759749, + "learning_rate": 1.8231438787893978e-05, + "loss": 1.0003, + "step": 9188 + }, + { + "epoch": 0.22, + "grad_norm": 1.9495249341241536, + "learning_rate": 1.8231005484145194e-05, + "loss": 1.0178, + "step": 9189 + }, + { + "epoch": 0.22, + "grad_norm": 2.0259151155541684, + "learning_rate": 1.8230572132472888e-05, + "loss": 1.1758, + "step": 9190 + }, + { + "epoch": 0.22, + "grad_norm": 1.165412828687258, + "learning_rate": 1.8230138732879588e-05, + "loss": 0.9805, + "step": 9191 + }, + { + "epoch": 0.22, + "grad_norm": 2.076480488838648, + "learning_rate": 1.8229705285367812e-05, + "loss": 1.0731, + "step": 9192 + }, + { + "epoch": 0.22, + "grad_norm": 1.9779311583775216, + "learning_rate": 1.822927178994009e-05, + "loss": 0.939, + "step": 9193 + }, + { + "epoch": 0.22, + "grad_norm": 2.1842282557773545, + "learning_rate": 1.822883824659894e-05, + "loss": 0.912, + "step": 9194 + }, + { + "epoch": 0.22, + "grad_norm": 2.175725112615454, + "learning_rate": 1.8228404655346885e-05, + "loss": 1.0145, + "step": 9195 + }, + { + "epoch": 0.22, + "grad_norm": 2.394466879424394, + "learning_rate": 1.8227971016186457e-05, + "loss": 0.9912, + "step": 9196 + }, + { + "epoch": 0.22, + "grad_norm": 1.912536958189515, + "learning_rate": 1.8227537329120177e-05, + "loss": 1.0114, + "step": 9197 + }, + { + "epoch": 0.22, + "grad_norm": 2.1311306699870727, + "learning_rate": 1.8227103594150568e-05, + "loss": 1.1063, + "step": 9198 + }, + { + "epoch": 0.22, + "grad_norm": 2.217480776391529, + "learning_rate": 1.822666981128016e-05, + "loss": 0.945, + "step": 9199 + }, + { + "epoch": 0.22, + "grad_norm": 2.345294426154099, + "learning_rate": 1.8226235980511476e-05, + "loss": 0.9793, + "step": 9200 + }, + { + "epoch": 0.22, + "grad_norm": 2.1293693736496495, + "learning_rate": 1.8225802101847038e-05, + "loss": 1.0717, + "step": 9201 + }, + { + "epoch": 0.22, + "grad_norm": 1.6841018615588164, + "learning_rate": 1.8225368175289378e-05, + "loss": 1.0939, + "step": 9202 + }, + { + "epoch": 0.22, + "grad_norm": 2.251622405461414, + "learning_rate": 1.822493420084102e-05, + "loss": 1.1344, + "step": 9203 + }, + { + "epoch": 0.22, + "grad_norm": 1.9172632704191803, + "learning_rate": 1.822450017850449e-05, + "loss": 0.9726, + "step": 9204 + }, + { + "epoch": 0.22, + "grad_norm": 1.972063484573936, + "learning_rate": 1.8224066108282318e-05, + "loss": 1.0425, + "step": 9205 + }, + { + "epoch": 0.22, + "grad_norm": 1.888651325770862, + "learning_rate": 1.8223631990177027e-05, + "loss": 1.0094, + "step": 9206 + }, + { + "epoch": 0.22, + "grad_norm": 2.1677723330582483, + "learning_rate": 1.822319782419115e-05, + "loss": 0.9625, + "step": 9207 + }, + { + "epoch": 0.22, + "grad_norm": 2.3143224447917072, + "learning_rate": 1.822276361032721e-05, + "loss": 1.0645, + "step": 9208 + }, + { + "epoch": 0.22, + "grad_norm": 2.153215608138036, + "learning_rate": 1.822232934858774e-05, + "loss": 1.0192, + "step": 9209 + }, + { + "epoch": 0.22, + "grad_norm": 1.098343867664016, + "learning_rate": 1.822189503897526e-05, + "loss": 1.0327, + "step": 9210 + }, + { + "epoch": 0.22, + "grad_norm": 2.364622650551758, + "learning_rate": 1.822146068149231e-05, + "loss": 0.9789, + "step": 9211 + }, + { + "epoch": 0.22, + "grad_norm": 1.9703574660179246, + "learning_rate": 1.822102627614141e-05, + "loss": 1.0724, + "step": 9212 + }, + { + "epoch": 0.22, + "grad_norm": 1.0973726569945792, + "learning_rate": 1.822059182292509e-05, + "loss": 0.9119, + "step": 9213 + }, + { + "epoch": 0.22, + "grad_norm": 2.1685538055593696, + "learning_rate": 1.8220157321845887e-05, + "loss": 1.1809, + "step": 9214 + }, + { + "epoch": 0.22, + "grad_norm": 2.188923471433385, + "learning_rate": 1.821972277290632e-05, + "loss": 1.0805, + "step": 9215 + }, + { + "epoch": 0.22, + "grad_norm": 1.6295309532654303, + "learning_rate": 1.8219288176108928e-05, + "loss": 0.987, + "step": 9216 + }, + { + "epoch": 0.22, + "grad_norm": 2.0482215195854656, + "learning_rate": 1.821885353145624e-05, + "loss": 0.9687, + "step": 9217 + }, + { + "epoch": 0.22, + "grad_norm": 2.0995516755803294, + "learning_rate": 1.821841883895078e-05, + "loss": 1.1984, + "step": 9218 + }, + { + "epoch": 0.22, + "grad_norm": 2.122605228449265, + "learning_rate": 1.8217984098595086e-05, + "loss": 1.0396, + "step": 9219 + }, + { + "epoch": 0.22, + "grad_norm": 1.9631383508251725, + "learning_rate": 1.8217549310391687e-05, + "loss": 1.0406, + "step": 9220 + }, + { + "epoch": 0.22, + "grad_norm": 2.126866609698409, + "learning_rate": 1.8217114474343113e-05, + "loss": 1.1583, + "step": 9221 + }, + { + "epoch": 0.22, + "grad_norm": 2.087261325467517, + "learning_rate": 1.82166795904519e-05, + "loss": 1.0108, + "step": 9222 + }, + { + "epoch": 0.22, + "grad_norm": 2.2979725377458897, + "learning_rate": 1.8216244658720573e-05, + "loss": 1.1386, + "step": 9223 + }, + { + "epoch": 0.22, + "grad_norm": 1.9645802760129913, + "learning_rate": 1.821580967915167e-05, + "loss": 1.0338, + "step": 9224 + }, + { + "epoch": 0.22, + "grad_norm": 1.8820767210869707, + "learning_rate": 1.8215374651747723e-05, + "loss": 0.9785, + "step": 9225 + }, + { + "epoch": 0.22, + "grad_norm": 2.1621413843418194, + "learning_rate": 1.8214939576511265e-05, + "loss": 1.032, + "step": 9226 + }, + { + "epoch": 0.22, + "grad_norm": 1.9564787801438372, + "learning_rate": 1.8214504453444825e-05, + "loss": 0.9253, + "step": 9227 + }, + { + "epoch": 0.22, + "grad_norm": 1.9408485035491443, + "learning_rate": 1.821406928255094e-05, + "loss": 1.2541, + "step": 9228 + }, + { + "epoch": 0.22, + "grad_norm": 1.9135243614235087, + "learning_rate": 1.8213634063832146e-05, + "loss": 1.0979, + "step": 9229 + }, + { + "epoch": 0.22, + "grad_norm": 1.112166642740128, + "learning_rate": 1.8213198797290973e-05, + "loss": 0.9829, + "step": 9230 + }, + { + "epoch": 0.22, + "grad_norm": 1.0607423815097536, + "learning_rate": 1.821276348292995e-05, + "loss": 0.9119, + "step": 9231 + }, + { + "epoch": 0.22, + "grad_norm": 1.9430644412974287, + "learning_rate": 1.8212328120751626e-05, + "loss": 1.1658, + "step": 9232 + }, + { + "epoch": 0.22, + "grad_norm": 2.286278548416821, + "learning_rate": 1.8211892710758525e-05, + "loss": 1.0422, + "step": 9233 + }, + { + "epoch": 0.22, + "grad_norm": 2.2399136382059592, + "learning_rate": 1.8211457252953185e-05, + "loss": 1.0207, + "step": 9234 + }, + { + "epoch": 0.22, + "grad_norm": 1.9833673078422143, + "learning_rate": 1.821102174733814e-05, + "loss": 1.1294, + "step": 9235 + }, + { + "epoch": 0.22, + "grad_norm": 1.8539589462525252, + "learning_rate": 1.8210586193915928e-05, + "loss": 1.0417, + "step": 9236 + }, + { + "epoch": 0.22, + "grad_norm": 1.0994744231794151, + "learning_rate": 1.8210150592689083e-05, + "loss": 1.0001, + "step": 9237 + }, + { + "epoch": 0.22, + "grad_norm": 2.267471968439665, + "learning_rate": 1.8209714943660143e-05, + "loss": 1.0545, + "step": 9238 + }, + { + "epoch": 0.22, + "grad_norm": 2.662589437813792, + "learning_rate": 1.8209279246831643e-05, + "loss": 1.0152, + "step": 9239 + }, + { + "epoch": 0.22, + "grad_norm": 2.3638838881394593, + "learning_rate": 1.8208843502206118e-05, + "loss": 1.0127, + "step": 9240 + }, + { + "epoch": 0.22, + "grad_norm": 5.148316226500136, + "learning_rate": 1.8208407709786107e-05, + "loss": 1.0903, + "step": 9241 + }, + { + "epoch": 0.22, + "grad_norm": 1.9024906025882677, + "learning_rate": 1.8207971869574154e-05, + "loss": 1.1217, + "step": 9242 + }, + { + "epoch": 0.22, + "grad_norm": 2.1348319677015417, + "learning_rate": 1.8207535981572785e-05, + "loss": 1.0686, + "step": 9243 + }, + { + "epoch": 0.22, + "grad_norm": 2.205366877241132, + "learning_rate": 1.8207100045784544e-05, + "loss": 1.0166, + "step": 9244 + }, + { + "epoch": 0.22, + "grad_norm": 2.157396972275273, + "learning_rate": 1.8206664062211966e-05, + "loss": 1.0987, + "step": 9245 + }, + { + "epoch": 0.22, + "grad_norm": 2.1961901355376563, + "learning_rate": 1.8206228030857595e-05, + "loss": 0.9095, + "step": 9246 + }, + { + "epoch": 0.22, + "grad_norm": 2.4940577084732247, + "learning_rate": 1.8205791951723966e-05, + "loss": 1.1487, + "step": 9247 + }, + { + "epoch": 0.22, + "grad_norm": 2.078141378328309, + "learning_rate": 1.8205355824813617e-05, + "loss": 1.1391, + "step": 9248 + }, + { + "epoch": 0.22, + "grad_norm": 2.090309128906493, + "learning_rate": 1.8204919650129088e-05, + "loss": 1.0428, + "step": 9249 + }, + { + "epoch": 0.22, + "grad_norm": 2.193703716539053, + "learning_rate": 1.8204483427672924e-05, + "loss": 0.9818, + "step": 9250 + }, + { + "epoch": 0.22, + "grad_norm": 1.9548850364302717, + "learning_rate": 1.8204047157447652e-05, + "loss": 0.9313, + "step": 9251 + }, + { + "epoch": 0.22, + "grad_norm": 1.9824809223575082, + "learning_rate": 1.8203610839455828e-05, + "loss": 1.1311, + "step": 9252 + }, + { + "epoch": 0.22, + "grad_norm": 2.480620500754852, + "learning_rate": 1.820317447369998e-05, + "loss": 1.0134, + "step": 9253 + }, + { + "epoch": 0.22, + "grad_norm": 1.0924843758666394, + "learning_rate": 1.8202738060182655e-05, + "loss": 1.0174, + "step": 9254 + }, + { + "epoch": 0.22, + "grad_norm": 2.1112459576539395, + "learning_rate": 1.8202301598906388e-05, + "loss": 1.0621, + "step": 9255 + }, + { + "epoch": 0.22, + "grad_norm": 1.1185532556804074, + "learning_rate": 1.820186508987373e-05, + "loss": 0.9606, + "step": 9256 + }, + { + "epoch": 0.22, + "grad_norm": 5.363173847736285, + "learning_rate": 1.8201428533087214e-05, + "loss": 1.0548, + "step": 9257 + }, + { + "epoch": 0.22, + "grad_norm": 1.9462735664269495, + "learning_rate": 1.820099192854938e-05, + "loss": 1.1146, + "step": 9258 + }, + { + "epoch": 0.22, + "grad_norm": 2.0796288114713746, + "learning_rate": 1.8200555276262784e-05, + "loss": 1.1674, + "step": 9259 + }, + { + "epoch": 0.22, + "grad_norm": 2.0401514668584615, + "learning_rate": 1.8200118576229953e-05, + "loss": 1.076, + "step": 9260 + }, + { + "epoch": 0.22, + "grad_norm": 1.987789070183668, + "learning_rate": 1.8199681828453438e-05, + "loss": 1.1956, + "step": 9261 + }, + { + "epoch": 0.22, + "grad_norm": 2.1257432578307824, + "learning_rate": 1.8199245032935775e-05, + "loss": 0.9789, + "step": 9262 + }, + { + "epoch": 0.22, + "grad_norm": 2.049008118234299, + "learning_rate": 1.8198808189679518e-05, + "loss": 1.1514, + "step": 9263 + }, + { + "epoch": 0.22, + "grad_norm": 2.0094689776091275, + "learning_rate": 1.81983712986872e-05, + "loss": 1.0549, + "step": 9264 + }, + { + "epoch": 0.22, + "grad_norm": 1.912781050021178, + "learning_rate": 1.819793435996137e-05, + "loss": 1.0303, + "step": 9265 + }, + { + "epoch": 0.22, + "grad_norm": 1.9514286178540836, + "learning_rate": 1.819749737350457e-05, + "loss": 1.0043, + "step": 9266 + }, + { + "epoch": 0.22, + "grad_norm": 1.1064096640343069, + "learning_rate": 1.8197060339319345e-05, + "loss": 1.0201, + "step": 9267 + }, + { + "epoch": 0.22, + "grad_norm": 2.4182853035095806, + "learning_rate": 1.8196623257408238e-05, + "loss": 1.0855, + "step": 9268 + }, + { + "epoch": 0.22, + "grad_norm": 2.0237505306576162, + "learning_rate": 1.8196186127773797e-05, + "loss": 1.084, + "step": 9269 + }, + { + "epoch": 0.22, + "grad_norm": 1.7876572894012372, + "learning_rate": 1.8195748950418567e-05, + "loss": 1.1263, + "step": 9270 + }, + { + "epoch": 0.22, + "grad_norm": 1.8603492447695975, + "learning_rate": 1.819531172534509e-05, + "loss": 1.0752, + "step": 9271 + }, + { + "epoch": 0.22, + "grad_norm": 2.1966409407707452, + "learning_rate": 1.8194874452555914e-05, + "loss": 1.1998, + "step": 9272 + }, + { + "epoch": 0.22, + "grad_norm": 1.123803191725516, + "learning_rate": 1.8194437132053586e-05, + "loss": 0.9605, + "step": 9273 + }, + { + "epoch": 0.22, + "grad_norm": 1.1053748293093235, + "learning_rate": 1.8193999763840647e-05, + "loss": 1.0015, + "step": 9274 + }, + { + "epoch": 0.22, + "grad_norm": 2.0828249464768493, + "learning_rate": 1.819356234791965e-05, + "loss": 1.1907, + "step": 9275 + }, + { + "epoch": 0.22, + "grad_norm": 2.7096492344927694, + "learning_rate": 1.819312488429314e-05, + "loss": 1.0685, + "step": 9276 + }, + { + "epoch": 0.22, + "grad_norm": 2.4149480007452127, + "learning_rate": 1.819268737296366e-05, + "loss": 1.0017, + "step": 9277 + }, + { + "epoch": 0.22, + "grad_norm": 2.2735373435758124, + "learning_rate": 1.8192249813933762e-05, + "loss": 1.0802, + "step": 9278 + }, + { + "epoch": 0.22, + "grad_norm": 2.019766851919176, + "learning_rate": 1.819181220720599e-05, + "loss": 0.982, + "step": 9279 + }, + { + "epoch": 0.22, + "grad_norm": 1.9576170895350093, + "learning_rate": 1.81913745527829e-05, + "loss": 1.0272, + "step": 9280 + }, + { + "epoch": 0.22, + "grad_norm": 1.9210595116476865, + "learning_rate": 1.8190936850667028e-05, + "loss": 1.1582, + "step": 9281 + }, + { + "epoch": 0.22, + "grad_norm": 2.593592384731463, + "learning_rate": 1.819049910086093e-05, + "loss": 1.0037, + "step": 9282 + }, + { + "epoch": 0.22, + "grad_norm": 1.160434876083422, + "learning_rate": 1.819006130336715e-05, + "loss": 1.0234, + "step": 9283 + }, + { + "epoch": 0.22, + "grad_norm": 2.357686406114834, + "learning_rate": 1.818962345818825e-05, + "loss": 1.1128, + "step": 9284 + }, + { + "epoch": 0.22, + "grad_norm": 1.9018973818458853, + "learning_rate": 1.818918556532676e-05, + "loss": 1.0926, + "step": 9285 + }, + { + "epoch": 0.22, + "grad_norm": 2.050964940677261, + "learning_rate": 1.818874762478524e-05, + "loss": 1.0353, + "step": 9286 + }, + { + "epoch": 0.22, + "grad_norm": 2.0230322887413545, + "learning_rate": 1.818830963656624e-05, + "loss": 1.0691, + "step": 9287 + }, + { + "epoch": 0.22, + "grad_norm": 2.0112776124471203, + "learning_rate": 1.818787160067231e-05, + "loss": 0.987, + "step": 9288 + }, + { + "epoch": 0.22, + "grad_norm": 2.2424029109632424, + "learning_rate": 1.8187433517106e-05, + "loss": 0.978, + "step": 9289 + }, + { + "epoch": 0.22, + "grad_norm": 2.0163859890624156, + "learning_rate": 1.8186995385869857e-05, + "loss": 0.9249, + "step": 9290 + }, + { + "epoch": 0.22, + "grad_norm": 5.117249492459037, + "learning_rate": 1.818655720696644e-05, + "loss": 1.0513, + "step": 9291 + }, + { + "epoch": 0.22, + "grad_norm": 1.18594674932868, + "learning_rate": 1.818611898039829e-05, + "loss": 0.9713, + "step": 9292 + }, + { + "epoch": 0.22, + "grad_norm": 1.9804740840207073, + "learning_rate": 1.8185680706167966e-05, + "loss": 0.9489, + "step": 9293 + }, + { + "epoch": 0.22, + "grad_norm": 2.772596277596428, + "learning_rate": 1.8185242384278016e-05, + "loss": 1.0835, + "step": 9294 + }, + { + "epoch": 0.22, + "grad_norm": 2.62836579686132, + "learning_rate": 1.8184804014730994e-05, + "loss": 1.0013, + "step": 9295 + }, + { + "epoch": 0.22, + "grad_norm": 2.0046742770259907, + "learning_rate": 1.8184365597529454e-05, + "loss": 1.0277, + "step": 9296 + }, + { + "epoch": 0.22, + "grad_norm": 1.9933245110448408, + "learning_rate": 1.8183927132675945e-05, + "loss": 1.0684, + "step": 9297 + }, + { + "epoch": 0.22, + "grad_norm": 2.0476467347797263, + "learning_rate": 1.818348862017302e-05, + "loss": 1.1694, + "step": 9298 + }, + { + "epoch": 0.22, + "grad_norm": 2.4073839958274172, + "learning_rate": 1.8183050060023233e-05, + "loss": 1.0483, + "step": 9299 + }, + { + "epoch": 0.22, + "grad_norm": 1.981370614398396, + "learning_rate": 1.818261145222914e-05, + "loss": 1.1553, + "step": 9300 + }, + { + "epoch": 0.22, + "grad_norm": 2.1658126525913737, + "learning_rate": 1.8182172796793292e-05, + "loss": 1.0764, + "step": 9301 + }, + { + "epoch": 0.22, + "grad_norm": 2.29023426968592, + "learning_rate": 1.8181734093718245e-05, + "loss": 1.0129, + "step": 9302 + }, + { + "epoch": 0.22, + "grad_norm": 2.0632946221470063, + "learning_rate": 1.818129534300655e-05, + "loss": 1.1238, + "step": 9303 + }, + { + "epoch": 0.22, + "grad_norm": 2.87119244443172, + "learning_rate": 1.8180856544660763e-05, + "loss": 1.0567, + "step": 9304 + }, + { + "epoch": 0.22, + "grad_norm": 2.029062295250186, + "learning_rate": 1.8180417698683437e-05, + "loss": 1.0923, + "step": 9305 + }, + { + "epoch": 0.22, + "grad_norm": 2.1318053149637715, + "learning_rate": 1.817997880507713e-05, + "loss": 1.0188, + "step": 9306 + }, + { + "epoch": 0.22, + "grad_norm": 2.0506073621416734, + "learning_rate": 1.81795398638444e-05, + "loss": 1.1072, + "step": 9307 + }, + { + "epoch": 0.22, + "grad_norm": 2.276550607408438, + "learning_rate": 1.81791008749878e-05, + "loss": 1.1312, + "step": 9308 + }, + { + "epoch": 0.22, + "grad_norm": 2.4465858634433295, + "learning_rate": 1.817866183850988e-05, + "loss": 1.1128, + "step": 9309 + }, + { + "epoch": 0.22, + "grad_norm": 2.7081949390836235, + "learning_rate": 1.8178222754413206e-05, + "loss": 1.1113, + "step": 9310 + }, + { + "epoch": 0.22, + "grad_norm": 2.038704458976023, + "learning_rate": 1.8177783622700328e-05, + "loss": 0.9808, + "step": 9311 + }, + { + "epoch": 0.22, + "grad_norm": 2.1149528255278542, + "learning_rate": 1.8177344443373806e-05, + "loss": 1.1051, + "step": 9312 + }, + { + "epoch": 0.22, + "grad_norm": 1.1014756341132228, + "learning_rate": 1.8176905216436192e-05, + "loss": 0.9698, + "step": 9313 + }, + { + "epoch": 0.22, + "grad_norm": 1.1419040603031305, + "learning_rate": 1.817646594189005e-05, + "loss": 1.0099, + "step": 9314 + }, + { + "epoch": 0.22, + "grad_norm": 2.3808345206135324, + "learning_rate": 1.8176026619737935e-05, + "loss": 1.0877, + "step": 9315 + }, + { + "epoch": 0.22, + "grad_norm": 2.3262356194675124, + "learning_rate": 1.8175587249982405e-05, + "loss": 1.2003, + "step": 9316 + }, + { + "epoch": 0.22, + "grad_norm": 1.9247619421183533, + "learning_rate": 1.8175147832626018e-05, + "loss": 1.0875, + "step": 9317 + }, + { + "epoch": 0.22, + "grad_norm": 2.0187278420915966, + "learning_rate": 1.8174708367671326e-05, + "loss": 1.0223, + "step": 9318 + }, + { + "epoch": 0.22, + "grad_norm": 1.9934889009099963, + "learning_rate": 1.81742688551209e-05, + "loss": 1.016, + "step": 9319 + }, + { + "epoch": 0.22, + "grad_norm": 2.5331071501450215, + "learning_rate": 1.8173829294977294e-05, + "loss": 0.926, + "step": 9320 + }, + { + "epoch": 0.22, + "grad_norm": 2.0088290122245733, + "learning_rate": 1.8173389687243063e-05, + "loss": 1.2088, + "step": 9321 + }, + { + "epoch": 0.22, + "grad_norm": 2.3696112383996, + "learning_rate": 1.8172950031920772e-05, + "loss": 1.0764, + "step": 9322 + }, + { + "epoch": 0.22, + "grad_norm": 2.3075869095543693, + "learning_rate": 1.8172510329012974e-05, + "loss": 0.9696, + "step": 9323 + }, + { + "epoch": 0.22, + "grad_norm": 2.6125141860166656, + "learning_rate": 1.8172070578522235e-05, + "loss": 0.9639, + "step": 9324 + }, + { + "epoch": 0.22, + "grad_norm": 1.9905217531089499, + "learning_rate": 1.8171630780451115e-05, + "loss": 1.1015, + "step": 9325 + }, + { + "epoch": 0.22, + "grad_norm": 2.2908660536183683, + "learning_rate": 1.8171190934802175e-05, + "loss": 1.0945, + "step": 9326 + }, + { + "epoch": 0.22, + "grad_norm": 2.1394984453017876, + "learning_rate": 1.8170751041577972e-05, + "loss": 1.0347, + "step": 9327 + }, + { + "epoch": 0.22, + "grad_norm": 1.2430420568084626, + "learning_rate": 1.817031110078107e-05, + "loss": 0.9682, + "step": 9328 + }, + { + "epoch": 0.22, + "grad_norm": 2.526969080227581, + "learning_rate": 1.816987111241403e-05, + "loss": 1.0806, + "step": 9329 + }, + { + "epoch": 0.22, + "grad_norm": 2.1594425575280805, + "learning_rate": 1.8169431076479415e-05, + "loss": 1.0811, + "step": 9330 + }, + { + "epoch": 0.22, + "grad_norm": 2.034519215408427, + "learning_rate": 1.8168990992979785e-05, + "loss": 1.0201, + "step": 9331 + }, + { + "epoch": 0.22, + "grad_norm": 2.1823938284265485, + "learning_rate": 1.8168550861917702e-05, + "loss": 1.0336, + "step": 9332 + }, + { + "epoch": 0.22, + "grad_norm": 2.667513346220684, + "learning_rate": 1.8168110683295733e-05, + "loss": 1.1252, + "step": 9333 + }, + { + "epoch": 0.22, + "grad_norm": 2.881731049756331, + "learning_rate": 1.8167670457116434e-05, + "loss": 0.9476, + "step": 9334 + }, + { + "epoch": 0.22, + "grad_norm": 1.0903990134955956, + "learning_rate": 1.8167230183382375e-05, + "loss": 0.9454, + "step": 9335 + }, + { + "epoch": 0.22, + "grad_norm": 1.9771996816769641, + "learning_rate": 1.8166789862096114e-05, + "loss": 1.0937, + "step": 9336 + }, + { + "epoch": 0.22, + "grad_norm": 1.8685112833408934, + "learning_rate": 1.8166349493260218e-05, + "loss": 1.0537, + "step": 9337 + }, + { + "epoch": 0.22, + "grad_norm": 2.159573271967504, + "learning_rate": 1.816590907687725e-05, + "loss": 1.0668, + "step": 9338 + }, + { + "epoch": 0.22, + "grad_norm": 2.010566883668991, + "learning_rate": 1.816546861294977e-05, + "loss": 1.0132, + "step": 9339 + }, + { + "epoch": 0.22, + "grad_norm": 2.0158142895724533, + "learning_rate": 1.8165028101480354e-05, + "loss": 1.176, + "step": 9340 + }, + { + "epoch": 0.22, + "grad_norm": 2.5480981607289133, + "learning_rate": 1.8164587542471555e-05, + "loss": 1.0797, + "step": 9341 + }, + { + "epoch": 0.22, + "grad_norm": 2.1284022421451656, + "learning_rate": 1.816414693592594e-05, + "loss": 1.1022, + "step": 9342 + }, + { + "epoch": 0.22, + "grad_norm": 1.9845699221373463, + "learning_rate": 1.816370628184608e-05, + "loss": 0.9635, + "step": 9343 + }, + { + "epoch": 0.22, + "grad_norm": 2.2031780908237772, + "learning_rate": 1.816326558023454e-05, + "loss": 0.9641, + "step": 9344 + }, + { + "epoch": 0.22, + "grad_norm": 2.1092473616218292, + "learning_rate": 1.816282483109388e-05, + "loss": 1.1499, + "step": 9345 + }, + { + "epoch": 0.22, + "grad_norm": 2.22283346399049, + "learning_rate": 1.816238403442667e-05, + "loss": 1.0792, + "step": 9346 + }, + { + "epoch": 0.22, + "grad_norm": 2.157931065780694, + "learning_rate": 1.8161943190235476e-05, + "loss": 1.0606, + "step": 9347 + }, + { + "epoch": 0.22, + "grad_norm": 1.1581830196661922, + "learning_rate": 1.8161502298522865e-05, + "loss": 0.9223, + "step": 9348 + }, + { + "epoch": 0.22, + "grad_norm": 2.0310455466959003, + "learning_rate": 1.81610613592914e-05, + "loss": 1.1659, + "step": 9349 + }, + { + "epoch": 0.22, + "grad_norm": 2.1251519642778898, + "learning_rate": 1.816062037254366e-05, + "loss": 1.0146, + "step": 9350 + }, + { + "epoch": 0.22, + "grad_norm": 1.0706987986899965, + "learning_rate": 1.8160179338282196e-05, + "loss": 0.9671, + "step": 9351 + }, + { + "epoch": 0.22, + "grad_norm": 1.9586593017347265, + "learning_rate": 1.815973825650959e-05, + "loss": 0.9908, + "step": 9352 + }, + { + "epoch": 0.22, + "grad_norm": 2.0453281076511867, + "learning_rate": 1.8159297127228403e-05, + "loss": 1.1384, + "step": 9353 + }, + { + "epoch": 0.22, + "grad_norm": 1.8812586671488578, + "learning_rate": 1.8158855950441206e-05, + "loss": 1.2548, + "step": 9354 + }, + { + "epoch": 0.22, + "grad_norm": 2.2417056152761003, + "learning_rate": 1.8158414726150565e-05, + "loss": 1.008, + "step": 9355 + }, + { + "epoch": 0.22, + "grad_norm": 2.230664790560623, + "learning_rate": 1.815797345435905e-05, + "loss": 1.0644, + "step": 9356 + }, + { + "epoch": 0.22, + "grad_norm": 2.0414226219730254, + "learning_rate": 1.815753213506923e-05, + "loss": 1.095, + "step": 9357 + }, + { + "epoch": 0.22, + "grad_norm": 1.1307951830934524, + "learning_rate": 1.815709076828368e-05, + "loss": 0.9543, + "step": 9358 + }, + { + "epoch": 0.22, + "grad_norm": 2.079205949981699, + "learning_rate": 1.815664935400496e-05, + "loss": 1.0635, + "step": 9359 + }, + { + "epoch": 0.22, + "grad_norm": 2.444075540346216, + "learning_rate": 1.815620789223565e-05, + "loss": 1.068, + "step": 9360 + }, + { + "epoch": 0.22, + "grad_norm": 1.878494202767731, + "learning_rate": 1.8155766382978314e-05, + "loss": 0.9307, + "step": 9361 + }, + { + "epoch": 0.22, + "grad_norm": 3.9419705320285203, + "learning_rate": 1.815532482623552e-05, + "loss": 1.2639, + "step": 9362 + }, + { + "epoch": 0.22, + "grad_norm": 2.21552564652686, + "learning_rate": 1.8154883222009844e-05, + "loss": 1.0944, + "step": 9363 + }, + { + "epoch": 0.22, + "grad_norm": 2.4310975776337354, + "learning_rate": 1.815444157030386e-05, + "loss": 1.051, + "step": 9364 + }, + { + "epoch": 0.22, + "grad_norm": 2.032937299191962, + "learning_rate": 1.8153999871120135e-05, + "loss": 0.9075, + "step": 9365 + }, + { + "epoch": 0.22, + "grad_norm": 2.4051657591416613, + "learning_rate": 1.8153558124461236e-05, + "loss": 1.0737, + "step": 9366 + }, + { + "epoch": 0.22, + "grad_norm": 1.1382567350697095, + "learning_rate": 1.8153116330329745e-05, + "loss": 1.0106, + "step": 9367 + }, + { + "epoch": 0.22, + "grad_norm": 2.0492271453208795, + "learning_rate": 1.815267448872823e-05, + "loss": 1.0142, + "step": 9368 + }, + { + "epoch": 0.22, + "grad_norm": 2.159979697060727, + "learning_rate": 1.815223259965926e-05, + "loss": 1.1949, + "step": 9369 + }, + { + "epoch": 0.22, + "grad_norm": 2.094951279130828, + "learning_rate": 1.8151790663125412e-05, + "loss": 1.2342, + "step": 9370 + }, + { + "epoch": 0.22, + "grad_norm": 2.4379871593024287, + "learning_rate": 1.8151348679129257e-05, + "loss": 1.0248, + "step": 9371 + }, + { + "epoch": 0.22, + "grad_norm": 2.134831517708591, + "learning_rate": 1.815090664767337e-05, + "loss": 1.1014, + "step": 9372 + }, + { + "epoch": 0.22, + "grad_norm": 2.3856651492412504, + "learning_rate": 1.8150464568760327e-05, + "loss": 1.0438, + "step": 9373 + }, + { + "epoch": 0.22, + "grad_norm": 2.1012999331695883, + "learning_rate": 1.8150022442392695e-05, + "loss": 1.1631, + "step": 9374 + }, + { + "epoch": 0.22, + "grad_norm": 2.2791585571778903, + "learning_rate": 1.814958026857305e-05, + "loss": 1.0177, + "step": 9375 + }, + { + "epoch": 0.22, + "grad_norm": 2.653248839316033, + "learning_rate": 1.8149138047303977e-05, + "loss": 1.0819, + "step": 9376 + }, + { + "epoch": 0.22, + "grad_norm": 1.8621772539729853, + "learning_rate": 1.8148695778588034e-05, + "loss": 0.9615, + "step": 9377 + }, + { + "epoch": 0.22, + "grad_norm": 2.2097879380847303, + "learning_rate": 1.814825346242781e-05, + "loss": 1.0216, + "step": 9378 + }, + { + "epoch": 0.22, + "grad_norm": 1.9931143329739534, + "learning_rate": 1.814781109882587e-05, + "loss": 1.1199, + "step": 9379 + }, + { + "epoch": 0.22, + "grad_norm": 2.431016225963388, + "learning_rate": 1.8147368687784798e-05, + "loss": 1.2527, + "step": 9380 + }, + { + "epoch": 0.22, + "grad_norm": 2.7314470857642443, + "learning_rate": 1.8146926229307166e-05, + "loss": 1.068, + "step": 9381 + }, + { + "epoch": 0.22, + "grad_norm": 2.017680194939548, + "learning_rate": 1.814648372339555e-05, + "loss": 1.0343, + "step": 9382 + }, + { + "epoch": 0.22, + "grad_norm": 2.367444464003743, + "learning_rate": 1.8146041170052524e-05, + "loss": 1.0787, + "step": 9383 + }, + { + "epoch": 0.22, + "grad_norm": 2.1211403105957616, + "learning_rate": 1.814559856928067e-05, + "loss": 0.9617, + "step": 9384 + }, + { + "epoch": 0.22, + "grad_norm": 2.0546419158790634, + "learning_rate": 1.814515592108256e-05, + "loss": 1.0009, + "step": 9385 + }, + { + "epoch": 0.22, + "grad_norm": 2.090262925476646, + "learning_rate": 1.814471322546077e-05, + "loss": 1.0483, + "step": 9386 + }, + { + "epoch": 0.22, + "grad_norm": 2.005165643271133, + "learning_rate": 1.8144270482417887e-05, + "loss": 1.0187, + "step": 9387 + }, + { + "epoch": 0.22, + "grad_norm": 2.2243505344903722, + "learning_rate": 1.8143827691956483e-05, + "loss": 1.1521, + "step": 9388 + }, + { + "epoch": 0.22, + "grad_norm": 2.1261512805460643, + "learning_rate": 1.8143384854079132e-05, + "loss": 1.0354, + "step": 9389 + }, + { + "epoch": 0.22, + "grad_norm": 2.0309600220762904, + "learning_rate": 1.8142941968788422e-05, + "loss": 0.9334, + "step": 9390 + }, + { + "epoch": 0.22, + "grad_norm": 2.29577272824293, + "learning_rate": 1.8142499036086922e-05, + "loss": 1.0938, + "step": 9391 + }, + { + "epoch": 0.22, + "grad_norm": 2.0496309185386252, + "learning_rate": 1.8142056055977212e-05, + "loss": 1.1871, + "step": 9392 + }, + { + "epoch": 0.22, + "grad_norm": 1.8786158795211623, + "learning_rate": 1.8141613028461877e-05, + "loss": 1.1137, + "step": 9393 + }, + { + "epoch": 0.22, + "grad_norm": 2.128313101919674, + "learning_rate": 1.8141169953543495e-05, + "loss": 1.0232, + "step": 9394 + }, + { + "epoch": 0.22, + "grad_norm": 1.745175725217137, + "learning_rate": 1.8140726831224644e-05, + "loss": 0.9953, + "step": 9395 + }, + { + "epoch": 0.22, + "grad_norm": 2.2124411981126353, + "learning_rate": 1.8140283661507904e-05, + "loss": 1.0492, + "step": 9396 + }, + { + "epoch": 0.22, + "grad_norm": 2.5432777100818944, + "learning_rate": 1.813984044439585e-05, + "loss": 1.0346, + "step": 9397 + }, + { + "epoch": 0.22, + "grad_norm": 2.4086651719437655, + "learning_rate": 1.8139397179891075e-05, + "loss": 1.1081, + "step": 9398 + }, + { + "epoch": 0.22, + "grad_norm": 2.194390880994411, + "learning_rate": 1.8138953867996147e-05, + "loss": 1.1123, + "step": 9399 + }, + { + "epoch": 0.22, + "grad_norm": 2.3278741202334396, + "learning_rate": 1.8138510508713656e-05, + "loss": 1.0312, + "step": 9400 + }, + { + "epoch": 0.22, + "grad_norm": 1.8795725070309213, + "learning_rate": 1.813806710204618e-05, + "loss": 1.074, + "step": 9401 + }, + { + "epoch": 0.22, + "grad_norm": 2.2886992147334304, + "learning_rate": 1.81376236479963e-05, + "loss": 1.1114, + "step": 9402 + }, + { + "epoch": 0.22, + "grad_norm": 2.371404635007417, + "learning_rate": 1.81371801465666e-05, + "loss": 1.0321, + "step": 9403 + }, + { + "epoch": 0.22, + "grad_norm": 2.1228917965843452, + "learning_rate": 1.813673659775966e-05, + "loss": 1.0091, + "step": 9404 + }, + { + "epoch": 0.22, + "grad_norm": 1.9107011734979442, + "learning_rate": 1.813629300157806e-05, + "loss": 1.0249, + "step": 9405 + }, + { + "epoch": 0.22, + "grad_norm": 1.8444705856721053, + "learning_rate": 1.8135849358024392e-05, + "loss": 1.073, + "step": 9406 + }, + { + "epoch": 0.22, + "grad_norm": 2.506369544122614, + "learning_rate": 1.813540566710123e-05, + "loss": 1.1038, + "step": 9407 + }, + { + "epoch": 0.22, + "grad_norm": 1.20579729934107, + "learning_rate": 1.8134961928811164e-05, + "loss": 1.0103, + "step": 9408 + }, + { + "epoch": 0.22, + "grad_norm": 2.1743982814392444, + "learning_rate": 1.813451814315677e-05, + "loss": 1.1507, + "step": 9409 + }, + { + "epoch": 0.22, + "grad_norm": 2.7011757183978524, + "learning_rate": 1.8134074310140638e-05, + "loss": 1.1219, + "step": 9410 + }, + { + "epoch": 0.22, + "grad_norm": 2.130533096594077, + "learning_rate": 1.813363042976535e-05, + "loss": 0.9615, + "step": 9411 + }, + { + "epoch": 0.22, + "grad_norm": 2.4535419808651313, + "learning_rate": 1.8133186502033493e-05, + "loss": 1.2296, + "step": 9412 + }, + { + "epoch": 0.22, + "grad_norm": 2.3097972630840014, + "learning_rate": 1.8132742526947648e-05, + "loss": 1.1946, + "step": 9413 + }, + { + "epoch": 0.22, + "grad_norm": 2.2657422633490287, + "learning_rate": 1.8132298504510398e-05, + "loss": 1.2019, + "step": 9414 + }, + { + "epoch": 0.22, + "grad_norm": 2.2781879858348093, + "learning_rate": 1.8131854434724335e-05, + "loss": 1.084, + "step": 9415 + }, + { + "epoch": 0.22, + "grad_norm": 1.9602835085181614, + "learning_rate": 1.8131410317592037e-05, + "loss": 0.9921, + "step": 9416 + }, + { + "epoch": 0.22, + "grad_norm": 2.116358628530211, + "learning_rate": 1.8130966153116096e-05, + "loss": 0.9695, + "step": 9417 + }, + { + "epoch": 0.22, + "grad_norm": 2.084775154470994, + "learning_rate": 1.81305219412991e-05, + "loss": 0.9398, + "step": 9418 + }, + { + "epoch": 0.22, + "grad_norm": 1.9250805889066422, + "learning_rate": 1.8130077682143627e-05, + "loss": 1.0923, + "step": 9419 + }, + { + "epoch": 0.22, + "grad_norm": 1.9981784384538852, + "learning_rate": 1.812963337565227e-05, + "loss": 1.0634, + "step": 9420 + }, + { + "epoch": 0.22, + "grad_norm": 2.044291145558896, + "learning_rate": 1.8129189021827608e-05, + "loss": 1.034, + "step": 9421 + }, + { + "epoch": 0.22, + "grad_norm": 1.9839646023450188, + "learning_rate": 1.8128744620672237e-05, + "loss": 1.0116, + "step": 9422 + }, + { + "epoch": 0.22, + "grad_norm": 1.9695010025360395, + "learning_rate": 1.8128300172188744e-05, + "loss": 1.0505, + "step": 9423 + }, + { + "epoch": 0.22, + "grad_norm": 1.1567826005932254, + "learning_rate": 1.812785567637971e-05, + "loss": 1.0131, + "step": 9424 + }, + { + "epoch": 0.22, + "grad_norm": 2.146230728202998, + "learning_rate": 1.812741113324773e-05, + "loss": 1.0214, + "step": 9425 + }, + { + "epoch": 0.22, + "grad_norm": 1.8851529660998616, + "learning_rate": 1.8126966542795387e-05, + "loss": 0.9826, + "step": 9426 + }, + { + "epoch": 0.22, + "grad_norm": 2.1281983598522936, + "learning_rate": 1.8126521905025272e-05, + "loss": 1.025, + "step": 9427 + }, + { + "epoch": 0.22, + "grad_norm": 2.343720176626241, + "learning_rate": 1.812607721993998e-05, + "loss": 1.0128, + "step": 9428 + }, + { + "epoch": 0.22, + "grad_norm": 2.086637554718816, + "learning_rate": 1.8125632487542084e-05, + "loss": 0.9653, + "step": 9429 + }, + { + "epoch": 0.22, + "grad_norm": 1.1428996190127196, + "learning_rate": 1.812518770783419e-05, + "loss": 0.9977, + "step": 9430 + }, + { + "epoch": 0.22, + "grad_norm": 2.149912211105052, + "learning_rate": 1.8124742880818877e-05, + "loss": 1.003, + "step": 9431 + }, + { + "epoch": 0.22, + "grad_norm": 1.9273277412324916, + "learning_rate": 1.812429800649874e-05, + "loss": 1.1702, + "step": 9432 + }, + { + "epoch": 0.22, + "grad_norm": 2.3152113183644865, + "learning_rate": 1.812385308487637e-05, + "loss": 1.1497, + "step": 9433 + }, + { + "epoch": 0.22, + "grad_norm": 2.52759255877034, + "learning_rate": 1.8123408115954354e-05, + "loss": 1.1412, + "step": 9434 + }, + { + "epoch": 0.22, + "grad_norm": 2.1338349948997926, + "learning_rate": 1.812296309973528e-05, + "loss": 0.9887, + "step": 9435 + }, + { + "epoch": 0.22, + "grad_norm": 2.119485881253422, + "learning_rate": 1.812251803622175e-05, + "loss": 1.0188, + "step": 9436 + }, + { + "epoch": 0.22, + "grad_norm": 1.9459442724139269, + "learning_rate": 1.8122072925416347e-05, + "loss": 1.0248, + "step": 9437 + }, + { + "epoch": 0.22, + "grad_norm": 2.017714644906583, + "learning_rate": 1.8121627767321658e-05, + "loss": 0.9391, + "step": 9438 + }, + { + "epoch": 0.22, + "grad_norm": 2.1378946855590297, + "learning_rate": 1.8121182561940288e-05, + "loss": 0.9308, + "step": 9439 + }, + { + "epoch": 0.22, + "grad_norm": 1.7023260789337444, + "learning_rate": 1.8120737309274818e-05, + "loss": 1.0356, + "step": 9440 + }, + { + "epoch": 0.22, + "grad_norm": 2.6543628811260658, + "learning_rate": 1.8120292009327843e-05, + "loss": 1.0788, + "step": 9441 + }, + { + "epoch": 0.22, + "grad_norm": 1.8191266210773482, + "learning_rate": 1.811984666210196e-05, + "loss": 1.0342, + "step": 9442 + }, + { + "epoch": 0.22, + "grad_norm": 2.118418796609978, + "learning_rate": 1.811940126759976e-05, + "loss": 1.1139, + "step": 9443 + }, + { + "epoch": 0.22, + "grad_norm": 2.040453996056635, + "learning_rate": 1.8118955825823833e-05, + "loss": 1.0935, + "step": 9444 + }, + { + "epoch": 0.22, + "grad_norm": 1.1397829109245083, + "learning_rate": 1.8118510336776776e-05, + "loss": 0.9808, + "step": 9445 + }, + { + "epoch": 0.22, + "grad_norm": 2.3678114105432675, + "learning_rate": 1.8118064800461182e-05, + "loss": 1.0622, + "step": 9446 + }, + { + "epoch": 0.22, + "grad_norm": 1.88327884605665, + "learning_rate": 1.8117619216879645e-05, + "loss": 0.9911, + "step": 9447 + }, + { + "epoch": 0.22, + "grad_norm": 1.837719395112089, + "learning_rate": 1.8117173586034757e-05, + "loss": 1.0184, + "step": 9448 + }, + { + "epoch": 0.22, + "grad_norm": 2.0900648859043782, + "learning_rate": 1.8116727907929113e-05, + "loss": 0.9945, + "step": 9449 + }, + { + "epoch": 0.22, + "grad_norm": 1.9688016530150507, + "learning_rate": 1.8116282182565313e-05, + "loss": 0.9649, + "step": 9450 + }, + { + "epoch": 0.22, + "grad_norm": 1.9924293612006823, + "learning_rate": 1.8115836409945947e-05, + "loss": 1.0493, + "step": 9451 + }, + { + "epoch": 0.22, + "grad_norm": 2.025650113550961, + "learning_rate": 1.8115390590073612e-05, + "loss": 1.0151, + "step": 9452 + }, + { + "epoch": 0.22, + "grad_norm": 2.0737292926561337, + "learning_rate": 1.8114944722950903e-05, + "loss": 1.0137, + "step": 9453 + }, + { + "epoch": 0.22, + "grad_norm": 1.9356138811292105, + "learning_rate": 1.8114498808580418e-05, + "loss": 1.0012, + "step": 9454 + }, + { + "epoch": 0.22, + "grad_norm": 2.422048074936893, + "learning_rate": 1.8114052846964753e-05, + "loss": 1.0118, + "step": 9455 + }, + { + "epoch": 0.22, + "grad_norm": 2.1059792401579767, + "learning_rate": 1.8113606838106503e-05, + "loss": 0.9817, + "step": 9456 + }, + { + "epoch": 0.22, + "grad_norm": 1.9667377548272378, + "learning_rate": 1.8113160782008263e-05, + "loss": 0.9232, + "step": 9457 + }, + { + "epoch": 0.22, + "grad_norm": 1.8410767672564126, + "learning_rate": 1.8112714678672632e-05, + "loss": 1.0989, + "step": 9458 + }, + { + "epoch": 0.22, + "grad_norm": 2.097243022963436, + "learning_rate": 1.811226852810221e-05, + "loss": 1.1153, + "step": 9459 + }, + { + "epoch": 0.22, + "grad_norm": 2.4784997694092272, + "learning_rate": 1.811182233029959e-05, + "loss": 1.1598, + "step": 9460 + }, + { + "epoch": 0.22, + "grad_norm": 2.0192984065963313, + "learning_rate": 1.8111376085267376e-05, + "loss": 1.0291, + "step": 9461 + }, + { + "epoch": 0.22, + "grad_norm": 2.017763807369592, + "learning_rate": 1.811092979300816e-05, + "loss": 1.0983, + "step": 9462 + }, + { + "epoch": 0.22, + "grad_norm": 2.12305090113366, + "learning_rate": 1.8110483453524543e-05, + "loss": 0.9604, + "step": 9463 + }, + { + "epoch": 0.22, + "grad_norm": 2.326455460717508, + "learning_rate": 1.811003706681912e-05, + "loss": 0.9837, + "step": 9464 + }, + { + "epoch": 0.22, + "grad_norm": 2.0502404020883556, + "learning_rate": 1.81095906328945e-05, + "loss": 1.0113, + "step": 9465 + }, + { + "epoch": 0.22, + "grad_norm": 1.102441427986239, + "learning_rate": 1.8109144151753274e-05, + "loss": 1.0085, + "step": 9466 + }, + { + "epoch": 0.22, + "grad_norm": 2.3766683597550866, + "learning_rate": 1.810869762339804e-05, + "loss": 0.9867, + "step": 9467 + }, + { + "epoch": 0.22, + "grad_norm": 2.018986040163824, + "learning_rate": 1.8108251047831405e-05, + "loss": 1.0441, + "step": 9468 + }, + { + "epoch": 0.22, + "grad_norm": 2.2302257212117746, + "learning_rate": 1.8107804425055967e-05, + "loss": 1.0015, + "step": 9469 + }, + { + "epoch": 0.22, + "grad_norm": 1.8865052327232357, + "learning_rate": 1.810735775507432e-05, + "loss": 1.1063, + "step": 9470 + }, + { + "epoch": 0.22, + "grad_norm": 3.3996348239498873, + "learning_rate": 1.8106911037889072e-05, + "loss": 1.079, + "step": 9471 + }, + { + "epoch": 0.22, + "grad_norm": 1.9335197916239928, + "learning_rate": 1.810646427350282e-05, + "loss": 1.1211, + "step": 9472 + }, + { + "epoch": 0.22, + "grad_norm": 3.0235531024231466, + "learning_rate": 1.8106017461918165e-05, + "loss": 1.1108, + "step": 9473 + }, + { + "epoch": 0.22, + "grad_norm": 1.911919667627577, + "learning_rate": 1.8105570603137714e-05, + "loss": 1.1086, + "step": 9474 + }, + { + "epoch": 0.22, + "grad_norm": 2.0125250259605245, + "learning_rate": 1.810512369716406e-05, + "loss": 1.0537, + "step": 9475 + }, + { + "epoch": 0.22, + "grad_norm": 1.9467587248263687, + "learning_rate": 1.8104676743999816e-05, + "loss": 0.9495, + "step": 9476 + }, + { + "epoch": 0.22, + "grad_norm": 2.3621190178704135, + "learning_rate": 1.8104229743647573e-05, + "loss": 0.9718, + "step": 9477 + }, + { + "epoch": 0.22, + "grad_norm": 1.993259474373815, + "learning_rate": 1.8103782696109943e-05, + "loss": 1.0621, + "step": 9478 + }, + { + "epoch": 0.22, + "grad_norm": 1.9645170221928152, + "learning_rate": 1.8103335601389522e-05, + "loss": 1.174, + "step": 9479 + }, + { + "epoch": 0.22, + "grad_norm": 1.1631829984380118, + "learning_rate": 1.8102888459488916e-05, + "loss": 0.8889, + "step": 9480 + }, + { + "epoch": 0.22, + "grad_norm": 2.1553487135839298, + "learning_rate": 1.810244127041073e-05, + "loss": 1.1589, + "step": 9481 + }, + { + "epoch": 0.22, + "grad_norm": 1.8628177837403843, + "learning_rate": 1.810199403415756e-05, + "loss": 1.015, + "step": 9482 + }, + { + "epoch": 0.22, + "grad_norm": 2.21421414526818, + "learning_rate": 1.8101546750732022e-05, + "loss": 1.1553, + "step": 9483 + }, + { + "epoch": 0.22, + "grad_norm": 2.2267227627565678, + "learning_rate": 1.810109942013671e-05, + "loss": 0.9917, + "step": 9484 + }, + { + "epoch": 0.22, + "grad_norm": 2.532007480920698, + "learning_rate": 1.8100652042374232e-05, + "loss": 1.0185, + "step": 9485 + }, + { + "epoch": 0.22, + "grad_norm": 1.927896065414452, + "learning_rate": 1.8100204617447195e-05, + "loss": 0.9797, + "step": 9486 + }, + { + "epoch": 0.22, + "grad_norm": 2.4197411756422955, + "learning_rate": 1.80997571453582e-05, + "loss": 1.103, + "step": 9487 + }, + { + "epoch": 0.22, + "grad_norm": 2.0968044898597253, + "learning_rate": 1.8099309626109857e-05, + "loss": 1.0312, + "step": 9488 + }, + { + "epoch": 0.22, + "grad_norm": 2.0066706132778305, + "learning_rate": 1.809886205970477e-05, + "loss": 1.0692, + "step": 9489 + }, + { + "epoch": 0.22, + "grad_norm": 1.892359931680797, + "learning_rate": 1.8098414446145543e-05, + "loss": 1.0944, + "step": 9490 + }, + { + "epoch": 0.22, + "grad_norm": 2.927266182643626, + "learning_rate": 1.8097966785434786e-05, + "loss": 1.1263, + "step": 9491 + }, + { + "epoch": 0.22, + "grad_norm": 2.0470646543309057, + "learning_rate": 1.8097519077575095e-05, + "loss": 1.0432, + "step": 9492 + }, + { + "epoch": 0.22, + "grad_norm": 2.386317877635069, + "learning_rate": 1.809707132256909e-05, + "loss": 1.0499, + "step": 9493 + }, + { + "epoch": 0.22, + "grad_norm": 2.762320896301531, + "learning_rate": 1.8096623520419372e-05, + "loss": 1.0612, + "step": 9494 + }, + { + "epoch": 0.22, + "grad_norm": 1.9581283030453933, + "learning_rate": 1.8096175671128547e-05, + "loss": 1.1932, + "step": 9495 + }, + { + "epoch": 0.22, + "grad_norm": 2.4386224216122545, + "learning_rate": 1.8095727774699224e-05, + "loss": 1.0714, + "step": 9496 + }, + { + "epoch": 0.22, + "grad_norm": 1.910111761684271, + "learning_rate": 1.8095279831134012e-05, + "loss": 1.1459, + "step": 9497 + }, + { + "epoch": 0.22, + "grad_norm": 2.314871431828754, + "learning_rate": 1.809483184043552e-05, + "loss": 0.9576, + "step": 9498 + }, + { + "epoch": 0.22, + "grad_norm": 2.2347549871573418, + "learning_rate": 1.809438380260635e-05, + "loss": 0.9875, + "step": 9499 + }, + { + "epoch": 0.22, + "grad_norm": 1.2317138125994689, + "learning_rate": 1.8093935717649117e-05, + "loss": 1.002, + "step": 9500 + }, + { + "epoch": 0.22, + "grad_norm": 2.3168387541975948, + "learning_rate": 1.8093487585566426e-05, + "loss": 1.0365, + "step": 9501 + }, + { + "epoch": 0.22, + "grad_norm": 1.9658693145466113, + "learning_rate": 1.809303940636089e-05, + "loss": 1.1254, + "step": 9502 + }, + { + "epoch": 0.22, + "grad_norm": 1.980938656303067, + "learning_rate": 1.8092591180035114e-05, + "loss": 1.1073, + "step": 9503 + }, + { + "epoch": 0.22, + "grad_norm": 1.1013952525226764, + "learning_rate": 1.809214290659171e-05, + "loss": 0.9871, + "step": 9504 + }, + { + "epoch": 0.22, + "grad_norm": 2.3013143988318236, + "learning_rate": 1.809169458603329e-05, + "loss": 1.0144, + "step": 9505 + }, + { + "epoch": 0.22, + "grad_norm": 2.099466653785835, + "learning_rate": 1.8091246218362464e-05, + "loss": 1.0691, + "step": 9506 + }, + { + "epoch": 0.22, + "grad_norm": 1.8737881403661536, + "learning_rate": 1.809079780358184e-05, + "loss": 1.032, + "step": 9507 + }, + { + "epoch": 0.22, + "grad_norm": 2.1540981607015106, + "learning_rate": 1.809034934169403e-05, + "loss": 0.9093, + "step": 9508 + }, + { + "epoch": 0.22, + "grad_norm": 2.081104998582328, + "learning_rate": 1.808990083270164e-05, + "loss": 1.0726, + "step": 9509 + }, + { + "epoch": 0.22, + "grad_norm": 1.9353078340655936, + "learning_rate": 1.808945227660729e-05, + "loss": 0.8443, + "step": 9510 + }, + { + "epoch": 0.22, + "grad_norm": 1.885948272740878, + "learning_rate": 1.808900367341359e-05, + "loss": 1.1226, + "step": 9511 + }, + { + "epoch": 0.22, + "grad_norm": 2.10561034652252, + "learning_rate": 1.8088555023123144e-05, + "loss": 1.0168, + "step": 9512 + }, + { + "epoch": 0.22, + "grad_norm": 1.793221729391815, + "learning_rate": 1.8088106325738572e-05, + "loss": 0.9501, + "step": 9513 + }, + { + "epoch": 0.22, + "grad_norm": 2.3239533514274244, + "learning_rate": 1.8087657581262488e-05, + "loss": 0.8187, + "step": 9514 + }, + { + "epoch": 0.22, + "grad_norm": 2.0921137698385306, + "learning_rate": 1.8087208789697498e-05, + "loss": 1.147, + "step": 9515 + }, + { + "epoch": 0.22, + "grad_norm": 2.515441345739787, + "learning_rate": 1.8086759951046217e-05, + "loss": 1.0493, + "step": 9516 + }, + { + "epoch": 0.22, + "grad_norm": 3.3753520174360263, + "learning_rate": 1.8086311065311258e-05, + "loss": 0.9062, + "step": 9517 + }, + { + "epoch": 0.22, + "grad_norm": 2.3122910708450135, + "learning_rate": 1.808586213249524e-05, + "loss": 1.1191, + "step": 9518 + }, + { + "epoch": 0.22, + "grad_norm": 2.137741908392925, + "learning_rate": 1.808541315260077e-05, + "loss": 1.0837, + "step": 9519 + }, + { + "epoch": 0.22, + "grad_norm": 2.0513544916339206, + "learning_rate": 1.808496412563047e-05, + "loss": 0.8646, + "step": 9520 + }, + { + "epoch": 0.22, + "grad_norm": 2.044821843321202, + "learning_rate": 1.8084515051586943e-05, + "loss": 1.0368, + "step": 9521 + }, + { + "epoch": 0.22, + "grad_norm": 2.5215649133011615, + "learning_rate": 1.808406593047281e-05, + "loss": 1.1006, + "step": 9522 + }, + { + "epoch": 0.22, + "grad_norm": 2.045207827788303, + "learning_rate": 1.808361676229069e-05, + "loss": 1.0632, + "step": 9523 + }, + { + "epoch": 0.22, + "grad_norm": 1.9232130402289147, + "learning_rate": 1.808316754704319e-05, + "loss": 1.0013, + "step": 9524 + }, + { + "epoch": 0.22, + "grad_norm": 2.5089282877553694, + "learning_rate": 1.8082718284732932e-05, + "loss": 1.1439, + "step": 9525 + }, + { + "epoch": 0.22, + "grad_norm": 1.2978460738379594, + "learning_rate": 1.808226897536253e-05, + "loss": 1.0291, + "step": 9526 + }, + { + "epoch": 0.22, + "grad_norm": 2.081463507908065, + "learning_rate": 1.8081819618934597e-05, + "loss": 1.0898, + "step": 9527 + }, + { + "epoch": 0.22, + "grad_norm": 1.8689041213534223, + "learning_rate": 1.8081370215451752e-05, + "loss": 1.1176, + "step": 9528 + }, + { + "epoch": 0.22, + "grad_norm": 2.759156198294267, + "learning_rate": 1.8080920764916613e-05, + "loss": 1.0976, + "step": 9529 + }, + { + "epoch": 0.22, + "grad_norm": 2.1063248109549133, + "learning_rate": 1.8080471267331792e-05, + "loss": 1.0768, + "step": 9530 + }, + { + "epoch": 0.22, + "grad_norm": 1.996348184729542, + "learning_rate": 1.808002172269991e-05, + "loss": 1.0919, + "step": 9531 + }, + { + "epoch": 0.22, + "grad_norm": 2.069463523173798, + "learning_rate": 1.8079572131023585e-05, + "loss": 1.1232, + "step": 9532 + }, + { + "epoch": 0.22, + "grad_norm": 2.049829671888308, + "learning_rate": 1.807912249230543e-05, + "loss": 1.092, + "step": 9533 + }, + { + "epoch": 0.22, + "grad_norm": 1.8647446639871321, + "learning_rate": 1.807867280654807e-05, + "loss": 1.17, + "step": 9534 + }, + { + "epoch": 0.22, + "grad_norm": 1.9570569572787566, + "learning_rate": 1.807822307375412e-05, + "loss": 1.0929, + "step": 9535 + }, + { + "epoch": 0.22, + "grad_norm": 2.069040765048612, + "learning_rate": 1.807777329392619e-05, + "loss": 0.9891, + "step": 9536 + }, + { + "epoch": 0.22, + "grad_norm": 1.9120801957770892, + "learning_rate": 1.8077323467066913e-05, + "loss": 1.1485, + "step": 9537 + }, + { + "epoch": 0.22, + "grad_norm": 2.1616929041039805, + "learning_rate": 1.80768735931789e-05, + "loss": 0.9556, + "step": 9538 + }, + { + "epoch": 0.22, + "grad_norm": 2.3212936160195783, + "learning_rate": 1.807642367226477e-05, + "loss": 1.0017, + "step": 9539 + }, + { + "epoch": 0.22, + "grad_norm": 2.23597586285305, + "learning_rate": 1.8075973704327148e-05, + "loss": 1.1378, + "step": 9540 + }, + { + "epoch": 0.22, + "grad_norm": 2.2544736973249573, + "learning_rate": 1.8075523689368646e-05, + "loss": 1.083, + "step": 9541 + }, + { + "epoch": 0.22, + "grad_norm": 1.9063643270744322, + "learning_rate": 1.8075073627391892e-05, + "loss": 1.0097, + "step": 9542 + }, + { + "epoch": 0.22, + "grad_norm": 2.1361399482077013, + "learning_rate": 1.8074623518399503e-05, + "loss": 1.0595, + "step": 9543 + }, + { + "epoch": 0.22, + "grad_norm": 2.004474455463132, + "learning_rate": 1.8074173362394095e-05, + "loss": 1.0395, + "step": 9544 + }, + { + "epoch": 0.22, + "grad_norm": 2.1507278734670416, + "learning_rate": 1.8073723159378298e-05, + "loss": 1.1311, + "step": 9545 + }, + { + "epoch": 0.22, + "grad_norm": 2.7877013860718, + "learning_rate": 1.8073272909354727e-05, + "loss": 0.9671, + "step": 9546 + }, + { + "epoch": 0.22, + "grad_norm": 1.9693883627754014, + "learning_rate": 1.8072822612326007e-05, + "loss": 1.0461, + "step": 9547 + }, + { + "epoch": 0.22, + "grad_norm": 2.1248080361048234, + "learning_rate": 1.8072372268294755e-05, + "loss": 1.1133, + "step": 9548 + }, + { + "epoch": 0.22, + "grad_norm": 1.1302375547735857, + "learning_rate": 1.8071921877263597e-05, + "loss": 0.9515, + "step": 9549 + }, + { + "epoch": 0.22, + "grad_norm": 1.9760532517071363, + "learning_rate": 1.8071471439235154e-05, + "loss": 1.1458, + "step": 9550 + }, + { + "epoch": 0.23, + "grad_norm": 1.7456827023899883, + "learning_rate": 1.807102095421205e-05, + "loss": 1.0415, + "step": 9551 + }, + { + "epoch": 0.23, + "grad_norm": 1.216025592716843, + "learning_rate": 1.807057042219691e-05, + "loss": 0.924, + "step": 9552 + }, + { + "epoch": 0.23, + "grad_norm": 2.012466468322516, + "learning_rate": 1.8070119843192348e-05, + "loss": 1.0917, + "step": 9553 + }, + { + "epoch": 0.23, + "grad_norm": 2.238683660642918, + "learning_rate": 1.8069669217200995e-05, + "loss": 1.0508, + "step": 9554 + }, + { + "epoch": 0.23, + "grad_norm": 2.1338462837193664, + "learning_rate": 1.8069218544225474e-05, + "loss": 1.1837, + "step": 9555 + }, + { + "epoch": 0.23, + "grad_norm": 2.4875465479728094, + "learning_rate": 1.8068767824268407e-05, + "loss": 1.004, + "step": 9556 + }, + { + "epoch": 0.23, + "grad_norm": 2.337395744521605, + "learning_rate": 1.8068317057332422e-05, + "loss": 1.0684, + "step": 9557 + }, + { + "epoch": 0.23, + "grad_norm": 2.06806905963081, + "learning_rate": 1.8067866243420136e-05, + "loss": 1.023, + "step": 9558 + }, + { + "epoch": 0.23, + "grad_norm": 2.0198842354422935, + "learning_rate": 1.806741538253418e-05, + "loss": 1.1367, + "step": 9559 + }, + { + "epoch": 0.23, + "grad_norm": 1.9261743295977287, + "learning_rate": 1.806696447467718e-05, + "loss": 1.1366, + "step": 9560 + }, + { + "epoch": 0.23, + "grad_norm": 1.1031207076382945, + "learning_rate": 1.806651351985176e-05, + "loss": 1.0353, + "step": 9561 + }, + { + "epoch": 0.23, + "grad_norm": 2.1299428939543916, + "learning_rate": 1.806606251806054e-05, + "loss": 1.0946, + "step": 9562 + }, + { + "epoch": 0.23, + "grad_norm": 1.9366478057664953, + "learning_rate": 1.8065611469306152e-05, + "loss": 1.0421, + "step": 9563 + }, + { + "epoch": 0.23, + "grad_norm": 1.9887721062626456, + "learning_rate": 1.806516037359122e-05, + "loss": 1.0642, + "step": 9564 + }, + { + "epoch": 0.23, + "grad_norm": 2.0806791631524297, + "learning_rate": 1.806470923091837e-05, + "loss": 1.1333, + "step": 9565 + }, + { + "epoch": 0.23, + "grad_norm": 2.9395377428922633, + "learning_rate": 1.806425804129023e-05, + "loss": 1.0278, + "step": 9566 + }, + { + "epoch": 0.23, + "grad_norm": 2.232231699297371, + "learning_rate": 1.8063806804709433e-05, + "loss": 1.1089, + "step": 9567 + }, + { + "epoch": 0.23, + "grad_norm": 1.8521085812334088, + "learning_rate": 1.8063355521178594e-05, + "loss": 1.049, + "step": 9568 + }, + { + "epoch": 0.23, + "grad_norm": 2.04526055834581, + "learning_rate": 1.8062904190700344e-05, + "loss": 1.0121, + "step": 9569 + }, + { + "epoch": 0.23, + "grad_norm": 1.8946347566600508, + "learning_rate": 1.8062452813277316e-05, + "loss": 0.8991, + "step": 9570 + }, + { + "epoch": 0.23, + "grad_norm": 1.1637446653450163, + "learning_rate": 1.8062001388912135e-05, + "loss": 0.9936, + "step": 9571 + }, + { + "epoch": 0.23, + "grad_norm": 2.767738190726928, + "learning_rate": 1.806154991760743e-05, + "loss": 1.0987, + "step": 9572 + }, + { + "epoch": 0.23, + "grad_norm": 2.15534612293685, + "learning_rate": 1.806109839936583e-05, + "loss": 0.9636, + "step": 9573 + }, + { + "epoch": 0.23, + "grad_norm": 2.4146958599727113, + "learning_rate": 1.806064683418996e-05, + "loss": 0.968, + "step": 9574 + }, + { + "epoch": 0.23, + "grad_norm": 2.378168343389946, + "learning_rate": 1.8060195222082457e-05, + "loss": 1.1332, + "step": 9575 + }, + { + "epoch": 0.23, + "grad_norm": 2.3453455993177026, + "learning_rate": 1.805974356304594e-05, + "loss": 1.0449, + "step": 9576 + }, + { + "epoch": 0.23, + "grad_norm": 2.021268069170878, + "learning_rate": 1.8059291857083048e-05, + "loss": 1.1425, + "step": 9577 + }, + { + "epoch": 0.23, + "grad_norm": 1.92600975484808, + "learning_rate": 1.8058840104196406e-05, + "loss": 1.0696, + "step": 9578 + }, + { + "epoch": 0.23, + "grad_norm": 2.3000359291400887, + "learning_rate": 1.8058388304388646e-05, + "loss": 1.0911, + "step": 9579 + }, + { + "epoch": 0.23, + "grad_norm": 2.1694976015670226, + "learning_rate": 1.80579364576624e-05, + "loss": 1.0566, + "step": 9580 + }, + { + "epoch": 0.23, + "grad_norm": 2.439008437779939, + "learning_rate": 1.8057484564020294e-05, + "loss": 1.0482, + "step": 9581 + }, + { + "epoch": 0.23, + "grad_norm": 2.1054183271842573, + "learning_rate": 1.805703262346496e-05, + "loss": 0.977, + "step": 9582 + }, + { + "epoch": 0.23, + "grad_norm": 2.0920930802082034, + "learning_rate": 1.8056580635999033e-05, + "loss": 0.9764, + "step": 9583 + }, + { + "epoch": 0.23, + "grad_norm": 1.8845672325119387, + "learning_rate": 1.8056128601625146e-05, + "loss": 1.0131, + "step": 9584 + }, + { + "epoch": 0.23, + "grad_norm": 1.8291707902823058, + "learning_rate": 1.8055676520345928e-05, + "loss": 0.972, + "step": 9585 + }, + { + "epoch": 0.23, + "grad_norm": 2.0494998592323137, + "learning_rate": 1.8055224392164005e-05, + "loss": 1.0504, + "step": 9586 + }, + { + "epoch": 0.23, + "grad_norm": 2.1334562983834697, + "learning_rate": 1.805477221708202e-05, + "loss": 1.0646, + "step": 9587 + }, + { + "epoch": 0.23, + "grad_norm": 1.8906576219317672, + "learning_rate": 1.80543199951026e-05, + "loss": 1.1972, + "step": 9588 + }, + { + "epoch": 0.23, + "grad_norm": 2.0905448404931484, + "learning_rate": 1.8053867726228376e-05, + "loss": 1.0936, + "step": 9589 + }, + { + "epoch": 0.23, + "grad_norm": 2.067897085771841, + "learning_rate": 1.8053415410461988e-05, + "loss": 1.1259, + "step": 9590 + }, + { + "epoch": 0.23, + "grad_norm": 2.130801308576751, + "learning_rate": 1.8052963047806065e-05, + "loss": 1.0384, + "step": 9591 + }, + { + "epoch": 0.23, + "grad_norm": 1.9629143425827584, + "learning_rate": 1.8052510638263244e-05, + "loss": 1.077, + "step": 9592 + }, + { + "epoch": 0.23, + "grad_norm": 2.3260969986503413, + "learning_rate": 1.8052058181836152e-05, + "loss": 0.9438, + "step": 9593 + }, + { + "epoch": 0.23, + "grad_norm": 2.1521595140018577, + "learning_rate": 1.805160567852743e-05, + "loss": 1.0979, + "step": 9594 + }, + { + "epoch": 0.23, + "grad_norm": 2.0205727054459444, + "learning_rate": 1.805115312833971e-05, + "loss": 0.9976, + "step": 9595 + }, + { + "epoch": 0.23, + "grad_norm": 1.8419124718946143, + "learning_rate": 1.8050700531275632e-05, + "loss": 0.9899, + "step": 9596 + }, + { + "epoch": 0.23, + "grad_norm": 2.2781305626518717, + "learning_rate": 1.8050247887337825e-05, + "loss": 0.964, + "step": 9597 + }, + { + "epoch": 0.23, + "grad_norm": 1.9578113752272068, + "learning_rate": 1.8049795196528924e-05, + "loss": 0.9662, + "step": 9598 + }, + { + "epoch": 0.23, + "grad_norm": 2.0631980506630434, + "learning_rate": 1.8049342458851563e-05, + "loss": 1.1074, + "step": 9599 + }, + { + "epoch": 0.23, + "grad_norm": 2.923360136616994, + "learning_rate": 1.8048889674308387e-05, + "loss": 1.0204, + "step": 9600 + }, + { + "epoch": 0.23, + "grad_norm": 2.493921921722148, + "learning_rate": 1.8048436842902025e-05, + "loss": 0.9682, + "step": 9601 + }, + { + "epoch": 0.23, + "grad_norm": 2.3572743560490377, + "learning_rate": 1.8047983964635118e-05, + "loss": 1.0238, + "step": 9602 + }, + { + "epoch": 0.23, + "grad_norm": 1.884051141929075, + "learning_rate": 1.80475310395103e-05, + "loss": 0.9916, + "step": 9603 + }, + { + "epoch": 0.23, + "grad_norm": 1.8125994207752698, + "learning_rate": 1.8047078067530205e-05, + "loss": 1.1141, + "step": 9604 + }, + { + "epoch": 0.23, + "grad_norm": 2.41627814447899, + "learning_rate": 1.804662504869748e-05, + "loss": 0.9857, + "step": 9605 + }, + { + "epoch": 0.23, + "grad_norm": 1.9124182845131474, + "learning_rate": 1.804617198301475e-05, + "loss": 1.0004, + "step": 9606 + }, + { + "epoch": 0.23, + "grad_norm": 2.2216014441661063, + "learning_rate": 1.8045718870484662e-05, + "loss": 1.0561, + "step": 9607 + }, + { + "epoch": 0.23, + "grad_norm": 2.382332218244304, + "learning_rate": 1.804526571110985e-05, + "loss": 1.0601, + "step": 9608 + }, + { + "epoch": 0.23, + "grad_norm": 1.8719407171316425, + "learning_rate": 1.8044812504892958e-05, + "loss": 0.9379, + "step": 9609 + }, + { + "epoch": 0.23, + "grad_norm": 4.00172142587524, + "learning_rate": 1.8044359251836615e-05, + "loss": 0.9414, + "step": 9610 + }, + { + "epoch": 0.23, + "grad_norm": 1.9485885655192543, + "learning_rate": 1.8043905951943472e-05, + "loss": 0.9923, + "step": 9611 + }, + { + "epoch": 0.23, + "grad_norm": 2.0279200157860555, + "learning_rate": 1.8043452605216157e-05, + "loss": 1.05, + "step": 9612 + }, + { + "epoch": 0.23, + "grad_norm": 2.039786261740472, + "learning_rate": 1.804299921165732e-05, + "loss": 1.1617, + "step": 9613 + }, + { + "epoch": 0.23, + "grad_norm": 2.1135263393418033, + "learning_rate": 1.804254577126959e-05, + "loss": 1.0273, + "step": 9614 + }, + { + "epoch": 0.23, + "grad_norm": 2.130261779496892, + "learning_rate": 1.8042092284055615e-05, + "loss": 1.019, + "step": 9615 + }, + { + "epoch": 0.23, + "grad_norm": 2.765711826440514, + "learning_rate": 1.804163875001803e-05, + "loss": 1.1212, + "step": 9616 + }, + { + "epoch": 0.23, + "grad_norm": 1.2379475235614834, + "learning_rate": 1.8041185169159478e-05, + "loss": 0.9503, + "step": 9617 + }, + { + "epoch": 0.23, + "grad_norm": 2.526080807054074, + "learning_rate": 1.8040731541482603e-05, + "loss": 1.0132, + "step": 9618 + }, + { + "epoch": 0.23, + "grad_norm": 2.125284380260682, + "learning_rate": 1.804027786699004e-05, + "loss": 1.095, + "step": 9619 + }, + { + "epoch": 0.23, + "grad_norm": 2.1999771581030902, + "learning_rate": 1.8039824145684438e-05, + "loss": 1.0895, + "step": 9620 + }, + { + "epoch": 0.23, + "grad_norm": 2.2030449925218845, + "learning_rate": 1.803937037756843e-05, + "loss": 0.8837, + "step": 9621 + }, + { + "epoch": 0.23, + "grad_norm": 1.8940932358038107, + "learning_rate": 1.8038916562644666e-05, + "loss": 1.0618, + "step": 9622 + }, + { + "epoch": 0.23, + "grad_norm": 2.609633254235112, + "learning_rate": 1.8038462700915782e-05, + "loss": 1.0575, + "step": 9623 + }, + { + "epoch": 0.23, + "grad_norm": 1.115549554898417, + "learning_rate": 1.8038008792384426e-05, + "loss": 1.0113, + "step": 9624 + }, + { + "epoch": 0.23, + "grad_norm": 2.873279636498928, + "learning_rate": 1.8037554837053238e-05, + "loss": 0.9956, + "step": 9625 + }, + { + "epoch": 0.23, + "grad_norm": 2.020140140594465, + "learning_rate": 1.803710083492486e-05, + "loss": 1.077, + "step": 9626 + }, + { + "epoch": 0.23, + "grad_norm": 2.0928567805452034, + "learning_rate": 1.8036646786001935e-05, + "loss": 1.0523, + "step": 9627 + }, + { + "epoch": 0.23, + "grad_norm": 2.2068368897898947, + "learning_rate": 1.803619269028711e-05, + "loss": 1.0434, + "step": 9628 + }, + { + "epoch": 0.23, + "grad_norm": 1.8241613491858297, + "learning_rate": 1.8035738547783025e-05, + "loss": 0.9841, + "step": 9629 + }, + { + "epoch": 0.23, + "grad_norm": 2.043952146091329, + "learning_rate": 1.803528435849233e-05, + "loss": 1.0246, + "step": 9630 + }, + { + "epoch": 0.23, + "grad_norm": 2.0390167431018162, + "learning_rate": 1.8034830122417658e-05, + "loss": 0.9153, + "step": 9631 + }, + { + "epoch": 0.23, + "grad_norm": 2.042689102564556, + "learning_rate": 1.803437583956167e-05, + "loss": 0.95, + "step": 9632 + }, + { + "epoch": 0.23, + "grad_norm": 3.4807231526592446, + "learning_rate": 1.8033921509926998e-05, + "loss": 1.0412, + "step": 9633 + }, + { + "epoch": 0.23, + "grad_norm": 1.1027456456212115, + "learning_rate": 1.8033467133516288e-05, + "loss": 1.0224, + "step": 9634 + }, + { + "epoch": 0.23, + "grad_norm": 2.1555378981295816, + "learning_rate": 1.8033012710332195e-05, + "loss": 0.9765, + "step": 9635 + }, + { + "epoch": 0.23, + "grad_norm": 2.526697121325709, + "learning_rate": 1.8032558240377355e-05, + "loss": 0.9884, + "step": 9636 + }, + { + "epoch": 0.23, + "grad_norm": 2.080926393651884, + "learning_rate": 1.803210372365442e-05, + "loss": 1.0268, + "step": 9637 + }, + { + "epoch": 0.23, + "grad_norm": 2.1820592721894125, + "learning_rate": 1.8031649160166035e-05, + "loss": 1.1892, + "step": 9638 + }, + { + "epoch": 0.23, + "grad_norm": 2.2303818897271808, + "learning_rate": 1.8031194549914845e-05, + "loss": 0.963, + "step": 9639 + }, + { + "epoch": 0.23, + "grad_norm": 2.296189162386637, + "learning_rate": 1.8030739892903498e-05, + "loss": 1.1232, + "step": 9640 + }, + { + "epoch": 0.23, + "grad_norm": 1.8945460568419659, + "learning_rate": 1.803028518913464e-05, + "loss": 0.981, + "step": 9641 + }, + { + "epoch": 0.23, + "grad_norm": 1.9115106696905568, + "learning_rate": 1.8029830438610916e-05, + "loss": 1.0099, + "step": 9642 + }, + { + "epoch": 0.23, + "grad_norm": 2.0835664505720284, + "learning_rate": 1.802937564133498e-05, + "loss": 1.1429, + "step": 9643 + }, + { + "epoch": 0.23, + "grad_norm": 2.164324654667175, + "learning_rate": 1.8028920797309478e-05, + "loss": 0.9691, + "step": 9644 + }, + { + "epoch": 0.23, + "grad_norm": 2.149773584818657, + "learning_rate": 1.8028465906537055e-05, + "loss": 1.0112, + "step": 9645 + }, + { + "epoch": 0.23, + "grad_norm": 1.9800065115310728, + "learning_rate": 1.8028010969020364e-05, + "loss": 0.969, + "step": 9646 + }, + { + "epoch": 0.23, + "grad_norm": 2.1576366425496327, + "learning_rate": 1.802755598476205e-05, + "loss": 0.8748, + "step": 9647 + }, + { + "epoch": 0.23, + "grad_norm": 2.0529199462122745, + "learning_rate": 1.8027100953764766e-05, + "loss": 1.1038, + "step": 9648 + }, + { + "epoch": 0.23, + "grad_norm": 2.3866911463538187, + "learning_rate": 1.802664587603116e-05, + "loss": 1.1147, + "step": 9649 + }, + { + "epoch": 0.23, + "grad_norm": 2.1932578122784316, + "learning_rate": 1.8026190751563874e-05, + "loss": 0.9262, + "step": 9650 + }, + { + "epoch": 0.23, + "grad_norm": 2.0118249498269876, + "learning_rate": 1.8025735580365568e-05, + "loss": 1.0294, + "step": 9651 + }, + { + "epoch": 0.23, + "grad_norm": 2.2462745039190892, + "learning_rate": 1.802528036243889e-05, + "loss": 1.0289, + "step": 9652 + }, + { + "epoch": 0.23, + "grad_norm": 1.041142854242808, + "learning_rate": 1.8024825097786487e-05, + "loss": 0.9683, + "step": 9653 + }, + { + "epoch": 0.23, + "grad_norm": 1.862441849145793, + "learning_rate": 1.802436978641101e-05, + "loss": 1.0111, + "step": 9654 + }, + { + "epoch": 0.23, + "grad_norm": 2.001140102700724, + "learning_rate": 1.8023914428315115e-05, + "loss": 1.0863, + "step": 9655 + }, + { + "epoch": 0.23, + "grad_norm": 2.3217562789843895, + "learning_rate": 1.802345902350145e-05, + "loss": 0.9089, + "step": 9656 + }, + { + "epoch": 0.23, + "grad_norm": 2.248557331605427, + "learning_rate": 1.8023003571972664e-05, + "loss": 1.0621, + "step": 9657 + }, + { + "epoch": 0.23, + "grad_norm": 2.2987484623411265, + "learning_rate": 1.8022548073731413e-05, + "loss": 0.9646, + "step": 9658 + }, + { + "epoch": 0.23, + "grad_norm": 2.278285689015234, + "learning_rate": 1.8022092528780348e-05, + "loss": 1.0148, + "step": 9659 + }, + { + "epoch": 0.23, + "grad_norm": 1.9408424098824208, + "learning_rate": 1.8021636937122117e-05, + "loss": 1.0834, + "step": 9660 + }, + { + "epoch": 0.23, + "grad_norm": 2.0015167389903166, + "learning_rate": 1.8021181298759377e-05, + "loss": 1.094, + "step": 9661 + }, + { + "epoch": 0.23, + "grad_norm": 1.1931202150541023, + "learning_rate": 1.802072561369478e-05, + "loss": 1.1212, + "step": 9662 + }, + { + "epoch": 0.23, + "grad_norm": 2.021825094695116, + "learning_rate": 1.8020269881930982e-05, + "loss": 1.0194, + "step": 9663 + }, + { + "epoch": 0.23, + "grad_norm": 1.930857601532497, + "learning_rate": 1.8019814103470633e-05, + "loss": 1.0957, + "step": 9664 + }, + { + "epoch": 0.23, + "grad_norm": 1.1947210787275757, + "learning_rate": 1.8019358278316386e-05, + "loss": 1.03, + "step": 9665 + }, + { + "epoch": 0.23, + "grad_norm": 2.0781065844359836, + "learning_rate": 1.8018902406470894e-05, + "loss": 1.1891, + "step": 9666 + }, + { + "epoch": 0.23, + "grad_norm": 1.9839925330530024, + "learning_rate": 1.8018446487936815e-05, + "loss": 0.972, + "step": 9667 + }, + { + "epoch": 0.23, + "grad_norm": 2.8568827759528266, + "learning_rate": 1.8017990522716804e-05, + "loss": 1.0121, + "step": 9668 + }, + { + "epoch": 0.23, + "grad_norm": 2.8059663473643814, + "learning_rate": 1.8017534510813513e-05, + "loss": 0.9351, + "step": 9669 + }, + { + "epoch": 0.23, + "grad_norm": 2.530484595684182, + "learning_rate": 1.8017078452229597e-05, + "loss": 1.0879, + "step": 9670 + }, + { + "epoch": 0.23, + "grad_norm": 2.218476713870038, + "learning_rate": 1.8016622346967714e-05, + "loss": 1.1456, + "step": 9671 + }, + { + "epoch": 0.23, + "grad_norm": 1.1100394291270879, + "learning_rate": 1.8016166195030514e-05, + "loss": 0.9913, + "step": 9672 + }, + { + "epoch": 0.23, + "grad_norm": 1.1037428364576611, + "learning_rate": 1.801570999642066e-05, + "loss": 0.9347, + "step": 9673 + }, + { + "epoch": 0.23, + "grad_norm": 1.9076546246251198, + "learning_rate": 1.8015253751140803e-05, + "loss": 1.0851, + "step": 9674 + }, + { + "epoch": 0.23, + "grad_norm": 2.1694552312952893, + "learning_rate": 1.80147974591936e-05, + "loss": 1.0269, + "step": 9675 + }, + { + "epoch": 0.23, + "grad_norm": 2.2928267710975536, + "learning_rate": 1.801434112058171e-05, + "loss": 1.2352, + "step": 9676 + }, + { + "epoch": 0.23, + "grad_norm": 1.9800291231811986, + "learning_rate": 1.801388473530779e-05, + "loss": 0.9998, + "step": 9677 + }, + { + "epoch": 0.23, + "grad_norm": 1.9738494777949418, + "learning_rate": 1.8013428303374495e-05, + "loss": 1.1568, + "step": 9678 + }, + { + "epoch": 0.23, + "grad_norm": 1.0710828101398011, + "learning_rate": 1.801297182478448e-05, + "loss": 0.9716, + "step": 9679 + }, + { + "epoch": 0.23, + "grad_norm": 2.7667114391769303, + "learning_rate": 1.8012515299540413e-05, + "loss": 1.123, + "step": 9680 + }, + { + "epoch": 0.23, + "grad_norm": 1.8356440668531162, + "learning_rate": 1.8012058727644943e-05, + "loss": 1.0648, + "step": 9681 + }, + { + "epoch": 0.23, + "grad_norm": 2.2695108086950686, + "learning_rate": 1.8011602109100724e-05, + "loss": 1.0045, + "step": 9682 + }, + { + "epoch": 0.23, + "grad_norm": 2.1289084245294436, + "learning_rate": 1.8011145443910426e-05, + "loss": 1.0816, + "step": 9683 + }, + { + "epoch": 0.23, + "grad_norm": 3.1261477944877685, + "learning_rate": 1.8010688732076706e-05, + "loss": 0.9714, + "step": 9684 + }, + { + "epoch": 0.23, + "grad_norm": 2.00908315468758, + "learning_rate": 1.8010231973602218e-05, + "loss": 0.9828, + "step": 9685 + }, + { + "epoch": 0.23, + "grad_norm": 2.200671935278069, + "learning_rate": 1.8009775168489622e-05, + "loss": 1.0508, + "step": 9686 + }, + { + "epoch": 0.23, + "grad_norm": 2.209125485411688, + "learning_rate": 1.8009318316741577e-05, + "loss": 1.0972, + "step": 9687 + }, + { + "epoch": 0.23, + "grad_norm": 1.9524861808578515, + "learning_rate": 1.8008861418360746e-05, + "loss": 0.9463, + "step": 9688 + }, + { + "epoch": 0.23, + "grad_norm": 2.061892299482143, + "learning_rate": 1.8008404473349793e-05, + "loss": 1.0911, + "step": 9689 + }, + { + "epoch": 0.23, + "grad_norm": 1.1460461636053392, + "learning_rate": 1.800794748171137e-05, + "loss": 1.0654, + "step": 9690 + }, + { + "epoch": 0.23, + "grad_norm": 1.932244693652624, + "learning_rate": 1.8007490443448144e-05, + "loss": 1.0754, + "step": 9691 + }, + { + "epoch": 0.23, + "grad_norm": 2.06763866457053, + "learning_rate": 1.800703335856277e-05, + "loss": 1.0408, + "step": 9692 + }, + { + "epoch": 0.23, + "grad_norm": 2.343327662940459, + "learning_rate": 1.8006576227057914e-05, + "loss": 1.0892, + "step": 9693 + }, + { + "epoch": 0.23, + "grad_norm": 1.8394432748181269, + "learning_rate": 1.8006119048936236e-05, + "loss": 1.1278, + "step": 9694 + }, + { + "epoch": 0.23, + "grad_norm": 1.8689393013074043, + "learning_rate": 1.80056618242004e-05, + "loss": 1.0361, + "step": 9695 + }, + { + "epoch": 0.23, + "grad_norm": 2.174398450836943, + "learning_rate": 1.8005204552853068e-05, + "loss": 1.1618, + "step": 9696 + }, + { + "epoch": 0.23, + "grad_norm": 2.0571628112090523, + "learning_rate": 1.8004747234896896e-05, + "loss": 1.0931, + "step": 9697 + }, + { + "epoch": 0.23, + "grad_norm": 1.9357680581461456, + "learning_rate": 1.8004289870334552e-05, + "loss": 1.0327, + "step": 9698 + }, + { + "epoch": 0.23, + "grad_norm": 2.007756198101086, + "learning_rate": 1.8003832459168703e-05, + "loss": 1.0733, + "step": 9699 + }, + { + "epoch": 0.23, + "grad_norm": 2.283317796766036, + "learning_rate": 1.8003375001402005e-05, + "loss": 1.1253, + "step": 9700 + }, + { + "epoch": 0.23, + "grad_norm": 2.562118013062494, + "learning_rate": 1.800291749703712e-05, + "loss": 1.0594, + "step": 9701 + }, + { + "epoch": 0.23, + "grad_norm": 4.933736314227696, + "learning_rate": 1.8002459946076717e-05, + "loss": 0.8699, + "step": 9702 + }, + { + "epoch": 0.23, + "grad_norm": 2.030142274604195, + "learning_rate": 1.8002002348523463e-05, + "loss": 1.0324, + "step": 9703 + }, + { + "epoch": 0.23, + "grad_norm": 1.8531709864130774, + "learning_rate": 1.8001544704380013e-05, + "loss": 1.1247, + "step": 9704 + }, + { + "epoch": 0.23, + "grad_norm": 2.076440726687659, + "learning_rate": 1.800108701364904e-05, + "loss": 1.0439, + "step": 9705 + }, + { + "epoch": 0.23, + "grad_norm": 2.1696802742141483, + "learning_rate": 1.8000629276333204e-05, + "loss": 1.1296, + "step": 9706 + }, + { + "epoch": 0.23, + "grad_norm": 2.655222263777793, + "learning_rate": 1.8000171492435172e-05, + "loss": 1.075, + "step": 9707 + }, + { + "epoch": 0.23, + "grad_norm": 2.1248046194830947, + "learning_rate": 1.799971366195761e-05, + "loss": 1.1131, + "step": 9708 + }, + { + "epoch": 0.23, + "grad_norm": 1.9144039653523164, + "learning_rate": 1.7999255784903178e-05, + "loss": 1.0977, + "step": 9709 + }, + { + "epoch": 0.23, + "grad_norm": 2.2169381324047404, + "learning_rate": 1.799879786127455e-05, + "loss": 1.0236, + "step": 9710 + }, + { + "epoch": 0.23, + "grad_norm": 1.15442559489517, + "learning_rate": 1.7998339891074385e-05, + "loss": 1.0317, + "step": 9711 + }, + { + "epoch": 0.23, + "grad_norm": 1.9828024989729616, + "learning_rate": 1.7997881874305356e-05, + "loss": 1.0423, + "step": 9712 + }, + { + "epoch": 0.23, + "grad_norm": 2.1023573614196915, + "learning_rate": 1.7997423810970128e-05, + "loss": 1.1509, + "step": 9713 + }, + { + "epoch": 0.23, + "grad_norm": 2.3036244957518432, + "learning_rate": 1.7996965701071364e-05, + "loss": 0.9691, + "step": 9714 + }, + { + "epoch": 0.23, + "grad_norm": 2.0355551286675366, + "learning_rate": 1.7996507544611735e-05, + "loss": 1.057, + "step": 9715 + }, + { + "epoch": 0.23, + "grad_norm": 2.261501851886056, + "learning_rate": 1.7996049341593903e-05, + "loss": 0.9983, + "step": 9716 + }, + { + "epoch": 0.23, + "grad_norm": 2.3176042147675977, + "learning_rate": 1.7995591092020544e-05, + "loss": 1.1096, + "step": 9717 + }, + { + "epoch": 0.23, + "grad_norm": 2.1151816996886788, + "learning_rate": 1.7995132795894323e-05, + "loss": 1.0629, + "step": 9718 + }, + { + "epoch": 0.23, + "grad_norm": 2.2963775696025706, + "learning_rate": 1.799467445321791e-05, + "loss": 1.1244, + "step": 9719 + }, + { + "epoch": 0.23, + "grad_norm": 1.8874235376302115, + "learning_rate": 1.7994216063993968e-05, + "loss": 0.9668, + "step": 9720 + }, + { + "epoch": 0.23, + "grad_norm": 2.2290067389638395, + "learning_rate": 1.7993757628225166e-05, + "loss": 1.0385, + "step": 9721 + }, + { + "epoch": 0.23, + "grad_norm": 2.1681104297995786, + "learning_rate": 1.799329914591418e-05, + "loss": 1.0151, + "step": 9722 + }, + { + "epoch": 0.23, + "grad_norm": 1.867335206126585, + "learning_rate": 1.799284061706368e-05, + "loss": 1.0375, + "step": 9723 + }, + { + "epoch": 0.23, + "grad_norm": 2.1373280923853963, + "learning_rate": 1.7992382041676326e-05, + "loss": 1.0041, + "step": 9724 + }, + { + "epoch": 0.23, + "grad_norm": 7.596140271338576, + "learning_rate": 1.7991923419754793e-05, + "loss": 0.9129, + "step": 9725 + }, + { + "epoch": 0.23, + "grad_norm": 2.149186709315785, + "learning_rate": 1.799146475130175e-05, + "loss": 1.0214, + "step": 9726 + }, + { + "epoch": 0.23, + "grad_norm": 1.8011019707768097, + "learning_rate": 1.7991006036319876e-05, + "loss": 0.9588, + "step": 9727 + }, + { + "epoch": 0.23, + "grad_norm": 2.218123105424997, + "learning_rate": 1.799054727481183e-05, + "loss": 0.8711, + "step": 9728 + }, + { + "epoch": 0.23, + "grad_norm": 2.2950330636299263, + "learning_rate": 1.799008846678029e-05, + "loss": 0.8904, + "step": 9729 + }, + { + "epoch": 0.23, + "grad_norm": 1.9755276810706688, + "learning_rate": 1.7989629612227924e-05, + "loss": 1.1291, + "step": 9730 + }, + { + "epoch": 0.23, + "grad_norm": 2.0356756645094767, + "learning_rate": 1.7989170711157406e-05, + "loss": 1.0066, + "step": 9731 + }, + { + "epoch": 0.23, + "grad_norm": 2.0148768175350655, + "learning_rate": 1.7988711763571406e-05, + "loss": 1.0539, + "step": 9732 + }, + { + "epoch": 0.23, + "grad_norm": 2.165245884771482, + "learning_rate": 1.7988252769472597e-05, + "loss": 0.9724, + "step": 9733 + }, + { + "epoch": 0.23, + "grad_norm": 1.0924144641902025, + "learning_rate": 1.798779372886365e-05, + "loss": 0.9673, + "step": 9734 + }, + { + "epoch": 0.23, + "grad_norm": 1.9604307553857745, + "learning_rate": 1.798733464174724e-05, + "loss": 1.074, + "step": 9735 + }, + { + "epoch": 0.23, + "grad_norm": 2.222099144606807, + "learning_rate": 1.798687550812604e-05, + "loss": 0.9649, + "step": 9736 + }, + { + "epoch": 0.23, + "grad_norm": 1.9775439476612173, + "learning_rate": 1.7986416328002723e-05, + "loss": 1.0095, + "step": 9737 + }, + { + "epoch": 0.23, + "grad_norm": 2.119792606001001, + "learning_rate": 1.798595710137996e-05, + "loss": 1.0555, + "step": 9738 + }, + { + "epoch": 0.23, + "grad_norm": 4.006937643436651, + "learning_rate": 1.7985497828260425e-05, + "loss": 1.0285, + "step": 9739 + }, + { + "epoch": 0.23, + "grad_norm": 2.939109138943109, + "learning_rate": 1.7985038508646794e-05, + "loss": 1.034, + "step": 9740 + }, + { + "epoch": 0.23, + "grad_norm": 2.1563177737351786, + "learning_rate": 1.7984579142541743e-05, + "loss": 1.0057, + "step": 9741 + }, + { + "epoch": 0.23, + "grad_norm": 1.95383644497903, + "learning_rate": 1.7984119729947944e-05, + "loss": 1.0704, + "step": 9742 + }, + { + "epoch": 0.23, + "grad_norm": 2.306635391998977, + "learning_rate": 1.7983660270868074e-05, + "loss": 1.109, + "step": 9743 + }, + { + "epoch": 0.23, + "grad_norm": 1.9367511847271202, + "learning_rate": 1.7983200765304802e-05, + "loss": 1.092, + "step": 9744 + }, + { + "epoch": 0.23, + "grad_norm": 2.045981386137383, + "learning_rate": 1.798274121326081e-05, + "loss": 1.0364, + "step": 9745 + }, + { + "epoch": 0.23, + "grad_norm": 2.4324857278300835, + "learning_rate": 1.798228161473877e-05, + "loss": 1.0752, + "step": 9746 + }, + { + "epoch": 0.23, + "grad_norm": 2.0934033179124376, + "learning_rate": 1.798182196974136e-05, + "loss": 0.9514, + "step": 9747 + }, + { + "epoch": 0.23, + "grad_norm": 1.9723184973912524, + "learning_rate": 1.7981362278271258e-05, + "loss": 1.0213, + "step": 9748 + }, + { + "epoch": 0.23, + "grad_norm": 1.8968452332878534, + "learning_rate": 1.7980902540331134e-05, + "loss": 1.083, + "step": 9749 + }, + { + "epoch": 0.23, + "grad_norm": 2.272658314204828, + "learning_rate": 1.798044275592367e-05, + "loss": 1.1911, + "step": 9750 + }, + { + "epoch": 0.23, + "grad_norm": 2.0668702344910486, + "learning_rate": 1.797998292505154e-05, + "loss": 1.1155, + "step": 9751 + }, + { + "epoch": 0.23, + "grad_norm": 2.141132279456399, + "learning_rate": 1.7979523047717427e-05, + "loss": 1.0439, + "step": 9752 + }, + { + "epoch": 0.23, + "grad_norm": 1.854391325537484, + "learning_rate": 1.7979063123924004e-05, + "loss": 1.0608, + "step": 9753 + }, + { + "epoch": 0.23, + "grad_norm": 2.055836436096399, + "learning_rate": 1.7978603153673944e-05, + "loss": 1.0392, + "step": 9754 + }, + { + "epoch": 0.23, + "grad_norm": 1.1373844290903072, + "learning_rate": 1.7978143136969938e-05, + "loss": 1.0284, + "step": 9755 + }, + { + "epoch": 0.23, + "grad_norm": 1.924371255716298, + "learning_rate": 1.7977683073814655e-05, + "loss": 1.0307, + "step": 9756 + }, + { + "epoch": 0.23, + "grad_norm": 1.9771329421534949, + "learning_rate": 1.797722296421077e-05, + "loss": 1.1321, + "step": 9757 + }, + { + "epoch": 0.23, + "grad_norm": 1.9505553678447314, + "learning_rate": 1.7976762808160972e-05, + "loss": 1.0442, + "step": 9758 + }, + { + "epoch": 0.23, + "grad_norm": 2.069474412993364, + "learning_rate": 1.7976302605667936e-05, + "loss": 1.0081, + "step": 9759 + }, + { + "epoch": 0.23, + "grad_norm": 2.005558044212582, + "learning_rate": 1.797584235673434e-05, + "loss": 0.9093, + "step": 9760 + }, + { + "epoch": 0.23, + "grad_norm": 2.250433801971129, + "learning_rate": 1.7975382061362863e-05, + "loss": 1.1453, + "step": 9761 + }, + { + "epoch": 0.23, + "grad_norm": 2.1555552008911407, + "learning_rate": 1.797492171955619e-05, + "loss": 1.0974, + "step": 9762 + }, + { + "epoch": 0.23, + "grad_norm": 2.170739886036269, + "learning_rate": 1.7974461331316994e-05, + "loss": 1.154, + "step": 9763 + }, + { + "epoch": 0.23, + "grad_norm": 1.398339699676504, + "learning_rate": 1.797400089664796e-05, + "loss": 1.0939, + "step": 9764 + }, + { + "epoch": 0.23, + "grad_norm": 1.922425649375852, + "learning_rate": 1.797354041555177e-05, + "loss": 1.0227, + "step": 9765 + }, + { + "epoch": 0.23, + "grad_norm": 2.145210576075981, + "learning_rate": 1.7973079888031102e-05, + "loss": 0.988, + "step": 9766 + }, + { + "epoch": 0.23, + "grad_norm": 1.8442331350121925, + "learning_rate": 1.7972619314088643e-05, + "loss": 1.0678, + "step": 9767 + }, + { + "epoch": 0.23, + "grad_norm": 2.253898759429491, + "learning_rate": 1.7972158693727064e-05, + "loss": 1.186, + "step": 9768 + }, + { + "epoch": 0.23, + "grad_norm": 2.5080683175150003, + "learning_rate": 1.7971698026949055e-05, + "loss": 1.1067, + "step": 9769 + }, + { + "epoch": 0.23, + "grad_norm": 2.250418403709301, + "learning_rate": 1.79712373137573e-05, + "loss": 0.9387, + "step": 9770 + }, + { + "epoch": 0.23, + "grad_norm": 2.189232770906075, + "learning_rate": 1.7970776554154472e-05, + "loss": 1.0047, + "step": 9771 + }, + { + "epoch": 0.23, + "grad_norm": 2.4384407023442516, + "learning_rate": 1.7970315748143264e-05, + "loss": 1.0931, + "step": 9772 + }, + { + "epoch": 0.23, + "grad_norm": 1.9868664514900185, + "learning_rate": 1.7969854895726354e-05, + "loss": 1.1057, + "step": 9773 + }, + { + "epoch": 0.23, + "grad_norm": 2.0553377549431042, + "learning_rate": 1.796939399690642e-05, + "loss": 1.0162, + "step": 9774 + }, + { + "epoch": 0.23, + "grad_norm": 2.0191970504631835, + "learning_rate": 1.7968933051686157e-05, + "loss": 1.0598, + "step": 9775 + }, + { + "epoch": 0.23, + "grad_norm": 2.526087932668897, + "learning_rate": 1.7968472060068242e-05, + "loss": 1.095, + "step": 9776 + }, + { + "epoch": 0.23, + "grad_norm": 1.8584551202455668, + "learning_rate": 1.796801102205536e-05, + "loss": 1.02, + "step": 9777 + }, + { + "epoch": 0.23, + "grad_norm": 2.6655565949592637, + "learning_rate": 1.7967549937650193e-05, + "loss": 0.9673, + "step": 9778 + }, + { + "epoch": 0.23, + "grad_norm": 1.8345660974536788, + "learning_rate": 1.796708880685543e-05, + "loss": 1.1004, + "step": 9779 + }, + { + "epoch": 0.23, + "grad_norm": 1.1728434737098887, + "learning_rate": 1.796662762967375e-05, + "loss": 1.042, + "step": 9780 + }, + { + "epoch": 0.23, + "grad_norm": 1.8312440525046232, + "learning_rate": 1.7966166406107847e-05, + "loss": 1.0737, + "step": 9781 + }, + { + "epoch": 0.23, + "grad_norm": 3.9960999903625507, + "learning_rate": 1.7965705136160397e-05, + "loss": 0.919, + "step": 9782 + }, + { + "epoch": 0.23, + "grad_norm": 1.8301579747800094, + "learning_rate": 1.7965243819834092e-05, + "loss": 1.0433, + "step": 9783 + }, + { + "epoch": 0.23, + "grad_norm": 2.215148104674685, + "learning_rate": 1.7964782457131615e-05, + "loss": 1.0392, + "step": 9784 + }, + { + "epoch": 0.23, + "grad_norm": 2.1178324855598047, + "learning_rate": 1.7964321048055656e-05, + "loss": 1.1376, + "step": 9785 + }, + { + "epoch": 0.23, + "grad_norm": 1.9236628257468862, + "learning_rate": 1.7963859592608896e-05, + "loss": 1.067, + "step": 9786 + }, + { + "epoch": 0.23, + "grad_norm": 1.8737234543422379, + "learning_rate": 1.796339809079402e-05, + "loss": 0.9546, + "step": 9787 + }, + { + "epoch": 0.23, + "grad_norm": 3.3493046430791678, + "learning_rate": 1.7962936542613723e-05, + "loss": 1.0942, + "step": 9788 + }, + { + "epoch": 0.23, + "grad_norm": 1.8327526549789082, + "learning_rate": 1.7962474948070688e-05, + "loss": 1.0677, + "step": 9789 + }, + { + "epoch": 0.23, + "grad_norm": 2.0867998630679843, + "learning_rate": 1.7962013307167602e-05, + "loss": 1.1886, + "step": 9790 + }, + { + "epoch": 0.23, + "grad_norm": 2.080787244776158, + "learning_rate": 1.7961551619907158e-05, + "loss": 1.0726, + "step": 9791 + }, + { + "epoch": 0.23, + "grad_norm": 2.22579482641545, + "learning_rate": 1.7961089886292036e-05, + "loss": 0.9615, + "step": 9792 + }, + { + "epoch": 0.23, + "grad_norm": 2.4722409912843415, + "learning_rate": 1.7960628106324927e-05, + "loss": 1.0564, + "step": 9793 + }, + { + "epoch": 0.23, + "grad_norm": 2.3122711912171865, + "learning_rate": 1.7960166280008528e-05, + "loss": 1.0281, + "step": 9794 + }, + { + "epoch": 0.23, + "grad_norm": 2.0964149554271896, + "learning_rate": 1.795970440734551e-05, + "loss": 1.0273, + "step": 9795 + }, + { + "epoch": 0.23, + "grad_norm": 1.1736727405805303, + "learning_rate": 1.795924248833858e-05, + "loss": 1.0315, + "step": 9796 + }, + { + "epoch": 0.23, + "grad_norm": 2.618188148535011, + "learning_rate": 1.795878052299042e-05, + "loss": 0.9075, + "step": 9797 + }, + { + "epoch": 0.23, + "grad_norm": 1.9488847394171285, + "learning_rate": 1.7958318511303717e-05, + "loss": 0.963, + "step": 9798 + }, + { + "epoch": 0.23, + "grad_norm": 2.601679862836145, + "learning_rate": 1.795785645328117e-05, + "loss": 1.0794, + "step": 9799 + }, + { + "epoch": 0.23, + "grad_norm": 1.827924980068249, + "learning_rate": 1.795739434892546e-05, + "loss": 1.1172, + "step": 9800 + }, + { + "epoch": 0.23, + "grad_norm": 2.2961234678520652, + "learning_rate": 1.7956932198239278e-05, + "loss": 0.9713, + "step": 9801 + }, + { + "epoch": 0.23, + "grad_norm": 2.0701018136262768, + "learning_rate": 1.7956470001225322e-05, + "loss": 1.0154, + "step": 9802 + }, + { + "epoch": 0.23, + "grad_norm": 1.9829335532467196, + "learning_rate": 1.7956007757886275e-05, + "loss": 0.9781, + "step": 9803 + }, + { + "epoch": 0.23, + "grad_norm": 2.031446828097528, + "learning_rate": 1.7955545468224832e-05, + "loss": 1.0663, + "step": 9804 + }, + { + "epoch": 0.23, + "grad_norm": 2.043957026758357, + "learning_rate": 1.7955083132243687e-05, + "loss": 0.9981, + "step": 9805 + }, + { + "epoch": 0.23, + "grad_norm": 1.074933631856972, + "learning_rate": 1.795462074994553e-05, + "loss": 0.984, + "step": 9806 + }, + { + "epoch": 0.23, + "grad_norm": 1.8320268759161822, + "learning_rate": 1.795415832133305e-05, + "loss": 1.0606, + "step": 9807 + }, + { + "epoch": 0.23, + "grad_norm": 2.3603821995817933, + "learning_rate": 1.7953695846408945e-05, + "loss": 1.1109, + "step": 9808 + }, + { + "epoch": 0.23, + "grad_norm": 1.9277048580949683, + "learning_rate": 1.7953233325175904e-05, + "loss": 1.0381, + "step": 9809 + }, + { + "epoch": 0.23, + "grad_norm": 2.374565413164265, + "learning_rate": 1.795277075763662e-05, + "loss": 1.0376, + "step": 9810 + }, + { + "epoch": 0.23, + "grad_norm": 2.29448966186934, + "learning_rate": 1.7952308143793783e-05, + "loss": 1.0663, + "step": 9811 + }, + { + "epoch": 0.23, + "grad_norm": 1.1035253094990884, + "learning_rate": 1.7951845483650093e-05, + "loss": 0.9583, + "step": 9812 + }, + { + "epoch": 0.23, + "grad_norm": 1.919214947153512, + "learning_rate": 1.7951382777208243e-05, + "loss": 1.035, + "step": 9813 + }, + { + "epoch": 0.23, + "grad_norm": 2.3819162100895435, + "learning_rate": 1.7950920024470923e-05, + "loss": 1.0687, + "step": 9814 + }, + { + "epoch": 0.23, + "grad_norm": 1.9313665200113819, + "learning_rate": 1.795045722544083e-05, + "loss": 1.113, + "step": 9815 + }, + { + "epoch": 0.23, + "grad_norm": 1.8895818867410847, + "learning_rate": 1.794999438012066e-05, + "loss": 0.9663, + "step": 9816 + }, + { + "epoch": 0.23, + "grad_norm": 2.1362664666548796, + "learning_rate": 1.7949531488513106e-05, + "loss": 1.0966, + "step": 9817 + }, + { + "epoch": 0.23, + "grad_norm": 2.1315206230304105, + "learning_rate": 1.794906855062086e-05, + "loss": 1.008, + "step": 9818 + }, + { + "epoch": 0.23, + "grad_norm": 1.8952963526594955, + "learning_rate": 1.794860556644662e-05, + "loss": 1.0734, + "step": 9819 + }, + { + "epoch": 0.23, + "grad_norm": 2.2520248972571175, + "learning_rate": 1.7948142535993084e-05, + "loss": 1.0856, + "step": 9820 + }, + { + "epoch": 0.23, + "grad_norm": 1.9658543477307562, + "learning_rate": 1.7947679459262945e-05, + "loss": 1.1804, + "step": 9821 + }, + { + "epoch": 0.23, + "grad_norm": 2.4873877605719734, + "learning_rate": 1.79472163362589e-05, + "loss": 0.9127, + "step": 9822 + }, + { + "epoch": 0.23, + "grad_norm": 1.9941179285196282, + "learning_rate": 1.7946753166983642e-05, + "loss": 0.9961, + "step": 9823 + }, + { + "epoch": 0.23, + "grad_norm": 2.231776030696977, + "learning_rate": 1.7946289951439876e-05, + "loss": 1.0166, + "step": 9824 + }, + { + "epoch": 0.23, + "grad_norm": 1.9507494919499282, + "learning_rate": 1.7945826689630293e-05, + "loss": 1.0047, + "step": 9825 + }, + { + "epoch": 0.23, + "grad_norm": 12.200381306810893, + "learning_rate": 1.7945363381557594e-05, + "loss": 1.137, + "step": 9826 + }, + { + "epoch": 0.23, + "grad_norm": 1.95933125485008, + "learning_rate": 1.794490002722447e-05, + "loss": 1.0123, + "step": 9827 + }, + { + "epoch": 0.23, + "grad_norm": 2.664171473852175, + "learning_rate": 1.7944436626633625e-05, + "loss": 1.0043, + "step": 9828 + }, + { + "epoch": 0.23, + "grad_norm": 2.173343762467296, + "learning_rate": 1.7943973179787753e-05, + "loss": 1.0466, + "step": 9829 + }, + { + "epoch": 0.23, + "grad_norm": 2.0031690380125178, + "learning_rate": 1.7943509686689556e-05, + "loss": 0.9792, + "step": 9830 + }, + { + "epoch": 0.23, + "grad_norm": 1.8113810449865526, + "learning_rate": 1.7943046147341727e-05, + "loss": 0.9963, + "step": 9831 + }, + { + "epoch": 0.23, + "grad_norm": 1.9135978923091501, + "learning_rate": 1.7942582561746973e-05, + "loss": 1.1841, + "step": 9832 + }, + { + "epoch": 0.23, + "grad_norm": 2.0559066987176546, + "learning_rate": 1.7942118929907987e-05, + "loss": 1.0909, + "step": 9833 + }, + { + "epoch": 0.23, + "grad_norm": 2.216982127212936, + "learning_rate": 1.7941655251827467e-05, + "loss": 1.1499, + "step": 9834 + }, + { + "epoch": 0.23, + "grad_norm": 1.9560597232081498, + "learning_rate": 1.794119152750812e-05, + "loss": 1.1089, + "step": 9835 + }, + { + "epoch": 0.23, + "grad_norm": 1.9289964491136562, + "learning_rate": 1.794072775695264e-05, + "loss": 1.1073, + "step": 9836 + }, + { + "epoch": 0.23, + "grad_norm": 2.256927743395529, + "learning_rate": 1.7940263940163732e-05, + "loss": 1.132, + "step": 9837 + }, + { + "epoch": 0.23, + "grad_norm": 2.111936128342833, + "learning_rate": 1.793980007714409e-05, + "loss": 1.1511, + "step": 9838 + }, + { + "epoch": 0.23, + "grad_norm": 1.8586131976336027, + "learning_rate": 1.7939336167896416e-05, + "loss": 1.1186, + "step": 9839 + }, + { + "epoch": 0.23, + "grad_norm": 2.090503535389274, + "learning_rate": 1.7938872212423417e-05, + "loss": 0.979, + "step": 9840 + }, + { + "epoch": 0.23, + "grad_norm": 1.9787848517482693, + "learning_rate": 1.793840821072779e-05, + "loss": 1.194, + "step": 9841 + }, + { + "epoch": 0.23, + "grad_norm": 2.061789510146145, + "learning_rate": 1.7937944162812238e-05, + "loss": 1.0176, + "step": 9842 + }, + { + "epoch": 0.23, + "grad_norm": 2.1100318549352806, + "learning_rate": 1.7937480068679462e-05, + "loss": 1.073, + "step": 9843 + }, + { + "epoch": 0.23, + "grad_norm": 1.9751799734853994, + "learning_rate": 1.7937015928332165e-05, + "loss": 0.9729, + "step": 9844 + }, + { + "epoch": 0.23, + "grad_norm": 2.135187705331175, + "learning_rate": 1.7936551741773043e-05, + "loss": 1.0398, + "step": 9845 + }, + { + "epoch": 0.23, + "grad_norm": 2.2063263554453005, + "learning_rate": 1.7936087509004812e-05, + "loss": 1.0536, + "step": 9846 + }, + { + "epoch": 0.23, + "grad_norm": 1.8879744148988273, + "learning_rate": 1.793562323003016e-05, + "loss": 0.8746, + "step": 9847 + }, + { + "epoch": 0.23, + "grad_norm": 3.270864337195682, + "learning_rate": 1.79351589048518e-05, + "loss": 1.0335, + "step": 9848 + }, + { + "epoch": 0.23, + "grad_norm": 2.271018650362847, + "learning_rate": 1.7934694533472433e-05, + "loss": 1.0608, + "step": 9849 + }, + { + "epoch": 0.23, + "grad_norm": 2.25935841213547, + "learning_rate": 1.7934230115894765e-05, + "loss": 1.0712, + "step": 9850 + }, + { + "epoch": 0.23, + "grad_norm": 1.8937273404210413, + "learning_rate": 1.7933765652121493e-05, + "loss": 1.0659, + "step": 9851 + }, + { + "epoch": 0.23, + "grad_norm": 2.0905237082148926, + "learning_rate": 1.793330114215533e-05, + "loss": 1.1224, + "step": 9852 + }, + { + "epoch": 0.23, + "grad_norm": 1.8832095910753377, + "learning_rate": 1.7932836585998973e-05, + "loss": 1.1936, + "step": 9853 + }, + { + "epoch": 0.23, + "grad_norm": 2.0178813341821793, + "learning_rate": 1.7932371983655125e-05, + "loss": 1.0331, + "step": 9854 + }, + { + "epoch": 0.23, + "grad_norm": 2.0573445084795803, + "learning_rate": 1.7931907335126505e-05, + "loss": 0.9091, + "step": 9855 + }, + { + "epoch": 0.23, + "grad_norm": 2.7247484040936345, + "learning_rate": 1.7931442640415805e-05, + "loss": 1.0669, + "step": 9856 + }, + { + "epoch": 0.23, + "grad_norm": 3.078940899120873, + "learning_rate": 1.7930977899525736e-05, + "loss": 0.9363, + "step": 9857 + }, + { + "epoch": 0.23, + "grad_norm": 2.0607928190866294, + "learning_rate": 1.7930513112459002e-05, + "loss": 1.0938, + "step": 9858 + }, + { + "epoch": 0.23, + "grad_norm": 2.2760260160403285, + "learning_rate": 1.7930048279218306e-05, + "loss": 1.0025, + "step": 9859 + }, + { + "epoch": 0.23, + "grad_norm": 3.148099762115766, + "learning_rate": 1.7929583399806363e-05, + "loss": 0.96, + "step": 9860 + }, + { + "epoch": 0.23, + "grad_norm": 1.9663221248902856, + "learning_rate": 1.7929118474225876e-05, + "loss": 1.047, + "step": 9861 + }, + { + "epoch": 0.23, + "grad_norm": 2.146785125150878, + "learning_rate": 1.7928653502479546e-05, + "loss": 0.9973, + "step": 9862 + }, + { + "epoch": 0.23, + "grad_norm": 1.8703900660413455, + "learning_rate": 1.792818848457009e-05, + "loss": 1.0399, + "step": 9863 + }, + { + "epoch": 0.23, + "grad_norm": 1.9576581209849568, + "learning_rate": 1.792772342050021e-05, + "loss": 1.1767, + "step": 9864 + }, + { + "epoch": 0.23, + "grad_norm": 1.1922531194082906, + "learning_rate": 1.7927258310272613e-05, + "loss": 1.0168, + "step": 9865 + }, + { + "epoch": 0.23, + "grad_norm": 2.09058871917541, + "learning_rate": 1.7926793153890006e-05, + "loss": 1.0602, + "step": 9866 + }, + { + "epoch": 0.23, + "grad_norm": 1.7710686505244595, + "learning_rate": 1.7926327951355106e-05, + "loss": 1.036, + "step": 9867 + }, + { + "epoch": 0.23, + "grad_norm": 2.282914690918528, + "learning_rate": 1.7925862702670608e-05, + "loss": 1.0889, + "step": 9868 + }, + { + "epoch": 0.23, + "grad_norm": 2.176650535507981, + "learning_rate": 1.7925397407839233e-05, + "loss": 0.9835, + "step": 9869 + }, + { + "epoch": 0.23, + "grad_norm": 1.8733143071079297, + "learning_rate": 1.7924932066863682e-05, + "loss": 1.2024, + "step": 9870 + }, + { + "epoch": 0.23, + "grad_norm": 2.0990732096135054, + "learning_rate": 1.792446667974667e-05, + "loss": 1.1195, + "step": 9871 + }, + { + "epoch": 0.23, + "grad_norm": 1.9473541683471418, + "learning_rate": 1.7924001246490907e-05, + "loss": 1.0616, + "step": 9872 + }, + { + "epoch": 0.23, + "grad_norm": 2.035268794757995, + "learning_rate": 1.7923535767099093e-05, + "loss": 1.0716, + "step": 9873 + }, + { + "epoch": 0.23, + "grad_norm": 1.9674144499058506, + "learning_rate": 1.792307024157395e-05, + "loss": 1.1212, + "step": 9874 + }, + { + "epoch": 0.23, + "grad_norm": 1.9750588935471647, + "learning_rate": 1.792260466991818e-05, + "loss": 1.1462, + "step": 9875 + }, + { + "epoch": 0.23, + "grad_norm": 2.0633440968122367, + "learning_rate": 1.79221390521345e-05, + "loss": 1.0795, + "step": 9876 + }, + { + "epoch": 0.23, + "grad_norm": 2.2418831803990407, + "learning_rate": 1.792167338822562e-05, + "loss": 1.0247, + "step": 9877 + }, + { + "epoch": 0.23, + "grad_norm": 2.244290957715122, + "learning_rate": 1.792120767819425e-05, + "loss": 1.1795, + "step": 9878 + }, + { + "epoch": 0.23, + "grad_norm": 2.4333098519208973, + "learning_rate": 1.7920741922043093e-05, + "loss": 1.0261, + "step": 9879 + }, + { + "epoch": 0.23, + "grad_norm": 2.0775439483738976, + "learning_rate": 1.7920276119774876e-05, + "loss": 0.9522, + "step": 9880 + }, + { + "epoch": 0.23, + "grad_norm": 2.0236082755196265, + "learning_rate": 1.7919810271392305e-05, + "loss": 1.1096, + "step": 9881 + }, + { + "epoch": 0.23, + "grad_norm": 1.2491946194434247, + "learning_rate": 1.7919344376898088e-05, + "loss": 1.0696, + "step": 9882 + }, + { + "epoch": 0.23, + "grad_norm": 2.0920476989025, + "learning_rate": 1.791887843629494e-05, + "loss": 1.022, + "step": 9883 + }, + { + "epoch": 0.23, + "grad_norm": 2.136342170192761, + "learning_rate": 1.7918412449585577e-05, + "loss": 1.0308, + "step": 9884 + }, + { + "epoch": 0.23, + "grad_norm": 2.1270728058986292, + "learning_rate": 1.7917946416772708e-05, + "loss": 1.1506, + "step": 9885 + }, + { + "epoch": 0.23, + "grad_norm": 2.2068080127746086, + "learning_rate": 1.791748033785905e-05, + "loss": 1.106, + "step": 9886 + }, + { + "epoch": 0.23, + "grad_norm": 2.0559593289909266, + "learning_rate": 1.7917014212847314e-05, + "loss": 0.9288, + "step": 9887 + }, + { + "epoch": 0.23, + "grad_norm": 1.1868861993937494, + "learning_rate": 1.7916548041740213e-05, + "loss": 1.0135, + "step": 9888 + }, + { + "epoch": 0.23, + "grad_norm": 2.3262485960728623, + "learning_rate": 1.7916081824540466e-05, + "loss": 0.9743, + "step": 9889 + }, + { + "epoch": 0.23, + "grad_norm": 2.2479953663647874, + "learning_rate": 1.7915615561250783e-05, + "loss": 1.1565, + "step": 9890 + }, + { + "epoch": 0.23, + "grad_norm": 2.125493226526468, + "learning_rate": 1.791514925187388e-05, + "loss": 1.0272, + "step": 9891 + }, + { + "epoch": 0.23, + "grad_norm": 2.1836435508721594, + "learning_rate": 1.791468289641247e-05, + "loss": 1.003, + "step": 9892 + }, + { + "epoch": 0.23, + "grad_norm": 2.1003828346219, + "learning_rate": 1.7914216494869278e-05, + "loss": 0.9782, + "step": 9893 + }, + { + "epoch": 0.23, + "grad_norm": 1.9947407276741642, + "learning_rate": 1.7913750047247004e-05, + "loss": 1.0873, + "step": 9894 + }, + { + "epoch": 0.23, + "grad_norm": 2.1670755815769405, + "learning_rate": 1.7913283553548374e-05, + "loss": 1.1367, + "step": 9895 + }, + { + "epoch": 0.23, + "grad_norm": 1.1095914037781833, + "learning_rate": 1.7912817013776102e-05, + "loss": 0.9968, + "step": 9896 + }, + { + "epoch": 0.23, + "grad_norm": 2.3769696215689518, + "learning_rate": 1.7912350427932905e-05, + "loss": 1.1578, + "step": 9897 + }, + { + "epoch": 0.23, + "grad_norm": 2.1805369493410915, + "learning_rate": 1.79118837960215e-05, + "loss": 0.9559, + "step": 9898 + }, + { + "epoch": 0.23, + "grad_norm": 2.3615708362468935, + "learning_rate": 1.79114171180446e-05, + "loss": 0.9748, + "step": 9899 + }, + { + "epoch": 0.23, + "grad_norm": 1.9861520916140663, + "learning_rate": 1.7910950394004926e-05, + "loss": 1.0849, + "step": 9900 + }, + { + "epoch": 0.23, + "grad_norm": 1.186203053276094, + "learning_rate": 1.7910483623905195e-05, + "loss": 0.9377, + "step": 9901 + }, + { + "epoch": 0.23, + "grad_norm": 2.987276588425011, + "learning_rate": 1.791001680774812e-05, + "loss": 1.1183, + "step": 9902 + }, + { + "epoch": 0.23, + "grad_norm": 1.9736253999215083, + "learning_rate": 1.7909549945536427e-05, + "loss": 1.0919, + "step": 9903 + }, + { + "epoch": 0.23, + "grad_norm": 1.985707527817092, + "learning_rate": 1.790908303727283e-05, + "loss": 1.0639, + "step": 9904 + }, + { + "epoch": 0.23, + "grad_norm": 1.9435797097988705, + "learning_rate": 1.7908616082960047e-05, + "loss": 1.1034, + "step": 9905 + }, + { + "epoch": 0.23, + "grad_norm": 2.1839685116869627, + "learning_rate": 1.7908149082600796e-05, + "loss": 0.927, + "step": 9906 + }, + { + "epoch": 0.23, + "grad_norm": 2.2161239542700115, + "learning_rate": 1.79076820361978e-05, + "loss": 1.218, + "step": 9907 + }, + { + "epoch": 0.23, + "grad_norm": 2.76184648797734, + "learning_rate": 1.790721494375377e-05, + "loss": 1.1129, + "step": 9908 + }, + { + "epoch": 0.23, + "grad_norm": 2.0729311251168534, + "learning_rate": 1.790674780527144e-05, + "loss": 1.0857, + "step": 9909 + }, + { + "epoch": 0.23, + "grad_norm": 1.8251550968764725, + "learning_rate": 1.7906280620753517e-05, + "loss": 1.0257, + "step": 9910 + }, + { + "epoch": 0.23, + "grad_norm": 2.414463651163108, + "learning_rate": 1.790581339020272e-05, + "loss": 0.9087, + "step": 9911 + }, + { + "epoch": 0.23, + "grad_norm": 1.9556670353509804, + "learning_rate": 1.790534611362178e-05, + "loss": 1.136, + "step": 9912 + }, + { + "epoch": 0.23, + "grad_norm": 2.337418281133477, + "learning_rate": 1.7904878791013412e-05, + "loss": 0.9556, + "step": 9913 + }, + { + "epoch": 0.23, + "grad_norm": 2.122279686348468, + "learning_rate": 1.7904411422380336e-05, + "loss": 1.0429, + "step": 9914 + }, + { + "epoch": 0.23, + "grad_norm": 2.0574619185293677, + "learning_rate": 1.7903944007725273e-05, + "loss": 1.0308, + "step": 9915 + }, + { + "epoch": 0.23, + "grad_norm": 1.9841505741202154, + "learning_rate": 1.790347654705095e-05, + "loss": 0.9988, + "step": 9916 + }, + { + "epoch": 0.23, + "grad_norm": 1.1760801572205504, + "learning_rate": 1.790300904036008e-05, + "loss": 1.016, + "step": 9917 + }, + { + "epoch": 0.23, + "grad_norm": 2.3027957555405667, + "learning_rate": 1.790254148765539e-05, + "loss": 1.1182, + "step": 9918 + }, + { + "epoch": 0.23, + "grad_norm": 2.2105448873616793, + "learning_rate": 1.7902073888939603e-05, + "loss": 1.1495, + "step": 9919 + }, + { + "epoch": 0.23, + "grad_norm": 2.5495374549390557, + "learning_rate": 1.7901606244215436e-05, + "loss": 0.9693, + "step": 9920 + }, + { + "epoch": 0.23, + "grad_norm": 2.3130142668669005, + "learning_rate": 1.7901138553485623e-05, + "loss": 1.0772, + "step": 9921 + }, + { + "epoch": 0.23, + "grad_norm": 1.8791081404077699, + "learning_rate": 1.7900670816752875e-05, + "loss": 1.0912, + "step": 9922 + }, + { + "epoch": 0.23, + "grad_norm": 4.123425226795879, + "learning_rate": 1.790020303401992e-05, + "loss": 0.9293, + "step": 9923 + }, + { + "epoch": 0.23, + "grad_norm": 1.8695724159467682, + "learning_rate": 1.7899735205289483e-05, + "loss": 1.1084, + "step": 9924 + }, + { + "epoch": 0.23, + "grad_norm": 1.216801526832261, + "learning_rate": 1.7899267330564287e-05, + "loss": 0.9244, + "step": 9925 + }, + { + "epoch": 0.23, + "grad_norm": 2.248028051246163, + "learning_rate": 1.7898799409847055e-05, + "loss": 1.099, + "step": 9926 + }, + { + "epoch": 0.23, + "grad_norm": 2.2072705304776155, + "learning_rate": 1.7898331443140512e-05, + "loss": 1.0823, + "step": 9927 + }, + { + "epoch": 0.23, + "grad_norm": 1.9636807814432717, + "learning_rate": 1.7897863430447384e-05, + "loss": 1.1239, + "step": 9928 + }, + { + "epoch": 0.23, + "grad_norm": 2.4497044249443376, + "learning_rate": 1.7897395371770393e-05, + "loss": 1.0278, + "step": 9929 + }, + { + "epoch": 0.23, + "grad_norm": 2.223267758140374, + "learning_rate": 1.789692726711227e-05, + "loss": 1.0922, + "step": 9930 + }, + { + "epoch": 0.23, + "grad_norm": 2.009488239437394, + "learning_rate": 1.789645911647573e-05, + "loss": 1.0761, + "step": 9931 + }, + { + "epoch": 0.23, + "grad_norm": 2.0627711263050195, + "learning_rate": 1.789599091986351e-05, + "loss": 1.2194, + "step": 9932 + }, + { + "epoch": 0.23, + "grad_norm": 2.0568999822489524, + "learning_rate": 1.7895522677278327e-05, + "loss": 0.9483, + "step": 9933 + }, + { + "epoch": 0.23, + "grad_norm": 2.1508143882947346, + "learning_rate": 1.7895054388722913e-05, + "loss": 0.9714, + "step": 9934 + }, + { + "epoch": 0.23, + "grad_norm": 1.8447832189212956, + "learning_rate": 1.7894586054199993e-05, + "loss": 1.0982, + "step": 9935 + }, + { + "epoch": 0.23, + "grad_norm": 2.0001908154004497, + "learning_rate": 1.7894117673712295e-05, + "loss": 1.1199, + "step": 9936 + }, + { + "epoch": 0.23, + "grad_norm": 2.141154825162036, + "learning_rate": 1.7893649247262543e-05, + "loss": 1.1275, + "step": 9937 + }, + { + "epoch": 0.23, + "grad_norm": 2.1271714401910304, + "learning_rate": 1.7893180774853467e-05, + "loss": 1.1826, + "step": 9938 + }, + { + "epoch": 0.23, + "grad_norm": 1.948314362099701, + "learning_rate": 1.7892712256487792e-05, + "loss": 1.0783, + "step": 9939 + }, + { + "epoch": 0.23, + "grad_norm": 2.1769790476019337, + "learning_rate": 1.789224369216825e-05, + "loss": 1.03, + "step": 9940 + }, + { + "epoch": 0.23, + "grad_norm": 1.0690207956531788, + "learning_rate": 1.7891775081897564e-05, + "loss": 0.9376, + "step": 9941 + }, + { + "epoch": 0.23, + "grad_norm": 2.095971187661656, + "learning_rate": 1.7891306425678467e-05, + "loss": 1.076, + "step": 9942 + }, + { + "epoch": 0.23, + "grad_norm": 2.2449467122622444, + "learning_rate": 1.7890837723513686e-05, + "loss": 1.0933, + "step": 9943 + }, + { + "epoch": 0.23, + "grad_norm": 2.245212555706157, + "learning_rate": 1.7890368975405947e-05, + "loss": 1.0372, + "step": 9944 + }, + { + "epoch": 0.23, + "grad_norm": 1.9509568427063377, + "learning_rate": 1.7889900181357983e-05, + "loss": 1.0056, + "step": 9945 + }, + { + "epoch": 0.23, + "grad_norm": 2.005234729824814, + "learning_rate": 1.7889431341372523e-05, + "loss": 0.9978, + "step": 9946 + }, + { + "epoch": 0.23, + "grad_norm": 2.103820092158699, + "learning_rate": 1.7888962455452296e-05, + "loss": 1.0175, + "step": 9947 + }, + { + "epoch": 0.23, + "grad_norm": 1.9643676694850947, + "learning_rate": 1.788849352360003e-05, + "loss": 1.1372, + "step": 9948 + }, + { + "epoch": 0.23, + "grad_norm": 2.130898544889449, + "learning_rate": 1.7888024545818458e-05, + "loss": 1.1069, + "step": 9949 + }, + { + "epoch": 0.23, + "grad_norm": 1.9917035846283597, + "learning_rate": 1.788755552211031e-05, + "loss": 1.0648, + "step": 9950 + }, + { + "epoch": 0.23, + "grad_norm": 1.1232913754551128, + "learning_rate": 1.7887086452478318e-05, + "loss": 1.0078, + "step": 9951 + }, + { + "epoch": 0.23, + "grad_norm": 1.888257283690763, + "learning_rate": 1.788661733692521e-05, + "loss": 1.0743, + "step": 9952 + }, + { + "epoch": 0.23, + "grad_norm": 2.243811158006545, + "learning_rate": 1.788614817545372e-05, + "loss": 1.1568, + "step": 9953 + }, + { + "epoch": 0.23, + "grad_norm": 1.9728417505254996, + "learning_rate": 1.7885678968066576e-05, + "loss": 1.1214, + "step": 9954 + }, + { + "epoch": 0.23, + "grad_norm": 2.0626470337733287, + "learning_rate": 1.7885209714766517e-05, + "loss": 0.9787, + "step": 9955 + }, + { + "epoch": 0.23, + "grad_norm": 1.0659598068400231, + "learning_rate": 1.788474041555627e-05, + "loss": 0.9681, + "step": 9956 + }, + { + "epoch": 0.23, + "grad_norm": 2.023968603534728, + "learning_rate": 1.7884271070438563e-05, + "loss": 0.9664, + "step": 9957 + }, + { + "epoch": 0.23, + "grad_norm": 2.1587232741821074, + "learning_rate": 1.7883801679416138e-05, + "loss": 1.0963, + "step": 9958 + }, + { + "epoch": 0.23, + "grad_norm": 1.9771975570310902, + "learning_rate": 1.7883332242491726e-05, + "loss": 1.0276, + "step": 9959 + }, + { + "epoch": 0.23, + "grad_norm": 2.068809660971635, + "learning_rate": 1.7882862759668052e-05, + "loss": 1.0463, + "step": 9960 + }, + { + "epoch": 0.23, + "grad_norm": 2.451058729569575, + "learning_rate": 1.7882393230947857e-05, + "loss": 1.0485, + "step": 9961 + }, + { + "epoch": 0.23, + "grad_norm": 2.038265947915572, + "learning_rate": 1.7881923656333876e-05, + "loss": 0.9372, + "step": 9962 + }, + { + "epoch": 0.23, + "grad_norm": 2.3776938255840427, + "learning_rate": 1.788145403582884e-05, + "loss": 0.8848, + "step": 9963 + }, + { + "epoch": 0.23, + "grad_norm": 2.297725003940382, + "learning_rate": 1.788098436943548e-05, + "loss": 1.1213, + "step": 9964 + }, + { + "epoch": 0.23, + "grad_norm": 1.8095932811472364, + "learning_rate": 1.7880514657156534e-05, + "loss": 1.0248, + "step": 9965 + }, + { + "epoch": 0.23, + "grad_norm": 1.9814968010786853, + "learning_rate": 1.788004489899474e-05, + "loss": 0.939, + "step": 9966 + }, + { + "epoch": 0.23, + "grad_norm": 2.7388296080774155, + "learning_rate": 1.7879575094952826e-05, + "loss": 1.1819, + "step": 9967 + }, + { + "epoch": 0.23, + "grad_norm": 1.13265652055424, + "learning_rate": 1.7879105245033533e-05, + "loss": 1.0384, + "step": 9968 + }, + { + "epoch": 0.23, + "grad_norm": 2.2095838885470642, + "learning_rate": 1.7878635349239593e-05, + "loss": 1.1172, + "step": 9969 + }, + { + "epoch": 0.23, + "grad_norm": 1.97105123635972, + "learning_rate": 1.7878165407573747e-05, + "loss": 0.9745, + "step": 9970 + }, + { + "epoch": 0.23, + "grad_norm": 2.1139627496899576, + "learning_rate": 1.7877695420038725e-05, + "loss": 1.0774, + "step": 9971 + }, + { + "epoch": 0.23, + "grad_norm": 2.2017640986391322, + "learning_rate": 1.7877225386637266e-05, + "loss": 0.956, + "step": 9972 + }, + { + "epoch": 0.23, + "grad_norm": 2.1478760607766114, + "learning_rate": 1.787675530737211e-05, + "loss": 1.0928, + "step": 9973 + }, + { + "epoch": 0.23, + "grad_norm": 1.7744385018762474, + "learning_rate": 1.7876285182245992e-05, + "loss": 1.0099, + "step": 9974 + }, + { + "epoch": 0.24, + "grad_norm": 1.9664276217627075, + "learning_rate": 1.7875815011261645e-05, + "loss": 1.0487, + "step": 9975 + }, + { + "epoch": 0.24, + "grad_norm": 2.1565851552736053, + "learning_rate": 1.7875344794421808e-05, + "loss": 0.9471, + "step": 9976 + }, + { + "epoch": 0.24, + "grad_norm": 2.1445582835216768, + "learning_rate": 1.7874874531729225e-05, + "loss": 0.9932, + "step": 9977 + }, + { + "epoch": 0.24, + "grad_norm": 2.0530866229730558, + "learning_rate": 1.7874404223186626e-05, + "loss": 1.0018, + "step": 9978 + }, + { + "epoch": 0.24, + "grad_norm": 2.5071069545272944, + "learning_rate": 1.7873933868796753e-05, + "loss": 0.9976, + "step": 9979 + }, + { + "epoch": 0.24, + "grad_norm": 2.0320205829267786, + "learning_rate": 1.7873463468562343e-05, + "loss": 1.1425, + "step": 9980 + }, + { + "epoch": 0.24, + "grad_norm": 2.187778277450633, + "learning_rate": 1.7872993022486137e-05, + "loss": 1.1629, + "step": 9981 + }, + { + "epoch": 0.24, + "grad_norm": 2.0305510757020575, + "learning_rate": 1.7872522530570876e-05, + "loss": 1.0403, + "step": 9982 + }, + { + "epoch": 0.24, + "grad_norm": 1.9782883988695035, + "learning_rate": 1.7872051992819295e-05, + "loss": 1.0252, + "step": 9983 + }, + { + "epoch": 0.24, + "grad_norm": 2.3076861979012713, + "learning_rate": 1.7871581409234134e-05, + "loss": 0.9978, + "step": 9984 + }, + { + "epoch": 0.24, + "grad_norm": 3.163930487380726, + "learning_rate": 1.787111077981813e-05, + "loss": 0.9348, + "step": 9985 + }, + { + "epoch": 0.24, + "grad_norm": 2.036944924404781, + "learning_rate": 1.787064010457403e-05, + "loss": 1.0444, + "step": 9986 + }, + { + "epoch": 0.24, + "grad_norm": 1.9966312263848343, + "learning_rate": 1.7870169383504574e-05, + "loss": 0.9725, + "step": 9987 + }, + { + "epoch": 0.24, + "grad_norm": 1.8852176930778792, + "learning_rate": 1.78696986166125e-05, + "loss": 1.0225, + "step": 9988 + }, + { + "epoch": 0.24, + "grad_norm": 1.9226463639690303, + "learning_rate": 1.7869227803900548e-05, + "loss": 1.1396, + "step": 9989 + }, + { + "epoch": 0.24, + "grad_norm": 1.866303686447702, + "learning_rate": 1.786875694537146e-05, + "loss": 1.0015, + "step": 9990 + }, + { + "epoch": 0.24, + "grad_norm": 1.8677186815159246, + "learning_rate": 1.7868286041027977e-05, + "loss": 0.9346, + "step": 9991 + }, + { + "epoch": 0.24, + "grad_norm": 1.8043409968046304, + "learning_rate": 1.7867815090872845e-05, + "loss": 0.9824, + "step": 9992 + }, + { + "epoch": 0.24, + "grad_norm": 2.011275201516736, + "learning_rate": 1.78673440949088e-05, + "loss": 0.9884, + "step": 9993 + }, + { + "epoch": 0.24, + "grad_norm": 2.437640391927626, + "learning_rate": 1.7866873053138588e-05, + "loss": 1.0555, + "step": 9994 + }, + { + "epoch": 0.24, + "grad_norm": 3.028134881674313, + "learning_rate": 1.786640196556495e-05, + "loss": 1.1072, + "step": 9995 + }, + { + "epoch": 0.24, + "grad_norm": 1.1069409956388014, + "learning_rate": 1.786593083219063e-05, + "loss": 0.9847, + "step": 9996 + }, + { + "epoch": 0.24, + "grad_norm": 2.115598692014627, + "learning_rate": 1.7865459653018368e-05, + "loss": 1.0764, + "step": 9997 + }, + { + "epoch": 0.24, + "grad_norm": 2.1353250024353243, + "learning_rate": 1.7864988428050912e-05, + "loss": 0.9485, + "step": 9998 + }, + { + "epoch": 0.24, + "grad_norm": 2.366763302553441, + "learning_rate": 1.7864517157291e-05, + "loss": 0.972, + "step": 9999 + }, + { + "epoch": 0.24, + "grad_norm": 1.9551085436157114, + "learning_rate": 1.7864045840741383e-05, + "loss": 1.0674, + "step": 10000 + }, + { + "epoch": 0.24, + "grad_norm": 1.0864895945463615, + "learning_rate": 1.78635744784048e-05, + "loss": 1.0027, + "step": 10001 + }, + { + "epoch": 0.24, + "grad_norm": 1.9429368907110482, + "learning_rate": 1.7863103070283995e-05, + "loss": 1.0517, + "step": 10002 + }, + { + "epoch": 0.24, + "grad_norm": 2.462145674498214, + "learning_rate": 1.7862631616381717e-05, + "loss": 1.0236, + "step": 10003 + }, + { + "epoch": 0.24, + "grad_norm": 1.9588604131463525, + "learning_rate": 1.786216011670071e-05, + "loss": 0.9176, + "step": 10004 + }, + { + "epoch": 0.24, + "grad_norm": 2.4787666931626635, + "learning_rate": 1.7861688571243715e-05, + "loss": 1.0688, + "step": 10005 + }, + { + "epoch": 0.24, + "grad_norm": 2.2835836240313587, + "learning_rate": 1.786121698001348e-05, + "loss": 0.9735, + "step": 10006 + }, + { + "epoch": 0.24, + "grad_norm": 1.1082842750195765, + "learning_rate": 1.786074534301275e-05, + "loss": 1.0344, + "step": 10007 + }, + { + "epoch": 0.24, + "grad_norm": 2.9614196644803976, + "learning_rate": 1.7860273660244275e-05, + "loss": 1.0297, + "step": 10008 + }, + { + "epoch": 0.24, + "grad_norm": 1.9058732928744027, + "learning_rate": 1.7859801931710798e-05, + "loss": 1.0263, + "step": 10009 + }, + { + "epoch": 0.24, + "grad_norm": 1.7981123871686397, + "learning_rate": 1.7859330157415065e-05, + "loss": 0.9886, + "step": 10010 + }, + { + "epoch": 0.24, + "grad_norm": 2.3155834274865277, + "learning_rate": 1.785885833735982e-05, + "loss": 1.0219, + "step": 10011 + }, + { + "epoch": 0.24, + "grad_norm": 2.200074996983524, + "learning_rate": 1.785838647154782e-05, + "loss": 1.1165, + "step": 10012 + }, + { + "epoch": 0.24, + "grad_norm": 2.79517953596028, + "learning_rate": 1.7857914559981805e-05, + "loss": 1.1895, + "step": 10013 + }, + { + "epoch": 0.24, + "grad_norm": 2.2677850560053376, + "learning_rate": 1.785744260266452e-05, + "loss": 1.0488, + "step": 10014 + }, + { + "epoch": 0.24, + "grad_norm": 2.098740350613825, + "learning_rate": 1.7856970599598716e-05, + "loss": 0.9987, + "step": 10015 + }, + { + "epoch": 0.24, + "grad_norm": 2.1495322617870687, + "learning_rate": 1.7856498550787144e-05, + "loss": 1.0773, + "step": 10016 + }, + { + "epoch": 0.24, + "grad_norm": 2.208342599563494, + "learning_rate": 1.7856026456232553e-05, + "loss": 1.0596, + "step": 10017 + }, + { + "epoch": 0.24, + "grad_norm": 2.0744769608369626, + "learning_rate": 1.7855554315937685e-05, + "loss": 1.0069, + "step": 10018 + }, + { + "epoch": 0.24, + "grad_norm": 2.2080341033433233, + "learning_rate": 1.785508212990529e-05, + "loss": 1.1583, + "step": 10019 + }, + { + "epoch": 0.24, + "grad_norm": 2.6375149951583126, + "learning_rate": 1.7854609898138126e-05, + "loss": 1.1987, + "step": 10020 + }, + { + "epoch": 0.24, + "grad_norm": 1.8032135068411859, + "learning_rate": 1.7854137620638933e-05, + "loss": 1.1957, + "step": 10021 + }, + { + "epoch": 0.24, + "grad_norm": 2.36414778700385, + "learning_rate": 1.7853665297410465e-05, + "loss": 1.0997, + "step": 10022 + }, + { + "epoch": 0.24, + "grad_norm": 2.4806100165416227, + "learning_rate": 1.7853192928455475e-05, + "loss": 1.0496, + "step": 10023 + }, + { + "epoch": 0.24, + "grad_norm": 1.9331385905821645, + "learning_rate": 1.78527205137767e-05, + "loss": 1.0494, + "step": 10024 + }, + { + "epoch": 0.24, + "grad_norm": 2.124681651083977, + "learning_rate": 1.785224805337691e-05, + "loss": 0.983, + "step": 10025 + }, + { + "epoch": 0.24, + "grad_norm": 2.306697787164841, + "learning_rate": 1.785177554725884e-05, + "loss": 0.9803, + "step": 10026 + }, + { + "epoch": 0.24, + "grad_norm": 2.0991653022671803, + "learning_rate": 1.7851302995425247e-05, + "loss": 1.0737, + "step": 10027 + }, + { + "epoch": 0.24, + "grad_norm": 2.1283245245559375, + "learning_rate": 1.7850830397878888e-05, + "loss": 1.0915, + "step": 10028 + }, + { + "epoch": 0.24, + "grad_norm": 1.0844804409482809, + "learning_rate": 1.7850357754622503e-05, + "loss": 0.989, + "step": 10029 + }, + { + "epoch": 0.24, + "grad_norm": 2.158703247439928, + "learning_rate": 1.7849885065658852e-05, + "loss": 1.0813, + "step": 10030 + }, + { + "epoch": 0.24, + "grad_norm": 1.1135346272810402, + "learning_rate": 1.7849412330990686e-05, + "loss": 1.0055, + "step": 10031 + }, + { + "epoch": 0.24, + "grad_norm": 2.1836487358187107, + "learning_rate": 1.7848939550620755e-05, + "loss": 0.8639, + "step": 10032 + }, + { + "epoch": 0.24, + "grad_norm": 2.0269081786159124, + "learning_rate": 1.784846672455181e-05, + "loss": 1.0546, + "step": 10033 + }, + { + "epoch": 0.24, + "grad_norm": 1.94635529412613, + "learning_rate": 1.7847993852786612e-05, + "loss": 1.152, + "step": 10034 + }, + { + "epoch": 0.24, + "grad_norm": 2.18856233054671, + "learning_rate": 1.7847520935327905e-05, + "loss": 1.0967, + "step": 10035 + }, + { + "epoch": 0.24, + "grad_norm": 2.4383111888301197, + "learning_rate": 1.784704797217845e-05, + "loss": 1.0076, + "step": 10036 + }, + { + "epoch": 0.24, + "grad_norm": 2.2049979596246994, + "learning_rate": 1.7846574963340995e-05, + "loss": 1.0265, + "step": 10037 + }, + { + "epoch": 0.24, + "grad_norm": 2.2808954115590128, + "learning_rate": 1.78461019088183e-05, + "loss": 1.0932, + "step": 10038 + }, + { + "epoch": 0.24, + "grad_norm": 2.060044610472657, + "learning_rate": 1.7845628808613112e-05, + "loss": 0.9936, + "step": 10039 + }, + { + "epoch": 0.24, + "grad_norm": 1.8904788278740594, + "learning_rate": 1.784515566272819e-05, + "loss": 1.042, + "step": 10040 + }, + { + "epoch": 0.24, + "grad_norm": 2.0747476743459594, + "learning_rate": 1.7844682471166288e-05, + "loss": 1.0917, + "step": 10041 + }, + { + "epoch": 0.24, + "grad_norm": 1.901360914766636, + "learning_rate": 1.784420923393016e-05, + "loss": 1.0004, + "step": 10042 + }, + { + "epoch": 0.24, + "grad_norm": 2.036722487086623, + "learning_rate": 1.7843735951022565e-05, + "loss": 1.0971, + "step": 10043 + }, + { + "epoch": 0.24, + "grad_norm": 1.8822255330742506, + "learning_rate": 1.7843262622446253e-05, + "loss": 1.0221, + "step": 10044 + }, + { + "epoch": 0.24, + "grad_norm": 2.0700290537068637, + "learning_rate": 1.7842789248203985e-05, + "loss": 1.145, + "step": 10045 + }, + { + "epoch": 0.24, + "grad_norm": 1.7386365891636089, + "learning_rate": 1.7842315828298514e-05, + "loss": 0.8566, + "step": 10046 + }, + { + "epoch": 0.24, + "grad_norm": 3.046905119759288, + "learning_rate": 1.78418423627326e-05, + "loss": 1.0241, + "step": 10047 + }, + { + "epoch": 0.24, + "grad_norm": 2.3718566104312044, + "learning_rate": 1.7841368851508995e-05, + "loss": 1.0485, + "step": 10048 + }, + { + "epoch": 0.24, + "grad_norm": 2.0269571889115947, + "learning_rate": 1.7840895294630457e-05, + "loss": 1.024, + "step": 10049 + }, + { + "epoch": 0.24, + "grad_norm": 2.0685627770213935, + "learning_rate": 1.7840421692099746e-05, + "loss": 0.978, + "step": 10050 + }, + { + "epoch": 0.24, + "grad_norm": 2.178816882540846, + "learning_rate": 1.7839948043919617e-05, + "loss": 1.0049, + "step": 10051 + }, + { + "epoch": 0.24, + "grad_norm": 2.018132160019157, + "learning_rate": 1.783947435009283e-05, + "loss": 1.0098, + "step": 10052 + }, + { + "epoch": 0.24, + "grad_norm": 1.966940929119767, + "learning_rate": 1.7839000610622138e-05, + "loss": 1.0681, + "step": 10053 + }, + { + "epoch": 0.24, + "grad_norm": 2.0597474018454096, + "learning_rate": 1.7838526825510304e-05, + "loss": 1.077, + "step": 10054 + }, + { + "epoch": 0.24, + "grad_norm": 3.769039174930597, + "learning_rate": 1.7838052994760087e-05, + "loss": 0.9153, + "step": 10055 + }, + { + "epoch": 0.24, + "grad_norm": 2.104034795974786, + "learning_rate": 1.783757911837424e-05, + "loss": 1.0233, + "step": 10056 + }, + { + "epoch": 0.24, + "grad_norm": 2.237611045690375, + "learning_rate": 1.783710519635553e-05, + "loss": 0.9595, + "step": 10057 + }, + { + "epoch": 0.24, + "grad_norm": 2.116222978539179, + "learning_rate": 1.783663122870671e-05, + "loss": 1.0001, + "step": 10058 + }, + { + "epoch": 0.24, + "grad_norm": 2.2482330297390125, + "learning_rate": 1.7836157215430544e-05, + "loss": 1.0733, + "step": 10059 + }, + { + "epoch": 0.24, + "grad_norm": 2.3063490949462224, + "learning_rate": 1.7835683156529787e-05, + "loss": 0.9366, + "step": 10060 + }, + { + "epoch": 0.24, + "grad_norm": 2.0779826131823707, + "learning_rate": 1.7835209052007206e-05, + "loss": 1.2618, + "step": 10061 + }, + { + "epoch": 0.24, + "grad_norm": 2.290966959348625, + "learning_rate": 1.7834734901865554e-05, + "loss": 1.1444, + "step": 10062 + }, + { + "epoch": 0.24, + "grad_norm": 2.2681915456045774, + "learning_rate": 1.7834260706107597e-05, + "loss": 0.9815, + "step": 10063 + }, + { + "epoch": 0.24, + "grad_norm": 2.381261814041505, + "learning_rate": 1.7833786464736092e-05, + "loss": 1.0572, + "step": 10064 + }, + { + "epoch": 0.24, + "grad_norm": 1.9585358882656823, + "learning_rate": 1.7833312177753805e-05, + "loss": 1.0684, + "step": 10065 + }, + { + "epoch": 0.24, + "grad_norm": 1.981889591612005, + "learning_rate": 1.7832837845163492e-05, + "loss": 1.1179, + "step": 10066 + }, + { + "epoch": 0.24, + "grad_norm": 2.3066923654908167, + "learning_rate": 1.7832363466967913e-05, + "loss": 1.0859, + "step": 10067 + }, + { + "epoch": 0.24, + "grad_norm": 2.232469243659278, + "learning_rate": 1.783188904316984e-05, + "loss": 1.0519, + "step": 10068 + }, + { + "epoch": 0.24, + "grad_norm": 1.9465144181741592, + "learning_rate": 1.783141457377203e-05, + "loss": 1.0329, + "step": 10069 + }, + { + "epoch": 0.24, + "grad_norm": 1.820871563019268, + "learning_rate": 1.7830940058777244e-05, + "loss": 1.1242, + "step": 10070 + }, + { + "epoch": 0.24, + "grad_norm": 2.090974373667089, + "learning_rate": 1.7830465498188244e-05, + "loss": 0.9819, + "step": 10071 + }, + { + "epoch": 0.24, + "grad_norm": 2.1824567540621236, + "learning_rate": 1.7829990892007794e-05, + "loss": 0.9939, + "step": 10072 + }, + { + "epoch": 0.24, + "grad_norm": 2.0389449175415453, + "learning_rate": 1.7829516240238662e-05, + "loss": 1.1068, + "step": 10073 + }, + { + "epoch": 0.24, + "grad_norm": 1.993929734133221, + "learning_rate": 1.7829041542883606e-05, + "loss": 1.1122, + "step": 10074 + }, + { + "epoch": 0.24, + "grad_norm": 1.9436131417101146, + "learning_rate": 1.7828566799945394e-05, + "loss": 1.0716, + "step": 10075 + }, + { + "epoch": 0.24, + "grad_norm": 2.0984591736858342, + "learning_rate": 1.782809201142678e-05, + "loss": 1.1179, + "step": 10076 + }, + { + "epoch": 0.24, + "grad_norm": 2.4270922197279345, + "learning_rate": 1.7827617177330544e-05, + "loss": 0.9626, + "step": 10077 + }, + { + "epoch": 0.24, + "grad_norm": 2.013010577812736, + "learning_rate": 1.7827142297659435e-05, + "loss": 0.9826, + "step": 10078 + }, + { + "epoch": 0.24, + "grad_norm": 2.888495770993541, + "learning_rate": 1.7826667372416233e-05, + "loss": 1.0648, + "step": 10079 + }, + { + "epoch": 0.24, + "grad_norm": 1.8390940170126615, + "learning_rate": 1.782619240160369e-05, + "loss": 1.0493, + "step": 10080 + }, + { + "epoch": 0.24, + "grad_norm": 2.2790877640396623, + "learning_rate": 1.782571738522458e-05, + "loss": 0.9752, + "step": 10081 + }, + { + "epoch": 0.24, + "grad_norm": 2.1758928104483877, + "learning_rate": 1.7825242323281666e-05, + "loss": 1.1037, + "step": 10082 + }, + { + "epoch": 0.24, + "grad_norm": 2.1806389310296606, + "learning_rate": 1.7824767215777715e-05, + "loss": 1.0019, + "step": 10083 + }, + { + "epoch": 0.24, + "grad_norm": 2.0415430358324613, + "learning_rate": 1.7824292062715488e-05, + "loss": 1.0721, + "step": 10084 + }, + { + "epoch": 0.24, + "grad_norm": 1.7503291624364892, + "learning_rate": 1.7823816864097758e-05, + "loss": 0.8221, + "step": 10085 + }, + { + "epoch": 0.24, + "grad_norm": 2.0895001798600887, + "learning_rate": 1.7823341619927288e-05, + "loss": 0.9235, + "step": 10086 + }, + { + "epoch": 0.24, + "grad_norm": 1.8121668517404068, + "learning_rate": 1.782286633020685e-05, + "loss": 1.0005, + "step": 10087 + }, + { + "epoch": 0.24, + "grad_norm": 3.2291410951473525, + "learning_rate": 1.7822390994939205e-05, + "loss": 1.042, + "step": 10088 + }, + { + "epoch": 0.24, + "grad_norm": 2.1849029637966986, + "learning_rate": 1.782191561412712e-05, + "loss": 1.0251, + "step": 10089 + }, + { + "epoch": 0.24, + "grad_norm": 1.967979841434825, + "learning_rate": 1.782144018777337e-05, + "loss": 1.0915, + "step": 10090 + }, + { + "epoch": 0.24, + "grad_norm": 2.0599920814086223, + "learning_rate": 1.7820964715880716e-05, + "loss": 1.1461, + "step": 10091 + }, + { + "epoch": 0.24, + "grad_norm": 1.9249916243535443, + "learning_rate": 1.7820489198451932e-05, + "loss": 1.02, + "step": 10092 + }, + { + "epoch": 0.24, + "grad_norm": 2.063960498401841, + "learning_rate": 1.7820013635489783e-05, + "loss": 1.0487, + "step": 10093 + }, + { + "epoch": 0.24, + "grad_norm": 1.8660616228584403, + "learning_rate": 1.781953802699704e-05, + "loss": 0.962, + "step": 10094 + }, + { + "epoch": 0.24, + "grad_norm": 2.1577847416146807, + "learning_rate": 1.7819062372976467e-05, + "loss": 0.975, + "step": 10095 + }, + { + "epoch": 0.24, + "grad_norm": 2.8521035838027196, + "learning_rate": 1.7818586673430842e-05, + "loss": 1.0629, + "step": 10096 + }, + { + "epoch": 0.24, + "grad_norm": 2.1217843543090615, + "learning_rate": 1.781811092836293e-05, + "loss": 1.1157, + "step": 10097 + }, + { + "epoch": 0.24, + "grad_norm": 2.0521843311201993, + "learning_rate": 1.78176351377755e-05, + "loss": 1.1279, + "step": 10098 + }, + { + "epoch": 0.24, + "grad_norm": 1.2204127433626113, + "learning_rate": 1.781715930167132e-05, + "loss": 0.9461, + "step": 10099 + }, + { + "epoch": 0.24, + "grad_norm": 2.295447315267468, + "learning_rate": 1.7816683420053164e-05, + "loss": 1.0273, + "step": 10100 + }, + { + "epoch": 0.24, + "grad_norm": 1.8325093274400366, + "learning_rate": 1.7816207492923807e-05, + "loss": 1.0545, + "step": 10101 + }, + { + "epoch": 0.24, + "grad_norm": 2.1707206034184052, + "learning_rate": 1.781573152028601e-05, + "loss": 1.0562, + "step": 10102 + }, + { + "epoch": 0.24, + "grad_norm": 2.3262710990530553, + "learning_rate": 1.7815255502142553e-05, + "loss": 1.1935, + "step": 10103 + }, + { + "epoch": 0.24, + "grad_norm": 2.034015424316312, + "learning_rate": 1.7814779438496207e-05, + "loss": 1.1042, + "step": 10104 + }, + { + "epoch": 0.24, + "grad_norm": 3.140251464390345, + "learning_rate": 1.781430332934974e-05, + "loss": 0.9272, + "step": 10105 + }, + { + "epoch": 0.24, + "grad_norm": 2.177633714380873, + "learning_rate": 1.781382717470592e-05, + "loss": 0.8562, + "step": 10106 + }, + { + "epoch": 0.24, + "grad_norm": 2.0152348626650647, + "learning_rate": 1.7813350974567528e-05, + "loss": 1.0229, + "step": 10107 + }, + { + "epoch": 0.24, + "grad_norm": 2.113739686474589, + "learning_rate": 1.781287472893733e-05, + "loss": 1.0547, + "step": 10108 + }, + { + "epoch": 0.24, + "grad_norm": 1.9549517763384578, + "learning_rate": 1.7812398437818107e-05, + "loss": 1.1508, + "step": 10109 + }, + { + "epoch": 0.24, + "grad_norm": 2.1914614491556983, + "learning_rate": 1.7811922101212622e-05, + "loss": 1.0744, + "step": 10110 + }, + { + "epoch": 0.24, + "grad_norm": 2.2598845040325553, + "learning_rate": 1.7811445719123654e-05, + "loss": 1.0676, + "step": 10111 + }, + { + "epoch": 0.24, + "grad_norm": 2.244554466969956, + "learning_rate": 1.781096929155398e-05, + "loss": 0.8958, + "step": 10112 + }, + { + "epoch": 0.24, + "grad_norm": 2.3159703695749787, + "learning_rate": 1.7810492818506365e-05, + "loss": 1.0319, + "step": 10113 + }, + { + "epoch": 0.24, + "grad_norm": 2.044730582621812, + "learning_rate": 1.7810016299983594e-05, + "loss": 0.9794, + "step": 10114 + }, + { + "epoch": 0.24, + "grad_norm": 2.1175120206601328, + "learning_rate": 1.780953973598843e-05, + "loss": 1.1239, + "step": 10115 + }, + { + "epoch": 0.24, + "grad_norm": 2.1325933302109026, + "learning_rate": 1.7809063126523653e-05, + "loss": 1.0255, + "step": 10116 + }, + { + "epoch": 0.24, + "grad_norm": 1.9656052318969681, + "learning_rate": 1.780858647159204e-05, + "loss": 1.023, + "step": 10117 + }, + { + "epoch": 0.24, + "grad_norm": 1.1101105979184385, + "learning_rate": 1.7808109771196364e-05, + "loss": 0.9831, + "step": 10118 + }, + { + "epoch": 0.24, + "grad_norm": 1.9190441807424747, + "learning_rate": 1.7807633025339403e-05, + "loss": 1.0841, + "step": 10119 + }, + { + "epoch": 0.24, + "grad_norm": 2.414988273506122, + "learning_rate": 1.780715623402393e-05, + "loss": 1.0937, + "step": 10120 + }, + { + "epoch": 0.24, + "grad_norm": 1.9116105510523262, + "learning_rate": 1.780667939725272e-05, + "loss": 0.9562, + "step": 10121 + }, + { + "epoch": 0.24, + "grad_norm": 2.08380631731455, + "learning_rate": 1.780620251502855e-05, + "loss": 0.8841, + "step": 10122 + }, + { + "epoch": 0.24, + "grad_norm": 4.612666344916118, + "learning_rate": 1.78057255873542e-05, + "loss": 0.9686, + "step": 10123 + }, + { + "epoch": 0.24, + "grad_norm": 1.9634124856869808, + "learning_rate": 1.780524861423244e-05, + "loss": 1.0589, + "step": 10124 + }, + { + "epoch": 0.24, + "grad_norm": 2.4967870084429524, + "learning_rate": 1.7804771595666057e-05, + "loss": 1.1288, + "step": 10125 + }, + { + "epoch": 0.24, + "grad_norm": 1.9288208641643867, + "learning_rate": 1.7804294531657822e-05, + "loss": 1.0609, + "step": 10126 + }, + { + "epoch": 0.24, + "grad_norm": 2.0624529729795675, + "learning_rate": 1.7803817422210512e-05, + "loss": 1.0593, + "step": 10127 + }, + { + "epoch": 0.24, + "grad_norm": 2.042647185442109, + "learning_rate": 1.7803340267326907e-05, + "loss": 1.0534, + "step": 10128 + }, + { + "epoch": 0.24, + "grad_norm": 1.9018990253362602, + "learning_rate": 1.7802863067009783e-05, + "loss": 0.9808, + "step": 10129 + }, + { + "epoch": 0.24, + "grad_norm": 2.538970720484689, + "learning_rate": 1.7802385821261922e-05, + "loss": 0.9175, + "step": 10130 + }, + { + "epoch": 0.24, + "grad_norm": 2.10067573946854, + "learning_rate": 1.78019085300861e-05, + "loss": 1.1374, + "step": 10131 + }, + { + "epoch": 0.24, + "grad_norm": 2.049553041063825, + "learning_rate": 1.7801431193485097e-05, + "loss": 1.1143, + "step": 10132 + }, + { + "epoch": 0.24, + "grad_norm": 2.1721111795683203, + "learning_rate": 1.780095381146169e-05, + "loss": 1.0981, + "step": 10133 + }, + { + "epoch": 0.24, + "grad_norm": 2.335269740336582, + "learning_rate": 1.7800476384018658e-05, + "loss": 1.1129, + "step": 10134 + }, + { + "epoch": 0.24, + "grad_norm": 2.922440028118999, + "learning_rate": 1.7799998911158786e-05, + "loss": 0.9503, + "step": 10135 + }, + { + "epoch": 0.24, + "grad_norm": 2.118184314901809, + "learning_rate": 1.7799521392884854e-05, + "loss": 1.0054, + "step": 10136 + }, + { + "epoch": 0.24, + "grad_norm": 2.1072846855006135, + "learning_rate": 1.7799043829199635e-05, + "loss": 1.1171, + "step": 10137 + }, + { + "epoch": 0.24, + "grad_norm": 2.220285093220894, + "learning_rate": 1.7798566220105912e-05, + "loss": 0.9953, + "step": 10138 + }, + { + "epoch": 0.24, + "grad_norm": 2.1302992018215967, + "learning_rate": 1.7798088565606473e-05, + "loss": 1.0742, + "step": 10139 + }, + { + "epoch": 0.24, + "grad_norm": 2.3259878474131632, + "learning_rate": 1.779761086570409e-05, + "loss": 0.9378, + "step": 10140 + }, + { + "epoch": 0.24, + "grad_norm": 3.4689343179026264, + "learning_rate": 1.779713312040155e-05, + "loss": 1.0222, + "step": 10141 + }, + { + "epoch": 0.24, + "grad_norm": 1.122883234333807, + "learning_rate": 1.7796655329701627e-05, + "loss": 1.047, + "step": 10142 + }, + { + "epoch": 0.24, + "grad_norm": 1.974757064480538, + "learning_rate": 1.7796177493607113e-05, + "loss": 1.1931, + "step": 10143 + }, + { + "epoch": 0.24, + "grad_norm": 1.1887682256720427, + "learning_rate": 1.7795699612120784e-05, + "loss": 1.0131, + "step": 10144 + }, + { + "epoch": 0.24, + "grad_norm": 2.455653778633533, + "learning_rate": 1.7795221685245425e-05, + "loss": 1.0769, + "step": 10145 + }, + { + "epoch": 0.24, + "grad_norm": 2.4311193922498577, + "learning_rate": 1.779474371298382e-05, + "loss": 1.0547, + "step": 10146 + }, + { + "epoch": 0.24, + "grad_norm": 2.4130725703917055, + "learning_rate": 1.7794265695338743e-05, + "loss": 0.9732, + "step": 10147 + }, + { + "epoch": 0.24, + "grad_norm": 2.208186030418714, + "learning_rate": 1.779378763231299e-05, + "loss": 0.967, + "step": 10148 + }, + { + "epoch": 0.24, + "grad_norm": 2.0322751044470713, + "learning_rate": 1.779330952390933e-05, + "loss": 1.0602, + "step": 10149 + }, + { + "epoch": 0.24, + "grad_norm": 2.3561739061774296, + "learning_rate": 1.779283137013056e-05, + "loss": 1.0974, + "step": 10150 + }, + { + "epoch": 0.24, + "grad_norm": 2.06165069194546, + "learning_rate": 1.779235317097946e-05, + "loss": 0.9454, + "step": 10151 + }, + { + "epoch": 0.24, + "grad_norm": 2.3610609903676347, + "learning_rate": 1.779187492645881e-05, + "loss": 0.9006, + "step": 10152 + }, + { + "epoch": 0.24, + "grad_norm": 2.493557416988799, + "learning_rate": 1.77913966365714e-05, + "loss": 1.1194, + "step": 10153 + }, + { + "epoch": 0.24, + "grad_norm": 1.9370056047191628, + "learning_rate": 1.779091830132001e-05, + "loss": 1.0049, + "step": 10154 + }, + { + "epoch": 0.24, + "grad_norm": 1.9779605127927053, + "learning_rate": 1.7790439920707424e-05, + "loss": 1.0763, + "step": 10155 + }, + { + "epoch": 0.24, + "grad_norm": 2.074611990680457, + "learning_rate": 1.7789961494736438e-05, + "loss": 1.0617, + "step": 10156 + }, + { + "epoch": 0.24, + "grad_norm": 2.9456821185694544, + "learning_rate": 1.7789483023409822e-05, + "loss": 1.0237, + "step": 10157 + }, + { + "epoch": 0.24, + "grad_norm": 1.9324014364604258, + "learning_rate": 1.7789004506730376e-05, + "loss": 0.9582, + "step": 10158 + }, + { + "epoch": 0.24, + "grad_norm": 2.082711234442624, + "learning_rate": 1.7788525944700873e-05, + "loss": 1.0514, + "step": 10159 + }, + { + "epoch": 0.24, + "grad_norm": 2.5451628287956356, + "learning_rate": 1.778804733732411e-05, + "loss": 0.9643, + "step": 10160 + }, + { + "epoch": 0.24, + "grad_norm": 1.1585409908346156, + "learning_rate": 1.778756868460287e-05, + "loss": 1.0017, + "step": 10161 + }, + { + "epoch": 0.24, + "grad_norm": 2.8834609159209452, + "learning_rate": 1.778708998653994e-05, + "loss": 0.8415, + "step": 10162 + }, + { + "epoch": 0.24, + "grad_norm": 1.9511174951169492, + "learning_rate": 1.7786611243138104e-05, + "loss": 1.0644, + "step": 10163 + }, + { + "epoch": 0.24, + "grad_norm": 2.781598348414138, + "learning_rate": 1.7786132454400155e-05, + "loss": 0.9862, + "step": 10164 + }, + { + "epoch": 0.24, + "grad_norm": 2.233904309021114, + "learning_rate": 1.778565362032888e-05, + "loss": 0.9829, + "step": 10165 + }, + { + "epoch": 0.24, + "grad_norm": 2.2876189243097445, + "learning_rate": 1.778517474092706e-05, + "loss": 1.0327, + "step": 10166 + }, + { + "epoch": 0.24, + "grad_norm": 2.0882145888584644, + "learning_rate": 1.778469581619749e-05, + "loss": 1.0318, + "step": 10167 + }, + { + "epoch": 0.24, + "grad_norm": 1.8764084138209751, + "learning_rate": 1.7784216846142955e-05, + "loss": 0.9437, + "step": 10168 + }, + { + "epoch": 0.24, + "grad_norm": 2.1279071548007695, + "learning_rate": 1.7783737830766247e-05, + "loss": 1.0334, + "step": 10169 + }, + { + "epoch": 0.24, + "grad_norm": 2.0683548612518337, + "learning_rate": 1.7783258770070153e-05, + "loss": 1.074, + "step": 10170 + }, + { + "epoch": 0.24, + "grad_norm": 2.087024794170287, + "learning_rate": 1.778277966405746e-05, + "loss": 1.051, + "step": 10171 + }, + { + "epoch": 0.24, + "grad_norm": 2.204398203174012, + "learning_rate": 1.7782300512730965e-05, + "loss": 1.1773, + "step": 10172 + }, + { + "epoch": 0.24, + "grad_norm": 1.1263174449189128, + "learning_rate": 1.7781821316093445e-05, + "loss": 1.0465, + "step": 10173 + }, + { + "epoch": 0.24, + "grad_norm": 1.8816995075753171, + "learning_rate": 1.7781342074147706e-05, + "loss": 1.1086, + "step": 10174 + }, + { + "epoch": 0.24, + "grad_norm": 2.160371586645001, + "learning_rate": 1.7780862786896526e-05, + "loss": 1.1091, + "step": 10175 + }, + { + "epoch": 0.24, + "grad_norm": 2.0702763748840884, + "learning_rate": 1.7780383454342696e-05, + "loss": 1.0393, + "step": 10176 + }, + { + "epoch": 0.24, + "grad_norm": 2.0066660861372894, + "learning_rate": 1.7779904076489013e-05, + "loss": 1.0715, + "step": 10177 + }, + { + "epoch": 0.24, + "grad_norm": 2.4797883484148, + "learning_rate": 1.7779424653338268e-05, + "loss": 1.1073, + "step": 10178 + }, + { + "epoch": 0.24, + "grad_norm": 2.057162694406809, + "learning_rate": 1.7778945184893245e-05, + "loss": 1.0137, + "step": 10179 + }, + { + "epoch": 0.24, + "grad_norm": 2.208247482870964, + "learning_rate": 1.7778465671156744e-05, + "loss": 1.0581, + "step": 10180 + }, + { + "epoch": 0.24, + "grad_norm": 1.1068056490742137, + "learning_rate": 1.777798611213155e-05, + "loss": 0.9258, + "step": 10181 + }, + { + "epoch": 0.24, + "grad_norm": 2.029629081000255, + "learning_rate": 1.777750650782046e-05, + "loss": 1.0489, + "step": 10182 + }, + { + "epoch": 0.24, + "grad_norm": 2.076466399405137, + "learning_rate": 1.7777026858226265e-05, + "loss": 1.1436, + "step": 10183 + }, + { + "epoch": 0.24, + "grad_norm": 2.1781898251068603, + "learning_rate": 1.7776547163351758e-05, + "loss": 1.0726, + "step": 10184 + }, + { + "epoch": 0.24, + "grad_norm": 1.8976027357929304, + "learning_rate": 1.777606742319973e-05, + "loss": 1.1429, + "step": 10185 + }, + { + "epoch": 0.24, + "grad_norm": 1.8954239906092325, + "learning_rate": 1.777558763777298e-05, + "loss": 1.0579, + "step": 10186 + }, + { + "epoch": 0.24, + "grad_norm": 2.116213242468395, + "learning_rate": 1.777510780707429e-05, + "loss": 1.1147, + "step": 10187 + }, + { + "epoch": 0.24, + "grad_norm": 1.7378458673936952, + "learning_rate": 1.7774627931106464e-05, + "loss": 1.1352, + "step": 10188 + }, + { + "epoch": 0.24, + "grad_norm": 2.1417723836321216, + "learning_rate": 1.7774148009872295e-05, + "loss": 1.0594, + "step": 10189 + }, + { + "epoch": 0.24, + "grad_norm": 2.343676215242216, + "learning_rate": 1.7773668043374573e-05, + "loss": 0.9591, + "step": 10190 + }, + { + "epoch": 0.24, + "grad_norm": 1.9964033742154943, + "learning_rate": 1.7773188031616096e-05, + "loss": 1.0029, + "step": 10191 + }, + { + "epoch": 0.24, + "grad_norm": 2.040985468591464, + "learning_rate": 1.7772707974599656e-05, + "loss": 1.0438, + "step": 10192 + }, + { + "epoch": 0.24, + "grad_norm": 2.099295305892311, + "learning_rate": 1.777222787232805e-05, + "loss": 1.0218, + "step": 10193 + }, + { + "epoch": 0.24, + "grad_norm": 1.9915515352052355, + "learning_rate": 1.7771747724804073e-05, + "loss": 1.1032, + "step": 10194 + }, + { + "epoch": 0.24, + "grad_norm": 2.32203301276581, + "learning_rate": 1.777126753203052e-05, + "loss": 0.9161, + "step": 10195 + }, + { + "epoch": 0.24, + "grad_norm": 2.433171906005344, + "learning_rate": 1.7770787294010185e-05, + "loss": 1.0458, + "step": 10196 + }, + { + "epoch": 0.24, + "grad_norm": 2.205684903803957, + "learning_rate": 1.7770307010745867e-05, + "loss": 0.9751, + "step": 10197 + }, + { + "epoch": 0.24, + "grad_norm": 1.916411155715982, + "learning_rate": 1.7769826682240365e-05, + "loss": 1.0997, + "step": 10198 + }, + { + "epoch": 0.24, + "grad_norm": 2.1399765012336425, + "learning_rate": 1.7769346308496466e-05, + "loss": 1.071, + "step": 10199 + }, + { + "epoch": 0.24, + "grad_norm": 2.0815775800391583, + "learning_rate": 1.776886588951698e-05, + "loss": 1.079, + "step": 10200 + }, + { + "epoch": 0.24, + "grad_norm": 2.434716431734932, + "learning_rate": 1.7768385425304692e-05, + "loss": 1.0732, + "step": 10201 + }, + { + "epoch": 0.24, + "grad_norm": 2.1167492373319425, + "learning_rate": 1.7767904915862405e-05, + "loss": 0.9443, + "step": 10202 + }, + { + "epoch": 0.24, + "grad_norm": 1.9939309874807616, + "learning_rate": 1.776742436119292e-05, + "loss": 1.0273, + "step": 10203 + }, + { + "epoch": 0.24, + "grad_norm": 1.16448468034711, + "learning_rate": 1.776694376129903e-05, + "loss": 1.0209, + "step": 10204 + }, + { + "epoch": 0.24, + "grad_norm": 2.1269634726476694, + "learning_rate": 1.776646311618353e-05, + "loss": 1.0187, + "step": 10205 + }, + { + "epoch": 0.24, + "grad_norm": 2.6314720346329463, + "learning_rate": 1.776598242584923e-05, + "loss": 1.0296, + "step": 10206 + }, + { + "epoch": 0.24, + "grad_norm": 2.087831581108751, + "learning_rate": 1.7765501690298913e-05, + "loss": 1.1253, + "step": 10207 + }, + { + "epoch": 0.24, + "grad_norm": 2.3081019310533732, + "learning_rate": 1.7765020909535395e-05, + "loss": 1.1308, + "step": 10208 + }, + { + "epoch": 0.24, + "grad_norm": 2.0245816411678828, + "learning_rate": 1.7764540083561462e-05, + "loss": 0.9797, + "step": 10209 + }, + { + "epoch": 0.24, + "grad_norm": 2.261293959174182, + "learning_rate": 1.776405921237992e-05, + "loss": 1.2639, + "step": 10210 + }, + { + "epoch": 0.24, + "grad_norm": 1.944922444963496, + "learning_rate": 1.776357829599357e-05, + "loss": 0.959, + "step": 10211 + }, + { + "epoch": 0.24, + "grad_norm": 2.207791547189052, + "learning_rate": 1.7763097334405204e-05, + "loss": 1.1341, + "step": 10212 + }, + { + "epoch": 0.24, + "grad_norm": 1.8363584933623167, + "learning_rate": 1.776261632761763e-05, + "loss": 0.9495, + "step": 10213 + }, + { + "epoch": 0.24, + "grad_norm": 1.1052472329441723, + "learning_rate": 1.7762135275633645e-05, + "loss": 1.0544, + "step": 10214 + }, + { + "epoch": 0.24, + "grad_norm": 2.05446433228312, + "learning_rate": 1.7761654178456053e-05, + "loss": 0.973, + "step": 10215 + }, + { + "epoch": 0.24, + "grad_norm": 2.0566562928674377, + "learning_rate": 1.776117303608765e-05, + "loss": 1.1141, + "step": 10216 + }, + { + "epoch": 0.24, + "grad_norm": 1.8092054814008538, + "learning_rate": 1.7760691848531246e-05, + "loss": 1.1727, + "step": 10217 + }, + { + "epoch": 0.24, + "grad_norm": 2.196839046396032, + "learning_rate": 1.7760210615789633e-05, + "loss": 1.1053, + "step": 10218 + }, + { + "epoch": 0.24, + "grad_norm": 2.255623561694293, + "learning_rate": 1.7759729337865618e-05, + "loss": 1.0033, + "step": 10219 + }, + { + "epoch": 0.24, + "grad_norm": 1.1643538948383745, + "learning_rate": 1.7759248014762e-05, + "loss": 0.9373, + "step": 10220 + }, + { + "epoch": 0.24, + "grad_norm": 2.3505284261145896, + "learning_rate": 1.7758766646481587e-05, + "loss": 1.0427, + "step": 10221 + }, + { + "epoch": 0.24, + "grad_norm": 2.0200787294218503, + "learning_rate": 1.7758285233027174e-05, + "loss": 1.0913, + "step": 10222 + }, + { + "epoch": 0.24, + "grad_norm": 1.9568575491217717, + "learning_rate": 1.7757803774401572e-05, + "loss": 0.97, + "step": 10223 + }, + { + "epoch": 0.24, + "grad_norm": 2.0819134087297337, + "learning_rate": 1.775732227060758e-05, + "loss": 0.985, + "step": 10224 + }, + { + "epoch": 0.24, + "grad_norm": 1.1471281390810835, + "learning_rate": 1.7756840721648e-05, + "loss": 0.9231, + "step": 10225 + }, + { + "epoch": 0.24, + "grad_norm": 1.9633726379837122, + "learning_rate": 1.7756359127525638e-05, + "loss": 1.1275, + "step": 10226 + }, + { + "epoch": 0.24, + "grad_norm": 1.9150260304061544, + "learning_rate": 1.7755877488243297e-05, + "loss": 1.1017, + "step": 10227 + }, + { + "epoch": 0.24, + "grad_norm": 1.8840967774153945, + "learning_rate": 1.7755395803803783e-05, + "loss": 1.0235, + "step": 10228 + }, + { + "epoch": 0.24, + "grad_norm": 2.163994396210545, + "learning_rate": 1.7754914074209898e-05, + "loss": 0.9554, + "step": 10229 + }, + { + "epoch": 0.24, + "grad_norm": 3.1452644307503452, + "learning_rate": 1.7754432299464448e-05, + "loss": 0.9744, + "step": 10230 + }, + { + "epoch": 0.24, + "grad_norm": 1.9306056506322502, + "learning_rate": 1.775395047957024e-05, + "loss": 1.0621, + "step": 10231 + }, + { + "epoch": 0.24, + "grad_norm": 1.8326644709275706, + "learning_rate": 1.7753468614530075e-05, + "loss": 0.9289, + "step": 10232 + }, + { + "epoch": 0.24, + "grad_norm": 1.216317245700712, + "learning_rate": 1.7752986704346764e-05, + "loss": 0.9906, + "step": 10233 + }, + { + "epoch": 0.24, + "grad_norm": 1.1253863944318725, + "learning_rate": 1.7752504749023108e-05, + "loss": 0.9102, + "step": 10234 + }, + { + "epoch": 0.24, + "grad_norm": 2.1651565332122447, + "learning_rate": 1.7752022748561917e-05, + "loss": 1.0303, + "step": 10235 + }, + { + "epoch": 0.24, + "grad_norm": 2.194253801239751, + "learning_rate": 1.7751540702965993e-05, + "loss": 1.0402, + "step": 10236 + }, + { + "epoch": 0.24, + "grad_norm": 1.0900385694890564, + "learning_rate": 1.7751058612238147e-05, + "loss": 1.0078, + "step": 10237 + }, + { + "epoch": 0.24, + "grad_norm": 2.1922257114357833, + "learning_rate": 1.775057647638118e-05, + "loss": 1.1579, + "step": 10238 + }, + { + "epoch": 0.24, + "grad_norm": 2.1681637363770356, + "learning_rate": 1.7750094295397907e-05, + "loss": 1.0754, + "step": 10239 + }, + { + "epoch": 0.24, + "grad_norm": 2.1221295447011235, + "learning_rate": 1.774961206929113e-05, + "loss": 1.1675, + "step": 10240 + }, + { + "epoch": 0.24, + "grad_norm": 2.171981350628483, + "learning_rate": 1.7749129798063657e-05, + "loss": 0.962, + "step": 10241 + }, + { + "epoch": 0.24, + "grad_norm": 2.1707676714279103, + "learning_rate": 1.77486474817183e-05, + "loss": 1.0371, + "step": 10242 + }, + { + "epoch": 0.24, + "grad_norm": 2.2760561839431084, + "learning_rate": 1.774816512025786e-05, + "loss": 1.0743, + "step": 10243 + }, + { + "epoch": 0.24, + "grad_norm": 1.8819700807202848, + "learning_rate": 1.7747682713685154e-05, + "loss": 0.9593, + "step": 10244 + }, + { + "epoch": 0.24, + "grad_norm": 2.1407000536578322, + "learning_rate": 1.7747200262002983e-05, + "loss": 0.9505, + "step": 10245 + }, + { + "epoch": 0.24, + "grad_norm": 2.0329188935821003, + "learning_rate": 1.774671776521416e-05, + "loss": 0.9754, + "step": 10246 + }, + { + "epoch": 0.24, + "grad_norm": 1.9952138188474178, + "learning_rate": 1.7746235223321492e-05, + "loss": 1.09, + "step": 10247 + }, + { + "epoch": 0.24, + "grad_norm": 2.003868056304197, + "learning_rate": 1.7745752636327793e-05, + "loss": 0.9136, + "step": 10248 + }, + { + "epoch": 0.24, + "grad_norm": 1.858470915961127, + "learning_rate": 1.774527000423587e-05, + "loss": 0.9422, + "step": 10249 + }, + { + "epoch": 0.24, + "grad_norm": 1.9972455684061248, + "learning_rate": 1.7744787327048533e-05, + "loss": 0.9257, + "step": 10250 + }, + { + "epoch": 0.24, + "grad_norm": 2.0621298686799636, + "learning_rate": 1.7744304604768588e-05, + "loss": 1.0849, + "step": 10251 + }, + { + "epoch": 0.24, + "grad_norm": 2.152671264395816, + "learning_rate": 1.7743821837398854e-05, + "loss": 1.0171, + "step": 10252 + }, + { + "epoch": 0.24, + "grad_norm": 2.797451615832383, + "learning_rate": 1.7743339024942135e-05, + "loss": 1.1, + "step": 10253 + }, + { + "epoch": 0.24, + "grad_norm": 2.070952450388677, + "learning_rate": 1.7742856167401248e-05, + "loss": 1.1687, + "step": 10254 + }, + { + "epoch": 0.24, + "grad_norm": 2.18647921344115, + "learning_rate": 1.7742373264778997e-05, + "loss": 1.035, + "step": 10255 + }, + { + "epoch": 0.24, + "grad_norm": 1.2447559606862542, + "learning_rate": 1.7741890317078205e-05, + "loss": 1.0639, + "step": 10256 + }, + { + "epoch": 0.24, + "grad_norm": 1.8795743546145238, + "learning_rate": 1.774140732430167e-05, + "loss": 0.9193, + "step": 10257 + }, + { + "epoch": 0.24, + "grad_norm": 2.2963371871742435, + "learning_rate": 1.774092428645221e-05, + "loss": 1.1122, + "step": 10258 + }, + { + "epoch": 0.24, + "grad_norm": 1.795710260931171, + "learning_rate": 1.7740441203532638e-05, + "loss": 0.9751, + "step": 10259 + }, + { + "epoch": 0.24, + "grad_norm": 2.128361168033682, + "learning_rate": 1.7739958075545772e-05, + "loss": 1.1045, + "step": 10260 + }, + { + "epoch": 0.24, + "grad_norm": 2.107199074760096, + "learning_rate": 1.7739474902494416e-05, + "loss": 0.9209, + "step": 10261 + }, + { + "epoch": 0.24, + "grad_norm": 4.01826123565959, + "learning_rate": 1.7738991684381387e-05, + "loss": 0.9601, + "step": 10262 + }, + { + "epoch": 0.24, + "grad_norm": 2.362768299072931, + "learning_rate": 1.77385084212095e-05, + "loss": 0.9656, + "step": 10263 + }, + { + "epoch": 0.24, + "grad_norm": 2.2322986377177534, + "learning_rate": 1.7738025112981565e-05, + "loss": 1.1194, + "step": 10264 + }, + { + "epoch": 0.24, + "grad_norm": 1.1709459804605646, + "learning_rate": 1.7737541759700398e-05, + "loss": 1.006, + "step": 10265 + }, + { + "epoch": 0.24, + "grad_norm": 2.4891606628901304, + "learning_rate": 1.7737058361368814e-05, + "loss": 0.8725, + "step": 10266 + }, + { + "epoch": 0.24, + "grad_norm": 1.858676058393647, + "learning_rate": 1.7736574917989627e-05, + "loss": 0.9854, + "step": 10267 + }, + { + "epoch": 0.24, + "grad_norm": 2.1992208281154877, + "learning_rate": 1.773609142956565e-05, + "loss": 0.9969, + "step": 10268 + }, + { + "epoch": 0.24, + "grad_norm": 2.4913777970659754, + "learning_rate": 1.77356078960997e-05, + "loss": 1.1379, + "step": 10269 + }, + { + "epoch": 0.24, + "grad_norm": 1.2735314144421501, + "learning_rate": 1.773512431759459e-05, + "loss": 1.0208, + "step": 10270 + }, + { + "epoch": 0.24, + "grad_norm": 1.8425688624983194, + "learning_rate": 1.7734640694053143e-05, + "loss": 1.154, + "step": 10271 + }, + { + "epoch": 0.24, + "grad_norm": 1.961984570234569, + "learning_rate": 1.7734157025478164e-05, + "loss": 1.0641, + "step": 10272 + }, + { + "epoch": 0.24, + "grad_norm": 1.0809445264732107, + "learning_rate": 1.7733673311872478e-05, + "loss": 0.9909, + "step": 10273 + }, + { + "epoch": 0.24, + "grad_norm": 2.3545844307412014, + "learning_rate": 1.7733189553238893e-05, + "loss": 1.0609, + "step": 10274 + }, + { + "epoch": 0.24, + "grad_norm": 2.0356943454261045, + "learning_rate": 1.7732705749580235e-05, + "loss": 1.0601, + "step": 10275 + }, + { + "epoch": 0.24, + "grad_norm": 1.1409946666880275, + "learning_rate": 1.7732221900899313e-05, + "loss": 1.0011, + "step": 10276 + }, + { + "epoch": 0.24, + "grad_norm": 2.0392408280432117, + "learning_rate": 1.773173800719895e-05, + "loss": 0.9145, + "step": 10277 + }, + { + "epoch": 0.24, + "grad_norm": 2.0624952538618113, + "learning_rate": 1.7731254068481955e-05, + "loss": 1.0143, + "step": 10278 + }, + { + "epoch": 0.24, + "grad_norm": 1.1426148098982207, + "learning_rate": 1.7730770084751156e-05, + "loss": 0.9923, + "step": 10279 + }, + { + "epoch": 0.24, + "grad_norm": 2.070064791444497, + "learning_rate": 1.7730286056009363e-05, + "loss": 1.0271, + "step": 10280 + }, + { + "epoch": 0.24, + "grad_norm": 2.2628422001250867, + "learning_rate": 1.77298019822594e-05, + "loss": 1.1916, + "step": 10281 + }, + { + "epoch": 0.24, + "grad_norm": 2.087459048566194, + "learning_rate": 1.7729317863504082e-05, + "loss": 0.9479, + "step": 10282 + }, + { + "epoch": 0.24, + "grad_norm": 2.404858852554622, + "learning_rate": 1.7728833699746226e-05, + "loss": 1.0263, + "step": 10283 + }, + { + "epoch": 0.24, + "grad_norm": 2.0566999261816337, + "learning_rate": 1.7728349490988653e-05, + "loss": 1.1262, + "step": 10284 + }, + { + "epoch": 0.24, + "grad_norm": 1.9134999291448642, + "learning_rate": 1.7727865237234187e-05, + "loss": 1.0814, + "step": 10285 + }, + { + "epoch": 0.24, + "grad_norm": 1.8925043374401866, + "learning_rate": 1.7727380938485642e-05, + "loss": 0.9219, + "step": 10286 + }, + { + "epoch": 0.24, + "grad_norm": 2.330230562514901, + "learning_rate": 1.7726896594745834e-05, + "loss": 1.0234, + "step": 10287 + }, + { + "epoch": 0.24, + "grad_norm": 2.1811462090130354, + "learning_rate": 1.7726412206017593e-05, + "loss": 0.9345, + "step": 10288 + }, + { + "epoch": 0.24, + "grad_norm": 2.002299169138728, + "learning_rate": 1.772592777230373e-05, + "loss": 0.9934, + "step": 10289 + }, + { + "epoch": 0.24, + "grad_norm": 2.0764394486331215, + "learning_rate": 1.7725443293607074e-05, + "loss": 0.9883, + "step": 10290 + }, + { + "epoch": 0.24, + "grad_norm": 1.1290870316746138, + "learning_rate": 1.7724958769930437e-05, + "loss": 1.0188, + "step": 10291 + }, + { + "epoch": 0.24, + "grad_norm": 2.187733021952562, + "learning_rate": 1.7724474201276646e-05, + "loss": 0.963, + "step": 10292 + }, + { + "epoch": 0.24, + "grad_norm": 2.047544175018691, + "learning_rate": 1.772398958764852e-05, + "loss": 1.0725, + "step": 10293 + }, + { + "epoch": 0.24, + "grad_norm": 1.080240253525224, + "learning_rate": 1.7723504929048886e-05, + "loss": 0.9934, + "step": 10294 + }, + { + "epoch": 0.24, + "grad_norm": 2.049009564982451, + "learning_rate": 1.7723020225480554e-05, + "loss": 1.0265, + "step": 10295 + }, + { + "epoch": 0.24, + "grad_norm": 2.0328297027645985, + "learning_rate": 1.7722535476946357e-05, + "loss": 0.9888, + "step": 10296 + }, + { + "epoch": 0.24, + "grad_norm": 2.0120079765457213, + "learning_rate": 1.7722050683449114e-05, + "loss": 1.0633, + "step": 10297 + }, + { + "epoch": 0.24, + "grad_norm": 2.3582315939771967, + "learning_rate": 1.7721565844991643e-05, + "loss": 1.0245, + "step": 10298 + }, + { + "epoch": 0.24, + "grad_norm": 2.0872664590297334, + "learning_rate": 1.7721080961576776e-05, + "loss": 1.1741, + "step": 10299 + }, + { + "epoch": 0.24, + "grad_norm": 1.1092961268191082, + "learning_rate": 1.7720596033207327e-05, + "loss": 1.0366, + "step": 10300 + }, + { + "epoch": 0.24, + "grad_norm": 1.9782881450645902, + "learning_rate": 1.7720111059886127e-05, + "loss": 1.1484, + "step": 10301 + }, + { + "epoch": 0.24, + "grad_norm": 2.23191169334955, + "learning_rate": 1.7719626041615992e-05, + "loss": 0.9586, + "step": 10302 + }, + { + "epoch": 0.24, + "grad_norm": 3.719200271118239, + "learning_rate": 1.7719140978399755e-05, + "loss": 0.983, + "step": 10303 + }, + { + "epoch": 0.24, + "grad_norm": 1.8513606970376784, + "learning_rate": 1.7718655870240234e-05, + "loss": 0.9333, + "step": 10304 + }, + { + "epoch": 0.24, + "grad_norm": 2.1426824913321454, + "learning_rate": 1.7718170717140256e-05, + "loss": 1.1123, + "step": 10305 + }, + { + "epoch": 0.24, + "grad_norm": 2.107820927934114, + "learning_rate": 1.771768551910264e-05, + "loss": 1.0897, + "step": 10306 + }, + { + "epoch": 0.24, + "grad_norm": 2.2986240133283546, + "learning_rate": 1.771720027613022e-05, + "loss": 1.0593, + "step": 10307 + }, + { + "epoch": 0.24, + "grad_norm": 3.1984597104479637, + "learning_rate": 1.7716714988225815e-05, + "loss": 1.2274, + "step": 10308 + }, + { + "epoch": 0.24, + "grad_norm": 2.7252527888942044, + "learning_rate": 1.7716229655392255e-05, + "loss": 1.1072, + "step": 10309 + }, + { + "epoch": 0.24, + "grad_norm": 2.09400511808032, + "learning_rate": 1.771574427763236e-05, + "loss": 1.1519, + "step": 10310 + }, + { + "epoch": 0.24, + "grad_norm": 1.9636161040591595, + "learning_rate": 1.771525885494896e-05, + "loss": 1.1363, + "step": 10311 + }, + { + "epoch": 0.24, + "grad_norm": 2.027818873636265, + "learning_rate": 1.7714773387344878e-05, + "loss": 0.9194, + "step": 10312 + }, + { + "epoch": 0.24, + "grad_norm": 1.9086644669561705, + "learning_rate": 1.7714287874822945e-05, + "loss": 0.9753, + "step": 10313 + }, + { + "epoch": 0.24, + "grad_norm": 2.165474733995206, + "learning_rate": 1.7713802317385985e-05, + "loss": 1.0064, + "step": 10314 + }, + { + "epoch": 0.24, + "grad_norm": 2.4524458786833976, + "learning_rate": 1.7713316715036827e-05, + "loss": 1.0052, + "step": 10315 + }, + { + "epoch": 0.24, + "grad_norm": 2.2320090576137783, + "learning_rate": 1.7712831067778298e-05, + "loss": 1.0938, + "step": 10316 + }, + { + "epoch": 0.24, + "grad_norm": 1.1626616744320033, + "learning_rate": 1.771234537561322e-05, + "loss": 0.936, + "step": 10317 + }, + { + "epoch": 0.24, + "grad_norm": 2.040465942083364, + "learning_rate": 1.771185963854443e-05, + "loss": 1.0695, + "step": 10318 + }, + { + "epoch": 0.24, + "grad_norm": 1.937098769204856, + "learning_rate": 1.771137385657475e-05, + "loss": 0.9688, + "step": 10319 + }, + { + "epoch": 0.24, + "grad_norm": 2.452254006233771, + "learning_rate": 1.771088802970701e-05, + "loss": 0.9821, + "step": 10320 + }, + { + "epoch": 0.24, + "grad_norm": 2.0496751518084024, + "learning_rate": 1.771040215794404e-05, + "loss": 1.0305, + "step": 10321 + }, + { + "epoch": 0.24, + "grad_norm": 2.211867059961923, + "learning_rate": 1.7709916241288667e-05, + "loss": 1.0612, + "step": 10322 + }, + { + "epoch": 0.24, + "grad_norm": 2.2007733606668345, + "learning_rate": 1.770943027974372e-05, + "loss": 1.1251, + "step": 10323 + }, + { + "epoch": 0.24, + "grad_norm": 2.0932636080787415, + "learning_rate": 1.770894427331203e-05, + "loss": 1.0817, + "step": 10324 + }, + { + "epoch": 0.24, + "grad_norm": 2.404111706887943, + "learning_rate": 1.7708458221996426e-05, + "loss": 0.9837, + "step": 10325 + }, + { + "epoch": 0.24, + "grad_norm": 2.2883824020234993, + "learning_rate": 1.7707972125799738e-05, + "loss": 0.9028, + "step": 10326 + }, + { + "epoch": 0.24, + "grad_norm": 1.9596196572672342, + "learning_rate": 1.7707485984724795e-05, + "loss": 1.1507, + "step": 10327 + }, + { + "epoch": 0.24, + "grad_norm": 2.454806204647665, + "learning_rate": 1.7706999798774426e-05, + "loss": 1.0751, + "step": 10328 + }, + { + "epoch": 0.24, + "grad_norm": 2.158612182686819, + "learning_rate": 1.7706513567951467e-05, + "loss": 0.9799, + "step": 10329 + }, + { + "epoch": 0.24, + "grad_norm": 2.5285655670185663, + "learning_rate": 1.7706027292258747e-05, + "loss": 1.1003, + "step": 10330 + }, + { + "epoch": 0.24, + "grad_norm": 2.209243245833991, + "learning_rate": 1.7705540971699095e-05, + "loss": 1.015, + "step": 10331 + }, + { + "epoch": 0.24, + "grad_norm": 2.3455775669616927, + "learning_rate": 1.7705054606275343e-05, + "loss": 1.0582, + "step": 10332 + }, + { + "epoch": 0.24, + "grad_norm": 2.332669143318528, + "learning_rate": 1.7704568195990324e-05, + "loss": 0.9055, + "step": 10333 + }, + { + "epoch": 0.24, + "grad_norm": 2.8672868937728797, + "learning_rate": 1.770408174084687e-05, + "loss": 1.0195, + "step": 10334 + }, + { + "epoch": 0.24, + "grad_norm": 2.4691990866123597, + "learning_rate": 1.7703595240847815e-05, + "loss": 1.0735, + "step": 10335 + }, + { + "epoch": 0.24, + "grad_norm": 1.1237809524921734, + "learning_rate": 1.770310869599599e-05, + "loss": 0.9066, + "step": 10336 + }, + { + "epoch": 0.24, + "grad_norm": 2.003657933341558, + "learning_rate": 1.770262210629422e-05, + "loss": 1.0415, + "step": 10337 + }, + { + "epoch": 0.24, + "grad_norm": 1.9606421236780687, + "learning_rate": 1.7702135471745358e-05, + "loss": 1.1105, + "step": 10338 + }, + { + "epoch": 0.24, + "grad_norm": 2.25232689755786, + "learning_rate": 1.7701648792352215e-05, + "loss": 1.0297, + "step": 10339 + }, + { + "epoch": 0.24, + "grad_norm": 1.1794742778767624, + "learning_rate": 1.7701162068117636e-05, + "loss": 0.977, + "step": 10340 + }, + { + "epoch": 0.24, + "grad_norm": 1.085370003190239, + "learning_rate": 1.7700675299044457e-05, + "loss": 0.9435, + "step": 10341 + }, + { + "epoch": 0.24, + "grad_norm": 2.0563562563163664, + "learning_rate": 1.7700188485135506e-05, + "loss": 0.9701, + "step": 10342 + }, + { + "epoch": 0.24, + "grad_norm": 2.1095172466555763, + "learning_rate": 1.769970162639362e-05, + "loss": 0.9893, + "step": 10343 + }, + { + "epoch": 0.24, + "grad_norm": 1.97867785501354, + "learning_rate": 1.7699214722821634e-05, + "loss": 0.8789, + "step": 10344 + }, + { + "epoch": 0.24, + "grad_norm": 2.277007350430963, + "learning_rate": 1.769872777442238e-05, + "loss": 0.9842, + "step": 10345 + }, + { + "epoch": 0.24, + "grad_norm": 2.1808446594977164, + "learning_rate": 1.7698240781198696e-05, + "loss": 1.0767, + "step": 10346 + }, + { + "epoch": 0.24, + "grad_norm": 2.198346584926225, + "learning_rate": 1.769775374315342e-05, + "loss": 1.0721, + "step": 10347 + }, + { + "epoch": 0.24, + "grad_norm": 2.1087648093062032, + "learning_rate": 1.7697266660289383e-05, + "loss": 1.0719, + "step": 10348 + }, + { + "epoch": 0.24, + "grad_norm": 1.8388519245325632, + "learning_rate": 1.769677953260942e-05, + "loss": 1.1257, + "step": 10349 + }, + { + "epoch": 0.24, + "grad_norm": 2.0312473071291386, + "learning_rate": 1.7696292360116373e-05, + "loss": 0.9254, + "step": 10350 + }, + { + "epoch": 0.24, + "grad_norm": 1.0729210084897525, + "learning_rate": 1.7695805142813073e-05, + "loss": 0.9209, + "step": 10351 + }, + { + "epoch": 0.24, + "grad_norm": 2.3167745923953507, + "learning_rate": 1.769531788070236e-05, + "loss": 1.0099, + "step": 10352 + }, + { + "epoch": 0.24, + "grad_norm": 1.1092586878210133, + "learning_rate": 1.7694830573787075e-05, + "loss": 0.969, + "step": 10353 + }, + { + "epoch": 0.24, + "grad_norm": 1.9805069451580268, + "learning_rate": 1.769434322207004e-05, + "loss": 1.0006, + "step": 10354 + }, + { + "epoch": 0.24, + "grad_norm": 2.162587251076178, + "learning_rate": 1.7693855825554113e-05, + "loss": 0.9811, + "step": 10355 + }, + { + "epoch": 0.24, + "grad_norm": 1.8859833882133015, + "learning_rate": 1.7693368384242115e-05, + "loss": 1.0179, + "step": 10356 + }, + { + "epoch": 0.24, + "grad_norm": 1.1330005405414925, + "learning_rate": 1.7692880898136894e-05, + "loss": 1.0009, + "step": 10357 + }, + { + "epoch": 0.24, + "grad_norm": 2.1332462578422233, + "learning_rate": 1.7692393367241283e-05, + "loss": 0.9694, + "step": 10358 + }, + { + "epoch": 0.24, + "grad_norm": 2.59460760505398, + "learning_rate": 1.7691905791558123e-05, + "loss": 1.2413, + "step": 10359 + }, + { + "epoch": 0.24, + "grad_norm": 1.847435991219326, + "learning_rate": 1.7691418171090254e-05, + "loss": 0.9753, + "step": 10360 + }, + { + "epoch": 0.24, + "grad_norm": 1.98881197995768, + "learning_rate": 1.7690930505840513e-05, + "loss": 1.118, + "step": 10361 + }, + { + "epoch": 0.24, + "grad_norm": 1.9816136974852816, + "learning_rate": 1.7690442795811734e-05, + "loss": 1.0478, + "step": 10362 + }, + { + "epoch": 0.24, + "grad_norm": 2.226238884957332, + "learning_rate": 1.7689955041006768e-05, + "loss": 1.1266, + "step": 10363 + }, + { + "epoch": 0.24, + "grad_norm": 1.0658862460470884, + "learning_rate": 1.7689467241428446e-05, + "loss": 1.0024, + "step": 10364 + }, + { + "epoch": 0.24, + "grad_norm": 1.9542057615686945, + "learning_rate": 1.7688979397079612e-05, + "loss": 0.9165, + "step": 10365 + }, + { + "epoch": 0.24, + "grad_norm": 1.8733859483718613, + "learning_rate": 1.7688491507963108e-05, + "loss": 1.0882, + "step": 10366 + }, + { + "epoch": 0.24, + "grad_norm": 2.249420088281896, + "learning_rate": 1.7688003574081773e-05, + "loss": 0.9983, + "step": 10367 + }, + { + "epoch": 0.24, + "grad_norm": 1.9642448301683784, + "learning_rate": 1.7687515595438443e-05, + "loss": 1.0943, + "step": 10368 + }, + { + "epoch": 0.24, + "grad_norm": 2.2553168195514903, + "learning_rate": 1.7687027572035965e-05, + "loss": 0.9501, + "step": 10369 + }, + { + "epoch": 0.24, + "grad_norm": 1.147465787025106, + "learning_rate": 1.768653950387718e-05, + "loss": 0.9902, + "step": 10370 + }, + { + "epoch": 0.24, + "grad_norm": 2.0386730922203613, + "learning_rate": 1.768605139096493e-05, + "loss": 1.0198, + "step": 10371 + }, + { + "epoch": 0.24, + "grad_norm": 2.209837363844782, + "learning_rate": 1.7685563233302053e-05, + "loss": 1.04, + "step": 10372 + }, + { + "epoch": 0.24, + "grad_norm": 1.8395969483875856, + "learning_rate": 1.7685075030891393e-05, + "loss": 1.0354, + "step": 10373 + }, + { + "epoch": 0.24, + "grad_norm": 2.080853802487499, + "learning_rate": 1.7684586783735793e-05, + "loss": 1.0396, + "step": 10374 + }, + { + "epoch": 0.24, + "grad_norm": 3.509446230670335, + "learning_rate": 1.7684098491838097e-05, + "loss": 1.0276, + "step": 10375 + }, + { + "epoch": 0.24, + "grad_norm": 1.0958044838007754, + "learning_rate": 1.7683610155201148e-05, + "loss": 1.0142, + "step": 10376 + }, + { + "epoch": 0.24, + "grad_norm": 1.91942595475743, + "learning_rate": 1.7683121773827787e-05, + "loss": 1.128, + "step": 10377 + }, + { + "epoch": 0.24, + "grad_norm": 1.1697960600444224, + "learning_rate": 1.7682633347720858e-05, + "loss": 1.0714, + "step": 10378 + }, + { + "epoch": 0.24, + "grad_norm": 2.0314643626396753, + "learning_rate": 1.7682144876883207e-05, + "loss": 1.1018, + "step": 10379 + }, + { + "epoch": 0.24, + "grad_norm": 1.95125855201073, + "learning_rate": 1.7681656361317676e-05, + "loss": 1.0942, + "step": 10380 + }, + { + "epoch": 0.24, + "grad_norm": 2.4647793930519133, + "learning_rate": 1.7681167801027107e-05, + "loss": 1.0792, + "step": 10381 + }, + { + "epoch": 0.24, + "grad_norm": 2.0874524098242477, + "learning_rate": 1.7680679196014348e-05, + "loss": 1.0275, + "step": 10382 + }, + { + "epoch": 0.24, + "grad_norm": 2.346776539601329, + "learning_rate": 1.7680190546282244e-05, + "loss": 1.0973, + "step": 10383 + }, + { + "epoch": 0.24, + "grad_norm": 2.1149067542448456, + "learning_rate": 1.767970185183364e-05, + "loss": 1.2454, + "step": 10384 + }, + { + "epoch": 0.24, + "grad_norm": 2.2019728539834795, + "learning_rate": 1.7679213112671378e-05, + "loss": 0.958, + "step": 10385 + }, + { + "epoch": 0.24, + "grad_norm": 1.8251849179178548, + "learning_rate": 1.7678724328798308e-05, + "loss": 0.9549, + "step": 10386 + }, + { + "epoch": 0.24, + "grad_norm": 2.0144140764340492, + "learning_rate": 1.7678235500217274e-05, + "loss": 1.0024, + "step": 10387 + }, + { + "epoch": 0.24, + "grad_norm": 1.803694643769044, + "learning_rate": 1.7677746626931122e-05, + "loss": 1.0656, + "step": 10388 + }, + { + "epoch": 0.24, + "grad_norm": 2.1091967221483636, + "learning_rate": 1.7677257708942697e-05, + "loss": 1.0081, + "step": 10389 + }, + { + "epoch": 0.24, + "grad_norm": 2.3013872493475516, + "learning_rate": 1.767676874625485e-05, + "loss": 1.1086, + "step": 10390 + }, + { + "epoch": 0.24, + "grad_norm": 2.243083335651148, + "learning_rate": 1.7676279738870423e-05, + "loss": 1.0884, + "step": 10391 + }, + { + "epoch": 0.24, + "grad_norm": 1.9082288136358836, + "learning_rate": 1.7675790686792264e-05, + "loss": 1.0307, + "step": 10392 + }, + { + "epoch": 0.24, + "grad_norm": 2.1007573385019818, + "learning_rate": 1.7675301590023225e-05, + "loss": 1.0207, + "step": 10393 + }, + { + "epoch": 0.24, + "grad_norm": 2.1196670320092217, + "learning_rate": 1.7674812448566147e-05, + "loss": 1.1575, + "step": 10394 + }, + { + "epoch": 0.24, + "grad_norm": 1.9970608965178402, + "learning_rate": 1.7674323262423883e-05, + "loss": 1.2161, + "step": 10395 + }, + { + "epoch": 0.24, + "grad_norm": 2.527724043241846, + "learning_rate": 1.767383403159928e-05, + "loss": 1.0384, + "step": 10396 + }, + { + "epoch": 0.24, + "grad_norm": 1.937931532171893, + "learning_rate": 1.767334475609518e-05, + "loss": 1.0889, + "step": 10397 + }, + { + "epoch": 0.24, + "grad_norm": 2.075818633207541, + "learning_rate": 1.7672855435914442e-05, + "loss": 1.0728, + "step": 10398 + }, + { + "epoch": 0.24, + "grad_norm": 1.9855494696097165, + "learning_rate": 1.767236607105991e-05, + "loss": 1.132, + "step": 10399 + }, + { + "epoch": 0.25, + "grad_norm": 1.3261033631775605, + "learning_rate": 1.7671876661534436e-05, + "loss": 0.9874, + "step": 10400 + }, + { + "epoch": 0.25, + "grad_norm": 2.3663159427612928, + "learning_rate": 1.7671387207340864e-05, + "loss": 1.0371, + "step": 10401 + }, + { + "epoch": 0.25, + "grad_norm": 2.08055749376282, + "learning_rate": 1.767089770848205e-05, + "loss": 1.0137, + "step": 10402 + }, + { + "epoch": 0.25, + "grad_norm": 2.0998034792268485, + "learning_rate": 1.7670408164960833e-05, + "loss": 1.0789, + "step": 10403 + }, + { + "epoch": 0.25, + "grad_norm": 2.0611429949927964, + "learning_rate": 1.766991857678008e-05, + "loss": 1.0358, + "step": 10404 + }, + { + "epoch": 0.25, + "grad_norm": 1.9536040029851274, + "learning_rate": 1.766942894394263e-05, + "loss": 0.9873, + "step": 10405 + }, + { + "epoch": 0.25, + "grad_norm": 2.380886055825799, + "learning_rate": 1.7668939266451337e-05, + "loss": 1.0436, + "step": 10406 + }, + { + "epoch": 0.25, + "grad_norm": 1.9264078738450714, + "learning_rate": 1.7668449544309053e-05, + "loss": 1.0724, + "step": 10407 + }, + { + "epoch": 0.25, + "grad_norm": 1.070724901223534, + "learning_rate": 1.7667959777518622e-05, + "loss": 0.9471, + "step": 10408 + }, + { + "epoch": 0.25, + "grad_norm": 2.397555293801915, + "learning_rate": 1.7667469966082908e-05, + "loss": 1.0368, + "step": 10409 + }, + { + "epoch": 0.25, + "grad_norm": 2.0978926587791915, + "learning_rate": 1.7666980110004756e-05, + "loss": 1.0958, + "step": 10410 + }, + { + "epoch": 0.25, + "grad_norm": 2.422010227483136, + "learning_rate": 1.7666490209287015e-05, + "loss": 1.0918, + "step": 10411 + }, + { + "epoch": 0.25, + "grad_norm": 1.8984615760765933, + "learning_rate": 1.7666000263932542e-05, + "loss": 0.9908, + "step": 10412 + }, + { + "epoch": 0.25, + "grad_norm": 2.153293347135801, + "learning_rate": 1.766551027394419e-05, + "loss": 1.0724, + "step": 10413 + }, + { + "epoch": 0.25, + "grad_norm": 1.9651417952333912, + "learning_rate": 1.7665020239324808e-05, + "loss": 0.9497, + "step": 10414 + }, + { + "epoch": 0.25, + "grad_norm": 1.9630130989976193, + "learning_rate": 1.7664530160077252e-05, + "loss": 1.0026, + "step": 10415 + }, + { + "epoch": 0.25, + "grad_norm": 2.027345455105566, + "learning_rate": 1.7664040036204375e-05, + "loss": 1.0438, + "step": 10416 + }, + { + "epoch": 0.25, + "grad_norm": 1.9484494592600543, + "learning_rate": 1.766354986770903e-05, + "loss": 1.094, + "step": 10417 + }, + { + "epoch": 0.25, + "grad_norm": 2.0205508089722333, + "learning_rate": 1.7663059654594073e-05, + "loss": 1.0098, + "step": 10418 + }, + { + "epoch": 0.25, + "grad_norm": 2.263221487223924, + "learning_rate": 1.766256939686236e-05, + "loss": 1.191, + "step": 10419 + }, + { + "epoch": 0.25, + "grad_norm": 2.2048482718082414, + "learning_rate": 1.7662079094516735e-05, + "loss": 0.979, + "step": 10420 + }, + { + "epoch": 0.25, + "grad_norm": 1.9438276360531748, + "learning_rate": 1.7661588747560062e-05, + "loss": 0.9933, + "step": 10421 + }, + { + "epoch": 0.25, + "grad_norm": 2.235013882853227, + "learning_rate": 1.7661098355995195e-05, + "loss": 1.1348, + "step": 10422 + }, + { + "epoch": 0.25, + "grad_norm": 2.7543708661284585, + "learning_rate": 1.7660607919824988e-05, + "loss": 1.0449, + "step": 10423 + }, + { + "epoch": 0.25, + "grad_norm": 1.9694654895021027, + "learning_rate": 1.76601174390523e-05, + "loss": 1.0577, + "step": 10424 + }, + { + "epoch": 0.25, + "grad_norm": 2.1313219439210758, + "learning_rate": 1.7659626913679977e-05, + "loss": 0.9894, + "step": 10425 + }, + { + "epoch": 0.25, + "grad_norm": 2.338243002017926, + "learning_rate": 1.765913634371088e-05, + "loss": 1.0018, + "step": 10426 + }, + { + "epoch": 0.25, + "grad_norm": 2.212168086856814, + "learning_rate": 1.765864572914787e-05, + "loss": 0.9025, + "step": 10427 + }, + { + "epoch": 0.25, + "grad_norm": 1.1660976473914586, + "learning_rate": 1.7658155069993802e-05, + "loss": 1.0016, + "step": 10428 + }, + { + "epoch": 0.25, + "grad_norm": 1.1826380472682443, + "learning_rate": 1.7657664366251527e-05, + "loss": 0.9812, + "step": 10429 + }, + { + "epoch": 0.25, + "grad_norm": 2.9762020166963996, + "learning_rate": 1.7657173617923905e-05, + "loss": 0.9888, + "step": 10430 + }, + { + "epoch": 0.25, + "grad_norm": 2.21568528014412, + "learning_rate": 1.7656682825013794e-05, + "loss": 1.0994, + "step": 10431 + }, + { + "epoch": 0.25, + "grad_norm": 1.948327979252038, + "learning_rate": 1.7656191987524055e-05, + "loss": 1.0291, + "step": 10432 + }, + { + "epoch": 0.25, + "grad_norm": 2.832893456965383, + "learning_rate": 1.765570110545754e-05, + "loss": 1.1483, + "step": 10433 + }, + { + "epoch": 0.25, + "grad_norm": 1.0679500482619784, + "learning_rate": 1.765521017881711e-05, + "loss": 1.0076, + "step": 10434 + }, + { + "epoch": 0.25, + "grad_norm": 2.110776470224837, + "learning_rate": 1.7654719207605625e-05, + "loss": 0.8806, + "step": 10435 + }, + { + "epoch": 0.25, + "grad_norm": 1.11994138500409, + "learning_rate": 1.7654228191825937e-05, + "loss": 0.918, + "step": 10436 + }, + { + "epoch": 0.25, + "grad_norm": 3.037943846207073, + "learning_rate": 1.7653737131480914e-05, + "loss": 1.0507, + "step": 10437 + }, + { + "epoch": 0.25, + "grad_norm": 2.7295326820194865, + "learning_rate": 1.7653246026573404e-05, + "loss": 0.9679, + "step": 10438 + }, + { + "epoch": 0.25, + "grad_norm": 1.0353822456934105, + "learning_rate": 1.7652754877106275e-05, + "loss": 0.9695, + "step": 10439 + }, + { + "epoch": 0.25, + "grad_norm": 2.1872792910668304, + "learning_rate": 1.7652263683082388e-05, + "loss": 1.1235, + "step": 10440 + }, + { + "epoch": 0.25, + "grad_norm": 2.135131630710223, + "learning_rate": 1.7651772444504594e-05, + "loss": 0.9878, + "step": 10441 + }, + { + "epoch": 0.25, + "grad_norm": 2.6620734962605046, + "learning_rate": 1.765128116137576e-05, + "loss": 0.9922, + "step": 10442 + }, + { + "epoch": 0.25, + "grad_norm": 2.354179752794891, + "learning_rate": 1.765078983369875e-05, + "loss": 0.9958, + "step": 10443 + }, + { + "epoch": 0.25, + "grad_norm": 2.34290804094667, + "learning_rate": 1.765029846147641e-05, + "loss": 1.0, + "step": 10444 + }, + { + "epoch": 0.25, + "grad_norm": 2.1076417519692705, + "learning_rate": 1.764980704471162e-05, + "loss": 0.9933, + "step": 10445 + }, + { + "epoch": 0.25, + "grad_norm": 2.257814082822443, + "learning_rate": 1.7649315583407226e-05, + "loss": 1.1351, + "step": 10446 + }, + { + "epoch": 0.25, + "grad_norm": 1.863905843727422, + "learning_rate": 1.7648824077566095e-05, + "loss": 1.0809, + "step": 10447 + }, + { + "epoch": 0.25, + "grad_norm": 3.0053444843369284, + "learning_rate": 1.7648332527191086e-05, + "loss": 1.0474, + "step": 10448 + }, + { + "epoch": 0.25, + "grad_norm": 2.0760716365200795, + "learning_rate": 1.764784093228507e-05, + "loss": 1.0626, + "step": 10449 + }, + { + "epoch": 0.25, + "grad_norm": 2.2343919093152422, + "learning_rate": 1.7647349292850898e-05, + "loss": 0.8645, + "step": 10450 + }, + { + "epoch": 0.25, + "grad_norm": 2.2770881987922116, + "learning_rate": 1.7646857608891438e-05, + "loss": 1.2026, + "step": 10451 + }, + { + "epoch": 0.25, + "grad_norm": 1.9286295267620697, + "learning_rate": 1.7646365880409554e-05, + "loss": 1.1156, + "step": 10452 + }, + { + "epoch": 0.25, + "grad_norm": 1.1634060995954782, + "learning_rate": 1.7645874107408107e-05, + "loss": 1.0063, + "step": 10453 + }, + { + "epoch": 0.25, + "grad_norm": 1.1317010939898409, + "learning_rate": 1.764538228988996e-05, + "loss": 1.0277, + "step": 10454 + }, + { + "epoch": 0.25, + "grad_norm": 1.8190805890873727, + "learning_rate": 1.7644890427857975e-05, + "loss": 1.0963, + "step": 10455 + }, + { + "epoch": 0.25, + "grad_norm": 2.058992684026767, + "learning_rate": 1.764439852131502e-05, + "loss": 1.1845, + "step": 10456 + }, + { + "epoch": 0.25, + "grad_norm": 2.4133821523293615, + "learning_rate": 1.7643906570263958e-05, + "loss": 1.0577, + "step": 10457 + }, + { + "epoch": 0.25, + "grad_norm": 2.4619588344729175, + "learning_rate": 1.764341457470765e-05, + "loss": 0.9422, + "step": 10458 + }, + { + "epoch": 0.25, + "grad_norm": 2.206862499528934, + "learning_rate": 1.7642922534648962e-05, + "loss": 0.9828, + "step": 10459 + }, + { + "epoch": 0.25, + "grad_norm": 2.149786717322508, + "learning_rate": 1.7642430450090758e-05, + "loss": 1.0269, + "step": 10460 + }, + { + "epoch": 0.25, + "grad_norm": 2.5035782259201667, + "learning_rate": 1.7641938321035904e-05, + "loss": 1.0132, + "step": 10461 + }, + { + "epoch": 0.25, + "grad_norm": 2.230785204727811, + "learning_rate": 1.7641446147487267e-05, + "loss": 1.0754, + "step": 10462 + }, + { + "epoch": 0.25, + "grad_norm": 3.342863953354232, + "learning_rate": 1.7640953929447712e-05, + "loss": 1.0433, + "step": 10463 + }, + { + "epoch": 0.25, + "grad_norm": 2.1100878741643165, + "learning_rate": 1.7640461666920104e-05, + "loss": 1.0615, + "step": 10464 + }, + { + "epoch": 0.25, + "grad_norm": 2.003977067664593, + "learning_rate": 1.763996935990731e-05, + "loss": 1.0214, + "step": 10465 + }, + { + "epoch": 0.25, + "grad_norm": 2.524196967099871, + "learning_rate": 1.7639477008412192e-05, + "loss": 0.9803, + "step": 10466 + }, + { + "epoch": 0.25, + "grad_norm": 1.9771534155761588, + "learning_rate": 1.7638984612437623e-05, + "loss": 1.086, + "step": 10467 + }, + { + "epoch": 0.25, + "grad_norm": 1.996597663144635, + "learning_rate": 1.7638492171986468e-05, + "loss": 1.0613, + "step": 10468 + }, + { + "epoch": 0.25, + "grad_norm": 2.7365883408053118, + "learning_rate": 1.763799968706159e-05, + "loss": 1.0898, + "step": 10469 + }, + { + "epoch": 0.25, + "grad_norm": 1.8589506694413755, + "learning_rate": 1.7637507157665867e-05, + "loss": 0.9833, + "step": 10470 + }, + { + "epoch": 0.25, + "grad_norm": 2.7035884599405735, + "learning_rate": 1.763701458380215e-05, + "loss": 0.9753, + "step": 10471 + }, + { + "epoch": 0.25, + "grad_norm": 2.3129745693980115, + "learning_rate": 1.7636521965473324e-05, + "loss": 0.96, + "step": 10472 + }, + { + "epoch": 0.25, + "grad_norm": 1.8668141721927671, + "learning_rate": 1.7636029302682244e-05, + "loss": 1.0469, + "step": 10473 + }, + { + "epoch": 0.25, + "grad_norm": 1.113364351040427, + "learning_rate": 1.7635536595431784e-05, + "loss": 1.0395, + "step": 10474 + }, + { + "epoch": 0.25, + "grad_norm": 2.0729071096588556, + "learning_rate": 1.7635043843724817e-05, + "loss": 0.9963, + "step": 10475 + }, + { + "epoch": 0.25, + "grad_norm": 1.859984163240235, + "learning_rate": 1.7634551047564205e-05, + "loss": 1.0546, + "step": 10476 + }, + { + "epoch": 0.25, + "grad_norm": 2.078566595265179, + "learning_rate": 1.763405820695282e-05, + "loss": 1.094, + "step": 10477 + }, + { + "epoch": 0.25, + "grad_norm": 3.081726634663616, + "learning_rate": 1.7633565321893527e-05, + "loss": 1.0198, + "step": 10478 + }, + { + "epoch": 0.25, + "grad_norm": 4.107562829545346, + "learning_rate": 1.7633072392389204e-05, + "loss": 1.1269, + "step": 10479 + }, + { + "epoch": 0.25, + "grad_norm": 2.1870945376960695, + "learning_rate": 1.763257941844272e-05, + "loss": 1.0591, + "step": 10480 + }, + { + "epoch": 0.25, + "grad_norm": 2.2107344380274156, + "learning_rate": 1.7632086400056936e-05, + "loss": 1.127, + "step": 10481 + }, + { + "epoch": 0.25, + "grad_norm": 2.4821742657070365, + "learning_rate": 1.7631593337234733e-05, + "loss": 1.1015, + "step": 10482 + }, + { + "epoch": 0.25, + "grad_norm": 2.6864393270312523, + "learning_rate": 1.7631100229978977e-05, + "loss": 0.9817, + "step": 10483 + }, + { + "epoch": 0.25, + "grad_norm": 2.611349033976025, + "learning_rate": 1.763060707829254e-05, + "loss": 0.9293, + "step": 10484 + }, + { + "epoch": 0.25, + "grad_norm": 2.288245662502507, + "learning_rate": 1.7630113882178287e-05, + "loss": 1.1225, + "step": 10485 + }, + { + "epoch": 0.25, + "grad_norm": 2.2354669591133014, + "learning_rate": 1.7629620641639102e-05, + "loss": 0.9909, + "step": 10486 + }, + { + "epoch": 0.25, + "grad_norm": 2.531045406239022, + "learning_rate": 1.7629127356677847e-05, + "loss": 1.0874, + "step": 10487 + }, + { + "epoch": 0.25, + "grad_norm": 2.127803639665407, + "learning_rate": 1.7628634027297397e-05, + "loss": 1.0763, + "step": 10488 + }, + { + "epoch": 0.25, + "grad_norm": 2.051262388101157, + "learning_rate": 1.7628140653500628e-05, + "loss": 0.9786, + "step": 10489 + }, + { + "epoch": 0.25, + "grad_norm": 2.072543912591214, + "learning_rate": 1.7627647235290407e-05, + "loss": 1.0242, + "step": 10490 + }, + { + "epoch": 0.25, + "grad_norm": 2.8222258654107026, + "learning_rate": 1.762715377266961e-05, + "loss": 0.9367, + "step": 10491 + }, + { + "epoch": 0.25, + "grad_norm": 2.323618256207561, + "learning_rate": 1.7626660265641105e-05, + "loss": 1.0298, + "step": 10492 + }, + { + "epoch": 0.25, + "grad_norm": 1.9452394597194553, + "learning_rate": 1.7626166714207774e-05, + "loss": 1.0099, + "step": 10493 + }, + { + "epoch": 0.25, + "grad_norm": 2.339833857896716, + "learning_rate": 1.762567311837248e-05, + "loss": 1.0634, + "step": 10494 + }, + { + "epoch": 0.25, + "grad_norm": 2.0760227998332046, + "learning_rate": 1.762517947813811e-05, + "loss": 1.0046, + "step": 10495 + }, + { + "epoch": 0.25, + "grad_norm": 2.1665123008030633, + "learning_rate": 1.7624685793507526e-05, + "loss": 1.1399, + "step": 10496 + }, + { + "epoch": 0.25, + "grad_norm": 2.156033220777348, + "learning_rate": 1.762419206448361e-05, + "loss": 1.0658, + "step": 10497 + }, + { + "epoch": 0.25, + "grad_norm": 2.1213567516486336, + "learning_rate": 1.7623698291069235e-05, + "loss": 1.1966, + "step": 10498 + }, + { + "epoch": 0.25, + "grad_norm": 1.9531518345419898, + "learning_rate": 1.762320447326727e-05, + "loss": 1.1448, + "step": 10499 + }, + { + "epoch": 0.25, + "grad_norm": 2.05817698210781, + "learning_rate": 1.7622710611080597e-05, + "loss": 1.0068, + "step": 10500 + }, + { + "epoch": 0.25, + "grad_norm": 2.1879064914541466, + "learning_rate": 1.762221670451209e-05, + "loss": 1.0361, + "step": 10501 + }, + { + "epoch": 0.25, + "grad_norm": 1.0604399790555856, + "learning_rate": 1.7621722753564628e-05, + "loss": 0.975, + "step": 10502 + }, + { + "epoch": 0.25, + "grad_norm": 2.1250956263790393, + "learning_rate": 1.762122875824108e-05, + "loss": 1.0524, + "step": 10503 + }, + { + "epoch": 0.25, + "grad_norm": 2.5492370325878757, + "learning_rate": 1.7620734718544326e-05, + "loss": 0.9704, + "step": 10504 + }, + { + "epoch": 0.25, + "grad_norm": 2.138088055619452, + "learning_rate": 1.762024063447724e-05, + "loss": 1.0095, + "step": 10505 + }, + { + "epoch": 0.25, + "grad_norm": 1.8127659597913997, + "learning_rate": 1.76197465060427e-05, + "loss": 1.1538, + "step": 10506 + }, + { + "epoch": 0.25, + "grad_norm": 1.867951413366416, + "learning_rate": 1.7619252333243586e-05, + "loss": 1.0562, + "step": 10507 + }, + { + "epoch": 0.25, + "grad_norm": 2.477660201926206, + "learning_rate": 1.7618758116082772e-05, + "loss": 1.0352, + "step": 10508 + }, + { + "epoch": 0.25, + "grad_norm": 2.0776427236396264, + "learning_rate": 1.7618263854563137e-05, + "loss": 1.0993, + "step": 10509 + }, + { + "epoch": 0.25, + "grad_norm": 2.1452342632965644, + "learning_rate": 1.7617769548687555e-05, + "loss": 0.955, + "step": 10510 + }, + { + "epoch": 0.25, + "grad_norm": 1.861106608137076, + "learning_rate": 1.761727519845891e-05, + "loss": 1.086, + "step": 10511 + }, + { + "epoch": 0.25, + "grad_norm": 1.8856263784468963, + "learning_rate": 1.7616780803880077e-05, + "loss": 1.0459, + "step": 10512 + }, + { + "epoch": 0.25, + "grad_norm": 1.9464701013493388, + "learning_rate": 1.7616286364953935e-05, + "loss": 0.9988, + "step": 10513 + }, + { + "epoch": 0.25, + "grad_norm": 1.1662284632461424, + "learning_rate": 1.7615791881683358e-05, + "loss": 0.9911, + "step": 10514 + }, + { + "epoch": 0.25, + "grad_norm": 2.14486522421293, + "learning_rate": 1.7615297354071233e-05, + "loss": 1.0861, + "step": 10515 + }, + { + "epoch": 0.25, + "grad_norm": 2.124166778463215, + "learning_rate": 1.7614802782120433e-05, + "loss": 1.0617, + "step": 10516 + }, + { + "epoch": 0.25, + "grad_norm": 2.57434653167981, + "learning_rate": 1.7614308165833845e-05, + "loss": 0.9897, + "step": 10517 + }, + { + "epoch": 0.25, + "grad_norm": 2.116169826564177, + "learning_rate": 1.7613813505214343e-05, + "loss": 1.0145, + "step": 10518 + }, + { + "epoch": 0.25, + "grad_norm": 2.211181372433697, + "learning_rate": 1.7613318800264808e-05, + "loss": 1.0616, + "step": 10519 + }, + { + "epoch": 0.25, + "grad_norm": 2.0778870503491698, + "learning_rate": 1.761282405098812e-05, + "loss": 1.2298, + "step": 10520 + }, + { + "epoch": 0.25, + "grad_norm": 2.270010748208463, + "learning_rate": 1.7612329257387163e-05, + "loss": 1.055, + "step": 10521 + }, + { + "epoch": 0.25, + "grad_norm": 1.0674621949731304, + "learning_rate": 1.761183441946481e-05, + "loss": 0.9681, + "step": 10522 + }, + { + "epoch": 0.25, + "grad_norm": 2.2829558323828314, + "learning_rate": 1.761133953722395e-05, + "loss": 0.9455, + "step": 10523 + }, + { + "epoch": 0.25, + "grad_norm": 1.9577628902876627, + "learning_rate": 1.7610844610667463e-05, + "loss": 0.9102, + "step": 10524 + }, + { + "epoch": 0.25, + "grad_norm": 1.8892693618052472, + "learning_rate": 1.7610349639798226e-05, + "loss": 1.0412, + "step": 10525 + }, + { + "epoch": 0.25, + "grad_norm": 1.092223299105586, + "learning_rate": 1.7609854624619124e-05, + "loss": 1.0074, + "step": 10526 + }, + { + "epoch": 0.25, + "grad_norm": 2.1636478907198815, + "learning_rate": 1.7609359565133043e-05, + "loss": 0.9568, + "step": 10527 + }, + { + "epoch": 0.25, + "grad_norm": 2.230842909166029, + "learning_rate": 1.7608864461342856e-05, + "loss": 1.1471, + "step": 10528 + }, + { + "epoch": 0.25, + "grad_norm": 1.8954639534762407, + "learning_rate": 1.7608369313251453e-05, + "loss": 1.0396, + "step": 10529 + }, + { + "epoch": 0.25, + "grad_norm": 2.974962448282697, + "learning_rate": 1.7607874120861715e-05, + "loss": 0.9771, + "step": 10530 + }, + { + "epoch": 0.25, + "grad_norm": 2.244652214929209, + "learning_rate": 1.7607378884176528e-05, + "loss": 1.0787, + "step": 10531 + }, + { + "epoch": 0.25, + "grad_norm": 2.3039636163087467, + "learning_rate": 1.760688360319877e-05, + "loss": 1.0366, + "step": 10532 + }, + { + "epoch": 0.25, + "grad_norm": 1.0677507580832128, + "learning_rate": 1.760638827793133e-05, + "loss": 0.9728, + "step": 10533 + }, + { + "epoch": 0.25, + "grad_norm": 2.156681508704702, + "learning_rate": 1.7605892908377088e-05, + "loss": 1.0235, + "step": 10534 + }, + { + "epoch": 0.25, + "grad_norm": 2.0615783069253104, + "learning_rate": 1.760539749453893e-05, + "loss": 1.1367, + "step": 10535 + }, + { + "epoch": 0.25, + "grad_norm": 2.41070489632059, + "learning_rate": 1.7604902036419737e-05, + "loss": 1.0218, + "step": 10536 + }, + { + "epoch": 0.25, + "grad_norm": 1.9407705936021469, + "learning_rate": 1.76044065340224e-05, + "loss": 0.9939, + "step": 10537 + }, + { + "epoch": 0.25, + "grad_norm": 1.9342903650030128, + "learning_rate": 1.7603910987349803e-05, + "loss": 0.9659, + "step": 10538 + }, + { + "epoch": 0.25, + "grad_norm": 2.1420121214157266, + "learning_rate": 1.7603415396404822e-05, + "loss": 0.9323, + "step": 10539 + }, + { + "epoch": 0.25, + "grad_norm": 1.9535531932425556, + "learning_rate": 1.7602919761190355e-05, + "loss": 1.005, + "step": 10540 + }, + { + "epoch": 0.25, + "grad_norm": 2.041411892624548, + "learning_rate": 1.760242408170928e-05, + "loss": 1.1681, + "step": 10541 + }, + { + "epoch": 0.25, + "grad_norm": 2.2283470063270334, + "learning_rate": 1.760192835796448e-05, + "loss": 1.0321, + "step": 10542 + }, + { + "epoch": 0.25, + "grad_norm": 1.9859522493889192, + "learning_rate": 1.7601432589958856e-05, + "loss": 1.146, + "step": 10543 + }, + { + "epoch": 0.25, + "grad_norm": 2.051548774575089, + "learning_rate": 1.760093677769528e-05, + "loss": 1.0432, + "step": 10544 + }, + { + "epoch": 0.25, + "grad_norm": 1.0755975305073797, + "learning_rate": 1.7600440921176643e-05, + "loss": 0.9669, + "step": 10545 + }, + { + "epoch": 0.25, + "grad_norm": 2.3811174480492547, + "learning_rate": 1.7599945020405834e-05, + "loss": 1.0927, + "step": 10546 + }, + { + "epoch": 0.25, + "grad_norm": 2.161171120628591, + "learning_rate": 1.7599449075385736e-05, + "loss": 1.0573, + "step": 10547 + }, + { + "epoch": 0.25, + "grad_norm": 2.9977682771682193, + "learning_rate": 1.7598953086119243e-05, + "loss": 0.8987, + "step": 10548 + }, + { + "epoch": 0.25, + "grad_norm": 2.8167309334153883, + "learning_rate": 1.759845705260924e-05, + "loss": 1.1496, + "step": 10549 + }, + { + "epoch": 0.25, + "grad_norm": 2.2207743652998335, + "learning_rate": 1.759796097485861e-05, + "loss": 1.0411, + "step": 10550 + }, + { + "epoch": 0.25, + "grad_norm": 2.2250720060928106, + "learning_rate": 1.759746485287025e-05, + "loss": 1.1053, + "step": 10551 + }, + { + "epoch": 0.25, + "grad_norm": 2.087413148238568, + "learning_rate": 1.7596968686647042e-05, + "loss": 1.1027, + "step": 10552 + }, + { + "epoch": 0.25, + "grad_norm": 2.036257980465069, + "learning_rate": 1.759647247619188e-05, + "loss": 0.9564, + "step": 10553 + }, + { + "epoch": 0.25, + "grad_norm": 2.1527310328456775, + "learning_rate": 1.7595976221507647e-05, + "loss": 1.0272, + "step": 10554 + }, + { + "epoch": 0.25, + "grad_norm": 1.9999203082828114, + "learning_rate": 1.7595479922597238e-05, + "loss": 0.9944, + "step": 10555 + }, + { + "epoch": 0.25, + "grad_norm": 2.0492559359983575, + "learning_rate": 1.7594983579463537e-05, + "loss": 1.1254, + "step": 10556 + }, + { + "epoch": 0.25, + "grad_norm": 2.1626847168449186, + "learning_rate": 1.759448719210944e-05, + "loss": 0.9998, + "step": 10557 + }, + { + "epoch": 0.25, + "grad_norm": 2.1861002846108195, + "learning_rate": 1.7593990760537836e-05, + "loss": 0.9861, + "step": 10558 + }, + { + "epoch": 0.25, + "grad_norm": 1.9915970321116294, + "learning_rate": 1.759349428475161e-05, + "loss": 1.0424, + "step": 10559 + }, + { + "epoch": 0.25, + "grad_norm": 2.1568040751378725, + "learning_rate": 1.7592997764753655e-05, + "loss": 1.0418, + "step": 10560 + }, + { + "epoch": 0.25, + "grad_norm": 2.075788294610987, + "learning_rate": 1.7592501200546865e-05, + "loss": 1.0189, + "step": 10561 + }, + { + "epoch": 0.25, + "grad_norm": 2.8965248014576734, + "learning_rate": 1.759200459213413e-05, + "loss": 1.141, + "step": 10562 + }, + { + "epoch": 0.25, + "grad_norm": 2.073240811847926, + "learning_rate": 1.7591507939518342e-05, + "loss": 1.0261, + "step": 10563 + }, + { + "epoch": 0.25, + "grad_norm": 2.198215572157077, + "learning_rate": 1.7591011242702388e-05, + "loss": 1.2151, + "step": 10564 + }, + { + "epoch": 0.25, + "grad_norm": 1.9263847444742341, + "learning_rate": 1.7590514501689163e-05, + "loss": 0.8945, + "step": 10565 + }, + { + "epoch": 0.25, + "grad_norm": 2.0252366283329013, + "learning_rate": 1.759001771648156e-05, + "loss": 0.9462, + "step": 10566 + }, + { + "epoch": 0.25, + "grad_norm": 2.0821250642553446, + "learning_rate": 1.7589520887082474e-05, + "loss": 1.1003, + "step": 10567 + }, + { + "epoch": 0.25, + "grad_norm": 1.9996682021981098, + "learning_rate": 1.7589024013494796e-05, + "loss": 0.9516, + "step": 10568 + }, + { + "epoch": 0.25, + "grad_norm": 2.0594939970182145, + "learning_rate": 1.758852709572141e-05, + "loss": 1.0858, + "step": 10569 + }, + { + "epoch": 0.25, + "grad_norm": 1.8716209289644816, + "learning_rate": 1.7588030133765222e-05, + "loss": 1.0813, + "step": 10570 + }, + { + "epoch": 0.25, + "grad_norm": 2.466051890480303, + "learning_rate": 1.7587533127629123e-05, + "loss": 0.9693, + "step": 10571 + }, + { + "epoch": 0.25, + "grad_norm": 2.024186908099451, + "learning_rate": 1.7587036077315996e-05, + "loss": 1.1128, + "step": 10572 + }, + { + "epoch": 0.25, + "grad_norm": 2.3308519566702226, + "learning_rate": 1.758653898282875e-05, + "loss": 1.0872, + "step": 10573 + }, + { + "epoch": 0.25, + "grad_norm": 2.0346669571729596, + "learning_rate": 1.758604184417027e-05, + "loss": 1.2033, + "step": 10574 + }, + { + "epoch": 0.25, + "grad_norm": 2.2390406810457084, + "learning_rate": 1.7585544661343454e-05, + "loss": 1.1128, + "step": 10575 + }, + { + "epoch": 0.25, + "grad_norm": 1.1771854074880819, + "learning_rate": 1.758504743435119e-05, + "loss": 0.959, + "step": 10576 + }, + { + "epoch": 0.25, + "grad_norm": 1.8369625565284273, + "learning_rate": 1.7584550163196382e-05, + "loss": 1.144, + "step": 10577 + }, + { + "epoch": 0.25, + "grad_norm": 2.159224524931766, + "learning_rate": 1.758405284788192e-05, + "loss": 0.8763, + "step": 10578 + }, + { + "epoch": 0.25, + "grad_norm": 1.1789037759859518, + "learning_rate": 1.7583555488410703e-05, + "loss": 0.9853, + "step": 10579 + }, + { + "epoch": 0.25, + "grad_norm": 2.6543690748917825, + "learning_rate": 1.7583058084785626e-05, + "loss": 0.9691, + "step": 10580 + }, + { + "epoch": 0.25, + "grad_norm": 1.2064627248663264, + "learning_rate": 1.7582560637009584e-05, + "loss": 0.9349, + "step": 10581 + }, + { + "epoch": 0.25, + "grad_norm": 2.0872030667221964, + "learning_rate": 1.7582063145085473e-05, + "loss": 1.0427, + "step": 10582 + }, + { + "epoch": 0.25, + "grad_norm": 2.134885607577768, + "learning_rate": 1.7581565609016186e-05, + "loss": 0.9218, + "step": 10583 + }, + { + "epoch": 0.25, + "grad_norm": 1.8960791048243506, + "learning_rate": 1.758106802880463e-05, + "loss": 1.0143, + "step": 10584 + }, + { + "epoch": 0.25, + "grad_norm": 1.0873278015051517, + "learning_rate": 1.7580570404453693e-05, + "loss": 0.9309, + "step": 10585 + }, + { + "epoch": 0.25, + "grad_norm": 2.2250687668670097, + "learning_rate": 1.7580072735966275e-05, + "loss": 1.0782, + "step": 10586 + }, + { + "epoch": 0.25, + "grad_norm": 2.3917087258917125, + "learning_rate": 1.7579575023345273e-05, + "loss": 0.953, + "step": 10587 + }, + { + "epoch": 0.25, + "grad_norm": 1.0766629943464245, + "learning_rate": 1.7579077266593587e-05, + "loss": 1.0126, + "step": 10588 + }, + { + "epoch": 0.25, + "grad_norm": 2.1605107379157538, + "learning_rate": 1.757857946571411e-05, + "loss": 1.1436, + "step": 10589 + }, + { + "epoch": 0.25, + "grad_norm": 2.249564537859059, + "learning_rate": 1.7578081620709747e-05, + "loss": 1.0804, + "step": 10590 + }, + { + "epoch": 0.25, + "grad_norm": 2.1174790935549717, + "learning_rate": 1.7577583731583396e-05, + "loss": 1.0333, + "step": 10591 + }, + { + "epoch": 0.25, + "grad_norm": 3.2598370159404437, + "learning_rate": 1.757708579833795e-05, + "loss": 1.015, + "step": 10592 + }, + { + "epoch": 0.25, + "grad_norm": 1.8507529350711764, + "learning_rate": 1.7576587820976315e-05, + "loss": 1.0195, + "step": 10593 + }, + { + "epoch": 0.25, + "grad_norm": 2.003553009990619, + "learning_rate": 1.7576089799501384e-05, + "loss": 0.9569, + "step": 10594 + }, + { + "epoch": 0.25, + "grad_norm": 3.630379241050449, + "learning_rate": 1.757559173391606e-05, + "loss": 1.0182, + "step": 10595 + }, + { + "epoch": 0.25, + "grad_norm": 1.9825703148397338, + "learning_rate": 1.7575093624223243e-05, + "loss": 1.098, + "step": 10596 + }, + { + "epoch": 0.25, + "grad_norm": 2.1732093439322773, + "learning_rate": 1.757459547042583e-05, + "loss": 0.9584, + "step": 10597 + }, + { + "epoch": 0.25, + "grad_norm": 2.433425601449306, + "learning_rate": 1.757409727252673e-05, + "loss": 0.976, + "step": 10598 + }, + { + "epoch": 0.25, + "grad_norm": 2.029552711380613, + "learning_rate": 1.7573599030528834e-05, + "loss": 1.0674, + "step": 10599 + }, + { + "epoch": 0.25, + "grad_norm": 2.257065429970066, + "learning_rate": 1.7573100744435044e-05, + "loss": 0.9302, + "step": 10600 + }, + { + "epoch": 0.25, + "grad_norm": 2.1020662147119333, + "learning_rate": 1.757260241424827e-05, + "loss": 1.0511, + "step": 10601 + }, + { + "epoch": 0.25, + "grad_norm": 1.9419625891796404, + "learning_rate": 1.7572104039971405e-05, + "loss": 1.0057, + "step": 10602 + }, + { + "epoch": 0.25, + "grad_norm": 1.9660905568403053, + "learning_rate": 1.757160562160735e-05, + "loss": 0.9815, + "step": 10603 + }, + { + "epoch": 0.25, + "grad_norm": 2.149035495851464, + "learning_rate": 1.7571107159159013e-05, + "loss": 1.0773, + "step": 10604 + }, + { + "epoch": 0.25, + "grad_norm": 1.9985558305813176, + "learning_rate": 1.757060865262929e-05, + "loss": 1.0805, + "step": 10605 + }, + { + "epoch": 0.25, + "grad_norm": 2.3826382599700944, + "learning_rate": 1.757011010202109e-05, + "loss": 1.0652, + "step": 10606 + }, + { + "epoch": 0.25, + "grad_norm": 1.1023367194766642, + "learning_rate": 1.756961150733731e-05, + "loss": 1.03, + "step": 10607 + }, + { + "epoch": 0.25, + "grad_norm": 3.283475123305788, + "learning_rate": 1.756911286858086e-05, + "loss": 1.0416, + "step": 10608 + }, + { + "epoch": 0.25, + "grad_norm": 2.368546273564232, + "learning_rate": 1.7568614185754635e-05, + "loss": 0.9749, + "step": 10609 + }, + { + "epoch": 0.25, + "grad_norm": 2.0425819722396703, + "learning_rate": 1.7568115458861542e-05, + "loss": 1.1276, + "step": 10610 + }, + { + "epoch": 0.25, + "grad_norm": 2.133057603249688, + "learning_rate": 1.7567616687904484e-05, + "loss": 1.0202, + "step": 10611 + }, + { + "epoch": 0.25, + "grad_norm": 1.858762150920178, + "learning_rate": 1.7567117872886368e-05, + "loss": 0.932, + "step": 10612 + }, + { + "epoch": 0.25, + "grad_norm": 2.5010148359764037, + "learning_rate": 1.7566619013810094e-05, + "loss": 1.0595, + "step": 10613 + }, + { + "epoch": 0.25, + "grad_norm": 1.935748183399237, + "learning_rate": 1.7566120110678572e-05, + "loss": 1.0357, + "step": 10614 + }, + { + "epoch": 0.25, + "grad_norm": 1.1486944240462775, + "learning_rate": 1.7565621163494697e-05, + "loss": 1.0013, + "step": 10615 + }, + { + "epoch": 0.25, + "grad_norm": 2.359329774492984, + "learning_rate": 1.7565122172261384e-05, + "loss": 1.0345, + "step": 10616 + }, + { + "epoch": 0.25, + "grad_norm": 1.8342227632322454, + "learning_rate": 1.7564623136981532e-05, + "loss": 1.0082, + "step": 10617 + }, + { + "epoch": 0.25, + "grad_norm": 1.3003401310826856, + "learning_rate": 1.7564124057658057e-05, + "loss": 1.0198, + "step": 10618 + }, + { + "epoch": 0.25, + "grad_norm": 2.227716386028186, + "learning_rate": 1.756362493429385e-05, + "loss": 0.9391, + "step": 10619 + }, + { + "epoch": 0.25, + "grad_norm": 2.026162528652987, + "learning_rate": 1.7563125766891826e-05, + "loss": 1.0604, + "step": 10620 + }, + { + "epoch": 0.25, + "grad_norm": 1.9263370638662503, + "learning_rate": 1.7562626555454887e-05, + "loss": 1.0031, + "step": 10621 + }, + { + "epoch": 0.25, + "grad_norm": 1.9464495367697618, + "learning_rate": 1.7562127299985944e-05, + "loss": 1.025, + "step": 10622 + }, + { + "epoch": 0.25, + "grad_norm": 2.164280974208502, + "learning_rate": 1.7561628000487902e-05, + "loss": 1.0834, + "step": 10623 + }, + { + "epoch": 0.25, + "grad_norm": 1.942003275733061, + "learning_rate": 1.7561128656963667e-05, + "loss": 1.0965, + "step": 10624 + }, + { + "epoch": 0.25, + "grad_norm": 2.5003986812349877, + "learning_rate": 1.756062926941615e-05, + "loss": 1.0533, + "step": 10625 + }, + { + "epoch": 0.25, + "grad_norm": 2.3912969693073385, + "learning_rate": 1.7560129837848255e-05, + "loss": 1.0299, + "step": 10626 + }, + { + "epoch": 0.25, + "grad_norm": 2.053526185060981, + "learning_rate": 1.755963036226289e-05, + "loss": 1.0018, + "step": 10627 + }, + { + "epoch": 0.25, + "grad_norm": 1.994381359350532, + "learning_rate": 1.755913084266296e-05, + "loss": 1.0905, + "step": 10628 + }, + { + "epoch": 0.25, + "grad_norm": 2.043843822037296, + "learning_rate": 1.755863127905138e-05, + "loss": 1.026, + "step": 10629 + }, + { + "epoch": 0.25, + "grad_norm": 1.2119047627949073, + "learning_rate": 1.7558131671431056e-05, + "loss": 1.0293, + "step": 10630 + }, + { + "epoch": 0.25, + "grad_norm": 2.247000772568123, + "learning_rate": 1.7557632019804894e-05, + "loss": 0.9702, + "step": 10631 + }, + { + "epoch": 0.25, + "grad_norm": 2.2551641432258216, + "learning_rate": 1.755713232417581e-05, + "loss": 1.0713, + "step": 10632 + }, + { + "epoch": 0.25, + "grad_norm": 1.9576578681328498, + "learning_rate": 1.7556632584546705e-05, + "loss": 1.0039, + "step": 10633 + }, + { + "epoch": 0.25, + "grad_norm": 1.204949129007613, + "learning_rate": 1.7556132800920494e-05, + "loss": 1.0201, + "step": 10634 + }, + { + "epoch": 0.25, + "grad_norm": 1.1335757469292875, + "learning_rate": 1.7555632973300087e-05, + "loss": 0.9547, + "step": 10635 + }, + { + "epoch": 0.25, + "grad_norm": 2.1579771253658615, + "learning_rate": 1.755513310168839e-05, + "loss": 1.131, + "step": 10636 + }, + { + "epoch": 0.25, + "grad_norm": 2.0724066100757503, + "learning_rate": 1.755463318608832e-05, + "loss": 0.8946, + "step": 10637 + }, + { + "epoch": 0.25, + "grad_norm": 1.9185951299229014, + "learning_rate": 1.7554133226502782e-05, + "loss": 0.9907, + "step": 10638 + }, + { + "epoch": 0.25, + "grad_norm": 2.0242573612583863, + "learning_rate": 1.755363322293469e-05, + "loss": 1.0975, + "step": 10639 + }, + { + "epoch": 0.25, + "grad_norm": 2.2240304711740597, + "learning_rate": 1.755313317538695e-05, + "loss": 1.0053, + "step": 10640 + }, + { + "epoch": 0.25, + "grad_norm": 2.0055239624736525, + "learning_rate": 1.755263308386248e-05, + "loss": 1.0072, + "step": 10641 + }, + { + "epoch": 0.25, + "grad_norm": 1.1309327633339672, + "learning_rate": 1.755213294836419e-05, + "loss": 1.0016, + "step": 10642 + }, + { + "epoch": 0.25, + "grad_norm": 2.091869500802354, + "learning_rate": 1.755163276889499e-05, + "loss": 0.9713, + "step": 10643 + }, + { + "epoch": 0.25, + "grad_norm": 1.1264547648100918, + "learning_rate": 1.7551132545457793e-05, + "loss": 0.8907, + "step": 10644 + }, + { + "epoch": 0.25, + "grad_norm": 2.0334372621533623, + "learning_rate": 1.7550632278055514e-05, + "loss": 0.9708, + "step": 10645 + }, + { + "epoch": 0.25, + "grad_norm": 1.988307550931785, + "learning_rate": 1.7550131966691057e-05, + "loss": 1.1353, + "step": 10646 + }, + { + "epoch": 0.25, + "grad_norm": 2.315864463745451, + "learning_rate": 1.7549631611367345e-05, + "loss": 1.0613, + "step": 10647 + }, + { + "epoch": 0.25, + "grad_norm": 2.103572606932735, + "learning_rate": 1.7549131212087288e-05, + "loss": 0.8959, + "step": 10648 + }, + { + "epoch": 0.25, + "grad_norm": 2.577399379222237, + "learning_rate": 1.75486307688538e-05, + "loss": 1.2561, + "step": 10649 + }, + { + "epoch": 0.25, + "grad_norm": 2.0926056544074076, + "learning_rate": 1.7548130281669793e-05, + "loss": 1.0407, + "step": 10650 + }, + { + "epoch": 0.25, + "grad_norm": 1.994089143202466, + "learning_rate": 1.7547629750538182e-05, + "loss": 1.1198, + "step": 10651 + }, + { + "epoch": 0.25, + "grad_norm": 6.85099469961765, + "learning_rate": 1.7547129175461882e-05, + "loss": 1.0833, + "step": 10652 + }, + { + "epoch": 0.25, + "grad_norm": 2.040833551782114, + "learning_rate": 1.7546628556443806e-05, + "loss": 0.9557, + "step": 10653 + }, + { + "epoch": 0.25, + "grad_norm": 1.8206072255857464, + "learning_rate": 1.754612789348687e-05, + "loss": 0.9473, + "step": 10654 + }, + { + "epoch": 0.25, + "grad_norm": 1.1562325613818738, + "learning_rate": 1.7545627186593986e-05, + "loss": 1.0253, + "step": 10655 + }, + { + "epoch": 0.25, + "grad_norm": 1.9207961011441983, + "learning_rate": 1.754512643576807e-05, + "loss": 0.9774, + "step": 10656 + }, + { + "epoch": 0.25, + "grad_norm": 2.095835906554673, + "learning_rate": 1.7544625641012044e-05, + "loss": 1.1052, + "step": 10657 + }, + { + "epoch": 0.25, + "grad_norm": 1.764293575732049, + "learning_rate": 1.7544124802328814e-05, + "loss": 1.0626, + "step": 10658 + }, + { + "epoch": 0.25, + "grad_norm": 1.996518349908888, + "learning_rate": 1.7543623919721305e-05, + "loss": 1.0099, + "step": 10659 + }, + { + "epoch": 0.25, + "grad_norm": 2.510813931920503, + "learning_rate": 1.754312299319243e-05, + "loss": 1.0407, + "step": 10660 + }, + { + "epoch": 0.25, + "grad_norm": 1.9150852690680449, + "learning_rate": 1.7542622022745102e-05, + "loss": 0.992, + "step": 10661 + }, + { + "epoch": 0.25, + "grad_norm": 2.278230633464764, + "learning_rate": 1.754212100838224e-05, + "loss": 1.1668, + "step": 10662 + }, + { + "epoch": 0.25, + "grad_norm": 2.558154673247854, + "learning_rate": 1.7541619950106764e-05, + "loss": 1.1429, + "step": 10663 + }, + { + "epoch": 0.25, + "grad_norm": 2.0979682752563864, + "learning_rate": 1.7541118847921586e-05, + "loss": 1.0548, + "step": 10664 + }, + { + "epoch": 0.25, + "grad_norm": 2.2650426153102656, + "learning_rate": 1.754061770182963e-05, + "loss": 1.0436, + "step": 10665 + }, + { + "epoch": 0.25, + "grad_norm": 2.5165408544023697, + "learning_rate": 1.754011651183381e-05, + "loss": 0.9176, + "step": 10666 + }, + { + "epoch": 0.25, + "grad_norm": 1.9093778275700655, + "learning_rate": 1.7539615277937045e-05, + "loss": 1.0802, + "step": 10667 + }, + { + "epoch": 0.25, + "grad_norm": 1.9903129969454516, + "learning_rate": 1.753911400014225e-05, + "loss": 1.0405, + "step": 10668 + }, + { + "epoch": 0.25, + "grad_norm": 2.73375152138456, + "learning_rate": 1.753861267845235e-05, + "loss": 0.9896, + "step": 10669 + }, + { + "epoch": 0.25, + "grad_norm": 1.8481697996672233, + "learning_rate": 1.7538111312870255e-05, + "loss": 1.0381, + "step": 10670 + }, + { + "epoch": 0.25, + "grad_norm": 2.16043683821094, + "learning_rate": 1.7537609903398895e-05, + "loss": 1.1436, + "step": 10671 + }, + { + "epoch": 0.25, + "grad_norm": 3.0035855524513537, + "learning_rate": 1.7537108450041182e-05, + "loss": 1.0254, + "step": 10672 + }, + { + "epoch": 0.25, + "grad_norm": 1.1116360744080835, + "learning_rate": 1.7536606952800034e-05, + "loss": 0.9402, + "step": 10673 + }, + { + "epoch": 0.25, + "grad_norm": 1.9180478857334982, + "learning_rate": 1.753610541167838e-05, + "loss": 1.0564, + "step": 10674 + }, + { + "epoch": 0.25, + "grad_norm": 2.0217497589226134, + "learning_rate": 1.7535603826679132e-05, + "loss": 1.0156, + "step": 10675 + }, + { + "epoch": 0.25, + "grad_norm": 1.9294939125019344, + "learning_rate": 1.7535102197805212e-05, + "loss": 0.853, + "step": 10676 + }, + { + "epoch": 0.25, + "grad_norm": 1.9295210537284657, + "learning_rate": 1.753460052505954e-05, + "loss": 0.9422, + "step": 10677 + }, + { + "epoch": 0.25, + "grad_norm": 2.1329122362161193, + "learning_rate": 1.753409880844504e-05, + "loss": 1.1308, + "step": 10678 + }, + { + "epoch": 0.25, + "grad_norm": 2.429052550043073, + "learning_rate": 1.7533597047964633e-05, + "loss": 1.0348, + "step": 10679 + }, + { + "epoch": 0.25, + "grad_norm": 2.0888067273962143, + "learning_rate": 1.7533095243621237e-05, + "loss": 1.0665, + "step": 10680 + }, + { + "epoch": 0.25, + "grad_norm": 1.932475033296009, + "learning_rate": 1.7532593395417775e-05, + "loss": 1.1607, + "step": 10681 + }, + { + "epoch": 0.25, + "grad_norm": 2.2394149189056605, + "learning_rate": 1.7532091503357172e-05, + "loss": 1.0456, + "step": 10682 + }, + { + "epoch": 0.25, + "grad_norm": 1.8449415110835001, + "learning_rate": 1.7531589567442347e-05, + "loss": 1.0492, + "step": 10683 + }, + { + "epoch": 0.25, + "grad_norm": 2.326683998636208, + "learning_rate": 1.7531087587676223e-05, + "loss": 1.0216, + "step": 10684 + }, + { + "epoch": 0.25, + "grad_norm": 2.0055559938984007, + "learning_rate": 1.753058556406172e-05, + "loss": 1.0201, + "step": 10685 + }, + { + "epoch": 0.25, + "grad_norm": 2.1188582432172036, + "learning_rate": 1.7530083496601763e-05, + "loss": 1.0743, + "step": 10686 + }, + { + "epoch": 0.25, + "grad_norm": 3.045926780972416, + "learning_rate": 1.7529581385299283e-05, + "loss": 1.1639, + "step": 10687 + }, + { + "epoch": 0.25, + "grad_norm": 2.3818824197738806, + "learning_rate": 1.7529079230157192e-05, + "loss": 1.0855, + "step": 10688 + }, + { + "epoch": 0.25, + "grad_norm": 2.03344399480228, + "learning_rate": 1.7528577031178416e-05, + "loss": 0.9382, + "step": 10689 + }, + { + "epoch": 0.25, + "grad_norm": 1.9694567822089835, + "learning_rate": 1.7528074788365885e-05, + "loss": 1.23, + "step": 10690 + }, + { + "epoch": 0.25, + "grad_norm": 2.0058877358154175, + "learning_rate": 1.7527572501722516e-05, + "loss": 1.1631, + "step": 10691 + }, + { + "epoch": 0.25, + "grad_norm": 2.237419143702097, + "learning_rate": 1.7527070171251236e-05, + "loss": 0.9823, + "step": 10692 + }, + { + "epoch": 0.25, + "grad_norm": 1.865777471376565, + "learning_rate": 1.7526567796954973e-05, + "loss": 0.9786, + "step": 10693 + }, + { + "epoch": 0.25, + "grad_norm": 2.416970249037928, + "learning_rate": 1.752606537883665e-05, + "loss": 1.0358, + "step": 10694 + }, + { + "epoch": 0.25, + "grad_norm": 1.8885962766756441, + "learning_rate": 1.752556291689919e-05, + "loss": 1.0937, + "step": 10695 + }, + { + "epoch": 0.25, + "grad_norm": 1.7750840122959615, + "learning_rate": 1.752506041114552e-05, + "loss": 1.0042, + "step": 10696 + }, + { + "epoch": 0.25, + "grad_norm": 2.2021035473174946, + "learning_rate": 1.7524557861578567e-05, + "loss": 1.1058, + "step": 10697 + }, + { + "epoch": 0.25, + "grad_norm": 2.1441540335294857, + "learning_rate": 1.7524055268201254e-05, + "loss": 1.0352, + "step": 10698 + }, + { + "epoch": 0.25, + "grad_norm": 2.3517800756011655, + "learning_rate": 1.752355263101651e-05, + "loss": 1.0454, + "step": 10699 + }, + { + "epoch": 0.25, + "grad_norm": 2.032599646799692, + "learning_rate": 1.7523049950027263e-05, + "loss": 1.1445, + "step": 10700 + }, + { + "epoch": 0.25, + "grad_norm": 1.9775830128577725, + "learning_rate": 1.7522547225236434e-05, + "loss": 1.1662, + "step": 10701 + }, + { + "epoch": 0.25, + "grad_norm": 1.9130252725797823, + "learning_rate": 1.7522044456646957e-05, + "loss": 1.1562, + "step": 10702 + }, + { + "epoch": 0.25, + "grad_norm": 2.0779195935871604, + "learning_rate": 1.7521541644261754e-05, + "loss": 1.155, + "step": 10703 + }, + { + "epoch": 0.25, + "grad_norm": 1.9623989255637615, + "learning_rate": 1.7521038788083755e-05, + "loss": 1.0254, + "step": 10704 + }, + { + "epoch": 0.25, + "grad_norm": 1.9819744764464167, + "learning_rate": 1.7520535888115887e-05, + "loss": 1.0535, + "step": 10705 + }, + { + "epoch": 0.25, + "grad_norm": 2.2362688079613324, + "learning_rate": 1.7520032944361074e-05, + "loss": 1.0122, + "step": 10706 + }, + { + "epoch": 0.25, + "grad_norm": 2.446187725461845, + "learning_rate": 1.751952995682225e-05, + "loss": 0.9664, + "step": 10707 + }, + { + "epoch": 0.25, + "grad_norm": 1.945692275215598, + "learning_rate": 1.7519026925502344e-05, + "loss": 1.0503, + "step": 10708 + }, + { + "epoch": 0.25, + "grad_norm": 1.9816011881725029, + "learning_rate": 1.7518523850404285e-05, + "loss": 1.1041, + "step": 10709 + }, + { + "epoch": 0.25, + "grad_norm": 2.1719911977546733, + "learning_rate": 1.7518020731530998e-05, + "loss": 1.1707, + "step": 10710 + }, + { + "epoch": 0.25, + "grad_norm": 1.996399942520444, + "learning_rate": 1.7517517568885413e-05, + "loss": 0.9229, + "step": 10711 + }, + { + "epoch": 0.25, + "grad_norm": 2.066577810510243, + "learning_rate": 1.751701436247046e-05, + "loss": 1.0837, + "step": 10712 + }, + { + "epoch": 0.25, + "grad_norm": 2.012837052849223, + "learning_rate": 1.7516511112289068e-05, + "loss": 1.0992, + "step": 10713 + }, + { + "epoch": 0.25, + "grad_norm": 2.1155700569135774, + "learning_rate": 1.751600781834417e-05, + "loss": 1.0023, + "step": 10714 + }, + { + "epoch": 0.25, + "grad_norm": 1.9311742477525102, + "learning_rate": 1.75155044806387e-05, + "loss": 1.1033, + "step": 10715 + }, + { + "epoch": 0.25, + "grad_norm": 1.0992076868054286, + "learning_rate": 1.7515001099175578e-05, + "loss": 0.9616, + "step": 10716 + }, + { + "epoch": 0.25, + "grad_norm": 2.379067766644611, + "learning_rate": 1.751449767395774e-05, + "loss": 1.0047, + "step": 10717 + }, + { + "epoch": 0.25, + "grad_norm": 3.3492608745440515, + "learning_rate": 1.751399420498812e-05, + "loss": 1.2373, + "step": 10718 + }, + { + "epoch": 0.25, + "grad_norm": 1.9999772120365908, + "learning_rate": 1.751349069226965e-05, + "loss": 1.0028, + "step": 10719 + }, + { + "epoch": 0.25, + "grad_norm": 2.2440481343229193, + "learning_rate": 1.7512987135805254e-05, + "loss": 1.0733, + "step": 10720 + }, + { + "epoch": 0.25, + "grad_norm": 2.439426700003769, + "learning_rate": 1.7512483535597868e-05, + "loss": 0.954, + "step": 10721 + }, + { + "epoch": 0.25, + "grad_norm": 1.967845208406824, + "learning_rate": 1.751197989165043e-05, + "loss": 1.2812, + "step": 10722 + }, + { + "epoch": 0.25, + "grad_norm": 2.089464443313059, + "learning_rate": 1.7511476203965857e-05, + "loss": 1.0242, + "step": 10723 + }, + { + "epoch": 0.25, + "grad_norm": 2.328719694544154, + "learning_rate": 1.75109724725471e-05, + "loss": 1.2584, + "step": 10724 + }, + { + "epoch": 0.25, + "grad_norm": 2.01828806647626, + "learning_rate": 1.751046869739708e-05, + "loss": 1.1345, + "step": 10725 + }, + { + "epoch": 0.25, + "grad_norm": 2.4469323896097266, + "learning_rate": 1.7509964878518736e-05, + "loss": 1.0491, + "step": 10726 + }, + { + "epoch": 0.25, + "grad_norm": 1.9813296609557993, + "learning_rate": 1.7509461015914992e-05, + "loss": 0.9801, + "step": 10727 + }, + { + "epoch": 0.25, + "grad_norm": 2.478450683737973, + "learning_rate": 1.7508957109588794e-05, + "loss": 1.1007, + "step": 10728 + }, + { + "epoch": 0.25, + "grad_norm": 2.163853534531804, + "learning_rate": 1.7508453159543072e-05, + "loss": 1.0331, + "step": 10729 + }, + { + "epoch": 0.25, + "grad_norm": 1.9666115225748932, + "learning_rate": 1.7507949165780753e-05, + "loss": 1.1526, + "step": 10730 + }, + { + "epoch": 0.25, + "grad_norm": 2.4027143174174443, + "learning_rate": 1.750744512830478e-05, + "loss": 1.0392, + "step": 10731 + }, + { + "epoch": 0.25, + "grad_norm": 2.597020526721709, + "learning_rate": 1.7506941047118084e-05, + "loss": 0.9833, + "step": 10732 + }, + { + "epoch": 0.25, + "grad_norm": 1.1268275790553008, + "learning_rate": 1.7506436922223602e-05, + "loss": 1.0012, + "step": 10733 + }, + { + "epoch": 0.25, + "grad_norm": 2.222385289309985, + "learning_rate": 1.750593275362427e-05, + "loss": 0.9659, + "step": 10734 + }, + { + "epoch": 0.25, + "grad_norm": 2.041171514453258, + "learning_rate": 1.750542854132302e-05, + "loss": 1.0148, + "step": 10735 + }, + { + "epoch": 0.25, + "grad_norm": 1.2097749417994446, + "learning_rate": 1.7504924285322782e-05, + "loss": 1.0207, + "step": 10736 + }, + { + "epoch": 0.25, + "grad_norm": 2.402207226382397, + "learning_rate": 1.7504419985626504e-05, + "loss": 0.9307, + "step": 10737 + }, + { + "epoch": 0.25, + "grad_norm": 2.06605902625055, + "learning_rate": 1.7503915642237116e-05, + "loss": 1.1447, + "step": 10738 + }, + { + "epoch": 0.25, + "grad_norm": 1.216199404050884, + "learning_rate": 1.7503411255157558e-05, + "loss": 1.0142, + "step": 10739 + }, + { + "epoch": 0.25, + "grad_norm": 2.329787128344677, + "learning_rate": 1.750290682439076e-05, + "loss": 1.1113, + "step": 10740 + }, + { + "epoch": 0.25, + "grad_norm": 2.2828343152079906, + "learning_rate": 1.7502402349939667e-05, + "loss": 0.9845, + "step": 10741 + }, + { + "epoch": 0.25, + "grad_norm": 2.53935686642547, + "learning_rate": 1.750189783180721e-05, + "loss": 1.0392, + "step": 10742 + }, + { + "epoch": 0.25, + "grad_norm": 2.2493478764728296, + "learning_rate": 1.7501393269996332e-05, + "loss": 0.9446, + "step": 10743 + }, + { + "epoch": 0.25, + "grad_norm": 1.9812949858222857, + "learning_rate": 1.7500888664509962e-05, + "loss": 1.055, + "step": 10744 + }, + { + "epoch": 0.25, + "grad_norm": 2.0133788291493366, + "learning_rate": 1.7500384015351048e-05, + "loss": 1.0554, + "step": 10745 + }, + { + "epoch": 0.25, + "grad_norm": 2.0981790366319446, + "learning_rate": 1.7499879322522524e-05, + "loss": 1.0777, + "step": 10746 + }, + { + "epoch": 0.25, + "grad_norm": 1.142756435855081, + "learning_rate": 1.7499374586027323e-05, + "loss": 0.9672, + "step": 10747 + }, + { + "epoch": 0.25, + "grad_norm": 2.823185226244508, + "learning_rate": 1.7498869805868394e-05, + "loss": 0.9442, + "step": 10748 + }, + { + "epoch": 0.25, + "grad_norm": 1.930674179163521, + "learning_rate": 1.749836498204867e-05, + "loss": 1.0259, + "step": 10749 + }, + { + "epoch": 0.25, + "grad_norm": 2.4220085061099748, + "learning_rate": 1.749786011457109e-05, + "loss": 1.0412, + "step": 10750 + }, + { + "epoch": 0.25, + "grad_norm": 2.128302692898056, + "learning_rate": 1.7497355203438594e-05, + "loss": 1.0649, + "step": 10751 + }, + { + "epoch": 0.25, + "grad_norm": 2.223660156126276, + "learning_rate": 1.7496850248654123e-05, + "loss": 1.0434, + "step": 10752 + }, + { + "epoch": 0.25, + "grad_norm": 2.5460838163316097, + "learning_rate": 1.7496345250220617e-05, + "loss": 1.1142, + "step": 10753 + }, + { + "epoch": 0.25, + "grad_norm": 2.6214390307335855, + "learning_rate": 1.7495840208141015e-05, + "loss": 0.9751, + "step": 10754 + }, + { + "epoch": 0.25, + "grad_norm": 2.0967750495426207, + "learning_rate": 1.749533512241826e-05, + "loss": 1.1338, + "step": 10755 + }, + { + "epoch": 0.25, + "grad_norm": 2.1264326489504524, + "learning_rate": 1.7494829993055288e-05, + "loss": 1.0812, + "step": 10756 + }, + { + "epoch": 0.25, + "grad_norm": 2.0630289542686224, + "learning_rate": 1.7494324820055045e-05, + "loss": 1.0178, + "step": 10757 + }, + { + "epoch": 0.25, + "grad_norm": 1.9824211492912955, + "learning_rate": 1.7493819603420468e-05, + "loss": 1.1714, + "step": 10758 + }, + { + "epoch": 0.25, + "grad_norm": 3.216309415094118, + "learning_rate": 1.7493314343154503e-05, + "loss": 1.0706, + "step": 10759 + }, + { + "epoch": 0.25, + "grad_norm": 1.8539326719152647, + "learning_rate": 1.7492809039260086e-05, + "loss": 0.9584, + "step": 10760 + }, + { + "epoch": 0.25, + "grad_norm": 1.8015620615051693, + "learning_rate": 1.7492303691740166e-05, + "loss": 0.9785, + "step": 10761 + }, + { + "epoch": 0.25, + "grad_norm": 2.3265325254076332, + "learning_rate": 1.7491798300597684e-05, + "loss": 1.0609, + "step": 10762 + }, + { + "epoch": 0.25, + "grad_norm": 1.9263824512589978, + "learning_rate": 1.7491292865835576e-05, + "loss": 1.0338, + "step": 10763 + }, + { + "epoch": 0.25, + "grad_norm": 2.484419586498304, + "learning_rate": 1.749078738745679e-05, + "loss": 1.0089, + "step": 10764 + }, + { + "epoch": 0.25, + "grad_norm": 1.1241707870949968, + "learning_rate": 1.7490281865464268e-05, + "loss": 0.9475, + "step": 10765 + }, + { + "epoch": 0.25, + "grad_norm": 2.6350481574298565, + "learning_rate": 1.748977629986096e-05, + "loss": 1.0051, + "step": 10766 + }, + { + "epoch": 0.25, + "grad_norm": 2.0158807112703716, + "learning_rate": 1.7489270690649794e-05, + "loss": 1.0045, + "step": 10767 + }, + { + "epoch": 0.25, + "grad_norm": 1.9349942489021523, + "learning_rate": 1.748876503783373e-05, + "loss": 1.1928, + "step": 10768 + }, + { + "epoch": 0.25, + "grad_norm": 2.0146361062638176, + "learning_rate": 1.7488259341415702e-05, + "loss": 1.0646, + "step": 10769 + }, + { + "epoch": 0.25, + "grad_norm": 2.182376249933842, + "learning_rate": 1.7487753601398655e-05, + "loss": 0.9476, + "step": 10770 + }, + { + "epoch": 0.25, + "grad_norm": 2.0651284599576503, + "learning_rate": 1.748724781778554e-05, + "loss": 1.0912, + "step": 10771 + }, + { + "epoch": 0.25, + "grad_norm": 2.2151120200409613, + "learning_rate": 1.7486741990579298e-05, + "loss": 1.2574, + "step": 10772 + }, + { + "epoch": 0.25, + "grad_norm": 1.121948224132954, + "learning_rate": 1.748623611978287e-05, + "loss": 0.9342, + "step": 10773 + }, + { + "epoch": 0.25, + "grad_norm": 2.2894185652033876, + "learning_rate": 1.7485730205399208e-05, + "loss": 1.015, + "step": 10774 + }, + { + "epoch": 0.25, + "grad_norm": 15.844040395899977, + "learning_rate": 1.7485224247431255e-05, + "loss": 0.9514, + "step": 10775 + }, + { + "epoch": 0.25, + "grad_norm": 2.1162111781190136, + "learning_rate": 1.7484718245881957e-05, + "loss": 1.0921, + "step": 10776 + }, + { + "epoch": 0.25, + "grad_norm": 2.077653874955129, + "learning_rate": 1.7484212200754256e-05, + "loss": 1.0574, + "step": 10777 + }, + { + "epoch": 0.25, + "grad_norm": 1.0497886890478692, + "learning_rate": 1.7483706112051105e-05, + "loss": 0.9681, + "step": 10778 + }, + { + "epoch": 0.25, + "grad_norm": 1.9337344521809159, + "learning_rate": 1.7483199979775447e-05, + "loss": 1.0729, + "step": 10779 + }, + { + "epoch": 0.25, + "grad_norm": 1.882099942750617, + "learning_rate": 1.7482693803930228e-05, + "loss": 1.1218, + "step": 10780 + }, + { + "epoch": 0.25, + "grad_norm": 1.0906965035850018, + "learning_rate": 1.7482187584518397e-05, + "loss": 0.9901, + "step": 10781 + }, + { + "epoch": 0.25, + "grad_norm": 2.1976692691499387, + "learning_rate": 1.7481681321542906e-05, + "loss": 0.9127, + "step": 10782 + }, + { + "epoch": 0.25, + "grad_norm": 2.0141929440575073, + "learning_rate": 1.7481175015006693e-05, + "loss": 0.9822, + "step": 10783 + }, + { + "epoch": 0.25, + "grad_norm": 2.1905484365043217, + "learning_rate": 1.748066866491271e-05, + "loss": 0.9887, + "step": 10784 + }, + { + "epoch": 0.25, + "grad_norm": 3.632769206260793, + "learning_rate": 1.748016227126391e-05, + "loss": 1.0491, + "step": 10785 + }, + { + "epoch": 0.25, + "grad_norm": 2.341131422004149, + "learning_rate": 1.7479655834063235e-05, + "loss": 0.9658, + "step": 10786 + }, + { + "epoch": 0.25, + "grad_norm": 2.554178366359913, + "learning_rate": 1.7479149353313635e-05, + "loss": 1.0888, + "step": 10787 + }, + { + "epoch": 0.25, + "grad_norm": 2.6425820623398097, + "learning_rate": 1.747864282901806e-05, + "loss": 0.9124, + "step": 10788 + }, + { + "epoch": 0.25, + "grad_norm": 2.228448999697808, + "learning_rate": 1.7478136261179458e-05, + "loss": 1.0396, + "step": 10789 + }, + { + "epoch": 0.25, + "grad_norm": 2.7005959154137686, + "learning_rate": 1.747762964980078e-05, + "loss": 0.8563, + "step": 10790 + }, + { + "epoch": 0.25, + "grad_norm": 2.5083258355433813, + "learning_rate": 1.7477122994884973e-05, + "loss": 0.843, + "step": 10791 + }, + { + "epoch": 0.25, + "grad_norm": 2.5552860577370873, + "learning_rate": 1.747661629643499e-05, + "loss": 0.948, + "step": 10792 + }, + { + "epoch": 0.25, + "grad_norm": 2.1409475880085074, + "learning_rate": 1.7476109554453783e-05, + "loss": 0.9824, + "step": 10793 + }, + { + "epoch": 0.25, + "grad_norm": 2.339368600695017, + "learning_rate": 1.7475602768944297e-05, + "loss": 0.9794, + "step": 10794 + }, + { + "epoch": 0.25, + "grad_norm": 2.341408802787782, + "learning_rate": 1.747509593990948e-05, + "loss": 1.0702, + "step": 10795 + }, + { + "epoch": 0.25, + "grad_norm": 2.033024221775151, + "learning_rate": 1.7474589067352296e-05, + "loss": 0.9163, + "step": 10796 + }, + { + "epoch": 0.25, + "grad_norm": 2.790916879253882, + "learning_rate": 1.7474082151275683e-05, + "loss": 1.0535, + "step": 10797 + }, + { + "epoch": 0.25, + "grad_norm": 1.9937073941493486, + "learning_rate": 1.74735751916826e-05, + "loss": 1.1986, + "step": 10798 + }, + { + "epoch": 0.25, + "grad_norm": 2.442731978004449, + "learning_rate": 1.747306818857599e-05, + "loss": 1.0192, + "step": 10799 + }, + { + "epoch": 0.25, + "grad_norm": 1.2297093134933244, + "learning_rate": 1.7472561141958818e-05, + "loss": 1.0213, + "step": 10800 + }, + { + "epoch": 0.25, + "grad_norm": 2.0497732079411786, + "learning_rate": 1.7472054051834028e-05, + "loss": 1.0657, + "step": 10801 + }, + { + "epoch": 0.25, + "grad_norm": 2.197524517860816, + "learning_rate": 1.7471546918204575e-05, + "loss": 1.0932, + "step": 10802 + }, + { + "epoch": 0.25, + "grad_norm": 2.039447951241773, + "learning_rate": 1.7471039741073407e-05, + "loss": 1.043, + "step": 10803 + }, + { + "epoch": 0.25, + "grad_norm": 2.0962647448101066, + "learning_rate": 1.7470532520443483e-05, + "loss": 1.1683, + "step": 10804 + }, + { + "epoch": 0.25, + "grad_norm": 4.364400287708829, + "learning_rate": 1.7470025256317752e-05, + "loss": 0.9745, + "step": 10805 + }, + { + "epoch": 0.25, + "grad_norm": 2.195332354832526, + "learning_rate": 1.7469517948699167e-05, + "loss": 1.1186, + "step": 10806 + }, + { + "epoch": 0.25, + "grad_norm": 3.044745043644592, + "learning_rate": 1.7469010597590688e-05, + "loss": 1.0471, + "step": 10807 + }, + { + "epoch": 0.25, + "grad_norm": 1.9304497081976464, + "learning_rate": 1.7468503202995263e-05, + "loss": 1.0734, + "step": 10808 + }, + { + "epoch": 0.25, + "grad_norm": 1.11247115610133, + "learning_rate": 1.7467995764915847e-05, + "loss": 0.9895, + "step": 10809 + }, + { + "epoch": 0.25, + "grad_norm": 2.063599960565915, + "learning_rate": 1.7467488283355396e-05, + "loss": 1.0152, + "step": 10810 + }, + { + "epoch": 0.25, + "grad_norm": 3.1225959986607963, + "learning_rate": 1.7466980758316864e-05, + "loss": 1.1562, + "step": 10811 + }, + { + "epoch": 0.25, + "grad_norm": 2.150779598965862, + "learning_rate": 1.7466473189803203e-05, + "loss": 1.025, + "step": 10812 + }, + { + "epoch": 0.25, + "grad_norm": 2.3160625360765774, + "learning_rate": 1.7465965577817375e-05, + "loss": 1.0001, + "step": 10813 + }, + { + "epoch": 0.25, + "grad_norm": 1.1252399232619401, + "learning_rate": 1.746545792236233e-05, + "loss": 1.0141, + "step": 10814 + }, + { + "epoch": 0.25, + "grad_norm": 2.1125794378399787, + "learning_rate": 1.7464950223441026e-05, + "loss": 1.0522, + "step": 10815 + }, + { + "epoch": 0.25, + "grad_norm": 1.1156463806699126, + "learning_rate": 1.7464442481056418e-05, + "loss": 0.8883, + "step": 10816 + }, + { + "epoch": 0.25, + "grad_norm": 1.8229681489974856, + "learning_rate": 1.7463934695211463e-05, + "loss": 1.0294, + "step": 10817 + }, + { + "epoch": 0.25, + "grad_norm": 2.0670484978446235, + "learning_rate": 1.7463426865909118e-05, + "loss": 1.0298, + "step": 10818 + }, + { + "epoch": 0.25, + "grad_norm": 2.1576221964770985, + "learning_rate": 1.746291899315234e-05, + "loss": 1.0426, + "step": 10819 + }, + { + "epoch": 0.25, + "grad_norm": 2.3458997346666277, + "learning_rate": 1.7462411076944084e-05, + "loss": 0.9476, + "step": 10820 + }, + { + "epoch": 0.25, + "grad_norm": 1.0817806309992661, + "learning_rate": 1.7461903117287305e-05, + "loss": 0.964, + "step": 10821 + }, + { + "epoch": 0.25, + "grad_norm": 2.0900435835672937, + "learning_rate": 1.746139511418497e-05, + "loss": 0.9822, + "step": 10822 + }, + { + "epoch": 0.25, + "grad_norm": 2.000539770387834, + "learning_rate": 1.7460887067640025e-05, + "loss": 1.0477, + "step": 10823 + }, + { + "epoch": 0.26, + "grad_norm": 1.8593185471581979, + "learning_rate": 1.7460378977655432e-05, + "loss": 1.0035, + "step": 10824 + }, + { + "epoch": 0.26, + "grad_norm": 2.200349526402022, + "learning_rate": 1.7459870844234154e-05, + "loss": 1.0287, + "step": 10825 + }, + { + "epoch": 0.26, + "grad_norm": 2.1237835869250814, + "learning_rate": 1.745936266737915e-05, + "loss": 1.1405, + "step": 10826 + }, + { + "epoch": 0.26, + "grad_norm": 1.853424434433512, + "learning_rate": 1.745885444709337e-05, + "loss": 1.0249, + "step": 10827 + }, + { + "epoch": 0.26, + "grad_norm": 1.8618771394067535, + "learning_rate": 1.7458346183379778e-05, + "loss": 0.9749, + "step": 10828 + }, + { + "epoch": 0.26, + "grad_norm": 2.2075739654271045, + "learning_rate": 1.7457837876241332e-05, + "loss": 0.9909, + "step": 10829 + }, + { + "epoch": 0.26, + "grad_norm": 1.9883680293072952, + "learning_rate": 1.7457329525680996e-05, + "loss": 1.1403, + "step": 10830 + }, + { + "epoch": 0.26, + "grad_norm": 2.226456660967553, + "learning_rate": 1.7456821131701724e-05, + "loss": 0.9971, + "step": 10831 + }, + { + "epoch": 0.26, + "grad_norm": 2.23102528408608, + "learning_rate": 1.745631269430648e-05, + "loss": 1.026, + "step": 10832 + }, + { + "epoch": 0.26, + "grad_norm": 2.3457187982396093, + "learning_rate": 1.7455804213498223e-05, + "loss": 1.1721, + "step": 10833 + }, + { + "epoch": 0.26, + "grad_norm": 2.1656918352480337, + "learning_rate": 1.7455295689279913e-05, + "loss": 0.9731, + "step": 10834 + }, + { + "epoch": 0.26, + "grad_norm": 1.8934780117498442, + "learning_rate": 1.745478712165451e-05, + "loss": 1.1079, + "step": 10835 + }, + { + "epoch": 0.26, + "grad_norm": 2.3029168190563016, + "learning_rate": 1.745427851062498e-05, + "loss": 1.0758, + "step": 10836 + }, + { + "epoch": 0.26, + "grad_norm": 1.9652598906835632, + "learning_rate": 1.7453769856194277e-05, + "loss": 1.0377, + "step": 10837 + }, + { + "epoch": 0.26, + "grad_norm": 2.3060368813495806, + "learning_rate": 1.7453261158365367e-05, + "loss": 0.9046, + "step": 10838 + }, + { + "epoch": 0.26, + "grad_norm": 2.0874863097943317, + "learning_rate": 1.745275241714121e-05, + "loss": 1.1173, + "step": 10839 + }, + { + "epoch": 0.26, + "grad_norm": 1.9884924734844023, + "learning_rate": 1.745224363252477e-05, + "loss": 0.9742, + "step": 10840 + }, + { + "epoch": 0.26, + "grad_norm": 2.135995458591457, + "learning_rate": 1.745173480451901e-05, + "loss": 1.0539, + "step": 10841 + }, + { + "epoch": 0.26, + "grad_norm": 5.072606710601921, + "learning_rate": 1.745122593312689e-05, + "loss": 1.027, + "step": 10842 + }, + { + "epoch": 0.26, + "grad_norm": 1.9515410636993193, + "learning_rate": 1.745071701835137e-05, + "loss": 1.033, + "step": 10843 + }, + { + "epoch": 0.26, + "grad_norm": 1.9924578445855998, + "learning_rate": 1.7450208060195423e-05, + "loss": 1.0469, + "step": 10844 + }, + { + "epoch": 0.26, + "grad_norm": 2.0117680344516624, + "learning_rate": 1.7449699058662e-05, + "loss": 1.1383, + "step": 10845 + }, + { + "epoch": 0.26, + "grad_norm": 2.6832325118223452, + "learning_rate": 1.7449190013754075e-05, + "loss": 1.129, + "step": 10846 + }, + { + "epoch": 0.26, + "grad_norm": 2.245030868359275, + "learning_rate": 1.7448680925474605e-05, + "loss": 1.0147, + "step": 10847 + }, + { + "epoch": 0.26, + "grad_norm": 6.208944682417092, + "learning_rate": 1.7448171793826553e-05, + "loss": 0.9798, + "step": 10848 + }, + { + "epoch": 0.26, + "grad_norm": 2.20856680684106, + "learning_rate": 1.744766261881289e-05, + "loss": 0.9531, + "step": 10849 + }, + { + "epoch": 0.26, + "grad_norm": 2.2094238262155015, + "learning_rate": 1.7447153400436577e-05, + "loss": 0.9825, + "step": 10850 + }, + { + "epoch": 0.26, + "grad_norm": 2.2462651329376255, + "learning_rate": 1.744664413870058e-05, + "loss": 0.9446, + "step": 10851 + }, + { + "epoch": 0.26, + "grad_norm": 2.0157911279327063, + "learning_rate": 1.7446134833607863e-05, + "loss": 1.099, + "step": 10852 + }, + { + "epoch": 0.26, + "grad_norm": 2.2109903889065685, + "learning_rate": 1.7445625485161393e-05, + "loss": 0.9169, + "step": 10853 + }, + { + "epoch": 0.26, + "grad_norm": 2.988991691793948, + "learning_rate": 1.7445116093364133e-05, + "loss": 1.1265, + "step": 10854 + }, + { + "epoch": 0.26, + "grad_norm": 1.0787081940036043, + "learning_rate": 1.744460665821905e-05, + "loss": 1.0028, + "step": 10855 + }, + { + "epoch": 0.26, + "grad_norm": 2.6285369970287844, + "learning_rate": 1.7444097179729112e-05, + "loss": 1.0443, + "step": 10856 + }, + { + "epoch": 0.26, + "grad_norm": 2.206862887674486, + "learning_rate": 1.7443587657897278e-05, + "loss": 1.0162, + "step": 10857 + }, + { + "epoch": 0.26, + "grad_norm": 1.14436094329902, + "learning_rate": 1.7443078092726525e-05, + "loss": 0.9758, + "step": 10858 + }, + { + "epoch": 0.26, + "grad_norm": 1.9909558255555813, + "learning_rate": 1.7442568484219812e-05, + "loss": 1.1461, + "step": 10859 + }, + { + "epoch": 0.26, + "grad_norm": 2.1682066324571294, + "learning_rate": 1.7442058832380113e-05, + "loss": 1.0396, + "step": 10860 + }, + { + "epoch": 0.26, + "grad_norm": 2.408824487312154, + "learning_rate": 1.7441549137210386e-05, + "loss": 1.1025, + "step": 10861 + }, + { + "epoch": 0.26, + "grad_norm": 2.2463446410659906, + "learning_rate": 1.744103939871361e-05, + "loss": 1.0115, + "step": 10862 + }, + { + "epoch": 0.26, + "grad_norm": 2.1559554876443894, + "learning_rate": 1.744052961689274e-05, + "loss": 1.1437, + "step": 10863 + }, + { + "epoch": 0.26, + "grad_norm": 1.1144194627799675, + "learning_rate": 1.7440019791750757e-05, + "loss": 0.886, + "step": 10864 + }, + { + "epoch": 0.26, + "grad_norm": 2.0879010841671475, + "learning_rate": 1.743950992329062e-05, + "loss": 1.1359, + "step": 10865 + }, + { + "epoch": 0.26, + "grad_norm": 2.4686203767715367, + "learning_rate": 1.74390000115153e-05, + "loss": 1.0264, + "step": 10866 + }, + { + "epoch": 0.26, + "grad_norm": 2.199815595553724, + "learning_rate": 1.743849005642777e-05, + "loss": 0.9455, + "step": 10867 + }, + { + "epoch": 0.26, + "grad_norm": 15.817217010446424, + "learning_rate": 1.7437980058030993e-05, + "loss": 1.1042, + "step": 10868 + }, + { + "epoch": 0.26, + "grad_norm": 2.3504969544757652, + "learning_rate": 1.7437470016327944e-05, + "loss": 1.023, + "step": 10869 + }, + { + "epoch": 0.26, + "grad_norm": 2.1313953970557895, + "learning_rate": 1.743695993132159e-05, + "loss": 1.1241, + "step": 10870 + }, + { + "epoch": 0.26, + "grad_norm": 2.1045292749256865, + "learning_rate": 1.74364498030149e-05, + "loss": 1.0116, + "step": 10871 + }, + { + "epoch": 0.26, + "grad_norm": 2.0551708427110658, + "learning_rate": 1.7435939631410845e-05, + "loss": 1.1012, + "step": 10872 + }, + { + "epoch": 0.26, + "grad_norm": 2.383780694002952, + "learning_rate": 1.7435429416512394e-05, + "loss": 1.0478, + "step": 10873 + }, + { + "epoch": 0.26, + "grad_norm": 1.1239829177289191, + "learning_rate": 1.7434919158322516e-05, + "loss": 0.9898, + "step": 10874 + }, + { + "epoch": 0.26, + "grad_norm": 2.1316921776626714, + "learning_rate": 1.743440885684419e-05, + "loss": 0.9954, + "step": 10875 + }, + { + "epoch": 0.26, + "grad_norm": 2.538421285065132, + "learning_rate": 1.743389851208038e-05, + "loss": 1.0863, + "step": 10876 + }, + { + "epoch": 0.26, + "grad_norm": 2.8385293732185164, + "learning_rate": 1.7433388124034062e-05, + "loss": 1.1239, + "step": 10877 + }, + { + "epoch": 0.26, + "grad_norm": 2.065788844966135, + "learning_rate": 1.74328776927082e-05, + "loss": 1.0531, + "step": 10878 + }, + { + "epoch": 0.26, + "grad_norm": 2.499100904914961, + "learning_rate": 1.7432367218105772e-05, + "loss": 1.1623, + "step": 10879 + }, + { + "epoch": 0.26, + "grad_norm": 3.166441950416185, + "learning_rate": 1.7431856700229752e-05, + "loss": 1.0487, + "step": 10880 + }, + { + "epoch": 0.26, + "grad_norm": 2.4266831942048874, + "learning_rate": 1.7431346139083108e-05, + "loss": 1.056, + "step": 10881 + }, + { + "epoch": 0.26, + "grad_norm": 1.1091138940026266, + "learning_rate": 1.7430835534668814e-05, + "loss": 1.039, + "step": 10882 + }, + { + "epoch": 0.26, + "grad_norm": 1.857416726618169, + "learning_rate": 1.7430324886989843e-05, + "loss": 1.0959, + "step": 10883 + }, + { + "epoch": 0.26, + "grad_norm": 2.1167977293646008, + "learning_rate": 1.7429814196049166e-05, + "loss": 1.0499, + "step": 10884 + }, + { + "epoch": 0.26, + "grad_norm": 1.970084688432539, + "learning_rate": 1.7429303461849764e-05, + "loss": 1.1067, + "step": 10885 + }, + { + "epoch": 0.26, + "grad_norm": 2.0347545765322144, + "learning_rate": 1.74287926843946e-05, + "loss": 0.9574, + "step": 10886 + }, + { + "epoch": 0.26, + "grad_norm": 2.1029050807871967, + "learning_rate": 1.7428281863686654e-05, + "loss": 0.9189, + "step": 10887 + }, + { + "epoch": 0.26, + "grad_norm": 1.136866823691819, + "learning_rate": 1.74277709997289e-05, + "loss": 0.9693, + "step": 10888 + }, + { + "epoch": 0.26, + "grad_norm": 2.205833919741198, + "learning_rate": 1.742726009252431e-05, + "loss": 0.9389, + "step": 10889 + }, + { + "epoch": 0.26, + "grad_norm": 1.1349085320382406, + "learning_rate": 1.7426749142075863e-05, + "loss": 0.9811, + "step": 10890 + }, + { + "epoch": 0.26, + "grad_norm": 2.0047050464388874, + "learning_rate": 1.742623814838653e-05, + "loss": 0.9741, + "step": 10891 + }, + { + "epoch": 0.26, + "grad_norm": 1.8367067959821168, + "learning_rate": 1.742572711145929e-05, + "loss": 1.0265, + "step": 10892 + }, + { + "epoch": 0.26, + "grad_norm": 2.0200917113725394, + "learning_rate": 1.7425216031297113e-05, + "loss": 1.0589, + "step": 10893 + }, + { + "epoch": 0.26, + "grad_norm": 2.1213072872160477, + "learning_rate": 1.7424704907902975e-05, + "loss": 1.031, + "step": 10894 + }, + { + "epoch": 0.26, + "grad_norm": 2.0241220169738883, + "learning_rate": 1.742419374127986e-05, + "loss": 0.9886, + "step": 10895 + }, + { + "epoch": 0.26, + "grad_norm": 2.2949514843674184, + "learning_rate": 1.7423682531430736e-05, + "loss": 1.0331, + "step": 10896 + }, + { + "epoch": 0.26, + "grad_norm": 2.4630795458546686, + "learning_rate": 1.742317127835858e-05, + "loss": 1.045, + "step": 10897 + }, + { + "epoch": 0.26, + "grad_norm": 2.018550531030555, + "learning_rate": 1.7422659982066373e-05, + "loss": 1.0772, + "step": 10898 + }, + { + "epoch": 0.26, + "grad_norm": 1.966276016633321, + "learning_rate": 1.7422148642557088e-05, + "loss": 1.0345, + "step": 10899 + }, + { + "epoch": 0.26, + "grad_norm": 1.1775993812273433, + "learning_rate": 1.7421637259833707e-05, + "loss": 0.9795, + "step": 10900 + }, + { + "epoch": 0.26, + "grad_norm": 2.8850321941144705, + "learning_rate": 1.7421125833899203e-05, + "loss": 0.9232, + "step": 10901 + }, + { + "epoch": 0.26, + "grad_norm": 2.014229294601763, + "learning_rate": 1.7420614364756553e-05, + "loss": 0.9598, + "step": 10902 + }, + { + "epoch": 0.26, + "grad_norm": 2.0717839457370526, + "learning_rate": 1.7420102852408735e-05, + "loss": 1.217, + "step": 10903 + }, + { + "epoch": 0.26, + "grad_norm": 2.4500734382820517, + "learning_rate": 1.7419591296858733e-05, + "loss": 1.0939, + "step": 10904 + }, + { + "epoch": 0.26, + "grad_norm": 1.125787329094812, + "learning_rate": 1.741907969810952e-05, + "loss": 0.9684, + "step": 10905 + }, + { + "epoch": 0.26, + "grad_norm": 2.1421721745649913, + "learning_rate": 1.7418568056164077e-05, + "loss": 1.0611, + "step": 10906 + }, + { + "epoch": 0.26, + "grad_norm": 2.214265085182148, + "learning_rate": 1.7418056371025383e-05, + "loss": 0.9063, + "step": 10907 + }, + { + "epoch": 0.26, + "grad_norm": 1.101703619329976, + "learning_rate": 1.7417544642696414e-05, + "loss": 0.9933, + "step": 10908 + }, + { + "epoch": 0.26, + "grad_norm": 2.4148710560118714, + "learning_rate": 1.741703287118015e-05, + "loss": 1.198, + "step": 10909 + }, + { + "epoch": 0.26, + "grad_norm": 1.9173781027886132, + "learning_rate": 1.7416521056479577e-05, + "loss": 1.1622, + "step": 10910 + }, + { + "epoch": 0.26, + "grad_norm": 2.458867765287154, + "learning_rate": 1.7416009198597668e-05, + "loss": 0.9134, + "step": 10911 + }, + { + "epoch": 0.26, + "grad_norm": 2.0313711447765512, + "learning_rate": 1.7415497297537407e-05, + "loss": 1.111, + "step": 10912 + }, + { + "epoch": 0.26, + "grad_norm": 1.892391009081199, + "learning_rate": 1.7414985353301773e-05, + "loss": 1.0426, + "step": 10913 + }, + { + "epoch": 0.26, + "grad_norm": 1.1264493864311, + "learning_rate": 1.7414473365893744e-05, + "loss": 1.0253, + "step": 10914 + }, + { + "epoch": 0.26, + "grad_norm": 2.194556280803421, + "learning_rate": 1.7413961335316305e-05, + "loss": 1.1611, + "step": 10915 + }, + { + "epoch": 0.26, + "grad_norm": 2.0118816008900957, + "learning_rate": 1.7413449261572437e-05, + "loss": 1.2153, + "step": 10916 + }, + { + "epoch": 0.26, + "grad_norm": 3.0666068073762607, + "learning_rate": 1.741293714466512e-05, + "loss": 0.999, + "step": 10917 + }, + { + "epoch": 0.26, + "grad_norm": 2.21381618489805, + "learning_rate": 1.7412424984597334e-05, + "loss": 1.0219, + "step": 10918 + }, + { + "epoch": 0.26, + "grad_norm": 1.1415907485652053, + "learning_rate": 1.7411912781372066e-05, + "loss": 0.9837, + "step": 10919 + }, + { + "epoch": 0.26, + "grad_norm": 2.0565764701560343, + "learning_rate": 1.7411400534992292e-05, + "loss": 0.9531, + "step": 10920 + }, + { + "epoch": 0.26, + "grad_norm": 2.073003629417032, + "learning_rate": 1.7410888245461e-05, + "loss": 0.869, + "step": 10921 + }, + { + "epoch": 0.26, + "grad_norm": 2.1520063118231274, + "learning_rate": 1.7410375912781173e-05, + "loss": 1.0759, + "step": 10922 + }, + { + "epoch": 0.26, + "grad_norm": 2.112809716169976, + "learning_rate": 1.7409863536955788e-05, + "loss": 1.1772, + "step": 10923 + }, + { + "epoch": 0.26, + "grad_norm": 2.249985384189964, + "learning_rate": 1.7409351117987833e-05, + "loss": 0.9578, + "step": 10924 + }, + { + "epoch": 0.26, + "grad_norm": 2.005963711406879, + "learning_rate": 1.7408838655880287e-05, + "loss": 1.1016, + "step": 10925 + }, + { + "epoch": 0.26, + "grad_norm": 2.096037329869894, + "learning_rate": 1.7408326150636138e-05, + "loss": 1.1852, + "step": 10926 + }, + { + "epoch": 0.26, + "grad_norm": 2.067290646798797, + "learning_rate": 1.740781360225837e-05, + "loss": 0.9921, + "step": 10927 + }, + { + "epoch": 0.26, + "grad_norm": 2.370969272988019, + "learning_rate": 1.7407301010749967e-05, + "loss": 0.9333, + "step": 10928 + }, + { + "epoch": 0.26, + "grad_norm": 2.1563022981223887, + "learning_rate": 1.740678837611391e-05, + "loss": 1.0165, + "step": 10929 + }, + { + "epoch": 0.26, + "grad_norm": 1.995901339032634, + "learning_rate": 1.7406275698353184e-05, + "loss": 1.0843, + "step": 10930 + }, + { + "epoch": 0.26, + "grad_norm": 2.014987857446612, + "learning_rate": 1.740576297747078e-05, + "loss": 1.0726, + "step": 10931 + }, + { + "epoch": 0.26, + "grad_norm": 2.0374567629672775, + "learning_rate": 1.7405250213469678e-05, + "loss": 1.0555, + "step": 10932 + }, + { + "epoch": 0.26, + "grad_norm": 2.2384553450680653, + "learning_rate": 1.7404737406352866e-05, + "loss": 0.9757, + "step": 10933 + }, + { + "epoch": 0.26, + "grad_norm": 1.9848392191855984, + "learning_rate": 1.7404224556123324e-05, + "loss": 0.9748, + "step": 10934 + }, + { + "epoch": 0.26, + "grad_norm": 2.5582030198176864, + "learning_rate": 1.740371166278405e-05, + "loss": 1.0738, + "step": 10935 + }, + { + "epoch": 0.26, + "grad_norm": 1.0629110586427128, + "learning_rate": 1.7403198726338017e-05, + "loss": 0.9753, + "step": 10936 + }, + { + "epoch": 0.26, + "grad_norm": 2.0186833821386556, + "learning_rate": 1.740268574678822e-05, + "loss": 1.0873, + "step": 10937 + }, + { + "epoch": 0.26, + "grad_norm": 2.34135886288518, + "learning_rate": 1.7402172724137643e-05, + "loss": 1.1473, + "step": 10938 + }, + { + "epoch": 0.26, + "grad_norm": 1.8916395636026866, + "learning_rate": 1.7401659658389268e-05, + "loss": 1.0038, + "step": 10939 + }, + { + "epoch": 0.26, + "grad_norm": 2.1951528758216137, + "learning_rate": 1.7401146549546093e-05, + "loss": 1.0614, + "step": 10940 + }, + { + "epoch": 0.26, + "grad_norm": 1.9176800816767416, + "learning_rate": 1.7400633397611097e-05, + "loss": 1.066, + "step": 10941 + }, + { + "epoch": 0.26, + "grad_norm": 2.1573580816844777, + "learning_rate": 1.7400120202587273e-05, + "loss": 1.0664, + "step": 10942 + }, + { + "epoch": 0.26, + "grad_norm": 1.9378656866117954, + "learning_rate": 1.7399606964477604e-05, + "loss": 1.0752, + "step": 10943 + }, + { + "epoch": 0.26, + "grad_norm": 2.345337941844526, + "learning_rate": 1.7399093683285082e-05, + "loss": 0.9551, + "step": 10944 + }, + { + "epoch": 0.26, + "grad_norm": 2.0360199826125367, + "learning_rate": 1.739858035901269e-05, + "loss": 1.0598, + "step": 10945 + }, + { + "epoch": 0.26, + "grad_norm": 1.8461115634949086, + "learning_rate": 1.7398066991663426e-05, + "loss": 1.0908, + "step": 10946 + }, + { + "epoch": 0.26, + "grad_norm": 2.273955548410859, + "learning_rate": 1.7397553581240272e-05, + "loss": 0.9118, + "step": 10947 + }, + { + "epoch": 0.26, + "grad_norm": 1.7456026994243077, + "learning_rate": 1.7397040127746215e-05, + "loss": 1.1054, + "step": 10948 + }, + { + "epoch": 0.26, + "grad_norm": 2.5306380321910993, + "learning_rate": 1.7396526631184256e-05, + "loss": 1.0038, + "step": 10949 + }, + { + "epoch": 0.26, + "grad_norm": 2.201308496664367, + "learning_rate": 1.739601309155737e-05, + "loss": 1.0111, + "step": 10950 + }, + { + "epoch": 0.26, + "grad_norm": 2.1135533936751614, + "learning_rate": 1.739549950886856e-05, + "loss": 0.983, + "step": 10951 + }, + { + "epoch": 0.26, + "grad_norm": 1.2668579900075532, + "learning_rate": 1.7394985883120807e-05, + "loss": 0.9706, + "step": 10952 + }, + { + "epoch": 0.26, + "grad_norm": 2.3358620170729956, + "learning_rate": 1.7394472214317107e-05, + "loss": 0.9613, + "step": 10953 + }, + { + "epoch": 0.26, + "grad_norm": 2.0902496994216393, + "learning_rate": 1.7393958502460447e-05, + "loss": 0.9855, + "step": 10954 + }, + { + "epoch": 0.26, + "grad_norm": 2.034942369859641, + "learning_rate": 1.739344474755382e-05, + "loss": 0.9799, + "step": 10955 + }, + { + "epoch": 0.26, + "grad_norm": 2.0440392280113544, + "learning_rate": 1.7392930949600217e-05, + "loss": 0.9451, + "step": 10956 + }, + { + "epoch": 0.26, + "grad_norm": 1.8811561461381872, + "learning_rate": 1.739241710860263e-05, + "loss": 1.0407, + "step": 10957 + }, + { + "epoch": 0.26, + "grad_norm": 2.0605328891604935, + "learning_rate": 1.739190322456405e-05, + "loss": 1.0817, + "step": 10958 + }, + { + "epoch": 0.26, + "grad_norm": 2.1572658244156804, + "learning_rate": 1.7391389297487472e-05, + "loss": 0.9934, + "step": 10959 + }, + { + "epoch": 0.26, + "grad_norm": 1.786864984717927, + "learning_rate": 1.739087532737588e-05, + "loss": 1.0719, + "step": 10960 + }, + { + "epoch": 0.26, + "grad_norm": 2.2057655828101543, + "learning_rate": 1.7390361314232275e-05, + "loss": 1.1369, + "step": 10961 + }, + { + "epoch": 0.26, + "grad_norm": 3.5062646551232857, + "learning_rate": 1.7389847258059646e-05, + "loss": 1.0412, + "step": 10962 + }, + { + "epoch": 0.26, + "grad_norm": 2.0366551745747516, + "learning_rate": 1.738933315886099e-05, + "loss": 1.0479, + "step": 10963 + }, + { + "epoch": 0.26, + "grad_norm": 2.441103382398027, + "learning_rate": 1.7388819016639292e-05, + "loss": 1.0647, + "step": 10964 + }, + { + "epoch": 0.26, + "grad_norm": 2.2155702010775733, + "learning_rate": 1.738830483139755e-05, + "loss": 0.9728, + "step": 10965 + }, + { + "epoch": 0.26, + "grad_norm": 1.1685146065733232, + "learning_rate": 1.7387790603138764e-05, + "loss": 0.9237, + "step": 10966 + }, + { + "epoch": 0.26, + "grad_norm": 2.170559746711686, + "learning_rate": 1.7387276331865916e-05, + "loss": 0.9515, + "step": 10967 + }, + { + "epoch": 0.26, + "grad_norm": 2.266505185107626, + "learning_rate": 1.738676201758201e-05, + "loss": 1.0472, + "step": 10968 + }, + { + "epoch": 0.26, + "grad_norm": 1.2988703534424786, + "learning_rate": 1.7386247660290038e-05, + "loss": 1.0122, + "step": 10969 + }, + { + "epoch": 0.26, + "grad_norm": 2.6995947275107977, + "learning_rate": 1.738573325999299e-05, + "loss": 1.0348, + "step": 10970 + }, + { + "epoch": 0.26, + "grad_norm": 2.0065281940632334, + "learning_rate": 1.7385218816693864e-05, + "loss": 1.0637, + "step": 10971 + }, + { + "epoch": 0.26, + "grad_norm": 1.0649006489016677, + "learning_rate": 1.738470433039566e-05, + "loss": 1.0634, + "step": 10972 + }, + { + "epoch": 0.26, + "grad_norm": 2.2515565918235048, + "learning_rate": 1.738418980110137e-05, + "loss": 0.8998, + "step": 10973 + }, + { + "epoch": 0.26, + "grad_norm": 2.0157368908689044, + "learning_rate": 1.7383675228813986e-05, + "loss": 1.0074, + "step": 10974 + }, + { + "epoch": 0.26, + "grad_norm": 2.0497063227665047, + "learning_rate": 1.7383160613536508e-05, + "loss": 0.989, + "step": 10975 + }, + { + "epoch": 0.26, + "grad_norm": 2.0482009316582994, + "learning_rate": 1.738264595527193e-05, + "loss": 1.2082, + "step": 10976 + }, + { + "epoch": 0.26, + "grad_norm": 2.2996370151165015, + "learning_rate": 1.7382131254023252e-05, + "loss": 1.174, + "step": 10977 + }, + { + "epoch": 0.26, + "grad_norm": 1.990996145503383, + "learning_rate": 1.7381616509793468e-05, + "loss": 1.0525, + "step": 10978 + }, + { + "epoch": 0.26, + "grad_norm": 2.217345364655596, + "learning_rate": 1.7381101722585577e-05, + "loss": 1.0261, + "step": 10979 + }, + { + "epoch": 0.26, + "grad_norm": 1.2081307319359842, + "learning_rate": 1.7380586892402572e-05, + "loss": 0.9424, + "step": 10980 + }, + { + "epoch": 0.26, + "grad_norm": 2.017182036855245, + "learning_rate": 1.7380072019247455e-05, + "loss": 1.0258, + "step": 10981 + }, + { + "epoch": 0.26, + "grad_norm": 2.244565934647815, + "learning_rate": 1.7379557103123223e-05, + "loss": 0.9702, + "step": 10982 + }, + { + "epoch": 0.26, + "grad_norm": 2.3220482689222437, + "learning_rate": 1.7379042144032873e-05, + "loss": 1.189, + "step": 10983 + }, + { + "epoch": 0.26, + "grad_norm": 1.0487136786149007, + "learning_rate": 1.7378527141979402e-05, + "loss": 0.8973, + "step": 10984 + }, + { + "epoch": 0.26, + "grad_norm": 1.8038343296741899, + "learning_rate": 1.7378012096965813e-05, + "loss": 1.1368, + "step": 10985 + }, + { + "epoch": 0.26, + "grad_norm": 1.9686045864379635, + "learning_rate": 1.73774970089951e-05, + "loss": 1.1142, + "step": 10986 + }, + { + "epoch": 0.26, + "grad_norm": 1.0939644043534646, + "learning_rate": 1.7376981878070265e-05, + "loss": 0.9589, + "step": 10987 + }, + { + "epoch": 0.26, + "grad_norm": 3.1221384307116975, + "learning_rate": 1.73764667041943e-05, + "loss": 1.0345, + "step": 10988 + }, + { + "epoch": 0.26, + "grad_norm": 2.113682062150583, + "learning_rate": 1.737595148737022e-05, + "loss": 1.1117, + "step": 10989 + }, + { + "epoch": 0.26, + "grad_norm": 2.027556389271701, + "learning_rate": 1.7375436227601007e-05, + "loss": 0.9668, + "step": 10990 + }, + { + "epoch": 0.26, + "grad_norm": 2.0504717137976916, + "learning_rate": 1.7374920924889675e-05, + "loss": 1.051, + "step": 10991 + }, + { + "epoch": 0.26, + "grad_norm": 1.1795729027280144, + "learning_rate": 1.7374405579239217e-05, + "loss": 0.9786, + "step": 10992 + }, + { + "epoch": 0.26, + "grad_norm": 1.9195437832147781, + "learning_rate": 1.737389019065263e-05, + "loss": 1.0537, + "step": 10993 + }, + { + "epoch": 0.26, + "grad_norm": 2.5258313140334425, + "learning_rate": 1.7373374759132927e-05, + "loss": 1.022, + "step": 10994 + }, + { + "epoch": 0.26, + "grad_norm": 1.9626152962841283, + "learning_rate": 1.73728592846831e-05, + "loss": 0.9272, + "step": 10995 + }, + { + "epoch": 0.26, + "grad_norm": 1.1750350002614134, + "learning_rate": 1.7372343767306148e-05, + "loss": 0.9895, + "step": 10996 + }, + { + "epoch": 0.26, + "grad_norm": 1.9004659558755872, + "learning_rate": 1.737182820700508e-05, + "loss": 1.0402, + "step": 10997 + }, + { + "epoch": 0.26, + "grad_norm": 2.514731050449079, + "learning_rate": 1.7371312603782893e-05, + "loss": 1.0235, + "step": 10998 + }, + { + "epoch": 0.26, + "grad_norm": 1.8846291480288047, + "learning_rate": 1.737079695764259e-05, + "loss": 1.153, + "step": 10999 + }, + { + "epoch": 0.26, + "grad_norm": 1.9898686849823823, + "learning_rate": 1.7370281268587175e-05, + "loss": 1.132, + "step": 11000 + }, + { + "epoch": 0.26, + "grad_norm": 2.3575622009967385, + "learning_rate": 1.736976553661965e-05, + "loss": 1.1178, + "step": 11001 + }, + { + "epoch": 0.26, + "grad_norm": 1.9739453750339562, + "learning_rate": 1.7369249761743015e-05, + "loss": 1.0561, + "step": 11002 + }, + { + "epoch": 0.26, + "grad_norm": 2.1815806278578167, + "learning_rate": 1.7368733943960278e-05, + "loss": 0.8915, + "step": 11003 + }, + { + "epoch": 0.26, + "grad_norm": 2.483054259431134, + "learning_rate": 1.736821808327443e-05, + "loss": 1.0613, + "step": 11004 + }, + { + "epoch": 0.26, + "grad_norm": 1.0719072645512655, + "learning_rate": 1.7367702179688494e-05, + "loss": 0.9433, + "step": 11005 + }, + { + "epoch": 0.26, + "grad_norm": 2.071422735920554, + "learning_rate": 1.7367186233205457e-05, + "loss": 0.9353, + "step": 11006 + }, + { + "epoch": 0.26, + "grad_norm": 2.4586127990289866, + "learning_rate": 1.7366670243828333e-05, + "loss": 1.1243, + "step": 11007 + }, + { + "epoch": 0.26, + "grad_norm": 1.9634485517599594, + "learning_rate": 1.736615421156012e-05, + "loss": 1.0562, + "step": 11008 + }, + { + "epoch": 0.26, + "grad_norm": 2.0269349806768, + "learning_rate": 1.7365638136403827e-05, + "loss": 1.0723, + "step": 11009 + }, + { + "epoch": 0.26, + "grad_norm": 2.075185354769882, + "learning_rate": 1.7365122018362456e-05, + "loss": 0.9904, + "step": 11010 + }, + { + "epoch": 0.26, + "grad_norm": 1.093286598121815, + "learning_rate": 1.736460585743901e-05, + "loss": 0.9751, + "step": 11011 + }, + { + "epoch": 0.26, + "grad_norm": 1.9267226361941714, + "learning_rate": 1.7364089653636503e-05, + "loss": 1.0761, + "step": 11012 + }, + { + "epoch": 0.26, + "grad_norm": 2.244572565474486, + "learning_rate": 1.736357340695793e-05, + "loss": 0.9347, + "step": 11013 + }, + { + "epoch": 0.26, + "grad_norm": 2.083851218494358, + "learning_rate": 1.7363057117406304e-05, + "loss": 1.0965, + "step": 11014 + }, + { + "epoch": 0.26, + "grad_norm": 3.508121218230643, + "learning_rate": 1.7362540784984627e-05, + "loss": 1.0954, + "step": 11015 + }, + { + "epoch": 0.26, + "grad_norm": 3.2290599940851274, + "learning_rate": 1.7362024409695908e-05, + "loss": 1.0064, + "step": 11016 + }, + { + "epoch": 0.26, + "grad_norm": 3.3721599567632223, + "learning_rate": 1.736150799154315e-05, + "loss": 1.0372, + "step": 11017 + }, + { + "epoch": 0.26, + "grad_norm": 2.2335769614704177, + "learning_rate": 1.7360991530529363e-05, + "loss": 1.0652, + "step": 11018 + }, + { + "epoch": 0.26, + "grad_norm": 2.206633103587952, + "learning_rate": 1.7360475026657552e-05, + "loss": 1.0563, + "step": 11019 + }, + { + "epoch": 0.26, + "grad_norm": 1.9710489613278552, + "learning_rate": 1.7359958479930726e-05, + "loss": 1.0273, + "step": 11020 + }, + { + "epoch": 0.26, + "grad_norm": 1.1461036718591742, + "learning_rate": 1.7359441890351893e-05, + "loss": 0.9612, + "step": 11021 + }, + { + "epoch": 0.26, + "grad_norm": 2.108302347543195, + "learning_rate": 1.735892525792406e-05, + "loss": 0.9443, + "step": 11022 + }, + { + "epoch": 0.26, + "grad_norm": 2.0531588979929984, + "learning_rate": 1.735840858265023e-05, + "loss": 1.1335, + "step": 11023 + }, + { + "epoch": 0.26, + "grad_norm": 2.0369157475098656, + "learning_rate": 1.7357891864533418e-05, + "loss": 1.1052, + "step": 11024 + }, + { + "epoch": 0.26, + "grad_norm": 2.4631194577292983, + "learning_rate": 1.7357375103576628e-05, + "loss": 1.0344, + "step": 11025 + }, + { + "epoch": 0.26, + "grad_norm": 1.7893005913736149, + "learning_rate": 1.735685829978287e-05, + "loss": 1.0113, + "step": 11026 + }, + { + "epoch": 0.26, + "grad_norm": 2.933126547940726, + "learning_rate": 1.735634145315516e-05, + "loss": 1.0615, + "step": 11027 + }, + { + "epoch": 0.26, + "grad_norm": 2.007225116064047, + "learning_rate": 1.7355824563696496e-05, + "loss": 0.978, + "step": 11028 + }, + { + "epoch": 0.26, + "grad_norm": 1.8505376577999306, + "learning_rate": 1.7355307631409894e-05, + "loss": 0.9051, + "step": 11029 + }, + { + "epoch": 0.26, + "grad_norm": 1.975417938834701, + "learning_rate": 1.7354790656298364e-05, + "loss": 0.9546, + "step": 11030 + }, + { + "epoch": 0.26, + "grad_norm": 2.054589411530013, + "learning_rate": 1.7354273638364914e-05, + "loss": 0.94, + "step": 11031 + }, + { + "epoch": 0.26, + "grad_norm": 1.975201831635428, + "learning_rate": 1.7353756577612552e-05, + "loss": 1.0276, + "step": 11032 + }, + { + "epoch": 0.26, + "grad_norm": 2.2352946561875164, + "learning_rate": 1.735323947404429e-05, + "loss": 1.0025, + "step": 11033 + }, + { + "epoch": 0.26, + "grad_norm": 3.5052063886722142, + "learning_rate": 1.7352722327663146e-05, + "loss": 1.0262, + "step": 11034 + }, + { + "epoch": 0.26, + "grad_norm": 3.462979006298333, + "learning_rate": 1.735220513847212e-05, + "loss": 0.9498, + "step": 11035 + }, + { + "epoch": 0.26, + "grad_norm": 2.1326451863106817, + "learning_rate": 1.735168790647423e-05, + "loss": 1.1468, + "step": 11036 + }, + { + "epoch": 0.26, + "grad_norm": 2.062838097923712, + "learning_rate": 1.735117063167248e-05, + "loss": 1.1237, + "step": 11037 + }, + { + "epoch": 0.26, + "grad_norm": 2.18831861837568, + "learning_rate": 1.7350653314069898e-05, + "loss": 0.9372, + "step": 11038 + }, + { + "epoch": 0.26, + "grad_norm": 2.0481642306230823, + "learning_rate": 1.7350135953669477e-05, + "loss": 1.0835, + "step": 11039 + }, + { + "epoch": 0.26, + "grad_norm": 2.1339049268555588, + "learning_rate": 1.7349618550474237e-05, + "loss": 1.0372, + "step": 11040 + }, + { + "epoch": 0.26, + "grad_norm": 2.1281151679766954, + "learning_rate": 1.7349101104487195e-05, + "loss": 1.1028, + "step": 11041 + }, + { + "epoch": 0.26, + "grad_norm": 2.380410821338621, + "learning_rate": 1.734858361571136e-05, + "loss": 1.0508, + "step": 11042 + }, + { + "epoch": 0.26, + "grad_norm": 2.0688387610852206, + "learning_rate": 1.7348066084149743e-05, + "loss": 1.0887, + "step": 11043 + }, + { + "epoch": 0.26, + "grad_norm": 2.127550994439838, + "learning_rate": 1.7347548509805357e-05, + "loss": 1.0463, + "step": 11044 + }, + { + "epoch": 0.26, + "grad_norm": 2.04857129891025, + "learning_rate": 1.7347030892681217e-05, + "loss": 1.0194, + "step": 11045 + }, + { + "epoch": 0.26, + "grad_norm": 2.2188827420792476, + "learning_rate": 1.734651323278034e-05, + "loss": 1.0104, + "step": 11046 + }, + { + "epoch": 0.26, + "grad_norm": 1.8255909904925505, + "learning_rate": 1.7345995530105737e-05, + "loss": 0.9758, + "step": 11047 + }, + { + "epoch": 0.26, + "grad_norm": 1.2349078227691201, + "learning_rate": 1.734547778466042e-05, + "loss": 1.0425, + "step": 11048 + }, + { + "epoch": 0.26, + "grad_norm": 2.469822173287242, + "learning_rate": 1.734495999644741e-05, + "loss": 1.0658, + "step": 11049 + }, + { + "epoch": 0.26, + "grad_norm": 1.9884644494340185, + "learning_rate": 1.7344442165469714e-05, + "loss": 1.2134, + "step": 11050 + }, + { + "epoch": 0.26, + "grad_norm": 2.268683969908533, + "learning_rate": 1.7343924291730352e-05, + "loss": 1.0352, + "step": 11051 + }, + { + "epoch": 0.26, + "grad_norm": 1.0824413766036876, + "learning_rate": 1.7343406375232333e-05, + "loss": 1.0388, + "step": 11052 + }, + { + "epoch": 0.26, + "grad_norm": 1.879645080892406, + "learning_rate": 1.7342888415978683e-05, + "loss": 1.0245, + "step": 11053 + }, + { + "epoch": 0.26, + "grad_norm": 1.9204534874437544, + "learning_rate": 1.734237041397241e-05, + "loss": 1.0212, + "step": 11054 + }, + { + "epoch": 0.26, + "grad_norm": 3.0574372808831756, + "learning_rate": 1.734185236921653e-05, + "loss": 1.1489, + "step": 11055 + }, + { + "epoch": 0.26, + "grad_norm": 2.243620678218504, + "learning_rate": 1.7341334281714067e-05, + "loss": 0.9937, + "step": 11056 + }, + { + "epoch": 0.26, + "grad_norm": 2.0437996186330682, + "learning_rate": 1.7340816151468026e-05, + "loss": 1.1087, + "step": 11057 + }, + { + "epoch": 0.26, + "grad_norm": 2.1082665035292494, + "learning_rate": 1.734029797848143e-05, + "loss": 1.022, + "step": 11058 + }, + { + "epoch": 0.26, + "grad_norm": 1.9737939030356664, + "learning_rate": 1.73397797627573e-05, + "loss": 1.0813, + "step": 11059 + }, + { + "epoch": 0.26, + "grad_norm": 1.9820017103358196, + "learning_rate": 1.7339261504298645e-05, + "loss": 0.9968, + "step": 11060 + }, + { + "epoch": 0.26, + "grad_norm": 1.9302466289285927, + "learning_rate": 1.7338743203108484e-05, + "loss": 0.9484, + "step": 11061 + }, + { + "epoch": 0.26, + "grad_norm": 1.1246382542327646, + "learning_rate": 1.7338224859189837e-05, + "loss": 0.8813, + "step": 11062 + }, + { + "epoch": 0.26, + "grad_norm": 1.1144490512522074, + "learning_rate": 1.7337706472545727e-05, + "loss": 1.0123, + "step": 11063 + }, + { + "epoch": 0.26, + "grad_norm": 2.0165181971722554, + "learning_rate": 1.733718804317916e-05, + "loss": 1.0312, + "step": 11064 + }, + { + "epoch": 0.26, + "grad_norm": 2.2873115967301, + "learning_rate": 1.7336669571093165e-05, + "loss": 1.0746, + "step": 11065 + }, + { + "epoch": 0.26, + "grad_norm": 2.050725918886254, + "learning_rate": 1.7336151056290758e-05, + "loss": 0.9123, + "step": 11066 + }, + { + "epoch": 0.26, + "grad_norm": 2.045784192320968, + "learning_rate": 1.7335632498774956e-05, + "loss": 0.9876, + "step": 11067 + }, + { + "epoch": 0.26, + "grad_norm": 1.1260286686777068, + "learning_rate": 1.7335113898548775e-05, + "loss": 0.891, + "step": 11068 + }, + { + "epoch": 0.26, + "grad_norm": 5.255313084193679, + "learning_rate": 1.7334595255615243e-05, + "loss": 1.0051, + "step": 11069 + }, + { + "epoch": 0.26, + "grad_norm": 2.414321577940359, + "learning_rate": 1.7334076569977372e-05, + "loss": 0.9795, + "step": 11070 + }, + { + "epoch": 0.26, + "grad_norm": 2.1935993080969163, + "learning_rate": 1.733355784163819e-05, + "loss": 1.0268, + "step": 11071 + }, + { + "epoch": 0.26, + "grad_norm": 2.0846015099977278, + "learning_rate": 1.733303907060071e-05, + "loss": 1.153, + "step": 11072 + }, + { + "epoch": 0.26, + "grad_norm": 2.0694475650158224, + "learning_rate": 1.7332520256867954e-05, + "loss": 1.1163, + "step": 11073 + }, + { + "epoch": 0.26, + "grad_norm": 2.195236477201593, + "learning_rate": 1.7332001400442946e-05, + "loss": 1.0127, + "step": 11074 + }, + { + "epoch": 0.26, + "grad_norm": 2.1356845018526776, + "learning_rate": 1.7331482501328704e-05, + "loss": 1.1382, + "step": 11075 + }, + { + "epoch": 0.26, + "grad_norm": 2.23379618330969, + "learning_rate": 1.733096355952825e-05, + "loss": 1.0989, + "step": 11076 + }, + { + "epoch": 0.26, + "grad_norm": 1.961465125262669, + "learning_rate": 1.7330444575044603e-05, + "loss": 1.129, + "step": 11077 + }, + { + "epoch": 0.26, + "grad_norm": 1.9739141757972876, + "learning_rate": 1.7329925547880785e-05, + "loss": 1.025, + "step": 11078 + }, + { + "epoch": 0.26, + "grad_norm": 2.166094910518074, + "learning_rate": 1.7329406478039825e-05, + "loss": 0.9368, + "step": 11079 + }, + { + "epoch": 0.26, + "grad_norm": 2.1290343941610153, + "learning_rate": 1.732888736552474e-05, + "loss": 1.0452, + "step": 11080 + }, + { + "epoch": 0.26, + "grad_norm": 1.9140482000624484, + "learning_rate": 1.732836821033855e-05, + "loss": 1.0196, + "step": 11081 + }, + { + "epoch": 0.26, + "grad_norm": 2.4360764111557907, + "learning_rate": 1.732784901248428e-05, + "loss": 0.9692, + "step": 11082 + }, + { + "epoch": 0.26, + "grad_norm": 3.1206264754787894, + "learning_rate": 1.7327329771964952e-05, + "loss": 1.1165, + "step": 11083 + }, + { + "epoch": 0.26, + "grad_norm": 2.3366699549561694, + "learning_rate": 1.7326810488783593e-05, + "loss": 1.1155, + "step": 11084 + }, + { + "epoch": 0.26, + "grad_norm": 2.158558749873151, + "learning_rate": 1.732629116294322e-05, + "loss": 1.0031, + "step": 11085 + }, + { + "epoch": 0.26, + "grad_norm": 3.20634651838473, + "learning_rate": 1.7325771794446865e-05, + "loss": 1.0991, + "step": 11086 + }, + { + "epoch": 0.26, + "grad_norm": 2.2468858736017827, + "learning_rate": 1.7325252383297547e-05, + "loss": 1.0525, + "step": 11087 + }, + { + "epoch": 0.26, + "grad_norm": 2.2910413827001013, + "learning_rate": 1.7324732929498287e-05, + "loss": 1.0688, + "step": 11088 + }, + { + "epoch": 0.26, + "grad_norm": 2.2074224562079685, + "learning_rate": 1.7324213433052115e-05, + "loss": 1.0352, + "step": 11089 + }, + { + "epoch": 0.26, + "grad_norm": 2.3062660370786996, + "learning_rate": 1.7323693893962055e-05, + "loss": 1.0817, + "step": 11090 + }, + { + "epoch": 0.26, + "grad_norm": 1.9625201805673191, + "learning_rate": 1.732317431223113e-05, + "loss": 0.9559, + "step": 11091 + }, + { + "epoch": 0.26, + "grad_norm": 2.0192601346591887, + "learning_rate": 1.732265468786236e-05, + "loss": 1.0646, + "step": 11092 + }, + { + "epoch": 0.26, + "grad_norm": 2.1132953743806366, + "learning_rate": 1.732213502085878e-05, + "loss": 0.9151, + "step": 11093 + }, + { + "epoch": 0.26, + "grad_norm": 2.0479048332949645, + "learning_rate": 1.7321615311223413e-05, + "loss": 1.1291, + "step": 11094 + }, + { + "epoch": 0.26, + "grad_norm": 2.038853819968916, + "learning_rate": 1.7321095558959284e-05, + "loss": 0.9898, + "step": 11095 + }, + { + "epoch": 0.26, + "grad_norm": 1.7915461859622765, + "learning_rate": 1.732057576406942e-05, + "loss": 0.9292, + "step": 11096 + }, + { + "epoch": 0.26, + "grad_norm": 2.1010051426047816, + "learning_rate": 1.7320055926556843e-05, + "loss": 1.0689, + "step": 11097 + }, + { + "epoch": 0.26, + "grad_norm": 2.3469264012486017, + "learning_rate": 1.7319536046424583e-05, + "loss": 0.9564, + "step": 11098 + }, + { + "epoch": 0.26, + "grad_norm": 2.3217363538443085, + "learning_rate": 1.731901612367567e-05, + "loss": 1.0692, + "step": 11099 + }, + { + "epoch": 0.26, + "grad_norm": 2.094701578529236, + "learning_rate": 1.7318496158313124e-05, + "loss": 0.8716, + "step": 11100 + }, + { + "epoch": 0.26, + "grad_norm": 3.596720306727462, + "learning_rate": 1.731797615033998e-05, + "loss": 1.0965, + "step": 11101 + }, + { + "epoch": 0.26, + "grad_norm": 1.0936377863711764, + "learning_rate": 1.731745609975926e-05, + "loss": 0.9617, + "step": 11102 + }, + { + "epoch": 0.26, + "grad_norm": 2.159178388544293, + "learning_rate": 1.7316936006573994e-05, + "loss": 1.0163, + "step": 11103 + }, + { + "epoch": 0.26, + "grad_norm": 2.695550571773757, + "learning_rate": 1.731641587078721e-05, + "loss": 1.0776, + "step": 11104 + }, + { + "epoch": 0.26, + "grad_norm": 2.5508972132663956, + "learning_rate": 1.7315895692401937e-05, + "loss": 1.0543, + "step": 11105 + }, + { + "epoch": 0.26, + "grad_norm": 2.287578470522162, + "learning_rate": 1.7315375471421203e-05, + "loss": 0.9859, + "step": 11106 + }, + { + "epoch": 0.26, + "grad_norm": 2.136106019782012, + "learning_rate": 1.7314855207848037e-05, + "loss": 1.0531, + "step": 11107 + }, + { + "epoch": 0.26, + "grad_norm": 2.1859539770774292, + "learning_rate": 1.731433490168547e-05, + "loss": 0.9835, + "step": 11108 + }, + { + "epoch": 0.26, + "grad_norm": 1.9406203860551123, + "learning_rate": 1.7313814552936527e-05, + "loss": 1.0763, + "step": 11109 + }, + { + "epoch": 0.26, + "grad_norm": 1.9496854465167786, + "learning_rate": 1.7313294161604236e-05, + "loss": 1.1125, + "step": 11110 + }, + { + "epoch": 0.26, + "grad_norm": 1.8259407530860006, + "learning_rate": 1.7312773727691637e-05, + "loss": 1.0298, + "step": 11111 + }, + { + "epoch": 0.26, + "grad_norm": 2.0894841621215066, + "learning_rate": 1.7312253251201754e-05, + "loss": 1.1138, + "step": 11112 + }, + { + "epoch": 0.26, + "grad_norm": 2.2436222782133775, + "learning_rate": 1.7311732732137617e-05, + "loss": 1.0703, + "step": 11113 + }, + { + "epoch": 0.26, + "grad_norm": 2.8400801183285473, + "learning_rate": 1.7311212170502254e-05, + "loss": 0.9673, + "step": 11114 + }, + { + "epoch": 0.26, + "grad_norm": 2.0731174944832005, + "learning_rate": 1.7310691566298702e-05, + "loss": 0.9825, + "step": 11115 + }, + { + "epoch": 0.26, + "grad_norm": 2.3777844473761935, + "learning_rate": 1.7310170919529993e-05, + "loss": 1.0075, + "step": 11116 + }, + { + "epoch": 0.26, + "grad_norm": 2.3052613156235977, + "learning_rate": 1.7309650230199148e-05, + "loss": 1.0475, + "step": 11117 + }, + { + "epoch": 0.26, + "grad_norm": 1.9020772268461712, + "learning_rate": 1.7309129498309208e-05, + "loss": 1.0099, + "step": 11118 + }, + { + "epoch": 0.26, + "grad_norm": 1.8670196536320802, + "learning_rate": 1.73086087238632e-05, + "loss": 0.9741, + "step": 11119 + }, + { + "epoch": 0.26, + "grad_norm": 1.060433756365779, + "learning_rate": 1.730808790686416e-05, + "loss": 0.9441, + "step": 11120 + }, + { + "epoch": 0.26, + "grad_norm": 2.27946929577768, + "learning_rate": 1.7307567047315118e-05, + "loss": 0.9075, + "step": 11121 + }, + { + "epoch": 0.26, + "grad_norm": 1.9106953655824221, + "learning_rate": 1.7307046145219108e-05, + "loss": 1.2334, + "step": 11122 + }, + { + "epoch": 0.26, + "grad_norm": 2.279040536709934, + "learning_rate": 1.730652520057916e-05, + "loss": 1.0146, + "step": 11123 + }, + { + "epoch": 0.26, + "grad_norm": 2.3903806335944697, + "learning_rate": 1.730600421339831e-05, + "loss": 1.0534, + "step": 11124 + }, + { + "epoch": 0.26, + "grad_norm": 2.114706335660681, + "learning_rate": 1.7305483183679592e-05, + "loss": 1.0153, + "step": 11125 + }, + { + "epoch": 0.26, + "grad_norm": 2.0056010827965864, + "learning_rate": 1.7304962111426036e-05, + "loss": 1.009, + "step": 11126 + }, + { + "epoch": 0.26, + "grad_norm": 2.3685492547390417, + "learning_rate": 1.730444099664068e-05, + "loss": 1.0501, + "step": 11127 + }, + { + "epoch": 0.26, + "grad_norm": 2.3567102740166406, + "learning_rate": 1.7303919839326554e-05, + "loss": 1.0596, + "step": 11128 + }, + { + "epoch": 0.26, + "grad_norm": 1.969924916297763, + "learning_rate": 1.7303398639486696e-05, + "loss": 0.9155, + "step": 11129 + }, + { + "epoch": 0.26, + "grad_norm": 1.9879964894376716, + "learning_rate": 1.7302877397124137e-05, + "loss": 0.9179, + "step": 11130 + }, + { + "epoch": 0.26, + "grad_norm": 1.9526203517437737, + "learning_rate": 1.7302356112241913e-05, + "loss": 1.0834, + "step": 11131 + }, + { + "epoch": 0.26, + "grad_norm": 2.418626786155992, + "learning_rate": 1.7301834784843064e-05, + "loss": 1.1267, + "step": 11132 + }, + { + "epoch": 0.26, + "grad_norm": 1.9884534630784045, + "learning_rate": 1.7301313414930618e-05, + "loss": 1.002, + "step": 11133 + }, + { + "epoch": 0.26, + "grad_norm": 2.219106432331183, + "learning_rate": 1.7300792002507614e-05, + "loss": 1.1111, + "step": 11134 + }, + { + "epoch": 0.26, + "grad_norm": 2.6140681253363867, + "learning_rate": 1.730027054757709e-05, + "loss": 0.9257, + "step": 11135 + }, + { + "epoch": 0.26, + "grad_norm": 1.1686782503958257, + "learning_rate": 1.7299749050142074e-05, + "loss": 0.9682, + "step": 11136 + }, + { + "epoch": 0.26, + "grad_norm": 2.6857485112007677, + "learning_rate": 1.7299227510205613e-05, + "loss": 1.0046, + "step": 11137 + }, + { + "epoch": 0.26, + "grad_norm": 1.100307734283632, + "learning_rate": 1.7298705927770737e-05, + "loss": 0.9162, + "step": 11138 + }, + { + "epoch": 0.26, + "grad_norm": 2.275775282603493, + "learning_rate": 1.7298184302840485e-05, + "loss": 0.9939, + "step": 11139 + }, + { + "epoch": 0.26, + "grad_norm": 2.8694542237256093, + "learning_rate": 1.729766263541789e-05, + "loss": 0.9917, + "step": 11140 + }, + { + "epoch": 0.26, + "grad_norm": 2.342772420816968, + "learning_rate": 1.7297140925505996e-05, + "loss": 1.098, + "step": 11141 + }, + { + "epoch": 0.26, + "grad_norm": 2.0213430117968363, + "learning_rate": 1.7296619173107837e-05, + "loss": 1.0129, + "step": 11142 + }, + { + "epoch": 0.26, + "grad_norm": 2.1208452594870995, + "learning_rate": 1.7296097378226452e-05, + "loss": 1.0044, + "step": 11143 + }, + { + "epoch": 0.26, + "grad_norm": 2.2693923224015493, + "learning_rate": 1.7295575540864878e-05, + "loss": 0.9647, + "step": 11144 + }, + { + "epoch": 0.26, + "grad_norm": 2.095327363518879, + "learning_rate": 1.729505366102615e-05, + "loss": 0.9205, + "step": 11145 + }, + { + "epoch": 0.26, + "grad_norm": 2.1565807018097836, + "learning_rate": 1.7294531738713313e-05, + "loss": 1.1097, + "step": 11146 + }, + { + "epoch": 0.26, + "grad_norm": 1.8556570283832536, + "learning_rate": 1.7294009773929405e-05, + "loss": 1.1741, + "step": 11147 + }, + { + "epoch": 0.26, + "grad_norm": 2.0357868121295004, + "learning_rate": 1.7293487766677457e-05, + "loss": 0.9465, + "step": 11148 + }, + { + "epoch": 0.26, + "grad_norm": 2.6053836260277237, + "learning_rate": 1.7292965716960518e-05, + "loss": 1.0822, + "step": 11149 + }, + { + "epoch": 0.26, + "grad_norm": 1.843999550790122, + "learning_rate": 1.7292443624781624e-05, + "loss": 1.0797, + "step": 11150 + }, + { + "epoch": 0.26, + "grad_norm": 1.9599076717847943, + "learning_rate": 1.729192149014381e-05, + "loss": 0.8936, + "step": 11151 + }, + { + "epoch": 0.26, + "grad_norm": 1.9753827494014884, + "learning_rate": 1.729139931305013e-05, + "loss": 0.9161, + "step": 11152 + }, + { + "epoch": 0.26, + "grad_norm": 2.036941663541986, + "learning_rate": 1.7290877093503606e-05, + "loss": 1.0112, + "step": 11153 + }, + { + "epoch": 0.26, + "grad_norm": 2.6149883679826784, + "learning_rate": 1.729035483150729e-05, + "loss": 1.0423, + "step": 11154 + }, + { + "epoch": 0.26, + "grad_norm": 2.9166413018884874, + "learning_rate": 1.728983252706422e-05, + "loss": 0.948, + "step": 11155 + }, + { + "epoch": 0.26, + "grad_norm": 2.0146451600797457, + "learning_rate": 1.7289310180177438e-05, + "loss": 1.0134, + "step": 11156 + }, + { + "epoch": 0.26, + "grad_norm": 2.127076834810739, + "learning_rate": 1.7288787790849984e-05, + "loss": 1.0579, + "step": 11157 + }, + { + "epoch": 0.26, + "grad_norm": 1.2253184832209414, + "learning_rate": 1.72882653590849e-05, + "loss": 0.956, + "step": 11158 + }, + { + "epoch": 0.26, + "grad_norm": 2.0577027492494535, + "learning_rate": 1.7287742884885225e-05, + "loss": 1.1057, + "step": 11159 + }, + { + "epoch": 0.26, + "grad_norm": 3.3326351221180266, + "learning_rate": 1.728722036825401e-05, + "loss": 0.9472, + "step": 11160 + }, + { + "epoch": 0.26, + "grad_norm": 2.3537027798035592, + "learning_rate": 1.7286697809194285e-05, + "loss": 1.0038, + "step": 11161 + }, + { + "epoch": 0.26, + "grad_norm": 2.0608062099587636, + "learning_rate": 1.72861752077091e-05, + "loss": 1.0513, + "step": 11162 + }, + { + "epoch": 0.26, + "grad_norm": 3.1889998360020773, + "learning_rate": 1.7285652563801493e-05, + "loss": 0.9512, + "step": 11163 + }, + { + "epoch": 0.26, + "grad_norm": 1.9628753240425472, + "learning_rate": 1.7285129877474516e-05, + "loss": 1.0364, + "step": 11164 + }, + { + "epoch": 0.26, + "grad_norm": 2.1657625282462774, + "learning_rate": 1.7284607148731206e-05, + "loss": 1.0378, + "step": 11165 + }, + { + "epoch": 0.26, + "grad_norm": 2.0728775051028863, + "learning_rate": 1.7284084377574605e-05, + "loss": 1.0882, + "step": 11166 + }, + { + "epoch": 0.26, + "grad_norm": 1.9991410379941563, + "learning_rate": 1.728356156400776e-05, + "loss": 1.1281, + "step": 11167 + }, + { + "epoch": 0.26, + "grad_norm": 2.5288566023900088, + "learning_rate": 1.7283038708033707e-05, + "loss": 1.0576, + "step": 11168 + }, + { + "epoch": 0.26, + "grad_norm": 2.055041221873008, + "learning_rate": 1.72825158096555e-05, + "loss": 1.0572, + "step": 11169 + }, + { + "epoch": 0.26, + "grad_norm": 1.9709813883944038, + "learning_rate": 1.728199286887619e-05, + "loss": 1.1787, + "step": 11170 + }, + { + "epoch": 0.26, + "grad_norm": 2.109759424918168, + "learning_rate": 1.72814698856988e-05, + "loss": 1.1621, + "step": 11171 + }, + { + "epoch": 0.26, + "grad_norm": 2.166574185125438, + "learning_rate": 1.7280946860126394e-05, + "loss": 1.0353, + "step": 11172 + }, + { + "epoch": 0.26, + "grad_norm": 1.9686070850489545, + "learning_rate": 1.7280423792162008e-05, + "loss": 1.1563, + "step": 11173 + }, + { + "epoch": 0.26, + "grad_norm": 2.0025247832876243, + "learning_rate": 1.727990068180869e-05, + "loss": 0.9706, + "step": 11174 + }, + { + "epoch": 0.26, + "grad_norm": 2.113158378118328, + "learning_rate": 1.7279377529069484e-05, + "loss": 1.0007, + "step": 11175 + }, + { + "epoch": 0.26, + "grad_norm": 1.7578075679074079, + "learning_rate": 1.7278854333947438e-05, + "loss": 1.0125, + "step": 11176 + }, + { + "epoch": 0.26, + "grad_norm": 1.9963467913127486, + "learning_rate": 1.72783310964456e-05, + "loss": 1.0932, + "step": 11177 + }, + { + "epoch": 0.26, + "grad_norm": 2.707302376833398, + "learning_rate": 1.727780781656701e-05, + "loss": 1.0684, + "step": 11178 + }, + { + "epoch": 0.26, + "grad_norm": 2.325351475931253, + "learning_rate": 1.727728449431472e-05, + "loss": 1.074, + "step": 11179 + }, + { + "epoch": 0.26, + "grad_norm": 1.8337512199439485, + "learning_rate": 1.7276761129691776e-05, + "loss": 1.0964, + "step": 11180 + }, + { + "epoch": 0.26, + "grad_norm": 1.8513975510706562, + "learning_rate": 1.7276237722701227e-05, + "loss": 0.9125, + "step": 11181 + }, + { + "epoch": 0.26, + "grad_norm": 3.19320109803393, + "learning_rate": 1.7275714273346117e-05, + "loss": 1.0136, + "step": 11182 + }, + { + "epoch": 0.26, + "grad_norm": 2.3320845079890096, + "learning_rate": 1.7275190781629494e-05, + "loss": 1.1198, + "step": 11183 + }, + { + "epoch": 0.26, + "grad_norm": 1.1829166222276486, + "learning_rate": 1.7274667247554407e-05, + "loss": 0.951, + "step": 11184 + }, + { + "epoch": 0.26, + "grad_norm": 1.9560352399819156, + "learning_rate": 1.7274143671123905e-05, + "loss": 1.1701, + "step": 11185 + }, + { + "epoch": 0.26, + "grad_norm": 3.6001548280086486, + "learning_rate": 1.7273620052341037e-05, + "loss": 1.1262, + "step": 11186 + }, + { + "epoch": 0.26, + "grad_norm": 1.9939249187771368, + "learning_rate": 1.727309639120885e-05, + "loss": 1.0348, + "step": 11187 + }, + { + "epoch": 0.26, + "grad_norm": 1.8377531446902284, + "learning_rate": 1.727257268773039e-05, + "loss": 1.0957, + "step": 11188 + }, + { + "epoch": 0.26, + "grad_norm": 2.0077162265656856, + "learning_rate": 1.7272048941908713e-05, + "loss": 1.0538, + "step": 11189 + }, + { + "epoch": 0.26, + "grad_norm": 1.941313298948569, + "learning_rate": 1.7271525153746865e-05, + "loss": 0.9588, + "step": 11190 + }, + { + "epoch": 0.26, + "grad_norm": 1.971908656769549, + "learning_rate": 1.7271001323247892e-05, + "loss": 1.1529, + "step": 11191 + }, + { + "epoch": 0.26, + "grad_norm": 2.0289110528525893, + "learning_rate": 1.727047745041485e-05, + "loss": 1.1082, + "step": 11192 + }, + { + "epoch": 0.26, + "grad_norm": 2.3712800088685575, + "learning_rate": 1.7269953535250782e-05, + "loss": 1.1027, + "step": 11193 + }, + { + "epoch": 0.26, + "grad_norm": 2.456942149769468, + "learning_rate": 1.726942957775875e-05, + "loss": 1.043, + "step": 11194 + }, + { + "epoch": 0.26, + "grad_norm": 3.5889127022588356, + "learning_rate": 1.726890557794179e-05, + "loss": 1.1798, + "step": 11195 + }, + { + "epoch": 0.26, + "grad_norm": 2.257771272802638, + "learning_rate": 1.7268381535802966e-05, + "loss": 1.054, + "step": 11196 + }, + { + "epoch": 0.26, + "grad_norm": 2.104716499327247, + "learning_rate": 1.726785745134532e-05, + "loss": 1.0182, + "step": 11197 + }, + { + "epoch": 0.26, + "grad_norm": 1.0616442659626724, + "learning_rate": 1.7267333324571912e-05, + "loss": 0.9577, + "step": 11198 + }, + { + "epoch": 0.26, + "grad_norm": 1.957961908239952, + "learning_rate": 1.7266809155485783e-05, + "loss": 1.0112, + "step": 11199 + }, + { + "epoch": 0.26, + "grad_norm": 3.999451609921155, + "learning_rate": 1.7266284944089994e-05, + "loss": 0.9834, + "step": 11200 + }, + { + "epoch": 0.26, + "grad_norm": 2.127913610542947, + "learning_rate": 1.7265760690387594e-05, + "loss": 1.0511, + "step": 11201 + }, + { + "epoch": 0.26, + "grad_norm": 1.9359035501330037, + "learning_rate": 1.7265236394381634e-05, + "loss": 1.1088, + "step": 11202 + }, + { + "epoch": 0.26, + "grad_norm": 2.106680637172741, + "learning_rate": 1.7264712056075167e-05, + "loss": 1.0974, + "step": 11203 + }, + { + "epoch": 0.26, + "grad_norm": 2.0949174371644257, + "learning_rate": 1.7264187675471246e-05, + "loss": 1.0365, + "step": 11204 + }, + { + "epoch": 0.26, + "grad_norm": 1.1198255049828774, + "learning_rate": 1.7263663252572923e-05, + "loss": 0.8949, + "step": 11205 + }, + { + "epoch": 0.26, + "grad_norm": 2.359646586847114, + "learning_rate": 1.7263138787383255e-05, + "loss": 1.0606, + "step": 11206 + }, + { + "epoch": 0.26, + "grad_norm": 1.960009917235001, + "learning_rate": 1.7262614279905295e-05, + "loss": 0.9283, + "step": 11207 + }, + { + "epoch": 0.26, + "grad_norm": 2.012944496430695, + "learning_rate": 1.7262089730142094e-05, + "loss": 1.001, + "step": 11208 + }, + { + "epoch": 0.26, + "grad_norm": 1.9953253773249087, + "learning_rate": 1.7261565138096706e-05, + "loss": 1.1098, + "step": 11209 + }, + { + "epoch": 0.26, + "grad_norm": 1.9818613804454415, + "learning_rate": 1.7261040503772187e-05, + "loss": 1.046, + "step": 11210 + }, + { + "epoch": 0.26, + "grad_norm": 2.150735062022872, + "learning_rate": 1.7260515827171596e-05, + "loss": 1.0315, + "step": 11211 + }, + { + "epoch": 0.26, + "grad_norm": 2.121726151126779, + "learning_rate": 1.7259991108297976e-05, + "loss": 1.1373, + "step": 11212 + }, + { + "epoch": 0.26, + "grad_norm": 2.8597037643102303, + "learning_rate": 1.7259466347154393e-05, + "loss": 1.0345, + "step": 11213 + }, + { + "epoch": 0.26, + "grad_norm": 2.3566819227116693, + "learning_rate": 1.72589415437439e-05, + "loss": 0.9595, + "step": 11214 + }, + { + "epoch": 0.26, + "grad_norm": 2.2111687666686812, + "learning_rate": 1.725841669806955e-05, + "loss": 0.9904, + "step": 11215 + }, + { + "epoch": 0.26, + "grad_norm": 2.1198217792587455, + "learning_rate": 1.72578918101344e-05, + "loss": 1.033, + "step": 11216 + }, + { + "epoch": 0.26, + "grad_norm": 2.026376791609559, + "learning_rate": 1.7257366879941508e-05, + "loss": 1.0285, + "step": 11217 + }, + { + "epoch": 0.26, + "grad_norm": 1.8038475200995805, + "learning_rate": 1.7256841907493928e-05, + "loss": 1.0582, + "step": 11218 + }, + { + "epoch": 0.26, + "grad_norm": 2.7813238782225578, + "learning_rate": 1.7256316892794712e-05, + "loss": 1.0983, + "step": 11219 + }, + { + "epoch": 0.26, + "grad_norm": 2.133938769930824, + "learning_rate": 1.7255791835846926e-05, + "loss": 1.0837, + "step": 11220 + }, + { + "epoch": 0.26, + "grad_norm": 1.956273299645381, + "learning_rate": 1.725526673665362e-05, + "loss": 1.0102, + "step": 11221 + }, + { + "epoch": 0.26, + "grad_norm": 2.016978359812213, + "learning_rate": 1.7254741595217856e-05, + "loss": 1.0384, + "step": 11222 + }, + { + "epoch": 0.26, + "grad_norm": 2.0115942026444014, + "learning_rate": 1.725421641154269e-05, + "loss": 1.0833, + "step": 11223 + }, + { + "epoch": 0.26, + "grad_norm": 2.0484858419926817, + "learning_rate": 1.725369118563118e-05, + "loss": 1.0024, + "step": 11224 + }, + { + "epoch": 0.26, + "grad_norm": 2.3049855228466116, + "learning_rate": 1.725316591748638e-05, + "loss": 1.0654, + "step": 11225 + }, + { + "epoch": 0.26, + "grad_norm": 2.399616755532091, + "learning_rate": 1.7252640607111354e-05, + "loss": 1.0178, + "step": 11226 + }, + { + "epoch": 0.26, + "grad_norm": 2.6245822261831786, + "learning_rate": 1.725211525450916e-05, + "loss": 1.0042, + "step": 11227 + }, + { + "epoch": 0.26, + "grad_norm": 2.0955240094432415, + "learning_rate": 1.725158985968285e-05, + "loss": 0.9663, + "step": 11228 + }, + { + "epoch": 0.26, + "grad_norm": 4.4262121644750065, + "learning_rate": 1.725106442263549e-05, + "loss": 1.0255, + "step": 11229 + }, + { + "epoch": 0.26, + "grad_norm": 1.1248293505490288, + "learning_rate": 1.7250538943370142e-05, + "loss": 0.9672, + "step": 11230 + }, + { + "epoch": 0.26, + "grad_norm": 2.0086541659079824, + "learning_rate": 1.7250013421889857e-05, + "loss": 1.1184, + "step": 11231 + }, + { + "epoch": 0.26, + "grad_norm": 2.7654406259119564, + "learning_rate": 1.7249487858197698e-05, + "loss": 1.1498, + "step": 11232 + }, + { + "epoch": 0.26, + "grad_norm": 1.7776045387849624, + "learning_rate": 1.7248962252296725e-05, + "loss": 1.0508, + "step": 11233 + }, + { + "epoch": 0.26, + "grad_norm": 1.9622462003606231, + "learning_rate": 1.724843660419e-05, + "loss": 1.0062, + "step": 11234 + }, + { + "epoch": 0.26, + "grad_norm": 2.209241231843165, + "learning_rate": 1.7247910913880583e-05, + "loss": 1.2203, + "step": 11235 + }, + { + "epoch": 0.26, + "grad_norm": 2.090521033113536, + "learning_rate": 1.7247385181371535e-05, + "loss": 1.1066, + "step": 11236 + }, + { + "epoch": 0.26, + "grad_norm": 1.9774344471939889, + "learning_rate": 1.7246859406665916e-05, + "loss": 1.1906, + "step": 11237 + }, + { + "epoch": 0.26, + "grad_norm": 2.089039860341394, + "learning_rate": 1.7246333589766786e-05, + "loss": 1.2319, + "step": 11238 + }, + { + "epoch": 0.26, + "grad_norm": 1.936424830992595, + "learning_rate": 1.7245807730677206e-05, + "loss": 1.0046, + "step": 11239 + }, + { + "epoch": 0.26, + "grad_norm": 1.9914669375798963, + "learning_rate": 1.724528182940024e-05, + "loss": 1.0797, + "step": 11240 + }, + { + "epoch": 0.26, + "grad_norm": 2.1121232691839253, + "learning_rate": 1.724475588593895e-05, + "loss": 0.8391, + "step": 11241 + }, + { + "epoch": 0.26, + "grad_norm": 2.0136321521304246, + "learning_rate": 1.7244229900296398e-05, + "loss": 1.0397, + "step": 11242 + }, + { + "epoch": 0.26, + "grad_norm": 1.085215785185062, + "learning_rate": 1.724370387247565e-05, + "loss": 1.0462, + "step": 11243 + }, + { + "epoch": 0.26, + "grad_norm": 2.493782058570405, + "learning_rate": 1.7243177802479758e-05, + "loss": 1.1106, + "step": 11244 + }, + { + "epoch": 0.26, + "grad_norm": 2.0935224405082486, + "learning_rate": 1.7242651690311794e-05, + "loss": 1.1426, + "step": 11245 + }, + { + "epoch": 0.26, + "grad_norm": 2.0933000016276635, + "learning_rate": 1.724212553597482e-05, + "loss": 1.0511, + "step": 11246 + }, + { + "epoch": 0.26, + "grad_norm": 2.9454337571458504, + "learning_rate": 1.7241599339471894e-05, + "loss": 1.1325, + "step": 11247 + }, + { + "epoch": 0.26, + "grad_norm": 2.2720881060148908, + "learning_rate": 1.724107310080609e-05, + "loss": 1.0429, + "step": 11248 + }, + { + "epoch": 0.27, + "grad_norm": 1.8253714742804923, + "learning_rate": 1.724054681998046e-05, + "loss": 1.0832, + "step": 11249 + }, + { + "epoch": 0.27, + "grad_norm": 1.8702158431283096, + "learning_rate": 1.724002049699808e-05, + "loss": 1.0629, + "step": 11250 + }, + { + "epoch": 0.27, + "grad_norm": 1.8552074772314444, + "learning_rate": 1.7239494131862003e-05, + "loss": 1.0949, + "step": 11251 + }, + { + "epoch": 0.27, + "grad_norm": 1.8661051745914643, + "learning_rate": 1.7238967724575305e-05, + "loss": 0.963, + "step": 11252 + }, + { + "epoch": 0.27, + "grad_norm": 2.3076354981356633, + "learning_rate": 1.723844127514104e-05, + "loss": 1.2781, + "step": 11253 + }, + { + "epoch": 0.27, + "grad_norm": 1.9506052612599605, + "learning_rate": 1.7237914783562283e-05, + "loss": 1.1073, + "step": 11254 + }, + { + "epoch": 0.27, + "grad_norm": 1.8264956721453685, + "learning_rate": 1.723738824984209e-05, + "loss": 0.9134, + "step": 11255 + }, + { + "epoch": 0.27, + "grad_norm": 2.2593082138757277, + "learning_rate": 1.7236861673983534e-05, + "loss": 1.0572, + "step": 11256 + }, + { + "epoch": 0.27, + "grad_norm": 2.0656822443597793, + "learning_rate": 1.7236335055989676e-05, + "loss": 1.0725, + "step": 11257 + }, + { + "epoch": 0.27, + "grad_norm": 3.0366526977911286, + "learning_rate": 1.7235808395863586e-05, + "loss": 1.0404, + "step": 11258 + }, + { + "epoch": 0.27, + "grad_norm": 2.030856727391825, + "learning_rate": 1.723528169360833e-05, + "loss": 1.0977, + "step": 11259 + }, + { + "epoch": 0.27, + "grad_norm": 1.972291537895372, + "learning_rate": 1.723475494922697e-05, + "loss": 1.0444, + "step": 11260 + }, + { + "epoch": 0.27, + "grad_norm": 1.155935342122281, + "learning_rate": 1.723422816272258e-05, + "loss": 1.0289, + "step": 11261 + }, + { + "epoch": 0.27, + "grad_norm": 1.779710047383735, + "learning_rate": 1.723370133409822e-05, + "loss": 1.0135, + "step": 11262 + }, + { + "epoch": 0.27, + "grad_norm": 1.1764858023474738, + "learning_rate": 1.723317446335696e-05, + "loss": 1.0223, + "step": 11263 + }, + { + "epoch": 0.27, + "grad_norm": 1.190432483976489, + "learning_rate": 1.7232647550501875e-05, + "loss": 0.9691, + "step": 11264 + }, + { + "epoch": 0.27, + "grad_norm": 2.0682434146558726, + "learning_rate": 1.723212059553602e-05, + "loss": 0.9854, + "step": 11265 + }, + { + "epoch": 0.27, + "grad_norm": 1.8495986722532785, + "learning_rate": 1.723159359846247e-05, + "loss": 1.0137, + "step": 11266 + }, + { + "epoch": 0.27, + "grad_norm": 1.159954831631891, + "learning_rate": 1.7231066559284293e-05, + "loss": 0.98, + "step": 11267 + }, + { + "epoch": 0.27, + "grad_norm": 2.380249252820152, + "learning_rate": 1.723053947800456e-05, + "loss": 1.0968, + "step": 11268 + }, + { + "epoch": 0.27, + "grad_norm": 2.265187961770074, + "learning_rate": 1.7230012354626334e-05, + "loss": 1.0262, + "step": 11269 + }, + { + "epoch": 0.27, + "grad_norm": 2.1331649974114786, + "learning_rate": 1.722948518915269e-05, + "loss": 1.0106, + "step": 11270 + }, + { + "epoch": 0.27, + "grad_norm": 2.102432323347419, + "learning_rate": 1.722895798158669e-05, + "loss": 1.1434, + "step": 11271 + }, + { + "epoch": 0.27, + "grad_norm": 2.1937349258868473, + "learning_rate": 1.7228430731931407e-05, + "loss": 1.0337, + "step": 11272 + }, + { + "epoch": 0.27, + "grad_norm": 2.288356800283505, + "learning_rate": 1.7227903440189914e-05, + "loss": 0.9725, + "step": 11273 + }, + { + "epoch": 0.27, + "grad_norm": 2.027425758360596, + "learning_rate": 1.722737610636528e-05, + "loss": 0.9158, + "step": 11274 + }, + { + "epoch": 0.27, + "grad_norm": 2.1922322145066544, + "learning_rate": 1.7226848730460574e-05, + "loss": 1.0701, + "step": 11275 + }, + { + "epoch": 0.27, + "grad_norm": 2.056444995062926, + "learning_rate": 1.722632131247887e-05, + "loss": 1.0854, + "step": 11276 + }, + { + "epoch": 0.27, + "grad_norm": 2.017432331560389, + "learning_rate": 1.722579385242323e-05, + "loss": 0.9902, + "step": 11277 + }, + { + "epoch": 0.27, + "grad_norm": 1.088959650103917, + "learning_rate": 1.7225266350296733e-05, + "loss": 0.9138, + "step": 11278 + }, + { + "epoch": 0.27, + "grad_norm": 2.724653943186114, + "learning_rate": 1.7224738806102447e-05, + "loss": 0.8962, + "step": 11279 + }, + { + "epoch": 0.27, + "grad_norm": 2.1081954950924713, + "learning_rate": 1.7224211219843444e-05, + "loss": 1.0518, + "step": 11280 + }, + { + "epoch": 0.27, + "grad_norm": 2.2136359707177498, + "learning_rate": 1.7223683591522798e-05, + "loss": 1.144, + "step": 11281 + }, + { + "epoch": 0.27, + "grad_norm": 1.886745289429305, + "learning_rate": 1.7223155921143578e-05, + "loss": 0.9478, + "step": 11282 + }, + { + "epoch": 0.27, + "grad_norm": 2.362018163926713, + "learning_rate": 1.7222628208708857e-05, + "loss": 1.0823, + "step": 11283 + }, + { + "epoch": 0.27, + "grad_norm": 2.1128333160395214, + "learning_rate": 1.722210045422171e-05, + "loss": 1.0703, + "step": 11284 + }, + { + "epoch": 0.27, + "grad_norm": 2.00505589923955, + "learning_rate": 1.7221572657685205e-05, + "loss": 1.0522, + "step": 11285 + }, + { + "epoch": 0.27, + "grad_norm": 2.185391967212235, + "learning_rate": 1.722104481910242e-05, + "loss": 1.0295, + "step": 11286 + }, + { + "epoch": 0.27, + "grad_norm": 1.8824445296404793, + "learning_rate": 1.722051693847642e-05, + "loss": 1.0358, + "step": 11287 + }, + { + "epoch": 0.27, + "grad_norm": 1.1128624425294142, + "learning_rate": 1.721998901581029e-05, + "loss": 0.9198, + "step": 11288 + }, + { + "epoch": 0.27, + "grad_norm": 1.9704844395516756, + "learning_rate": 1.7219461051107102e-05, + "loss": 1.1363, + "step": 11289 + }, + { + "epoch": 0.27, + "grad_norm": 2.1380233288619515, + "learning_rate": 1.721893304436992e-05, + "loss": 1.1202, + "step": 11290 + }, + { + "epoch": 0.27, + "grad_norm": 3.033133315912205, + "learning_rate": 1.7218404995601823e-05, + "loss": 1.1749, + "step": 11291 + }, + { + "epoch": 0.27, + "grad_norm": 2.1277914948821315, + "learning_rate": 1.721787690480589e-05, + "loss": 1.0617, + "step": 11292 + }, + { + "epoch": 0.27, + "grad_norm": 2.02045541608251, + "learning_rate": 1.7217348771985194e-05, + "loss": 0.9931, + "step": 11293 + }, + { + "epoch": 0.27, + "grad_norm": 2.139577968578247, + "learning_rate": 1.7216820597142804e-05, + "loss": 0.9852, + "step": 11294 + }, + { + "epoch": 0.27, + "grad_norm": 1.9443037587486687, + "learning_rate": 1.7216292380281802e-05, + "loss": 1.0681, + "step": 11295 + }, + { + "epoch": 0.27, + "grad_norm": 2.149975959139988, + "learning_rate": 1.721576412140526e-05, + "loss": 1.0351, + "step": 11296 + }, + { + "epoch": 0.27, + "grad_norm": 2.1535488592339402, + "learning_rate": 1.721523582051626e-05, + "loss": 0.996, + "step": 11297 + }, + { + "epoch": 0.27, + "grad_norm": 2.6789726837571877, + "learning_rate": 1.7214707477617866e-05, + "loss": 0.9987, + "step": 11298 + }, + { + "epoch": 0.27, + "grad_norm": 2.072500826748613, + "learning_rate": 1.7214179092713165e-05, + "loss": 0.8957, + "step": 11299 + }, + { + "epoch": 0.27, + "grad_norm": 1.8527267799074985, + "learning_rate": 1.7213650665805224e-05, + "loss": 1.0251, + "step": 11300 + }, + { + "epoch": 0.27, + "grad_norm": 2.18519945708212, + "learning_rate": 1.7213122196897135e-05, + "loss": 1.014, + "step": 11301 + }, + { + "epoch": 0.27, + "grad_norm": 1.9464492975933068, + "learning_rate": 1.7212593685991956e-05, + "loss": 1.1634, + "step": 11302 + }, + { + "epoch": 0.27, + "grad_norm": 1.085344611981941, + "learning_rate": 1.7212065133092778e-05, + "loss": 0.9517, + "step": 11303 + }, + { + "epoch": 0.27, + "grad_norm": 1.0965133021263944, + "learning_rate": 1.7211536538202673e-05, + "loss": 0.9893, + "step": 11304 + }, + { + "epoch": 0.27, + "grad_norm": 2.1401374052300333, + "learning_rate": 1.7211007901324714e-05, + "loss": 1.0455, + "step": 11305 + }, + { + "epoch": 0.27, + "grad_norm": 1.9258353488579472, + "learning_rate": 1.721047922246199e-05, + "loss": 0.9784, + "step": 11306 + }, + { + "epoch": 0.27, + "grad_norm": 2.0892123768200497, + "learning_rate": 1.720995050161757e-05, + "loss": 1.0695, + "step": 11307 + }, + { + "epoch": 0.27, + "grad_norm": 1.988627025706952, + "learning_rate": 1.7209421738794536e-05, + "loss": 1.0671, + "step": 11308 + }, + { + "epoch": 0.27, + "grad_norm": 2.1157767987493217, + "learning_rate": 1.720889293399597e-05, + "loss": 0.9769, + "step": 11309 + }, + { + "epoch": 0.27, + "grad_norm": 2.258792379329834, + "learning_rate": 1.720836408722494e-05, + "loss": 1.1512, + "step": 11310 + }, + { + "epoch": 0.27, + "grad_norm": 1.1105534274410787, + "learning_rate": 1.7207835198484537e-05, + "loss": 0.9505, + "step": 11311 + }, + { + "epoch": 0.27, + "grad_norm": 2.2831766167383494, + "learning_rate": 1.720730626777783e-05, + "loss": 1.0226, + "step": 11312 + }, + { + "epoch": 0.27, + "grad_norm": 1.8949729219025713, + "learning_rate": 1.720677729510791e-05, + "loss": 1.0011, + "step": 11313 + }, + { + "epoch": 0.27, + "grad_norm": 2.029624164387295, + "learning_rate": 1.720624828047785e-05, + "loss": 1.0217, + "step": 11314 + }, + { + "epoch": 0.27, + "grad_norm": 1.9213621768086875, + "learning_rate": 1.7205719223890732e-05, + "loss": 0.9545, + "step": 11315 + }, + { + "epoch": 0.27, + "grad_norm": 2.110635318627009, + "learning_rate": 1.7205190125349633e-05, + "loss": 0.9311, + "step": 11316 + }, + { + "epoch": 0.27, + "grad_norm": 2.034105483485876, + "learning_rate": 1.7204660984857638e-05, + "loss": 0.9339, + "step": 11317 + }, + { + "epoch": 0.27, + "grad_norm": 2.0399167441671042, + "learning_rate": 1.7204131802417826e-05, + "loss": 1.0784, + "step": 11318 + }, + { + "epoch": 0.27, + "grad_norm": 2.173987709106016, + "learning_rate": 1.7203602578033275e-05, + "loss": 0.9781, + "step": 11319 + }, + { + "epoch": 0.27, + "grad_norm": 1.9308832636992863, + "learning_rate": 1.7203073311707072e-05, + "loss": 0.9856, + "step": 11320 + }, + { + "epoch": 0.27, + "grad_norm": 2.4037995397353455, + "learning_rate": 1.720254400344229e-05, + "loss": 0.93, + "step": 11321 + }, + { + "epoch": 0.27, + "grad_norm": 2.1969493894832515, + "learning_rate": 1.720201465324202e-05, + "loss": 1.0253, + "step": 11322 + }, + { + "epoch": 0.27, + "grad_norm": 2.0984228584074507, + "learning_rate": 1.7201485261109345e-05, + "loss": 1.0797, + "step": 11323 + }, + { + "epoch": 0.27, + "grad_norm": 2.1948213063815847, + "learning_rate": 1.720095582704734e-05, + "loss": 1.0809, + "step": 11324 + }, + { + "epoch": 0.27, + "grad_norm": 2.288883503711603, + "learning_rate": 1.7200426351059093e-05, + "loss": 1.0467, + "step": 11325 + }, + { + "epoch": 0.27, + "grad_norm": 1.9139385632123203, + "learning_rate": 1.719989683314768e-05, + "loss": 1.0682, + "step": 11326 + }, + { + "epoch": 0.27, + "grad_norm": 1.8822279570292229, + "learning_rate": 1.7199367273316187e-05, + "loss": 1.0519, + "step": 11327 + }, + { + "epoch": 0.27, + "grad_norm": 1.9244736055143559, + "learning_rate": 1.71988376715677e-05, + "loss": 1.1154, + "step": 11328 + }, + { + "epoch": 0.27, + "grad_norm": 1.97732567036143, + "learning_rate": 1.71983080279053e-05, + "loss": 1.1315, + "step": 11329 + }, + { + "epoch": 0.27, + "grad_norm": 1.901053674590005, + "learning_rate": 1.7197778342332075e-05, + "loss": 0.9393, + "step": 11330 + }, + { + "epoch": 0.27, + "grad_norm": 2.075084945436561, + "learning_rate": 1.7197248614851103e-05, + "loss": 1.0261, + "step": 11331 + }, + { + "epoch": 0.27, + "grad_norm": 2.066093311158311, + "learning_rate": 1.7196718845465472e-05, + "loss": 1.0373, + "step": 11332 + }, + { + "epoch": 0.27, + "grad_norm": 2.3272789299445096, + "learning_rate": 1.7196189034178267e-05, + "loss": 1.0512, + "step": 11333 + }, + { + "epoch": 0.27, + "grad_norm": 2.317245377846998, + "learning_rate": 1.719565918099257e-05, + "loss": 1.0383, + "step": 11334 + }, + { + "epoch": 0.27, + "grad_norm": 2.010144281618102, + "learning_rate": 1.7195129285911465e-05, + "loss": 1.1284, + "step": 11335 + }, + { + "epoch": 0.27, + "grad_norm": 2.0431343067603556, + "learning_rate": 1.7194599348938042e-05, + "loss": 1.0164, + "step": 11336 + }, + { + "epoch": 0.27, + "grad_norm": 1.8607643229605506, + "learning_rate": 1.719406937007538e-05, + "loss": 1.0485, + "step": 11337 + }, + { + "epoch": 0.27, + "grad_norm": 1.896331980241498, + "learning_rate": 1.7193539349326573e-05, + "loss": 1.0125, + "step": 11338 + }, + { + "epoch": 0.27, + "grad_norm": 2.482950466807774, + "learning_rate": 1.71930092866947e-05, + "loss": 1.0107, + "step": 11339 + }, + { + "epoch": 0.27, + "grad_norm": 2.034985367864195, + "learning_rate": 1.7192479182182853e-05, + "loss": 1.0873, + "step": 11340 + }, + { + "epoch": 0.27, + "grad_norm": 2.3660215869999752, + "learning_rate": 1.7191949035794113e-05, + "loss": 1.0341, + "step": 11341 + }, + { + "epoch": 0.27, + "grad_norm": 2.0590104429755645, + "learning_rate": 1.7191418847531567e-05, + "loss": 1.0354, + "step": 11342 + }, + { + "epoch": 0.27, + "grad_norm": 2.276794947564913, + "learning_rate": 1.7190888617398306e-05, + "loss": 1.2021, + "step": 11343 + }, + { + "epoch": 0.27, + "grad_norm": 1.926698068928972, + "learning_rate": 1.7190358345397412e-05, + "loss": 1.0917, + "step": 11344 + }, + { + "epoch": 0.27, + "grad_norm": 2.045733767223607, + "learning_rate": 1.718982803153198e-05, + "loss": 1.0586, + "step": 11345 + }, + { + "epoch": 0.27, + "grad_norm": 2.040261850043331, + "learning_rate": 1.7189297675805088e-05, + "loss": 1.1442, + "step": 11346 + }, + { + "epoch": 0.27, + "grad_norm": 2.0398968844038787, + "learning_rate": 1.718876727821983e-05, + "loss": 1.0773, + "step": 11347 + }, + { + "epoch": 0.27, + "grad_norm": 2.1410750578705744, + "learning_rate": 1.7188236838779297e-05, + "loss": 1.0495, + "step": 11348 + }, + { + "epoch": 0.27, + "grad_norm": 1.9756912137632658, + "learning_rate": 1.718770635748657e-05, + "loss": 1.0653, + "step": 11349 + }, + { + "epoch": 0.27, + "grad_norm": 2.3603002561778235, + "learning_rate": 1.718717583434474e-05, + "loss": 1.1696, + "step": 11350 + }, + { + "epoch": 0.27, + "grad_norm": 1.7707527030602421, + "learning_rate": 1.71866452693569e-05, + "loss": 1.0168, + "step": 11351 + }, + { + "epoch": 0.27, + "grad_norm": 1.9318416124950295, + "learning_rate": 1.718611466252613e-05, + "loss": 0.9675, + "step": 11352 + }, + { + "epoch": 0.27, + "grad_norm": 2.4726447678400323, + "learning_rate": 1.718558401385553e-05, + "loss": 1.0566, + "step": 11353 + }, + { + "epoch": 0.27, + "grad_norm": 2.360315763535839, + "learning_rate": 1.7185053323348187e-05, + "loss": 0.9409, + "step": 11354 + }, + { + "epoch": 0.27, + "grad_norm": 2.5170144289968026, + "learning_rate": 1.7184522591007184e-05, + "loss": 1.0792, + "step": 11355 + }, + { + "epoch": 0.27, + "grad_norm": 1.8841170996335215, + "learning_rate": 1.718399181683562e-05, + "loss": 0.9646, + "step": 11356 + }, + { + "epoch": 0.27, + "grad_norm": 2.246140437470217, + "learning_rate": 1.718346100083658e-05, + "loss": 1.0253, + "step": 11357 + }, + { + "epoch": 0.27, + "grad_norm": 1.141665259885603, + "learning_rate": 1.7182930143013154e-05, + "loss": 0.9842, + "step": 11358 + }, + { + "epoch": 0.27, + "grad_norm": 2.0970933417963367, + "learning_rate": 1.7182399243368436e-05, + "loss": 0.9487, + "step": 11359 + }, + { + "epoch": 0.27, + "grad_norm": 1.9663935600449296, + "learning_rate": 1.7181868301905515e-05, + "loss": 1.0104, + "step": 11360 + }, + { + "epoch": 0.27, + "grad_norm": 1.0781268617797062, + "learning_rate": 1.7181337318627485e-05, + "loss": 0.9105, + "step": 11361 + }, + { + "epoch": 0.27, + "grad_norm": 2.047134329182755, + "learning_rate": 1.7180806293537435e-05, + "loss": 1.0611, + "step": 11362 + }, + { + "epoch": 0.27, + "grad_norm": 1.8857507203594064, + "learning_rate": 1.7180275226638455e-05, + "loss": 1.0256, + "step": 11363 + }, + { + "epoch": 0.27, + "grad_norm": 2.3930317106224464, + "learning_rate": 1.7179744117933642e-05, + "loss": 1.0734, + "step": 11364 + }, + { + "epoch": 0.27, + "grad_norm": 1.0789081382873054, + "learning_rate": 1.7179212967426083e-05, + "loss": 0.9736, + "step": 11365 + }, + { + "epoch": 0.27, + "grad_norm": 2.3944925020973766, + "learning_rate": 1.7178681775118874e-05, + "loss": 1.0424, + "step": 11366 + }, + { + "epoch": 0.27, + "grad_norm": 2.129005880714937, + "learning_rate": 1.717815054101511e-05, + "loss": 1.0865, + "step": 11367 + }, + { + "epoch": 0.27, + "grad_norm": 1.9220918681684516, + "learning_rate": 1.717761926511788e-05, + "loss": 1.0996, + "step": 11368 + }, + { + "epoch": 0.27, + "grad_norm": 2.45025692667867, + "learning_rate": 1.7177087947430275e-05, + "loss": 1.0377, + "step": 11369 + }, + { + "epoch": 0.27, + "grad_norm": 2.5257712717719047, + "learning_rate": 1.717655658795539e-05, + "loss": 1.135, + "step": 11370 + }, + { + "epoch": 0.27, + "grad_norm": 1.1001179129652618, + "learning_rate": 1.717602518669633e-05, + "loss": 1.0154, + "step": 11371 + }, + { + "epoch": 0.27, + "grad_norm": 1.9834675701114701, + "learning_rate": 1.717549374365617e-05, + "loss": 1.0506, + "step": 11372 + }, + { + "epoch": 0.27, + "grad_norm": 1.1527241100326389, + "learning_rate": 1.717496225883802e-05, + "loss": 1.0255, + "step": 11373 + }, + { + "epoch": 0.27, + "grad_norm": 2.266475051975796, + "learning_rate": 1.717443073224496e-05, + "loss": 0.9742, + "step": 11374 + }, + { + "epoch": 0.27, + "grad_norm": 2.6175267654410406, + "learning_rate": 1.7173899163880105e-05, + "loss": 0.9536, + "step": 11375 + }, + { + "epoch": 0.27, + "grad_norm": 1.9737379879180914, + "learning_rate": 1.7173367553746526e-05, + "loss": 1.0684, + "step": 11376 + }, + { + "epoch": 0.27, + "grad_norm": 2.0696942457781127, + "learning_rate": 1.7172835901847338e-05, + "loss": 1.1245, + "step": 11377 + }, + { + "epoch": 0.27, + "grad_norm": 1.0947649440281424, + "learning_rate": 1.7172304208185627e-05, + "loss": 1.0608, + "step": 11378 + }, + { + "epoch": 0.27, + "grad_norm": 1.932497478871857, + "learning_rate": 1.7171772472764488e-05, + "loss": 1.1132, + "step": 11379 + }, + { + "epoch": 0.27, + "grad_norm": 2.1493090698608874, + "learning_rate": 1.717124069558702e-05, + "loss": 1.0877, + "step": 11380 + }, + { + "epoch": 0.27, + "grad_norm": 2.119203775757007, + "learning_rate": 1.717070887665632e-05, + "loss": 1.1095, + "step": 11381 + }, + { + "epoch": 0.27, + "grad_norm": 2.411708910513906, + "learning_rate": 1.717017701597548e-05, + "loss": 1.0321, + "step": 11382 + }, + { + "epoch": 0.27, + "grad_norm": 2.203533930863918, + "learning_rate": 1.7169645113547598e-05, + "loss": 1.0614, + "step": 11383 + }, + { + "epoch": 0.27, + "grad_norm": 2.1819428575646302, + "learning_rate": 1.7169113169375778e-05, + "loss": 1.1243, + "step": 11384 + }, + { + "epoch": 0.27, + "grad_norm": 4.6241057551749645, + "learning_rate": 1.7168581183463106e-05, + "loss": 1.0517, + "step": 11385 + }, + { + "epoch": 0.27, + "grad_norm": 2.2641243260296045, + "learning_rate": 1.7168049155812686e-05, + "loss": 0.9436, + "step": 11386 + }, + { + "epoch": 0.27, + "grad_norm": 1.1452525485345728, + "learning_rate": 1.7167517086427618e-05, + "loss": 0.9896, + "step": 11387 + }, + { + "epoch": 0.27, + "grad_norm": 2.1392805821606986, + "learning_rate": 1.716698497531099e-05, + "loss": 0.9982, + "step": 11388 + }, + { + "epoch": 0.27, + "grad_norm": 1.1349491535831164, + "learning_rate": 1.7166452822465913e-05, + "loss": 0.977, + "step": 11389 + }, + { + "epoch": 0.27, + "grad_norm": 1.9469096716102858, + "learning_rate": 1.7165920627895477e-05, + "loss": 1.1101, + "step": 11390 + }, + { + "epoch": 0.27, + "grad_norm": 2.094490297704888, + "learning_rate": 1.7165388391602783e-05, + "loss": 1.0584, + "step": 11391 + }, + { + "epoch": 0.27, + "grad_norm": 2.250467273211033, + "learning_rate": 1.7164856113590925e-05, + "loss": 1.1464, + "step": 11392 + }, + { + "epoch": 0.27, + "grad_norm": 2.3621111574098435, + "learning_rate": 1.716432379386301e-05, + "loss": 1.0935, + "step": 11393 + }, + { + "epoch": 0.27, + "grad_norm": 1.9263416371671471, + "learning_rate": 1.7163791432422133e-05, + "loss": 1.089, + "step": 11394 + }, + { + "epoch": 0.27, + "grad_norm": 2.2224178027080406, + "learning_rate": 1.7163259029271392e-05, + "loss": 1.0896, + "step": 11395 + }, + { + "epoch": 0.27, + "grad_norm": 2.1652133657049366, + "learning_rate": 1.716272658441389e-05, + "loss": 0.965, + "step": 11396 + }, + { + "epoch": 0.27, + "grad_norm": 2.2486931900628377, + "learning_rate": 1.716219409785273e-05, + "loss": 0.9486, + "step": 11397 + }, + { + "epoch": 0.27, + "grad_norm": 2.243465006222617, + "learning_rate": 1.7161661569591004e-05, + "loss": 1.0062, + "step": 11398 + }, + { + "epoch": 0.27, + "grad_norm": 2.768071066784666, + "learning_rate": 1.716112899963182e-05, + "loss": 1.086, + "step": 11399 + }, + { + "epoch": 0.27, + "grad_norm": 2.59629695685663, + "learning_rate": 1.7160596387978273e-05, + "loss": 1.1186, + "step": 11400 + }, + { + "epoch": 0.27, + "grad_norm": 1.7935946619055545, + "learning_rate": 1.7160063734633468e-05, + "loss": 1.0527, + "step": 11401 + }, + { + "epoch": 0.27, + "grad_norm": 2.046613006372864, + "learning_rate": 1.7159531039600507e-05, + "loss": 1.0703, + "step": 11402 + }, + { + "epoch": 0.27, + "grad_norm": 2.026490775096797, + "learning_rate": 1.715899830288249e-05, + "loss": 1.1059, + "step": 11403 + }, + { + "epoch": 0.27, + "grad_norm": 2.032687070344351, + "learning_rate": 1.7158465524482512e-05, + "loss": 1.1035, + "step": 11404 + }, + { + "epoch": 0.27, + "grad_norm": 2.5649889107914046, + "learning_rate": 1.715793270440369e-05, + "loss": 1.0533, + "step": 11405 + }, + { + "epoch": 0.27, + "grad_norm": 1.8702581561669793, + "learning_rate": 1.715739984264911e-05, + "loss": 1.1036, + "step": 11406 + }, + { + "epoch": 0.27, + "grad_norm": 2.0612689650761338, + "learning_rate": 1.7156866939221888e-05, + "loss": 1.0055, + "step": 11407 + }, + { + "epoch": 0.27, + "grad_norm": 1.7908051064422223, + "learning_rate": 1.715633399412512e-05, + "loss": 1.036, + "step": 11408 + }, + { + "epoch": 0.27, + "grad_norm": 2.2106582844193428, + "learning_rate": 1.7155801007361908e-05, + "loss": 0.9481, + "step": 11409 + }, + { + "epoch": 0.27, + "grad_norm": 2.689157348930156, + "learning_rate": 1.7155267978935355e-05, + "loss": 1.0144, + "step": 11410 + }, + { + "epoch": 0.27, + "grad_norm": 2.0676100682123084, + "learning_rate": 1.7154734908848568e-05, + "loss": 1.0706, + "step": 11411 + }, + { + "epoch": 0.27, + "grad_norm": 2.1553317432333894, + "learning_rate": 1.715420179710465e-05, + "loss": 1.1277, + "step": 11412 + }, + { + "epoch": 0.27, + "grad_norm": 1.2264617408532767, + "learning_rate": 1.7153668643706704e-05, + "loss": 0.9771, + "step": 11413 + }, + { + "epoch": 0.27, + "grad_norm": 2.0579300888291683, + "learning_rate": 1.7153135448657834e-05, + "loss": 1.1748, + "step": 11414 + }, + { + "epoch": 0.27, + "grad_norm": 2.0000714758677565, + "learning_rate": 1.7152602211961146e-05, + "loss": 1.1462, + "step": 11415 + }, + { + "epoch": 0.27, + "grad_norm": 2.0845729856121897, + "learning_rate": 1.7152068933619742e-05, + "loss": 1.0162, + "step": 11416 + }, + { + "epoch": 0.27, + "grad_norm": 1.9350943220355028, + "learning_rate": 1.715153561363673e-05, + "loss": 0.9491, + "step": 11417 + }, + { + "epoch": 0.27, + "grad_norm": 1.162376372335557, + "learning_rate": 1.7151002252015213e-05, + "loss": 1.0321, + "step": 11418 + }, + { + "epoch": 0.27, + "grad_norm": 1.8817326458474353, + "learning_rate": 1.7150468848758296e-05, + "loss": 1.0474, + "step": 11419 + }, + { + "epoch": 0.27, + "grad_norm": 2.3331504579969176, + "learning_rate": 1.7149935403869086e-05, + "loss": 1.0373, + "step": 11420 + }, + { + "epoch": 0.27, + "grad_norm": 1.1758569042763298, + "learning_rate": 1.714940191735069e-05, + "loss": 0.9478, + "step": 11421 + }, + { + "epoch": 0.27, + "grad_norm": 1.909703912009016, + "learning_rate": 1.714886838920621e-05, + "loss": 0.8911, + "step": 11422 + }, + { + "epoch": 0.27, + "grad_norm": 1.981226400425097, + "learning_rate": 1.7148334819438758e-05, + "loss": 1.0817, + "step": 11423 + }, + { + "epoch": 0.27, + "grad_norm": 2.359181374951844, + "learning_rate": 1.7147801208051432e-05, + "loss": 1.0473, + "step": 11424 + }, + { + "epoch": 0.27, + "grad_norm": 2.1699688706343383, + "learning_rate": 1.7147267555047352e-05, + "loss": 1.0991, + "step": 11425 + }, + { + "epoch": 0.27, + "grad_norm": 2.10325651445779, + "learning_rate": 1.7146733860429614e-05, + "loss": 0.9024, + "step": 11426 + }, + { + "epoch": 0.27, + "grad_norm": 2.311890681118661, + "learning_rate": 1.7146200124201327e-05, + "loss": 0.9471, + "step": 11427 + }, + { + "epoch": 0.27, + "grad_norm": 1.1548557042192333, + "learning_rate": 1.71456663463656e-05, + "loss": 0.937, + "step": 11428 + }, + { + "epoch": 0.27, + "grad_norm": 2.5060439303859012, + "learning_rate": 1.7145132526925543e-05, + "loss": 0.9007, + "step": 11429 + }, + { + "epoch": 0.27, + "grad_norm": 2.121925502943398, + "learning_rate": 1.714459866588426e-05, + "loss": 1.0415, + "step": 11430 + }, + { + "epoch": 0.27, + "grad_norm": 1.9185198380739978, + "learning_rate": 1.7144064763244864e-05, + "loss": 0.9954, + "step": 11431 + }, + { + "epoch": 0.27, + "grad_norm": 1.8718268869751171, + "learning_rate": 1.7143530819010463e-05, + "loss": 1.0314, + "step": 11432 + }, + { + "epoch": 0.27, + "grad_norm": 1.918073650699132, + "learning_rate": 1.7142996833184163e-05, + "loss": 1.1007, + "step": 11433 + }, + { + "epoch": 0.27, + "grad_norm": 1.812505165621471, + "learning_rate": 1.7142462805769073e-05, + "loss": 1.0215, + "step": 11434 + }, + { + "epoch": 0.27, + "grad_norm": 2.0253164418962344, + "learning_rate": 1.71419287367683e-05, + "loss": 1.1316, + "step": 11435 + }, + { + "epoch": 0.27, + "grad_norm": 2.161219900322657, + "learning_rate": 1.7141394626184965e-05, + "loss": 0.9621, + "step": 11436 + }, + { + "epoch": 0.27, + "grad_norm": 2.2791141972684814, + "learning_rate": 1.7140860474022163e-05, + "loss": 1.0886, + "step": 11437 + }, + { + "epoch": 0.27, + "grad_norm": 2.196273226954396, + "learning_rate": 1.7140326280283013e-05, + "loss": 1.0535, + "step": 11438 + }, + { + "epoch": 0.27, + "grad_norm": 2.0184471812730678, + "learning_rate": 1.7139792044970622e-05, + "loss": 0.9041, + "step": 11439 + }, + { + "epoch": 0.27, + "grad_norm": 2.8925289571493327, + "learning_rate": 1.71392577680881e-05, + "loss": 1.0308, + "step": 11440 + }, + { + "epoch": 0.27, + "grad_norm": 1.7749339074656303, + "learning_rate": 1.7138723449638562e-05, + "loss": 1.1094, + "step": 11441 + }, + { + "epoch": 0.27, + "grad_norm": 2.0173906902093317, + "learning_rate": 1.7138189089625113e-05, + "loss": 1.0692, + "step": 11442 + }, + { + "epoch": 0.27, + "grad_norm": 2.1811685289329397, + "learning_rate": 1.7137654688050868e-05, + "loss": 1.0984, + "step": 11443 + }, + { + "epoch": 0.27, + "grad_norm": 1.8828042572284256, + "learning_rate": 1.713712024491894e-05, + "loss": 1.1713, + "step": 11444 + }, + { + "epoch": 0.27, + "grad_norm": 1.8056174863334318, + "learning_rate": 1.7136585760232436e-05, + "loss": 0.8708, + "step": 11445 + }, + { + "epoch": 0.27, + "grad_norm": 2.1596995557734333, + "learning_rate": 1.713605123399447e-05, + "loss": 0.9775, + "step": 11446 + }, + { + "epoch": 0.27, + "grad_norm": 2.283086268454677, + "learning_rate": 1.7135516666208156e-05, + "loss": 1.1693, + "step": 11447 + }, + { + "epoch": 0.27, + "grad_norm": 1.967085149288139, + "learning_rate": 1.7134982056876604e-05, + "loss": 1.0313, + "step": 11448 + }, + { + "epoch": 0.27, + "grad_norm": 2.33635289612654, + "learning_rate": 1.713444740600293e-05, + "loss": 1.1464, + "step": 11449 + }, + { + "epoch": 0.27, + "grad_norm": 2.356289795614112, + "learning_rate": 1.7133912713590243e-05, + "loss": 1.1023, + "step": 11450 + }, + { + "epoch": 0.27, + "grad_norm": 1.9061016314253139, + "learning_rate": 1.7133377979641655e-05, + "loss": 0.9555, + "step": 11451 + }, + { + "epoch": 0.27, + "grad_norm": 1.7561416284415308, + "learning_rate": 1.7132843204160284e-05, + "loss": 1.0536, + "step": 11452 + }, + { + "epoch": 0.27, + "grad_norm": 1.9471100311114995, + "learning_rate": 1.713230838714924e-05, + "loss": 1.0492, + "step": 11453 + }, + { + "epoch": 0.27, + "grad_norm": 2.0551452659377065, + "learning_rate": 1.7131773528611643e-05, + "loss": 1.0239, + "step": 11454 + }, + { + "epoch": 0.27, + "grad_norm": 2.0072551165373493, + "learning_rate": 1.71312386285506e-05, + "loss": 1.0106, + "step": 11455 + }, + { + "epoch": 0.27, + "grad_norm": 2.0441077594864163, + "learning_rate": 1.713070368696923e-05, + "loss": 1.2394, + "step": 11456 + }, + { + "epoch": 0.27, + "grad_norm": 2.0548124473668117, + "learning_rate": 1.713016870387064e-05, + "loss": 1.0508, + "step": 11457 + }, + { + "epoch": 0.27, + "grad_norm": 2.115223497537135, + "learning_rate": 1.7129633679257956e-05, + "loss": 1.0284, + "step": 11458 + }, + { + "epoch": 0.27, + "grad_norm": 2.103439028482324, + "learning_rate": 1.7129098613134285e-05, + "loss": 1.0584, + "step": 11459 + }, + { + "epoch": 0.27, + "grad_norm": 1.126590952781773, + "learning_rate": 1.7128563505502744e-05, + "loss": 0.9626, + "step": 11460 + }, + { + "epoch": 0.27, + "grad_norm": 2.1853566272833564, + "learning_rate": 1.712802835636645e-05, + "loss": 1.0371, + "step": 11461 + }, + { + "epoch": 0.27, + "grad_norm": 2.0370281937992414, + "learning_rate": 1.712749316572852e-05, + "loss": 0.9908, + "step": 11462 + }, + { + "epoch": 0.27, + "grad_norm": 2.0435410137078165, + "learning_rate": 1.7126957933592067e-05, + "loss": 0.8656, + "step": 11463 + }, + { + "epoch": 0.27, + "grad_norm": 2.0367574979913883, + "learning_rate": 1.7126422659960208e-05, + "loss": 0.9867, + "step": 11464 + }, + { + "epoch": 0.27, + "grad_norm": 2.0839772919227904, + "learning_rate": 1.712588734483606e-05, + "loss": 1.045, + "step": 11465 + }, + { + "epoch": 0.27, + "grad_norm": 2.140570363053151, + "learning_rate": 1.712535198822274e-05, + "loss": 0.9452, + "step": 11466 + }, + { + "epoch": 0.27, + "grad_norm": 1.9929816084491434, + "learning_rate": 1.7124816590123368e-05, + "loss": 1.0441, + "step": 11467 + }, + { + "epoch": 0.27, + "grad_norm": 2.043268860364763, + "learning_rate": 1.7124281150541054e-05, + "loss": 0.9704, + "step": 11468 + }, + { + "epoch": 0.27, + "grad_norm": 1.1423875562168582, + "learning_rate": 1.712374566947892e-05, + "loss": 0.9905, + "step": 11469 + }, + { + "epoch": 0.27, + "grad_norm": 2.24334338590936, + "learning_rate": 1.7123210146940087e-05, + "loss": 1.0828, + "step": 11470 + }, + { + "epoch": 0.27, + "grad_norm": 2.0751353986730297, + "learning_rate": 1.7122674582927668e-05, + "loss": 1.1021, + "step": 11471 + }, + { + "epoch": 0.27, + "grad_norm": 1.1727268339674741, + "learning_rate": 1.712213897744478e-05, + "loss": 1.0006, + "step": 11472 + }, + { + "epoch": 0.27, + "grad_norm": 2.063611047634633, + "learning_rate": 1.7121603330494547e-05, + "loss": 0.9687, + "step": 11473 + }, + { + "epoch": 0.27, + "grad_norm": 1.967215269073031, + "learning_rate": 1.712106764208008e-05, + "loss": 1.0784, + "step": 11474 + }, + { + "epoch": 0.27, + "grad_norm": 2.091762927235801, + "learning_rate": 1.7120531912204507e-05, + "loss": 0.986, + "step": 11475 + }, + { + "epoch": 0.27, + "grad_norm": 2.1456373343398276, + "learning_rate": 1.711999614087094e-05, + "loss": 1.1351, + "step": 11476 + }, + { + "epoch": 0.27, + "grad_norm": 2.204519690118738, + "learning_rate": 1.7119460328082506e-05, + "loss": 1.071, + "step": 11477 + }, + { + "epoch": 0.27, + "grad_norm": 1.866257465721732, + "learning_rate": 1.7118924473842318e-05, + "loss": 1.0491, + "step": 11478 + }, + { + "epoch": 0.27, + "grad_norm": 2.109555729766268, + "learning_rate": 1.7118388578153498e-05, + "loss": 1.0699, + "step": 11479 + }, + { + "epoch": 0.27, + "grad_norm": 1.8738446793969092, + "learning_rate": 1.7117852641019162e-05, + "loss": 0.9003, + "step": 11480 + }, + { + "epoch": 0.27, + "grad_norm": 2.123562148197397, + "learning_rate": 1.711731666244244e-05, + "loss": 1.0179, + "step": 11481 + }, + { + "epoch": 0.27, + "grad_norm": 2.194791764516916, + "learning_rate": 1.7116780642426444e-05, + "loss": 1.0258, + "step": 11482 + }, + { + "epoch": 0.27, + "grad_norm": 1.688141493234554, + "learning_rate": 1.71162445809743e-05, + "loss": 0.9356, + "step": 11483 + }, + { + "epoch": 0.27, + "grad_norm": 1.9577121984139785, + "learning_rate": 1.7115708478089124e-05, + "loss": 1.0777, + "step": 11484 + }, + { + "epoch": 0.27, + "grad_norm": 2.0777340978221344, + "learning_rate": 1.7115172333774043e-05, + "loss": 1.0488, + "step": 11485 + }, + { + "epoch": 0.27, + "grad_norm": 2.2614748234189177, + "learning_rate": 1.7114636148032176e-05, + "loss": 0.9439, + "step": 11486 + }, + { + "epoch": 0.27, + "grad_norm": 2.2804139679034505, + "learning_rate": 1.7114099920866644e-05, + "loss": 1.0353, + "step": 11487 + }, + { + "epoch": 0.27, + "grad_norm": 2.2478373708093855, + "learning_rate": 1.7113563652280568e-05, + "loss": 0.9839, + "step": 11488 + }, + { + "epoch": 0.27, + "grad_norm": 2.112937442934706, + "learning_rate": 1.7113027342277077e-05, + "loss": 0.996, + "step": 11489 + }, + { + "epoch": 0.27, + "grad_norm": 2.2732697714629797, + "learning_rate": 1.7112490990859286e-05, + "loss": 1.0668, + "step": 11490 + }, + { + "epoch": 0.27, + "grad_norm": 2.13031185014437, + "learning_rate": 1.711195459803032e-05, + "loss": 0.8741, + "step": 11491 + }, + { + "epoch": 0.27, + "grad_norm": 1.9693916085177567, + "learning_rate": 1.71114181637933e-05, + "loss": 1.0859, + "step": 11492 + }, + { + "epoch": 0.27, + "grad_norm": 2.398832760872218, + "learning_rate": 1.711088168815136e-05, + "loss": 1.0773, + "step": 11493 + }, + { + "epoch": 0.27, + "grad_norm": 2.2069833144746505, + "learning_rate": 1.711034517110761e-05, + "loss": 0.9807, + "step": 11494 + }, + { + "epoch": 0.27, + "grad_norm": 2.3858606581954755, + "learning_rate": 1.710980861266518e-05, + "loss": 0.9806, + "step": 11495 + }, + { + "epoch": 0.27, + "grad_norm": 1.8452424443268485, + "learning_rate": 1.7109272012827195e-05, + "loss": 0.9168, + "step": 11496 + }, + { + "epoch": 0.27, + "grad_norm": 2.227420351656535, + "learning_rate": 1.7108735371596776e-05, + "loss": 1.1039, + "step": 11497 + }, + { + "epoch": 0.27, + "grad_norm": 1.8967586241939811, + "learning_rate": 1.7108198688977048e-05, + "loss": 1.0619, + "step": 11498 + }, + { + "epoch": 0.27, + "grad_norm": 2.1630767345232695, + "learning_rate": 1.7107661964971137e-05, + "loss": 1.0468, + "step": 11499 + }, + { + "epoch": 0.27, + "grad_norm": 2.0062196376934023, + "learning_rate": 1.710712519958217e-05, + "loss": 1.0721, + "step": 11500 + }, + { + "epoch": 0.27, + "grad_norm": 1.940764239838969, + "learning_rate": 1.710658839281327e-05, + "loss": 1.0013, + "step": 11501 + }, + { + "epoch": 0.27, + "grad_norm": 1.9961897207357242, + "learning_rate": 1.710605154466756e-05, + "loss": 1.207, + "step": 11502 + }, + { + "epoch": 0.27, + "grad_norm": 2.023193057863368, + "learning_rate": 1.7105514655148173e-05, + "loss": 1.0147, + "step": 11503 + }, + { + "epoch": 0.27, + "grad_norm": 2.084778316504559, + "learning_rate": 1.710497772425823e-05, + "loss": 0.9993, + "step": 11504 + }, + { + "epoch": 0.27, + "grad_norm": 1.8401767912448457, + "learning_rate": 1.710444075200085e-05, + "loss": 1.1026, + "step": 11505 + }, + { + "epoch": 0.27, + "grad_norm": 2.7052301224290436, + "learning_rate": 1.710390373837917e-05, + "loss": 0.958, + "step": 11506 + }, + { + "epoch": 0.27, + "grad_norm": 1.1789569106238826, + "learning_rate": 1.7103366683396316e-05, + "loss": 0.9536, + "step": 11507 + }, + { + "epoch": 0.27, + "grad_norm": 1.8582628841395892, + "learning_rate": 1.7102829587055413e-05, + "loss": 1.0085, + "step": 11508 + }, + { + "epoch": 0.27, + "grad_norm": 1.1408930454978305, + "learning_rate": 1.710229244935958e-05, + "loss": 0.9573, + "step": 11509 + }, + { + "epoch": 0.27, + "grad_norm": 1.7039812184881247, + "learning_rate": 1.7101755270311957e-05, + "loss": 1.0645, + "step": 11510 + }, + { + "epoch": 0.27, + "grad_norm": 1.9342631269930035, + "learning_rate": 1.7101218049915666e-05, + "loss": 1.0236, + "step": 11511 + }, + { + "epoch": 0.27, + "grad_norm": 2.159476282594298, + "learning_rate": 1.7100680788173836e-05, + "loss": 1.0226, + "step": 11512 + }, + { + "epoch": 0.27, + "grad_norm": 2.2866486047364205, + "learning_rate": 1.7100143485089596e-05, + "loss": 0.9308, + "step": 11513 + }, + { + "epoch": 0.27, + "grad_norm": 1.8697800759485985, + "learning_rate": 1.7099606140666068e-05, + "loss": 1.035, + "step": 11514 + }, + { + "epoch": 0.27, + "grad_norm": 2.360847935667638, + "learning_rate": 1.709906875490639e-05, + "loss": 1.0998, + "step": 11515 + }, + { + "epoch": 0.27, + "grad_norm": 1.8836683295930396, + "learning_rate": 1.7098531327813684e-05, + "loss": 1.0054, + "step": 11516 + }, + { + "epoch": 0.27, + "grad_norm": 1.8563975835220132, + "learning_rate": 1.7097993859391082e-05, + "loss": 1.0222, + "step": 11517 + }, + { + "epoch": 0.27, + "grad_norm": 2.2816708623369597, + "learning_rate": 1.709745634964171e-05, + "loss": 0.9635, + "step": 11518 + }, + { + "epoch": 0.27, + "grad_norm": 2.079155283592682, + "learning_rate": 1.7096918798568705e-05, + "loss": 0.9015, + "step": 11519 + }, + { + "epoch": 0.27, + "grad_norm": 1.9734009099691994, + "learning_rate": 1.7096381206175188e-05, + "loss": 0.9104, + "step": 11520 + }, + { + "epoch": 0.27, + "grad_norm": 2.173566690804838, + "learning_rate": 1.7095843572464294e-05, + "loss": 1.1656, + "step": 11521 + }, + { + "epoch": 0.27, + "grad_norm": 2.905097644760719, + "learning_rate": 1.7095305897439152e-05, + "loss": 1.0136, + "step": 11522 + }, + { + "epoch": 0.27, + "grad_norm": 2.257876514687244, + "learning_rate": 1.7094768181102895e-05, + "loss": 1.0135, + "step": 11523 + }, + { + "epoch": 0.27, + "grad_norm": 1.9367632928971807, + "learning_rate": 1.709423042345865e-05, + "loss": 1.1441, + "step": 11524 + }, + { + "epoch": 0.27, + "grad_norm": 2.0956984739438624, + "learning_rate": 1.7093692624509545e-05, + "loss": 0.9215, + "step": 11525 + }, + { + "epoch": 0.27, + "grad_norm": 1.930390542403674, + "learning_rate": 1.709315478425872e-05, + "loss": 1.0213, + "step": 11526 + }, + { + "epoch": 0.27, + "grad_norm": 2.0782850995274416, + "learning_rate": 1.7092616902709304e-05, + "loss": 0.9267, + "step": 11527 + }, + { + "epoch": 0.27, + "grad_norm": 1.951023519783222, + "learning_rate": 1.7092078979864425e-05, + "loss": 1.0704, + "step": 11528 + }, + { + "epoch": 0.27, + "grad_norm": 2.2911829363949434, + "learning_rate": 1.7091541015727216e-05, + "loss": 0.9287, + "step": 11529 + }, + { + "epoch": 0.27, + "grad_norm": 2.0048803480656225, + "learning_rate": 1.709100301030081e-05, + "loss": 1.0575, + "step": 11530 + }, + { + "epoch": 0.27, + "grad_norm": 2.0749720144745494, + "learning_rate": 1.709046496358834e-05, + "loss": 1.1788, + "step": 11531 + }, + { + "epoch": 0.27, + "grad_norm": 1.129746026036533, + "learning_rate": 1.7089926875592938e-05, + "loss": 0.9746, + "step": 11532 + }, + { + "epoch": 0.27, + "grad_norm": 2.081727581665055, + "learning_rate": 1.7089388746317738e-05, + "loss": 1.0629, + "step": 11533 + }, + { + "epoch": 0.27, + "grad_norm": 1.0972306482332768, + "learning_rate": 1.708885057576587e-05, + "loss": 1.0231, + "step": 11534 + }, + { + "epoch": 0.27, + "grad_norm": 2.0293714661744766, + "learning_rate": 1.7088312363940472e-05, + "loss": 1.01, + "step": 11535 + }, + { + "epoch": 0.27, + "grad_norm": 1.970020791132176, + "learning_rate": 1.7087774110844674e-05, + "loss": 1.1164, + "step": 11536 + }, + { + "epoch": 0.27, + "grad_norm": 2.175254139516192, + "learning_rate": 1.7087235816481614e-05, + "loss": 1.0613, + "step": 11537 + }, + { + "epoch": 0.27, + "grad_norm": 2.201564187851449, + "learning_rate": 1.7086697480854423e-05, + "loss": 1.0877, + "step": 11538 + }, + { + "epoch": 0.27, + "grad_norm": 2.2512339999181967, + "learning_rate": 1.708615910396623e-05, + "loss": 0.8623, + "step": 11539 + }, + { + "epoch": 0.27, + "grad_norm": 2.207245441609531, + "learning_rate": 1.708562068582018e-05, + "loss": 1.0387, + "step": 11540 + }, + { + "epoch": 0.27, + "grad_norm": 1.882709105052183, + "learning_rate": 1.7085082226419402e-05, + "loss": 1.0934, + "step": 11541 + }, + { + "epoch": 0.27, + "grad_norm": 1.217367757021441, + "learning_rate": 1.7084543725767035e-05, + "loss": 1.0584, + "step": 11542 + }, + { + "epoch": 0.27, + "grad_norm": 3.0254868726953936, + "learning_rate": 1.7084005183866206e-05, + "loss": 1.1124, + "step": 11543 + }, + { + "epoch": 0.27, + "grad_norm": 1.9982049209233776, + "learning_rate": 1.708346660072006e-05, + "loss": 0.9235, + "step": 11544 + }, + { + "epoch": 0.27, + "grad_norm": 1.8303001995774058, + "learning_rate": 1.7082927976331726e-05, + "loss": 1.0867, + "step": 11545 + }, + { + "epoch": 0.27, + "grad_norm": 1.9197043736989217, + "learning_rate": 1.7082389310704347e-05, + "loss": 0.9512, + "step": 11546 + }, + { + "epoch": 0.27, + "grad_norm": 1.9607281414604814, + "learning_rate": 1.7081850603841053e-05, + "loss": 1.0333, + "step": 11547 + }, + { + "epoch": 0.27, + "grad_norm": 2.0438673210255183, + "learning_rate": 1.708131185574498e-05, + "loss": 1.0636, + "step": 11548 + }, + { + "epoch": 0.27, + "grad_norm": 2.3518452671371333, + "learning_rate": 1.7080773066419272e-05, + "loss": 1.0229, + "step": 11549 + }, + { + "epoch": 0.27, + "grad_norm": 1.9132579254675341, + "learning_rate": 1.708023423586706e-05, + "loss": 1.0278, + "step": 11550 + }, + { + "epoch": 0.27, + "grad_norm": 2.075290724075616, + "learning_rate": 1.7079695364091483e-05, + "loss": 1.0749, + "step": 11551 + }, + { + "epoch": 0.27, + "grad_norm": 1.9856750406061932, + "learning_rate": 1.7079156451095677e-05, + "loss": 1.0696, + "step": 11552 + }, + { + "epoch": 0.27, + "grad_norm": 2.0631738644705138, + "learning_rate": 1.707861749688278e-05, + "loss": 1.0382, + "step": 11553 + }, + { + "epoch": 0.27, + "grad_norm": 2.3051937748385103, + "learning_rate": 1.707807850145593e-05, + "loss": 1.0431, + "step": 11554 + }, + { + "epoch": 0.27, + "grad_norm": 2.1788385271375104, + "learning_rate": 1.707753946481827e-05, + "loss": 1.0081, + "step": 11555 + }, + { + "epoch": 0.27, + "grad_norm": 2.312284400553544, + "learning_rate": 1.7077000386972934e-05, + "loss": 1.1253, + "step": 11556 + }, + { + "epoch": 0.27, + "grad_norm": 1.996294597750179, + "learning_rate": 1.707646126792306e-05, + "loss": 1.0136, + "step": 11557 + }, + { + "epoch": 0.27, + "grad_norm": 2.1555709075077116, + "learning_rate": 1.7075922107671786e-05, + "loss": 1.1592, + "step": 11558 + }, + { + "epoch": 0.27, + "grad_norm": 2.121929469922821, + "learning_rate": 1.7075382906222254e-05, + "loss": 1.0925, + "step": 11559 + }, + { + "epoch": 0.27, + "grad_norm": 1.0988773506128235, + "learning_rate": 1.7074843663577607e-05, + "loss": 0.9941, + "step": 11560 + }, + { + "epoch": 0.27, + "grad_norm": 2.2133801580392296, + "learning_rate": 1.7074304379740975e-05, + "loss": 1.0502, + "step": 11561 + }, + { + "epoch": 0.27, + "grad_norm": 2.2343189114294866, + "learning_rate": 1.7073765054715506e-05, + "loss": 0.9233, + "step": 11562 + }, + { + "epoch": 0.27, + "grad_norm": 2.0929117826784074, + "learning_rate": 1.7073225688504337e-05, + "loss": 1.0085, + "step": 11563 + }, + { + "epoch": 0.27, + "grad_norm": 2.1653697531728837, + "learning_rate": 1.707268628111061e-05, + "loss": 1.0951, + "step": 11564 + }, + { + "epoch": 0.27, + "grad_norm": 2.4921631218613283, + "learning_rate": 1.7072146832537464e-05, + "loss": 1.0429, + "step": 11565 + }, + { + "epoch": 0.27, + "grad_norm": 2.6146634916050404, + "learning_rate": 1.707160734278804e-05, + "loss": 1.1442, + "step": 11566 + }, + { + "epoch": 0.27, + "grad_norm": 2.6470580560467583, + "learning_rate": 1.7071067811865477e-05, + "loss": 1.2077, + "step": 11567 + }, + { + "epoch": 0.27, + "grad_norm": 2.380854738252864, + "learning_rate": 1.707052823977292e-05, + "loss": 0.9649, + "step": 11568 + }, + { + "epoch": 0.27, + "grad_norm": 2.9172000201155313, + "learning_rate": 1.7069988626513508e-05, + "loss": 1.1247, + "step": 11569 + }, + { + "epoch": 0.27, + "grad_norm": 2.5875767734785704, + "learning_rate": 1.7069448972090387e-05, + "loss": 0.9927, + "step": 11570 + }, + { + "epoch": 0.27, + "grad_norm": 1.764628179193226, + "learning_rate": 1.7068909276506692e-05, + "loss": 1.0252, + "step": 11571 + }, + { + "epoch": 0.27, + "grad_norm": 2.9658958136865348, + "learning_rate": 1.7068369539765572e-05, + "loss": 1.0166, + "step": 11572 + }, + { + "epoch": 0.27, + "grad_norm": 1.9773431760997384, + "learning_rate": 1.7067829761870166e-05, + "loss": 1.1126, + "step": 11573 + }, + { + "epoch": 0.27, + "grad_norm": 2.8369732878472758, + "learning_rate": 1.7067289942823617e-05, + "loss": 1.085, + "step": 11574 + }, + { + "epoch": 0.27, + "grad_norm": 1.1439126298039948, + "learning_rate": 1.7066750082629067e-05, + "loss": 0.9912, + "step": 11575 + }, + { + "epoch": 0.27, + "grad_norm": 2.177748424544581, + "learning_rate": 1.7066210181289666e-05, + "loss": 1.0811, + "step": 11576 + }, + { + "epoch": 0.27, + "grad_norm": 2.195052368327289, + "learning_rate": 1.7065670238808545e-05, + "loss": 1.1935, + "step": 11577 + }, + { + "epoch": 0.27, + "grad_norm": 2.317389839156084, + "learning_rate": 1.706513025518886e-05, + "loss": 0.9839, + "step": 11578 + }, + { + "epoch": 0.27, + "grad_norm": 2.0551088261726513, + "learning_rate": 1.706459023043375e-05, + "loss": 1.0406, + "step": 11579 + }, + { + "epoch": 0.27, + "grad_norm": 1.9039804604111965, + "learning_rate": 1.7064050164546356e-05, + "loss": 0.9602, + "step": 11580 + }, + { + "epoch": 0.27, + "grad_norm": 2.242224215433691, + "learning_rate": 1.7063510057529825e-05, + "loss": 1.0656, + "step": 11581 + }, + { + "epoch": 0.27, + "grad_norm": 1.8672544277089331, + "learning_rate": 1.706296990938731e-05, + "loss": 1.0803, + "step": 11582 + }, + { + "epoch": 0.27, + "grad_norm": 2.7120820368377636, + "learning_rate": 1.7062429720121937e-05, + "loss": 1.0414, + "step": 11583 + }, + { + "epoch": 0.27, + "grad_norm": 2.0077277988535585, + "learning_rate": 1.706188948973687e-05, + "loss": 1.1251, + "step": 11584 + }, + { + "epoch": 0.27, + "grad_norm": 1.0829541228094224, + "learning_rate": 1.7061349218235243e-05, + "loss": 1.0379, + "step": 11585 + }, + { + "epoch": 0.27, + "grad_norm": 4.1199720238558335, + "learning_rate": 1.7060808905620207e-05, + "loss": 1.0795, + "step": 11586 + }, + { + "epoch": 0.27, + "grad_norm": 2.3120618791587035, + "learning_rate": 1.7060268551894904e-05, + "loss": 1.0869, + "step": 11587 + }, + { + "epoch": 0.27, + "grad_norm": 2.20497625973465, + "learning_rate": 1.7059728157062485e-05, + "loss": 1.0243, + "step": 11588 + }, + { + "epoch": 0.27, + "grad_norm": 2.0374591710857435, + "learning_rate": 1.7059187721126093e-05, + "loss": 1.1115, + "step": 11589 + }, + { + "epoch": 0.27, + "grad_norm": 1.9813287933621562, + "learning_rate": 1.7058647244088873e-05, + "loss": 0.9669, + "step": 11590 + }, + { + "epoch": 0.27, + "grad_norm": 2.007060215940709, + "learning_rate": 1.7058106725953977e-05, + "loss": 1.1454, + "step": 11591 + }, + { + "epoch": 0.27, + "grad_norm": 2.031054473599616, + "learning_rate": 1.7057566166724545e-05, + "loss": 1.0353, + "step": 11592 + }, + { + "epoch": 0.27, + "grad_norm": 1.869919009577036, + "learning_rate": 1.7057025566403733e-05, + "loss": 1.0089, + "step": 11593 + }, + { + "epoch": 0.27, + "grad_norm": 1.1626931913857295, + "learning_rate": 1.705648492499468e-05, + "loss": 0.9899, + "step": 11594 + }, + { + "epoch": 0.27, + "grad_norm": 1.9820359842195383, + "learning_rate": 1.7055944242500543e-05, + "loss": 1.0457, + "step": 11595 + }, + { + "epoch": 0.27, + "grad_norm": 1.9900825230110284, + "learning_rate": 1.705540351892446e-05, + "loss": 1.1545, + "step": 11596 + }, + { + "epoch": 0.27, + "grad_norm": 2.2252663107293214, + "learning_rate": 1.7054862754269586e-05, + "loss": 1.0087, + "step": 11597 + }, + { + "epoch": 0.27, + "grad_norm": 2.249361434364591, + "learning_rate": 1.7054321948539066e-05, + "loss": 1.193, + "step": 11598 + }, + { + "epoch": 0.27, + "grad_norm": 2.177234083708451, + "learning_rate": 1.705378110173605e-05, + "loss": 0.9938, + "step": 11599 + }, + { + "epoch": 0.27, + "grad_norm": 1.9680805088577689, + "learning_rate": 1.705324021386369e-05, + "loss": 0.9066, + "step": 11600 + }, + { + "epoch": 0.27, + "grad_norm": 2.0319207976297626, + "learning_rate": 1.705269928492513e-05, + "loss": 1.2062, + "step": 11601 + }, + { + "epoch": 0.27, + "grad_norm": 1.8105679737639442, + "learning_rate": 1.7052158314923524e-05, + "loss": 0.9974, + "step": 11602 + }, + { + "epoch": 0.27, + "grad_norm": 2.2983090714399816, + "learning_rate": 1.705161730386202e-05, + "loss": 1.0008, + "step": 11603 + }, + { + "epoch": 0.27, + "grad_norm": 1.1602840937150236, + "learning_rate": 1.7051076251743763e-05, + "loss": 1.0293, + "step": 11604 + }, + { + "epoch": 0.27, + "grad_norm": 5.54131369579852, + "learning_rate": 1.705053515857191e-05, + "loss": 1.134, + "step": 11605 + }, + { + "epoch": 0.27, + "grad_norm": 2.078161544160297, + "learning_rate": 1.7049994024349613e-05, + "loss": 1.1904, + "step": 11606 + }, + { + "epoch": 0.27, + "grad_norm": 2.073143016756322, + "learning_rate": 1.7049452849080014e-05, + "loss": 1.1134, + "step": 11607 + }, + { + "epoch": 0.27, + "grad_norm": 2.041333856083233, + "learning_rate": 1.704891163276627e-05, + "loss": 1.0118, + "step": 11608 + }, + { + "epoch": 0.27, + "grad_norm": 2.715891630696486, + "learning_rate": 1.7048370375411533e-05, + "loss": 1.1237, + "step": 11609 + }, + { + "epoch": 0.27, + "grad_norm": 2.9747364913407353, + "learning_rate": 1.704782907701895e-05, + "loss": 0.9637, + "step": 11610 + }, + { + "epoch": 0.27, + "grad_norm": 2.077232915095736, + "learning_rate": 1.7047287737591672e-05, + "loss": 0.9551, + "step": 11611 + }, + { + "epoch": 0.27, + "grad_norm": 2.0271795770014136, + "learning_rate": 1.7046746357132858e-05, + "loss": 1.0582, + "step": 11612 + }, + { + "epoch": 0.27, + "grad_norm": 2.083695094419127, + "learning_rate": 1.7046204935645655e-05, + "loss": 1.1007, + "step": 11613 + }, + { + "epoch": 0.27, + "grad_norm": 1.1337705556458135, + "learning_rate": 1.7045663473133215e-05, + "loss": 1.0179, + "step": 11614 + }, + { + "epoch": 0.27, + "grad_norm": 1.887429792783678, + "learning_rate": 1.7045121969598694e-05, + "loss": 0.9686, + "step": 11615 + }, + { + "epoch": 0.27, + "grad_norm": 1.9392305739381326, + "learning_rate": 1.7044580425045244e-05, + "loss": 1.0965, + "step": 11616 + }, + { + "epoch": 0.27, + "grad_norm": 1.9715394182270447, + "learning_rate": 1.7044038839476013e-05, + "loss": 0.9557, + "step": 11617 + }, + { + "epoch": 0.27, + "grad_norm": 1.906374016009853, + "learning_rate": 1.7043497212894154e-05, + "loss": 1.0076, + "step": 11618 + }, + { + "epoch": 0.27, + "grad_norm": 1.9561438787735275, + "learning_rate": 1.704295554530283e-05, + "loss": 1.0932, + "step": 11619 + }, + { + "epoch": 0.27, + "grad_norm": 1.9901089435720523, + "learning_rate": 1.7042413836705187e-05, + "loss": 1.1028, + "step": 11620 + }, + { + "epoch": 0.27, + "grad_norm": 2.0988813461567455, + "learning_rate": 1.7041872087104383e-05, + "loss": 1.1287, + "step": 11621 + }, + { + "epoch": 0.27, + "grad_norm": 2.120644734274048, + "learning_rate": 1.7041330296503566e-05, + "loss": 0.9653, + "step": 11622 + }, + { + "epoch": 0.27, + "grad_norm": 1.9003295169436403, + "learning_rate": 1.70407884649059e-05, + "loss": 0.9474, + "step": 11623 + }, + { + "epoch": 0.27, + "grad_norm": 1.9577550609583354, + "learning_rate": 1.704024659231453e-05, + "loss": 1.0202, + "step": 11624 + }, + { + "epoch": 0.27, + "grad_norm": 2.21770288138067, + "learning_rate": 1.703970467873262e-05, + "loss": 1.0453, + "step": 11625 + }, + { + "epoch": 0.27, + "grad_norm": 1.1496845897693202, + "learning_rate": 1.7039162724163317e-05, + "loss": 1.027, + "step": 11626 + }, + { + "epoch": 0.27, + "grad_norm": 2.0950450905126408, + "learning_rate": 1.703862072860978e-05, + "loss": 1.0916, + "step": 11627 + }, + { + "epoch": 0.27, + "grad_norm": 1.9585979813688665, + "learning_rate": 1.7038078692075166e-05, + "loss": 0.9502, + "step": 11628 + }, + { + "epoch": 0.27, + "grad_norm": 2.076202566174795, + "learning_rate": 1.703753661456263e-05, + "loss": 1.1422, + "step": 11629 + }, + { + "epoch": 0.27, + "grad_norm": 2.3918543131911734, + "learning_rate": 1.703699449607533e-05, + "loss": 1.0067, + "step": 11630 + }, + { + "epoch": 0.27, + "grad_norm": 3.650631529776571, + "learning_rate": 1.7036452336616413e-05, + "loss": 1.0554, + "step": 11631 + }, + { + "epoch": 0.27, + "grad_norm": 1.1716606472590285, + "learning_rate": 1.703591013618905e-05, + "loss": 1.0205, + "step": 11632 + }, + { + "epoch": 0.27, + "grad_norm": 2.764553785219716, + "learning_rate": 1.703536789479639e-05, + "loss": 1.1344, + "step": 11633 + }, + { + "epoch": 0.27, + "grad_norm": 1.8624555369019629, + "learning_rate": 1.7034825612441585e-05, + "loss": 1.0995, + "step": 11634 + }, + { + "epoch": 0.27, + "grad_norm": 2.439552414964068, + "learning_rate": 1.7034283289127806e-05, + "loss": 1.0937, + "step": 11635 + }, + { + "epoch": 0.27, + "grad_norm": 2.0592113698041525, + "learning_rate": 1.7033740924858194e-05, + "loss": 1.0834, + "step": 11636 + }, + { + "epoch": 0.27, + "grad_norm": 2.3643673515805084, + "learning_rate": 1.7033198519635922e-05, + "loss": 0.996, + "step": 11637 + }, + { + "epoch": 0.27, + "grad_norm": 2.0143626334403746, + "learning_rate": 1.7032656073464137e-05, + "loss": 1.0309, + "step": 11638 + }, + { + "epoch": 0.27, + "grad_norm": 1.8789317333531506, + "learning_rate": 1.7032113586346003e-05, + "loss": 0.9772, + "step": 11639 + }, + { + "epoch": 0.27, + "grad_norm": 1.0610727055791773, + "learning_rate": 1.7031571058284678e-05, + "loss": 0.8539, + "step": 11640 + }, + { + "epoch": 0.27, + "grad_norm": 2.1915942459291977, + "learning_rate": 1.7031028489283324e-05, + "loss": 0.8697, + "step": 11641 + }, + { + "epoch": 0.27, + "grad_norm": 2.5526215052040953, + "learning_rate": 1.703048587934509e-05, + "loss": 0.9891, + "step": 11642 + }, + { + "epoch": 0.27, + "grad_norm": 2.1925393557892816, + "learning_rate": 1.7029943228473147e-05, + "loss": 1.1037, + "step": 11643 + }, + { + "epoch": 0.27, + "grad_norm": 2.032242800905719, + "learning_rate": 1.7029400536670646e-05, + "loss": 1.0169, + "step": 11644 + }, + { + "epoch": 0.27, + "grad_norm": 2.1929898497431424, + "learning_rate": 1.702885780394075e-05, + "loss": 1.0597, + "step": 11645 + }, + { + "epoch": 0.27, + "grad_norm": 2.0907342179010038, + "learning_rate": 1.7028315030286617e-05, + "loss": 1.0309, + "step": 11646 + }, + { + "epoch": 0.27, + "grad_norm": 2.1840138089973995, + "learning_rate": 1.7027772215711414e-05, + "loss": 0.8733, + "step": 11647 + }, + { + "epoch": 0.27, + "grad_norm": 1.9010286500019202, + "learning_rate": 1.702722936021829e-05, + "loss": 0.9571, + "step": 11648 + }, + { + "epoch": 0.27, + "grad_norm": 2.158474050644101, + "learning_rate": 1.7026686463810413e-05, + "loss": 1.0908, + "step": 11649 + }, + { + "epoch": 0.27, + "grad_norm": 2.1958203799437896, + "learning_rate": 1.7026143526490947e-05, + "loss": 0.9833, + "step": 11650 + }, + { + "epoch": 0.27, + "grad_norm": 1.1525099046581122, + "learning_rate": 1.7025600548263043e-05, + "loss": 0.9189, + "step": 11651 + }, + { + "epoch": 0.27, + "grad_norm": 1.8753802159243391, + "learning_rate": 1.702505752912987e-05, + "loss": 1.0082, + "step": 11652 + }, + { + "epoch": 0.27, + "grad_norm": 3.2190799076775516, + "learning_rate": 1.7024514469094592e-05, + "loss": 0.8877, + "step": 11653 + }, + { + "epoch": 0.27, + "grad_norm": 2.4631667558073134, + "learning_rate": 1.7023971368160362e-05, + "loss": 0.9976, + "step": 11654 + }, + { + "epoch": 0.27, + "grad_norm": 2.1264195373867936, + "learning_rate": 1.7023428226330348e-05, + "loss": 1.0901, + "step": 11655 + }, + { + "epoch": 0.27, + "grad_norm": 2.246745551892031, + "learning_rate": 1.7022885043607714e-05, + "loss": 1.0248, + "step": 11656 + }, + { + "epoch": 0.27, + "grad_norm": 1.8931121686580157, + "learning_rate": 1.7022341819995622e-05, + "loss": 0.9853, + "step": 11657 + }, + { + "epoch": 0.27, + "grad_norm": 2.2990506832369197, + "learning_rate": 1.7021798555497225e-05, + "loss": 0.9679, + "step": 11658 + }, + { + "epoch": 0.27, + "grad_norm": 2.4167666879134053, + "learning_rate": 1.70212552501157e-05, + "loss": 1.0612, + "step": 11659 + }, + { + "epoch": 0.27, + "grad_norm": 2.284400305872004, + "learning_rate": 1.7020711903854202e-05, + "loss": 1.1078, + "step": 11660 + }, + { + "epoch": 0.27, + "grad_norm": 1.0776751320841103, + "learning_rate": 1.7020168516715894e-05, + "loss": 1.0436, + "step": 11661 + }, + { + "epoch": 0.27, + "grad_norm": 1.261888354043564, + "learning_rate": 1.7019625088703944e-05, + "loss": 1.0178, + "step": 11662 + }, + { + "epoch": 0.27, + "grad_norm": 1.2363358293236022, + "learning_rate": 1.701908161982152e-05, + "loss": 1.0165, + "step": 11663 + }, + { + "epoch": 0.27, + "grad_norm": 2.126885978130996, + "learning_rate": 1.7018538110071773e-05, + "loss": 1.0611, + "step": 11664 + }, + { + "epoch": 0.27, + "grad_norm": 2.0659471999795973, + "learning_rate": 1.7017994559457878e-05, + "loss": 1.0232, + "step": 11665 + }, + { + "epoch": 0.27, + "grad_norm": 1.1140596786569632, + "learning_rate": 1.7017450967982995e-05, + "loss": 0.9791, + "step": 11666 + }, + { + "epoch": 0.27, + "grad_norm": 2.5569299655453377, + "learning_rate": 1.7016907335650292e-05, + "loss": 0.9556, + "step": 11667 + }, + { + "epoch": 0.27, + "grad_norm": 1.0541979834782127, + "learning_rate": 1.701636366246293e-05, + "loss": 1.0207, + "step": 11668 + }, + { + "epoch": 0.27, + "grad_norm": 1.0556958054384988, + "learning_rate": 1.7015819948424085e-05, + "loss": 1.0237, + "step": 11669 + }, + { + "epoch": 0.27, + "grad_norm": 1.9195690137583925, + "learning_rate": 1.7015276193536912e-05, + "loss": 0.9928, + "step": 11670 + }, + { + "epoch": 0.27, + "grad_norm": 2.1968042147214595, + "learning_rate": 1.7014732397804578e-05, + "loss": 1.0845, + "step": 11671 + }, + { + "epoch": 0.27, + "grad_norm": 2.1605329376971927, + "learning_rate": 1.7014188561230252e-05, + "loss": 0.9628, + "step": 11672 + }, + { + "epoch": 0.28, + "grad_norm": 2.092285678600838, + "learning_rate": 1.7013644683817098e-05, + "loss": 0.9854, + "step": 11673 + }, + { + "epoch": 0.28, + "grad_norm": 2.0673746357514178, + "learning_rate": 1.7013100765568284e-05, + "loss": 1.0102, + "step": 11674 + }, + { + "epoch": 0.28, + "grad_norm": 2.150604279876166, + "learning_rate": 1.701255680648698e-05, + "loss": 1.0665, + "step": 11675 + }, + { + "epoch": 0.28, + "grad_norm": 2.0084550618817554, + "learning_rate": 1.701201280657635e-05, + "loss": 0.9535, + "step": 11676 + }, + { + "epoch": 0.28, + "grad_norm": 1.9098090085227102, + "learning_rate": 1.7011468765839558e-05, + "loss": 0.9585, + "step": 11677 + }, + { + "epoch": 0.28, + "grad_norm": 2.0965729948032363, + "learning_rate": 1.701092468427978e-05, + "loss": 0.999, + "step": 11678 + }, + { + "epoch": 0.28, + "grad_norm": 2.2157062697242833, + "learning_rate": 1.7010380561900174e-05, + "loss": 1.0126, + "step": 11679 + }, + { + "epoch": 0.28, + "grad_norm": 2.3323693256035836, + "learning_rate": 1.7009836398703913e-05, + "loss": 1.0979, + "step": 11680 + }, + { + "epoch": 0.28, + "grad_norm": 3.1028154862846744, + "learning_rate": 1.700929219469417e-05, + "loss": 1.0835, + "step": 11681 + }, + { + "epoch": 0.28, + "grad_norm": 2.0046205760235223, + "learning_rate": 1.7008747949874105e-05, + "loss": 0.9548, + "step": 11682 + }, + { + "epoch": 0.28, + "grad_norm": 2.1348183437575545, + "learning_rate": 1.7008203664246892e-05, + "loss": 1.0152, + "step": 11683 + }, + { + "epoch": 0.28, + "grad_norm": 2.1315269069299796, + "learning_rate": 1.70076593378157e-05, + "loss": 1.0818, + "step": 11684 + }, + { + "epoch": 0.28, + "grad_norm": 2.176966345498946, + "learning_rate": 1.7007114970583694e-05, + "loss": 1.0535, + "step": 11685 + }, + { + "epoch": 0.28, + "grad_norm": 1.1076494585211893, + "learning_rate": 1.7006570562554047e-05, + "loss": 0.9558, + "step": 11686 + }, + { + "epoch": 0.28, + "grad_norm": 1.7417383911437518, + "learning_rate": 1.700602611372993e-05, + "loss": 0.9019, + "step": 11687 + }, + { + "epoch": 0.28, + "grad_norm": 2.0835565344136953, + "learning_rate": 1.7005481624114506e-05, + "loss": 1.0893, + "step": 11688 + }, + { + "epoch": 0.28, + "grad_norm": 2.2711357736763413, + "learning_rate": 1.7004937093710953e-05, + "loss": 1.04, + "step": 11689 + }, + { + "epoch": 0.28, + "grad_norm": 1.922230710740813, + "learning_rate": 1.700439252252244e-05, + "loss": 1.1726, + "step": 11690 + }, + { + "epoch": 0.28, + "grad_norm": 1.9244106321561727, + "learning_rate": 1.7003847910552134e-05, + "loss": 1.1584, + "step": 11691 + }, + { + "epoch": 0.28, + "grad_norm": 2.136611380376949, + "learning_rate": 1.7003303257803207e-05, + "loss": 1.1484, + "step": 11692 + }, + { + "epoch": 0.28, + "grad_norm": 2.878272228013401, + "learning_rate": 1.7002758564278832e-05, + "loss": 1.065, + "step": 11693 + }, + { + "epoch": 0.28, + "grad_norm": 2.0545392080944054, + "learning_rate": 1.700221382998218e-05, + "loss": 1.1314, + "step": 11694 + }, + { + "epoch": 0.28, + "grad_norm": 1.0600208701200005, + "learning_rate": 1.7001669054916424e-05, + "loss": 0.9517, + "step": 11695 + }, + { + "epoch": 0.28, + "grad_norm": 1.1391123820104634, + "learning_rate": 1.700112423908473e-05, + "loss": 0.9698, + "step": 11696 + }, + { + "epoch": 0.28, + "grad_norm": 1.1517431785358279, + "learning_rate": 1.700057938249028e-05, + "loss": 0.9478, + "step": 11697 + }, + { + "epoch": 0.28, + "grad_norm": 2.129081177086798, + "learning_rate": 1.700003448513624e-05, + "loss": 1.1032, + "step": 11698 + }, + { + "epoch": 0.28, + "grad_norm": 2.3453462286122737, + "learning_rate": 1.6999489547025782e-05, + "loss": 0.9437, + "step": 11699 + }, + { + "epoch": 0.28, + "grad_norm": 2.2920204664914876, + "learning_rate": 1.6998944568162075e-05, + "loss": 0.8891, + "step": 11700 + }, + { + "epoch": 0.28, + "grad_norm": 2.347856265444068, + "learning_rate": 1.69983995485483e-05, + "loss": 1.0733, + "step": 11701 + }, + { + "epoch": 0.28, + "grad_norm": 2.5022971305195023, + "learning_rate": 1.6997854488187632e-05, + "loss": 1.0893, + "step": 11702 + }, + { + "epoch": 0.28, + "grad_norm": 2.4235620385263195, + "learning_rate": 1.699730938708323e-05, + "loss": 1.0193, + "step": 11703 + }, + { + "epoch": 0.28, + "grad_norm": 1.159151003301111, + "learning_rate": 1.6996764245238282e-05, + "loss": 0.9894, + "step": 11704 + }, + { + "epoch": 0.28, + "grad_norm": 2.032614329764838, + "learning_rate": 1.6996219062655963e-05, + "loss": 1.135, + "step": 11705 + }, + { + "epoch": 0.28, + "grad_norm": 1.9396865861237405, + "learning_rate": 1.6995673839339435e-05, + "loss": 1.1579, + "step": 11706 + }, + { + "epoch": 0.28, + "grad_norm": 1.9476083974644298, + "learning_rate": 1.6995128575291882e-05, + "loss": 0.9763, + "step": 11707 + }, + { + "epoch": 0.28, + "grad_norm": 1.9375752330601712, + "learning_rate": 1.699458327051647e-05, + "loss": 1.1014, + "step": 11708 + }, + { + "epoch": 0.28, + "grad_norm": 2.015399008564644, + "learning_rate": 1.6994037925016386e-05, + "loss": 1.0532, + "step": 11709 + }, + { + "epoch": 0.28, + "grad_norm": 2.323061882240458, + "learning_rate": 1.69934925387948e-05, + "loss": 1.0295, + "step": 11710 + }, + { + "epoch": 0.28, + "grad_norm": 2.7553247900477054, + "learning_rate": 1.699294711185488e-05, + "loss": 1.1036, + "step": 11711 + }, + { + "epoch": 0.28, + "grad_norm": 2.2331432866898653, + "learning_rate": 1.6992401644199815e-05, + "loss": 1.0223, + "step": 11712 + }, + { + "epoch": 0.28, + "grad_norm": 2.139421379907461, + "learning_rate": 1.6991856135832767e-05, + "loss": 1.0345, + "step": 11713 + }, + { + "epoch": 0.28, + "grad_norm": 1.921300606149045, + "learning_rate": 1.6991310586756923e-05, + "loss": 1.1307, + "step": 11714 + }, + { + "epoch": 0.28, + "grad_norm": 2.062603274314674, + "learning_rate": 1.6990764996975456e-05, + "loss": 0.9501, + "step": 11715 + }, + { + "epoch": 0.28, + "grad_norm": 1.9909200315520443, + "learning_rate": 1.699021936649154e-05, + "loss": 1.0373, + "step": 11716 + }, + { + "epoch": 0.28, + "grad_norm": 2.1811918855425754, + "learning_rate": 1.6989673695308354e-05, + "loss": 0.9091, + "step": 11717 + }, + { + "epoch": 0.28, + "grad_norm": 1.954822689846546, + "learning_rate": 1.6989127983429077e-05, + "loss": 1.1014, + "step": 11718 + }, + { + "epoch": 0.28, + "grad_norm": 2.123356857827274, + "learning_rate": 1.698858223085688e-05, + "loss": 1.0115, + "step": 11719 + }, + { + "epoch": 0.28, + "grad_norm": 1.1998847797533652, + "learning_rate": 1.698803643759495e-05, + "loss": 0.9335, + "step": 11720 + }, + { + "epoch": 0.28, + "grad_norm": 2.690814710714929, + "learning_rate": 1.698749060364646e-05, + "loss": 0.9888, + "step": 11721 + }, + { + "epoch": 0.28, + "grad_norm": 1.040425209996611, + "learning_rate": 1.698694472901458e-05, + "loss": 0.977, + "step": 11722 + }, + { + "epoch": 0.28, + "grad_norm": 1.9717659609002565, + "learning_rate": 1.69863988137025e-05, + "loss": 0.9869, + "step": 11723 + }, + { + "epoch": 0.28, + "grad_norm": 2.107257191261908, + "learning_rate": 1.6985852857713397e-05, + "loss": 0.9526, + "step": 11724 + }, + { + "epoch": 0.28, + "grad_norm": 2.337703087821101, + "learning_rate": 1.6985306861050443e-05, + "loss": 1.1347, + "step": 11725 + }, + { + "epoch": 0.28, + "grad_norm": 2.3273601733806637, + "learning_rate": 1.6984760823716818e-05, + "loss": 1.0976, + "step": 11726 + }, + { + "epoch": 0.28, + "grad_norm": 2.075969052411086, + "learning_rate": 1.6984214745715712e-05, + "loss": 0.9379, + "step": 11727 + }, + { + "epoch": 0.28, + "grad_norm": 2.3360990273882507, + "learning_rate": 1.6983668627050292e-05, + "loss": 1.1199, + "step": 11728 + }, + { + "epoch": 0.28, + "grad_norm": 1.7921758765473124, + "learning_rate": 1.6983122467723743e-05, + "loss": 1.0735, + "step": 11729 + }, + { + "epoch": 0.28, + "grad_norm": 2.201562281476417, + "learning_rate": 1.6982576267739247e-05, + "loss": 1.0563, + "step": 11730 + }, + { + "epoch": 0.28, + "grad_norm": 2.1047452716374098, + "learning_rate": 1.6982030027099975e-05, + "loss": 0.922, + "step": 11731 + }, + { + "epoch": 0.28, + "grad_norm": 2.6406009324870237, + "learning_rate": 1.698148374580912e-05, + "loss": 1.0223, + "step": 11732 + }, + { + "epoch": 0.28, + "grad_norm": 2.0397032847272407, + "learning_rate": 1.6980937423869852e-05, + "loss": 0.9624, + "step": 11733 + }, + { + "epoch": 0.28, + "grad_norm": 1.8895992718690402, + "learning_rate": 1.698039106128536e-05, + "loss": 0.9596, + "step": 11734 + }, + { + "epoch": 0.28, + "grad_norm": 2.137736601973182, + "learning_rate": 1.697984465805882e-05, + "loss": 1.0272, + "step": 11735 + }, + { + "epoch": 0.28, + "grad_norm": 2.0993830408595935, + "learning_rate": 1.6979298214193412e-05, + "loss": 1.0132, + "step": 11736 + }, + { + "epoch": 0.28, + "grad_norm": 2.12261750245031, + "learning_rate": 1.697875172969232e-05, + "loss": 1.0298, + "step": 11737 + }, + { + "epoch": 0.28, + "grad_norm": 5.302360333024298, + "learning_rate": 1.6978205204558728e-05, + "loss": 1.027, + "step": 11738 + }, + { + "epoch": 0.28, + "grad_norm": 2.283522900238567, + "learning_rate": 1.6977658638795816e-05, + "loss": 1.1187, + "step": 11739 + }, + { + "epoch": 0.28, + "grad_norm": 2.0033587073464205, + "learning_rate": 1.697711203240677e-05, + "loss": 1.1059, + "step": 11740 + }, + { + "epoch": 0.28, + "grad_norm": 2.0959733761986734, + "learning_rate": 1.6976565385394763e-05, + "loss": 0.9976, + "step": 11741 + }, + { + "epoch": 0.28, + "grad_norm": 1.9713442067342348, + "learning_rate": 1.6976018697762985e-05, + "loss": 1.0837, + "step": 11742 + }, + { + "epoch": 0.28, + "grad_norm": 1.882010777855668, + "learning_rate": 1.6975471969514617e-05, + "loss": 0.909, + "step": 11743 + }, + { + "epoch": 0.28, + "grad_norm": 2.2068783593179235, + "learning_rate": 1.6974925200652846e-05, + "loss": 0.8573, + "step": 11744 + }, + { + "epoch": 0.28, + "grad_norm": 2.016407276058712, + "learning_rate": 1.697437839118085e-05, + "loss": 1.0695, + "step": 11745 + }, + { + "epoch": 0.28, + "grad_norm": 1.2083083818092257, + "learning_rate": 1.6973831541101816e-05, + "loss": 0.9691, + "step": 11746 + }, + { + "epoch": 0.28, + "grad_norm": 2.180197547342593, + "learning_rate": 1.6973284650418923e-05, + "loss": 1.1359, + "step": 11747 + }, + { + "epoch": 0.28, + "grad_norm": 1.7253378083606834, + "learning_rate": 1.6972737719135365e-05, + "loss": 0.9989, + "step": 11748 + }, + { + "epoch": 0.28, + "grad_norm": 1.8595496411086616, + "learning_rate": 1.697219074725432e-05, + "loss": 0.9944, + "step": 11749 + }, + { + "epoch": 0.28, + "grad_norm": 2.8305887507987477, + "learning_rate": 1.6971643734778967e-05, + "loss": 1.1138, + "step": 11750 + }, + { + "epoch": 0.28, + "grad_norm": 2.042512951327448, + "learning_rate": 1.6971096681712502e-05, + "loss": 1.0451, + "step": 11751 + }, + { + "epoch": 0.28, + "grad_norm": 2.114754710491473, + "learning_rate": 1.6970549588058106e-05, + "loss": 0.9934, + "step": 11752 + }, + { + "epoch": 0.28, + "grad_norm": 2.299655033818676, + "learning_rate": 1.6970002453818963e-05, + "loss": 0.941, + "step": 11753 + }, + { + "epoch": 0.28, + "grad_norm": 2.0373237110561195, + "learning_rate": 1.6969455278998254e-05, + "loss": 1.0981, + "step": 11754 + }, + { + "epoch": 0.28, + "grad_norm": 2.088428145544581, + "learning_rate": 1.6968908063599175e-05, + "loss": 0.9927, + "step": 11755 + }, + { + "epoch": 0.28, + "grad_norm": 2.3082386922186067, + "learning_rate": 1.6968360807624904e-05, + "loss": 1.0436, + "step": 11756 + }, + { + "epoch": 0.28, + "grad_norm": 1.9095288568360673, + "learning_rate": 1.6967813511078633e-05, + "loss": 0.9748, + "step": 11757 + }, + { + "epoch": 0.28, + "grad_norm": 2.0309027410513587, + "learning_rate": 1.6967266173963545e-05, + "loss": 1.0326, + "step": 11758 + }, + { + "epoch": 0.28, + "grad_norm": 2.393386129746025, + "learning_rate": 1.6966718796282827e-05, + "loss": 1.0666, + "step": 11759 + }, + { + "epoch": 0.28, + "grad_norm": 1.8624944966480135, + "learning_rate": 1.6966171378039665e-05, + "loss": 1.0709, + "step": 11760 + }, + { + "epoch": 0.28, + "grad_norm": 2.039828133897678, + "learning_rate": 1.696562391923725e-05, + "loss": 0.9247, + "step": 11761 + }, + { + "epoch": 0.28, + "grad_norm": 2.060362757461543, + "learning_rate": 1.6965076419878764e-05, + "loss": 0.8721, + "step": 11762 + }, + { + "epoch": 0.28, + "grad_norm": 1.712241939203409, + "learning_rate": 1.69645288799674e-05, + "loss": 0.8681, + "step": 11763 + }, + { + "epoch": 0.28, + "grad_norm": 1.8625177888237747, + "learning_rate": 1.6963981299506348e-05, + "loss": 1.2112, + "step": 11764 + }, + { + "epoch": 0.28, + "grad_norm": 1.9544004841262432, + "learning_rate": 1.6963433678498786e-05, + "loss": 1.0929, + "step": 11765 + }, + { + "epoch": 0.28, + "grad_norm": 1.86053580350503, + "learning_rate": 1.696288601694791e-05, + "loss": 1.0118, + "step": 11766 + }, + { + "epoch": 0.28, + "grad_norm": 1.1017715696606283, + "learning_rate": 1.696233831485691e-05, + "loss": 1.0229, + "step": 11767 + }, + { + "epoch": 0.28, + "grad_norm": 1.1285536031284398, + "learning_rate": 1.6961790572228967e-05, + "loss": 0.9699, + "step": 11768 + }, + { + "epoch": 0.28, + "grad_norm": 2.3115541901696286, + "learning_rate": 1.6961242789067282e-05, + "loss": 1.0153, + "step": 11769 + }, + { + "epoch": 0.28, + "grad_norm": 1.1010887410417844, + "learning_rate": 1.6960694965375033e-05, + "loss": 0.9775, + "step": 11770 + }, + { + "epoch": 0.28, + "grad_norm": 1.9419159180641115, + "learning_rate": 1.6960147101155415e-05, + "loss": 1.0141, + "step": 11771 + }, + { + "epoch": 0.28, + "grad_norm": 2.0126910840072894, + "learning_rate": 1.695959919641162e-05, + "loss": 1.0056, + "step": 11772 + }, + { + "epoch": 0.28, + "grad_norm": 1.0846427965862404, + "learning_rate": 1.695905125114683e-05, + "loss": 0.9881, + "step": 11773 + }, + { + "epoch": 0.28, + "grad_norm": 2.080910312714087, + "learning_rate": 1.6958503265364243e-05, + "loss": 1.0949, + "step": 11774 + }, + { + "epoch": 0.28, + "grad_norm": 2.209392399357716, + "learning_rate": 1.6957955239067046e-05, + "loss": 0.9421, + "step": 11775 + }, + { + "epoch": 0.28, + "grad_norm": 2.6620283759970316, + "learning_rate": 1.695740717225843e-05, + "loss": 1.096, + "step": 11776 + }, + { + "epoch": 0.28, + "grad_norm": 2.0831422841742118, + "learning_rate": 1.695685906494159e-05, + "loss": 1.0161, + "step": 11777 + }, + { + "epoch": 0.28, + "grad_norm": 2.1589344261483157, + "learning_rate": 1.6956310917119712e-05, + "loss": 1.0058, + "step": 11778 + }, + { + "epoch": 0.28, + "grad_norm": 2.1281369470742004, + "learning_rate": 1.695576272879599e-05, + "loss": 1.0211, + "step": 11779 + }, + { + "epoch": 0.28, + "grad_norm": 2.0442812368357024, + "learning_rate": 1.6955214499973616e-05, + "loss": 1.0833, + "step": 11780 + }, + { + "epoch": 0.28, + "grad_norm": 1.8672685010799783, + "learning_rate": 1.695466623065578e-05, + "loss": 1.0033, + "step": 11781 + }, + { + "epoch": 0.28, + "grad_norm": 1.865714754975417, + "learning_rate": 1.6954117920845672e-05, + "loss": 1.137, + "step": 11782 + }, + { + "epoch": 0.28, + "grad_norm": 2.1447183051027214, + "learning_rate": 1.6953569570546494e-05, + "loss": 1.0672, + "step": 11783 + }, + { + "epoch": 0.28, + "grad_norm": 2.30145446864679, + "learning_rate": 1.695302117976143e-05, + "loss": 0.8926, + "step": 11784 + }, + { + "epoch": 0.28, + "grad_norm": 1.970742753715477, + "learning_rate": 1.6952472748493675e-05, + "loss": 0.953, + "step": 11785 + }, + { + "epoch": 0.28, + "grad_norm": 2.1182657877359907, + "learning_rate": 1.6951924276746425e-05, + "loss": 1.0694, + "step": 11786 + }, + { + "epoch": 0.28, + "grad_norm": 1.9996968151071324, + "learning_rate": 1.6951375764522868e-05, + "loss": 0.9321, + "step": 11787 + }, + { + "epoch": 0.28, + "grad_norm": 2.2575078707372125, + "learning_rate": 1.69508272118262e-05, + "loss": 0.941, + "step": 11788 + }, + { + "epoch": 0.28, + "grad_norm": 2.594303625067559, + "learning_rate": 1.695027861865962e-05, + "loss": 1.0629, + "step": 11789 + }, + { + "epoch": 0.28, + "grad_norm": 1.9979617354962436, + "learning_rate": 1.6949729985026314e-05, + "loss": 1.0936, + "step": 11790 + }, + { + "epoch": 0.28, + "grad_norm": 1.114472741896338, + "learning_rate": 1.6949181310929476e-05, + "loss": 0.9723, + "step": 11791 + }, + { + "epoch": 0.28, + "grad_norm": 1.9909489230319002, + "learning_rate": 1.694863259637231e-05, + "loss": 0.8552, + "step": 11792 + }, + { + "epoch": 0.28, + "grad_norm": 1.936361352432181, + "learning_rate": 1.6948083841358004e-05, + "loss": 1.0105, + "step": 11793 + }, + { + "epoch": 0.28, + "grad_norm": 1.0957323933583503, + "learning_rate": 1.6947535045889754e-05, + "loss": 1.0209, + "step": 11794 + }, + { + "epoch": 0.28, + "grad_norm": 1.9459700057919533, + "learning_rate": 1.6946986209970754e-05, + "loss": 1.0787, + "step": 11795 + }, + { + "epoch": 0.28, + "grad_norm": 2.140143732150382, + "learning_rate": 1.6946437333604203e-05, + "loss": 0.9671, + "step": 11796 + }, + { + "epoch": 0.28, + "grad_norm": 2.062200434538221, + "learning_rate": 1.6945888416793296e-05, + "loss": 0.8782, + "step": 11797 + }, + { + "epoch": 0.28, + "grad_norm": 1.9840608565442661, + "learning_rate": 1.694533945954122e-05, + "loss": 1.0016, + "step": 11798 + }, + { + "epoch": 0.28, + "grad_norm": 2.0066880291128664, + "learning_rate": 1.6944790461851186e-05, + "loss": 0.951, + "step": 11799 + }, + { + "epoch": 0.28, + "grad_norm": 2.202873360288045, + "learning_rate": 1.694424142372638e-05, + "loss": 1.0319, + "step": 11800 + }, + { + "epoch": 0.28, + "grad_norm": 1.168521817934316, + "learning_rate": 1.6943692345170005e-05, + "loss": 1.0837, + "step": 11801 + }, + { + "epoch": 0.28, + "grad_norm": 1.1617823218692442, + "learning_rate": 1.6943143226185252e-05, + "loss": 1.0309, + "step": 11802 + }, + { + "epoch": 0.28, + "grad_norm": 2.2470201005821364, + "learning_rate": 1.694259406677532e-05, + "loss": 1.0223, + "step": 11803 + }, + { + "epoch": 0.28, + "grad_norm": 2.1248138974939796, + "learning_rate": 1.6942044866943414e-05, + "loss": 0.931, + "step": 11804 + }, + { + "epoch": 0.28, + "grad_norm": 1.780918866334448, + "learning_rate": 1.6941495626692718e-05, + "loss": 1.0695, + "step": 11805 + }, + { + "epoch": 0.28, + "grad_norm": 2.4856723403206833, + "learning_rate": 1.694094634602644e-05, + "loss": 0.9649, + "step": 11806 + }, + { + "epoch": 0.28, + "grad_norm": 1.9845398417662963, + "learning_rate": 1.6940397024947773e-05, + "loss": 1.131, + "step": 11807 + }, + { + "epoch": 0.28, + "grad_norm": 1.773634989322515, + "learning_rate": 1.693984766345992e-05, + "loss": 1.1405, + "step": 11808 + }, + { + "epoch": 0.28, + "grad_norm": 2.159617419107301, + "learning_rate": 1.6939298261566075e-05, + "loss": 0.9837, + "step": 11809 + }, + { + "epoch": 0.28, + "grad_norm": 1.9611045428573948, + "learning_rate": 1.6938748819269436e-05, + "loss": 1.0514, + "step": 11810 + }, + { + "epoch": 0.28, + "grad_norm": 2.3506143213229596, + "learning_rate": 1.6938199336573208e-05, + "loss": 0.9746, + "step": 11811 + }, + { + "epoch": 0.28, + "grad_norm": 1.9205255540433512, + "learning_rate": 1.6937649813480586e-05, + "loss": 1.1745, + "step": 11812 + }, + { + "epoch": 0.28, + "grad_norm": 2.1940629749648366, + "learning_rate": 1.6937100249994772e-05, + "loss": 1.0623, + "step": 11813 + }, + { + "epoch": 0.28, + "grad_norm": 1.9829926614521638, + "learning_rate": 1.693655064611896e-05, + "loss": 1.0208, + "step": 11814 + }, + { + "epoch": 0.28, + "grad_norm": 2.1011213546215632, + "learning_rate": 1.693600100185636e-05, + "loss": 1.018, + "step": 11815 + }, + { + "epoch": 0.28, + "grad_norm": 1.0813481735619954, + "learning_rate": 1.693545131721016e-05, + "loss": 0.9489, + "step": 11816 + }, + { + "epoch": 0.28, + "grad_norm": 2.3629299356872204, + "learning_rate": 1.693490159218357e-05, + "loss": 1.077, + "step": 11817 + }, + { + "epoch": 0.28, + "grad_norm": 4.524818758111742, + "learning_rate": 1.6934351826779787e-05, + "loss": 0.991, + "step": 11818 + }, + { + "epoch": 0.28, + "grad_norm": 2.117629731715276, + "learning_rate": 1.6933802021002012e-05, + "loss": 1.1719, + "step": 11819 + }, + { + "epoch": 0.28, + "grad_norm": 1.1066166264496626, + "learning_rate": 1.6933252174853446e-05, + "loss": 0.9896, + "step": 11820 + }, + { + "epoch": 0.28, + "grad_norm": 1.983271568756488, + "learning_rate": 1.6932702288337293e-05, + "loss": 0.975, + "step": 11821 + }, + { + "epoch": 0.28, + "grad_norm": 2.24154203593389, + "learning_rate": 1.693215236145675e-05, + "loss": 1.1042, + "step": 11822 + }, + { + "epoch": 0.28, + "grad_norm": 2.0368032713500988, + "learning_rate": 1.693160239421502e-05, + "loss": 0.9143, + "step": 11823 + }, + { + "epoch": 0.28, + "grad_norm": 3.3462257440548324, + "learning_rate": 1.693105238661531e-05, + "loss": 1.0239, + "step": 11824 + }, + { + "epoch": 0.28, + "grad_norm": 1.0448205548924685, + "learning_rate": 1.6930502338660817e-05, + "loss": 0.9724, + "step": 11825 + }, + { + "epoch": 0.28, + "grad_norm": 3.256372899921319, + "learning_rate": 1.6929952250354747e-05, + "loss": 1.0289, + "step": 11826 + }, + { + "epoch": 0.28, + "grad_norm": 1.1135604440340106, + "learning_rate": 1.6929402121700298e-05, + "loss": 0.9976, + "step": 11827 + }, + { + "epoch": 0.28, + "grad_norm": 2.1850723634850726, + "learning_rate": 1.692885195270068e-05, + "loss": 1.1137, + "step": 11828 + }, + { + "epoch": 0.28, + "grad_norm": 2.3331574205010774, + "learning_rate": 1.6928301743359087e-05, + "loss": 0.9904, + "step": 11829 + }, + { + "epoch": 0.28, + "grad_norm": 2.042327892943648, + "learning_rate": 1.692775149367873e-05, + "loss": 1.0922, + "step": 11830 + }, + { + "epoch": 0.28, + "grad_norm": 2.42018932372481, + "learning_rate": 1.692720120366281e-05, + "loss": 1.0163, + "step": 11831 + }, + { + "epoch": 0.28, + "grad_norm": 2.184809343197975, + "learning_rate": 1.6926650873314535e-05, + "loss": 0.9954, + "step": 11832 + }, + { + "epoch": 0.28, + "grad_norm": 2.151310761318237, + "learning_rate": 1.69261005026371e-05, + "loss": 0.9508, + "step": 11833 + }, + { + "epoch": 0.28, + "grad_norm": 2.404629498686817, + "learning_rate": 1.692555009163372e-05, + "loss": 0.964, + "step": 11834 + }, + { + "epoch": 0.28, + "grad_norm": 2.7794941849212518, + "learning_rate": 1.6924999640307596e-05, + "loss": 0.9919, + "step": 11835 + }, + { + "epoch": 0.28, + "grad_norm": 2.12332286780122, + "learning_rate": 1.6924449148661927e-05, + "loss": 1.1141, + "step": 11836 + }, + { + "epoch": 0.28, + "grad_norm": 2.0805983393227727, + "learning_rate": 1.6923898616699925e-05, + "loss": 1.1061, + "step": 11837 + }, + { + "epoch": 0.28, + "grad_norm": 1.8918547721780303, + "learning_rate": 1.6923348044424793e-05, + "loss": 1.0604, + "step": 11838 + }, + { + "epoch": 0.28, + "grad_norm": 2.3659423569623548, + "learning_rate": 1.6922797431839736e-05, + "loss": 1.0082, + "step": 11839 + }, + { + "epoch": 0.28, + "grad_norm": 2.184827295986635, + "learning_rate": 1.692224677894796e-05, + "loss": 1.1747, + "step": 11840 + }, + { + "epoch": 0.28, + "grad_norm": 1.901816994641488, + "learning_rate": 1.6921696085752674e-05, + "loss": 0.9831, + "step": 11841 + }, + { + "epoch": 0.28, + "grad_norm": 2.175140356667959, + "learning_rate": 1.692114535225708e-05, + "loss": 0.855, + "step": 11842 + }, + { + "epoch": 0.28, + "grad_norm": 2.2016270344836455, + "learning_rate": 1.6920594578464386e-05, + "loss": 1.0843, + "step": 11843 + }, + { + "epoch": 0.28, + "grad_norm": 2.0485289228375136, + "learning_rate": 1.6920043764377804e-05, + "loss": 1.1964, + "step": 11844 + }, + { + "epoch": 0.28, + "grad_norm": 1.9971044374269185, + "learning_rate": 1.6919492910000534e-05, + "loss": 0.9517, + "step": 11845 + }, + { + "epoch": 0.28, + "grad_norm": 1.9972555508313818, + "learning_rate": 1.6918942015335785e-05, + "loss": 1.0584, + "step": 11846 + }, + { + "epoch": 0.28, + "grad_norm": 2.192353139540578, + "learning_rate": 1.6918391080386763e-05, + "loss": 1.0652, + "step": 11847 + }, + { + "epoch": 0.28, + "grad_norm": 2.2316194571733603, + "learning_rate": 1.691784010515668e-05, + "loss": 1.0056, + "step": 11848 + }, + { + "epoch": 0.28, + "grad_norm": 1.1149780639458655, + "learning_rate": 1.6917289089648742e-05, + "loss": 0.966, + "step": 11849 + }, + { + "epoch": 0.28, + "grad_norm": 2.146111377835305, + "learning_rate": 1.6916738033866153e-05, + "loss": 0.9998, + "step": 11850 + }, + { + "epoch": 0.28, + "grad_norm": 2.139528617219998, + "learning_rate": 1.6916186937812132e-05, + "loss": 1.0479, + "step": 11851 + }, + { + "epoch": 0.28, + "grad_norm": 2.3071183760468292, + "learning_rate": 1.6915635801489876e-05, + "loss": 1.0376, + "step": 11852 + }, + { + "epoch": 0.28, + "grad_norm": 2.2485004768223824, + "learning_rate": 1.6915084624902602e-05, + "loss": 0.9889, + "step": 11853 + }, + { + "epoch": 0.28, + "grad_norm": 2.4620859699121804, + "learning_rate": 1.6914533408053513e-05, + "loss": 1.0201, + "step": 11854 + }, + { + "epoch": 0.28, + "grad_norm": 2.381618209254431, + "learning_rate": 1.6913982150945823e-05, + "loss": 0.8741, + "step": 11855 + }, + { + "epoch": 0.28, + "grad_norm": 2.102203194368395, + "learning_rate": 1.691343085358274e-05, + "loss": 1.1259, + "step": 11856 + }, + { + "epoch": 0.28, + "grad_norm": 1.881389915081795, + "learning_rate": 1.691287951596747e-05, + "loss": 1.0602, + "step": 11857 + }, + { + "epoch": 0.28, + "grad_norm": 4.244128326346848, + "learning_rate": 1.6912328138103237e-05, + "loss": 1.1169, + "step": 11858 + }, + { + "epoch": 0.28, + "grad_norm": 1.8129448564426243, + "learning_rate": 1.6911776719993232e-05, + "loss": 1.0916, + "step": 11859 + }, + { + "epoch": 0.28, + "grad_norm": 2.0198245885771473, + "learning_rate": 1.6911225261640678e-05, + "loss": 1.1618, + "step": 11860 + }, + { + "epoch": 0.28, + "grad_norm": 2.0756979645469618, + "learning_rate": 1.691067376304878e-05, + "loss": 1.0693, + "step": 11861 + }, + { + "epoch": 0.28, + "grad_norm": 1.1103891160413628, + "learning_rate": 1.6910122224220757e-05, + "loss": 0.9467, + "step": 11862 + }, + { + "epoch": 0.28, + "grad_norm": 1.962153015459737, + "learning_rate": 1.6909570645159808e-05, + "loss": 0.9917, + "step": 11863 + }, + { + "epoch": 0.28, + "grad_norm": 2.061441369842374, + "learning_rate": 1.6909019025869158e-05, + "loss": 1.0491, + "step": 11864 + }, + { + "epoch": 0.28, + "grad_norm": 2.1159675463240157, + "learning_rate": 1.6908467366352004e-05, + "loss": 1.1395, + "step": 11865 + }, + { + "epoch": 0.28, + "grad_norm": 2.208477730857082, + "learning_rate": 1.690791566661157e-05, + "loss": 1.0293, + "step": 11866 + }, + { + "epoch": 0.28, + "grad_norm": 2.1726595719844286, + "learning_rate": 1.690736392665107e-05, + "loss": 0.9933, + "step": 11867 + }, + { + "epoch": 0.28, + "grad_norm": 2.4059066574919794, + "learning_rate": 1.6906812146473703e-05, + "loss": 0.9535, + "step": 11868 + }, + { + "epoch": 0.28, + "grad_norm": 2.0450110545838194, + "learning_rate": 1.6906260326082692e-05, + "loss": 0.9896, + "step": 11869 + }, + { + "epoch": 0.28, + "grad_norm": 2.9919445490747187, + "learning_rate": 1.6905708465481242e-05, + "loss": 1.0537, + "step": 11870 + }, + { + "epoch": 0.28, + "grad_norm": 2.004991533185624, + "learning_rate": 1.6905156564672575e-05, + "loss": 0.892, + "step": 11871 + }, + { + "epoch": 0.28, + "grad_norm": 1.8621690018817858, + "learning_rate": 1.6904604623659898e-05, + "loss": 1.067, + "step": 11872 + }, + { + "epoch": 0.28, + "grad_norm": 1.9757667300619182, + "learning_rate": 1.690405264244643e-05, + "loss": 1.102, + "step": 11873 + }, + { + "epoch": 0.28, + "grad_norm": 2.1280682058272564, + "learning_rate": 1.690350062103538e-05, + "loss": 1.0729, + "step": 11874 + }, + { + "epoch": 0.28, + "grad_norm": 1.92319665701704, + "learning_rate": 1.6902948559429962e-05, + "loss": 0.9239, + "step": 11875 + }, + { + "epoch": 0.28, + "grad_norm": 2.0014855557552176, + "learning_rate": 1.6902396457633397e-05, + "loss": 1.0737, + "step": 11876 + }, + { + "epoch": 0.28, + "grad_norm": 2.072118641137539, + "learning_rate": 1.6901844315648893e-05, + "loss": 1.2043, + "step": 11877 + }, + { + "epoch": 0.28, + "grad_norm": 2.1449210101756164, + "learning_rate": 1.690129213347966e-05, + "loss": 1.0646, + "step": 11878 + }, + { + "epoch": 0.28, + "grad_norm": 1.9866418610658267, + "learning_rate": 1.690073991112893e-05, + "loss": 0.9629, + "step": 11879 + }, + { + "epoch": 0.28, + "grad_norm": 2.0665966320988494, + "learning_rate": 1.69001876485999e-05, + "loss": 1.0353, + "step": 11880 + }, + { + "epoch": 0.28, + "grad_norm": 2.1045642845965205, + "learning_rate": 1.6899635345895796e-05, + "loss": 0.9557, + "step": 11881 + }, + { + "epoch": 0.28, + "grad_norm": 2.317202912319242, + "learning_rate": 1.689908300301983e-05, + "loss": 1.0294, + "step": 11882 + }, + { + "epoch": 0.28, + "grad_norm": 2.096268912236916, + "learning_rate": 1.6898530619975218e-05, + "loss": 0.9049, + "step": 11883 + }, + { + "epoch": 0.28, + "grad_norm": 2.3914544014303067, + "learning_rate": 1.6897978196765174e-05, + "loss": 0.9146, + "step": 11884 + }, + { + "epoch": 0.28, + "grad_norm": 1.888592892941132, + "learning_rate": 1.689742573339292e-05, + "loss": 0.9476, + "step": 11885 + }, + { + "epoch": 0.28, + "grad_norm": 2.201178137339609, + "learning_rate": 1.689687322986167e-05, + "loss": 1.1209, + "step": 11886 + }, + { + "epoch": 0.28, + "grad_norm": 1.8436105725816887, + "learning_rate": 1.689632068617464e-05, + "loss": 1.0972, + "step": 11887 + }, + { + "epoch": 0.28, + "grad_norm": 2.6121092887243478, + "learning_rate": 1.6895768102335045e-05, + "loss": 1.1295, + "step": 11888 + }, + { + "epoch": 0.28, + "grad_norm": 1.997924686170141, + "learning_rate": 1.689521547834611e-05, + "loss": 1.2648, + "step": 11889 + }, + { + "epoch": 0.28, + "grad_norm": 1.9997595028011015, + "learning_rate": 1.6894662814211046e-05, + "loss": 1.0792, + "step": 11890 + }, + { + "epoch": 0.28, + "grad_norm": 1.1402292136467234, + "learning_rate": 1.689411010993307e-05, + "loss": 0.9675, + "step": 11891 + }, + { + "epoch": 0.28, + "grad_norm": 2.293525996987453, + "learning_rate": 1.6893557365515404e-05, + "loss": 1.1904, + "step": 11892 + }, + { + "epoch": 0.28, + "grad_norm": 2.4334445554792166, + "learning_rate": 1.6893004580961264e-05, + "loss": 0.9251, + "step": 11893 + }, + { + "epoch": 0.28, + "grad_norm": 1.794115857130484, + "learning_rate": 1.689245175627387e-05, + "loss": 0.8706, + "step": 11894 + }, + { + "epoch": 0.28, + "grad_norm": 1.9620673521813516, + "learning_rate": 1.689189889145644e-05, + "loss": 0.9751, + "step": 11895 + }, + { + "epoch": 0.28, + "grad_norm": 1.1260138900008743, + "learning_rate": 1.689134598651219e-05, + "loss": 0.9867, + "step": 11896 + }, + { + "epoch": 0.28, + "grad_norm": 2.159566787852436, + "learning_rate": 1.6890793041444345e-05, + "loss": 0.8536, + "step": 11897 + }, + { + "epoch": 0.28, + "grad_norm": 2.0395079451380735, + "learning_rate": 1.689024005625612e-05, + "loss": 0.897, + "step": 11898 + }, + { + "epoch": 0.28, + "grad_norm": 2.079318737380556, + "learning_rate": 1.6889687030950736e-05, + "loss": 1.0416, + "step": 11899 + }, + { + "epoch": 0.28, + "grad_norm": 2.237095119305339, + "learning_rate": 1.688913396553141e-05, + "loss": 0.9529, + "step": 11900 + }, + { + "epoch": 0.28, + "grad_norm": 2.0480231554062778, + "learning_rate": 1.688858086000137e-05, + "loss": 1.0638, + "step": 11901 + }, + { + "epoch": 0.28, + "grad_norm": 2.272922155749751, + "learning_rate": 1.688802771436383e-05, + "loss": 1.047, + "step": 11902 + }, + { + "epoch": 0.28, + "grad_norm": 2.3302865102072094, + "learning_rate": 1.688747452862201e-05, + "loss": 0.9061, + "step": 11903 + }, + { + "epoch": 0.28, + "grad_norm": 2.3670765885015355, + "learning_rate": 1.6886921302779134e-05, + "loss": 1.066, + "step": 11904 + }, + { + "epoch": 0.28, + "grad_norm": 2.0723898376940917, + "learning_rate": 1.6886368036838422e-05, + "loss": 1.062, + "step": 11905 + }, + { + "epoch": 0.28, + "grad_norm": 2.4731025709187047, + "learning_rate": 1.6885814730803095e-05, + "loss": 1.0464, + "step": 11906 + }, + { + "epoch": 0.28, + "grad_norm": 2.2187666334113034, + "learning_rate": 1.6885261384676373e-05, + "loss": 1.0233, + "step": 11907 + }, + { + "epoch": 0.28, + "grad_norm": 2.6928584967895284, + "learning_rate": 1.688470799846148e-05, + "loss": 1.0194, + "step": 11908 + }, + { + "epoch": 0.28, + "grad_norm": 2.115533241845209, + "learning_rate": 1.6884154572161636e-05, + "loss": 0.8869, + "step": 11909 + }, + { + "epoch": 0.28, + "grad_norm": 1.9650999247281653, + "learning_rate": 1.6883601105780068e-05, + "loss": 0.9457, + "step": 11910 + }, + { + "epoch": 0.28, + "grad_norm": 1.114648094184713, + "learning_rate": 1.6883047599319994e-05, + "loss": 0.9342, + "step": 11911 + }, + { + "epoch": 0.28, + "grad_norm": 2.21269847969147, + "learning_rate": 1.6882494052784637e-05, + "loss": 0.9918, + "step": 11912 + }, + { + "epoch": 0.28, + "grad_norm": 1.9127763651774154, + "learning_rate": 1.688194046617722e-05, + "loss": 1.0729, + "step": 11913 + }, + { + "epoch": 0.28, + "grad_norm": 2.194844769168412, + "learning_rate": 1.688138683950097e-05, + "loss": 1.0241, + "step": 11914 + }, + { + "epoch": 0.28, + "grad_norm": 2.087172134320347, + "learning_rate": 1.6880833172759105e-05, + "loss": 1.1848, + "step": 11915 + }, + { + "epoch": 0.28, + "grad_norm": 2.0705144387851546, + "learning_rate": 1.688027946595485e-05, + "loss": 1.0467, + "step": 11916 + }, + { + "epoch": 0.28, + "grad_norm": 2.3954483204454973, + "learning_rate": 1.687972571909143e-05, + "loss": 1.1101, + "step": 11917 + }, + { + "epoch": 0.28, + "grad_norm": 2.005886410950722, + "learning_rate": 1.6879171932172074e-05, + "loss": 1.035, + "step": 11918 + }, + { + "epoch": 0.28, + "grad_norm": 2.087835397137106, + "learning_rate": 1.6878618105199995e-05, + "loss": 0.9918, + "step": 11919 + }, + { + "epoch": 0.28, + "grad_norm": 2.154179479940117, + "learning_rate": 1.6878064238178423e-05, + "loss": 1.0696, + "step": 11920 + }, + { + "epoch": 0.28, + "grad_norm": 1.9441756389440523, + "learning_rate": 1.6877510331110586e-05, + "loss": 1.0095, + "step": 11921 + }, + { + "epoch": 0.28, + "grad_norm": 1.9879768407653342, + "learning_rate": 1.6876956383999707e-05, + "loss": 0.9447, + "step": 11922 + }, + { + "epoch": 0.28, + "grad_norm": 2.0621462705118496, + "learning_rate": 1.687640239684901e-05, + "loss": 1.0888, + "step": 11923 + }, + { + "epoch": 0.28, + "grad_norm": 1.9659560057704868, + "learning_rate": 1.6875848369661722e-05, + "loss": 0.9581, + "step": 11924 + }, + { + "epoch": 0.28, + "grad_norm": 2.251756413274713, + "learning_rate": 1.6875294302441067e-05, + "loss": 0.9462, + "step": 11925 + }, + { + "epoch": 0.28, + "grad_norm": 2.1937723809120144, + "learning_rate": 1.687474019519027e-05, + "loss": 0.8912, + "step": 11926 + }, + { + "epoch": 0.28, + "grad_norm": 2.2912161351939138, + "learning_rate": 1.6874186047912565e-05, + "loss": 1.1669, + "step": 11927 + }, + { + "epoch": 0.28, + "grad_norm": 1.7751704782179416, + "learning_rate": 1.6873631860611165e-05, + "loss": 1.1075, + "step": 11928 + }, + { + "epoch": 0.28, + "grad_norm": 1.95483928263787, + "learning_rate": 1.687307763328931e-05, + "loss": 1.0474, + "step": 11929 + }, + { + "epoch": 0.28, + "grad_norm": 2.069380886396209, + "learning_rate": 1.6872523365950218e-05, + "loss": 1.0346, + "step": 11930 + }, + { + "epoch": 0.28, + "grad_norm": 2.162507046409573, + "learning_rate": 1.6871969058597118e-05, + "loss": 1.0107, + "step": 11931 + }, + { + "epoch": 0.28, + "grad_norm": 2.3373772039270775, + "learning_rate": 1.687141471123324e-05, + "loss": 1.1854, + "step": 11932 + }, + { + "epoch": 0.28, + "grad_norm": 1.9645833051148847, + "learning_rate": 1.6870860323861813e-05, + "loss": 0.9356, + "step": 11933 + }, + { + "epoch": 0.28, + "grad_norm": 1.1111762002888121, + "learning_rate": 1.6870305896486058e-05, + "loss": 0.934, + "step": 11934 + }, + { + "epoch": 0.28, + "grad_norm": 2.046310839933549, + "learning_rate": 1.686975142910921e-05, + "loss": 1.0431, + "step": 11935 + }, + { + "epoch": 0.28, + "grad_norm": 2.209683311599812, + "learning_rate": 1.686919692173449e-05, + "loss": 0.9055, + "step": 11936 + }, + { + "epoch": 0.28, + "grad_norm": 2.112839880211575, + "learning_rate": 1.6868642374365135e-05, + "loss": 1.0754, + "step": 11937 + }, + { + "epoch": 0.28, + "grad_norm": 1.9035951500110728, + "learning_rate": 1.686808778700437e-05, + "loss": 0.9854, + "step": 11938 + }, + { + "epoch": 0.28, + "grad_norm": 1.9254785002568209, + "learning_rate": 1.686753315965542e-05, + "loss": 1.0098, + "step": 11939 + }, + { + "epoch": 0.28, + "grad_norm": 2.2638982179025033, + "learning_rate": 1.686697849232152e-05, + "loss": 0.9621, + "step": 11940 + }, + { + "epoch": 0.28, + "grad_norm": 1.9790524058066292, + "learning_rate": 1.6866423785005895e-05, + "loss": 1.0914, + "step": 11941 + }, + { + "epoch": 0.28, + "grad_norm": 2.165781614090476, + "learning_rate": 1.6865869037711776e-05, + "loss": 1.0273, + "step": 11942 + }, + { + "epoch": 0.28, + "grad_norm": 2.0192606681750185, + "learning_rate": 1.68653142504424e-05, + "loss": 1.0823, + "step": 11943 + }, + { + "epoch": 0.28, + "grad_norm": 2.222575348787935, + "learning_rate": 1.6864759423200984e-05, + "loss": 1.0191, + "step": 11944 + }, + { + "epoch": 0.28, + "grad_norm": 2.404352746127627, + "learning_rate": 1.6864204555990768e-05, + "loss": 1.0866, + "step": 11945 + }, + { + "epoch": 0.28, + "grad_norm": 2.1291221896839745, + "learning_rate": 1.6863649648814982e-05, + "loss": 1.1501, + "step": 11946 + }, + { + "epoch": 0.28, + "grad_norm": 2.288494290005413, + "learning_rate": 1.686309470167685e-05, + "loss": 1.0342, + "step": 11947 + }, + { + "epoch": 0.28, + "grad_norm": 2.032134672000274, + "learning_rate": 1.686253971457961e-05, + "loss": 0.9745, + "step": 11948 + }, + { + "epoch": 0.28, + "grad_norm": 1.9632494102735598, + "learning_rate": 1.6861984687526494e-05, + "loss": 1.0699, + "step": 11949 + }, + { + "epoch": 0.28, + "grad_norm": 2.207696578290733, + "learning_rate": 1.6861429620520725e-05, + "loss": 1.1987, + "step": 11950 + }, + { + "epoch": 0.28, + "grad_norm": 1.080680812346168, + "learning_rate": 1.6860874513565544e-05, + "loss": 0.9431, + "step": 11951 + }, + { + "epoch": 0.28, + "grad_norm": 2.5667758924755484, + "learning_rate": 1.6860319366664182e-05, + "loss": 1.1011, + "step": 11952 + }, + { + "epoch": 0.28, + "grad_norm": 2.0828322904679295, + "learning_rate": 1.6859764179819866e-05, + "loss": 1.079, + "step": 11953 + }, + { + "epoch": 0.28, + "grad_norm": 2.062777456311837, + "learning_rate": 1.6859208953035832e-05, + "loss": 1.0562, + "step": 11954 + }, + { + "epoch": 0.28, + "grad_norm": 1.819615075831626, + "learning_rate": 1.685865368631531e-05, + "loss": 1.0177, + "step": 11955 + }, + { + "epoch": 0.28, + "grad_norm": 2.0210809272791397, + "learning_rate": 1.6858098379661538e-05, + "loss": 1.1111, + "step": 11956 + }, + { + "epoch": 0.28, + "grad_norm": 2.1137019857302115, + "learning_rate": 1.685754303307774e-05, + "loss": 1.0224, + "step": 11957 + }, + { + "epoch": 0.28, + "grad_norm": 1.057016351271301, + "learning_rate": 1.6856987646567163e-05, + "loss": 0.9715, + "step": 11958 + }, + { + "epoch": 0.28, + "grad_norm": 1.1278171902244178, + "learning_rate": 1.685643222013303e-05, + "loss": 0.9237, + "step": 11959 + }, + { + "epoch": 0.28, + "grad_norm": 2.0536158082110134, + "learning_rate": 1.6855876753778574e-05, + "loss": 1.0915, + "step": 11960 + }, + { + "epoch": 0.28, + "grad_norm": 2.0243941911972523, + "learning_rate": 1.6855321247507034e-05, + "loss": 1.0304, + "step": 11961 + }, + { + "epoch": 0.28, + "grad_norm": 1.1297417495342892, + "learning_rate": 1.6854765701321648e-05, + "loss": 0.9094, + "step": 11962 + }, + { + "epoch": 0.28, + "grad_norm": 2.2376518278873703, + "learning_rate": 1.6854210115225642e-05, + "loss": 1.0612, + "step": 11963 + }, + { + "epoch": 0.28, + "grad_norm": 1.8089300856313606, + "learning_rate": 1.6853654489222258e-05, + "loss": 0.9603, + "step": 11964 + }, + { + "epoch": 0.28, + "grad_norm": 2.430838249144781, + "learning_rate": 1.6853098823314728e-05, + "loss": 1.0741, + "step": 11965 + }, + { + "epoch": 0.28, + "grad_norm": 1.9267014218743672, + "learning_rate": 1.685254311750628e-05, + "loss": 1.1996, + "step": 11966 + }, + { + "epoch": 0.28, + "grad_norm": 1.9899113878971073, + "learning_rate": 1.6851987371800164e-05, + "loss": 1.0319, + "step": 11967 + }, + { + "epoch": 0.28, + "grad_norm": 2.0562243206685675, + "learning_rate": 1.6851431586199604e-05, + "loss": 1.0545, + "step": 11968 + }, + { + "epoch": 0.28, + "grad_norm": 2.6269249154990564, + "learning_rate": 1.6850875760707842e-05, + "loss": 0.969, + "step": 11969 + }, + { + "epoch": 0.28, + "grad_norm": 2.1213713010118735, + "learning_rate": 1.6850319895328112e-05, + "loss": 1.0209, + "step": 11970 + }, + { + "epoch": 0.28, + "grad_norm": 1.9868400831421906, + "learning_rate": 1.6849763990063654e-05, + "loss": 0.9372, + "step": 11971 + }, + { + "epoch": 0.28, + "grad_norm": 1.8613612777088142, + "learning_rate": 1.6849208044917696e-05, + "loss": 0.9134, + "step": 11972 + }, + { + "epoch": 0.28, + "grad_norm": 1.8592715443702477, + "learning_rate": 1.6848652059893485e-05, + "loss": 1.0498, + "step": 11973 + }, + { + "epoch": 0.28, + "grad_norm": 2.1040456205116187, + "learning_rate": 1.6848096034994253e-05, + "loss": 0.927, + "step": 11974 + }, + { + "epoch": 0.28, + "grad_norm": 1.9697454879033116, + "learning_rate": 1.684753997022323e-05, + "loss": 1.0222, + "step": 11975 + }, + { + "epoch": 0.28, + "grad_norm": 2.3521993040101705, + "learning_rate": 1.6846983865583674e-05, + "loss": 0.9106, + "step": 11976 + }, + { + "epoch": 0.28, + "grad_norm": 1.1731396661910156, + "learning_rate": 1.6846427721078798e-05, + "loss": 1.0316, + "step": 11977 + }, + { + "epoch": 0.28, + "grad_norm": 2.075540413379249, + "learning_rate": 1.684587153671186e-05, + "loss": 1.0994, + "step": 11978 + }, + { + "epoch": 0.28, + "grad_norm": 2.5544601605036092, + "learning_rate": 1.6845315312486088e-05, + "loss": 1.0964, + "step": 11979 + }, + { + "epoch": 0.28, + "grad_norm": 2.10508980563309, + "learning_rate": 1.6844759048404724e-05, + "loss": 1.1182, + "step": 11980 + }, + { + "epoch": 0.28, + "grad_norm": 2.3687142502344907, + "learning_rate": 1.6844202744471006e-05, + "loss": 1.0871, + "step": 11981 + }, + { + "epoch": 0.28, + "grad_norm": 1.8956572772455287, + "learning_rate": 1.684364640068817e-05, + "loss": 0.9963, + "step": 11982 + }, + { + "epoch": 0.28, + "grad_norm": 1.8376634046992673, + "learning_rate": 1.684309001705946e-05, + "loss": 1.0246, + "step": 11983 + }, + { + "epoch": 0.28, + "grad_norm": 1.8274676901874376, + "learning_rate": 1.6842533593588116e-05, + "loss": 1.0687, + "step": 11984 + }, + { + "epoch": 0.28, + "grad_norm": 1.9408135012936796, + "learning_rate": 1.684197713027737e-05, + "loss": 1.1438, + "step": 11985 + }, + { + "epoch": 0.28, + "grad_norm": 2.2779825126111324, + "learning_rate": 1.684142062713047e-05, + "loss": 1.0582, + "step": 11986 + }, + { + "epoch": 0.28, + "grad_norm": 2.3355840113313318, + "learning_rate": 1.6840864084150658e-05, + "loss": 1.0691, + "step": 11987 + }, + { + "epoch": 0.28, + "grad_norm": 2.030811568774458, + "learning_rate": 1.684030750134116e-05, + "loss": 1.044, + "step": 11988 + }, + { + "epoch": 0.28, + "grad_norm": 2.480433072937433, + "learning_rate": 1.683975087870523e-05, + "loss": 1.0608, + "step": 11989 + }, + { + "epoch": 0.28, + "grad_norm": 1.9597037991432145, + "learning_rate": 1.683919421624611e-05, + "loss": 0.9581, + "step": 11990 + }, + { + "epoch": 0.28, + "grad_norm": 1.8946412151046552, + "learning_rate": 1.683863751396703e-05, + "loss": 1.0738, + "step": 11991 + }, + { + "epoch": 0.28, + "grad_norm": 2.6419971041246932, + "learning_rate": 1.683808077187124e-05, + "loss": 1.0623, + "step": 11992 + }, + { + "epoch": 0.28, + "grad_norm": 2.20918304125916, + "learning_rate": 1.6837523989961973e-05, + "loss": 0.9823, + "step": 11993 + }, + { + "epoch": 0.28, + "grad_norm": 2.0799263907774956, + "learning_rate": 1.6836967168242483e-05, + "loss": 1.0987, + "step": 11994 + }, + { + "epoch": 0.28, + "grad_norm": 2.000912540153012, + "learning_rate": 1.6836410306716005e-05, + "loss": 1.0603, + "step": 11995 + }, + { + "epoch": 0.28, + "grad_norm": 2.0517567519285502, + "learning_rate": 1.6835853405385778e-05, + "loss": 1.1624, + "step": 11996 + }, + { + "epoch": 0.28, + "grad_norm": 1.9687251272030584, + "learning_rate": 1.6835296464255052e-05, + "loss": 1.0568, + "step": 11997 + }, + { + "epoch": 0.28, + "grad_norm": 1.91767040494175, + "learning_rate": 1.6834739483327066e-05, + "loss": 1.1722, + "step": 11998 + }, + { + "epoch": 0.28, + "grad_norm": 2.452550254868237, + "learning_rate": 1.6834182462605062e-05, + "loss": 0.9624, + "step": 11999 + }, + { + "epoch": 0.28, + "grad_norm": 2.4323664929995545, + "learning_rate": 1.6833625402092286e-05, + "loss": 1.0103, + "step": 12000 + }, + { + "epoch": 0.28, + "grad_norm": 1.9024232559166583, + "learning_rate": 1.6833068301791977e-05, + "loss": 1.0401, + "step": 12001 + }, + { + "epoch": 0.28, + "grad_norm": 1.9510889937866078, + "learning_rate": 1.683251116170738e-05, + "loss": 1.0601, + "step": 12002 + }, + { + "epoch": 0.28, + "grad_norm": 1.0776721884123508, + "learning_rate": 1.683195398184174e-05, + "loss": 1.005, + "step": 12003 + }, + { + "epoch": 0.28, + "grad_norm": 1.9657269311524217, + "learning_rate": 1.6831396762198303e-05, + "loss": 1.0222, + "step": 12004 + }, + { + "epoch": 0.28, + "grad_norm": 2.0780514170994184, + "learning_rate": 1.683083950278031e-05, + "loss": 1.0605, + "step": 12005 + }, + { + "epoch": 0.28, + "grad_norm": 2.024101562367118, + "learning_rate": 1.6830282203591007e-05, + "loss": 1.0187, + "step": 12006 + }, + { + "epoch": 0.28, + "grad_norm": 2.0111828380694976, + "learning_rate": 1.682972486463364e-05, + "loss": 0.9766, + "step": 12007 + }, + { + "epoch": 0.28, + "grad_norm": 2.08732309235602, + "learning_rate": 1.682916748591145e-05, + "loss": 1.0612, + "step": 12008 + }, + { + "epoch": 0.28, + "grad_norm": 2.0708236266946822, + "learning_rate": 1.6828610067427686e-05, + "loss": 1.1376, + "step": 12009 + }, + { + "epoch": 0.28, + "grad_norm": 2.174507511357728, + "learning_rate": 1.6828052609185592e-05, + "loss": 1.0241, + "step": 12010 + }, + { + "epoch": 0.28, + "grad_norm": 2.357428628339789, + "learning_rate": 1.6827495111188413e-05, + "loss": 0.991, + "step": 12011 + }, + { + "epoch": 0.28, + "grad_norm": 2.651851453955443, + "learning_rate": 1.6826937573439395e-05, + "loss": 0.9591, + "step": 12012 + }, + { + "epoch": 0.28, + "grad_norm": 2.1317080342730095, + "learning_rate": 1.682637999594179e-05, + "loss": 1.0697, + "step": 12013 + }, + { + "epoch": 0.28, + "grad_norm": 1.1398896964684864, + "learning_rate": 1.6825822378698836e-05, + "loss": 1.0004, + "step": 12014 + }, + { + "epoch": 0.28, + "grad_norm": 1.9382533064972485, + "learning_rate": 1.6825264721713787e-05, + "loss": 1.0748, + "step": 12015 + }, + { + "epoch": 0.28, + "grad_norm": 1.9564246795707252, + "learning_rate": 1.6824707024989885e-05, + "loss": 1.0857, + "step": 12016 + }, + { + "epoch": 0.28, + "grad_norm": 2.158419140301442, + "learning_rate": 1.6824149288530374e-05, + "loss": 1.0427, + "step": 12017 + }, + { + "epoch": 0.28, + "grad_norm": 2.2036182725555267, + "learning_rate": 1.6823591512338508e-05, + "loss": 1.0731, + "step": 12018 + }, + { + "epoch": 0.28, + "grad_norm": 2.391311640152724, + "learning_rate": 1.682303369641753e-05, + "loss": 0.9901, + "step": 12019 + }, + { + "epoch": 0.28, + "grad_norm": 2.029678020031178, + "learning_rate": 1.6822475840770692e-05, + "loss": 1.0785, + "step": 12020 + }, + { + "epoch": 0.28, + "grad_norm": 1.9541956858864893, + "learning_rate": 1.6821917945401237e-05, + "loss": 0.9926, + "step": 12021 + }, + { + "epoch": 0.28, + "grad_norm": 2.0004523004378494, + "learning_rate": 1.6821360010312418e-05, + "loss": 1.0809, + "step": 12022 + }, + { + "epoch": 0.28, + "grad_norm": 2.2140075889796442, + "learning_rate": 1.682080203550748e-05, + "loss": 0.9842, + "step": 12023 + }, + { + "epoch": 0.28, + "grad_norm": 2.043830647375693, + "learning_rate": 1.6820244020989672e-05, + "loss": 1.0576, + "step": 12024 + }, + { + "epoch": 0.28, + "grad_norm": 5.85964799566886, + "learning_rate": 1.681968596676225e-05, + "loss": 1.0509, + "step": 12025 + }, + { + "epoch": 0.28, + "grad_norm": 2.109279043831964, + "learning_rate": 1.681912787282845e-05, + "loss": 1.1178, + "step": 12026 + }, + { + "epoch": 0.28, + "grad_norm": 2.088221463279416, + "learning_rate": 1.6818569739191535e-05, + "loss": 0.9772, + "step": 12027 + }, + { + "epoch": 0.28, + "grad_norm": 2.100066945872083, + "learning_rate": 1.6818011565854744e-05, + "loss": 1.0479, + "step": 12028 + }, + { + "epoch": 0.28, + "grad_norm": 1.884480358682704, + "learning_rate": 1.6817453352821333e-05, + "loss": 1.095, + "step": 12029 + }, + { + "epoch": 0.28, + "grad_norm": 2.0159457597450983, + "learning_rate": 1.681689510009455e-05, + "loss": 1.0647, + "step": 12030 + }, + { + "epoch": 0.28, + "grad_norm": 1.9321496640397324, + "learning_rate": 1.6816336807677644e-05, + "loss": 0.999, + "step": 12031 + }, + { + "epoch": 0.28, + "grad_norm": 2.41834682457646, + "learning_rate": 1.681577847557387e-05, + "loss": 0.8457, + "step": 12032 + }, + { + "epoch": 0.28, + "grad_norm": 2.160206965304691, + "learning_rate": 1.6815220103786472e-05, + "loss": 1.071, + "step": 12033 + }, + { + "epoch": 0.28, + "grad_norm": 1.914890627747316, + "learning_rate": 1.6814661692318707e-05, + "loss": 1.0499, + "step": 12034 + }, + { + "epoch": 0.28, + "grad_norm": 2.1253516271361526, + "learning_rate": 1.6814103241173825e-05, + "loss": 1.0031, + "step": 12035 + }, + { + "epoch": 0.28, + "grad_norm": 2.126497965006185, + "learning_rate": 1.6813544750355074e-05, + "loss": 1.0967, + "step": 12036 + }, + { + "epoch": 0.28, + "grad_norm": 2.2991678902219306, + "learning_rate": 1.6812986219865712e-05, + "loss": 1.0318, + "step": 12037 + }, + { + "epoch": 0.28, + "grad_norm": 2.198495017611177, + "learning_rate": 1.6812427649708984e-05, + "loss": 1.2147, + "step": 12038 + }, + { + "epoch": 0.28, + "grad_norm": 2.7857231908854043, + "learning_rate": 1.681186903988815e-05, + "loss": 1.1359, + "step": 12039 + }, + { + "epoch": 0.28, + "grad_norm": 2.4070287245957256, + "learning_rate": 1.6811310390406454e-05, + "loss": 0.902, + "step": 12040 + }, + { + "epoch": 0.28, + "grad_norm": 2.1034812334475967, + "learning_rate": 1.681075170126715e-05, + "loss": 0.9596, + "step": 12041 + }, + { + "epoch": 0.28, + "grad_norm": 2.9724248875650767, + "learning_rate": 1.6810192972473497e-05, + "loss": 1.0366, + "step": 12042 + }, + { + "epoch": 0.28, + "grad_norm": 2.205041546028269, + "learning_rate": 1.6809634204028745e-05, + "loss": 0.9683, + "step": 12043 + }, + { + "epoch": 0.28, + "grad_norm": 2.7391784610576635, + "learning_rate": 1.6809075395936142e-05, + "loss": 0.878, + "step": 12044 + }, + { + "epoch": 0.28, + "grad_norm": 2.1139413630384687, + "learning_rate": 1.6808516548198955e-05, + "loss": 1.1493, + "step": 12045 + }, + { + "epoch": 0.28, + "grad_norm": 2.23616843757036, + "learning_rate": 1.6807957660820423e-05, + "loss": 1.0614, + "step": 12046 + }, + { + "epoch": 0.28, + "grad_norm": 2.0617592356881755, + "learning_rate": 1.6807398733803804e-05, + "loss": 1.0829, + "step": 12047 + }, + { + "epoch": 0.28, + "grad_norm": 2.0511796891211276, + "learning_rate": 1.680683976715236e-05, + "loss": 1.0287, + "step": 12048 + }, + { + "epoch": 0.28, + "grad_norm": 2.091215302370247, + "learning_rate": 1.6806280760869338e-05, + "loss": 1.0577, + "step": 12049 + }, + { + "epoch": 0.28, + "grad_norm": 2.1624188754069085, + "learning_rate": 1.6805721714957995e-05, + "loss": 1.0025, + "step": 12050 + }, + { + "epoch": 0.28, + "grad_norm": 1.8699098091082853, + "learning_rate": 1.6805162629421583e-05, + "loss": 1.1795, + "step": 12051 + }, + { + "epoch": 0.28, + "grad_norm": 1.8712836657093823, + "learning_rate": 1.6804603504263362e-05, + "loss": 1.0629, + "step": 12052 + }, + { + "epoch": 0.28, + "grad_norm": 1.6901997736853216, + "learning_rate": 1.6804044339486584e-05, + "loss": 1.097, + "step": 12053 + }, + { + "epoch": 0.28, + "grad_norm": 2.132575774657574, + "learning_rate": 1.6803485135094506e-05, + "loss": 1.048, + "step": 12054 + }, + { + "epoch": 0.28, + "grad_norm": 3.120484885412497, + "learning_rate": 1.6802925891090383e-05, + "loss": 0.9363, + "step": 12055 + }, + { + "epoch": 0.28, + "grad_norm": 1.09870751741521, + "learning_rate": 1.680236660747747e-05, + "loss": 0.9907, + "step": 12056 + }, + { + "epoch": 0.28, + "grad_norm": 2.1322980984070035, + "learning_rate": 1.680180728425903e-05, + "loss": 0.9483, + "step": 12057 + }, + { + "epoch": 0.28, + "grad_norm": 2.060571954447995, + "learning_rate": 1.6801247921438313e-05, + "loss": 1.1401, + "step": 12058 + }, + { + "epoch": 0.28, + "grad_norm": 2.2512021032445553, + "learning_rate": 1.6800688519018575e-05, + "loss": 1.0477, + "step": 12059 + }, + { + "epoch": 0.28, + "grad_norm": 2.1237746634533106, + "learning_rate": 1.680012907700308e-05, + "loss": 1.0145, + "step": 12060 + }, + { + "epoch": 0.28, + "grad_norm": 2.2581367437223325, + "learning_rate": 1.6799569595395074e-05, + "loss": 0.9948, + "step": 12061 + }, + { + "epoch": 0.28, + "grad_norm": 2.3743109467181265, + "learning_rate": 1.6799010074197827e-05, + "loss": 1.0663, + "step": 12062 + }, + { + "epoch": 0.28, + "grad_norm": 2.408875518287504, + "learning_rate": 1.6798450513414586e-05, + "loss": 1.09, + "step": 12063 + }, + { + "epoch": 0.28, + "grad_norm": 2.565335529492297, + "learning_rate": 1.6797890913048615e-05, + "loss": 1.0039, + "step": 12064 + }, + { + "epoch": 0.28, + "grad_norm": 2.1166077704073265, + "learning_rate": 1.6797331273103172e-05, + "loss": 1.0752, + "step": 12065 + }, + { + "epoch": 0.28, + "grad_norm": 2.1575978431480234, + "learning_rate": 1.6796771593581515e-05, + "loss": 1.004, + "step": 12066 + }, + { + "epoch": 0.28, + "grad_norm": 2.1487164200363953, + "learning_rate": 1.67962118744869e-05, + "loss": 1.0519, + "step": 12067 + }, + { + "epoch": 0.28, + "grad_norm": 2.0060183952817683, + "learning_rate": 1.6795652115822592e-05, + "loss": 1.002, + "step": 12068 + }, + { + "epoch": 0.28, + "grad_norm": 1.976221642533918, + "learning_rate": 1.6795092317591845e-05, + "loss": 1.0783, + "step": 12069 + }, + { + "epoch": 0.28, + "grad_norm": 2.032568030447027, + "learning_rate": 1.6794532479797917e-05, + "loss": 1.0541, + "step": 12070 + }, + { + "epoch": 0.28, + "grad_norm": 1.9795138033725028, + "learning_rate": 1.6793972602444068e-05, + "loss": 0.964, + "step": 12071 + }, + { + "epoch": 0.28, + "grad_norm": 2.2685282218570024, + "learning_rate": 1.679341268553356e-05, + "loss": 0.9351, + "step": 12072 + }, + { + "epoch": 0.28, + "grad_norm": 2.2663925038841812, + "learning_rate": 1.6792852729069655e-05, + "loss": 0.9961, + "step": 12073 + }, + { + "epoch": 0.28, + "grad_norm": 1.9781342690325625, + "learning_rate": 1.679229273305561e-05, + "loss": 0.9272, + "step": 12074 + }, + { + "epoch": 0.28, + "grad_norm": 2.3254735527642256, + "learning_rate": 1.6791732697494686e-05, + "loss": 1.0766, + "step": 12075 + }, + { + "epoch": 0.28, + "grad_norm": 1.1742695369804848, + "learning_rate": 1.6791172622390143e-05, + "loss": 0.9488, + "step": 12076 + }, + { + "epoch": 0.28, + "grad_norm": 3.8441212537093494, + "learning_rate": 1.6790612507745248e-05, + "loss": 1.2027, + "step": 12077 + }, + { + "epoch": 0.28, + "grad_norm": 1.9208614748708512, + "learning_rate": 1.6790052353563254e-05, + "loss": 0.9954, + "step": 12078 + }, + { + "epoch": 0.28, + "grad_norm": 2.355024676962908, + "learning_rate": 1.6789492159847423e-05, + "loss": 1.0636, + "step": 12079 + }, + { + "epoch": 0.28, + "grad_norm": 1.774905927043168, + "learning_rate": 1.6788931926601023e-05, + "loss": 1.0784, + "step": 12080 + }, + { + "epoch": 0.28, + "grad_norm": 1.0578644337796708, + "learning_rate": 1.6788371653827308e-05, + "loss": 0.9427, + "step": 12081 + }, + { + "epoch": 0.28, + "grad_norm": 2.153097619978163, + "learning_rate": 1.6787811341529545e-05, + "loss": 1.0492, + "step": 12082 + }, + { + "epoch": 0.28, + "grad_norm": 2.008182388820785, + "learning_rate": 1.6787250989711e-05, + "loss": 1.0536, + "step": 12083 + }, + { + "epoch": 0.28, + "grad_norm": 1.9434112047707208, + "learning_rate": 1.6786690598374925e-05, + "loss": 1.067, + "step": 12084 + }, + { + "epoch": 0.28, + "grad_norm": 2.167379477894447, + "learning_rate": 1.678613016752459e-05, + "loss": 1.1937, + "step": 12085 + }, + { + "epoch": 0.28, + "grad_norm": 1.9209615952038817, + "learning_rate": 1.678556969716326e-05, + "loss": 1.125, + "step": 12086 + }, + { + "epoch": 0.28, + "grad_norm": 1.969093692897438, + "learning_rate": 1.678500918729419e-05, + "loss": 1.118, + "step": 12087 + }, + { + "epoch": 0.28, + "grad_norm": 1.9992108491290155, + "learning_rate": 1.6784448637920654e-05, + "loss": 1.0439, + "step": 12088 + }, + { + "epoch": 0.28, + "grad_norm": 2.320317850564317, + "learning_rate": 1.6783888049045905e-05, + "loss": 1.0061, + "step": 12089 + }, + { + "epoch": 0.28, + "grad_norm": 1.041677270974383, + "learning_rate": 1.6783327420673217e-05, + "loss": 0.9912, + "step": 12090 + }, + { + "epoch": 0.28, + "grad_norm": 2.133362205810267, + "learning_rate": 1.6782766752805843e-05, + "loss": 1.0721, + "step": 12091 + }, + { + "epoch": 0.28, + "grad_norm": 2.245636007790171, + "learning_rate": 1.678220604544706e-05, + "loss": 0.9581, + "step": 12092 + }, + { + "epoch": 0.28, + "grad_norm": 2.8125208978903085, + "learning_rate": 1.6781645298600122e-05, + "loss": 1.0177, + "step": 12093 + }, + { + "epoch": 0.28, + "grad_norm": 2.1582231380432315, + "learning_rate": 1.67810845122683e-05, + "loss": 1.0318, + "step": 12094 + }, + { + "epoch": 0.28, + "grad_norm": 1.7711439621813492, + "learning_rate": 1.6780523686454856e-05, + "loss": 1.0739, + "step": 12095 + }, + { + "epoch": 0.28, + "grad_norm": 2.029302881349444, + "learning_rate": 1.6779962821163055e-05, + "loss": 1.1506, + "step": 12096 + }, + { + "epoch": 0.28, + "grad_norm": 2.1397008874210997, + "learning_rate": 1.6779401916396167e-05, + "loss": 1.1132, + "step": 12097 + }, + { + "epoch": 0.29, + "grad_norm": 2.636959725158984, + "learning_rate": 1.6778840972157453e-05, + "loss": 1.0034, + "step": 12098 + }, + { + "epoch": 0.29, + "grad_norm": 2.0015856628916575, + "learning_rate": 1.677827998845018e-05, + "loss": 1.1139, + "step": 12099 + }, + { + "epoch": 0.29, + "grad_norm": 1.812013609008091, + "learning_rate": 1.6777718965277618e-05, + "loss": 0.9921, + "step": 12100 + }, + { + "epoch": 0.29, + "grad_norm": 1.9985872784268885, + "learning_rate": 1.677715790264303e-05, + "loss": 1.0667, + "step": 12101 + }, + { + "epoch": 0.29, + "grad_norm": 1.8924162524071335, + "learning_rate": 1.677659680054968e-05, + "loss": 1.0213, + "step": 12102 + }, + { + "epoch": 0.29, + "grad_norm": 2.0869313178199214, + "learning_rate": 1.677603565900084e-05, + "loss": 1.0141, + "step": 12103 + }, + { + "epoch": 0.29, + "grad_norm": 2.0327990173388253, + "learning_rate": 1.6775474477999775e-05, + "loss": 1.1036, + "step": 12104 + }, + { + "epoch": 0.29, + "grad_norm": 2.158509638324907, + "learning_rate": 1.6774913257549753e-05, + "loss": 0.9559, + "step": 12105 + }, + { + "epoch": 0.29, + "grad_norm": 2.0408427350041594, + "learning_rate": 1.677435199765404e-05, + "loss": 0.9568, + "step": 12106 + }, + { + "epoch": 0.29, + "grad_norm": 2.1988816405225546, + "learning_rate": 1.6773790698315904e-05, + "loss": 1.0483, + "step": 12107 + }, + { + "epoch": 0.29, + "grad_norm": 2.5289687026626146, + "learning_rate": 1.6773229359538615e-05, + "loss": 1.2173, + "step": 12108 + }, + { + "epoch": 0.29, + "grad_norm": 2.1176388949303555, + "learning_rate": 1.677266798132544e-05, + "loss": 1.1102, + "step": 12109 + }, + { + "epoch": 0.29, + "grad_norm": 2.027391169591576, + "learning_rate": 1.6772106563679652e-05, + "loss": 0.94, + "step": 12110 + }, + { + "epoch": 0.29, + "grad_norm": 2.247907562976913, + "learning_rate": 1.677154510660451e-05, + "loss": 1.1076, + "step": 12111 + }, + { + "epoch": 0.29, + "grad_norm": 2.4569936538694583, + "learning_rate": 1.677098361010329e-05, + "loss": 1.0711, + "step": 12112 + }, + { + "epoch": 0.29, + "grad_norm": 2.3369562498533787, + "learning_rate": 1.677042207417926e-05, + "loss": 0.9704, + "step": 12113 + }, + { + "epoch": 0.29, + "grad_norm": 2.131785956712957, + "learning_rate": 1.676986049883569e-05, + "loss": 1.0, + "step": 12114 + }, + { + "epoch": 0.29, + "grad_norm": 1.9582890243217301, + "learning_rate": 1.676929888407585e-05, + "loss": 1.125, + "step": 12115 + }, + { + "epoch": 0.29, + "grad_norm": 2.9983192210644796, + "learning_rate": 1.6768737229903006e-05, + "loss": 1.0057, + "step": 12116 + }, + { + "epoch": 0.29, + "grad_norm": 2.5863431446895038, + "learning_rate": 1.6768175536320433e-05, + "loss": 1.1178, + "step": 12117 + }, + { + "epoch": 0.29, + "grad_norm": 2.4450966166340864, + "learning_rate": 1.67676138033314e-05, + "loss": 1.0077, + "step": 12118 + }, + { + "epoch": 0.29, + "grad_norm": 2.6151415834780853, + "learning_rate": 1.6767052030939176e-05, + "loss": 1.1167, + "step": 12119 + }, + { + "epoch": 0.29, + "grad_norm": 2.329426049375207, + "learning_rate": 1.676649021914703e-05, + "loss": 1.0516, + "step": 12120 + }, + { + "epoch": 0.29, + "grad_norm": 1.1761941113645717, + "learning_rate": 1.676592836795824e-05, + "loss": 1.0223, + "step": 12121 + }, + { + "epoch": 0.29, + "grad_norm": 2.592444595758775, + "learning_rate": 1.6765366477376067e-05, + "loss": 1.0612, + "step": 12122 + }, + { + "epoch": 0.29, + "grad_norm": 2.406700470155845, + "learning_rate": 1.6764804547403794e-05, + "loss": 1.1658, + "step": 12123 + }, + { + "epoch": 0.29, + "grad_norm": 2.184354438590315, + "learning_rate": 1.6764242578044685e-05, + "loss": 1.0996, + "step": 12124 + }, + { + "epoch": 0.29, + "grad_norm": 1.212486916118441, + "learning_rate": 1.6763680569302013e-05, + "loss": 0.9421, + "step": 12125 + }, + { + "epoch": 0.29, + "grad_norm": 2.2167782423521922, + "learning_rate": 1.6763118521179055e-05, + "loss": 1.0065, + "step": 12126 + }, + { + "epoch": 0.29, + "grad_norm": 2.3194191576433085, + "learning_rate": 1.6762556433679077e-05, + "loss": 1.0123, + "step": 12127 + }, + { + "epoch": 0.29, + "grad_norm": 1.902452288308113, + "learning_rate": 1.6761994306805352e-05, + "loss": 1.1401, + "step": 12128 + }, + { + "epoch": 0.29, + "grad_norm": 2.3840299821158983, + "learning_rate": 1.676143214056116e-05, + "loss": 0.9782, + "step": 12129 + }, + { + "epoch": 0.29, + "grad_norm": 2.2438114271747645, + "learning_rate": 1.676086993494977e-05, + "loss": 0.9873, + "step": 12130 + }, + { + "epoch": 0.29, + "grad_norm": 1.9098526797775968, + "learning_rate": 1.676030768997445e-05, + "loss": 0.9999, + "step": 12131 + }, + { + "epoch": 0.29, + "grad_norm": 2.222294813872344, + "learning_rate": 1.6759745405638483e-05, + "loss": 0.8696, + "step": 12132 + }, + { + "epoch": 0.29, + "grad_norm": 2.0557012817384117, + "learning_rate": 1.6759183081945132e-05, + "loss": 1.181, + "step": 12133 + }, + { + "epoch": 0.29, + "grad_norm": 1.9577735468806508, + "learning_rate": 1.6758620718897686e-05, + "loss": 1.0417, + "step": 12134 + }, + { + "epoch": 0.29, + "grad_norm": 2.1728130552544425, + "learning_rate": 1.6758058316499404e-05, + "loss": 0.9696, + "step": 12135 + }, + { + "epoch": 0.29, + "grad_norm": 1.0997355744001402, + "learning_rate": 1.675749587475357e-05, + "loss": 1.0758, + "step": 12136 + }, + { + "epoch": 0.29, + "grad_norm": 2.149064650249182, + "learning_rate": 1.6756933393663454e-05, + "loss": 1.0966, + "step": 12137 + }, + { + "epoch": 0.29, + "grad_norm": 2.202475288869052, + "learning_rate": 1.6756370873232333e-05, + "loss": 0.9587, + "step": 12138 + }, + { + "epoch": 0.29, + "grad_norm": 2.161334950112095, + "learning_rate": 1.6755808313463483e-05, + "loss": 1.0533, + "step": 12139 + }, + { + "epoch": 0.29, + "grad_norm": 2.264193086670548, + "learning_rate": 1.6755245714360176e-05, + "loss": 1.0318, + "step": 12140 + }, + { + "epoch": 0.29, + "grad_norm": 2.6624538824235633, + "learning_rate": 1.675468307592569e-05, + "loss": 0.9072, + "step": 12141 + }, + { + "epoch": 0.29, + "grad_norm": 2.323669545917449, + "learning_rate": 1.67541203981633e-05, + "loss": 1.0595, + "step": 12142 + }, + { + "epoch": 0.29, + "grad_norm": 2.4508655750916652, + "learning_rate": 1.6753557681076283e-05, + "loss": 0.9735, + "step": 12143 + }, + { + "epoch": 0.29, + "grad_norm": 2.0025380326840168, + "learning_rate": 1.675299492466792e-05, + "loss": 1.0152, + "step": 12144 + }, + { + "epoch": 0.29, + "grad_norm": 1.9845316057285223, + "learning_rate": 1.6752432128941478e-05, + "loss": 0.984, + "step": 12145 + }, + { + "epoch": 0.29, + "grad_norm": 1.9423063322116727, + "learning_rate": 1.675186929390024e-05, + "loss": 1.0531, + "step": 12146 + }, + { + "epoch": 0.29, + "grad_norm": 1.9393007444513999, + "learning_rate": 1.6751306419547477e-05, + "loss": 0.9829, + "step": 12147 + }, + { + "epoch": 0.29, + "grad_norm": 2.103211974809277, + "learning_rate": 1.6750743505886477e-05, + "loss": 1.0691, + "step": 12148 + }, + { + "epoch": 0.29, + "grad_norm": 1.8682991819352837, + "learning_rate": 1.6750180552920505e-05, + "loss": 1.0361, + "step": 12149 + }, + { + "epoch": 0.29, + "grad_norm": 1.8411375009026931, + "learning_rate": 1.674961756065285e-05, + "loss": 0.9548, + "step": 12150 + }, + { + "epoch": 0.29, + "grad_norm": 2.4650753999816586, + "learning_rate": 1.674905452908678e-05, + "loss": 1.0727, + "step": 12151 + }, + { + "epoch": 0.29, + "grad_norm": 2.2493155986802824, + "learning_rate": 1.6748491458225583e-05, + "loss": 1.0982, + "step": 12152 + }, + { + "epoch": 0.29, + "grad_norm": 2.2480122474900157, + "learning_rate": 1.674792834807253e-05, + "loss": 1.0467, + "step": 12153 + }, + { + "epoch": 0.29, + "grad_norm": 2.1337836106184875, + "learning_rate": 1.67473651986309e-05, + "loss": 1.1407, + "step": 12154 + }, + { + "epoch": 0.29, + "grad_norm": 2.3530200834255295, + "learning_rate": 1.6746802009903973e-05, + "loss": 0.9535, + "step": 12155 + }, + { + "epoch": 0.29, + "grad_norm": 2.1204436079783586, + "learning_rate": 1.674623878189503e-05, + "loss": 1.0302, + "step": 12156 + }, + { + "epoch": 0.29, + "grad_norm": 2.312011451781599, + "learning_rate": 1.6745675514607352e-05, + "loss": 1.0391, + "step": 12157 + }, + { + "epoch": 0.29, + "grad_norm": 2.4259085116006704, + "learning_rate": 1.674511220804421e-05, + "loss": 1.0703, + "step": 12158 + }, + { + "epoch": 0.29, + "grad_norm": 1.1283106369067912, + "learning_rate": 1.674454886220889e-05, + "loss": 0.9723, + "step": 12159 + }, + { + "epoch": 0.29, + "grad_norm": 1.9882536645097035, + "learning_rate": 1.6743985477104674e-05, + "loss": 0.9472, + "step": 12160 + }, + { + "epoch": 0.29, + "grad_norm": 2.368221249430645, + "learning_rate": 1.674342205273484e-05, + "loss": 1.0632, + "step": 12161 + }, + { + "epoch": 0.29, + "grad_norm": 2.2475814073346654, + "learning_rate": 1.6742858589102667e-05, + "loss": 0.9114, + "step": 12162 + }, + { + "epoch": 0.29, + "grad_norm": 1.8289982009885073, + "learning_rate": 1.6742295086211437e-05, + "loss": 1.0286, + "step": 12163 + }, + { + "epoch": 0.29, + "grad_norm": 1.8923408150503491, + "learning_rate": 1.674173154406443e-05, + "loss": 1.1059, + "step": 12164 + }, + { + "epoch": 0.29, + "grad_norm": 1.8906642247448189, + "learning_rate": 1.6741167962664926e-05, + "loss": 1.1024, + "step": 12165 + }, + { + "epoch": 0.29, + "grad_norm": 4.448897482448346, + "learning_rate": 1.674060434201621e-05, + "loss": 1.1651, + "step": 12166 + }, + { + "epoch": 0.29, + "grad_norm": 2.0116215997218703, + "learning_rate": 1.674004068212156e-05, + "loss": 1.0663, + "step": 12167 + }, + { + "epoch": 0.29, + "grad_norm": 3.7716895314163166, + "learning_rate": 1.6739476982984262e-05, + "loss": 0.9404, + "step": 12168 + }, + { + "epoch": 0.29, + "grad_norm": 2.0611132216338097, + "learning_rate": 1.6738913244607596e-05, + "loss": 0.9793, + "step": 12169 + }, + { + "epoch": 0.29, + "grad_norm": 2.1846485006709684, + "learning_rate": 1.6738349466994837e-05, + "loss": 1.1138, + "step": 12170 + }, + { + "epoch": 0.29, + "grad_norm": 1.9599582056630915, + "learning_rate": 1.673778565014928e-05, + "loss": 1.0565, + "step": 12171 + }, + { + "epoch": 0.29, + "grad_norm": 2.225679332121829, + "learning_rate": 1.6737221794074198e-05, + "loss": 0.9909, + "step": 12172 + }, + { + "epoch": 0.29, + "grad_norm": 1.097525449146791, + "learning_rate": 1.673665789877288e-05, + "loss": 0.9786, + "step": 12173 + }, + { + "epoch": 0.29, + "grad_norm": 2.2594200372205555, + "learning_rate": 1.6736093964248607e-05, + "loss": 1.178, + "step": 12174 + }, + { + "epoch": 0.29, + "grad_norm": 2.0537130610454573, + "learning_rate": 1.673552999050466e-05, + "loss": 1.0175, + "step": 12175 + }, + { + "epoch": 0.29, + "grad_norm": 1.9830321918087173, + "learning_rate": 1.673496597754433e-05, + "loss": 0.9663, + "step": 12176 + }, + { + "epoch": 0.29, + "grad_norm": 1.9787734165932727, + "learning_rate": 1.673440192537089e-05, + "loss": 1.0047, + "step": 12177 + }, + { + "epoch": 0.29, + "grad_norm": 2.2099639086170257, + "learning_rate": 1.6733837833987634e-05, + "loss": 1.0185, + "step": 12178 + }, + { + "epoch": 0.29, + "grad_norm": 2.1022788847795684, + "learning_rate": 1.6733273703397837e-05, + "loss": 1.027, + "step": 12179 + }, + { + "epoch": 0.29, + "grad_norm": 2.062540884712243, + "learning_rate": 1.6732709533604795e-05, + "loss": 0.9606, + "step": 12180 + }, + { + "epoch": 0.29, + "grad_norm": 2.0033137101191705, + "learning_rate": 1.6732145324611782e-05, + "loss": 1.1614, + "step": 12181 + }, + { + "epoch": 0.29, + "grad_norm": 1.8116583939687705, + "learning_rate": 1.6731581076422093e-05, + "loss": 1.0466, + "step": 12182 + }, + { + "epoch": 0.29, + "grad_norm": 2.09279768062404, + "learning_rate": 1.6731016789039004e-05, + "loss": 1.0273, + "step": 12183 + }, + { + "epoch": 0.29, + "grad_norm": 2.4217007801242763, + "learning_rate": 1.6730452462465804e-05, + "loss": 1.1304, + "step": 12184 + }, + { + "epoch": 0.29, + "grad_norm": 1.9847026329437016, + "learning_rate": 1.672988809670578e-05, + "loss": 1.0283, + "step": 12185 + }, + { + "epoch": 0.29, + "grad_norm": 2.123269617223535, + "learning_rate": 1.6729323691762215e-05, + "loss": 1.0496, + "step": 12186 + }, + { + "epoch": 0.29, + "grad_norm": 2.3139974216913592, + "learning_rate": 1.6728759247638403e-05, + "loss": 1.1814, + "step": 12187 + }, + { + "epoch": 0.29, + "grad_norm": 1.7623348531267833, + "learning_rate": 1.6728194764337616e-05, + "loss": 0.9161, + "step": 12188 + }, + { + "epoch": 0.29, + "grad_norm": 1.1350240454182676, + "learning_rate": 1.6727630241863154e-05, + "loss": 0.9455, + "step": 12189 + }, + { + "epoch": 0.29, + "grad_norm": 2.102715803412831, + "learning_rate": 1.6727065680218298e-05, + "loss": 0.9772, + "step": 12190 + }, + { + "epoch": 0.29, + "grad_norm": 2.139940975270811, + "learning_rate": 1.6726501079406333e-05, + "loss": 1.101, + "step": 12191 + }, + { + "epoch": 0.29, + "grad_norm": 2.0212420241118396, + "learning_rate": 1.6725936439430556e-05, + "loss": 1.1021, + "step": 12192 + }, + { + "epoch": 0.29, + "grad_norm": 2.162410703779081, + "learning_rate": 1.672537176029424e-05, + "loss": 1.1689, + "step": 12193 + }, + { + "epoch": 0.29, + "grad_norm": 1.1184682319917614, + "learning_rate": 1.6724807042000683e-05, + "loss": 0.9691, + "step": 12194 + }, + { + "epoch": 0.29, + "grad_norm": 1.084717850574065, + "learning_rate": 1.6724242284553173e-05, + "loss": 1.0022, + "step": 12195 + }, + { + "epoch": 0.29, + "grad_norm": 2.465082401537009, + "learning_rate": 1.6723677487954994e-05, + "loss": 0.9472, + "step": 12196 + }, + { + "epoch": 0.29, + "grad_norm": 1.8978095816593417, + "learning_rate": 1.6723112652209435e-05, + "loss": 0.9591, + "step": 12197 + }, + { + "epoch": 0.29, + "grad_norm": 2.1273303450548373, + "learning_rate": 1.6722547777319788e-05, + "loss": 1.2216, + "step": 12198 + }, + { + "epoch": 0.29, + "grad_norm": 1.978288907597315, + "learning_rate": 1.6721982863289337e-05, + "loss": 0.9994, + "step": 12199 + }, + { + "epoch": 0.29, + "grad_norm": 1.9736534097050678, + "learning_rate": 1.6721417910121373e-05, + "loss": 0.9094, + "step": 12200 + }, + { + "epoch": 0.29, + "grad_norm": 2.2648433694554604, + "learning_rate": 1.672085291781919e-05, + "loss": 1.0968, + "step": 12201 + }, + { + "epoch": 0.29, + "grad_norm": 2.102105630320757, + "learning_rate": 1.6720287886386065e-05, + "loss": 1.05, + "step": 12202 + }, + { + "epoch": 0.29, + "grad_norm": 2.287014154419632, + "learning_rate": 1.6719722815825304e-05, + "loss": 1.0975, + "step": 12203 + }, + { + "epoch": 0.29, + "grad_norm": 2.4989681825682073, + "learning_rate": 1.6719157706140187e-05, + "loss": 1.0732, + "step": 12204 + }, + { + "epoch": 0.29, + "grad_norm": 1.840816642241525, + "learning_rate": 1.671859255733401e-05, + "loss": 1.0111, + "step": 12205 + }, + { + "epoch": 0.29, + "grad_norm": 2.10282168801032, + "learning_rate": 1.6718027369410057e-05, + "loss": 0.9053, + "step": 12206 + }, + { + "epoch": 0.29, + "grad_norm": 2.1368848491034584, + "learning_rate": 1.6717462142371622e-05, + "loss": 0.9607, + "step": 12207 + }, + { + "epoch": 0.29, + "grad_norm": 2.1275801321191157, + "learning_rate": 1.6716896876221996e-05, + "loss": 1.0789, + "step": 12208 + }, + { + "epoch": 0.29, + "grad_norm": 1.7317499328086428, + "learning_rate": 1.6716331570964465e-05, + "loss": 1.0116, + "step": 12209 + }, + { + "epoch": 0.29, + "grad_norm": 1.9391947579705466, + "learning_rate": 1.6715766226602334e-05, + "loss": 1.1388, + "step": 12210 + }, + { + "epoch": 0.29, + "grad_norm": 3.6107211148909997, + "learning_rate": 1.671520084313888e-05, + "loss": 0.9862, + "step": 12211 + }, + { + "epoch": 0.29, + "grad_norm": 1.9945781384967856, + "learning_rate": 1.6714635420577405e-05, + "loss": 1.0541, + "step": 12212 + }, + { + "epoch": 0.29, + "grad_norm": 2.091363097542061, + "learning_rate": 1.6714069958921193e-05, + "loss": 1.0692, + "step": 12213 + }, + { + "epoch": 0.29, + "grad_norm": 2.114814875538248, + "learning_rate": 1.6713504458173538e-05, + "loss": 0.9656, + "step": 12214 + }, + { + "epoch": 0.29, + "grad_norm": 1.929581714662041, + "learning_rate": 1.671293891833774e-05, + "loss": 1.0267, + "step": 12215 + }, + { + "epoch": 0.29, + "grad_norm": 2.001472689270516, + "learning_rate": 1.6712373339417082e-05, + "loss": 1.0675, + "step": 12216 + }, + { + "epoch": 0.29, + "grad_norm": 2.3795637102492777, + "learning_rate": 1.6711807721414865e-05, + "loss": 1.0466, + "step": 12217 + }, + { + "epoch": 0.29, + "grad_norm": 2.3346997597198524, + "learning_rate": 1.6711242064334377e-05, + "loss": 1.1171, + "step": 12218 + }, + { + "epoch": 0.29, + "grad_norm": 1.8652452260096217, + "learning_rate": 1.6710676368178913e-05, + "loss": 1.0103, + "step": 12219 + }, + { + "epoch": 0.29, + "grad_norm": 1.0795338201845943, + "learning_rate": 1.6710110632951765e-05, + "loss": 1.025, + "step": 12220 + }, + { + "epoch": 0.29, + "grad_norm": 2.098844734060859, + "learning_rate": 1.6709544858656232e-05, + "loss": 1.0609, + "step": 12221 + }, + { + "epoch": 0.29, + "grad_norm": 1.923511348945394, + "learning_rate": 1.67089790452956e-05, + "loss": 1.0316, + "step": 12222 + }, + { + "epoch": 0.29, + "grad_norm": 2.073825791999466, + "learning_rate": 1.670841319287317e-05, + "loss": 1.0715, + "step": 12223 + }, + { + "epoch": 0.29, + "grad_norm": 2.0104091582946624, + "learning_rate": 1.6707847301392237e-05, + "loss": 1.203, + "step": 12224 + }, + { + "epoch": 0.29, + "grad_norm": 2.038875164500934, + "learning_rate": 1.670728137085609e-05, + "loss": 1.0308, + "step": 12225 + }, + { + "epoch": 0.29, + "grad_norm": 2.2405098111379025, + "learning_rate": 1.670671540126803e-05, + "loss": 1.0564, + "step": 12226 + }, + { + "epoch": 0.29, + "grad_norm": 1.9374423547214028, + "learning_rate": 1.670614939263135e-05, + "loss": 1.0278, + "step": 12227 + }, + { + "epoch": 0.29, + "grad_norm": 2.025782026000376, + "learning_rate": 1.670558334494934e-05, + "loss": 0.921, + "step": 12228 + }, + { + "epoch": 0.29, + "grad_norm": 1.1060965165856587, + "learning_rate": 1.6705017258225306e-05, + "loss": 0.9622, + "step": 12229 + }, + { + "epoch": 0.29, + "grad_norm": 1.0981341438260146, + "learning_rate": 1.6704451132462534e-05, + "loss": 0.9817, + "step": 12230 + }, + { + "epoch": 0.29, + "grad_norm": 2.031253625220035, + "learning_rate": 1.670388496766433e-05, + "loss": 0.9354, + "step": 12231 + }, + { + "epoch": 0.29, + "grad_norm": 2.211996991348197, + "learning_rate": 1.6703318763833985e-05, + "loss": 0.926, + "step": 12232 + }, + { + "epoch": 0.29, + "grad_norm": 1.9410856773687744, + "learning_rate": 1.670275252097479e-05, + "loss": 0.9481, + "step": 12233 + }, + { + "epoch": 0.29, + "grad_norm": 1.798921127263098, + "learning_rate": 1.6702186239090058e-05, + "loss": 0.9706, + "step": 12234 + }, + { + "epoch": 0.29, + "grad_norm": 2.5629739248593517, + "learning_rate": 1.6701619918183066e-05, + "loss": 0.9305, + "step": 12235 + }, + { + "epoch": 0.29, + "grad_norm": 2.0219534891132525, + "learning_rate": 1.6701053558257125e-05, + "loss": 0.8936, + "step": 12236 + }, + { + "epoch": 0.29, + "grad_norm": 2.3336112725244464, + "learning_rate": 1.670048715931553e-05, + "loss": 1.0496, + "step": 12237 + }, + { + "epoch": 0.29, + "grad_norm": 2.018998375276794, + "learning_rate": 1.6699920721361576e-05, + "loss": 0.9955, + "step": 12238 + }, + { + "epoch": 0.29, + "grad_norm": 2.0377882530199702, + "learning_rate": 1.669935424439856e-05, + "loss": 0.9895, + "step": 12239 + }, + { + "epoch": 0.29, + "grad_norm": 1.83921416635852, + "learning_rate": 1.6698787728429786e-05, + "loss": 1.1034, + "step": 12240 + }, + { + "epoch": 0.29, + "grad_norm": 2.065278580393435, + "learning_rate": 1.6698221173458547e-05, + "loss": 1.0887, + "step": 12241 + }, + { + "epoch": 0.29, + "grad_norm": 1.801465219036598, + "learning_rate": 1.6697654579488147e-05, + "loss": 1.0663, + "step": 12242 + }, + { + "epoch": 0.29, + "grad_norm": 2.9859101705932334, + "learning_rate": 1.669708794652188e-05, + "loss": 1.1148, + "step": 12243 + }, + { + "epoch": 0.29, + "grad_norm": 1.7824171785422949, + "learning_rate": 1.6696521274563046e-05, + "loss": 1.079, + "step": 12244 + }, + { + "epoch": 0.29, + "grad_norm": 2.073106135644819, + "learning_rate": 1.6695954563614945e-05, + "loss": 1.0665, + "step": 12245 + }, + { + "epoch": 0.29, + "grad_norm": 1.8098789585936899, + "learning_rate": 1.6695387813680873e-05, + "loss": 1.1283, + "step": 12246 + }, + { + "epoch": 0.29, + "grad_norm": 1.9359874200528007, + "learning_rate": 1.6694821024764142e-05, + "loss": 1.0857, + "step": 12247 + }, + { + "epoch": 0.29, + "grad_norm": 1.946578225648942, + "learning_rate": 1.6694254196868036e-05, + "loss": 0.957, + "step": 12248 + }, + { + "epoch": 0.29, + "grad_norm": 1.8957956366163882, + "learning_rate": 1.6693687329995864e-05, + "loss": 1.1785, + "step": 12249 + }, + { + "epoch": 0.29, + "grad_norm": 1.1772749174221404, + "learning_rate": 1.669312042415093e-05, + "loss": 0.9662, + "step": 12250 + }, + { + "epoch": 0.29, + "grad_norm": 1.8217058434842253, + "learning_rate": 1.6692553479336522e-05, + "loss": 1.0219, + "step": 12251 + }, + { + "epoch": 0.29, + "grad_norm": 2.2116533663679903, + "learning_rate": 1.6691986495555955e-05, + "loss": 1.0417, + "step": 12252 + }, + { + "epoch": 0.29, + "grad_norm": 2.1565558415960524, + "learning_rate": 1.6691419472812523e-05, + "loss": 1.0914, + "step": 12253 + }, + { + "epoch": 0.29, + "grad_norm": 2.076271408614077, + "learning_rate": 1.669085241110953e-05, + "loss": 0.9301, + "step": 12254 + }, + { + "epoch": 0.29, + "grad_norm": 2.0036040386919765, + "learning_rate": 1.669028531045027e-05, + "loss": 0.945, + "step": 12255 + }, + { + "epoch": 0.29, + "grad_norm": 1.961460279349369, + "learning_rate": 1.6689718170838058e-05, + "loss": 1.1493, + "step": 12256 + }, + { + "epoch": 0.29, + "grad_norm": 2.0430395123660996, + "learning_rate": 1.6689150992276183e-05, + "loss": 1.1066, + "step": 12257 + }, + { + "epoch": 0.29, + "grad_norm": 1.9894728805331472, + "learning_rate": 1.6688583774767954e-05, + "loss": 1.0032, + "step": 12258 + }, + { + "epoch": 0.29, + "grad_norm": 2.037045450958604, + "learning_rate": 1.6688016518316676e-05, + "loss": 1.1061, + "step": 12259 + }, + { + "epoch": 0.29, + "grad_norm": 2.06898655075173, + "learning_rate": 1.6687449222925644e-05, + "loss": 1.002, + "step": 12260 + }, + { + "epoch": 0.29, + "grad_norm": 2.3504582959002347, + "learning_rate": 1.668688188859817e-05, + "loss": 1.0302, + "step": 12261 + }, + { + "epoch": 0.29, + "grad_norm": 1.1138856510369206, + "learning_rate": 1.668631451533755e-05, + "loss": 1.0016, + "step": 12262 + }, + { + "epoch": 0.29, + "grad_norm": 2.269769945227365, + "learning_rate": 1.668574710314709e-05, + "loss": 0.9896, + "step": 12263 + }, + { + "epoch": 0.29, + "grad_norm": 1.1247130434064858, + "learning_rate": 1.6685179652030096e-05, + "loss": 1.0003, + "step": 12264 + }, + { + "epoch": 0.29, + "grad_norm": 1.075757344055783, + "learning_rate": 1.6684612161989866e-05, + "loss": 0.9916, + "step": 12265 + }, + { + "epoch": 0.29, + "grad_norm": 1.9384768640964765, + "learning_rate": 1.668404463302971e-05, + "loss": 0.9962, + "step": 12266 + }, + { + "epoch": 0.29, + "grad_norm": 1.0690535909625565, + "learning_rate": 1.6683477065152928e-05, + "loss": 1.0317, + "step": 12267 + }, + { + "epoch": 0.29, + "grad_norm": 2.492519298278362, + "learning_rate": 1.668290945836283e-05, + "loss": 1.2034, + "step": 12268 + }, + { + "epoch": 0.29, + "grad_norm": 2.1975581040165912, + "learning_rate": 1.6682341812662715e-05, + "loss": 1.1029, + "step": 12269 + }, + { + "epoch": 0.29, + "grad_norm": 2.233620546617761, + "learning_rate": 1.6681774128055893e-05, + "loss": 0.9912, + "step": 12270 + }, + { + "epoch": 0.29, + "grad_norm": 1.8911758331740882, + "learning_rate": 1.6681206404545665e-05, + "loss": 1.0963, + "step": 12271 + }, + { + "epoch": 0.29, + "grad_norm": 1.9492655721709387, + "learning_rate": 1.6680638642135337e-05, + "loss": 1.1312, + "step": 12272 + }, + { + "epoch": 0.29, + "grad_norm": 2.341615571431869, + "learning_rate": 1.6680070840828215e-05, + "loss": 1.0509, + "step": 12273 + }, + { + "epoch": 0.29, + "grad_norm": 1.6622648694508244, + "learning_rate": 1.6679503000627607e-05, + "loss": 0.9838, + "step": 12274 + }, + { + "epoch": 0.29, + "grad_norm": 1.9121142860145937, + "learning_rate": 1.6678935121536817e-05, + "loss": 0.9757, + "step": 12275 + }, + { + "epoch": 0.29, + "grad_norm": 2.2945987573371966, + "learning_rate": 1.6678367203559154e-05, + "loss": 1.003, + "step": 12276 + }, + { + "epoch": 0.29, + "grad_norm": 2.3951403904283772, + "learning_rate": 1.6677799246697924e-05, + "loss": 1.0433, + "step": 12277 + }, + { + "epoch": 0.29, + "grad_norm": 2.0548254017097367, + "learning_rate": 1.667723125095643e-05, + "loss": 1.1227, + "step": 12278 + }, + { + "epoch": 0.29, + "grad_norm": 2.2316944252462054, + "learning_rate": 1.667666321633798e-05, + "loss": 1.0066, + "step": 12279 + }, + { + "epoch": 0.29, + "grad_norm": 1.8385782426311996, + "learning_rate": 1.6676095142845883e-05, + "loss": 0.9967, + "step": 12280 + }, + { + "epoch": 0.29, + "grad_norm": 2.170147773643691, + "learning_rate": 1.6675527030483452e-05, + "loss": 1.1476, + "step": 12281 + }, + { + "epoch": 0.29, + "grad_norm": 1.9768932709162772, + "learning_rate": 1.6674958879253986e-05, + "loss": 1.0758, + "step": 12282 + }, + { + "epoch": 0.29, + "grad_norm": 2.028761609769406, + "learning_rate": 1.6674390689160795e-05, + "loss": 1.1449, + "step": 12283 + }, + { + "epoch": 0.29, + "grad_norm": 2.084587245256921, + "learning_rate": 1.667382246020719e-05, + "loss": 1.0552, + "step": 12284 + }, + { + "epoch": 0.29, + "grad_norm": 2.1010508111566946, + "learning_rate": 1.667325419239648e-05, + "loss": 1.058, + "step": 12285 + }, + { + "epoch": 0.29, + "grad_norm": 2.1074852358288103, + "learning_rate": 1.6672685885731964e-05, + "loss": 1.0476, + "step": 12286 + }, + { + "epoch": 0.29, + "grad_norm": 2.102156491119303, + "learning_rate": 1.6672117540216962e-05, + "loss": 0.9645, + "step": 12287 + }, + { + "epoch": 0.29, + "grad_norm": 2.169158183866925, + "learning_rate": 1.6671549155854783e-05, + "loss": 1.046, + "step": 12288 + }, + { + "epoch": 0.29, + "grad_norm": 2.04848597554303, + "learning_rate": 1.6670980732648726e-05, + "loss": 0.9222, + "step": 12289 + }, + { + "epoch": 0.29, + "grad_norm": 2.4543232535877366, + "learning_rate": 1.6670412270602115e-05, + "loss": 1.1811, + "step": 12290 + }, + { + "epoch": 0.29, + "grad_norm": 2.4131831603416285, + "learning_rate": 1.6669843769718245e-05, + "loss": 1.1736, + "step": 12291 + }, + { + "epoch": 0.29, + "grad_norm": 2.532497765523263, + "learning_rate": 1.6669275230000438e-05, + "loss": 0.9702, + "step": 12292 + }, + { + "epoch": 0.29, + "grad_norm": 2.0614712034047473, + "learning_rate": 1.6668706651451994e-05, + "loss": 0.953, + "step": 12293 + }, + { + "epoch": 0.29, + "grad_norm": 2.6274192949494317, + "learning_rate": 1.6668138034076235e-05, + "loss": 1.0809, + "step": 12294 + }, + { + "epoch": 0.29, + "grad_norm": 2.3557881564556937, + "learning_rate": 1.666756937787646e-05, + "loss": 0.9723, + "step": 12295 + }, + { + "epoch": 0.29, + "grad_norm": 1.9865443038543076, + "learning_rate": 1.6667000682855988e-05, + "loss": 1.0779, + "step": 12296 + }, + { + "epoch": 0.29, + "grad_norm": 2.033525199727177, + "learning_rate": 1.666643194901812e-05, + "loss": 1.068, + "step": 12297 + }, + { + "epoch": 0.29, + "grad_norm": 1.987240083801914, + "learning_rate": 1.6665863176366186e-05, + "loss": 1.0908, + "step": 12298 + }, + { + "epoch": 0.29, + "grad_norm": 1.9741857106731648, + "learning_rate": 1.6665294364903482e-05, + "loss": 1.0161, + "step": 12299 + }, + { + "epoch": 0.29, + "grad_norm": 2.937240624009778, + "learning_rate": 1.666472551463332e-05, + "loss": 1.0247, + "step": 12300 + }, + { + "epoch": 0.29, + "grad_norm": 2.0584778192077158, + "learning_rate": 1.666415662555902e-05, + "loss": 1.1091, + "step": 12301 + }, + { + "epoch": 0.29, + "grad_norm": 1.916682370253148, + "learning_rate": 1.666358769768389e-05, + "loss": 1.0011, + "step": 12302 + }, + { + "epoch": 0.29, + "grad_norm": 2.0757389376424698, + "learning_rate": 1.666301873101124e-05, + "loss": 1.0212, + "step": 12303 + }, + { + "epoch": 0.29, + "grad_norm": 2.181998110880447, + "learning_rate": 1.6662449725544386e-05, + "loss": 0.9326, + "step": 12304 + }, + { + "epoch": 0.29, + "grad_norm": 2.375140916591057, + "learning_rate": 1.6661880681286647e-05, + "loss": 0.9082, + "step": 12305 + }, + { + "epoch": 0.29, + "grad_norm": 2.4294705562393717, + "learning_rate": 1.6661311598241325e-05, + "loss": 1.0323, + "step": 12306 + }, + { + "epoch": 0.29, + "grad_norm": 2.1489101197847167, + "learning_rate": 1.6660742476411736e-05, + "loss": 1.091, + "step": 12307 + }, + { + "epoch": 0.29, + "grad_norm": 2.083684357361555, + "learning_rate": 1.6660173315801194e-05, + "loss": 0.9687, + "step": 12308 + }, + { + "epoch": 0.29, + "grad_norm": 2.49891161337701, + "learning_rate": 1.6659604116413016e-05, + "loss": 1.1019, + "step": 12309 + }, + { + "epoch": 0.29, + "grad_norm": 1.8896309621382357, + "learning_rate": 1.665903487825052e-05, + "loss": 1.0555, + "step": 12310 + }, + { + "epoch": 0.29, + "grad_norm": 1.9991933980296213, + "learning_rate": 1.6658465601317008e-05, + "loss": 0.9711, + "step": 12311 + }, + { + "epoch": 0.29, + "grad_norm": 2.2897401828163257, + "learning_rate": 1.6657896285615804e-05, + "loss": 1.1429, + "step": 12312 + }, + { + "epoch": 0.29, + "grad_norm": 2.209527325243641, + "learning_rate": 1.6657326931150222e-05, + "loss": 1.0341, + "step": 12313 + }, + { + "epoch": 0.29, + "grad_norm": 1.9854052014406476, + "learning_rate": 1.6656757537923572e-05, + "loss": 1.1118, + "step": 12314 + }, + { + "epoch": 0.29, + "grad_norm": 1.942246740848264, + "learning_rate": 1.6656188105939172e-05, + "loss": 1.0764, + "step": 12315 + }, + { + "epoch": 0.29, + "grad_norm": 1.962192759008441, + "learning_rate": 1.665561863520034e-05, + "loss": 1.0661, + "step": 12316 + }, + { + "epoch": 0.29, + "grad_norm": 2.0329672293405925, + "learning_rate": 1.665504912571039e-05, + "loss": 1.087, + "step": 12317 + }, + { + "epoch": 0.29, + "grad_norm": 1.888205641143555, + "learning_rate": 1.665447957747263e-05, + "loss": 1.0877, + "step": 12318 + }, + { + "epoch": 0.29, + "grad_norm": 2.1395854630554663, + "learning_rate": 1.665390999049039e-05, + "loss": 1.1065, + "step": 12319 + }, + { + "epoch": 0.29, + "grad_norm": 2.106005843751289, + "learning_rate": 1.6653340364766978e-05, + "loss": 1.0543, + "step": 12320 + }, + { + "epoch": 0.29, + "grad_norm": 1.8166935152631665, + "learning_rate": 1.6652770700305708e-05, + "loss": 1.1046, + "step": 12321 + }, + { + "epoch": 0.29, + "grad_norm": 1.847363374613488, + "learning_rate": 1.6652200997109906e-05, + "loss": 1.0108, + "step": 12322 + }, + { + "epoch": 0.29, + "grad_norm": 2.051598285204428, + "learning_rate": 1.665163125518288e-05, + "loss": 1.1561, + "step": 12323 + }, + { + "epoch": 0.29, + "grad_norm": 2.158121207568374, + "learning_rate": 1.6651061474527953e-05, + "loss": 1.0895, + "step": 12324 + }, + { + "epoch": 0.29, + "grad_norm": 1.9036090964727932, + "learning_rate": 1.665049165514844e-05, + "loss": 1.0416, + "step": 12325 + }, + { + "epoch": 0.29, + "grad_norm": 2.4964213834384634, + "learning_rate": 1.6649921797047658e-05, + "loss": 1.1296, + "step": 12326 + }, + { + "epoch": 0.29, + "grad_norm": 2.0518286009437356, + "learning_rate": 1.664935190022893e-05, + "loss": 1.0768, + "step": 12327 + }, + { + "epoch": 0.29, + "grad_norm": 1.877955745242482, + "learning_rate": 1.6648781964695565e-05, + "loss": 1.0275, + "step": 12328 + }, + { + "epoch": 0.29, + "grad_norm": 2.099899514023473, + "learning_rate": 1.664821199045089e-05, + "loss": 1.037, + "step": 12329 + }, + { + "epoch": 0.29, + "grad_norm": 1.9969928760252038, + "learning_rate": 1.6647641977498215e-05, + "loss": 1.0152, + "step": 12330 + }, + { + "epoch": 0.29, + "grad_norm": 2.018039575472721, + "learning_rate": 1.6647071925840868e-05, + "loss": 1.0028, + "step": 12331 + }, + { + "epoch": 0.29, + "grad_norm": 1.12379086946574, + "learning_rate": 1.6646501835482162e-05, + "loss": 1.0521, + "step": 12332 + }, + { + "epoch": 0.29, + "grad_norm": 1.7976463394739426, + "learning_rate": 1.664593170642542e-05, + "loss": 0.9878, + "step": 12333 + }, + { + "epoch": 0.29, + "grad_norm": 2.0022319889069493, + "learning_rate": 1.6645361538673962e-05, + "loss": 0.9571, + "step": 12334 + }, + { + "epoch": 0.29, + "grad_norm": 1.8102084960382516, + "learning_rate": 1.66447913322311e-05, + "loss": 1.1997, + "step": 12335 + }, + { + "epoch": 0.29, + "grad_norm": 2.1697570134489124, + "learning_rate": 1.6644221087100164e-05, + "loss": 1.0671, + "step": 12336 + }, + { + "epoch": 0.29, + "grad_norm": 1.1111451307976135, + "learning_rate": 1.6643650803284462e-05, + "loss": 1.0104, + "step": 12337 + }, + { + "epoch": 0.29, + "grad_norm": 2.2133447687380667, + "learning_rate": 1.6643080480787328e-05, + "loss": 1.0105, + "step": 12338 + }, + { + "epoch": 0.29, + "grad_norm": 1.7971200306862596, + "learning_rate": 1.6642510119612073e-05, + "loss": 0.9012, + "step": 12339 + }, + { + "epoch": 0.29, + "grad_norm": 2.061091954669361, + "learning_rate": 1.6641939719762023e-05, + "loss": 0.9387, + "step": 12340 + }, + { + "epoch": 0.29, + "grad_norm": 2.1765973886423935, + "learning_rate": 1.6641369281240497e-05, + "loss": 1.1197, + "step": 12341 + }, + { + "epoch": 0.29, + "grad_norm": 1.1264179493428104, + "learning_rate": 1.6640798804050816e-05, + "loss": 0.9938, + "step": 12342 + }, + { + "epoch": 0.29, + "grad_norm": 2.003389989672408, + "learning_rate": 1.6640228288196303e-05, + "loss": 1.0572, + "step": 12343 + }, + { + "epoch": 0.29, + "grad_norm": 2.0952345721139, + "learning_rate": 1.6639657733680276e-05, + "loss": 0.9984, + "step": 12344 + }, + { + "epoch": 0.29, + "grad_norm": 1.8741748671852418, + "learning_rate": 1.663908714050606e-05, + "loss": 1.0349, + "step": 12345 + }, + { + "epoch": 0.29, + "grad_norm": 2.0310432038815716, + "learning_rate": 1.6638516508676977e-05, + "loss": 1.1915, + "step": 12346 + }, + { + "epoch": 0.29, + "grad_norm": 1.9868994893739824, + "learning_rate": 1.6637945838196356e-05, + "loss": 1.0212, + "step": 12347 + }, + { + "epoch": 0.29, + "grad_norm": 1.1549896772866963, + "learning_rate": 1.6637375129067502e-05, + "loss": 0.9945, + "step": 12348 + }, + { + "epoch": 0.29, + "grad_norm": 2.132820110276901, + "learning_rate": 1.6636804381293755e-05, + "loss": 1.0474, + "step": 12349 + }, + { + "epoch": 0.29, + "grad_norm": 2.3075239643561427, + "learning_rate": 1.6636233594878434e-05, + "loss": 1.1019, + "step": 12350 + }, + { + "epoch": 0.29, + "grad_norm": 2.2739926523521325, + "learning_rate": 1.6635662769824854e-05, + "loss": 1.0833, + "step": 12351 + }, + { + "epoch": 0.29, + "grad_norm": 2.289346517634783, + "learning_rate": 1.6635091906136346e-05, + "loss": 1.1393, + "step": 12352 + }, + { + "epoch": 0.29, + "grad_norm": 1.938539525482054, + "learning_rate": 1.6634521003816235e-05, + "loss": 1.0647, + "step": 12353 + }, + { + "epoch": 0.29, + "grad_norm": 1.1702154860613738, + "learning_rate": 1.663395006286784e-05, + "loss": 1.0479, + "step": 12354 + }, + { + "epoch": 0.29, + "grad_norm": 2.138356734163179, + "learning_rate": 1.6633379083294487e-05, + "loss": 1.0583, + "step": 12355 + }, + { + "epoch": 0.29, + "grad_norm": 1.1100211601811873, + "learning_rate": 1.6632808065099503e-05, + "loss": 1.0129, + "step": 12356 + }, + { + "epoch": 0.29, + "grad_norm": 1.9216929811471863, + "learning_rate": 1.663223700828621e-05, + "loss": 1.0219, + "step": 12357 + }, + { + "epoch": 0.29, + "grad_norm": 1.9626345455557728, + "learning_rate": 1.663166591285793e-05, + "loss": 1.039, + "step": 12358 + }, + { + "epoch": 0.29, + "grad_norm": 2.7969236562801454, + "learning_rate": 1.6631094778817997e-05, + "loss": 0.9123, + "step": 12359 + }, + { + "epoch": 0.29, + "grad_norm": 2.093886747938712, + "learning_rate": 1.6630523606169726e-05, + "loss": 1.0866, + "step": 12360 + }, + { + "epoch": 0.29, + "grad_norm": 1.7538478680045133, + "learning_rate": 1.6629952394916447e-05, + "loss": 1.0451, + "step": 12361 + }, + { + "epoch": 0.29, + "grad_norm": 1.9987868791946362, + "learning_rate": 1.6629381145061492e-05, + "loss": 1.1542, + "step": 12362 + }, + { + "epoch": 0.29, + "grad_norm": 2.05525029202118, + "learning_rate": 1.6628809856608177e-05, + "loss": 1.0475, + "step": 12363 + }, + { + "epoch": 0.29, + "grad_norm": 1.9207118154127054, + "learning_rate": 1.6628238529559834e-05, + "loss": 1.1175, + "step": 12364 + }, + { + "epoch": 0.29, + "grad_norm": 2.7466606951998678, + "learning_rate": 1.6627667163919788e-05, + "loss": 1.0486, + "step": 12365 + }, + { + "epoch": 0.29, + "grad_norm": 2.380264068447194, + "learning_rate": 1.6627095759691364e-05, + "loss": 1.1094, + "step": 12366 + }, + { + "epoch": 0.29, + "grad_norm": 1.9526685021502097, + "learning_rate": 1.662652431687789e-05, + "loss": 1.112, + "step": 12367 + }, + { + "epoch": 0.29, + "grad_norm": 1.8360973937837854, + "learning_rate": 1.6625952835482696e-05, + "loss": 1.1412, + "step": 12368 + }, + { + "epoch": 0.29, + "grad_norm": 2.0012784824823595, + "learning_rate": 1.6625381315509103e-05, + "loss": 1.0255, + "step": 12369 + }, + { + "epoch": 0.29, + "grad_norm": 2.4051183703041503, + "learning_rate": 1.6624809756960445e-05, + "loss": 1.0574, + "step": 12370 + }, + { + "epoch": 0.29, + "grad_norm": 2.330225815433078, + "learning_rate": 1.6624238159840043e-05, + "loss": 1.085, + "step": 12371 + }, + { + "epoch": 0.29, + "grad_norm": 1.9438655093536727, + "learning_rate": 1.6623666524151233e-05, + "loss": 1.1455, + "step": 12372 + }, + { + "epoch": 0.29, + "grad_norm": 2.048212601570863, + "learning_rate": 1.6623094849897342e-05, + "loss": 1.0221, + "step": 12373 + }, + { + "epoch": 0.29, + "grad_norm": 2.050251384471474, + "learning_rate": 1.6622523137081693e-05, + "loss": 0.99, + "step": 12374 + }, + { + "epoch": 0.29, + "grad_norm": 2.0297967274042135, + "learning_rate": 1.6621951385707617e-05, + "loss": 1.0857, + "step": 12375 + }, + { + "epoch": 0.29, + "grad_norm": 1.1184937742698395, + "learning_rate": 1.6621379595778444e-05, + "loss": 0.9931, + "step": 12376 + }, + { + "epoch": 0.29, + "grad_norm": 1.7960108713526595, + "learning_rate": 1.6620807767297505e-05, + "loss": 1.0906, + "step": 12377 + }, + { + "epoch": 0.29, + "grad_norm": 1.8768461056237604, + "learning_rate": 1.6620235900268127e-05, + "loss": 1.1207, + "step": 12378 + }, + { + "epoch": 0.29, + "grad_norm": 2.0077959538447803, + "learning_rate": 1.6619663994693634e-05, + "loss": 0.9719, + "step": 12379 + }, + { + "epoch": 0.29, + "grad_norm": 2.158591441458884, + "learning_rate": 1.6619092050577367e-05, + "loss": 0.9773, + "step": 12380 + }, + { + "epoch": 0.29, + "grad_norm": 1.8401412360167229, + "learning_rate": 1.661852006792265e-05, + "loss": 1.0465, + "step": 12381 + }, + { + "epoch": 0.29, + "grad_norm": 2.245378322066238, + "learning_rate": 1.6617948046732814e-05, + "loss": 1.0631, + "step": 12382 + }, + { + "epoch": 0.29, + "grad_norm": 2.7735191103184085, + "learning_rate": 1.6617375987011186e-05, + "loss": 1.044, + "step": 12383 + }, + { + "epoch": 0.29, + "grad_norm": 1.7860222087301951, + "learning_rate": 1.6616803888761105e-05, + "loss": 1.0219, + "step": 12384 + }, + { + "epoch": 0.29, + "grad_norm": 1.9389156446624911, + "learning_rate": 1.6616231751985893e-05, + "loss": 1.212, + "step": 12385 + }, + { + "epoch": 0.29, + "grad_norm": 1.9792644244314364, + "learning_rate": 1.6615659576688886e-05, + "loss": 0.9876, + "step": 12386 + }, + { + "epoch": 0.29, + "grad_norm": 2.7565321556459264, + "learning_rate": 1.6615087362873415e-05, + "loss": 1.1145, + "step": 12387 + }, + { + "epoch": 0.29, + "grad_norm": 1.162196613291038, + "learning_rate": 1.6614515110542807e-05, + "loss": 1.0076, + "step": 12388 + }, + { + "epoch": 0.29, + "grad_norm": 1.8281410418574835, + "learning_rate": 1.6613942819700402e-05, + "loss": 0.9812, + "step": 12389 + }, + { + "epoch": 0.29, + "grad_norm": 2.0597733081159424, + "learning_rate": 1.661337049034953e-05, + "loss": 0.9923, + "step": 12390 + }, + { + "epoch": 0.29, + "grad_norm": 1.939133660770905, + "learning_rate": 1.661279812249352e-05, + "loss": 0.9961, + "step": 12391 + }, + { + "epoch": 0.29, + "grad_norm": 2.314718444118976, + "learning_rate": 1.6612225716135703e-05, + "loss": 0.8706, + "step": 12392 + }, + { + "epoch": 0.29, + "grad_norm": 2.2552040287634547, + "learning_rate": 1.6611653271279415e-05, + "loss": 1.0799, + "step": 12393 + }, + { + "epoch": 0.29, + "grad_norm": 2.1118857231230224, + "learning_rate": 1.661108078792799e-05, + "loss": 1.0272, + "step": 12394 + }, + { + "epoch": 0.29, + "grad_norm": 2.0524987864574227, + "learning_rate": 1.661050826608476e-05, + "loss": 0.9832, + "step": 12395 + }, + { + "epoch": 0.29, + "grad_norm": 2.0577797823204396, + "learning_rate": 1.6609935705753057e-05, + "loss": 1.0909, + "step": 12396 + }, + { + "epoch": 0.29, + "grad_norm": 2.260632629705963, + "learning_rate": 1.6609363106936215e-05, + "loss": 1.0033, + "step": 12397 + }, + { + "epoch": 0.29, + "grad_norm": 1.9860779731110878, + "learning_rate": 1.660879046963757e-05, + "loss": 1.0022, + "step": 12398 + }, + { + "epoch": 0.29, + "grad_norm": 2.446072086355108, + "learning_rate": 1.6608217793860455e-05, + "loss": 1.105, + "step": 12399 + }, + { + "epoch": 0.29, + "grad_norm": 1.9985049165301476, + "learning_rate": 1.66076450796082e-05, + "loss": 0.8683, + "step": 12400 + }, + { + "epoch": 0.29, + "grad_norm": 1.8827835280862415, + "learning_rate": 1.660707232688415e-05, + "loss": 1.1076, + "step": 12401 + }, + { + "epoch": 0.29, + "grad_norm": 2.1340249716810873, + "learning_rate": 1.6606499535691623e-05, + "loss": 1.0008, + "step": 12402 + }, + { + "epoch": 0.29, + "grad_norm": 1.9697293664281554, + "learning_rate": 1.6605926706033973e-05, + "loss": 1.02, + "step": 12403 + }, + { + "epoch": 0.29, + "grad_norm": 1.2079783611824393, + "learning_rate": 1.6605353837914522e-05, + "loss": 1.0397, + "step": 12404 + }, + { + "epoch": 0.29, + "grad_norm": 2.2010861551878547, + "learning_rate": 1.660478093133661e-05, + "loss": 1.0736, + "step": 12405 + }, + { + "epoch": 0.29, + "grad_norm": 1.917148983995198, + "learning_rate": 1.6604207986303577e-05, + "loss": 0.9911, + "step": 12406 + }, + { + "epoch": 0.29, + "grad_norm": 2.2521767128934442, + "learning_rate": 1.660363500281875e-05, + "loss": 1.0767, + "step": 12407 + }, + { + "epoch": 0.29, + "grad_norm": 2.2163234044370355, + "learning_rate": 1.660306198088547e-05, + "loss": 1.0982, + "step": 12408 + }, + { + "epoch": 0.29, + "grad_norm": 1.872325472886967, + "learning_rate": 1.6602488920507075e-05, + "loss": 1.0946, + "step": 12409 + }, + { + "epoch": 0.29, + "grad_norm": 2.348161933262913, + "learning_rate": 1.6601915821686895e-05, + "loss": 0.944, + "step": 12410 + }, + { + "epoch": 0.29, + "grad_norm": 1.9492751555437553, + "learning_rate": 1.6601342684428274e-05, + "loss": 1.1241, + "step": 12411 + }, + { + "epoch": 0.29, + "grad_norm": 2.0161781928246003, + "learning_rate": 1.6600769508734545e-05, + "loss": 1.0458, + "step": 12412 + }, + { + "epoch": 0.29, + "grad_norm": 2.910896511062526, + "learning_rate": 1.6600196294609047e-05, + "loss": 1.1219, + "step": 12413 + }, + { + "epoch": 0.29, + "grad_norm": 1.9093537515307235, + "learning_rate": 1.6599623042055114e-05, + "loss": 1.0481, + "step": 12414 + }, + { + "epoch": 0.29, + "grad_norm": 1.9441744028012782, + "learning_rate": 1.659904975107609e-05, + "loss": 1.0827, + "step": 12415 + }, + { + "epoch": 0.29, + "grad_norm": 1.9671287550787384, + "learning_rate": 1.6598476421675303e-05, + "loss": 1.0116, + "step": 12416 + }, + { + "epoch": 0.29, + "grad_norm": 2.102671444761018, + "learning_rate": 1.6597903053856104e-05, + "loss": 1.1048, + "step": 12417 + }, + { + "epoch": 0.29, + "grad_norm": 2.0592032499321204, + "learning_rate": 1.6597329647621822e-05, + "loss": 1.0379, + "step": 12418 + }, + { + "epoch": 0.29, + "grad_norm": 1.975371703431985, + "learning_rate": 1.65967562029758e-05, + "loss": 1.2465, + "step": 12419 + }, + { + "epoch": 0.29, + "grad_norm": 2.0273058555622594, + "learning_rate": 1.659618271992137e-05, + "loss": 1.1196, + "step": 12420 + }, + { + "epoch": 0.29, + "grad_norm": 2.0654435637528734, + "learning_rate": 1.6595609198461878e-05, + "loss": 1.1795, + "step": 12421 + }, + { + "epoch": 0.29, + "grad_norm": 1.1544900874059243, + "learning_rate": 1.6595035638600662e-05, + "loss": 0.9564, + "step": 12422 + }, + { + "epoch": 0.29, + "grad_norm": 3.3659626235875173, + "learning_rate": 1.659446204034106e-05, + "loss": 0.9834, + "step": 12423 + }, + { + "epoch": 0.29, + "grad_norm": 2.1661338357899096, + "learning_rate": 1.6593888403686417e-05, + "loss": 0.9031, + "step": 12424 + }, + { + "epoch": 0.29, + "grad_norm": 1.9344697063302616, + "learning_rate": 1.6593314728640063e-05, + "loss": 1.1076, + "step": 12425 + }, + { + "epoch": 0.29, + "grad_norm": 2.0348529978143217, + "learning_rate": 1.6592741015205343e-05, + "loss": 1.0202, + "step": 12426 + }, + { + "epoch": 0.29, + "grad_norm": 2.540171547543471, + "learning_rate": 1.65921672633856e-05, + "loss": 1.0615, + "step": 12427 + }, + { + "epoch": 0.29, + "grad_norm": 2.667834616803764, + "learning_rate": 1.659159347318417e-05, + "loss": 0.9637, + "step": 12428 + }, + { + "epoch": 0.29, + "grad_norm": 2.078992197838872, + "learning_rate": 1.65910196446044e-05, + "loss": 1.0773, + "step": 12429 + }, + { + "epoch": 0.29, + "grad_norm": 1.8340252307042075, + "learning_rate": 1.659044577764962e-05, + "loss": 0.965, + "step": 12430 + }, + { + "epoch": 0.29, + "grad_norm": 1.8557328656093528, + "learning_rate": 1.6589871872323185e-05, + "loss": 0.9471, + "step": 12431 + }, + { + "epoch": 0.29, + "grad_norm": 1.9591943456219136, + "learning_rate": 1.6589297928628423e-05, + "loss": 1.1481, + "step": 12432 + }, + { + "epoch": 0.29, + "grad_norm": 2.050309197615306, + "learning_rate": 1.658872394656869e-05, + "loss": 1.142, + "step": 12433 + }, + { + "epoch": 0.29, + "grad_norm": 2.350924581200395, + "learning_rate": 1.658814992614731e-05, + "loss": 0.9425, + "step": 12434 + }, + { + "epoch": 0.29, + "grad_norm": 2.4466223038620445, + "learning_rate": 1.6587575867367645e-05, + "loss": 0.9942, + "step": 12435 + }, + { + "epoch": 0.29, + "grad_norm": 2.0269899545948076, + "learning_rate": 1.6587001770233024e-05, + "loss": 1.0937, + "step": 12436 + }, + { + "epoch": 0.29, + "grad_norm": 2.4901041875729657, + "learning_rate": 1.6586427634746794e-05, + "loss": 1.0158, + "step": 12437 + }, + { + "epoch": 0.29, + "grad_norm": 2.1786506105660406, + "learning_rate": 1.6585853460912295e-05, + "loss": 0.9312, + "step": 12438 + }, + { + "epoch": 0.29, + "grad_norm": 1.9910170580930082, + "learning_rate": 1.6585279248732873e-05, + "loss": 1.0118, + "step": 12439 + }, + { + "epoch": 0.29, + "grad_norm": 2.006356240693047, + "learning_rate": 1.658470499821187e-05, + "loss": 1.0136, + "step": 12440 + }, + { + "epoch": 0.29, + "grad_norm": 2.430345932290753, + "learning_rate": 1.658413070935263e-05, + "loss": 0.9772, + "step": 12441 + }, + { + "epoch": 0.29, + "grad_norm": 1.1612414645283216, + "learning_rate": 1.65835563821585e-05, + "loss": 0.9559, + "step": 12442 + }, + { + "epoch": 0.29, + "grad_norm": 2.0218324904002576, + "learning_rate": 1.6582982016632818e-05, + "loss": 1.0993, + "step": 12443 + }, + { + "epoch": 0.29, + "grad_norm": 2.1713417041908665, + "learning_rate": 1.658240761277893e-05, + "loss": 1.1303, + "step": 12444 + }, + { + "epoch": 0.29, + "grad_norm": 2.072759249178197, + "learning_rate": 1.6581833170600182e-05, + "loss": 1.192, + "step": 12445 + }, + { + "epoch": 0.29, + "grad_norm": 2.0571017360862336, + "learning_rate": 1.6581258690099916e-05, + "loss": 1.0549, + "step": 12446 + }, + { + "epoch": 0.29, + "grad_norm": 2.0913461472058485, + "learning_rate": 1.6580684171281478e-05, + "loss": 0.9229, + "step": 12447 + }, + { + "epoch": 0.29, + "grad_norm": 2.1999008661702835, + "learning_rate": 1.6580109614148213e-05, + "loss": 1.0791, + "step": 12448 + }, + { + "epoch": 0.29, + "grad_norm": 1.7940368351253448, + "learning_rate": 1.6579535018703472e-05, + "loss": 1.0412, + "step": 12449 + }, + { + "epoch": 0.29, + "grad_norm": 1.8437218068506083, + "learning_rate": 1.6578960384950588e-05, + "loss": 1.1028, + "step": 12450 + }, + { + "epoch": 0.29, + "grad_norm": 3.1079671406522396, + "learning_rate": 1.657838571289292e-05, + "loss": 0.9987, + "step": 12451 + }, + { + "epoch": 0.29, + "grad_norm": 1.8447274711156316, + "learning_rate": 1.6577811002533803e-05, + "loss": 1.1009, + "step": 12452 + }, + { + "epoch": 0.29, + "grad_norm": 2.066870890389123, + "learning_rate": 1.657723625387659e-05, + "loss": 1.0191, + "step": 12453 + }, + { + "epoch": 0.29, + "grad_norm": 2.1156816770202025, + "learning_rate": 1.6576661466924625e-05, + "loss": 1.1008, + "step": 12454 + }, + { + "epoch": 0.29, + "grad_norm": 2.0100328106916128, + "learning_rate": 1.657608664168125e-05, + "loss": 1.136, + "step": 12455 + }, + { + "epoch": 0.29, + "grad_norm": 2.0671602459546845, + "learning_rate": 1.6575511778149823e-05, + "loss": 1.0403, + "step": 12456 + }, + { + "epoch": 0.29, + "grad_norm": 1.1028386567204818, + "learning_rate": 1.657493687633368e-05, + "loss": 0.9895, + "step": 12457 + }, + { + "epoch": 0.29, + "grad_norm": 1.8460256972071043, + "learning_rate": 1.6574361936236177e-05, + "loss": 1.1783, + "step": 12458 + }, + { + "epoch": 0.29, + "grad_norm": 1.8907743377266026, + "learning_rate": 1.6573786957860658e-05, + "loss": 1.0166, + "step": 12459 + }, + { + "epoch": 0.29, + "grad_norm": 2.0349367899962676, + "learning_rate": 1.6573211941210467e-05, + "loss": 1.1409, + "step": 12460 + }, + { + "epoch": 0.29, + "grad_norm": 1.8221020806591852, + "learning_rate": 1.6572636886288957e-05, + "loss": 1.0098, + "step": 12461 + }, + { + "epoch": 0.29, + "grad_norm": 2.8521321442970753, + "learning_rate": 1.6572061793099473e-05, + "loss": 1.0735, + "step": 12462 + }, + { + "epoch": 0.29, + "grad_norm": 2.0789419173822568, + "learning_rate": 1.6571486661645367e-05, + "loss": 1.0851, + "step": 12463 + }, + { + "epoch": 0.29, + "grad_norm": 1.1239474690978464, + "learning_rate": 1.6570911491929984e-05, + "loss": 1.0165, + "step": 12464 + }, + { + "epoch": 0.29, + "grad_norm": 2.4077480049302036, + "learning_rate": 1.657033628395667e-05, + "loss": 1.1938, + "step": 12465 + }, + { + "epoch": 0.29, + "grad_norm": 1.9494912381341563, + "learning_rate": 1.656976103772878e-05, + "loss": 1.088, + "step": 12466 + }, + { + "epoch": 0.29, + "grad_norm": 2.676233938568939, + "learning_rate": 1.6569185753249662e-05, + "loss": 1.0125, + "step": 12467 + }, + { + "epoch": 0.29, + "grad_norm": 1.9417303089984763, + "learning_rate": 1.656861043052267e-05, + "loss": 0.9064, + "step": 12468 + }, + { + "epoch": 0.29, + "grad_norm": 2.105722373498766, + "learning_rate": 1.656803506955114e-05, + "loss": 0.9563, + "step": 12469 + }, + { + "epoch": 0.29, + "grad_norm": 2.2793659649807614, + "learning_rate": 1.6567459670338434e-05, + "loss": 1.0199, + "step": 12470 + }, + { + "epoch": 0.29, + "grad_norm": 2.0230067428164786, + "learning_rate": 1.65668842328879e-05, + "loss": 0.9451, + "step": 12471 + }, + { + "epoch": 0.29, + "grad_norm": 2.138097181303258, + "learning_rate": 1.6566308757202883e-05, + "loss": 0.9348, + "step": 12472 + }, + { + "epoch": 0.29, + "grad_norm": 2.1365457595849477, + "learning_rate": 1.656573324328674e-05, + "loss": 1.1433, + "step": 12473 + }, + { + "epoch": 0.29, + "grad_norm": 2.2642793974657445, + "learning_rate": 1.656515769114282e-05, + "loss": 0.9722, + "step": 12474 + }, + { + "epoch": 0.29, + "grad_norm": 1.0994735828939772, + "learning_rate": 1.6564582100774472e-05, + "loss": 0.9825, + "step": 12475 + }, + { + "epoch": 0.29, + "grad_norm": 2.3692938061470494, + "learning_rate": 1.656400647218505e-05, + "loss": 1.0581, + "step": 12476 + }, + { + "epoch": 0.29, + "grad_norm": 2.5773177868231176, + "learning_rate": 1.6563430805377903e-05, + "loss": 0.9557, + "step": 12477 + }, + { + "epoch": 0.29, + "grad_norm": 2.0730469332546386, + "learning_rate": 1.6562855100356386e-05, + "loss": 1.0142, + "step": 12478 + }, + { + "epoch": 0.29, + "grad_norm": 2.379912323782443, + "learning_rate": 1.656227935712385e-05, + "loss": 1.0273, + "step": 12479 + }, + { + "epoch": 0.29, + "grad_norm": 1.9664382345371298, + "learning_rate": 1.6561703575683645e-05, + "loss": 1.1158, + "step": 12480 + }, + { + "epoch": 0.29, + "grad_norm": 2.6808944102892065, + "learning_rate": 1.6561127756039122e-05, + "loss": 1.0394, + "step": 12481 + }, + { + "epoch": 0.29, + "grad_norm": 2.404482927065188, + "learning_rate": 1.6560551898193637e-05, + "loss": 0.9756, + "step": 12482 + }, + { + "epoch": 0.29, + "grad_norm": 2.0900906451500885, + "learning_rate": 1.655997600215054e-05, + "loss": 0.97, + "step": 12483 + }, + { + "epoch": 0.29, + "grad_norm": 2.268276959245407, + "learning_rate": 1.6559400067913187e-05, + "loss": 1.1033, + "step": 12484 + }, + { + "epoch": 0.29, + "grad_norm": 2.0869613460473384, + "learning_rate": 1.655882409548493e-05, + "loss": 1.0344, + "step": 12485 + }, + { + "epoch": 0.29, + "grad_norm": 2.1628454692554056, + "learning_rate": 1.6558248084869125e-05, + "loss": 1.1686, + "step": 12486 + }, + { + "epoch": 0.29, + "grad_norm": 1.9366712939964892, + "learning_rate": 1.6557672036069122e-05, + "loss": 1.1425, + "step": 12487 + }, + { + "epoch": 0.29, + "grad_norm": 2.0175138233731356, + "learning_rate": 1.6557095949088275e-05, + "loss": 1.019, + "step": 12488 + }, + { + "epoch": 0.29, + "grad_norm": 2.116147167050777, + "learning_rate": 1.655651982392994e-05, + "loss": 1.0662, + "step": 12489 + }, + { + "epoch": 0.29, + "grad_norm": 1.9824289648276896, + "learning_rate": 1.655594366059747e-05, + "loss": 1.0163, + "step": 12490 + }, + { + "epoch": 0.29, + "grad_norm": 2.0786507399044387, + "learning_rate": 1.6555367459094223e-05, + "loss": 1.1307, + "step": 12491 + }, + { + "epoch": 0.29, + "grad_norm": 2.0237224415741086, + "learning_rate": 1.6554791219423548e-05, + "loss": 1.0244, + "step": 12492 + }, + { + "epoch": 0.29, + "grad_norm": 2.030265949950143, + "learning_rate": 1.6554214941588807e-05, + "loss": 1.0341, + "step": 12493 + }, + { + "epoch": 0.29, + "grad_norm": 2.1831335292259326, + "learning_rate": 1.655363862559335e-05, + "loss": 0.9714, + "step": 12494 + }, + { + "epoch": 0.29, + "grad_norm": 2.632422379036428, + "learning_rate": 1.6553062271440536e-05, + "loss": 0.9283, + "step": 12495 + }, + { + "epoch": 0.29, + "grad_norm": 2.384808923061129, + "learning_rate": 1.6552485879133714e-05, + "loss": 0.9538, + "step": 12496 + }, + { + "epoch": 0.29, + "grad_norm": 2.139043080326731, + "learning_rate": 1.655190944867625e-05, + "loss": 1.0847, + "step": 12497 + }, + { + "epoch": 0.29, + "grad_norm": 2.526741618941007, + "learning_rate": 1.6551332980071492e-05, + "loss": 1.0585, + "step": 12498 + }, + { + "epoch": 0.29, + "grad_norm": 2.19240845583466, + "learning_rate": 1.6550756473322797e-05, + "loss": 1.1286, + "step": 12499 + }, + { + "epoch": 0.29, + "grad_norm": 1.8326410892475438, + "learning_rate": 1.6550179928433527e-05, + "loss": 1.0454, + "step": 12500 + }, + { + "epoch": 0.29, + "grad_norm": 1.9039217554994257, + "learning_rate": 1.6549603345407035e-05, + "loss": 1.0612, + "step": 12501 + }, + { + "epoch": 0.29, + "grad_norm": 2.2881673539102176, + "learning_rate": 1.654902672424668e-05, + "loss": 0.9185, + "step": 12502 + }, + { + "epoch": 0.29, + "grad_norm": 2.1053651061407077, + "learning_rate": 1.6548450064955815e-05, + "loss": 1.1742, + "step": 12503 + }, + { + "epoch": 0.29, + "grad_norm": 2.0334273480582485, + "learning_rate": 1.6547873367537805e-05, + "loss": 1.119, + "step": 12504 + }, + { + "epoch": 0.29, + "grad_norm": 2.235489781736471, + "learning_rate": 1.6547296631996e-05, + "loss": 1.0895, + "step": 12505 + }, + { + "epoch": 0.29, + "grad_norm": 2.3056632517656084, + "learning_rate": 1.654671985833376e-05, + "loss": 1.0018, + "step": 12506 + }, + { + "epoch": 0.29, + "grad_norm": 4.351000050113906, + "learning_rate": 1.6546143046554447e-05, + "loss": 1.0797, + "step": 12507 + }, + { + "epoch": 0.29, + "grad_norm": 1.9658357093291456, + "learning_rate": 1.6545566196661416e-05, + "loss": 1.0525, + "step": 12508 + }, + { + "epoch": 0.29, + "grad_norm": 2.0366847437138422, + "learning_rate": 1.6544989308658025e-05, + "loss": 1.0644, + "step": 12509 + }, + { + "epoch": 0.29, + "grad_norm": 2.147342157309868, + "learning_rate": 1.6544412382547638e-05, + "loss": 0.9874, + "step": 12510 + }, + { + "epoch": 0.29, + "grad_norm": 2.081848453764507, + "learning_rate": 1.6543835418333606e-05, + "loss": 1.0227, + "step": 12511 + }, + { + "epoch": 0.29, + "grad_norm": 2.0122262307676393, + "learning_rate": 1.6543258416019293e-05, + "loss": 1.0986, + "step": 12512 + }, + { + "epoch": 0.29, + "grad_norm": 1.959681490766269, + "learning_rate": 1.654268137560806e-05, + "loss": 1.0125, + "step": 12513 + }, + { + "epoch": 0.29, + "grad_norm": 2.30884393556297, + "learning_rate": 1.6542104297103264e-05, + "loss": 1.0887, + "step": 12514 + }, + { + "epoch": 0.29, + "grad_norm": 2.182783496950895, + "learning_rate": 1.6541527180508265e-05, + "loss": 0.975, + "step": 12515 + }, + { + "epoch": 0.29, + "grad_norm": 2.04993662970642, + "learning_rate": 1.6540950025826424e-05, + "loss": 1.0064, + "step": 12516 + }, + { + "epoch": 0.29, + "grad_norm": 2.090552816078991, + "learning_rate": 1.65403728330611e-05, + "loss": 1.1271, + "step": 12517 + }, + { + "epoch": 0.29, + "grad_norm": 2.0712314342079003, + "learning_rate": 1.6539795602215655e-05, + "loss": 1.1394, + "step": 12518 + }, + { + "epoch": 0.29, + "grad_norm": 2.3148615837818345, + "learning_rate": 1.6539218333293453e-05, + "loss": 1.0063, + "step": 12519 + }, + { + "epoch": 0.29, + "grad_norm": 2.1006923071789747, + "learning_rate": 1.653864102629785e-05, + "loss": 1.183, + "step": 12520 + }, + { + "epoch": 0.29, + "grad_norm": 2.4590968372557773, + "learning_rate": 1.6538063681232203e-05, + "loss": 0.9054, + "step": 12521 + }, + { + "epoch": 0.3, + "grad_norm": 2.3198523192027594, + "learning_rate": 1.6537486298099884e-05, + "loss": 0.9785, + "step": 12522 + }, + { + "epoch": 0.3, + "grad_norm": 1.9087417828393385, + "learning_rate": 1.653690887690425e-05, + "loss": 1.0838, + "step": 12523 + }, + { + "epoch": 0.3, + "grad_norm": 1.9728366228770458, + "learning_rate": 1.6536331417648664e-05, + "loss": 0.9913, + "step": 12524 + }, + { + "epoch": 0.3, + "grad_norm": 2.1121065219429873, + "learning_rate": 1.653575392033649e-05, + "loss": 0.9963, + "step": 12525 + }, + { + "epoch": 0.3, + "grad_norm": 1.9851328671407633, + "learning_rate": 1.6535176384971082e-05, + "loss": 0.9737, + "step": 12526 + }, + { + "epoch": 0.3, + "grad_norm": 3.670819017376195, + "learning_rate": 1.653459881155581e-05, + "loss": 0.9244, + "step": 12527 + }, + { + "epoch": 0.3, + "grad_norm": 1.7422865260264055, + "learning_rate": 1.6534021200094033e-05, + "loss": 1.0531, + "step": 12528 + }, + { + "epoch": 0.3, + "grad_norm": 2.293198233660848, + "learning_rate": 1.6533443550589116e-05, + "loss": 1.0875, + "step": 12529 + }, + { + "epoch": 0.3, + "grad_norm": 2.238766495090036, + "learning_rate": 1.6532865863044424e-05, + "loss": 1.1437, + "step": 12530 + }, + { + "epoch": 0.3, + "grad_norm": 1.1063874778589733, + "learning_rate": 1.653228813746332e-05, + "loss": 0.9557, + "step": 12531 + }, + { + "epoch": 0.3, + "grad_norm": 2.0082539091059197, + "learning_rate": 1.6531710373849166e-05, + "loss": 0.9556, + "step": 12532 + }, + { + "epoch": 0.3, + "grad_norm": 1.9678918937733687, + "learning_rate": 1.6531132572205324e-05, + "loss": 1.115, + "step": 12533 + }, + { + "epoch": 0.3, + "grad_norm": 3.4352447036397376, + "learning_rate": 1.653055473253516e-05, + "loss": 0.9465, + "step": 12534 + }, + { + "epoch": 0.3, + "grad_norm": 1.9491234233581773, + "learning_rate": 1.6529976854842042e-05, + "loss": 1.0142, + "step": 12535 + }, + { + "epoch": 0.3, + "grad_norm": 1.135184558575648, + "learning_rate": 1.6529398939129328e-05, + "loss": 0.9826, + "step": 12536 + }, + { + "epoch": 0.3, + "grad_norm": 2.686290989782479, + "learning_rate": 1.652882098540039e-05, + "loss": 0.9899, + "step": 12537 + }, + { + "epoch": 0.3, + "grad_norm": 2.7636645316137605, + "learning_rate": 1.6528242993658586e-05, + "loss": 1.1564, + "step": 12538 + }, + { + "epoch": 0.3, + "grad_norm": 1.8350451113486899, + "learning_rate": 1.6527664963907288e-05, + "loss": 1.0001, + "step": 12539 + }, + { + "epoch": 0.3, + "grad_norm": 1.1699728463091472, + "learning_rate": 1.652708689614985e-05, + "loss": 1.0007, + "step": 12540 + }, + { + "epoch": 0.3, + "grad_norm": 6.2100357378698785, + "learning_rate": 1.6526508790389655e-05, + "loss": 1.232, + "step": 12541 + }, + { + "epoch": 0.3, + "grad_norm": 2.133110913193886, + "learning_rate": 1.6525930646630057e-05, + "loss": 0.9575, + "step": 12542 + }, + { + "epoch": 0.3, + "grad_norm": 3.9201204193693573, + "learning_rate": 1.652535246487442e-05, + "loss": 1.0229, + "step": 12543 + }, + { + "epoch": 0.3, + "grad_norm": 2.2072829618852596, + "learning_rate": 1.652477424512612e-05, + "loss": 0.9469, + "step": 12544 + }, + { + "epoch": 0.3, + "grad_norm": 2.1363344611877655, + "learning_rate": 1.652419598738852e-05, + "loss": 0.9424, + "step": 12545 + }, + { + "epoch": 0.3, + "grad_norm": 2.043457199652698, + "learning_rate": 1.6523617691664982e-05, + "loss": 1.0829, + "step": 12546 + }, + { + "epoch": 0.3, + "grad_norm": 1.8649860292894895, + "learning_rate": 1.6523039357958876e-05, + "loss": 1.0181, + "step": 12547 + }, + { + "epoch": 0.3, + "grad_norm": 2.074469317843445, + "learning_rate": 1.652246098627357e-05, + "loss": 1.1173, + "step": 12548 + }, + { + "epoch": 0.3, + "grad_norm": 10.263539842750864, + "learning_rate": 1.6521882576612433e-05, + "loss": 1.0849, + "step": 12549 + }, + { + "epoch": 0.3, + "grad_norm": 2.946264871647178, + "learning_rate": 1.6521304128978826e-05, + "loss": 1.016, + "step": 12550 + }, + { + "epoch": 0.3, + "grad_norm": 2.0717907162751033, + "learning_rate": 1.652072564337613e-05, + "loss": 0.9696, + "step": 12551 + }, + { + "epoch": 0.3, + "grad_norm": 1.0581660810043865, + "learning_rate": 1.6520147119807698e-05, + "loss": 1.0139, + "step": 12552 + }, + { + "epoch": 0.3, + "grad_norm": 2.0348617096662505, + "learning_rate": 1.6519568558276908e-05, + "loss": 1.1001, + "step": 12553 + }, + { + "epoch": 0.3, + "grad_norm": 5.864487199938265, + "learning_rate": 1.6518989958787126e-05, + "loss": 1.0537, + "step": 12554 + }, + { + "epoch": 0.3, + "grad_norm": 1.947837574969875, + "learning_rate": 1.651841132134172e-05, + "loss": 1.1559, + "step": 12555 + }, + { + "epoch": 0.3, + "grad_norm": 1.8646111139827688, + "learning_rate": 1.651783264594406e-05, + "loss": 1.0713, + "step": 12556 + }, + { + "epoch": 0.3, + "grad_norm": 1.936778396521893, + "learning_rate": 1.6517253932597515e-05, + "loss": 1.1242, + "step": 12557 + }, + { + "epoch": 0.3, + "grad_norm": 2.233766004709038, + "learning_rate": 1.6516675181305456e-05, + "loss": 1.1915, + "step": 12558 + }, + { + "epoch": 0.3, + "grad_norm": 2.3105213769697293, + "learning_rate": 1.6516096392071248e-05, + "loss": 0.9508, + "step": 12559 + }, + { + "epoch": 0.3, + "grad_norm": 8.92421183144205, + "learning_rate": 1.651551756489826e-05, + "loss": 1.131, + "step": 12560 + }, + { + "epoch": 0.3, + "grad_norm": 1.9449447267191242, + "learning_rate": 1.6514938699789877e-05, + "loss": 0.986, + "step": 12561 + }, + { + "epoch": 0.3, + "grad_norm": 2.0093033079738656, + "learning_rate": 1.651435979674945e-05, + "loss": 1.0558, + "step": 12562 + }, + { + "epoch": 0.3, + "grad_norm": 2.062948017291744, + "learning_rate": 1.6513780855780358e-05, + "loss": 1.0012, + "step": 12563 + }, + { + "epoch": 0.3, + "grad_norm": 1.9582994267685325, + "learning_rate": 1.6513201876885972e-05, + "loss": 1.0854, + "step": 12564 + }, + { + "epoch": 0.3, + "grad_norm": 2.3579661315677254, + "learning_rate": 1.6512622860069667e-05, + "loss": 1.0762, + "step": 12565 + }, + { + "epoch": 0.3, + "grad_norm": 2.2312080528452873, + "learning_rate": 1.6512043805334808e-05, + "loss": 0.9964, + "step": 12566 + }, + { + "epoch": 0.3, + "grad_norm": 2.1196128158731926, + "learning_rate": 1.6511464712684764e-05, + "loss": 0.9904, + "step": 12567 + }, + { + "epoch": 0.3, + "grad_norm": 1.9807849469144356, + "learning_rate": 1.6510885582122912e-05, + "loss": 1.0776, + "step": 12568 + }, + { + "epoch": 0.3, + "grad_norm": 2.290483082953766, + "learning_rate": 1.651030641365263e-05, + "loss": 1.1315, + "step": 12569 + }, + { + "epoch": 0.3, + "grad_norm": 2.1982723642427646, + "learning_rate": 1.6509727207277273e-05, + "loss": 1.0863, + "step": 12570 + }, + { + "epoch": 0.3, + "grad_norm": 2.162412970053851, + "learning_rate": 1.650914796300023e-05, + "loss": 1.0721, + "step": 12571 + }, + { + "epoch": 0.3, + "grad_norm": 1.1331364690453491, + "learning_rate": 1.6508568680824865e-05, + "loss": 0.9791, + "step": 12572 + }, + { + "epoch": 0.3, + "grad_norm": 2.096584621405455, + "learning_rate": 1.6507989360754553e-05, + "loss": 0.9976, + "step": 12573 + }, + { + "epoch": 0.3, + "grad_norm": 2.0799500339765022, + "learning_rate": 1.6507410002792665e-05, + "loss": 1.0221, + "step": 12574 + }, + { + "epoch": 0.3, + "grad_norm": 2.2093431660342984, + "learning_rate": 1.6506830606942578e-05, + "loss": 1.0474, + "step": 12575 + }, + { + "epoch": 0.3, + "grad_norm": 2.967768143528016, + "learning_rate": 1.650625117320766e-05, + "loss": 1.1253, + "step": 12576 + }, + { + "epoch": 0.3, + "grad_norm": 2.0016139630542362, + "learning_rate": 1.650567170159129e-05, + "loss": 1.0561, + "step": 12577 + }, + { + "epoch": 0.3, + "grad_norm": 1.9393657278554748, + "learning_rate": 1.6505092192096837e-05, + "loss": 1.1293, + "step": 12578 + }, + { + "epoch": 0.3, + "grad_norm": 2.175582065488077, + "learning_rate": 1.650451264472768e-05, + "loss": 1.1092, + "step": 12579 + }, + { + "epoch": 0.3, + "grad_norm": 2.327128643971419, + "learning_rate": 1.6503933059487195e-05, + "loss": 1.0053, + "step": 12580 + }, + { + "epoch": 0.3, + "grad_norm": 2.1730387215428784, + "learning_rate": 1.6503353436378747e-05, + "loss": 0.9841, + "step": 12581 + }, + { + "epoch": 0.3, + "grad_norm": 2.0059437100270063, + "learning_rate": 1.6502773775405717e-05, + "loss": 1.1263, + "step": 12582 + }, + { + "epoch": 0.3, + "grad_norm": 1.846283948047422, + "learning_rate": 1.650219407657148e-05, + "loss": 1.0744, + "step": 12583 + }, + { + "epoch": 0.3, + "grad_norm": 2.3056304834070365, + "learning_rate": 1.6501614339879413e-05, + "loss": 1.1264, + "step": 12584 + }, + { + "epoch": 0.3, + "grad_norm": 1.7332664176203463, + "learning_rate": 1.650103456533289e-05, + "loss": 1.1118, + "step": 12585 + }, + { + "epoch": 0.3, + "grad_norm": 1.1205848280557806, + "learning_rate": 1.650045475293528e-05, + "loss": 1.0258, + "step": 12586 + }, + { + "epoch": 0.3, + "grad_norm": 2.473288195618248, + "learning_rate": 1.6499874902689968e-05, + "loss": 1.0727, + "step": 12587 + }, + { + "epoch": 0.3, + "grad_norm": 2.1683217228474247, + "learning_rate": 1.6499295014600327e-05, + "loss": 1.1192, + "step": 12588 + }, + { + "epoch": 0.3, + "grad_norm": 1.807586620846233, + "learning_rate": 1.649871508866973e-05, + "loss": 1.0016, + "step": 12589 + }, + { + "epoch": 0.3, + "grad_norm": 2.0225088366183117, + "learning_rate": 1.6498135124901557e-05, + "loss": 1.093, + "step": 12590 + }, + { + "epoch": 0.3, + "grad_norm": 2.070598448886872, + "learning_rate": 1.6497555123299184e-05, + "loss": 0.974, + "step": 12591 + }, + { + "epoch": 0.3, + "grad_norm": 2.3405574909450317, + "learning_rate": 1.649697508386599e-05, + "loss": 0.9927, + "step": 12592 + }, + { + "epoch": 0.3, + "grad_norm": 1.92045917867907, + "learning_rate": 1.6496395006605346e-05, + "loss": 1.0467, + "step": 12593 + }, + { + "epoch": 0.3, + "grad_norm": 2.588378423565897, + "learning_rate": 1.6495814891520636e-05, + "loss": 0.9498, + "step": 12594 + }, + { + "epoch": 0.3, + "grad_norm": 2.065309374373533, + "learning_rate": 1.6495234738615238e-05, + "loss": 1.1724, + "step": 12595 + }, + { + "epoch": 0.3, + "grad_norm": 2.1067069524068507, + "learning_rate": 1.6494654547892522e-05, + "loss": 0.9377, + "step": 12596 + }, + { + "epoch": 0.3, + "grad_norm": 2.509263568880449, + "learning_rate": 1.6494074319355874e-05, + "loss": 1.0896, + "step": 12597 + }, + { + "epoch": 0.3, + "grad_norm": 1.9260493725016072, + "learning_rate": 1.649349405300867e-05, + "loss": 1.0038, + "step": 12598 + }, + { + "epoch": 0.3, + "grad_norm": 2.1852959005329744, + "learning_rate": 1.6492913748854285e-05, + "loss": 1.2501, + "step": 12599 + }, + { + "epoch": 0.3, + "grad_norm": 2.019030700563293, + "learning_rate": 1.6492333406896103e-05, + "loss": 0.9953, + "step": 12600 + }, + { + "epoch": 0.3, + "grad_norm": 2.867461106736425, + "learning_rate": 1.6491753027137497e-05, + "loss": 1.0419, + "step": 12601 + }, + { + "epoch": 0.3, + "grad_norm": 2.184280120975586, + "learning_rate": 1.6491172609581856e-05, + "loss": 1.2356, + "step": 12602 + }, + { + "epoch": 0.3, + "grad_norm": 2.3177972089299517, + "learning_rate": 1.649059215423255e-05, + "loss": 0.9616, + "step": 12603 + }, + { + "epoch": 0.3, + "grad_norm": 2.0955042445825924, + "learning_rate": 1.649001166109296e-05, + "loss": 1.1869, + "step": 12604 + }, + { + "epoch": 0.3, + "grad_norm": 2.1875122719233153, + "learning_rate": 1.648943113016647e-05, + "loss": 1.1446, + "step": 12605 + }, + { + "epoch": 0.3, + "grad_norm": 2.0066840748977426, + "learning_rate": 1.648885056145645e-05, + "loss": 0.9581, + "step": 12606 + }, + { + "epoch": 0.3, + "grad_norm": 1.722480680910757, + "learning_rate": 1.6488269954966295e-05, + "loss": 1.0747, + "step": 12607 + }, + { + "epoch": 0.3, + "grad_norm": 2.177925641059794, + "learning_rate": 1.6487689310699377e-05, + "loss": 0.9504, + "step": 12608 + }, + { + "epoch": 0.3, + "grad_norm": 2.7184824590956276, + "learning_rate": 1.648710862865908e-05, + "loss": 1.0749, + "step": 12609 + }, + { + "epoch": 0.3, + "grad_norm": 2.323445721305817, + "learning_rate": 1.648652790884878e-05, + "loss": 1.0751, + "step": 12610 + }, + { + "epoch": 0.3, + "grad_norm": 2.1394721557719056, + "learning_rate": 1.648594715127186e-05, + "loss": 1.0245, + "step": 12611 + }, + { + "epoch": 0.3, + "grad_norm": 2.0406235089500164, + "learning_rate": 1.64853663559317e-05, + "loss": 0.9355, + "step": 12612 + }, + { + "epoch": 0.3, + "grad_norm": 2.1158547636403777, + "learning_rate": 1.648478552283169e-05, + "loss": 0.9558, + "step": 12613 + }, + { + "epoch": 0.3, + "grad_norm": 2.0694422183753334, + "learning_rate": 1.6484204651975203e-05, + "loss": 0.9782, + "step": 12614 + }, + { + "epoch": 0.3, + "grad_norm": 1.1010789632166378, + "learning_rate": 1.6483623743365624e-05, + "loss": 0.9594, + "step": 12615 + }, + { + "epoch": 0.3, + "grad_norm": 2.117319219748138, + "learning_rate": 1.6483042797006334e-05, + "loss": 1.0557, + "step": 12616 + }, + { + "epoch": 0.3, + "grad_norm": 2.099503218939113, + "learning_rate": 1.6482461812900718e-05, + "loss": 1.118, + "step": 12617 + }, + { + "epoch": 0.3, + "grad_norm": 2.075203965000739, + "learning_rate": 1.6481880791052158e-05, + "loss": 1.0623, + "step": 12618 + }, + { + "epoch": 0.3, + "grad_norm": 2.0570176639128146, + "learning_rate": 1.648129973146403e-05, + "loss": 1.0899, + "step": 12619 + }, + { + "epoch": 0.3, + "grad_norm": 2.306820157751432, + "learning_rate": 1.6480718634139726e-05, + "loss": 1.0763, + "step": 12620 + }, + { + "epoch": 0.3, + "grad_norm": 1.0917398842215076, + "learning_rate": 1.648013749908263e-05, + "loss": 1.0388, + "step": 12621 + }, + { + "epoch": 0.3, + "grad_norm": 1.0286681312898658, + "learning_rate": 1.6479556326296116e-05, + "loss": 0.9394, + "step": 12622 + }, + { + "epoch": 0.3, + "grad_norm": 1.996961498560448, + "learning_rate": 1.6478975115783577e-05, + "loss": 0.8874, + "step": 12623 + }, + { + "epoch": 0.3, + "grad_norm": 2.385999586922745, + "learning_rate": 1.6478393867548396e-05, + "loss": 0.9819, + "step": 12624 + }, + { + "epoch": 0.3, + "grad_norm": 1.0828598181904996, + "learning_rate": 1.647781258159395e-05, + "loss": 0.9991, + "step": 12625 + }, + { + "epoch": 0.3, + "grad_norm": 2.2516834841188733, + "learning_rate": 1.647723125792363e-05, + "loss": 1.0584, + "step": 12626 + }, + { + "epoch": 0.3, + "grad_norm": 2.1330981920733194, + "learning_rate": 1.647664989654082e-05, + "loss": 1.0312, + "step": 12627 + }, + { + "epoch": 0.3, + "grad_norm": 2.466911473523172, + "learning_rate": 1.6476068497448903e-05, + "loss": 1.0576, + "step": 12628 + }, + { + "epoch": 0.3, + "grad_norm": 2.3286771136625943, + "learning_rate": 1.6475487060651265e-05, + "loss": 1.1511, + "step": 12629 + }, + { + "epoch": 0.3, + "grad_norm": 2.194168317743814, + "learning_rate": 1.6474905586151293e-05, + "loss": 1.002, + "step": 12630 + }, + { + "epoch": 0.3, + "grad_norm": 2.342943347658531, + "learning_rate": 1.647432407395237e-05, + "loss": 0.9003, + "step": 12631 + }, + { + "epoch": 0.3, + "grad_norm": 1.9092462533556251, + "learning_rate": 1.647374252405788e-05, + "loss": 1.0907, + "step": 12632 + }, + { + "epoch": 0.3, + "grad_norm": 1.9677531418215157, + "learning_rate": 1.6473160936471214e-05, + "loss": 1.1137, + "step": 12633 + }, + { + "epoch": 0.3, + "grad_norm": 2.3047544837366347, + "learning_rate": 1.6472579311195752e-05, + "loss": 0.9988, + "step": 12634 + }, + { + "epoch": 0.3, + "grad_norm": 1.1077854169072558, + "learning_rate": 1.6471997648234886e-05, + "loss": 0.9412, + "step": 12635 + }, + { + "epoch": 0.3, + "grad_norm": 2.794745657871797, + "learning_rate": 1.6471415947592e-05, + "loss": 1.0613, + "step": 12636 + }, + { + "epoch": 0.3, + "grad_norm": 2.00644721735516, + "learning_rate": 1.647083420927049e-05, + "loss": 1.1475, + "step": 12637 + }, + { + "epoch": 0.3, + "grad_norm": 2.257496900222676, + "learning_rate": 1.6470252433273723e-05, + "loss": 1.0307, + "step": 12638 + }, + { + "epoch": 0.3, + "grad_norm": 2.110418321514744, + "learning_rate": 1.6469670619605103e-05, + "loss": 1.0823, + "step": 12639 + }, + { + "epoch": 0.3, + "grad_norm": 2.0406052344518346, + "learning_rate": 1.6469088768268013e-05, + "loss": 1.0061, + "step": 12640 + }, + { + "epoch": 0.3, + "grad_norm": 2.164811816106034, + "learning_rate": 1.646850687926584e-05, + "loss": 1.2055, + "step": 12641 + }, + { + "epoch": 0.3, + "grad_norm": 2.0863489333363923, + "learning_rate": 1.646792495260197e-05, + "loss": 1.0787, + "step": 12642 + }, + { + "epoch": 0.3, + "grad_norm": 2.0554519991892635, + "learning_rate": 1.6467342988279793e-05, + "loss": 1.1289, + "step": 12643 + }, + { + "epoch": 0.3, + "grad_norm": 2.4040684948882345, + "learning_rate": 1.64667609863027e-05, + "loss": 0.9068, + "step": 12644 + }, + { + "epoch": 0.3, + "grad_norm": 1.9865649268197727, + "learning_rate": 1.646617894667408e-05, + "loss": 1.0738, + "step": 12645 + }, + { + "epoch": 0.3, + "grad_norm": 2.1022034567482457, + "learning_rate": 1.646559686939731e-05, + "loss": 1.031, + "step": 12646 + }, + { + "epoch": 0.3, + "grad_norm": 2.230257671301142, + "learning_rate": 1.6465014754475794e-05, + "loss": 1.0029, + "step": 12647 + }, + { + "epoch": 0.3, + "grad_norm": 2.170793759996507, + "learning_rate": 1.6464432601912914e-05, + "loss": 0.8908, + "step": 12648 + }, + { + "epoch": 0.3, + "grad_norm": 1.997837465626046, + "learning_rate": 1.6463850411712063e-05, + "loss": 0.9323, + "step": 12649 + }, + { + "epoch": 0.3, + "grad_norm": 2.291170147040962, + "learning_rate": 1.6463268183876627e-05, + "loss": 0.9991, + "step": 12650 + }, + { + "epoch": 0.3, + "grad_norm": 2.1552233106801584, + "learning_rate": 1.6462685918409998e-05, + "loss": 0.9738, + "step": 12651 + }, + { + "epoch": 0.3, + "grad_norm": 2.0632982729892415, + "learning_rate": 1.6462103615315563e-05, + "loss": 1.0469, + "step": 12652 + }, + { + "epoch": 0.3, + "grad_norm": 2.101872622021105, + "learning_rate": 1.6461521274596717e-05, + "loss": 1.0992, + "step": 12653 + }, + { + "epoch": 0.3, + "grad_norm": 2.3384635962330145, + "learning_rate": 1.6460938896256848e-05, + "loss": 1.0868, + "step": 12654 + }, + { + "epoch": 0.3, + "grad_norm": 2.024653265816062, + "learning_rate": 1.6460356480299345e-05, + "loss": 1.1309, + "step": 12655 + }, + { + "epoch": 0.3, + "grad_norm": 1.1073500932282296, + "learning_rate": 1.6459774026727606e-05, + "loss": 0.9046, + "step": 12656 + }, + { + "epoch": 0.3, + "grad_norm": 2.2826311268474275, + "learning_rate": 1.6459191535545013e-05, + "loss": 1.0678, + "step": 12657 + }, + { + "epoch": 0.3, + "grad_norm": 2.113543373649195, + "learning_rate": 1.6458609006754966e-05, + "loss": 1.0617, + "step": 12658 + }, + { + "epoch": 0.3, + "grad_norm": 2.230103883832146, + "learning_rate": 1.6458026440360847e-05, + "loss": 0.9576, + "step": 12659 + }, + { + "epoch": 0.3, + "grad_norm": 2.075488063190043, + "learning_rate": 1.6457443836366062e-05, + "loss": 1.0885, + "step": 12660 + }, + { + "epoch": 0.3, + "grad_norm": 2.2569052800296623, + "learning_rate": 1.6456861194773988e-05, + "loss": 1.0633, + "step": 12661 + }, + { + "epoch": 0.3, + "grad_norm": 2.504035131452364, + "learning_rate": 1.6456278515588023e-05, + "loss": 1.2346, + "step": 12662 + }, + { + "epoch": 0.3, + "grad_norm": 2.204353321310279, + "learning_rate": 1.6455695798811567e-05, + "loss": 1.0375, + "step": 12663 + }, + { + "epoch": 0.3, + "grad_norm": 2.20280335138939, + "learning_rate": 1.6455113044448e-05, + "loss": 0.9698, + "step": 12664 + }, + { + "epoch": 0.3, + "grad_norm": 2.061940114417344, + "learning_rate": 1.6454530252500723e-05, + "loss": 1.1221, + "step": 12665 + }, + { + "epoch": 0.3, + "grad_norm": 1.1529461344922773, + "learning_rate": 1.645394742297313e-05, + "loss": 1.0106, + "step": 12666 + }, + { + "epoch": 0.3, + "grad_norm": 2.0251691051275214, + "learning_rate": 1.645336455586861e-05, + "loss": 0.9681, + "step": 12667 + }, + { + "epoch": 0.3, + "grad_norm": 2.1234919640566434, + "learning_rate": 1.645278165119056e-05, + "loss": 1.1084, + "step": 12668 + }, + { + "epoch": 0.3, + "grad_norm": 1.9735593139151077, + "learning_rate": 1.6452198708942368e-05, + "loss": 0.9508, + "step": 12669 + }, + { + "epoch": 0.3, + "grad_norm": 1.9656171620304417, + "learning_rate": 1.6451615729127436e-05, + "loss": 1.0156, + "step": 12670 + }, + { + "epoch": 0.3, + "grad_norm": 1.1045318941252602, + "learning_rate": 1.6451032711749154e-05, + "loss": 1.0484, + "step": 12671 + }, + { + "epoch": 0.3, + "grad_norm": 2.009343299154024, + "learning_rate": 1.6450449656810914e-05, + "loss": 1.0006, + "step": 12672 + }, + { + "epoch": 0.3, + "grad_norm": 1.1548455702400622, + "learning_rate": 1.6449866564316118e-05, + "loss": 1.0407, + "step": 12673 + }, + { + "epoch": 0.3, + "grad_norm": 2.495341269906352, + "learning_rate": 1.6449283434268153e-05, + "loss": 1.0089, + "step": 12674 + }, + { + "epoch": 0.3, + "grad_norm": 2.635764814023892, + "learning_rate": 1.644870026667042e-05, + "loss": 0.9859, + "step": 12675 + }, + { + "epoch": 0.3, + "grad_norm": 1.1654218570835502, + "learning_rate": 1.6448117061526314e-05, + "loss": 0.9948, + "step": 12676 + }, + { + "epoch": 0.3, + "grad_norm": 2.42115701561674, + "learning_rate": 1.644753381883923e-05, + "loss": 0.8801, + "step": 12677 + }, + { + "epoch": 0.3, + "grad_norm": 2.035221742895967, + "learning_rate": 1.644695053861256e-05, + "loss": 1.1253, + "step": 12678 + }, + { + "epoch": 0.3, + "grad_norm": 2.333628711974921, + "learning_rate": 1.6446367220849704e-05, + "loss": 1.1198, + "step": 12679 + }, + { + "epoch": 0.3, + "grad_norm": 2.0265989043564168, + "learning_rate": 1.6445783865554056e-05, + "loss": 1.1122, + "step": 12680 + }, + { + "epoch": 0.3, + "grad_norm": 2.287618929653121, + "learning_rate": 1.6445200472729016e-05, + "loss": 1.0833, + "step": 12681 + }, + { + "epoch": 0.3, + "grad_norm": 1.9809006016404007, + "learning_rate": 1.6444617042377978e-05, + "loss": 1.0787, + "step": 12682 + }, + { + "epoch": 0.3, + "grad_norm": 2.276975419228537, + "learning_rate": 1.6444033574504336e-05, + "loss": 1.0526, + "step": 12683 + }, + { + "epoch": 0.3, + "grad_norm": 2.0230179867529756, + "learning_rate": 1.644345006911149e-05, + "loss": 1.0751, + "step": 12684 + }, + { + "epoch": 0.3, + "grad_norm": 1.9825890568064182, + "learning_rate": 1.6442866526202842e-05, + "loss": 1.0968, + "step": 12685 + }, + { + "epoch": 0.3, + "grad_norm": 3.0608650117029343, + "learning_rate": 1.6442282945781784e-05, + "loss": 0.9396, + "step": 12686 + }, + { + "epoch": 0.3, + "grad_norm": 1.9351139608750463, + "learning_rate": 1.6441699327851715e-05, + "loss": 0.9247, + "step": 12687 + }, + { + "epoch": 0.3, + "grad_norm": 2.1086710755914777, + "learning_rate": 1.6441115672416033e-05, + "loss": 1.2134, + "step": 12688 + }, + { + "epoch": 0.3, + "grad_norm": 1.830477195017424, + "learning_rate": 1.6440531979478137e-05, + "loss": 0.9122, + "step": 12689 + }, + { + "epoch": 0.3, + "grad_norm": 2.2369764594877504, + "learning_rate": 1.643994824904142e-05, + "loss": 1.0482, + "step": 12690 + }, + { + "epoch": 0.3, + "grad_norm": 2.745691940742681, + "learning_rate": 1.6439364481109292e-05, + "loss": 1.0316, + "step": 12691 + }, + { + "epoch": 0.3, + "grad_norm": 2.230734169766575, + "learning_rate": 1.643878067568514e-05, + "loss": 1.2019, + "step": 12692 + }, + { + "epoch": 0.3, + "grad_norm": 2.1434328073357793, + "learning_rate": 1.6438196832772372e-05, + "loss": 1.0543, + "step": 12693 + }, + { + "epoch": 0.3, + "grad_norm": 2.057280082861162, + "learning_rate": 1.6437612952374382e-05, + "loss": 0.9983, + "step": 12694 + }, + { + "epoch": 0.3, + "grad_norm": 2.2031333892320974, + "learning_rate": 1.6437029034494574e-05, + "loss": 1.0106, + "step": 12695 + }, + { + "epoch": 0.3, + "grad_norm": 2.2331825823612226, + "learning_rate": 1.6436445079136344e-05, + "loss": 1.011, + "step": 12696 + }, + { + "epoch": 0.3, + "grad_norm": 2.184447130712976, + "learning_rate": 1.6435861086303092e-05, + "loss": 1.0772, + "step": 12697 + }, + { + "epoch": 0.3, + "grad_norm": 1.862013440619761, + "learning_rate": 1.643527705599822e-05, + "loss": 0.9911, + "step": 12698 + }, + { + "epoch": 0.3, + "grad_norm": 2.230933716828121, + "learning_rate": 1.6434692988225125e-05, + "loss": 0.9767, + "step": 12699 + }, + { + "epoch": 0.3, + "grad_norm": 1.9083048156233757, + "learning_rate": 1.6434108882987213e-05, + "loss": 1.189, + "step": 12700 + }, + { + "epoch": 0.3, + "grad_norm": 2.24489180726618, + "learning_rate": 1.643352474028788e-05, + "loss": 1.0591, + "step": 12701 + }, + { + "epoch": 0.3, + "grad_norm": 1.9827332996647473, + "learning_rate": 1.643294056013053e-05, + "loss": 1.1597, + "step": 12702 + }, + { + "epoch": 0.3, + "grad_norm": 1.8543623775370635, + "learning_rate": 1.6432356342518564e-05, + "loss": 1.0537, + "step": 12703 + }, + { + "epoch": 0.3, + "grad_norm": 2.0262499496716724, + "learning_rate": 1.6431772087455384e-05, + "loss": 1.104, + "step": 12704 + }, + { + "epoch": 0.3, + "grad_norm": 2.0196329809221245, + "learning_rate": 1.643118779494439e-05, + "loss": 1.098, + "step": 12705 + }, + { + "epoch": 0.3, + "grad_norm": 1.8079417023445896, + "learning_rate": 1.6430603464988982e-05, + "loss": 0.948, + "step": 12706 + }, + { + "epoch": 0.3, + "grad_norm": 2.680368655505225, + "learning_rate": 1.6430019097592567e-05, + "loss": 0.9993, + "step": 12707 + }, + { + "epoch": 0.3, + "grad_norm": 1.857107399154848, + "learning_rate": 1.6429434692758542e-05, + "loss": 1.1331, + "step": 12708 + }, + { + "epoch": 0.3, + "grad_norm": 1.7389351129360204, + "learning_rate": 1.6428850250490317e-05, + "loss": 0.9272, + "step": 12709 + }, + { + "epoch": 0.3, + "grad_norm": 2.702551265337363, + "learning_rate": 1.642826577079129e-05, + "loss": 1.0352, + "step": 12710 + }, + { + "epoch": 0.3, + "grad_norm": 2.0429587904156716, + "learning_rate": 1.6427681253664863e-05, + "loss": 1.0273, + "step": 12711 + }, + { + "epoch": 0.3, + "grad_norm": 1.9950675990868962, + "learning_rate": 1.642709669911444e-05, + "loss": 0.8766, + "step": 12712 + }, + { + "epoch": 0.3, + "grad_norm": 1.8635394219412877, + "learning_rate": 1.642651210714343e-05, + "loss": 0.9761, + "step": 12713 + }, + { + "epoch": 0.3, + "grad_norm": 1.9328591256504344, + "learning_rate": 1.6425927477755227e-05, + "loss": 0.9766, + "step": 12714 + }, + { + "epoch": 0.3, + "grad_norm": 1.9766880970138199, + "learning_rate": 1.642534281095324e-05, + "loss": 1.0251, + "step": 12715 + }, + { + "epoch": 0.3, + "grad_norm": 2.028320255965984, + "learning_rate": 1.6424758106740874e-05, + "loss": 1.1022, + "step": 12716 + }, + { + "epoch": 0.3, + "grad_norm": 1.869752648309183, + "learning_rate": 1.642417336512153e-05, + "loss": 1.0389, + "step": 12717 + }, + { + "epoch": 0.3, + "grad_norm": 1.2125234952397206, + "learning_rate": 1.642358858609862e-05, + "loss": 0.9191, + "step": 12718 + }, + { + "epoch": 0.3, + "grad_norm": 2.045118559161176, + "learning_rate": 1.642300376967554e-05, + "loss": 1.1045, + "step": 12719 + }, + { + "epoch": 0.3, + "grad_norm": 2.069145704259214, + "learning_rate": 1.64224189158557e-05, + "loss": 1.0495, + "step": 12720 + }, + { + "epoch": 0.3, + "grad_norm": 1.1711622536306467, + "learning_rate": 1.64218340246425e-05, + "loss": 0.9199, + "step": 12721 + }, + { + "epoch": 0.3, + "grad_norm": 1.9638181344284005, + "learning_rate": 1.6421249096039352e-05, + "loss": 0.9726, + "step": 12722 + }, + { + "epoch": 0.3, + "grad_norm": 2.262046236879407, + "learning_rate": 1.6420664130049663e-05, + "loss": 1.055, + "step": 12723 + }, + { + "epoch": 0.3, + "grad_norm": 1.976327690281556, + "learning_rate": 1.6420079126676828e-05, + "loss": 0.838, + "step": 12724 + }, + { + "epoch": 0.3, + "grad_norm": 2.3941139654726666, + "learning_rate": 1.6419494085924267e-05, + "loss": 0.9981, + "step": 12725 + }, + { + "epoch": 0.3, + "grad_norm": 1.1799516192717485, + "learning_rate": 1.6418909007795374e-05, + "loss": 1.0808, + "step": 12726 + }, + { + "epoch": 0.3, + "grad_norm": 2.085731048538166, + "learning_rate": 1.641832389229356e-05, + "loss": 1.1091, + "step": 12727 + }, + { + "epoch": 0.3, + "grad_norm": 2.051876077189455, + "learning_rate": 1.6417738739422234e-05, + "loss": 1.0938, + "step": 12728 + }, + { + "epoch": 0.3, + "grad_norm": 2.0428987745441747, + "learning_rate": 1.64171535491848e-05, + "loss": 1.0252, + "step": 12729 + }, + { + "epoch": 0.3, + "grad_norm": 2.3396259573283116, + "learning_rate": 1.6416568321584668e-05, + "loss": 0.9598, + "step": 12730 + }, + { + "epoch": 0.3, + "grad_norm": 1.0439245068704468, + "learning_rate": 1.641598305662524e-05, + "loss": 1.0312, + "step": 12731 + }, + { + "epoch": 0.3, + "grad_norm": 2.094480518972055, + "learning_rate": 1.641539775430993e-05, + "loss": 1.0425, + "step": 12732 + }, + { + "epoch": 0.3, + "grad_norm": 2.0864476960038076, + "learning_rate": 1.6414812414642144e-05, + "loss": 1.0233, + "step": 12733 + }, + { + "epoch": 0.3, + "grad_norm": 1.1099156790253828, + "learning_rate": 1.6414227037625288e-05, + "loss": 0.9841, + "step": 12734 + }, + { + "epoch": 0.3, + "grad_norm": 2.3759519409749377, + "learning_rate": 1.6413641623262774e-05, + "loss": 1.0598, + "step": 12735 + }, + { + "epoch": 0.3, + "grad_norm": 4.564997860737631, + "learning_rate": 1.6413056171558006e-05, + "loss": 1.074, + "step": 12736 + }, + { + "epoch": 0.3, + "grad_norm": 2.0986162096700802, + "learning_rate": 1.6412470682514396e-05, + "loss": 1.0221, + "step": 12737 + }, + { + "epoch": 0.3, + "grad_norm": 2.1845597781463226, + "learning_rate": 1.641188515613535e-05, + "loss": 1.0407, + "step": 12738 + }, + { + "epoch": 0.3, + "grad_norm": 1.8419880669165178, + "learning_rate": 1.6411299592424276e-05, + "loss": 1.0209, + "step": 12739 + }, + { + "epoch": 0.3, + "grad_norm": 2.019530696841151, + "learning_rate": 1.641071399138459e-05, + "loss": 1.048, + "step": 12740 + }, + { + "epoch": 0.3, + "grad_norm": 2.0809580513308403, + "learning_rate": 1.6410128353019697e-05, + "loss": 1.0185, + "step": 12741 + }, + { + "epoch": 0.3, + "grad_norm": 1.9718317870186304, + "learning_rate": 1.6409542677333007e-05, + "loss": 1.0137, + "step": 12742 + }, + { + "epoch": 0.3, + "grad_norm": 1.987027405742386, + "learning_rate": 1.640895696432793e-05, + "loss": 1.0022, + "step": 12743 + }, + { + "epoch": 0.3, + "grad_norm": 2.307902470840082, + "learning_rate": 1.6408371214007876e-05, + "loss": 1.1045, + "step": 12744 + }, + { + "epoch": 0.3, + "grad_norm": 2.029103675331806, + "learning_rate": 1.6407785426376258e-05, + "loss": 0.9987, + "step": 12745 + }, + { + "epoch": 0.3, + "grad_norm": 2.0833336371431526, + "learning_rate": 1.640719960143648e-05, + "loss": 1.0575, + "step": 12746 + }, + { + "epoch": 0.3, + "grad_norm": 2.1012262076732884, + "learning_rate": 1.6406613739191964e-05, + "loss": 1.0946, + "step": 12747 + }, + { + "epoch": 0.3, + "grad_norm": 2.145724441251253, + "learning_rate": 1.640602783964611e-05, + "loss": 1.0848, + "step": 12748 + }, + { + "epoch": 0.3, + "grad_norm": 1.9212657731051286, + "learning_rate": 1.6405441902802334e-05, + "loss": 1.1099, + "step": 12749 + }, + { + "epoch": 0.3, + "grad_norm": 2.139047214480088, + "learning_rate": 1.640485592866405e-05, + "loss": 1.0349, + "step": 12750 + }, + { + "epoch": 0.3, + "grad_norm": 1.1838043606998019, + "learning_rate": 1.6404269917234662e-05, + "loss": 0.974, + "step": 12751 + }, + { + "epoch": 0.3, + "grad_norm": 1.1068717747241943, + "learning_rate": 1.6403683868517593e-05, + "loss": 1.0263, + "step": 12752 + }, + { + "epoch": 0.3, + "grad_norm": 1.8830889959089119, + "learning_rate": 1.6403097782516244e-05, + "loss": 1.1404, + "step": 12753 + }, + { + "epoch": 0.3, + "grad_norm": 1.9535042126401747, + "learning_rate": 1.6402511659234038e-05, + "loss": 0.9467, + "step": 12754 + }, + { + "epoch": 0.3, + "grad_norm": 2.0926279059891093, + "learning_rate": 1.640192549867438e-05, + "loss": 0.9396, + "step": 12755 + }, + { + "epoch": 0.3, + "grad_norm": 2.4041493288827263, + "learning_rate": 1.6401339300840683e-05, + "loss": 1.0003, + "step": 12756 + }, + { + "epoch": 0.3, + "grad_norm": 3.608152177999633, + "learning_rate": 1.640075306573636e-05, + "loss": 1.0469, + "step": 12757 + }, + { + "epoch": 0.3, + "grad_norm": 2.195912445042005, + "learning_rate": 1.6400166793364828e-05, + "loss": 1.0707, + "step": 12758 + }, + { + "epoch": 0.3, + "grad_norm": 2.05326421743234, + "learning_rate": 1.63995804837295e-05, + "loss": 1.1416, + "step": 12759 + }, + { + "epoch": 0.3, + "grad_norm": 2.282329562051871, + "learning_rate": 1.639899413683379e-05, + "loss": 0.9633, + "step": 12760 + }, + { + "epoch": 0.3, + "grad_norm": 2.167885394899601, + "learning_rate": 1.6398407752681108e-05, + "loss": 1.0782, + "step": 12761 + }, + { + "epoch": 0.3, + "grad_norm": 1.9272355357707973, + "learning_rate": 1.6397821331274872e-05, + "loss": 1.1209, + "step": 12762 + }, + { + "epoch": 0.3, + "grad_norm": 2.2465385730455947, + "learning_rate": 1.639723487261849e-05, + "loss": 1.0199, + "step": 12763 + }, + { + "epoch": 0.3, + "grad_norm": 1.9815157464887132, + "learning_rate": 1.639664837671539e-05, + "loss": 1.0424, + "step": 12764 + }, + { + "epoch": 0.3, + "grad_norm": 2.1183500331389196, + "learning_rate": 1.6396061843568973e-05, + "loss": 1.1199, + "step": 12765 + }, + { + "epoch": 0.3, + "grad_norm": 1.8960440012184032, + "learning_rate": 1.6395475273182655e-05, + "loss": 1.1805, + "step": 12766 + }, + { + "epoch": 0.3, + "grad_norm": 2.3125305983252424, + "learning_rate": 1.6394888665559857e-05, + "loss": 1.0142, + "step": 12767 + }, + { + "epoch": 0.3, + "grad_norm": 2.1621637480746885, + "learning_rate": 1.6394302020703995e-05, + "loss": 0.9673, + "step": 12768 + }, + { + "epoch": 0.3, + "grad_norm": 2.103656163187422, + "learning_rate": 1.6393715338618483e-05, + "loss": 1.0896, + "step": 12769 + }, + { + "epoch": 0.3, + "grad_norm": 2.437726732974108, + "learning_rate": 1.6393128619306734e-05, + "loss": 1.1365, + "step": 12770 + }, + { + "epoch": 0.3, + "grad_norm": 2.0595960185312947, + "learning_rate": 1.6392541862772166e-05, + "loss": 0.9607, + "step": 12771 + }, + { + "epoch": 0.3, + "grad_norm": 1.0925981957005744, + "learning_rate": 1.6391955069018198e-05, + "loss": 1.0162, + "step": 12772 + }, + { + "epoch": 0.3, + "grad_norm": 1.9715382886988655, + "learning_rate": 1.639136823804824e-05, + "loss": 0.9862, + "step": 12773 + }, + { + "epoch": 0.3, + "grad_norm": 2.166408618828723, + "learning_rate": 1.6390781369865715e-05, + "loss": 1.0317, + "step": 12774 + }, + { + "epoch": 0.3, + "grad_norm": 2.049463484841742, + "learning_rate": 1.6390194464474038e-05, + "loss": 1.0315, + "step": 12775 + }, + { + "epoch": 0.3, + "grad_norm": 2.2584069484808307, + "learning_rate": 1.6389607521876625e-05, + "loss": 0.9684, + "step": 12776 + }, + { + "epoch": 0.3, + "grad_norm": 2.6504787936932797, + "learning_rate": 1.6389020542076898e-05, + "loss": 0.991, + "step": 12777 + }, + { + "epoch": 0.3, + "grad_norm": 1.0836334930037015, + "learning_rate": 1.6388433525078266e-05, + "loss": 0.9957, + "step": 12778 + }, + { + "epoch": 0.3, + "grad_norm": 2.068350728537414, + "learning_rate": 1.6387846470884154e-05, + "loss": 1.0804, + "step": 12779 + }, + { + "epoch": 0.3, + "grad_norm": 1.9797223188596362, + "learning_rate": 1.6387259379497973e-05, + "loss": 1.0199, + "step": 12780 + }, + { + "epoch": 0.3, + "grad_norm": 1.927252071541445, + "learning_rate": 1.638667225092315e-05, + "loss": 1.1235, + "step": 12781 + }, + { + "epoch": 0.3, + "grad_norm": 2.1698051278737114, + "learning_rate": 1.6386085085163096e-05, + "loss": 1.0034, + "step": 12782 + }, + { + "epoch": 0.3, + "grad_norm": 2.1884598372823345, + "learning_rate": 1.6385497882221236e-05, + "loss": 0.9776, + "step": 12783 + }, + { + "epoch": 0.3, + "grad_norm": 2.531568712921704, + "learning_rate": 1.6384910642100983e-05, + "loss": 0.9458, + "step": 12784 + }, + { + "epoch": 0.3, + "grad_norm": 1.1056600216072525, + "learning_rate": 1.6384323364805762e-05, + "loss": 0.9975, + "step": 12785 + }, + { + "epoch": 0.3, + "grad_norm": 2.420688835124641, + "learning_rate": 1.6383736050338984e-05, + "loss": 1.0655, + "step": 12786 + }, + { + "epoch": 0.3, + "grad_norm": 1.9726134725219298, + "learning_rate": 1.638314869870408e-05, + "loss": 1.0362, + "step": 12787 + }, + { + "epoch": 0.3, + "grad_norm": 2.0506509799680934, + "learning_rate": 1.6382561309904457e-05, + "loss": 1.1093, + "step": 12788 + }, + { + "epoch": 0.3, + "grad_norm": 2.6302854972991176, + "learning_rate": 1.6381973883943546e-05, + "loss": 1.0639, + "step": 12789 + }, + { + "epoch": 0.3, + "grad_norm": 1.8721086813298558, + "learning_rate": 1.6381386420824763e-05, + "loss": 1.0078, + "step": 12790 + }, + { + "epoch": 0.3, + "grad_norm": 2.3024783144769634, + "learning_rate": 1.6380798920551526e-05, + "loss": 0.9615, + "step": 12791 + }, + { + "epoch": 0.3, + "grad_norm": 2.409124434206318, + "learning_rate": 1.6380211383127255e-05, + "loss": 1.0424, + "step": 12792 + }, + { + "epoch": 0.3, + "grad_norm": 2.1609255582087465, + "learning_rate": 1.637962380855538e-05, + "loss": 0.9366, + "step": 12793 + }, + { + "epoch": 0.3, + "grad_norm": 2.0380017934253734, + "learning_rate": 1.637903619683931e-05, + "loss": 0.9328, + "step": 12794 + }, + { + "epoch": 0.3, + "grad_norm": 1.9644841328147993, + "learning_rate": 1.6378448547982473e-05, + "loss": 0.9485, + "step": 12795 + }, + { + "epoch": 0.3, + "grad_norm": 2.112831563665398, + "learning_rate": 1.6377860861988288e-05, + "loss": 1.189, + "step": 12796 + }, + { + "epoch": 0.3, + "grad_norm": 2.0199819059466333, + "learning_rate": 1.6377273138860182e-05, + "loss": 1.1972, + "step": 12797 + }, + { + "epoch": 0.3, + "grad_norm": 1.916386291150766, + "learning_rate": 1.637668537860157e-05, + "loss": 1.0666, + "step": 12798 + }, + { + "epoch": 0.3, + "grad_norm": 3.761195303439192, + "learning_rate": 1.6376097581215874e-05, + "loss": 1.0942, + "step": 12799 + }, + { + "epoch": 0.3, + "grad_norm": 2.0306103024599524, + "learning_rate": 1.6375509746706523e-05, + "loss": 1.0372, + "step": 12800 + }, + { + "epoch": 0.3, + "grad_norm": 2.452879228721419, + "learning_rate": 1.6374921875076933e-05, + "loss": 1.0146, + "step": 12801 + }, + { + "epoch": 0.3, + "grad_norm": 2.190607679455213, + "learning_rate": 1.6374333966330535e-05, + "loss": 1.0122, + "step": 12802 + }, + { + "epoch": 0.3, + "grad_norm": 2.001579459487593, + "learning_rate": 1.637374602047074e-05, + "loss": 1.0359, + "step": 12803 + }, + { + "epoch": 0.3, + "grad_norm": 2.239601015789991, + "learning_rate": 1.6373158037500984e-05, + "loss": 1.0388, + "step": 12804 + }, + { + "epoch": 0.3, + "grad_norm": 2.764452784088699, + "learning_rate": 1.6372570017424682e-05, + "loss": 1.0742, + "step": 12805 + }, + { + "epoch": 0.3, + "grad_norm": 1.9049295005486135, + "learning_rate": 1.6371981960245257e-05, + "loss": 1.1858, + "step": 12806 + }, + { + "epoch": 0.3, + "grad_norm": 1.0972868823524735, + "learning_rate": 1.6371393865966137e-05, + "loss": 1.0496, + "step": 12807 + }, + { + "epoch": 0.3, + "grad_norm": 2.1116442057792204, + "learning_rate": 1.6370805734590746e-05, + "loss": 1.0888, + "step": 12808 + }, + { + "epoch": 0.3, + "grad_norm": 2.057031256895395, + "learning_rate": 1.6370217566122505e-05, + "loss": 1.0313, + "step": 12809 + }, + { + "epoch": 0.3, + "grad_norm": 2.2705331539462623, + "learning_rate": 1.6369629360564845e-05, + "loss": 0.9498, + "step": 12810 + }, + { + "epoch": 0.3, + "grad_norm": 2.6379343445855086, + "learning_rate": 1.6369041117921183e-05, + "loss": 0.9893, + "step": 12811 + }, + { + "epoch": 0.3, + "grad_norm": 1.9119807632729913, + "learning_rate": 1.6368452838194947e-05, + "loss": 1.0167, + "step": 12812 + }, + { + "epoch": 0.3, + "grad_norm": 1.8858154760211345, + "learning_rate": 1.6367864521389563e-05, + "loss": 1.1629, + "step": 12813 + }, + { + "epoch": 0.3, + "grad_norm": 2.3451877528295175, + "learning_rate": 1.6367276167508456e-05, + "loss": 1.1357, + "step": 12814 + }, + { + "epoch": 0.3, + "grad_norm": 1.9983409919618027, + "learning_rate": 1.636668777655505e-05, + "loss": 0.9133, + "step": 12815 + }, + { + "epoch": 0.3, + "grad_norm": 2.0659340677736293, + "learning_rate": 1.6366099348532777e-05, + "loss": 1.0166, + "step": 12816 + }, + { + "epoch": 0.3, + "grad_norm": 2.250371577321403, + "learning_rate": 1.636551088344505e-05, + "loss": 1.0343, + "step": 12817 + }, + { + "epoch": 0.3, + "grad_norm": 4.00230052936715, + "learning_rate": 1.636492238129531e-05, + "loss": 0.9255, + "step": 12818 + }, + { + "epoch": 0.3, + "grad_norm": 2.3097595740148233, + "learning_rate": 1.636433384208697e-05, + "loss": 1.0156, + "step": 12819 + }, + { + "epoch": 0.3, + "grad_norm": 2.0197071599715435, + "learning_rate": 1.636374526582347e-05, + "loss": 1.018, + "step": 12820 + }, + { + "epoch": 0.3, + "grad_norm": 3.019903178583351, + "learning_rate": 1.636315665250823e-05, + "loss": 1.153, + "step": 12821 + }, + { + "epoch": 0.3, + "grad_norm": 1.9834458969031477, + "learning_rate": 1.636256800214467e-05, + "loss": 1.0175, + "step": 12822 + }, + { + "epoch": 0.3, + "grad_norm": 1.9968968036238075, + "learning_rate": 1.6361979314736233e-05, + "loss": 1.0632, + "step": 12823 + }, + { + "epoch": 0.3, + "grad_norm": 2.3109757469966845, + "learning_rate": 1.6361390590286336e-05, + "loss": 1.0329, + "step": 12824 + }, + { + "epoch": 0.3, + "grad_norm": 1.898087983226967, + "learning_rate": 1.636080182879841e-05, + "loss": 1.1123, + "step": 12825 + }, + { + "epoch": 0.3, + "grad_norm": 2.9564043851262314, + "learning_rate": 1.636021303027588e-05, + "loss": 1.0364, + "step": 12826 + }, + { + "epoch": 0.3, + "grad_norm": 2.327583094207822, + "learning_rate": 1.6359624194722176e-05, + "loss": 1.0986, + "step": 12827 + }, + { + "epoch": 0.3, + "grad_norm": 2.6664205434034955, + "learning_rate": 1.6359035322140727e-05, + "loss": 1.0209, + "step": 12828 + }, + { + "epoch": 0.3, + "grad_norm": 2.0404657340670105, + "learning_rate": 1.635844641253496e-05, + "loss": 0.904, + "step": 12829 + }, + { + "epoch": 0.3, + "grad_norm": 2.52112583989847, + "learning_rate": 1.6357857465908308e-05, + "loss": 1.0208, + "step": 12830 + }, + { + "epoch": 0.3, + "grad_norm": 2.2666181146057602, + "learning_rate": 1.6357268482264195e-05, + "loss": 1.0387, + "step": 12831 + }, + { + "epoch": 0.3, + "grad_norm": 2.0359415683766073, + "learning_rate": 1.6356679461606052e-05, + "loss": 1.0086, + "step": 12832 + }, + { + "epoch": 0.3, + "grad_norm": 1.9827630902714415, + "learning_rate": 1.6356090403937307e-05, + "loss": 1.0589, + "step": 12833 + }, + { + "epoch": 0.3, + "grad_norm": 1.9204297372080157, + "learning_rate": 1.6355501309261398e-05, + "loss": 0.975, + "step": 12834 + }, + { + "epoch": 0.3, + "grad_norm": 2.3165380795607895, + "learning_rate": 1.6354912177581743e-05, + "loss": 1.0818, + "step": 12835 + }, + { + "epoch": 0.3, + "grad_norm": 2.204220530054668, + "learning_rate": 1.6354323008901774e-05, + "loss": 1.1194, + "step": 12836 + }, + { + "epoch": 0.3, + "grad_norm": 2.1522793322086216, + "learning_rate": 1.635373380322493e-05, + "loss": 1.0937, + "step": 12837 + }, + { + "epoch": 0.3, + "grad_norm": 2.2454022070641506, + "learning_rate": 1.6353144560554636e-05, + "loss": 1.0683, + "step": 12838 + }, + { + "epoch": 0.3, + "grad_norm": 2.0635041046831972, + "learning_rate": 1.6352555280894323e-05, + "loss": 0.9984, + "step": 12839 + }, + { + "epoch": 0.3, + "grad_norm": 1.9951346393602414, + "learning_rate": 1.635196596424742e-05, + "loss": 1.0796, + "step": 12840 + }, + { + "epoch": 0.3, + "grad_norm": 1.6734410809942517, + "learning_rate": 1.635137661061736e-05, + "loss": 0.9978, + "step": 12841 + }, + { + "epoch": 0.3, + "grad_norm": 2.2119877734544335, + "learning_rate": 1.6350787220007576e-05, + "loss": 1.2522, + "step": 12842 + }, + { + "epoch": 0.3, + "grad_norm": 1.9316130080240104, + "learning_rate": 1.6350197792421497e-05, + "loss": 1.1677, + "step": 12843 + }, + { + "epoch": 0.3, + "grad_norm": 2.3061856585052305, + "learning_rate": 1.634960832786256e-05, + "loss": 0.9718, + "step": 12844 + }, + { + "epoch": 0.3, + "grad_norm": 1.1396817990644352, + "learning_rate": 1.6349018826334188e-05, + "loss": 0.9756, + "step": 12845 + }, + { + "epoch": 0.3, + "grad_norm": 2.9346635404589367, + "learning_rate": 1.634842928783982e-05, + "loss": 1.0635, + "step": 12846 + }, + { + "epoch": 0.3, + "grad_norm": 2.1362362557469052, + "learning_rate": 1.6347839712382887e-05, + "loss": 0.993, + "step": 12847 + }, + { + "epoch": 0.3, + "grad_norm": 2.042471673177284, + "learning_rate": 1.6347250099966823e-05, + "loss": 1.0575, + "step": 12848 + }, + { + "epoch": 0.3, + "grad_norm": 2.024165913471251, + "learning_rate": 1.6346660450595058e-05, + "loss": 1.1018, + "step": 12849 + }, + { + "epoch": 0.3, + "grad_norm": 2.4216685725307063, + "learning_rate": 1.6346070764271025e-05, + "loss": 1.1256, + "step": 12850 + }, + { + "epoch": 0.3, + "grad_norm": 2.586822453871635, + "learning_rate": 1.6345481040998163e-05, + "loss": 0.9546, + "step": 12851 + }, + { + "epoch": 0.3, + "grad_norm": 1.9863077197109866, + "learning_rate": 1.6344891280779894e-05, + "loss": 1.0691, + "step": 12852 + }, + { + "epoch": 0.3, + "grad_norm": 2.1635109404253288, + "learning_rate": 1.6344301483619666e-05, + "loss": 0.9955, + "step": 12853 + }, + { + "epoch": 0.3, + "grad_norm": 1.2033822386290598, + "learning_rate": 1.6343711649520904e-05, + "loss": 0.9977, + "step": 12854 + }, + { + "epoch": 0.3, + "grad_norm": 2.067182567347086, + "learning_rate": 1.6343121778487047e-05, + "loss": 1.0313, + "step": 12855 + }, + { + "epoch": 0.3, + "grad_norm": 2.135210416470007, + "learning_rate": 1.6342531870521523e-05, + "loss": 1.0904, + "step": 12856 + }, + { + "epoch": 0.3, + "grad_norm": 2.027707708080901, + "learning_rate": 1.634194192562777e-05, + "loss": 1.0635, + "step": 12857 + }, + { + "epoch": 0.3, + "grad_norm": 2.230165733955035, + "learning_rate": 1.6341351943809228e-05, + "loss": 0.8186, + "step": 12858 + }, + { + "epoch": 0.3, + "grad_norm": 2.2889875765490704, + "learning_rate": 1.634076192506932e-05, + "loss": 1.1036, + "step": 12859 + }, + { + "epoch": 0.3, + "grad_norm": 2.508157844380539, + "learning_rate": 1.6340171869411497e-05, + "loss": 1.0901, + "step": 12860 + }, + { + "epoch": 0.3, + "grad_norm": 1.877116067226415, + "learning_rate": 1.633958177683918e-05, + "loss": 1.0955, + "step": 12861 + }, + { + "epoch": 0.3, + "grad_norm": 1.935750948380271, + "learning_rate": 1.6338991647355812e-05, + "loss": 1.1165, + "step": 12862 + }, + { + "epoch": 0.3, + "grad_norm": 2.150152158767595, + "learning_rate": 1.6338401480964832e-05, + "loss": 1.0944, + "step": 12863 + }, + { + "epoch": 0.3, + "grad_norm": 2.214577181776867, + "learning_rate": 1.633781127766967e-05, + "loss": 1.2334, + "step": 12864 + }, + { + "epoch": 0.3, + "grad_norm": 2.2199700451074538, + "learning_rate": 1.633722103747376e-05, + "loss": 1.0704, + "step": 12865 + }, + { + "epoch": 0.3, + "grad_norm": 4.284378505829866, + "learning_rate": 1.6336630760380546e-05, + "loss": 1.0185, + "step": 12866 + }, + { + "epoch": 0.3, + "grad_norm": 2.0726885766201923, + "learning_rate": 1.633604044639346e-05, + "loss": 1.1391, + "step": 12867 + }, + { + "epoch": 0.3, + "grad_norm": 1.9804006475245557, + "learning_rate": 1.6335450095515944e-05, + "loss": 1.075, + "step": 12868 + }, + { + "epoch": 0.3, + "grad_norm": 1.946093622066558, + "learning_rate": 1.633485970775143e-05, + "loss": 1.146, + "step": 12869 + }, + { + "epoch": 0.3, + "grad_norm": 2.006870612601346, + "learning_rate": 1.6334269283103353e-05, + "loss": 0.9838, + "step": 12870 + }, + { + "epoch": 0.3, + "grad_norm": 1.92179073523599, + "learning_rate": 1.633367882157516e-05, + "loss": 1.0471, + "step": 12871 + }, + { + "epoch": 0.3, + "grad_norm": 2.4043433068120326, + "learning_rate": 1.6333088323170285e-05, + "loss": 1.1191, + "step": 12872 + }, + { + "epoch": 0.3, + "grad_norm": 2.056618922228658, + "learning_rate": 1.633249778789216e-05, + "loss": 0.9268, + "step": 12873 + }, + { + "epoch": 0.3, + "grad_norm": 1.9079977781460677, + "learning_rate": 1.633190721574423e-05, + "loss": 0.9936, + "step": 12874 + }, + { + "epoch": 0.3, + "grad_norm": 1.985065784658076, + "learning_rate": 1.633131660672993e-05, + "loss": 1.1459, + "step": 12875 + }, + { + "epoch": 0.3, + "grad_norm": 3.4827434212967128, + "learning_rate": 1.6330725960852706e-05, + "loss": 1.0458, + "step": 12876 + }, + { + "epoch": 0.3, + "grad_norm": 3.0847521877636206, + "learning_rate": 1.6330135278115988e-05, + "loss": 0.9982, + "step": 12877 + }, + { + "epoch": 0.3, + "grad_norm": 2.016606475186772, + "learning_rate": 1.6329544558523212e-05, + "loss": 1.1022, + "step": 12878 + }, + { + "epoch": 0.3, + "grad_norm": 2.0883163254244494, + "learning_rate": 1.6328953802077833e-05, + "loss": 1.0958, + "step": 12879 + }, + { + "epoch": 0.3, + "grad_norm": 2.0365943245007623, + "learning_rate": 1.6328363008783274e-05, + "loss": 1.0588, + "step": 12880 + }, + { + "epoch": 0.3, + "grad_norm": 2.0287708203211188, + "learning_rate": 1.6327772178642986e-05, + "loss": 0.9196, + "step": 12881 + }, + { + "epoch": 0.3, + "grad_norm": 2.121674196328712, + "learning_rate": 1.6327181311660406e-05, + "loss": 1.0153, + "step": 12882 + }, + { + "epoch": 0.3, + "grad_norm": 1.1557808105115104, + "learning_rate": 1.632659040783897e-05, + "loss": 0.9698, + "step": 12883 + }, + { + "epoch": 0.3, + "grad_norm": 2.128710671942943, + "learning_rate": 1.6325999467182123e-05, + "loss": 1.2053, + "step": 12884 + }, + { + "epoch": 0.3, + "grad_norm": 2.4396111631473922, + "learning_rate": 1.6325408489693308e-05, + "loss": 1.0341, + "step": 12885 + }, + { + "epoch": 0.3, + "grad_norm": 2.214580062119279, + "learning_rate": 1.6324817475375957e-05, + "loss": 1.0601, + "step": 12886 + }, + { + "epoch": 0.3, + "grad_norm": 1.099662785287429, + "learning_rate": 1.6324226424233517e-05, + "loss": 1.0025, + "step": 12887 + }, + { + "epoch": 0.3, + "grad_norm": 1.9745801988119607, + "learning_rate": 1.632363533626943e-05, + "loss": 1.1187, + "step": 12888 + }, + { + "epoch": 0.3, + "grad_norm": 1.8934384502025674, + "learning_rate": 1.6323044211487135e-05, + "loss": 1.0204, + "step": 12889 + }, + { + "epoch": 0.3, + "grad_norm": 2.2154293363690782, + "learning_rate": 1.6322453049890078e-05, + "loss": 1.1966, + "step": 12890 + }, + { + "epoch": 0.3, + "grad_norm": 2.198858168767432, + "learning_rate": 1.6321861851481693e-05, + "loss": 1.0231, + "step": 12891 + }, + { + "epoch": 0.3, + "grad_norm": 1.1559685262998134, + "learning_rate": 1.632127061626543e-05, + "loss": 0.927, + "step": 12892 + }, + { + "epoch": 0.3, + "grad_norm": 1.8768991535516928, + "learning_rate": 1.6320679344244727e-05, + "loss": 1.1144, + "step": 12893 + }, + { + "epoch": 0.3, + "grad_norm": 2.0637106170326596, + "learning_rate": 1.6320088035423027e-05, + "loss": 1.0793, + "step": 12894 + }, + { + "epoch": 0.3, + "grad_norm": 2.0686088362384156, + "learning_rate": 1.6319496689803775e-05, + "loss": 0.9658, + "step": 12895 + }, + { + "epoch": 0.3, + "grad_norm": 1.9759371837007582, + "learning_rate": 1.6318905307390412e-05, + "loss": 1.0069, + "step": 12896 + }, + { + "epoch": 0.3, + "grad_norm": 2.305355574976199, + "learning_rate": 1.6318313888186383e-05, + "loss": 1.0815, + "step": 12897 + }, + { + "epoch": 0.3, + "grad_norm": 1.9232971616901162, + "learning_rate": 1.6317722432195127e-05, + "loss": 1.0946, + "step": 12898 + }, + { + "epoch": 0.3, + "grad_norm": 1.7965158358834425, + "learning_rate": 1.6317130939420093e-05, + "loss": 0.9911, + "step": 12899 + }, + { + "epoch": 0.3, + "grad_norm": 1.1207892684797265, + "learning_rate": 1.6316539409864724e-05, + "loss": 0.9749, + "step": 12900 + }, + { + "epoch": 0.3, + "grad_norm": 1.9807792508107769, + "learning_rate": 1.631594784353246e-05, + "loss": 1.058, + "step": 12901 + }, + { + "epoch": 0.3, + "grad_norm": 1.9276132690512306, + "learning_rate": 1.631535624042675e-05, + "loss": 1.1233, + "step": 12902 + }, + { + "epoch": 0.3, + "grad_norm": 1.9868744787530945, + "learning_rate": 1.6314764600551035e-05, + "loss": 1.0046, + "step": 12903 + }, + { + "epoch": 0.3, + "grad_norm": 2.2015967511175436, + "learning_rate": 1.6314172923908764e-05, + "loss": 1.1171, + "step": 12904 + }, + { + "epoch": 0.3, + "grad_norm": 2.2904615563545043, + "learning_rate": 1.6313581210503375e-05, + "loss": 1.0847, + "step": 12905 + }, + { + "epoch": 0.3, + "grad_norm": 1.8813479510400395, + "learning_rate": 1.631298946033832e-05, + "loss": 0.8877, + "step": 12906 + }, + { + "epoch": 0.3, + "grad_norm": 2.0693860077212363, + "learning_rate": 1.6312397673417043e-05, + "loss": 1.0738, + "step": 12907 + }, + { + "epoch": 0.3, + "grad_norm": 2.076092081519305, + "learning_rate": 1.6311805849742984e-05, + "loss": 0.932, + "step": 12908 + }, + { + "epoch": 0.3, + "grad_norm": 1.9776482338264225, + "learning_rate": 1.6311213989319598e-05, + "loss": 0.9967, + "step": 12909 + }, + { + "epoch": 0.3, + "grad_norm": 1.8871229731652768, + "learning_rate": 1.6310622092150323e-05, + "loss": 1.0583, + "step": 12910 + }, + { + "epoch": 0.3, + "grad_norm": 2.0888459294607555, + "learning_rate": 1.631003015823861e-05, + "loss": 1.1148, + "step": 12911 + }, + { + "epoch": 0.3, + "grad_norm": 1.8521796314183547, + "learning_rate": 1.63094381875879e-05, + "loss": 0.9136, + "step": 12912 + }, + { + "epoch": 0.3, + "grad_norm": 2.0736468301962905, + "learning_rate": 1.630884618020165e-05, + "loss": 1.042, + "step": 12913 + }, + { + "epoch": 0.3, + "grad_norm": 2.228268724600524, + "learning_rate": 1.6308254136083295e-05, + "loss": 1.171, + "step": 12914 + }, + { + "epoch": 0.3, + "grad_norm": 2.0142519193611355, + "learning_rate": 1.6307662055236288e-05, + "loss": 1.1414, + "step": 12915 + }, + { + "epoch": 0.3, + "grad_norm": 1.9451632452011731, + "learning_rate": 1.6307069937664077e-05, + "loss": 0.9945, + "step": 12916 + }, + { + "epoch": 0.3, + "grad_norm": 1.1839186008968212, + "learning_rate": 1.6306477783370108e-05, + "loss": 0.9942, + "step": 12917 + }, + { + "epoch": 0.3, + "grad_norm": 1.840924708988101, + "learning_rate": 1.6305885592357826e-05, + "loss": 1.021, + "step": 12918 + }, + { + "epoch": 0.3, + "grad_norm": 1.885259706154324, + "learning_rate": 1.6305293364630684e-05, + "loss": 1.1115, + "step": 12919 + }, + { + "epoch": 0.3, + "grad_norm": 2.5714943407614, + "learning_rate": 1.630470110019213e-05, + "loss": 1.0419, + "step": 12920 + }, + { + "epoch": 0.3, + "grad_norm": 1.847421557368391, + "learning_rate": 1.6304108799045607e-05, + "loss": 1.1868, + "step": 12921 + }, + { + "epoch": 0.3, + "grad_norm": 3.3073832287247646, + "learning_rate": 1.6303516461194564e-05, + "loss": 1.0997, + "step": 12922 + }, + { + "epoch": 0.3, + "grad_norm": 2.063967264878816, + "learning_rate": 1.6302924086642456e-05, + "loss": 0.9687, + "step": 12923 + }, + { + "epoch": 0.3, + "grad_norm": 1.8928228712505315, + "learning_rate": 1.6302331675392728e-05, + "loss": 1.0844, + "step": 12924 + }, + { + "epoch": 0.3, + "grad_norm": 2.052001498356757, + "learning_rate": 1.630173922744883e-05, + "loss": 1.1107, + "step": 12925 + }, + { + "epoch": 0.3, + "grad_norm": 1.241001604842712, + "learning_rate": 1.6301146742814215e-05, + "loss": 1.0002, + "step": 12926 + }, + { + "epoch": 0.3, + "grad_norm": 2.1481266233550387, + "learning_rate": 1.630055422149232e-05, + "loss": 1.0054, + "step": 12927 + }, + { + "epoch": 0.3, + "grad_norm": 1.9640912213194284, + "learning_rate": 1.6299961663486613e-05, + "loss": 1.0957, + "step": 12928 + }, + { + "epoch": 0.3, + "grad_norm": 1.9737667517724249, + "learning_rate": 1.629936906880053e-05, + "loss": 1.0059, + "step": 12929 + }, + { + "epoch": 0.3, + "grad_norm": 1.0441649346887956, + "learning_rate": 1.6298776437437526e-05, + "loss": 0.9558, + "step": 12930 + }, + { + "epoch": 0.3, + "grad_norm": 1.9258134872589643, + "learning_rate": 1.629818376940105e-05, + "loss": 1.0865, + "step": 12931 + }, + { + "epoch": 0.3, + "grad_norm": 1.821336894199846, + "learning_rate": 1.6297591064694557e-05, + "loss": 1.0668, + "step": 12932 + }, + { + "epoch": 0.3, + "grad_norm": 2.0716578416300986, + "learning_rate": 1.629699832332149e-05, + "loss": 1.0785, + "step": 12933 + }, + { + "epoch": 0.3, + "grad_norm": 1.8918326307020608, + "learning_rate": 1.629640554528531e-05, + "loss": 1.0263, + "step": 12934 + }, + { + "epoch": 0.3, + "grad_norm": 2.214400676341528, + "learning_rate": 1.629581273058946e-05, + "loss": 1.0158, + "step": 12935 + }, + { + "epoch": 0.3, + "grad_norm": 2.7985150747127467, + "learning_rate": 1.62952198792374e-05, + "loss": 1.1754, + "step": 12936 + }, + { + "epoch": 0.3, + "grad_norm": 1.1439175947160145, + "learning_rate": 1.6294626991232574e-05, + "loss": 0.9349, + "step": 12937 + }, + { + "epoch": 0.3, + "grad_norm": 1.956441738969491, + "learning_rate": 1.6294034066578436e-05, + "loss": 1.0667, + "step": 12938 + }, + { + "epoch": 0.3, + "grad_norm": 2.077770202082732, + "learning_rate": 1.629344110527844e-05, + "loss": 1.0909, + "step": 12939 + }, + { + "epoch": 0.3, + "grad_norm": 2.0035157118235936, + "learning_rate": 1.6292848107336038e-05, + "loss": 1.0393, + "step": 12940 + }, + { + "epoch": 0.3, + "grad_norm": 2.128054833222583, + "learning_rate": 1.6292255072754682e-05, + "loss": 0.9843, + "step": 12941 + }, + { + "epoch": 0.3, + "grad_norm": 3.0315834243319397, + "learning_rate": 1.6291662001537827e-05, + "loss": 1.0536, + "step": 12942 + }, + { + "epoch": 0.3, + "grad_norm": 2.392972465929309, + "learning_rate": 1.629106889368892e-05, + "loss": 1.0645, + "step": 12943 + }, + { + "epoch": 0.3, + "grad_norm": 2.0854849875939014, + "learning_rate": 1.6290475749211417e-05, + "loss": 0.9555, + "step": 12944 + }, + { + "epoch": 0.3, + "grad_norm": 2.906344894733311, + "learning_rate": 1.6289882568108778e-05, + "loss": 1.1585, + "step": 12945 + }, + { + "epoch": 0.3, + "grad_norm": 1.1510182736511376, + "learning_rate": 1.628928935038445e-05, + "loss": 0.9524, + "step": 12946 + }, + { + "epoch": 0.31, + "grad_norm": 1.133562950962034, + "learning_rate": 1.6288696096041886e-05, + "loss": 0.9191, + "step": 12947 + }, + { + "epoch": 0.31, + "grad_norm": 2.1311418057035953, + "learning_rate": 1.628810280508454e-05, + "loss": 0.9189, + "step": 12948 + }, + { + "epoch": 0.31, + "grad_norm": 2.174540812890105, + "learning_rate": 1.6287509477515872e-05, + "loss": 1.2191, + "step": 12949 + }, + { + "epoch": 0.31, + "grad_norm": 2.0341240933192957, + "learning_rate": 1.6286916113339333e-05, + "loss": 1.048, + "step": 12950 + }, + { + "epoch": 0.31, + "grad_norm": 1.0325307128426715, + "learning_rate": 1.628632271255838e-05, + "loss": 0.9691, + "step": 12951 + }, + { + "epoch": 0.31, + "grad_norm": 2.015696850097197, + "learning_rate": 1.6285729275176463e-05, + "loss": 0.9845, + "step": 12952 + }, + { + "epoch": 0.31, + "grad_norm": 2.394036077310134, + "learning_rate": 1.628513580119704e-05, + "loss": 1.1324, + "step": 12953 + }, + { + "epoch": 0.31, + "grad_norm": 2.104024185066118, + "learning_rate": 1.6284542290623568e-05, + "loss": 1.075, + "step": 12954 + }, + { + "epoch": 0.31, + "grad_norm": 2.0799995729899887, + "learning_rate": 1.62839487434595e-05, + "loss": 1.0713, + "step": 12955 + }, + { + "epoch": 0.31, + "grad_norm": 1.9334473713463594, + "learning_rate": 1.6283355159708297e-05, + "loss": 0.9179, + "step": 12956 + }, + { + "epoch": 0.31, + "grad_norm": 1.82225152106822, + "learning_rate": 1.6282761539373402e-05, + "loss": 1.1373, + "step": 12957 + }, + { + "epoch": 0.31, + "grad_norm": 2.0556986457017836, + "learning_rate": 1.628216788245829e-05, + "loss": 1.0267, + "step": 12958 + }, + { + "epoch": 0.31, + "grad_norm": 2.1151527552386646, + "learning_rate": 1.6281574188966402e-05, + "loss": 1.0259, + "step": 12959 + }, + { + "epoch": 0.31, + "grad_norm": 1.8691043470314075, + "learning_rate": 1.6280980458901205e-05, + "loss": 1.1262, + "step": 12960 + }, + { + "epoch": 0.31, + "grad_norm": 2.132574763340279, + "learning_rate": 1.6280386692266142e-05, + "loss": 1.0924, + "step": 12961 + }, + { + "epoch": 0.31, + "grad_norm": 1.7974795889640622, + "learning_rate": 1.6279792889064688e-05, + "loss": 1.0129, + "step": 12962 + }, + { + "epoch": 0.31, + "grad_norm": 2.0916497451932976, + "learning_rate": 1.6279199049300288e-05, + "loss": 1.0547, + "step": 12963 + }, + { + "epoch": 0.31, + "grad_norm": 2.2098557820800395, + "learning_rate": 1.6278605172976404e-05, + "loss": 0.97, + "step": 12964 + }, + { + "epoch": 0.31, + "grad_norm": 2.066226470852638, + "learning_rate": 1.6278011260096495e-05, + "loss": 1.1255, + "step": 12965 + }, + { + "epoch": 0.31, + "grad_norm": 2.116382538998671, + "learning_rate": 1.6277417310664015e-05, + "loss": 1.1244, + "step": 12966 + }, + { + "epoch": 0.31, + "grad_norm": 1.9307594929362446, + "learning_rate": 1.6276823324682426e-05, + "loss": 0.965, + "step": 12967 + }, + { + "epoch": 0.31, + "grad_norm": 2.6557226541987724, + "learning_rate": 1.6276229302155183e-05, + "loss": 1.0159, + "step": 12968 + }, + { + "epoch": 0.31, + "grad_norm": 2.0891792751727176, + "learning_rate": 1.6275635243085743e-05, + "loss": 0.9338, + "step": 12969 + }, + { + "epoch": 0.31, + "grad_norm": 2.0407112382853327, + "learning_rate": 1.6275041147477572e-05, + "loss": 1.0211, + "step": 12970 + }, + { + "epoch": 0.31, + "grad_norm": 2.405673295721323, + "learning_rate": 1.6274447015334122e-05, + "loss": 1.0253, + "step": 12971 + }, + { + "epoch": 0.31, + "grad_norm": 2.3899746492123866, + "learning_rate": 1.6273852846658856e-05, + "loss": 1.0853, + "step": 12972 + }, + { + "epoch": 0.31, + "grad_norm": 2.066518531227215, + "learning_rate": 1.6273258641455233e-05, + "loss": 1.0776, + "step": 12973 + }, + { + "epoch": 0.31, + "grad_norm": 2.6685032546497705, + "learning_rate": 1.6272664399726712e-05, + "loss": 1.1215, + "step": 12974 + }, + { + "epoch": 0.31, + "grad_norm": 2.381275916744638, + "learning_rate": 1.6272070121476753e-05, + "loss": 0.9626, + "step": 12975 + }, + { + "epoch": 0.31, + "grad_norm": 2.387282567244375, + "learning_rate": 1.6271475806708815e-05, + "loss": 0.9299, + "step": 12976 + }, + { + "epoch": 0.31, + "grad_norm": 2.3337453785638176, + "learning_rate": 1.6270881455426358e-05, + "loss": 1.0025, + "step": 12977 + }, + { + "epoch": 0.31, + "grad_norm": 1.9906896794184035, + "learning_rate": 1.6270287067632847e-05, + "loss": 0.9883, + "step": 12978 + }, + { + "epoch": 0.31, + "grad_norm": 2.304964885950719, + "learning_rate": 1.626969264333174e-05, + "loss": 1.0176, + "step": 12979 + }, + { + "epoch": 0.31, + "grad_norm": 2.240480884273491, + "learning_rate": 1.6269098182526494e-05, + "loss": 1.116, + "step": 12980 + }, + { + "epoch": 0.31, + "grad_norm": 2.2357643697445853, + "learning_rate": 1.6268503685220577e-05, + "loss": 0.9118, + "step": 12981 + }, + { + "epoch": 0.31, + "grad_norm": 2.1624056700090986, + "learning_rate": 1.6267909151417444e-05, + "loss": 0.9198, + "step": 12982 + }, + { + "epoch": 0.31, + "grad_norm": 1.9576634001869608, + "learning_rate": 1.6267314581120564e-05, + "loss": 1.1568, + "step": 12983 + }, + { + "epoch": 0.31, + "grad_norm": 2.153772227664132, + "learning_rate": 1.6266719974333387e-05, + "loss": 1.0691, + "step": 12984 + }, + { + "epoch": 0.31, + "grad_norm": 1.1199958661983802, + "learning_rate": 1.6266125331059387e-05, + "loss": 0.9795, + "step": 12985 + }, + { + "epoch": 0.31, + "grad_norm": 2.1322640594943607, + "learning_rate": 1.626553065130202e-05, + "loss": 1.0415, + "step": 12986 + }, + { + "epoch": 0.31, + "grad_norm": 1.9842232799632271, + "learning_rate": 1.6264935935064753e-05, + "loss": 1.0428, + "step": 12987 + }, + { + "epoch": 0.31, + "grad_norm": 1.9813700276714266, + "learning_rate": 1.6264341182351043e-05, + "loss": 1.0106, + "step": 12988 + }, + { + "epoch": 0.31, + "grad_norm": 2.5007096710108674, + "learning_rate": 1.6263746393164356e-05, + "loss": 1.0172, + "step": 12989 + }, + { + "epoch": 0.31, + "grad_norm": 2.1335323325511553, + "learning_rate": 1.6263151567508154e-05, + "loss": 0.9544, + "step": 12990 + }, + { + "epoch": 0.31, + "grad_norm": 2.5887783190692373, + "learning_rate": 1.62625567053859e-05, + "loss": 1.004, + "step": 12991 + }, + { + "epoch": 0.31, + "grad_norm": 2.5079209775502944, + "learning_rate": 1.6261961806801062e-05, + "loss": 1.1543, + "step": 12992 + }, + { + "epoch": 0.31, + "grad_norm": 2.284694664850735, + "learning_rate": 1.6261366871757097e-05, + "loss": 0.9996, + "step": 12993 + }, + { + "epoch": 0.31, + "grad_norm": 1.9718460190688587, + "learning_rate": 1.626077190025747e-05, + "loss": 0.9494, + "step": 12994 + }, + { + "epoch": 0.31, + "grad_norm": 1.8806837581485059, + "learning_rate": 1.6260176892305654e-05, + "loss": 0.9481, + "step": 12995 + }, + { + "epoch": 0.31, + "grad_norm": 2.064387252752776, + "learning_rate": 1.6259581847905097e-05, + "loss": 1.0778, + "step": 12996 + }, + { + "epoch": 0.31, + "grad_norm": 1.9597177223915279, + "learning_rate": 1.6258986767059278e-05, + "loss": 1.0248, + "step": 12997 + }, + { + "epoch": 0.31, + "grad_norm": 2.091701753953095, + "learning_rate": 1.6258391649771654e-05, + "loss": 1.2123, + "step": 12998 + }, + { + "epoch": 0.31, + "grad_norm": 2.120430117838047, + "learning_rate": 1.6257796496045694e-05, + "loss": 1.012, + "step": 12999 + }, + { + "epoch": 0.31, + "grad_norm": 2.2194891079705537, + "learning_rate": 1.625720130588486e-05, + "loss": 0.9793, + "step": 13000 + }, + { + "epoch": 0.31, + "grad_norm": 2.0439268406153133, + "learning_rate": 1.625660607929262e-05, + "loss": 0.8296, + "step": 13001 + }, + { + "epoch": 0.31, + "grad_norm": 1.1605174185002423, + "learning_rate": 1.625601081627244e-05, + "loss": 0.9841, + "step": 13002 + }, + { + "epoch": 0.31, + "grad_norm": 2.2326675861367717, + "learning_rate": 1.6255415516827783e-05, + "loss": 1.0509, + "step": 13003 + }, + { + "epoch": 0.31, + "grad_norm": 2.293871778220604, + "learning_rate": 1.6254820180962115e-05, + "loss": 1.0341, + "step": 13004 + }, + { + "epoch": 0.31, + "grad_norm": 1.8896562503436327, + "learning_rate": 1.6254224808678906e-05, + "loss": 0.9953, + "step": 13005 + }, + { + "epoch": 0.31, + "grad_norm": 2.183845792620688, + "learning_rate": 1.625362939998162e-05, + "loss": 1.1309, + "step": 13006 + }, + { + "epoch": 0.31, + "grad_norm": 1.9493489548048042, + "learning_rate": 1.6253033954873725e-05, + "loss": 1.0666, + "step": 13007 + }, + { + "epoch": 0.31, + "grad_norm": 1.914561533891642, + "learning_rate": 1.6252438473358682e-05, + "loss": 1.067, + "step": 13008 + }, + { + "epoch": 0.31, + "grad_norm": 1.7081038850561374, + "learning_rate": 1.625184295543997e-05, + "loss": 1.1796, + "step": 13009 + }, + { + "epoch": 0.31, + "grad_norm": 2.1026676824547974, + "learning_rate": 1.625124740112104e-05, + "loss": 0.9742, + "step": 13010 + }, + { + "epoch": 0.31, + "grad_norm": 2.112193399968502, + "learning_rate": 1.6250651810405375e-05, + "loss": 1.1154, + "step": 13011 + }, + { + "epoch": 0.31, + "grad_norm": 2.1436459156948113, + "learning_rate": 1.6250056183296434e-05, + "loss": 1.0602, + "step": 13012 + }, + { + "epoch": 0.31, + "grad_norm": 2.292787131211667, + "learning_rate": 1.6249460519797687e-05, + "loss": 1.0031, + "step": 13013 + }, + { + "epoch": 0.31, + "grad_norm": 1.95782246853988, + "learning_rate": 1.6248864819912603e-05, + "loss": 1.1211, + "step": 13014 + }, + { + "epoch": 0.31, + "grad_norm": 2.2825643439837684, + "learning_rate": 1.624826908364465e-05, + "loss": 1.1265, + "step": 13015 + }, + { + "epoch": 0.31, + "grad_norm": 2.1133034081584228, + "learning_rate": 1.6247673310997294e-05, + "loss": 1.0038, + "step": 13016 + }, + { + "epoch": 0.31, + "grad_norm": 1.8628324621971124, + "learning_rate": 1.6247077501974008e-05, + "loss": 1.0966, + "step": 13017 + }, + { + "epoch": 0.31, + "grad_norm": 2.1266974330715827, + "learning_rate": 1.624648165657826e-05, + "loss": 1.0351, + "step": 13018 + }, + { + "epoch": 0.31, + "grad_norm": 1.904684945936814, + "learning_rate": 1.6245885774813514e-05, + "loss": 1.1414, + "step": 13019 + }, + { + "epoch": 0.31, + "grad_norm": 2.43226153745955, + "learning_rate": 1.624528985668325e-05, + "loss": 1.1911, + "step": 13020 + }, + { + "epoch": 0.31, + "grad_norm": 1.8693442174617323, + "learning_rate": 1.6244693902190928e-05, + "loss": 1.03, + "step": 13021 + }, + { + "epoch": 0.31, + "grad_norm": 2.1066764925528103, + "learning_rate": 1.6244097911340018e-05, + "loss": 1.0647, + "step": 13022 + }, + { + "epoch": 0.31, + "grad_norm": 2.371366226551694, + "learning_rate": 1.6243501884133997e-05, + "loss": 0.9835, + "step": 13023 + }, + { + "epoch": 0.31, + "grad_norm": 1.9848205320158252, + "learning_rate": 1.6242905820576327e-05, + "loss": 1.0694, + "step": 13024 + }, + { + "epoch": 0.31, + "grad_norm": 1.9505172572635792, + "learning_rate": 1.6242309720670492e-05, + "loss": 1.0161, + "step": 13025 + }, + { + "epoch": 0.31, + "grad_norm": 1.8825698656999499, + "learning_rate": 1.6241713584419946e-05, + "loss": 1.1143, + "step": 13026 + }, + { + "epoch": 0.31, + "grad_norm": 2.0649918074014537, + "learning_rate": 1.6241117411828167e-05, + "loss": 1.0066, + "step": 13027 + }, + { + "epoch": 0.31, + "grad_norm": 1.8832266110271396, + "learning_rate": 1.624052120289863e-05, + "loss": 1.0807, + "step": 13028 + }, + { + "epoch": 0.31, + "grad_norm": 1.959161194449844, + "learning_rate": 1.6239924957634802e-05, + "loss": 1.0016, + "step": 13029 + }, + { + "epoch": 0.31, + "grad_norm": 1.9930791106663295, + "learning_rate": 1.6239328676040155e-05, + "loss": 1.0758, + "step": 13030 + }, + { + "epoch": 0.31, + "grad_norm": 2.369199607115698, + "learning_rate": 1.623873235811816e-05, + "loss": 1.0633, + "step": 13031 + }, + { + "epoch": 0.31, + "grad_norm": 2.3918013452254194, + "learning_rate": 1.623813600387229e-05, + "loss": 1.0666, + "step": 13032 + }, + { + "epoch": 0.31, + "grad_norm": 1.8316361815592301, + "learning_rate": 1.6237539613306018e-05, + "loss": 0.9954, + "step": 13033 + }, + { + "epoch": 0.31, + "grad_norm": 2.0562015006458774, + "learning_rate": 1.6236943186422815e-05, + "loss": 1.0148, + "step": 13034 + }, + { + "epoch": 0.31, + "grad_norm": 1.0953884035067474, + "learning_rate": 1.6236346723226154e-05, + "loss": 0.956, + "step": 13035 + }, + { + "epoch": 0.31, + "grad_norm": 2.30344738378466, + "learning_rate": 1.623575022371951e-05, + "loss": 0.9707, + "step": 13036 + }, + { + "epoch": 0.31, + "grad_norm": 2.174645103711322, + "learning_rate": 1.6235153687906354e-05, + "loss": 1.0747, + "step": 13037 + }, + { + "epoch": 0.31, + "grad_norm": 1.9995566066132933, + "learning_rate": 1.6234557115790157e-05, + "loss": 0.9548, + "step": 13038 + }, + { + "epoch": 0.31, + "grad_norm": 1.8330908286806766, + "learning_rate": 1.6233960507374397e-05, + "loss": 0.9902, + "step": 13039 + }, + { + "epoch": 0.31, + "grad_norm": 1.7586894530700068, + "learning_rate": 1.6233363862662543e-05, + "loss": 0.9848, + "step": 13040 + }, + { + "epoch": 0.31, + "grad_norm": 2.3246286530633613, + "learning_rate": 1.6232767181658072e-05, + "loss": 0.9978, + "step": 13041 + }, + { + "epoch": 0.31, + "grad_norm": 2.0411517024878525, + "learning_rate": 1.623217046436446e-05, + "loss": 0.9847, + "step": 13042 + }, + { + "epoch": 0.31, + "grad_norm": 1.8165352610179966, + "learning_rate": 1.6231573710785172e-05, + "loss": 0.9009, + "step": 13043 + }, + { + "epoch": 0.31, + "grad_norm": 1.9367128406773426, + "learning_rate": 1.6230976920923696e-05, + "loss": 0.8242, + "step": 13044 + }, + { + "epoch": 0.31, + "grad_norm": 2.0524944168344055, + "learning_rate": 1.6230380094783497e-05, + "loss": 1.0025, + "step": 13045 + }, + { + "epoch": 0.31, + "grad_norm": 2.100373333626918, + "learning_rate": 1.622978323236805e-05, + "loss": 1.0341, + "step": 13046 + }, + { + "epoch": 0.31, + "grad_norm": 1.9131535212188235, + "learning_rate": 1.6229186333680832e-05, + "loss": 0.9641, + "step": 13047 + }, + { + "epoch": 0.31, + "grad_norm": 1.985808294278732, + "learning_rate": 1.6228589398725326e-05, + "loss": 1.0982, + "step": 13048 + }, + { + "epoch": 0.31, + "grad_norm": 1.9591464585800673, + "learning_rate": 1.6227992427504995e-05, + "loss": 0.968, + "step": 13049 + }, + { + "epoch": 0.31, + "grad_norm": 1.185554432884145, + "learning_rate": 1.6227395420023323e-05, + "loss": 1.0537, + "step": 13050 + }, + { + "epoch": 0.31, + "grad_norm": 2.0115008165541544, + "learning_rate": 1.6226798376283782e-05, + "loss": 0.9444, + "step": 13051 + }, + { + "epoch": 0.31, + "grad_norm": 1.8466686818840374, + "learning_rate": 1.6226201296289848e-05, + "loss": 1.1124, + "step": 13052 + }, + { + "epoch": 0.31, + "grad_norm": 2.0186593369327803, + "learning_rate": 1.6225604180045e-05, + "loss": 1.0419, + "step": 13053 + }, + { + "epoch": 0.31, + "grad_norm": 2.024978845002995, + "learning_rate": 1.6225007027552715e-05, + "loss": 1.0805, + "step": 13054 + }, + { + "epoch": 0.31, + "grad_norm": 2.3768729644617967, + "learning_rate": 1.6224409838816465e-05, + "loss": 1.0414, + "step": 13055 + }, + { + "epoch": 0.31, + "grad_norm": 2.6057162956323485, + "learning_rate": 1.6223812613839734e-05, + "loss": 1.0228, + "step": 13056 + }, + { + "epoch": 0.31, + "grad_norm": 1.9433253173862974, + "learning_rate": 1.6223215352625994e-05, + "loss": 1.0476, + "step": 13057 + }, + { + "epoch": 0.31, + "grad_norm": 2.1366748911010602, + "learning_rate": 1.6222618055178723e-05, + "loss": 1.1244, + "step": 13058 + }, + { + "epoch": 0.31, + "grad_norm": 2.054171522224453, + "learning_rate": 1.62220207215014e-05, + "loss": 1.0196, + "step": 13059 + }, + { + "epoch": 0.31, + "grad_norm": 2.6897834894030583, + "learning_rate": 1.6221423351597505e-05, + "loss": 1.1287, + "step": 13060 + }, + { + "epoch": 0.31, + "grad_norm": 2.0077712794281553, + "learning_rate": 1.622082594547051e-05, + "loss": 0.9675, + "step": 13061 + }, + { + "epoch": 0.31, + "grad_norm": 2.034606099589869, + "learning_rate": 1.62202285031239e-05, + "loss": 1.0591, + "step": 13062 + }, + { + "epoch": 0.31, + "grad_norm": 2.7352643159961545, + "learning_rate": 1.6219631024561148e-05, + "loss": 0.9684, + "step": 13063 + }, + { + "epoch": 0.31, + "grad_norm": 1.9107727911443202, + "learning_rate": 1.6219033509785738e-05, + "loss": 0.9164, + "step": 13064 + }, + { + "epoch": 0.31, + "grad_norm": 1.9966256799728885, + "learning_rate": 1.6218435958801143e-05, + "loss": 1.036, + "step": 13065 + }, + { + "epoch": 0.31, + "grad_norm": 1.804794297139185, + "learning_rate": 1.6217838371610847e-05, + "loss": 1.0082, + "step": 13066 + }, + { + "epoch": 0.31, + "grad_norm": 1.1005125344972342, + "learning_rate": 1.6217240748218325e-05, + "loss": 0.95, + "step": 13067 + }, + { + "epoch": 0.31, + "grad_norm": 2.1318505024991268, + "learning_rate": 1.621664308862706e-05, + "loss": 0.9166, + "step": 13068 + }, + { + "epoch": 0.31, + "grad_norm": 1.9824587467950474, + "learning_rate": 1.621604539284053e-05, + "loss": 1.068, + "step": 13069 + }, + { + "epoch": 0.31, + "grad_norm": 1.9789706577851207, + "learning_rate": 1.6215447660862216e-05, + "loss": 1.0103, + "step": 13070 + }, + { + "epoch": 0.31, + "grad_norm": 2.0672827682020007, + "learning_rate": 1.6214849892695602e-05, + "loss": 1.1305, + "step": 13071 + }, + { + "epoch": 0.31, + "grad_norm": 2.943339581840679, + "learning_rate": 1.6214252088344164e-05, + "loss": 1.0013, + "step": 13072 + }, + { + "epoch": 0.31, + "grad_norm": 2.5748652115341844, + "learning_rate": 1.621365424781138e-05, + "loss": 0.9815, + "step": 13073 + }, + { + "epoch": 0.31, + "grad_norm": 2.1365266125993077, + "learning_rate": 1.6213056371100732e-05, + "loss": 0.9766, + "step": 13074 + }, + { + "epoch": 0.31, + "grad_norm": 2.050228854669338, + "learning_rate": 1.6212458458215706e-05, + "loss": 1.0232, + "step": 13075 + }, + { + "epoch": 0.31, + "grad_norm": 2.087670497959157, + "learning_rate": 1.621186050915978e-05, + "loss": 1.0579, + "step": 13076 + }, + { + "epoch": 0.31, + "grad_norm": 1.9139409261406386, + "learning_rate": 1.6211262523936432e-05, + "loss": 1.1551, + "step": 13077 + }, + { + "epoch": 0.31, + "grad_norm": 2.1528044265384088, + "learning_rate": 1.6210664502549152e-05, + "loss": 1.0457, + "step": 13078 + }, + { + "epoch": 0.31, + "grad_norm": 2.13943212465591, + "learning_rate": 1.6210066445001413e-05, + "loss": 1.1018, + "step": 13079 + }, + { + "epoch": 0.31, + "grad_norm": 2.2810057600573153, + "learning_rate": 1.6209468351296703e-05, + "loss": 1.0574, + "step": 13080 + }, + { + "epoch": 0.31, + "grad_norm": 2.0867099358669923, + "learning_rate": 1.6208870221438503e-05, + "loss": 0.9635, + "step": 13081 + }, + { + "epoch": 0.31, + "grad_norm": 1.8729158281435516, + "learning_rate": 1.6208272055430292e-05, + "loss": 0.9098, + "step": 13082 + }, + { + "epoch": 0.31, + "grad_norm": 1.9210197773306803, + "learning_rate": 1.620767385327556e-05, + "loss": 1.017, + "step": 13083 + }, + { + "epoch": 0.31, + "grad_norm": 2.2042193480605956, + "learning_rate": 1.620707561497778e-05, + "loss": 1.0324, + "step": 13084 + }, + { + "epoch": 0.31, + "grad_norm": 2.514083821185611, + "learning_rate": 1.6206477340540444e-05, + "loss": 0.881, + "step": 13085 + }, + { + "epoch": 0.31, + "grad_norm": 1.1717556382028833, + "learning_rate": 1.6205879029967032e-05, + "loss": 1.0128, + "step": 13086 + }, + { + "epoch": 0.31, + "grad_norm": 2.413040937913417, + "learning_rate": 1.620528068326103e-05, + "loss": 1.0283, + "step": 13087 + }, + { + "epoch": 0.31, + "grad_norm": 1.9317793367220482, + "learning_rate": 1.6204682300425915e-05, + "loss": 0.9798, + "step": 13088 + }, + { + "epoch": 0.31, + "grad_norm": 2.664527782343026, + "learning_rate": 1.6204083881465175e-05, + "loss": 1.0023, + "step": 13089 + }, + { + "epoch": 0.31, + "grad_norm": 2.0098236793172286, + "learning_rate": 1.62034854263823e-05, + "loss": 1.0715, + "step": 13090 + }, + { + "epoch": 0.31, + "grad_norm": 1.9009594641170011, + "learning_rate": 1.620288693518076e-05, + "loss": 0.9979, + "step": 13091 + }, + { + "epoch": 0.31, + "grad_norm": 2.1867899077462463, + "learning_rate": 1.6202288407864056e-05, + "loss": 0.9128, + "step": 13092 + }, + { + "epoch": 0.31, + "grad_norm": 2.7819578560535017, + "learning_rate": 1.620168984443566e-05, + "loss": 1.068, + "step": 13093 + }, + { + "epoch": 0.31, + "grad_norm": 1.7956036740784367, + "learning_rate": 1.6201091244899067e-05, + "loss": 0.9776, + "step": 13094 + }, + { + "epoch": 0.31, + "grad_norm": 1.8205364449214583, + "learning_rate": 1.6200492609257758e-05, + "loss": 1.1527, + "step": 13095 + }, + { + "epoch": 0.31, + "grad_norm": 2.123473172764469, + "learning_rate": 1.6199893937515214e-05, + "loss": 1.1062, + "step": 13096 + }, + { + "epoch": 0.31, + "grad_norm": 1.8717033464846589, + "learning_rate": 1.6199295229674927e-05, + "loss": 1.0786, + "step": 13097 + }, + { + "epoch": 0.31, + "grad_norm": 2.4976517138829233, + "learning_rate": 1.6198696485740378e-05, + "loss": 1.0671, + "step": 13098 + }, + { + "epoch": 0.31, + "grad_norm": 2.191275202667054, + "learning_rate": 1.619809770571506e-05, + "loss": 1.0721, + "step": 13099 + }, + { + "epoch": 0.31, + "grad_norm": 2.1857748286665046, + "learning_rate": 1.619749888960245e-05, + "loss": 1.1945, + "step": 13100 + }, + { + "epoch": 0.31, + "grad_norm": 1.9191162409379559, + "learning_rate": 1.6196900037406043e-05, + "loss": 1.1467, + "step": 13101 + }, + { + "epoch": 0.31, + "grad_norm": 2.6429208718596855, + "learning_rate": 1.619630114912932e-05, + "loss": 1.0658, + "step": 13102 + }, + { + "epoch": 0.31, + "grad_norm": 2.0542741240021174, + "learning_rate": 1.6195702224775768e-05, + "loss": 1.0187, + "step": 13103 + }, + { + "epoch": 0.31, + "grad_norm": 1.1757338999880689, + "learning_rate": 1.619510326434888e-05, + "loss": 0.8753, + "step": 13104 + }, + { + "epoch": 0.31, + "grad_norm": 2.1593186200283747, + "learning_rate": 1.619450426785214e-05, + "loss": 0.9119, + "step": 13105 + }, + { + "epoch": 0.31, + "grad_norm": 2.111416715713905, + "learning_rate": 1.619390523528903e-05, + "loss": 1.0702, + "step": 13106 + }, + { + "epoch": 0.31, + "grad_norm": 1.951779419636138, + "learning_rate": 1.619330616666305e-05, + "loss": 1.0739, + "step": 13107 + }, + { + "epoch": 0.31, + "grad_norm": 1.964983429586646, + "learning_rate": 1.6192707061977676e-05, + "loss": 1.0178, + "step": 13108 + }, + { + "epoch": 0.31, + "grad_norm": 1.9598997388787798, + "learning_rate": 1.61921079212364e-05, + "loss": 1.0089, + "step": 13109 + }, + { + "epoch": 0.31, + "grad_norm": 1.9240008957976802, + "learning_rate": 1.6191508744442713e-05, + "loss": 0.9508, + "step": 13110 + }, + { + "epoch": 0.31, + "grad_norm": 1.9911277878983233, + "learning_rate": 1.6190909531600103e-05, + "loss": 1.1821, + "step": 13111 + }, + { + "epoch": 0.31, + "grad_norm": 1.9837375996645612, + "learning_rate": 1.6190310282712056e-05, + "loss": 0.9834, + "step": 13112 + }, + { + "epoch": 0.31, + "grad_norm": 2.10712428230128, + "learning_rate": 1.6189710997782063e-05, + "loss": 1.1387, + "step": 13113 + }, + { + "epoch": 0.31, + "grad_norm": 1.9889445088578936, + "learning_rate": 1.6189111676813613e-05, + "loss": 1.0492, + "step": 13114 + }, + { + "epoch": 0.31, + "grad_norm": 2.530127770689355, + "learning_rate": 1.6188512319810197e-05, + "loss": 1.0864, + "step": 13115 + }, + { + "epoch": 0.31, + "grad_norm": 2.553877446615938, + "learning_rate": 1.6187912926775303e-05, + "loss": 1.1227, + "step": 13116 + }, + { + "epoch": 0.31, + "grad_norm": 2.1345769398215793, + "learning_rate": 1.618731349771242e-05, + "loss": 0.9929, + "step": 13117 + }, + { + "epoch": 0.31, + "grad_norm": 1.9019727014590644, + "learning_rate": 1.6186714032625036e-05, + "loss": 1.075, + "step": 13118 + }, + { + "epoch": 0.31, + "grad_norm": 1.9635116897020386, + "learning_rate": 1.6186114531516646e-05, + "loss": 1.0659, + "step": 13119 + }, + { + "epoch": 0.31, + "grad_norm": 2.3345684885782085, + "learning_rate": 1.6185514994390738e-05, + "loss": 1.0956, + "step": 13120 + }, + { + "epoch": 0.31, + "grad_norm": 2.0999914321829296, + "learning_rate": 1.6184915421250805e-05, + "loss": 0.9761, + "step": 13121 + }, + { + "epoch": 0.31, + "grad_norm": 1.8717329358540873, + "learning_rate": 1.6184315812100337e-05, + "loss": 1.0, + "step": 13122 + }, + { + "epoch": 0.31, + "grad_norm": 2.05005787625621, + "learning_rate": 1.6183716166942824e-05, + "loss": 1.0813, + "step": 13123 + }, + { + "epoch": 0.31, + "grad_norm": 2.851546944043435, + "learning_rate": 1.6183116485781755e-05, + "loss": 1.0599, + "step": 13124 + }, + { + "epoch": 0.31, + "grad_norm": 2.2628083864514643, + "learning_rate": 1.6182516768620627e-05, + "loss": 1.1027, + "step": 13125 + }, + { + "epoch": 0.31, + "grad_norm": 1.1885611150234103, + "learning_rate": 1.618191701546293e-05, + "loss": 0.9304, + "step": 13126 + }, + { + "epoch": 0.31, + "grad_norm": 2.2770441190048962, + "learning_rate": 1.6181317226312154e-05, + "loss": 0.9158, + "step": 13127 + }, + { + "epoch": 0.31, + "grad_norm": 2.0481068059147636, + "learning_rate": 1.618071740117179e-05, + "loss": 0.9999, + "step": 13128 + }, + { + "epoch": 0.31, + "grad_norm": 1.8193201369215277, + "learning_rate": 1.6180117540045338e-05, + "loss": 1.0324, + "step": 13129 + }, + { + "epoch": 0.31, + "grad_norm": 1.1969630092706098, + "learning_rate": 1.617951764293628e-05, + "loss": 1.0341, + "step": 13130 + }, + { + "epoch": 0.31, + "grad_norm": 1.992336277104622, + "learning_rate": 1.6178917709848113e-05, + "loss": 1.0031, + "step": 13131 + }, + { + "epoch": 0.31, + "grad_norm": 2.2548032372171973, + "learning_rate": 1.6178317740784333e-05, + "loss": 1.0672, + "step": 13132 + }, + { + "epoch": 0.31, + "grad_norm": 1.8991122777075808, + "learning_rate": 1.6177717735748433e-05, + "loss": 1.111, + "step": 13133 + }, + { + "epoch": 0.31, + "grad_norm": 2.196472256904503, + "learning_rate": 1.61771176947439e-05, + "loss": 1.0211, + "step": 13134 + }, + { + "epoch": 0.31, + "grad_norm": 2.0144964413667226, + "learning_rate": 1.6176517617774234e-05, + "loss": 1.1498, + "step": 13135 + }, + { + "epoch": 0.31, + "grad_norm": 1.1442995083965075, + "learning_rate": 1.617591750484293e-05, + "loss": 0.88, + "step": 13136 + }, + { + "epoch": 0.31, + "grad_norm": 1.8458913637729444, + "learning_rate": 1.6175317355953477e-05, + "loss": 1.0618, + "step": 13137 + }, + { + "epoch": 0.31, + "grad_norm": 1.9207274569819939, + "learning_rate": 1.6174717171109372e-05, + "loss": 1.0673, + "step": 13138 + }, + { + "epoch": 0.31, + "grad_norm": 1.8666503476025107, + "learning_rate": 1.6174116950314106e-05, + "loss": 1.0198, + "step": 13139 + }, + { + "epoch": 0.31, + "grad_norm": 1.898277775153866, + "learning_rate": 1.6173516693571178e-05, + "loss": 0.9411, + "step": 13140 + }, + { + "epoch": 0.31, + "grad_norm": 2.0920375193356215, + "learning_rate": 1.617291640088408e-05, + "loss": 1.0591, + "step": 13141 + }, + { + "epoch": 0.31, + "grad_norm": 2.0353614324790343, + "learning_rate": 1.617231607225631e-05, + "loss": 1.0662, + "step": 13142 + }, + { + "epoch": 0.31, + "grad_norm": 1.0490833500051715, + "learning_rate": 1.6171715707691366e-05, + "loss": 0.9949, + "step": 13143 + }, + { + "epoch": 0.31, + "grad_norm": 2.0019641010634825, + "learning_rate": 1.6171115307192734e-05, + "loss": 1.0234, + "step": 13144 + }, + { + "epoch": 0.31, + "grad_norm": 1.1087139116560225, + "learning_rate": 1.6170514870763912e-05, + "loss": 0.9445, + "step": 13145 + }, + { + "epoch": 0.31, + "grad_norm": 2.461740712849775, + "learning_rate": 1.6169914398408405e-05, + "loss": 1.0598, + "step": 13146 + }, + { + "epoch": 0.31, + "grad_norm": 2.0064047369887397, + "learning_rate": 1.6169313890129702e-05, + "loss": 1.1605, + "step": 13147 + }, + { + "epoch": 0.31, + "grad_norm": 1.8497209843825815, + "learning_rate": 1.61687133459313e-05, + "loss": 1.0344, + "step": 13148 + }, + { + "epoch": 0.31, + "grad_norm": 1.9102424958520472, + "learning_rate": 1.6168112765816692e-05, + "loss": 1.0635, + "step": 13149 + }, + { + "epoch": 0.31, + "grad_norm": 2.0232396999128044, + "learning_rate": 1.616751214978938e-05, + "loss": 1.0322, + "step": 13150 + }, + { + "epoch": 0.31, + "grad_norm": 2.70142892406946, + "learning_rate": 1.6166911497852864e-05, + "loss": 1.0737, + "step": 13151 + }, + { + "epoch": 0.31, + "grad_norm": 2.06342038392033, + "learning_rate": 1.6166310810010633e-05, + "loss": 1.0716, + "step": 13152 + }, + { + "epoch": 0.31, + "grad_norm": 1.9580809970744073, + "learning_rate": 1.616571008626619e-05, + "loss": 1.0939, + "step": 13153 + }, + { + "epoch": 0.31, + "grad_norm": 1.1739586939607711, + "learning_rate": 1.616510932662303e-05, + "loss": 1.0551, + "step": 13154 + }, + { + "epoch": 0.31, + "grad_norm": 1.9466977129471301, + "learning_rate": 1.6164508531084653e-05, + "loss": 1.1262, + "step": 13155 + }, + { + "epoch": 0.31, + "grad_norm": 2.467981097100525, + "learning_rate": 1.616390769965455e-05, + "loss": 0.9031, + "step": 13156 + }, + { + "epoch": 0.31, + "grad_norm": 2.284246437814783, + "learning_rate": 1.6163306832336232e-05, + "loss": 1.0714, + "step": 13157 + }, + { + "epoch": 0.31, + "grad_norm": 1.813803076815122, + "learning_rate": 1.616270592913319e-05, + "loss": 1.1022, + "step": 13158 + }, + { + "epoch": 0.31, + "grad_norm": 2.0764043747666996, + "learning_rate": 1.616210499004892e-05, + "loss": 1.0133, + "step": 13159 + }, + { + "epoch": 0.31, + "grad_norm": 2.037887337256784, + "learning_rate": 1.6161504015086923e-05, + "loss": 1.032, + "step": 13160 + }, + { + "epoch": 0.31, + "grad_norm": 2.1949377345644057, + "learning_rate": 1.61609030042507e-05, + "loss": 1.1308, + "step": 13161 + }, + { + "epoch": 0.31, + "grad_norm": 2.1384938433763585, + "learning_rate": 1.616030195754375e-05, + "loss": 1.057, + "step": 13162 + }, + { + "epoch": 0.31, + "grad_norm": 2.883215452380276, + "learning_rate": 1.615970087496957e-05, + "loss": 1.0981, + "step": 13163 + }, + { + "epoch": 0.31, + "grad_norm": 2.1609878974107417, + "learning_rate": 1.6159099756531668e-05, + "loss": 1.1568, + "step": 13164 + }, + { + "epoch": 0.31, + "grad_norm": 1.9308954062540022, + "learning_rate": 1.615849860223353e-05, + "loss": 0.9907, + "step": 13165 + }, + { + "epoch": 0.31, + "grad_norm": 2.2187391952676356, + "learning_rate": 1.6157897412078668e-05, + "loss": 1.004, + "step": 13166 + }, + { + "epoch": 0.31, + "grad_norm": 2.388024449432378, + "learning_rate": 1.615729618607057e-05, + "loss": 1.0944, + "step": 13167 + }, + { + "epoch": 0.31, + "grad_norm": 1.8785060030727856, + "learning_rate": 1.615669492421275e-05, + "loss": 0.9259, + "step": 13168 + }, + { + "epoch": 0.31, + "grad_norm": 1.9534473575169553, + "learning_rate": 1.6156093626508706e-05, + "loss": 1.0919, + "step": 13169 + }, + { + "epoch": 0.31, + "grad_norm": 1.949491936688817, + "learning_rate": 1.615549229296193e-05, + "loss": 1.0098, + "step": 13170 + }, + { + "epoch": 0.31, + "grad_norm": 3.0073526750089408, + "learning_rate": 1.6154890923575933e-05, + "loss": 1.0953, + "step": 13171 + }, + { + "epoch": 0.31, + "grad_norm": 1.95426037123774, + "learning_rate": 1.6154289518354208e-05, + "loss": 1.0767, + "step": 13172 + }, + { + "epoch": 0.31, + "grad_norm": 2.032121296137897, + "learning_rate": 1.6153688077300265e-05, + "loss": 0.9765, + "step": 13173 + }, + { + "epoch": 0.31, + "grad_norm": 2.053305348920804, + "learning_rate": 1.6153086600417602e-05, + "loss": 1.0804, + "step": 13174 + }, + { + "epoch": 0.31, + "grad_norm": 1.9757607871033336, + "learning_rate": 1.615248508770972e-05, + "loss": 1.1711, + "step": 13175 + }, + { + "epoch": 0.31, + "grad_norm": 2.2764377286958872, + "learning_rate": 1.615188353918012e-05, + "loss": 1.1349, + "step": 13176 + }, + { + "epoch": 0.31, + "grad_norm": 1.9832321060387486, + "learning_rate": 1.6151281954832307e-05, + "loss": 1.0691, + "step": 13177 + }, + { + "epoch": 0.31, + "grad_norm": 1.9760583853835478, + "learning_rate": 1.6150680334669785e-05, + "loss": 1.0377, + "step": 13178 + }, + { + "epoch": 0.31, + "grad_norm": 2.1085107923814004, + "learning_rate": 1.615007867869605e-05, + "loss": 1.1325, + "step": 13179 + }, + { + "epoch": 0.31, + "grad_norm": 2.267120848223299, + "learning_rate": 1.6149476986914617e-05, + "loss": 0.9433, + "step": 13180 + }, + { + "epoch": 0.31, + "grad_norm": 2.067190387194265, + "learning_rate": 1.6148875259328978e-05, + "loss": 1.05, + "step": 13181 + }, + { + "epoch": 0.31, + "grad_norm": 2.0711022156836822, + "learning_rate": 1.614827349594264e-05, + "loss": 1.0686, + "step": 13182 + }, + { + "epoch": 0.31, + "grad_norm": 1.1508157109473205, + "learning_rate": 1.614767169675911e-05, + "loss": 0.994, + "step": 13183 + }, + { + "epoch": 0.31, + "grad_norm": 2.0066209703481284, + "learning_rate": 1.6147069861781887e-05, + "loss": 0.8922, + "step": 13184 + }, + { + "epoch": 0.31, + "grad_norm": 1.9720712902142796, + "learning_rate": 1.614646799101448e-05, + "loss": 0.9896, + "step": 13185 + }, + { + "epoch": 0.31, + "grad_norm": 1.9698554573902989, + "learning_rate": 1.6145866084460384e-05, + "loss": 0.9629, + "step": 13186 + }, + { + "epoch": 0.31, + "grad_norm": 1.8414659857720856, + "learning_rate": 1.6145264142123114e-05, + "loss": 1.0433, + "step": 13187 + }, + { + "epoch": 0.31, + "grad_norm": 1.9272790492342304, + "learning_rate": 1.6144662164006172e-05, + "loss": 1.0385, + "step": 13188 + }, + { + "epoch": 0.31, + "grad_norm": 2.085153600068265, + "learning_rate": 1.614406015011306e-05, + "loss": 1.1119, + "step": 13189 + }, + { + "epoch": 0.31, + "grad_norm": 1.9038694197845663, + "learning_rate": 1.6143458100447283e-05, + "loss": 0.9905, + "step": 13190 + }, + { + "epoch": 0.31, + "grad_norm": 1.8695037059651876, + "learning_rate": 1.614285601501235e-05, + "loss": 1.1463, + "step": 13191 + }, + { + "epoch": 0.31, + "grad_norm": 1.8930071363205638, + "learning_rate": 1.6142253893811766e-05, + "loss": 0.9869, + "step": 13192 + }, + { + "epoch": 0.31, + "grad_norm": 2.1747881845251493, + "learning_rate": 1.6141651736849032e-05, + "loss": 1.0318, + "step": 13193 + }, + { + "epoch": 0.31, + "grad_norm": 1.8873054293065832, + "learning_rate": 1.6141049544127657e-05, + "loss": 1.0149, + "step": 13194 + }, + { + "epoch": 0.31, + "grad_norm": 1.901731107389521, + "learning_rate": 1.6140447315651152e-05, + "loss": 1.0516, + "step": 13195 + }, + { + "epoch": 0.31, + "grad_norm": 2.0793894344047548, + "learning_rate": 1.6139845051423012e-05, + "loss": 1.1233, + "step": 13196 + }, + { + "epoch": 0.31, + "grad_norm": 1.0972414989474029, + "learning_rate": 1.6139242751446754e-05, + "loss": 0.9831, + "step": 13197 + }, + { + "epoch": 0.31, + "grad_norm": 2.1676977909722135, + "learning_rate": 1.6138640415725882e-05, + "loss": 0.9906, + "step": 13198 + }, + { + "epoch": 0.31, + "grad_norm": 2.036464750096065, + "learning_rate": 1.61380380442639e-05, + "loss": 1.1322, + "step": 13199 + }, + { + "epoch": 0.31, + "grad_norm": 1.906642090939101, + "learning_rate": 1.613743563706432e-05, + "loss": 1.1276, + "step": 13200 + }, + { + "epoch": 0.31, + "grad_norm": 1.896741537488138, + "learning_rate": 1.6136833194130642e-05, + "loss": 1.0281, + "step": 13201 + }, + { + "epoch": 0.31, + "grad_norm": 2.1130208565204263, + "learning_rate": 1.6136230715466384e-05, + "loss": 1.1283, + "step": 13202 + }, + { + "epoch": 0.31, + "grad_norm": 2.178673274733338, + "learning_rate": 1.6135628201075042e-05, + "loss": 1.0637, + "step": 13203 + }, + { + "epoch": 0.31, + "grad_norm": 2.096247410264908, + "learning_rate": 1.6135025650960137e-05, + "loss": 1.0433, + "step": 13204 + }, + { + "epoch": 0.31, + "grad_norm": 2.0508524284044976, + "learning_rate": 1.6134423065125164e-05, + "loss": 0.9529, + "step": 13205 + }, + { + "epoch": 0.31, + "grad_norm": 2.9993408295874064, + "learning_rate": 1.613382044357364e-05, + "loss": 0.9699, + "step": 13206 + }, + { + "epoch": 0.31, + "grad_norm": 2.2224826585730857, + "learning_rate": 1.613321778630907e-05, + "loss": 1.0789, + "step": 13207 + }, + { + "epoch": 0.31, + "grad_norm": 2.494661576661363, + "learning_rate": 1.613261509333497e-05, + "loss": 1.1593, + "step": 13208 + }, + { + "epoch": 0.31, + "grad_norm": 2.027104308251438, + "learning_rate": 1.613201236465484e-05, + "loss": 1.1387, + "step": 13209 + }, + { + "epoch": 0.31, + "grad_norm": 2.1347777571555913, + "learning_rate": 1.6131409600272186e-05, + "loss": 1.0611, + "step": 13210 + }, + { + "epoch": 0.31, + "grad_norm": 3.1919509205936727, + "learning_rate": 1.6130806800190532e-05, + "loss": 1.2197, + "step": 13211 + }, + { + "epoch": 0.31, + "grad_norm": 2.0175451339921397, + "learning_rate": 1.613020396441338e-05, + "loss": 0.9649, + "step": 13212 + }, + { + "epoch": 0.31, + "grad_norm": 1.1144127146748068, + "learning_rate": 1.612960109294424e-05, + "loss": 0.9417, + "step": 13213 + }, + { + "epoch": 0.31, + "grad_norm": 2.2667714058507515, + "learning_rate": 1.6128998185786616e-05, + "loss": 1.153, + "step": 13214 + }, + { + "epoch": 0.31, + "grad_norm": 2.21420211087452, + "learning_rate": 1.6128395242944027e-05, + "loss": 0.9563, + "step": 13215 + }, + { + "epoch": 0.31, + "grad_norm": 1.9398833516875191, + "learning_rate": 1.6127792264419983e-05, + "loss": 1.138, + "step": 13216 + }, + { + "epoch": 0.31, + "grad_norm": 1.9599532400810675, + "learning_rate": 1.6127189250217988e-05, + "loss": 1.052, + "step": 13217 + }, + { + "epoch": 0.31, + "grad_norm": 1.8680740260967346, + "learning_rate": 1.612658620034156e-05, + "loss": 1.0576, + "step": 13218 + }, + { + "epoch": 0.31, + "grad_norm": 1.9504645708634356, + "learning_rate": 1.612598311479421e-05, + "loss": 0.9869, + "step": 13219 + }, + { + "epoch": 0.31, + "grad_norm": 2.8479726772397163, + "learning_rate": 1.6125379993579448e-05, + "loss": 0.9843, + "step": 13220 + }, + { + "epoch": 0.31, + "grad_norm": 2.070137376508675, + "learning_rate": 1.6124776836700778e-05, + "loss": 1.0084, + "step": 13221 + }, + { + "epoch": 0.31, + "grad_norm": 2.3996976203370664, + "learning_rate": 1.612417364416172e-05, + "loss": 0.9287, + "step": 13222 + }, + { + "epoch": 0.31, + "grad_norm": 1.9047860870051023, + "learning_rate": 1.6123570415965785e-05, + "loss": 1.0179, + "step": 13223 + }, + { + "epoch": 0.31, + "grad_norm": 2.02990458088734, + "learning_rate": 1.6122967152116483e-05, + "loss": 1.1646, + "step": 13224 + }, + { + "epoch": 0.31, + "grad_norm": 2.1779404176010346, + "learning_rate": 1.6122363852617333e-05, + "loss": 1.0483, + "step": 13225 + }, + { + "epoch": 0.31, + "grad_norm": 1.890595443993074, + "learning_rate": 1.6121760517471837e-05, + "loss": 0.9406, + "step": 13226 + }, + { + "epoch": 0.31, + "grad_norm": 2.24203190062176, + "learning_rate": 1.6121157146683514e-05, + "loss": 1.0501, + "step": 13227 + }, + { + "epoch": 0.31, + "grad_norm": 1.8999420460641159, + "learning_rate": 1.6120553740255877e-05, + "loss": 0.9255, + "step": 13228 + }, + { + "epoch": 0.31, + "grad_norm": 2.391126559331451, + "learning_rate": 1.6119950298192438e-05, + "loss": 1.0308, + "step": 13229 + }, + { + "epoch": 0.31, + "grad_norm": 2.284378696573524, + "learning_rate": 1.6119346820496712e-05, + "loss": 1.0985, + "step": 13230 + }, + { + "epoch": 0.31, + "grad_norm": 1.4236753889403795, + "learning_rate": 1.6118743307172212e-05, + "loss": 0.9511, + "step": 13231 + }, + { + "epoch": 0.31, + "grad_norm": 2.0207843215652543, + "learning_rate": 1.611813975822245e-05, + "loss": 0.9999, + "step": 13232 + }, + { + "epoch": 0.31, + "grad_norm": 1.9779815940444059, + "learning_rate": 1.611753617365094e-05, + "loss": 0.9249, + "step": 13233 + }, + { + "epoch": 0.31, + "grad_norm": 2.1978912605108047, + "learning_rate": 1.61169325534612e-05, + "loss": 1.0401, + "step": 13234 + }, + { + "epoch": 0.31, + "grad_norm": 2.084188340380651, + "learning_rate": 1.611632889765674e-05, + "loss": 0.9142, + "step": 13235 + }, + { + "epoch": 0.31, + "grad_norm": 1.9742320604153025, + "learning_rate": 1.6115725206241078e-05, + "loss": 1.2479, + "step": 13236 + }, + { + "epoch": 0.31, + "grad_norm": 1.7939505711988728, + "learning_rate": 1.6115121479217724e-05, + "loss": 1.0583, + "step": 13237 + }, + { + "epoch": 0.31, + "grad_norm": 2.022682389182158, + "learning_rate": 1.6114517716590205e-05, + "loss": 1.1144, + "step": 13238 + }, + { + "epoch": 0.31, + "grad_norm": 1.888643698620216, + "learning_rate": 1.6113913918362023e-05, + "loss": 1.025, + "step": 13239 + }, + { + "epoch": 0.31, + "grad_norm": 2.325231576374018, + "learning_rate": 1.6113310084536698e-05, + "loss": 0.9551, + "step": 13240 + }, + { + "epoch": 0.31, + "grad_norm": 1.8500192090853347, + "learning_rate": 1.6112706215117748e-05, + "loss": 1.1108, + "step": 13241 + }, + { + "epoch": 0.31, + "grad_norm": 1.9797232255562065, + "learning_rate": 1.6112102310108686e-05, + "loss": 1.1403, + "step": 13242 + }, + { + "epoch": 0.31, + "grad_norm": 1.9447186582292038, + "learning_rate": 1.611149836951303e-05, + "loss": 1.0283, + "step": 13243 + }, + { + "epoch": 0.31, + "grad_norm": 1.143365856000128, + "learning_rate": 1.6110894393334293e-05, + "loss": 0.987, + "step": 13244 + }, + { + "epoch": 0.31, + "grad_norm": 2.432141867072558, + "learning_rate": 1.6110290381575996e-05, + "loss": 1.0435, + "step": 13245 + }, + { + "epoch": 0.31, + "grad_norm": 1.950336203717413, + "learning_rate": 1.6109686334241655e-05, + "loss": 1.0286, + "step": 13246 + }, + { + "epoch": 0.31, + "grad_norm": 1.9141295732696264, + "learning_rate": 1.6109082251334786e-05, + "loss": 1.0247, + "step": 13247 + }, + { + "epoch": 0.31, + "grad_norm": 2.277995193868803, + "learning_rate": 1.6108478132858903e-05, + "loss": 0.9834, + "step": 13248 + }, + { + "epoch": 0.31, + "grad_norm": 2.125982384965044, + "learning_rate": 1.610787397881753e-05, + "loss": 1.0392, + "step": 13249 + }, + { + "epoch": 0.31, + "grad_norm": 2.2333837817982647, + "learning_rate": 1.610726978921418e-05, + "loss": 0.986, + "step": 13250 + }, + { + "epoch": 0.31, + "grad_norm": 2.0864767007664846, + "learning_rate": 1.610666556405237e-05, + "loss": 1.0038, + "step": 13251 + }, + { + "epoch": 0.31, + "grad_norm": 1.9829311440190096, + "learning_rate": 1.6106061303335622e-05, + "loss": 1.2248, + "step": 13252 + }, + { + "epoch": 0.31, + "grad_norm": 1.9721356300947819, + "learning_rate": 1.610545700706745e-05, + "loss": 1.004, + "step": 13253 + }, + { + "epoch": 0.31, + "grad_norm": 1.7360683394054235, + "learning_rate": 1.6104852675251373e-05, + "loss": 1.0241, + "step": 13254 + }, + { + "epoch": 0.31, + "grad_norm": 2.1357556079600704, + "learning_rate": 1.6104248307890915e-05, + "loss": 0.9881, + "step": 13255 + }, + { + "epoch": 0.31, + "grad_norm": 1.779475221086188, + "learning_rate": 1.6103643904989588e-05, + "loss": 1.0222, + "step": 13256 + }, + { + "epoch": 0.31, + "grad_norm": 1.7961806959805506, + "learning_rate": 1.6103039466550916e-05, + "loss": 1.0457, + "step": 13257 + }, + { + "epoch": 0.31, + "grad_norm": 2.1420044604802304, + "learning_rate": 1.6102434992578415e-05, + "loss": 1.0525, + "step": 13258 + }, + { + "epoch": 0.31, + "grad_norm": 2.133371093108181, + "learning_rate": 1.6101830483075603e-05, + "loss": 1.1613, + "step": 13259 + }, + { + "epoch": 0.31, + "grad_norm": 2.1144717853756223, + "learning_rate": 1.6101225938046007e-05, + "loss": 1.0313, + "step": 13260 + }, + { + "epoch": 0.31, + "grad_norm": 1.887816670439543, + "learning_rate": 1.610062135749314e-05, + "loss": 1.1177, + "step": 13261 + }, + { + "epoch": 0.31, + "grad_norm": 2.196579221642177, + "learning_rate": 1.610001674142052e-05, + "loss": 1.0942, + "step": 13262 + }, + { + "epoch": 0.31, + "grad_norm": 2.30132171007958, + "learning_rate": 1.6099412089831676e-05, + "loss": 0.9819, + "step": 13263 + }, + { + "epoch": 0.31, + "grad_norm": 1.9181216577393658, + "learning_rate": 1.6098807402730124e-05, + "loss": 0.959, + "step": 13264 + }, + { + "epoch": 0.31, + "grad_norm": 2.103135951762571, + "learning_rate": 1.6098202680119382e-05, + "loss": 1.0587, + "step": 13265 + }, + { + "epoch": 0.31, + "grad_norm": 2.5266579959941096, + "learning_rate": 1.6097597922002974e-05, + "loss": 1.0635, + "step": 13266 + }, + { + "epoch": 0.31, + "grad_norm": 2.0417678748343384, + "learning_rate": 1.609699312838442e-05, + "loss": 1.082, + "step": 13267 + }, + { + "epoch": 0.31, + "grad_norm": 1.9964654538724247, + "learning_rate": 1.609638829926724e-05, + "loss": 1.0303, + "step": 13268 + }, + { + "epoch": 0.31, + "grad_norm": 2.0788458392077613, + "learning_rate": 1.6095783434654963e-05, + "loss": 1.021, + "step": 13269 + }, + { + "epoch": 0.31, + "grad_norm": 2.311735818104713, + "learning_rate": 1.60951785345511e-05, + "loss": 0.9839, + "step": 13270 + }, + { + "epoch": 0.31, + "grad_norm": 3.6739890907723836, + "learning_rate": 1.609457359895918e-05, + "loss": 0.9491, + "step": 13271 + }, + { + "epoch": 0.31, + "grad_norm": 1.8637849388854866, + "learning_rate": 1.6093968627882723e-05, + "loss": 1.0495, + "step": 13272 + }, + { + "epoch": 0.31, + "grad_norm": 1.9733208009053331, + "learning_rate": 1.609336362132525e-05, + "loss": 1.1576, + "step": 13273 + }, + { + "epoch": 0.31, + "grad_norm": 2.440713577379924, + "learning_rate": 1.6092758579290286e-05, + "loss": 0.9652, + "step": 13274 + }, + { + "epoch": 0.31, + "grad_norm": 2.0269506796711645, + "learning_rate": 1.6092153501781352e-05, + "loss": 0.9739, + "step": 13275 + }, + { + "epoch": 0.31, + "grad_norm": 1.1741138220064287, + "learning_rate": 1.6091548388801975e-05, + "loss": 0.9934, + "step": 13276 + }, + { + "epoch": 0.31, + "grad_norm": 1.9832981620507184, + "learning_rate": 1.609094324035567e-05, + "loss": 1.0957, + "step": 13277 + }, + { + "epoch": 0.31, + "grad_norm": 2.1072509181330563, + "learning_rate": 1.6090338056445965e-05, + "loss": 1.0495, + "step": 13278 + }, + { + "epoch": 0.31, + "grad_norm": 1.9231538825528578, + "learning_rate": 1.6089732837076387e-05, + "loss": 1.052, + "step": 13279 + }, + { + "epoch": 0.31, + "grad_norm": 2.1330635939112863, + "learning_rate": 1.6089127582250453e-05, + "loss": 1.1147, + "step": 13280 + }, + { + "epoch": 0.31, + "grad_norm": 2.193288610353727, + "learning_rate": 1.6088522291971695e-05, + "loss": 1.1056, + "step": 13281 + }, + { + "epoch": 0.31, + "grad_norm": 2.021463350748759, + "learning_rate": 1.608791696624363e-05, + "loss": 1.0593, + "step": 13282 + }, + { + "epoch": 0.31, + "grad_norm": 2.9022311115613033, + "learning_rate": 1.608731160506978e-05, + "loss": 1.1341, + "step": 13283 + }, + { + "epoch": 0.31, + "grad_norm": 1.12446787413685, + "learning_rate": 1.6086706208453683e-05, + "loss": 0.9619, + "step": 13284 + }, + { + "epoch": 0.31, + "grad_norm": 2.243852261404361, + "learning_rate": 1.608610077639885e-05, + "loss": 1.0218, + "step": 13285 + }, + { + "epoch": 0.31, + "grad_norm": 1.970266199447108, + "learning_rate": 1.6085495308908813e-05, + "loss": 0.9346, + "step": 13286 + }, + { + "epoch": 0.31, + "grad_norm": 1.1235497735658775, + "learning_rate": 1.60848898059871e-05, + "loss": 0.9439, + "step": 13287 + }, + { + "epoch": 0.31, + "grad_norm": 1.156193640622492, + "learning_rate": 1.6084284267637228e-05, + "loss": 1.0452, + "step": 13288 + }, + { + "epoch": 0.31, + "grad_norm": 1.1153575848780979, + "learning_rate": 1.6083678693862725e-05, + "loss": 0.9723, + "step": 13289 + }, + { + "epoch": 0.31, + "grad_norm": 2.366287301354902, + "learning_rate": 1.608307308466712e-05, + "loss": 1.055, + "step": 13290 + }, + { + "epoch": 0.31, + "grad_norm": 2.2369941307940513, + "learning_rate": 1.608246744005394e-05, + "loss": 1.0855, + "step": 13291 + }, + { + "epoch": 0.31, + "grad_norm": 1.0939863432827894, + "learning_rate": 1.6081861760026708e-05, + "loss": 0.9188, + "step": 13292 + }, + { + "epoch": 0.31, + "grad_norm": 1.9184751071646347, + "learning_rate": 1.608125604458895e-05, + "loss": 1.1407, + "step": 13293 + }, + { + "epoch": 0.31, + "grad_norm": 2.119714928849655, + "learning_rate": 1.6080650293744192e-05, + "loss": 1.0292, + "step": 13294 + }, + { + "epoch": 0.31, + "grad_norm": 2.003168828099047, + "learning_rate": 1.6080044507495967e-05, + "loss": 1.0585, + "step": 13295 + }, + { + "epoch": 0.31, + "grad_norm": 1.9450147715249488, + "learning_rate": 1.6079438685847796e-05, + "loss": 1.0875, + "step": 13296 + }, + { + "epoch": 0.31, + "grad_norm": 1.8564805490143683, + "learning_rate": 1.607883282880321e-05, + "loss": 1.1394, + "step": 13297 + }, + { + "epoch": 0.31, + "grad_norm": 1.9957777176021012, + "learning_rate": 1.607822693636573e-05, + "loss": 0.9788, + "step": 13298 + }, + { + "epoch": 0.31, + "grad_norm": 2.0441500573284173, + "learning_rate": 1.607762100853889e-05, + "loss": 1.0468, + "step": 13299 + }, + { + "epoch": 0.31, + "grad_norm": 1.2776471377032983, + "learning_rate": 1.607701504532622e-05, + "loss": 1.007, + "step": 13300 + }, + { + "epoch": 0.31, + "grad_norm": 2.2621164185906397, + "learning_rate": 1.6076409046731243e-05, + "loss": 1.0392, + "step": 13301 + }, + { + "epoch": 0.31, + "grad_norm": 2.421623766754406, + "learning_rate": 1.607580301275749e-05, + "loss": 1.0166, + "step": 13302 + }, + { + "epoch": 0.31, + "grad_norm": 2.148033370025672, + "learning_rate": 1.6075196943408485e-05, + "loss": 1.0014, + "step": 13303 + }, + { + "epoch": 0.31, + "grad_norm": 2.1369031802233223, + "learning_rate": 1.607459083868776e-05, + "loss": 0.8979, + "step": 13304 + }, + { + "epoch": 0.31, + "grad_norm": 2.155143299157904, + "learning_rate": 1.6073984698598845e-05, + "loss": 0.9977, + "step": 13305 + }, + { + "epoch": 0.31, + "grad_norm": 1.9560299089512578, + "learning_rate": 1.6073378523145272e-05, + "loss": 1.0314, + "step": 13306 + }, + { + "epoch": 0.31, + "grad_norm": 1.8641239359730777, + "learning_rate": 1.607277231233056e-05, + "loss": 1.1234, + "step": 13307 + }, + { + "epoch": 0.31, + "grad_norm": 2.044602262328176, + "learning_rate": 1.607216606615825e-05, + "loss": 1.1352, + "step": 13308 + }, + { + "epoch": 0.31, + "grad_norm": 1.9345283923789667, + "learning_rate": 1.6071559784631862e-05, + "loss": 0.8882, + "step": 13309 + }, + { + "epoch": 0.31, + "grad_norm": 1.9680250119436944, + "learning_rate": 1.6070953467754935e-05, + "loss": 1.0777, + "step": 13310 + }, + { + "epoch": 0.31, + "grad_norm": 1.8434836583700849, + "learning_rate": 1.6070347115530994e-05, + "loss": 0.9763, + "step": 13311 + }, + { + "epoch": 0.31, + "grad_norm": 1.9246703775519582, + "learning_rate": 1.606974072796357e-05, + "loss": 0.9596, + "step": 13312 + }, + { + "epoch": 0.31, + "grad_norm": 2.251907406791171, + "learning_rate": 1.6069134305056193e-05, + "loss": 1.0294, + "step": 13313 + }, + { + "epoch": 0.31, + "grad_norm": 2.193611030268548, + "learning_rate": 1.6068527846812398e-05, + "loss": 0.9816, + "step": 13314 + }, + { + "epoch": 0.31, + "grad_norm": 1.9516010520633216, + "learning_rate": 1.6067921353235707e-05, + "loss": 1.0506, + "step": 13315 + }, + { + "epoch": 0.31, + "grad_norm": 2.3109398772750236, + "learning_rate": 1.606731482432966e-05, + "loss": 1.1665, + "step": 13316 + }, + { + "epoch": 0.31, + "grad_norm": 2.0508922004205257, + "learning_rate": 1.6066708260097782e-05, + "loss": 1.1123, + "step": 13317 + }, + { + "epoch": 0.31, + "grad_norm": 2.0216497101411743, + "learning_rate": 1.6066101660543612e-05, + "loss": 1.1402, + "step": 13318 + }, + { + "epoch": 0.31, + "grad_norm": 1.9907374748130373, + "learning_rate": 1.6065495025670675e-05, + "loss": 1.0185, + "step": 13319 + }, + { + "epoch": 0.31, + "grad_norm": 2.0352328803385435, + "learning_rate": 1.606488835548251e-05, + "loss": 1.0797, + "step": 13320 + }, + { + "epoch": 0.31, + "grad_norm": 1.8965907170174556, + "learning_rate": 1.6064281649982637e-05, + "loss": 1.1255, + "step": 13321 + }, + { + "epoch": 0.31, + "grad_norm": 1.9272825229430826, + "learning_rate": 1.6063674909174604e-05, + "loss": 1.0914, + "step": 13322 + }, + { + "epoch": 0.31, + "grad_norm": 2.1929867663397657, + "learning_rate": 1.606306813306193e-05, + "loss": 1.0158, + "step": 13323 + }, + { + "epoch": 0.31, + "grad_norm": 2.9840965572533817, + "learning_rate": 1.606246132164816e-05, + "loss": 1.0055, + "step": 13324 + }, + { + "epoch": 0.31, + "grad_norm": 1.9424263872319276, + "learning_rate": 1.606185447493682e-05, + "loss": 1.1519, + "step": 13325 + }, + { + "epoch": 0.31, + "grad_norm": 2.1128459052764224, + "learning_rate": 1.606124759293144e-05, + "loss": 1.0186, + "step": 13326 + }, + { + "epoch": 0.31, + "grad_norm": 1.8214775278131257, + "learning_rate": 1.606064067563556e-05, + "loss": 1.0383, + "step": 13327 + }, + { + "epoch": 0.31, + "grad_norm": 2.0691977252682783, + "learning_rate": 1.6060033723052707e-05, + "loss": 1.1738, + "step": 13328 + }, + { + "epoch": 0.31, + "grad_norm": 2.041829700882996, + "learning_rate": 1.6059426735186426e-05, + "loss": 1.0121, + "step": 13329 + }, + { + "epoch": 0.31, + "grad_norm": 2.0425605794123314, + "learning_rate": 1.6058819712040237e-05, + "loss": 1.1286, + "step": 13330 + }, + { + "epoch": 0.31, + "grad_norm": 2.0020142870957316, + "learning_rate": 1.6058212653617688e-05, + "loss": 1.0689, + "step": 13331 + }, + { + "epoch": 0.31, + "grad_norm": 1.9433555195640773, + "learning_rate": 1.60576055599223e-05, + "loss": 1.0186, + "step": 13332 + }, + { + "epoch": 0.31, + "grad_norm": 2.0923394124931733, + "learning_rate": 1.605699843095762e-05, + "loss": 0.9757, + "step": 13333 + }, + { + "epoch": 0.31, + "grad_norm": 2.3030192400800216, + "learning_rate": 1.6056391266727178e-05, + "loss": 0.9009, + "step": 13334 + }, + { + "epoch": 0.31, + "grad_norm": 2.227907634029692, + "learning_rate": 1.6055784067234507e-05, + "loss": 1.0353, + "step": 13335 + }, + { + "epoch": 0.31, + "grad_norm": 2.1738395572264917, + "learning_rate": 1.6055176832483143e-05, + "loss": 1.0058, + "step": 13336 + }, + { + "epoch": 0.31, + "grad_norm": 2.396228852538927, + "learning_rate": 1.6054569562476622e-05, + "loss": 1.1383, + "step": 13337 + }, + { + "epoch": 0.31, + "grad_norm": 1.158364187463314, + "learning_rate": 1.6053962257218483e-05, + "loss": 0.9602, + "step": 13338 + }, + { + "epoch": 0.31, + "grad_norm": 1.1932588876786971, + "learning_rate": 1.6053354916712258e-05, + "loss": 1.0016, + "step": 13339 + }, + { + "epoch": 0.31, + "grad_norm": 1.8586463505400868, + "learning_rate": 1.6052747540961483e-05, + "loss": 0.9333, + "step": 13340 + }, + { + "epoch": 0.31, + "grad_norm": 2.0389580297929473, + "learning_rate": 1.60521401299697e-05, + "loss": 1.011, + "step": 13341 + }, + { + "epoch": 0.31, + "grad_norm": 2.387501676680409, + "learning_rate": 1.6051532683740433e-05, + "loss": 1.0757, + "step": 13342 + }, + { + "epoch": 0.31, + "grad_norm": 2.0570459113683333, + "learning_rate": 1.6050925202277232e-05, + "loss": 0.9854, + "step": 13343 + }, + { + "epoch": 0.31, + "grad_norm": 2.0948845309157083, + "learning_rate": 1.6050317685583627e-05, + "loss": 1.0605, + "step": 13344 + }, + { + "epoch": 0.31, + "grad_norm": 1.9855014429686513, + "learning_rate": 1.604971013366316e-05, + "loss": 0.9371, + "step": 13345 + }, + { + "epoch": 0.31, + "grad_norm": 1.9247470819997041, + "learning_rate": 1.6049102546519363e-05, + "loss": 0.9691, + "step": 13346 + }, + { + "epoch": 0.31, + "grad_norm": 1.7951633513529552, + "learning_rate": 1.6048494924155776e-05, + "loss": 0.9723, + "step": 13347 + }, + { + "epoch": 0.31, + "grad_norm": 2.0547041295152804, + "learning_rate": 1.6047887266575936e-05, + "loss": 1.116, + "step": 13348 + }, + { + "epoch": 0.31, + "grad_norm": 1.9321179261906622, + "learning_rate": 1.6047279573783382e-05, + "loss": 1.0102, + "step": 13349 + }, + { + "epoch": 0.31, + "grad_norm": 1.975562718894217, + "learning_rate": 1.6046671845781653e-05, + "loss": 1.0985, + "step": 13350 + }, + { + "epoch": 0.31, + "grad_norm": 1.9677290244095427, + "learning_rate": 1.6046064082574284e-05, + "loss": 1.0951, + "step": 13351 + }, + { + "epoch": 0.31, + "grad_norm": 2.2100351244151564, + "learning_rate": 1.604545628416482e-05, + "loss": 1.0682, + "step": 13352 + }, + { + "epoch": 0.31, + "grad_norm": 1.943680380243022, + "learning_rate": 1.6044848450556788e-05, + "loss": 0.9635, + "step": 13353 + }, + { + "epoch": 0.31, + "grad_norm": 1.0907133335121881, + "learning_rate": 1.6044240581753742e-05, + "loss": 0.9766, + "step": 13354 + }, + { + "epoch": 0.31, + "grad_norm": 2.2921501760639207, + "learning_rate": 1.604363267775921e-05, + "loss": 0.9373, + "step": 13355 + }, + { + "epoch": 0.31, + "grad_norm": 1.8240160681427304, + "learning_rate": 1.604302473857674e-05, + "loss": 0.872, + "step": 13356 + }, + { + "epoch": 0.31, + "grad_norm": 2.2554209720788916, + "learning_rate": 1.604241676420986e-05, + "loss": 0.851, + "step": 13357 + }, + { + "epoch": 0.31, + "grad_norm": 2.70113608887551, + "learning_rate": 1.604180875466212e-05, + "loss": 1.0333, + "step": 13358 + }, + { + "epoch": 0.31, + "grad_norm": 2.0007605099757475, + "learning_rate": 1.6041200709937058e-05, + "loss": 1.0309, + "step": 13359 + }, + { + "epoch": 0.31, + "grad_norm": 2.1853842002532633, + "learning_rate": 1.604059263003821e-05, + "loss": 1.1826, + "step": 13360 + }, + { + "epoch": 0.31, + "grad_norm": 3.8689685985001736, + "learning_rate": 1.603998451496912e-05, + "loss": 1.01, + "step": 13361 + }, + { + "epoch": 0.31, + "grad_norm": 2.176678847615145, + "learning_rate": 1.6039376364733328e-05, + "loss": 1.0017, + "step": 13362 + }, + { + "epoch": 0.31, + "grad_norm": 2.162994873571255, + "learning_rate": 1.6038768179334373e-05, + "loss": 0.9299, + "step": 13363 + }, + { + "epoch": 0.31, + "grad_norm": 2.5172769904286314, + "learning_rate": 1.60381599587758e-05, + "loss": 0.9476, + "step": 13364 + }, + { + "epoch": 0.31, + "grad_norm": 1.0980763747087343, + "learning_rate": 1.6037551703061148e-05, + "loss": 0.9759, + "step": 13365 + }, + { + "epoch": 0.31, + "grad_norm": 2.1001159801983103, + "learning_rate": 1.603694341219396e-05, + "loss": 1.1422, + "step": 13366 + }, + { + "epoch": 0.31, + "grad_norm": 2.0052401251667176, + "learning_rate": 1.6036335086177772e-05, + "loss": 1.0564, + "step": 13367 + }, + { + "epoch": 0.31, + "grad_norm": 2.2480937470654423, + "learning_rate": 1.6035726725016134e-05, + "loss": 0.8924, + "step": 13368 + }, + { + "epoch": 0.31, + "grad_norm": 2.038722777064655, + "learning_rate": 1.6035118328712583e-05, + "loss": 1.0082, + "step": 13369 + }, + { + "epoch": 0.31, + "grad_norm": 2.3288180598571553, + "learning_rate": 1.603450989727066e-05, + "loss": 1.1327, + "step": 13370 + }, + { + "epoch": 0.32, + "grad_norm": 1.0909436878846623, + "learning_rate": 1.603390143069391e-05, + "loss": 1.0123, + "step": 13371 + }, + { + "epoch": 0.32, + "grad_norm": 2.020671272764214, + "learning_rate": 1.603329292898588e-05, + "loss": 0.9793, + "step": 13372 + }, + { + "epoch": 0.32, + "grad_norm": 1.9558419696615474, + "learning_rate": 1.6032684392150108e-05, + "loss": 1.0465, + "step": 13373 + }, + { + "epoch": 0.32, + "grad_norm": 1.0284612323380624, + "learning_rate": 1.6032075820190134e-05, + "loss": 0.9581, + "step": 13374 + }, + { + "epoch": 0.32, + "grad_norm": 1.9278886728703513, + "learning_rate": 1.6031467213109505e-05, + "loss": 1.0695, + "step": 13375 + }, + { + "epoch": 0.32, + "grad_norm": 2.28055193265369, + "learning_rate": 1.6030858570911768e-05, + "loss": 0.9305, + "step": 13376 + }, + { + "epoch": 0.32, + "grad_norm": 2.342505777826367, + "learning_rate": 1.603024989360046e-05, + "loss": 0.9355, + "step": 13377 + }, + { + "epoch": 0.32, + "grad_norm": 3.1841385379627365, + "learning_rate": 1.602964118117913e-05, + "loss": 1.1356, + "step": 13378 + }, + { + "epoch": 0.32, + "grad_norm": 2.281876320870544, + "learning_rate": 1.6029032433651317e-05, + "loss": 0.8847, + "step": 13379 + }, + { + "epoch": 0.32, + "grad_norm": 2.0421973299406115, + "learning_rate": 1.6028423651020572e-05, + "loss": 1.155, + "step": 13380 + }, + { + "epoch": 0.32, + "grad_norm": 4.010164028086236, + "learning_rate": 1.6027814833290435e-05, + "loss": 0.97, + "step": 13381 + }, + { + "epoch": 0.32, + "grad_norm": 2.6140809537252743, + "learning_rate": 1.602720598046445e-05, + "loss": 0.8529, + "step": 13382 + }, + { + "epoch": 0.32, + "grad_norm": 4.748836683981976, + "learning_rate": 1.6026597092546164e-05, + "loss": 1.0683, + "step": 13383 + }, + { + "epoch": 0.32, + "grad_norm": 1.9281428682956694, + "learning_rate": 1.6025988169539122e-05, + "loss": 1.0577, + "step": 13384 + }, + { + "epoch": 0.32, + "grad_norm": 2.4182381398030026, + "learning_rate": 1.602537921144687e-05, + "loss": 1.1459, + "step": 13385 + }, + { + "epoch": 0.32, + "grad_norm": 1.7771259947362845, + "learning_rate": 1.602477021827295e-05, + "loss": 1.1618, + "step": 13386 + }, + { + "epoch": 0.32, + "grad_norm": 2.113466397438514, + "learning_rate": 1.6024161190020913e-05, + "loss": 1.1297, + "step": 13387 + }, + { + "epoch": 0.32, + "grad_norm": 2.0052951959068777, + "learning_rate": 1.60235521266943e-05, + "loss": 1.0728, + "step": 13388 + }, + { + "epoch": 0.32, + "grad_norm": 2.044060440028905, + "learning_rate": 1.6022943028296663e-05, + "loss": 1.0444, + "step": 13389 + }, + { + "epoch": 0.32, + "grad_norm": 1.9890537470971839, + "learning_rate": 1.6022333894831543e-05, + "loss": 1.0331, + "step": 13390 + }, + { + "epoch": 0.32, + "grad_norm": 2.485860034551167, + "learning_rate": 1.6021724726302487e-05, + "loss": 1.0474, + "step": 13391 + }, + { + "epoch": 0.32, + "grad_norm": 1.1423381993671946, + "learning_rate": 1.6021115522713047e-05, + "loss": 0.9874, + "step": 13392 + }, + { + "epoch": 0.32, + "grad_norm": 2.1279403943164046, + "learning_rate": 1.6020506284066763e-05, + "loss": 1.1242, + "step": 13393 + }, + { + "epoch": 0.32, + "grad_norm": 3.084882830388473, + "learning_rate": 1.6019897010367186e-05, + "loss": 1.0063, + "step": 13394 + }, + { + "epoch": 0.32, + "grad_norm": 2.0897712884355095, + "learning_rate": 1.6019287701617863e-05, + "loss": 1.19, + "step": 13395 + }, + { + "epoch": 0.32, + "grad_norm": 2.2044281858018793, + "learning_rate": 1.6018678357822343e-05, + "loss": 0.8794, + "step": 13396 + }, + { + "epoch": 0.32, + "grad_norm": 1.872245864417672, + "learning_rate": 1.601806897898417e-05, + "loss": 1.0543, + "step": 13397 + }, + { + "epoch": 0.32, + "grad_norm": 1.069989415598193, + "learning_rate": 1.6017459565106896e-05, + "loss": 0.9426, + "step": 13398 + }, + { + "epoch": 0.32, + "grad_norm": 2.324073570289248, + "learning_rate": 1.601685011619406e-05, + "loss": 1.0409, + "step": 13399 + }, + { + "epoch": 0.32, + "grad_norm": 1.8504861246130968, + "learning_rate": 1.6016240632249224e-05, + "loss": 0.9758, + "step": 13400 + }, + { + "epoch": 0.32, + "grad_norm": 1.12417219380718, + "learning_rate": 1.601563111327593e-05, + "loss": 1.0159, + "step": 13401 + }, + { + "epoch": 0.32, + "grad_norm": 1.9162377120377225, + "learning_rate": 1.601502155927773e-05, + "loss": 1.0457, + "step": 13402 + }, + { + "epoch": 0.32, + "grad_norm": 2.104171132061675, + "learning_rate": 1.6014411970258166e-05, + "loss": 1.0977, + "step": 13403 + }, + { + "epoch": 0.32, + "grad_norm": 1.9957750368331426, + "learning_rate": 1.601380234622079e-05, + "loss": 1.0465, + "step": 13404 + }, + { + "epoch": 0.32, + "grad_norm": 1.1040260273920208, + "learning_rate": 1.601319268716916e-05, + "loss": 0.9087, + "step": 13405 + }, + { + "epoch": 0.32, + "grad_norm": 2.0366762285147932, + "learning_rate": 1.6012582993106812e-05, + "loss": 1.1002, + "step": 13406 + }, + { + "epoch": 0.32, + "grad_norm": 1.0824549147678664, + "learning_rate": 1.6011973264037305e-05, + "loss": 0.9955, + "step": 13407 + }, + { + "epoch": 0.32, + "grad_norm": 1.1605719109116728, + "learning_rate": 1.6011363499964185e-05, + "loss": 0.9218, + "step": 13408 + }, + { + "epoch": 0.32, + "grad_norm": 1.8835569767000677, + "learning_rate": 1.6010753700891003e-05, + "loss": 0.8883, + "step": 13409 + }, + { + "epoch": 0.32, + "grad_norm": 1.8772266777325175, + "learning_rate": 1.601014386682131e-05, + "loss": 1.0012, + "step": 13410 + }, + { + "epoch": 0.32, + "grad_norm": 2.122641570788477, + "learning_rate": 1.600953399775866e-05, + "loss": 1.1215, + "step": 13411 + }, + { + "epoch": 0.32, + "grad_norm": 2.066029155144697, + "learning_rate": 1.60089240937066e-05, + "loss": 1.0524, + "step": 13412 + }, + { + "epoch": 0.32, + "grad_norm": 2.2285572388401143, + "learning_rate": 1.6008314154668676e-05, + "loss": 0.9368, + "step": 13413 + }, + { + "epoch": 0.32, + "grad_norm": 1.935624158367221, + "learning_rate": 1.6007704180648448e-05, + "loss": 1.0164, + "step": 13414 + }, + { + "epoch": 0.32, + "grad_norm": 1.8118856512636123, + "learning_rate": 1.6007094171649463e-05, + "loss": 0.9738, + "step": 13415 + }, + { + "epoch": 0.32, + "grad_norm": 2.165586904837744, + "learning_rate": 1.6006484127675278e-05, + "loss": 0.9698, + "step": 13416 + }, + { + "epoch": 0.32, + "grad_norm": 2.1641553058847913, + "learning_rate": 1.6005874048729437e-05, + "loss": 1.143, + "step": 13417 + }, + { + "epoch": 0.32, + "grad_norm": 2.2661976793416376, + "learning_rate": 1.6005263934815498e-05, + "loss": 1.075, + "step": 13418 + }, + { + "epoch": 0.32, + "grad_norm": 1.804354637998654, + "learning_rate": 1.600465378593701e-05, + "loss": 0.9938, + "step": 13419 + }, + { + "epoch": 0.32, + "grad_norm": 1.8478866255468382, + "learning_rate": 1.6004043602097527e-05, + "loss": 1.057, + "step": 13420 + }, + { + "epoch": 0.32, + "grad_norm": 1.8535096240147482, + "learning_rate": 1.6003433383300603e-05, + "loss": 1.0688, + "step": 13421 + }, + { + "epoch": 0.32, + "grad_norm": 2.330993062574358, + "learning_rate": 1.6002823129549784e-05, + "loss": 0.9535, + "step": 13422 + }, + { + "epoch": 0.32, + "grad_norm": 2.1610055083660904, + "learning_rate": 1.600221284084863e-05, + "loss": 0.9801, + "step": 13423 + }, + { + "epoch": 0.32, + "grad_norm": 2.499208640506369, + "learning_rate": 1.6001602517200698e-05, + "loss": 1.0197, + "step": 13424 + }, + { + "epoch": 0.32, + "grad_norm": 2.2111387234954094, + "learning_rate": 1.600099215860953e-05, + "loss": 0.9838, + "step": 13425 + }, + { + "epoch": 0.32, + "grad_norm": 2.156027700576848, + "learning_rate": 1.600038176507869e-05, + "loss": 0.9681, + "step": 13426 + }, + { + "epoch": 0.32, + "grad_norm": 1.8104944896038053, + "learning_rate": 1.5999771336611726e-05, + "loss": 0.9221, + "step": 13427 + }, + { + "epoch": 0.32, + "grad_norm": 2.0160886298077063, + "learning_rate": 1.5999160873212193e-05, + "loss": 0.9685, + "step": 13428 + }, + { + "epoch": 0.32, + "grad_norm": 5.8176344497631085, + "learning_rate": 1.599855037488365e-05, + "loss": 1.0035, + "step": 13429 + }, + { + "epoch": 0.32, + "grad_norm": 2.3772756053932627, + "learning_rate": 1.599793984162964e-05, + "loss": 1.1102, + "step": 13430 + }, + { + "epoch": 0.32, + "grad_norm": 1.9222962572701237, + "learning_rate": 1.599732927345373e-05, + "loss": 1.1799, + "step": 13431 + }, + { + "epoch": 0.32, + "grad_norm": 2.0485677800306514, + "learning_rate": 1.599671867035947e-05, + "loss": 1.0716, + "step": 13432 + }, + { + "epoch": 0.32, + "grad_norm": 2.287767527022717, + "learning_rate": 1.5996108032350414e-05, + "loss": 1.0694, + "step": 13433 + }, + { + "epoch": 0.32, + "grad_norm": 2.1422158935799454, + "learning_rate": 1.599549735943012e-05, + "loss": 1.017, + "step": 13434 + }, + { + "epoch": 0.32, + "grad_norm": 2.5184390215570174, + "learning_rate": 1.599488665160214e-05, + "loss": 1.1215, + "step": 13435 + }, + { + "epoch": 0.32, + "grad_norm": 2.590983826807916, + "learning_rate": 1.5994275908870037e-05, + "loss": 1.1292, + "step": 13436 + }, + { + "epoch": 0.32, + "grad_norm": 1.9563349567306383, + "learning_rate": 1.599366513123736e-05, + "loss": 1.1237, + "step": 13437 + }, + { + "epoch": 0.32, + "grad_norm": 2.23102003757059, + "learning_rate": 1.5993054318707665e-05, + "loss": 1.0757, + "step": 13438 + }, + { + "epoch": 0.32, + "grad_norm": 1.732591436742164, + "learning_rate": 1.599244347128451e-05, + "loss": 1.025, + "step": 13439 + }, + { + "epoch": 0.32, + "grad_norm": 1.9527120579252961, + "learning_rate": 1.5991832588971454e-05, + "loss": 1.0349, + "step": 13440 + }, + { + "epoch": 0.32, + "grad_norm": 2.1151093449761493, + "learning_rate": 1.5991221671772053e-05, + "loss": 0.974, + "step": 13441 + }, + { + "epoch": 0.32, + "grad_norm": 1.8951837301562773, + "learning_rate": 1.5990610719689858e-05, + "loss": 1.0207, + "step": 13442 + }, + { + "epoch": 0.32, + "grad_norm": 2.1084264808157362, + "learning_rate": 1.5989999732728434e-05, + "loss": 1.0591, + "step": 13443 + }, + { + "epoch": 0.32, + "grad_norm": 1.9338843052079495, + "learning_rate": 1.5989388710891334e-05, + "loss": 0.9362, + "step": 13444 + }, + { + "epoch": 0.32, + "grad_norm": 4.0751827052625496, + "learning_rate": 1.5988777654182118e-05, + "loss": 1.0346, + "step": 13445 + }, + { + "epoch": 0.32, + "grad_norm": 2.3140234653012666, + "learning_rate": 1.5988166562604343e-05, + "loss": 0.9846, + "step": 13446 + }, + { + "epoch": 0.32, + "grad_norm": 1.9478279542223829, + "learning_rate": 1.5987555436161564e-05, + "loss": 0.9502, + "step": 13447 + }, + { + "epoch": 0.32, + "grad_norm": 2.1540787625115714, + "learning_rate": 1.5986944274857344e-05, + "loss": 1.0189, + "step": 13448 + }, + { + "epoch": 0.32, + "grad_norm": 2.111375958371062, + "learning_rate": 1.598633307869524e-05, + "loss": 1.1499, + "step": 13449 + }, + { + "epoch": 0.32, + "grad_norm": 2.007237314195633, + "learning_rate": 1.5985721847678806e-05, + "loss": 1.2641, + "step": 13450 + }, + { + "epoch": 0.32, + "grad_norm": 2.887925524157605, + "learning_rate": 1.59851105818116e-05, + "loss": 1.0767, + "step": 13451 + }, + { + "epoch": 0.32, + "grad_norm": 1.9865573079892853, + "learning_rate": 1.5984499281097194e-05, + "loss": 1.1742, + "step": 13452 + }, + { + "epoch": 0.32, + "grad_norm": 2.220529763392324, + "learning_rate": 1.5983887945539138e-05, + "loss": 1.1192, + "step": 13453 + }, + { + "epoch": 0.32, + "grad_norm": 1.7786265192872879, + "learning_rate": 1.5983276575140986e-05, + "loss": 1.0697, + "step": 13454 + }, + { + "epoch": 0.32, + "grad_norm": 1.938524956303301, + "learning_rate": 1.598266516990631e-05, + "loss": 1.0317, + "step": 13455 + }, + { + "epoch": 0.32, + "grad_norm": 2.4484405187128404, + "learning_rate": 1.5982053729838657e-05, + "loss": 0.9096, + "step": 13456 + }, + { + "epoch": 0.32, + "grad_norm": 2.892265105278621, + "learning_rate": 1.59814422549416e-05, + "loss": 1.0503, + "step": 13457 + }, + { + "epoch": 0.32, + "grad_norm": 1.9124746961172154, + "learning_rate": 1.5980830745218686e-05, + "loss": 1.0043, + "step": 13458 + }, + { + "epoch": 0.32, + "grad_norm": 1.911613815118372, + "learning_rate": 1.5980219200673488e-05, + "loss": 1.1267, + "step": 13459 + }, + { + "epoch": 0.32, + "grad_norm": 1.9624372163247756, + "learning_rate": 1.5979607621309557e-05, + "loss": 1.0015, + "step": 13460 + }, + { + "epoch": 0.32, + "grad_norm": 2.362321067361393, + "learning_rate": 1.597899600713046e-05, + "loss": 1.0049, + "step": 13461 + }, + { + "epoch": 0.32, + "grad_norm": 2.020516521903857, + "learning_rate": 1.5978384358139752e-05, + "loss": 1.0513, + "step": 13462 + }, + { + "epoch": 0.32, + "grad_norm": 2.0298502062573083, + "learning_rate": 1.5977772674340996e-05, + "loss": 1.0973, + "step": 13463 + }, + { + "epoch": 0.32, + "grad_norm": 2.0114574157307814, + "learning_rate": 1.5977160955737758e-05, + "loss": 1.0416, + "step": 13464 + }, + { + "epoch": 0.32, + "grad_norm": 1.0910491436532739, + "learning_rate": 1.59765492023336e-05, + "loss": 0.9517, + "step": 13465 + }, + { + "epoch": 0.32, + "grad_norm": 2.037073773857524, + "learning_rate": 1.5975937414132076e-05, + "loss": 1.0505, + "step": 13466 + }, + { + "epoch": 0.32, + "grad_norm": 1.9316636122472657, + "learning_rate": 1.5975325591136753e-05, + "loss": 1.1862, + "step": 13467 + }, + { + "epoch": 0.32, + "grad_norm": 2.2025426486406343, + "learning_rate": 1.5974713733351192e-05, + "loss": 0.9883, + "step": 13468 + }, + { + "epoch": 0.32, + "grad_norm": 1.1802176456065268, + "learning_rate": 1.5974101840778954e-05, + "loss": 1.067, + "step": 13469 + }, + { + "epoch": 0.32, + "grad_norm": 1.72499091822045, + "learning_rate": 1.597348991342361e-05, + "loss": 1.1242, + "step": 13470 + }, + { + "epoch": 0.32, + "grad_norm": 1.9556315977919658, + "learning_rate": 1.597287795128871e-05, + "loss": 1.0723, + "step": 13471 + }, + { + "epoch": 0.32, + "grad_norm": 2.0517895042235987, + "learning_rate": 1.597226595437783e-05, + "loss": 1.0687, + "step": 13472 + }, + { + "epoch": 0.32, + "grad_norm": 1.8907026250820522, + "learning_rate": 1.597165392269452e-05, + "loss": 1.0444, + "step": 13473 + }, + { + "epoch": 0.32, + "grad_norm": 2.1204577430102884, + "learning_rate": 1.5971041856242356e-05, + "loss": 1.033, + "step": 13474 + }, + { + "epoch": 0.32, + "grad_norm": 1.0599870697168523, + "learning_rate": 1.5970429755024893e-05, + "loss": 0.9713, + "step": 13475 + }, + { + "epoch": 0.32, + "grad_norm": 1.8709383568515665, + "learning_rate": 1.59698176190457e-05, + "loss": 1.0907, + "step": 13476 + }, + { + "epoch": 0.32, + "grad_norm": 1.9409669142975412, + "learning_rate": 1.5969205448308335e-05, + "loss": 1.0475, + "step": 13477 + }, + { + "epoch": 0.32, + "grad_norm": 1.1199729584624925, + "learning_rate": 1.596859324281637e-05, + "loss": 0.9868, + "step": 13478 + }, + { + "epoch": 0.32, + "grad_norm": 1.7880482590245768, + "learning_rate": 1.5967981002573362e-05, + "loss": 1.0968, + "step": 13479 + }, + { + "epoch": 0.32, + "grad_norm": 1.0973698006389014, + "learning_rate": 1.5967368727582882e-05, + "loss": 0.9777, + "step": 13480 + }, + { + "epoch": 0.32, + "grad_norm": 1.9077523654736368, + "learning_rate": 1.5966756417848492e-05, + "loss": 1.093, + "step": 13481 + }, + { + "epoch": 0.32, + "grad_norm": 1.9410823884334552, + "learning_rate": 1.5966144073373755e-05, + "loss": 0.9999, + "step": 13482 + }, + { + "epoch": 0.32, + "grad_norm": 1.1479788343973276, + "learning_rate": 1.596553169416224e-05, + "loss": 1.0052, + "step": 13483 + }, + { + "epoch": 0.32, + "grad_norm": 1.8959013999744085, + "learning_rate": 1.596491928021751e-05, + "loss": 0.9458, + "step": 13484 + }, + { + "epoch": 0.32, + "grad_norm": 1.8985949008180347, + "learning_rate": 1.5964306831543132e-05, + "loss": 1.0464, + "step": 13485 + }, + { + "epoch": 0.32, + "grad_norm": 1.7809343198144003, + "learning_rate": 1.5963694348142672e-05, + "loss": 0.9886, + "step": 13486 + }, + { + "epoch": 0.32, + "grad_norm": 1.1743007307929596, + "learning_rate": 1.5963081830019692e-05, + "loss": 1.0495, + "step": 13487 + }, + { + "epoch": 0.32, + "grad_norm": 2.1058116154401874, + "learning_rate": 1.5962469277177766e-05, + "loss": 1.0247, + "step": 13488 + }, + { + "epoch": 0.32, + "grad_norm": 2.554926574435909, + "learning_rate": 1.5961856689620453e-05, + "loss": 1.0065, + "step": 13489 + }, + { + "epoch": 0.32, + "grad_norm": 2.017496949457271, + "learning_rate": 1.5961244067351326e-05, + "loss": 0.9889, + "step": 13490 + }, + { + "epoch": 0.32, + "grad_norm": 1.0723298467015223, + "learning_rate": 1.5960631410373947e-05, + "loss": 0.977, + "step": 13491 + }, + { + "epoch": 0.32, + "grad_norm": 2.686622554360871, + "learning_rate": 1.5960018718691886e-05, + "loss": 1.1462, + "step": 13492 + }, + { + "epoch": 0.32, + "grad_norm": 2.568350614326098, + "learning_rate": 1.5959405992308705e-05, + "loss": 1.0259, + "step": 13493 + }, + { + "epoch": 0.32, + "grad_norm": 2.028086352870961, + "learning_rate": 1.595879323122798e-05, + "loss": 1.1486, + "step": 13494 + }, + { + "epoch": 0.32, + "grad_norm": 2.166353259696404, + "learning_rate": 1.5958180435453273e-05, + "loss": 1.1354, + "step": 13495 + }, + { + "epoch": 0.32, + "grad_norm": 1.8367703743861197, + "learning_rate": 1.5957567604988154e-05, + "loss": 0.9927, + "step": 13496 + }, + { + "epoch": 0.32, + "grad_norm": 2.1862182491518363, + "learning_rate": 1.595695473983619e-05, + "loss": 1.001, + "step": 13497 + }, + { + "epoch": 0.32, + "grad_norm": 2.0408243130726738, + "learning_rate": 1.5956341840000948e-05, + "loss": 1.0826, + "step": 13498 + }, + { + "epoch": 0.32, + "grad_norm": 2.07738076907267, + "learning_rate": 1.5955728905486e-05, + "loss": 1.0503, + "step": 13499 + }, + { + "epoch": 0.32, + "grad_norm": 1.9468616345836185, + "learning_rate": 1.595511593629491e-05, + "loss": 1.0085, + "step": 13500 + }, + { + "epoch": 0.32, + "grad_norm": 2.210002448808444, + "learning_rate": 1.595450293243125e-05, + "loss": 1.0393, + "step": 13501 + }, + { + "epoch": 0.32, + "grad_norm": 1.9302882640149166, + "learning_rate": 1.5953889893898592e-05, + "loss": 1.1007, + "step": 13502 + }, + { + "epoch": 0.32, + "grad_norm": 2.6565631515321853, + "learning_rate": 1.5953276820700502e-05, + "loss": 0.9794, + "step": 13503 + }, + { + "epoch": 0.32, + "grad_norm": 2.016931268846158, + "learning_rate": 1.5952663712840542e-05, + "loss": 1.0323, + "step": 13504 + }, + { + "epoch": 0.32, + "grad_norm": 2.0378648519668556, + "learning_rate": 1.59520505703223e-05, + "loss": 1.0456, + "step": 13505 + }, + { + "epoch": 0.32, + "grad_norm": 5.165419500643052, + "learning_rate": 1.595143739314933e-05, + "loss": 1.1592, + "step": 13506 + }, + { + "epoch": 0.32, + "grad_norm": 1.9478499672631777, + "learning_rate": 1.5950824181325206e-05, + "loss": 1.1805, + "step": 13507 + }, + { + "epoch": 0.32, + "grad_norm": 2.035666826782138, + "learning_rate": 1.5950210934853502e-05, + "loss": 1.1266, + "step": 13508 + }, + { + "epoch": 0.32, + "grad_norm": 1.9822469939531, + "learning_rate": 1.5949597653737785e-05, + "loss": 0.995, + "step": 13509 + }, + { + "epoch": 0.32, + "grad_norm": 2.1621778903958315, + "learning_rate": 1.594898433798163e-05, + "loss": 1.1473, + "step": 13510 + }, + { + "epoch": 0.32, + "grad_norm": 1.9301863797801766, + "learning_rate": 1.59483709875886e-05, + "loss": 0.9168, + "step": 13511 + }, + { + "epoch": 0.32, + "grad_norm": 2.348765883472765, + "learning_rate": 1.5947757602562273e-05, + "loss": 1.0775, + "step": 13512 + }, + { + "epoch": 0.32, + "grad_norm": 2.1710161167772792, + "learning_rate": 1.5947144182906217e-05, + "loss": 1.0996, + "step": 13513 + }, + { + "epoch": 0.32, + "grad_norm": 1.9368871975043758, + "learning_rate": 1.5946530728624005e-05, + "loss": 1.0466, + "step": 13514 + }, + { + "epoch": 0.32, + "grad_norm": 2.303769527792281, + "learning_rate": 1.594591723971921e-05, + "loss": 0.9498, + "step": 13515 + }, + { + "epoch": 0.32, + "grad_norm": 2.0181429397585515, + "learning_rate": 1.5945303716195403e-05, + "loss": 1.161, + "step": 13516 + }, + { + "epoch": 0.32, + "grad_norm": 1.911687586043285, + "learning_rate": 1.5944690158056154e-05, + "loss": 0.9782, + "step": 13517 + }, + { + "epoch": 0.32, + "grad_norm": 1.9476886574810564, + "learning_rate": 1.5944076565305032e-05, + "loss": 1.1452, + "step": 13518 + }, + { + "epoch": 0.32, + "grad_norm": 2.0393229494980605, + "learning_rate": 1.5943462937945624e-05, + "loss": 1.1441, + "step": 13519 + }, + { + "epoch": 0.32, + "grad_norm": 2.568539681197537, + "learning_rate": 1.5942849275981488e-05, + "loss": 1.049, + "step": 13520 + }, + { + "epoch": 0.32, + "grad_norm": 1.9355334942222366, + "learning_rate": 1.59422355794162e-05, + "loss": 1.068, + "step": 13521 + }, + { + "epoch": 0.32, + "grad_norm": 3.4612132614067925, + "learning_rate": 1.5941621848253343e-05, + "loss": 1.0502, + "step": 13522 + }, + { + "epoch": 0.32, + "grad_norm": 1.9319720454020142, + "learning_rate": 1.5941008082496473e-05, + "loss": 1.0114, + "step": 13523 + }, + { + "epoch": 0.32, + "grad_norm": 2.049292615834668, + "learning_rate": 1.594039428214918e-05, + "loss": 1.0213, + "step": 13524 + }, + { + "epoch": 0.32, + "grad_norm": 2.1853201273819547, + "learning_rate": 1.5939780447215024e-05, + "loss": 0.9552, + "step": 13525 + }, + { + "epoch": 0.32, + "grad_norm": 1.958723259778236, + "learning_rate": 1.5939166577697593e-05, + "loss": 1.0146, + "step": 13526 + }, + { + "epoch": 0.32, + "grad_norm": 2.300215474371803, + "learning_rate": 1.593855267360045e-05, + "loss": 0.9693, + "step": 13527 + }, + { + "epoch": 0.32, + "grad_norm": 1.7715589672953158, + "learning_rate": 1.5937938734927174e-05, + "loss": 0.9354, + "step": 13528 + }, + { + "epoch": 0.32, + "grad_norm": 1.8882166062037182, + "learning_rate": 1.593732476168134e-05, + "loss": 1.0269, + "step": 13529 + }, + { + "epoch": 0.32, + "grad_norm": 2.074079969804793, + "learning_rate": 1.5936710753866516e-05, + "loss": 1.009, + "step": 13530 + }, + { + "epoch": 0.32, + "grad_norm": 1.2434983141244755, + "learning_rate": 1.5936096711486287e-05, + "loss": 0.9992, + "step": 13531 + }, + { + "epoch": 0.32, + "grad_norm": 1.154004723833564, + "learning_rate": 1.5935482634544227e-05, + "loss": 0.9624, + "step": 13532 + }, + { + "epoch": 0.32, + "grad_norm": 3.1348540253519066, + "learning_rate": 1.5934868523043903e-05, + "loss": 0.8975, + "step": 13533 + }, + { + "epoch": 0.32, + "grad_norm": 2.010693250535836, + "learning_rate": 1.5934254376988895e-05, + "loss": 1.0228, + "step": 13534 + }, + { + "epoch": 0.32, + "grad_norm": 2.1795055327761954, + "learning_rate": 1.5933640196382782e-05, + "loss": 1.1, + "step": 13535 + }, + { + "epoch": 0.32, + "grad_norm": 2.060224185553297, + "learning_rate": 1.5933025981229136e-05, + "loss": 1.1076, + "step": 13536 + }, + { + "epoch": 0.32, + "grad_norm": 1.95365450941182, + "learning_rate": 1.5932411731531535e-05, + "loss": 1.057, + "step": 13537 + }, + { + "epoch": 0.32, + "grad_norm": 1.2444883899902583, + "learning_rate": 1.5931797447293553e-05, + "loss": 0.9809, + "step": 13538 + }, + { + "epoch": 0.32, + "grad_norm": 1.0850974384643883, + "learning_rate": 1.593118312851877e-05, + "loss": 0.9295, + "step": 13539 + }, + { + "epoch": 0.32, + "grad_norm": 2.2762760123024317, + "learning_rate": 1.5930568775210762e-05, + "loss": 1.0325, + "step": 13540 + }, + { + "epoch": 0.32, + "grad_norm": 2.1801941341170186, + "learning_rate": 1.5929954387373103e-05, + "loss": 0.9853, + "step": 13541 + }, + { + "epoch": 0.32, + "grad_norm": 2.9249174447367285, + "learning_rate": 1.5929339965009373e-05, + "loss": 1.0879, + "step": 13542 + }, + { + "epoch": 0.32, + "grad_norm": 1.8398349124196072, + "learning_rate": 1.592872550812315e-05, + "loss": 0.8861, + "step": 13543 + }, + { + "epoch": 0.32, + "grad_norm": 1.1223815358950286, + "learning_rate": 1.5928111016718005e-05, + "loss": 0.9, + "step": 13544 + }, + { + "epoch": 0.32, + "grad_norm": 2.391956054043611, + "learning_rate": 1.5927496490797526e-05, + "loss": 1.1006, + "step": 13545 + }, + { + "epoch": 0.32, + "grad_norm": 2.1988933548643734, + "learning_rate": 1.5926881930365285e-05, + "loss": 1.0438, + "step": 13546 + }, + { + "epoch": 0.32, + "grad_norm": 2.3259079406687264, + "learning_rate": 1.5926267335424857e-05, + "loss": 0.8393, + "step": 13547 + }, + { + "epoch": 0.32, + "grad_norm": 1.066743443045628, + "learning_rate": 1.5925652705979828e-05, + "loss": 0.9692, + "step": 13548 + }, + { + "epoch": 0.32, + "grad_norm": 2.1466110224020163, + "learning_rate": 1.592503804203377e-05, + "loss": 1.0328, + "step": 13549 + }, + { + "epoch": 0.32, + "grad_norm": 1.9112852945300167, + "learning_rate": 1.592442334359027e-05, + "loss": 0.9686, + "step": 13550 + }, + { + "epoch": 0.32, + "grad_norm": 2.167882247543665, + "learning_rate": 1.59238086106529e-05, + "loss": 1.0113, + "step": 13551 + }, + { + "epoch": 0.32, + "grad_norm": 2.097284252397891, + "learning_rate": 1.592319384322524e-05, + "loss": 1.1491, + "step": 13552 + }, + { + "epoch": 0.32, + "grad_norm": 2.661745293329267, + "learning_rate": 1.5922579041310866e-05, + "loss": 1.0151, + "step": 13553 + }, + { + "epoch": 0.32, + "grad_norm": 2.062577612705318, + "learning_rate": 1.5921964204913366e-05, + "loss": 1.1258, + "step": 13554 + }, + { + "epoch": 0.32, + "grad_norm": 2.0304081179703655, + "learning_rate": 1.5921349334036316e-05, + "loss": 1.0225, + "step": 13555 + }, + { + "epoch": 0.32, + "grad_norm": 1.848718198259266, + "learning_rate": 1.5920734428683295e-05, + "loss": 0.9989, + "step": 13556 + }, + { + "epoch": 0.32, + "grad_norm": 1.9585270992136798, + "learning_rate": 1.5920119488857885e-05, + "loss": 1.0484, + "step": 13557 + }, + { + "epoch": 0.32, + "grad_norm": 1.9859703929470873, + "learning_rate": 1.591950451456366e-05, + "loss": 0.9707, + "step": 13558 + }, + { + "epoch": 0.32, + "grad_norm": 2.0683263107616208, + "learning_rate": 1.591888950580421e-05, + "loss": 1.0885, + "step": 13559 + }, + { + "epoch": 0.32, + "grad_norm": 2.3530700209318094, + "learning_rate": 1.591827446258311e-05, + "loss": 1.01, + "step": 13560 + }, + { + "epoch": 0.32, + "grad_norm": 2.4225129064183157, + "learning_rate": 1.5917659384903944e-05, + "loss": 1.0926, + "step": 13561 + }, + { + "epoch": 0.32, + "grad_norm": 1.9628942193050594, + "learning_rate": 1.5917044272770288e-05, + "loss": 1.0692, + "step": 13562 + }, + { + "epoch": 0.32, + "grad_norm": 2.999880496981542, + "learning_rate": 1.5916429126185732e-05, + "loss": 1.0194, + "step": 13563 + }, + { + "epoch": 0.32, + "grad_norm": 2.0465254874249292, + "learning_rate": 1.5915813945153847e-05, + "loss": 0.9869, + "step": 13564 + }, + { + "epoch": 0.32, + "grad_norm": 2.944123494730258, + "learning_rate": 1.5915198729678224e-05, + "loss": 0.9915, + "step": 13565 + }, + { + "epoch": 0.32, + "grad_norm": 2.214742688807542, + "learning_rate": 1.591458347976244e-05, + "loss": 1.1387, + "step": 13566 + }, + { + "epoch": 0.32, + "grad_norm": 1.9354271146771935, + "learning_rate": 1.5913968195410077e-05, + "loss": 1.0578, + "step": 13567 + }, + { + "epoch": 0.32, + "grad_norm": 2.3276744473966446, + "learning_rate": 1.5913352876624722e-05, + "loss": 1.0595, + "step": 13568 + }, + { + "epoch": 0.32, + "grad_norm": 2.008966758775216, + "learning_rate": 1.591273752340995e-05, + "loss": 0.9431, + "step": 13569 + }, + { + "epoch": 0.32, + "grad_norm": 1.9809870529579672, + "learning_rate": 1.5912122135769356e-05, + "loss": 1.0215, + "step": 13570 + }, + { + "epoch": 0.32, + "grad_norm": 2.2925251115467966, + "learning_rate": 1.5911506713706507e-05, + "loss": 0.9862, + "step": 13571 + }, + { + "epoch": 0.32, + "grad_norm": 2.004783289492861, + "learning_rate": 1.5910891257225e-05, + "loss": 1.0595, + "step": 13572 + }, + { + "epoch": 0.32, + "grad_norm": 2.0184984042437097, + "learning_rate": 1.5910275766328408e-05, + "loss": 1.0623, + "step": 13573 + }, + { + "epoch": 0.32, + "grad_norm": 1.9106630324543488, + "learning_rate": 1.5909660241020323e-05, + "loss": 1.1714, + "step": 13574 + }, + { + "epoch": 0.32, + "grad_norm": 1.138407548941847, + "learning_rate": 1.5909044681304324e-05, + "loss": 0.9646, + "step": 13575 + }, + { + "epoch": 0.32, + "grad_norm": 1.9628923547200985, + "learning_rate": 1.5908429087183998e-05, + "loss": 0.9236, + "step": 13576 + }, + { + "epoch": 0.32, + "grad_norm": 8.61887208434684, + "learning_rate": 1.5907813458662923e-05, + "loss": 1.1945, + "step": 13577 + }, + { + "epoch": 0.32, + "grad_norm": 2.0102299572303566, + "learning_rate": 1.590719779574469e-05, + "loss": 1.0739, + "step": 13578 + }, + { + "epoch": 0.32, + "grad_norm": 2.303170890481123, + "learning_rate": 1.590658209843288e-05, + "loss": 0.9836, + "step": 13579 + }, + { + "epoch": 0.32, + "grad_norm": 1.0751381917595833, + "learning_rate": 1.590596636673108e-05, + "loss": 0.9723, + "step": 13580 + }, + { + "epoch": 0.32, + "grad_norm": 2.258176776574398, + "learning_rate": 1.5905350600642876e-05, + "loss": 1.0352, + "step": 13581 + }, + { + "epoch": 0.32, + "grad_norm": 1.1061653976700716, + "learning_rate": 1.590473480017185e-05, + "loss": 0.9501, + "step": 13582 + }, + { + "epoch": 0.32, + "grad_norm": 1.9258125691449615, + "learning_rate": 1.5904118965321585e-05, + "loss": 1.078, + "step": 13583 + }, + { + "epoch": 0.32, + "grad_norm": 2.6002301322070127, + "learning_rate": 1.5903503096095674e-05, + "loss": 1.04, + "step": 13584 + }, + { + "epoch": 0.32, + "grad_norm": 1.9915301505320728, + "learning_rate": 1.5902887192497698e-05, + "loss": 0.9569, + "step": 13585 + }, + { + "epoch": 0.32, + "grad_norm": 1.06940022976396, + "learning_rate": 1.5902271254531245e-05, + "loss": 0.9543, + "step": 13586 + }, + { + "epoch": 0.32, + "grad_norm": 2.3839413866917325, + "learning_rate": 1.59016552821999e-05, + "loss": 1.0414, + "step": 13587 + }, + { + "epoch": 0.32, + "grad_norm": 1.7493574256476734, + "learning_rate": 1.5901039275507247e-05, + "loss": 0.9784, + "step": 13588 + }, + { + "epoch": 0.32, + "grad_norm": 2.2023443058814314, + "learning_rate": 1.5900423234456877e-05, + "loss": 0.9086, + "step": 13589 + }, + { + "epoch": 0.32, + "grad_norm": 1.833673952177243, + "learning_rate": 1.5899807159052373e-05, + "loss": 1.1211, + "step": 13590 + }, + { + "epoch": 0.32, + "grad_norm": 2.129735825161606, + "learning_rate": 1.5899191049297328e-05, + "loss": 1.0042, + "step": 13591 + }, + { + "epoch": 0.32, + "grad_norm": 2.04558529608267, + "learning_rate": 1.589857490519532e-05, + "loss": 0.9818, + "step": 13592 + }, + { + "epoch": 0.32, + "grad_norm": 2.012089735957092, + "learning_rate": 1.5897958726749947e-05, + "loss": 0.9646, + "step": 13593 + }, + { + "epoch": 0.32, + "grad_norm": 2.187586803927898, + "learning_rate": 1.589734251396479e-05, + "loss": 1.0313, + "step": 13594 + }, + { + "epoch": 0.32, + "grad_norm": 2.12423159708359, + "learning_rate": 1.5896726266843433e-05, + "loss": 1.1268, + "step": 13595 + }, + { + "epoch": 0.32, + "grad_norm": 2.31122496641699, + "learning_rate": 1.589610998538947e-05, + "loss": 0.9749, + "step": 13596 + }, + { + "epoch": 0.32, + "grad_norm": 2.5260467808117006, + "learning_rate": 1.589549366960649e-05, + "loss": 0.935, + "step": 13597 + }, + { + "epoch": 0.32, + "grad_norm": 2.0012652826271085, + "learning_rate": 1.589487731949808e-05, + "loss": 1.1224, + "step": 13598 + }, + { + "epoch": 0.32, + "grad_norm": 2.0169222823767825, + "learning_rate": 1.5894260935067828e-05, + "loss": 1.0332, + "step": 13599 + }, + { + "epoch": 0.32, + "grad_norm": 1.124361478816913, + "learning_rate": 1.589364451631932e-05, + "loss": 0.9964, + "step": 13600 + }, + { + "epoch": 0.32, + "grad_norm": 5.327802589683086, + "learning_rate": 1.589302806325615e-05, + "loss": 1.0707, + "step": 13601 + }, + { + "epoch": 0.32, + "grad_norm": 2.143095165780237, + "learning_rate": 1.5892411575881906e-05, + "loss": 0.9218, + "step": 13602 + }, + { + "epoch": 0.32, + "grad_norm": 1.981280950560607, + "learning_rate": 1.5891795054200173e-05, + "loss": 1.19, + "step": 13603 + }, + { + "epoch": 0.32, + "grad_norm": 2.060892406753705, + "learning_rate": 1.5891178498214545e-05, + "loss": 1.2133, + "step": 13604 + }, + { + "epoch": 0.32, + "grad_norm": 2.059476293395754, + "learning_rate": 1.5890561907928612e-05, + "loss": 1.0741, + "step": 13605 + }, + { + "epoch": 0.32, + "grad_norm": 1.8459626425344944, + "learning_rate": 1.5889945283345963e-05, + "loss": 1.0321, + "step": 13606 + }, + { + "epoch": 0.32, + "grad_norm": 1.9354964070712684, + "learning_rate": 1.5889328624470186e-05, + "loss": 1.1089, + "step": 13607 + }, + { + "epoch": 0.32, + "grad_norm": 2.4754787994653733, + "learning_rate": 1.5888711931304874e-05, + "loss": 1.0969, + "step": 13608 + }, + { + "epoch": 0.32, + "grad_norm": 2.8846049088059003, + "learning_rate": 1.5888095203853614e-05, + "loss": 1.0285, + "step": 13609 + }, + { + "epoch": 0.32, + "grad_norm": 2.031629425670093, + "learning_rate": 1.5887478442120007e-05, + "loss": 1.1135, + "step": 13610 + }, + { + "epoch": 0.32, + "grad_norm": 1.1222294088948837, + "learning_rate": 1.588686164610763e-05, + "loss": 0.9799, + "step": 13611 + }, + { + "epoch": 0.32, + "grad_norm": 2.1228316967015606, + "learning_rate": 1.5886244815820082e-05, + "loss": 1.084, + "step": 13612 + }, + { + "epoch": 0.32, + "grad_norm": 1.1350481121620006, + "learning_rate": 1.5885627951260954e-05, + "loss": 1.0215, + "step": 13613 + }, + { + "epoch": 0.32, + "grad_norm": 2.0371319839506663, + "learning_rate": 1.5885011052433834e-05, + "loss": 1.0182, + "step": 13614 + }, + { + "epoch": 0.32, + "grad_norm": 1.814696362180984, + "learning_rate": 1.588439411934232e-05, + "loss": 1.0303, + "step": 13615 + }, + { + "epoch": 0.32, + "grad_norm": 2.633756718350325, + "learning_rate": 1.5883777151990002e-05, + "loss": 0.9043, + "step": 13616 + }, + { + "epoch": 0.32, + "grad_norm": 2.1311484030695134, + "learning_rate": 1.588316015038047e-05, + "loss": 1.0513, + "step": 13617 + }, + { + "epoch": 0.32, + "grad_norm": 2.0429308172747014, + "learning_rate": 1.5882543114517313e-05, + "loss": 1.0217, + "step": 13618 + }, + { + "epoch": 0.32, + "grad_norm": 2.1794731442140587, + "learning_rate": 1.588192604440413e-05, + "loss": 1.0688, + "step": 13619 + }, + { + "epoch": 0.32, + "grad_norm": 2.0472106565106376, + "learning_rate": 1.588130894004451e-05, + "loss": 0.9567, + "step": 13620 + }, + { + "epoch": 0.32, + "grad_norm": 2.1304538871048577, + "learning_rate": 1.5880691801442047e-05, + "loss": 1.0347, + "step": 13621 + }, + { + "epoch": 0.32, + "grad_norm": 1.9623062961448758, + "learning_rate": 1.5880074628600338e-05, + "loss": 0.9991, + "step": 13622 + }, + { + "epoch": 0.32, + "grad_norm": 2.065639656748186, + "learning_rate": 1.587945742152297e-05, + "loss": 0.9463, + "step": 13623 + }, + { + "epoch": 0.32, + "grad_norm": 1.9381373774121897, + "learning_rate": 1.587884018021354e-05, + "loss": 1.14, + "step": 13624 + }, + { + "epoch": 0.32, + "grad_norm": 2.0293579241534547, + "learning_rate": 1.5878222904675643e-05, + "loss": 1.0036, + "step": 13625 + }, + { + "epoch": 0.32, + "grad_norm": 1.872685025931619, + "learning_rate": 1.587760559491287e-05, + "loss": 1.0915, + "step": 13626 + }, + { + "epoch": 0.32, + "grad_norm": 2.092526856052594, + "learning_rate": 1.5876988250928812e-05, + "loss": 1.0561, + "step": 13627 + }, + { + "epoch": 0.32, + "grad_norm": 2.1329462738449574, + "learning_rate": 1.5876370872727072e-05, + "loss": 1.0965, + "step": 13628 + }, + { + "epoch": 0.32, + "grad_norm": 1.9209580287358865, + "learning_rate": 1.5875753460311245e-05, + "loss": 0.8699, + "step": 13629 + }, + { + "epoch": 0.32, + "grad_norm": 2.3905337964336306, + "learning_rate": 1.5875136013684917e-05, + "loss": 0.9443, + "step": 13630 + }, + { + "epoch": 0.32, + "grad_norm": 2.232193092515321, + "learning_rate": 1.5874518532851688e-05, + "loss": 0.9527, + "step": 13631 + }, + { + "epoch": 0.32, + "grad_norm": 2.0878294458739677, + "learning_rate": 1.5873901017815146e-05, + "loss": 1.0768, + "step": 13632 + }, + { + "epoch": 0.32, + "grad_norm": 2.4095915292367147, + "learning_rate": 1.58732834685789e-05, + "loss": 1.049, + "step": 13633 + }, + { + "epoch": 0.32, + "grad_norm": 1.846520471240235, + "learning_rate": 1.5872665885146535e-05, + "loss": 0.9827, + "step": 13634 + }, + { + "epoch": 0.32, + "grad_norm": 2.0959928463182567, + "learning_rate": 1.587204826752165e-05, + "loss": 0.9554, + "step": 13635 + }, + { + "epoch": 0.32, + "grad_norm": 1.963968265958768, + "learning_rate": 1.5871430615707843e-05, + "loss": 1.1783, + "step": 13636 + }, + { + "epoch": 0.32, + "grad_norm": 2.1283860688270475, + "learning_rate": 1.5870812929708706e-05, + "loss": 0.9844, + "step": 13637 + }, + { + "epoch": 0.32, + "grad_norm": 2.327954680698067, + "learning_rate": 1.587019520952784e-05, + "loss": 1.1184, + "step": 13638 + }, + { + "epoch": 0.32, + "grad_norm": 2.245328070910769, + "learning_rate": 1.5869577455168833e-05, + "loss": 1.0039, + "step": 13639 + }, + { + "epoch": 0.32, + "grad_norm": 1.170251411924956, + "learning_rate": 1.5868959666635297e-05, + "loss": 0.9005, + "step": 13640 + }, + { + "epoch": 0.32, + "grad_norm": 2.0596987081509828, + "learning_rate": 1.5868341843930812e-05, + "loss": 1.0471, + "step": 13641 + }, + { + "epoch": 0.32, + "grad_norm": 1.9775074504751013, + "learning_rate": 1.5867723987058986e-05, + "loss": 1.1212, + "step": 13642 + }, + { + "epoch": 0.32, + "grad_norm": 1.817361024037221, + "learning_rate": 1.586710609602341e-05, + "loss": 1.0892, + "step": 13643 + }, + { + "epoch": 0.32, + "grad_norm": 1.914536172294967, + "learning_rate": 1.586648817082769e-05, + "loss": 1.048, + "step": 13644 + }, + { + "epoch": 0.32, + "grad_norm": 1.0791049138401172, + "learning_rate": 1.5865870211475417e-05, + "loss": 0.991, + "step": 13645 + }, + { + "epoch": 0.32, + "grad_norm": 1.9450487023566068, + "learning_rate": 1.5865252217970187e-05, + "loss": 0.9344, + "step": 13646 + }, + { + "epoch": 0.32, + "grad_norm": 2.3591792684840964, + "learning_rate": 1.5864634190315608e-05, + "loss": 1.0333, + "step": 13647 + }, + { + "epoch": 0.32, + "grad_norm": 1.8110546170539243, + "learning_rate": 1.5864016128515267e-05, + "loss": 1.0413, + "step": 13648 + }, + { + "epoch": 0.32, + "grad_norm": 2.0328122080609186, + "learning_rate": 1.586339803257277e-05, + "loss": 1.0601, + "step": 13649 + }, + { + "epoch": 0.32, + "grad_norm": 2.345062774477049, + "learning_rate": 1.586277990249171e-05, + "loss": 1.0149, + "step": 13650 + }, + { + "epoch": 0.32, + "grad_norm": 2.0913963971887375, + "learning_rate": 1.586216173827569e-05, + "loss": 1.1116, + "step": 13651 + }, + { + "epoch": 0.32, + "grad_norm": 2.0833962876772, + "learning_rate": 1.5861543539928313e-05, + "loss": 0.9832, + "step": 13652 + }, + { + "epoch": 0.32, + "grad_norm": 3.377928652767344, + "learning_rate": 1.586092530745317e-05, + "loss": 1.0566, + "step": 13653 + }, + { + "epoch": 0.32, + "grad_norm": 1.9700297269867928, + "learning_rate": 1.5860307040853867e-05, + "loss": 1.0002, + "step": 13654 + }, + { + "epoch": 0.32, + "grad_norm": 1.9801744883405565, + "learning_rate": 1.5859688740133998e-05, + "loss": 1.1128, + "step": 13655 + }, + { + "epoch": 0.32, + "grad_norm": 2.1413658695578395, + "learning_rate": 1.5859070405297168e-05, + "loss": 1.181, + "step": 13656 + }, + { + "epoch": 0.32, + "grad_norm": 1.1549885911366444, + "learning_rate": 1.5858452036346973e-05, + "loss": 0.9426, + "step": 13657 + }, + { + "epoch": 0.32, + "grad_norm": 2.0390912963763097, + "learning_rate": 1.5857833633287015e-05, + "loss": 1.2143, + "step": 13658 + }, + { + "epoch": 0.32, + "grad_norm": 2.187691188231447, + "learning_rate": 1.58572151961209e-05, + "loss": 0.9039, + "step": 13659 + }, + { + "epoch": 0.32, + "grad_norm": 2.565393964306728, + "learning_rate": 1.585659672485222e-05, + "loss": 1.0001, + "step": 13660 + }, + { + "epoch": 0.32, + "grad_norm": 2.2245965704829898, + "learning_rate": 1.5855978219484582e-05, + "loss": 1.1718, + "step": 13661 + }, + { + "epoch": 0.32, + "grad_norm": 1.9736697987672869, + "learning_rate": 1.5855359680021583e-05, + "loss": 1.0157, + "step": 13662 + }, + { + "epoch": 0.32, + "grad_norm": 1.93093647515926, + "learning_rate": 1.5854741106466825e-05, + "loss": 1.0877, + "step": 13663 + }, + { + "epoch": 0.32, + "grad_norm": 1.8712789263708067, + "learning_rate": 1.585412249882391e-05, + "loss": 1.065, + "step": 13664 + }, + { + "epoch": 0.32, + "grad_norm": 2.1980430344089092, + "learning_rate": 1.5853503857096446e-05, + "loss": 1.1255, + "step": 13665 + }, + { + "epoch": 0.32, + "grad_norm": 2.141443450573122, + "learning_rate": 1.5852885181288023e-05, + "loss": 1.0743, + "step": 13666 + }, + { + "epoch": 0.32, + "grad_norm": 2.1375827766220197, + "learning_rate": 1.5852266471402256e-05, + "loss": 1.0945, + "step": 13667 + }, + { + "epoch": 0.32, + "grad_norm": 1.7551568340910373, + "learning_rate": 1.5851647727442735e-05, + "loss": 1.0363, + "step": 13668 + }, + { + "epoch": 0.32, + "grad_norm": 2.1223419733565927, + "learning_rate": 1.585102894941307e-05, + "loss": 1.0096, + "step": 13669 + }, + { + "epoch": 0.32, + "grad_norm": 2.536261017312712, + "learning_rate": 1.5850410137316865e-05, + "loss": 0.8969, + "step": 13670 + }, + { + "epoch": 0.32, + "grad_norm": 2.1962601401959203, + "learning_rate": 1.5849791291157717e-05, + "loss": 1.073, + "step": 13671 + }, + { + "epoch": 0.32, + "grad_norm": 1.9302088332729896, + "learning_rate": 1.584917241093923e-05, + "loss": 0.9068, + "step": 13672 + }, + { + "epoch": 0.32, + "grad_norm": 2.3533507016277957, + "learning_rate": 1.5848553496665013e-05, + "loss": 1.1002, + "step": 13673 + }, + { + "epoch": 0.32, + "grad_norm": 2.4480207998957555, + "learning_rate": 1.5847934548338665e-05, + "loss": 1.0839, + "step": 13674 + }, + { + "epoch": 0.32, + "grad_norm": 2.1581814436880724, + "learning_rate": 1.584731556596379e-05, + "loss": 1.1266, + "step": 13675 + }, + { + "epoch": 0.32, + "grad_norm": 1.06524793781834, + "learning_rate": 1.5846696549543993e-05, + "loss": 0.9617, + "step": 13676 + }, + { + "epoch": 0.32, + "grad_norm": 2.086631062660072, + "learning_rate": 1.5846077499082883e-05, + "loss": 1.0585, + "step": 13677 + }, + { + "epoch": 0.32, + "grad_norm": 2.0494484146552487, + "learning_rate": 1.5845458414584048e-05, + "loss": 1.0704, + "step": 13678 + }, + { + "epoch": 0.32, + "grad_norm": 1.8813971032298566, + "learning_rate": 1.5844839296051115e-05, + "loss": 1.1351, + "step": 13679 + }, + { + "epoch": 0.32, + "grad_norm": 2.3030518770134463, + "learning_rate": 1.5844220143487667e-05, + "loss": 1.0265, + "step": 13680 + }, + { + "epoch": 0.32, + "grad_norm": 2.4773701845922296, + "learning_rate": 1.5843600956897324e-05, + "loss": 1.0867, + "step": 13681 + }, + { + "epoch": 0.32, + "grad_norm": 2.193905553471922, + "learning_rate": 1.5842981736283686e-05, + "loss": 0.9236, + "step": 13682 + }, + { + "epoch": 0.32, + "grad_norm": 2.1405237481307076, + "learning_rate": 1.5842362481650356e-05, + "loss": 1.045, + "step": 13683 + }, + { + "epoch": 0.32, + "grad_norm": 2.0826919185243127, + "learning_rate": 1.5841743193000945e-05, + "loss": 1.0134, + "step": 13684 + }, + { + "epoch": 0.32, + "grad_norm": 2.241710713601559, + "learning_rate": 1.5841123870339055e-05, + "loss": 1.1354, + "step": 13685 + }, + { + "epoch": 0.32, + "grad_norm": 2.4610829128719196, + "learning_rate": 1.5840504513668294e-05, + "loss": 1.0261, + "step": 13686 + }, + { + "epoch": 0.32, + "grad_norm": 2.2071418512954746, + "learning_rate": 1.5839885122992263e-05, + "loss": 1.0856, + "step": 13687 + }, + { + "epoch": 0.32, + "grad_norm": 2.0956023888405553, + "learning_rate": 1.5839265698314575e-05, + "loss": 0.9535, + "step": 13688 + }, + { + "epoch": 0.32, + "grad_norm": 2.3517224802564924, + "learning_rate": 1.583864623963883e-05, + "loss": 0.9305, + "step": 13689 + }, + { + "epoch": 0.32, + "grad_norm": 1.8575388829413655, + "learning_rate": 1.5838026746968643e-05, + "loss": 0.9055, + "step": 13690 + }, + { + "epoch": 0.32, + "grad_norm": 2.287202471488348, + "learning_rate": 1.5837407220307612e-05, + "loss": 1.0085, + "step": 13691 + }, + { + "epoch": 0.32, + "grad_norm": 1.9396768914575697, + "learning_rate": 1.583678765965935e-05, + "loss": 1.1404, + "step": 13692 + }, + { + "epoch": 0.32, + "grad_norm": 2.343095881258778, + "learning_rate": 1.5836168065027463e-05, + "loss": 1.0793, + "step": 13693 + }, + { + "epoch": 0.32, + "grad_norm": 6.263917668655307, + "learning_rate": 1.5835548436415558e-05, + "loss": 1.1316, + "step": 13694 + }, + { + "epoch": 0.32, + "grad_norm": 1.8599118119736022, + "learning_rate": 1.583492877382724e-05, + "loss": 1.0015, + "step": 13695 + }, + { + "epoch": 0.32, + "grad_norm": 1.7936535643717093, + "learning_rate": 1.583430907726612e-05, + "loss": 1.0824, + "step": 13696 + }, + { + "epoch": 0.32, + "grad_norm": 2.1021009198853893, + "learning_rate": 1.5833689346735807e-05, + "loss": 1.0861, + "step": 13697 + }, + { + "epoch": 0.32, + "grad_norm": 2.089064594774544, + "learning_rate": 1.5833069582239905e-05, + "loss": 1.1128, + "step": 13698 + }, + { + "epoch": 0.32, + "grad_norm": 2.085147553887516, + "learning_rate": 1.5832449783782027e-05, + "loss": 1.0133, + "step": 13699 + }, + { + "epoch": 0.32, + "grad_norm": 3.3102799440546526, + "learning_rate": 1.583182995136578e-05, + "loss": 0.9828, + "step": 13700 + }, + { + "epoch": 0.32, + "grad_norm": 3.3007678992379508, + "learning_rate": 1.5831210084994775e-05, + "loss": 0.9502, + "step": 13701 + }, + { + "epoch": 0.32, + "grad_norm": 2.0876587657814807, + "learning_rate": 1.5830590184672613e-05, + "loss": 1.1471, + "step": 13702 + }, + { + "epoch": 0.32, + "grad_norm": 2.0706421507567465, + "learning_rate": 1.5829970250402916e-05, + "loss": 0.9608, + "step": 13703 + }, + { + "epoch": 0.32, + "grad_norm": 2.047205688066472, + "learning_rate": 1.5829350282189278e-05, + "loss": 1.0662, + "step": 13704 + }, + { + "epoch": 0.32, + "grad_norm": 1.8603004102482879, + "learning_rate": 1.582873028003532e-05, + "loss": 1.2178, + "step": 13705 + }, + { + "epoch": 0.32, + "grad_norm": 2.0315293804248427, + "learning_rate": 1.582811024394465e-05, + "loss": 1.0395, + "step": 13706 + }, + { + "epoch": 0.32, + "grad_norm": 2.0695212155461005, + "learning_rate": 1.5827490173920882e-05, + "loss": 1.0833, + "step": 13707 + }, + { + "epoch": 0.32, + "grad_norm": 2.145071591944474, + "learning_rate": 1.5826870069967615e-05, + "loss": 1.0559, + "step": 13708 + }, + { + "epoch": 0.32, + "grad_norm": 2.2105483546271447, + "learning_rate": 1.5826249932088467e-05, + "loss": 1.0846, + "step": 13709 + }, + { + "epoch": 0.32, + "grad_norm": 2.121694923860802, + "learning_rate": 1.5825629760287052e-05, + "loss": 1.0362, + "step": 13710 + }, + { + "epoch": 0.32, + "grad_norm": 1.9457673796870483, + "learning_rate": 1.582500955456697e-05, + "loss": 0.9327, + "step": 13711 + }, + { + "epoch": 0.32, + "grad_norm": 2.927687196688635, + "learning_rate": 1.582438931493184e-05, + "loss": 1.0334, + "step": 13712 + }, + { + "epoch": 0.32, + "grad_norm": 1.8043907169719389, + "learning_rate": 1.5823769041385272e-05, + "loss": 1.0042, + "step": 13713 + }, + { + "epoch": 0.32, + "grad_norm": 1.9563991582077551, + "learning_rate": 1.5823148733930876e-05, + "loss": 0.9361, + "step": 13714 + }, + { + "epoch": 0.32, + "grad_norm": 2.4643095160338686, + "learning_rate": 1.5822528392572266e-05, + "loss": 1.0412, + "step": 13715 + }, + { + "epoch": 0.32, + "grad_norm": 2.8247313311297373, + "learning_rate": 1.5821908017313052e-05, + "loss": 0.932, + "step": 13716 + }, + { + "epoch": 0.32, + "grad_norm": 2.2056080055451934, + "learning_rate": 1.5821287608156847e-05, + "loss": 1.0886, + "step": 13717 + }, + { + "epoch": 0.32, + "grad_norm": 2.8179662522651796, + "learning_rate": 1.582066716510726e-05, + "loss": 0.9465, + "step": 13718 + }, + { + "epoch": 0.32, + "grad_norm": 2.211914296287036, + "learning_rate": 1.582004668816791e-05, + "loss": 1.0519, + "step": 13719 + }, + { + "epoch": 0.32, + "grad_norm": 1.0878620100067595, + "learning_rate": 1.58194261773424e-05, + "loss": 0.969, + "step": 13720 + }, + { + "epoch": 0.32, + "grad_norm": 1.1730178613668014, + "learning_rate": 1.581880563263435e-05, + "loss": 0.97, + "step": 13721 + }, + { + "epoch": 0.32, + "grad_norm": 1.9668500783215388, + "learning_rate": 1.5818185054047376e-05, + "loss": 1.1212, + "step": 13722 + }, + { + "epoch": 0.32, + "grad_norm": 1.1495606807774608, + "learning_rate": 1.5817564441585083e-05, + "loss": 1.058, + "step": 13723 + }, + { + "epoch": 0.32, + "grad_norm": 2.4764814899459306, + "learning_rate": 1.581694379525109e-05, + "loss": 1.0351, + "step": 13724 + }, + { + "epoch": 0.32, + "grad_norm": 2.005767969571845, + "learning_rate": 1.5816323115049005e-05, + "loss": 0.9268, + "step": 13725 + }, + { + "epoch": 0.32, + "grad_norm": 2.046249005115038, + "learning_rate": 1.5815702400982447e-05, + "loss": 1.0006, + "step": 13726 + }, + { + "epoch": 0.32, + "grad_norm": 2.987941990010126, + "learning_rate": 1.5815081653055028e-05, + "loss": 1.0356, + "step": 13727 + }, + { + "epoch": 0.32, + "grad_norm": 2.2694031498477862, + "learning_rate": 1.5814460871270364e-05, + "loss": 1.0176, + "step": 13728 + }, + { + "epoch": 0.32, + "grad_norm": 1.9044912132110663, + "learning_rate": 1.5813840055632067e-05, + "loss": 0.9952, + "step": 13729 + }, + { + "epoch": 0.32, + "grad_norm": 2.407829816491094, + "learning_rate": 1.5813219206143755e-05, + "loss": 0.9875, + "step": 13730 + }, + { + "epoch": 0.32, + "grad_norm": 1.9214683194264208, + "learning_rate": 1.5812598322809036e-05, + "loss": 1.0834, + "step": 13731 + }, + { + "epoch": 0.32, + "grad_norm": 1.86894239647082, + "learning_rate": 1.581197740563153e-05, + "loss": 0.9594, + "step": 13732 + }, + { + "epoch": 0.32, + "grad_norm": 1.9716979432642492, + "learning_rate": 1.5811356454614853e-05, + "loss": 1.0503, + "step": 13733 + }, + { + "epoch": 0.32, + "grad_norm": 1.1966527331569552, + "learning_rate": 1.5810735469762623e-05, + "loss": 1.0111, + "step": 13734 + }, + { + "epoch": 0.32, + "grad_norm": 2.089803352923646, + "learning_rate": 1.5810114451078446e-05, + "loss": 1.0335, + "step": 13735 + }, + { + "epoch": 0.32, + "grad_norm": 1.104892644518991, + "learning_rate": 1.5809493398565945e-05, + "loss": 1.0454, + "step": 13736 + }, + { + "epoch": 0.32, + "grad_norm": 1.1810569888304314, + "learning_rate": 1.5808872312228735e-05, + "loss": 1.0021, + "step": 13737 + }, + { + "epoch": 0.32, + "grad_norm": 2.465034310490675, + "learning_rate": 1.5808251192070428e-05, + "loss": 1.117, + "step": 13738 + }, + { + "epoch": 0.32, + "grad_norm": 1.97358142431027, + "learning_rate": 1.5807630038094646e-05, + "loss": 0.9912, + "step": 13739 + }, + { + "epoch": 0.32, + "grad_norm": 2.0993064818249936, + "learning_rate": 1.5807008850305003e-05, + "loss": 1.0593, + "step": 13740 + }, + { + "epoch": 0.32, + "grad_norm": 2.107714999869259, + "learning_rate": 1.580638762870512e-05, + "loss": 1.0291, + "step": 13741 + }, + { + "epoch": 0.32, + "grad_norm": 1.1628944051496592, + "learning_rate": 1.5805766373298605e-05, + "loss": 0.9243, + "step": 13742 + }, + { + "epoch": 0.32, + "grad_norm": 1.8945240733413837, + "learning_rate": 1.580514508408908e-05, + "loss": 1.0434, + "step": 13743 + }, + { + "epoch": 0.32, + "grad_norm": 2.164471272760119, + "learning_rate": 1.580452376108016e-05, + "loss": 1.0281, + "step": 13744 + }, + { + "epoch": 0.32, + "grad_norm": 2.1658515131680445, + "learning_rate": 1.580390240427547e-05, + "loss": 1.0401, + "step": 13745 + }, + { + "epoch": 0.32, + "grad_norm": 2.030211953933566, + "learning_rate": 1.5803281013678623e-05, + "loss": 1.2381, + "step": 13746 + }, + { + "epoch": 0.32, + "grad_norm": 1.853249933860991, + "learning_rate": 1.5802659589293234e-05, + "loss": 0.984, + "step": 13747 + }, + { + "epoch": 0.32, + "grad_norm": 1.9740201985383667, + "learning_rate": 1.5802038131122926e-05, + "loss": 0.998, + "step": 13748 + }, + { + "epoch": 0.32, + "grad_norm": 2.179900172616641, + "learning_rate": 1.5801416639171314e-05, + "loss": 0.9919, + "step": 13749 + }, + { + "epoch": 0.32, + "grad_norm": 2.1832685004431744, + "learning_rate": 1.5800795113442017e-05, + "loss": 1.0062, + "step": 13750 + }, + { + "epoch": 0.32, + "grad_norm": 1.9063768482655228, + "learning_rate": 1.5800173553938654e-05, + "loss": 1.019, + "step": 13751 + }, + { + "epoch": 0.32, + "grad_norm": 2.0424382815061204, + "learning_rate": 1.5799551960664846e-05, + "loss": 0.9577, + "step": 13752 + }, + { + "epoch": 0.32, + "grad_norm": 2.5409778274023727, + "learning_rate": 1.579893033362421e-05, + "loss": 0.9477, + "step": 13753 + }, + { + "epoch": 0.32, + "grad_norm": 2.945595106766127, + "learning_rate": 1.5798308672820363e-05, + "loss": 1.0131, + "step": 13754 + }, + { + "epoch": 0.32, + "grad_norm": 2.035270881490521, + "learning_rate": 1.579768697825693e-05, + "loss": 0.9272, + "step": 13755 + }, + { + "epoch": 0.32, + "grad_norm": 1.1795253150732896, + "learning_rate": 1.5797065249937528e-05, + "loss": 1.047, + "step": 13756 + }, + { + "epoch": 0.32, + "grad_norm": 2.7269389832272983, + "learning_rate": 1.5796443487865774e-05, + "loss": 1.1227, + "step": 13757 + }, + { + "epoch": 0.32, + "grad_norm": 1.9194643989299671, + "learning_rate": 1.5795821692045294e-05, + "loss": 1.0625, + "step": 13758 + }, + { + "epoch": 0.32, + "grad_norm": 1.9949702225320947, + "learning_rate": 1.5795199862479706e-05, + "loss": 1.0322, + "step": 13759 + }, + { + "epoch": 0.32, + "grad_norm": 1.8066884054489853, + "learning_rate": 1.5794577999172628e-05, + "loss": 1.0443, + "step": 13760 + }, + { + "epoch": 0.32, + "grad_norm": 1.9381927341300047, + "learning_rate": 1.5793956102127682e-05, + "loss": 1.0089, + "step": 13761 + }, + { + "epoch": 0.32, + "grad_norm": 1.999610389294044, + "learning_rate": 1.579333417134849e-05, + "loss": 1.1591, + "step": 13762 + }, + { + "epoch": 0.32, + "grad_norm": 1.9156212448210495, + "learning_rate": 1.579271220683867e-05, + "loss": 1.0311, + "step": 13763 + }, + { + "epoch": 0.32, + "grad_norm": 2.2938533339648797, + "learning_rate": 1.5792090208601847e-05, + "loss": 0.9505, + "step": 13764 + }, + { + "epoch": 0.32, + "grad_norm": 1.8191896448632627, + "learning_rate": 1.5791468176641642e-05, + "loss": 1.2028, + "step": 13765 + }, + { + "epoch": 0.32, + "grad_norm": 1.920225953635639, + "learning_rate": 1.5790846110961676e-05, + "loss": 1.0476, + "step": 13766 + }, + { + "epoch": 0.32, + "grad_norm": 2.077747826681504, + "learning_rate": 1.5790224011565566e-05, + "loss": 0.9715, + "step": 13767 + }, + { + "epoch": 0.32, + "grad_norm": 2.2204612985675447, + "learning_rate": 1.5789601878456946e-05, + "loss": 0.9577, + "step": 13768 + }, + { + "epoch": 0.32, + "grad_norm": 2.1200847465110066, + "learning_rate": 1.5788979711639427e-05, + "loss": 0.9712, + "step": 13769 + }, + { + "epoch": 0.32, + "grad_norm": 1.892564601894168, + "learning_rate": 1.5788357511116633e-05, + "loss": 1.0622, + "step": 13770 + }, + { + "epoch": 0.32, + "grad_norm": 2.596093923416686, + "learning_rate": 1.5787735276892194e-05, + "loss": 0.9816, + "step": 13771 + }, + { + "epoch": 0.32, + "grad_norm": 2.0382270620186174, + "learning_rate": 1.5787113008969725e-05, + "loss": 1.0418, + "step": 13772 + }, + { + "epoch": 0.32, + "grad_norm": 2.4603005515922494, + "learning_rate": 1.5786490707352852e-05, + "loss": 1.0937, + "step": 13773 + }, + { + "epoch": 0.32, + "grad_norm": 1.7853199641921615, + "learning_rate": 1.57858683720452e-05, + "loss": 1.0168, + "step": 13774 + }, + { + "epoch": 0.32, + "grad_norm": 1.9972283063009986, + "learning_rate": 1.578524600305039e-05, + "loss": 1.0416, + "step": 13775 + }, + { + "epoch": 0.32, + "grad_norm": 1.9839089278767283, + "learning_rate": 1.578462360037204e-05, + "loss": 1.0683, + "step": 13776 + }, + { + "epoch": 0.32, + "grad_norm": 1.9065804116983822, + "learning_rate": 1.5784001164013788e-05, + "loss": 1.0534, + "step": 13777 + }, + { + "epoch": 0.32, + "grad_norm": 1.9210168700660382, + "learning_rate": 1.578337869397925e-05, + "loss": 1.0331, + "step": 13778 + }, + { + "epoch": 0.32, + "grad_norm": 2.0011845797519507, + "learning_rate": 1.5782756190272047e-05, + "loss": 1.0132, + "step": 13779 + }, + { + "epoch": 0.32, + "grad_norm": 2.08296800415301, + "learning_rate": 1.5782133652895804e-05, + "loss": 0.9886, + "step": 13780 + }, + { + "epoch": 0.32, + "grad_norm": 1.9438724529416276, + "learning_rate": 1.5781511081854152e-05, + "loss": 1.0268, + "step": 13781 + }, + { + "epoch": 0.32, + "grad_norm": 2.3815456994003075, + "learning_rate": 1.5780888477150712e-05, + "loss": 1.0185, + "step": 13782 + }, + { + "epoch": 0.32, + "grad_norm": 2.55014283494069, + "learning_rate": 1.5780265838789107e-05, + "loss": 1.0101, + "step": 13783 + }, + { + "epoch": 0.32, + "grad_norm": 2.1356990400603473, + "learning_rate": 1.5779643166772966e-05, + "loss": 0.9075, + "step": 13784 + }, + { + "epoch": 0.32, + "grad_norm": 2.156627293169219, + "learning_rate": 1.5779020461105913e-05, + "loss": 1.0497, + "step": 13785 + }, + { + "epoch": 0.32, + "grad_norm": 1.9648499981684149, + "learning_rate": 1.5778397721791574e-05, + "loss": 0.9605, + "step": 13786 + }, + { + "epoch": 0.32, + "grad_norm": 1.9679285925379393, + "learning_rate": 1.5777774948833574e-05, + "loss": 1.1555, + "step": 13787 + }, + { + "epoch": 0.32, + "grad_norm": 2.046745456011433, + "learning_rate": 1.5777152142235535e-05, + "loss": 1.0644, + "step": 13788 + }, + { + "epoch": 0.32, + "grad_norm": 2.340370100910734, + "learning_rate": 1.5776529302001093e-05, + "loss": 0.9415, + "step": 13789 + }, + { + "epoch": 0.32, + "grad_norm": 2.1009383479030745, + "learning_rate": 1.5775906428133866e-05, + "loss": 1.1219, + "step": 13790 + }, + { + "epoch": 0.32, + "grad_norm": 4.281116451426316, + "learning_rate": 1.5775283520637483e-05, + "loss": 1.0613, + "step": 13791 + }, + { + "epoch": 0.32, + "grad_norm": 1.9534100516447326, + "learning_rate": 1.577466057951557e-05, + "loss": 0.994, + "step": 13792 + }, + { + "epoch": 0.32, + "grad_norm": 2.7350297955317964, + "learning_rate": 1.5774037604771756e-05, + "loss": 1.0362, + "step": 13793 + }, + { + "epoch": 0.32, + "grad_norm": 1.9446464287788283, + "learning_rate": 1.5773414596409666e-05, + "loss": 1.0608, + "step": 13794 + }, + { + "epoch": 0.32, + "grad_norm": 2.4330452014518147, + "learning_rate": 1.577279155443293e-05, + "loss": 1.1403, + "step": 13795 + }, + { + "epoch": 0.33, + "grad_norm": 2.7581884902709826, + "learning_rate": 1.5772168478845174e-05, + "loss": 0.9854, + "step": 13796 + }, + { + "epoch": 0.33, + "grad_norm": 1.1955979416825362, + "learning_rate": 1.5771545369650026e-05, + "loss": 0.9321, + "step": 13797 + }, + { + "epoch": 0.33, + "grad_norm": 2.2054799313976283, + "learning_rate": 1.577092222685111e-05, + "loss": 1.1441, + "step": 13798 + }, + { + "epoch": 0.33, + "grad_norm": 2.042865248528193, + "learning_rate": 1.577029905045206e-05, + "loss": 0.9917, + "step": 13799 + }, + { + "epoch": 0.33, + "grad_norm": 2.234902006196756, + "learning_rate": 1.5769675840456503e-05, + "loss": 1.012, + "step": 13800 + }, + { + "epoch": 0.33, + "grad_norm": 2.1052496079800203, + "learning_rate": 1.5769052596868065e-05, + "loss": 1.0586, + "step": 13801 + }, + { + "epoch": 0.33, + "grad_norm": 2.0383187995157237, + "learning_rate": 1.5768429319690376e-05, + "loss": 1.0484, + "step": 13802 + }, + { + "epoch": 0.33, + "grad_norm": 2.1583176516658944, + "learning_rate": 1.5767806008927068e-05, + "loss": 1.001, + "step": 13803 + }, + { + "epoch": 0.33, + "grad_norm": 2.1572918227976365, + "learning_rate": 1.5767182664581767e-05, + "loss": 1.0338, + "step": 13804 + }, + { + "epoch": 0.33, + "grad_norm": 2.150262940859931, + "learning_rate": 1.57665592866581e-05, + "loss": 1.1185, + "step": 13805 + }, + { + "epoch": 0.33, + "grad_norm": 1.1028832805490605, + "learning_rate": 1.5765935875159703e-05, + "loss": 1.0119, + "step": 13806 + }, + { + "epoch": 0.33, + "grad_norm": 1.1414066986421554, + "learning_rate": 1.5765312430090197e-05, + "loss": 0.9641, + "step": 13807 + }, + { + "epoch": 0.33, + "grad_norm": 2.166002923530061, + "learning_rate": 1.5764688951453216e-05, + "loss": 0.958, + "step": 13808 + }, + { + "epoch": 0.33, + "grad_norm": 1.9738754357383432, + "learning_rate": 1.57640654392524e-05, + "loss": 1.0063, + "step": 13809 + }, + { + "epoch": 0.33, + "grad_norm": 2.046262953976628, + "learning_rate": 1.5763441893491364e-05, + "loss": 1.0742, + "step": 13810 + }, + { + "epoch": 0.33, + "grad_norm": 2.001540021142915, + "learning_rate": 1.5762818314173743e-05, + "loss": 0.9654, + "step": 13811 + }, + { + "epoch": 0.33, + "grad_norm": 2.419631091543845, + "learning_rate": 1.576219470130317e-05, + "loss": 1.0568, + "step": 13812 + }, + { + "epoch": 0.33, + "grad_norm": 2.1054673559807036, + "learning_rate": 1.576157105488328e-05, + "loss": 1.12, + "step": 13813 + }, + { + "epoch": 0.33, + "grad_norm": 2.292797405528066, + "learning_rate": 1.5760947374917693e-05, + "loss": 0.9468, + "step": 13814 + }, + { + "epoch": 0.33, + "grad_norm": 1.0668733771472576, + "learning_rate": 1.576032366141005e-05, + "loss": 0.9453, + "step": 13815 + }, + { + "epoch": 0.33, + "grad_norm": 1.8083345467429515, + "learning_rate": 1.5759699914363977e-05, + "loss": 1.121, + "step": 13816 + }, + { + "epoch": 0.33, + "grad_norm": 2.16059921971455, + "learning_rate": 1.575907613378311e-05, + "loss": 0.9627, + "step": 13817 + }, + { + "epoch": 0.33, + "grad_norm": 2.6647796462665734, + "learning_rate": 1.5758452319671075e-05, + "loss": 1.0565, + "step": 13818 + }, + { + "epoch": 0.33, + "grad_norm": 1.9002284580395878, + "learning_rate": 1.5757828472031512e-05, + "loss": 1.0831, + "step": 13819 + }, + { + "epoch": 0.33, + "grad_norm": 2.747316530139246, + "learning_rate": 1.5757204590868042e-05, + "loss": 0.9463, + "step": 13820 + }, + { + "epoch": 0.33, + "grad_norm": 2.6140561820496537, + "learning_rate": 1.575658067618431e-05, + "loss": 0.9868, + "step": 13821 + }, + { + "epoch": 0.33, + "grad_norm": 2.074325785843103, + "learning_rate": 1.5755956727983943e-05, + "loss": 0.906, + "step": 13822 + }, + { + "epoch": 0.33, + "grad_norm": 2.4612110954666164, + "learning_rate": 1.5755332746270574e-05, + "loss": 1.0793, + "step": 13823 + }, + { + "epoch": 0.33, + "grad_norm": 1.8599375391742898, + "learning_rate": 1.5754708731047833e-05, + "loss": 1.0627, + "step": 13824 + }, + { + "epoch": 0.33, + "grad_norm": 2.374458628619297, + "learning_rate": 1.5754084682319356e-05, + "loss": 1.1042, + "step": 13825 + }, + { + "epoch": 0.33, + "grad_norm": 2.260775054832788, + "learning_rate": 1.5753460600088777e-05, + "loss": 1.0535, + "step": 13826 + }, + { + "epoch": 0.33, + "grad_norm": 1.9700194289095343, + "learning_rate": 1.5752836484359726e-05, + "loss": 1.1083, + "step": 13827 + }, + { + "epoch": 0.33, + "grad_norm": 2.1062621345156325, + "learning_rate": 1.575221233513584e-05, + "loss": 1.0587, + "step": 13828 + }, + { + "epoch": 0.33, + "grad_norm": 2.1994534195992346, + "learning_rate": 1.5751588152420757e-05, + "loss": 1.0006, + "step": 13829 + }, + { + "epoch": 0.33, + "grad_norm": 2.0015743976659293, + "learning_rate": 1.5750963936218104e-05, + "loss": 1.1517, + "step": 13830 + }, + { + "epoch": 0.33, + "grad_norm": 4.2773007913287335, + "learning_rate": 1.575033968653152e-05, + "loss": 1.0896, + "step": 13831 + }, + { + "epoch": 0.33, + "grad_norm": 1.9732387623355547, + "learning_rate": 1.5749715403364635e-05, + "loss": 0.9752, + "step": 13832 + }, + { + "epoch": 0.33, + "grad_norm": 2.362388627828658, + "learning_rate": 1.574909108672109e-05, + "loss": 0.965, + "step": 13833 + }, + { + "epoch": 0.33, + "grad_norm": 1.9962273600707865, + "learning_rate": 1.5748466736604514e-05, + "loss": 1.0445, + "step": 13834 + }, + { + "epoch": 0.33, + "grad_norm": 2.1719997041307146, + "learning_rate": 1.574784235301855e-05, + "loss": 1.0603, + "step": 13835 + }, + { + "epoch": 0.33, + "grad_norm": 1.0777106791214475, + "learning_rate": 1.574721793596682e-05, + "loss": 0.9324, + "step": 13836 + }, + { + "epoch": 0.33, + "grad_norm": 2.1297089929066395, + "learning_rate": 1.5746593485452973e-05, + "loss": 0.9253, + "step": 13837 + }, + { + "epoch": 0.33, + "grad_norm": 1.9584016941234081, + "learning_rate": 1.5745969001480637e-05, + "loss": 0.9329, + "step": 13838 + }, + { + "epoch": 0.33, + "grad_norm": 2.4768199993617452, + "learning_rate": 1.574534448405345e-05, + "loss": 0.918, + "step": 13839 + }, + { + "epoch": 0.33, + "grad_norm": 2.00501631467861, + "learning_rate": 1.5744719933175053e-05, + "loss": 1.0164, + "step": 13840 + }, + { + "epoch": 0.33, + "grad_norm": 1.9595775460916056, + "learning_rate": 1.5744095348849073e-05, + "loss": 0.9428, + "step": 13841 + }, + { + "epoch": 0.33, + "grad_norm": 1.8955305770998718, + "learning_rate": 1.5743470731079152e-05, + "loss": 1.0741, + "step": 13842 + }, + { + "epoch": 0.33, + "grad_norm": 1.8594566409185287, + "learning_rate": 1.5742846079868926e-05, + "loss": 0.9453, + "step": 13843 + }, + { + "epoch": 0.33, + "grad_norm": 2.3095537697197583, + "learning_rate": 1.5742221395222033e-05, + "loss": 1.0242, + "step": 13844 + }, + { + "epoch": 0.33, + "grad_norm": 1.137669143048339, + "learning_rate": 1.574159667714211e-05, + "loss": 1.0148, + "step": 13845 + }, + { + "epoch": 0.33, + "grad_norm": 1.8497282161720963, + "learning_rate": 1.574097192563279e-05, + "loss": 1.0265, + "step": 13846 + }, + { + "epoch": 0.33, + "grad_norm": 2.075995167912564, + "learning_rate": 1.574034714069772e-05, + "loss": 1.1269, + "step": 13847 + }, + { + "epoch": 0.33, + "grad_norm": 2.2482641836531436, + "learning_rate": 1.5739722322340524e-05, + "loss": 0.9659, + "step": 13848 + }, + { + "epoch": 0.33, + "grad_norm": 1.9147822317510053, + "learning_rate": 1.5739097470564853e-05, + "loss": 1.1781, + "step": 13849 + }, + { + "epoch": 0.33, + "grad_norm": 1.859260377235967, + "learning_rate": 1.5738472585374334e-05, + "loss": 1.0303, + "step": 13850 + }, + { + "epoch": 0.33, + "grad_norm": 1.8777382634556923, + "learning_rate": 1.5737847666772615e-05, + "loss": 1.0873, + "step": 13851 + }, + { + "epoch": 0.33, + "grad_norm": 2.153928824813373, + "learning_rate": 1.573722271476333e-05, + "loss": 1.0867, + "step": 13852 + }, + { + "epoch": 0.33, + "grad_norm": 2.7654562977142945, + "learning_rate": 1.573659772935012e-05, + "loss": 0.8572, + "step": 13853 + }, + { + "epoch": 0.33, + "grad_norm": 2.2292789754329427, + "learning_rate": 1.5735972710536623e-05, + "loss": 0.9727, + "step": 13854 + }, + { + "epoch": 0.33, + "grad_norm": 2.0520119692338423, + "learning_rate": 1.5735347658326472e-05, + "loss": 0.9282, + "step": 13855 + }, + { + "epoch": 0.33, + "grad_norm": 2.2331009602268677, + "learning_rate": 1.5734722572723313e-05, + "loss": 0.9718, + "step": 13856 + }, + { + "epoch": 0.33, + "grad_norm": 1.997357487375205, + "learning_rate": 1.573409745373078e-05, + "loss": 0.868, + "step": 13857 + }, + { + "epoch": 0.33, + "grad_norm": 2.309222276405067, + "learning_rate": 1.573347230135252e-05, + "loss": 1.0595, + "step": 13858 + }, + { + "epoch": 0.33, + "grad_norm": 1.9588615386337616, + "learning_rate": 1.573284711559217e-05, + "loss": 1.0251, + "step": 13859 + }, + { + "epoch": 0.33, + "grad_norm": 1.93181679527858, + "learning_rate": 1.5732221896453368e-05, + "loss": 1.0712, + "step": 13860 + }, + { + "epoch": 0.33, + "grad_norm": 1.9380041733539726, + "learning_rate": 1.5731596643939755e-05, + "loss": 1.1011, + "step": 13861 + }, + { + "epoch": 0.33, + "grad_norm": 1.963332645504045, + "learning_rate": 1.5730971358054972e-05, + "loss": 0.9943, + "step": 13862 + }, + { + "epoch": 0.33, + "grad_norm": 2.1290132678326406, + "learning_rate": 1.5730346038802657e-05, + "loss": 1.0284, + "step": 13863 + }, + { + "epoch": 0.33, + "grad_norm": 2.0744293112975996, + "learning_rate": 1.5729720686186457e-05, + "loss": 1.0423, + "step": 13864 + }, + { + "epoch": 0.33, + "grad_norm": 2.7251447956510373, + "learning_rate": 1.5729095300210006e-05, + "loss": 1.16, + "step": 13865 + }, + { + "epoch": 0.33, + "grad_norm": 1.9694304801642788, + "learning_rate": 1.572846988087695e-05, + "loss": 1.1765, + "step": 13866 + }, + { + "epoch": 0.33, + "grad_norm": 1.1932497427796993, + "learning_rate": 1.572784442819093e-05, + "loss": 1.0194, + "step": 13867 + }, + { + "epoch": 0.33, + "grad_norm": 2.082281708302431, + "learning_rate": 1.5727218942155588e-05, + "loss": 1.0047, + "step": 13868 + }, + { + "epoch": 0.33, + "grad_norm": 2.0829238850102985, + "learning_rate": 1.5726593422774558e-05, + "loss": 0.9705, + "step": 13869 + }, + { + "epoch": 0.33, + "grad_norm": 1.9468164987790915, + "learning_rate": 1.572596787005149e-05, + "loss": 1.178, + "step": 13870 + }, + { + "epoch": 0.33, + "grad_norm": 2.30185473626036, + "learning_rate": 1.5725342283990027e-05, + "loss": 0.9418, + "step": 13871 + }, + { + "epoch": 0.33, + "grad_norm": 1.8025697836281784, + "learning_rate": 1.572471666459381e-05, + "loss": 1.0382, + "step": 13872 + }, + { + "epoch": 0.33, + "grad_norm": 1.1292381862840455, + "learning_rate": 1.5724091011866474e-05, + "loss": 1.0421, + "step": 13873 + }, + { + "epoch": 0.33, + "grad_norm": 1.122700121463035, + "learning_rate": 1.572346532581167e-05, + "loss": 0.9869, + "step": 13874 + }, + { + "epoch": 0.33, + "grad_norm": 2.1424068169562283, + "learning_rate": 1.5722839606433043e-05, + "loss": 0.8809, + "step": 13875 + }, + { + "epoch": 0.33, + "grad_norm": 2.199396095993485, + "learning_rate": 1.5722213853734227e-05, + "loss": 0.9556, + "step": 13876 + }, + { + "epoch": 0.33, + "grad_norm": 2.3380429373429252, + "learning_rate": 1.572158806771887e-05, + "loss": 1.009, + "step": 13877 + }, + { + "epoch": 0.33, + "grad_norm": 2.1740479089356763, + "learning_rate": 1.572096224839062e-05, + "loss": 0.9426, + "step": 13878 + }, + { + "epoch": 0.33, + "grad_norm": 2.687264634602649, + "learning_rate": 1.5720336395753117e-05, + "loss": 1.1591, + "step": 13879 + }, + { + "epoch": 0.33, + "grad_norm": 1.9910007089174386, + "learning_rate": 1.571971050981e-05, + "loss": 1.0053, + "step": 13880 + }, + { + "epoch": 0.33, + "grad_norm": 1.8499365286132132, + "learning_rate": 1.571908459056492e-05, + "loss": 1.0646, + "step": 13881 + }, + { + "epoch": 0.33, + "grad_norm": 2.2244389341435813, + "learning_rate": 1.571845863802152e-05, + "loss": 0.9562, + "step": 13882 + }, + { + "epoch": 0.33, + "grad_norm": 1.936976492460255, + "learning_rate": 1.571783265218344e-05, + "loss": 1.1061, + "step": 13883 + }, + { + "epoch": 0.33, + "grad_norm": 2.0901256843485583, + "learning_rate": 1.5717206633054328e-05, + "loss": 1.0082, + "step": 13884 + }, + { + "epoch": 0.33, + "grad_norm": 2.2185258878095784, + "learning_rate": 1.5716580580637833e-05, + "loss": 0.9799, + "step": 13885 + }, + { + "epoch": 0.33, + "grad_norm": 1.8942959880264338, + "learning_rate": 1.5715954494937595e-05, + "loss": 0.9617, + "step": 13886 + }, + { + "epoch": 0.33, + "grad_norm": 2.0976688865906126, + "learning_rate": 1.5715328375957256e-05, + "loss": 1.0175, + "step": 13887 + }, + { + "epoch": 0.33, + "grad_norm": 1.1524751545855647, + "learning_rate": 1.5714702223700472e-05, + "loss": 0.982, + "step": 13888 + }, + { + "epoch": 0.33, + "grad_norm": 2.14223101012496, + "learning_rate": 1.571407603817088e-05, + "loss": 0.9118, + "step": 13889 + }, + { + "epoch": 0.33, + "grad_norm": 1.9117736611802805, + "learning_rate": 1.571344981937213e-05, + "loss": 1.1148, + "step": 13890 + }, + { + "epoch": 0.33, + "grad_norm": 1.8429845915503043, + "learning_rate": 1.571282356730786e-05, + "loss": 1.0563, + "step": 13891 + }, + { + "epoch": 0.33, + "grad_norm": 2.587272183098982, + "learning_rate": 1.5712197281981727e-05, + "loss": 1.0282, + "step": 13892 + }, + { + "epoch": 0.33, + "grad_norm": 2.0460670398304774, + "learning_rate": 1.5711570963397376e-05, + "loss": 1.0185, + "step": 13893 + }, + { + "epoch": 0.33, + "grad_norm": 2.474263077891535, + "learning_rate": 1.5710944611558445e-05, + "loss": 1.0633, + "step": 13894 + }, + { + "epoch": 0.33, + "grad_norm": 2.1339030181912038, + "learning_rate": 1.5710318226468587e-05, + "loss": 1.1078, + "step": 13895 + }, + { + "epoch": 0.33, + "grad_norm": 1.9007993125860858, + "learning_rate": 1.570969180813145e-05, + "loss": 1.0103, + "step": 13896 + }, + { + "epoch": 0.33, + "grad_norm": 2.031374516246958, + "learning_rate": 1.570906535655068e-05, + "loss": 1.066, + "step": 13897 + }, + { + "epoch": 0.33, + "grad_norm": 1.9997230456536679, + "learning_rate": 1.5708438871729925e-05, + "loss": 1.2236, + "step": 13898 + }, + { + "epoch": 0.33, + "grad_norm": 2.169744493125567, + "learning_rate": 1.570781235367283e-05, + "loss": 0.9297, + "step": 13899 + }, + { + "epoch": 0.33, + "grad_norm": 2.192903594882743, + "learning_rate": 1.5707185802383045e-05, + "loss": 1.0033, + "step": 13900 + }, + { + "epoch": 0.33, + "grad_norm": 2.059480698853028, + "learning_rate": 1.5706559217864214e-05, + "loss": 1.1425, + "step": 13901 + }, + { + "epoch": 0.33, + "grad_norm": 2.05507135839677, + "learning_rate": 1.5705932600119993e-05, + "loss": 1.1612, + "step": 13902 + }, + { + "epoch": 0.33, + "grad_norm": 1.097862383596232, + "learning_rate": 1.5705305949154025e-05, + "loss": 1.0006, + "step": 13903 + }, + { + "epoch": 0.33, + "grad_norm": 2.166035505234525, + "learning_rate": 1.570467926496996e-05, + "loss": 0.9856, + "step": 13904 + }, + { + "epoch": 0.33, + "grad_norm": 2.002689192179276, + "learning_rate": 1.5704052547571445e-05, + "loss": 1.0724, + "step": 13905 + }, + { + "epoch": 0.33, + "grad_norm": 2.076239048460125, + "learning_rate": 1.570342579696213e-05, + "loss": 1.0036, + "step": 13906 + }, + { + "epoch": 0.33, + "grad_norm": 1.1588280533320126, + "learning_rate": 1.5702799013145665e-05, + "loss": 0.9421, + "step": 13907 + }, + { + "epoch": 0.33, + "grad_norm": 1.936482434360478, + "learning_rate": 1.57021721961257e-05, + "loss": 0.8749, + "step": 13908 + }, + { + "epoch": 0.33, + "grad_norm": 1.9631250488207768, + "learning_rate": 1.570154534590588e-05, + "loss": 1.0595, + "step": 13909 + }, + { + "epoch": 0.33, + "grad_norm": 2.038234874554895, + "learning_rate": 1.570091846248986e-05, + "loss": 1.1126, + "step": 13910 + }, + { + "epoch": 0.33, + "grad_norm": 2.115562403402952, + "learning_rate": 1.570029154588129e-05, + "loss": 1.0768, + "step": 13911 + }, + { + "epoch": 0.33, + "grad_norm": 1.176481740470326, + "learning_rate": 1.5699664596083814e-05, + "loss": 0.9837, + "step": 13912 + }, + { + "epoch": 0.33, + "grad_norm": 2.168380051444522, + "learning_rate": 1.5699037613101087e-05, + "loss": 0.8771, + "step": 13913 + }, + { + "epoch": 0.33, + "grad_norm": 2.138105797316045, + "learning_rate": 1.569841059693676e-05, + "loss": 1.2382, + "step": 13914 + }, + { + "epoch": 0.33, + "grad_norm": 2.0745546245573467, + "learning_rate": 1.5697783547594482e-05, + "loss": 0.9474, + "step": 13915 + }, + { + "epoch": 0.33, + "grad_norm": 2.3047852880333126, + "learning_rate": 1.5697156465077903e-05, + "loss": 0.9661, + "step": 13916 + }, + { + "epoch": 0.33, + "grad_norm": 1.882408434100907, + "learning_rate": 1.5696529349390676e-05, + "loss": 0.991, + "step": 13917 + }, + { + "epoch": 0.33, + "grad_norm": 1.9650316554409457, + "learning_rate": 1.569590220053645e-05, + "loss": 1.0647, + "step": 13918 + }, + { + "epoch": 0.33, + "grad_norm": 1.9070357251876633, + "learning_rate": 1.5695275018518885e-05, + "loss": 1.1554, + "step": 13919 + }, + { + "epoch": 0.33, + "grad_norm": 2.0712746094613532, + "learning_rate": 1.5694647803341618e-05, + "loss": 0.9147, + "step": 13920 + }, + { + "epoch": 0.33, + "grad_norm": 2.453603531357861, + "learning_rate": 1.5694020555008314e-05, + "loss": 0.96, + "step": 13921 + }, + { + "epoch": 0.33, + "grad_norm": 2.3597249516002, + "learning_rate": 1.5693393273522618e-05, + "loss": 1.0613, + "step": 13922 + }, + { + "epoch": 0.33, + "grad_norm": 1.9239906290812132, + "learning_rate": 1.5692765958888182e-05, + "loss": 0.9597, + "step": 13923 + }, + { + "epoch": 0.33, + "grad_norm": 2.1340375643175977, + "learning_rate": 1.569213861110866e-05, + "loss": 1.0689, + "step": 13924 + }, + { + "epoch": 0.33, + "grad_norm": 2.1085738344990452, + "learning_rate": 1.5691511230187706e-05, + "loss": 0.9926, + "step": 13925 + }, + { + "epoch": 0.33, + "grad_norm": 2.2334171421775633, + "learning_rate": 1.5690883816128973e-05, + "loss": 1.0797, + "step": 13926 + }, + { + "epoch": 0.33, + "grad_norm": 2.0373290730742095, + "learning_rate": 1.569025636893611e-05, + "loss": 1.0823, + "step": 13927 + }, + { + "epoch": 0.33, + "grad_norm": 1.9460513447789662, + "learning_rate": 1.5689628888612774e-05, + "loss": 0.9255, + "step": 13928 + }, + { + "epoch": 0.33, + "grad_norm": 2.1102838406074156, + "learning_rate": 1.568900137516262e-05, + "loss": 0.9506, + "step": 13929 + }, + { + "epoch": 0.33, + "grad_norm": 1.9033621293627223, + "learning_rate": 1.5688373828589296e-05, + "loss": 1.0484, + "step": 13930 + }, + { + "epoch": 0.33, + "grad_norm": 1.2267852435926772, + "learning_rate": 1.568774624889646e-05, + "loss": 0.9945, + "step": 13931 + }, + { + "epoch": 0.33, + "grad_norm": 1.906607735984602, + "learning_rate": 1.5687118636087764e-05, + "loss": 1.0358, + "step": 13932 + }, + { + "epoch": 0.33, + "grad_norm": 1.8131972722090106, + "learning_rate": 1.5686490990166864e-05, + "loss": 0.9412, + "step": 13933 + }, + { + "epoch": 0.33, + "grad_norm": 1.8089369009725782, + "learning_rate": 1.568586331113741e-05, + "loss": 1.0237, + "step": 13934 + }, + { + "epoch": 0.33, + "grad_norm": 2.094883776035908, + "learning_rate": 1.5685235599003065e-05, + "loss": 0.9442, + "step": 13935 + }, + { + "epoch": 0.33, + "grad_norm": 2.237919102721618, + "learning_rate": 1.5684607853767474e-05, + "loss": 1.0376, + "step": 13936 + }, + { + "epoch": 0.33, + "grad_norm": 1.154011641316099, + "learning_rate": 1.5683980075434297e-05, + "loss": 1.0399, + "step": 13937 + }, + { + "epoch": 0.33, + "grad_norm": 2.1418084005244533, + "learning_rate": 1.5683352264007188e-05, + "loss": 1.0554, + "step": 13938 + }, + { + "epoch": 0.33, + "grad_norm": 1.9974118496795843, + "learning_rate": 1.5682724419489805e-05, + "loss": 1.1132, + "step": 13939 + }, + { + "epoch": 0.33, + "grad_norm": 2.074793420807898, + "learning_rate": 1.5682096541885797e-05, + "loss": 1.0798, + "step": 13940 + }, + { + "epoch": 0.33, + "grad_norm": 1.920959344663353, + "learning_rate": 1.568146863119883e-05, + "loss": 1.1001, + "step": 13941 + }, + { + "epoch": 0.33, + "grad_norm": 2.1110607949733895, + "learning_rate": 1.568084068743255e-05, + "loss": 0.9707, + "step": 13942 + }, + { + "epoch": 0.33, + "grad_norm": 2.032567852493613, + "learning_rate": 1.568021271059062e-05, + "loss": 1.1242, + "step": 13943 + }, + { + "epoch": 0.33, + "grad_norm": 2.073822751247538, + "learning_rate": 1.567958470067669e-05, + "loss": 1.1181, + "step": 13944 + }, + { + "epoch": 0.33, + "grad_norm": 1.906707335427351, + "learning_rate": 1.567895665769442e-05, + "loss": 1.0458, + "step": 13945 + }, + { + "epoch": 0.33, + "grad_norm": 2.941487728500066, + "learning_rate": 1.5678328581647467e-05, + "loss": 1.02, + "step": 13946 + }, + { + "epoch": 0.33, + "grad_norm": 1.9635129220899752, + "learning_rate": 1.567770047253949e-05, + "loss": 0.9833, + "step": 13947 + }, + { + "epoch": 0.33, + "grad_norm": 1.936268866552634, + "learning_rate": 1.5677072330374138e-05, + "loss": 1.1262, + "step": 13948 + }, + { + "epoch": 0.33, + "grad_norm": 2.874279589471173, + "learning_rate": 1.567644415515508e-05, + "loss": 0.9912, + "step": 13949 + }, + { + "epoch": 0.33, + "grad_norm": 2.0136983731660347, + "learning_rate": 1.567581594688596e-05, + "loss": 1.0403, + "step": 13950 + }, + { + "epoch": 0.33, + "grad_norm": 2.1814201483913225, + "learning_rate": 1.567518770557045e-05, + "loss": 1.1082, + "step": 13951 + }, + { + "epoch": 0.33, + "grad_norm": 1.9820260231708855, + "learning_rate": 1.5674559431212193e-05, + "loss": 1.0105, + "step": 13952 + }, + { + "epoch": 0.33, + "grad_norm": 2.1294785782362444, + "learning_rate": 1.5673931123814862e-05, + "loss": 1.0763, + "step": 13953 + }, + { + "epoch": 0.33, + "grad_norm": 1.9868239290479852, + "learning_rate": 1.5673302783382104e-05, + "loss": 1.1712, + "step": 13954 + }, + { + "epoch": 0.33, + "grad_norm": 2.0639490023623908, + "learning_rate": 1.5672674409917575e-05, + "loss": 0.9777, + "step": 13955 + }, + { + "epoch": 0.33, + "grad_norm": 3.471975617297109, + "learning_rate": 1.5672046003424948e-05, + "loss": 0.9775, + "step": 13956 + }, + { + "epoch": 0.33, + "grad_norm": 2.0240786648838878, + "learning_rate": 1.5671417563907872e-05, + "loss": 1.0777, + "step": 13957 + }, + { + "epoch": 0.33, + "grad_norm": 1.9843418963836263, + "learning_rate": 1.5670789091370005e-05, + "loss": 1.04, + "step": 13958 + }, + { + "epoch": 0.33, + "grad_norm": 2.0572706343937828, + "learning_rate": 1.567016058581501e-05, + "loss": 1.0717, + "step": 13959 + }, + { + "epoch": 0.33, + "grad_norm": 2.016128156805281, + "learning_rate": 1.5669532047246546e-05, + "loss": 1.0772, + "step": 13960 + }, + { + "epoch": 0.33, + "grad_norm": 2.184449026230754, + "learning_rate": 1.566890347566827e-05, + "loss": 1.1995, + "step": 13961 + }, + { + "epoch": 0.33, + "grad_norm": 2.033397596780634, + "learning_rate": 1.5668274871083844e-05, + "loss": 0.9879, + "step": 13962 + }, + { + "epoch": 0.33, + "grad_norm": 2.0358784783187405, + "learning_rate": 1.5667646233496925e-05, + "loss": 1.1212, + "step": 13963 + }, + { + "epoch": 0.33, + "grad_norm": 2.305947811519586, + "learning_rate": 1.566701756291118e-05, + "loss": 1.1068, + "step": 13964 + }, + { + "epoch": 0.33, + "grad_norm": 1.907897300540713, + "learning_rate": 1.566638885933026e-05, + "loss": 1.0419, + "step": 13965 + }, + { + "epoch": 0.33, + "grad_norm": 2.5092428240839197, + "learning_rate": 1.566576012275783e-05, + "loss": 1.058, + "step": 13966 + }, + { + "epoch": 0.33, + "grad_norm": 2.1666621418431533, + "learning_rate": 1.5665131353197555e-05, + "loss": 1.0643, + "step": 13967 + }, + { + "epoch": 0.33, + "grad_norm": 2.027978098157946, + "learning_rate": 1.5664502550653088e-05, + "loss": 0.9535, + "step": 13968 + }, + { + "epoch": 0.33, + "grad_norm": 1.9365181049359108, + "learning_rate": 1.566387371512809e-05, + "loss": 0.9527, + "step": 13969 + }, + { + "epoch": 0.33, + "grad_norm": 1.9312121977615158, + "learning_rate": 1.566324484662624e-05, + "loss": 1.1326, + "step": 13970 + }, + { + "epoch": 0.33, + "grad_norm": 1.9611958313614581, + "learning_rate": 1.5662615945151173e-05, + "loss": 1.128, + "step": 13971 + }, + { + "epoch": 0.33, + "grad_norm": 2.1072939987734625, + "learning_rate": 1.566198701070657e-05, + "loss": 1.0807, + "step": 13972 + }, + { + "epoch": 0.33, + "grad_norm": 2.017357956278625, + "learning_rate": 1.566135804329608e-05, + "loss": 1.0578, + "step": 13973 + }, + { + "epoch": 0.33, + "grad_norm": 2.210326025327306, + "learning_rate": 1.5660729042923374e-05, + "loss": 1.0211, + "step": 13974 + }, + { + "epoch": 0.33, + "grad_norm": 1.7836074107056754, + "learning_rate": 1.566010000959211e-05, + "loss": 0.9765, + "step": 13975 + }, + { + "epoch": 0.33, + "grad_norm": 2.033993730364186, + "learning_rate": 1.5659470943305956e-05, + "loss": 1.1677, + "step": 13976 + }, + { + "epoch": 0.33, + "grad_norm": 2.1873650351680283, + "learning_rate": 1.5658841844068568e-05, + "loss": 1.1165, + "step": 13977 + }, + { + "epoch": 0.33, + "grad_norm": 2.0187120772473337, + "learning_rate": 1.565821271188361e-05, + "loss": 1.1938, + "step": 13978 + }, + { + "epoch": 0.33, + "grad_norm": 2.0122922385451876, + "learning_rate": 1.5657583546754744e-05, + "loss": 1.193, + "step": 13979 + }, + { + "epoch": 0.33, + "grad_norm": 2.0775980989517873, + "learning_rate": 1.565695434868564e-05, + "loss": 1.1103, + "step": 13980 + }, + { + "epoch": 0.33, + "grad_norm": 2.0305609376797293, + "learning_rate": 1.5656325117679957e-05, + "loss": 1.1257, + "step": 13981 + }, + { + "epoch": 0.33, + "grad_norm": 2.070694443383698, + "learning_rate": 1.5655695853741353e-05, + "loss": 0.9617, + "step": 13982 + }, + { + "epoch": 0.33, + "grad_norm": 1.8458905058056572, + "learning_rate": 1.5655066556873502e-05, + "loss": 1.1684, + "step": 13983 + }, + { + "epoch": 0.33, + "grad_norm": 2.3605042949909696, + "learning_rate": 1.565443722708006e-05, + "loss": 0.9484, + "step": 13984 + }, + { + "epoch": 0.33, + "grad_norm": 2.16215127757457, + "learning_rate": 1.5653807864364696e-05, + "loss": 0.9669, + "step": 13985 + }, + { + "epoch": 0.33, + "grad_norm": 2.0698992593422085, + "learning_rate": 1.5653178468731076e-05, + "loss": 1.0302, + "step": 13986 + }, + { + "epoch": 0.33, + "grad_norm": 1.6751653723572424, + "learning_rate": 1.5652549040182857e-05, + "loss": 1.0702, + "step": 13987 + }, + { + "epoch": 0.33, + "grad_norm": 1.8839114213927317, + "learning_rate": 1.5651919578723708e-05, + "loss": 0.9653, + "step": 13988 + }, + { + "epoch": 0.33, + "grad_norm": 2.240237641664732, + "learning_rate": 1.5651290084357295e-05, + "loss": 1.1333, + "step": 13989 + }, + { + "epoch": 0.33, + "grad_norm": 2.207199322896385, + "learning_rate": 1.5650660557087284e-05, + "loss": 1.0238, + "step": 13990 + }, + { + "epoch": 0.33, + "grad_norm": 2.249879181452605, + "learning_rate": 1.5650030996917334e-05, + "loss": 1.177, + "step": 13991 + }, + { + "epoch": 0.33, + "grad_norm": 1.9928746744627401, + "learning_rate": 1.5649401403851115e-05, + "loss": 0.9846, + "step": 13992 + }, + { + "epoch": 0.33, + "grad_norm": 1.1279283079741658, + "learning_rate": 1.5648771777892296e-05, + "loss": 0.9248, + "step": 13993 + }, + { + "epoch": 0.33, + "grad_norm": 1.9926717080913812, + "learning_rate": 1.564814211904454e-05, + "loss": 1.1236, + "step": 13994 + }, + { + "epoch": 0.33, + "grad_norm": 2.0255306855545787, + "learning_rate": 1.5647512427311508e-05, + "loss": 1.0661, + "step": 13995 + }, + { + "epoch": 0.33, + "grad_norm": 2.0604384597171257, + "learning_rate": 1.564688270269687e-05, + "loss": 0.9555, + "step": 13996 + }, + { + "epoch": 0.33, + "grad_norm": 1.1783321259015642, + "learning_rate": 1.5646252945204298e-05, + "loss": 0.9508, + "step": 13997 + }, + { + "epoch": 0.33, + "grad_norm": 1.8362375470762315, + "learning_rate": 1.5645623154837454e-05, + "loss": 1.0508, + "step": 13998 + }, + { + "epoch": 0.33, + "grad_norm": 1.9715536484906595, + "learning_rate": 1.56449933316e-05, + "loss": 1.083, + "step": 13999 + }, + { + "epoch": 0.33, + "grad_norm": 1.9456184551870384, + "learning_rate": 1.5644363475495614e-05, + "loss": 1.0602, + "step": 14000 + }, + { + "epoch": 0.33, + "grad_norm": 1.0722524756353167, + "learning_rate": 1.564373358652795e-05, + "loss": 0.9443, + "step": 14001 + }, + { + "epoch": 0.33, + "grad_norm": 1.0723894469887891, + "learning_rate": 1.5643103664700687e-05, + "loss": 0.9107, + "step": 14002 + }, + { + "epoch": 0.33, + "grad_norm": 2.194504991830406, + "learning_rate": 1.564247371001748e-05, + "loss": 1.1025, + "step": 14003 + }, + { + "epoch": 0.33, + "grad_norm": 2.092428597748992, + "learning_rate": 1.5641843722482015e-05, + "loss": 0.8988, + "step": 14004 + }, + { + "epoch": 0.33, + "grad_norm": 1.9406881254487494, + "learning_rate": 1.5641213702097947e-05, + "loss": 1.0458, + "step": 14005 + }, + { + "epoch": 0.33, + "grad_norm": 1.9857000342069357, + "learning_rate": 1.5640583648868943e-05, + "loss": 1.0104, + "step": 14006 + }, + { + "epoch": 0.33, + "grad_norm": 2.2604680024838824, + "learning_rate": 1.5639953562798676e-05, + "loss": 1.1308, + "step": 14007 + }, + { + "epoch": 0.33, + "grad_norm": 2.06905509319778, + "learning_rate": 1.563932344389082e-05, + "loss": 1.0991, + "step": 14008 + }, + { + "epoch": 0.33, + "grad_norm": 2.056166954663502, + "learning_rate": 1.563869329214903e-05, + "loss": 1.0377, + "step": 14009 + }, + { + "epoch": 0.33, + "grad_norm": 2.1637539996587303, + "learning_rate": 1.5638063107576983e-05, + "loss": 0.985, + "step": 14010 + }, + { + "epoch": 0.33, + "grad_norm": 1.9960059575428277, + "learning_rate": 1.5637432890178354e-05, + "loss": 1.008, + "step": 14011 + }, + { + "epoch": 0.33, + "grad_norm": 1.867488725432827, + "learning_rate": 1.5636802639956798e-05, + "loss": 1.1743, + "step": 14012 + }, + { + "epoch": 0.33, + "grad_norm": 2.2814203238141317, + "learning_rate": 1.5636172356915995e-05, + "loss": 1.0593, + "step": 14013 + }, + { + "epoch": 0.33, + "grad_norm": 2.0732777473317308, + "learning_rate": 1.5635542041059615e-05, + "loss": 0.8367, + "step": 14014 + }, + { + "epoch": 0.33, + "grad_norm": 1.815287821363382, + "learning_rate": 1.5634911692391323e-05, + "loss": 1.0319, + "step": 14015 + }, + { + "epoch": 0.33, + "grad_norm": 2.078736171613605, + "learning_rate": 1.5634281310914792e-05, + "loss": 1.0198, + "step": 14016 + }, + { + "epoch": 0.33, + "grad_norm": 1.9031710246258007, + "learning_rate": 1.563365089663369e-05, + "loss": 1.0798, + "step": 14017 + }, + { + "epoch": 0.33, + "grad_norm": 1.83274849261239, + "learning_rate": 1.5633020449551694e-05, + "loss": 0.9776, + "step": 14018 + }, + { + "epoch": 0.33, + "grad_norm": 2.4831277341231104, + "learning_rate": 1.5632389969672462e-05, + "loss": 1.0686, + "step": 14019 + }, + { + "epoch": 0.33, + "grad_norm": 1.9441772766942962, + "learning_rate": 1.5631759456999675e-05, + "loss": 0.9925, + "step": 14020 + }, + { + "epoch": 0.33, + "grad_norm": 2.0092331475858494, + "learning_rate": 1.5631128911537007e-05, + "loss": 1.0018, + "step": 14021 + }, + { + "epoch": 0.33, + "grad_norm": 1.9829189393807838, + "learning_rate": 1.5630498333288115e-05, + "loss": 1.1349, + "step": 14022 + }, + { + "epoch": 0.33, + "grad_norm": 2.623239827178082, + "learning_rate": 1.5629867722256684e-05, + "loss": 1.1046, + "step": 14023 + }, + { + "epoch": 0.33, + "grad_norm": 1.0954850182905238, + "learning_rate": 1.5629237078446378e-05, + "loss": 0.9758, + "step": 14024 + }, + { + "epoch": 0.33, + "grad_norm": 2.0082958848811474, + "learning_rate": 1.5628606401860875e-05, + "loss": 1.0013, + "step": 14025 + }, + { + "epoch": 0.33, + "grad_norm": 1.8953927011043348, + "learning_rate": 1.562797569250384e-05, + "loss": 1.1062, + "step": 14026 + }, + { + "epoch": 0.33, + "grad_norm": 2.1939646578039187, + "learning_rate": 1.562734495037895e-05, + "loss": 0.8988, + "step": 14027 + }, + { + "epoch": 0.33, + "grad_norm": 2.102968623816445, + "learning_rate": 1.5626714175489877e-05, + "loss": 1.1059, + "step": 14028 + }, + { + "epoch": 0.33, + "grad_norm": 2.1814149953977875, + "learning_rate": 1.5626083367840296e-05, + "loss": 1.0847, + "step": 14029 + }, + { + "epoch": 0.33, + "grad_norm": 1.149339662050484, + "learning_rate": 1.562545252743387e-05, + "loss": 1.0226, + "step": 14030 + }, + { + "epoch": 0.33, + "grad_norm": 2.666062827705481, + "learning_rate": 1.562482165427428e-05, + "loss": 0.9998, + "step": 14031 + }, + { + "epoch": 0.33, + "grad_norm": 2.8011380307379663, + "learning_rate": 1.56241907483652e-05, + "loss": 0.9562, + "step": 14032 + }, + { + "epoch": 0.33, + "grad_norm": 2.655034982521722, + "learning_rate": 1.56235598097103e-05, + "loss": 1.0385, + "step": 14033 + }, + { + "epoch": 0.33, + "grad_norm": 2.003033828259898, + "learning_rate": 1.5622928838313254e-05, + "loss": 1.0489, + "step": 14034 + }, + { + "epoch": 0.33, + "grad_norm": 2.178728471935062, + "learning_rate": 1.5622297834177738e-05, + "loss": 1.0026, + "step": 14035 + }, + { + "epoch": 0.33, + "grad_norm": 3.2447861441034704, + "learning_rate": 1.562166679730742e-05, + "loss": 1.12, + "step": 14036 + }, + { + "epoch": 0.33, + "grad_norm": 2.0018625194701367, + "learning_rate": 1.5621035727705982e-05, + "loss": 0.9971, + "step": 14037 + }, + { + "epoch": 0.33, + "grad_norm": 2.367298507998981, + "learning_rate": 1.562040462537709e-05, + "loss": 1.024, + "step": 14038 + }, + { + "epoch": 0.33, + "grad_norm": 1.1845367647815548, + "learning_rate": 1.5619773490324427e-05, + "loss": 0.9296, + "step": 14039 + }, + { + "epoch": 0.33, + "grad_norm": 2.1497848570910563, + "learning_rate": 1.561914232255166e-05, + "loss": 1.045, + "step": 14040 + }, + { + "epoch": 0.33, + "grad_norm": 2.151526274450581, + "learning_rate": 1.5618511122062475e-05, + "loss": 1.005, + "step": 14041 + }, + { + "epoch": 0.33, + "grad_norm": 1.8897854903148184, + "learning_rate": 1.5617879888860533e-05, + "loss": 1.067, + "step": 14042 + }, + { + "epoch": 0.33, + "grad_norm": 2.097270186378974, + "learning_rate": 1.5617248622949515e-05, + "loss": 1.0376, + "step": 14043 + }, + { + "epoch": 0.33, + "grad_norm": 2.1594529884745874, + "learning_rate": 1.56166173243331e-05, + "loss": 0.9731, + "step": 14044 + }, + { + "epoch": 0.33, + "grad_norm": 2.289084667124517, + "learning_rate": 1.5615985993014956e-05, + "loss": 1.0379, + "step": 14045 + }, + { + "epoch": 0.33, + "grad_norm": 1.1387184192442694, + "learning_rate": 1.561535462899877e-05, + "loss": 0.8862, + "step": 14046 + }, + { + "epoch": 0.33, + "grad_norm": 2.209513435514339, + "learning_rate": 1.561472323228821e-05, + "loss": 0.973, + "step": 14047 + }, + { + "epoch": 0.33, + "grad_norm": 1.1176874398472962, + "learning_rate": 1.561409180288695e-05, + "loss": 0.9572, + "step": 14048 + }, + { + "epoch": 0.33, + "grad_norm": 2.32510678302227, + "learning_rate": 1.5613460340798674e-05, + "loss": 1.0976, + "step": 14049 + }, + { + "epoch": 0.33, + "grad_norm": 1.907401488525454, + "learning_rate": 1.5612828846027052e-05, + "loss": 1.0105, + "step": 14050 + }, + { + "epoch": 0.33, + "grad_norm": 1.0764177317615378, + "learning_rate": 1.5612197318575767e-05, + "loss": 0.9188, + "step": 14051 + }, + { + "epoch": 0.33, + "grad_norm": 2.2300076227224213, + "learning_rate": 1.561156575844849e-05, + "loss": 1.0855, + "step": 14052 + }, + { + "epoch": 0.33, + "grad_norm": 2.2241790696930135, + "learning_rate": 1.56109341656489e-05, + "loss": 0.986, + "step": 14053 + }, + { + "epoch": 0.33, + "grad_norm": 2.0599947552962683, + "learning_rate": 1.5610302540180674e-05, + "loss": 0.9985, + "step": 14054 + }, + { + "epoch": 0.33, + "grad_norm": 2.1513156008065795, + "learning_rate": 1.5609670882047492e-05, + "loss": 1.0635, + "step": 14055 + }, + { + "epoch": 0.33, + "grad_norm": 1.8319321835758158, + "learning_rate": 1.5609039191253032e-05, + "loss": 0.9093, + "step": 14056 + }, + { + "epoch": 0.33, + "grad_norm": 3.46495304652956, + "learning_rate": 1.560840746780097e-05, + "loss": 1.0592, + "step": 14057 + }, + { + "epoch": 0.33, + "grad_norm": 2.0729146737789192, + "learning_rate": 1.560777571169498e-05, + "loss": 0.9881, + "step": 14058 + }, + { + "epoch": 0.33, + "grad_norm": 2.3022516680353666, + "learning_rate": 1.5607143922938747e-05, + "loss": 1.0259, + "step": 14059 + }, + { + "epoch": 0.33, + "grad_norm": 1.9310809194568153, + "learning_rate": 1.5606512101535947e-05, + "loss": 1.0164, + "step": 14060 + }, + { + "epoch": 0.33, + "grad_norm": 1.758227142199821, + "learning_rate": 1.560588024749026e-05, + "loss": 0.9828, + "step": 14061 + }, + { + "epoch": 0.33, + "grad_norm": 2.5464704763109545, + "learning_rate": 1.5605248360805358e-05, + "loss": 0.9425, + "step": 14062 + }, + { + "epoch": 0.33, + "grad_norm": 1.8402838624124447, + "learning_rate": 1.560461644148493e-05, + "loss": 1.0195, + "step": 14063 + }, + { + "epoch": 0.33, + "grad_norm": 2.92690354149255, + "learning_rate": 1.560398448953265e-05, + "loss": 0.9594, + "step": 14064 + }, + { + "epoch": 0.33, + "grad_norm": 1.8606683674092745, + "learning_rate": 1.56033525049522e-05, + "loss": 1.0145, + "step": 14065 + }, + { + "epoch": 0.33, + "grad_norm": 2.021443206718047, + "learning_rate": 1.560272048774725e-05, + "loss": 1.1911, + "step": 14066 + }, + { + "epoch": 0.33, + "grad_norm": 1.7557934040824092, + "learning_rate": 1.5602088437921497e-05, + "loss": 0.9376, + "step": 14067 + }, + { + "epoch": 0.33, + "grad_norm": 2.004428673370946, + "learning_rate": 1.5601456355478606e-05, + "loss": 0.957, + "step": 14068 + }, + { + "epoch": 0.33, + "grad_norm": 2.0582336026145325, + "learning_rate": 1.5600824240422266e-05, + "loss": 1.097, + "step": 14069 + }, + { + "epoch": 0.33, + "grad_norm": 1.9290352634975434, + "learning_rate": 1.5600192092756154e-05, + "loss": 1.0043, + "step": 14070 + }, + { + "epoch": 0.33, + "grad_norm": 2.1653956445258067, + "learning_rate": 1.5599559912483945e-05, + "loss": 1.0482, + "step": 14071 + }, + { + "epoch": 0.33, + "grad_norm": 1.9004813879035156, + "learning_rate": 1.559892769960933e-05, + "loss": 0.9944, + "step": 14072 + }, + { + "epoch": 0.33, + "grad_norm": 1.0795622308375235, + "learning_rate": 1.5598295454135985e-05, + "loss": 0.9625, + "step": 14073 + }, + { + "epoch": 0.33, + "grad_norm": 1.1290839250952507, + "learning_rate": 1.5597663176067596e-05, + "loss": 0.9395, + "step": 14074 + }, + { + "epoch": 0.33, + "grad_norm": 1.1199257156115283, + "learning_rate": 1.5597030865407828e-05, + "loss": 1.0628, + "step": 14075 + }, + { + "epoch": 0.33, + "grad_norm": 2.088776430643579, + "learning_rate": 1.5596398522160385e-05, + "loss": 1.0615, + "step": 14076 + }, + { + "epoch": 0.33, + "grad_norm": 2.0430881893731105, + "learning_rate": 1.5595766146328934e-05, + "loss": 1.0919, + "step": 14077 + }, + { + "epoch": 0.33, + "grad_norm": 2.2498547394125334, + "learning_rate": 1.5595133737917163e-05, + "loss": 1.1361, + "step": 14078 + }, + { + "epoch": 0.33, + "grad_norm": 2.151777697984869, + "learning_rate": 1.559450129692875e-05, + "loss": 1.028, + "step": 14079 + }, + { + "epoch": 0.33, + "grad_norm": 2.9915692191194574, + "learning_rate": 1.5593868823367378e-05, + "loss": 1.1064, + "step": 14080 + }, + { + "epoch": 0.33, + "grad_norm": 2.030711371849527, + "learning_rate": 1.5593236317236733e-05, + "loss": 0.8772, + "step": 14081 + }, + { + "epoch": 0.33, + "grad_norm": 1.9354378323581192, + "learning_rate": 1.5592603778540493e-05, + "loss": 1.023, + "step": 14082 + }, + { + "epoch": 0.33, + "grad_norm": 2.1247941365126004, + "learning_rate": 1.5591971207282345e-05, + "loss": 0.9895, + "step": 14083 + }, + { + "epoch": 0.33, + "grad_norm": 2.619005074262986, + "learning_rate": 1.5591338603465972e-05, + "loss": 1.1123, + "step": 14084 + }, + { + "epoch": 0.33, + "grad_norm": 1.9446485807873435, + "learning_rate": 1.5590705967095053e-05, + "loss": 1.0497, + "step": 14085 + }, + { + "epoch": 0.33, + "grad_norm": 1.937783448321784, + "learning_rate": 1.5590073298173275e-05, + "loss": 0.9398, + "step": 14086 + }, + { + "epoch": 0.33, + "grad_norm": 1.8945195561470862, + "learning_rate": 1.558944059670432e-05, + "loss": 0.9567, + "step": 14087 + }, + { + "epoch": 0.33, + "grad_norm": 1.9286431819748346, + "learning_rate": 1.5588807862691874e-05, + "loss": 1.0011, + "step": 14088 + }, + { + "epoch": 0.33, + "grad_norm": 2.8136300356607515, + "learning_rate": 1.5588175096139614e-05, + "loss": 0.9312, + "step": 14089 + }, + { + "epoch": 0.33, + "grad_norm": 1.8732678868603398, + "learning_rate": 1.5587542297051233e-05, + "loss": 1.0921, + "step": 14090 + }, + { + "epoch": 0.33, + "grad_norm": 2.375548998457617, + "learning_rate": 1.5586909465430414e-05, + "loss": 1.0418, + "step": 14091 + }, + { + "epoch": 0.33, + "grad_norm": 1.893798145111126, + "learning_rate": 1.5586276601280833e-05, + "loss": 0.8446, + "step": 14092 + }, + { + "epoch": 0.33, + "grad_norm": 2.548601300918623, + "learning_rate": 1.558564370460619e-05, + "loss": 1.099, + "step": 14093 + }, + { + "epoch": 0.33, + "grad_norm": 1.9379470917651307, + "learning_rate": 1.5585010775410153e-05, + "loss": 1.0698, + "step": 14094 + }, + { + "epoch": 0.33, + "grad_norm": 1.8591921324723661, + "learning_rate": 1.5584377813696417e-05, + "loss": 1.0325, + "step": 14095 + }, + { + "epoch": 0.33, + "grad_norm": 1.8376261673381047, + "learning_rate": 1.5583744819468667e-05, + "loss": 0.9526, + "step": 14096 + }, + { + "epoch": 0.33, + "grad_norm": 2.0039377829908505, + "learning_rate": 1.5583111792730587e-05, + "loss": 1.1298, + "step": 14097 + }, + { + "epoch": 0.33, + "grad_norm": 1.88853700391963, + "learning_rate": 1.558247873348586e-05, + "loss": 0.9695, + "step": 14098 + }, + { + "epoch": 0.33, + "grad_norm": 1.9117011618623565, + "learning_rate": 1.558184564173818e-05, + "loss": 1.0635, + "step": 14099 + }, + { + "epoch": 0.33, + "grad_norm": 2.01207195602479, + "learning_rate": 1.5581212517491224e-05, + "loss": 1.1292, + "step": 14100 + }, + { + "epoch": 0.33, + "grad_norm": 2.1521876871582304, + "learning_rate": 1.5580579360748676e-05, + "loss": 1.0581, + "step": 14101 + }, + { + "epoch": 0.33, + "grad_norm": 1.986126657120144, + "learning_rate": 1.5579946171514237e-05, + "loss": 0.9841, + "step": 14102 + }, + { + "epoch": 0.33, + "grad_norm": 2.0329818216371724, + "learning_rate": 1.557931294979158e-05, + "loss": 0.9054, + "step": 14103 + }, + { + "epoch": 0.33, + "grad_norm": 1.120065937037954, + "learning_rate": 1.5578679695584398e-05, + "loss": 0.941, + "step": 14104 + }, + { + "epoch": 0.33, + "grad_norm": 2.3043355493214923, + "learning_rate": 1.5578046408896378e-05, + "loss": 0.9491, + "step": 14105 + }, + { + "epoch": 0.33, + "grad_norm": 2.052416991969691, + "learning_rate": 1.5577413089731204e-05, + "loss": 0.9687, + "step": 14106 + }, + { + "epoch": 0.33, + "grad_norm": 2.2759959069684284, + "learning_rate": 1.5576779738092566e-05, + "loss": 1.0725, + "step": 14107 + }, + { + "epoch": 0.33, + "grad_norm": 2.0688102153598025, + "learning_rate": 1.557614635398415e-05, + "loss": 1.2171, + "step": 14108 + }, + { + "epoch": 0.33, + "grad_norm": 1.9988574909315997, + "learning_rate": 1.5575512937409648e-05, + "loss": 0.9824, + "step": 14109 + }, + { + "epoch": 0.33, + "grad_norm": 2.264887306953161, + "learning_rate": 1.5574879488372738e-05, + "loss": 1.0021, + "step": 14110 + }, + { + "epoch": 0.33, + "grad_norm": 1.984218720675397, + "learning_rate": 1.5574246006877118e-05, + "loss": 1.101, + "step": 14111 + }, + { + "epoch": 0.33, + "grad_norm": 2.083535531955648, + "learning_rate": 1.5573612492926472e-05, + "loss": 0.9922, + "step": 14112 + }, + { + "epoch": 0.33, + "grad_norm": 1.8336950775750085, + "learning_rate": 1.5572978946524493e-05, + "loss": 0.9972, + "step": 14113 + }, + { + "epoch": 0.33, + "grad_norm": 2.1197170632700564, + "learning_rate": 1.5572345367674864e-05, + "loss": 1.0178, + "step": 14114 + }, + { + "epoch": 0.33, + "grad_norm": 1.934698022895201, + "learning_rate": 1.5571711756381274e-05, + "loss": 1.0824, + "step": 14115 + }, + { + "epoch": 0.33, + "grad_norm": 1.958541890861675, + "learning_rate": 1.557107811264742e-05, + "loss": 1.0038, + "step": 14116 + }, + { + "epoch": 0.33, + "grad_norm": 2.3469782451645154, + "learning_rate": 1.557044443647698e-05, + "loss": 0.9219, + "step": 14117 + }, + { + "epoch": 0.33, + "grad_norm": 3.01211973315172, + "learning_rate": 1.5569810727873652e-05, + "loss": 1.0829, + "step": 14118 + }, + { + "epoch": 0.33, + "grad_norm": 2.0585106401636666, + "learning_rate": 1.556917698684112e-05, + "loss": 1.027, + "step": 14119 + }, + { + "epoch": 0.33, + "grad_norm": 2.1199556797765835, + "learning_rate": 1.5568543213383076e-05, + "loss": 0.9911, + "step": 14120 + }, + { + "epoch": 0.33, + "grad_norm": 1.7328076596983615, + "learning_rate": 1.5567909407503213e-05, + "loss": 1.1727, + "step": 14121 + }, + { + "epoch": 0.33, + "grad_norm": 1.2109977824449736, + "learning_rate": 1.5567275569205216e-05, + "loss": 0.9741, + "step": 14122 + }, + { + "epoch": 0.33, + "grad_norm": 2.20374638992019, + "learning_rate": 1.556664169849278e-05, + "loss": 0.9841, + "step": 14123 + }, + { + "epoch": 0.33, + "grad_norm": 1.2033198317116764, + "learning_rate": 1.5566007795369594e-05, + "loss": 1.0119, + "step": 14124 + }, + { + "epoch": 0.33, + "grad_norm": 1.8445068555939128, + "learning_rate": 1.5565373859839348e-05, + "loss": 0.9502, + "step": 14125 + }, + { + "epoch": 0.33, + "grad_norm": 2.0770889981243443, + "learning_rate": 1.5564739891905734e-05, + "loss": 0.974, + "step": 14126 + }, + { + "epoch": 0.33, + "grad_norm": 2.186910039246124, + "learning_rate": 1.556410589157244e-05, + "loss": 0.9853, + "step": 14127 + }, + { + "epoch": 0.33, + "grad_norm": 1.052820787240618, + "learning_rate": 1.5563471858843163e-05, + "loss": 0.9137, + "step": 14128 + }, + { + "epoch": 0.33, + "grad_norm": 1.8470419528271869, + "learning_rate": 1.556283779372159e-05, + "loss": 1.0266, + "step": 14129 + }, + { + "epoch": 0.33, + "grad_norm": 1.9807485732159686, + "learning_rate": 1.5562203696211416e-05, + "loss": 1.0419, + "step": 14130 + }, + { + "epoch": 0.33, + "grad_norm": 1.0622340364259524, + "learning_rate": 1.5561569566316327e-05, + "loss": 0.9091, + "step": 14131 + }, + { + "epoch": 0.33, + "grad_norm": 2.3483046756486923, + "learning_rate": 1.5560935404040023e-05, + "loss": 1.0995, + "step": 14132 + }, + { + "epoch": 0.33, + "grad_norm": 1.8997149854252409, + "learning_rate": 1.556030120938619e-05, + "loss": 1.0468, + "step": 14133 + }, + { + "epoch": 0.33, + "grad_norm": 1.9853007050430225, + "learning_rate": 1.5559666982358525e-05, + "loss": 1.0384, + "step": 14134 + }, + { + "epoch": 0.33, + "grad_norm": 1.991188074792642, + "learning_rate": 1.555903272296072e-05, + "loss": 1.0308, + "step": 14135 + }, + { + "epoch": 0.33, + "grad_norm": 2.273506610842398, + "learning_rate": 1.5558398431196465e-05, + "loss": 0.9832, + "step": 14136 + }, + { + "epoch": 0.33, + "grad_norm": 2.000393012981092, + "learning_rate": 1.5557764107069454e-05, + "loss": 1.0041, + "step": 14137 + }, + { + "epoch": 0.33, + "grad_norm": 2.348137838267249, + "learning_rate": 1.555712975058338e-05, + "loss": 1.0634, + "step": 14138 + }, + { + "epoch": 0.33, + "grad_norm": 2.0382351943459764, + "learning_rate": 1.5556495361741936e-05, + "loss": 1.0668, + "step": 14139 + }, + { + "epoch": 0.33, + "grad_norm": 1.8573348262367024, + "learning_rate": 1.555586094054882e-05, + "loss": 0.9955, + "step": 14140 + }, + { + "epoch": 0.33, + "grad_norm": 2.0566729605769174, + "learning_rate": 1.5555226487007723e-05, + "loss": 1.1049, + "step": 14141 + }, + { + "epoch": 0.33, + "grad_norm": 2.3555849023548827, + "learning_rate": 1.5554592001122335e-05, + "loss": 1.0276, + "step": 14142 + }, + { + "epoch": 0.33, + "grad_norm": 1.9617068387529804, + "learning_rate": 1.5553957482896356e-05, + "loss": 1.0969, + "step": 14143 + }, + { + "epoch": 0.33, + "grad_norm": 2.057156180634002, + "learning_rate": 1.5553322932333477e-05, + "loss": 0.9783, + "step": 14144 + }, + { + "epoch": 0.33, + "grad_norm": 2.0644912815447483, + "learning_rate": 1.5552688349437396e-05, + "loss": 1.0002, + "step": 14145 + }, + { + "epoch": 0.33, + "grad_norm": 3.33989916199175, + "learning_rate": 1.5552053734211805e-05, + "loss": 1.0015, + "step": 14146 + }, + { + "epoch": 0.33, + "grad_norm": 2.186049067453203, + "learning_rate": 1.55514190866604e-05, + "loss": 1.0801, + "step": 14147 + }, + { + "epoch": 0.33, + "grad_norm": 1.1283131863991547, + "learning_rate": 1.5550784406786875e-05, + "loss": 0.9445, + "step": 14148 + }, + { + "epoch": 0.33, + "grad_norm": 2.310289754773649, + "learning_rate": 1.5550149694594924e-05, + "loss": 1.0008, + "step": 14149 + }, + { + "epoch": 0.33, + "grad_norm": 2.0222460321986344, + "learning_rate": 1.5549514950088243e-05, + "loss": 0.9854, + "step": 14150 + }, + { + "epoch": 0.33, + "grad_norm": 2.878366991777512, + "learning_rate": 1.554888017327053e-05, + "loss": 1.1151, + "step": 14151 + }, + { + "epoch": 0.33, + "grad_norm": 1.9059432352189944, + "learning_rate": 1.5548245364145486e-05, + "loss": 0.9133, + "step": 14152 + }, + { + "epoch": 0.33, + "grad_norm": 2.2640805927674865, + "learning_rate": 1.5547610522716796e-05, + "loss": 0.9036, + "step": 14153 + }, + { + "epoch": 0.33, + "grad_norm": 2.497331881930807, + "learning_rate": 1.5546975648988157e-05, + "loss": 1.0046, + "step": 14154 + }, + { + "epoch": 0.33, + "grad_norm": 2.142022668656059, + "learning_rate": 1.5546340742963278e-05, + "loss": 1.0561, + "step": 14155 + }, + { + "epoch": 0.33, + "grad_norm": 2.0792412287654485, + "learning_rate": 1.554570580464584e-05, + "loss": 1.1951, + "step": 14156 + }, + { + "epoch": 0.33, + "grad_norm": 1.989300460292721, + "learning_rate": 1.5545070834039556e-05, + "loss": 0.9894, + "step": 14157 + }, + { + "epoch": 0.33, + "grad_norm": 1.9478069820808623, + "learning_rate": 1.5544435831148107e-05, + "loss": 1.0671, + "step": 14158 + }, + { + "epoch": 0.33, + "grad_norm": 2.078302132469454, + "learning_rate": 1.55438007959752e-05, + "loss": 0.933, + "step": 14159 + }, + { + "epoch": 0.33, + "grad_norm": 1.9621090737467342, + "learning_rate": 1.554316572852453e-05, + "loss": 0.8559, + "step": 14160 + }, + { + "epoch": 0.33, + "grad_norm": 1.0956761741256067, + "learning_rate": 1.554253062879979e-05, + "loss": 1.0032, + "step": 14161 + }, + { + "epoch": 0.33, + "grad_norm": 2.4431027675793566, + "learning_rate": 1.5541895496804686e-05, + "loss": 0.9742, + "step": 14162 + }, + { + "epoch": 0.33, + "grad_norm": 2.8039256005902886, + "learning_rate": 1.5541260332542906e-05, + "loss": 1.1191, + "step": 14163 + }, + { + "epoch": 0.33, + "grad_norm": 2.101396160340833, + "learning_rate": 1.554062513601816e-05, + "loss": 1.0608, + "step": 14164 + }, + { + "epoch": 0.33, + "grad_norm": 1.9865183694613742, + "learning_rate": 1.553998990723414e-05, + "loss": 0.9592, + "step": 14165 + }, + { + "epoch": 0.33, + "grad_norm": 2.0350510411928036, + "learning_rate": 1.5539354646194544e-05, + "loss": 1.0046, + "step": 14166 + }, + { + "epoch": 0.33, + "grad_norm": 1.1572228842085313, + "learning_rate": 1.553871935290307e-05, + "loss": 1.0759, + "step": 14167 + }, + { + "epoch": 0.33, + "grad_norm": 2.0279471675639678, + "learning_rate": 1.553808402736342e-05, + "loss": 1.0439, + "step": 14168 + }, + { + "epoch": 0.33, + "grad_norm": 2.1633447671995776, + "learning_rate": 1.553744866957929e-05, + "loss": 0.9217, + "step": 14169 + }, + { + "epoch": 0.33, + "grad_norm": 1.914369407931541, + "learning_rate": 1.553681327955438e-05, + "loss": 1.1093, + "step": 14170 + }, + { + "epoch": 0.33, + "grad_norm": 2.482815697372639, + "learning_rate": 1.5536177857292392e-05, + "loss": 1.0024, + "step": 14171 + }, + { + "epoch": 0.33, + "grad_norm": 2.39570988282184, + "learning_rate": 1.5535542402797027e-05, + "loss": 1.0176, + "step": 14172 + }, + { + "epoch": 0.33, + "grad_norm": 2.7365044878011613, + "learning_rate": 1.5534906916071976e-05, + "loss": 1.1878, + "step": 14173 + }, + { + "epoch": 0.33, + "grad_norm": 1.1302366388904983, + "learning_rate": 1.5534271397120947e-05, + "loss": 0.9532, + "step": 14174 + }, + { + "epoch": 0.33, + "grad_norm": 1.9332473696297208, + "learning_rate": 1.553363584594764e-05, + "loss": 1.0518, + "step": 14175 + }, + { + "epoch": 0.33, + "grad_norm": 2.1462243495800712, + "learning_rate": 1.553300026255575e-05, + "loss": 1.064, + "step": 14176 + }, + { + "epoch": 0.33, + "grad_norm": 2.175260701031158, + "learning_rate": 1.553236464694898e-05, + "loss": 1.0806, + "step": 14177 + }, + { + "epoch": 0.33, + "grad_norm": 2.0141069402311436, + "learning_rate": 1.5531728999131034e-05, + "loss": 1.0701, + "step": 14178 + }, + { + "epoch": 0.33, + "grad_norm": 1.126117906586408, + "learning_rate": 1.553109331910561e-05, + "loss": 0.9394, + "step": 14179 + }, + { + "epoch": 0.33, + "grad_norm": 1.9609332481182402, + "learning_rate": 1.553045760687641e-05, + "loss": 0.9742, + "step": 14180 + }, + { + "epoch": 0.33, + "grad_norm": 1.1114793102257334, + "learning_rate": 1.5529821862447133e-05, + "loss": 1.0415, + "step": 14181 + }, + { + "epoch": 0.33, + "grad_norm": 2.139123277187378, + "learning_rate": 1.5529186085821485e-05, + "loss": 0.9948, + "step": 14182 + }, + { + "epoch": 0.33, + "grad_norm": 1.1878557425247314, + "learning_rate": 1.552855027700316e-05, + "loss": 0.9956, + "step": 14183 + }, + { + "epoch": 0.33, + "grad_norm": 1.9913032069771077, + "learning_rate": 1.552791443599587e-05, + "loss": 1.1019, + "step": 14184 + }, + { + "epoch": 0.33, + "grad_norm": 2.3576403784644286, + "learning_rate": 1.5527278562803313e-05, + "loss": 1.1175, + "step": 14185 + }, + { + "epoch": 0.33, + "grad_norm": 3.045983883028296, + "learning_rate": 1.552664265742919e-05, + "loss": 1.0761, + "step": 14186 + }, + { + "epoch": 0.33, + "grad_norm": 1.8786263196633224, + "learning_rate": 1.5526006719877198e-05, + "loss": 1.1267, + "step": 14187 + }, + { + "epoch": 0.33, + "grad_norm": 2.2869670545314205, + "learning_rate": 1.552537075015105e-05, + "loss": 0.9389, + "step": 14188 + }, + { + "epoch": 0.33, + "grad_norm": 1.9646421094639477, + "learning_rate": 1.5524734748254442e-05, + "loss": 1.0528, + "step": 14189 + }, + { + "epoch": 0.33, + "grad_norm": 1.8494577875678002, + "learning_rate": 1.552409871419108e-05, + "loss": 0.993, + "step": 14190 + }, + { + "epoch": 0.33, + "grad_norm": 1.1670572968998343, + "learning_rate": 1.5523462647964667e-05, + "loss": 0.9615, + "step": 14191 + }, + { + "epoch": 0.33, + "grad_norm": 1.1029611682080727, + "learning_rate": 1.5522826549578908e-05, + "loss": 0.9428, + "step": 14192 + }, + { + "epoch": 0.33, + "grad_norm": 1.1246088319785745, + "learning_rate": 1.55221904190375e-05, + "loss": 0.9467, + "step": 14193 + }, + { + "epoch": 0.33, + "grad_norm": 2.754503453293265, + "learning_rate": 1.5521554256344152e-05, + "loss": 0.9808, + "step": 14194 + }, + { + "epoch": 0.33, + "grad_norm": 2.5257159073020734, + "learning_rate": 1.552091806150257e-05, + "loss": 1.037, + "step": 14195 + }, + { + "epoch": 0.33, + "grad_norm": 2.0479395893635077, + "learning_rate": 1.5520281834516452e-05, + "loss": 1.1471, + "step": 14196 + }, + { + "epoch": 0.33, + "grad_norm": 2.2493532330720347, + "learning_rate": 1.5519645575389508e-05, + "loss": 1.1149, + "step": 14197 + }, + { + "epoch": 0.33, + "grad_norm": 1.9845160275567688, + "learning_rate": 1.5519009284125438e-05, + "loss": 1.0024, + "step": 14198 + }, + { + "epoch": 0.33, + "grad_norm": 1.0780378717654306, + "learning_rate": 1.5518372960727947e-05, + "loss": 0.9714, + "step": 14199 + }, + { + "epoch": 0.33, + "grad_norm": 2.298044195776241, + "learning_rate": 1.5517736605200745e-05, + "loss": 1.0224, + "step": 14200 + }, + { + "epoch": 0.33, + "grad_norm": 2.1850877737515706, + "learning_rate": 1.5517100217547532e-05, + "loss": 0.9493, + "step": 14201 + }, + { + "epoch": 0.33, + "grad_norm": 2.0417110473289073, + "learning_rate": 1.5516463797772017e-05, + "loss": 0.9381, + "step": 14202 + }, + { + "epoch": 0.33, + "grad_norm": 1.857146864757735, + "learning_rate": 1.55158273458779e-05, + "loss": 1.1527, + "step": 14203 + }, + { + "epoch": 0.33, + "grad_norm": 1.9674459501989212, + "learning_rate": 1.551519086186889e-05, + "loss": 1.0028, + "step": 14204 + }, + { + "epoch": 0.33, + "grad_norm": 2.0187264153930493, + "learning_rate": 1.5514554345748694e-05, + "loss": 1.048, + "step": 14205 + }, + { + "epoch": 0.33, + "grad_norm": 1.9209305445902145, + "learning_rate": 1.551391779752102e-05, + "loss": 1.0928, + "step": 14206 + }, + { + "epoch": 0.33, + "grad_norm": 2.110483734708597, + "learning_rate": 1.5513281217189565e-05, + "loss": 1.0625, + "step": 14207 + }, + { + "epoch": 0.33, + "grad_norm": 3.1313420061000694, + "learning_rate": 1.5512644604758046e-05, + "loss": 1.0719, + "step": 14208 + }, + { + "epoch": 0.33, + "grad_norm": 1.9422466605306963, + "learning_rate": 1.551200796023016e-05, + "loss": 0.972, + "step": 14209 + }, + { + "epoch": 0.33, + "grad_norm": 1.9515630121894696, + "learning_rate": 1.5511371283609622e-05, + "loss": 1.0148, + "step": 14210 + }, + { + "epoch": 0.33, + "grad_norm": 1.8986700807376224, + "learning_rate": 1.5510734574900137e-05, + "loss": 1.1055, + "step": 14211 + }, + { + "epoch": 0.33, + "grad_norm": 2.0012750081894977, + "learning_rate": 1.5510097834105406e-05, + "loss": 1.0761, + "step": 14212 + }, + { + "epoch": 0.33, + "grad_norm": 2.122915419713052, + "learning_rate": 1.5509461061229147e-05, + "loss": 1.1242, + "step": 14213 + }, + { + "epoch": 0.33, + "grad_norm": 2.4186224269553094, + "learning_rate": 1.5508824256275055e-05, + "loss": 1.1302, + "step": 14214 + }, + { + "epoch": 0.33, + "grad_norm": 2.476576925316174, + "learning_rate": 1.550818741924685e-05, + "loss": 1.0221, + "step": 14215 + }, + { + "epoch": 0.33, + "grad_norm": 2.226556572153332, + "learning_rate": 1.550755055014823e-05, + "loss": 1.0621, + "step": 14216 + }, + { + "epoch": 0.33, + "grad_norm": 2.021383960275451, + "learning_rate": 1.550691364898291e-05, + "loss": 1.1028, + "step": 14217 + }, + { + "epoch": 0.33, + "grad_norm": 2.0987290608385063, + "learning_rate": 1.5506276715754592e-05, + "loss": 1.0021, + "step": 14218 + }, + { + "epoch": 0.33, + "grad_norm": 2.0106774702884618, + "learning_rate": 1.5505639750466986e-05, + "loss": 1.1574, + "step": 14219 + }, + { + "epoch": 0.34, + "grad_norm": 2.404008462056017, + "learning_rate": 1.5505002753123807e-05, + "loss": 1.1638, + "step": 14220 + }, + { + "epoch": 0.34, + "grad_norm": 4.746574720591517, + "learning_rate": 1.550436572372876e-05, + "loss": 1.1625, + "step": 14221 + }, + { + "epoch": 0.34, + "grad_norm": 1.8333050879788195, + "learning_rate": 1.550372866228555e-05, + "loss": 0.9138, + "step": 14222 + }, + { + "epoch": 0.34, + "grad_norm": 2.1446243130879763, + "learning_rate": 1.550309156879789e-05, + "loss": 1.1083, + "step": 14223 + }, + { + "epoch": 0.34, + "grad_norm": 2.0066512642419263, + "learning_rate": 1.5502454443269485e-05, + "loss": 1.0522, + "step": 14224 + }, + { + "epoch": 0.34, + "grad_norm": 1.1447432597494405, + "learning_rate": 1.5501817285704055e-05, + "loss": 1.0371, + "step": 14225 + }, + { + "epoch": 0.34, + "grad_norm": 2.046332069705819, + "learning_rate": 1.55011800961053e-05, + "loss": 1.0887, + "step": 14226 + }, + { + "epoch": 0.34, + "grad_norm": 2.2330020671449593, + "learning_rate": 1.550054287447693e-05, + "loss": 0.9651, + "step": 14227 + }, + { + "epoch": 0.34, + "grad_norm": 2.0634087025821244, + "learning_rate": 1.5499905620822662e-05, + "loss": 0.958, + "step": 14228 + }, + { + "epoch": 0.34, + "grad_norm": 2.3241552024656826, + "learning_rate": 1.54992683351462e-05, + "loss": 1.0027, + "step": 14229 + }, + { + "epoch": 0.34, + "grad_norm": 1.9580174142644902, + "learning_rate": 1.5498631017451256e-05, + "loss": 0.9707, + "step": 14230 + }, + { + "epoch": 0.34, + "grad_norm": 2.4825398724159555, + "learning_rate": 1.5497993667741545e-05, + "loss": 0.9279, + "step": 14231 + }, + { + "epoch": 0.34, + "grad_norm": 1.9883025388149393, + "learning_rate": 1.549735628602077e-05, + "loss": 1.0988, + "step": 14232 + }, + { + "epoch": 0.34, + "grad_norm": 1.9229646536145786, + "learning_rate": 1.5496718872292648e-05, + "loss": 0.9318, + "step": 14233 + }, + { + "epoch": 0.34, + "grad_norm": 2.0094160711108926, + "learning_rate": 1.549608142656089e-05, + "loss": 1.105, + "step": 14234 + }, + { + "epoch": 0.34, + "grad_norm": 2.3092874682427897, + "learning_rate": 1.5495443948829203e-05, + "loss": 1.0133, + "step": 14235 + }, + { + "epoch": 0.34, + "grad_norm": 2.2853471863177988, + "learning_rate": 1.5494806439101306e-05, + "loss": 1.0065, + "step": 14236 + }, + { + "epoch": 0.34, + "grad_norm": 2.134416864513068, + "learning_rate": 1.5494168897380905e-05, + "loss": 0.9934, + "step": 14237 + }, + { + "epoch": 0.34, + "grad_norm": 2.099071529386258, + "learning_rate": 1.5493531323671707e-05, + "loss": 0.9803, + "step": 14238 + }, + { + "epoch": 0.34, + "grad_norm": 2.1434555570625244, + "learning_rate": 1.5492893717977438e-05, + "loss": 1.0015, + "step": 14239 + }, + { + "epoch": 0.34, + "grad_norm": 2.580851154487873, + "learning_rate": 1.54922560803018e-05, + "loss": 0.9708, + "step": 14240 + }, + { + "epoch": 0.34, + "grad_norm": 1.873864235459473, + "learning_rate": 1.5491618410648507e-05, + "loss": 0.9541, + "step": 14241 + }, + { + "epoch": 0.34, + "grad_norm": 1.983683761110486, + "learning_rate": 1.549098070902127e-05, + "loss": 1.101, + "step": 14242 + }, + { + "epoch": 0.34, + "grad_norm": 1.9280147899897568, + "learning_rate": 1.5490342975423814e-05, + "loss": 1.0703, + "step": 14243 + }, + { + "epoch": 0.34, + "grad_norm": 2.223380955864614, + "learning_rate": 1.5489705209859838e-05, + "loss": 1.1107, + "step": 14244 + }, + { + "epoch": 0.34, + "grad_norm": 1.987863693651742, + "learning_rate": 1.548906741233306e-05, + "loss": 1.09, + "step": 14245 + }, + { + "epoch": 0.34, + "grad_norm": 2.0075308348451637, + "learning_rate": 1.5488429582847194e-05, + "loss": 1.0103, + "step": 14246 + }, + { + "epoch": 0.34, + "grad_norm": 1.8130761120647194, + "learning_rate": 1.5487791721405953e-05, + "loss": 0.9258, + "step": 14247 + }, + { + "epoch": 0.34, + "grad_norm": 2.00668341493291, + "learning_rate": 1.548715382801305e-05, + "loss": 1.115, + "step": 14248 + }, + { + "epoch": 0.34, + "grad_norm": 2.2930070236811857, + "learning_rate": 1.54865159026722e-05, + "loss": 1.1358, + "step": 14249 + }, + { + "epoch": 0.34, + "grad_norm": 2.1907583117522105, + "learning_rate": 1.548587794538712e-05, + "loss": 1.0008, + "step": 14250 + }, + { + "epoch": 0.34, + "grad_norm": 2.264630266312426, + "learning_rate": 1.5485239956161523e-05, + "loss": 1.0668, + "step": 14251 + }, + { + "epoch": 0.34, + "grad_norm": 1.1347207529975807, + "learning_rate": 1.548460193499912e-05, + "loss": 1.0125, + "step": 14252 + }, + { + "epoch": 0.34, + "grad_norm": 2.289235924383911, + "learning_rate": 1.548396388190363e-05, + "loss": 1.0977, + "step": 14253 + }, + { + "epoch": 0.34, + "grad_norm": 1.9728884570914207, + "learning_rate": 1.5483325796878764e-05, + "loss": 0.9694, + "step": 14254 + }, + { + "epoch": 0.34, + "grad_norm": 2.106240069533251, + "learning_rate": 1.548268767992824e-05, + "loss": 1.0644, + "step": 14255 + }, + { + "epoch": 0.34, + "grad_norm": 2.538054697685511, + "learning_rate": 1.5482049531055772e-05, + "loss": 0.9073, + "step": 14256 + }, + { + "epoch": 0.34, + "grad_norm": 1.9655736791836325, + "learning_rate": 1.5481411350265075e-05, + "loss": 1.0897, + "step": 14257 + }, + { + "epoch": 0.34, + "grad_norm": 1.9671341545067693, + "learning_rate": 1.5480773137559867e-05, + "loss": 1.0971, + "step": 14258 + }, + { + "epoch": 0.34, + "grad_norm": 2.1451977497937, + "learning_rate": 1.5480134892943865e-05, + "loss": 1.0275, + "step": 14259 + }, + { + "epoch": 0.34, + "grad_norm": 1.15350496411201, + "learning_rate": 1.5479496616420782e-05, + "loss": 0.9779, + "step": 14260 + }, + { + "epoch": 0.34, + "grad_norm": 2.168547292734278, + "learning_rate": 1.5478858307994333e-05, + "loss": 0.9728, + "step": 14261 + }, + { + "epoch": 0.34, + "grad_norm": 1.9973810511454029, + "learning_rate": 1.5478219967668236e-05, + "loss": 0.924, + "step": 14262 + }, + { + "epoch": 0.34, + "grad_norm": 2.1750658762310056, + "learning_rate": 1.547758159544621e-05, + "loss": 1.0783, + "step": 14263 + }, + { + "epoch": 0.34, + "grad_norm": 1.9572472441235054, + "learning_rate": 1.5476943191331965e-05, + "loss": 1.0052, + "step": 14264 + }, + { + "epoch": 0.34, + "grad_norm": 1.8802745082187864, + "learning_rate": 1.5476304755329227e-05, + "loss": 1.0956, + "step": 14265 + }, + { + "epoch": 0.34, + "grad_norm": 2.2264347772815665, + "learning_rate": 1.547566628744171e-05, + "loss": 1.0095, + "step": 14266 + }, + { + "epoch": 0.34, + "grad_norm": 2.0774718815963897, + "learning_rate": 1.5475027787673126e-05, + "loss": 1.1036, + "step": 14267 + }, + { + "epoch": 0.34, + "grad_norm": 3.9781239652248384, + "learning_rate": 1.54743892560272e-05, + "loss": 1.195, + "step": 14268 + }, + { + "epoch": 0.34, + "grad_norm": 2.0050625962031914, + "learning_rate": 1.5473750692507647e-05, + "loss": 1.1052, + "step": 14269 + }, + { + "epoch": 0.34, + "grad_norm": 1.0563719363982709, + "learning_rate": 1.547311209711818e-05, + "loss": 0.9923, + "step": 14270 + }, + { + "epoch": 0.34, + "grad_norm": 2.0267917921838445, + "learning_rate": 1.547247346986253e-05, + "loss": 0.9696, + "step": 14271 + }, + { + "epoch": 0.34, + "grad_norm": 1.8915791456049293, + "learning_rate": 1.54718348107444e-05, + "loss": 1.0199, + "step": 14272 + }, + { + "epoch": 0.34, + "grad_norm": 2.2732350163601085, + "learning_rate": 1.5471196119767515e-05, + "loss": 1.1404, + "step": 14273 + }, + { + "epoch": 0.34, + "grad_norm": 1.9088846861692401, + "learning_rate": 1.5470557396935597e-05, + "loss": 0.8785, + "step": 14274 + }, + { + "epoch": 0.34, + "grad_norm": 2.025334881072442, + "learning_rate": 1.5469918642252363e-05, + "loss": 1.0335, + "step": 14275 + }, + { + "epoch": 0.34, + "grad_norm": 1.859506550811478, + "learning_rate": 1.5469279855721527e-05, + "loss": 1.078, + "step": 14276 + }, + { + "epoch": 0.34, + "grad_norm": 1.924650770359275, + "learning_rate": 1.5468641037346816e-05, + "loss": 1.0455, + "step": 14277 + }, + { + "epoch": 0.34, + "grad_norm": 2.276913989098539, + "learning_rate": 1.5468002187131944e-05, + "loss": 0.9924, + "step": 14278 + }, + { + "epoch": 0.34, + "grad_norm": 2.1491834193088266, + "learning_rate": 1.546736330508063e-05, + "loss": 1.016, + "step": 14279 + }, + { + "epoch": 0.34, + "grad_norm": 1.0604484975264172, + "learning_rate": 1.5466724391196598e-05, + "loss": 1.0336, + "step": 14280 + }, + { + "epoch": 0.34, + "grad_norm": 2.1021859648895664, + "learning_rate": 1.5466085445483568e-05, + "loss": 1.0676, + "step": 14281 + }, + { + "epoch": 0.34, + "grad_norm": 2.0679460645909624, + "learning_rate": 1.5465446467945254e-05, + "loss": 1.1446, + "step": 14282 + }, + { + "epoch": 0.34, + "grad_norm": 1.1197586398235184, + "learning_rate": 1.5464807458585383e-05, + "loss": 0.9529, + "step": 14283 + }, + { + "epoch": 0.34, + "grad_norm": 2.2485508539229677, + "learning_rate": 1.546416841740767e-05, + "loss": 1.0516, + "step": 14284 + }, + { + "epoch": 0.34, + "grad_norm": 1.806968944010283, + "learning_rate": 1.5463529344415842e-05, + "loss": 0.9854, + "step": 14285 + }, + { + "epoch": 0.34, + "grad_norm": 2.059570789337103, + "learning_rate": 1.5462890239613613e-05, + "loss": 0.797, + "step": 14286 + }, + { + "epoch": 0.34, + "grad_norm": 1.0923693447128566, + "learning_rate": 1.546225110300471e-05, + "loss": 0.9602, + "step": 14287 + }, + { + "epoch": 0.34, + "grad_norm": 2.055317950276473, + "learning_rate": 1.5461611934592852e-05, + "loss": 1.1879, + "step": 14288 + }, + { + "epoch": 0.34, + "grad_norm": 1.9423928307293608, + "learning_rate": 1.546097273438176e-05, + "loss": 1.0302, + "step": 14289 + }, + { + "epoch": 0.34, + "grad_norm": 2.2043840623839452, + "learning_rate": 1.5460333502375155e-05, + "loss": 1.1644, + "step": 14290 + }, + { + "epoch": 0.34, + "grad_norm": 2.0924805949392633, + "learning_rate": 1.5459694238576757e-05, + "loss": 1.1706, + "step": 14291 + }, + { + "epoch": 0.34, + "grad_norm": 1.0745055086853834, + "learning_rate": 1.5459054942990298e-05, + "loss": 0.9726, + "step": 14292 + }, + { + "epoch": 0.34, + "grad_norm": 2.089473091437662, + "learning_rate": 1.5458415615619486e-05, + "loss": 1.0477, + "step": 14293 + }, + { + "epoch": 0.34, + "grad_norm": 2.1671825037842267, + "learning_rate": 1.545777625646805e-05, + "loss": 1.0444, + "step": 14294 + }, + { + "epoch": 0.34, + "grad_norm": 2.0687976112642152, + "learning_rate": 1.5457136865539716e-05, + "loss": 1.0693, + "step": 14295 + }, + { + "epoch": 0.34, + "grad_norm": 2.01007900016863, + "learning_rate": 1.5456497442838202e-05, + "loss": 1.0048, + "step": 14296 + }, + { + "epoch": 0.34, + "grad_norm": 1.7712525732739788, + "learning_rate": 1.5455857988367236e-05, + "loss": 1.0626, + "step": 14297 + }, + { + "epoch": 0.34, + "grad_norm": 2.878938621286593, + "learning_rate": 1.5455218502130533e-05, + "loss": 1.0327, + "step": 14298 + }, + { + "epoch": 0.34, + "grad_norm": 1.1745703289835414, + "learning_rate": 1.5454578984131822e-05, + "loss": 0.9823, + "step": 14299 + }, + { + "epoch": 0.34, + "grad_norm": 1.8674320181548127, + "learning_rate": 1.5453939434374824e-05, + "loss": 1.0101, + "step": 14300 + }, + { + "epoch": 0.34, + "grad_norm": 2.114018138356969, + "learning_rate": 1.5453299852863265e-05, + "loss": 1.0151, + "step": 14301 + }, + { + "epoch": 0.34, + "grad_norm": 1.910624801264319, + "learning_rate": 1.5452660239600873e-05, + "loss": 0.9681, + "step": 14302 + }, + { + "epoch": 0.34, + "grad_norm": 2.005679595214096, + "learning_rate": 1.545202059459136e-05, + "loss": 1.1131, + "step": 14303 + }, + { + "epoch": 0.34, + "grad_norm": 1.9501093777814287, + "learning_rate": 1.5451380917838462e-05, + "loss": 0.9718, + "step": 14304 + }, + { + "epoch": 0.34, + "grad_norm": 2.114194883484855, + "learning_rate": 1.5450741209345894e-05, + "loss": 1.0608, + "step": 14305 + }, + { + "epoch": 0.34, + "grad_norm": 1.8083236296425165, + "learning_rate": 1.5450101469117386e-05, + "loss": 0.9853, + "step": 14306 + }, + { + "epoch": 0.34, + "grad_norm": 2.2800879096660545, + "learning_rate": 1.544946169715666e-05, + "loss": 0.9996, + "step": 14307 + }, + { + "epoch": 0.34, + "grad_norm": 1.9607183998455457, + "learning_rate": 1.5448821893467444e-05, + "loss": 0.985, + "step": 14308 + }, + { + "epoch": 0.34, + "grad_norm": 2.558135762520407, + "learning_rate": 1.5448182058053466e-05, + "loss": 1.0276, + "step": 14309 + }, + { + "epoch": 0.34, + "grad_norm": 1.9341448976743456, + "learning_rate": 1.5447542190918444e-05, + "loss": 0.99, + "step": 14310 + }, + { + "epoch": 0.34, + "grad_norm": 1.8521881919974272, + "learning_rate": 1.5446902292066106e-05, + "loss": 0.9531, + "step": 14311 + }, + { + "epoch": 0.34, + "grad_norm": 2.216090493155858, + "learning_rate": 1.5446262361500178e-05, + "loss": 1.0905, + "step": 14312 + }, + { + "epoch": 0.34, + "grad_norm": 2.5791832725609267, + "learning_rate": 1.544562239922439e-05, + "loss": 1.0182, + "step": 14313 + }, + { + "epoch": 0.34, + "grad_norm": 3.7271526052388406, + "learning_rate": 1.544498240524246e-05, + "loss": 1.0577, + "step": 14314 + }, + { + "epoch": 0.34, + "grad_norm": 2.0819922860903004, + "learning_rate": 1.544434237955812e-05, + "loss": 0.989, + "step": 14315 + }, + { + "epoch": 0.34, + "grad_norm": 2.0627303209359966, + "learning_rate": 1.5443702322175095e-05, + "loss": 1.0426, + "step": 14316 + }, + { + "epoch": 0.34, + "grad_norm": 1.8649576243957677, + "learning_rate": 1.544306223309711e-05, + "loss": 1.1242, + "step": 14317 + }, + { + "epoch": 0.34, + "grad_norm": 1.9392312528966211, + "learning_rate": 1.54424221123279e-05, + "loss": 1.1211, + "step": 14318 + }, + { + "epoch": 0.34, + "grad_norm": 2.151847115914424, + "learning_rate": 1.544178195987118e-05, + "loss": 1.0894, + "step": 14319 + }, + { + "epoch": 0.34, + "grad_norm": 2.0890091231001273, + "learning_rate": 1.5441141775730685e-05, + "loss": 1.0855, + "step": 14320 + }, + { + "epoch": 0.34, + "grad_norm": 2.0349434609334995, + "learning_rate": 1.5440501559910135e-05, + "loss": 1.0318, + "step": 14321 + }, + { + "epoch": 0.34, + "grad_norm": 2.5468119284544684, + "learning_rate": 1.5439861312413267e-05, + "loss": 1.0279, + "step": 14322 + }, + { + "epoch": 0.34, + "grad_norm": 2.474474089645877, + "learning_rate": 1.5439221033243804e-05, + "loss": 1.1398, + "step": 14323 + }, + { + "epoch": 0.34, + "grad_norm": 2.1432923297605426, + "learning_rate": 1.5438580722405473e-05, + "loss": 0.9999, + "step": 14324 + }, + { + "epoch": 0.34, + "grad_norm": 3.0146258168442293, + "learning_rate": 1.5437940379902e-05, + "loss": 1.0016, + "step": 14325 + }, + { + "epoch": 0.34, + "grad_norm": 2.1351074049201113, + "learning_rate": 1.5437300005737122e-05, + "loss": 1.1305, + "step": 14326 + }, + { + "epoch": 0.34, + "grad_norm": 1.9486216534079686, + "learning_rate": 1.543665959991456e-05, + "loss": 1.0471, + "step": 14327 + }, + { + "epoch": 0.34, + "grad_norm": 1.9578432585949868, + "learning_rate": 1.5436019162438044e-05, + "loss": 1.1423, + "step": 14328 + }, + { + "epoch": 0.34, + "grad_norm": 2.0860554207731474, + "learning_rate": 1.5435378693311305e-05, + "loss": 1.0393, + "step": 14329 + }, + { + "epoch": 0.34, + "grad_norm": 1.1989937794154288, + "learning_rate": 1.5434738192538067e-05, + "loss": 1.0061, + "step": 14330 + }, + { + "epoch": 0.34, + "grad_norm": 2.1249429758875817, + "learning_rate": 1.5434097660122066e-05, + "loss": 0.8988, + "step": 14331 + }, + { + "epoch": 0.34, + "grad_norm": 2.281219057426868, + "learning_rate": 1.5433457096067027e-05, + "loss": 1.1394, + "step": 14332 + }, + { + "epoch": 0.34, + "grad_norm": 1.9892658746482952, + "learning_rate": 1.543281650037668e-05, + "loss": 1.0132, + "step": 14333 + }, + { + "epoch": 0.34, + "grad_norm": 1.95082363776574, + "learning_rate": 1.5432175873054758e-05, + "loss": 1.0148, + "step": 14334 + }, + { + "epoch": 0.34, + "grad_norm": 2.041655857568871, + "learning_rate": 1.543153521410498e-05, + "loss": 1.1443, + "step": 14335 + }, + { + "epoch": 0.34, + "grad_norm": 2.8777830056910925, + "learning_rate": 1.5430894523531094e-05, + "loss": 0.9872, + "step": 14336 + }, + { + "epoch": 0.34, + "grad_norm": 2.055534788674453, + "learning_rate": 1.5430253801336816e-05, + "loss": 1.1563, + "step": 14337 + }, + { + "epoch": 0.34, + "grad_norm": 1.9863130341836188, + "learning_rate": 1.5429613047525885e-05, + "loss": 1.0252, + "step": 14338 + }, + { + "epoch": 0.34, + "grad_norm": 1.9882239555271, + "learning_rate": 1.5428972262102027e-05, + "loss": 1.1458, + "step": 14339 + }, + { + "epoch": 0.34, + "grad_norm": 2.018283497562314, + "learning_rate": 1.542833144506897e-05, + "loss": 0.9733, + "step": 14340 + }, + { + "epoch": 0.34, + "grad_norm": 2.2125739580592474, + "learning_rate": 1.542769059643045e-05, + "loss": 1.0317, + "step": 14341 + }, + { + "epoch": 0.34, + "grad_norm": 4.375977508725296, + "learning_rate": 1.5427049716190196e-05, + "loss": 0.9941, + "step": 14342 + }, + { + "epoch": 0.34, + "grad_norm": 1.8445516045684127, + "learning_rate": 1.5426408804351943e-05, + "loss": 1.1238, + "step": 14343 + }, + { + "epoch": 0.34, + "grad_norm": 2.0042265786815228, + "learning_rate": 1.542576786091942e-05, + "loss": 1.1081, + "step": 14344 + }, + { + "epoch": 0.34, + "grad_norm": 2.0676699744359075, + "learning_rate": 1.5425126885896357e-05, + "loss": 1.093, + "step": 14345 + }, + { + "epoch": 0.34, + "grad_norm": 2.1191079519539384, + "learning_rate": 1.5424485879286488e-05, + "loss": 1.0394, + "step": 14346 + }, + { + "epoch": 0.34, + "grad_norm": 1.953903240485049, + "learning_rate": 1.5423844841093545e-05, + "loss": 1.0068, + "step": 14347 + }, + { + "epoch": 0.34, + "grad_norm": 1.9704556884000737, + "learning_rate": 1.5423203771321263e-05, + "loss": 0.971, + "step": 14348 + }, + { + "epoch": 0.34, + "grad_norm": 1.9601713406406487, + "learning_rate": 1.5422562669973367e-05, + "loss": 1.1007, + "step": 14349 + }, + { + "epoch": 0.34, + "grad_norm": 1.080998729835825, + "learning_rate": 1.5421921537053596e-05, + "loss": 0.9912, + "step": 14350 + }, + { + "epoch": 0.34, + "grad_norm": 2.3322982942695862, + "learning_rate": 1.5421280372565683e-05, + "loss": 0.9735, + "step": 14351 + }, + { + "epoch": 0.34, + "grad_norm": 1.9962572389620254, + "learning_rate": 1.5420639176513358e-05, + "loss": 0.9697, + "step": 14352 + }, + { + "epoch": 0.34, + "grad_norm": 2.2585920088344658, + "learning_rate": 1.5419997948900358e-05, + "loss": 1.2049, + "step": 14353 + }, + { + "epoch": 0.34, + "grad_norm": 1.9225694424950006, + "learning_rate": 1.5419356689730408e-05, + "loss": 1.0533, + "step": 14354 + }, + { + "epoch": 0.34, + "grad_norm": 2.2814501679062396, + "learning_rate": 1.5418715399007253e-05, + "loss": 0.9153, + "step": 14355 + }, + { + "epoch": 0.34, + "grad_norm": 1.1008625560833247, + "learning_rate": 1.541807407673462e-05, + "loss": 0.9528, + "step": 14356 + }, + { + "epoch": 0.34, + "grad_norm": 2.846734544428308, + "learning_rate": 1.5417432722916244e-05, + "loss": 1.1223, + "step": 14357 + }, + { + "epoch": 0.34, + "grad_norm": 2.132589582699011, + "learning_rate": 1.541679133755586e-05, + "loss": 1.0012, + "step": 14358 + }, + { + "epoch": 0.34, + "grad_norm": 1.859601367165867, + "learning_rate": 1.54161499206572e-05, + "loss": 0.9522, + "step": 14359 + }, + { + "epoch": 0.34, + "grad_norm": 2.5881622400436735, + "learning_rate": 1.5415508472224002e-05, + "loss": 0.9647, + "step": 14360 + }, + { + "epoch": 0.34, + "grad_norm": 1.9053435939608736, + "learning_rate": 1.541486699226e-05, + "loss": 1.212, + "step": 14361 + }, + { + "epoch": 0.34, + "grad_norm": 2.0189679390182897, + "learning_rate": 1.541422548076893e-05, + "loss": 1.088, + "step": 14362 + }, + { + "epoch": 0.34, + "grad_norm": 1.0216785104198234, + "learning_rate": 1.5413583937754517e-05, + "loss": 0.9029, + "step": 14363 + }, + { + "epoch": 0.34, + "grad_norm": 1.8983132364376798, + "learning_rate": 1.5412942363220512e-05, + "loss": 1.0827, + "step": 14364 + }, + { + "epoch": 0.34, + "grad_norm": 1.9859007573264735, + "learning_rate": 1.541230075717064e-05, + "loss": 0.9952, + "step": 14365 + }, + { + "epoch": 0.34, + "grad_norm": 2.116799356694411, + "learning_rate": 1.5411659119608643e-05, + "loss": 0.9005, + "step": 14366 + }, + { + "epoch": 0.34, + "grad_norm": 2.480242470873061, + "learning_rate": 1.541101745053825e-05, + "loss": 1.0724, + "step": 14367 + }, + { + "epoch": 0.34, + "grad_norm": 1.971281494758984, + "learning_rate": 1.5410375749963197e-05, + "loss": 1.0528, + "step": 14368 + }, + { + "epoch": 0.34, + "grad_norm": 2.042122923124427, + "learning_rate": 1.540973401788723e-05, + "loss": 1.0229, + "step": 14369 + }, + { + "epoch": 0.34, + "grad_norm": 1.9750363313281032, + "learning_rate": 1.5409092254314076e-05, + "loss": 1.0295, + "step": 14370 + }, + { + "epoch": 0.34, + "grad_norm": 1.8943708144174016, + "learning_rate": 1.5408450459247474e-05, + "loss": 1.0038, + "step": 14371 + }, + { + "epoch": 0.34, + "grad_norm": 2.1290558983872656, + "learning_rate": 1.5407808632691162e-05, + "loss": 1.1306, + "step": 14372 + }, + { + "epoch": 0.34, + "grad_norm": 2.2796307649580907, + "learning_rate": 1.5407166774648877e-05, + "loss": 0.9569, + "step": 14373 + }, + { + "epoch": 0.34, + "grad_norm": 2.194662853457606, + "learning_rate": 1.5406524885124352e-05, + "loss": 1.0964, + "step": 14374 + }, + { + "epoch": 0.34, + "grad_norm": 2.108686159387344, + "learning_rate": 1.540588296412133e-05, + "loss": 1.1269, + "step": 14375 + }, + { + "epoch": 0.34, + "grad_norm": 2.0390224399184924, + "learning_rate": 1.5405241011643544e-05, + "loss": 1.0643, + "step": 14376 + }, + { + "epoch": 0.34, + "grad_norm": 3.4909420781547973, + "learning_rate": 1.5404599027694734e-05, + "loss": 1.0232, + "step": 14377 + }, + { + "epoch": 0.34, + "grad_norm": 3.552212264761805, + "learning_rate": 1.540395701227864e-05, + "loss": 0.9312, + "step": 14378 + }, + { + "epoch": 0.34, + "grad_norm": 2.4415810488113827, + "learning_rate": 1.5403314965398995e-05, + "loss": 0.8863, + "step": 14379 + }, + { + "epoch": 0.34, + "grad_norm": 2.036872278155827, + "learning_rate": 1.540267288705954e-05, + "loss": 1.1494, + "step": 14380 + }, + { + "epoch": 0.34, + "grad_norm": 2.1867406075123887, + "learning_rate": 1.5402030777264014e-05, + "loss": 1.0524, + "step": 14381 + }, + { + "epoch": 0.34, + "grad_norm": 2.187637582990752, + "learning_rate": 1.5401388636016154e-05, + "loss": 1.0526, + "step": 14382 + }, + { + "epoch": 0.34, + "grad_norm": 2.2324374180433733, + "learning_rate": 1.54007464633197e-05, + "loss": 1.1181, + "step": 14383 + }, + { + "epoch": 0.34, + "grad_norm": 1.161028338252746, + "learning_rate": 1.5400104259178386e-05, + "loss": 1.0483, + "step": 14384 + }, + { + "epoch": 0.34, + "grad_norm": 2.1976058414707906, + "learning_rate": 1.539946202359596e-05, + "loss": 1.0664, + "step": 14385 + }, + { + "epoch": 0.34, + "grad_norm": 1.9387224454623335, + "learning_rate": 1.5398819756576155e-05, + "loss": 1.1192, + "step": 14386 + }, + { + "epoch": 0.34, + "grad_norm": 1.1228535846559953, + "learning_rate": 1.539817745812271e-05, + "loss": 0.9408, + "step": 14387 + }, + { + "epoch": 0.34, + "grad_norm": 1.192706749861581, + "learning_rate": 1.539753512823937e-05, + "loss": 0.9413, + "step": 14388 + }, + { + "epoch": 0.34, + "grad_norm": 2.0947380881448145, + "learning_rate": 1.539689276692987e-05, + "loss": 0.9924, + "step": 14389 + }, + { + "epoch": 0.34, + "grad_norm": 2.1741823128934374, + "learning_rate": 1.5396250374197953e-05, + "loss": 0.8173, + "step": 14390 + }, + { + "epoch": 0.34, + "grad_norm": 2.47314668996101, + "learning_rate": 1.5395607950047357e-05, + "loss": 1.0994, + "step": 14391 + }, + { + "epoch": 0.34, + "grad_norm": 2.164519645915322, + "learning_rate": 1.5394965494481824e-05, + "loss": 1.1155, + "step": 14392 + }, + { + "epoch": 0.34, + "grad_norm": 1.224962920012707, + "learning_rate": 1.5394323007505094e-05, + "loss": 0.9966, + "step": 14393 + }, + { + "epoch": 0.34, + "grad_norm": 2.2198392632375588, + "learning_rate": 1.5393680489120907e-05, + "loss": 0.9964, + "step": 14394 + }, + { + "epoch": 0.34, + "grad_norm": 1.1074347964126703, + "learning_rate": 1.5393037939333004e-05, + "loss": 0.9847, + "step": 14395 + }, + { + "epoch": 0.34, + "grad_norm": 2.0033004290681946, + "learning_rate": 1.5392395358145125e-05, + "loss": 1.0628, + "step": 14396 + }, + { + "epoch": 0.34, + "grad_norm": 1.8972522956610822, + "learning_rate": 1.5391752745561022e-05, + "loss": 0.9911, + "step": 14397 + }, + { + "epoch": 0.34, + "grad_norm": 1.832465736482348, + "learning_rate": 1.539111010158442e-05, + "loss": 0.9846, + "step": 14398 + }, + { + "epoch": 0.34, + "grad_norm": 2.3769194762011807, + "learning_rate": 1.539046742621907e-05, + "loss": 1.0219, + "step": 14399 + }, + { + "epoch": 0.34, + "grad_norm": 2.0136347567470034, + "learning_rate": 1.538982471946871e-05, + "loss": 1.0755, + "step": 14400 + }, + { + "epoch": 0.34, + "grad_norm": 1.140232665478698, + "learning_rate": 1.5389181981337085e-05, + "loss": 0.9188, + "step": 14401 + }, + { + "epoch": 0.34, + "grad_norm": 1.915250890859193, + "learning_rate": 1.5388539211827934e-05, + "loss": 1.0445, + "step": 14402 + }, + { + "epoch": 0.34, + "grad_norm": 1.0539069634174283, + "learning_rate": 1.538789641094501e-05, + "loss": 0.9438, + "step": 14403 + }, + { + "epoch": 0.34, + "grad_norm": 1.9569227264441782, + "learning_rate": 1.538725357869204e-05, + "loss": 1.0819, + "step": 14404 + }, + { + "epoch": 0.34, + "grad_norm": 2.1251135340638645, + "learning_rate": 1.5386610715072773e-05, + "loss": 1.1124, + "step": 14405 + }, + { + "epoch": 0.34, + "grad_norm": 2.126791488904747, + "learning_rate": 1.5385967820090958e-05, + "loss": 0.9468, + "step": 14406 + }, + { + "epoch": 0.34, + "grad_norm": 2.0047818594905498, + "learning_rate": 1.538532489375033e-05, + "loss": 0.8727, + "step": 14407 + }, + { + "epoch": 0.34, + "grad_norm": 2.2850953440415385, + "learning_rate": 1.5384681936054636e-05, + "loss": 1.0916, + "step": 14408 + }, + { + "epoch": 0.34, + "grad_norm": 1.704035135876379, + "learning_rate": 1.5384038947007617e-05, + "loss": 1.0501, + "step": 14409 + }, + { + "epoch": 0.34, + "grad_norm": 2.0605856019489295, + "learning_rate": 1.5383395926613026e-05, + "loss": 0.9776, + "step": 14410 + }, + { + "epoch": 0.34, + "grad_norm": 1.87000109309707, + "learning_rate": 1.5382752874874595e-05, + "loss": 1.0952, + "step": 14411 + }, + { + "epoch": 0.34, + "grad_norm": 2.316040161608113, + "learning_rate": 1.5382109791796073e-05, + "loss": 1.0274, + "step": 14412 + }, + { + "epoch": 0.34, + "grad_norm": 2.0238416203876475, + "learning_rate": 1.53814666773812e-05, + "loss": 1.0853, + "step": 14413 + }, + { + "epoch": 0.34, + "grad_norm": 2.0184712927948114, + "learning_rate": 1.5380823531633727e-05, + "loss": 0.9924, + "step": 14414 + }, + { + "epoch": 0.34, + "grad_norm": 2.569495482966396, + "learning_rate": 1.5380180354557398e-05, + "loss": 1.0232, + "step": 14415 + }, + { + "epoch": 0.34, + "grad_norm": 2.3081380090503774, + "learning_rate": 1.5379537146155954e-05, + "loss": 1.0699, + "step": 14416 + }, + { + "epoch": 0.34, + "grad_norm": 2.416508415435327, + "learning_rate": 1.5378893906433142e-05, + "loss": 1.0866, + "step": 14417 + }, + { + "epoch": 0.34, + "grad_norm": 1.8249299699451178, + "learning_rate": 1.5378250635392707e-05, + "loss": 1.0482, + "step": 14418 + }, + { + "epoch": 0.34, + "grad_norm": 2.1891574147356114, + "learning_rate": 1.5377607333038392e-05, + "loss": 1.076, + "step": 14419 + }, + { + "epoch": 0.34, + "grad_norm": 1.8934363578546456, + "learning_rate": 1.5376963999373945e-05, + "loss": 1.0399, + "step": 14420 + }, + { + "epoch": 0.34, + "grad_norm": 2.1360286768836496, + "learning_rate": 1.5376320634403112e-05, + "loss": 1.0785, + "step": 14421 + }, + { + "epoch": 0.34, + "grad_norm": 2.0487189450188312, + "learning_rate": 1.537567723812964e-05, + "loss": 1.0341, + "step": 14422 + }, + { + "epoch": 0.34, + "grad_norm": 2.098333322861352, + "learning_rate": 1.537503381055727e-05, + "loss": 1.0199, + "step": 14423 + }, + { + "epoch": 0.34, + "grad_norm": 2.036036077781226, + "learning_rate": 1.5374390351689752e-05, + "loss": 0.9163, + "step": 14424 + }, + { + "epoch": 0.34, + "grad_norm": 1.9111467325856608, + "learning_rate": 1.5373746861530835e-05, + "loss": 1.0619, + "step": 14425 + }, + { + "epoch": 0.34, + "grad_norm": 1.1068386943549562, + "learning_rate": 1.5373103340084262e-05, + "loss": 0.9949, + "step": 14426 + }, + { + "epoch": 0.34, + "grad_norm": 1.971292354946095, + "learning_rate": 1.537245978735378e-05, + "loss": 1.124, + "step": 14427 + }, + { + "epoch": 0.34, + "grad_norm": 2.0038129571480194, + "learning_rate": 1.5371816203343132e-05, + "loss": 1.0735, + "step": 14428 + }, + { + "epoch": 0.34, + "grad_norm": 2.132311059008383, + "learning_rate": 1.5371172588056076e-05, + "loss": 1.0749, + "step": 14429 + }, + { + "epoch": 0.34, + "grad_norm": 1.126184175958713, + "learning_rate": 1.5370528941496346e-05, + "loss": 0.9662, + "step": 14430 + }, + { + "epoch": 0.34, + "grad_norm": 2.019835875241963, + "learning_rate": 1.53698852636677e-05, + "loss": 0.8706, + "step": 14431 + }, + { + "epoch": 0.34, + "grad_norm": 1.81836804972791, + "learning_rate": 1.536924155457388e-05, + "loss": 1.0232, + "step": 14432 + }, + { + "epoch": 0.34, + "grad_norm": 1.1147050891775288, + "learning_rate": 1.536859781421864e-05, + "loss": 0.9682, + "step": 14433 + }, + { + "epoch": 0.34, + "grad_norm": 3.270940468325566, + "learning_rate": 1.536795404260572e-05, + "loss": 0.9683, + "step": 14434 + }, + { + "epoch": 0.34, + "grad_norm": 3.485448599982556, + "learning_rate": 1.536731023973887e-05, + "loss": 0.9936, + "step": 14435 + }, + { + "epoch": 0.34, + "grad_norm": 2.376938116581426, + "learning_rate": 1.5366666405621846e-05, + "loss": 1.0516, + "step": 14436 + }, + { + "epoch": 0.34, + "grad_norm": 1.1703203725454538, + "learning_rate": 1.5366022540258388e-05, + "loss": 1.0615, + "step": 14437 + }, + { + "epoch": 0.34, + "grad_norm": 2.3242771849579276, + "learning_rate": 1.536537864365225e-05, + "loss": 1.0353, + "step": 14438 + }, + { + "epoch": 0.34, + "grad_norm": 2.335861596936342, + "learning_rate": 1.5364734715807176e-05, + "loss": 1.0397, + "step": 14439 + }, + { + "epoch": 0.34, + "grad_norm": 1.1539323528910597, + "learning_rate": 1.536409075672692e-05, + "loss": 1.0243, + "step": 14440 + }, + { + "epoch": 0.34, + "grad_norm": 1.0988356512164879, + "learning_rate": 1.5363446766415226e-05, + "loss": 0.982, + "step": 14441 + }, + { + "epoch": 0.34, + "grad_norm": 2.0752618396417435, + "learning_rate": 1.536280274487585e-05, + "loss": 1.0836, + "step": 14442 + }, + { + "epoch": 0.34, + "grad_norm": 2.2170707312916442, + "learning_rate": 1.536215869211254e-05, + "loss": 1.0478, + "step": 14443 + }, + { + "epoch": 0.34, + "grad_norm": 1.9874229079659624, + "learning_rate": 1.5361514608129043e-05, + "loss": 1.0525, + "step": 14444 + }, + { + "epoch": 0.34, + "grad_norm": 2.016781604710144, + "learning_rate": 1.536087049292911e-05, + "loss": 0.9661, + "step": 14445 + }, + { + "epoch": 0.34, + "grad_norm": 2.6140916763179205, + "learning_rate": 1.5360226346516493e-05, + "loss": 1.096, + "step": 14446 + }, + { + "epoch": 0.34, + "grad_norm": 2.040860888898913, + "learning_rate": 1.535958216889494e-05, + "loss": 1.0503, + "step": 14447 + }, + { + "epoch": 0.34, + "grad_norm": 1.8751129665219568, + "learning_rate": 1.5358937960068204e-05, + "loss": 0.9893, + "step": 14448 + }, + { + "epoch": 0.34, + "grad_norm": 1.934731606716395, + "learning_rate": 1.535829372004003e-05, + "loss": 0.9888, + "step": 14449 + }, + { + "epoch": 0.34, + "grad_norm": 1.8731489204686247, + "learning_rate": 1.5357649448814177e-05, + "loss": 0.9838, + "step": 14450 + }, + { + "epoch": 0.34, + "grad_norm": 1.9491390030277644, + "learning_rate": 1.5357005146394394e-05, + "loss": 1.0379, + "step": 14451 + }, + { + "epoch": 0.34, + "grad_norm": 6.565010838196415, + "learning_rate": 1.535636081278443e-05, + "loss": 0.9494, + "step": 14452 + }, + { + "epoch": 0.34, + "grad_norm": 1.8572904214863304, + "learning_rate": 1.5355716447988038e-05, + "loss": 1.0385, + "step": 14453 + }, + { + "epoch": 0.34, + "grad_norm": 2.131795963678404, + "learning_rate": 1.535507205200897e-05, + "loss": 1.0346, + "step": 14454 + }, + { + "epoch": 0.34, + "grad_norm": 2.136173482516283, + "learning_rate": 1.5354427624850978e-05, + "loss": 1.0139, + "step": 14455 + }, + { + "epoch": 0.34, + "grad_norm": 1.8456802837911248, + "learning_rate": 1.5353783166517806e-05, + "loss": 1.0904, + "step": 14456 + }, + { + "epoch": 0.34, + "grad_norm": 2.1432214899416175, + "learning_rate": 1.535313867701322e-05, + "loss": 0.8941, + "step": 14457 + }, + { + "epoch": 0.34, + "grad_norm": 1.8931354825452038, + "learning_rate": 1.5352494156340963e-05, + "loss": 1.1816, + "step": 14458 + }, + { + "epoch": 0.34, + "grad_norm": 1.775051317036668, + "learning_rate": 1.5351849604504794e-05, + "loss": 0.997, + "step": 14459 + }, + { + "epoch": 0.34, + "grad_norm": 2.03711204457771, + "learning_rate": 1.5351205021508457e-05, + "loss": 1.1051, + "step": 14460 + }, + { + "epoch": 0.34, + "grad_norm": 2.16430255160131, + "learning_rate": 1.5350560407355714e-05, + "loss": 1.1604, + "step": 14461 + }, + { + "epoch": 0.34, + "grad_norm": 1.8773741919660363, + "learning_rate": 1.5349915762050316e-05, + "loss": 1.1188, + "step": 14462 + }, + { + "epoch": 0.34, + "grad_norm": 2.261144609042413, + "learning_rate": 1.5349271085596013e-05, + "loss": 1.0314, + "step": 14463 + }, + { + "epoch": 0.34, + "grad_norm": 2.024228957606696, + "learning_rate": 1.534862637799656e-05, + "loss": 1.0202, + "step": 14464 + }, + { + "epoch": 0.34, + "grad_norm": 1.8203806605363473, + "learning_rate": 1.534798163925571e-05, + "loss": 1.1571, + "step": 14465 + }, + { + "epoch": 0.34, + "grad_norm": 2.7810432196147006, + "learning_rate": 1.5347336869377215e-05, + "loss": 0.9603, + "step": 14466 + }, + { + "epoch": 0.34, + "grad_norm": 2.239546214487324, + "learning_rate": 1.5346692068364834e-05, + "loss": 0.8856, + "step": 14467 + }, + { + "epoch": 0.34, + "grad_norm": 2.3372327535054858, + "learning_rate": 1.534604723622232e-05, + "loss": 1.0735, + "step": 14468 + }, + { + "epoch": 0.34, + "grad_norm": 2.7314234503579304, + "learning_rate": 1.5345402372953427e-05, + "loss": 1.0742, + "step": 14469 + }, + { + "epoch": 0.34, + "grad_norm": 1.1724781137428704, + "learning_rate": 1.534475747856191e-05, + "loss": 0.9614, + "step": 14470 + }, + { + "epoch": 0.34, + "grad_norm": 2.025632128162651, + "learning_rate": 1.534411255305152e-05, + "loss": 1.0456, + "step": 14471 + }, + { + "epoch": 0.34, + "grad_norm": 2.078525844703704, + "learning_rate": 1.5343467596426015e-05, + "loss": 1.0818, + "step": 14472 + }, + { + "epoch": 0.34, + "grad_norm": 2.1135609943131453, + "learning_rate": 1.5342822608689153e-05, + "loss": 1.0959, + "step": 14473 + }, + { + "epoch": 0.34, + "grad_norm": 2.0605817383939993, + "learning_rate": 1.5342177589844682e-05, + "loss": 1.0317, + "step": 14474 + }, + { + "epoch": 0.34, + "grad_norm": 2.3142847014965846, + "learning_rate": 1.5341532539896364e-05, + "loss": 1.0429, + "step": 14475 + }, + { + "epoch": 0.34, + "grad_norm": 1.0759683025304114, + "learning_rate": 1.5340887458847954e-05, + "loss": 0.9615, + "step": 14476 + }, + { + "epoch": 0.34, + "grad_norm": 1.1209556424118352, + "learning_rate": 1.5340242346703204e-05, + "loss": 1.0234, + "step": 14477 + }, + { + "epoch": 0.34, + "grad_norm": 3.1712478457520796, + "learning_rate": 1.5339597203465876e-05, + "loss": 1.0377, + "step": 14478 + }, + { + "epoch": 0.34, + "grad_norm": 1.8604935739536177, + "learning_rate": 1.5338952029139718e-05, + "loss": 1.0097, + "step": 14479 + }, + { + "epoch": 0.34, + "grad_norm": 1.859784318645314, + "learning_rate": 1.5338306823728494e-05, + "loss": 0.9467, + "step": 14480 + }, + { + "epoch": 0.34, + "grad_norm": 1.8040776924815454, + "learning_rate": 1.5337661587235952e-05, + "loss": 0.9861, + "step": 14481 + }, + { + "epoch": 0.34, + "grad_norm": 1.875550502501258, + "learning_rate": 1.533701631966586e-05, + "loss": 1.0216, + "step": 14482 + }, + { + "epoch": 0.34, + "grad_norm": 2.558915066926896, + "learning_rate": 1.533637102102197e-05, + "loss": 1.0347, + "step": 14483 + }, + { + "epoch": 0.34, + "grad_norm": 2.066926911202041, + "learning_rate": 1.5335725691308035e-05, + "loss": 1.0704, + "step": 14484 + }, + { + "epoch": 0.34, + "grad_norm": 3.0031535969970604, + "learning_rate": 1.5335080330527818e-05, + "loss": 0.9762, + "step": 14485 + }, + { + "epoch": 0.34, + "grad_norm": 1.7944628902560962, + "learning_rate": 1.5334434938685073e-05, + "loss": 1.0424, + "step": 14486 + }, + { + "epoch": 0.34, + "grad_norm": 2.1129178890090263, + "learning_rate": 1.5333789515783558e-05, + "loss": 0.9954, + "step": 14487 + }, + { + "epoch": 0.34, + "grad_norm": 2.32030360514368, + "learning_rate": 1.5333144061827033e-05, + "loss": 1.0254, + "step": 14488 + }, + { + "epoch": 0.34, + "grad_norm": 2.1536965091298477, + "learning_rate": 1.5332498576819256e-05, + "loss": 1.0114, + "step": 14489 + }, + { + "epoch": 0.34, + "grad_norm": 1.889414044069276, + "learning_rate": 1.5331853060763985e-05, + "loss": 1.0883, + "step": 14490 + }, + { + "epoch": 0.34, + "grad_norm": 2.1030921720118556, + "learning_rate": 1.5331207513664973e-05, + "loss": 1.0494, + "step": 14491 + }, + { + "epoch": 0.34, + "grad_norm": 1.9537834859303356, + "learning_rate": 1.5330561935525988e-05, + "loss": 1.1077, + "step": 14492 + }, + { + "epoch": 0.34, + "grad_norm": 2.0194390852813973, + "learning_rate": 1.532991632635078e-05, + "loss": 1.1761, + "step": 14493 + }, + { + "epoch": 0.34, + "grad_norm": 2.0740634349102725, + "learning_rate": 1.5329270686143112e-05, + "loss": 1.0866, + "step": 14494 + }, + { + "epoch": 0.34, + "grad_norm": 1.170439189643089, + "learning_rate": 1.5328625014906746e-05, + "loss": 1.0034, + "step": 14495 + }, + { + "epoch": 0.34, + "grad_norm": 2.0085180721758076, + "learning_rate": 1.5327979312645433e-05, + "loss": 1.1391, + "step": 14496 + }, + { + "epoch": 0.34, + "grad_norm": 2.187234655320687, + "learning_rate": 1.5327333579362945e-05, + "loss": 1.0367, + "step": 14497 + }, + { + "epoch": 0.34, + "grad_norm": 1.8918373989159758, + "learning_rate": 1.5326687815063025e-05, + "loss": 0.9919, + "step": 14498 + }, + { + "epoch": 0.34, + "grad_norm": 1.8873903170727628, + "learning_rate": 1.532604201974945e-05, + "loss": 1.0112, + "step": 14499 + }, + { + "epoch": 0.34, + "grad_norm": 1.8601058116805236, + "learning_rate": 1.5325396193425967e-05, + "loss": 1.0159, + "step": 14500 + }, + { + "epoch": 0.34, + "grad_norm": 1.1846060631507664, + "learning_rate": 1.5324750336096347e-05, + "loss": 1.0001, + "step": 14501 + }, + { + "epoch": 0.34, + "grad_norm": 1.0981060363108297, + "learning_rate": 1.5324104447764342e-05, + "loss": 0.973, + "step": 14502 + }, + { + "epoch": 0.34, + "grad_norm": 2.238670799445315, + "learning_rate": 1.5323458528433713e-05, + "loss": 1.103, + "step": 14503 + }, + { + "epoch": 0.34, + "grad_norm": 1.866126113453016, + "learning_rate": 1.5322812578108224e-05, + "loss": 0.9975, + "step": 14504 + }, + { + "epoch": 0.34, + "grad_norm": 1.1652853068810891, + "learning_rate": 1.5322166596791636e-05, + "loss": 0.9745, + "step": 14505 + }, + { + "epoch": 0.34, + "grad_norm": 1.1159705032841218, + "learning_rate": 1.532152058448771e-05, + "loss": 0.8841, + "step": 14506 + }, + { + "epoch": 0.34, + "grad_norm": 1.8777467380165893, + "learning_rate": 1.5320874541200202e-05, + "loss": 1.0143, + "step": 14507 + }, + { + "epoch": 0.34, + "grad_norm": 2.2091509715524063, + "learning_rate": 1.5320228466932885e-05, + "loss": 0.9604, + "step": 14508 + }, + { + "epoch": 0.34, + "grad_norm": 2.4245998480022566, + "learning_rate": 1.5319582361689508e-05, + "loss": 1.1747, + "step": 14509 + }, + { + "epoch": 0.34, + "grad_norm": 2.2369113452093496, + "learning_rate": 1.531893622547384e-05, + "loss": 1.0639, + "step": 14510 + }, + { + "epoch": 0.34, + "grad_norm": 2.0089316472813357, + "learning_rate": 1.531829005828964e-05, + "loss": 1.1716, + "step": 14511 + }, + { + "epoch": 0.34, + "grad_norm": 1.773102216335952, + "learning_rate": 1.5317643860140675e-05, + "loss": 1.1142, + "step": 14512 + }, + { + "epoch": 0.34, + "grad_norm": 1.9758448881260977, + "learning_rate": 1.53169976310307e-05, + "loss": 1.0634, + "step": 14513 + }, + { + "epoch": 0.34, + "grad_norm": 2.0383176439970385, + "learning_rate": 1.5316351370963484e-05, + "loss": 1.0185, + "step": 14514 + }, + { + "epoch": 0.34, + "grad_norm": 1.8134662614272015, + "learning_rate": 1.5315705079942787e-05, + "loss": 1.046, + "step": 14515 + }, + { + "epoch": 0.34, + "grad_norm": 2.2619733668649316, + "learning_rate": 1.5315058757972373e-05, + "loss": 1.149, + "step": 14516 + }, + { + "epoch": 0.34, + "grad_norm": 2.1493999682977827, + "learning_rate": 1.5314412405056e-05, + "loss": 0.9989, + "step": 14517 + }, + { + "epoch": 0.34, + "grad_norm": 2.238391522689895, + "learning_rate": 1.531376602119744e-05, + "loss": 1.0772, + "step": 14518 + }, + { + "epoch": 0.34, + "grad_norm": 2.3867751458633495, + "learning_rate": 1.531311960640045e-05, + "loss": 1.1186, + "step": 14519 + }, + { + "epoch": 0.34, + "grad_norm": 1.9749661871037287, + "learning_rate": 1.5312473160668796e-05, + "loss": 1.1176, + "step": 14520 + }, + { + "epoch": 0.34, + "grad_norm": 2.1441050924233065, + "learning_rate": 1.5311826684006243e-05, + "loss": 1.0392, + "step": 14521 + }, + { + "epoch": 0.34, + "grad_norm": 1.9460059676401065, + "learning_rate": 1.5311180176416548e-05, + "loss": 1.0611, + "step": 14522 + }, + { + "epoch": 0.34, + "grad_norm": 1.7962255998619274, + "learning_rate": 1.5310533637903483e-05, + "loss": 1.1311, + "step": 14523 + }, + { + "epoch": 0.34, + "grad_norm": 2.318515330253657, + "learning_rate": 1.530988706847081e-05, + "loss": 1.0451, + "step": 14524 + }, + { + "epoch": 0.34, + "grad_norm": 1.990912288525379, + "learning_rate": 1.5309240468122294e-05, + "loss": 1.0395, + "step": 14525 + }, + { + "epoch": 0.34, + "grad_norm": 1.8244366639461385, + "learning_rate": 1.53085938368617e-05, + "loss": 1.0829, + "step": 14526 + }, + { + "epoch": 0.34, + "grad_norm": 3.2660867628942025, + "learning_rate": 1.5307947174692795e-05, + "loss": 0.9343, + "step": 14527 + }, + { + "epoch": 0.34, + "grad_norm": 2.1924515888624043, + "learning_rate": 1.5307300481619334e-05, + "loss": 0.9399, + "step": 14528 + }, + { + "epoch": 0.34, + "grad_norm": 2.008991099509901, + "learning_rate": 1.5306653757645095e-05, + "loss": 1.0368, + "step": 14529 + }, + { + "epoch": 0.34, + "grad_norm": 2.0378012695980705, + "learning_rate": 1.5306007002773834e-05, + "loss": 1.0251, + "step": 14530 + }, + { + "epoch": 0.34, + "grad_norm": 1.1740303617088235, + "learning_rate": 1.5305360217009324e-05, + "loss": 0.9175, + "step": 14531 + }, + { + "epoch": 0.34, + "grad_norm": 2.17315975971652, + "learning_rate": 1.530471340035532e-05, + "loss": 1.0208, + "step": 14532 + }, + { + "epoch": 0.34, + "grad_norm": 2.0574251640013608, + "learning_rate": 1.5304066552815604e-05, + "loss": 1.0642, + "step": 14533 + }, + { + "epoch": 0.34, + "grad_norm": 2.200402736992365, + "learning_rate": 1.530341967439393e-05, + "loss": 0.9114, + "step": 14534 + }, + { + "epoch": 0.34, + "grad_norm": 2.200233361786753, + "learning_rate": 1.5302772765094065e-05, + "loss": 1.0034, + "step": 14535 + }, + { + "epoch": 0.34, + "grad_norm": 1.9422531499397209, + "learning_rate": 1.530212582491978e-05, + "loss": 1.1055, + "step": 14536 + }, + { + "epoch": 0.34, + "grad_norm": 2.0151640666254886, + "learning_rate": 1.5301478853874838e-05, + "loss": 1.0865, + "step": 14537 + }, + { + "epoch": 0.34, + "grad_norm": 1.862158996922351, + "learning_rate": 1.5300831851963013e-05, + "loss": 0.9468, + "step": 14538 + }, + { + "epoch": 0.34, + "grad_norm": 2.024837939353733, + "learning_rate": 1.530018481918806e-05, + "loss": 1.0494, + "step": 14539 + }, + { + "epoch": 0.34, + "grad_norm": 1.1751445790298884, + "learning_rate": 1.529953775555376e-05, + "loss": 1.0343, + "step": 14540 + }, + { + "epoch": 0.34, + "grad_norm": 1.092097004125297, + "learning_rate": 1.529889066106387e-05, + "loss": 0.9424, + "step": 14541 + }, + { + "epoch": 0.34, + "grad_norm": 2.0275790079417173, + "learning_rate": 1.5298243535722162e-05, + "loss": 1.0097, + "step": 14542 + }, + { + "epoch": 0.34, + "grad_norm": 1.8354421982322682, + "learning_rate": 1.52975963795324e-05, + "loss": 0.968, + "step": 14543 + }, + { + "epoch": 0.34, + "grad_norm": 2.131031648296133, + "learning_rate": 1.5296949192498358e-05, + "loss": 0.9998, + "step": 14544 + }, + { + "epoch": 0.34, + "grad_norm": 1.9536238000038824, + "learning_rate": 1.5296301974623794e-05, + "loss": 0.9966, + "step": 14545 + }, + { + "epoch": 0.34, + "grad_norm": 1.9110701215175867, + "learning_rate": 1.529565472591249e-05, + "loss": 1.0684, + "step": 14546 + }, + { + "epoch": 0.34, + "grad_norm": 2.0419416656246296, + "learning_rate": 1.5295007446368207e-05, + "loss": 0.9967, + "step": 14547 + }, + { + "epoch": 0.34, + "grad_norm": 2.0369451204358713, + "learning_rate": 1.5294360135994712e-05, + "loss": 1.0435, + "step": 14548 + }, + { + "epoch": 0.34, + "grad_norm": 1.8911013496336722, + "learning_rate": 1.5293712794795777e-05, + "loss": 1.1508, + "step": 14549 + }, + { + "epoch": 0.34, + "grad_norm": 2.2981807159493277, + "learning_rate": 1.529306542277517e-05, + "loss": 0.9261, + "step": 14550 + }, + { + "epoch": 0.34, + "grad_norm": 2.133518458094342, + "learning_rate": 1.529241801993666e-05, + "loss": 1.0076, + "step": 14551 + }, + { + "epoch": 0.34, + "grad_norm": 2.025466028702525, + "learning_rate": 1.5291770586284018e-05, + "loss": 1.1069, + "step": 14552 + }, + { + "epoch": 0.34, + "grad_norm": 2.2337717246073763, + "learning_rate": 1.5291123121821013e-05, + "loss": 0.9098, + "step": 14553 + }, + { + "epoch": 0.34, + "grad_norm": 1.995926560834554, + "learning_rate": 1.529047562655141e-05, + "loss": 1.013, + "step": 14554 + }, + { + "epoch": 0.34, + "grad_norm": 2.049938475484983, + "learning_rate": 1.528982810047899e-05, + "loss": 1.0396, + "step": 14555 + }, + { + "epoch": 0.34, + "grad_norm": 1.1444362470350473, + "learning_rate": 1.528918054360751e-05, + "loss": 0.9618, + "step": 14556 + }, + { + "epoch": 0.34, + "grad_norm": 2.0568722574549843, + "learning_rate": 1.528853295594075e-05, + "loss": 0.987, + "step": 14557 + }, + { + "epoch": 0.34, + "grad_norm": 2.1574232881444004, + "learning_rate": 1.5287885337482472e-05, + "loss": 1.0823, + "step": 14558 + }, + { + "epoch": 0.34, + "grad_norm": 1.790385959786622, + "learning_rate": 1.5287237688236455e-05, + "loss": 1.145, + "step": 14559 + }, + { + "epoch": 0.34, + "grad_norm": 1.914205249321696, + "learning_rate": 1.5286590008206465e-05, + "loss": 1.0574, + "step": 14560 + }, + { + "epoch": 0.34, + "grad_norm": 1.0705509085594187, + "learning_rate": 1.5285942297396277e-05, + "loss": 0.9699, + "step": 14561 + }, + { + "epoch": 0.34, + "grad_norm": 2.291866529743826, + "learning_rate": 1.5285294555809658e-05, + "loss": 0.9419, + "step": 14562 + }, + { + "epoch": 0.34, + "grad_norm": 1.8932564544452053, + "learning_rate": 1.528464678345038e-05, + "loss": 0.9206, + "step": 14563 + }, + { + "epoch": 0.34, + "grad_norm": 2.1309842510481345, + "learning_rate": 1.5283998980322214e-05, + "loss": 1.0403, + "step": 14564 + }, + { + "epoch": 0.34, + "grad_norm": 1.900522172859999, + "learning_rate": 1.5283351146428933e-05, + "loss": 1.0627, + "step": 14565 + }, + { + "epoch": 0.34, + "grad_norm": 1.9890940804345563, + "learning_rate": 1.5282703281774312e-05, + "loss": 1.0193, + "step": 14566 + }, + { + "epoch": 0.34, + "grad_norm": 2.218628456294064, + "learning_rate": 1.5282055386362116e-05, + "loss": 1.1192, + "step": 14567 + }, + { + "epoch": 0.34, + "grad_norm": 2.021153191720681, + "learning_rate": 1.5281407460196126e-05, + "loss": 0.9484, + "step": 14568 + }, + { + "epoch": 0.34, + "grad_norm": 2.1377215902265236, + "learning_rate": 1.528075950328011e-05, + "loss": 0.9103, + "step": 14569 + }, + { + "epoch": 0.34, + "grad_norm": 2.046827237012384, + "learning_rate": 1.5280111515617835e-05, + "loss": 1.0543, + "step": 14570 + }, + { + "epoch": 0.34, + "grad_norm": 2.2567185344968648, + "learning_rate": 1.527946349721308e-05, + "loss": 1.0509, + "step": 14571 + }, + { + "epoch": 0.34, + "grad_norm": 2.373299726508088, + "learning_rate": 1.5278815448069622e-05, + "loss": 1.0062, + "step": 14572 + }, + { + "epoch": 0.34, + "grad_norm": 1.8229632540977019, + "learning_rate": 1.5278167368191224e-05, + "loss": 1.0794, + "step": 14573 + }, + { + "epoch": 0.34, + "grad_norm": 1.8982527779359197, + "learning_rate": 1.5277519257581668e-05, + "loss": 1.0679, + "step": 14574 + }, + { + "epoch": 0.34, + "grad_norm": 1.8662465038234075, + "learning_rate": 1.527687111624472e-05, + "loss": 1.1342, + "step": 14575 + }, + { + "epoch": 0.34, + "grad_norm": 2.0033832575398676, + "learning_rate": 1.527622294418416e-05, + "loss": 0.8857, + "step": 14576 + }, + { + "epoch": 0.34, + "grad_norm": 1.9639024765708089, + "learning_rate": 1.5275574741403763e-05, + "loss": 0.9863, + "step": 14577 + }, + { + "epoch": 0.34, + "grad_norm": 2.0338949132580137, + "learning_rate": 1.52749265079073e-05, + "loss": 1.0634, + "step": 14578 + }, + { + "epoch": 0.34, + "grad_norm": 2.3352114354299824, + "learning_rate": 1.527427824369854e-05, + "loss": 0.9872, + "step": 14579 + }, + { + "epoch": 0.34, + "grad_norm": 1.8455050363407737, + "learning_rate": 1.5273629948781264e-05, + "loss": 1.1523, + "step": 14580 + }, + { + "epoch": 0.34, + "grad_norm": 2.3146144306733643, + "learning_rate": 1.527298162315924e-05, + "loss": 1.0486, + "step": 14581 + }, + { + "epoch": 0.34, + "grad_norm": 1.8656170533610434, + "learning_rate": 1.5272333266836256e-05, + "loss": 0.9999, + "step": 14582 + }, + { + "epoch": 0.34, + "grad_norm": 2.292917613010832, + "learning_rate": 1.5271684879816074e-05, + "loss": 1.2482, + "step": 14583 + }, + { + "epoch": 0.34, + "grad_norm": 1.850971626109317, + "learning_rate": 1.5271036462102472e-05, + "loss": 1.0163, + "step": 14584 + }, + { + "epoch": 0.34, + "grad_norm": 1.9570385767771954, + "learning_rate": 1.5270388013699232e-05, + "loss": 1.0405, + "step": 14585 + }, + { + "epoch": 0.34, + "grad_norm": 2.2425361277741414, + "learning_rate": 1.526973953461012e-05, + "loss": 1.1012, + "step": 14586 + }, + { + "epoch": 0.34, + "grad_norm": 2.0412114436152895, + "learning_rate": 1.526909102483892e-05, + "loss": 1.0236, + "step": 14587 + }, + { + "epoch": 0.34, + "grad_norm": 1.8257032700952835, + "learning_rate": 1.52684424843894e-05, + "loss": 1.0994, + "step": 14588 + }, + { + "epoch": 0.34, + "grad_norm": 2.04053164030044, + "learning_rate": 1.5267793913265344e-05, + "loss": 1.0583, + "step": 14589 + }, + { + "epoch": 0.34, + "grad_norm": 2.4564784844688226, + "learning_rate": 1.526714531147052e-05, + "loss": 1.0644, + "step": 14590 + }, + { + "epoch": 0.34, + "grad_norm": 2.102509797495441, + "learning_rate": 1.526649667900871e-05, + "loss": 0.9921, + "step": 14591 + }, + { + "epoch": 0.34, + "grad_norm": 2.2691835831784224, + "learning_rate": 1.5265848015883687e-05, + "loss": 0.9911, + "step": 14592 + }, + { + "epoch": 0.34, + "grad_norm": 2.311998236253333, + "learning_rate": 1.5265199322099235e-05, + "loss": 1.0583, + "step": 14593 + }, + { + "epoch": 0.34, + "grad_norm": 1.1235131922029324, + "learning_rate": 1.5264550597659122e-05, + "loss": 0.9718, + "step": 14594 + }, + { + "epoch": 0.34, + "grad_norm": 1.9409788248130366, + "learning_rate": 1.526390184256713e-05, + "loss": 1.1249, + "step": 14595 + }, + { + "epoch": 0.34, + "grad_norm": 1.0368875483623894, + "learning_rate": 1.5263253056827032e-05, + "loss": 0.9217, + "step": 14596 + }, + { + "epoch": 0.34, + "grad_norm": 2.4067251233526554, + "learning_rate": 1.526260424044261e-05, + "loss": 1.1296, + "step": 14597 + }, + { + "epoch": 0.34, + "grad_norm": 2.1354783167265046, + "learning_rate": 1.5261955393417643e-05, + "loss": 1.0297, + "step": 14598 + }, + { + "epoch": 0.34, + "grad_norm": 2.6567597860718477, + "learning_rate": 1.5261306515755904e-05, + "loss": 1.0858, + "step": 14599 + }, + { + "epoch": 0.34, + "grad_norm": 2.4515208411962397, + "learning_rate": 1.5260657607461167e-05, + "loss": 1.1388, + "step": 14600 + }, + { + "epoch": 0.34, + "grad_norm": 1.9409284776127222, + "learning_rate": 1.5260008668537225e-05, + "loss": 1.0396, + "step": 14601 + }, + { + "epoch": 0.34, + "grad_norm": 2.113193190290319, + "learning_rate": 1.525935969898784e-05, + "loss": 0.9096, + "step": 14602 + }, + { + "epoch": 0.34, + "grad_norm": 2.389003746495805, + "learning_rate": 1.52587106988168e-05, + "loss": 1.0069, + "step": 14603 + }, + { + "epoch": 0.34, + "grad_norm": 2.350986759593175, + "learning_rate": 1.5258061668027884e-05, + "loss": 0.9876, + "step": 14604 + }, + { + "epoch": 0.34, + "grad_norm": 1.9202992907747611, + "learning_rate": 1.5257412606624868e-05, + "loss": 1.082, + "step": 14605 + }, + { + "epoch": 0.34, + "grad_norm": 2.1406725983479005, + "learning_rate": 1.525676351461153e-05, + "loss": 1.0188, + "step": 14606 + }, + { + "epoch": 0.34, + "grad_norm": 2.25164662212436, + "learning_rate": 1.525611439199165e-05, + "loss": 0.9952, + "step": 14607 + }, + { + "epoch": 0.34, + "grad_norm": 1.895545233652407, + "learning_rate": 1.5255465238769009e-05, + "loss": 0.9286, + "step": 14608 + }, + { + "epoch": 0.34, + "grad_norm": 2.0950058769747013, + "learning_rate": 1.5254816054947385e-05, + "loss": 1.0859, + "step": 14609 + }, + { + "epoch": 0.34, + "grad_norm": 2.065813229618948, + "learning_rate": 1.525416684053056e-05, + "loss": 0.8822, + "step": 14610 + }, + { + "epoch": 0.34, + "grad_norm": 2.4851849119751517, + "learning_rate": 1.5253517595522308e-05, + "loss": 1.1936, + "step": 14611 + }, + { + "epoch": 0.34, + "grad_norm": 1.8464020992479189, + "learning_rate": 1.5252868319926418e-05, + "loss": 0.9797, + "step": 14612 + }, + { + "epoch": 0.34, + "grad_norm": 2.4247171230229827, + "learning_rate": 1.5252219013746663e-05, + "loss": 1.0484, + "step": 14613 + }, + { + "epoch": 0.34, + "grad_norm": 1.8413824680129476, + "learning_rate": 1.5251569676986826e-05, + "loss": 1.0514, + "step": 14614 + }, + { + "epoch": 0.34, + "grad_norm": 1.8994910647466834, + "learning_rate": 1.5250920309650688e-05, + "loss": 1.0774, + "step": 14615 + }, + { + "epoch": 0.34, + "grad_norm": 3.232395387019513, + "learning_rate": 1.5250270911742031e-05, + "loss": 1.0862, + "step": 14616 + }, + { + "epoch": 0.34, + "grad_norm": 1.9786943894352234, + "learning_rate": 1.5249621483264632e-05, + "loss": 1.0415, + "step": 14617 + }, + { + "epoch": 0.34, + "grad_norm": 1.970307199447764, + "learning_rate": 1.5248972024222277e-05, + "loss": 1.1737, + "step": 14618 + }, + { + "epoch": 0.34, + "grad_norm": 1.8292773336572974, + "learning_rate": 1.5248322534618742e-05, + "loss": 0.9901, + "step": 14619 + }, + { + "epoch": 0.34, + "grad_norm": 2.58332079152226, + "learning_rate": 1.5247673014457815e-05, + "loss": 0.9164, + "step": 14620 + }, + { + "epoch": 0.34, + "grad_norm": 2.188441918495517, + "learning_rate": 1.5247023463743274e-05, + "loss": 1.0438, + "step": 14621 + }, + { + "epoch": 0.34, + "grad_norm": 2.015178065087258, + "learning_rate": 1.5246373882478899e-05, + "loss": 1.1353, + "step": 14622 + }, + { + "epoch": 0.34, + "grad_norm": 3.398050194284366, + "learning_rate": 1.5245724270668476e-05, + "loss": 1.1599, + "step": 14623 + }, + { + "epoch": 0.34, + "grad_norm": 1.9501165847856856, + "learning_rate": 1.5245074628315786e-05, + "loss": 0.9785, + "step": 14624 + }, + { + "epoch": 0.34, + "grad_norm": 2.106677746626958, + "learning_rate": 1.5244424955424611e-05, + "loss": 1.0665, + "step": 14625 + }, + { + "epoch": 0.34, + "grad_norm": 1.961972204441507, + "learning_rate": 1.5243775251998729e-05, + "loss": 1.061, + "step": 14626 + }, + { + "epoch": 0.34, + "grad_norm": 2.102539848406335, + "learning_rate": 1.5243125518041935e-05, + "loss": 1.054, + "step": 14627 + }, + { + "epoch": 0.34, + "grad_norm": 2.201896124910374, + "learning_rate": 1.5242475753558e-05, + "loss": 0.9983, + "step": 14628 + }, + { + "epoch": 0.34, + "grad_norm": 1.936506325390764, + "learning_rate": 1.5241825958550712e-05, + "loss": 0.9767, + "step": 14629 + }, + { + "epoch": 0.34, + "grad_norm": 2.3347328712630437, + "learning_rate": 1.524117613302385e-05, + "loss": 0.9853, + "step": 14630 + }, + { + "epoch": 0.34, + "grad_norm": 2.265338410757727, + "learning_rate": 1.5240526276981207e-05, + "loss": 1.089, + "step": 14631 + }, + { + "epoch": 0.34, + "grad_norm": 2.2494438522166824, + "learning_rate": 1.5239876390426558e-05, + "loss": 0.9578, + "step": 14632 + }, + { + "epoch": 0.34, + "grad_norm": 2.018163389863231, + "learning_rate": 1.523922647336369e-05, + "loss": 1.0397, + "step": 14633 + }, + { + "epoch": 0.34, + "grad_norm": 1.1455683730800892, + "learning_rate": 1.523857652579639e-05, + "loss": 1.0335, + "step": 14634 + }, + { + "epoch": 0.34, + "grad_norm": 1.09444992424351, + "learning_rate": 1.5237926547728432e-05, + "loss": 0.9692, + "step": 14635 + }, + { + "epoch": 0.34, + "grad_norm": 1.9796426864768486, + "learning_rate": 1.5237276539163613e-05, + "loss": 1.0282, + "step": 14636 + }, + { + "epoch": 0.34, + "grad_norm": 2.691451236856856, + "learning_rate": 1.5236626500105706e-05, + "loss": 1.0025, + "step": 14637 + }, + { + "epoch": 0.34, + "grad_norm": 1.121023016785025, + "learning_rate": 1.523597643055851e-05, + "loss": 0.9547, + "step": 14638 + }, + { + "epoch": 0.34, + "grad_norm": 2.083812108445612, + "learning_rate": 1.5235326330525794e-05, + "loss": 1.0553, + "step": 14639 + }, + { + "epoch": 0.34, + "grad_norm": 2.1245311095826396, + "learning_rate": 1.5234676200011356e-05, + "loss": 0.9889, + "step": 14640 + }, + { + "epoch": 0.34, + "grad_norm": 2.1919079616795454, + "learning_rate": 1.5234026039018974e-05, + "loss": 1.0464, + "step": 14641 + }, + { + "epoch": 0.34, + "grad_norm": 1.2244551377458495, + "learning_rate": 1.5233375847552436e-05, + "loss": 0.9603, + "step": 14642 + }, + { + "epoch": 0.34, + "grad_norm": 2.26648536133213, + "learning_rate": 1.5232725625615525e-05, + "loss": 1.0522, + "step": 14643 + }, + { + "epoch": 0.35, + "grad_norm": 2.507475533047222, + "learning_rate": 1.523207537321203e-05, + "loss": 1.0562, + "step": 14644 + }, + { + "epoch": 0.35, + "grad_norm": 2.2443171133789486, + "learning_rate": 1.5231425090345737e-05, + "loss": 1.0399, + "step": 14645 + }, + { + "epoch": 0.35, + "grad_norm": 2.8559385335117433, + "learning_rate": 1.5230774777020428e-05, + "loss": 1.0696, + "step": 14646 + }, + { + "epoch": 0.35, + "grad_norm": 2.085015376237937, + "learning_rate": 1.5230124433239894e-05, + "loss": 1.1011, + "step": 14647 + }, + { + "epoch": 0.35, + "grad_norm": 2.006303931001415, + "learning_rate": 1.5229474059007921e-05, + "loss": 1.1063, + "step": 14648 + }, + { + "epoch": 0.35, + "grad_norm": 1.113536078061581, + "learning_rate": 1.5228823654328292e-05, + "loss": 0.9969, + "step": 14649 + }, + { + "epoch": 0.35, + "grad_norm": 1.8776466326908905, + "learning_rate": 1.5228173219204795e-05, + "loss": 0.966, + "step": 14650 + }, + { + "epoch": 0.35, + "grad_norm": 1.2713331959982013, + "learning_rate": 1.5227522753641222e-05, + "loss": 1.0026, + "step": 14651 + }, + { + "epoch": 0.35, + "grad_norm": 2.4174753239884526, + "learning_rate": 1.5226872257641354e-05, + "loss": 1.0148, + "step": 14652 + }, + { + "epoch": 0.35, + "grad_norm": 1.1344986189460375, + "learning_rate": 1.5226221731208984e-05, + "loss": 0.9881, + "step": 14653 + }, + { + "epoch": 0.35, + "grad_norm": 2.309292455627321, + "learning_rate": 1.5225571174347893e-05, + "loss": 1.0203, + "step": 14654 + }, + { + "epoch": 0.35, + "grad_norm": 1.9275212934672064, + "learning_rate": 1.5224920587061873e-05, + "loss": 1.1644, + "step": 14655 + }, + { + "epoch": 0.35, + "grad_norm": 1.8714870091128937, + "learning_rate": 1.5224269969354713e-05, + "loss": 1.0144, + "step": 14656 + }, + { + "epoch": 0.35, + "grad_norm": 2.3121259281968776, + "learning_rate": 1.52236193212302e-05, + "loss": 1.1415, + "step": 14657 + }, + { + "epoch": 0.35, + "grad_norm": 1.9641084297590283, + "learning_rate": 1.5222968642692118e-05, + "loss": 1.0646, + "step": 14658 + }, + { + "epoch": 0.35, + "grad_norm": 2.088475535504973, + "learning_rate": 1.5222317933744262e-05, + "loss": 1.0549, + "step": 14659 + }, + { + "epoch": 0.35, + "grad_norm": 1.8530065911655078, + "learning_rate": 1.5221667194390412e-05, + "loss": 1.1427, + "step": 14660 + }, + { + "epoch": 0.35, + "grad_norm": 2.409825947258591, + "learning_rate": 1.522101642463437e-05, + "loss": 0.9261, + "step": 14661 + }, + { + "epoch": 0.35, + "grad_norm": 2.2875019289585747, + "learning_rate": 1.5220365624479914e-05, + "loss": 0.9853, + "step": 14662 + }, + { + "epoch": 0.35, + "grad_norm": 2.0695901339814164, + "learning_rate": 1.5219714793930838e-05, + "loss": 0.9219, + "step": 14663 + }, + { + "epoch": 0.35, + "grad_norm": 2.2295215618748268, + "learning_rate": 1.5219063932990927e-05, + "loss": 1.0021, + "step": 14664 + }, + { + "epoch": 0.35, + "grad_norm": 1.179388522221819, + "learning_rate": 1.5218413041663976e-05, + "loss": 1.0192, + "step": 14665 + }, + { + "epoch": 0.35, + "grad_norm": 2.124154090973618, + "learning_rate": 1.521776211995377e-05, + "loss": 1.1041, + "step": 14666 + }, + { + "epoch": 0.35, + "grad_norm": 2.4747277028682992, + "learning_rate": 1.5217111167864102e-05, + "loss": 0.9737, + "step": 14667 + }, + { + "epoch": 0.35, + "grad_norm": 1.835133877223415, + "learning_rate": 1.5216460185398761e-05, + "loss": 1.1088, + "step": 14668 + }, + { + "epoch": 0.35, + "grad_norm": 1.8734530274825056, + "learning_rate": 1.521580917256154e-05, + "loss": 0.9596, + "step": 14669 + }, + { + "epoch": 0.35, + "grad_norm": 3.1967951952685065, + "learning_rate": 1.5215158129356222e-05, + "loss": 0.9699, + "step": 14670 + }, + { + "epoch": 0.35, + "grad_norm": 2.247399350758602, + "learning_rate": 1.5214507055786605e-05, + "loss": 1.0274, + "step": 14671 + }, + { + "epoch": 0.35, + "grad_norm": 2.3077821495027178, + "learning_rate": 1.5213855951856475e-05, + "loss": 0.9471, + "step": 14672 + }, + { + "epoch": 0.35, + "grad_norm": 1.0658217808070387, + "learning_rate": 1.5213204817569627e-05, + "loss": 0.9371, + "step": 14673 + }, + { + "epoch": 0.35, + "grad_norm": 1.9775574581444462, + "learning_rate": 1.521255365292985e-05, + "loss": 1.0811, + "step": 14674 + }, + { + "epoch": 0.35, + "grad_norm": 1.7958068366645898, + "learning_rate": 1.5211902457940934e-05, + "loss": 1.0569, + "step": 14675 + }, + { + "epoch": 0.35, + "grad_norm": 2.019694468475916, + "learning_rate": 1.5211251232606674e-05, + "loss": 1.0537, + "step": 14676 + }, + { + "epoch": 0.35, + "grad_norm": 1.9762594738505623, + "learning_rate": 1.5210599976930857e-05, + "loss": 0.9375, + "step": 14677 + }, + { + "epoch": 0.35, + "grad_norm": 1.0687109627586526, + "learning_rate": 1.5209948690917282e-05, + "loss": 1.061, + "step": 14678 + }, + { + "epoch": 0.35, + "grad_norm": 0.9679319994500614, + "learning_rate": 1.5209297374569732e-05, + "loss": 0.9905, + "step": 14679 + }, + { + "epoch": 0.35, + "grad_norm": 2.2824794031806297, + "learning_rate": 1.5208646027892005e-05, + "loss": 1.1436, + "step": 14680 + }, + { + "epoch": 0.35, + "grad_norm": 2.3893800708715447, + "learning_rate": 1.520799465088789e-05, + "loss": 0.9534, + "step": 14681 + }, + { + "epoch": 0.35, + "grad_norm": 2.165786139475511, + "learning_rate": 1.5207343243561182e-05, + "loss": 0.8732, + "step": 14682 + }, + { + "epoch": 0.35, + "grad_norm": 2.1552377110386405, + "learning_rate": 1.5206691805915674e-05, + "loss": 1.0515, + "step": 14683 + }, + { + "epoch": 0.35, + "grad_norm": 2.208788711742792, + "learning_rate": 1.5206040337955157e-05, + "loss": 1.0729, + "step": 14684 + }, + { + "epoch": 0.35, + "grad_norm": 2.320499020823827, + "learning_rate": 1.5205388839683428e-05, + "loss": 1.0965, + "step": 14685 + }, + { + "epoch": 0.35, + "grad_norm": 2.1743334636323275, + "learning_rate": 1.5204737311104271e-05, + "loss": 1.084, + "step": 14686 + }, + { + "epoch": 0.35, + "grad_norm": 1.0506324099843038, + "learning_rate": 1.5204085752221491e-05, + "loss": 0.9937, + "step": 14687 + }, + { + "epoch": 0.35, + "grad_norm": 2.0687626470540446, + "learning_rate": 1.5203434163038872e-05, + "loss": 0.9998, + "step": 14688 + }, + { + "epoch": 0.35, + "grad_norm": 1.9552305039036246, + "learning_rate": 1.5202782543560214e-05, + "loss": 1.0748, + "step": 14689 + }, + { + "epoch": 0.35, + "grad_norm": 1.9954247046286462, + "learning_rate": 1.520213089378931e-05, + "loss": 0.9386, + "step": 14690 + }, + { + "epoch": 0.35, + "grad_norm": 2.171080143463103, + "learning_rate": 1.5201479213729954e-05, + "loss": 1.0276, + "step": 14691 + }, + { + "epoch": 0.35, + "grad_norm": 2.0139453450875564, + "learning_rate": 1.5200827503385935e-05, + "loss": 0.9376, + "step": 14692 + }, + { + "epoch": 0.35, + "grad_norm": 2.0124655370999025, + "learning_rate": 1.5200175762761055e-05, + "loss": 1.0668, + "step": 14693 + }, + { + "epoch": 0.35, + "grad_norm": 1.7881556801499805, + "learning_rate": 1.5199523991859107e-05, + "loss": 1.04, + "step": 14694 + }, + { + "epoch": 0.35, + "grad_norm": 2.1575290815672705, + "learning_rate": 1.519887219068388e-05, + "loss": 1.0425, + "step": 14695 + }, + { + "epoch": 0.35, + "grad_norm": 2.3161525537135503, + "learning_rate": 1.5198220359239173e-05, + "loss": 1.0494, + "step": 14696 + }, + { + "epoch": 0.35, + "grad_norm": 2.0990387252531484, + "learning_rate": 1.5197568497528785e-05, + "loss": 1.1352, + "step": 14697 + }, + { + "epoch": 0.35, + "grad_norm": 1.9377398037270164, + "learning_rate": 1.5196916605556504e-05, + "loss": 1.0434, + "step": 14698 + }, + { + "epoch": 0.35, + "grad_norm": 2.227237512303768, + "learning_rate": 1.5196264683326131e-05, + "loss": 1.0589, + "step": 14699 + }, + { + "epoch": 0.35, + "grad_norm": 2.522753731105138, + "learning_rate": 1.519561273084146e-05, + "loss": 1.0664, + "step": 14700 + }, + { + "epoch": 0.35, + "grad_norm": 2.43489474576613, + "learning_rate": 1.5194960748106287e-05, + "loss": 0.849, + "step": 14701 + }, + { + "epoch": 0.35, + "grad_norm": 2.59442641266974, + "learning_rate": 1.5194308735124406e-05, + "loss": 1.0313, + "step": 14702 + }, + { + "epoch": 0.35, + "grad_norm": 1.9244694976774526, + "learning_rate": 1.5193656691899618e-05, + "loss": 1.0593, + "step": 14703 + }, + { + "epoch": 0.35, + "grad_norm": 1.987803784233005, + "learning_rate": 1.5193004618435713e-05, + "loss": 1.0379, + "step": 14704 + }, + { + "epoch": 0.35, + "grad_norm": 1.793621668436738, + "learning_rate": 1.5192352514736495e-05, + "loss": 1.0452, + "step": 14705 + }, + { + "epoch": 0.35, + "grad_norm": 1.128605521114262, + "learning_rate": 1.5191700380805754e-05, + "loss": 1.0146, + "step": 14706 + }, + { + "epoch": 0.35, + "grad_norm": 2.0068884332199706, + "learning_rate": 1.5191048216647291e-05, + "loss": 1.0331, + "step": 14707 + }, + { + "epoch": 0.35, + "grad_norm": 2.132349848456735, + "learning_rate": 1.5190396022264903e-05, + "loss": 1.0552, + "step": 14708 + }, + { + "epoch": 0.35, + "grad_norm": 1.8648866546527196, + "learning_rate": 1.5189743797662382e-05, + "loss": 1.0976, + "step": 14709 + }, + { + "epoch": 0.35, + "grad_norm": 2.148984093276885, + "learning_rate": 1.5189091542843533e-05, + "loss": 1.179, + "step": 14710 + }, + { + "epoch": 0.35, + "grad_norm": 1.729275676415547, + "learning_rate": 1.5188439257812148e-05, + "loss": 1.0561, + "step": 14711 + }, + { + "epoch": 0.35, + "grad_norm": 2.0397179856735423, + "learning_rate": 1.5187786942572032e-05, + "loss": 1.0246, + "step": 14712 + }, + { + "epoch": 0.35, + "grad_norm": 2.3222275491004067, + "learning_rate": 1.518713459712697e-05, + "loss": 1.0777, + "step": 14713 + }, + { + "epoch": 0.35, + "grad_norm": 2.25188738564867, + "learning_rate": 1.5186482221480774e-05, + "loss": 0.8761, + "step": 14714 + }, + { + "epoch": 0.35, + "grad_norm": 2.0847965257174366, + "learning_rate": 1.5185829815637236e-05, + "loss": 1.0448, + "step": 14715 + }, + { + "epoch": 0.35, + "grad_norm": 1.8605036319196016, + "learning_rate": 1.5185177379600153e-05, + "loss": 1.1379, + "step": 14716 + }, + { + "epoch": 0.35, + "grad_norm": 2.64010573429462, + "learning_rate": 1.5184524913373326e-05, + "loss": 1.0972, + "step": 14717 + }, + { + "epoch": 0.35, + "grad_norm": 1.0908640322978207, + "learning_rate": 1.5183872416960554e-05, + "loss": 0.9316, + "step": 14718 + }, + { + "epoch": 0.35, + "grad_norm": 2.0720940426401544, + "learning_rate": 1.5183219890365638e-05, + "loss": 1.0465, + "step": 14719 + }, + { + "epoch": 0.35, + "grad_norm": 2.047564639122912, + "learning_rate": 1.5182567333592368e-05, + "loss": 0.9989, + "step": 14720 + }, + { + "epoch": 0.35, + "grad_norm": 2.0255936349134553, + "learning_rate": 1.5181914746644557e-05, + "loss": 0.8797, + "step": 14721 + }, + { + "epoch": 0.35, + "grad_norm": 1.7702192676579611, + "learning_rate": 1.5181262129525996e-05, + "loss": 1.1478, + "step": 14722 + }, + { + "epoch": 0.35, + "grad_norm": 1.7189908300540462, + "learning_rate": 1.5180609482240485e-05, + "loss": 1.0054, + "step": 14723 + }, + { + "epoch": 0.35, + "grad_norm": 1.8616038095271896, + "learning_rate": 1.5179956804791826e-05, + "loss": 1.0786, + "step": 14724 + }, + { + "epoch": 0.35, + "grad_norm": 2.3901149599723395, + "learning_rate": 1.517930409718382e-05, + "loss": 1.1013, + "step": 14725 + }, + { + "epoch": 0.35, + "grad_norm": 1.951045642240445, + "learning_rate": 1.5178651359420264e-05, + "loss": 1.1527, + "step": 14726 + }, + { + "epoch": 0.35, + "grad_norm": 2.4275207484523027, + "learning_rate": 1.5177998591504963e-05, + "loss": 0.9579, + "step": 14727 + }, + { + "epoch": 0.35, + "grad_norm": 2.063231904130554, + "learning_rate": 1.517734579344171e-05, + "loss": 0.9927, + "step": 14728 + }, + { + "epoch": 0.35, + "grad_norm": 2.440921725926309, + "learning_rate": 1.5176692965234316e-05, + "loss": 1.0726, + "step": 14729 + }, + { + "epoch": 0.35, + "grad_norm": 1.9193133028038283, + "learning_rate": 1.5176040106886574e-05, + "loss": 1.0079, + "step": 14730 + }, + { + "epoch": 0.35, + "grad_norm": 3.3284438991636214, + "learning_rate": 1.5175387218402287e-05, + "loss": 1.0124, + "step": 14731 + }, + { + "epoch": 0.35, + "grad_norm": 1.986136269657826, + "learning_rate": 1.5174734299785259e-05, + "loss": 1.0108, + "step": 14732 + }, + { + "epoch": 0.35, + "grad_norm": 2.3163794814295158, + "learning_rate": 1.5174081351039287e-05, + "loss": 0.9735, + "step": 14733 + }, + { + "epoch": 0.35, + "grad_norm": 1.957732570892089, + "learning_rate": 1.5173428372168177e-05, + "loss": 1.0983, + "step": 14734 + }, + { + "epoch": 0.35, + "grad_norm": 2.1244489572417016, + "learning_rate": 1.5172775363175732e-05, + "loss": 0.9953, + "step": 14735 + }, + { + "epoch": 0.35, + "grad_norm": 1.1212951327883702, + "learning_rate": 1.5172122324065748e-05, + "loss": 0.9423, + "step": 14736 + }, + { + "epoch": 0.35, + "grad_norm": 1.940470603272479, + "learning_rate": 1.5171469254842031e-05, + "loss": 1.1182, + "step": 14737 + }, + { + "epoch": 0.35, + "grad_norm": 2.044029354346551, + "learning_rate": 1.517081615550838e-05, + "loss": 1.0889, + "step": 14738 + }, + { + "epoch": 0.35, + "grad_norm": 1.8562377664926744, + "learning_rate": 1.5170163026068602e-05, + "loss": 1.0791, + "step": 14739 + }, + { + "epoch": 0.35, + "grad_norm": 1.981855347738875, + "learning_rate": 1.5169509866526501e-05, + "loss": 0.9114, + "step": 14740 + }, + { + "epoch": 0.35, + "grad_norm": 2.051944163165083, + "learning_rate": 1.5168856676885874e-05, + "loss": 1.133, + "step": 14741 + }, + { + "epoch": 0.35, + "grad_norm": 2.468251087852548, + "learning_rate": 1.5168203457150527e-05, + "loss": 1.2142, + "step": 14742 + }, + { + "epoch": 0.35, + "grad_norm": 2.1830653929439485, + "learning_rate": 1.5167550207324263e-05, + "loss": 0.9693, + "step": 14743 + }, + { + "epoch": 0.35, + "grad_norm": 2.4782273141387496, + "learning_rate": 1.516689692741089e-05, + "loss": 0.918, + "step": 14744 + }, + { + "epoch": 0.35, + "grad_norm": 1.9429758556084187, + "learning_rate": 1.5166243617414202e-05, + "loss": 0.9665, + "step": 14745 + }, + { + "epoch": 0.35, + "grad_norm": 1.8861673009023636, + "learning_rate": 1.5165590277338011e-05, + "loss": 1.1099, + "step": 14746 + }, + { + "epoch": 0.35, + "grad_norm": 1.8573040202804696, + "learning_rate": 1.5164936907186116e-05, + "loss": 0.9878, + "step": 14747 + }, + { + "epoch": 0.35, + "grad_norm": 1.927583090180483, + "learning_rate": 1.5164283506962326e-05, + "loss": 1.0251, + "step": 14748 + }, + { + "epoch": 0.35, + "grad_norm": 1.9907302585542728, + "learning_rate": 1.516363007667044e-05, + "loss": 1.0433, + "step": 14749 + }, + { + "epoch": 0.35, + "grad_norm": 2.301553558828664, + "learning_rate": 1.5162976616314264e-05, + "loss": 1.0343, + "step": 14750 + }, + { + "epoch": 0.35, + "grad_norm": 2.123330936887782, + "learning_rate": 1.5162323125897604e-05, + "loss": 1.0705, + "step": 14751 + }, + { + "epoch": 0.35, + "grad_norm": 1.915031215096268, + "learning_rate": 1.5161669605424268e-05, + "loss": 1.0152, + "step": 14752 + }, + { + "epoch": 0.35, + "grad_norm": 2.0839504517486307, + "learning_rate": 1.5161016054898055e-05, + "loss": 0.8876, + "step": 14753 + }, + { + "epoch": 0.35, + "grad_norm": 2.247681570811693, + "learning_rate": 1.5160362474322771e-05, + "loss": 1.1386, + "step": 14754 + }, + { + "epoch": 0.35, + "grad_norm": 2.1413434126785704, + "learning_rate": 1.5159708863702225e-05, + "loss": 1.139, + "step": 14755 + }, + { + "epoch": 0.35, + "grad_norm": 2.04495289639204, + "learning_rate": 1.515905522304022e-05, + "loss": 1.1135, + "step": 14756 + }, + { + "epoch": 0.35, + "grad_norm": 2.076045662281468, + "learning_rate": 1.5158401552340562e-05, + "loss": 1.0219, + "step": 14757 + }, + { + "epoch": 0.35, + "grad_norm": 1.9215575571222656, + "learning_rate": 1.5157747851607055e-05, + "loss": 0.9735, + "step": 14758 + }, + { + "epoch": 0.35, + "grad_norm": 2.0440515501058996, + "learning_rate": 1.5157094120843512e-05, + "loss": 1.0033, + "step": 14759 + }, + { + "epoch": 0.35, + "grad_norm": 2.4224527205434203, + "learning_rate": 1.515644036005373e-05, + "loss": 1.0995, + "step": 14760 + }, + { + "epoch": 0.35, + "grad_norm": 1.9932622267423545, + "learning_rate": 1.5155786569241521e-05, + "loss": 0.9532, + "step": 14761 + }, + { + "epoch": 0.35, + "grad_norm": 1.9122215079576868, + "learning_rate": 1.515513274841069e-05, + "loss": 1.0151, + "step": 14762 + }, + { + "epoch": 0.35, + "grad_norm": 2.194871545627199, + "learning_rate": 1.5154478897565046e-05, + "loss": 0.9832, + "step": 14763 + }, + { + "epoch": 0.35, + "grad_norm": 1.8300895974268852, + "learning_rate": 1.515382501670839e-05, + "loss": 1.0141, + "step": 14764 + }, + { + "epoch": 0.35, + "grad_norm": 2.077841284763514, + "learning_rate": 1.5153171105844537e-05, + "loss": 1.0898, + "step": 14765 + }, + { + "epoch": 0.35, + "grad_norm": 2.2767581714388663, + "learning_rate": 1.5152517164977289e-05, + "loss": 1.0493, + "step": 14766 + }, + { + "epoch": 0.35, + "grad_norm": 2.5155934363659624, + "learning_rate": 1.5151863194110455e-05, + "loss": 1.08, + "step": 14767 + }, + { + "epoch": 0.35, + "grad_norm": 1.9213600108363786, + "learning_rate": 1.515120919324784e-05, + "loss": 1.0627, + "step": 14768 + }, + { + "epoch": 0.35, + "grad_norm": 1.9882651707330696, + "learning_rate": 1.5150555162393254e-05, + "loss": 1.1443, + "step": 14769 + }, + { + "epoch": 0.35, + "grad_norm": 1.9883151682666136, + "learning_rate": 1.514990110155051e-05, + "loss": 1.0942, + "step": 14770 + }, + { + "epoch": 0.35, + "grad_norm": 2.030357711443408, + "learning_rate": 1.5149247010723409e-05, + "loss": 0.8945, + "step": 14771 + }, + { + "epoch": 0.35, + "grad_norm": 1.8389542045506309, + "learning_rate": 1.5148592889915763e-05, + "loss": 1.0038, + "step": 14772 + }, + { + "epoch": 0.35, + "grad_norm": 1.8588430464288606, + "learning_rate": 1.5147938739131375e-05, + "loss": 1.0848, + "step": 14773 + }, + { + "epoch": 0.35, + "grad_norm": 1.9275945724600316, + "learning_rate": 1.5147284558374062e-05, + "loss": 1.0453, + "step": 14774 + }, + { + "epoch": 0.35, + "grad_norm": 2.151220995742834, + "learning_rate": 1.5146630347647628e-05, + "loss": 1.1055, + "step": 14775 + }, + { + "epoch": 0.35, + "grad_norm": 1.8990407027617409, + "learning_rate": 1.5145976106955882e-05, + "loss": 1.0308, + "step": 14776 + }, + { + "epoch": 0.35, + "grad_norm": 1.9833321078374055, + "learning_rate": 1.5145321836302635e-05, + "loss": 1.1677, + "step": 14777 + }, + { + "epoch": 0.35, + "grad_norm": 3.145744007901397, + "learning_rate": 1.5144667535691694e-05, + "loss": 0.9432, + "step": 14778 + }, + { + "epoch": 0.35, + "grad_norm": 2.0323056594463496, + "learning_rate": 1.5144013205126868e-05, + "loss": 1.0614, + "step": 14779 + }, + { + "epoch": 0.35, + "grad_norm": 2.0218610746579952, + "learning_rate": 1.5143358844611973e-05, + "loss": 1.0829, + "step": 14780 + }, + { + "epoch": 0.35, + "grad_norm": 1.120986808766982, + "learning_rate": 1.5142704454150812e-05, + "loss": 0.9783, + "step": 14781 + }, + { + "epoch": 0.35, + "grad_norm": 2.210915634359706, + "learning_rate": 1.5142050033747197e-05, + "loss": 1.1024, + "step": 14782 + }, + { + "epoch": 0.35, + "grad_norm": 2.0470365027647524, + "learning_rate": 1.5141395583404939e-05, + "loss": 0.9783, + "step": 14783 + }, + { + "epoch": 0.35, + "grad_norm": 1.175342535591028, + "learning_rate": 1.5140741103127848e-05, + "loss": 0.9536, + "step": 14784 + }, + { + "epoch": 0.35, + "grad_norm": 2.8906038871522495, + "learning_rate": 1.5140086592919735e-05, + "loss": 1.0626, + "step": 14785 + }, + { + "epoch": 0.35, + "grad_norm": 2.0372167793956106, + "learning_rate": 1.513943205278441e-05, + "loss": 1.0563, + "step": 14786 + }, + { + "epoch": 0.35, + "grad_norm": 1.0793886952702756, + "learning_rate": 1.5138777482725688e-05, + "loss": 0.9678, + "step": 14787 + }, + { + "epoch": 0.35, + "grad_norm": 2.4960930062010824, + "learning_rate": 1.5138122882747373e-05, + "loss": 1.1388, + "step": 14788 + }, + { + "epoch": 0.35, + "grad_norm": 1.156378374029193, + "learning_rate": 1.513746825285328e-05, + "loss": 0.9772, + "step": 14789 + }, + { + "epoch": 0.35, + "grad_norm": 1.888264541503719, + "learning_rate": 1.5136813593047222e-05, + "loss": 1.029, + "step": 14790 + }, + { + "epoch": 0.35, + "grad_norm": 1.8742798695143226, + "learning_rate": 1.5136158903333007e-05, + "loss": 1.0556, + "step": 14791 + }, + { + "epoch": 0.35, + "grad_norm": 2.146732359765267, + "learning_rate": 1.5135504183714448e-05, + "loss": 1.2364, + "step": 14792 + }, + { + "epoch": 0.35, + "grad_norm": 2.0363588965749284, + "learning_rate": 1.5134849434195362e-05, + "loss": 0.9963, + "step": 14793 + }, + { + "epoch": 0.35, + "grad_norm": 2.2024022768054667, + "learning_rate": 1.5134194654779555e-05, + "loss": 1.0045, + "step": 14794 + }, + { + "epoch": 0.35, + "grad_norm": 1.0848377314055433, + "learning_rate": 1.5133539845470839e-05, + "loss": 0.9268, + "step": 14795 + }, + { + "epoch": 0.35, + "grad_norm": 1.8542950403236769, + "learning_rate": 1.5132885006273032e-05, + "loss": 1.1386, + "step": 14796 + }, + { + "epoch": 0.35, + "grad_norm": 2.177811041314258, + "learning_rate": 1.5132230137189941e-05, + "loss": 1.0331, + "step": 14797 + }, + { + "epoch": 0.35, + "grad_norm": 1.980164175295391, + "learning_rate": 1.5131575238225385e-05, + "loss": 1.0262, + "step": 14798 + }, + { + "epoch": 0.35, + "grad_norm": 1.9297036553249087, + "learning_rate": 1.5130920309383167e-05, + "loss": 1.0282, + "step": 14799 + }, + { + "epoch": 0.35, + "grad_norm": 1.9894106016908126, + "learning_rate": 1.5130265350667111e-05, + "loss": 1.1158, + "step": 14800 + }, + { + "epoch": 0.35, + "grad_norm": 2.031990554730962, + "learning_rate": 1.5129610362081027e-05, + "loss": 0.9016, + "step": 14801 + }, + { + "epoch": 0.35, + "grad_norm": 2.1642387061045554, + "learning_rate": 1.5128955343628724e-05, + "loss": 1.0457, + "step": 14802 + }, + { + "epoch": 0.35, + "grad_norm": 3.829083623360074, + "learning_rate": 1.5128300295314021e-05, + "loss": 0.9084, + "step": 14803 + }, + { + "epoch": 0.35, + "grad_norm": 2.7168887419015237, + "learning_rate": 1.512764521714073e-05, + "loss": 0.9602, + "step": 14804 + }, + { + "epoch": 0.35, + "grad_norm": 2.1080187866093523, + "learning_rate": 1.5126990109112665e-05, + "loss": 0.8739, + "step": 14805 + }, + { + "epoch": 0.35, + "grad_norm": 1.9232062573997624, + "learning_rate": 1.512633497123364e-05, + "loss": 1.1009, + "step": 14806 + }, + { + "epoch": 0.35, + "grad_norm": 1.9618728493567128, + "learning_rate": 1.5125679803507472e-05, + "loss": 1.154, + "step": 14807 + }, + { + "epoch": 0.35, + "grad_norm": 1.9556196648421165, + "learning_rate": 1.512502460593797e-05, + "loss": 1.001, + "step": 14808 + }, + { + "epoch": 0.35, + "grad_norm": 2.5863859242160157, + "learning_rate": 1.5124369378528954e-05, + "loss": 0.9952, + "step": 14809 + }, + { + "epoch": 0.35, + "grad_norm": 1.9064646480583378, + "learning_rate": 1.512371412128424e-05, + "loss": 1.1459, + "step": 14810 + }, + { + "epoch": 0.35, + "grad_norm": 2.84178703952018, + "learning_rate": 1.5123058834207633e-05, + "loss": 1.0445, + "step": 14811 + }, + { + "epoch": 0.35, + "grad_norm": 2.015575143447917, + "learning_rate": 1.5122403517302964e-05, + "loss": 1.0085, + "step": 14812 + }, + { + "epoch": 0.35, + "grad_norm": 1.959301353360337, + "learning_rate": 1.512174817057403e-05, + "loss": 0.9664, + "step": 14813 + }, + { + "epoch": 0.35, + "grad_norm": 2.172428330339046, + "learning_rate": 1.5121092794024668e-05, + "loss": 1.0229, + "step": 14814 + }, + { + "epoch": 0.35, + "grad_norm": 2.14391697785127, + "learning_rate": 1.5120437387658675e-05, + "loss": 1.1525, + "step": 14815 + }, + { + "epoch": 0.35, + "grad_norm": 1.8801339428846913, + "learning_rate": 1.5119781951479877e-05, + "loss": 1.135, + "step": 14816 + }, + { + "epoch": 0.35, + "grad_norm": 1.8963987602336996, + "learning_rate": 1.5119126485492086e-05, + "loss": 0.9857, + "step": 14817 + }, + { + "epoch": 0.35, + "grad_norm": 1.1520425140111388, + "learning_rate": 1.5118470989699119e-05, + "loss": 0.9851, + "step": 14818 + }, + { + "epoch": 0.35, + "grad_norm": 2.2128601518234756, + "learning_rate": 1.5117815464104797e-05, + "loss": 0.9884, + "step": 14819 + }, + { + "epoch": 0.35, + "grad_norm": 2.00256031638387, + "learning_rate": 1.5117159908712929e-05, + "loss": 1.0166, + "step": 14820 + }, + { + "epoch": 0.35, + "grad_norm": 3.5901973467720647, + "learning_rate": 1.511650432352734e-05, + "loss": 0.8784, + "step": 14821 + }, + { + "epoch": 0.35, + "grad_norm": 2.1886874323505943, + "learning_rate": 1.511584870855184e-05, + "loss": 1.0541, + "step": 14822 + }, + { + "epoch": 0.35, + "grad_norm": 2.246746600266492, + "learning_rate": 1.5115193063790253e-05, + "loss": 1.11, + "step": 14823 + }, + { + "epoch": 0.35, + "grad_norm": 2.00464979339867, + "learning_rate": 1.5114537389246388e-05, + "loss": 1.0775, + "step": 14824 + }, + { + "epoch": 0.35, + "grad_norm": 1.9196033854533947, + "learning_rate": 1.511388168492407e-05, + "loss": 1.1526, + "step": 14825 + }, + { + "epoch": 0.35, + "grad_norm": 1.9332258189914453, + "learning_rate": 1.511322595082711e-05, + "loss": 0.9602, + "step": 14826 + }, + { + "epoch": 0.35, + "grad_norm": 1.155813154777088, + "learning_rate": 1.5112570186959334e-05, + "loss": 1.0224, + "step": 14827 + }, + { + "epoch": 0.35, + "grad_norm": 1.9662860884940359, + "learning_rate": 1.5111914393324553e-05, + "loss": 1.0492, + "step": 14828 + }, + { + "epoch": 0.35, + "grad_norm": 1.9740291064992372, + "learning_rate": 1.511125856992659e-05, + "loss": 0.9698, + "step": 14829 + }, + { + "epoch": 0.35, + "grad_norm": 1.9911935361417064, + "learning_rate": 1.511060271676926e-05, + "loss": 1.0994, + "step": 14830 + }, + { + "epoch": 0.35, + "grad_norm": 2.3878104170453933, + "learning_rate": 1.5109946833856386e-05, + "loss": 0.9744, + "step": 14831 + }, + { + "epoch": 0.35, + "grad_norm": 2.4927299318960854, + "learning_rate": 1.510929092119178e-05, + "loss": 0.9694, + "step": 14832 + }, + { + "epoch": 0.35, + "grad_norm": 1.1039440456410918, + "learning_rate": 1.5108634978779267e-05, + "loss": 0.9753, + "step": 14833 + }, + { + "epoch": 0.35, + "grad_norm": 2.898984332599311, + "learning_rate": 1.5107979006622662e-05, + "loss": 1.0212, + "step": 14834 + }, + { + "epoch": 0.35, + "grad_norm": 2.013244908373138, + "learning_rate": 1.5107323004725786e-05, + "loss": 1.0677, + "step": 14835 + }, + { + "epoch": 0.35, + "grad_norm": 1.1866853874262024, + "learning_rate": 1.510666697309246e-05, + "loss": 1.0375, + "step": 14836 + }, + { + "epoch": 0.35, + "grad_norm": 3.183052377694001, + "learning_rate": 1.5106010911726501e-05, + "loss": 1.0342, + "step": 14837 + }, + { + "epoch": 0.35, + "grad_norm": 1.8568856042909547, + "learning_rate": 1.5105354820631732e-05, + "loss": 1.0667, + "step": 14838 + }, + { + "epoch": 0.35, + "grad_norm": 1.911788352260281, + "learning_rate": 1.5104698699811968e-05, + "loss": 1.062, + "step": 14839 + }, + { + "epoch": 0.35, + "grad_norm": 2.800287115993885, + "learning_rate": 1.5104042549271033e-05, + "loss": 0.898, + "step": 14840 + }, + { + "epoch": 0.35, + "grad_norm": 2.131070813817326, + "learning_rate": 1.5103386369012746e-05, + "loss": 1.0627, + "step": 14841 + }, + { + "epoch": 0.35, + "grad_norm": 2.113183934197468, + "learning_rate": 1.510273015904093e-05, + "loss": 1.0721, + "step": 14842 + }, + { + "epoch": 0.35, + "grad_norm": 1.8701667759149958, + "learning_rate": 1.5102073919359402e-05, + "loss": 1.0438, + "step": 14843 + }, + { + "epoch": 0.35, + "grad_norm": 1.86347867599042, + "learning_rate": 1.5101417649971986e-05, + "loss": 0.9971, + "step": 14844 + }, + { + "epoch": 0.35, + "grad_norm": 2.186510904675745, + "learning_rate": 1.51007613508825e-05, + "loss": 1.1744, + "step": 14845 + }, + { + "epoch": 0.35, + "grad_norm": 2.0881630829704383, + "learning_rate": 1.5100105022094766e-05, + "loss": 0.9491, + "step": 14846 + }, + { + "epoch": 0.35, + "grad_norm": 2.280843112614497, + "learning_rate": 1.5099448663612605e-05, + "loss": 1.0056, + "step": 14847 + }, + { + "epoch": 0.35, + "grad_norm": 1.8149774873393705, + "learning_rate": 1.5098792275439842e-05, + "loss": 0.9929, + "step": 14848 + }, + { + "epoch": 0.35, + "grad_norm": 2.1027486050464606, + "learning_rate": 1.5098135857580296e-05, + "loss": 1.1641, + "step": 14849 + }, + { + "epoch": 0.35, + "grad_norm": 1.9435931617442823, + "learning_rate": 1.5097479410037787e-05, + "loss": 1.0025, + "step": 14850 + }, + { + "epoch": 0.35, + "grad_norm": 2.027518141842718, + "learning_rate": 1.509682293281614e-05, + "loss": 1.0565, + "step": 14851 + }, + { + "epoch": 0.35, + "grad_norm": 2.0458968624573246, + "learning_rate": 1.5096166425919176e-05, + "loss": 1.0365, + "step": 14852 + }, + { + "epoch": 0.35, + "grad_norm": 2.3572011655080503, + "learning_rate": 1.5095509889350716e-05, + "loss": 0.9276, + "step": 14853 + }, + { + "epoch": 0.35, + "grad_norm": 2.144261911073247, + "learning_rate": 1.5094853323114587e-05, + "loss": 0.9687, + "step": 14854 + }, + { + "epoch": 0.35, + "grad_norm": 2.046580468733422, + "learning_rate": 1.5094196727214607e-05, + "loss": 1.0045, + "step": 14855 + }, + { + "epoch": 0.35, + "grad_norm": 2.215165165066564, + "learning_rate": 1.5093540101654602e-05, + "loss": 1.1636, + "step": 14856 + }, + { + "epoch": 0.35, + "grad_norm": 1.848386946795637, + "learning_rate": 1.5092883446438396e-05, + "loss": 1.0319, + "step": 14857 + }, + { + "epoch": 0.35, + "grad_norm": 2.002519628921971, + "learning_rate": 1.5092226761569805e-05, + "loss": 0.9525, + "step": 14858 + }, + { + "epoch": 0.35, + "grad_norm": 2.246141536453094, + "learning_rate": 1.5091570047052666e-05, + "loss": 1.0925, + "step": 14859 + }, + { + "epoch": 0.35, + "grad_norm": 2.1730734250566317, + "learning_rate": 1.5090913302890786e-05, + "loss": 1.1041, + "step": 14860 + }, + { + "epoch": 0.35, + "grad_norm": 1.9438141739919466, + "learning_rate": 1.5090256529088e-05, + "loss": 1.0119, + "step": 14861 + }, + { + "epoch": 0.35, + "grad_norm": 2.0681644734877476, + "learning_rate": 1.5089599725648128e-05, + "loss": 1.1008, + "step": 14862 + }, + { + "epoch": 0.35, + "grad_norm": 2.1363517453237035, + "learning_rate": 1.5088942892574995e-05, + "loss": 0.9894, + "step": 14863 + }, + { + "epoch": 0.35, + "grad_norm": 1.9232773616933385, + "learning_rate": 1.5088286029872427e-05, + "loss": 0.9746, + "step": 14864 + }, + { + "epoch": 0.35, + "grad_norm": 2.180237918066535, + "learning_rate": 1.5087629137544246e-05, + "loss": 0.9483, + "step": 14865 + }, + { + "epoch": 0.35, + "grad_norm": 1.9777493406278057, + "learning_rate": 1.508697221559428e-05, + "loss": 1.131, + "step": 14866 + }, + { + "epoch": 0.35, + "grad_norm": 2.6229925639103513, + "learning_rate": 1.5086315264026345e-05, + "loss": 1.077, + "step": 14867 + }, + { + "epoch": 0.35, + "grad_norm": 2.529235520410358, + "learning_rate": 1.5085658282844278e-05, + "loss": 1.1333, + "step": 14868 + }, + { + "epoch": 0.35, + "grad_norm": 2.0768810316652426, + "learning_rate": 1.5085001272051893e-05, + "loss": 1.0231, + "step": 14869 + }, + { + "epoch": 0.35, + "grad_norm": 2.0037845187458516, + "learning_rate": 1.5084344231653027e-05, + "loss": 1.1276, + "step": 14870 + }, + { + "epoch": 0.35, + "grad_norm": 1.9937454953271698, + "learning_rate": 1.5083687161651495e-05, + "loss": 1.0062, + "step": 14871 + }, + { + "epoch": 0.35, + "grad_norm": 1.8867777312677292, + "learning_rate": 1.508303006205113e-05, + "loss": 0.8921, + "step": 14872 + }, + { + "epoch": 0.35, + "grad_norm": 2.0917163201367703, + "learning_rate": 1.5082372932855753e-05, + "loss": 1.0446, + "step": 14873 + }, + { + "epoch": 0.35, + "grad_norm": 1.1132023171626337, + "learning_rate": 1.5081715774069191e-05, + "loss": 0.9918, + "step": 14874 + }, + { + "epoch": 0.35, + "grad_norm": 1.9719410316757775, + "learning_rate": 1.5081058585695272e-05, + "loss": 1.1933, + "step": 14875 + }, + { + "epoch": 0.35, + "grad_norm": 2.5971186688422425, + "learning_rate": 1.508040136773782e-05, + "loss": 0.9517, + "step": 14876 + }, + { + "epoch": 0.35, + "grad_norm": 2.2301468748462736, + "learning_rate": 1.5079744120200663e-05, + "loss": 1.1596, + "step": 14877 + }, + { + "epoch": 0.35, + "grad_norm": 1.8889329185237405, + "learning_rate": 1.5079086843087626e-05, + "loss": 1.0763, + "step": 14878 + }, + { + "epoch": 0.35, + "grad_norm": 1.0747627736063703, + "learning_rate": 1.507842953640254e-05, + "loss": 0.9951, + "step": 14879 + }, + { + "epoch": 0.35, + "grad_norm": 2.141297869408198, + "learning_rate": 1.507777220014923e-05, + "loss": 1.0553, + "step": 14880 + }, + { + "epoch": 0.35, + "grad_norm": 2.123932340685398, + "learning_rate": 1.5077114834331524e-05, + "loss": 0.9319, + "step": 14881 + }, + { + "epoch": 0.35, + "grad_norm": 2.891258460627549, + "learning_rate": 1.5076457438953242e-05, + "loss": 0.9971, + "step": 14882 + }, + { + "epoch": 0.35, + "grad_norm": 2.323787024615931, + "learning_rate": 1.5075800014018223e-05, + "loss": 0.9495, + "step": 14883 + }, + { + "epoch": 0.35, + "grad_norm": 2.275632814894144, + "learning_rate": 1.5075142559530287e-05, + "loss": 0.9149, + "step": 14884 + }, + { + "epoch": 0.35, + "grad_norm": 1.9224608006351838, + "learning_rate": 1.5074485075493263e-05, + "loss": 1.1049, + "step": 14885 + }, + { + "epoch": 0.35, + "grad_norm": 2.232844909312798, + "learning_rate": 1.5073827561910983e-05, + "loss": 0.9581, + "step": 14886 + }, + { + "epoch": 0.35, + "grad_norm": 2.4169237379534856, + "learning_rate": 1.5073170018787272e-05, + "loss": 1.0851, + "step": 14887 + }, + { + "epoch": 0.35, + "grad_norm": 1.129310415114859, + "learning_rate": 1.5072512446125958e-05, + "loss": 0.9494, + "step": 14888 + }, + { + "epoch": 0.35, + "grad_norm": 1.1364262768294509, + "learning_rate": 1.5071854843930873e-05, + "loss": 0.9565, + "step": 14889 + }, + { + "epoch": 0.35, + "grad_norm": 1.842105819088064, + "learning_rate": 1.5071197212205842e-05, + "loss": 0.9997, + "step": 14890 + }, + { + "epoch": 0.35, + "grad_norm": 1.8589503202487334, + "learning_rate": 1.5070539550954696e-05, + "loss": 1.0294, + "step": 14891 + }, + { + "epoch": 0.35, + "grad_norm": 1.9157565626013204, + "learning_rate": 1.5069881860181261e-05, + "loss": 1.0936, + "step": 14892 + }, + { + "epoch": 0.35, + "grad_norm": 2.1671466439269844, + "learning_rate": 1.5069224139889372e-05, + "loss": 1.0732, + "step": 14893 + }, + { + "epoch": 0.35, + "grad_norm": 1.9967864556495747, + "learning_rate": 1.506856639008285e-05, + "loss": 0.9227, + "step": 14894 + }, + { + "epoch": 0.35, + "grad_norm": 1.081341591142526, + "learning_rate": 1.5067908610765536e-05, + "loss": 0.9796, + "step": 14895 + }, + { + "epoch": 0.35, + "grad_norm": 2.010799235543915, + "learning_rate": 1.506725080194125e-05, + "loss": 1.1552, + "step": 14896 + }, + { + "epoch": 0.35, + "grad_norm": 1.7059458542038974, + "learning_rate": 1.5066592963613826e-05, + "loss": 1.029, + "step": 14897 + }, + { + "epoch": 0.35, + "grad_norm": 2.026660293729273, + "learning_rate": 1.5065935095787093e-05, + "loss": 0.996, + "step": 14898 + }, + { + "epoch": 0.35, + "grad_norm": 1.9345855710483588, + "learning_rate": 1.5065277198464885e-05, + "loss": 1.0849, + "step": 14899 + }, + { + "epoch": 0.35, + "grad_norm": 1.948854166132352, + "learning_rate": 1.5064619271651027e-05, + "loss": 1.0423, + "step": 14900 + }, + { + "epoch": 0.35, + "grad_norm": 2.181146183714015, + "learning_rate": 1.5063961315349352e-05, + "loss": 1.0838, + "step": 14901 + }, + { + "epoch": 0.35, + "grad_norm": 1.8584745584955742, + "learning_rate": 1.5063303329563693e-05, + "loss": 1.0493, + "step": 14902 + }, + { + "epoch": 0.35, + "grad_norm": 2.0759305910218453, + "learning_rate": 1.5062645314297877e-05, + "loss": 1.0256, + "step": 14903 + }, + { + "epoch": 0.35, + "grad_norm": 1.8191511337510404, + "learning_rate": 1.5061987269555735e-05, + "loss": 1.0151, + "step": 14904 + }, + { + "epoch": 0.35, + "grad_norm": 1.9939770137545394, + "learning_rate": 1.5061329195341105e-05, + "loss": 1.034, + "step": 14905 + }, + { + "epoch": 0.35, + "grad_norm": 2.094902643170229, + "learning_rate": 1.5060671091657813e-05, + "loss": 1.1328, + "step": 14906 + }, + { + "epoch": 0.35, + "grad_norm": 2.0874427328608163, + "learning_rate": 1.506001295850969e-05, + "loss": 1.103, + "step": 14907 + }, + { + "epoch": 0.35, + "grad_norm": 2.4336901783590883, + "learning_rate": 1.5059354795900571e-05, + "loss": 0.9998, + "step": 14908 + }, + { + "epoch": 0.35, + "grad_norm": 3.09976123001859, + "learning_rate": 1.5058696603834287e-05, + "loss": 0.9939, + "step": 14909 + }, + { + "epoch": 0.35, + "grad_norm": 1.922511291522678, + "learning_rate": 1.5058038382314673e-05, + "loss": 0.98, + "step": 14910 + }, + { + "epoch": 0.35, + "grad_norm": 2.3588607893655547, + "learning_rate": 1.505738013134555e-05, + "loss": 0.9596, + "step": 14911 + }, + { + "epoch": 0.35, + "grad_norm": 1.1816707658470316, + "learning_rate": 1.5056721850930766e-05, + "loss": 1.0376, + "step": 14912 + }, + { + "epoch": 0.35, + "grad_norm": 2.3086105832585115, + "learning_rate": 1.505606354107414e-05, + "loss": 0.9336, + "step": 14913 + }, + { + "epoch": 0.35, + "grad_norm": 2.6502859643808248, + "learning_rate": 1.5055405201779518e-05, + "loss": 0.935, + "step": 14914 + }, + { + "epoch": 0.35, + "grad_norm": 2.0786120690505094, + "learning_rate": 1.5054746833050721e-05, + "loss": 0.9775, + "step": 14915 + }, + { + "epoch": 0.35, + "grad_norm": 2.3352335180852952, + "learning_rate": 1.505408843489159e-05, + "loss": 1.1271, + "step": 14916 + }, + { + "epoch": 0.35, + "grad_norm": 2.18272610025668, + "learning_rate": 1.5053430007305958e-05, + "loss": 1.0284, + "step": 14917 + }, + { + "epoch": 0.35, + "grad_norm": 1.9182078606752118, + "learning_rate": 1.5052771550297651e-05, + "loss": 1.1033, + "step": 14918 + }, + { + "epoch": 0.35, + "grad_norm": 2.3261687218070577, + "learning_rate": 1.5052113063870514e-05, + "loss": 0.9968, + "step": 14919 + }, + { + "epoch": 0.35, + "grad_norm": 1.9747212368383589, + "learning_rate": 1.5051454548028373e-05, + "loss": 0.9659, + "step": 14920 + }, + { + "epoch": 0.35, + "grad_norm": 2.0304252968866923, + "learning_rate": 1.5050796002775065e-05, + "loss": 0.9239, + "step": 14921 + }, + { + "epoch": 0.35, + "grad_norm": 1.9646956260449242, + "learning_rate": 1.5050137428114422e-05, + "loss": 1.056, + "step": 14922 + }, + { + "epoch": 0.35, + "grad_norm": 2.1040988201556194, + "learning_rate": 1.5049478824050282e-05, + "loss": 1.0892, + "step": 14923 + }, + { + "epoch": 0.35, + "grad_norm": 2.0270443122224093, + "learning_rate": 1.5048820190586478e-05, + "loss": 1.1553, + "step": 14924 + }, + { + "epoch": 0.35, + "grad_norm": 2.061938523681059, + "learning_rate": 1.5048161527726843e-05, + "loss": 0.9886, + "step": 14925 + }, + { + "epoch": 0.35, + "grad_norm": 1.793443956088587, + "learning_rate": 1.5047502835475211e-05, + "loss": 1.0731, + "step": 14926 + }, + { + "epoch": 0.35, + "grad_norm": 1.9410939428340426, + "learning_rate": 1.5046844113835426e-05, + "loss": 1.1225, + "step": 14927 + }, + { + "epoch": 0.35, + "grad_norm": 2.263083962322508, + "learning_rate": 1.504618536281131e-05, + "loss": 1.1039, + "step": 14928 + }, + { + "epoch": 0.35, + "grad_norm": 2.0210307633964653, + "learning_rate": 1.5045526582406708e-05, + "loss": 0.96, + "step": 14929 + }, + { + "epoch": 0.35, + "grad_norm": 2.3165055514210344, + "learning_rate": 1.5044867772625455e-05, + "loss": 1.1012, + "step": 14930 + }, + { + "epoch": 0.35, + "grad_norm": 2.280056181774779, + "learning_rate": 1.5044208933471383e-05, + "loss": 1.0231, + "step": 14931 + }, + { + "epoch": 0.35, + "grad_norm": 2.301063040468324, + "learning_rate": 1.5043550064948327e-05, + "loss": 0.8845, + "step": 14932 + }, + { + "epoch": 0.35, + "grad_norm": 3.3221924477013474, + "learning_rate": 1.504289116706013e-05, + "loss": 1.0808, + "step": 14933 + }, + { + "epoch": 0.35, + "grad_norm": 2.112230075937281, + "learning_rate": 1.5042232239810618e-05, + "loss": 1.06, + "step": 14934 + }, + { + "epoch": 0.35, + "grad_norm": 2.681454435766169, + "learning_rate": 1.5041573283203641e-05, + "loss": 1.069, + "step": 14935 + }, + { + "epoch": 0.35, + "grad_norm": 2.2933057617791244, + "learning_rate": 1.5040914297243023e-05, + "loss": 1.0701, + "step": 14936 + }, + { + "epoch": 0.35, + "grad_norm": 2.05597542128639, + "learning_rate": 1.5040255281932608e-05, + "loss": 1.1086, + "step": 14937 + }, + { + "epoch": 0.35, + "grad_norm": 1.927733724723482, + "learning_rate": 1.5039596237276234e-05, + "loss": 0.9513, + "step": 14938 + }, + { + "epoch": 0.35, + "grad_norm": 2.0155651235881784, + "learning_rate": 1.5038937163277729e-05, + "loss": 0.9994, + "step": 14939 + }, + { + "epoch": 0.35, + "grad_norm": 1.1986977071616527, + "learning_rate": 1.503827805994094e-05, + "loss": 1.0374, + "step": 14940 + }, + { + "epoch": 0.35, + "grad_norm": 2.0217916000643728, + "learning_rate": 1.50376189272697e-05, + "loss": 1.0452, + "step": 14941 + }, + { + "epoch": 0.35, + "grad_norm": 2.174158013230859, + "learning_rate": 1.5036959765267852e-05, + "loss": 0.9776, + "step": 14942 + }, + { + "epoch": 0.35, + "grad_norm": 2.1093860146888384, + "learning_rate": 1.5036300573939224e-05, + "loss": 0.9798, + "step": 14943 + }, + { + "epoch": 0.35, + "grad_norm": 2.117196446176399, + "learning_rate": 1.5035641353287662e-05, + "loss": 1.0007, + "step": 14944 + }, + { + "epoch": 0.35, + "grad_norm": 2.0583792295667336, + "learning_rate": 1.5034982103317001e-05, + "loss": 0.9244, + "step": 14945 + }, + { + "epoch": 0.35, + "grad_norm": 2.0337865952588805, + "learning_rate": 1.5034322824031083e-05, + "loss": 1.1123, + "step": 14946 + }, + { + "epoch": 0.35, + "grad_norm": 2.0647844985382005, + "learning_rate": 1.5033663515433738e-05, + "loss": 1.0296, + "step": 14947 + }, + { + "epoch": 0.35, + "grad_norm": 2.0889208713399356, + "learning_rate": 1.5033004177528816e-05, + "loss": 1.0422, + "step": 14948 + }, + { + "epoch": 0.35, + "grad_norm": 1.7909119119952048, + "learning_rate": 1.5032344810320148e-05, + "loss": 0.9286, + "step": 14949 + }, + { + "epoch": 0.35, + "grad_norm": 2.067126357196292, + "learning_rate": 1.5031685413811574e-05, + "loss": 1.0098, + "step": 14950 + }, + { + "epoch": 0.35, + "grad_norm": 2.3569795229846306, + "learning_rate": 1.5031025988006935e-05, + "loss": 1.1569, + "step": 14951 + }, + { + "epoch": 0.35, + "grad_norm": 1.9293383867591707, + "learning_rate": 1.503036653291007e-05, + "loss": 1.111, + "step": 14952 + }, + { + "epoch": 0.35, + "grad_norm": 1.925631047026715, + "learning_rate": 1.502970704852482e-05, + "loss": 1.0445, + "step": 14953 + }, + { + "epoch": 0.35, + "grad_norm": 2.1373843404821535, + "learning_rate": 1.502904753485502e-05, + "loss": 1.035, + "step": 14954 + }, + { + "epoch": 0.35, + "grad_norm": 2.038792996687895, + "learning_rate": 1.5028387991904517e-05, + "loss": 1.1872, + "step": 14955 + }, + { + "epoch": 0.35, + "grad_norm": 2.0780246093484975, + "learning_rate": 1.5027728419677146e-05, + "loss": 1.0559, + "step": 14956 + }, + { + "epoch": 0.35, + "grad_norm": 2.0809532388728393, + "learning_rate": 1.5027068818176747e-05, + "loss": 0.9502, + "step": 14957 + }, + { + "epoch": 0.35, + "grad_norm": 2.3080116114791327, + "learning_rate": 1.502640918740716e-05, + "loss": 1.081, + "step": 14958 + }, + { + "epoch": 0.35, + "grad_norm": 2.802985608407815, + "learning_rate": 1.502574952737223e-05, + "loss": 1.0001, + "step": 14959 + }, + { + "epoch": 0.35, + "grad_norm": 2.2958362116386533, + "learning_rate": 1.5025089838075797e-05, + "loss": 1.0695, + "step": 14960 + }, + { + "epoch": 0.35, + "grad_norm": 2.0110054745190484, + "learning_rate": 1.50244301195217e-05, + "loss": 1.0644, + "step": 14961 + }, + { + "epoch": 0.35, + "grad_norm": 2.011612543524675, + "learning_rate": 1.5023770371713778e-05, + "loss": 1.2069, + "step": 14962 + }, + { + "epoch": 0.35, + "grad_norm": 1.8945670685503713, + "learning_rate": 1.5023110594655875e-05, + "loss": 0.9865, + "step": 14963 + }, + { + "epoch": 0.35, + "grad_norm": 2.0856131148393526, + "learning_rate": 1.5022450788351828e-05, + "loss": 1.0531, + "step": 14964 + }, + { + "epoch": 0.35, + "grad_norm": 1.1369206871422304, + "learning_rate": 1.5021790952805489e-05, + "loss": 1.024, + "step": 14965 + }, + { + "epoch": 0.35, + "grad_norm": 2.015346598319527, + "learning_rate": 1.5021131088020688e-05, + "loss": 1.0076, + "step": 14966 + }, + { + "epoch": 0.35, + "grad_norm": 2.322568281573979, + "learning_rate": 1.5020471194001274e-05, + "loss": 0.8919, + "step": 14967 + }, + { + "epoch": 0.35, + "grad_norm": 1.9345028115444447, + "learning_rate": 1.501981127075109e-05, + "loss": 1.0498, + "step": 14968 + }, + { + "epoch": 0.35, + "grad_norm": 2.362166399000907, + "learning_rate": 1.501915131827397e-05, + "loss": 1.0243, + "step": 14969 + }, + { + "epoch": 0.35, + "grad_norm": 1.9591875222929547, + "learning_rate": 1.5018491336573768e-05, + "loss": 0.9819, + "step": 14970 + }, + { + "epoch": 0.35, + "grad_norm": 2.3855754873690835, + "learning_rate": 1.5017831325654316e-05, + "loss": 1.0619, + "step": 14971 + }, + { + "epoch": 0.35, + "grad_norm": 1.879865780790386, + "learning_rate": 1.5017171285519463e-05, + "loss": 1.0185, + "step": 14972 + }, + { + "epoch": 0.35, + "grad_norm": 1.9209972006149947, + "learning_rate": 1.5016511216173052e-05, + "loss": 0.9246, + "step": 14973 + }, + { + "epoch": 0.35, + "grad_norm": 1.97225065700214, + "learning_rate": 1.5015851117618923e-05, + "loss": 1.1128, + "step": 14974 + }, + { + "epoch": 0.35, + "grad_norm": 2.281645200697352, + "learning_rate": 1.501519098986092e-05, + "loss": 1.1193, + "step": 14975 + }, + { + "epoch": 0.35, + "grad_norm": 3.807555395555874, + "learning_rate": 1.5014530832902886e-05, + "loss": 0.9895, + "step": 14976 + }, + { + "epoch": 0.35, + "grad_norm": 1.9841106686480676, + "learning_rate": 1.501387064674867e-05, + "loss": 1.1034, + "step": 14977 + }, + { + "epoch": 0.35, + "grad_norm": 1.981990131803845, + "learning_rate": 1.501321043140211e-05, + "loss": 1.091, + "step": 14978 + }, + { + "epoch": 0.35, + "grad_norm": 2.66457941037471, + "learning_rate": 1.5012550186867051e-05, + "loss": 1.1634, + "step": 14979 + }, + { + "epoch": 0.35, + "grad_norm": 2.0425921028192953, + "learning_rate": 1.501188991314734e-05, + "loss": 0.9903, + "step": 14980 + }, + { + "epoch": 0.35, + "grad_norm": 1.9278514367797386, + "learning_rate": 1.5011229610246815e-05, + "loss": 1.0065, + "step": 14981 + }, + { + "epoch": 0.35, + "grad_norm": 1.7708692885864303, + "learning_rate": 1.5010569278169327e-05, + "loss": 1.1172, + "step": 14982 + }, + { + "epoch": 0.35, + "grad_norm": 1.9095253782651835, + "learning_rate": 1.5009908916918719e-05, + "loss": 0.9452, + "step": 14983 + }, + { + "epoch": 0.35, + "grad_norm": 2.34046047211617, + "learning_rate": 1.5009248526498834e-05, + "loss": 1.1663, + "step": 14984 + }, + { + "epoch": 0.35, + "grad_norm": 2.1232346188074596, + "learning_rate": 1.500858810691352e-05, + "loss": 0.9895, + "step": 14985 + }, + { + "epoch": 0.35, + "grad_norm": 2.768179747928592, + "learning_rate": 1.5007927658166619e-05, + "loss": 1.0234, + "step": 14986 + }, + { + "epoch": 0.35, + "grad_norm": 2.131783357855166, + "learning_rate": 1.5007267180261975e-05, + "loss": 1.0252, + "step": 14987 + }, + { + "epoch": 0.35, + "grad_norm": 1.9241333944394772, + "learning_rate": 1.500660667320344e-05, + "loss": 0.9318, + "step": 14988 + }, + { + "epoch": 0.35, + "grad_norm": 2.105070040610404, + "learning_rate": 1.5005946136994857e-05, + "loss": 0.9063, + "step": 14989 + }, + { + "epoch": 0.35, + "grad_norm": 1.9481714916428925, + "learning_rate": 1.5005285571640067e-05, + "loss": 0.9748, + "step": 14990 + }, + { + "epoch": 0.35, + "grad_norm": 2.093775434626433, + "learning_rate": 1.5004624977142923e-05, + "loss": 0.8528, + "step": 14991 + }, + { + "epoch": 0.35, + "grad_norm": 2.056215851454542, + "learning_rate": 1.5003964353507265e-05, + "loss": 1.0279, + "step": 14992 + }, + { + "epoch": 0.35, + "grad_norm": 2.091183082001444, + "learning_rate": 1.5003303700736944e-05, + "loss": 1.0741, + "step": 14993 + }, + { + "epoch": 0.35, + "grad_norm": 4.127030648053116, + "learning_rate": 1.5002643018835806e-05, + "loss": 1.1481, + "step": 14994 + }, + { + "epoch": 0.35, + "grad_norm": 1.1159813539268686, + "learning_rate": 1.5001982307807694e-05, + "loss": 1.0381, + "step": 14995 + }, + { + "epoch": 0.35, + "grad_norm": 1.7799872560968688, + "learning_rate": 1.5001321567656459e-05, + "loss": 0.9791, + "step": 14996 + }, + { + "epoch": 0.35, + "grad_norm": 1.0736089435335285, + "learning_rate": 1.5000660798385943e-05, + "loss": 0.9524, + "step": 14997 + }, + { + "epoch": 0.35, + "grad_norm": 2.154429424951408, + "learning_rate": 1.5000000000000002e-05, + "loss": 1.0715, + "step": 14998 + }, + { + "epoch": 0.35, + "grad_norm": 1.908535690669817, + "learning_rate": 1.4999339172502477e-05, + "loss": 0.9698, + "step": 14999 + }, + { + "epoch": 0.35, + "grad_norm": 1.9577646504047812, + "learning_rate": 1.4998678315897213e-05, + "loss": 1.0058, + "step": 15000 + }, + { + "epoch": 0.35, + "grad_norm": 1.0827205524653634, + "learning_rate": 1.4998017430188063e-05, + "loss": 0.9804, + "step": 15001 + }, + { + "epoch": 0.35, + "grad_norm": 1.9495781264287728, + "learning_rate": 1.4997356515378877e-05, + "loss": 1.0483, + "step": 15002 + }, + { + "epoch": 0.35, + "grad_norm": 2.0998314365236976, + "learning_rate": 1.4996695571473495e-05, + "loss": 1.0005, + "step": 15003 + }, + { + "epoch": 0.35, + "grad_norm": 1.9934534937893023, + "learning_rate": 1.4996034598475769e-05, + "loss": 1.0006, + "step": 15004 + }, + { + "epoch": 0.35, + "grad_norm": 3.7709829366519387, + "learning_rate": 1.4995373596389549e-05, + "loss": 0.8327, + "step": 15005 + }, + { + "epoch": 0.35, + "grad_norm": 2.0547866953721696, + "learning_rate": 1.499471256521868e-05, + "loss": 0.9205, + "step": 15006 + }, + { + "epoch": 0.35, + "grad_norm": 2.0084327170506993, + "learning_rate": 1.4994051504967018e-05, + "loss": 1.0444, + "step": 15007 + }, + { + "epoch": 0.35, + "grad_norm": 1.8874024248076275, + "learning_rate": 1.4993390415638405e-05, + "loss": 0.9378, + "step": 15008 + }, + { + "epoch": 0.35, + "grad_norm": 2.0588133402350977, + "learning_rate": 1.4992729297236693e-05, + "loss": 0.9181, + "step": 15009 + }, + { + "epoch": 0.35, + "grad_norm": 2.0125795319047577, + "learning_rate": 1.4992068149765726e-05, + "loss": 0.9768, + "step": 15010 + }, + { + "epoch": 0.35, + "grad_norm": 1.1925079083255434, + "learning_rate": 1.499140697322936e-05, + "loss": 1.0302, + "step": 15011 + }, + { + "epoch": 0.35, + "grad_norm": 3.0222149129458455, + "learning_rate": 1.4990745767631445e-05, + "loss": 1.0443, + "step": 15012 + }, + { + "epoch": 0.35, + "grad_norm": 2.4823531012608977, + "learning_rate": 1.4990084532975826e-05, + "loss": 1.0594, + "step": 15013 + }, + { + "epoch": 0.35, + "grad_norm": 2.0837353071077414, + "learning_rate": 1.4989423269266357e-05, + "loss": 1.1417, + "step": 15014 + }, + { + "epoch": 0.35, + "grad_norm": 1.9124968861685248, + "learning_rate": 1.4988761976506885e-05, + "loss": 1.0451, + "step": 15015 + }, + { + "epoch": 0.35, + "grad_norm": 1.8439610254927177, + "learning_rate": 1.4988100654701258e-05, + "loss": 1.0752, + "step": 15016 + }, + { + "epoch": 0.35, + "grad_norm": 2.2106254643041248, + "learning_rate": 1.4987439303853336e-05, + "loss": 1.0144, + "step": 15017 + }, + { + "epoch": 0.35, + "grad_norm": 2.2133266684542447, + "learning_rate": 1.4986777923966958e-05, + "loss": 1.1497, + "step": 15018 + }, + { + "epoch": 0.35, + "grad_norm": 2.2485677038523137, + "learning_rate": 1.4986116515045982e-05, + "loss": 1.0525, + "step": 15019 + }, + { + "epoch": 0.35, + "grad_norm": 1.7934305301118671, + "learning_rate": 1.4985455077094258e-05, + "loss": 1.2265, + "step": 15020 + }, + { + "epoch": 0.35, + "grad_norm": 4.302391818746349, + "learning_rate": 1.4984793610115635e-05, + "loss": 1.0638, + "step": 15021 + }, + { + "epoch": 0.35, + "grad_norm": 1.0973289110260505, + "learning_rate": 1.4984132114113965e-05, + "loss": 0.9898, + "step": 15022 + }, + { + "epoch": 0.35, + "grad_norm": 2.4124161795430292, + "learning_rate": 1.4983470589093103e-05, + "loss": 0.9499, + "step": 15023 + }, + { + "epoch": 0.35, + "grad_norm": 2.2072595955024377, + "learning_rate": 1.4982809035056894e-05, + "loss": 1.1127, + "step": 15024 + }, + { + "epoch": 0.35, + "grad_norm": 2.1005150180258765, + "learning_rate": 1.4982147452009197e-05, + "loss": 1.1493, + "step": 15025 + }, + { + "epoch": 0.35, + "grad_norm": 1.8956053889643405, + "learning_rate": 1.4981485839953855e-05, + "loss": 1.0461, + "step": 15026 + }, + { + "epoch": 0.35, + "grad_norm": 2.1096486868840443, + "learning_rate": 1.498082419889473e-05, + "loss": 0.8495, + "step": 15027 + }, + { + "epoch": 0.35, + "grad_norm": 2.0489380275625244, + "learning_rate": 1.4980162528835666e-05, + "loss": 1.0409, + "step": 15028 + }, + { + "epoch": 0.35, + "grad_norm": 2.02052746021569, + "learning_rate": 1.4979500829780522e-05, + "loss": 1.0293, + "step": 15029 + }, + { + "epoch": 0.35, + "grad_norm": 2.1430167558341617, + "learning_rate": 1.4978839101733144e-05, + "loss": 1.0315, + "step": 15030 + }, + { + "epoch": 0.35, + "grad_norm": 2.3451932039995906, + "learning_rate": 1.4978177344697394e-05, + "loss": 1.1849, + "step": 15031 + }, + { + "epoch": 0.35, + "grad_norm": 2.172897496273705, + "learning_rate": 1.4977515558677116e-05, + "loss": 0.9998, + "step": 15032 + }, + { + "epoch": 0.35, + "grad_norm": 1.7668638435329664, + "learning_rate": 1.4976853743676165e-05, + "loss": 1.1408, + "step": 15033 + }, + { + "epoch": 0.35, + "grad_norm": 1.9825926726503273, + "learning_rate": 1.4976191899698398e-05, + "loss": 1.0142, + "step": 15034 + }, + { + "epoch": 0.35, + "grad_norm": 2.1888394843196934, + "learning_rate": 1.4975530026747667e-05, + "loss": 1.1251, + "step": 15035 + }, + { + "epoch": 0.35, + "grad_norm": 1.9898099361311063, + "learning_rate": 1.4974868124827823e-05, + "loss": 1.0827, + "step": 15036 + }, + { + "epoch": 0.35, + "grad_norm": 1.1165722195975374, + "learning_rate": 1.4974206193942724e-05, + "loss": 0.9086, + "step": 15037 + }, + { + "epoch": 0.35, + "grad_norm": 2.3375004855746067, + "learning_rate": 1.497354423409622e-05, + "loss": 1.0003, + "step": 15038 + }, + { + "epoch": 0.35, + "grad_norm": 1.8490916938046158, + "learning_rate": 1.4972882245292165e-05, + "loss": 0.9469, + "step": 15039 + }, + { + "epoch": 0.35, + "grad_norm": 2.1451014571041633, + "learning_rate": 1.4972220227534422e-05, + "loss": 1.0797, + "step": 15040 + }, + { + "epoch": 0.35, + "grad_norm": 1.9982365680610346, + "learning_rate": 1.4971558180826832e-05, + "loss": 0.9926, + "step": 15041 + }, + { + "epoch": 0.35, + "grad_norm": 2.201279824985369, + "learning_rate": 1.497089610517326e-05, + "loss": 1.1266, + "step": 15042 + }, + { + "epoch": 0.35, + "grad_norm": 1.9502142235264521, + "learning_rate": 1.4970234000577555e-05, + "loss": 1.1728, + "step": 15043 + }, + { + "epoch": 0.35, + "grad_norm": 2.142861354560887, + "learning_rate": 1.4969571867043576e-05, + "loss": 1.0935, + "step": 15044 + }, + { + "epoch": 0.35, + "grad_norm": 1.8237805331557568, + "learning_rate": 1.4968909704575171e-05, + "loss": 1.0718, + "step": 15045 + }, + { + "epoch": 0.35, + "grad_norm": 2.131879789794342, + "learning_rate": 1.4968247513176209e-05, + "loss": 0.9672, + "step": 15046 + }, + { + "epoch": 0.35, + "grad_norm": 1.77049815869106, + "learning_rate": 1.4967585292850531e-05, + "loss": 1.1462, + "step": 15047 + }, + { + "epoch": 0.35, + "grad_norm": 1.1062574718505542, + "learning_rate": 1.4966923043602002e-05, + "loss": 0.9492, + "step": 15048 + }, + { + "epoch": 0.35, + "grad_norm": 2.3150968767970532, + "learning_rate": 1.4966260765434472e-05, + "loss": 0.9747, + "step": 15049 + }, + { + "epoch": 0.35, + "grad_norm": 2.258714587008007, + "learning_rate": 1.4965598458351797e-05, + "loss": 1.0007, + "step": 15050 + }, + { + "epoch": 0.35, + "grad_norm": 2.1755678607932145, + "learning_rate": 1.4964936122357843e-05, + "loss": 1.1121, + "step": 15051 + }, + { + "epoch": 0.35, + "grad_norm": 2.0367671770801388, + "learning_rate": 1.4964273757456451e-05, + "loss": 1.057, + "step": 15052 + }, + { + "epoch": 0.35, + "grad_norm": 2.094361569654992, + "learning_rate": 1.4963611363651494e-05, + "loss": 1.0799, + "step": 15053 + }, + { + "epoch": 0.35, + "grad_norm": 2.13966663906516, + "learning_rate": 1.4962948940946814e-05, + "loss": 1.0206, + "step": 15054 + }, + { + "epoch": 0.35, + "grad_norm": 3.162348438021199, + "learning_rate": 1.4962286489346276e-05, + "loss": 0.9798, + "step": 15055 + }, + { + "epoch": 0.35, + "grad_norm": 1.9582575620916431, + "learning_rate": 1.4961624008853735e-05, + "loss": 1.0743, + "step": 15056 + }, + { + "epoch": 0.35, + "grad_norm": 2.0265159913085093, + "learning_rate": 1.4960961499473048e-05, + "loss": 1.0332, + "step": 15057 + }, + { + "epoch": 0.35, + "grad_norm": 2.1922674962240336, + "learning_rate": 1.4960298961208073e-05, + "loss": 1.0813, + "step": 15058 + }, + { + "epoch": 0.35, + "grad_norm": 2.1907030532173093, + "learning_rate": 1.4959636394062666e-05, + "loss": 1.0094, + "step": 15059 + }, + { + "epoch": 0.35, + "grad_norm": 2.5393964736712054, + "learning_rate": 1.4958973798040685e-05, + "loss": 1.1044, + "step": 15060 + }, + { + "epoch": 0.35, + "grad_norm": 1.1045044094939058, + "learning_rate": 1.4958311173145991e-05, + "loss": 1.0012, + "step": 15061 + }, + { + "epoch": 0.35, + "grad_norm": 1.9157468048699502, + "learning_rate": 1.495764851938244e-05, + "loss": 1.1213, + "step": 15062 + }, + { + "epoch": 0.35, + "grad_norm": 2.104899424450819, + "learning_rate": 1.4956985836753887e-05, + "loss": 1.0559, + "step": 15063 + }, + { + "epoch": 0.35, + "grad_norm": 1.1779324410870775, + "learning_rate": 1.4956323125264193e-05, + "loss": 1.0017, + "step": 15064 + }, + { + "epoch": 0.35, + "grad_norm": 1.135515908392162, + "learning_rate": 1.4955660384917218e-05, + "loss": 0.9347, + "step": 15065 + }, + { + "epoch": 0.35, + "grad_norm": 2.50603198810147, + "learning_rate": 1.495499761571682e-05, + "loss": 1.2522, + "step": 15066 + }, + { + "epoch": 0.35, + "grad_norm": 2.0358283415026754, + "learning_rate": 1.4954334817666855e-05, + "loss": 1.1553, + "step": 15067 + }, + { + "epoch": 0.35, + "grad_norm": 2.1729377364097457, + "learning_rate": 1.4953671990771189e-05, + "loss": 0.9646, + "step": 15068 + }, + { + "epoch": 0.36, + "grad_norm": 2.0070934463415857, + "learning_rate": 1.4953009135033671e-05, + "loss": 1.02, + "step": 15069 + }, + { + "epoch": 0.36, + "grad_norm": 2.01243769740562, + "learning_rate": 1.495234625045817e-05, + "loss": 0.9455, + "step": 15070 + }, + { + "epoch": 0.36, + "grad_norm": 2.4036742381475045, + "learning_rate": 1.4951683337048536e-05, + "loss": 1.0416, + "step": 15071 + }, + { + "epoch": 0.36, + "grad_norm": 1.832429074355303, + "learning_rate": 1.495102039480864e-05, + "loss": 0.9452, + "step": 15072 + }, + { + "epoch": 0.36, + "grad_norm": 2.011957794924091, + "learning_rate": 1.495035742374233e-05, + "loss": 0.9781, + "step": 15073 + }, + { + "epoch": 0.36, + "grad_norm": 2.066348298036289, + "learning_rate": 1.4949694423853478e-05, + "loss": 1.0724, + "step": 15074 + }, + { + "epoch": 0.36, + "grad_norm": 2.0371127009707517, + "learning_rate": 1.4949031395145932e-05, + "loss": 1.0594, + "step": 15075 + }, + { + "epoch": 0.36, + "grad_norm": 2.0067849308223833, + "learning_rate": 1.4948368337623562e-05, + "loss": 1.0046, + "step": 15076 + }, + { + "epoch": 0.36, + "grad_norm": 3.438506027077561, + "learning_rate": 1.4947705251290223e-05, + "loss": 1.1091, + "step": 15077 + }, + { + "epoch": 0.36, + "grad_norm": 1.9432650509064837, + "learning_rate": 1.4947042136149778e-05, + "loss": 1.0296, + "step": 15078 + }, + { + "epoch": 0.36, + "grad_norm": 2.4118806147497667, + "learning_rate": 1.494637899220609e-05, + "loss": 1.0513, + "step": 15079 + }, + { + "epoch": 0.36, + "grad_norm": 2.694948240040044, + "learning_rate": 1.4945715819463013e-05, + "loss": 1.0578, + "step": 15080 + }, + { + "epoch": 0.36, + "grad_norm": 2.657660908916564, + "learning_rate": 1.4945052617924415e-05, + "loss": 1.0774, + "step": 15081 + }, + { + "epoch": 0.36, + "grad_norm": 1.8244244922276112, + "learning_rate": 1.4944389387594156e-05, + "loss": 1.0319, + "step": 15082 + }, + { + "epoch": 0.36, + "grad_norm": 2.900789003033709, + "learning_rate": 1.4943726128476097e-05, + "loss": 0.998, + "step": 15083 + }, + { + "epoch": 0.36, + "grad_norm": 1.776987256922836, + "learning_rate": 1.4943062840574096e-05, + "loss": 1.0508, + "step": 15084 + }, + { + "epoch": 0.36, + "grad_norm": 2.0888173013555735, + "learning_rate": 1.4942399523892021e-05, + "loss": 1.0771, + "step": 15085 + }, + { + "epoch": 0.36, + "grad_norm": 2.0953461523210732, + "learning_rate": 1.494173617843373e-05, + "loss": 0.9778, + "step": 15086 + }, + { + "epoch": 0.36, + "grad_norm": 1.998437266977981, + "learning_rate": 1.4941072804203089e-05, + "loss": 1.0599, + "step": 15087 + }, + { + "epoch": 0.36, + "grad_norm": 1.7000897905852643, + "learning_rate": 1.4940409401203955e-05, + "loss": 1.1126, + "step": 15088 + }, + { + "epoch": 0.36, + "grad_norm": 2.0226476655408083, + "learning_rate": 1.4939745969440194e-05, + "loss": 0.9499, + "step": 15089 + }, + { + "epoch": 0.36, + "grad_norm": 1.1367933641685182, + "learning_rate": 1.4939082508915668e-05, + "loss": 0.9555, + "step": 15090 + }, + { + "epoch": 0.36, + "grad_norm": 1.1115547629557117, + "learning_rate": 1.4938419019634243e-05, + "loss": 0.9557, + "step": 15091 + }, + { + "epoch": 0.36, + "grad_norm": 2.505571636313883, + "learning_rate": 1.4937755501599774e-05, + "loss": 1.0443, + "step": 15092 + }, + { + "epoch": 0.36, + "grad_norm": 1.0563646343244286, + "learning_rate": 1.4937091954816131e-05, + "loss": 1.0081, + "step": 15093 + }, + { + "epoch": 0.36, + "grad_norm": 2.8403988050455347, + "learning_rate": 1.4936428379287176e-05, + "loss": 1.0066, + "step": 15094 + }, + { + "epoch": 0.36, + "grad_norm": 2.0491497691448326, + "learning_rate": 1.4935764775016775e-05, + "loss": 1.0054, + "step": 15095 + }, + { + "epoch": 0.36, + "grad_norm": 2.116243990451648, + "learning_rate": 1.4935101142008783e-05, + "loss": 0.9545, + "step": 15096 + }, + { + "epoch": 0.36, + "grad_norm": 2.212716002278755, + "learning_rate": 1.4934437480267075e-05, + "loss": 1.1362, + "step": 15097 + }, + { + "epoch": 0.36, + "grad_norm": 1.8466883120292543, + "learning_rate": 1.4933773789795508e-05, + "loss": 1.0679, + "step": 15098 + }, + { + "epoch": 0.36, + "grad_norm": 1.819319933758456, + "learning_rate": 1.4933110070597945e-05, + "loss": 1.0255, + "step": 15099 + }, + { + "epoch": 0.36, + "grad_norm": 2.0090148558493337, + "learning_rate": 1.4932446322678255e-05, + "loss": 1.0032, + "step": 15100 + }, + { + "epoch": 0.36, + "grad_norm": 1.9855464537312586, + "learning_rate": 1.4931782546040302e-05, + "loss": 1.1176, + "step": 15101 + }, + { + "epoch": 0.36, + "grad_norm": 2.2553193717587305, + "learning_rate": 1.4931118740687952e-05, + "loss": 1.0325, + "step": 15102 + }, + { + "epoch": 0.36, + "grad_norm": 1.8369480096457578, + "learning_rate": 1.4930454906625062e-05, + "loss": 1.03, + "step": 15103 + }, + { + "epoch": 0.36, + "grad_norm": 2.5234378905672474, + "learning_rate": 1.4929791043855506e-05, + "loss": 1.1511, + "step": 15104 + }, + { + "epoch": 0.36, + "grad_norm": 1.9295805323375284, + "learning_rate": 1.4929127152383148e-05, + "loss": 1.1225, + "step": 15105 + }, + { + "epoch": 0.36, + "grad_norm": 1.9638277477982873, + "learning_rate": 1.4928463232211848e-05, + "loss": 1.073, + "step": 15106 + }, + { + "epoch": 0.36, + "grad_norm": 2.0087804970305982, + "learning_rate": 1.4927799283345474e-05, + "loss": 1.0149, + "step": 15107 + }, + { + "epoch": 0.36, + "grad_norm": 2.1545763245193754, + "learning_rate": 1.4927135305787895e-05, + "loss": 1.0345, + "step": 15108 + }, + { + "epoch": 0.36, + "grad_norm": 2.082898500246154, + "learning_rate": 1.492647129954297e-05, + "loss": 1.064, + "step": 15109 + }, + { + "epoch": 0.36, + "grad_norm": 2.069242555645467, + "learning_rate": 1.4925807264614573e-05, + "loss": 1.0281, + "step": 15110 + }, + { + "epoch": 0.36, + "grad_norm": 2.094539375989284, + "learning_rate": 1.4925143201006566e-05, + "loss": 0.998, + "step": 15111 + }, + { + "epoch": 0.36, + "grad_norm": 1.8600618531082445, + "learning_rate": 1.4924479108722816e-05, + "loss": 1.1305, + "step": 15112 + }, + { + "epoch": 0.36, + "grad_norm": 2.364020682153823, + "learning_rate": 1.4923814987767191e-05, + "loss": 1.09, + "step": 15113 + }, + { + "epoch": 0.36, + "grad_norm": 2.26327776152283, + "learning_rate": 1.4923150838143555e-05, + "loss": 1.0388, + "step": 15114 + }, + { + "epoch": 0.36, + "grad_norm": 2.2450068033103086, + "learning_rate": 1.4922486659855774e-05, + "loss": 1.1892, + "step": 15115 + }, + { + "epoch": 0.36, + "grad_norm": 3.1569241704294804, + "learning_rate": 1.492182245290772e-05, + "loss": 0.9495, + "step": 15116 + }, + { + "epoch": 0.36, + "grad_norm": 1.8462473057010196, + "learning_rate": 1.4921158217303257e-05, + "loss": 0.9929, + "step": 15117 + }, + { + "epoch": 0.36, + "grad_norm": 2.1920372650282305, + "learning_rate": 1.4920493953046254e-05, + "loss": 1.0341, + "step": 15118 + }, + { + "epoch": 0.36, + "grad_norm": 1.8876584258992362, + "learning_rate": 1.4919829660140577e-05, + "loss": 1.0017, + "step": 15119 + }, + { + "epoch": 0.36, + "grad_norm": 1.836087687162765, + "learning_rate": 1.491916533859009e-05, + "loss": 1.051, + "step": 15120 + }, + { + "epoch": 0.36, + "grad_norm": 2.1558635082816955, + "learning_rate": 1.491850098839867e-05, + "loss": 1.0179, + "step": 15121 + }, + { + "epoch": 0.36, + "grad_norm": 2.0130878382228126, + "learning_rate": 1.4917836609570176e-05, + "loss": 0.9966, + "step": 15122 + }, + { + "epoch": 0.36, + "grad_norm": 1.1862785537978044, + "learning_rate": 1.4917172202108485e-05, + "loss": 1.0224, + "step": 15123 + }, + { + "epoch": 0.36, + "grad_norm": 1.9626552994662163, + "learning_rate": 1.4916507766017455e-05, + "loss": 0.9612, + "step": 15124 + }, + { + "epoch": 0.36, + "grad_norm": 2.4546119577205814, + "learning_rate": 1.4915843301300968e-05, + "loss": 1.0019, + "step": 15125 + }, + { + "epoch": 0.36, + "grad_norm": 2.0928241387160815, + "learning_rate": 1.491517880796288e-05, + "loss": 0.9482, + "step": 15126 + }, + { + "epoch": 0.36, + "grad_norm": 3.881304688509178, + "learning_rate": 1.4914514286007064e-05, + "loss": 0.9905, + "step": 15127 + }, + { + "epoch": 0.36, + "grad_norm": 2.6175906616315476, + "learning_rate": 1.4913849735437392e-05, + "loss": 1.0221, + "step": 15128 + }, + { + "epoch": 0.36, + "grad_norm": 2.2513640795264553, + "learning_rate": 1.491318515625773e-05, + "loss": 1.0675, + "step": 15129 + }, + { + "epoch": 0.36, + "grad_norm": 1.961116728022405, + "learning_rate": 1.491252054847195e-05, + "loss": 1.023, + "step": 15130 + }, + { + "epoch": 0.36, + "grad_norm": 2.0408037065384748, + "learning_rate": 1.4911855912083916e-05, + "loss": 1.0048, + "step": 15131 + }, + { + "epoch": 0.36, + "grad_norm": 1.081838268240329, + "learning_rate": 1.4911191247097504e-05, + "loss": 0.9654, + "step": 15132 + }, + { + "epoch": 0.36, + "grad_norm": 1.924311741806568, + "learning_rate": 1.4910526553516585e-05, + "loss": 1.0784, + "step": 15133 + }, + { + "epoch": 0.36, + "grad_norm": 1.8993597467839873, + "learning_rate": 1.4909861831345022e-05, + "loss": 0.9847, + "step": 15134 + }, + { + "epoch": 0.36, + "grad_norm": 2.166369818973685, + "learning_rate": 1.490919708058669e-05, + "loss": 1.0054, + "step": 15135 + }, + { + "epoch": 0.36, + "grad_norm": 2.0999152511863834, + "learning_rate": 1.4908532301245459e-05, + "loss": 1.074, + "step": 15136 + }, + { + "epoch": 0.36, + "grad_norm": 2.0955763075316973, + "learning_rate": 1.4907867493325199e-05, + "loss": 1.0199, + "step": 15137 + }, + { + "epoch": 0.36, + "grad_norm": 4.386870029287577, + "learning_rate": 1.490720265682978e-05, + "loss": 1.0283, + "step": 15138 + }, + { + "epoch": 0.36, + "grad_norm": 2.854240108143194, + "learning_rate": 1.4906537791763073e-05, + "loss": 1.0258, + "step": 15139 + }, + { + "epoch": 0.36, + "grad_norm": 2.8027320470534436, + "learning_rate": 1.490587289812895e-05, + "loss": 0.9581, + "step": 15140 + }, + { + "epoch": 0.36, + "grad_norm": 1.1628976173185854, + "learning_rate": 1.4905207975931282e-05, + "loss": 1.049, + "step": 15141 + }, + { + "epoch": 0.36, + "grad_norm": 2.225973748179959, + "learning_rate": 1.4904543025173943e-05, + "loss": 1.0647, + "step": 15142 + }, + { + "epoch": 0.36, + "grad_norm": 1.8612124726350006, + "learning_rate": 1.4903878045860798e-05, + "loss": 1.0059, + "step": 15143 + }, + { + "epoch": 0.36, + "grad_norm": 2.020266515378305, + "learning_rate": 1.4903213037995725e-05, + "loss": 1.1191, + "step": 15144 + }, + { + "epoch": 0.36, + "grad_norm": 1.8834157006913408, + "learning_rate": 1.4902548001582591e-05, + "loss": 1.0341, + "step": 15145 + }, + { + "epoch": 0.36, + "grad_norm": 2.0196932167638617, + "learning_rate": 1.4901882936625272e-05, + "loss": 1.0919, + "step": 15146 + }, + { + "epoch": 0.36, + "grad_norm": 2.0789917982511206, + "learning_rate": 1.4901217843127636e-05, + "loss": 1.0538, + "step": 15147 + }, + { + "epoch": 0.36, + "grad_norm": 2.376223857018817, + "learning_rate": 1.4900552721093559e-05, + "loss": 0.9621, + "step": 15148 + }, + { + "epoch": 0.36, + "grad_norm": 1.8455055158784313, + "learning_rate": 1.4899887570526916e-05, + "loss": 0.961, + "step": 15149 + }, + { + "epoch": 0.36, + "grad_norm": 1.8501894363800084, + "learning_rate": 1.4899222391431572e-05, + "loss": 0.927, + "step": 15150 + }, + { + "epoch": 0.36, + "grad_norm": 1.9476227581083982, + "learning_rate": 1.4898557183811409e-05, + "loss": 1.0079, + "step": 15151 + }, + { + "epoch": 0.36, + "grad_norm": 1.9736741356343523, + "learning_rate": 1.489789194767029e-05, + "loss": 1.0893, + "step": 15152 + }, + { + "epoch": 0.36, + "grad_norm": 2.273492528861438, + "learning_rate": 1.4897226683012096e-05, + "loss": 0.9404, + "step": 15153 + }, + { + "epoch": 0.36, + "grad_norm": 1.9398060547874114, + "learning_rate": 1.4896561389840696e-05, + "loss": 0.8731, + "step": 15154 + }, + { + "epoch": 0.36, + "grad_norm": 2.040225242479822, + "learning_rate": 1.4895896068159967e-05, + "loss": 1.1121, + "step": 15155 + }, + { + "epoch": 0.36, + "grad_norm": 1.8146762132032939, + "learning_rate": 1.4895230717973783e-05, + "loss": 1.0767, + "step": 15156 + }, + { + "epoch": 0.36, + "grad_norm": 1.8091666030678324, + "learning_rate": 1.4894565339286014e-05, + "loss": 0.9388, + "step": 15157 + }, + { + "epoch": 0.36, + "grad_norm": 1.9645286685673604, + "learning_rate": 1.4893899932100534e-05, + "loss": 1.1391, + "step": 15158 + }, + { + "epoch": 0.36, + "grad_norm": 1.0764609348245526, + "learning_rate": 1.4893234496421222e-05, + "loss": 0.9842, + "step": 15159 + }, + { + "epoch": 0.36, + "grad_norm": 1.9854607504571273, + "learning_rate": 1.4892569032251949e-05, + "loss": 1.113, + "step": 15160 + }, + { + "epoch": 0.36, + "grad_norm": 2.6467672029737366, + "learning_rate": 1.4891903539596588e-05, + "loss": 1.0811, + "step": 15161 + }, + { + "epoch": 0.36, + "grad_norm": 2.658591293870636, + "learning_rate": 1.489123801845902e-05, + "loss": 1.1441, + "step": 15162 + }, + { + "epoch": 0.36, + "grad_norm": 1.1523714524915585, + "learning_rate": 1.4890572468843112e-05, + "loss": 0.9136, + "step": 15163 + }, + { + "epoch": 0.36, + "grad_norm": 1.9776626399363948, + "learning_rate": 1.4889906890752745e-05, + "loss": 0.8588, + "step": 15164 + }, + { + "epoch": 0.36, + "grad_norm": 1.9842186651729996, + "learning_rate": 1.488924128419179e-05, + "loss": 1.1251, + "step": 15165 + }, + { + "epoch": 0.36, + "grad_norm": 1.9192447050240695, + "learning_rate": 1.4888575649164125e-05, + "loss": 0.9733, + "step": 15166 + }, + { + "epoch": 0.36, + "grad_norm": 2.0375778377592066, + "learning_rate": 1.4887909985673626e-05, + "loss": 1.0596, + "step": 15167 + }, + { + "epoch": 0.36, + "grad_norm": 1.1108357565668392, + "learning_rate": 1.4887244293724167e-05, + "loss": 0.992, + "step": 15168 + }, + { + "epoch": 0.36, + "grad_norm": 3.085327426521749, + "learning_rate": 1.4886578573319623e-05, + "loss": 1.0327, + "step": 15169 + }, + { + "epoch": 0.36, + "grad_norm": 2.1729844041999575, + "learning_rate": 1.4885912824463875e-05, + "loss": 1.0076, + "step": 15170 + }, + { + "epoch": 0.36, + "grad_norm": 2.1421669768312697, + "learning_rate": 1.4885247047160792e-05, + "loss": 1.0482, + "step": 15171 + }, + { + "epoch": 0.36, + "grad_norm": 1.9545133307940992, + "learning_rate": 1.4884581241414257e-05, + "loss": 1.0824, + "step": 15172 + }, + { + "epoch": 0.36, + "grad_norm": 2.671218779808595, + "learning_rate": 1.488391540722814e-05, + "loss": 1.0243, + "step": 15173 + }, + { + "epoch": 0.36, + "grad_norm": 2.056815147919692, + "learning_rate": 1.4883249544606326e-05, + "loss": 1.0621, + "step": 15174 + }, + { + "epoch": 0.36, + "grad_norm": 1.9157690351671566, + "learning_rate": 1.4882583653552683e-05, + "loss": 1.0862, + "step": 15175 + }, + { + "epoch": 0.36, + "grad_norm": 1.104345863318829, + "learning_rate": 1.4881917734071094e-05, + "loss": 0.9053, + "step": 15176 + }, + { + "epoch": 0.36, + "grad_norm": 1.8878430459430047, + "learning_rate": 1.4881251786165436e-05, + "loss": 1.0235, + "step": 15177 + }, + { + "epoch": 0.36, + "grad_norm": 1.92072063446404, + "learning_rate": 1.488058580983958e-05, + "loss": 0.9191, + "step": 15178 + }, + { + "epoch": 0.36, + "grad_norm": 1.917283050857229, + "learning_rate": 1.4879919805097412e-05, + "loss": 0.987, + "step": 15179 + }, + { + "epoch": 0.36, + "grad_norm": 3.3081886846907578, + "learning_rate": 1.4879253771942805e-05, + "loss": 1.0301, + "step": 15180 + }, + { + "epoch": 0.36, + "grad_norm": 1.9200329003873398, + "learning_rate": 1.4878587710379639e-05, + "loss": 1.1284, + "step": 15181 + }, + { + "epoch": 0.36, + "grad_norm": 2.1052223990458736, + "learning_rate": 1.4877921620411789e-05, + "loss": 0.9607, + "step": 15182 + }, + { + "epoch": 0.36, + "grad_norm": 1.9227695108540261, + "learning_rate": 1.4877255502043136e-05, + "loss": 1.06, + "step": 15183 + }, + { + "epoch": 0.36, + "grad_norm": 2.010448405400493, + "learning_rate": 1.4876589355277555e-05, + "loss": 1.1199, + "step": 15184 + }, + { + "epoch": 0.36, + "grad_norm": 2.090882455418119, + "learning_rate": 1.487592318011893e-05, + "loss": 1.0575, + "step": 15185 + }, + { + "epoch": 0.36, + "grad_norm": 2.0094988796986666, + "learning_rate": 1.4875256976571135e-05, + "loss": 1.0913, + "step": 15186 + }, + { + "epoch": 0.36, + "grad_norm": 2.4145935037457074, + "learning_rate": 1.487459074463805e-05, + "loss": 0.9871, + "step": 15187 + }, + { + "epoch": 0.36, + "grad_norm": 1.8176880545686749, + "learning_rate": 1.4873924484323557e-05, + "loss": 0.9658, + "step": 15188 + }, + { + "epoch": 0.36, + "grad_norm": 1.1374481113725907, + "learning_rate": 1.487325819563153e-05, + "loss": 1.0153, + "step": 15189 + }, + { + "epoch": 0.36, + "grad_norm": 1.9856004529638454, + "learning_rate": 1.4872591878565852e-05, + "loss": 1.0372, + "step": 15190 + }, + { + "epoch": 0.36, + "grad_norm": 2.1361238274250596, + "learning_rate": 1.4871925533130402e-05, + "loss": 0.9814, + "step": 15191 + }, + { + "epoch": 0.36, + "grad_norm": 3.3969733259261528, + "learning_rate": 1.4871259159329056e-05, + "loss": 0.9125, + "step": 15192 + }, + { + "epoch": 0.36, + "grad_norm": 2.1697422049988093, + "learning_rate": 1.48705927571657e-05, + "loss": 0.9981, + "step": 15193 + }, + { + "epoch": 0.36, + "grad_norm": 2.6000678951727214, + "learning_rate": 1.4869926326644208e-05, + "loss": 0.9294, + "step": 15194 + }, + { + "epoch": 0.36, + "grad_norm": 1.098081983868769, + "learning_rate": 1.486925986776847e-05, + "loss": 0.9725, + "step": 15195 + }, + { + "epoch": 0.36, + "grad_norm": 2.096048992156674, + "learning_rate": 1.4868593380542351e-05, + "loss": 1.0271, + "step": 15196 + }, + { + "epoch": 0.36, + "grad_norm": 1.1477865254322261, + "learning_rate": 1.4867926864969744e-05, + "loss": 0.9728, + "step": 15197 + }, + { + "epoch": 0.36, + "grad_norm": 1.0552856362972332, + "learning_rate": 1.4867260321054526e-05, + "loss": 0.9547, + "step": 15198 + }, + { + "epoch": 0.36, + "grad_norm": 1.9227516479891609, + "learning_rate": 1.4866593748800578e-05, + "loss": 1.0664, + "step": 15199 + }, + { + "epoch": 0.36, + "grad_norm": 1.8412850145010238, + "learning_rate": 1.4865927148211779e-05, + "loss": 1.0273, + "step": 15200 + }, + { + "epoch": 0.36, + "grad_norm": 2.0441281841276004, + "learning_rate": 1.486526051929201e-05, + "loss": 1.0602, + "step": 15201 + }, + { + "epoch": 0.36, + "grad_norm": 1.805619220037566, + "learning_rate": 1.4864593862045159e-05, + "loss": 0.9238, + "step": 15202 + }, + { + "epoch": 0.36, + "grad_norm": 1.8537509005020723, + "learning_rate": 1.4863927176475097e-05, + "loss": 0.9246, + "step": 15203 + }, + { + "epoch": 0.36, + "grad_norm": 1.9358412362881572, + "learning_rate": 1.4863260462585715e-05, + "loss": 1.1189, + "step": 15204 + }, + { + "epoch": 0.36, + "grad_norm": 2.2599674586415546, + "learning_rate": 1.4862593720380886e-05, + "loss": 1.0749, + "step": 15205 + }, + { + "epoch": 0.36, + "grad_norm": 1.7348982174402443, + "learning_rate": 1.4861926949864502e-05, + "loss": 1.0266, + "step": 15206 + }, + { + "epoch": 0.36, + "grad_norm": 2.0488525094616747, + "learning_rate": 1.486126015104044e-05, + "loss": 0.9274, + "step": 15207 + }, + { + "epoch": 0.36, + "grad_norm": 1.9540001401074234, + "learning_rate": 1.486059332391258e-05, + "loss": 1.1693, + "step": 15208 + }, + { + "epoch": 0.36, + "grad_norm": 2.2810530542821335, + "learning_rate": 1.4859926468484807e-05, + "loss": 1.1425, + "step": 15209 + }, + { + "epoch": 0.36, + "grad_norm": 2.0300140250681564, + "learning_rate": 1.4859259584761002e-05, + "loss": 0.9033, + "step": 15210 + }, + { + "epoch": 0.36, + "grad_norm": 1.907003323722846, + "learning_rate": 1.4858592672745053e-05, + "loss": 1.1282, + "step": 15211 + }, + { + "epoch": 0.36, + "grad_norm": 2.1720336938202762, + "learning_rate": 1.4857925732440839e-05, + "loss": 1.0221, + "step": 15212 + }, + { + "epoch": 0.36, + "grad_norm": 1.8839045329025599, + "learning_rate": 1.4857258763852241e-05, + "loss": 1.1756, + "step": 15213 + }, + { + "epoch": 0.36, + "grad_norm": 1.9739397993209502, + "learning_rate": 1.4856591766983145e-05, + "loss": 1.0734, + "step": 15214 + }, + { + "epoch": 0.36, + "grad_norm": 2.200747149290112, + "learning_rate": 1.4855924741837435e-05, + "loss": 1.0046, + "step": 15215 + }, + { + "epoch": 0.36, + "grad_norm": 1.8457500482074416, + "learning_rate": 1.4855257688418991e-05, + "loss": 1.0223, + "step": 15216 + }, + { + "epoch": 0.36, + "grad_norm": 2.0936597384405813, + "learning_rate": 1.4854590606731702e-05, + "loss": 1.1072, + "step": 15217 + }, + { + "epoch": 0.36, + "grad_norm": 2.0405716711529895, + "learning_rate": 1.485392349677945e-05, + "loss": 1.0246, + "step": 15218 + }, + { + "epoch": 0.36, + "grad_norm": 2.0896918917902227, + "learning_rate": 1.485325635856612e-05, + "loss": 1.0266, + "step": 15219 + }, + { + "epoch": 0.36, + "grad_norm": 1.9300783340469043, + "learning_rate": 1.4852589192095592e-05, + "loss": 0.9616, + "step": 15220 + }, + { + "epoch": 0.36, + "grad_norm": 2.0723586031263834, + "learning_rate": 1.4851921997371756e-05, + "loss": 1.0345, + "step": 15221 + }, + { + "epoch": 0.36, + "grad_norm": 2.150853419396451, + "learning_rate": 1.4851254774398492e-05, + "loss": 1.0968, + "step": 15222 + }, + { + "epoch": 0.36, + "grad_norm": 2.0065199138507674, + "learning_rate": 1.4850587523179689e-05, + "loss": 1.0465, + "step": 15223 + }, + { + "epoch": 0.36, + "grad_norm": 1.8477709873318822, + "learning_rate": 1.4849920243719227e-05, + "loss": 1.0741, + "step": 15224 + }, + { + "epoch": 0.36, + "grad_norm": 4.0482838092160955, + "learning_rate": 1.4849252936020998e-05, + "loss": 1.071, + "step": 15225 + }, + { + "epoch": 0.36, + "grad_norm": 2.047208704550184, + "learning_rate": 1.4848585600088876e-05, + "loss": 0.9899, + "step": 15226 + }, + { + "epoch": 0.36, + "grad_norm": 3.628066978064567, + "learning_rate": 1.484791823592676e-05, + "loss": 1.023, + "step": 15227 + }, + { + "epoch": 0.36, + "grad_norm": 2.1443156666435566, + "learning_rate": 1.4847250843538527e-05, + "loss": 0.9349, + "step": 15228 + }, + { + "epoch": 0.36, + "grad_norm": 1.9858746200810609, + "learning_rate": 1.4846583422928067e-05, + "loss": 1.0682, + "step": 15229 + }, + { + "epoch": 0.36, + "grad_norm": 2.425355876742655, + "learning_rate": 1.484591597409926e-05, + "loss": 0.9885, + "step": 15230 + }, + { + "epoch": 0.36, + "grad_norm": 1.1122716125530892, + "learning_rate": 1.4845248497055997e-05, + "loss": 1.056, + "step": 15231 + }, + { + "epoch": 0.36, + "grad_norm": 2.1254813635398273, + "learning_rate": 1.4844580991802166e-05, + "loss": 0.985, + "step": 15232 + }, + { + "epoch": 0.36, + "grad_norm": 2.1120106898226054, + "learning_rate": 1.4843913458341646e-05, + "loss": 1.0988, + "step": 15233 + }, + { + "epoch": 0.36, + "grad_norm": 1.0985024881418433, + "learning_rate": 1.4843245896678333e-05, + "loss": 0.932, + "step": 15234 + }, + { + "epoch": 0.36, + "grad_norm": 1.9065277850396367, + "learning_rate": 1.4842578306816108e-05, + "loss": 1.0346, + "step": 15235 + }, + { + "epoch": 0.36, + "grad_norm": 1.9758788965614833, + "learning_rate": 1.4841910688758858e-05, + "loss": 1.0436, + "step": 15236 + }, + { + "epoch": 0.36, + "grad_norm": 1.7460316831269862, + "learning_rate": 1.4841243042510472e-05, + "loss": 1.0757, + "step": 15237 + }, + { + "epoch": 0.36, + "grad_norm": 2.1574837126409245, + "learning_rate": 1.4840575368074835e-05, + "loss": 0.9912, + "step": 15238 + }, + { + "epoch": 0.36, + "grad_norm": 1.9980168022202696, + "learning_rate": 1.4839907665455837e-05, + "loss": 1.1596, + "step": 15239 + }, + { + "epoch": 0.36, + "grad_norm": 2.007470301330532, + "learning_rate": 1.4839239934657364e-05, + "loss": 0.9835, + "step": 15240 + }, + { + "epoch": 0.36, + "grad_norm": 1.8108934242613448, + "learning_rate": 1.4838572175683304e-05, + "loss": 0.9274, + "step": 15241 + }, + { + "epoch": 0.36, + "grad_norm": 1.8662354464196627, + "learning_rate": 1.4837904388537545e-05, + "loss": 0.999, + "step": 15242 + }, + { + "epoch": 0.36, + "grad_norm": 2.0007126149098426, + "learning_rate": 1.4837236573223973e-05, + "loss": 1.0643, + "step": 15243 + }, + { + "epoch": 0.36, + "grad_norm": 2.0976196407741083, + "learning_rate": 1.483656872974648e-05, + "loss": 0.932, + "step": 15244 + }, + { + "epoch": 0.36, + "grad_norm": 2.211345106596063, + "learning_rate": 1.4835900858108951e-05, + "loss": 1.1163, + "step": 15245 + }, + { + "epoch": 0.36, + "grad_norm": 1.8722624427683605, + "learning_rate": 1.4835232958315278e-05, + "loss": 1.0606, + "step": 15246 + }, + { + "epoch": 0.36, + "grad_norm": 2.0813701767312294, + "learning_rate": 1.4834565030369348e-05, + "loss": 1.0748, + "step": 15247 + }, + { + "epoch": 0.36, + "grad_norm": 1.882919885216426, + "learning_rate": 1.4833897074275049e-05, + "loss": 1.0478, + "step": 15248 + }, + { + "epoch": 0.36, + "grad_norm": 2.1813611228907344, + "learning_rate": 1.4833229090036273e-05, + "loss": 1.1734, + "step": 15249 + }, + { + "epoch": 0.36, + "grad_norm": 2.2263298654536086, + "learning_rate": 1.4832561077656905e-05, + "loss": 1.0214, + "step": 15250 + }, + { + "epoch": 0.36, + "grad_norm": 1.9454164976305124, + "learning_rate": 1.4831893037140836e-05, + "loss": 1.1202, + "step": 15251 + }, + { + "epoch": 0.36, + "grad_norm": 3.2526972035416257, + "learning_rate": 1.4831224968491955e-05, + "loss": 1.0139, + "step": 15252 + }, + { + "epoch": 0.36, + "grad_norm": 2.067879209085909, + "learning_rate": 1.4830556871714157e-05, + "loss": 1.0016, + "step": 15253 + }, + { + "epoch": 0.36, + "grad_norm": 2.1541811469068954, + "learning_rate": 1.4829888746811323e-05, + "loss": 0.9463, + "step": 15254 + }, + { + "epoch": 0.36, + "grad_norm": 2.4424171243245643, + "learning_rate": 1.482922059378735e-05, + "loss": 1.037, + "step": 15255 + }, + { + "epoch": 0.36, + "grad_norm": 2.010252532722239, + "learning_rate": 1.4828552412646126e-05, + "loss": 1.0654, + "step": 15256 + }, + { + "epoch": 0.36, + "grad_norm": 1.883179696757189, + "learning_rate": 1.482788420339154e-05, + "loss": 1.0236, + "step": 15257 + }, + { + "epoch": 0.36, + "grad_norm": 2.040239151614552, + "learning_rate": 1.4827215966027483e-05, + "loss": 1.0863, + "step": 15258 + }, + { + "epoch": 0.36, + "grad_norm": 2.18919277887099, + "learning_rate": 1.482654770055785e-05, + "loss": 1.0031, + "step": 15259 + }, + { + "epoch": 0.36, + "grad_norm": 1.9556321858119066, + "learning_rate": 1.4825879406986525e-05, + "loss": 1.0465, + "step": 15260 + }, + { + "epoch": 0.36, + "grad_norm": 1.7505658888006612, + "learning_rate": 1.4825211085317404e-05, + "loss": 1.0494, + "step": 15261 + }, + { + "epoch": 0.36, + "grad_norm": 2.1285333729030462, + "learning_rate": 1.4824542735554375e-05, + "loss": 1.04, + "step": 15262 + }, + { + "epoch": 0.36, + "grad_norm": 1.862016990820414, + "learning_rate": 1.482387435770133e-05, + "loss": 1.0574, + "step": 15263 + }, + { + "epoch": 0.36, + "grad_norm": 2.438571243296539, + "learning_rate": 1.4823205951762164e-05, + "loss": 1.0746, + "step": 15264 + }, + { + "epoch": 0.36, + "grad_norm": 1.754009630081236, + "learning_rate": 1.4822537517740763e-05, + "loss": 1.0473, + "step": 15265 + }, + { + "epoch": 0.36, + "grad_norm": 1.8349048162834265, + "learning_rate": 1.4821869055641023e-05, + "loss": 1.0554, + "step": 15266 + }, + { + "epoch": 0.36, + "grad_norm": 2.0352965991120806, + "learning_rate": 1.4821200565466835e-05, + "loss": 1.0832, + "step": 15267 + }, + { + "epoch": 0.36, + "grad_norm": 2.2003493403258103, + "learning_rate": 1.4820532047222091e-05, + "loss": 1.0925, + "step": 15268 + }, + { + "epoch": 0.36, + "grad_norm": 1.8040376423093412, + "learning_rate": 1.4819863500910683e-05, + "loss": 1.0422, + "step": 15269 + }, + { + "epoch": 0.36, + "grad_norm": 1.8304008501617086, + "learning_rate": 1.4819194926536504e-05, + "loss": 0.9632, + "step": 15270 + }, + { + "epoch": 0.36, + "grad_norm": 2.0695814020409804, + "learning_rate": 1.4818526324103445e-05, + "loss": 0.9529, + "step": 15271 + }, + { + "epoch": 0.36, + "grad_norm": 2.121684242152519, + "learning_rate": 1.4817857693615406e-05, + "loss": 1.1808, + "step": 15272 + }, + { + "epoch": 0.36, + "grad_norm": 1.9099827698733323, + "learning_rate": 1.4817189035076265e-05, + "loss": 0.9642, + "step": 15273 + }, + { + "epoch": 0.36, + "grad_norm": 1.1517307232945462, + "learning_rate": 1.4816520348489932e-05, + "loss": 0.9883, + "step": 15274 + }, + { + "epoch": 0.36, + "grad_norm": 2.093858547141449, + "learning_rate": 1.4815851633860288e-05, + "loss": 1.0274, + "step": 15275 + }, + { + "epoch": 0.36, + "grad_norm": 2.181780237769679, + "learning_rate": 1.4815182891191234e-05, + "loss": 0.9623, + "step": 15276 + }, + { + "epoch": 0.36, + "grad_norm": 1.9165884411142158, + "learning_rate": 1.4814514120486657e-05, + "loss": 1.1176, + "step": 15277 + }, + { + "epoch": 0.36, + "grad_norm": 2.0164973816273526, + "learning_rate": 1.481384532175046e-05, + "loss": 1.0175, + "step": 15278 + }, + { + "epoch": 0.36, + "grad_norm": 2.304003466861686, + "learning_rate": 1.481317649498653e-05, + "loss": 0.9825, + "step": 15279 + }, + { + "epoch": 0.36, + "grad_norm": 1.9451289839625783, + "learning_rate": 1.481250764019876e-05, + "loss": 1.1609, + "step": 15280 + }, + { + "epoch": 0.36, + "grad_norm": 2.4148211861985995, + "learning_rate": 1.4811838757391049e-05, + "loss": 1.147, + "step": 15281 + }, + { + "epoch": 0.36, + "grad_norm": 2.0110584094212407, + "learning_rate": 1.4811169846567286e-05, + "loss": 1.0499, + "step": 15282 + }, + { + "epoch": 0.36, + "grad_norm": 2.157702374621911, + "learning_rate": 1.4810500907731374e-05, + "loss": 0.8905, + "step": 15283 + }, + { + "epoch": 0.36, + "grad_norm": 2.0747698574379942, + "learning_rate": 1.4809831940887196e-05, + "loss": 1.0612, + "step": 15284 + }, + { + "epoch": 0.36, + "grad_norm": 1.884091693829867, + "learning_rate": 1.4809162946038659e-05, + "loss": 0.9189, + "step": 15285 + }, + { + "epoch": 0.36, + "grad_norm": 1.9778791185860551, + "learning_rate": 1.4808493923189652e-05, + "loss": 1.0553, + "step": 15286 + }, + { + "epoch": 0.36, + "grad_norm": 1.8857470491461212, + "learning_rate": 1.4807824872344072e-05, + "loss": 1.0316, + "step": 15287 + }, + { + "epoch": 0.36, + "grad_norm": 1.9907441349756188, + "learning_rate": 1.4807155793505811e-05, + "loss": 1.0462, + "step": 15288 + }, + { + "epoch": 0.36, + "grad_norm": 2.1718236535292568, + "learning_rate": 1.480648668667877e-05, + "loss": 0.926, + "step": 15289 + }, + { + "epoch": 0.36, + "grad_norm": 1.8416856367321484, + "learning_rate": 1.4805817551866839e-05, + "loss": 1.1119, + "step": 15290 + }, + { + "epoch": 0.36, + "grad_norm": 2.0023390272995876, + "learning_rate": 1.4805148389073918e-05, + "loss": 1.0397, + "step": 15291 + }, + { + "epoch": 0.36, + "grad_norm": 2.117902226662713, + "learning_rate": 1.48044791983039e-05, + "loss": 0.9748, + "step": 15292 + }, + { + "epoch": 0.36, + "grad_norm": 2.493918173439391, + "learning_rate": 1.4803809979560688e-05, + "loss": 1.0904, + "step": 15293 + }, + { + "epoch": 0.36, + "grad_norm": 2.0206684345546946, + "learning_rate": 1.480314073284817e-05, + "loss": 1.041, + "step": 15294 + }, + { + "epoch": 0.36, + "grad_norm": 2.209771063659527, + "learning_rate": 1.4802471458170245e-05, + "loss": 0.9819, + "step": 15295 + }, + { + "epoch": 0.36, + "grad_norm": 2.108792626710109, + "learning_rate": 1.480180215553081e-05, + "loss": 1.1275, + "step": 15296 + }, + { + "epoch": 0.36, + "grad_norm": 2.184159196061468, + "learning_rate": 1.4801132824933767e-05, + "loss": 0.9728, + "step": 15297 + }, + { + "epoch": 0.36, + "grad_norm": 1.9288583048522085, + "learning_rate": 1.4800463466383004e-05, + "loss": 1.0703, + "step": 15298 + }, + { + "epoch": 0.36, + "grad_norm": 2.1606252285800935, + "learning_rate": 1.4799794079882426e-05, + "loss": 1.0289, + "step": 15299 + }, + { + "epoch": 0.36, + "grad_norm": 2.3069531905657947, + "learning_rate": 1.4799124665435927e-05, + "loss": 1.0696, + "step": 15300 + }, + { + "epoch": 0.36, + "grad_norm": 2.038672067441107, + "learning_rate": 1.4798455223047401e-05, + "loss": 1.0222, + "step": 15301 + }, + { + "epoch": 0.36, + "grad_norm": 2.081850500518442, + "learning_rate": 1.4797785752720755e-05, + "loss": 1.0334, + "step": 15302 + }, + { + "epoch": 0.36, + "grad_norm": 2.208566988271624, + "learning_rate": 1.4797116254459878e-05, + "loss": 1.0246, + "step": 15303 + }, + { + "epoch": 0.36, + "grad_norm": 1.8991228576053094, + "learning_rate": 1.4796446728268674e-05, + "loss": 1.0129, + "step": 15304 + }, + { + "epoch": 0.36, + "grad_norm": 2.0677189942171372, + "learning_rate": 1.4795777174151033e-05, + "loss": 1.1217, + "step": 15305 + }, + { + "epoch": 0.36, + "grad_norm": 1.9939558220725657, + "learning_rate": 1.4795107592110866e-05, + "loss": 1.119, + "step": 15306 + }, + { + "epoch": 0.36, + "grad_norm": 2.152865017443331, + "learning_rate": 1.4794437982152061e-05, + "loss": 0.8499, + "step": 15307 + }, + { + "epoch": 0.36, + "grad_norm": 2.098998399042789, + "learning_rate": 1.4793768344278521e-05, + "loss": 1.1684, + "step": 15308 + }, + { + "epoch": 0.36, + "grad_norm": 2.1140693761731533, + "learning_rate": 1.4793098678494145e-05, + "loss": 1.0366, + "step": 15309 + }, + { + "epoch": 0.36, + "grad_norm": 2.320912329169842, + "learning_rate": 1.479242898480283e-05, + "loss": 1.0645, + "step": 15310 + }, + { + "epoch": 0.36, + "grad_norm": 2.1193686903128692, + "learning_rate": 1.4791759263208474e-05, + "loss": 1.1072, + "step": 15311 + }, + { + "epoch": 0.36, + "grad_norm": 4.148823050218942, + "learning_rate": 1.479108951371498e-05, + "loss": 1.042, + "step": 15312 + }, + { + "epoch": 0.36, + "grad_norm": 2.0094365011706428, + "learning_rate": 1.4790419736326248e-05, + "loss": 1.1108, + "step": 15313 + }, + { + "epoch": 0.36, + "grad_norm": 1.1136883735772662, + "learning_rate": 1.4789749931046173e-05, + "loss": 0.9669, + "step": 15314 + }, + { + "epoch": 0.36, + "grad_norm": 2.323840979721545, + "learning_rate": 1.4789080097878659e-05, + "loss": 0.9995, + "step": 15315 + }, + { + "epoch": 0.36, + "grad_norm": 2.4016682925482216, + "learning_rate": 1.4788410236827605e-05, + "loss": 0.9927, + "step": 15316 + }, + { + "epoch": 0.36, + "grad_norm": 1.7739310712076573, + "learning_rate": 1.4787740347896908e-05, + "loss": 1.0335, + "step": 15317 + }, + { + "epoch": 0.36, + "grad_norm": 1.875248195940303, + "learning_rate": 1.4787070431090474e-05, + "loss": 0.9947, + "step": 15318 + }, + { + "epoch": 0.36, + "grad_norm": 2.081244858016841, + "learning_rate": 1.4786400486412198e-05, + "loss": 0.8951, + "step": 15319 + }, + { + "epoch": 0.36, + "grad_norm": 1.95706815753806, + "learning_rate": 1.4785730513865983e-05, + "loss": 0.993, + "step": 15320 + }, + { + "epoch": 0.36, + "grad_norm": 2.22203612455094, + "learning_rate": 1.4785060513455732e-05, + "loss": 0.929, + "step": 15321 + }, + { + "epoch": 0.36, + "grad_norm": 2.7060917678310665, + "learning_rate": 1.4784390485185342e-05, + "loss": 1.1019, + "step": 15322 + }, + { + "epoch": 0.36, + "grad_norm": 2.0788849367929307, + "learning_rate": 1.4783720429058719e-05, + "loss": 0.9961, + "step": 15323 + }, + { + "epoch": 0.36, + "grad_norm": 2.490452763200198, + "learning_rate": 1.4783050345079755e-05, + "loss": 1.1102, + "step": 15324 + }, + { + "epoch": 0.36, + "grad_norm": 2.389356470998469, + "learning_rate": 1.4782380233252365e-05, + "loss": 0.8849, + "step": 15325 + }, + { + "epoch": 0.36, + "grad_norm": 1.9257846104658338, + "learning_rate": 1.4781710093580437e-05, + "loss": 0.9663, + "step": 15326 + }, + { + "epoch": 0.36, + "grad_norm": 2.0649213344539277, + "learning_rate": 1.4781039926067884e-05, + "loss": 1.1222, + "step": 15327 + }, + { + "epoch": 0.36, + "grad_norm": 2.013770988552254, + "learning_rate": 1.4780369730718595e-05, + "loss": 1.0542, + "step": 15328 + }, + { + "epoch": 0.36, + "grad_norm": 1.9624590701788314, + "learning_rate": 1.477969950753649e-05, + "loss": 0.9714, + "step": 15329 + }, + { + "epoch": 0.36, + "grad_norm": 2.13104557024565, + "learning_rate": 1.4779029256525457e-05, + "loss": 0.9483, + "step": 15330 + }, + { + "epoch": 0.36, + "grad_norm": 2.01473457306691, + "learning_rate": 1.47783589776894e-05, + "loss": 0.9605, + "step": 15331 + }, + { + "epoch": 0.36, + "grad_norm": 1.959805475412062, + "learning_rate": 1.4777688671032226e-05, + "loss": 1.0757, + "step": 15332 + }, + { + "epoch": 0.36, + "grad_norm": 2.0870906957466584, + "learning_rate": 1.4777018336557835e-05, + "loss": 0.9899, + "step": 15333 + }, + { + "epoch": 0.36, + "grad_norm": 2.271853026001936, + "learning_rate": 1.4776347974270135e-05, + "loss": 1.0242, + "step": 15334 + }, + { + "epoch": 0.36, + "grad_norm": 2.0935804533389395, + "learning_rate": 1.477567758417302e-05, + "loss": 1.0019, + "step": 15335 + }, + { + "epoch": 0.36, + "grad_norm": 2.0241866629251817, + "learning_rate": 1.4775007166270402e-05, + "loss": 1.0864, + "step": 15336 + }, + { + "epoch": 0.36, + "grad_norm": 1.9395509848343186, + "learning_rate": 1.4774336720566178e-05, + "loss": 0.9992, + "step": 15337 + }, + { + "epoch": 0.36, + "grad_norm": 2.06483167644064, + "learning_rate": 1.4773666247064255e-05, + "loss": 1.043, + "step": 15338 + }, + { + "epoch": 0.36, + "grad_norm": 1.0828454365226223, + "learning_rate": 1.4772995745768538e-05, + "loss": 1.0118, + "step": 15339 + }, + { + "epoch": 0.36, + "grad_norm": 6.256749964181881, + "learning_rate": 1.4772325216682926e-05, + "loss": 1.0281, + "step": 15340 + }, + { + "epoch": 0.36, + "grad_norm": 2.216997344031448, + "learning_rate": 1.4771654659811326e-05, + "loss": 1.106, + "step": 15341 + }, + { + "epoch": 0.36, + "grad_norm": 2.4366169234059685, + "learning_rate": 1.4770984075157641e-05, + "loss": 1.0661, + "step": 15342 + }, + { + "epoch": 0.36, + "grad_norm": 2.0288585085185518, + "learning_rate": 1.4770313462725777e-05, + "loss": 1.0159, + "step": 15343 + }, + { + "epoch": 0.36, + "grad_norm": 2.0129091509220602, + "learning_rate": 1.476964282251964e-05, + "loss": 1.0558, + "step": 15344 + }, + { + "epoch": 0.36, + "grad_norm": 2.191396968974045, + "learning_rate": 1.4768972154543129e-05, + "loss": 1.0136, + "step": 15345 + }, + { + "epoch": 0.36, + "grad_norm": 2.071158767650591, + "learning_rate": 1.4768301458800154e-05, + "loss": 0.9565, + "step": 15346 + }, + { + "epoch": 0.36, + "grad_norm": 1.8494197211721346, + "learning_rate": 1.4767630735294618e-05, + "loss": 0.8706, + "step": 15347 + }, + { + "epoch": 0.36, + "grad_norm": 2.0992723558209296, + "learning_rate": 1.4766959984030425e-05, + "loss": 1.0135, + "step": 15348 + }, + { + "epoch": 0.36, + "grad_norm": 2.3328176929321223, + "learning_rate": 1.4766289205011483e-05, + "loss": 0.9574, + "step": 15349 + }, + { + "epoch": 0.36, + "grad_norm": 2.0279699072477055, + "learning_rate": 1.4765618398241697e-05, + "loss": 1.0643, + "step": 15350 + }, + { + "epoch": 0.36, + "grad_norm": 2.25644078776567, + "learning_rate": 1.4764947563724974e-05, + "loss": 1.0275, + "step": 15351 + }, + { + "epoch": 0.36, + "grad_norm": 1.940598575175825, + "learning_rate": 1.4764276701465212e-05, + "loss": 1.065, + "step": 15352 + }, + { + "epoch": 0.36, + "grad_norm": 1.9787174443891113, + "learning_rate": 1.4763605811466328e-05, + "loss": 1.0397, + "step": 15353 + }, + { + "epoch": 0.36, + "grad_norm": 2.212930365053673, + "learning_rate": 1.4762934893732218e-05, + "loss": 0.8244, + "step": 15354 + }, + { + "epoch": 0.36, + "grad_norm": 2.1877280060584288, + "learning_rate": 1.47622639482668e-05, + "loss": 0.9281, + "step": 15355 + }, + { + "epoch": 0.36, + "grad_norm": 1.22147596651885, + "learning_rate": 1.4761592975073965e-05, + "loss": 0.9606, + "step": 15356 + }, + { + "epoch": 0.36, + "grad_norm": 1.2594927656770856, + "learning_rate": 1.4760921974157636e-05, + "loss": 0.9462, + "step": 15357 + }, + { + "epoch": 0.36, + "grad_norm": 2.300426354181807, + "learning_rate": 1.4760250945521708e-05, + "loss": 0.9617, + "step": 15358 + }, + { + "epoch": 0.36, + "grad_norm": 1.9914744460840754, + "learning_rate": 1.4759579889170093e-05, + "loss": 1.0995, + "step": 15359 + }, + { + "epoch": 0.36, + "grad_norm": 1.8710542909357804, + "learning_rate": 1.4758908805106696e-05, + "loss": 1.0383, + "step": 15360 + }, + { + "epoch": 0.36, + "grad_norm": 1.9309688278368058, + "learning_rate": 1.4758237693335424e-05, + "loss": 1.182, + "step": 15361 + }, + { + "epoch": 0.36, + "grad_norm": 2.470985209529213, + "learning_rate": 1.475756655386019e-05, + "loss": 1.0655, + "step": 15362 + }, + { + "epoch": 0.36, + "grad_norm": 1.897071557676397, + "learning_rate": 1.4756895386684893e-05, + "loss": 0.9772, + "step": 15363 + }, + { + "epoch": 0.36, + "grad_norm": 1.0458880258357173, + "learning_rate": 1.4756224191813449e-05, + "loss": 0.9391, + "step": 15364 + }, + { + "epoch": 0.36, + "grad_norm": 1.7978388990746739, + "learning_rate": 1.4755552969249758e-05, + "loss": 1.1641, + "step": 15365 + }, + { + "epoch": 0.36, + "grad_norm": 1.9344840184548737, + "learning_rate": 1.4754881718997735e-05, + "loss": 1.0786, + "step": 15366 + }, + { + "epoch": 0.36, + "grad_norm": 2.0573189972078656, + "learning_rate": 1.4754210441061283e-05, + "loss": 0.9658, + "step": 15367 + }, + { + "epoch": 0.36, + "grad_norm": 2.2154652262460974, + "learning_rate": 1.4753539135444314e-05, + "loss": 1.0457, + "step": 15368 + }, + { + "epoch": 0.36, + "grad_norm": 2.070499617514771, + "learning_rate": 1.4752867802150735e-05, + "loss": 1.0893, + "step": 15369 + }, + { + "epoch": 0.36, + "grad_norm": 2.0709991300494703, + "learning_rate": 1.4752196441184454e-05, + "loss": 0.992, + "step": 15370 + }, + { + "epoch": 0.36, + "grad_norm": 2.1129114331798626, + "learning_rate": 1.4751525052549383e-05, + "loss": 1.0209, + "step": 15371 + }, + { + "epoch": 0.36, + "grad_norm": 3.0952508382393833, + "learning_rate": 1.475085363624943e-05, + "loss": 0.9347, + "step": 15372 + }, + { + "epoch": 0.36, + "grad_norm": 2.170209617490011, + "learning_rate": 1.4750182192288497e-05, + "loss": 1.046, + "step": 15373 + }, + { + "epoch": 0.36, + "grad_norm": 2.2084219818069664, + "learning_rate": 1.4749510720670506e-05, + "loss": 1.0566, + "step": 15374 + }, + { + "epoch": 0.36, + "grad_norm": 2.0927357338153705, + "learning_rate": 1.4748839221399354e-05, + "loss": 1.06, + "step": 15375 + }, + { + "epoch": 0.36, + "grad_norm": 2.012846994676711, + "learning_rate": 1.4748167694478963e-05, + "loss": 1.1024, + "step": 15376 + }, + { + "epoch": 0.36, + "grad_norm": 2.0431679374170164, + "learning_rate": 1.4747496139913229e-05, + "loss": 0.8965, + "step": 15377 + }, + { + "epoch": 0.36, + "grad_norm": 2.1302708166777307, + "learning_rate": 1.4746824557706076e-05, + "loss": 1.0248, + "step": 15378 + }, + { + "epoch": 0.36, + "grad_norm": 1.1593411354478096, + "learning_rate": 1.4746152947861405e-05, + "loss": 1.0163, + "step": 15379 + }, + { + "epoch": 0.36, + "grad_norm": 2.257881227956129, + "learning_rate": 1.4745481310383127e-05, + "loss": 1.012, + "step": 15380 + }, + { + "epoch": 0.36, + "grad_norm": 2.1852668780147986, + "learning_rate": 1.4744809645275157e-05, + "loss": 1.0299, + "step": 15381 + }, + { + "epoch": 0.36, + "grad_norm": 2.3561647800895718, + "learning_rate": 1.4744137952541402e-05, + "loss": 0.9851, + "step": 15382 + }, + { + "epoch": 0.36, + "grad_norm": 1.078516605037215, + "learning_rate": 1.4743466232185775e-05, + "loss": 1.0183, + "step": 15383 + }, + { + "epoch": 0.36, + "grad_norm": 1.9330479973952925, + "learning_rate": 1.4742794484212184e-05, + "loss": 1.1463, + "step": 15384 + }, + { + "epoch": 0.36, + "grad_norm": 2.028139198584883, + "learning_rate": 1.4742122708624544e-05, + "loss": 1.0715, + "step": 15385 + }, + { + "epoch": 0.36, + "grad_norm": 2.030616444007098, + "learning_rate": 1.4741450905426763e-05, + "loss": 0.8845, + "step": 15386 + }, + { + "epoch": 0.36, + "grad_norm": 2.159720315305302, + "learning_rate": 1.4740779074622756e-05, + "loss": 1.0514, + "step": 15387 + }, + { + "epoch": 0.36, + "grad_norm": 2.1653853512888412, + "learning_rate": 1.4740107216216428e-05, + "loss": 0.9395, + "step": 15388 + }, + { + "epoch": 0.36, + "grad_norm": 1.8801224874075133, + "learning_rate": 1.4739435330211697e-05, + "loss": 0.9988, + "step": 15389 + }, + { + "epoch": 0.36, + "grad_norm": 2.2723350792299186, + "learning_rate": 1.4738763416612476e-05, + "loss": 1.0071, + "step": 15390 + }, + { + "epoch": 0.36, + "grad_norm": 2.104265458468366, + "learning_rate": 1.473809147542267e-05, + "loss": 1.0754, + "step": 15391 + }, + { + "epoch": 0.36, + "grad_norm": 2.4415158027864745, + "learning_rate": 1.4737419506646197e-05, + "loss": 0.9866, + "step": 15392 + }, + { + "epoch": 0.36, + "grad_norm": 2.608480072136517, + "learning_rate": 1.4736747510286968e-05, + "loss": 1.0391, + "step": 15393 + }, + { + "epoch": 0.36, + "grad_norm": 2.065120568094819, + "learning_rate": 1.4736075486348896e-05, + "loss": 1.0116, + "step": 15394 + }, + { + "epoch": 0.36, + "grad_norm": 2.0722032414240634, + "learning_rate": 1.4735403434835891e-05, + "loss": 1.1265, + "step": 15395 + }, + { + "epoch": 0.36, + "grad_norm": 3.1287126891281707, + "learning_rate": 1.473473135575187e-05, + "loss": 1.0919, + "step": 15396 + }, + { + "epoch": 0.36, + "grad_norm": 2.4279006693324736, + "learning_rate": 1.4734059249100743e-05, + "loss": 1.0313, + "step": 15397 + }, + { + "epoch": 0.36, + "grad_norm": 2.2469281748877945, + "learning_rate": 1.4733387114886424e-05, + "loss": 0.9933, + "step": 15398 + }, + { + "epoch": 0.36, + "grad_norm": 1.8790206964784584, + "learning_rate": 1.4732714953112828e-05, + "loss": 0.9711, + "step": 15399 + }, + { + "epoch": 0.36, + "grad_norm": 2.6191644271022225, + "learning_rate": 1.4732042763783866e-05, + "loss": 0.9227, + "step": 15400 + }, + { + "epoch": 0.36, + "grad_norm": 1.8561064731620598, + "learning_rate": 1.4731370546903454e-05, + "loss": 1.0223, + "step": 15401 + }, + { + "epoch": 0.36, + "grad_norm": 2.5405074123276705, + "learning_rate": 1.4730698302475505e-05, + "loss": 1.0868, + "step": 15402 + }, + { + "epoch": 0.36, + "grad_norm": 1.9010703288477602, + "learning_rate": 1.473002603050393e-05, + "loss": 0.9891, + "step": 15403 + }, + { + "epoch": 0.36, + "grad_norm": 1.8869931473627783, + "learning_rate": 1.472935373099265e-05, + "loss": 1.0092, + "step": 15404 + }, + { + "epoch": 0.36, + "grad_norm": 2.1347580047534236, + "learning_rate": 1.472868140394557e-05, + "loss": 1.0182, + "step": 15405 + }, + { + "epoch": 0.36, + "grad_norm": 1.1281099776059662, + "learning_rate": 1.4728009049366617e-05, + "loss": 0.9488, + "step": 15406 + }, + { + "epoch": 0.36, + "grad_norm": 1.8840010916966623, + "learning_rate": 1.472733666725969e-05, + "loss": 1.0618, + "step": 15407 + }, + { + "epoch": 0.36, + "grad_norm": 1.8816448326451838, + "learning_rate": 1.472666425762872e-05, + "loss": 1.0512, + "step": 15408 + }, + { + "epoch": 0.36, + "grad_norm": 1.168325036644149, + "learning_rate": 1.4725991820477612e-05, + "loss": 1.0319, + "step": 15409 + }, + { + "epoch": 0.36, + "grad_norm": 2.140978887900857, + "learning_rate": 1.4725319355810282e-05, + "loss": 1.0243, + "step": 15410 + }, + { + "epoch": 0.36, + "grad_norm": 2.107796032232743, + "learning_rate": 1.4724646863630646e-05, + "loss": 0.9805, + "step": 15411 + }, + { + "epoch": 0.36, + "grad_norm": 2.0051405304844216, + "learning_rate": 1.472397434394262e-05, + "loss": 1.0449, + "step": 15412 + }, + { + "epoch": 0.36, + "grad_norm": 1.8133641188477378, + "learning_rate": 1.4723301796750122e-05, + "loss": 1.0545, + "step": 15413 + }, + { + "epoch": 0.36, + "grad_norm": 1.901482567570609, + "learning_rate": 1.4722629222057063e-05, + "loss": 1.0208, + "step": 15414 + }, + { + "epoch": 0.36, + "grad_norm": 2.1111951853639694, + "learning_rate": 1.4721956619867364e-05, + "loss": 1.1808, + "step": 15415 + }, + { + "epoch": 0.36, + "grad_norm": 2.0526958947086595, + "learning_rate": 1.4721283990184938e-05, + "loss": 1.1241, + "step": 15416 + }, + { + "epoch": 0.36, + "grad_norm": 1.9384293125865593, + "learning_rate": 1.4720611333013701e-05, + "loss": 1.1168, + "step": 15417 + }, + { + "epoch": 0.36, + "grad_norm": 2.18268917314925, + "learning_rate": 1.471993864835757e-05, + "loss": 1.0793, + "step": 15418 + }, + { + "epoch": 0.36, + "grad_norm": 2.0327718474985694, + "learning_rate": 1.4719265936220464e-05, + "loss": 0.9919, + "step": 15419 + }, + { + "epoch": 0.36, + "grad_norm": 3.014123347719036, + "learning_rate": 1.4718593196606294e-05, + "loss": 0.9622, + "step": 15420 + }, + { + "epoch": 0.36, + "grad_norm": 2.1573138089839947, + "learning_rate": 1.4717920429518984e-05, + "loss": 1.0968, + "step": 15421 + }, + { + "epoch": 0.36, + "grad_norm": 3.3917655811801426, + "learning_rate": 1.4717247634962446e-05, + "loss": 1.0612, + "step": 15422 + }, + { + "epoch": 0.36, + "grad_norm": 1.9650125495613833, + "learning_rate": 1.4716574812940603e-05, + "loss": 0.9185, + "step": 15423 + }, + { + "epoch": 0.36, + "grad_norm": 1.9683769457794227, + "learning_rate": 1.4715901963457362e-05, + "loss": 1.0215, + "step": 15424 + }, + { + "epoch": 0.36, + "grad_norm": 2.1050945759636743, + "learning_rate": 1.4715229086516649e-05, + "loss": 1.0509, + "step": 15425 + }, + { + "epoch": 0.36, + "grad_norm": 2.450493530622331, + "learning_rate": 1.471455618212238e-05, + "loss": 1.0867, + "step": 15426 + }, + { + "epoch": 0.36, + "grad_norm": 1.9521740616638064, + "learning_rate": 1.4713883250278471e-05, + "loss": 0.9345, + "step": 15427 + }, + { + "epoch": 0.36, + "grad_norm": 1.9961976622196105, + "learning_rate": 1.4713210290988842e-05, + "loss": 1.0632, + "step": 15428 + }, + { + "epoch": 0.36, + "grad_norm": 1.1326631717873945, + "learning_rate": 1.471253730425741e-05, + "loss": 1.0217, + "step": 15429 + }, + { + "epoch": 0.36, + "grad_norm": 1.8907147795812256, + "learning_rate": 1.4711864290088099e-05, + "loss": 0.9743, + "step": 15430 + }, + { + "epoch": 0.36, + "grad_norm": 1.8214598859392013, + "learning_rate": 1.4711191248484812e-05, + "loss": 1.0531, + "step": 15431 + }, + { + "epoch": 0.36, + "grad_norm": 1.859992391101434, + "learning_rate": 1.4710518179451487e-05, + "loss": 0.9944, + "step": 15432 + }, + { + "epoch": 0.36, + "grad_norm": 1.9783733929670468, + "learning_rate": 1.470984508299203e-05, + "loss": 0.9867, + "step": 15433 + }, + { + "epoch": 0.36, + "grad_norm": 2.0954125525992464, + "learning_rate": 1.4709171959110367e-05, + "loss": 1.137, + "step": 15434 + }, + { + "epoch": 0.36, + "grad_norm": 1.8003892928402228, + "learning_rate": 1.470849880781041e-05, + "loss": 1.0191, + "step": 15435 + }, + { + "epoch": 0.36, + "grad_norm": 2.003907013125209, + "learning_rate": 1.4707825629096085e-05, + "loss": 1.1482, + "step": 15436 + }, + { + "epoch": 0.36, + "grad_norm": 2.6621129799634407, + "learning_rate": 1.4707152422971307e-05, + "loss": 0.9789, + "step": 15437 + }, + { + "epoch": 0.36, + "grad_norm": 1.9068646205338802, + "learning_rate": 1.4706479189439998e-05, + "loss": 1.0299, + "step": 15438 + }, + { + "epoch": 0.36, + "grad_norm": 1.9040109793923004, + "learning_rate": 1.4705805928506076e-05, + "loss": 0.9205, + "step": 15439 + }, + { + "epoch": 0.36, + "grad_norm": 2.063200905694237, + "learning_rate": 1.4705132640173463e-05, + "loss": 1.043, + "step": 15440 + }, + { + "epoch": 0.36, + "grad_norm": 2.0793529721776074, + "learning_rate": 1.4704459324446078e-05, + "loss": 1.0804, + "step": 15441 + }, + { + "epoch": 0.36, + "grad_norm": 1.0672725631895197, + "learning_rate": 1.4703785981327841e-05, + "loss": 0.9053, + "step": 15442 + }, + { + "epoch": 0.36, + "grad_norm": 1.0952017085401249, + "learning_rate": 1.4703112610822674e-05, + "loss": 0.9014, + "step": 15443 + }, + { + "epoch": 0.36, + "grad_norm": 1.063836264357841, + "learning_rate": 1.4702439212934496e-05, + "loss": 0.9265, + "step": 15444 + }, + { + "epoch": 0.36, + "grad_norm": 2.026413189009658, + "learning_rate": 1.4701765787667228e-05, + "loss": 1.0765, + "step": 15445 + }, + { + "epoch": 0.36, + "grad_norm": 1.8309886469057783, + "learning_rate": 1.4701092335024792e-05, + "loss": 1.1373, + "step": 15446 + }, + { + "epoch": 0.36, + "grad_norm": 2.144887933942763, + "learning_rate": 1.4700418855011105e-05, + "loss": 0.9806, + "step": 15447 + }, + { + "epoch": 0.36, + "grad_norm": 2.174939833554269, + "learning_rate": 1.4699745347630095e-05, + "loss": 0.9486, + "step": 15448 + }, + { + "epoch": 0.36, + "grad_norm": 1.8726648232170677, + "learning_rate": 1.4699071812885677e-05, + "loss": 1.0011, + "step": 15449 + }, + { + "epoch": 0.36, + "grad_norm": 1.960842072897734, + "learning_rate": 1.4698398250781776e-05, + "loss": 1.0702, + "step": 15450 + }, + { + "epoch": 0.36, + "grad_norm": 1.8933757762137566, + "learning_rate": 1.4697724661322312e-05, + "loss": 1.0887, + "step": 15451 + }, + { + "epoch": 0.36, + "grad_norm": 1.9777290650913717, + "learning_rate": 1.469705104451121e-05, + "loss": 0.8862, + "step": 15452 + }, + { + "epoch": 0.36, + "grad_norm": 1.123695423341061, + "learning_rate": 1.469637740035239e-05, + "loss": 1.0294, + "step": 15453 + }, + { + "epoch": 0.36, + "grad_norm": 1.1622748888180163, + "learning_rate": 1.4695703728849771e-05, + "loss": 0.9244, + "step": 15454 + }, + { + "epoch": 0.36, + "grad_norm": 1.8653348667226475, + "learning_rate": 1.4695030030007282e-05, + "loss": 1.0538, + "step": 15455 + }, + { + "epoch": 0.36, + "grad_norm": 1.7776257245428408, + "learning_rate": 1.4694356303828837e-05, + "loss": 1.1072, + "step": 15456 + }, + { + "epoch": 0.36, + "grad_norm": 1.9997152383278878, + "learning_rate": 1.4693682550318368e-05, + "loss": 1.0465, + "step": 15457 + }, + { + "epoch": 0.36, + "grad_norm": 2.091648149390086, + "learning_rate": 1.4693008769479788e-05, + "loss": 1.0539, + "step": 15458 + }, + { + "epoch": 0.36, + "grad_norm": 1.856283657915967, + "learning_rate": 1.4692334961317033e-05, + "loss": 1.0083, + "step": 15459 + }, + { + "epoch": 0.36, + "grad_norm": 2.2481904999989624, + "learning_rate": 1.4691661125834012e-05, + "loss": 1.0744, + "step": 15460 + }, + { + "epoch": 0.36, + "grad_norm": 1.987961261950928, + "learning_rate": 1.4690987263034655e-05, + "loss": 1.033, + "step": 15461 + }, + { + "epoch": 0.36, + "grad_norm": 2.007373815813191, + "learning_rate": 1.4690313372922887e-05, + "loss": 1.0319, + "step": 15462 + }, + { + "epoch": 0.36, + "grad_norm": 1.9602834173591415, + "learning_rate": 1.4689639455502629e-05, + "loss": 0.9998, + "step": 15463 + }, + { + "epoch": 0.36, + "grad_norm": 1.8167963834403895, + "learning_rate": 1.4688965510777803e-05, + "loss": 1.0861, + "step": 15464 + }, + { + "epoch": 0.36, + "grad_norm": 1.9303126196771876, + "learning_rate": 1.468829153875234e-05, + "loss": 1.036, + "step": 15465 + }, + { + "epoch": 0.36, + "grad_norm": 1.1765780620286683, + "learning_rate": 1.4687617539430154e-05, + "loss": 0.9381, + "step": 15466 + }, + { + "epoch": 0.36, + "grad_norm": 2.082953534121051, + "learning_rate": 1.4686943512815179e-05, + "loss": 0.9893, + "step": 15467 + }, + { + "epoch": 0.36, + "grad_norm": 1.8315971623790321, + "learning_rate": 1.4686269458911333e-05, + "loss": 1.0774, + "step": 15468 + }, + { + "epoch": 0.36, + "grad_norm": 2.048982338742692, + "learning_rate": 1.4685595377722544e-05, + "loss": 0.9785, + "step": 15469 + }, + { + "epoch": 0.36, + "grad_norm": 1.9330662458159433, + "learning_rate": 1.4684921269252734e-05, + "loss": 1.0486, + "step": 15470 + }, + { + "epoch": 0.36, + "grad_norm": 2.1209905313436184, + "learning_rate": 1.468424713350583e-05, + "loss": 0.9065, + "step": 15471 + }, + { + "epoch": 0.36, + "grad_norm": 2.2815019524748226, + "learning_rate": 1.4683572970485756e-05, + "loss": 0.9682, + "step": 15472 + }, + { + "epoch": 0.36, + "grad_norm": 1.827640380442949, + "learning_rate": 1.4682898780196434e-05, + "loss": 0.9894, + "step": 15473 + }, + { + "epoch": 0.36, + "grad_norm": 1.932749266156726, + "learning_rate": 1.4682224562641798e-05, + "loss": 0.9257, + "step": 15474 + }, + { + "epoch": 0.36, + "grad_norm": 1.0932584466923572, + "learning_rate": 1.4681550317825766e-05, + "loss": 0.9491, + "step": 15475 + }, + { + "epoch": 0.36, + "grad_norm": 2.088598967323642, + "learning_rate": 1.4680876045752266e-05, + "loss": 1.0104, + "step": 15476 + }, + { + "epoch": 0.36, + "grad_norm": 2.0211410837382435, + "learning_rate": 1.4680201746425221e-05, + "loss": 1.0527, + "step": 15477 + }, + { + "epoch": 0.36, + "grad_norm": 2.0382602743436244, + "learning_rate": 1.4679527419848563e-05, + "loss": 1.1071, + "step": 15478 + }, + { + "epoch": 0.36, + "grad_norm": 1.8164453818223245, + "learning_rate": 1.4678853066026216e-05, + "loss": 1.0541, + "step": 15479 + }, + { + "epoch": 0.36, + "grad_norm": 1.9756192955485639, + "learning_rate": 1.4678178684962101e-05, + "loss": 1.1912, + "step": 15480 + }, + { + "epoch": 0.36, + "grad_norm": 1.8107862746346497, + "learning_rate": 1.4677504276660153e-05, + "loss": 0.8791, + "step": 15481 + }, + { + "epoch": 0.36, + "grad_norm": 1.888506539949966, + "learning_rate": 1.4676829841124288e-05, + "loss": 1.0596, + "step": 15482 + }, + { + "epoch": 0.36, + "grad_norm": 1.8550877149747307, + "learning_rate": 1.4676155378358447e-05, + "loss": 1.0771, + "step": 15483 + }, + { + "epoch": 0.36, + "grad_norm": 1.7778728004250495, + "learning_rate": 1.4675480888366545e-05, + "loss": 0.9651, + "step": 15484 + }, + { + "epoch": 0.36, + "grad_norm": 2.133196809007005, + "learning_rate": 1.4674806371152514e-05, + "loss": 1.0853, + "step": 15485 + }, + { + "epoch": 0.36, + "grad_norm": 1.1123134335968619, + "learning_rate": 1.4674131826720278e-05, + "loss": 1.0138, + "step": 15486 + }, + { + "epoch": 0.36, + "grad_norm": 2.0382541401788488, + "learning_rate": 1.4673457255073769e-05, + "loss": 1.113, + "step": 15487 + }, + { + "epoch": 0.36, + "grad_norm": 1.8510475264938886, + "learning_rate": 1.4672782656216911e-05, + "loss": 1.081, + "step": 15488 + }, + { + "epoch": 0.36, + "grad_norm": 1.7600109400899404, + "learning_rate": 1.4672108030153634e-05, + "loss": 0.9942, + "step": 15489 + }, + { + "epoch": 0.36, + "grad_norm": 1.9893168202816693, + "learning_rate": 1.4671433376887864e-05, + "loss": 1.0751, + "step": 15490 + }, + { + "epoch": 0.36, + "grad_norm": 1.9237722710668226, + "learning_rate": 1.467075869642353e-05, + "loss": 1.0313, + "step": 15491 + }, + { + "epoch": 0.36, + "grad_norm": 2.1370403280558343, + "learning_rate": 1.4670083988764559e-05, + "loss": 1.0682, + "step": 15492 + }, + { + "epoch": 0.37, + "grad_norm": 2.0688663111114174, + "learning_rate": 1.4669409253914882e-05, + "loss": 0.9715, + "step": 15493 + }, + { + "epoch": 0.37, + "grad_norm": 2.1137567926429686, + "learning_rate": 1.4668734491878426e-05, + "loss": 1.0374, + "step": 15494 + }, + { + "epoch": 0.37, + "grad_norm": 1.9847161957600061, + "learning_rate": 1.4668059702659121e-05, + "loss": 1.1719, + "step": 15495 + }, + { + "epoch": 0.37, + "grad_norm": 2.8816117093934155, + "learning_rate": 1.4667384886260892e-05, + "loss": 1.0455, + "step": 15496 + }, + { + "epoch": 0.37, + "grad_norm": 2.0651439533358564, + "learning_rate": 1.466671004268767e-05, + "loss": 1.084, + "step": 15497 + }, + { + "epoch": 0.37, + "grad_norm": 1.9987825806827113, + "learning_rate": 1.4666035171943388e-05, + "loss": 1.0149, + "step": 15498 + }, + { + "epoch": 0.37, + "grad_norm": 2.09020373717713, + "learning_rate": 1.466536027403197e-05, + "loss": 1.017, + "step": 15499 + }, + { + "epoch": 0.37, + "grad_norm": 2.245312907918473, + "learning_rate": 1.4664685348957348e-05, + "loss": 0.9695, + "step": 15500 + }, + { + "epoch": 0.37, + "grad_norm": 1.9151929195853366, + "learning_rate": 1.4664010396723449e-05, + "loss": 1.0246, + "step": 15501 + }, + { + "epoch": 0.37, + "grad_norm": 1.899539449434049, + "learning_rate": 1.466333541733421e-05, + "loss": 1.0978, + "step": 15502 + }, + { + "epoch": 0.37, + "grad_norm": 1.1239589225017073, + "learning_rate": 1.466266041079355e-05, + "loss": 0.9948, + "step": 15503 + }, + { + "epoch": 0.37, + "grad_norm": 1.8049267107425269, + "learning_rate": 1.466198537710541e-05, + "loss": 1.0798, + "step": 15504 + }, + { + "epoch": 0.37, + "grad_norm": 1.9232527157954498, + "learning_rate": 1.4661310316273709e-05, + "loss": 0.9067, + "step": 15505 + }, + { + "epoch": 0.37, + "grad_norm": 4.267212345380696, + "learning_rate": 1.4660635228302388e-05, + "loss": 0.9488, + "step": 15506 + }, + { + "epoch": 0.37, + "grad_norm": 1.7701631115523992, + "learning_rate": 1.4659960113195371e-05, + "loss": 0.9985, + "step": 15507 + }, + { + "epoch": 0.37, + "grad_norm": 2.4811725639203694, + "learning_rate": 1.4659284970956593e-05, + "loss": 0.9593, + "step": 15508 + }, + { + "epoch": 0.37, + "grad_norm": 1.9338396616218354, + "learning_rate": 1.4658609801589982e-05, + "loss": 0.8598, + "step": 15509 + }, + { + "epoch": 0.37, + "grad_norm": 1.9102218804732443, + "learning_rate": 1.465793460509947e-05, + "loss": 1.0641, + "step": 15510 + }, + { + "epoch": 0.37, + "grad_norm": 2.3638665051436822, + "learning_rate": 1.4657259381488988e-05, + "loss": 1.0028, + "step": 15511 + }, + { + "epoch": 0.37, + "grad_norm": 1.9369641966023097, + "learning_rate": 1.4656584130762467e-05, + "loss": 1.0166, + "step": 15512 + }, + { + "epoch": 0.37, + "grad_norm": 2.128417522979956, + "learning_rate": 1.4655908852923841e-05, + "loss": 1.0785, + "step": 15513 + }, + { + "epoch": 0.37, + "grad_norm": 2.11503476586215, + "learning_rate": 1.4655233547977037e-05, + "loss": 1.0454, + "step": 15514 + }, + { + "epoch": 0.37, + "grad_norm": 1.9656610230115172, + "learning_rate": 1.465455821592599e-05, + "loss": 1.0571, + "step": 15515 + }, + { + "epoch": 0.37, + "grad_norm": 1.8933506042773345, + "learning_rate": 1.465388285677463e-05, + "loss": 1.0699, + "step": 15516 + }, + { + "epoch": 0.37, + "grad_norm": 1.8324075705816272, + "learning_rate": 1.4653207470526893e-05, + "loss": 1.0994, + "step": 15517 + }, + { + "epoch": 0.37, + "grad_norm": 2.4990001497938916, + "learning_rate": 1.465253205718671e-05, + "loss": 0.9859, + "step": 15518 + }, + { + "epoch": 0.37, + "grad_norm": 1.8162347484450865, + "learning_rate": 1.4651856616758009e-05, + "loss": 0.9056, + "step": 15519 + }, + { + "epoch": 0.37, + "grad_norm": 2.3001654471571844, + "learning_rate": 1.4651181149244728e-05, + "loss": 1.0299, + "step": 15520 + }, + { + "epoch": 0.37, + "grad_norm": 1.917019532810735, + "learning_rate": 1.46505056546508e-05, + "loss": 0.9467, + "step": 15521 + }, + { + "epoch": 0.37, + "grad_norm": 2.3476261132617964, + "learning_rate": 1.4649830132980151e-05, + "loss": 0.9866, + "step": 15522 + }, + { + "epoch": 0.37, + "grad_norm": 1.9690577949621675, + "learning_rate": 1.4649154584236723e-05, + "loss": 0.9966, + "step": 15523 + }, + { + "epoch": 0.37, + "grad_norm": 1.052886977153862, + "learning_rate": 1.464847900842444e-05, + "loss": 0.9235, + "step": 15524 + }, + { + "epoch": 0.37, + "grad_norm": 1.8611330486016828, + "learning_rate": 1.4647803405547248e-05, + "loss": 1.1213, + "step": 15525 + }, + { + "epoch": 0.37, + "grad_norm": 2.2401894150822796, + "learning_rate": 1.4647127775609067e-05, + "loss": 1.0229, + "step": 15526 + }, + { + "epoch": 0.37, + "grad_norm": 1.0955461281864878, + "learning_rate": 1.4646452118613838e-05, + "loss": 1.0112, + "step": 15527 + }, + { + "epoch": 0.37, + "grad_norm": 2.0606334094096366, + "learning_rate": 1.4645776434565492e-05, + "loss": 0.993, + "step": 15528 + }, + { + "epoch": 0.37, + "grad_norm": 2.0672675914555207, + "learning_rate": 1.4645100723467969e-05, + "loss": 1.0199, + "step": 15529 + }, + { + "epoch": 0.37, + "grad_norm": 2.0626315214702857, + "learning_rate": 1.4644424985325198e-05, + "loss": 0.9522, + "step": 15530 + }, + { + "epoch": 0.37, + "grad_norm": 2.071479521523979, + "learning_rate": 1.4643749220141113e-05, + "loss": 1.038, + "step": 15531 + }, + { + "epoch": 0.37, + "grad_norm": 2.0056304891590093, + "learning_rate": 1.4643073427919652e-05, + "loss": 0.8437, + "step": 15532 + }, + { + "epoch": 0.37, + "grad_norm": 1.9888660815235972, + "learning_rate": 1.4642397608664744e-05, + "loss": 1.0502, + "step": 15533 + }, + { + "epoch": 0.37, + "grad_norm": 1.9608340440298915, + "learning_rate": 1.4641721762380332e-05, + "loss": 1.1152, + "step": 15534 + }, + { + "epoch": 0.37, + "grad_norm": 2.0411538671602663, + "learning_rate": 1.4641045889070343e-05, + "loss": 1.17, + "step": 15535 + }, + { + "epoch": 0.37, + "grad_norm": 3.23101285227865, + "learning_rate": 1.4640369988738717e-05, + "loss": 1.0191, + "step": 15536 + }, + { + "epoch": 0.37, + "grad_norm": 2.2303292576166456, + "learning_rate": 1.4639694061389384e-05, + "loss": 0.9776, + "step": 15537 + }, + { + "epoch": 0.37, + "grad_norm": 1.9736860357513872, + "learning_rate": 1.4639018107026288e-05, + "loss": 0.9801, + "step": 15538 + }, + { + "epoch": 0.37, + "grad_norm": 2.050260447096353, + "learning_rate": 1.4638342125653358e-05, + "loss": 1.2402, + "step": 15539 + }, + { + "epoch": 0.37, + "grad_norm": 1.0900442712734892, + "learning_rate": 1.4637666117274531e-05, + "loss": 0.9772, + "step": 15540 + }, + { + "epoch": 0.37, + "grad_norm": 2.005721902535674, + "learning_rate": 1.4636990081893746e-05, + "loss": 1.1787, + "step": 15541 + }, + { + "epoch": 0.37, + "grad_norm": 2.6590278647290204, + "learning_rate": 1.4636314019514935e-05, + "loss": 1.0189, + "step": 15542 + }, + { + "epoch": 0.37, + "grad_norm": 2.071764822775293, + "learning_rate": 1.4635637930142035e-05, + "loss": 1.175, + "step": 15543 + }, + { + "epoch": 0.37, + "grad_norm": 1.8843265954174555, + "learning_rate": 1.4634961813778982e-05, + "loss": 1.0912, + "step": 15544 + }, + { + "epoch": 0.37, + "grad_norm": 2.0680532181580453, + "learning_rate": 1.4634285670429717e-05, + "loss": 1.1248, + "step": 15545 + }, + { + "epoch": 0.37, + "grad_norm": 1.9909498943146977, + "learning_rate": 1.4633609500098174e-05, + "loss": 1.0144, + "step": 15546 + }, + { + "epoch": 0.37, + "grad_norm": 2.037481658910041, + "learning_rate": 1.4632933302788289e-05, + "loss": 1.0092, + "step": 15547 + }, + { + "epoch": 0.37, + "grad_norm": 2.0485401035210367, + "learning_rate": 1.4632257078503997e-05, + "loss": 1.0079, + "step": 15548 + }, + { + "epoch": 0.37, + "grad_norm": 2.211855077483365, + "learning_rate": 1.4631580827249242e-05, + "loss": 0.9953, + "step": 15549 + }, + { + "epoch": 0.37, + "grad_norm": 2.140759484846252, + "learning_rate": 1.4630904549027954e-05, + "loss": 0.9398, + "step": 15550 + }, + { + "epoch": 0.37, + "grad_norm": 1.1448522348944976, + "learning_rate": 1.4630228243844073e-05, + "loss": 1.0102, + "step": 15551 + }, + { + "epoch": 0.37, + "grad_norm": 2.032421809000714, + "learning_rate": 1.462955191170154e-05, + "loss": 0.8692, + "step": 15552 + }, + { + "epoch": 0.37, + "grad_norm": 2.0810495951705237, + "learning_rate": 1.4628875552604291e-05, + "loss": 1.0327, + "step": 15553 + }, + { + "epoch": 0.37, + "grad_norm": 2.10954164601724, + "learning_rate": 1.4628199166556258e-05, + "loss": 0.9186, + "step": 15554 + }, + { + "epoch": 0.37, + "grad_norm": 2.0800293944699604, + "learning_rate": 1.462752275356139e-05, + "loss": 0.9577, + "step": 15555 + }, + { + "epoch": 0.37, + "grad_norm": 2.1850732363844254, + "learning_rate": 1.4626846313623615e-05, + "loss": 1.0811, + "step": 15556 + }, + { + "epoch": 0.37, + "grad_norm": 1.915603293056507, + "learning_rate": 1.4626169846746878e-05, + "loss": 1.0612, + "step": 15557 + }, + { + "epoch": 0.37, + "grad_norm": 4.16945281316373, + "learning_rate": 1.4625493352935114e-05, + "loss": 1.1353, + "step": 15558 + }, + { + "epoch": 0.37, + "grad_norm": 1.643639874743905, + "learning_rate": 1.4624816832192266e-05, + "loss": 1.0057, + "step": 15559 + }, + { + "epoch": 0.37, + "grad_norm": 1.9135258369919332, + "learning_rate": 1.4624140284522267e-05, + "loss": 1.1049, + "step": 15560 + }, + { + "epoch": 0.37, + "grad_norm": 2.2175942881500608, + "learning_rate": 1.462346370992906e-05, + "loss": 1.0103, + "step": 15561 + }, + { + "epoch": 0.37, + "grad_norm": 1.873218571890204, + "learning_rate": 1.4622787108416585e-05, + "loss": 0.9574, + "step": 15562 + }, + { + "epoch": 0.37, + "grad_norm": 1.1724143565267873, + "learning_rate": 1.4622110479988778e-05, + "loss": 1.0012, + "step": 15563 + }, + { + "epoch": 0.37, + "grad_norm": 1.893231127688526, + "learning_rate": 1.4621433824649582e-05, + "loss": 1.0627, + "step": 15564 + }, + { + "epoch": 0.37, + "grad_norm": 1.866915647740783, + "learning_rate": 1.4620757142402933e-05, + "loss": 1.0299, + "step": 15565 + }, + { + "epoch": 0.37, + "grad_norm": 1.887217180760222, + "learning_rate": 1.4620080433252776e-05, + "loss": 0.9624, + "step": 15566 + }, + { + "epoch": 0.37, + "grad_norm": 1.9345756784751722, + "learning_rate": 1.4619403697203045e-05, + "loss": 0.9468, + "step": 15567 + }, + { + "epoch": 0.37, + "grad_norm": 2.231251557951338, + "learning_rate": 1.4618726934257687e-05, + "loss": 0.8941, + "step": 15568 + }, + { + "epoch": 0.37, + "grad_norm": 1.9895615025183095, + "learning_rate": 1.4618050144420637e-05, + "loss": 1.2546, + "step": 15569 + }, + { + "epoch": 0.37, + "grad_norm": 2.0380457396397538, + "learning_rate": 1.4617373327695837e-05, + "loss": 1.014, + "step": 15570 + }, + { + "epoch": 0.37, + "grad_norm": 2.083368659117336, + "learning_rate": 1.4616696484087228e-05, + "loss": 0.9837, + "step": 15571 + }, + { + "epoch": 0.37, + "grad_norm": 1.9685183901638128, + "learning_rate": 1.4616019613598748e-05, + "loss": 1.065, + "step": 15572 + }, + { + "epoch": 0.37, + "grad_norm": 1.1423019050843677, + "learning_rate": 1.4615342716234342e-05, + "loss": 0.9799, + "step": 15573 + }, + { + "epoch": 0.37, + "grad_norm": 1.846497823040005, + "learning_rate": 1.461466579199795e-05, + "loss": 0.9597, + "step": 15574 + }, + { + "epoch": 0.37, + "grad_norm": 2.046925325835148, + "learning_rate": 1.4613988840893512e-05, + "loss": 1.1307, + "step": 15575 + }, + { + "epoch": 0.37, + "grad_norm": 1.9406377332140257, + "learning_rate": 1.461331186292497e-05, + "loss": 0.9271, + "step": 15576 + }, + { + "epoch": 0.37, + "grad_norm": 1.9830663609993822, + "learning_rate": 1.4612634858096266e-05, + "loss": 1.0482, + "step": 15577 + }, + { + "epoch": 0.37, + "grad_norm": 2.0352627316558993, + "learning_rate": 1.4611957826411343e-05, + "loss": 1.0693, + "step": 15578 + }, + { + "epoch": 0.37, + "grad_norm": 2.782074871259549, + "learning_rate": 1.461128076787414e-05, + "loss": 0.9464, + "step": 15579 + }, + { + "epoch": 0.37, + "grad_norm": 2.0112528890011614, + "learning_rate": 1.46106036824886e-05, + "loss": 1.088, + "step": 15580 + }, + { + "epoch": 0.37, + "grad_norm": 2.132471333825656, + "learning_rate": 1.4609926570258665e-05, + "loss": 0.9389, + "step": 15581 + }, + { + "epoch": 0.37, + "grad_norm": 1.750208879595332, + "learning_rate": 1.460924943118828e-05, + "loss": 0.9937, + "step": 15582 + }, + { + "epoch": 0.37, + "grad_norm": 1.7959766123301408, + "learning_rate": 1.460857226528139e-05, + "loss": 1.0657, + "step": 15583 + }, + { + "epoch": 0.37, + "grad_norm": 1.9815568595263242, + "learning_rate": 1.4607895072541925e-05, + "loss": 1.018, + "step": 15584 + }, + { + "epoch": 0.37, + "grad_norm": 2.39222523393138, + "learning_rate": 1.4607217852973841e-05, + "loss": 0.9589, + "step": 15585 + }, + { + "epoch": 0.37, + "grad_norm": 1.8403200603084633, + "learning_rate": 1.460654060658107e-05, + "loss": 1.0097, + "step": 15586 + }, + { + "epoch": 0.37, + "grad_norm": 1.9629586884692523, + "learning_rate": 1.4605863333367567e-05, + "loss": 0.8329, + "step": 15587 + }, + { + "epoch": 0.37, + "grad_norm": 1.891052996277587, + "learning_rate": 1.4605186033337266e-05, + "loss": 1.1623, + "step": 15588 + }, + { + "epoch": 0.37, + "grad_norm": 2.063087017134341, + "learning_rate": 1.4604508706494119e-05, + "loss": 1.1275, + "step": 15589 + }, + { + "epoch": 0.37, + "grad_norm": 2.9515251834795153, + "learning_rate": 1.4603831352842061e-05, + "loss": 1.1145, + "step": 15590 + }, + { + "epoch": 0.37, + "grad_norm": 1.1562033365755282, + "learning_rate": 1.460315397238504e-05, + "loss": 1.0193, + "step": 15591 + }, + { + "epoch": 0.37, + "grad_norm": 5.361605321166392, + "learning_rate": 1.4602476565127e-05, + "loss": 1.119, + "step": 15592 + }, + { + "epoch": 0.37, + "grad_norm": 1.0642507109104935, + "learning_rate": 1.4601799131071883e-05, + "loss": 0.9483, + "step": 15593 + }, + { + "epoch": 0.37, + "grad_norm": 2.0353521069298166, + "learning_rate": 1.4601121670223633e-05, + "loss": 1.0285, + "step": 15594 + }, + { + "epoch": 0.37, + "grad_norm": 2.552042337806227, + "learning_rate": 1.4600444182586197e-05, + "loss": 1.1463, + "step": 15595 + }, + { + "epoch": 0.37, + "grad_norm": 1.9907419742014638, + "learning_rate": 1.4599766668163522e-05, + "loss": 1.0052, + "step": 15596 + }, + { + "epoch": 0.37, + "grad_norm": 1.0687743205794518, + "learning_rate": 1.4599089126959545e-05, + "loss": 0.9665, + "step": 15597 + }, + { + "epoch": 0.37, + "grad_norm": 2.0014016748004972, + "learning_rate": 1.4598411558978217e-05, + "loss": 1.0079, + "step": 15598 + }, + { + "epoch": 0.37, + "grad_norm": 2.259296146843869, + "learning_rate": 1.4597733964223479e-05, + "loss": 1.0872, + "step": 15599 + }, + { + "epoch": 0.37, + "grad_norm": 2.627247850200346, + "learning_rate": 1.459705634269928e-05, + "loss": 0.9995, + "step": 15600 + }, + { + "epoch": 0.37, + "grad_norm": 1.0336696649948993, + "learning_rate": 1.4596378694409565e-05, + "loss": 0.9196, + "step": 15601 + }, + { + "epoch": 0.37, + "grad_norm": 1.073034991082929, + "learning_rate": 1.4595701019358275e-05, + "loss": 0.95, + "step": 15602 + }, + { + "epoch": 0.37, + "grad_norm": 1.9666392747152082, + "learning_rate": 1.4595023317549362e-05, + "loss": 0.9929, + "step": 15603 + }, + { + "epoch": 0.37, + "grad_norm": 1.878545639512617, + "learning_rate": 1.4594345588986767e-05, + "loss": 1.1374, + "step": 15604 + }, + { + "epoch": 0.37, + "grad_norm": 2.018214579601281, + "learning_rate": 1.4593667833674437e-05, + "loss": 0.9328, + "step": 15605 + }, + { + "epoch": 0.37, + "grad_norm": 2.4965398376595442, + "learning_rate": 1.4592990051616318e-05, + "loss": 1.0679, + "step": 15606 + }, + { + "epoch": 0.37, + "grad_norm": 2.01403832569996, + "learning_rate": 1.4592312242816355e-05, + "loss": 1.0309, + "step": 15607 + }, + { + "epoch": 0.37, + "grad_norm": 2.013954325071241, + "learning_rate": 1.45916344072785e-05, + "loss": 0.9701, + "step": 15608 + }, + { + "epoch": 0.37, + "grad_norm": 2.1197102551293, + "learning_rate": 1.4590956545006693e-05, + "loss": 1.1137, + "step": 15609 + }, + { + "epoch": 0.37, + "grad_norm": 2.376879584611589, + "learning_rate": 1.4590278656004888e-05, + "loss": 0.9783, + "step": 15610 + }, + { + "epoch": 0.37, + "grad_norm": 2.552419445035897, + "learning_rate": 1.4589600740277023e-05, + "loss": 1.0005, + "step": 15611 + }, + { + "epoch": 0.37, + "grad_norm": 3.404322527538741, + "learning_rate": 1.4588922797827052e-05, + "loss": 1.0749, + "step": 15612 + }, + { + "epoch": 0.37, + "grad_norm": 2.1103867715332716, + "learning_rate": 1.4588244828658918e-05, + "loss": 1.0654, + "step": 15613 + }, + { + "epoch": 0.37, + "grad_norm": 1.1482118065285154, + "learning_rate": 1.4587566832776569e-05, + "loss": 0.9209, + "step": 15614 + }, + { + "epoch": 0.37, + "grad_norm": 1.1065120807343263, + "learning_rate": 1.4586888810183956e-05, + "loss": 1.0163, + "step": 15615 + }, + { + "epoch": 0.37, + "grad_norm": 1.1155629937979816, + "learning_rate": 1.4586210760885023e-05, + "loss": 0.9355, + "step": 15616 + }, + { + "epoch": 0.37, + "grad_norm": 2.0498200589094795, + "learning_rate": 1.4585532684883718e-05, + "loss": 0.9683, + "step": 15617 + }, + { + "epoch": 0.37, + "grad_norm": 2.2660814880815474, + "learning_rate": 1.4584854582183992e-05, + "loss": 1.0932, + "step": 15618 + }, + { + "epoch": 0.37, + "grad_norm": 2.076916422751922, + "learning_rate": 1.4584176452789791e-05, + "loss": 1.0259, + "step": 15619 + }, + { + "epoch": 0.37, + "grad_norm": 1.8445708812200898, + "learning_rate": 1.4583498296705063e-05, + "loss": 1.1211, + "step": 15620 + }, + { + "epoch": 0.37, + "grad_norm": 2.116320480245352, + "learning_rate": 1.4582820113933755e-05, + "loss": 1.1568, + "step": 15621 + }, + { + "epoch": 0.37, + "grad_norm": 2.450728763845049, + "learning_rate": 1.4582141904479819e-05, + "loss": 0.9125, + "step": 15622 + }, + { + "epoch": 0.37, + "grad_norm": 2.1884889107602192, + "learning_rate": 1.4581463668347201e-05, + "loss": 1.0695, + "step": 15623 + }, + { + "epoch": 0.37, + "grad_norm": 2.0471509071164653, + "learning_rate": 1.4580785405539853e-05, + "loss": 0.9724, + "step": 15624 + }, + { + "epoch": 0.37, + "grad_norm": 2.2747895068899098, + "learning_rate": 1.4580107116061722e-05, + "loss": 1.0833, + "step": 15625 + }, + { + "epoch": 0.37, + "grad_norm": 1.9620816504304224, + "learning_rate": 1.4579428799916754e-05, + "loss": 1.0747, + "step": 15626 + }, + { + "epoch": 0.37, + "grad_norm": 1.9985743733623467, + "learning_rate": 1.4578750457108904e-05, + "loss": 0.9502, + "step": 15627 + }, + { + "epoch": 0.37, + "grad_norm": 2.163336268765807, + "learning_rate": 1.4578072087642118e-05, + "loss": 0.9309, + "step": 15628 + }, + { + "epoch": 0.37, + "grad_norm": 1.9414093326490403, + "learning_rate": 1.4577393691520349e-05, + "loss": 1.1287, + "step": 15629 + }, + { + "epoch": 0.37, + "grad_norm": 2.2207407704922666, + "learning_rate": 1.4576715268747545e-05, + "loss": 1.0878, + "step": 15630 + }, + { + "epoch": 0.37, + "grad_norm": 2.5290920863223323, + "learning_rate": 1.4576036819327654e-05, + "loss": 1.0378, + "step": 15631 + }, + { + "epoch": 0.37, + "grad_norm": 1.8054798587627958, + "learning_rate": 1.4575358343264628e-05, + "loss": 1.0801, + "step": 15632 + }, + { + "epoch": 0.37, + "grad_norm": 2.0763782744479005, + "learning_rate": 1.4574679840562417e-05, + "loss": 0.985, + "step": 15633 + }, + { + "epoch": 0.37, + "grad_norm": 2.3789702468126652, + "learning_rate": 1.4574001311224976e-05, + "loss": 0.9693, + "step": 15634 + }, + { + "epoch": 0.37, + "grad_norm": 2.1332313563845564, + "learning_rate": 1.4573322755256245e-05, + "loss": 1.0325, + "step": 15635 + }, + { + "epoch": 0.37, + "grad_norm": 1.7822271023710294, + "learning_rate": 1.4572644172660186e-05, + "loss": 1.0872, + "step": 15636 + }, + { + "epoch": 0.37, + "grad_norm": 2.048981885633714, + "learning_rate": 1.457196556344074e-05, + "loss": 1.044, + "step": 15637 + }, + { + "epoch": 0.37, + "grad_norm": 2.0162320037396433, + "learning_rate": 1.4571286927601868e-05, + "loss": 1.0114, + "step": 15638 + }, + { + "epoch": 0.37, + "grad_norm": 1.1654797241845223, + "learning_rate": 1.4570608265147512e-05, + "loss": 0.9174, + "step": 15639 + }, + { + "epoch": 0.37, + "grad_norm": 2.1739609972042673, + "learning_rate": 1.4569929576081633e-05, + "loss": 1.0111, + "step": 15640 + }, + { + "epoch": 0.37, + "grad_norm": 1.879194119373213, + "learning_rate": 1.4569250860408173e-05, + "loss": 0.9734, + "step": 15641 + }, + { + "epoch": 0.37, + "grad_norm": 2.253108618963747, + "learning_rate": 1.4568572118131087e-05, + "loss": 1.0094, + "step": 15642 + }, + { + "epoch": 0.37, + "grad_norm": 2.0176417285004935, + "learning_rate": 1.456789334925433e-05, + "loss": 0.8693, + "step": 15643 + }, + { + "epoch": 0.37, + "grad_norm": 1.0403920306739585, + "learning_rate": 1.456721455378185e-05, + "loss": 0.9062, + "step": 15644 + }, + { + "epoch": 0.37, + "grad_norm": 1.8997886951418945, + "learning_rate": 1.4566535731717603e-05, + "loss": 1.0499, + "step": 15645 + }, + { + "epoch": 0.37, + "grad_norm": 2.135737088912055, + "learning_rate": 1.456585688306554e-05, + "loss": 1.0476, + "step": 15646 + }, + { + "epoch": 0.37, + "grad_norm": 2.0352850797604214, + "learning_rate": 1.456517800782961e-05, + "loss": 1.1211, + "step": 15647 + }, + { + "epoch": 0.37, + "grad_norm": 2.2098666357249015, + "learning_rate": 1.4564499106013768e-05, + "loss": 0.9578, + "step": 15648 + }, + { + "epoch": 0.37, + "grad_norm": 1.9562816121629107, + "learning_rate": 1.4563820177621968e-05, + "loss": 1.078, + "step": 15649 + }, + { + "epoch": 0.37, + "grad_norm": 2.130924658950968, + "learning_rate": 1.4563141222658163e-05, + "loss": 0.9648, + "step": 15650 + }, + { + "epoch": 0.37, + "grad_norm": 2.1121761475365695, + "learning_rate": 1.4562462241126305e-05, + "loss": 1.04, + "step": 15651 + }, + { + "epoch": 0.37, + "grad_norm": 2.0620254769748128, + "learning_rate": 1.4561783233030346e-05, + "loss": 0.9598, + "step": 15652 + }, + { + "epoch": 0.37, + "grad_norm": 2.1473525116641867, + "learning_rate": 1.4561104198374243e-05, + "loss": 1.0121, + "step": 15653 + }, + { + "epoch": 0.37, + "grad_norm": 1.7903723697782028, + "learning_rate": 1.4560425137161946e-05, + "loss": 0.982, + "step": 15654 + }, + { + "epoch": 0.37, + "grad_norm": 2.1592073681401494, + "learning_rate": 1.4559746049397411e-05, + "loss": 1.024, + "step": 15655 + }, + { + "epoch": 0.37, + "grad_norm": 2.749320030474236, + "learning_rate": 1.455906693508459e-05, + "loss": 1.0332, + "step": 15656 + }, + { + "epoch": 0.37, + "grad_norm": 1.9925757926143164, + "learning_rate": 1.4558387794227438e-05, + "loss": 0.9545, + "step": 15657 + }, + { + "epoch": 0.37, + "grad_norm": 1.972025519513378, + "learning_rate": 1.4557708626829907e-05, + "loss": 1.0019, + "step": 15658 + }, + { + "epoch": 0.37, + "grad_norm": 1.07917498709876, + "learning_rate": 1.455702943289596e-05, + "loss": 0.9644, + "step": 15659 + }, + { + "epoch": 0.37, + "grad_norm": 2.0292303138777226, + "learning_rate": 1.4556350212429538e-05, + "loss": 1.0024, + "step": 15660 + }, + { + "epoch": 0.37, + "grad_norm": 3.001047274305441, + "learning_rate": 1.4555670965434602e-05, + "loss": 0.9001, + "step": 15661 + }, + { + "epoch": 0.37, + "grad_norm": 2.253885383629197, + "learning_rate": 1.4554991691915115e-05, + "loss": 1.0969, + "step": 15662 + }, + { + "epoch": 0.37, + "grad_norm": 2.0108748049955274, + "learning_rate": 1.4554312391875016e-05, + "loss": 1.0589, + "step": 15663 + }, + { + "epoch": 0.37, + "grad_norm": 2.8138481153490655, + "learning_rate": 1.4553633065318274e-05, + "loss": 1.1393, + "step": 15664 + }, + { + "epoch": 0.37, + "grad_norm": 3.037851184762099, + "learning_rate": 1.4552953712248834e-05, + "loss": 1.1525, + "step": 15665 + }, + { + "epoch": 0.37, + "grad_norm": 2.095358420877509, + "learning_rate": 1.4552274332670662e-05, + "loss": 0.9804, + "step": 15666 + }, + { + "epoch": 0.37, + "grad_norm": 1.085827105085742, + "learning_rate": 1.4551594926587703e-05, + "loss": 1.0295, + "step": 15667 + }, + { + "epoch": 0.37, + "grad_norm": 1.2215620471381887, + "learning_rate": 1.4550915494003919e-05, + "loss": 0.9571, + "step": 15668 + }, + { + "epoch": 0.37, + "grad_norm": 1.9195054314231224, + "learning_rate": 1.4550236034923263e-05, + "loss": 1.0637, + "step": 15669 + }, + { + "epoch": 0.37, + "grad_norm": 2.0067537155733843, + "learning_rate": 1.4549556549349694e-05, + "loss": 0.9558, + "step": 15670 + }, + { + "epoch": 0.37, + "grad_norm": 2.1147334830155398, + "learning_rate": 1.4548877037287166e-05, + "loss": 1.0146, + "step": 15671 + }, + { + "epoch": 0.37, + "grad_norm": 2.131721871199417, + "learning_rate": 1.4548197498739636e-05, + "loss": 1.039, + "step": 15672 + }, + { + "epoch": 0.37, + "grad_norm": 1.8425265790212402, + "learning_rate": 1.4547517933711058e-05, + "loss": 1.001, + "step": 15673 + }, + { + "epoch": 0.37, + "grad_norm": 2.543570419207365, + "learning_rate": 1.4546838342205392e-05, + "loss": 1.0664, + "step": 15674 + }, + { + "epoch": 0.37, + "grad_norm": 2.015916274765823, + "learning_rate": 1.4546158724226594e-05, + "loss": 0.8954, + "step": 15675 + }, + { + "epoch": 0.37, + "grad_norm": 1.112148720357971, + "learning_rate": 1.4545479079778622e-05, + "loss": 0.9933, + "step": 15676 + }, + { + "epoch": 0.37, + "grad_norm": 2.2909502186142143, + "learning_rate": 1.4544799408865428e-05, + "loss": 1.0654, + "step": 15677 + }, + { + "epoch": 0.37, + "grad_norm": 1.9459700818280026, + "learning_rate": 1.4544119711490976e-05, + "loss": 1.1068, + "step": 15678 + }, + { + "epoch": 0.37, + "grad_norm": 1.0567346485210518, + "learning_rate": 1.4543439987659221e-05, + "loss": 0.994, + "step": 15679 + }, + { + "epoch": 0.37, + "grad_norm": 2.18318223365015, + "learning_rate": 1.4542760237374119e-05, + "loss": 1.0845, + "step": 15680 + }, + { + "epoch": 0.37, + "grad_norm": 1.9448490242402132, + "learning_rate": 1.4542080460639627e-05, + "loss": 1.0576, + "step": 15681 + }, + { + "epoch": 0.37, + "grad_norm": 2.0333313361024397, + "learning_rate": 1.4541400657459705e-05, + "loss": 0.8747, + "step": 15682 + }, + { + "epoch": 0.37, + "grad_norm": 2.118136948890792, + "learning_rate": 1.4540720827838313e-05, + "loss": 0.9162, + "step": 15683 + }, + { + "epoch": 0.37, + "grad_norm": 1.9610200071811135, + "learning_rate": 1.4540040971779403e-05, + "loss": 0.9415, + "step": 15684 + }, + { + "epoch": 0.37, + "grad_norm": 2.355948813330198, + "learning_rate": 1.4539361089286942e-05, + "loss": 1.0532, + "step": 15685 + }, + { + "epoch": 0.37, + "grad_norm": 1.8348100035768, + "learning_rate": 1.4538681180364878e-05, + "loss": 1.0312, + "step": 15686 + }, + { + "epoch": 0.37, + "grad_norm": 2.3566924684648103, + "learning_rate": 1.4538001245017182e-05, + "loss": 1.0786, + "step": 15687 + }, + { + "epoch": 0.37, + "grad_norm": 1.8863116181342172, + "learning_rate": 1.4537321283247799e-05, + "loss": 1.015, + "step": 15688 + }, + { + "epoch": 0.37, + "grad_norm": 2.03945529879542, + "learning_rate": 1.4536641295060701e-05, + "loss": 1.0453, + "step": 15689 + }, + { + "epoch": 0.37, + "grad_norm": 1.8754439037711854, + "learning_rate": 1.4535961280459838e-05, + "loss": 0.9728, + "step": 15690 + }, + { + "epoch": 0.37, + "grad_norm": 2.0287334120645775, + "learning_rate": 1.4535281239449173e-05, + "loss": 1.1857, + "step": 15691 + }, + { + "epoch": 0.37, + "grad_norm": 2.0406705733485446, + "learning_rate": 1.4534601172032663e-05, + "loss": 0.9584, + "step": 15692 + }, + { + "epoch": 0.37, + "grad_norm": 2.177296448198224, + "learning_rate": 1.4533921078214271e-05, + "loss": 1.0742, + "step": 15693 + }, + { + "epoch": 0.37, + "grad_norm": 2.262932471146134, + "learning_rate": 1.4533240957997956e-05, + "loss": 1.1317, + "step": 15694 + }, + { + "epoch": 0.37, + "grad_norm": 1.8004178850173571, + "learning_rate": 1.4532560811387674e-05, + "loss": 1.0158, + "step": 15695 + }, + { + "epoch": 0.37, + "grad_norm": 1.834628230055565, + "learning_rate": 1.453188063838739e-05, + "loss": 1.043, + "step": 15696 + }, + { + "epoch": 0.37, + "grad_norm": 1.140899186480533, + "learning_rate": 1.4531200439001063e-05, + "loss": 1.0084, + "step": 15697 + }, + { + "epoch": 0.37, + "grad_norm": 2.5096361040117663, + "learning_rate": 1.453052021323265e-05, + "loss": 1.1578, + "step": 15698 + }, + { + "epoch": 0.37, + "grad_norm": 2.161257294979651, + "learning_rate": 1.4529839961086116e-05, + "loss": 1.0342, + "step": 15699 + }, + { + "epoch": 0.37, + "grad_norm": 1.8704319239001876, + "learning_rate": 1.452915968256542e-05, + "loss": 1.0142, + "step": 15700 + }, + { + "epoch": 0.37, + "grad_norm": 2.124434750731433, + "learning_rate": 1.4528479377674521e-05, + "loss": 1.0271, + "step": 15701 + }, + { + "epoch": 0.37, + "grad_norm": 2.328774461429299, + "learning_rate": 1.452779904641738e-05, + "loss": 1.1126, + "step": 15702 + }, + { + "epoch": 0.37, + "grad_norm": 1.9591786673553355, + "learning_rate": 1.4527118688797963e-05, + "loss": 0.9408, + "step": 15703 + }, + { + "epoch": 0.37, + "grad_norm": 1.966696417003545, + "learning_rate": 1.4526438304820225e-05, + "loss": 1.0448, + "step": 15704 + }, + { + "epoch": 0.37, + "grad_norm": 2.091647523023635, + "learning_rate": 1.452575789448813e-05, + "loss": 1.127, + "step": 15705 + }, + { + "epoch": 0.37, + "grad_norm": 2.2998433551250863, + "learning_rate": 1.4525077457805646e-05, + "loss": 0.9585, + "step": 15706 + }, + { + "epoch": 0.37, + "grad_norm": 1.843218972045184, + "learning_rate": 1.452439699477672e-05, + "loss": 1.0002, + "step": 15707 + }, + { + "epoch": 0.37, + "grad_norm": 1.854830247289457, + "learning_rate": 1.4523716505405327e-05, + "loss": 0.9755, + "step": 15708 + }, + { + "epoch": 0.37, + "grad_norm": 2.215006471010181, + "learning_rate": 1.4523035989695422e-05, + "loss": 1.1926, + "step": 15709 + }, + { + "epoch": 0.37, + "grad_norm": 2.1637133451577766, + "learning_rate": 1.452235544765097e-05, + "loss": 1.0446, + "step": 15710 + }, + { + "epoch": 0.37, + "grad_norm": 2.163885994949197, + "learning_rate": 1.4521674879275933e-05, + "loss": 1.1025, + "step": 15711 + }, + { + "epoch": 0.37, + "grad_norm": 2.037503456623087, + "learning_rate": 1.4520994284574275e-05, + "loss": 1.0035, + "step": 15712 + }, + { + "epoch": 0.37, + "grad_norm": 2.2219178945568827, + "learning_rate": 1.4520313663549957e-05, + "loss": 0.9918, + "step": 15713 + }, + { + "epoch": 0.37, + "grad_norm": 2.0219984896966428, + "learning_rate": 1.4519633016206938e-05, + "loss": 1.0943, + "step": 15714 + }, + { + "epoch": 0.37, + "grad_norm": 1.8568707600188536, + "learning_rate": 1.4518952342549189e-05, + "loss": 1.0719, + "step": 15715 + }, + { + "epoch": 0.37, + "grad_norm": 1.1182666352546808, + "learning_rate": 1.4518271642580663e-05, + "loss": 0.998, + "step": 15716 + }, + { + "epoch": 0.37, + "grad_norm": 1.9000856097421257, + "learning_rate": 1.4517590916305335e-05, + "loss": 1.0394, + "step": 15717 + }, + { + "epoch": 0.37, + "grad_norm": 2.033681902914752, + "learning_rate": 1.4516910163727158e-05, + "loss": 1.1092, + "step": 15718 + }, + { + "epoch": 0.37, + "grad_norm": 2.5065518172878485, + "learning_rate": 1.4516229384850105e-05, + "loss": 1.0832, + "step": 15719 + }, + { + "epoch": 0.37, + "grad_norm": 2.1997589483601616, + "learning_rate": 1.451554857967813e-05, + "loss": 1.0362, + "step": 15720 + }, + { + "epoch": 0.37, + "grad_norm": 2.023075354761978, + "learning_rate": 1.4514867748215204e-05, + "loss": 1.0006, + "step": 15721 + }, + { + "epoch": 0.37, + "grad_norm": 1.818308854015868, + "learning_rate": 1.4514186890465289e-05, + "loss": 0.9668, + "step": 15722 + }, + { + "epoch": 0.37, + "grad_norm": 1.9018763610808804, + "learning_rate": 1.4513506006432347e-05, + "loss": 0.9585, + "step": 15723 + }, + { + "epoch": 0.37, + "grad_norm": 2.0501723277844204, + "learning_rate": 1.4512825096120346e-05, + "loss": 1.1348, + "step": 15724 + }, + { + "epoch": 0.37, + "grad_norm": 1.970443269541463, + "learning_rate": 1.4512144159533248e-05, + "loss": 1.0248, + "step": 15725 + }, + { + "epoch": 0.37, + "grad_norm": 2.850071794844477, + "learning_rate": 1.4511463196675018e-05, + "loss": 0.891, + "step": 15726 + }, + { + "epoch": 0.37, + "grad_norm": 2.2087093323491476, + "learning_rate": 1.4510782207549619e-05, + "loss": 1.0344, + "step": 15727 + }, + { + "epoch": 0.37, + "grad_norm": 2.397031031811012, + "learning_rate": 1.451010119216102e-05, + "loss": 1.0923, + "step": 15728 + }, + { + "epoch": 0.37, + "grad_norm": 2.2804849106915834, + "learning_rate": 1.4509420150513183e-05, + "loss": 0.9545, + "step": 15729 + }, + { + "epoch": 0.37, + "grad_norm": 3.2658428670307273, + "learning_rate": 1.4508739082610075e-05, + "loss": 0.939, + "step": 15730 + }, + { + "epoch": 0.37, + "grad_norm": 2.0609074567489603, + "learning_rate": 1.4508057988455659e-05, + "loss": 0.9553, + "step": 15731 + }, + { + "epoch": 0.37, + "grad_norm": 2.2115295464354583, + "learning_rate": 1.4507376868053903e-05, + "loss": 1.1041, + "step": 15732 + }, + { + "epoch": 0.37, + "grad_norm": 2.2977081492999907, + "learning_rate": 1.4506695721408775e-05, + "loss": 0.9053, + "step": 15733 + }, + { + "epoch": 0.37, + "grad_norm": 2.1840198543920746, + "learning_rate": 1.4506014548524237e-05, + "loss": 0.9547, + "step": 15734 + }, + { + "epoch": 0.37, + "grad_norm": 2.1736837886051865, + "learning_rate": 1.4505333349404252e-05, + "loss": 0.9804, + "step": 15735 + }, + { + "epoch": 0.37, + "grad_norm": 1.9543434507606008, + "learning_rate": 1.4504652124052792e-05, + "loss": 0.9958, + "step": 15736 + }, + { + "epoch": 0.37, + "grad_norm": 1.9348111325304675, + "learning_rate": 1.450397087247382e-05, + "loss": 0.9313, + "step": 15737 + }, + { + "epoch": 0.37, + "grad_norm": 1.93210492741019, + "learning_rate": 1.4503289594671306e-05, + "loss": 1.0993, + "step": 15738 + }, + { + "epoch": 0.37, + "grad_norm": 2.092746742570468, + "learning_rate": 1.450260829064921e-05, + "loss": 0.9655, + "step": 15739 + }, + { + "epoch": 0.37, + "grad_norm": 2.2381170151824272, + "learning_rate": 1.4501926960411508e-05, + "loss": 0.9006, + "step": 15740 + }, + { + "epoch": 0.37, + "grad_norm": 1.888981592322705, + "learning_rate": 1.450124560396216e-05, + "loss": 0.9922, + "step": 15741 + }, + { + "epoch": 0.37, + "grad_norm": 2.4153542968457273, + "learning_rate": 1.4500564221305133e-05, + "loss": 0.9466, + "step": 15742 + }, + { + "epoch": 0.37, + "grad_norm": 2.070647069600114, + "learning_rate": 1.44998828124444e-05, + "loss": 0.9583, + "step": 15743 + }, + { + "epoch": 0.37, + "grad_norm": 1.8644583947874283, + "learning_rate": 1.4499201377383923e-05, + "loss": 0.9601, + "step": 15744 + }, + { + "epoch": 0.37, + "grad_norm": 2.338819126798084, + "learning_rate": 1.449851991612767e-05, + "loss": 0.9065, + "step": 15745 + }, + { + "epoch": 0.37, + "grad_norm": 2.1560839752362995, + "learning_rate": 1.449783842867961e-05, + "loss": 1.0808, + "step": 15746 + }, + { + "epoch": 0.37, + "grad_norm": 1.924307766899572, + "learning_rate": 1.4497156915043712e-05, + "loss": 0.9854, + "step": 15747 + }, + { + "epoch": 0.37, + "grad_norm": 1.9203426329572206, + "learning_rate": 1.4496475375223941e-05, + "loss": 1.1455, + "step": 15748 + }, + { + "epoch": 0.37, + "grad_norm": 1.8672914014098079, + "learning_rate": 1.4495793809224268e-05, + "loss": 0.9924, + "step": 15749 + }, + { + "epoch": 0.37, + "grad_norm": 1.9582366845268433, + "learning_rate": 1.449511221704866e-05, + "loss": 1.0749, + "step": 15750 + }, + { + "epoch": 0.37, + "grad_norm": 2.5766276276541022, + "learning_rate": 1.4494430598701087e-05, + "loss": 0.9794, + "step": 15751 + }, + { + "epoch": 0.37, + "grad_norm": 1.9826551277839675, + "learning_rate": 1.4493748954185513e-05, + "loss": 0.9956, + "step": 15752 + }, + { + "epoch": 0.37, + "grad_norm": 2.2030893853556313, + "learning_rate": 1.449306728350591e-05, + "loss": 1.0275, + "step": 15753 + }, + { + "epoch": 0.37, + "grad_norm": 2.097844144284023, + "learning_rate": 1.4492385586666249e-05, + "loss": 1.1271, + "step": 15754 + }, + { + "epoch": 0.37, + "grad_norm": 2.0067322032459014, + "learning_rate": 1.4491703863670496e-05, + "loss": 1.0378, + "step": 15755 + }, + { + "epoch": 0.37, + "grad_norm": 2.4420551825204684, + "learning_rate": 1.4491022114522621e-05, + "loss": 1.0063, + "step": 15756 + }, + { + "epoch": 0.37, + "grad_norm": 2.2412753816867848, + "learning_rate": 1.4490340339226596e-05, + "loss": 1.0218, + "step": 15757 + }, + { + "epoch": 0.37, + "grad_norm": 2.106672530504711, + "learning_rate": 1.4489658537786385e-05, + "loss": 0.9551, + "step": 15758 + }, + { + "epoch": 0.37, + "grad_norm": 2.001200757655011, + "learning_rate": 1.448897671020596e-05, + "loss": 1.0893, + "step": 15759 + }, + { + "epoch": 0.37, + "grad_norm": 2.0819708957879572, + "learning_rate": 1.4488294856489291e-05, + "loss": 1.0111, + "step": 15760 + }, + { + "epoch": 0.37, + "grad_norm": 2.148347590147388, + "learning_rate": 1.448761297664035e-05, + "loss": 1.1411, + "step": 15761 + }, + { + "epoch": 0.37, + "grad_norm": 1.843096966738463, + "learning_rate": 1.4486931070663107e-05, + "loss": 0.9171, + "step": 15762 + }, + { + "epoch": 0.37, + "grad_norm": 1.9739076416777064, + "learning_rate": 1.4486249138561528e-05, + "loss": 1.0364, + "step": 15763 + }, + { + "epoch": 0.37, + "grad_norm": 2.0175877848529646, + "learning_rate": 1.448556718033959e-05, + "loss": 1.0136, + "step": 15764 + }, + { + "epoch": 0.37, + "grad_norm": 2.652204857533796, + "learning_rate": 1.4484885196001255e-05, + "loss": 0.979, + "step": 15765 + }, + { + "epoch": 0.37, + "grad_norm": 1.1489242680560328, + "learning_rate": 1.4484203185550503e-05, + "loss": 1.0363, + "step": 15766 + }, + { + "epoch": 0.37, + "grad_norm": 1.8860283136478226, + "learning_rate": 1.4483521148991295e-05, + "loss": 1.0163, + "step": 15767 + }, + { + "epoch": 0.37, + "grad_norm": 2.2079673931685773, + "learning_rate": 1.4482839086327614e-05, + "loss": 1.0422, + "step": 15768 + }, + { + "epoch": 0.37, + "grad_norm": 1.0812525408775595, + "learning_rate": 1.4482156997563417e-05, + "loss": 0.9109, + "step": 15769 + }, + { + "epoch": 0.37, + "grad_norm": 1.170672798806086, + "learning_rate": 1.4481474882702688e-05, + "loss": 1.0831, + "step": 15770 + }, + { + "epoch": 0.37, + "grad_norm": 1.0387616065824834, + "learning_rate": 1.4480792741749395e-05, + "loss": 0.9642, + "step": 15771 + }, + { + "epoch": 0.37, + "grad_norm": 1.9731757564618042, + "learning_rate": 1.4480110574707504e-05, + "loss": 1.0869, + "step": 15772 + }, + { + "epoch": 0.37, + "grad_norm": 2.023648168956871, + "learning_rate": 1.4479428381580993e-05, + "loss": 1.1273, + "step": 15773 + }, + { + "epoch": 0.37, + "grad_norm": 1.9776087869314518, + "learning_rate": 1.447874616237383e-05, + "loss": 1.0502, + "step": 15774 + }, + { + "epoch": 0.37, + "grad_norm": 1.7998660838293552, + "learning_rate": 1.447806391708999e-05, + "loss": 1.0191, + "step": 15775 + }, + { + "epoch": 0.37, + "grad_norm": 2.13389137380685, + "learning_rate": 1.4477381645733446e-05, + "loss": 0.9021, + "step": 15776 + }, + { + "epoch": 0.37, + "grad_norm": 2.2969727166954597, + "learning_rate": 1.4476699348308165e-05, + "loss": 1.0766, + "step": 15777 + }, + { + "epoch": 0.37, + "grad_norm": 1.2030710159458193, + "learning_rate": 1.4476017024818126e-05, + "loss": 0.9717, + "step": 15778 + }, + { + "epoch": 0.37, + "grad_norm": 1.929944068149653, + "learning_rate": 1.4475334675267296e-05, + "loss": 0.9998, + "step": 15779 + }, + { + "epoch": 0.37, + "grad_norm": 1.9141851762948598, + "learning_rate": 1.4474652299659653e-05, + "loss": 0.9389, + "step": 15780 + }, + { + "epoch": 0.37, + "grad_norm": 1.8472881934720697, + "learning_rate": 1.4473969897999168e-05, + "loss": 1.0973, + "step": 15781 + }, + { + "epoch": 0.37, + "grad_norm": 1.0480172631985298, + "learning_rate": 1.4473287470289811e-05, + "loss": 0.9295, + "step": 15782 + }, + { + "epoch": 0.37, + "grad_norm": 2.328836544921858, + "learning_rate": 1.447260501653556e-05, + "loss": 0.9662, + "step": 15783 + }, + { + "epoch": 0.37, + "grad_norm": 2.0531870131284835, + "learning_rate": 1.4471922536740385e-05, + "loss": 1.0339, + "step": 15784 + }, + { + "epoch": 0.37, + "grad_norm": 1.717694247430802, + "learning_rate": 1.4471240030908265e-05, + "loss": 1.0571, + "step": 15785 + }, + { + "epoch": 0.37, + "grad_norm": 1.973946429818704, + "learning_rate": 1.4470557499043167e-05, + "loss": 1.0449, + "step": 15786 + }, + { + "epoch": 0.37, + "grad_norm": 2.3428750687928352, + "learning_rate": 1.4469874941149069e-05, + "loss": 0.8773, + "step": 15787 + }, + { + "epoch": 0.37, + "grad_norm": 2.247631162055752, + "learning_rate": 1.4469192357229942e-05, + "loss": 1.0723, + "step": 15788 + }, + { + "epoch": 0.37, + "grad_norm": 1.9563088047254498, + "learning_rate": 1.4468509747289764e-05, + "loss": 0.9679, + "step": 15789 + }, + { + "epoch": 0.37, + "grad_norm": 1.9475154281850466, + "learning_rate": 1.4467827111332506e-05, + "loss": 0.9451, + "step": 15790 + }, + { + "epoch": 0.37, + "grad_norm": 1.2199994723745016, + "learning_rate": 1.4467144449362147e-05, + "loss": 0.9671, + "step": 15791 + }, + { + "epoch": 0.37, + "grad_norm": 2.4754922822771483, + "learning_rate": 1.4466461761382657e-05, + "loss": 0.8918, + "step": 15792 + }, + { + "epoch": 0.37, + "grad_norm": 1.109946066785129, + "learning_rate": 1.4465779047398014e-05, + "loss": 0.9435, + "step": 15793 + }, + { + "epoch": 0.37, + "grad_norm": 5.61688181328217, + "learning_rate": 1.446509630741219e-05, + "loss": 1.0277, + "step": 15794 + }, + { + "epoch": 0.37, + "grad_norm": 1.9290208190947473, + "learning_rate": 1.4464413541429161e-05, + "loss": 1.1105, + "step": 15795 + }, + { + "epoch": 0.37, + "grad_norm": 2.243375826743258, + "learning_rate": 1.4463730749452904e-05, + "loss": 1.0845, + "step": 15796 + }, + { + "epoch": 0.37, + "grad_norm": 1.1118648428034854, + "learning_rate": 1.4463047931487391e-05, + "loss": 0.9553, + "step": 15797 + }, + { + "epoch": 0.37, + "grad_norm": 1.8252428676488226, + "learning_rate": 1.4462365087536603e-05, + "loss": 1.0603, + "step": 15798 + }, + { + "epoch": 0.37, + "grad_norm": 1.8175698886237297, + "learning_rate": 1.446168221760451e-05, + "loss": 1.0625, + "step": 15799 + }, + { + "epoch": 0.37, + "grad_norm": 3.75483448706832, + "learning_rate": 1.4460999321695095e-05, + "loss": 1.065, + "step": 15800 + }, + { + "epoch": 0.37, + "grad_norm": 1.1220174308845408, + "learning_rate": 1.4460316399812326e-05, + "loss": 0.9165, + "step": 15801 + }, + { + "epoch": 0.37, + "grad_norm": 1.9572539939200941, + "learning_rate": 1.4459633451960184e-05, + "loss": 1.0143, + "step": 15802 + }, + { + "epoch": 0.37, + "grad_norm": 2.153229249923107, + "learning_rate": 1.4458950478142643e-05, + "loss": 1.0373, + "step": 15803 + }, + { + "epoch": 0.37, + "grad_norm": 2.09039897472898, + "learning_rate": 1.4458267478363683e-05, + "loss": 1.028, + "step": 15804 + }, + { + "epoch": 0.37, + "grad_norm": 1.940876174812667, + "learning_rate": 1.4457584452627276e-05, + "loss": 1.0606, + "step": 15805 + }, + { + "epoch": 0.37, + "grad_norm": 1.0592766475683995, + "learning_rate": 1.4456901400937404e-05, + "loss": 0.8751, + "step": 15806 + }, + { + "epoch": 0.37, + "grad_norm": 2.1849019314104345, + "learning_rate": 1.4456218323298037e-05, + "loss": 0.9117, + "step": 15807 + }, + { + "epoch": 0.37, + "grad_norm": 2.8425496258410488, + "learning_rate": 1.4455535219713157e-05, + "loss": 0.951, + "step": 15808 + }, + { + "epoch": 0.37, + "grad_norm": 2.5307600753494746, + "learning_rate": 1.4454852090186742e-05, + "loss": 1.0519, + "step": 15809 + }, + { + "epoch": 0.37, + "grad_norm": 2.2151745978208908, + "learning_rate": 1.4454168934722765e-05, + "loss": 1.1421, + "step": 15810 + }, + { + "epoch": 0.37, + "grad_norm": 2.1320018697180614, + "learning_rate": 1.445348575332521e-05, + "loss": 1.0518, + "step": 15811 + }, + { + "epoch": 0.37, + "grad_norm": 1.0940038409885253, + "learning_rate": 1.4452802545998048e-05, + "loss": 0.9485, + "step": 15812 + }, + { + "epoch": 0.37, + "grad_norm": 1.084365557758656, + "learning_rate": 1.4452119312745266e-05, + "loss": 1.0472, + "step": 15813 + }, + { + "epoch": 0.37, + "grad_norm": 2.0921383454573648, + "learning_rate": 1.4451436053570826e-05, + "loss": 0.9745, + "step": 15814 + }, + { + "epoch": 0.37, + "grad_norm": 2.5824069611481244, + "learning_rate": 1.4450752768478725e-05, + "loss": 1.0588, + "step": 15815 + }, + { + "epoch": 0.37, + "grad_norm": 2.263652644989085, + "learning_rate": 1.4450069457472926e-05, + "loss": 0.9825, + "step": 15816 + }, + { + "epoch": 0.37, + "grad_norm": 2.364796702339441, + "learning_rate": 1.4449386120557418e-05, + "loss": 1.0057, + "step": 15817 + }, + { + "epoch": 0.37, + "grad_norm": 2.408055728209296, + "learning_rate": 1.4448702757736172e-05, + "loss": 0.9441, + "step": 15818 + }, + { + "epoch": 0.37, + "grad_norm": 2.138256305715726, + "learning_rate": 1.4448019369013173e-05, + "loss": 0.9147, + "step": 15819 + }, + { + "epoch": 0.37, + "grad_norm": 2.2121911346927994, + "learning_rate": 1.4447335954392393e-05, + "loss": 1.1179, + "step": 15820 + }, + { + "epoch": 0.37, + "grad_norm": 1.089175782715359, + "learning_rate": 1.4446652513877818e-05, + "loss": 0.9857, + "step": 15821 + }, + { + "epoch": 0.37, + "grad_norm": 1.145972406055561, + "learning_rate": 1.4445969047473427e-05, + "loss": 1.0214, + "step": 15822 + }, + { + "epoch": 0.37, + "grad_norm": 1.844605200421973, + "learning_rate": 1.4445285555183193e-05, + "loss": 0.994, + "step": 15823 + }, + { + "epoch": 0.37, + "grad_norm": 2.079934401091317, + "learning_rate": 1.4444602037011099e-05, + "loss": 0.941, + "step": 15824 + }, + { + "epoch": 0.37, + "grad_norm": 1.1343923546134145, + "learning_rate": 1.4443918492961127e-05, + "loss": 0.9959, + "step": 15825 + }, + { + "epoch": 0.37, + "grad_norm": 2.16539595838651, + "learning_rate": 1.4443234923037253e-05, + "loss": 1.1177, + "step": 15826 + }, + { + "epoch": 0.37, + "grad_norm": 2.1982403264361943, + "learning_rate": 1.4442551327243456e-05, + "loss": 0.9832, + "step": 15827 + }, + { + "epoch": 0.37, + "grad_norm": 1.9210404886308456, + "learning_rate": 1.4441867705583723e-05, + "loss": 1.0205, + "step": 15828 + }, + { + "epoch": 0.37, + "grad_norm": 2.149363524424723, + "learning_rate": 1.4441184058062028e-05, + "loss": 1.0709, + "step": 15829 + }, + { + "epoch": 0.37, + "grad_norm": 1.9065221590749895, + "learning_rate": 1.4440500384682353e-05, + "loss": 1.0047, + "step": 15830 + }, + { + "epoch": 0.37, + "grad_norm": 1.9417851102856665, + "learning_rate": 1.443981668544868e-05, + "loss": 1.006, + "step": 15831 + }, + { + "epoch": 0.37, + "grad_norm": 2.174593045651493, + "learning_rate": 1.4439132960364988e-05, + "loss": 1.2107, + "step": 15832 + }, + { + "epoch": 0.37, + "grad_norm": 2.139291060316537, + "learning_rate": 1.4438449209435259e-05, + "loss": 1.038, + "step": 15833 + }, + { + "epoch": 0.37, + "grad_norm": 3.370970031279258, + "learning_rate": 1.4437765432663472e-05, + "loss": 0.9185, + "step": 15834 + }, + { + "epoch": 0.37, + "grad_norm": 2.249637047770816, + "learning_rate": 1.4437081630053614e-05, + "loss": 1.0115, + "step": 15835 + }, + { + "epoch": 0.37, + "grad_norm": 2.586080490020488, + "learning_rate": 1.4436397801609658e-05, + "loss": 0.9042, + "step": 15836 + }, + { + "epoch": 0.37, + "grad_norm": 1.9359229304244179, + "learning_rate": 1.443571394733559e-05, + "loss": 0.9534, + "step": 15837 + }, + { + "epoch": 0.37, + "grad_norm": 2.378888413463398, + "learning_rate": 1.4435030067235394e-05, + "loss": 0.9658, + "step": 15838 + }, + { + "epoch": 0.37, + "grad_norm": 1.1523483857166668, + "learning_rate": 1.4434346161313041e-05, + "loss": 0.9538, + "step": 15839 + }, + { + "epoch": 0.37, + "grad_norm": 1.8865124970298657, + "learning_rate": 1.4433662229572529e-05, + "loss": 1.028, + "step": 15840 + }, + { + "epoch": 0.37, + "grad_norm": 1.8574315390751612, + "learning_rate": 1.4432978272017828e-05, + "loss": 0.9299, + "step": 15841 + }, + { + "epoch": 0.37, + "grad_norm": 2.0729689742617112, + "learning_rate": 1.4432294288652926e-05, + "loss": 1.085, + "step": 15842 + }, + { + "epoch": 0.37, + "grad_norm": 2.208383475815794, + "learning_rate": 1.4431610279481802e-05, + "loss": 0.9843, + "step": 15843 + }, + { + "epoch": 0.37, + "grad_norm": 2.6091792988343188, + "learning_rate": 1.4430926244508441e-05, + "loss": 0.9191, + "step": 15844 + }, + { + "epoch": 0.37, + "grad_norm": 2.34837338408588, + "learning_rate": 1.4430242183736824e-05, + "loss": 0.9603, + "step": 15845 + }, + { + "epoch": 0.37, + "grad_norm": 2.026218805108373, + "learning_rate": 1.4429558097170932e-05, + "loss": 0.9531, + "step": 15846 + }, + { + "epoch": 0.37, + "grad_norm": 1.9934391367393227, + "learning_rate": 1.4428873984814755e-05, + "loss": 0.9968, + "step": 15847 + }, + { + "epoch": 0.37, + "grad_norm": 2.110528278321852, + "learning_rate": 1.4428189846672267e-05, + "loss": 0.9881, + "step": 15848 + }, + { + "epoch": 0.37, + "grad_norm": 1.8703422308891569, + "learning_rate": 1.442750568274746e-05, + "loss": 1.1222, + "step": 15849 + }, + { + "epoch": 0.37, + "grad_norm": 2.1613044903288716, + "learning_rate": 1.4426821493044312e-05, + "loss": 1.0876, + "step": 15850 + }, + { + "epoch": 0.37, + "grad_norm": 1.7755074070510448, + "learning_rate": 1.4426137277566805e-05, + "loss": 1.0199, + "step": 15851 + }, + { + "epoch": 0.37, + "grad_norm": 2.2171477680782945, + "learning_rate": 1.4425453036318929e-05, + "loss": 1.0766, + "step": 15852 + }, + { + "epoch": 0.37, + "grad_norm": 1.8723758508869857, + "learning_rate": 1.4424768769304663e-05, + "loss": 1.0546, + "step": 15853 + }, + { + "epoch": 0.37, + "grad_norm": 2.131835313221119, + "learning_rate": 1.4424084476527991e-05, + "loss": 1.0664, + "step": 15854 + }, + { + "epoch": 0.37, + "grad_norm": 1.7844950175988858, + "learning_rate": 1.44234001579929e-05, + "loss": 1.0225, + "step": 15855 + }, + { + "epoch": 0.37, + "grad_norm": 2.1280580305108683, + "learning_rate": 1.4422715813703372e-05, + "loss": 0.896, + "step": 15856 + }, + { + "epoch": 0.37, + "grad_norm": 2.020569487990621, + "learning_rate": 1.4422031443663391e-05, + "loss": 1.0071, + "step": 15857 + }, + { + "epoch": 0.37, + "grad_norm": 2.1663863502348364, + "learning_rate": 1.4421347047876945e-05, + "loss": 0.9899, + "step": 15858 + }, + { + "epoch": 0.37, + "grad_norm": 2.0681641387921426, + "learning_rate": 1.4420662626348017e-05, + "loss": 1.1311, + "step": 15859 + }, + { + "epoch": 0.37, + "grad_norm": 2.130000183738218, + "learning_rate": 1.441997817908059e-05, + "loss": 1.0993, + "step": 15860 + }, + { + "epoch": 0.37, + "grad_norm": 1.9626730667247294, + "learning_rate": 1.4419293706078654e-05, + "loss": 0.9891, + "step": 15861 + }, + { + "epoch": 0.37, + "grad_norm": 2.0430420345907523, + "learning_rate": 1.4418609207346189e-05, + "loss": 0.9032, + "step": 15862 + }, + { + "epoch": 0.37, + "grad_norm": 2.157774195401881, + "learning_rate": 1.4417924682887181e-05, + "loss": 1.0953, + "step": 15863 + }, + { + "epoch": 0.37, + "grad_norm": 2.0044594724568365, + "learning_rate": 1.4417240132705618e-05, + "loss": 1.0203, + "step": 15864 + }, + { + "epoch": 0.37, + "grad_norm": 2.164404953338973, + "learning_rate": 1.4416555556805482e-05, + "loss": 0.9295, + "step": 15865 + }, + { + "epoch": 0.37, + "grad_norm": 1.040064216699111, + "learning_rate": 1.4415870955190767e-05, + "loss": 0.9972, + "step": 15866 + }, + { + "epoch": 0.37, + "grad_norm": 1.9538495405506031, + "learning_rate": 1.441518632786545e-05, + "loss": 1.1553, + "step": 15867 + }, + { + "epoch": 0.37, + "grad_norm": 2.1858817721549477, + "learning_rate": 1.441450167483352e-05, + "loss": 1.0239, + "step": 15868 + }, + { + "epoch": 0.37, + "grad_norm": 1.1624690509016795, + "learning_rate": 1.4413816996098962e-05, + "loss": 0.9452, + "step": 15869 + }, + { + "epoch": 0.37, + "grad_norm": 1.9736476912496819, + "learning_rate": 1.4413132291665768e-05, + "loss": 1.0059, + "step": 15870 + }, + { + "epoch": 0.37, + "grad_norm": 1.8988663044774408, + "learning_rate": 1.4412447561537919e-05, + "loss": 1.0641, + "step": 15871 + }, + { + "epoch": 0.37, + "grad_norm": 2.3943744102635436, + "learning_rate": 1.4411762805719402e-05, + "loss": 1.1057, + "step": 15872 + }, + { + "epoch": 0.37, + "grad_norm": 2.0260258542256304, + "learning_rate": 1.4411078024214207e-05, + "loss": 1.0253, + "step": 15873 + }, + { + "epoch": 0.37, + "grad_norm": 1.999909364543725, + "learning_rate": 1.4410393217026317e-05, + "loss": 0.9645, + "step": 15874 + }, + { + "epoch": 0.37, + "grad_norm": 2.166911737625649, + "learning_rate": 1.4409708384159724e-05, + "loss": 1.1218, + "step": 15875 + }, + { + "epoch": 0.37, + "grad_norm": 1.9449651057983268, + "learning_rate": 1.4409023525618411e-05, + "loss": 1.0716, + "step": 15876 + }, + { + "epoch": 0.37, + "grad_norm": 2.0154926347611126, + "learning_rate": 1.4408338641406368e-05, + "loss": 0.9505, + "step": 15877 + }, + { + "epoch": 0.37, + "grad_norm": 2.235928924507948, + "learning_rate": 1.440765373152758e-05, + "loss": 1.1072, + "step": 15878 + }, + { + "epoch": 0.37, + "grad_norm": 2.06098617801884, + "learning_rate": 1.4406968795986038e-05, + "loss": 0.9453, + "step": 15879 + }, + { + "epoch": 0.37, + "grad_norm": 2.254552718570386, + "learning_rate": 1.4406283834785728e-05, + "loss": 0.9684, + "step": 15880 + }, + { + "epoch": 0.37, + "grad_norm": 1.9212051269436092, + "learning_rate": 1.4405598847930639e-05, + "loss": 0.8379, + "step": 15881 + }, + { + "epoch": 0.37, + "grad_norm": 2.177424310840447, + "learning_rate": 1.440491383542476e-05, + "loss": 1.0434, + "step": 15882 + }, + { + "epoch": 0.37, + "grad_norm": 2.0854490309625646, + "learning_rate": 1.4404228797272076e-05, + "loss": 1.1374, + "step": 15883 + }, + { + "epoch": 0.37, + "grad_norm": 1.833785658471705, + "learning_rate": 1.4403543733476578e-05, + "loss": 1.0262, + "step": 15884 + }, + { + "epoch": 0.37, + "grad_norm": 1.169718040404054, + "learning_rate": 1.4402858644042254e-05, + "loss": 0.9873, + "step": 15885 + }, + { + "epoch": 0.37, + "grad_norm": 2.0874265548908153, + "learning_rate": 1.4402173528973094e-05, + "loss": 1.0922, + "step": 15886 + }, + { + "epoch": 0.37, + "grad_norm": 1.9626199140623344, + "learning_rate": 1.4401488388273086e-05, + "loss": 1.0295, + "step": 15887 + }, + { + "epoch": 0.37, + "grad_norm": 2.2454534783201376, + "learning_rate": 1.4400803221946218e-05, + "loss": 1.0771, + "step": 15888 + }, + { + "epoch": 0.37, + "grad_norm": 2.346997865735057, + "learning_rate": 1.4400118029996481e-05, + "loss": 0.9394, + "step": 15889 + }, + { + "epoch": 0.37, + "grad_norm": 1.952096982240108, + "learning_rate": 1.4399432812427862e-05, + "loss": 0.9652, + "step": 15890 + }, + { + "epoch": 0.37, + "grad_norm": 2.0672195205205237, + "learning_rate": 1.4398747569244355e-05, + "loss": 0.9573, + "step": 15891 + }, + { + "epoch": 0.37, + "grad_norm": 1.886540904002754, + "learning_rate": 1.4398062300449947e-05, + "loss": 1.0305, + "step": 15892 + }, + { + "epoch": 0.37, + "grad_norm": 1.9075340663034095, + "learning_rate": 1.4397377006048628e-05, + "loss": 1.0292, + "step": 15893 + }, + { + "epoch": 0.37, + "grad_norm": 2.6845705700857208, + "learning_rate": 1.4396691686044386e-05, + "loss": 0.9664, + "step": 15894 + }, + { + "epoch": 0.37, + "grad_norm": 2.164728492067333, + "learning_rate": 1.4396006340441214e-05, + "loss": 1.0549, + "step": 15895 + }, + { + "epoch": 0.37, + "grad_norm": 2.2378446333504463, + "learning_rate": 1.43953209692431e-05, + "loss": 1.1655, + "step": 15896 + }, + { + "epoch": 0.37, + "grad_norm": 2.048402741413356, + "learning_rate": 1.4394635572454036e-05, + "loss": 1.0885, + "step": 15897 + }, + { + "epoch": 0.37, + "grad_norm": 2.2216972913673425, + "learning_rate": 1.4393950150078016e-05, + "loss": 0.9825, + "step": 15898 + }, + { + "epoch": 0.37, + "grad_norm": 1.1177907564196936, + "learning_rate": 1.4393264702119023e-05, + "loss": 0.9459, + "step": 15899 + }, + { + "epoch": 0.37, + "grad_norm": 2.2544276299297317, + "learning_rate": 1.4392579228581054e-05, + "loss": 1.0364, + "step": 15900 + }, + { + "epoch": 0.37, + "grad_norm": 2.329036613574431, + "learning_rate": 1.4391893729468099e-05, + "loss": 1.0714, + "step": 15901 + }, + { + "epoch": 0.37, + "grad_norm": 2.1462089732434864, + "learning_rate": 1.4391208204784146e-05, + "loss": 1.0235, + "step": 15902 + }, + { + "epoch": 0.37, + "grad_norm": 2.3472446907634805, + "learning_rate": 1.439052265453319e-05, + "loss": 0.9709, + "step": 15903 + }, + { + "epoch": 0.37, + "grad_norm": 1.9492958126238749, + "learning_rate": 1.438983707871922e-05, + "loss": 0.9801, + "step": 15904 + }, + { + "epoch": 0.37, + "grad_norm": 2.184486120022972, + "learning_rate": 1.4389151477346231e-05, + "loss": 0.9224, + "step": 15905 + }, + { + "epoch": 0.37, + "grad_norm": 2.254518653079277, + "learning_rate": 1.438846585041821e-05, + "loss": 0.929, + "step": 15906 + }, + { + "epoch": 0.37, + "grad_norm": 2.0844963063006223, + "learning_rate": 1.4387780197939152e-05, + "loss": 1.1182, + "step": 15907 + }, + { + "epoch": 0.37, + "grad_norm": 1.9092697590608072, + "learning_rate": 1.4387094519913048e-05, + "loss": 0.9723, + "step": 15908 + }, + { + "epoch": 0.37, + "grad_norm": 1.9935923334218402, + "learning_rate": 1.438640881634389e-05, + "loss": 1.0841, + "step": 15909 + }, + { + "epoch": 0.37, + "grad_norm": 1.710612547484478, + "learning_rate": 1.4385723087235674e-05, + "loss": 1.0119, + "step": 15910 + }, + { + "epoch": 0.37, + "grad_norm": 2.2446538217608536, + "learning_rate": 1.438503733259239e-05, + "loss": 1.0325, + "step": 15911 + }, + { + "epoch": 0.37, + "grad_norm": 2.4880012482110083, + "learning_rate": 1.4384351552418027e-05, + "loss": 1.0148, + "step": 15912 + }, + { + "epoch": 0.37, + "grad_norm": 2.2760561142797706, + "learning_rate": 1.4383665746716582e-05, + "loss": 1.0639, + "step": 15913 + }, + { + "epoch": 0.37, + "grad_norm": 1.131500841495494, + "learning_rate": 1.4382979915492051e-05, + "loss": 0.939, + "step": 15914 + }, + { + "epoch": 0.37, + "grad_norm": 1.949271175581583, + "learning_rate": 1.4382294058748421e-05, + "loss": 0.9985, + "step": 15915 + }, + { + "epoch": 0.37, + "grad_norm": 1.0607214989397986, + "learning_rate": 1.4381608176489686e-05, + "loss": 1.0541, + "step": 15916 + }, + { + "epoch": 0.37, + "grad_norm": 1.7274456563629135, + "learning_rate": 1.438092226871984e-05, + "loss": 1.0259, + "step": 15917 + }, + { + "epoch": 0.38, + "grad_norm": 2.2721525555959827, + "learning_rate": 1.438023633544288e-05, + "loss": 0.9895, + "step": 15918 + }, + { + "epoch": 0.38, + "grad_norm": 2.307149755393792, + "learning_rate": 1.4379550376662799e-05, + "loss": 0.985, + "step": 15919 + }, + { + "epoch": 0.38, + "grad_norm": 1.9497568158529528, + "learning_rate": 1.4378864392383585e-05, + "loss": 0.9166, + "step": 15920 + }, + { + "epoch": 0.38, + "grad_norm": 2.1172083389789225, + "learning_rate": 1.437817838260924e-05, + "loss": 1.0511, + "step": 15921 + }, + { + "epoch": 0.38, + "grad_norm": 1.1042144747304559, + "learning_rate": 1.4377492347343753e-05, + "loss": 0.9914, + "step": 15922 + }, + { + "epoch": 0.38, + "grad_norm": 1.138635765667831, + "learning_rate": 1.4376806286591118e-05, + "loss": 1.0057, + "step": 15923 + }, + { + "epoch": 0.38, + "grad_norm": 2.199073674602808, + "learning_rate": 1.4376120200355334e-05, + "loss": 0.9281, + "step": 15924 + }, + { + "epoch": 0.38, + "grad_norm": 2.15978021166159, + "learning_rate": 1.437543408864039e-05, + "loss": 1.1113, + "step": 15925 + }, + { + "epoch": 0.38, + "grad_norm": 2.5448779843776768, + "learning_rate": 1.4374747951450285e-05, + "loss": 0.965, + "step": 15926 + }, + { + "epoch": 0.38, + "grad_norm": 1.0969945168233584, + "learning_rate": 1.4374061788789014e-05, + "loss": 1.0107, + "step": 15927 + }, + { + "epoch": 0.38, + "grad_norm": 2.0646756876864094, + "learning_rate": 1.4373375600660567e-05, + "loss": 1.1175, + "step": 15928 + }, + { + "epoch": 0.38, + "grad_norm": 2.2321158989572676, + "learning_rate": 1.4372689387068943e-05, + "loss": 1.1051, + "step": 15929 + }, + { + "epoch": 0.38, + "grad_norm": 1.1375497610282619, + "learning_rate": 1.4372003148018139e-05, + "loss": 0.957, + "step": 15930 + }, + { + "epoch": 0.38, + "grad_norm": 1.2203174388033975, + "learning_rate": 1.4371316883512147e-05, + "loss": 1.0349, + "step": 15931 + }, + { + "epoch": 0.38, + "grad_norm": 1.9781467157596782, + "learning_rate": 1.4370630593554965e-05, + "loss": 1.0761, + "step": 15932 + }, + { + "epoch": 0.38, + "grad_norm": 2.2071007654331964, + "learning_rate": 1.4369944278150587e-05, + "loss": 1.0381, + "step": 15933 + }, + { + "epoch": 0.38, + "grad_norm": 2.332992871521268, + "learning_rate": 1.4369257937303012e-05, + "loss": 1.0202, + "step": 15934 + }, + { + "epoch": 0.38, + "grad_norm": 1.9012142525459217, + "learning_rate": 1.4368571571016232e-05, + "loss": 1.0833, + "step": 15935 + }, + { + "epoch": 0.38, + "grad_norm": 4.258025728883071, + "learning_rate": 1.4367885179294248e-05, + "loss": 0.9771, + "step": 15936 + }, + { + "epoch": 0.38, + "grad_norm": 1.8751843193598332, + "learning_rate": 1.436719876214105e-05, + "loss": 1.0522, + "step": 15937 + }, + { + "epoch": 0.38, + "grad_norm": 2.2320070653125232, + "learning_rate": 1.4366512319560642e-05, + "loss": 1.1168, + "step": 15938 + }, + { + "epoch": 0.38, + "grad_norm": 2.225647321852972, + "learning_rate": 1.4365825851557012e-05, + "loss": 0.9898, + "step": 15939 + }, + { + "epoch": 0.38, + "grad_norm": 1.0887589072585833, + "learning_rate": 1.4365139358134166e-05, + "loss": 0.9446, + "step": 15940 + }, + { + "epoch": 0.38, + "grad_norm": 1.1100098925941866, + "learning_rate": 1.4364452839296092e-05, + "loss": 1.0515, + "step": 15941 + }, + { + "epoch": 0.38, + "grad_norm": 1.8834554241819506, + "learning_rate": 1.4363766295046797e-05, + "loss": 1.1275, + "step": 15942 + }, + { + "epoch": 0.38, + "grad_norm": 2.2264380380494355, + "learning_rate": 1.436307972539027e-05, + "loss": 1.0275, + "step": 15943 + }, + { + "epoch": 0.38, + "grad_norm": 2.0497642088732992, + "learning_rate": 1.4362393130330512e-05, + "loss": 0.9197, + "step": 15944 + }, + { + "epoch": 0.38, + "grad_norm": 2.327697930252345, + "learning_rate": 1.4361706509871523e-05, + "loss": 1.0183, + "step": 15945 + }, + { + "epoch": 0.38, + "grad_norm": 1.9445013591879803, + "learning_rate": 1.4361019864017293e-05, + "loss": 0.9939, + "step": 15946 + }, + { + "epoch": 0.38, + "grad_norm": 2.4479149668980167, + "learning_rate": 1.436033319277183e-05, + "loss": 1.0677, + "step": 15947 + }, + { + "epoch": 0.38, + "grad_norm": 1.8842797249069452, + "learning_rate": 1.4359646496139122e-05, + "loss": 0.9681, + "step": 15948 + }, + { + "epoch": 0.38, + "grad_norm": 2.125492686813279, + "learning_rate": 1.4358959774123174e-05, + "loss": 1.0716, + "step": 15949 + }, + { + "epoch": 0.38, + "grad_norm": 2.045355801214131, + "learning_rate": 1.435827302672798e-05, + "loss": 0.9039, + "step": 15950 + }, + { + "epoch": 0.38, + "grad_norm": 2.1119686263134874, + "learning_rate": 1.4357586253957545e-05, + "loss": 1.0752, + "step": 15951 + }, + { + "epoch": 0.38, + "grad_norm": 1.0869673756996605, + "learning_rate": 1.435689945581586e-05, + "loss": 0.9533, + "step": 15952 + }, + { + "epoch": 0.38, + "grad_norm": 2.2407427881641144, + "learning_rate": 1.4356212632306927e-05, + "loss": 0.9558, + "step": 15953 + }, + { + "epoch": 0.38, + "grad_norm": 1.9736322061310319, + "learning_rate": 1.4355525783434746e-05, + "loss": 1.0158, + "step": 15954 + }, + { + "epoch": 0.38, + "grad_norm": 2.2941049050964084, + "learning_rate": 1.4354838909203315e-05, + "loss": 1.0301, + "step": 15955 + }, + { + "epoch": 0.38, + "grad_norm": 1.9777149559021117, + "learning_rate": 1.4354152009616634e-05, + "loss": 1.0926, + "step": 15956 + }, + { + "epoch": 0.38, + "grad_norm": 2.3644774446019774, + "learning_rate": 1.4353465084678697e-05, + "loss": 1.0017, + "step": 15957 + }, + { + "epoch": 0.38, + "grad_norm": 1.9102134588758877, + "learning_rate": 1.4352778134393513e-05, + "loss": 0.9495, + "step": 15958 + }, + { + "epoch": 0.38, + "grad_norm": 2.1666154096598143, + "learning_rate": 1.4352091158765074e-05, + "loss": 1.0538, + "step": 15959 + }, + { + "epoch": 0.38, + "grad_norm": 1.9574339447808389, + "learning_rate": 1.4351404157797383e-05, + "loss": 1.2208, + "step": 15960 + }, + { + "epoch": 0.38, + "grad_norm": 2.1034031462017597, + "learning_rate": 1.4350717131494438e-05, + "loss": 1.0885, + "step": 15961 + }, + { + "epoch": 0.38, + "grad_norm": 2.871532213537521, + "learning_rate": 1.4350030079860245e-05, + "loss": 0.9859, + "step": 15962 + }, + { + "epoch": 0.38, + "grad_norm": 2.0821844312096185, + "learning_rate": 1.4349343002898795e-05, + "loss": 1.0564, + "step": 15963 + }, + { + "epoch": 0.38, + "grad_norm": 1.9889448692930307, + "learning_rate": 1.4348655900614094e-05, + "loss": 1.1432, + "step": 15964 + }, + { + "epoch": 0.38, + "grad_norm": 2.501473086978511, + "learning_rate": 1.4347968773010143e-05, + "loss": 1.0099, + "step": 15965 + }, + { + "epoch": 0.38, + "grad_norm": 2.2752549675380696, + "learning_rate": 1.4347281620090942e-05, + "loss": 0.9585, + "step": 15966 + }, + { + "epoch": 0.38, + "grad_norm": 3.3580088177268275, + "learning_rate": 1.434659444186049e-05, + "loss": 1.0127, + "step": 15967 + }, + { + "epoch": 0.38, + "grad_norm": 2.2482318012907987, + "learning_rate": 1.434590723832279e-05, + "loss": 1.0965, + "step": 15968 + }, + { + "epoch": 0.38, + "grad_norm": 2.2667336319647338, + "learning_rate": 1.4345220009481837e-05, + "loss": 1.0299, + "step": 15969 + }, + { + "epoch": 0.38, + "grad_norm": 2.175089745174681, + "learning_rate": 1.4344532755341645e-05, + "loss": 1.0099, + "step": 15970 + }, + { + "epoch": 0.38, + "grad_norm": 2.3109459672917656, + "learning_rate": 1.4343845475906201e-05, + "loss": 1.017, + "step": 15971 + }, + { + "epoch": 0.38, + "grad_norm": 1.785517714597069, + "learning_rate": 1.434315817117952e-05, + "loss": 1.0731, + "step": 15972 + }, + { + "epoch": 0.38, + "grad_norm": 2.3444398976252874, + "learning_rate": 1.4342470841165592e-05, + "loss": 0.966, + "step": 15973 + }, + { + "epoch": 0.38, + "grad_norm": 1.894071613551469, + "learning_rate": 1.4341783485868426e-05, + "loss": 1.0108, + "step": 15974 + }, + { + "epoch": 0.38, + "grad_norm": 2.050739759752096, + "learning_rate": 1.4341096105292021e-05, + "loss": 1.0149, + "step": 15975 + }, + { + "epoch": 0.38, + "grad_norm": 2.1022933507095436, + "learning_rate": 1.434040869944038e-05, + "loss": 0.8458, + "step": 15976 + }, + { + "epoch": 0.38, + "grad_norm": 2.497193281243752, + "learning_rate": 1.4339721268317508e-05, + "loss": 1.0486, + "step": 15977 + }, + { + "epoch": 0.38, + "grad_norm": 2.054269388214077, + "learning_rate": 1.43390338119274e-05, + "loss": 1.0712, + "step": 15978 + }, + { + "epoch": 0.38, + "grad_norm": 2.3463359561972763, + "learning_rate": 1.4338346330274067e-05, + "loss": 1.004, + "step": 15979 + }, + { + "epoch": 0.38, + "grad_norm": 2.2485741381004987, + "learning_rate": 1.4337658823361509e-05, + "loss": 0.9858, + "step": 15980 + }, + { + "epoch": 0.38, + "grad_norm": 2.6674561708926334, + "learning_rate": 1.4336971291193726e-05, + "loss": 0.9459, + "step": 15981 + }, + { + "epoch": 0.38, + "grad_norm": 2.116746201098284, + "learning_rate": 1.4336283733774722e-05, + "loss": 0.9736, + "step": 15982 + }, + { + "epoch": 0.38, + "grad_norm": 1.9439933049615026, + "learning_rate": 1.4335596151108504e-05, + "loss": 0.9214, + "step": 15983 + }, + { + "epoch": 0.38, + "grad_norm": 1.9714817077145523, + "learning_rate": 1.433490854319907e-05, + "loss": 1.0708, + "step": 15984 + }, + { + "epoch": 0.38, + "grad_norm": 2.188213960756337, + "learning_rate": 1.4334220910050429e-05, + "loss": 1.0484, + "step": 15985 + }, + { + "epoch": 0.38, + "grad_norm": 2.087064686397089, + "learning_rate": 1.433353325166658e-05, + "loss": 0.9453, + "step": 15986 + }, + { + "epoch": 0.38, + "grad_norm": 1.9191763630784424, + "learning_rate": 1.4332845568051532e-05, + "loss": 1.0037, + "step": 15987 + }, + { + "epoch": 0.38, + "grad_norm": 1.1678877699967842, + "learning_rate": 1.4332157859209278e-05, + "loss": 1.0706, + "step": 15988 + }, + { + "epoch": 0.38, + "grad_norm": 2.4998044934977943, + "learning_rate": 1.4331470125143835e-05, + "loss": 1.0522, + "step": 15989 + }, + { + "epoch": 0.38, + "grad_norm": 1.8976061110686713, + "learning_rate": 1.43307823658592e-05, + "loss": 1.0203, + "step": 15990 + }, + { + "epoch": 0.38, + "grad_norm": 2.425448991723235, + "learning_rate": 1.433009458135938e-05, + "loss": 0.9426, + "step": 15991 + }, + { + "epoch": 0.38, + "grad_norm": 2.646383486123252, + "learning_rate": 1.4329406771648375e-05, + "loss": 1.0706, + "step": 15992 + }, + { + "epoch": 0.38, + "grad_norm": 2.2098744829986234, + "learning_rate": 1.4328718936730197e-05, + "loss": 1.0575, + "step": 15993 + }, + { + "epoch": 0.38, + "grad_norm": 2.049687757968959, + "learning_rate": 1.4328031076608847e-05, + "loss": 0.9299, + "step": 15994 + }, + { + "epoch": 0.38, + "grad_norm": 1.1795188762767694, + "learning_rate": 1.4327343191288325e-05, + "loss": 0.9433, + "step": 15995 + }, + { + "epoch": 0.38, + "grad_norm": 2.416314680241138, + "learning_rate": 1.4326655280772648e-05, + "loss": 1.0284, + "step": 15996 + }, + { + "epoch": 0.38, + "grad_norm": 3.2091833633084663, + "learning_rate": 1.4325967345065808e-05, + "loss": 1.1101, + "step": 15997 + }, + { + "epoch": 0.38, + "grad_norm": 2.1664698557040296, + "learning_rate": 1.432527938417182e-05, + "loss": 0.9997, + "step": 15998 + }, + { + "epoch": 0.38, + "grad_norm": 2.4046521545073385, + "learning_rate": 1.4324591398094683e-05, + "loss": 1.0047, + "step": 15999 + }, + { + "epoch": 0.38, + "grad_norm": 1.965764306076468, + "learning_rate": 1.4323903386838409e-05, + "loss": 1.0012, + "step": 16000 + }, + { + "epoch": 0.38, + "grad_norm": 1.1326357100299838, + "learning_rate": 1.4323215350406998e-05, + "loss": 0.9762, + "step": 16001 + }, + { + "epoch": 0.38, + "grad_norm": 2.0554777483424185, + "learning_rate": 1.4322527288804461e-05, + "loss": 0.9886, + "step": 16002 + }, + { + "epoch": 0.38, + "grad_norm": 1.7605106391398206, + "learning_rate": 1.4321839202034799e-05, + "loss": 1.0125, + "step": 16003 + }, + { + "epoch": 0.38, + "grad_norm": 2.0170024513764004, + "learning_rate": 1.4321151090102024e-05, + "loss": 1.066, + "step": 16004 + }, + { + "epoch": 0.38, + "grad_norm": 2.180481043629414, + "learning_rate": 1.4320462953010137e-05, + "loss": 0.9776, + "step": 16005 + }, + { + "epoch": 0.38, + "grad_norm": 1.9536851565447517, + "learning_rate": 1.4319774790763148e-05, + "loss": 1.0576, + "step": 16006 + }, + { + "epoch": 0.38, + "grad_norm": 1.930991558297194, + "learning_rate": 1.4319086603365062e-05, + "loss": 1.0912, + "step": 16007 + }, + { + "epoch": 0.38, + "grad_norm": 2.2570649825809084, + "learning_rate": 1.4318398390819885e-05, + "loss": 0.9772, + "step": 16008 + }, + { + "epoch": 0.38, + "grad_norm": 1.0133605232641056, + "learning_rate": 1.4317710153131628e-05, + "loss": 0.9545, + "step": 16009 + }, + { + "epoch": 0.38, + "grad_norm": 2.2011947931046025, + "learning_rate": 1.4317021890304294e-05, + "loss": 1.11, + "step": 16010 + }, + { + "epoch": 0.38, + "grad_norm": 1.9028751472656988, + "learning_rate": 1.4316333602341892e-05, + "loss": 0.9883, + "step": 16011 + }, + { + "epoch": 0.38, + "grad_norm": 2.2509203474551294, + "learning_rate": 1.431564528924843e-05, + "loss": 1.1392, + "step": 16012 + }, + { + "epoch": 0.38, + "grad_norm": 2.0897817840926165, + "learning_rate": 1.4314956951027914e-05, + "loss": 1.0451, + "step": 16013 + }, + { + "epoch": 0.38, + "grad_norm": 2.0735490769917146, + "learning_rate": 1.4314268587684353e-05, + "loss": 0.9742, + "step": 16014 + }, + { + "epoch": 0.38, + "grad_norm": 2.7353243973726955, + "learning_rate": 1.4313580199221754e-05, + "loss": 1.0391, + "step": 16015 + }, + { + "epoch": 0.38, + "grad_norm": 1.8786571730659778, + "learning_rate": 1.4312891785644124e-05, + "loss": 1.1732, + "step": 16016 + }, + { + "epoch": 0.38, + "grad_norm": 2.19864899426968, + "learning_rate": 1.4312203346955478e-05, + "loss": 1.0006, + "step": 16017 + }, + { + "epoch": 0.38, + "grad_norm": 1.1586024215631598, + "learning_rate": 1.4311514883159812e-05, + "loss": 0.9706, + "step": 16018 + }, + { + "epoch": 0.38, + "grad_norm": 1.7602185364157672, + "learning_rate": 1.4310826394261148e-05, + "loss": 1.0093, + "step": 16019 + }, + { + "epoch": 0.38, + "grad_norm": 2.9550879547468734, + "learning_rate": 1.4310137880263481e-05, + "loss": 1.1021, + "step": 16020 + }, + { + "epoch": 0.38, + "grad_norm": 2.2372744241621634, + "learning_rate": 1.4309449341170832e-05, + "loss": 1.0345, + "step": 16021 + }, + { + "epoch": 0.38, + "grad_norm": 2.04072108151162, + "learning_rate": 1.43087607769872e-05, + "loss": 1.0445, + "step": 16022 + }, + { + "epoch": 0.38, + "grad_norm": 1.077843139005707, + "learning_rate": 1.4308072187716604e-05, + "loss": 0.9794, + "step": 16023 + }, + { + "epoch": 0.38, + "grad_norm": 2.0031342299090826, + "learning_rate": 1.4307383573363044e-05, + "loss": 0.9248, + "step": 16024 + }, + { + "epoch": 0.38, + "grad_norm": 2.4073393168553734, + "learning_rate": 1.4306694933930537e-05, + "loss": 0.9588, + "step": 16025 + }, + { + "epoch": 0.38, + "grad_norm": 2.0349794493381603, + "learning_rate": 1.4306006269423082e-05, + "loss": 1.1436, + "step": 16026 + }, + { + "epoch": 0.38, + "grad_norm": 1.1913865840179918, + "learning_rate": 1.4305317579844701e-05, + "loss": 0.957, + "step": 16027 + }, + { + "epoch": 0.38, + "grad_norm": 2.4328019952692723, + "learning_rate": 1.4304628865199397e-05, + "loss": 1.0693, + "step": 16028 + }, + { + "epoch": 0.38, + "grad_norm": 1.8909250397666464, + "learning_rate": 1.4303940125491175e-05, + "loss": 1.0287, + "step": 16029 + }, + { + "epoch": 0.38, + "grad_norm": 2.1690547471947994, + "learning_rate": 1.430325136072406e-05, + "loss": 1.1279, + "step": 16030 + }, + { + "epoch": 0.38, + "grad_norm": 1.1235138527625623, + "learning_rate": 1.4302562570902048e-05, + "loss": 0.9628, + "step": 16031 + }, + { + "epoch": 0.38, + "grad_norm": 2.004233772659056, + "learning_rate": 1.4301873756029156e-05, + "loss": 0.9071, + "step": 16032 + }, + { + "epoch": 0.38, + "grad_norm": 1.76129070565783, + "learning_rate": 1.4301184916109395e-05, + "loss": 0.9355, + "step": 16033 + }, + { + "epoch": 0.38, + "grad_norm": 2.1894107865766737, + "learning_rate": 1.430049605114677e-05, + "loss": 1.1216, + "step": 16034 + }, + { + "epoch": 0.38, + "grad_norm": 2.480913731974786, + "learning_rate": 1.4299807161145297e-05, + "loss": 0.9718, + "step": 16035 + }, + { + "epoch": 0.38, + "grad_norm": 2.0441621870607287, + "learning_rate": 1.4299118246108985e-05, + "loss": 1.06, + "step": 16036 + }, + { + "epoch": 0.38, + "grad_norm": 2.538116166988634, + "learning_rate": 1.4298429306041846e-05, + "loss": 1.132, + "step": 16037 + }, + { + "epoch": 0.38, + "grad_norm": 2.1051561280748974, + "learning_rate": 1.4297740340947891e-05, + "loss": 0.8928, + "step": 16038 + }, + { + "epoch": 0.38, + "grad_norm": 2.521161071422563, + "learning_rate": 1.429705135083113e-05, + "loss": 1.0432, + "step": 16039 + }, + { + "epoch": 0.38, + "grad_norm": 1.8161451195600222, + "learning_rate": 1.4296362335695578e-05, + "loss": 0.8597, + "step": 16040 + }, + { + "epoch": 0.38, + "grad_norm": 2.1589671879673173, + "learning_rate": 1.429567329554524e-05, + "loss": 1.0062, + "step": 16041 + }, + { + "epoch": 0.38, + "grad_norm": 1.9271671613283539, + "learning_rate": 1.4294984230384135e-05, + "loss": 1.0934, + "step": 16042 + }, + { + "epoch": 0.38, + "grad_norm": 1.07546037270402, + "learning_rate": 1.429429514021627e-05, + "loss": 0.9386, + "step": 16043 + }, + { + "epoch": 0.38, + "grad_norm": 1.928137940929181, + "learning_rate": 1.4293606025045663e-05, + "loss": 0.928, + "step": 16044 + }, + { + "epoch": 0.38, + "grad_norm": 2.0124848193014797, + "learning_rate": 1.429291688487632e-05, + "loss": 0.9745, + "step": 16045 + }, + { + "epoch": 0.38, + "grad_norm": 2.216405735609965, + "learning_rate": 1.4292227719712254e-05, + "loss": 0.8845, + "step": 16046 + }, + { + "epoch": 0.38, + "grad_norm": 2.1826824964122684, + "learning_rate": 1.4291538529557485e-05, + "loss": 1.0773, + "step": 16047 + }, + { + "epoch": 0.38, + "grad_norm": 2.4651927420384796, + "learning_rate": 1.4290849314416014e-05, + "loss": 1.0015, + "step": 16048 + }, + { + "epoch": 0.38, + "grad_norm": 2.639822549560536, + "learning_rate": 1.4290160074291864e-05, + "loss": 1.1148, + "step": 16049 + }, + { + "epoch": 0.38, + "grad_norm": 1.9986091924976386, + "learning_rate": 1.4289470809189039e-05, + "loss": 1.0055, + "step": 16050 + }, + { + "epoch": 0.38, + "grad_norm": 2.5141344512711505, + "learning_rate": 1.4288781519111563e-05, + "loss": 1.0621, + "step": 16051 + }, + { + "epoch": 0.38, + "grad_norm": 1.9162416875246258, + "learning_rate": 1.428809220406344e-05, + "loss": 1.1149, + "step": 16052 + }, + { + "epoch": 0.38, + "grad_norm": 1.8716043241563, + "learning_rate": 1.4287402864048686e-05, + "loss": 0.9983, + "step": 16053 + }, + { + "epoch": 0.38, + "grad_norm": 1.9619481613415346, + "learning_rate": 1.4286713499071313e-05, + "loss": 1.126, + "step": 16054 + }, + { + "epoch": 0.38, + "grad_norm": 1.8914506785095917, + "learning_rate": 1.428602410913534e-05, + "loss": 1.175, + "step": 16055 + }, + { + "epoch": 0.38, + "grad_norm": 1.9162729380467491, + "learning_rate": 1.4285334694244778e-05, + "loss": 1.0793, + "step": 16056 + }, + { + "epoch": 0.38, + "grad_norm": 1.9858369334818382, + "learning_rate": 1.4284645254403642e-05, + "loss": 1.1279, + "step": 16057 + }, + { + "epoch": 0.38, + "grad_norm": 2.306849025724394, + "learning_rate": 1.4283955789615942e-05, + "loss": 0.9815, + "step": 16058 + }, + { + "epoch": 0.38, + "grad_norm": 2.2373492964098314, + "learning_rate": 1.4283266299885696e-05, + "loss": 1.0443, + "step": 16059 + }, + { + "epoch": 0.38, + "grad_norm": 2.5738913380043935, + "learning_rate": 1.4282576785216917e-05, + "loss": 1.0496, + "step": 16060 + }, + { + "epoch": 0.38, + "grad_norm": 1.9072890772562459, + "learning_rate": 1.4281887245613621e-05, + "loss": 1.1269, + "step": 16061 + }, + { + "epoch": 0.38, + "grad_norm": 2.4766665449609446, + "learning_rate": 1.4281197681079821e-05, + "loss": 1.0464, + "step": 16062 + }, + { + "epoch": 0.38, + "grad_norm": 1.9965347212482674, + "learning_rate": 1.4280508091619534e-05, + "loss": 0.8198, + "step": 16063 + }, + { + "epoch": 0.38, + "grad_norm": 2.3194572941000184, + "learning_rate": 1.4279818477236773e-05, + "loss": 1.0403, + "step": 16064 + }, + { + "epoch": 0.38, + "grad_norm": 1.9155427510101117, + "learning_rate": 1.4279128837935554e-05, + "loss": 0.9813, + "step": 16065 + }, + { + "epoch": 0.38, + "grad_norm": 2.0354661503084714, + "learning_rate": 1.4278439173719892e-05, + "loss": 1.0048, + "step": 16066 + }, + { + "epoch": 0.38, + "grad_norm": 1.9516682437350268, + "learning_rate": 1.4277749484593801e-05, + "loss": 1.0543, + "step": 16067 + }, + { + "epoch": 0.38, + "grad_norm": 2.089191488948374, + "learning_rate": 1.4277059770561304e-05, + "loss": 1.1615, + "step": 16068 + }, + { + "epoch": 0.38, + "grad_norm": 1.9197205593683977, + "learning_rate": 1.4276370031626403e-05, + "loss": 1.1523, + "step": 16069 + }, + { + "epoch": 0.38, + "grad_norm": 1.901453454740407, + "learning_rate": 1.4275680267793125e-05, + "loss": 0.8985, + "step": 16070 + }, + { + "epoch": 0.38, + "grad_norm": 1.9156678621296788, + "learning_rate": 1.4274990479065483e-05, + "loss": 1.0234, + "step": 16071 + }, + { + "epoch": 0.38, + "grad_norm": 2.2611791739373848, + "learning_rate": 1.4274300665447496e-05, + "loss": 0.9753, + "step": 16072 + }, + { + "epoch": 0.38, + "grad_norm": 2.1944055807218397, + "learning_rate": 1.4273610826943171e-05, + "loss": 1.058, + "step": 16073 + }, + { + "epoch": 0.38, + "grad_norm": 2.3773474173301303, + "learning_rate": 1.4272920963556537e-05, + "loss": 0.8716, + "step": 16074 + }, + { + "epoch": 0.38, + "grad_norm": 2.1095528947026203, + "learning_rate": 1.4272231075291603e-05, + "loss": 1.1892, + "step": 16075 + }, + { + "epoch": 0.38, + "grad_norm": 2.0814358130874777, + "learning_rate": 1.4271541162152385e-05, + "loss": 1.0738, + "step": 16076 + }, + { + "epoch": 0.38, + "grad_norm": 1.1582084268863517, + "learning_rate": 1.4270851224142904e-05, + "loss": 1.0001, + "step": 16077 + }, + { + "epoch": 0.38, + "grad_norm": 2.0658254734567043, + "learning_rate": 1.427016126126717e-05, + "loss": 1.0935, + "step": 16078 + }, + { + "epoch": 0.38, + "grad_norm": 1.095888122624947, + "learning_rate": 1.4269471273529212e-05, + "loss": 1.0056, + "step": 16079 + }, + { + "epoch": 0.38, + "grad_norm": 2.0651170912821115, + "learning_rate": 1.4268781260933034e-05, + "loss": 1.0887, + "step": 16080 + }, + { + "epoch": 0.38, + "grad_norm": 1.819789320895416, + "learning_rate": 1.4268091223482667e-05, + "loss": 0.9349, + "step": 16081 + }, + { + "epoch": 0.38, + "grad_norm": 2.0555417299681067, + "learning_rate": 1.4267401161182117e-05, + "loss": 1.0246, + "step": 16082 + }, + { + "epoch": 0.38, + "grad_norm": 2.343115165783265, + "learning_rate": 1.4266711074035408e-05, + "loss": 1.0735, + "step": 16083 + }, + { + "epoch": 0.38, + "grad_norm": 1.833608813065484, + "learning_rate": 1.4266020962046555e-05, + "loss": 0.9496, + "step": 16084 + }, + { + "epoch": 0.38, + "grad_norm": 1.971951449301022, + "learning_rate": 1.4265330825219578e-05, + "loss": 0.9626, + "step": 16085 + }, + { + "epoch": 0.38, + "grad_norm": 2.673856307355601, + "learning_rate": 1.4264640663558496e-05, + "loss": 1.0241, + "step": 16086 + }, + { + "epoch": 0.38, + "grad_norm": 1.0234895099872336, + "learning_rate": 1.4263950477067326e-05, + "loss": 0.9895, + "step": 16087 + }, + { + "epoch": 0.38, + "grad_norm": 2.0395963450030337, + "learning_rate": 1.4263260265750085e-05, + "loss": 1.0501, + "step": 16088 + }, + { + "epoch": 0.38, + "grad_norm": 1.9794091983340028, + "learning_rate": 1.4262570029610792e-05, + "loss": 1.15, + "step": 16089 + }, + { + "epoch": 0.38, + "grad_norm": 1.8484213141165382, + "learning_rate": 1.4261879768653469e-05, + "loss": 1.2176, + "step": 16090 + }, + { + "epoch": 0.38, + "grad_norm": 1.9841194452910547, + "learning_rate": 1.4261189482882131e-05, + "loss": 1.0394, + "step": 16091 + }, + { + "epoch": 0.38, + "grad_norm": 2.0113372546327914, + "learning_rate": 1.4260499172300798e-05, + "loss": 1.0634, + "step": 16092 + }, + { + "epoch": 0.38, + "grad_norm": 2.1251791136125764, + "learning_rate": 1.425980883691349e-05, + "loss": 1.081, + "step": 16093 + }, + { + "epoch": 0.38, + "grad_norm": 2.4922280875831504, + "learning_rate": 1.425911847672423e-05, + "loss": 1.1059, + "step": 16094 + }, + { + "epoch": 0.38, + "grad_norm": 2.1086223623851, + "learning_rate": 1.4258428091737031e-05, + "loss": 1.1121, + "step": 16095 + }, + { + "epoch": 0.38, + "grad_norm": 1.997639980478818, + "learning_rate": 1.425773768195592e-05, + "loss": 0.9662, + "step": 16096 + }, + { + "epoch": 0.38, + "grad_norm": 1.9925810988407195, + "learning_rate": 1.4257047247384904e-05, + "loss": 1.1254, + "step": 16097 + }, + { + "epoch": 0.38, + "grad_norm": 1.9385815354181724, + "learning_rate": 1.4256356788028018e-05, + "loss": 1.1736, + "step": 16098 + }, + { + "epoch": 0.38, + "grad_norm": 2.2962553970023194, + "learning_rate": 1.425566630388927e-05, + "loss": 1.0456, + "step": 16099 + }, + { + "epoch": 0.38, + "grad_norm": 2.1649733987476822, + "learning_rate": 1.425497579497269e-05, + "loss": 0.9332, + "step": 16100 + }, + { + "epoch": 0.38, + "grad_norm": 5.473582246812261, + "learning_rate": 1.425428526128229e-05, + "loss": 1.0254, + "step": 16101 + }, + { + "epoch": 0.38, + "grad_norm": 2.0860887122843432, + "learning_rate": 1.42535947028221e-05, + "loss": 1.0619, + "step": 16102 + }, + { + "epoch": 0.38, + "grad_norm": 2.2250627752913905, + "learning_rate": 1.4252904119596129e-05, + "loss": 0.9387, + "step": 16103 + }, + { + "epoch": 0.38, + "grad_norm": 1.9301113006432693, + "learning_rate": 1.4252213511608406e-05, + "loss": 0.985, + "step": 16104 + }, + { + "epoch": 0.38, + "grad_norm": 1.9084937306047745, + "learning_rate": 1.425152287886295e-05, + "loss": 0.9899, + "step": 16105 + }, + { + "epoch": 0.38, + "grad_norm": 1.9046262771637745, + "learning_rate": 1.4250832221363782e-05, + "loss": 1.0375, + "step": 16106 + }, + { + "epoch": 0.38, + "grad_norm": 2.070382357939234, + "learning_rate": 1.425014153911492e-05, + "loss": 1.0958, + "step": 16107 + }, + { + "epoch": 0.38, + "grad_norm": 2.072916080198562, + "learning_rate": 1.4249450832120393e-05, + "loss": 0.9836, + "step": 16108 + }, + { + "epoch": 0.38, + "grad_norm": 2.6625317328935303, + "learning_rate": 1.4248760100384216e-05, + "loss": 0.9807, + "step": 16109 + }, + { + "epoch": 0.38, + "grad_norm": 1.813902672424854, + "learning_rate": 1.4248069343910412e-05, + "loss": 1.0186, + "step": 16110 + }, + { + "epoch": 0.38, + "grad_norm": 1.9652019957652251, + "learning_rate": 1.4247378562703004e-05, + "loss": 1.121, + "step": 16111 + }, + { + "epoch": 0.38, + "grad_norm": 2.221795209878453, + "learning_rate": 1.4246687756766013e-05, + "loss": 1.0756, + "step": 16112 + }, + { + "epoch": 0.38, + "grad_norm": 2.0484703699838405, + "learning_rate": 1.424599692610346e-05, + "loss": 1.2184, + "step": 16113 + }, + { + "epoch": 0.38, + "grad_norm": 2.0407530260899898, + "learning_rate": 1.424530607071937e-05, + "loss": 1.0634, + "step": 16114 + }, + { + "epoch": 0.38, + "grad_norm": 2.0429795495809078, + "learning_rate": 1.4244615190617765e-05, + "loss": 1.0366, + "step": 16115 + }, + { + "epoch": 0.38, + "grad_norm": 1.908772965008818, + "learning_rate": 1.4243924285802665e-05, + "loss": 1.0188, + "step": 16116 + }, + { + "epoch": 0.38, + "grad_norm": 2.151367390745606, + "learning_rate": 1.4243233356278097e-05, + "loss": 0.968, + "step": 16117 + }, + { + "epoch": 0.38, + "grad_norm": 1.904991838115522, + "learning_rate": 1.4242542402048076e-05, + "loss": 1.038, + "step": 16118 + }, + { + "epoch": 0.38, + "grad_norm": 2.023242086861994, + "learning_rate": 1.4241851423116634e-05, + "loss": 1.095, + "step": 16119 + }, + { + "epoch": 0.38, + "grad_norm": 2.2050777882786066, + "learning_rate": 1.4241160419487789e-05, + "loss": 1.0117, + "step": 16120 + }, + { + "epoch": 0.38, + "grad_norm": 2.4387708736685285, + "learning_rate": 1.4240469391165569e-05, + "loss": 1.0385, + "step": 16121 + }, + { + "epoch": 0.38, + "grad_norm": 2.114420685320199, + "learning_rate": 1.4239778338153984e-05, + "loss": 1.0117, + "step": 16122 + }, + { + "epoch": 0.38, + "grad_norm": 2.1241185800575115, + "learning_rate": 1.4239087260457076e-05, + "loss": 0.9654, + "step": 16123 + }, + { + "epoch": 0.38, + "grad_norm": 1.938687608266105, + "learning_rate": 1.4238396158078859e-05, + "loss": 1.0497, + "step": 16124 + }, + { + "epoch": 0.38, + "grad_norm": 2.046938993230777, + "learning_rate": 1.4237705031023354e-05, + "loss": 1.0294, + "step": 16125 + }, + { + "epoch": 0.38, + "grad_norm": 2.494126027373699, + "learning_rate": 1.423701387929459e-05, + "loss": 1.2063, + "step": 16126 + }, + { + "epoch": 0.38, + "grad_norm": 2.3242006808067424, + "learning_rate": 1.4236322702896588e-05, + "loss": 0.9536, + "step": 16127 + }, + { + "epoch": 0.38, + "grad_norm": 2.0667536795908883, + "learning_rate": 1.4235631501833378e-05, + "loss": 0.9643, + "step": 16128 + }, + { + "epoch": 0.38, + "grad_norm": 1.8443256339306797, + "learning_rate": 1.4234940276108977e-05, + "loss": 1.0933, + "step": 16129 + }, + { + "epoch": 0.38, + "grad_norm": 2.040119310173008, + "learning_rate": 1.4234249025727419e-05, + "loss": 0.9663, + "step": 16130 + }, + { + "epoch": 0.38, + "grad_norm": 2.1318648018435957, + "learning_rate": 1.4233557750692715e-05, + "loss": 0.8594, + "step": 16131 + }, + { + "epoch": 0.38, + "grad_norm": 2.2121417155807563, + "learning_rate": 1.4232866451008902e-05, + "loss": 1.0294, + "step": 16132 + }, + { + "epoch": 0.38, + "grad_norm": 2.2549416493900605, + "learning_rate": 1.4232175126679999e-05, + "loss": 0.9668, + "step": 16133 + }, + { + "epoch": 0.38, + "grad_norm": 1.9791012149918905, + "learning_rate": 1.423148377771003e-05, + "loss": 1.0234, + "step": 16134 + }, + { + "epoch": 0.38, + "grad_norm": 1.1546848793929203, + "learning_rate": 1.4230792404103026e-05, + "loss": 0.9411, + "step": 16135 + }, + { + "epoch": 0.38, + "grad_norm": 1.979973777084708, + "learning_rate": 1.4230101005863008e-05, + "loss": 1.0897, + "step": 16136 + }, + { + "epoch": 0.38, + "grad_norm": 2.05784108306633, + "learning_rate": 1.4229409582994005e-05, + "loss": 0.8619, + "step": 16137 + }, + { + "epoch": 0.38, + "grad_norm": 2.134563427198553, + "learning_rate": 1.422871813550004e-05, + "loss": 0.9859, + "step": 16138 + }, + { + "epoch": 0.38, + "grad_norm": 2.0533977567690576, + "learning_rate": 1.4228026663385136e-05, + "loss": 0.9516, + "step": 16139 + }, + { + "epoch": 0.38, + "grad_norm": 1.9306455038700139, + "learning_rate": 1.4227335166653325e-05, + "loss": 1.0546, + "step": 16140 + }, + { + "epoch": 0.38, + "grad_norm": 1.084605341544342, + "learning_rate": 1.4226643645308628e-05, + "loss": 0.9471, + "step": 16141 + }, + { + "epoch": 0.38, + "grad_norm": 2.0113836400532525, + "learning_rate": 1.4225952099355075e-05, + "loss": 0.9556, + "step": 16142 + }, + { + "epoch": 0.38, + "grad_norm": 2.1293954193193856, + "learning_rate": 1.4225260528796692e-05, + "loss": 1.0759, + "step": 16143 + }, + { + "epoch": 0.38, + "grad_norm": 2.060052096028195, + "learning_rate": 1.42245689336375e-05, + "loss": 1.0552, + "step": 16144 + }, + { + "epoch": 0.38, + "grad_norm": 1.8262833623080361, + "learning_rate": 1.4223877313881535e-05, + "loss": 0.9682, + "step": 16145 + }, + { + "epoch": 0.38, + "grad_norm": 2.7536733850139536, + "learning_rate": 1.422318566953282e-05, + "loss": 0.9458, + "step": 16146 + }, + { + "epoch": 0.38, + "grad_norm": 1.9508663637335035, + "learning_rate": 1.422249400059538e-05, + "loss": 0.9535, + "step": 16147 + }, + { + "epoch": 0.38, + "grad_norm": 1.9389895957754069, + "learning_rate": 1.422180230707324e-05, + "loss": 1.0587, + "step": 16148 + }, + { + "epoch": 0.38, + "grad_norm": 1.95103615879723, + "learning_rate": 1.4221110588970432e-05, + "loss": 1.1929, + "step": 16149 + }, + { + "epoch": 0.38, + "grad_norm": 2.192658349396316, + "learning_rate": 1.4220418846290981e-05, + "loss": 1.026, + "step": 16150 + }, + { + "epoch": 0.38, + "grad_norm": 2.42253372545834, + "learning_rate": 1.4219727079038919e-05, + "loss": 1.0365, + "step": 16151 + }, + { + "epoch": 0.38, + "grad_norm": 1.076501739507801, + "learning_rate": 1.4219035287218264e-05, + "loss": 0.9274, + "step": 16152 + }, + { + "epoch": 0.38, + "grad_norm": 2.2319806902850075, + "learning_rate": 1.4218343470833055e-05, + "loss": 0.9848, + "step": 16153 + }, + { + "epoch": 0.38, + "grad_norm": 1.8407548694209348, + "learning_rate": 1.4217651629887312e-05, + "loss": 0.9738, + "step": 16154 + }, + { + "epoch": 0.38, + "grad_norm": 1.788626467258992, + "learning_rate": 1.4216959764385067e-05, + "loss": 0.9582, + "step": 16155 + }, + { + "epoch": 0.38, + "grad_norm": 1.064383031348385, + "learning_rate": 1.4216267874330347e-05, + "loss": 1.022, + "step": 16156 + }, + { + "epoch": 0.38, + "grad_norm": 2.0691181921611577, + "learning_rate": 1.421557595972718e-05, + "loss": 1.0115, + "step": 16157 + }, + { + "epoch": 0.38, + "grad_norm": 1.937638318456271, + "learning_rate": 1.4214884020579595e-05, + "loss": 1.1678, + "step": 16158 + }, + { + "epoch": 0.38, + "grad_norm": 2.810515872460917, + "learning_rate": 1.4214192056891622e-05, + "loss": 0.9341, + "step": 16159 + }, + { + "epoch": 0.38, + "grad_norm": 1.9463625598898355, + "learning_rate": 1.421350006866729e-05, + "loss": 0.9855, + "step": 16160 + }, + { + "epoch": 0.38, + "grad_norm": 1.8151980451603804, + "learning_rate": 1.4212808055910623e-05, + "loss": 1.0088, + "step": 16161 + }, + { + "epoch": 0.38, + "grad_norm": 2.537904074117146, + "learning_rate": 1.4212116018625658e-05, + "loss": 1.1145, + "step": 16162 + }, + { + "epoch": 0.38, + "grad_norm": 1.0994852260206986, + "learning_rate": 1.4211423956816415e-05, + "loss": 1.0014, + "step": 16163 + }, + { + "epoch": 0.38, + "grad_norm": 2.2998957896360923, + "learning_rate": 1.4210731870486931e-05, + "loss": 1.0475, + "step": 16164 + }, + { + "epoch": 0.38, + "grad_norm": 2.672642838266164, + "learning_rate": 1.4210039759641232e-05, + "loss": 0.9889, + "step": 16165 + }, + { + "epoch": 0.38, + "grad_norm": 2.1062079613353872, + "learning_rate": 1.4209347624283352e-05, + "loss": 0.9365, + "step": 16166 + }, + { + "epoch": 0.38, + "grad_norm": 2.0005529963282047, + "learning_rate": 1.4208655464417313e-05, + "loss": 1.0777, + "step": 16167 + }, + { + "epoch": 0.38, + "grad_norm": 2.297940081874513, + "learning_rate": 1.4207963280047155e-05, + "loss": 1.127, + "step": 16168 + }, + { + "epoch": 0.38, + "grad_norm": 2.116067734022216, + "learning_rate": 1.4207271071176896e-05, + "loss": 1.1295, + "step": 16169 + }, + { + "epoch": 0.38, + "grad_norm": 1.9903286683087065, + "learning_rate": 1.4206578837810578e-05, + "loss": 1.0354, + "step": 16170 + }, + { + "epoch": 0.38, + "grad_norm": 1.1391178980468668, + "learning_rate": 1.4205886579952221e-05, + "loss": 0.9117, + "step": 16171 + }, + { + "epoch": 0.38, + "grad_norm": 1.8105357972467688, + "learning_rate": 1.4205194297605867e-05, + "loss": 1.0143, + "step": 16172 + }, + { + "epoch": 0.38, + "grad_norm": 1.943207558196281, + "learning_rate": 1.4204501990775535e-05, + "loss": 1.0836, + "step": 16173 + }, + { + "epoch": 0.38, + "grad_norm": 1.7274945887759778, + "learning_rate": 1.4203809659465266e-05, + "loss": 0.9529, + "step": 16174 + }, + { + "epoch": 0.38, + "grad_norm": 1.9382042476154995, + "learning_rate": 1.4203117303679083e-05, + "loss": 0.923, + "step": 16175 + }, + { + "epoch": 0.38, + "grad_norm": 2.3502632164432127, + "learning_rate": 1.4202424923421018e-05, + "loss": 0.9817, + "step": 16176 + }, + { + "epoch": 0.38, + "grad_norm": 1.1016343221010025, + "learning_rate": 1.420173251869511e-05, + "loss": 0.988, + "step": 16177 + }, + { + "epoch": 0.38, + "grad_norm": 2.1446967925258513, + "learning_rate": 1.4201040089505383e-05, + "loss": 1.0503, + "step": 16178 + }, + { + "epoch": 0.38, + "grad_norm": 1.2278987238893508, + "learning_rate": 1.420034763585587e-05, + "loss": 1.0647, + "step": 16179 + }, + { + "epoch": 0.38, + "grad_norm": 2.0836541651943823, + "learning_rate": 1.4199655157750602e-05, + "loss": 1.1533, + "step": 16180 + }, + { + "epoch": 0.38, + "grad_norm": 1.9869053785090642, + "learning_rate": 1.4198962655193617e-05, + "loss": 1.2068, + "step": 16181 + }, + { + "epoch": 0.38, + "grad_norm": 1.837198456674915, + "learning_rate": 1.4198270128188938e-05, + "loss": 1.0579, + "step": 16182 + }, + { + "epoch": 0.38, + "grad_norm": 2.9958191104493213, + "learning_rate": 1.4197577576740602e-05, + "loss": 1.055, + "step": 16183 + }, + { + "epoch": 0.38, + "grad_norm": 2.106740924163548, + "learning_rate": 1.419688500085264e-05, + "loss": 1.0352, + "step": 16184 + }, + { + "epoch": 0.38, + "grad_norm": 2.4057964091531945, + "learning_rate": 1.4196192400529088e-05, + "loss": 0.9011, + "step": 16185 + }, + { + "epoch": 0.38, + "grad_norm": 1.9574765449013694, + "learning_rate": 1.4195499775773972e-05, + "loss": 1.0335, + "step": 16186 + }, + { + "epoch": 0.38, + "grad_norm": 2.33361066455956, + "learning_rate": 1.4194807126591328e-05, + "loss": 0.9141, + "step": 16187 + }, + { + "epoch": 0.38, + "grad_norm": 1.8470125448724735, + "learning_rate": 1.4194114452985189e-05, + "loss": 0.9561, + "step": 16188 + }, + { + "epoch": 0.38, + "grad_norm": 1.9676691695008566, + "learning_rate": 1.4193421754959586e-05, + "loss": 0.937, + "step": 16189 + }, + { + "epoch": 0.38, + "grad_norm": 2.415265667089266, + "learning_rate": 1.4192729032518557e-05, + "loss": 0.9538, + "step": 16190 + }, + { + "epoch": 0.38, + "grad_norm": 2.062032119781939, + "learning_rate": 1.4192036285666133e-05, + "loss": 1.0237, + "step": 16191 + }, + { + "epoch": 0.38, + "grad_norm": 1.8358836626468316, + "learning_rate": 1.4191343514406345e-05, + "loss": 1.0683, + "step": 16192 + }, + { + "epoch": 0.38, + "grad_norm": 2.3298449715406515, + "learning_rate": 1.4190650718743229e-05, + "loss": 1.0832, + "step": 16193 + }, + { + "epoch": 0.38, + "grad_norm": 1.9039509072621406, + "learning_rate": 1.4189957898680815e-05, + "loss": 1.0648, + "step": 16194 + }, + { + "epoch": 0.38, + "grad_norm": 2.2165457286024424, + "learning_rate": 1.4189265054223142e-05, + "loss": 1.1501, + "step": 16195 + }, + { + "epoch": 0.38, + "grad_norm": 1.9737973420667938, + "learning_rate": 1.4188572185374241e-05, + "loss": 1.0043, + "step": 16196 + }, + { + "epoch": 0.38, + "grad_norm": 1.8317156024882797, + "learning_rate": 1.4187879292138144e-05, + "loss": 1.0343, + "step": 16197 + }, + { + "epoch": 0.38, + "grad_norm": 2.324016724449437, + "learning_rate": 1.4187186374518893e-05, + "loss": 0.9849, + "step": 16198 + }, + { + "epoch": 0.38, + "grad_norm": 1.9785724265285127, + "learning_rate": 1.4186493432520513e-05, + "loss": 0.9445, + "step": 16199 + }, + { + "epoch": 0.38, + "grad_norm": 2.185248485817072, + "learning_rate": 1.4185800466147048e-05, + "loss": 1.044, + "step": 16200 + }, + { + "epoch": 0.38, + "grad_norm": 1.9136531213231573, + "learning_rate": 1.4185107475402522e-05, + "loss": 0.8702, + "step": 16201 + }, + { + "epoch": 0.38, + "grad_norm": 2.300097471074484, + "learning_rate": 1.4184414460290979e-05, + "loss": 0.9856, + "step": 16202 + }, + { + "epoch": 0.38, + "grad_norm": 1.926284787561221, + "learning_rate": 1.4183721420816448e-05, + "loss": 0.9235, + "step": 16203 + }, + { + "epoch": 0.38, + "grad_norm": 1.8802416487461677, + "learning_rate": 1.4183028356982969e-05, + "loss": 1.0403, + "step": 16204 + }, + { + "epoch": 0.38, + "grad_norm": 2.4573441357952492, + "learning_rate": 1.4182335268794571e-05, + "loss": 0.9723, + "step": 16205 + }, + { + "epoch": 0.38, + "grad_norm": 1.8061298614921877, + "learning_rate": 1.4181642156255295e-05, + "loss": 1.078, + "step": 16206 + }, + { + "epoch": 0.38, + "grad_norm": 2.042490989519802, + "learning_rate": 1.4180949019369174e-05, + "loss": 0.9203, + "step": 16207 + }, + { + "epoch": 0.38, + "grad_norm": 1.816060921377852, + "learning_rate": 1.4180255858140243e-05, + "loss": 0.9591, + "step": 16208 + }, + { + "epoch": 0.38, + "grad_norm": 1.9097368496893963, + "learning_rate": 1.417956267257254e-05, + "loss": 0.961, + "step": 16209 + }, + { + "epoch": 0.38, + "grad_norm": 2.0492347525371004, + "learning_rate": 1.4178869462670102e-05, + "loss": 1.0166, + "step": 16210 + }, + { + "epoch": 0.38, + "grad_norm": 1.86447704566282, + "learning_rate": 1.417817622843696e-05, + "loss": 1.0801, + "step": 16211 + }, + { + "epoch": 0.38, + "grad_norm": 2.0903140969276746, + "learning_rate": 1.4177482969877155e-05, + "loss": 1.0631, + "step": 16212 + }, + { + "epoch": 0.38, + "grad_norm": 1.2319199615749383, + "learning_rate": 1.4176789686994718e-05, + "loss": 0.927, + "step": 16213 + }, + { + "epoch": 0.38, + "grad_norm": 1.8879332041615415, + "learning_rate": 1.4176096379793694e-05, + "loss": 1.038, + "step": 16214 + }, + { + "epoch": 0.38, + "grad_norm": 2.0848361980814922, + "learning_rate": 1.4175403048278113e-05, + "loss": 1.0554, + "step": 16215 + }, + { + "epoch": 0.38, + "grad_norm": 2.1852952093805205, + "learning_rate": 1.4174709692452014e-05, + "loss": 0.9336, + "step": 16216 + }, + { + "epoch": 0.38, + "grad_norm": 1.9090201490481655, + "learning_rate": 1.4174016312319432e-05, + "loss": 1.0571, + "step": 16217 + }, + { + "epoch": 0.38, + "grad_norm": 2.133486856189222, + "learning_rate": 1.4173322907884406e-05, + "loss": 0.933, + "step": 16218 + }, + { + "epoch": 0.38, + "grad_norm": 1.825050048127075, + "learning_rate": 1.4172629479150975e-05, + "loss": 1.0167, + "step": 16219 + }, + { + "epoch": 0.38, + "grad_norm": 1.0299819233306193, + "learning_rate": 1.417193602612317e-05, + "loss": 1.018, + "step": 16220 + }, + { + "epoch": 0.38, + "grad_norm": 1.9399611022970964, + "learning_rate": 1.4171242548805036e-05, + "loss": 1.0476, + "step": 16221 + }, + { + "epoch": 0.38, + "grad_norm": 1.942264828160083, + "learning_rate": 1.4170549047200604e-05, + "loss": 1.0392, + "step": 16222 + }, + { + "epoch": 0.38, + "grad_norm": 1.9707443953201156, + "learning_rate": 1.416985552131392e-05, + "loss": 0.9681, + "step": 16223 + }, + { + "epoch": 0.38, + "grad_norm": 1.1157601576623297, + "learning_rate": 1.416916197114901e-05, + "loss": 0.9569, + "step": 16224 + }, + { + "epoch": 0.38, + "grad_norm": 2.158860205397234, + "learning_rate": 1.4168468396709925e-05, + "loss": 1.0403, + "step": 16225 + }, + { + "epoch": 0.38, + "grad_norm": 1.5014359939718835, + "learning_rate": 1.4167774798000697e-05, + "loss": 0.9993, + "step": 16226 + }, + { + "epoch": 0.38, + "grad_norm": 2.227446782668706, + "learning_rate": 1.416708117502536e-05, + "loss": 0.993, + "step": 16227 + }, + { + "epoch": 0.38, + "grad_norm": 2.04060641731393, + "learning_rate": 1.4166387527787964e-05, + "loss": 1.0518, + "step": 16228 + }, + { + "epoch": 0.38, + "grad_norm": 2.01734094123421, + "learning_rate": 1.4165693856292535e-05, + "loss": 0.9458, + "step": 16229 + }, + { + "epoch": 0.38, + "grad_norm": 1.7995383511360767, + "learning_rate": 1.416500016054312e-05, + "loss": 0.9649, + "step": 16230 + }, + { + "epoch": 0.38, + "grad_norm": 1.9182718444403168, + "learning_rate": 1.4164306440543754e-05, + "loss": 1.1018, + "step": 16231 + }, + { + "epoch": 0.38, + "grad_norm": 4.095455274566016, + "learning_rate": 1.4163612696298481e-05, + "loss": 1.0518, + "step": 16232 + }, + { + "epoch": 0.38, + "grad_norm": 2.0156235983662576, + "learning_rate": 1.4162918927811334e-05, + "loss": 0.9708, + "step": 16233 + }, + { + "epoch": 0.38, + "grad_norm": 1.9585053313633534, + "learning_rate": 1.4162225135086354e-05, + "loss": 0.8756, + "step": 16234 + }, + { + "epoch": 0.38, + "grad_norm": 2.010301985037155, + "learning_rate": 1.4161531318127584e-05, + "loss": 1.0017, + "step": 16235 + }, + { + "epoch": 0.38, + "grad_norm": 2.249481114952915, + "learning_rate": 1.416083747693906e-05, + "loss": 1.0492, + "step": 16236 + }, + { + "epoch": 0.38, + "grad_norm": 1.9197282497042698, + "learning_rate": 1.4160143611524821e-05, + "loss": 1.1211, + "step": 16237 + }, + { + "epoch": 0.38, + "grad_norm": 2.091012436252627, + "learning_rate": 1.4159449721888911e-05, + "loss": 1.0408, + "step": 16238 + }, + { + "epoch": 0.38, + "grad_norm": 1.9129413425467285, + "learning_rate": 1.4158755808035367e-05, + "loss": 0.9526, + "step": 16239 + }, + { + "epoch": 0.38, + "grad_norm": 1.1228693720234113, + "learning_rate": 1.4158061869968232e-05, + "loss": 0.9284, + "step": 16240 + }, + { + "epoch": 0.38, + "grad_norm": 1.9716423576934377, + "learning_rate": 1.4157367907691542e-05, + "loss": 0.9858, + "step": 16241 + }, + { + "epoch": 0.38, + "grad_norm": 2.563601159564567, + "learning_rate": 1.4156673921209341e-05, + "loss": 0.9561, + "step": 16242 + }, + { + "epoch": 0.38, + "grad_norm": 2.1438062792831962, + "learning_rate": 1.4155979910525668e-05, + "loss": 1.0821, + "step": 16243 + }, + { + "epoch": 0.38, + "grad_norm": 1.9497145290311793, + "learning_rate": 1.4155285875644564e-05, + "loss": 1.0531, + "step": 16244 + }, + { + "epoch": 0.38, + "grad_norm": 2.1340095566496804, + "learning_rate": 1.4154591816570072e-05, + "loss": 1.0608, + "step": 16245 + }, + { + "epoch": 0.38, + "grad_norm": 2.241449586362752, + "learning_rate": 1.4153897733306228e-05, + "loss": 1.1081, + "step": 16246 + }, + { + "epoch": 0.38, + "grad_norm": 1.9692880724460753, + "learning_rate": 1.415320362585708e-05, + "loss": 0.9681, + "step": 16247 + }, + { + "epoch": 0.38, + "grad_norm": 2.016529844685217, + "learning_rate": 1.4152509494226662e-05, + "loss": 1.1654, + "step": 16248 + }, + { + "epoch": 0.38, + "grad_norm": 1.9582002640320328, + "learning_rate": 1.4151815338419022e-05, + "loss": 0.8401, + "step": 16249 + }, + { + "epoch": 0.38, + "grad_norm": 2.1595011004546967, + "learning_rate": 1.4151121158438195e-05, + "loss": 1.0239, + "step": 16250 + }, + { + "epoch": 0.38, + "grad_norm": 2.0253513863132193, + "learning_rate": 1.415042695428823e-05, + "loss": 1.0405, + "step": 16251 + }, + { + "epoch": 0.38, + "grad_norm": 2.083019718249259, + "learning_rate": 1.4149732725973162e-05, + "loss": 0.9324, + "step": 16252 + }, + { + "epoch": 0.38, + "grad_norm": 1.7377475012833266, + "learning_rate": 1.414903847349704e-05, + "loss": 1.0623, + "step": 16253 + }, + { + "epoch": 0.38, + "grad_norm": 2.322381193074211, + "learning_rate": 1.41483441968639e-05, + "loss": 1.0543, + "step": 16254 + }, + { + "epoch": 0.38, + "grad_norm": 2.8189977375968494, + "learning_rate": 1.4147649896077789e-05, + "loss": 0.9151, + "step": 16255 + }, + { + "epoch": 0.38, + "grad_norm": 1.9425869064211199, + "learning_rate": 1.4146955571142745e-05, + "loss": 1.1086, + "step": 16256 + }, + { + "epoch": 0.38, + "grad_norm": 1.908215203107895, + "learning_rate": 1.4146261222062813e-05, + "loss": 1.0558, + "step": 16257 + }, + { + "epoch": 0.38, + "grad_norm": 2.3940732194660903, + "learning_rate": 1.4145566848842036e-05, + "loss": 0.9876, + "step": 16258 + }, + { + "epoch": 0.38, + "grad_norm": 1.925745811145996, + "learning_rate": 1.4144872451484455e-05, + "loss": 1.1297, + "step": 16259 + }, + { + "epoch": 0.38, + "grad_norm": 2.0070614862039133, + "learning_rate": 1.4144178029994118e-05, + "loss": 1.0307, + "step": 16260 + }, + { + "epoch": 0.38, + "grad_norm": 3.0665623653968805, + "learning_rate": 1.414348358437506e-05, + "loss": 1.1171, + "step": 16261 + }, + { + "epoch": 0.38, + "grad_norm": 1.8101901472737072, + "learning_rate": 1.414278911463133e-05, + "loss": 1.1449, + "step": 16262 + }, + { + "epoch": 0.38, + "grad_norm": 1.9866485950226167, + "learning_rate": 1.4142094620766972e-05, + "loss": 1.1291, + "step": 16263 + }, + { + "epoch": 0.38, + "grad_norm": 1.8541824719107765, + "learning_rate": 1.4141400102786027e-05, + "loss": 1.0196, + "step": 16264 + }, + { + "epoch": 0.38, + "grad_norm": 1.07590256731442, + "learning_rate": 1.4140705560692538e-05, + "loss": 0.9541, + "step": 16265 + }, + { + "epoch": 0.38, + "grad_norm": 1.082492661669617, + "learning_rate": 1.4140010994490552e-05, + "loss": 0.9745, + "step": 16266 + }, + { + "epoch": 0.38, + "grad_norm": 1.9999721411537967, + "learning_rate": 1.4139316404184111e-05, + "loss": 0.8708, + "step": 16267 + }, + { + "epoch": 0.38, + "grad_norm": 2.3472740018648786, + "learning_rate": 1.4138621789777258e-05, + "loss": 1.099, + "step": 16268 + }, + { + "epoch": 0.38, + "grad_norm": 1.1208332775573808, + "learning_rate": 1.413792715127404e-05, + "loss": 0.9699, + "step": 16269 + }, + { + "epoch": 0.38, + "grad_norm": 2.023313733958919, + "learning_rate": 1.4137232488678503e-05, + "loss": 1.0506, + "step": 16270 + }, + { + "epoch": 0.38, + "grad_norm": 2.5081158816391715, + "learning_rate": 1.4136537801994683e-05, + "loss": 1.1318, + "step": 16271 + }, + { + "epoch": 0.38, + "grad_norm": 1.9195849602287656, + "learning_rate": 1.4135843091226635e-05, + "loss": 1.0385, + "step": 16272 + }, + { + "epoch": 0.38, + "grad_norm": 1.2236720641691374, + "learning_rate": 1.4135148356378393e-05, + "loss": 0.9093, + "step": 16273 + }, + { + "epoch": 0.38, + "grad_norm": 1.894931662290851, + "learning_rate": 1.4134453597454015e-05, + "loss": 1.0818, + "step": 16274 + }, + { + "epoch": 0.38, + "grad_norm": 2.0499173471215144, + "learning_rate": 1.4133758814457534e-05, + "loss": 0.9928, + "step": 16275 + }, + { + "epoch": 0.38, + "grad_norm": 2.0969330221959015, + "learning_rate": 1.4133064007393005e-05, + "loss": 1.0311, + "step": 16276 + }, + { + "epoch": 0.38, + "grad_norm": 2.190050501429109, + "learning_rate": 1.4132369176264467e-05, + "loss": 1.0435, + "step": 16277 + }, + { + "epoch": 0.38, + "grad_norm": 1.7447130193380063, + "learning_rate": 1.4131674321075966e-05, + "loss": 0.8724, + "step": 16278 + }, + { + "epoch": 0.38, + "grad_norm": 2.3255228280968065, + "learning_rate": 1.4130979441831552e-05, + "loss": 1.0214, + "step": 16279 + }, + { + "epoch": 0.38, + "grad_norm": 2.2062013753795453, + "learning_rate": 1.4130284538535261e-05, + "loss": 0.8522, + "step": 16280 + }, + { + "epoch": 0.38, + "grad_norm": 1.8213186507723544, + "learning_rate": 1.4129589611191154e-05, + "loss": 1.0353, + "step": 16281 + }, + { + "epoch": 0.38, + "grad_norm": 1.9054280342504353, + "learning_rate": 1.4128894659803263e-05, + "loss": 1.0853, + "step": 16282 + }, + { + "epoch": 0.38, + "grad_norm": 2.095867155229277, + "learning_rate": 1.4128199684375645e-05, + "loss": 0.9413, + "step": 16283 + }, + { + "epoch": 0.38, + "grad_norm": 1.8174439878073279, + "learning_rate": 1.412750468491234e-05, + "loss": 1.1367, + "step": 16284 + }, + { + "epoch": 0.38, + "grad_norm": 1.9737563255031583, + "learning_rate": 1.4126809661417396e-05, + "loss": 1.1737, + "step": 16285 + }, + { + "epoch": 0.38, + "grad_norm": 2.386838723623014, + "learning_rate": 1.4126114613894857e-05, + "loss": 1.0687, + "step": 16286 + }, + { + "epoch": 0.38, + "grad_norm": 2.035098473173569, + "learning_rate": 1.4125419542348775e-05, + "loss": 0.995, + "step": 16287 + }, + { + "epoch": 0.38, + "grad_norm": 2.1006743248673003, + "learning_rate": 1.4124724446783194e-05, + "loss": 1.0262, + "step": 16288 + }, + { + "epoch": 0.38, + "grad_norm": 1.9157448937282018, + "learning_rate": 1.412402932720216e-05, + "loss": 1.0506, + "step": 16289 + }, + { + "epoch": 0.38, + "grad_norm": 1.8988311693247168, + "learning_rate": 1.412333418360972e-05, + "loss": 1.0482, + "step": 16290 + }, + { + "epoch": 0.38, + "grad_norm": 1.0719484402290622, + "learning_rate": 1.4122639016009927e-05, + "loss": 1.0294, + "step": 16291 + }, + { + "epoch": 0.38, + "grad_norm": 1.997295457418338, + "learning_rate": 1.4121943824406822e-05, + "loss": 1.1098, + "step": 16292 + }, + { + "epoch": 0.38, + "grad_norm": 2.038239541249354, + "learning_rate": 1.4121248608804456e-05, + "loss": 1.1404, + "step": 16293 + }, + { + "epoch": 0.38, + "grad_norm": 1.0859261720567364, + "learning_rate": 1.4120553369206876e-05, + "loss": 0.8915, + "step": 16294 + }, + { + "epoch": 0.38, + "grad_norm": 1.8811830720410228, + "learning_rate": 1.4119858105618129e-05, + "loss": 0.9925, + "step": 16295 + }, + { + "epoch": 0.38, + "grad_norm": 2.2367343885469304, + "learning_rate": 1.4119162818042265e-05, + "loss": 1.0712, + "step": 16296 + }, + { + "epoch": 0.38, + "grad_norm": 2.0251847605568787, + "learning_rate": 1.411846750648333e-05, + "loss": 1.0515, + "step": 16297 + }, + { + "epoch": 0.38, + "grad_norm": 2.1744949950448262, + "learning_rate": 1.4117772170945375e-05, + "loss": 1.0376, + "step": 16298 + }, + { + "epoch": 0.38, + "grad_norm": 2.507334664201383, + "learning_rate": 1.4117076811432444e-05, + "loss": 0.9539, + "step": 16299 + }, + { + "epoch": 0.38, + "grad_norm": 1.9361374062575334, + "learning_rate": 1.4116381427948591e-05, + "loss": 0.9279, + "step": 16300 + }, + { + "epoch": 0.38, + "grad_norm": 1.9901051757208414, + "learning_rate": 1.4115686020497858e-05, + "loss": 1.1258, + "step": 16301 + }, + { + "epoch": 0.38, + "grad_norm": 2.1815530993996832, + "learning_rate": 1.4114990589084304e-05, + "loss": 0.9805, + "step": 16302 + }, + { + "epoch": 0.38, + "grad_norm": 2.125735584321084, + "learning_rate": 1.4114295133711966e-05, + "loss": 0.9701, + "step": 16303 + }, + { + "epoch": 0.38, + "grad_norm": 2.1219411198199634, + "learning_rate": 1.4113599654384904e-05, + "loss": 0.9205, + "step": 16304 + }, + { + "epoch": 0.38, + "grad_norm": 2.2826804573100654, + "learning_rate": 1.411290415110716e-05, + "loss": 1.0146, + "step": 16305 + }, + { + "epoch": 0.38, + "grad_norm": 2.078137847114053, + "learning_rate": 1.4112208623882787e-05, + "loss": 1.0544, + "step": 16306 + }, + { + "epoch": 0.38, + "grad_norm": 1.9826206393029104, + "learning_rate": 1.4111513072715833e-05, + "loss": 0.9648, + "step": 16307 + }, + { + "epoch": 0.38, + "grad_norm": 2.945930077870307, + "learning_rate": 1.4110817497610347e-05, + "loss": 1.1529, + "step": 16308 + }, + { + "epoch": 0.38, + "grad_norm": 2.052787294962418, + "learning_rate": 1.4110121898570382e-05, + "loss": 0.94, + "step": 16309 + }, + { + "epoch": 0.38, + "grad_norm": 2.7843012586271487, + "learning_rate": 1.4109426275599983e-05, + "loss": 0.9397, + "step": 16310 + }, + { + "epoch": 0.38, + "grad_norm": 2.0446022874947194, + "learning_rate": 1.4108730628703207e-05, + "loss": 0.9641, + "step": 16311 + }, + { + "epoch": 0.38, + "grad_norm": 2.2338718557123904, + "learning_rate": 1.4108034957884096e-05, + "loss": 1.0335, + "step": 16312 + }, + { + "epoch": 0.38, + "grad_norm": 2.2245535131813963, + "learning_rate": 1.4107339263146709e-05, + "loss": 0.9841, + "step": 16313 + }, + { + "epoch": 0.38, + "grad_norm": 2.3965566406802945, + "learning_rate": 1.4106643544495092e-05, + "loss": 0.9734, + "step": 16314 + }, + { + "epoch": 0.38, + "grad_norm": 1.1227959279938378, + "learning_rate": 1.4105947801933295e-05, + "loss": 0.9843, + "step": 16315 + }, + { + "epoch": 0.38, + "grad_norm": 2.065322299853001, + "learning_rate": 1.4105252035465368e-05, + "loss": 1.0237, + "step": 16316 + }, + { + "epoch": 0.38, + "grad_norm": 2.0298003593492133, + "learning_rate": 1.4104556245095368e-05, + "loss": 1.0257, + "step": 16317 + }, + { + "epoch": 0.38, + "grad_norm": 1.957776106052703, + "learning_rate": 1.4103860430827342e-05, + "loss": 1.0959, + "step": 16318 + }, + { + "epoch": 0.38, + "grad_norm": 2.073894734778262, + "learning_rate": 1.4103164592665337e-05, + "loss": 1.1094, + "step": 16319 + }, + { + "epoch": 0.38, + "grad_norm": 1.944481938909533, + "learning_rate": 1.4102468730613412e-05, + "loss": 1.02, + "step": 16320 + }, + { + "epoch": 0.38, + "grad_norm": 2.3265735497522106, + "learning_rate": 1.4101772844675617e-05, + "loss": 1.0408, + "step": 16321 + }, + { + "epoch": 0.38, + "grad_norm": 1.9973384382222727, + "learning_rate": 1.4101076934855997e-05, + "loss": 1.2351, + "step": 16322 + }, + { + "epoch": 0.38, + "grad_norm": 1.8052965207771783, + "learning_rate": 1.4100381001158612e-05, + "loss": 1.0821, + "step": 16323 + }, + { + "epoch": 0.38, + "grad_norm": 2.0058586942184986, + "learning_rate": 1.409968504358751e-05, + "loss": 1.0203, + "step": 16324 + }, + { + "epoch": 0.38, + "grad_norm": 2.051342696469808, + "learning_rate": 1.4098989062146741e-05, + "loss": 0.9965, + "step": 16325 + }, + { + "epoch": 0.38, + "grad_norm": 2.1261005065405087, + "learning_rate": 1.4098293056840364e-05, + "loss": 1.0414, + "step": 16326 + }, + { + "epoch": 0.38, + "grad_norm": 1.0944759114276241, + "learning_rate": 1.4097597027672425e-05, + "loss": 0.9644, + "step": 16327 + }, + { + "epoch": 0.38, + "grad_norm": 1.127784156286918, + "learning_rate": 1.409690097464698e-05, + "loss": 1.0209, + "step": 16328 + }, + { + "epoch": 0.38, + "grad_norm": 2.115496665405673, + "learning_rate": 1.409620489776808e-05, + "loss": 0.9496, + "step": 16329 + }, + { + "epoch": 0.38, + "grad_norm": 6.6963684341719105, + "learning_rate": 1.4095508797039781e-05, + "loss": 0.9686, + "step": 16330 + }, + { + "epoch": 0.38, + "grad_norm": 1.9184634587711074, + "learning_rate": 1.4094812672466127e-05, + "loss": 0.9514, + "step": 16331 + }, + { + "epoch": 0.38, + "grad_norm": 2.2328102362444757, + "learning_rate": 1.4094116524051182e-05, + "loss": 1.0114, + "step": 16332 + }, + { + "epoch": 0.38, + "grad_norm": 2.130970662433837, + "learning_rate": 1.4093420351798991e-05, + "loss": 0.9228, + "step": 16333 + }, + { + "epoch": 0.38, + "grad_norm": 2.0411873801073535, + "learning_rate": 1.4092724155713614e-05, + "loss": 1.0217, + "step": 16334 + }, + { + "epoch": 0.38, + "grad_norm": 1.9963461227405876, + "learning_rate": 1.4092027935799102e-05, + "loss": 1.0423, + "step": 16335 + }, + { + "epoch": 0.38, + "grad_norm": 2.377098875290727, + "learning_rate": 1.4091331692059506e-05, + "loss": 0.9631, + "step": 16336 + }, + { + "epoch": 0.38, + "grad_norm": 2.121972308570749, + "learning_rate": 1.4090635424498882e-05, + "loss": 1.0991, + "step": 16337 + }, + { + "epoch": 0.38, + "grad_norm": 1.952927873498349, + "learning_rate": 1.4089939133121282e-05, + "loss": 1.118, + "step": 16338 + }, + { + "epoch": 0.38, + "grad_norm": 1.9404828355169534, + "learning_rate": 1.4089242817930764e-05, + "loss": 1.0884, + "step": 16339 + }, + { + "epoch": 0.38, + "grad_norm": 1.7838476909075558, + "learning_rate": 1.4088546478931378e-05, + "loss": 0.9553, + "step": 16340 + }, + { + "epoch": 0.38, + "grad_norm": 2.358363634646446, + "learning_rate": 1.408785011612718e-05, + "loss": 1.0357, + "step": 16341 + }, + { + "epoch": 0.39, + "grad_norm": 2.119246159949295, + "learning_rate": 1.4087153729522224e-05, + "loss": 0.9853, + "step": 16342 + }, + { + "epoch": 0.39, + "grad_norm": 1.9743277431912953, + "learning_rate": 1.4086457319120566e-05, + "loss": 0.9482, + "step": 16343 + }, + { + "epoch": 0.39, + "grad_norm": 2.1178950359960163, + "learning_rate": 1.408576088492626e-05, + "loss": 1.0083, + "step": 16344 + }, + { + "epoch": 0.39, + "grad_norm": 1.748048231987453, + "learning_rate": 1.408506442694336e-05, + "loss": 0.9576, + "step": 16345 + }, + { + "epoch": 0.39, + "grad_norm": 2.2157986292801573, + "learning_rate": 1.408436794517592e-05, + "loss": 0.9567, + "step": 16346 + }, + { + "epoch": 0.39, + "grad_norm": 2.229567114474939, + "learning_rate": 1.4083671439628e-05, + "loss": 1.1127, + "step": 16347 + }, + { + "epoch": 0.39, + "grad_norm": 1.8672105384247302, + "learning_rate": 1.4082974910303648e-05, + "loss": 1.1181, + "step": 16348 + }, + { + "epoch": 0.39, + "grad_norm": 1.1544383457712124, + "learning_rate": 1.4082278357206928e-05, + "loss": 0.9964, + "step": 16349 + }, + { + "epoch": 0.39, + "grad_norm": 2.9915189442798154, + "learning_rate": 1.4081581780341887e-05, + "loss": 1.2024, + "step": 16350 + }, + { + "epoch": 0.39, + "grad_norm": 2.1622812804532536, + "learning_rate": 1.4080885179712588e-05, + "loss": 1.0604, + "step": 16351 + }, + { + "epoch": 0.39, + "grad_norm": 1.6606212872036827, + "learning_rate": 1.4080188555323079e-05, + "loss": 1.0455, + "step": 16352 + }, + { + "epoch": 0.39, + "grad_norm": 1.95820931702224, + "learning_rate": 1.4079491907177428e-05, + "loss": 0.99, + "step": 16353 + }, + { + "epoch": 0.39, + "grad_norm": 1.1440865237288826, + "learning_rate": 1.4078795235279677e-05, + "loss": 1.0422, + "step": 16354 + }, + { + "epoch": 0.39, + "grad_norm": 2.6645565245232494, + "learning_rate": 1.407809853963389e-05, + "loss": 1.0377, + "step": 16355 + }, + { + "epoch": 0.39, + "grad_norm": 2.3021797736962415, + "learning_rate": 1.4077401820244125e-05, + "loss": 1.0369, + "step": 16356 + }, + { + "epoch": 0.39, + "grad_norm": 2.142293178408823, + "learning_rate": 1.4076705077114432e-05, + "loss": 1.1263, + "step": 16357 + }, + { + "epoch": 0.39, + "grad_norm": 2.4098215954708957, + "learning_rate": 1.4076008310248874e-05, + "loss": 1.0028, + "step": 16358 + }, + { + "epoch": 0.39, + "grad_norm": 2.7122985237125987, + "learning_rate": 1.4075311519651502e-05, + "loss": 1.0925, + "step": 16359 + }, + { + "epoch": 0.39, + "grad_norm": 1.95582208307407, + "learning_rate": 1.4074614705326379e-05, + "loss": 0.9778, + "step": 16360 + }, + { + "epoch": 0.39, + "grad_norm": 1.9377352729475263, + "learning_rate": 1.4073917867277557e-05, + "loss": 0.987, + "step": 16361 + }, + { + "epoch": 0.39, + "grad_norm": 2.4904589748074053, + "learning_rate": 1.40732210055091e-05, + "loss": 0.9407, + "step": 16362 + }, + { + "epoch": 0.39, + "grad_norm": 1.9782182379016646, + "learning_rate": 1.4072524120025057e-05, + "loss": 1.0545, + "step": 16363 + }, + { + "epoch": 0.39, + "grad_norm": 2.0640805466734355, + "learning_rate": 1.4071827210829493e-05, + "loss": 1.0611, + "step": 16364 + }, + { + "epoch": 0.39, + "grad_norm": 2.0179095703410743, + "learning_rate": 1.4071130277926457e-05, + "loss": 0.8731, + "step": 16365 + }, + { + "epoch": 0.39, + "grad_norm": 3.6066406021781288, + "learning_rate": 1.4070433321320012e-05, + "loss": 1.1148, + "step": 16366 + }, + { + "epoch": 0.39, + "grad_norm": 2.6256996541126725, + "learning_rate": 1.4069736341014218e-05, + "loss": 0.921, + "step": 16367 + }, + { + "epoch": 0.39, + "grad_norm": 1.8943901031554242, + "learning_rate": 1.406903933701313e-05, + "loss": 1.0685, + "step": 16368 + }, + { + "epoch": 0.39, + "grad_norm": 1.7991303015754228, + "learning_rate": 1.4068342309320807e-05, + "loss": 0.9596, + "step": 16369 + }, + { + "epoch": 0.39, + "grad_norm": 2.0266152747083295, + "learning_rate": 1.4067645257941308e-05, + "loss": 1.0623, + "step": 16370 + }, + { + "epoch": 0.39, + "grad_norm": 2.3192124857924266, + "learning_rate": 1.4066948182878686e-05, + "loss": 1.0979, + "step": 16371 + }, + { + "epoch": 0.39, + "grad_norm": 2.057664758415686, + "learning_rate": 1.4066251084137006e-05, + "loss": 1.0295, + "step": 16372 + }, + { + "epoch": 0.39, + "grad_norm": 1.9996350296945358, + "learning_rate": 1.4065553961720327e-05, + "loss": 1.0238, + "step": 16373 + }, + { + "epoch": 0.39, + "grad_norm": 1.19278725666363, + "learning_rate": 1.4064856815632703e-05, + "loss": 0.9648, + "step": 16374 + }, + { + "epoch": 0.39, + "grad_norm": 1.1006747633470062, + "learning_rate": 1.4064159645878197e-05, + "loss": 0.9093, + "step": 16375 + }, + { + "epoch": 0.39, + "grad_norm": 1.8443473453405244, + "learning_rate": 1.4063462452460868e-05, + "loss": 1.113, + "step": 16376 + }, + { + "epoch": 0.39, + "grad_norm": 2.144895205084876, + "learning_rate": 1.4062765235384772e-05, + "loss": 1.0563, + "step": 16377 + }, + { + "epoch": 0.39, + "grad_norm": 2.2270099382726363, + "learning_rate": 1.4062067994653971e-05, + "loss": 0.8852, + "step": 16378 + }, + { + "epoch": 0.39, + "grad_norm": 2.4768799471547065, + "learning_rate": 1.4061370730272527e-05, + "loss": 1.0294, + "step": 16379 + }, + { + "epoch": 0.39, + "grad_norm": 1.9400437142274023, + "learning_rate": 1.4060673442244493e-05, + "loss": 1.0213, + "step": 16380 + }, + { + "epoch": 0.39, + "grad_norm": 2.343667133674383, + "learning_rate": 1.4059976130573936e-05, + "loss": 1.0836, + "step": 16381 + }, + { + "epoch": 0.39, + "grad_norm": 1.9972322484608662, + "learning_rate": 1.4059278795264908e-05, + "loss": 1.1452, + "step": 16382 + }, + { + "epoch": 0.39, + "grad_norm": 1.9897769203021392, + "learning_rate": 1.4058581436321478e-05, + "loss": 0.997, + "step": 16383 + }, + { + "epoch": 0.39, + "grad_norm": 2.253526411567548, + "learning_rate": 1.4057884053747699e-05, + "loss": 1.0187, + "step": 16384 + }, + { + "epoch": 0.39, + "grad_norm": 2.0830969118537426, + "learning_rate": 1.4057186647547638e-05, + "loss": 1.096, + "step": 16385 + }, + { + "epoch": 0.39, + "grad_norm": 2.4398390022478074, + "learning_rate": 1.405648921772535e-05, + "loss": 0.9229, + "step": 16386 + }, + { + "epoch": 0.39, + "grad_norm": 1.8895488381320362, + "learning_rate": 1.4055791764284897e-05, + "loss": 0.8811, + "step": 16387 + }, + { + "epoch": 0.39, + "grad_norm": 2.079547992723167, + "learning_rate": 1.405509428723034e-05, + "loss": 0.9719, + "step": 16388 + }, + { + "epoch": 0.39, + "grad_norm": 1.9265613615872044, + "learning_rate": 1.405439678656574e-05, + "loss": 1.0091, + "step": 16389 + }, + { + "epoch": 0.39, + "grad_norm": 2.309090635074325, + "learning_rate": 1.405369926229516e-05, + "loss": 1.0388, + "step": 16390 + }, + { + "epoch": 0.39, + "grad_norm": 2.0969667049460834, + "learning_rate": 1.4053001714422658e-05, + "loss": 0.9509, + "step": 16391 + }, + { + "epoch": 0.39, + "grad_norm": 1.8698172125752572, + "learning_rate": 1.4052304142952297e-05, + "loss": 0.9674, + "step": 16392 + }, + { + "epoch": 0.39, + "grad_norm": 2.2296190513979104, + "learning_rate": 1.405160654788814e-05, + "loss": 0.8659, + "step": 16393 + }, + { + "epoch": 0.39, + "grad_norm": 2.2145107482965294, + "learning_rate": 1.4050908929234245e-05, + "loss": 0.9415, + "step": 16394 + }, + { + "epoch": 0.39, + "grad_norm": 2.1120529277740374, + "learning_rate": 1.4050211286994678e-05, + "loss": 1.1114, + "step": 16395 + }, + { + "epoch": 0.39, + "grad_norm": 1.915155130046633, + "learning_rate": 1.4049513621173496e-05, + "loss": 1.0439, + "step": 16396 + }, + { + "epoch": 0.39, + "grad_norm": 2.43687392410642, + "learning_rate": 1.4048815931774764e-05, + "loss": 1.0577, + "step": 16397 + }, + { + "epoch": 0.39, + "grad_norm": 2.0848705285546867, + "learning_rate": 1.4048118218802545e-05, + "loss": 1.0808, + "step": 16398 + }, + { + "epoch": 0.39, + "grad_norm": 1.9583123271739498, + "learning_rate": 1.40474204822609e-05, + "loss": 0.9325, + "step": 16399 + }, + { + "epoch": 0.39, + "grad_norm": 1.921210880061593, + "learning_rate": 1.4046722722153891e-05, + "loss": 0.9738, + "step": 16400 + }, + { + "epoch": 0.39, + "grad_norm": 2.1593259959596804, + "learning_rate": 1.4046024938485581e-05, + "loss": 1.0719, + "step": 16401 + }, + { + "epoch": 0.39, + "grad_norm": 2.1831133523332418, + "learning_rate": 1.4045327131260033e-05, + "loss": 1.0847, + "step": 16402 + }, + { + "epoch": 0.39, + "grad_norm": 1.7897554942731475, + "learning_rate": 1.404462930048131e-05, + "loss": 0.9136, + "step": 16403 + }, + { + "epoch": 0.39, + "grad_norm": 1.1593834035978852, + "learning_rate": 1.4043931446153476e-05, + "loss": 0.9465, + "step": 16404 + }, + { + "epoch": 0.39, + "grad_norm": 1.1554040794638998, + "learning_rate": 1.404323356828059e-05, + "loss": 0.9124, + "step": 16405 + }, + { + "epoch": 0.39, + "grad_norm": 1.7680193763910659, + "learning_rate": 1.404253566686672e-05, + "loss": 1.0354, + "step": 16406 + }, + { + "epoch": 0.39, + "grad_norm": 2.0980217544626387, + "learning_rate": 1.404183774191593e-05, + "loss": 1.0043, + "step": 16407 + }, + { + "epoch": 0.39, + "grad_norm": 2.096532208484453, + "learning_rate": 1.4041139793432274e-05, + "loss": 0.9792, + "step": 16408 + }, + { + "epoch": 0.39, + "grad_norm": 1.057504766174281, + "learning_rate": 1.4040441821419829e-05, + "loss": 0.9234, + "step": 16409 + }, + { + "epoch": 0.39, + "grad_norm": 2.2319134110644194, + "learning_rate": 1.4039743825882648e-05, + "loss": 1.0335, + "step": 16410 + }, + { + "epoch": 0.39, + "grad_norm": 2.645374331340337, + "learning_rate": 1.4039045806824802e-05, + "loss": 0.9523, + "step": 16411 + }, + { + "epoch": 0.39, + "grad_norm": 1.9404350886872084, + "learning_rate": 1.4038347764250353e-05, + "loss": 1.0226, + "step": 16412 + }, + { + "epoch": 0.39, + "grad_norm": 1.9855643414973614, + "learning_rate": 1.4037649698163364e-05, + "loss": 1.0626, + "step": 16413 + }, + { + "epoch": 0.39, + "grad_norm": 2.4996820562439668, + "learning_rate": 1.4036951608567901e-05, + "loss": 0.9491, + "step": 16414 + }, + { + "epoch": 0.39, + "grad_norm": 1.8991621421618334, + "learning_rate": 1.4036253495468026e-05, + "loss": 0.9397, + "step": 16415 + }, + { + "epoch": 0.39, + "grad_norm": 2.1249982861466368, + "learning_rate": 1.4035555358867807e-05, + "loss": 0.9956, + "step": 16416 + }, + { + "epoch": 0.39, + "grad_norm": 2.3348246751622104, + "learning_rate": 1.4034857198771308e-05, + "loss": 0.9111, + "step": 16417 + }, + { + "epoch": 0.39, + "grad_norm": 1.8059730034682284, + "learning_rate": 1.4034159015182591e-05, + "loss": 1.0548, + "step": 16418 + }, + { + "epoch": 0.39, + "grad_norm": 1.9734101957964088, + "learning_rate": 1.4033460808105723e-05, + "loss": 0.9282, + "step": 16419 + }, + { + "epoch": 0.39, + "grad_norm": 2.0562959037511592, + "learning_rate": 1.403276257754477e-05, + "loss": 1.1264, + "step": 16420 + }, + { + "epoch": 0.39, + "grad_norm": 1.8562032942162086, + "learning_rate": 1.4032064323503796e-05, + "loss": 0.9676, + "step": 16421 + }, + { + "epoch": 0.39, + "grad_norm": 1.7704603284953122, + "learning_rate": 1.4031366045986867e-05, + "loss": 1.1245, + "step": 16422 + }, + { + "epoch": 0.39, + "grad_norm": 1.7596565989719724, + "learning_rate": 1.4030667744998048e-05, + "loss": 1.0049, + "step": 16423 + }, + { + "epoch": 0.39, + "grad_norm": 1.9412107756612986, + "learning_rate": 1.4029969420541406e-05, + "loss": 1.02, + "step": 16424 + }, + { + "epoch": 0.39, + "grad_norm": 2.2213973566015994, + "learning_rate": 1.4029271072621007e-05, + "loss": 0.9761, + "step": 16425 + }, + { + "epoch": 0.39, + "grad_norm": 2.360611969484124, + "learning_rate": 1.4028572701240915e-05, + "loss": 1.2161, + "step": 16426 + }, + { + "epoch": 0.39, + "grad_norm": 2.505339833073663, + "learning_rate": 1.4027874306405196e-05, + "loss": 1.1419, + "step": 16427 + }, + { + "epoch": 0.39, + "grad_norm": 2.0385288164941997, + "learning_rate": 1.4027175888117921e-05, + "loss": 0.9464, + "step": 16428 + }, + { + "epoch": 0.39, + "grad_norm": 1.8284781868285704, + "learning_rate": 1.4026477446383149e-05, + "loss": 1.0161, + "step": 16429 + }, + { + "epoch": 0.39, + "grad_norm": 2.07119447133783, + "learning_rate": 1.4025778981204954e-05, + "loss": 0.986, + "step": 16430 + }, + { + "epoch": 0.39, + "grad_norm": 1.9271602194688715, + "learning_rate": 1.4025080492587397e-05, + "loss": 0.874, + "step": 16431 + }, + { + "epoch": 0.39, + "grad_norm": 1.8415478834371286, + "learning_rate": 1.4024381980534549e-05, + "loss": 1.1524, + "step": 16432 + }, + { + "epoch": 0.39, + "grad_norm": 2.3293219234179134, + "learning_rate": 1.4023683445050473e-05, + "loss": 0.9025, + "step": 16433 + }, + { + "epoch": 0.39, + "grad_norm": 2.2266593385016242, + "learning_rate": 1.4022984886139241e-05, + "loss": 1.0375, + "step": 16434 + }, + { + "epoch": 0.39, + "grad_norm": 1.9797015819212436, + "learning_rate": 1.4022286303804912e-05, + "loss": 0.9143, + "step": 16435 + }, + { + "epoch": 0.39, + "grad_norm": 1.1435685125381794, + "learning_rate": 1.4021587698051567e-05, + "loss": 0.9276, + "step": 16436 + }, + { + "epoch": 0.39, + "grad_norm": 1.8862466932404278, + "learning_rate": 1.4020889068883258e-05, + "loss": 1.0532, + "step": 16437 + }, + { + "epoch": 0.39, + "grad_norm": 2.172293605638813, + "learning_rate": 1.4020190416304062e-05, + "loss": 1.0308, + "step": 16438 + }, + { + "epoch": 0.39, + "grad_norm": 2.222785289053138, + "learning_rate": 1.4019491740318045e-05, + "loss": 1.1118, + "step": 16439 + }, + { + "epoch": 0.39, + "grad_norm": 2.3136692416747544, + "learning_rate": 1.4018793040929275e-05, + "loss": 0.9908, + "step": 16440 + }, + { + "epoch": 0.39, + "grad_norm": 2.1794033500359467, + "learning_rate": 1.401809431814182e-05, + "loss": 1.0779, + "step": 16441 + }, + { + "epoch": 0.39, + "grad_norm": 2.0375858789780525, + "learning_rate": 1.4017395571959745e-05, + "loss": 0.8819, + "step": 16442 + }, + { + "epoch": 0.39, + "grad_norm": 2.346988346931674, + "learning_rate": 1.4016696802387121e-05, + "loss": 1.0228, + "step": 16443 + }, + { + "epoch": 0.39, + "grad_norm": 1.14859153746859, + "learning_rate": 1.401599800942802e-05, + "loss": 1.0185, + "step": 16444 + }, + { + "epoch": 0.39, + "grad_norm": 2.1456792821068147, + "learning_rate": 1.4015299193086507e-05, + "loss": 0.9932, + "step": 16445 + }, + { + "epoch": 0.39, + "grad_norm": 1.9340942026577301, + "learning_rate": 1.4014600353366648e-05, + "loss": 1.0785, + "step": 16446 + }, + { + "epoch": 0.39, + "grad_norm": 2.4042066189940905, + "learning_rate": 1.4013901490272517e-05, + "loss": 1.0516, + "step": 16447 + }, + { + "epoch": 0.39, + "grad_norm": 2.2073144198311847, + "learning_rate": 1.401320260380818e-05, + "loss": 1.0824, + "step": 16448 + }, + { + "epoch": 0.39, + "grad_norm": 2.035350549750428, + "learning_rate": 1.4012503693977705e-05, + "loss": 0.9937, + "step": 16449 + }, + { + "epoch": 0.39, + "grad_norm": 2.0300917294112293, + "learning_rate": 1.4011804760785165e-05, + "loss": 1.0333, + "step": 16450 + }, + { + "epoch": 0.39, + "grad_norm": 1.9185732069396417, + "learning_rate": 1.401110580423463e-05, + "loss": 0.9678, + "step": 16451 + }, + { + "epoch": 0.39, + "grad_norm": 2.024473610488459, + "learning_rate": 1.4010406824330162e-05, + "loss": 1.0126, + "step": 16452 + }, + { + "epoch": 0.39, + "grad_norm": 3.256689362108634, + "learning_rate": 1.4009707821075843e-05, + "loss": 1.1332, + "step": 16453 + }, + { + "epoch": 0.39, + "grad_norm": 1.1163976207406998, + "learning_rate": 1.4009008794475726e-05, + "loss": 1.0165, + "step": 16454 + }, + { + "epoch": 0.39, + "grad_norm": 2.194467965535429, + "learning_rate": 1.40083097445339e-05, + "loss": 0.9433, + "step": 16455 + }, + { + "epoch": 0.39, + "grad_norm": 1.1455503603813213, + "learning_rate": 1.4007610671254417e-05, + "loss": 1.0056, + "step": 16456 + }, + { + "epoch": 0.39, + "grad_norm": 1.8878577476253646, + "learning_rate": 1.4006911574641363e-05, + "loss": 1.1303, + "step": 16457 + }, + { + "epoch": 0.39, + "grad_norm": 1.9284444276555714, + "learning_rate": 1.4006212454698798e-05, + "loss": 0.9698, + "step": 16458 + }, + { + "epoch": 0.39, + "grad_norm": 1.9039397901358015, + "learning_rate": 1.4005513311430795e-05, + "loss": 0.9903, + "step": 16459 + }, + { + "epoch": 0.39, + "grad_norm": 2.053468091721898, + "learning_rate": 1.4004814144841428e-05, + "loss": 1.0261, + "step": 16460 + }, + { + "epoch": 0.39, + "grad_norm": 1.8758839659349809, + "learning_rate": 1.4004114954934762e-05, + "loss": 1.171, + "step": 16461 + }, + { + "epoch": 0.39, + "grad_norm": 2.700643861051089, + "learning_rate": 1.4003415741714874e-05, + "loss": 1.0149, + "step": 16462 + }, + { + "epoch": 0.39, + "grad_norm": 1.0620396017382425, + "learning_rate": 1.4002716505185828e-05, + "loss": 0.9646, + "step": 16463 + }, + { + "epoch": 0.39, + "grad_norm": 2.5067520873390903, + "learning_rate": 1.4002017245351705e-05, + "loss": 1.1561, + "step": 16464 + }, + { + "epoch": 0.39, + "grad_norm": 1.1488906076383387, + "learning_rate": 1.4001317962216568e-05, + "loss": 1.0363, + "step": 16465 + }, + { + "epoch": 0.39, + "grad_norm": 2.0271984928132993, + "learning_rate": 1.400061865578449e-05, + "loss": 0.9806, + "step": 16466 + }, + { + "epoch": 0.39, + "grad_norm": 2.4628536766496465, + "learning_rate": 1.3999919326059543e-05, + "loss": 1.0758, + "step": 16467 + }, + { + "epoch": 0.39, + "grad_norm": 1.8841216400995138, + "learning_rate": 1.3999219973045802e-05, + "loss": 0.982, + "step": 16468 + }, + { + "epoch": 0.39, + "grad_norm": 2.2095027603290442, + "learning_rate": 1.3998520596747333e-05, + "loss": 1.2026, + "step": 16469 + }, + { + "epoch": 0.39, + "grad_norm": 2.071808247354842, + "learning_rate": 1.3997821197168214e-05, + "loss": 1.0031, + "step": 16470 + }, + { + "epoch": 0.39, + "grad_norm": 1.1236328569926441, + "learning_rate": 1.3997121774312512e-05, + "loss": 1.0394, + "step": 16471 + }, + { + "epoch": 0.39, + "grad_norm": 1.906505820260193, + "learning_rate": 1.3996422328184305e-05, + "loss": 1.053, + "step": 16472 + }, + { + "epoch": 0.39, + "grad_norm": 2.0703844001718616, + "learning_rate": 1.3995722858787658e-05, + "loss": 1.0028, + "step": 16473 + }, + { + "epoch": 0.39, + "grad_norm": 2.0768478211441326, + "learning_rate": 1.399502336612665e-05, + "loss": 1.0086, + "step": 16474 + }, + { + "epoch": 0.39, + "grad_norm": 1.8838733309535611, + "learning_rate": 1.3994323850205349e-05, + "loss": 0.9914, + "step": 16475 + }, + { + "epoch": 0.39, + "grad_norm": 1.8950243055744964, + "learning_rate": 1.3993624311027833e-05, + "loss": 1.0127, + "step": 16476 + }, + { + "epoch": 0.39, + "grad_norm": 2.073645801760407, + "learning_rate": 1.399292474859817e-05, + "loss": 1.094, + "step": 16477 + }, + { + "epoch": 0.39, + "grad_norm": 2.2389957300186265, + "learning_rate": 1.3992225162920434e-05, + "loss": 0.9204, + "step": 16478 + }, + { + "epoch": 0.39, + "grad_norm": 2.354098350257964, + "learning_rate": 1.3991525553998704e-05, + "loss": 1.1181, + "step": 16479 + }, + { + "epoch": 0.39, + "grad_norm": 2.026399300734438, + "learning_rate": 1.399082592183704e-05, + "loss": 0.9967, + "step": 16480 + }, + { + "epoch": 0.39, + "grad_norm": 1.777830734844411, + "learning_rate": 1.3990126266439532e-05, + "loss": 1.0651, + "step": 16481 + }, + { + "epoch": 0.39, + "grad_norm": 1.9546328007906555, + "learning_rate": 1.3989426587810241e-05, + "loss": 0.9025, + "step": 16482 + }, + { + "epoch": 0.39, + "grad_norm": 1.789509914472809, + "learning_rate": 1.398872688595325e-05, + "loss": 0.9762, + "step": 16483 + }, + { + "epoch": 0.39, + "grad_norm": 1.901333303901494, + "learning_rate": 1.3988027160872624e-05, + "loss": 0.961, + "step": 16484 + }, + { + "epoch": 0.39, + "grad_norm": 1.892373649170146, + "learning_rate": 1.3987327412572442e-05, + "loss": 1.078, + "step": 16485 + }, + { + "epoch": 0.39, + "grad_norm": 1.0649151232561849, + "learning_rate": 1.3986627641056777e-05, + "loss": 0.9974, + "step": 16486 + }, + { + "epoch": 0.39, + "grad_norm": 1.923282389401258, + "learning_rate": 1.3985927846329705e-05, + "loss": 1.0823, + "step": 16487 + }, + { + "epoch": 0.39, + "grad_norm": 2.1696420814501143, + "learning_rate": 1.3985228028395298e-05, + "loss": 0.9732, + "step": 16488 + }, + { + "epoch": 0.39, + "grad_norm": 2.1711614728693323, + "learning_rate": 1.3984528187257636e-05, + "loss": 0.9852, + "step": 16489 + }, + { + "epoch": 0.39, + "grad_norm": 4.1688546722993065, + "learning_rate": 1.3983828322920786e-05, + "loss": 1.035, + "step": 16490 + }, + { + "epoch": 0.39, + "grad_norm": 2.0131433336264863, + "learning_rate": 1.3983128435388824e-05, + "loss": 0.9983, + "step": 16491 + }, + { + "epoch": 0.39, + "grad_norm": 2.003809355601732, + "learning_rate": 1.398242852466583e-05, + "loss": 1.0673, + "step": 16492 + }, + { + "epoch": 0.39, + "grad_norm": 2.2101730296740896, + "learning_rate": 1.3981728590755877e-05, + "loss": 0.9665, + "step": 16493 + }, + { + "epoch": 0.39, + "grad_norm": 2.2198223170141036, + "learning_rate": 1.3981028633663036e-05, + "loss": 1.0584, + "step": 16494 + }, + { + "epoch": 0.39, + "grad_norm": 2.013530322926224, + "learning_rate": 1.3980328653391389e-05, + "loss": 1.0449, + "step": 16495 + }, + { + "epoch": 0.39, + "grad_norm": 1.9871202634018532, + "learning_rate": 1.3979628649945007e-05, + "loss": 0.8519, + "step": 16496 + }, + { + "epoch": 0.39, + "grad_norm": 1.944658586386399, + "learning_rate": 1.3978928623327969e-05, + "loss": 0.9839, + "step": 16497 + }, + { + "epoch": 0.39, + "grad_norm": 1.0450862418786546, + "learning_rate": 1.3978228573544348e-05, + "loss": 0.8881, + "step": 16498 + }, + { + "epoch": 0.39, + "grad_norm": 2.081479420475983, + "learning_rate": 1.3977528500598218e-05, + "loss": 1.026, + "step": 16499 + }, + { + "epoch": 0.39, + "grad_norm": 2.294271298984707, + "learning_rate": 1.3976828404493662e-05, + "loss": 0.9147, + "step": 16500 + }, + { + "epoch": 0.39, + "grad_norm": 1.928301348110476, + "learning_rate": 1.3976128285234748e-05, + "loss": 1.0604, + "step": 16501 + }, + { + "epoch": 0.39, + "grad_norm": 2.1300978035558233, + "learning_rate": 1.3975428142825562e-05, + "loss": 1.1068, + "step": 16502 + }, + { + "epoch": 0.39, + "grad_norm": 1.9554774404617263, + "learning_rate": 1.3974727977270167e-05, + "loss": 1.1366, + "step": 16503 + }, + { + "epoch": 0.39, + "grad_norm": 2.324781521715652, + "learning_rate": 1.3974027788572655e-05, + "loss": 1.0744, + "step": 16504 + }, + { + "epoch": 0.39, + "grad_norm": 1.9354075354362354, + "learning_rate": 1.397332757673709e-05, + "loss": 1.0786, + "step": 16505 + }, + { + "epoch": 0.39, + "grad_norm": 1.1042273589276743, + "learning_rate": 1.3972627341767555e-05, + "loss": 0.9947, + "step": 16506 + }, + { + "epoch": 0.39, + "grad_norm": 2.224292294312575, + "learning_rate": 1.3971927083668127e-05, + "loss": 1.1583, + "step": 16507 + }, + { + "epoch": 0.39, + "grad_norm": 1.080423703900135, + "learning_rate": 1.397122680244288e-05, + "loss": 1.0198, + "step": 16508 + }, + { + "epoch": 0.39, + "grad_norm": 2.014210629033092, + "learning_rate": 1.3970526498095897e-05, + "loss": 0.9535, + "step": 16509 + }, + { + "epoch": 0.39, + "grad_norm": 1.9277394677179929, + "learning_rate": 1.3969826170631247e-05, + "loss": 1.0624, + "step": 16510 + }, + { + "epoch": 0.39, + "grad_norm": 2.09974480924051, + "learning_rate": 1.3969125820053018e-05, + "loss": 1.0149, + "step": 16511 + }, + { + "epoch": 0.39, + "grad_norm": 1.9726897347790386, + "learning_rate": 1.3968425446365275e-05, + "loss": 1.0682, + "step": 16512 + }, + { + "epoch": 0.39, + "grad_norm": 2.9183594862247704, + "learning_rate": 1.3967725049572109e-05, + "loss": 0.8857, + "step": 16513 + }, + { + "epoch": 0.39, + "grad_norm": 2.0985144300651974, + "learning_rate": 1.3967024629677585e-05, + "loss": 0.9939, + "step": 16514 + }, + { + "epoch": 0.39, + "grad_norm": 1.1661704719782524, + "learning_rate": 1.3966324186685795e-05, + "loss": 0.9695, + "step": 16515 + }, + { + "epoch": 0.39, + "grad_norm": 1.9448849232552303, + "learning_rate": 1.3965623720600805e-05, + "loss": 1.0447, + "step": 16516 + }, + { + "epoch": 0.39, + "grad_norm": 2.0740868325319917, + "learning_rate": 1.3964923231426701e-05, + "loss": 1.0872, + "step": 16517 + }, + { + "epoch": 0.39, + "grad_norm": 1.883180661558087, + "learning_rate": 1.3964222719167557e-05, + "loss": 0.9938, + "step": 16518 + }, + { + "epoch": 0.39, + "grad_norm": 2.57931659824688, + "learning_rate": 1.3963522183827454e-05, + "loss": 0.8015, + "step": 16519 + }, + { + "epoch": 0.39, + "grad_norm": 2.117410255472803, + "learning_rate": 1.3962821625410468e-05, + "loss": 1.0193, + "step": 16520 + }, + { + "epoch": 0.39, + "grad_norm": 2.2373895736227674, + "learning_rate": 1.3962121043920682e-05, + "loss": 1.0215, + "step": 16521 + }, + { + "epoch": 0.39, + "grad_norm": 2.0914278261377737, + "learning_rate": 1.3961420439362173e-05, + "loss": 1.1573, + "step": 16522 + }, + { + "epoch": 0.39, + "grad_norm": 1.8322450255356417, + "learning_rate": 1.396071981173902e-05, + "loss": 0.9919, + "step": 16523 + }, + { + "epoch": 0.39, + "grad_norm": 1.8413910337741144, + "learning_rate": 1.3960019161055297e-05, + "loss": 1.014, + "step": 16524 + }, + { + "epoch": 0.39, + "grad_norm": 2.205593385018361, + "learning_rate": 1.3959318487315096e-05, + "loss": 0.9558, + "step": 16525 + }, + { + "epoch": 0.39, + "grad_norm": 1.2046893773167884, + "learning_rate": 1.3958617790522486e-05, + "loss": 1.0282, + "step": 16526 + }, + { + "epoch": 0.39, + "grad_norm": 2.171671460647535, + "learning_rate": 1.395791707068155e-05, + "loss": 0.9208, + "step": 16527 + }, + { + "epoch": 0.39, + "grad_norm": 1.9456518518274217, + "learning_rate": 1.3957216327796368e-05, + "loss": 1.0994, + "step": 16528 + }, + { + "epoch": 0.39, + "grad_norm": 2.184597974848495, + "learning_rate": 1.395651556187102e-05, + "loss": 1.0297, + "step": 16529 + }, + { + "epoch": 0.39, + "grad_norm": 1.997115158706075, + "learning_rate": 1.3955814772909588e-05, + "loss": 1.0344, + "step": 16530 + }, + { + "epoch": 0.39, + "grad_norm": 2.27331213373997, + "learning_rate": 1.3955113960916146e-05, + "loss": 1.0213, + "step": 16531 + }, + { + "epoch": 0.39, + "grad_norm": 2.151380581647413, + "learning_rate": 1.3954413125894783e-05, + "loss": 0.9508, + "step": 16532 + }, + { + "epoch": 0.39, + "grad_norm": 1.9300979485899845, + "learning_rate": 1.3953712267849568e-05, + "loss": 1.0286, + "step": 16533 + }, + { + "epoch": 0.39, + "grad_norm": 2.0136106557054796, + "learning_rate": 1.3953011386784597e-05, + "loss": 0.9627, + "step": 16534 + }, + { + "epoch": 0.39, + "grad_norm": 2.244748574146236, + "learning_rate": 1.3952310482703935e-05, + "loss": 1.0469, + "step": 16535 + }, + { + "epoch": 0.39, + "grad_norm": 2.0691056498060973, + "learning_rate": 1.3951609555611673e-05, + "loss": 0.9275, + "step": 16536 + }, + { + "epoch": 0.39, + "grad_norm": 2.1309089534684165, + "learning_rate": 1.3950908605511888e-05, + "loss": 0.9223, + "step": 16537 + }, + { + "epoch": 0.39, + "grad_norm": 1.9219699543953868, + "learning_rate": 1.3950207632408662e-05, + "loss": 0.9503, + "step": 16538 + }, + { + "epoch": 0.39, + "grad_norm": 1.9786170440872641, + "learning_rate": 1.394950663630608e-05, + "loss": 1.1318, + "step": 16539 + }, + { + "epoch": 0.39, + "grad_norm": 2.071472435140288, + "learning_rate": 1.3948805617208216e-05, + "loss": 0.9841, + "step": 16540 + }, + { + "epoch": 0.39, + "grad_norm": 1.93860090178859, + "learning_rate": 1.394810457511916e-05, + "loss": 1.1164, + "step": 16541 + }, + { + "epoch": 0.39, + "grad_norm": 3.7559160085942085, + "learning_rate": 1.3947403510042984e-05, + "loss": 1.0313, + "step": 16542 + }, + { + "epoch": 0.39, + "grad_norm": 2.0195708944240485, + "learning_rate": 1.3946702421983778e-05, + "loss": 1.001, + "step": 16543 + }, + { + "epoch": 0.39, + "grad_norm": 2.375598332520866, + "learning_rate": 1.394600131094562e-05, + "loss": 0.9744, + "step": 16544 + }, + { + "epoch": 0.39, + "grad_norm": 1.9911940618597825, + "learning_rate": 1.3945300176932594e-05, + "loss": 1.0375, + "step": 16545 + }, + { + "epoch": 0.39, + "grad_norm": 1.899489784692633, + "learning_rate": 1.3944599019948779e-05, + "loss": 1.0393, + "step": 16546 + }, + { + "epoch": 0.39, + "grad_norm": 1.9686640447962085, + "learning_rate": 1.3943897839998265e-05, + "loss": 1.1092, + "step": 16547 + }, + { + "epoch": 0.39, + "grad_norm": 1.9235074441034614, + "learning_rate": 1.3943196637085123e-05, + "loss": 0.9759, + "step": 16548 + }, + { + "epoch": 0.39, + "grad_norm": 1.9814017821240446, + "learning_rate": 1.3942495411213446e-05, + "loss": 1.083, + "step": 16549 + }, + { + "epoch": 0.39, + "grad_norm": 2.2451883677473377, + "learning_rate": 1.394179416238731e-05, + "loss": 0.9982, + "step": 16550 + }, + { + "epoch": 0.39, + "grad_norm": 2.108243208370374, + "learning_rate": 1.3941092890610803e-05, + "loss": 0.8901, + "step": 16551 + }, + { + "epoch": 0.39, + "grad_norm": 2.30774237247389, + "learning_rate": 1.3940391595888e-05, + "loss": 1.1388, + "step": 16552 + }, + { + "epoch": 0.39, + "grad_norm": 2.249412332077699, + "learning_rate": 1.3939690278222997e-05, + "loss": 1.011, + "step": 16553 + }, + { + "epoch": 0.39, + "grad_norm": 2.022897057006939, + "learning_rate": 1.3938988937619864e-05, + "loss": 1.1525, + "step": 16554 + }, + { + "epoch": 0.39, + "grad_norm": 1.9544186810193833, + "learning_rate": 1.3938287574082692e-05, + "loss": 1.1213, + "step": 16555 + }, + { + "epoch": 0.39, + "grad_norm": 1.0329947676171256, + "learning_rate": 1.3937586187615564e-05, + "loss": 0.8684, + "step": 16556 + }, + { + "epoch": 0.39, + "grad_norm": 2.309245932685589, + "learning_rate": 1.3936884778222562e-05, + "loss": 0.9851, + "step": 16557 + }, + { + "epoch": 0.39, + "grad_norm": 1.9093822262056748, + "learning_rate": 1.3936183345907772e-05, + "loss": 1.0017, + "step": 16558 + }, + { + "epoch": 0.39, + "grad_norm": 1.9851235433155392, + "learning_rate": 1.3935481890675273e-05, + "loss": 1.0757, + "step": 16559 + }, + { + "epoch": 0.39, + "grad_norm": 2.191156038153693, + "learning_rate": 1.3934780412529158e-05, + "loss": 0.9923, + "step": 16560 + }, + { + "epoch": 0.39, + "grad_norm": 2.0191448699117265, + "learning_rate": 1.3934078911473499e-05, + "loss": 1.1248, + "step": 16561 + }, + { + "epoch": 0.39, + "grad_norm": 2.1963769782540123, + "learning_rate": 1.3933377387512393e-05, + "loss": 1.0638, + "step": 16562 + }, + { + "epoch": 0.39, + "grad_norm": 2.0886920110485634, + "learning_rate": 1.3932675840649915e-05, + "loss": 0.9565, + "step": 16563 + }, + { + "epoch": 0.39, + "grad_norm": 2.128220784224084, + "learning_rate": 1.3931974270890154e-05, + "loss": 1.1441, + "step": 16564 + }, + { + "epoch": 0.39, + "grad_norm": 2.109909711281429, + "learning_rate": 1.393127267823719e-05, + "loss": 0.9829, + "step": 16565 + }, + { + "epoch": 0.39, + "grad_norm": 1.8372215919077821, + "learning_rate": 1.393057106269512e-05, + "loss": 1.0485, + "step": 16566 + }, + { + "epoch": 0.39, + "grad_norm": 1.9752260290778159, + "learning_rate": 1.3929869424268016e-05, + "loss": 1.0245, + "step": 16567 + }, + { + "epoch": 0.39, + "grad_norm": 2.0272351428343063, + "learning_rate": 1.392916776295997e-05, + "loss": 0.9671, + "step": 16568 + }, + { + "epoch": 0.39, + "grad_norm": 1.827392052425495, + "learning_rate": 1.3928466078775063e-05, + "loss": 1.0727, + "step": 16569 + }, + { + "epoch": 0.39, + "grad_norm": 1.9510204038139298, + "learning_rate": 1.3927764371717386e-05, + "loss": 1.1754, + "step": 16570 + }, + { + "epoch": 0.39, + "grad_norm": 2.1417869030186165, + "learning_rate": 1.3927062641791018e-05, + "loss": 1.1379, + "step": 16571 + }, + { + "epoch": 0.39, + "grad_norm": 1.8224625292620722, + "learning_rate": 1.3926360889000048e-05, + "loss": 0.9926, + "step": 16572 + }, + { + "epoch": 0.39, + "grad_norm": 2.0651597397839283, + "learning_rate": 1.3925659113348564e-05, + "loss": 1.0135, + "step": 16573 + }, + { + "epoch": 0.39, + "grad_norm": 1.9488698629439012, + "learning_rate": 1.3924957314840648e-05, + "loss": 0.9791, + "step": 16574 + }, + { + "epoch": 0.39, + "grad_norm": 1.892661356227617, + "learning_rate": 1.3924255493480387e-05, + "loss": 1.0504, + "step": 16575 + }, + { + "epoch": 0.39, + "grad_norm": 2.216203759477555, + "learning_rate": 1.392355364927187e-05, + "loss": 1.0505, + "step": 16576 + }, + { + "epoch": 0.39, + "grad_norm": 2.2757563603024242, + "learning_rate": 1.3922851782219182e-05, + "loss": 1.0738, + "step": 16577 + }, + { + "epoch": 0.39, + "grad_norm": 2.044782176581411, + "learning_rate": 1.3922149892326409e-05, + "loss": 0.9538, + "step": 16578 + }, + { + "epoch": 0.39, + "grad_norm": 3.3632669938679185, + "learning_rate": 1.3921447979597636e-05, + "loss": 0.9788, + "step": 16579 + }, + { + "epoch": 0.39, + "grad_norm": 1.875119600604194, + "learning_rate": 1.3920746044036953e-05, + "loss": 1.043, + "step": 16580 + }, + { + "epoch": 0.39, + "grad_norm": 1.120131305241436, + "learning_rate": 1.3920044085648446e-05, + "loss": 0.9378, + "step": 16581 + }, + { + "epoch": 0.39, + "grad_norm": 1.0615190577833409, + "learning_rate": 1.3919342104436198e-05, + "loss": 0.9545, + "step": 16582 + }, + { + "epoch": 0.39, + "grad_norm": 3.0370011607820415, + "learning_rate": 1.3918640100404301e-05, + "loss": 0.9041, + "step": 16583 + }, + { + "epoch": 0.39, + "grad_norm": 2.22252220380633, + "learning_rate": 1.3917938073556841e-05, + "loss": 1.0017, + "step": 16584 + }, + { + "epoch": 0.39, + "grad_norm": 2.0526377225737558, + "learning_rate": 1.3917236023897906e-05, + "loss": 0.9604, + "step": 16585 + }, + { + "epoch": 0.39, + "grad_norm": 1.919366739676508, + "learning_rate": 1.3916533951431579e-05, + "loss": 1.1122, + "step": 16586 + }, + { + "epoch": 0.39, + "grad_norm": 2.170910377808517, + "learning_rate": 1.3915831856161956e-05, + "loss": 1.0394, + "step": 16587 + }, + { + "epoch": 0.39, + "grad_norm": 2.770545059585025, + "learning_rate": 1.3915129738093119e-05, + "loss": 0.9228, + "step": 16588 + }, + { + "epoch": 0.39, + "grad_norm": 2.444983826279168, + "learning_rate": 1.3914427597229156e-05, + "loss": 1.0982, + "step": 16589 + }, + { + "epoch": 0.39, + "grad_norm": 2.1232574730432465, + "learning_rate": 1.3913725433574157e-05, + "loss": 1.174, + "step": 16590 + }, + { + "epoch": 0.39, + "grad_norm": 2.0356135112067038, + "learning_rate": 1.391302324713221e-05, + "loss": 1.0028, + "step": 16591 + }, + { + "epoch": 0.39, + "grad_norm": 2.454651579015072, + "learning_rate": 1.39123210379074e-05, + "loss": 1.0858, + "step": 16592 + }, + { + "epoch": 0.39, + "grad_norm": 1.9278324524009574, + "learning_rate": 1.391161880590382e-05, + "loss": 1.0543, + "step": 16593 + }, + { + "epoch": 0.39, + "grad_norm": 2.002657590292186, + "learning_rate": 1.3910916551125558e-05, + "loss": 1.2602, + "step": 16594 + }, + { + "epoch": 0.39, + "grad_norm": 2.4931125656096014, + "learning_rate": 1.3910214273576701e-05, + "loss": 1.136, + "step": 16595 + }, + { + "epoch": 0.39, + "grad_norm": 1.0865078565048485, + "learning_rate": 1.3909511973261339e-05, + "loss": 0.9985, + "step": 16596 + }, + { + "epoch": 0.39, + "grad_norm": 2.0785783882832676, + "learning_rate": 1.3908809650183559e-05, + "loss": 0.9965, + "step": 16597 + }, + { + "epoch": 0.39, + "grad_norm": 2.1663479432360737, + "learning_rate": 1.3908107304347453e-05, + "loss": 0.9252, + "step": 16598 + }, + { + "epoch": 0.39, + "grad_norm": 2.2667420763642525, + "learning_rate": 1.3907404935757107e-05, + "loss": 1.0381, + "step": 16599 + }, + { + "epoch": 0.39, + "grad_norm": 1.991512148524445, + "learning_rate": 1.3906702544416614e-05, + "loss": 1.0397, + "step": 16600 + }, + { + "epoch": 0.39, + "grad_norm": 2.7620302617979053, + "learning_rate": 1.3906000130330061e-05, + "loss": 0.966, + "step": 16601 + }, + { + "epoch": 0.39, + "grad_norm": 2.2959698023427952, + "learning_rate": 1.3905297693501543e-05, + "loss": 1.0516, + "step": 16602 + }, + { + "epoch": 0.39, + "grad_norm": 2.0743989356019816, + "learning_rate": 1.3904595233935139e-05, + "loss": 1.0537, + "step": 16603 + }, + { + "epoch": 0.39, + "grad_norm": 1.920483920983984, + "learning_rate": 1.3903892751634949e-05, + "loss": 0.9764, + "step": 16604 + }, + { + "epoch": 0.39, + "grad_norm": 2.090519016678031, + "learning_rate": 1.3903190246605059e-05, + "loss": 1.0231, + "step": 16605 + }, + { + "epoch": 0.39, + "grad_norm": 2.0959024453411437, + "learning_rate": 1.390248771884956e-05, + "loss": 1.1153, + "step": 16606 + }, + { + "epoch": 0.39, + "grad_norm": 3.262668319280493, + "learning_rate": 1.3901785168372542e-05, + "loss": 0.9903, + "step": 16607 + }, + { + "epoch": 0.39, + "grad_norm": 2.496449362077747, + "learning_rate": 1.3901082595178092e-05, + "loss": 1.1554, + "step": 16608 + }, + { + "epoch": 0.39, + "grad_norm": 1.905856937013319, + "learning_rate": 1.390037999927031e-05, + "loss": 1.0971, + "step": 16609 + }, + { + "epoch": 0.39, + "grad_norm": 1.8859920375700518, + "learning_rate": 1.3899677380653276e-05, + "loss": 1.042, + "step": 16610 + }, + { + "epoch": 0.39, + "grad_norm": 2.3202926047071037, + "learning_rate": 1.3898974739331091e-05, + "loss": 1.0344, + "step": 16611 + }, + { + "epoch": 0.39, + "grad_norm": 1.9261633392865092, + "learning_rate": 1.3898272075307833e-05, + "loss": 1.1536, + "step": 16612 + }, + { + "epoch": 0.39, + "grad_norm": 2.4423527101814586, + "learning_rate": 1.3897569388587609e-05, + "loss": 1.0614, + "step": 16613 + }, + { + "epoch": 0.39, + "grad_norm": 2.000275737923082, + "learning_rate": 1.3896866679174495e-05, + "loss": 0.9733, + "step": 16614 + }, + { + "epoch": 0.39, + "grad_norm": 1.7917128646666933, + "learning_rate": 1.3896163947072595e-05, + "loss": 1.0383, + "step": 16615 + }, + { + "epoch": 0.39, + "grad_norm": 2.40724959110983, + "learning_rate": 1.3895461192285992e-05, + "loss": 1.1363, + "step": 16616 + }, + { + "epoch": 0.39, + "grad_norm": 2.0452570127097864, + "learning_rate": 1.3894758414818781e-05, + "loss": 1.0207, + "step": 16617 + }, + { + "epoch": 0.39, + "grad_norm": 1.9248780563650751, + "learning_rate": 1.3894055614675052e-05, + "loss": 1.1125, + "step": 16618 + }, + { + "epoch": 0.39, + "grad_norm": 2.4544873827566764, + "learning_rate": 1.38933527918589e-05, + "loss": 1.1475, + "step": 16619 + }, + { + "epoch": 0.39, + "grad_norm": 2.1305205649881134, + "learning_rate": 1.3892649946374417e-05, + "loss": 1.0951, + "step": 16620 + }, + { + "epoch": 0.39, + "grad_norm": 2.3171664732086263, + "learning_rate": 1.3891947078225692e-05, + "loss": 1.1148, + "step": 16621 + }, + { + "epoch": 0.39, + "grad_norm": 1.8658521469263154, + "learning_rate": 1.389124418741682e-05, + "loss": 1.0967, + "step": 16622 + }, + { + "epoch": 0.39, + "grad_norm": 2.218326737897137, + "learning_rate": 1.389054127395189e-05, + "loss": 1.055, + "step": 16623 + }, + { + "epoch": 0.39, + "grad_norm": 1.9665585658095297, + "learning_rate": 1.3889838337834997e-05, + "loss": 1.1316, + "step": 16624 + }, + { + "epoch": 0.39, + "grad_norm": 2.7365708390694254, + "learning_rate": 1.3889135379070235e-05, + "loss": 0.9265, + "step": 16625 + }, + { + "epoch": 0.39, + "grad_norm": 1.7727590638590385, + "learning_rate": 1.3888432397661697e-05, + "loss": 0.9615, + "step": 16626 + }, + { + "epoch": 0.39, + "grad_norm": 2.278384208754046, + "learning_rate": 1.3887729393613471e-05, + "loss": 1.0566, + "step": 16627 + }, + { + "epoch": 0.39, + "grad_norm": 1.9536181639086159, + "learning_rate": 1.3887026366929657e-05, + "loss": 1.0199, + "step": 16628 + }, + { + "epoch": 0.39, + "grad_norm": 2.0483234393493897, + "learning_rate": 1.3886323317614344e-05, + "loss": 0.8921, + "step": 16629 + }, + { + "epoch": 0.39, + "grad_norm": 1.8336984164665757, + "learning_rate": 1.3885620245671624e-05, + "loss": 1.0564, + "step": 16630 + }, + { + "epoch": 0.39, + "grad_norm": 1.0885624933950238, + "learning_rate": 1.3884917151105595e-05, + "loss": 0.9738, + "step": 16631 + }, + { + "epoch": 0.39, + "grad_norm": 2.1161337763815613, + "learning_rate": 1.388421403392035e-05, + "loss": 0.9908, + "step": 16632 + }, + { + "epoch": 0.39, + "grad_norm": 1.9183129695611278, + "learning_rate": 1.3883510894119979e-05, + "loss": 0.8903, + "step": 16633 + }, + { + "epoch": 0.39, + "grad_norm": 1.8747648796418936, + "learning_rate": 1.3882807731708579e-05, + "loss": 1.1386, + "step": 16634 + }, + { + "epoch": 0.39, + "grad_norm": 1.0633885354399477, + "learning_rate": 1.3882104546690241e-05, + "loss": 0.9804, + "step": 16635 + }, + { + "epoch": 0.39, + "grad_norm": 2.283281030765679, + "learning_rate": 1.3881401339069064e-05, + "loss": 1.0557, + "step": 16636 + }, + { + "epoch": 0.39, + "grad_norm": 2.0263208026632715, + "learning_rate": 1.3880698108849138e-05, + "loss": 1.0593, + "step": 16637 + }, + { + "epoch": 0.39, + "grad_norm": 1.1442421311592645, + "learning_rate": 1.387999485603456e-05, + "loss": 0.9018, + "step": 16638 + }, + { + "epoch": 0.39, + "grad_norm": 2.119484176672805, + "learning_rate": 1.3879291580629421e-05, + "loss": 1.0699, + "step": 16639 + }, + { + "epoch": 0.39, + "grad_norm": 2.0646549568744548, + "learning_rate": 1.3878588282637822e-05, + "loss": 1.1446, + "step": 16640 + }, + { + "epoch": 0.39, + "grad_norm": 2.486438955142733, + "learning_rate": 1.3877884962063853e-05, + "loss": 0.9415, + "step": 16641 + }, + { + "epoch": 0.39, + "grad_norm": 1.8876291580172713, + "learning_rate": 1.3877181618911606e-05, + "loss": 1.028, + "step": 16642 + }, + { + "epoch": 0.39, + "grad_norm": 2.0266430140252703, + "learning_rate": 1.3876478253185183e-05, + "loss": 1.0384, + "step": 16643 + }, + { + "epoch": 0.39, + "grad_norm": 1.0581649588394129, + "learning_rate": 1.3875774864888675e-05, + "loss": 0.9797, + "step": 16644 + }, + { + "epoch": 0.39, + "grad_norm": 1.782467730678645, + "learning_rate": 1.3875071454026183e-05, + "loss": 0.9635, + "step": 16645 + }, + { + "epoch": 0.39, + "grad_norm": 2.51752694744167, + "learning_rate": 1.3874368020601796e-05, + "loss": 1.054, + "step": 16646 + }, + { + "epoch": 0.39, + "grad_norm": 1.975232926724146, + "learning_rate": 1.3873664564619612e-05, + "loss": 0.996, + "step": 16647 + }, + { + "epoch": 0.39, + "grad_norm": 1.8489929272657015, + "learning_rate": 1.3872961086083725e-05, + "loss": 0.9486, + "step": 16648 + }, + { + "epoch": 0.39, + "grad_norm": 1.7443750494652341, + "learning_rate": 1.3872257584998235e-05, + "loss": 1.0407, + "step": 16649 + }, + { + "epoch": 0.39, + "grad_norm": 2.095787919529703, + "learning_rate": 1.3871554061367232e-05, + "loss": 1.1133, + "step": 16650 + }, + { + "epoch": 0.39, + "grad_norm": 2.165839805123141, + "learning_rate": 1.3870850515194817e-05, + "loss": 1.127, + "step": 16651 + }, + { + "epoch": 0.39, + "grad_norm": 1.9178606979992172, + "learning_rate": 1.3870146946485083e-05, + "loss": 1.1244, + "step": 16652 + }, + { + "epoch": 0.39, + "grad_norm": 1.9584381782211342, + "learning_rate": 1.386944335524213e-05, + "loss": 0.9495, + "step": 16653 + }, + { + "epoch": 0.39, + "grad_norm": 2.03690375454301, + "learning_rate": 1.3868739741470055e-05, + "loss": 0.9779, + "step": 16654 + }, + { + "epoch": 0.39, + "grad_norm": 1.1026579629731519, + "learning_rate": 1.386803610517295e-05, + "loss": 1.0027, + "step": 16655 + }, + { + "epoch": 0.39, + "grad_norm": 1.9355889549406222, + "learning_rate": 1.3867332446354916e-05, + "loss": 0.9961, + "step": 16656 + }, + { + "epoch": 0.39, + "grad_norm": 1.84315818925346, + "learning_rate": 1.3866628765020045e-05, + "loss": 1.0681, + "step": 16657 + }, + { + "epoch": 0.39, + "grad_norm": 2.298616619455803, + "learning_rate": 1.3865925061172439e-05, + "loss": 1.0493, + "step": 16658 + }, + { + "epoch": 0.39, + "grad_norm": 2.356960830211041, + "learning_rate": 1.3865221334816193e-05, + "loss": 0.9056, + "step": 16659 + }, + { + "epoch": 0.39, + "grad_norm": 2.0748581904170873, + "learning_rate": 1.3864517585955406e-05, + "loss": 1.0303, + "step": 16660 + }, + { + "epoch": 0.39, + "grad_norm": 1.0427064753379933, + "learning_rate": 1.3863813814594172e-05, + "loss": 0.9386, + "step": 16661 + }, + { + "epoch": 0.39, + "grad_norm": 2.0346325754987595, + "learning_rate": 1.3863110020736592e-05, + "loss": 1.0157, + "step": 16662 + }, + { + "epoch": 0.39, + "grad_norm": 2.2163871464210247, + "learning_rate": 1.3862406204386761e-05, + "loss": 0.933, + "step": 16663 + }, + { + "epoch": 0.39, + "grad_norm": 2.5020628042621933, + "learning_rate": 1.3861702365548781e-05, + "loss": 1.0997, + "step": 16664 + }, + { + "epoch": 0.39, + "grad_norm": 2.0547384673449414, + "learning_rate": 1.3860998504226744e-05, + "loss": 1.0407, + "step": 16665 + }, + { + "epoch": 0.39, + "grad_norm": 2.2158333462558364, + "learning_rate": 1.3860294620424754e-05, + "loss": 1.0884, + "step": 16666 + }, + { + "epoch": 0.39, + "grad_norm": 1.9803659654125119, + "learning_rate": 1.3859590714146906e-05, + "loss": 1.055, + "step": 16667 + }, + { + "epoch": 0.39, + "grad_norm": 1.8921127215202116, + "learning_rate": 1.38588867853973e-05, + "loss": 0.9018, + "step": 16668 + }, + { + "epoch": 0.39, + "grad_norm": 1.8945259046847844, + "learning_rate": 1.385818283418003e-05, + "loss": 0.9356, + "step": 16669 + }, + { + "epoch": 0.39, + "grad_norm": 1.9928046431140924, + "learning_rate": 1.3857478860499201e-05, + "loss": 0.9933, + "step": 16670 + }, + { + "epoch": 0.39, + "grad_norm": 2.4875029254912153, + "learning_rate": 1.3856774864358906e-05, + "loss": 0.9485, + "step": 16671 + }, + { + "epoch": 0.39, + "grad_norm": 2.1602369566304596, + "learning_rate": 1.385607084576325e-05, + "loss": 0.967, + "step": 16672 + }, + { + "epoch": 0.39, + "grad_norm": 2.5155677936528984, + "learning_rate": 1.3855366804716326e-05, + "loss": 0.9089, + "step": 16673 + }, + { + "epoch": 0.39, + "grad_norm": 1.125736124466528, + "learning_rate": 1.3854662741222238e-05, + "loss": 0.9303, + "step": 16674 + }, + { + "epoch": 0.39, + "grad_norm": 1.9991720571907874, + "learning_rate": 1.385395865528508e-05, + "loss": 1.0398, + "step": 16675 + }, + { + "epoch": 0.39, + "grad_norm": 1.9417281082501954, + "learning_rate": 1.3853254546908958e-05, + "loss": 1.0342, + "step": 16676 + }, + { + "epoch": 0.39, + "grad_norm": 2.065681398556165, + "learning_rate": 1.3852550416097968e-05, + "loss": 1.0077, + "step": 16677 + }, + { + "epoch": 0.39, + "grad_norm": 1.8039139329784664, + "learning_rate": 1.3851846262856208e-05, + "loss": 1.089, + "step": 16678 + }, + { + "epoch": 0.39, + "grad_norm": 2.212961122756828, + "learning_rate": 1.3851142087187778e-05, + "loss": 1.0332, + "step": 16679 + }, + { + "epoch": 0.39, + "grad_norm": 1.9423436424137812, + "learning_rate": 1.3850437889096781e-05, + "loss": 0.9394, + "step": 16680 + }, + { + "epoch": 0.39, + "grad_norm": 2.245121202656099, + "learning_rate": 1.384973366858732e-05, + "loss": 0.9828, + "step": 16681 + }, + { + "epoch": 0.39, + "grad_norm": 2.226797194199987, + "learning_rate": 1.3849029425663485e-05, + "loss": 1.0194, + "step": 16682 + }, + { + "epoch": 0.39, + "grad_norm": 1.9637469256524283, + "learning_rate": 1.3848325160329386e-05, + "loss": 0.9354, + "step": 16683 + }, + { + "epoch": 0.39, + "grad_norm": 1.8956128168845552, + "learning_rate": 1.3847620872589116e-05, + "loss": 0.9662, + "step": 16684 + }, + { + "epoch": 0.39, + "grad_norm": 2.0426330096785796, + "learning_rate": 1.3846916562446784e-05, + "loss": 0.9917, + "step": 16685 + }, + { + "epoch": 0.39, + "grad_norm": 2.181542414185262, + "learning_rate": 1.384621222990648e-05, + "loss": 0.8657, + "step": 16686 + }, + { + "epoch": 0.39, + "grad_norm": 1.0673030573058957, + "learning_rate": 1.3845507874972314e-05, + "loss": 0.9116, + "step": 16687 + }, + { + "epoch": 0.39, + "grad_norm": 2.3568268529107232, + "learning_rate": 1.3844803497648382e-05, + "loss": 1.0041, + "step": 16688 + }, + { + "epoch": 0.39, + "grad_norm": 2.091212963613817, + "learning_rate": 1.384409909793879e-05, + "loss": 1.0146, + "step": 16689 + }, + { + "epoch": 0.39, + "grad_norm": 1.8563214994592816, + "learning_rate": 1.3843394675847635e-05, + "loss": 1.0504, + "step": 16690 + }, + { + "epoch": 0.39, + "grad_norm": 1.123294478881412, + "learning_rate": 1.3842690231379014e-05, + "loss": 0.973, + "step": 16691 + }, + { + "epoch": 0.39, + "grad_norm": 1.037252077335408, + "learning_rate": 1.384198576453704e-05, + "loss": 1.022, + "step": 16692 + }, + { + "epoch": 0.39, + "grad_norm": 2.598172639219087, + "learning_rate": 1.3841281275325807e-05, + "loss": 0.8811, + "step": 16693 + }, + { + "epoch": 0.39, + "grad_norm": 2.0031592562073985, + "learning_rate": 1.3840576763749419e-05, + "loss": 1.1049, + "step": 16694 + }, + { + "epoch": 0.39, + "grad_norm": 2.146167572621815, + "learning_rate": 1.3839872229811974e-05, + "loss": 1.0618, + "step": 16695 + }, + { + "epoch": 0.39, + "grad_norm": 1.095565547780243, + "learning_rate": 1.3839167673517582e-05, + "loss": 0.9595, + "step": 16696 + }, + { + "epoch": 0.39, + "grad_norm": 1.7018416910609238, + "learning_rate": 1.3838463094870338e-05, + "loss": 1.0764, + "step": 16697 + }, + { + "epoch": 0.39, + "grad_norm": 2.104118564535339, + "learning_rate": 1.3837758493874347e-05, + "loss": 1.1089, + "step": 16698 + }, + { + "epoch": 0.39, + "grad_norm": 2.0159666650648105, + "learning_rate": 1.383705387053371e-05, + "loss": 0.9051, + "step": 16699 + }, + { + "epoch": 0.39, + "grad_norm": 1.896741436892082, + "learning_rate": 1.383634922485253e-05, + "loss": 1.0443, + "step": 16700 + }, + { + "epoch": 0.39, + "grad_norm": 2.8750295499440783, + "learning_rate": 1.3835644556834913e-05, + "loss": 0.9873, + "step": 16701 + }, + { + "epoch": 0.39, + "grad_norm": 2.008613530655649, + "learning_rate": 1.3834939866484956e-05, + "loss": 1.1036, + "step": 16702 + }, + { + "epoch": 0.39, + "grad_norm": 2.0275178598646315, + "learning_rate": 1.3834235153806766e-05, + "loss": 0.9414, + "step": 16703 + }, + { + "epoch": 0.39, + "grad_norm": 2.0546346968585145, + "learning_rate": 1.3833530418804446e-05, + "loss": 0.9363, + "step": 16704 + }, + { + "epoch": 0.39, + "grad_norm": 4.327372606309394, + "learning_rate": 1.3832825661482097e-05, + "loss": 0.9933, + "step": 16705 + }, + { + "epoch": 0.39, + "grad_norm": 2.328008518209286, + "learning_rate": 1.3832120881843825e-05, + "loss": 1.0084, + "step": 16706 + }, + { + "epoch": 0.39, + "grad_norm": 2.4234014893730267, + "learning_rate": 1.383141607989373e-05, + "loss": 1.1238, + "step": 16707 + }, + { + "epoch": 0.39, + "grad_norm": 2.0066882131902086, + "learning_rate": 1.3830711255635919e-05, + "loss": 1.1104, + "step": 16708 + }, + { + "epoch": 0.39, + "grad_norm": 2.127007750614506, + "learning_rate": 1.3830006409074495e-05, + "loss": 1.0412, + "step": 16709 + }, + { + "epoch": 0.39, + "grad_norm": 2.114805858184178, + "learning_rate": 1.3829301540213557e-05, + "loss": 0.9949, + "step": 16710 + }, + { + "epoch": 0.39, + "grad_norm": 2.098251139209723, + "learning_rate": 1.382859664905722e-05, + "loss": 0.9384, + "step": 16711 + }, + { + "epoch": 0.39, + "grad_norm": 1.8727790587103184, + "learning_rate": 1.3827891735609575e-05, + "loss": 0.9013, + "step": 16712 + }, + { + "epoch": 0.39, + "grad_norm": 1.0639432333317966, + "learning_rate": 1.3827186799874737e-05, + "loss": 0.9005, + "step": 16713 + }, + { + "epoch": 0.39, + "grad_norm": 2.0621022661396937, + "learning_rate": 1.3826481841856802e-05, + "loss": 1.1115, + "step": 16714 + }, + { + "epoch": 0.39, + "grad_norm": 1.8220157877498075, + "learning_rate": 1.382577686155988e-05, + "loss": 1.0804, + "step": 16715 + }, + { + "epoch": 0.39, + "grad_norm": 2.594986345167011, + "learning_rate": 1.3825071858988073e-05, + "loss": 0.9399, + "step": 16716 + }, + { + "epoch": 0.39, + "grad_norm": 2.122889503089857, + "learning_rate": 1.3824366834145488e-05, + "loss": 1.0687, + "step": 16717 + }, + { + "epoch": 0.39, + "grad_norm": 1.8420860892987987, + "learning_rate": 1.382366178703623e-05, + "loss": 1.0193, + "step": 16718 + }, + { + "epoch": 0.39, + "grad_norm": 2.283926820559935, + "learning_rate": 1.38229567176644e-05, + "loss": 0.952, + "step": 16719 + }, + { + "epoch": 0.39, + "grad_norm": 2.2522860459401257, + "learning_rate": 1.3822251626034105e-05, + "loss": 1.0088, + "step": 16720 + }, + { + "epoch": 0.39, + "grad_norm": 2.0356167605860382, + "learning_rate": 1.3821546512149453e-05, + "loss": 1.0969, + "step": 16721 + }, + { + "epoch": 0.39, + "grad_norm": 2.1524355796515993, + "learning_rate": 1.3820841376014546e-05, + "loss": 1.0895, + "step": 16722 + }, + { + "epoch": 0.39, + "grad_norm": 1.8368784143303625, + "learning_rate": 1.3820136217633491e-05, + "loss": 0.9636, + "step": 16723 + }, + { + "epoch": 0.39, + "grad_norm": 2.1221450014834558, + "learning_rate": 1.3819431037010393e-05, + "loss": 0.9967, + "step": 16724 + }, + { + "epoch": 0.39, + "grad_norm": 2.2306747710903894, + "learning_rate": 1.381872583414936e-05, + "loss": 1.0005, + "step": 16725 + }, + { + "epoch": 0.39, + "grad_norm": 1.913621013192823, + "learning_rate": 1.3818020609054494e-05, + "loss": 0.9019, + "step": 16726 + }, + { + "epoch": 0.39, + "grad_norm": 2.7294070238226524, + "learning_rate": 1.3817315361729904e-05, + "loss": 0.9158, + "step": 16727 + }, + { + "epoch": 0.39, + "grad_norm": 2.3431397827816145, + "learning_rate": 1.3816610092179697e-05, + "loss": 1.0076, + "step": 16728 + }, + { + "epoch": 0.39, + "grad_norm": 2.9376428110011483, + "learning_rate": 1.3815904800407975e-05, + "loss": 1.2389, + "step": 16729 + }, + { + "epoch": 0.39, + "grad_norm": 1.8016696080660795, + "learning_rate": 1.3815199486418851e-05, + "loss": 1.018, + "step": 16730 + }, + { + "epoch": 0.39, + "grad_norm": 1.9751591886217579, + "learning_rate": 1.3814494150216424e-05, + "loss": 1.1675, + "step": 16731 + }, + { + "epoch": 0.39, + "grad_norm": 2.51931045719, + "learning_rate": 1.3813788791804807e-05, + "loss": 0.9011, + "step": 16732 + }, + { + "epoch": 0.39, + "grad_norm": 1.9765951498236534, + "learning_rate": 1.3813083411188104e-05, + "loss": 1.2106, + "step": 16733 + }, + { + "epoch": 0.39, + "grad_norm": 2.1678849480086306, + "learning_rate": 1.3812378008370422e-05, + "loss": 0.9422, + "step": 16734 + }, + { + "epoch": 0.39, + "grad_norm": 2.0061231728025115, + "learning_rate": 1.3811672583355864e-05, + "loss": 1.0337, + "step": 16735 + }, + { + "epoch": 0.39, + "grad_norm": 1.9771674810210487, + "learning_rate": 1.3810967136148547e-05, + "loss": 0.9991, + "step": 16736 + }, + { + "epoch": 0.39, + "grad_norm": 2.3297895339136074, + "learning_rate": 1.3810261666752571e-05, + "loss": 1.1484, + "step": 16737 + }, + { + "epoch": 0.39, + "grad_norm": 2.117705502938087, + "learning_rate": 1.3809556175172043e-05, + "loss": 1.0891, + "step": 16738 + }, + { + "epoch": 0.39, + "grad_norm": 2.016110201265325, + "learning_rate": 1.3808850661411077e-05, + "loss": 1.0544, + "step": 16739 + }, + { + "epoch": 0.39, + "grad_norm": 1.8886115316990044, + "learning_rate": 1.3808145125473773e-05, + "loss": 1.0529, + "step": 16740 + }, + { + "epoch": 0.39, + "grad_norm": 2.2132156619243353, + "learning_rate": 1.3807439567364246e-05, + "loss": 1.128, + "step": 16741 + }, + { + "epoch": 0.39, + "grad_norm": 2.206782377187114, + "learning_rate": 1.3806733987086597e-05, + "loss": 1.1126, + "step": 16742 + }, + { + "epoch": 0.39, + "grad_norm": 1.930330574743975, + "learning_rate": 1.380602838464494e-05, + "loss": 0.9757, + "step": 16743 + }, + { + "epoch": 0.39, + "grad_norm": 1.9511311322045193, + "learning_rate": 1.3805322760043378e-05, + "loss": 1.0888, + "step": 16744 + }, + { + "epoch": 0.39, + "grad_norm": 2.2921716612074907, + "learning_rate": 1.3804617113286027e-05, + "loss": 1.0663, + "step": 16745 + }, + { + "epoch": 0.39, + "grad_norm": 1.179482688161236, + "learning_rate": 1.3803911444376984e-05, + "loss": 0.9473, + "step": 16746 + }, + { + "epoch": 0.39, + "grad_norm": 2.0449485733799704, + "learning_rate": 1.380320575332037e-05, + "loss": 0.9862, + "step": 16747 + }, + { + "epoch": 0.39, + "grad_norm": 1.9998675841309717, + "learning_rate": 1.3802500040120285e-05, + "loss": 1.0086, + "step": 16748 + }, + { + "epoch": 0.39, + "grad_norm": 2.24363331511671, + "learning_rate": 1.3801794304780843e-05, + "loss": 0.9413, + "step": 16749 + }, + { + "epoch": 0.39, + "grad_norm": 1.971273933140176, + "learning_rate": 1.3801088547306149e-05, + "loss": 1.0735, + "step": 16750 + }, + { + "epoch": 0.39, + "grad_norm": 1.8971914478803307, + "learning_rate": 1.3800382767700316e-05, + "loss": 1.0686, + "step": 16751 + }, + { + "epoch": 0.39, + "grad_norm": 1.9817380516759109, + "learning_rate": 1.3799676965967451e-05, + "loss": 1.1231, + "step": 16752 + }, + { + "epoch": 0.39, + "grad_norm": 1.7772600942223005, + "learning_rate": 1.3798971142111661e-05, + "loss": 1.0321, + "step": 16753 + }, + { + "epoch": 0.39, + "grad_norm": 1.8311003808995943, + "learning_rate": 1.379826529613706e-05, + "loss": 0.947, + "step": 16754 + }, + { + "epoch": 0.39, + "grad_norm": 2.102559054018754, + "learning_rate": 1.3797559428047755e-05, + "loss": 1.0545, + "step": 16755 + }, + { + "epoch": 0.39, + "grad_norm": 1.9040199974098362, + "learning_rate": 1.3796853537847859e-05, + "loss": 1.0371, + "step": 16756 + }, + { + "epoch": 0.39, + "grad_norm": 1.9638417277637263, + "learning_rate": 1.3796147625541479e-05, + "loss": 0.9769, + "step": 16757 + }, + { + "epoch": 0.39, + "grad_norm": 1.9474169979354996, + "learning_rate": 1.3795441691132726e-05, + "loss": 1.0363, + "step": 16758 + }, + { + "epoch": 0.39, + "grad_norm": 1.8985215240102113, + "learning_rate": 1.3794735734625708e-05, + "loss": 0.9195, + "step": 16759 + }, + { + "epoch": 0.39, + "grad_norm": 1.847688490475938, + "learning_rate": 1.3794029756024536e-05, + "loss": 1.1297, + "step": 16760 + }, + { + "epoch": 0.39, + "grad_norm": 1.8769067945006974, + "learning_rate": 1.3793323755333323e-05, + "loss": 1.0104, + "step": 16761 + }, + { + "epoch": 0.39, + "grad_norm": 1.9330539513429792, + "learning_rate": 1.3792617732556182e-05, + "loss": 1.0211, + "step": 16762 + }, + { + "epoch": 0.39, + "grad_norm": 2.2799678311802625, + "learning_rate": 1.3791911687697215e-05, + "loss": 1.0019, + "step": 16763 + }, + { + "epoch": 0.39, + "grad_norm": 1.073509025637424, + "learning_rate": 1.3791205620760541e-05, + "loss": 1.01, + "step": 16764 + }, + { + "epoch": 0.39, + "grad_norm": 1.9876454625865163, + "learning_rate": 1.3790499531750262e-05, + "loss": 1.1012, + "step": 16765 + }, + { + "epoch": 0.39, + "grad_norm": 2.040727980663763, + "learning_rate": 1.3789793420670503e-05, + "loss": 0.906, + "step": 16766 + }, + { + "epoch": 0.4, + "grad_norm": 2.3632077396798232, + "learning_rate": 1.378908728752536e-05, + "loss": 1.0542, + "step": 16767 + }, + { + "epoch": 0.4, + "grad_norm": 2.465633276559412, + "learning_rate": 1.3788381132318955e-05, + "loss": 1.0626, + "step": 16768 + }, + { + "epoch": 0.4, + "grad_norm": 2.5499614022198025, + "learning_rate": 1.3787674955055395e-05, + "loss": 0.9755, + "step": 16769 + }, + { + "epoch": 0.4, + "grad_norm": 1.7182037412705593, + "learning_rate": 1.3786968755738792e-05, + "loss": 0.9644, + "step": 16770 + }, + { + "epoch": 0.4, + "grad_norm": 2.0747599556839313, + "learning_rate": 1.3786262534373256e-05, + "loss": 0.9952, + "step": 16771 + }, + { + "epoch": 0.4, + "grad_norm": 2.1529967741855707, + "learning_rate": 1.3785556290962902e-05, + "loss": 0.9748, + "step": 16772 + }, + { + "epoch": 0.4, + "grad_norm": 2.062129548699101, + "learning_rate": 1.3784850025511842e-05, + "loss": 1.0549, + "step": 16773 + }, + { + "epoch": 0.4, + "grad_norm": 2.0132053275419004, + "learning_rate": 1.3784143738024188e-05, + "loss": 0.9983, + "step": 16774 + }, + { + "epoch": 0.4, + "grad_norm": 1.8330783642425392, + "learning_rate": 1.3783437428504048e-05, + "loss": 1.0372, + "step": 16775 + }, + { + "epoch": 0.4, + "grad_norm": 2.036968383134246, + "learning_rate": 1.3782731096955537e-05, + "loss": 1.0347, + "step": 16776 + }, + { + "epoch": 0.4, + "grad_norm": 3.901464367612236, + "learning_rate": 1.378202474338277e-05, + "loss": 1.1498, + "step": 16777 + }, + { + "epoch": 0.4, + "grad_norm": 2.4025055323958147, + "learning_rate": 1.3781318367789858e-05, + "loss": 1.0248, + "step": 16778 + }, + { + "epoch": 0.4, + "grad_norm": 2.471155518175292, + "learning_rate": 1.3780611970180913e-05, + "loss": 1.0269, + "step": 16779 + }, + { + "epoch": 0.4, + "grad_norm": 2.418557565360524, + "learning_rate": 1.3779905550560047e-05, + "loss": 0.9407, + "step": 16780 + }, + { + "epoch": 0.4, + "grad_norm": 2.4314980146229104, + "learning_rate": 1.3779199108931374e-05, + "loss": 0.9381, + "step": 16781 + }, + { + "epoch": 0.4, + "grad_norm": 2.2484089419886293, + "learning_rate": 1.3778492645299008e-05, + "loss": 0.8727, + "step": 16782 + }, + { + "epoch": 0.4, + "grad_norm": 2.0797504568807867, + "learning_rate": 1.3777786159667063e-05, + "loss": 1.1608, + "step": 16783 + }, + { + "epoch": 0.4, + "grad_norm": 1.919402960569471, + "learning_rate": 1.3777079652039649e-05, + "loss": 0.9853, + "step": 16784 + }, + { + "epoch": 0.4, + "grad_norm": 1.9274975310870603, + "learning_rate": 1.3776373122420884e-05, + "loss": 0.9794, + "step": 16785 + }, + { + "epoch": 0.4, + "grad_norm": 1.8348938018379681, + "learning_rate": 1.3775666570814876e-05, + "loss": 0.9173, + "step": 16786 + }, + { + "epoch": 0.4, + "grad_norm": 1.0981032077304382, + "learning_rate": 1.3774959997225743e-05, + "loss": 0.9347, + "step": 16787 + }, + { + "epoch": 0.4, + "grad_norm": 1.7989149800729034, + "learning_rate": 1.3774253401657599e-05, + "loss": 1.0323, + "step": 16788 + }, + { + "epoch": 0.4, + "grad_norm": 2.2947682109581407, + "learning_rate": 1.3773546784114556e-05, + "loss": 1.0204, + "step": 16789 + }, + { + "epoch": 0.4, + "grad_norm": 1.0764898151363993, + "learning_rate": 1.377284014460073e-05, + "loss": 0.8906, + "step": 16790 + }, + { + "epoch": 0.4, + "grad_norm": 2.186256574707549, + "learning_rate": 1.3772133483120229e-05, + "loss": 0.9488, + "step": 16791 + }, + { + "epoch": 0.4, + "grad_norm": 2.0108117306463154, + "learning_rate": 1.377142679967718e-05, + "loss": 1.0581, + "step": 16792 + }, + { + "epoch": 0.4, + "grad_norm": 2.172449643006168, + "learning_rate": 1.3770720094275685e-05, + "loss": 1.0206, + "step": 16793 + }, + { + "epoch": 0.4, + "grad_norm": 1.9185875000915176, + "learning_rate": 1.3770013366919867e-05, + "loss": 0.9975, + "step": 16794 + }, + { + "epoch": 0.4, + "grad_norm": 2.0790427868980332, + "learning_rate": 1.3769306617613833e-05, + "loss": 1.0462, + "step": 16795 + }, + { + "epoch": 0.4, + "grad_norm": 1.1185547641669529, + "learning_rate": 1.3768599846361707e-05, + "loss": 1.0259, + "step": 16796 + }, + { + "epoch": 0.4, + "grad_norm": 2.2146907424300832, + "learning_rate": 1.37678930531676e-05, + "loss": 1.0641, + "step": 16797 + }, + { + "epoch": 0.4, + "grad_norm": 1.9367657171788875, + "learning_rate": 1.3767186238035624e-05, + "loss": 0.9334, + "step": 16798 + }, + { + "epoch": 0.4, + "grad_norm": 1.8436983952937935, + "learning_rate": 1.3766479400969899e-05, + "loss": 1.0891, + "step": 16799 + }, + { + "epoch": 0.4, + "grad_norm": 1.9929677551419203, + "learning_rate": 1.3765772541974536e-05, + "loss": 0.9746, + "step": 16800 + }, + { + "epoch": 0.4, + "grad_norm": 2.086112864896908, + "learning_rate": 1.3765065661053654e-05, + "loss": 1.078, + "step": 16801 + }, + { + "epoch": 0.4, + "grad_norm": 2.1729230622518307, + "learning_rate": 1.376435875821137e-05, + "loss": 0.9881, + "step": 16802 + }, + { + "epoch": 0.4, + "grad_norm": 1.9491579998560253, + "learning_rate": 1.3763651833451795e-05, + "loss": 1.0407, + "step": 16803 + }, + { + "epoch": 0.4, + "grad_norm": 1.9957685321123628, + "learning_rate": 1.3762944886779049e-05, + "loss": 0.9776, + "step": 16804 + }, + { + "epoch": 0.4, + "grad_norm": 2.010303715889575, + "learning_rate": 1.3762237918197245e-05, + "loss": 1.1094, + "step": 16805 + }, + { + "epoch": 0.4, + "grad_norm": 1.9974009241359834, + "learning_rate": 1.37615309277105e-05, + "loss": 1.0967, + "step": 16806 + }, + { + "epoch": 0.4, + "grad_norm": 1.8395314999472234, + "learning_rate": 1.3760823915322934e-05, + "loss": 0.9384, + "step": 16807 + }, + { + "epoch": 0.4, + "grad_norm": 2.129374779647023, + "learning_rate": 1.3760116881038658e-05, + "loss": 0.993, + "step": 16808 + }, + { + "epoch": 0.4, + "grad_norm": 2.0648299636138825, + "learning_rate": 1.375940982486179e-05, + "loss": 1.0439, + "step": 16809 + }, + { + "epoch": 0.4, + "grad_norm": 3.2614370075910646, + "learning_rate": 1.3758702746796454e-05, + "loss": 1.04, + "step": 16810 + }, + { + "epoch": 0.4, + "grad_norm": 1.8908343704725739, + "learning_rate": 1.3757995646846755e-05, + "loss": 1.0127, + "step": 16811 + }, + { + "epoch": 0.4, + "grad_norm": 2.3615597454738086, + "learning_rate": 1.3757288525016816e-05, + "loss": 1.1489, + "step": 16812 + }, + { + "epoch": 0.4, + "grad_norm": 2.044497725007275, + "learning_rate": 1.375658138131076e-05, + "loss": 0.9301, + "step": 16813 + }, + { + "epoch": 0.4, + "grad_norm": 1.9622945440316286, + "learning_rate": 1.3755874215732687e-05, + "loss": 1.0087, + "step": 16814 + }, + { + "epoch": 0.4, + "grad_norm": 1.8582938419671347, + "learning_rate": 1.3755167028286733e-05, + "loss": 0.937, + "step": 16815 + }, + { + "epoch": 0.4, + "grad_norm": 2.340418360363675, + "learning_rate": 1.3754459818977007e-05, + "loss": 1.0395, + "step": 16816 + }, + { + "epoch": 0.4, + "grad_norm": 2.763453556172355, + "learning_rate": 1.3753752587807627e-05, + "loss": 1.0987, + "step": 16817 + }, + { + "epoch": 0.4, + "grad_norm": 1.223657463172498, + "learning_rate": 1.3753045334782707e-05, + "loss": 1.0152, + "step": 16818 + }, + { + "epoch": 0.4, + "grad_norm": 2.260665405423753, + "learning_rate": 1.3752338059906374e-05, + "loss": 1.1411, + "step": 16819 + }, + { + "epoch": 0.4, + "grad_norm": 1.9949072593413504, + "learning_rate": 1.3751630763182738e-05, + "loss": 1.0544, + "step": 16820 + }, + { + "epoch": 0.4, + "grad_norm": 2.0242951086062075, + "learning_rate": 1.375092344461592e-05, + "loss": 0.9775, + "step": 16821 + }, + { + "epoch": 0.4, + "grad_norm": 1.8867689366902285, + "learning_rate": 1.375021610421004e-05, + "loss": 0.8949, + "step": 16822 + }, + { + "epoch": 0.4, + "grad_norm": 1.9709571153916057, + "learning_rate": 1.3749508741969213e-05, + "loss": 1.1424, + "step": 16823 + }, + { + "epoch": 0.4, + "grad_norm": 1.998829068106704, + "learning_rate": 1.3748801357897559e-05, + "loss": 0.9984, + "step": 16824 + }, + { + "epoch": 0.4, + "grad_norm": 2.1012715412782166, + "learning_rate": 1.3748093951999195e-05, + "loss": 1.1045, + "step": 16825 + }, + { + "epoch": 0.4, + "grad_norm": 1.871710050246647, + "learning_rate": 1.3747386524278246e-05, + "loss": 1.09, + "step": 16826 + }, + { + "epoch": 0.4, + "grad_norm": 1.9147067744636905, + "learning_rate": 1.3746679074738822e-05, + "loss": 0.9999, + "step": 16827 + }, + { + "epoch": 0.4, + "grad_norm": 2.0612318935317973, + "learning_rate": 1.3745971603385049e-05, + "loss": 1.1186, + "step": 16828 + }, + { + "epoch": 0.4, + "grad_norm": 2.9604254371975163, + "learning_rate": 1.3745264110221042e-05, + "loss": 0.9873, + "step": 16829 + }, + { + "epoch": 0.4, + "grad_norm": 1.998736206802606, + "learning_rate": 1.3744556595250922e-05, + "loss": 1.1507, + "step": 16830 + }, + { + "epoch": 0.4, + "grad_norm": 1.727916488355046, + "learning_rate": 1.3743849058478809e-05, + "loss": 1.0615, + "step": 16831 + }, + { + "epoch": 0.4, + "grad_norm": 1.7984896160393227, + "learning_rate": 1.374314149990882e-05, + "loss": 1.0316, + "step": 16832 + }, + { + "epoch": 0.4, + "grad_norm": 2.063642240061853, + "learning_rate": 1.3742433919545076e-05, + "loss": 1.1183, + "step": 16833 + }, + { + "epoch": 0.4, + "grad_norm": 2.0309390022544043, + "learning_rate": 1.3741726317391699e-05, + "loss": 1.0922, + "step": 16834 + }, + { + "epoch": 0.4, + "grad_norm": 2.018195400942582, + "learning_rate": 1.3741018693452803e-05, + "loss": 1.0918, + "step": 16835 + }, + { + "epoch": 0.4, + "grad_norm": 2.1228943542311116, + "learning_rate": 1.3740311047732514e-05, + "loss": 1.0581, + "step": 16836 + }, + { + "epoch": 0.4, + "grad_norm": 2.010972083628172, + "learning_rate": 1.373960338023495e-05, + "loss": 1.0918, + "step": 16837 + }, + { + "epoch": 0.4, + "grad_norm": 1.9252335377541276, + "learning_rate": 1.3738895690964231e-05, + "loss": 1.068, + "step": 16838 + }, + { + "epoch": 0.4, + "grad_norm": 2.0821307585270423, + "learning_rate": 1.3738187979924478e-05, + "loss": 0.9982, + "step": 16839 + }, + { + "epoch": 0.4, + "grad_norm": 2.144043055082392, + "learning_rate": 1.373748024711981e-05, + "loss": 1.0406, + "step": 16840 + }, + { + "epoch": 0.4, + "grad_norm": 1.8667545032272328, + "learning_rate": 1.3736772492554352e-05, + "loss": 1.0441, + "step": 16841 + }, + { + "epoch": 0.4, + "grad_norm": 1.81939060338422, + "learning_rate": 1.3736064716232216e-05, + "loss": 0.9002, + "step": 16842 + }, + { + "epoch": 0.4, + "grad_norm": 1.8080072863693313, + "learning_rate": 1.3735356918157532e-05, + "loss": 0.989, + "step": 16843 + }, + { + "epoch": 0.4, + "grad_norm": 1.0877600703670025, + "learning_rate": 1.3734649098334416e-05, + "loss": 0.8953, + "step": 16844 + }, + { + "epoch": 0.4, + "grad_norm": 2.173840390990783, + "learning_rate": 1.3733941256766994e-05, + "loss": 1.0115, + "step": 16845 + }, + { + "epoch": 0.4, + "grad_norm": 2.219728249575657, + "learning_rate": 1.3733233393459377e-05, + "loss": 1.044, + "step": 16846 + }, + { + "epoch": 0.4, + "grad_norm": 2.0594962079819075, + "learning_rate": 1.3732525508415702e-05, + "loss": 0.9642, + "step": 16847 + }, + { + "epoch": 0.4, + "grad_norm": 1.9875986495742037, + "learning_rate": 1.3731817601640077e-05, + "loss": 1.0464, + "step": 16848 + }, + { + "epoch": 0.4, + "grad_norm": 1.8777191597188374, + "learning_rate": 1.3731109673136627e-05, + "loss": 0.9179, + "step": 16849 + }, + { + "epoch": 0.4, + "grad_norm": 1.9651176255126106, + "learning_rate": 1.3730401722909479e-05, + "loss": 1.0111, + "step": 16850 + }, + { + "epoch": 0.4, + "grad_norm": 2.0223129206554273, + "learning_rate": 1.3729693750962747e-05, + "loss": 0.9872, + "step": 16851 + }, + { + "epoch": 0.4, + "grad_norm": 2.148052267736203, + "learning_rate": 1.372898575730056e-05, + "loss": 1.0838, + "step": 16852 + }, + { + "epoch": 0.4, + "grad_norm": 2.1632388738984547, + "learning_rate": 1.3728277741927037e-05, + "loss": 1.0926, + "step": 16853 + }, + { + "epoch": 0.4, + "grad_norm": 2.1056682967840543, + "learning_rate": 1.37275697048463e-05, + "loss": 0.9556, + "step": 16854 + }, + { + "epoch": 0.4, + "grad_norm": 1.7923046677653456, + "learning_rate": 1.3726861646062472e-05, + "loss": 1.0029, + "step": 16855 + }, + { + "epoch": 0.4, + "grad_norm": 2.0868843038112503, + "learning_rate": 1.3726153565579676e-05, + "loss": 0.9642, + "step": 16856 + }, + { + "epoch": 0.4, + "grad_norm": 2.027272769617037, + "learning_rate": 1.3725445463402034e-05, + "loss": 0.9953, + "step": 16857 + }, + { + "epoch": 0.4, + "grad_norm": 2.000971861029036, + "learning_rate": 1.3724737339533669e-05, + "loss": 1.0776, + "step": 16858 + }, + { + "epoch": 0.4, + "grad_norm": 3.1224452488797936, + "learning_rate": 1.3724029193978706e-05, + "loss": 0.9623, + "step": 16859 + }, + { + "epoch": 0.4, + "grad_norm": 2.761259278207, + "learning_rate": 1.3723321026741262e-05, + "loss": 0.8598, + "step": 16860 + }, + { + "epoch": 0.4, + "grad_norm": 1.9936163551829742, + "learning_rate": 1.3722612837825466e-05, + "loss": 1.0353, + "step": 16861 + }, + { + "epoch": 0.4, + "grad_norm": 2.1904084051104267, + "learning_rate": 1.3721904627235444e-05, + "loss": 1.0385, + "step": 16862 + }, + { + "epoch": 0.4, + "grad_norm": 2.2384834145200694, + "learning_rate": 1.3721196394975306e-05, + "loss": 0.9693, + "step": 16863 + }, + { + "epoch": 0.4, + "grad_norm": 1.8718414157377568, + "learning_rate": 1.3720488141049193e-05, + "loss": 1.1351, + "step": 16864 + }, + { + "epoch": 0.4, + "grad_norm": 2.003218684313179, + "learning_rate": 1.3719779865461214e-05, + "loss": 0.959, + "step": 16865 + }, + { + "epoch": 0.4, + "grad_norm": 1.8763410776572518, + "learning_rate": 1.3719071568215503e-05, + "loss": 1.057, + "step": 16866 + }, + { + "epoch": 0.4, + "grad_norm": 1.102609431324679, + "learning_rate": 1.3718363249316179e-05, + "loss": 1.003, + "step": 16867 + }, + { + "epoch": 0.4, + "grad_norm": 1.8740483836948274, + "learning_rate": 1.3717654908767366e-05, + "loss": 0.9721, + "step": 16868 + }, + { + "epoch": 0.4, + "grad_norm": 1.9182929437048957, + "learning_rate": 1.3716946546573186e-05, + "loss": 1.0152, + "step": 16869 + }, + { + "epoch": 0.4, + "grad_norm": 1.9741244146237498, + "learning_rate": 1.3716238162737772e-05, + "loss": 1.0224, + "step": 16870 + }, + { + "epoch": 0.4, + "grad_norm": 2.147633400449799, + "learning_rate": 1.3715529757265243e-05, + "loss": 0.9355, + "step": 16871 + }, + { + "epoch": 0.4, + "grad_norm": 1.7058803611763405, + "learning_rate": 1.371482133015972e-05, + "loss": 1.0001, + "step": 16872 + }, + { + "epoch": 0.4, + "grad_norm": 2.109098382258878, + "learning_rate": 1.3714112881425332e-05, + "loss": 0.9804, + "step": 16873 + }, + { + "epoch": 0.4, + "grad_norm": 1.981977108862765, + "learning_rate": 1.3713404411066203e-05, + "loss": 0.9222, + "step": 16874 + }, + { + "epoch": 0.4, + "grad_norm": 2.271564078285674, + "learning_rate": 1.3712695919086458e-05, + "loss": 1.1364, + "step": 16875 + }, + { + "epoch": 0.4, + "grad_norm": 1.773488643986382, + "learning_rate": 1.371198740549022e-05, + "loss": 1.0868, + "step": 16876 + }, + { + "epoch": 0.4, + "grad_norm": 2.5180568162292754, + "learning_rate": 1.3711278870281622e-05, + "loss": 1.0331, + "step": 16877 + }, + { + "epoch": 0.4, + "grad_norm": 1.826261492432496, + "learning_rate": 1.3710570313464778e-05, + "loss": 1.0223, + "step": 16878 + }, + { + "epoch": 0.4, + "grad_norm": 2.3044998472403124, + "learning_rate": 1.3709861735043818e-05, + "loss": 0.8397, + "step": 16879 + }, + { + "epoch": 0.4, + "grad_norm": 1.1009497629273672, + "learning_rate": 1.3709153135022873e-05, + "loss": 0.9265, + "step": 16880 + }, + { + "epoch": 0.4, + "grad_norm": 1.9202807957803845, + "learning_rate": 1.3708444513406062e-05, + "loss": 1.0187, + "step": 16881 + }, + { + "epoch": 0.4, + "grad_norm": 1.8903137281573958, + "learning_rate": 1.3707735870197514e-05, + "loss": 1.0699, + "step": 16882 + }, + { + "epoch": 0.4, + "grad_norm": 2.0021291380940003, + "learning_rate": 1.3707027205401353e-05, + "loss": 1.0943, + "step": 16883 + }, + { + "epoch": 0.4, + "grad_norm": 1.9178295484978027, + "learning_rate": 1.3706318519021704e-05, + "loss": 0.9921, + "step": 16884 + }, + { + "epoch": 0.4, + "grad_norm": 1.1520492834136582, + "learning_rate": 1.37056098110627e-05, + "loss": 1.0065, + "step": 16885 + }, + { + "epoch": 0.4, + "grad_norm": 2.1276365879661285, + "learning_rate": 1.3704901081528457e-05, + "loss": 1.0197, + "step": 16886 + }, + { + "epoch": 0.4, + "grad_norm": 1.897285835945043, + "learning_rate": 1.3704192330423109e-05, + "loss": 0.9689, + "step": 16887 + }, + { + "epoch": 0.4, + "grad_norm": 1.880279755670887, + "learning_rate": 1.3703483557750783e-05, + "loss": 0.9695, + "step": 16888 + }, + { + "epoch": 0.4, + "grad_norm": 1.9915319744508357, + "learning_rate": 1.3702774763515601e-05, + "loss": 1.0246, + "step": 16889 + }, + { + "epoch": 0.4, + "grad_norm": 1.9151493397717434, + "learning_rate": 1.3702065947721691e-05, + "loss": 0.9543, + "step": 16890 + }, + { + "epoch": 0.4, + "grad_norm": 1.7112701716674847, + "learning_rate": 1.3701357110373184e-05, + "loss": 1.1061, + "step": 16891 + }, + { + "epoch": 0.4, + "grad_norm": 1.7344054297836369, + "learning_rate": 1.3700648251474204e-05, + "loss": 1.1351, + "step": 16892 + }, + { + "epoch": 0.4, + "grad_norm": 1.8502899937916093, + "learning_rate": 1.3699939371028873e-05, + "loss": 0.9364, + "step": 16893 + }, + { + "epoch": 0.4, + "grad_norm": 1.8745981681158463, + "learning_rate": 1.3699230469041331e-05, + "loss": 1.013, + "step": 16894 + }, + { + "epoch": 0.4, + "grad_norm": 1.8844006965214233, + "learning_rate": 1.3698521545515693e-05, + "loss": 1.113, + "step": 16895 + }, + { + "epoch": 0.4, + "grad_norm": 2.022426383548973, + "learning_rate": 1.3697812600456094e-05, + "loss": 0.951, + "step": 16896 + }, + { + "epoch": 0.4, + "grad_norm": 1.959730651538463, + "learning_rate": 1.3697103633866657e-05, + "loss": 0.9595, + "step": 16897 + }, + { + "epoch": 0.4, + "grad_norm": 1.0788598237547216, + "learning_rate": 1.3696394645751514e-05, + "loss": 0.9764, + "step": 16898 + }, + { + "epoch": 0.4, + "grad_norm": 2.425882194556434, + "learning_rate": 1.369568563611479e-05, + "loss": 1.1514, + "step": 16899 + }, + { + "epoch": 0.4, + "grad_norm": 2.032422402985046, + "learning_rate": 1.3694976604960615e-05, + "loss": 0.9524, + "step": 16900 + }, + { + "epoch": 0.4, + "grad_norm": 2.2515718624514642, + "learning_rate": 1.3694267552293117e-05, + "loss": 0.8933, + "step": 16901 + }, + { + "epoch": 0.4, + "grad_norm": 2.4496311115531713, + "learning_rate": 1.3693558478116425e-05, + "loss": 1.1561, + "step": 16902 + }, + { + "epoch": 0.4, + "grad_norm": 1.0856145133589907, + "learning_rate": 1.369284938243466e-05, + "loss": 0.9531, + "step": 16903 + }, + { + "epoch": 0.4, + "grad_norm": 2.054115965624857, + "learning_rate": 1.3692140265251962e-05, + "loss": 1.0406, + "step": 16904 + }, + { + "epoch": 0.4, + "grad_norm": 2.037385533117242, + "learning_rate": 1.3691431126572454e-05, + "loss": 1.0134, + "step": 16905 + }, + { + "epoch": 0.4, + "grad_norm": 2.0219346727217977, + "learning_rate": 1.3690721966400264e-05, + "loss": 1.0093, + "step": 16906 + }, + { + "epoch": 0.4, + "grad_norm": 1.7911006632093618, + "learning_rate": 1.3690012784739524e-05, + "loss": 1.0733, + "step": 16907 + }, + { + "epoch": 0.4, + "grad_norm": 2.250637598644022, + "learning_rate": 1.368930358159436e-05, + "loss": 0.9172, + "step": 16908 + }, + { + "epoch": 0.4, + "grad_norm": 2.003455572812036, + "learning_rate": 1.3688594356968903e-05, + "loss": 0.9871, + "step": 16909 + }, + { + "epoch": 0.4, + "grad_norm": 2.0219798260671977, + "learning_rate": 1.368788511086728e-05, + "loss": 0.8956, + "step": 16910 + }, + { + "epoch": 0.4, + "grad_norm": 1.9454376399960343, + "learning_rate": 1.3687175843293625e-05, + "loss": 0.9507, + "step": 16911 + }, + { + "epoch": 0.4, + "grad_norm": 1.8366319658027912, + "learning_rate": 1.3686466554252062e-05, + "loss": 1.0774, + "step": 16912 + }, + { + "epoch": 0.4, + "grad_norm": 2.0554126765314558, + "learning_rate": 1.3685757243746728e-05, + "loss": 1.0363, + "step": 16913 + }, + { + "epoch": 0.4, + "grad_norm": 1.8613884588707383, + "learning_rate": 1.3685047911781744e-05, + "loss": 1.0482, + "step": 16914 + }, + { + "epoch": 0.4, + "grad_norm": 2.4729576522248897, + "learning_rate": 1.368433855836125e-05, + "loss": 1.1171, + "step": 16915 + }, + { + "epoch": 0.4, + "grad_norm": 1.088980593356627, + "learning_rate": 1.3683629183489364e-05, + "loss": 0.9704, + "step": 16916 + }, + { + "epoch": 0.4, + "grad_norm": 2.1283564784576865, + "learning_rate": 1.3682919787170228e-05, + "loss": 0.8957, + "step": 16917 + }, + { + "epoch": 0.4, + "grad_norm": 1.7617967237051093, + "learning_rate": 1.3682210369407961e-05, + "loss": 1.0345, + "step": 16918 + }, + { + "epoch": 0.4, + "grad_norm": 2.4467775920174284, + "learning_rate": 1.3681500930206707e-05, + "loss": 1.0802, + "step": 16919 + }, + { + "epoch": 0.4, + "grad_norm": 1.9233526838701547, + "learning_rate": 1.3680791469570585e-05, + "loss": 1.0493, + "step": 16920 + }, + { + "epoch": 0.4, + "grad_norm": 1.794057626847611, + "learning_rate": 1.3680081987503728e-05, + "loss": 0.9775, + "step": 16921 + }, + { + "epoch": 0.4, + "grad_norm": 4.919097974796497, + "learning_rate": 1.3679372484010269e-05, + "loss": 1.1442, + "step": 16922 + }, + { + "epoch": 0.4, + "grad_norm": 1.8353980891450183, + "learning_rate": 1.3678662959094338e-05, + "loss": 1.0991, + "step": 16923 + }, + { + "epoch": 0.4, + "grad_norm": 1.8466798468659815, + "learning_rate": 1.3677953412760072e-05, + "loss": 1.0688, + "step": 16924 + }, + { + "epoch": 0.4, + "grad_norm": 2.1074404358819834, + "learning_rate": 1.3677243845011589e-05, + "loss": 1.0012, + "step": 16925 + }, + { + "epoch": 0.4, + "grad_norm": 1.905139370785974, + "learning_rate": 1.3676534255853036e-05, + "loss": 0.9688, + "step": 16926 + }, + { + "epoch": 0.4, + "grad_norm": 2.040525926474001, + "learning_rate": 1.3675824645288533e-05, + "loss": 0.9344, + "step": 16927 + }, + { + "epoch": 0.4, + "grad_norm": 1.115894222006856, + "learning_rate": 1.3675115013322215e-05, + "loss": 1.0092, + "step": 16928 + }, + { + "epoch": 0.4, + "grad_norm": 1.1528960934813863, + "learning_rate": 1.3674405359958214e-05, + "loss": 1.0029, + "step": 16929 + }, + { + "epoch": 0.4, + "grad_norm": 2.02961293410528, + "learning_rate": 1.367369568520066e-05, + "loss": 0.9105, + "step": 16930 + }, + { + "epoch": 0.4, + "grad_norm": 1.9093694168218345, + "learning_rate": 1.367298598905369e-05, + "loss": 1.0687, + "step": 16931 + }, + { + "epoch": 0.4, + "grad_norm": 1.0590370893286172, + "learning_rate": 1.367227627152143e-05, + "loss": 1.0091, + "step": 16932 + }, + { + "epoch": 0.4, + "grad_norm": 1.8342141841783994, + "learning_rate": 1.3671566532608015e-05, + "loss": 0.9247, + "step": 16933 + }, + { + "epoch": 0.4, + "grad_norm": 1.1797539965806945, + "learning_rate": 1.367085677231758e-05, + "loss": 0.9985, + "step": 16934 + }, + { + "epoch": 0.4, + "grad_norm": 1.9911157659889047, + "learning_rate": 1.3670146990654252e-05, + "loss": 1.0099, + "step": 16935 + }, + { + "epoch": 0.4, + "grad_norm": 2.0298465362052096, + "learning_rate": 1.3669437187622168e-05, + "loss": 1.1095, + "step": 16936 + }, + { + "epoch": 0.4, + "grad_norm": 2.043029851368991, + "learning_rate": 1.3668727363225459e-05, + "loss": 1.0846, + "step": 16937 + }, + { + "epoch": 0.4, + "grad_norm": 2.100734673616271, + "learning_rate": 1.3668017517468255e-05, + "loss": 0.9982, + "step": 16938 + }, + { + "epoch": 0.4, + "grad_norm": 1.8693944219105934, + "learning_rate": 1.3667307650354693e-05, + "loss": 1.1492, + "step": 16939 + }, + { + "epoch": 0.4, + "grad_norm": 1.8649453856749396, + "learning_rate": 1.3666597761888906e-05, + "loss": 1.0735, + "step": 16940 + }, + { + "epoch": 0.4, + "grad_norm": 2.029167400861331, + "learning_rate": 1.3665887852075026e-05, + "loss": 1.0391, + "step": 16941 + }, + { + "epoch": 0.4, + "grad_norm": 2.1369422633551527, + "learning_rate": 1.3665177920917185e-05, + "loss": 0.9945, + "step": 16942 + }, + { + "epoch": 0.4, + "grad_norm": 2.002189858616735, + "learning_rate": 1.3664467968419522e-05, + "loss": 0.8657, + "step": 16943 + }, + { + "epoch": 0.4, + "grad_norm": 1.8672025903886051, + "learning_rate": 1.366375799458616e-05, + "loss": 1.0122, + "step": 16944 + }, + { + "epoch": 0.4, + "grad_norm": 2.078434307385652, + "learning_rate": 1.3663047999421243e-05, + "loss": 1.0921, + "step": 16945 + }, + { + "epoch": 0.4, + "grad_norm": 2.204776276238839, + "learning_rate": 1.36623379829289e-05, + "loss": 1.0089, + "step": 16946 + }, + { + "epoch": 0.4, + "grad_norm": 1.9167955716025487, + "learning_rate": 1.3661627945113268e-05, + "loss": 1.1841, + "step": 16947 + }, + { + "epoch": 0.4, + "grad_norm": 1.8776017155928337, + "learning_rate": 1.3660917885978473e-05, + "loss": 1.0778, + "step": 16948 + }, + { + "epoch": 0.4, + "grad_norm": 2.0458128132803863, + "learning_rate": 1.366020780552866e-05, + "loss": 0.9757, + "step": 16949 + }, + { + "epoch": 0.4, + "grad_norm": 2.0872579559783078, + "learning_rate": 1.3659497703767957e-05, + "loss": 0.9885, + "step": 16950 + }, + { + "epoch": 0.4, + "grad_norm": 2.0909506877632014, + "learning_rate": 1.36587875807005e-05, + "loss": 1.0491, + "step": 16951 + }, + { + "epoch": 0.4, + "grad_norm": 1.8948263970145147, + "learning_rate": 1.3658077436330421e-05, + "loss": 0.9632, + "step": 16952 + }, + { + "epoch": 0.4, + "grad_norm": 1.9015932515091185, + "learning_rate": 1.365736727066186e-05, + "loss": 0.8798, + "step": 16953 + }, + { + "epoch": 0.4, + "grad_norm": 2.0212889029145904, + "learning_rate": 1.3656657083698947e-05, + "loss": 0.96, + "step": 16954 + }, + { + "epoch": 0.4, + "grad_norm": 2.2687270488609297, + "learning_rate": 1.365594687544582e-05, + "loss": 1.0067, + "step": 16955 + }, + { + "epoch": 0.4, + "grad_norm": 1.9803543303475404, + "learning_rate": 1.3655236645906613e-05, + "loss": 1.0683, + "step": 16956 + }, + { + "epoch": 0.4, + "grad_norm": 1.9392740093933698, + "learning_rate": 1.3654526395085461e-05, + "loss": 0.9576, + "step": 16957 + }, + { + "epoch": 0.4, + "grad_norm": 2.169004891851267, + "learning_rate": 1.3653816122986499e-05, + "loss": 1.139, + "step": 16958 + }, + { + "epoch": 0.4, + "grad_norm": 1.9291689236028378, + "learning_rate": 1.3653105829613862e-05, + "loss": 0.9812, + "step": 16959 + }, + { + "epoch": 0.4, + "grad_norm": 2.010731931422174, + "learning_rate": 1.3652395514971686e-05, + "loss": 1.0622, + "step": 16960 + }, + { + "epoch": 0.4, + "grad_norm": 2.0779363364611596, + "learning_rate": 1.365168517906411e-05, + "loss": 0.9964, + "step": 16961 + }, + { + "epoch": 0.4, + "grad_norm": 1.8473101878039444, + "learning_rate": 1.3650974821895262e-05, + "loss": 1.1194, + "step": 16962 + }, + { + "epoch": 0.4, + "grad_norm": 1.8657772961087962, + "learning_rate": 1.3650264443469285e-05, + "loss": 0.9569, + "step": 16963 + }, + { + "epoch": 0.4, + "grad_norm": 2.3407042000705935, + "learning_rate": 1.3649554043790316e-05, + "loss": 1.0819, + "step": 16964 + }, + { + "epoch": 0.4, + "grad_norm": 1.9891077794055225, + "learning_rate": 1.3648843622862483e-05, + "loss": 1.0454, + "step": 16965 + }, + { + "epoch": 0.4, + "grad_norm": 2.1537421363699396, + "learning_rate": 1.3648133180689932e-05, + "loss": 1.0686, + "step": 16966 + }, + { + "epoch": 0.4, + "grad_norm": 2.1091122575105823, + "learning_rate": 1.3647422717276788e-05, + "loss": 0.9579, + "step": 16967 + }, + { + "epoch": 0.4, + "grad_norm": 1.9490964807938045, + "learning_rate": 1.3646712232627201e-05, + "loss": 1.0691, + "step": 16968 + }, + { + "epoch": 0.4, + "grad_norm": 1.9308743429886814, + "learning_rate": 1.3646001726745298e-05, + "loss": 1.1032, + "step": 16969 + }, + { + "epoch": 0.4, + "grad_norm": 1.9166011938683483, + "learning_rate": 1.3645291199635218e-05, + "loss": 0.9611, + "step": 16970 + }, + { + "epoch": 0.4, + "grad_norm": 2.0906229339140006, + "learning_rate": 1.36445806513011e-05, + "loss": 1.0961, + "step": 16971 + }, + { + "epoch": 0.4, + "grad_norm": 1.971836953825256, + "learning_rate": 1.3643870081747076e-05, + "loss": 0.9248, + "step": 16972 + }, + { + "epoch": 0.4, + "grad_norm": 2.3730094054093573, + "learning_rate": 1.364315949097729e-05, + "loss": 0.9175, + "step": 16973 + }, + { + "epoch": 0.4, + "grad_norm": 1.9939067705322786, + "learning_rate": 1.3642448878995876e-05, + "loss": 0.9947, + "step": 16974 + }, + { + "epoch": 0.4, + "grad_norm": 1.1493679354794053, + "learning_rate": 1.3641738245806973e-05, + "loss": 1.0357, + "step": 16975 + }, + { + "epoch": 0.4, + "grad_norm": 1.89876730223978, + "learning_rate": 1.364102759141471e-05, + "loss": 1.0079, + "step": 16976 + }, + { + "epoch": 0.4, + "grad_norm": 1.98401324133707, + "learning_rate": 1.3640316915823242e-05, + "loss": 1.0441, + "step": 16977 + }, + { + "epoch": 0.4, + "grad_norm": 2.2414782905893245, + "learning_rate": 1.363960621903669e-05, + "loss": 0.9967, + "step": 16978 + }, + { + "epoch": 0.4, + "grad_norm": 2.2644248524475734, + "learning_rate": 1.36388955010592e-05, + "loss": 1.0332, + "step": 16979 + }, + { + "epoch": 0.4, + "grad_norm": 2.9845763737397486, + "learning_rate": 1.3638184761894907e-05, + "loss": 0.9567, + "step": 16980 + }, + { + "epoch": 0.4, + "grad_norm": 1.8657553161272356, + "learning_rate": 1.363747400154795e-05, + "loss": 1.0674, + "step": 16981 + }, + { + "epoch": 0.4, + "grad_norm": 1.1125679487435753, + "learning_rate": 1.3636763220022468e-05, + "loss": 0.9944, + "step": 16982 + }, + { + "epoch": 0.4, + "grad_norm": 2.2028088744957506, + "learning_rate": 1.3636052417322601e-05, + "loss": 1.0961, + "step": 16983 + }, + { + "epoch": 0.4, + "grad_norm": 2.0652062100157917, + "learning_rate": 1.3635341593452484e-05, + "loss": 0.9968, + "step": 16984 + }, + { + "epoch": 0.4, + "grad_norm": 2.0786427850737454, + "learning_rate": 1.3634630748416257e-05, + "loss": 1.0082, + "step": 16985 + }, + { + "epoch": 0.4, + "grad_norm": 2.063749920123637, + "learning_rate": 1.3633919882218063e-05, + "loss": 1.0744, + "step": 16986 + }, + { + "epoch": 0.4, + "grad_norm": 2.0975166301862282, + "learning_rate": 1.3633208994862033e-05, + "loss": 0.9687, + "step": 16987 + }, + { + "epoch": 0.4, + "grad_norm": 1.9972461434165303, + "learning_rate": 1.363249808635231e-05, + "loss": 0.9599, + "step": 16988 + }, + { + "epoch": 0.4, + "grad_norm": 2.252385680735083, + "learning_rate": 1.3631787156693035e-05, + "loss": 1.1676, + "step": 16989 + }, + { + "epoch": 0.4, + "grad_norm": 1.9681916010278542, + "learning_rate": 1.3631076205888344e-05, + "loss": 0.9755, + "step": 16990 + }, + { + "epoch": 0.4, + "grad_norm": 1.8649124904452736, + "learning_rate": 1.363036523394238e-05, + "loss": 1.1996, + "step": 16991 + }, + { + "epoch": 0.4, + "grad_norm": 1.9957003888217644, + "learning_rate": 1.3629654240859279e-05, + "loss": 1.0889, + "step": 16992 + }, + { + "epoch": 0.4, + "grad_norm": 1.1259043499585186, + "learning_rate": 1.362894322664318e-05, + "loss": 0.914, + "step": 16993 + }, + { + "epoch": 0.4, + "grad_norm": 2.373096142167173, + "learning_rate": 1.3628232191298227e-05, + "loss": 1.0597, + "step": 16994 + }, + { + "epoch": 0.4, + "grad_norm": 2.1469745244488063, + "learning_rate": 1.3627521134828553e-05, + "loss": 1.0412, + "step": 16995 + }, + { + "epoch": 0.4, + "grad_norm": 1.9282312586194728, + "learning_rate": 1.3626810057238311e-05, + "loss": 1.0142, + "step": 16996 + }, + { + "epoch": 0.4, + "grad_norm": 1.9410438938197356, + "learning_rate": 1.3626098958531624e-05, + "loss": 0.9744, + "step": 16997 + }, + { + "epoch": 0.4, + "grad_norm": 1.9457749399745146, + "learning_rate": 1.3625387838712647e-05, + "loss": 0.9954, + "step": 16998 + }, + { + "epoch": 0.4, + "grad_norm": 1.0781380682391652, + "learning_rate": 1.3624676697785511e-05, + "loss": 1.0408, + "step": 16999 + }, + { + "epoch": 0.4, + "grad_norm": 1.9837035980229303, + "learning_rate": 1.3623965535754363e-05, + "loss": 1.0065, + "step": 17000 + }, + { + "epoch": 0.4, + "grad_norm": 3.211068837453916, + "learning_rate": 1.3623254352623336e-05, + "loss": 1.005, + "step": 17001 + }, + { + "epoch": 0.4, + "grad_norm": 1.1304111963824879, + "learning_rate": 1.3622543148396577e-05, + "loss": 1.0017, + "step": 17002 + }, + { + "epoch": 0.4, + "grad_norm": 2.1661523844805872, + "learning_rate": 1.3621831923078224e-05, + "loss": 1.0282, + "step": 17003 + }, + { + "epoch": 0.4, + "grad_norm": 2.398670501100722, + "learning_rate": 1.362112067667242e-05, + "loss": 1.0303, + "step": 17004 + }, + { + "epoch": 0.4, + "grad_norm": 2.0714525858984416, + "learning_rate": 1.3620409409183304e-05, + "loss": 1.1159, + "step": 17005 + }, + { + "epoch": 0.4, + "grad_norm": 2.0899587604035164, + "learning_rate": 1.3619698120615019e-05, + "loss": 1.0314, + "step": 17006 + }, + { + "epoch": 0.4, + "grad_norm": 2.076803272651677, + "learning_rate": 1.3618986810971707e-05, + "loss": 1.0222, + "step": 17007 + }, + { + "epoch": 0.4, + "grad_norm": 2.2777891840896403, + "learning_rate": 1.3618275480257506e-05, + "loss": 0.8789, + "step": 17008 + }, + { + "epoch": 0.4, + "grad_norm": 1.9020586003252835, + "learning_rate": 1.361756412847656e-05, + "loss": 0.8944, + "step": 17009 + }, + { + "epoch": 0.4, + "grad_norm": 1.0842541074449, + "learning_rate": 1.361685275563301e-05, + "loss": 0.9998, + "step": 17010 + }, + { + "epoch": 0.4, + "grad_norm": 2.2297339043102333, + "learning_rate": 1.3616141361730999e-05, + "loss": 1.1003, + "step": 17011 + }, + { + "epoch": 0.4, + "grad_norm": 1.1295315503316785, + "learning_rate": 1.3615429946774668e-05, + "loss": 1.0036, + "step": 17012 + }, + { + "epoch": 0.4, + "grad_norm": 2.060799553822623, + "learning_rate": 1.3614718510768156e-05, + "loss": 1.0295, + "step": 17013 + }, + { + "epoch": 0.4, + "grad_norm": 2.23934569209259, + "learning_rate": 1.3614007053715612e-05, + "loss": 1.0471, + "step": 17014 + }, + { + "epoch": 0.4, + "grad_norm": 2.347581426978046, + "learning_rate": 1.3613295575621177e-05, + "loss": 0.8958, + "step": 17015 + }, + { + "epoch": 0.4, + "grad_norm": 2.270755010704719, + "learning_rate": 1.3612584076488986e-05, + "loss": 1.0626, + "step": 17016 + }, + { + "epoch": 0.4, + "grad_norm": 2.38071920011199, + "learning_rate": 1.361187255632319e-05, + "loss": 0.8927, + "step": 17017 + }, + { + "epoch": 0.4, + "grad_norm": 1.9744063589257292, + "learning_rate": 1.3611161015127926e-05, + "loss": 1.1098, + "step": 17018 + }, + { + "epoch": 0.4, + "grad_norm": 1.9619612044902988, + "learning_rate": 1.3610449452907345e-05, + "loss": 0.9377, + "step": 17019 + }, + { + "epoch": 0.4, + "grad_norm": 1.9734112654382134, + "learning_rate": 1.3609737869665579e-05, + "loss": 1.1948, + "step": 17020 + }, + { + "epoch": 0.4, + "grad_norm": 2.605296886536463, + "learning_rate": 1.3609026265406775e-05, + "loss": 0.9512, + "step": 17021 + }, + { + "epoch": 0.4, + "grad_norm": 2.0471562768323586, + "learning_rate": 1.3608314640135084e-05, + "loss": 1.0264, + "step": 17022 + }, + { + "epoch": 0.4, + "grad_norm": 2.1005940601204247, + "learning_rate": 1.3607602993854637e-05, + "loss": 1.0019, + "step": 17023 + }, + { + "epoch": 0.4, + "grad_norm": 1.0810143907784902, + "learning_rate": 1.3606891326569584e-05, + "loss": 1.0339, + "step": 17024 + }, + { + "epoch": 0.4, + "grad_norm": 1.9733686643164576, + "learning_rate": 1.3606179638284067e-05, + "loss": 0.9186, + "step": 17025 + }, + { + "epoch": 0.4, + "grad_norm": 1.9494141365958397, + "learning_rate": 1.3605467929002233e-05, + "loss": 1.0956, + "step": 17026 + }, + { + "epoch": 0.4, + "grad_norm": 2.3209642194820055, + "learning_rate": 1.3604756198728222e-05, + "loss": 0.9623, + "step": 17027 + }, + { + "epoch": 0.4, + "grad_norm": 2.010723666494405, + "learning_rate": 1.3604044447466182e-05, + "loss": 0.8985, + "step": 17028 + }, + { + "epoch": 0.4, + "grad_norm": 2.2862876976501973, + "learning_rate": 1.360333267522025e-05, + "loss": 1.0198, + "step": 17029 + }, + { + "epoch": 0.4, + "grad_norm": 2.107005502550033, + "learning_rate": 1.3602620881994577e-05, + "loss": 1.0201, + "step": 17030 + }, + { + "epoch": 0.4, + "grad_norm": 1.110288072983564, + "learning_rate": 1.3601909067793304e-05, + "loss": 0.9821, + "step": 17031 + }, + { + "epoch": 0.4, + "grad_norm": 2.1827536751487684, + "learning_rate": 1.3601197232620575e-05, + "loss": 0.9165, + "step": 17032 + }, + { + "epoch": 0.4, + "grad_norm": 1.0446183170369598, + "learning_rate": 1.3600485376480536e-05, + "loss": 0.9234, + "step": 17033 + }, + { + "epoch": 0.4, + "grad_norm": 1.8596924909689523, + "learning_rate": 1.3599773499377333e-05, + "loss": 1.0523, + "step": 17034 + }, + { + "epoch": 0.4, + "grad_norm": 1.9724834560027142, + "learning_rate": 1.359906160131511e-05, + "loss": 0.9025, + "step": 17035 + }, + { + "epoch": 0.4, + "grad_norm": 1.786129400725571, + "learning_rate": 1.3598349682298006e-05, + "loss": 0.9756, + "step": 17036 + }, + { + "epoch": 0.4, + "grad_norm": 1.8646916925666215, + "learning_rate": 1.3597637742330174e-05, + "loss": 0.9424, + "step": 17037 + }, + { + "epoch": 0.4, + "grad_norm": 1.862500100092797, + "learning_rate": 1.3596925781415755e-05, + "loss": 0.9919, + "step": 17038 + }, + { + "epoch": 0.4, + "grad_norm": 2.3884949283785106, + "learning_rate": 1.3596213799558898e-05, + "loss": 1.1002, + "step": 17039 + }, + { + "epoch": 0.4, + "grad_norm": 1.032739828474857, + "learning_rate": 1.359550179676374e-05, + "loss": 0.9826, + "step": 17040 + }, + { + "epoch": 0.4, + "grad_norm": 2.717859570522428, + "learning_rate": 1.3594789773034439e-05, + "loss": 1.0507, + "step": 17041 + }, + { + "epoch": 0.4, + "grad_norm": 1.0850669969305937, + "learning_rate": 1.3594077728375129e-05, + "loss": 0.961, + "step": 17042 + }, + { + "epoch": 0.4, + "grad_norm": 2.1153432904105802, + "learning_rate": 1.3593365662789966e-05, + "loss": 1.0314, + "step": 17043 + }, + { + "epoch": 0.4, + "grad_norm": 3.3636101590688976, + "learning_rate": 1.3592653576283083e-05, + "loss": 1.0663, + "step": 17044 + }, + { + "epoch": 0.4, + "grad_norm": 2.394736193064776, + "learning_rate": 1.359194146885864e-05, + "loss": 1.1272, + "step": 17045 + }, + { + "epoch": 0.4, + "grad_norm": 1.8477731840600702, + "learning_rate": 1.359122934052077e-05, + "loss": 1.0814, + "step": 17046 + }, + { + "epoch": 0.4, + "grad_norm": 2.0958689500044176, + "learning_rate": 1.3590517191273634e-05, + "loss": 1.02, + "step": 17047 + }, + { + "epoch": 0.4, + "grad_norm": 2.1777809499455807, + "learning_rate": 1.3589805021121363e-05, + "loss": 1.047, + "step": 17048 + }, + { + "epoch": 0.4, + "grad_norm": 1.8834402719080592, + "learning_rate": 1.3589092830068116e-05, + "loss": 1.0265, + "step": 17049 + }, + { + "epoch": 0.4, + "grad_norm": 1.0910630925004234, + "learning_rate": 1.3588380618118031e-05, + "loss": 0.8912, + "step": 17050 + }, + { + "epoch": 0.4, + "grad_norm": 2.211071414523007, + "learning_rate": 1.3587668385275258e-05, + "loss": 1.0362, + "step": 17051 + }, + { + "epoch": 0.4, + "grad_norm": 1.9800442224391417, + "learning_rate": 1.3586956131543944e-05, + "loss": 1.1035, + "step": 17052 + }, + { + "epoch": 0.4, + "grad_norm": 2.4831761137082977, + "learning_rate": 1.3586243856928238e-05, + "loss": 0.9974, + "step": 17053 + }, + { + "epoch": 0.4, + "grad_norm": 1.098971027130805, + "learning_rate": 1.3585531561432284e-05, + "loss": 1.0205, + "step": 17054 + }, + { + "epoch": 0.4, + "grad_norm": 2.114189611792986, + "learning_rate": 1.3584819245060228e-05, + "loss": 1.0968, + "step": 17055 + }, + { + "epoch": 0.4, + "grad_norm": 2.02837013746821, + "learning_rate": 1.3584106907816222e-05, + "loss": 1.114, + "step": 17056 + }, + { + "epoch": 0.4, + "grad_norm": 1.9442493992098202, + "learning_rate": 1.3583394549704404e-05, + "loss": 1.0471, + "step": 17057 + }, + { + "epoch": 0.4, + "grad_norm": 2.1877691885603334, + "learning_rate": 1.3582682170728939e-05, + "loss": 1.0514, + "step": 17058 + }, + { + "epoch": 0.4, + "grad_norm": 2.691202824895486, + "learning_rate": 1.3581969770893959e-05, + "loss": 0.9027, + "step": 17059 + }, + { + "epoch": 0.4, + "grad_norm": 2.081837914532396, + "learning_rate": 1.3581257350203616e-05, + "loss": 1.0813, + "step": 17060 + }, + { + "epoch": 0.4, + "grad_norm": 1.9928373096292562, + "learning_rate": 1.3580544908662062e-05, + "loss": 0.9617, + "step": 17061 + }, + { + "epoch": 0.4, + "grad_norm": 2.5979587967620814, + "learning_rate": 1.3579832446273438e-05, + "loss": 1.052, + "step": 17062 + }, + { + "epoch": 0.4, + "grad_norm": 2.227182853019757, + "learning_rate": 1.35791199630419e-05, + "loss": 1.1274, + "step": 17063 + }, + { + "epoch": 0.4, + "grad_norm": 1.1120432279726804, + "learning_rate": 1.357840745897159e-05, + "loss": 1.0113, + "step": 17064 + }, + { + "epoch": 0.4, + "grad_norm": 2.1619217028962314, + "learning_rate": 1.3577694934066659e-05, + "loss": 0.9975, + "step": 17065 + }, + { + "epoch": 0.4, + "grad_norm": 2.051707879938183, + "learning_rate": 1.3576982388331258e-05, + "loss": 0.9883, + "step": 17066 + }, + { + "epoch": 0.4, + "grad_norm": 2.181976228512659, + "learning_rate": 1.357626982176953e-05, + "loss": 1.162, + "step": 17067 + }, + { + "epoch": 0.4, + "grad_norm": 2.5695253528557456, + "learning_rate": 1.3575557234385628e-05, + "loss": 1.0243, + "step": 17068 + }, + { + "epoch": 0.4, + "grad_norm": 2.0563777153153144, + "learning_rate": 1.3574844626183698e-05, + "loss": 0.87, + "step": 17069 + }, + { + "epoch": 0.4, + "grad_norm": 2.327940426250899, + "learning_rate": 1.3574131997167891e-05, + "loss": 1.074, + "step": 17070 + }, + { + "epoch": 0.4, + "grad_norm": 1.0754290383231906, + "learning_rate": 1.3573419347342358e-05, + "loss": 1.0158, + "step": 17071 + }, + { + "epoch": 0.4, + "grad_norm": 1.9800203783832007, + "learning_rate": 1.3572706676711247e-05, + "loss": 0.9595, + "step": 17072 + }, + { + "epoch": 0.4, + "grad_norm": 1.8588576865678397, + "learning_rate": 1.3571993985278706e-05, + "loss": 1.0874, + "step": 17073 + }, + { + "epoch": 0.4, + "grad_norm": 2.1787670403775827, + "learning_rate": 1.3571281273048885e-05, + "loss": 1.0691, + "step": 17074 + }, + { + "epoch": 0.4, + "grad_norm": 2.0333655853116555, + "learning_rate": 1.3570568540025933e-05, + "loss": 1.0095, + "step": 17075 + }, + { + "epoch": 0.4, + "grad_norm": 2.46590850332446, + "learning_rate": 1.3569855786213998e-05, + "loss": 1.094, + "step": 17076 + }, + { + "epoch": 0.4, + "grad_norm": 1.908383917272601, + "learning_rate": 1.3569143011617238e-05, + "loss": 1.1193, + "step": 17077 + }, + { + "epoch": 0.4, + "grad_norm": 2.047564578628654, + "learning_rate": 1.3568430216239792e-05, + "loss": 1.0296, + "step": 17078 + }, + { + "epoch": 0.4, + "grad_norm": 2.132716708599725, + "learning_rate": 1.356771740008582e-05, + "loss": 1.1777, + "step": 17079 + }, + { + "epoch": 0.4, + "grad_norm": 1.996046597386684, + "learning_rate": 1.3567004563159467e-05, + "loss": 0.9679, + "step": 17080 + }, + { + "epoch": 0.4, + "grad_norm": 2.410268264618157, + "learning_rate": 1.3566291705464883e-05, + "loss": 1.0856, + "step": 17081 + }, + { + "epoch": 0.4, + "grad_norm": 2.0921540351652763, + "learning_rate": 1.3565578827006218e-05, + "loss": 0.9992, + "step": 17082 + }, + { + "epoch": 0.4, + "grad_norm": 2.1955043937518033, + "learning_rate": 1.3564865927787627e-05, + "loss": 1.0862, + "step": 17083 + }, + { + "epoch": 0.4, + "grad_norm": 2.00720877502179, + "learning_rate": 1.3564153007813257e-05, + "loss": 1.1846, + "step": 17084 + }, + { + "epoch": 0.4, + "grad_norm": 2.072477739041491, + "learning_rate": 1.356344006708726e-05, + "loss": 0.9928, + "step": 17085 + }, + { + "epoch": 0.4, + "grad_norm": 2.084300846718376, + "learning_rate": 1.3562727105613784e-05, + "loss": 0.9652, + "step": 17086 + }, + { + "epoch": 0.4, + "grad_norm": 2.1143098935987417, + "learning_rate": 1.3562014123396988e-05, + "loss": 0.9487, + "step": 17087 + }, + { + "epoch": 0.4, + "grad_norm": 3.5634544720256414, + "learning_rate": 1.3561301120441015e-05, + "loss": 0.9391, + "step": 17088 + }, + { + "epoch": 0.4, + "grad_norm": 2.2011464453009335, + "learning_rate": 1.3560588096750018e-05, + "loss": 1.0248, + "step": 17089 + }, + { + "epoch": 0.4, + "grad_norm": 1.9123482289527918, + "learning_rate": 1.355987505232815e-05, + "loss": 1.1052, + "step": 17090 + }, + { + "epoch": 0.4, + "grad_norm": 1.926668356328337, + "learning_rate": 1.3559161987179565e-05, + "loss": 1.2257, + "step": 17091 + }, + { + "epoch": 0.4, + "grad_norm": 2.1328973097191866, + "learning_rate": 1.3558448901308411e-05, + "loss": 1.1105, + "step": 17092 + }, + { + "epoch": 0.4, + "grad_norm": 1.9103326828752714, + "learning_rate": 1.355773579471884e-05, + "loss": 1.0391, + "step": 17093 + }, + { + "epoch": 0.4, + "grad_norm": 2.8355427913456377, + "learning_rate": 1.3557022667415009e-05, + "loss": 1.0183, + "step": 17094 + }, + { + "epoch": 0.4, + "grad_norm": 2.0496192150114227, + "learning_rate": 1.3556309519401061e-05, + "loss": 0.9848, + "step": 17095 + }, + { + "epoch": 0.4, + "grad_norm": 1.9765670994471385, + "learning_rate": 1.3555596350681158e-05, + "loss": 1.1404, + "step": 17096 + }, + { + "epoch": 0.4, + "grad_norm": 2.243430082469992, + "learning_rate": 1.355488316125944e-05, + "loss": 1.0913, + "step": 17097 + }, + { + "epoch": 0.4, + "grad_norm": 1.867664518197699, + "learning_rate": 1.3554169951140073e-05, + "loss": 1.1289, + "step": 17098 + }, + { + "epoch": 0.4, + "grad_norm": 2.331135403926056, + "learning_rate": 1.35534567203272e-05, + "loss": 1.0259, + "step": 17099 + }, + { + "epoch": 0.4, + "grad_norm": 7.508952670591522, + "learning_rate": 1.355274346882498e-05, + "loss": 0.9712, + "step": 17100 + }, + { + "epoch": 0.4, + "grad_norm": 1.9428668725561558, + "learning_rate": 1.355203019663756e-05, + "loss": 0.96, + "step": 17101 + }, + { + "epoch": 0.4, + "grad_norm": 1.808076876322305, + "learning_rate": 1.3551316903769097e-05, + "loss": 0.9359, + "step": 17102 + }, + { + "epoch": 0.4, + "grad_norm": 1.817060374077716, + "learning_rate": 1.3550603590223744e-05, + "loss": 1.0914, + "step": 17103 + }, + { + "epoch": 0.4, + "grad_norm": 2.1145317350597512, + "learning_rate": 1.3549890256005649e-05, + "loss": 0.9965, + "step": 17104 + }, + { + "epoch": 0.4, + "grad_norm": 2.5077995270433626, + "learning_rate": 1.354917690111897e-05, + "loss": 1.0048, + "step": 17105 + }, + { + "epoch": 0.4, + "grad_norm": 2.011863521695279, + "learning_rate": 1.3548463525567858e-05, + "loss": 1.0598, + "step": 17106 + }, + { + "epoch": 0.4, + "grad_norm": 1.8791033731619764, + "learning_rate": 1.3547750129356474e-05, + "loss": 0.985, + "step": 17107 + }, + { + "epoch": 0.4, + "grad_norm": 1.8768520651781053, + "learning_rate": 1.3547036712488961e-05, + "loss": 1.0522, + "step": 17108 + }, + { + "epoch": 0.4, + "grad_norm": 1.103189648106326, + "learning_rate": 1.3546323274969479e-05, + "loss": 1.0443, + "step": 17109 + }, + { + "epoch": 0.4, + "grad_norm": 2.1519277779912214, + "learning_rate": 1.354560981680218e-05, + "loss": 1.0377, + "step": 17110 + }, + { + "epoch": 0.4, + "grad_norm": 1.9581700829737907, + "learning_rate": 1.3544896337991219e-05, + "loss": 1.1725, + "step": 17111 + }, + { + "epoch": 0.4, + "grad_norm": 1.8426614889326207, + "learning_rate": 1.3544182838540745e-05, + "loss": 1.085, + "step": 17112 + }, + { + "epoch": 0.4, + "grad_norm": 2.319322608909456, + "learning_rate": 1.354346931845492e-05, + "loss": 1.0134, + "step": 17113 + }, + { + "epoch": 0.4, + "grad_norm": 1.769388638627512, + "learning_rate": 1.3542755777737892e-05, + "loss": 1.047, + "step": 17114 + }, + { + "epoch": 0.4, + "grad_norm": 2.3180386485927102, + "learning_rate": 1.3542042216393825e-05, + "loss": 1.0697, + "step": 17115 + }, + { + "epoch": 0.4, + "grad_norm": 2.3939558496093154, + "learning_rate": 1.3541328634426858e-05, + "loss": 1.0821, + "step": 17116 + }, + { + "epoch": 0.4, + "grad_norm": 2.400693551318208, + "learning_rate": 1.3540615031841163e-05, + "loss": 1.1304, + "step": 17117 + }, + { + "epoch": 0.4, + "grad_norm": 2.0783061478742177, + "learning_rate": 1.353990140864088e-05, + "loss": 1.0316, + "step": 17118 + }, + { + "epoch": 0.4, + "grad_norm": 2.2115742715048214, + "learning_rate": 1.353918776483017e-05, + "loss": 0.9573, + "step": 17119 + }, + { + "epoch": 0.4, + "grad_norm": 2.1003475162526226, + "learning_rate": 1.353847410041319e-05, + "loss": 0.9724, + "step": 17120 + }, + { + "epoch": 0.4, + "grad_norm": 2.0136477550353336, + "learning_rate": 1.3537760415394095e-05, + "loss": 0.8462, + "step": 17121 + }, + { + "epoch": 0.4, + "grad_norm": 2.3032962050591927, + "learning_rate": 1.3537046709777037e-05, + "loss": 0.9374, + "step": 17122 + }, + { + "epoch": 0.4, + "grad_norm": 1.1624077740404364, + "learning_rate": 1.3536332983566177e-05, + "loss": 0.9461, + "step": 17123 + }, + { + "epoch": 0.4, + "grad_norm": 2.5057020344716356, + "learning_rate": 1.3535619236765665e-05, + "loss": 1.0878, + "step": 17124 + }, + { + "epoch": 0.4, + "grad_norm": 2.2379088328100725, + "learning_rate": 1.3534905469379654e-05, + "loss": 1.0432, + "step": 17125 + }, + { + "epoch": 0.4, + "grad_norm": 2.242962362212403, + "learning_rate": 1.3534191681412309e-05, + "loss": 1.0466, + "step": 17126 + }, + { + "epoch": 0.4, + "grad_norm": 2.0069435897085577, + "learning_rate": 1.353347787286778e-05, + "loss": 0.9844, + "step": 17127 + }, + { + "epoch": 0.4, + "grad_norm": 2.2401860592761134, + "learning_rate": 1.3532764043750224e-05, + "loss": 1.0885, + "step": 17128 + }, + { + "epoch": 0.4, + "grad_norm": 2.2242015279720175, + "learning_rate": 1.3532050194063795e-05, + "loss": 1.0384, + "step": 17129 + }, + { + "epoch": 0.4, + "grad_norm": 2.2154209398897193, + "learning_rate": 1.3531336323812656e-05, + "loss": 1.0198, + "step": 17130 + }, + { + "epoch": 0.4, + "grad_norm": 2.432858865509854, + "learning_rate": 1.3530622433000958e-05, + "loss": 1.0056, + "step": 17131 + }, + { + "epoch": 0.4, + "grad_norm": 1.9482590391391097, + "learning_rate": 1.3529908521632854e-05, + "loss": 1.1443, + "step": 17132 + }, + { + "epoch": 0.4, + "grad_norm": 2.371472677117314, + "learning_rate": 1.352919458971251e-05, + "loss": 1.0673, + "step": 17133 + }, + { + "epoch": 0.4, + "grad_norm": 2.093267297970896, + "learning_rate": 1.3528480637244075e-05, + "loss": 0.9867, + "step": 17134 + }, + { + "epoch": 0.4, + "grad_norm": 1.9486683392591095, + "learning_rate": 1.3527766664231709e-05, + "loss": 0.9962, + "step": 17135 + }, + { + "epoch": 0.4, + "grad_norm": 2.159284698463796, + "learning_rate": 1.352705267067957e-05, + "loss": 1.11, + "step": 17136 + }, + { + "epoch": 0.4, + "grad_norm": 1.1001780999491446, + "learning_rate": 1.3526338656591813e-05, + "loss": 0.9844, + "step": 17137 + }, + { + "epoch": 0.4, + "grad_norm": 1.8900029706184582, + "learning_rate": 1.3525624621972595e-05, + "loss": 1.1171, + "step": 17138 + }, + { + "epoch": 0.4, + "grad_norm": 2.0054193401036287, + "learning_rate": 1.3524910566826075e-05, + "loss": 0.9992, + "step": 17139 + }, + { + "epoch": 0.4, + "grad_norm": 2.217797608560883, + "learning_rate": 1.3524196491156409e-05, + "loss": 1.0708, + "step": 17140 + }, + { + "epoch": 0.4, + "grad_norm": 1.794908164452458, + "learning_rate": 1.3523482394967756e-05, + "loss": 0.8746, + "step": 17141 + }, + { + "epoch": 0.4, + "grad_norm": 2.0540555347664116, + "learning_rate": 1.3522768278264273e-05, + "loss": 1.068, + "step": 17142 + }, + { + "epoch": 0.4, + "grad_norm": 1.9744320217571059, + "learning_rate": 1.3522054141050118e-05, + "loss": 0.9054, + "step": 17143 + }, + { + "epoch": 0.4, + "grad_norm": 1.8927129541463419, + "learning_rate": 1.352133998332945e-05, + "loss": 1.0789, + "step": 17144 + }, + { + "epoch": 0.4, + "grad_norm": 2.723467139460599, + "learning_rate": 1.3520625805106425e-05, + "loss": 1.042, + "step": 17145 + }, + { + "epoch": 0.4, + "grad_norm": 2.4135080498622274, + "learning_rate": 1.3519911606385199e-05, + "loss": 1.0094, + "step": 17146 + }, + { + "epoch": 0.4, + "grad_norm": 1.7584107004981853, + "learning_rate": 1.3519197387169939e-05, + "loss": 0.9869, + "step": 17147 + }, + { + "epoch": 0.4, + "grad_norm": 2.251308888554377, + "learning_rate": 1.3518483147464792e-05, + "loss": 0.9385, + "step": 17148 + }, + { + "epoch": 0.4, + "grad_norm": 2.0082914809909673, + "learning_rate": 1.3517768887273923e-05, + "loss": 1.1305, + "step": 17149 + }, + { + "epoch": 0.4, + "grad_norm": 1.9365247028133166, + "learning_rate": 1.3517054606601492e-05, + "loss": 1.0607, + "step": 17150 + }, + { + "epoch": 0.4, + "grad_norm": 1.8779131762899033, + "learning_rate": 1.3516340305451654e-05, + "loss": 1.0091, + "step": 17151 + }, + { + "epoch": 0.4, + "grad_norm": 1.9949828221443435, + "learning_rate": 1.3515625983828572e-05, + "loss": 0.8543, + "step": 17152 + }, + { + "epoch": 0.4, + "grad_norm": 2.081154112161527, + "learning_rate": 1.35149116417364e-05, + "loss": 0.9707, + "step": 17153 + }, + { + "epoch": 0.4, + "grad_norm": 1.9408988008812238, + "learning_rate": 1.35141972791793e-05, + "loss": 1.1051, + "step": 17154 + }, + { + "epoch": 0.4, + "grad_norm": 2.1517471366708207, + "learning_rate": 1.351348289616143e-05, + "loss": 0.8843, + "step": 17155 + }, + { + "epoch": 0.4, + "grad_norm": 1.79978438066834, + "learning_rate": 1.3512768492686954e-05, + "loss": 1.1293, + "step": 17156 + }, + { + "epoch": 0.4, + "grad_norm": 2.0101966343611815, + "learning_rate": 1.351205406876002e-05, + "loss": 0.9453, + "step": 17157 + }, + { + "epoch": 0.4, + "grad_norm": 2.0548466402783623, + "learning_rate": 1.3511339624384805e-05, + "loss": 1.0531, + "step": 17158 + }, + { + "epoch": 0.4, + "grad_norm": 1.9309207051531159, + "learning_rate": 1.3510625159565453e-05, + "loss": 1.1147, + "step": 17159 + }, + { + "epoch": 0.4, + "grad_norm": 3.020671376629157, + "learning_rate": 1.3509910674306133e-05, + "loss": 0.9956, + "step": 17160 + }, + { + "epoch": 0.4, + "grad_norm": 1.8895410164036566, + "learning_rate": 1.3509196168610999e-05, + "loss": 1.0268, + "step": 17161 + }, + { + "epoch": 0.4, + "grad_norm": 2.1235315881915366, + "learning_rate": 1.3508481642484215e-05, + "loss": 1.2205, + "step": 17162 + }, + { + "epoch": 0.4, + "grad_norm": 1.1368698729590963, + "learning_rate": 1.350776709592994e-05, + "loss": 0.9828, + "step": 17163 + }, + { + "epoch": 0.4, + "grad_norm": 1.9465014697810716, + "learning_rate": 1.3507052528952333e-05, + "loss": 1.1063, + "step": 17164 + }, + { + "epoch": 0.4, + "grad_norm": 2.9619866883265846, + "learning_rate": 1.3506337941555559e-05, + "loss": 0.9275, + "step": 17165 + }, + { + "epoch": 0.4, + "grad_norm": 2.207713855618826, + "learning_rate": 1.3505623333743774e-05, + "loss": 1.165, + "step": 17166 + }, + { + "epoch": 0.4, + "grad_norm": 1.924929293947075, + "learning_rate": 1.350490870552114e-05, + "loss": 1.1621, + "step": 17167 + }, + { + "epoch": 0.4, + "grad_norm": 2.3195113491182395, + "learning_rate": 1.3504194056891817e-05, + "loss": 1.0297, + "step": 17168 + }, + { + "epoch": 0.4, + "grad_norm": 2.1271807935819718, + "learning_rate": 1.3503479387859967e-05, + "loss": 0.8886, + "step": 17169 + }, + { + "epoch": 0.4, + "grad_norm": 1.9460683084704313, + "learning_rate": 1.3502764698429753e-05, + "loss": 1.0462, + "step": 17170 + }, + { + "epoch": 0.4, + "grad_norm": 1.8910034138169598, + "learning_rate": 1.3502049988605332e-05, + "loss": 0.8387, + "step": 17171 + }, + { + "epoch": 0.4, + "grad_norm": 2.519135336856317, + "learning_rate": 1.3501335258390868e-05, + "loss": 1.0106, + "step": 17172 + }, + { + "epoch": 0.4, + "grad_norm": 1.9557718426240518, + "learning_rate": 1.3500620507790523e-05, + "loss": 0.9842, + "step": 17173 + }, + { + "epoch": 0.4, + "grad_norm": 2.0100475158538114, + "learning_rate": 1.3499905736808455e-05, + "loss": 1.1569, + "step": 17174 + }, + { + "epoch": 0.4, + "grad_norm": 2.548535683268564, + "learning_rate": 1.3499190945448828e-05, + "loss": 1.0279, + "step": 17175 + }, + { + "epoch": 0.4, + "grad_norm": 2.030273304586512, + "learning_rate": 1.3498476133715803e-05, + "loss": 0.9861, + "step": 17176 + }, + { + "epoch": 0.4, + "grad_norm": 1.8846768077864648, + "learning_rate": 1.3497761301613543e-05, + "loss": 1.1346, + "step": 17177 + }, + { + "epoch": 0.4, + "grad_norm": 1.767573040617696, + "learning_rate": 1.3497046449146209e-05, + "loss": 1.05, + "step": 17178 + }, + { + "epoch": 0.4, + "grad_norm": 3.671484084385015, + "learning_rate": 1.3496331576317964e-05, + "loss": 0.9714, + "step": 17179 + }, + { + "epoch": 0.4, + "grad_norm": 1.0703812954713725, + "learning_rate": 1.3495616683132969e-05, + "loss": 0.9624, + "step": 17180 + }, + { + "epoch": 0.4, + "grad_norm": 1.0700841057646056, + "learning_rate": 1.349490176959539e-05, + "loss": 0.9834, + "step": 17181 + }, + { + "epoch": 0.4, + "grad_norm": 1.9281024790359385, + "learning_rate": 1.3494186835709383e-05, + "loss": 1.0055, + "step": 17182 + }, + { + "epoch": 0.4, + "grad_norm": 2.047176795652562, + "learning_rate": 1.3493471881479114e-05, + "loss": 0.9699, + "step": 17183 + }, + { + "epoch": 0.4, + "grad_norm": 2.0193461586047574, + "learning_rate": 1.3492756906908748e-05, + "loss": 0.9593, + "step": 17184 + }, + { + "epoch": 0.4, + "grad_norm": 1.8548072738857186, + "learning_rate": 1.3492041912002442e-05, + "loss": 0.9653, + "step": 17185 + }, + { + "epoch": 0.4, + "grad_norm": 1.7672561752743259, + "learning_rate": 1.3491326896764366e-05, + "loss": 0.9173, + "step": 17186 + }, + { + "epoch": 0.4, + "grad_norm": 2.180669441805883, + "learning_rate": 1.3490611861198676e-05, + "loss": 1.0574, + "step": 17187 + }, + { + "epoch": 0.4, + "grad_norm": 2.15416282365785, + "learning_rate": 1.3489896805309541e-05, + "loss": 1.0889, + "step": 17188 + }, + { + "epoch": 0.4, + "grad_norm": 1.9449895378585436, + "learning_rate": 1.3489181729101122e-05, + "loss": 1.0675, + "step": 17189 + }, + { + "epoch": 0.4, + "grad_norm": 1.8632625072439273, + "learning_rate": 1.3488466632577581e-05, + "loss": 0.9351, + "step": 17190 + }, + { + "epoch": 0.41, + "grad_norm": 2.0677334356821118, + "learning_rate": 1.3487751515743084e-05, + "loss": 1.0698, + "step": 17191 + }, + { + "epoch": 0.41, + "grad_norm": 1.7643694073838077, + "learning_rate": 1.3487036378601794e-05, + "loss": 0.942, + "step": 17192 + }, + { + "epoch": 0.41, + "grad_norm": 2.124006418181887, + "learning_rate": 1.3486321221157872e-05, + "loss": 0.9168, + "step": 17193 + }, + { + "epoch": 0.41, + "grad_norm": 1.9256344791526958, + "learning_rate": 1.3485606043415487e-05, + "loss": 0.9974, + "step": 17194 + }, + { + "epoch": 0.41, + "grad_norm": 2.307903692949641, + "learning_rate": 1.3484890845378798e-05, + "loss": 0.9388, + "step": 17195 + }, + { + "epoch": 0.41, + "grad_norm": 2.1496033499595324, + "learning_rate": 1.3484175627051974e-05, + "loss": 0.9863, + "step": 17196 + }, + { + "epoch": 0.41, + "grad_norm": 1.236976636875644, + "learning_rate": 1.3483460388439172e-05, + "loss": 1.003, + "step": 17197 + }, + { + "epoch": 0.41, + "grad_norm": 1.1833525083682261, + "learning_rate": 1.3482745129544564e-05, + "loss": 0.9939, + "step": 17198 + }, + { + "epoch": 0.41, + "grad_norm": 2.308519726198217, + "learning_rate": 1.3482029850372309e-05, + "loss": 1.1076, + "step": 17199 + }, + { + "epoch": 0.41, + "grad_norm": 2.0778781327005738, + "learning_rate": 1.3481314550926576e-05, + "loss": 1.0246, + "step": 17200 + }, + { + "epoch": 0.41, + "grad_norm": 2.4786526157030586, + "learning_rate": 1.3480599231211524e-05, + "loss": 1.0424, + "step": 17201 + }, + { + "epoch": 0.41, + "grad_norm": 2.088487998658536, + "learning_rate": 1.3479883891231326e-05, + "loss": 0.9635, + "step": 17202 + }, + { + "epoch": 0.41, + "grad_norm": 1.8895109241188939, + "learning_rate": 1.347916853099014e-05, + "loss": 0.9725, + "step": 17203 + }, + { + "epoch": 0.41, + "grad_norm": 1.9330489984335395, + "learning_rate": 1.3478453150492132e-05, + "loss": 1.0025, + "step": 17204 + }, + { + "epoch": 0.41, + "grad_norm": 1.8879048622896062, + "learning_rate": 1.3477737749741473e-05, + "loss": 0.9326, + "step": 17205 + }, + { + "epoch": 0.41, + "grad_norm": 1.8593293867982765, + "learning_rate": 1.3477022328742318e-05, + "loss": 1.0256, + "step": 17206 + }, + { + "epoch": 0.41, + "grad_norm": 2.338529173082693, + "learning_rate": 1.3476306887498842e-05, + "loss": 1.0673, + "step": 17207 + }, + { + "epoch": 0.41, + "grad_norm": 1.8244024474990514, + "learning_rate": 1.3475591426015204e-05, + "loss": 1.0576, + "step": 17208 + }, + { + "epoch": 0.41, + "grad_norm": 2.074605370533662, + "learning_rate": 1.3474875944295576e-05, + "loss": 0.8925, + "step": 17209 + }, + { + "epoch": 0.41, + "grad_norm": 2.164327837694462, + "learning_rate": 1.3474160442344118e-05, + "loss": 0.9148, + "step": 17210 + }, + { + "epoch": 0.41, + "grad_norm": 1.031290593689426, + "learning_rate": 1.3473444920164997e-05, + "loss": 0.9117, + "step": 17211 + }, + { + "epoch": 0.41, + "grad_norm": 1.9678411571502383, + "learning_rate": 1.3472729377762381e-05, + "loss": 1.0516, + "step": 17212 + }, + { + "epoch": 0.41, + "grad_norm": 2.312507171776805, + "learning_rate": 1.3472013815140437e-05, + "loss": 0.9597, + "step": 17213 + }, + { + "epoch": 0.41, + "grad_norm": 2.019940846687804, + "learning_rate": 1.3471298232303326e-05, + "loss": 1.082, + "step": 17214 + }, + { + "epoch": 0.41, + "grad_norm": 2.192505141819406, + "learning_rate": 1.347058262925522e-05, + "loss": 0.9647, + "step": 17215 + }, + { + "epoch": 0.41, + "grad_norm": 1.8633145117190084, + "learning_rate": 1.346986700600028e-05, + "loss": 1.0125, + "step": 17216 + }, + { + "epoch": 0.41, + "grad_norm": 2.0823177956588186, + "learning_rate": 1.346915136254268e-05, + "loss": 1.1244, + "step": 17217 + }, + { + "epoch": 0.41, + "grad_norm": 1.8639255995638067, + "learning_rate": 1.3468435698886582e-05, + "loss": 1.1605, + "step": 17218 + }, + { + "epoch": 0.41, + "grad_norm": 1.1365804452917665, + "learning_rate": 1.3467720015036152e-05, + "loss": 0.9455, + "step": 17219 + }, + { + "epoch": 0.41, + "grad_norm": 1.2384131418499538, + "learning_rate": 1.3467004310995557e-05, + "loss": 1.0547, + "step": 17220 + }, + { + "epoch": 0.41, + "grad_norm": 2.759683759147323, + "learning_rate": 1.3466288586768968e-05, + "loss": 0.9416, + "step": 17221 + }, + { + "epoch": 0.41, + "grad_norm": 2.061266648650093, + "learning_rate": 1.3465572842360546e-05, + "loss": 1.1379, + "step": 17222 + }, + { + "epoch": 0.41, + "grad_norm": 1.8595492602882415, + "learning_rate": 1.3464857077774465e-05, + "loss": 1.1248, + "step": 17223 + }, + { + "epoch": 0.41, + "grad_norm": 2.097440440115682, + "learning_rate": 1.346414129301489e-05, + "loss": 0.958, + "step": 17224 + }, + { + "epoch": 0.41, + "grad_norm": 2.184269162895787, + "learning_rate": 1.3463425488085982e-05, + "loss": 1.0795, + "step": 17225 + }, + { + "epoch": 0.41, + "grad_norm": 1.936867443271658, + "learning_rate": 1.346270966299192e-05, + "loss": 0.9802, + "step": 17226 + }, + { + "epoch": 0.41, + "grad_norm": 2.776153965007002, + "learning_rate": 1.3461993817736861e-05, + "loss": 1.0979, + "step": 17227 + }, + { + "epoch": 0.41, + "grad_norm": 2.1260968736311763, + "learning_rate": 1.3461277952324982e-05, + "loss": 1.0347, + "step": 17228 + }, + { + "epoch": 0.41, + "grad_norm": 1.7360307054708837, + "learning_rate": 1.3460562066760444e-05, + "loss": 0.9801, + "step": 17229 + }, + { + "epoch": 0.41, + "grad_norm": 2.210768553867864, + "learning_rate": 1.3459846161047421e-05, + "loss": 0.9159, + "step": 17230 + }, + { + "epoch": 0.41, + "grad_norm": 1.8868695864200167, + "learning_rate": 1.3459130235190077e-05, + "loss": 1.2095, + "step": 17231 + }, + { + "epoch": 0.41, + "grad_norm": 2.165992131398184, + "learning_rate": 1.3458414289192582e-05, + "loss": 1.0334, + "step": 17232 + }, + { + "epoch": 0.41, + "grad_norm": 1.0643178535611404, + "learning_rate": 1.3457698323059102e-05, + "loss": 0.962, + "step": 17233 + }, + { + "epoch": 0.41, + "grad_norm": 2.4946780256403462, + "learning_rate": 1.3456982336793811e-05, + "loss": 1.0833, + "step": 17234 + }, + { + "epoch": 0.41, + "grad_norm": 2.238593852244075, + "learning_rate": 1.3456266330400873e-05, + "loss": 0.8533, + "step": 17235 + }, + { + "epoch": 0.41, + "grad_norm": 1.9649667054396094, + "learning_rate": 1.3455550303884458e-05, + "loss": 1.1097, + "step": 17236 + }, + { + "epoch": 0.41, + "grad_norm": 1.9050807218737305, + "learning_rate": 1.3454834257248734e-05, + "loss": 0.8865, + "step": 17237 + }, + { + "epoch": 0.41, + "grad_norm": 2.1034425371457495, + "learning_rate": 1.3454118190497874e-05, + "loss": 1.0937, + "step": 17238 + }, + { + "epoch": 0.41, + "grad_norm": 2.2574173474683112, + "learning_rate": 1.3453402103636042e-05, + "loss": 1.0867, + "step": 17239 + }, + { + "epoch": 0.41, + "grad_norm": 1.9564261974011372, + "learning_rate": 1.3452685996667409e-05, + "loss": 0.9959, + "step": 17240 + }, + { + "epoch": 0.41, + "grad_norm": 2.181850866865631, + "learning_rate": 1.3451969869596146e-05, + "loss": 0.9856, + "step": 17241 + }, + { + "epoch": 0.41, + "grad_norm": 2.0431378251976455, + "learning_rate": 1.3451253722426423e-05, + "loss": 1.0593, + "step": 17242 + }, + { + "epoch": 0.41, + "grad_norm": 2.635791294556305, + "learning_rate": 1.3450537555162406e-05, + "loss": 1.0077, + "step": 17243 + }, + { + "epoch": 0.41, + "grad_norm": 1.9073361722615443, + "learning_rate": 1.3449821367808269e-05, + "loss": 1.0218, + "step": 17244 + }, + { + "epoch": 0.41, + "grad_norm": 1.1635928202529713, + "learning_rate": 1.3449105160368179e-05, + "loss": 0.9543, + "step": 17245 + }, + { + "epoch": 0.41, + "grad_norm": 2.0109492091617533, + "learning_rate": 1.3448388932846306e-05, + "loss": 1.0476, + "step": 17246 + }, + { + "epoch": 0.41, + "grad_norm": 1.9198719369364152, + "learning_rate": 1.3447672685246822e-05, + "loss": 0.9951, + "step": 17247 + }, + { + "epoch": 0.41, + "grad_norm": 1.8999876801211357, + "learning_rate": 1.3446956417573894e-05, + "loss": 1.1131, + "step": 17248 + }, + { + "epoch": 0.41, + "grad_norm": 2.0931308234692763, + "learning_rate": 1.3446240129831699e-05, + "loss": 0.9619, + "step": 17249 + }, + { + "epoch": 0.41, + "grad_norm": 1.8975925163111795, + "learning_rate": 1.3445523822024398e-05, + "loss": 1.0094, + "step": 17250 + }, + { + "epoch": 0.41, + "grad_norm": 1.9653185387853134, + "learning_rate": 1.344480749415617e-05, + "loss": 0.8889, + "step": 17251 + }, + { + "epoch": 0.41, + "grad_norm": 2.314594841543717, + "learning_rate": 1.3444091146231178e-05, + "loss": 0.988, + "step": 17252 + }, + { + "epoch": 0.41, + "grad_norm": 2.040860985756672, + "learning_rate": 1.3443374778253597e-05, + "loss": 1.0137, + "step": 17253 + }, + { + "epoch": 0.41, + "grad_norm": 1.8469346378310112, + "learning_rate": 1.3442658390227604e-05, + "loss": 1.0992, + "step": 17254 + }, + { + "epoch": 0.41, + "grad_norm": 1.8619357031332184, + "learning_rate": 1.3441941982157358e-05, + "loss": 1.0114, + "step": 17255 + }, + { + "epoch": 0.41, + "grad_norm": 2.0096998040595313, + "learning_rate": 1.3441225554047039e-05, + "loss": 0.968, + "step": 17256 + }, + { + "epoch": 0.41, + "grad_norm": 2.0642524172129213, + "learning_rate": 1.3440509105900813e-05, + "loss": 0.9571, + "step": 17257 + }, + { + "epoch": 0.41, + "grad_norm": 2.037866195801372, + "learning_rate": 1.3439792637722859e-05, + "loss": 1.0494, + "step": 17258 + }, + { + "epoch": 0.41, + "grad_norm": 2.2561214858281553, + "learning_rate": 1.3439076149517337e-05, + "loss": 0.9739, + "step": 17259 + }, + { + "epoch": 0.41, + "grad_norm": 1.9494785779967954, + "learning_rate": 1.343835964128843e-05, + "loss": 1.0231, + "step": 17260 + }, + { + "epoch": 0.41, + "grad_norm": 2.1125298265223975, + "learning_rate": 1.3437643113040302e-05, + "loss": 0.8962, + "step": 17261 + }, + { + "epoch": 0.41, + "grad_norm": 2.077765313213279, + "learning_rate": 1.3436926564777127e-05, + "loss": 1.0467, + "step": 17262 + }, + { + "epoch": 0.41, + "grad_norm": 2.0212495011445406, + "learning_rate": 1.3436209996503078e-05, + "loss": 0.9948, + "step": 17263 + }, + { + "epoch": 0.41, + "grad_norm": 1.7660792584526273, + "learning_rate": 1.3435493408222326e-05, + "loss": 1.0485, + "step": 17264 + }, + { + "epoch": 0.41, + "grad_norm": 2.5553007303060515, + "learning_rate": 1.3434776799939045e-05, + "loss": 1.0388, + "step": 17265 + }, + { + "epoch": 0.41, + "grad_norm": 2.3940553435321483, + "learning_rate": 1.3434060171657403e-05, + "loss": 1.0749, + "step": 17266 + }, + { + "epoch": 0.41, + "grad_norm": 1.9156457807015046, + "learning_rate": 1.3433343523381578e-05, + "loss": 1.003, + "step": 17267 + }, + { + "epoch": 0.41, + "grad_norm": 1.8489005886200711, + "learning_rate": 1.343262685511574e-05, + "loss": 1.0663, + "step": 17268 + }, + { + "epoch": 0.41, + "grad_norm": 3.0540688895892187, + "learning_rate": 1.3431910166864062e-05, + "loss": 0.9755, + "step": 17269 + }, + { + "epoch": 0.41, + "grad_norm": 1.7416235711764925, + "learning_rate": 1.3431193458630716e-05, + "loss": 0.946, + "step": 17270 + }, + { + "epoch": 0.41, + "grad_norm": 2.1001798146498287, + "learning_rate": 1.3430476730419874e-05, + "loss": 1.0741, + "step": 17271 + }, + { + "epoch": 0.41, + "grad_norm": 1.816832331256786, + "learning_rate": 1.3429759982235712e-05, + "loss": 1.0759, + "step": 17272 + }, + { + "epoch": 0.41, + "grad_norm": 2.0074519158200683, + "learning_rate": 1.3429043214082403e-05, + "loss": 0.9247, + "step": 17273 + }, + { + "epoch": 0.41, + "grad_norm": 1.785746159281043, + "learning_rate": 1.3428326425964115e-05, + "loss": 0.8434, + "step": 17274 + }, + { + "epoch": 0.41, + "grad_norm": 2.1105270316103493, + "learning_rate": 1.3427609617885029e-05, + "loss": 0.9624, + "step": 17275 + }, + { + "epoch": 0.41, + "grad_norm": 2.1319021027074703, + "learning_rate": 1.3426892789849312e-05, + "loss": 0.9774, + "step": 17276 + }, + { + "epoch": 0.41, + "grad_norm": 2.099041058728691, + "learning_rate": 1.3426175941861144e-05, + "loss": 0.9737, + "step": 17277 + }, + { + "epoch": 0.41, + "grad_norm": 2.4390565765206293, + "learning_rate": 1.3425459073924689e-05, + "loss": 0.929, + "step": 17278 + }, + { + "epoch": 0.41, + "grad_norm": 1.8192251975940374, + "learning_rate": 1.3424742186044131e-05, + "loss": 0.9831, + "step": 17279 + }, + { + "epoch": 0.41, + "grad_norm": 1.9689287861846694, + "learning_rate": 1.3424025278223637e-05, + "loss": 0.9374, + "step": 17280 + }, + { + "epoch": 0.41, + "grad_norm": 1.955771182145871, + "learning_rate": 1.3423308350467388e-05, + "loss": 1.0487, + "step": 17281 + }, + { + "epoch": 0.41, + "grad_norm": 1.9633480262652248, + "learning_rate": 1.342259140277955e-05, + "loss": 0.9883, + "step": 17282 + }, + { + "epoch": 0.41, + "grad_norm": 1.8257857986831398, + "learning_rate": 1.3421874435164303e-05, + "loss": 1.076, + "step": 17283 + }, + { + "epoch": 0.41, + "grad_norm": 2.189587321835093, + "learning_rate": 1.3421157447625822e-05, + "loss": 1.0085, + "step": 17284 + }, + { + "epoch": 0.41, + "grad_norm": 7.643991874358363, + "learning_rate": 1.3420440440168275e-05, + "loss": 0.8863, + "step": 17285 + }, + { + "epoch": 0.41, + "grad_norm": 1.995799368829699, + "learning_rate": 1.3419723412795843e-05, + "loss": 0.9248, + "step": 17286 + }, + { + "epoch": 0.41, + "grad_norm": 2.4412755824407424, + "learning_rate": 1.3419006365512697e-05, + "loss": 1.0617, + "step": 17287 + }, + { + "epoch": 0.41, + "grad_norm": 2.22794009045563, + "learning_rate": 1.3418289298323015e-05, + "loss": 1.0273, + "step": 17288 + }, + { + "epoch": 0.41, + "grad_norm": 1.8440469899401635, + "learning_rate": 1.3417572211230969e-05, + "loss": 0.9475, + "step": 17289 + }, + { + "epoch": 0.41, + "grad_norm": 2.119377626952598, + "learning_rate": 1.3416855104240739e-05, + "loss": 1.0894, + "step": 17290 + }, + { + "epoch": 0.41, + "grad_norm": 1.8085422969424074, + "learning_rate": 1.3416137977356491e-05, + "loss": 1.1612, + "step": 17291 + }, + { + "epoch": 0.41, + "grad_norm": 2.3124811487460506, + "learning_rate": 1.341542083058241e-05, + "loss": 0.9546, + "step": 17292 + }, + { + "epoch": 0.41, + "grad_norm": 1.1342962241274377, + "learning_rate": 1.3414703663922668e-05, + "loss": 0.9875, + "step": 17293 + }, + { + "epoch": 0.41, + "grad_norm": 1.2023791479919475, + "learning_rate": 1.3413986477381439e-05, + "loss": 1.0031, + "step": 17294 + }, + { + "epoch": 0.41, + "grad_norm": 1.9918797874755978, + "learning_rate": 1.3413269270962902e-05, + "loss": 0.8574, + "step": 17295 + }, + { + "epoch": 0.41, + "grad_norm": 1.8170159322284443, + "learning_rate": 1.341255204467123e-05, + "loss": 0.9194, + "step": 17296 + }, + { + "epoch": 0.41, + "grad_norm": 2.1730204515506686, + "learning_rate": 1.3411834798510596e-05, + "loss": 1.0125, + "step": 17297 + }, + { + "epoch": 0.41, + "grad_norm": 2.6082140804910234, + "learning_rate": 1.3411117532485185e-05, + "loss": 0.9939, + "step": 17298 + }, + { + "epoch": 0.41, + "grad_norm": 2.3815442525340313, + "learning_rate": 1.3410400246599163e-05, + "loss": 1.1406, + "step": 17299 + }, + { + "epoch": 0.41, + "grad_norm": 2.4083819295892677, + "learning_rate": 1.3409682940856714e-05, + "loss": 1.1435, + "step": 17300 + }, + { + "epoch": 0.41, + "grad_norm": 2.1376418474725196, + "learning_rate": 1.3408965615262008e-05, + "loss": 1.089, + "step": 17301 + }, + { + "epoch": 0.41, + "grad_norm": 1.8905622768933363, + "learning_rate": 1.3408248269819228e-05, + "loss": 0.9327, + "step": 17302 + }, + { + "epoch": 0.41, + "grad_norm": 1.9533107834313588, + "learning_rate": 1.3407530904532546e-05, + "loss": 0.9092, + "step": 17303 + }, + { + "epoch": 0.41, + "grad_norm": 2.091319005986135, + "learning_rate": 1.3406813519406142e-05, + "loss": 1.1256, + "step": 17304 + }, + { + "epoch": 0.41, + "grad_norm": 1.8650898989800866, + "learning_rate": 1.3406096114444193e-05, + "loss": 1.1897, + "step": 17305 + }, + { + "epoch": 0.41, + "grad_norm": 2.0035832137281733, + "learning_rate": 1.340537868965087e-05, + "loss": 1.1327, + "step": 17306 + }, + { + "epoch": 0.41, + "grad_norm": 2.0284088905867415, + "learning_rate": 1.3404661245030357e-05, + "loss": 1.0863, + "step": 17307 + }, + { + "epoch": 0.41, + "grad_norm": 2.053378060160292, + "learning_rate": 1.3403943780586826e-05, + "loss": 1.0054, + "step": 17308 + }, + { + "epoch": 0.41, + "grad_norm": 2.037227355406173, + "learning_rate": 1.3403226296324458e-05, + "loss": 0.893, + "step": 17309 + }, + { + "epoch": 0.41, + "grad_norm": 3.0270834015600716, + "learning_rate": 1.3402508792247427e-05, + "loss": 1.0971, + "step": 17310 + }, + { + "epoch": 0.41, + "grad_norm": 1.1494922696410188, + "learning_rate": 1.3401791268359916e-05, + "loss": 0.9431, + "step": 17311 + }, + { + "epoch": 0.41, + "grad_norm": 2.2419383553572994, + "learning_rate": 1.3401073724666098e-05, + "loss": 1.0379, + "step": 17312 + }, + { + "epoch": 0.41, + "grad_norm": 1.1355630724374708, + "learning_rate": 1.3400356161170152e-05, + "loss": 1.0211, + "step": 17313 + }, + { + "epoch": 0.41, + "grad_norm": 1.0830444119282747, + "learning_rate": 1.3399638577876255e-05, + "loss": 0.9502, + "step": 17314 + }, + { + "epoch": 0.41, + "grad_norm": 2.3730772316721764, + "learning_rate": 1.3398920974788585e-05, + "loss": 1.0806, + "step": 17315 + }, + { + "epoch": 0.41, + "grad_norm": 1.8756882286103886, + "learning_rate": 1.3398203351911323e-05, + "loss": 0.9783, + "step": 17316 + }, + { + "epoch": 0.41, + "grad_norm": 2.0962570820144912, + "learning_rate": 1.3397485709248643e-05, + "loss": 1.046, + "step": 17317 + }, + { + "epoch": 0.41, + "grad_norm": 2.087239888991482, + "learning_rate": 1.3396768046804728e-05, + "loss": 0.8726, + "step": 17318 + }, + { + "epoch": 0.41, + "grad_norm": 2.163159005640289, + "learning_rate": 1.3396050364583752e-05, + "loss": 1.0481, + "step": 17319 + }, + { + "epoch": 0.41, + "grad_norm": 1.9902034305898741, + "learning_rate": 1.3395332662589898e-05, + "loss": 1.1343, + "step": 17320 + }, + { + "epoch": 0.41, + "grad_norm": 2.102305996473277, + "learning_rate": 1.339461494082734e-05, + "loss": 0.9734, + "step": 17321 + }, + { + "epoch": 0.41, + "grad_norm": 1.7617873389410716, + "learning_rate": 1.3393897199300262e-05, + "loss": 1.0115, + "step": 17322 + }, + { + "epoch": 0.41, + "grad_norm": 2.5776354773833314, + "learning_rate": 1.3393179438012836e-05, + "loss": 0.9651, + "step": 17323 + }, + { + "epoch": 0.41, + "grad_norm": 2.0820932579245133, + "learning_rate": 1.3392461656969246e-05, + "loss": 0.9838, + "step": 17324 + }, + { + "epoch": 0.41, + "grad_norm": 2.0236023937884906, + "learning_rate": 1.3391743856173673e-05, + "loss": 1.0236, + "step": 17325 + }, + { + "epoch": 0.41, + "grad_norm": 2.152005622421293, + "learning_rate": 1.3391026035630294e-05, + "loss": 1.0835, + "step": 17326 + }, + { + "epoch": 0.41, + "grad_norm": 2.321932044024828, + "learning_rate": 1.3390308195343282e-05, + "loss": 0.9897, + "step": 17327 + }, + { + "epoch": 0.41, + "grad_norm": 2.218381187644411, + "learning_rate": 1.338959033531683e-05, + "loss": 0.9426, + "step": 17328 + }, + { + "epoch": 0.41, + "grad_norm": 2.076292946166142, + "learning_rate": 1.3388872455555102e-05, + "loss": 1.0874, + "step": 17329 + }, + { + "epoch": 0.41, + "grad_norm": 1.9319680575731175, + "learning_rate": 1.3388154556062292e-05, + "loss": 0.939, + "step": 17330 + }, + { + "epoch": 0.41, + "grad_norm": 1.9092611238179802, + "learning_rate": 1.338743663684257e-05, + "loss": 1.2379, + "step": 17331 + }, + { + "epoch": 0.41, + "grad_norm": 1.8084197689334638, + "learning_rate": 1.3386718697900123e-05, + "loss": 1.1163, + "step": 17332 + }, + { + "epoch": 0.41, + "grad_norm": 2.3221932877967824, + "learning_rate": 1.3386000739239125e-05, + "loss": 1.0043, + "step": 17333 + }, + { + "epoch": 0.41, + "grad_norm": 1.1625474713103008, + "learning_rate": 1.3385282760863758e-05, + "loss": 0.9535, + "step": 17334 + }, + { + "epoch": 0.41, + "grad_norm": 2.0819067571128813, + "learning_rate": 1.3384564762778203e-05, + "loss": 0.9956, + "step": 17335 + }, + { + "epoch": 0.41, + "grad_norm": 1.168049874685869, + "learning_rate": 1.3383846744986642e-05, + "loss": 0.9955, + "step": 17336 + }, + { + "epoch": 0.41, + "grad_norm": 2.534944945575258, + "learning_rate": 1.3383128707493253e-05, + "loss": 1.0625, + "step": 17337 + }, + { + "epoch": 0.41, + "grad_norm": 2.1808874831095606, + "learning_rate": 1.3382410650302218e-05, + "loss": 1.0803, + "step": 17338 + }, + { + "epoch": 0.41, + "grad_norm": 1.912791895626299, + "learning_rate": 1.3381692573417718e-05, + "loss": 1.0167, + "step": 17339 + }, + { + "epoch": 0.41, + "grad_norm": 2.0431748570236454, + "learning_rate": 1.3380974476843932e-05, + "loss": 1.0287, + "step": 17340 + }, + { + "epoch": 0.41, + "grad_norm": 1.798335502235005, + "learning_rate": 1.338025636058504e-05, + "loss": 1.0841, + "step": 17341 + }, + { + "epoch": 0.41, + "grad_norm": 2.0982564559390307, + "learning_rate": 1.3379538224645228e-05, + "loss": 1.2377, + "step": 17342 + }, + { + "epoch": 0.41, + "grad_norm": 1.0484472756313807, + "learning_rate": 1.3378820069028675e-05, + "loss": 0.9656, + "step": 17343 + }, + { + "epoch": 0.41, + "grad_norm": 1.7287327689293353, + "learning_rate": 1.3378101893739561e-05, + "loss": 0.9593, + "step": 17344 + }, + { + "epoch": 0.41, + "grad_norm": 2.308322568540381, + "learning_rate": 1.3377383698782066e-05, + "loss": 1.029, + "step": 17345 + }, + { + "epoch": 0.41, + "grad_norm": 2.2356080069166384, + "learning_rate": 1.3376665484160378e-05, + "loss": 0.8396, + "step": 17346 + }, + { + "epoch": 0.41, + "grad_norm": 1.9744658098421644, + "learning_rate": 1.3375947249878675e-05, + "loss": 1.016, + "step": 17347 + }, + { + "epoch": 0.41, + "grad_norm": 2.5006113439236275, + "learning_rate": 1.3375228995941135e-05, + "loss": 0.8865, + "step": 17348 + }, + { + "epoch": 0.41, + "grad_norm": 2.0502533975840245, + "learning_rate": 1.3374510722351945e-05, + "loss": 1.2051, + "step": 17349 + }, + { + "epoch": 0.41, + "grad_norm": 2.114924111555148, + "learning_rate": 1.3373792429115282e-05, + "loss": 1.0072, + "step": 17350 + }, + { + "epoch": 0.41, + "grad_norm": 2.2357604987841815, + "learning_rate": 1.3373074116235334e-05, + "loss": 0.906, + "step": 17351 + }, + { + "epoch": 0.41, + "grad_norm": 1.9642575778531475, + "learning_rate": 1.3372355783716282e-05, + "loss": 0.8797, + "step": 17352 + }, + { + "epoch": 0.41, + "grad_norm": 2.1352076534170306, + "learning_rate": 1.3371637431562303e-05, + "loss": 1.0801, + "step": 17353 + }, + { + "epoch": 0.41, + "grad_norm": 1.937089898753273, + "learning_rate": 1.337091905977759e-05, + "loss": 1.0218, + "step": 17354 + }, + { + "epoch": 0.41, + "grad_norm": 2.0302397104519736, + "learning_rate": 1.337020066836631e-05, + "loss": 1.0072, + "step": 17355 + }, + { + "epoch": 0.41, + "grad_norm": 2.121811590626321, + "learning_rate": 1.3369482257332664e-05, + "loss": 0.9004, + "step": 17356 + }, + { + "epoch": 0.41, + "grad_norm": 3.06446129261093, + "learning_rate": 1.336876382668082e-05, + "loss": 1.066, + "step": 17357 + }, + { + "epoch": 0.41, + "grad_norm": 2.4784106848323475, + "learning_rate": 1.3368045376414968e-05, + "loss": 1.1086, + "step": 17358 + }, + { + "epoch": 0.41, + "grad_norm": 2.069478468309185, + "learning_rate": 1.3367326906539286e-05, + "loss": 1.0666, + "step": 17359 + }, + { + "epoch": 0.41, + "grad_norm": 1.908463798144876, + "learning_rate": 1.3366608417057966e-05, + "loss": 0.9945, + "step": 17360 + }, + { + "epoch": 0.41, + "grad_norm": 1.795709408420675, + "learning_rate": 1.3365889907975184e-05, + "loss": 0.9914, + "step": 17361 + }, + { + "epoch": 0.41, + "grad_norm": 2.365212429462215, + "learning_rate": 1.3365171379295126e-05, + "loss": 0.957, + "step": 17362 + }, + { + "epoch": 0.41, + "grad_norm": 1.9890290230809127, + "learning_rate": 1.3364452831021972e-05, + "loss": 1.0301, + "step": 17363 + }, + { + "epoch": 0.41, + "grad_norm": 2.030051580502899, + "learning_rate": 1.3363734263159911e-05, + "loss": 1.0307, + "step": 17364 + }, + { + "epoch": 0.41, + "grad_norm": 2.0275298307214347, + "learning_rate": 1.3363015675713124e-05, + "loss": 1.0888, + "step": 17365 + }, + { + "epoch": 0.41, + "grad_norm": 2.432722188087768, + "learning_rate": 1.3362297068685794e-05, + "loss": 1.1045, + "step": 17366 + }, + { + "epoch": 0.41, + "grad_norm": 2.2133088597827792, + "learning_rate": 1.3361578442082107e-05, + "loss": 1.0545, + "step": 17367 + }, + { + "epoch": 0.41, + "grad_norm": 1.9040737404675232, + "learning_rate": 1.3360859795906244e-05, + "loss": 0.9096, + "step": 17368 + }, + { + "epoch": 0.41, + "grad_norm": 1.1451927674202746, + "learning_rate": 1.3360141130162394e-05, + "loss": 1.024, + "step": 17369 + }, + { + "epoch": 0.41, + "grad_norm": 1.8733297632615646, + "learning_rate": 1.3359422444854738e-05, + "loss": 0.8989, + "step": 17370 + }, + { + "epoch": 0.41, + "grad_norm": 2.0749390958925416, + "learning_rate": 1.3358703739987459e-05, + "loss": 1.1609, + "step": 17371 + }, + { + "epoch": 0.41, + "grad_norm": 1.9575111087342882, + "learning_rate": 1.3357985015564744e-05, + "loss": 1.1666, + "step": 17372 + }, + { + "epoch": 0.41, + "grad_norm": 1.9396233878881943, + "learning_rate": 1.335726627159078e-05, + "loss": 1.0351, + "step": 17373 + }, + { + "epoch": 0.41, + "grad_norm": 1.9467665039566329, + "learning_rate": 1.3356547508069745e-05, + "loss": 1.1941, + "step": 17374 + }, + { + "epoch": 0.41, + "grad_norm": 2.066625679328031, + "learning_rate": 1.335582872500583e-05, + "loss": 0.9468, + "step": 17375 + }, + { + "epoch": 0.41, + "grad_norm": 1.9690262252076058, + "learning_rate": 1.3355109922403218e-05, + "loss": 0.9764, + "step": 17376 + }, + { + "epoch": 0.41, + "grad_norm": 1.8722444138936083, + "learning_rate": 1.3354391100266096e-05, + "loss": 0.9276, + "step": 17377 + }, + { + "epoch": 0.41, + "grad_norm": 1.9699426759042027, + "learning_rate": 1.335367225859864e-05, + "loss": 0.9999, + "step": 17378 + }, + { + "epoch": 0.41, + "grad_norm": 1.9961262381162403, + "learning_rate": 1.3352953397405049e-05, + "loss": 0.9796, + "step": 17379 + }, + { + "epoch": 0.41, + "grad_norm": 2.132570729748392, + "learning_rate": 1.3352234516689497e-05, + "loss": 1.1082, + "step": 17380 + }, + { + "epoch": 0.41, + "grad_norm": 1.9752325957310295, + "learning_rate": 1.3351515616456178e-05, + "loss": 1.0186, + "step": 17381 + }, + { + "epoch": 0.41, + "grad_norm": 2.051593385832172, + "learning_rate": 1.3350796696709272e-05, + "loss": 1.0675, + "step": 17382 + }, + { + "epoch": 0.41, + "grad_norm": 1.858371592977765, + "learning_rate": 1.3350077757452969e-05, + "loss": 1.0359, + "step": 17383 + }, + { + "epoch": 0.41, + "grad_norm": 2.209538810961987, + "learning_rate": 1.334935879869145e-05, + "loss": 1.0225, + "step": 17384 + }, + { + "epoch": 0.41, + "grad_norm": 2.5747813320417263, + "learning_rate": 1.3348639820428906e-05, + "loss": 0.9612, + "step": 17385 + }, + { + "epoch": 0.41, + "grad_norm": 1.8054086789289854, + "learning_rate": 1.3347920822669518e-05, + "loss": 0.9807, + "step": 17386 + }, + { + "epoch": 0.41, + "grad_norm": 1.925623764341149, + "learning_rate": 1.3347201805417476e-05, + "loss": 1.0753, + "step": 17387 + }, + { + "epoch": 0.41, + "grad_norm": 1.8732737535183226, + "learning_rate": 1.3346482768676968e-05, + "loss": 1.1011, + "step": 17388 + }, + { + "epoch": 0.41, + "grad_norm": 1.9432773724943475, + "learning_rate": 1.3345763712452174e-05, + "loss": 0.9864, + "step": 17389 + }, + { + "epoch": 0.41, + "grad_norm": 2.022576350557366, + "learning_rate": 1.3345044636747288e-05, + "loss": 1.0618, + "step": 17390 + }, + { + "epoch": 0.41, + "grad_norm": 2.0065274339075376, + "learning_rate": 1.3344325541566492e-05, + "loss": 1.0594, + "step": 17391 + }, + { + "epoch": 0.41, + "grad_norm": 1.9710431515653137, + "learning_rate": 1.334360642691397e-05, + "loss": 1.1295, + "step": 17392 + }, + { + "epoch": 0.41, + "grad_norm": 1.9591846031544284, + "learning_rate": 1.3342887292793918e-05, + "loss": 1.0413, + "step": 17393 + }, + { + "epoch": 0.41, + "grad_norm": 1.9030264207455319, + "learning_rate": 1.3342168139210517e-05, + "loss": 1.0138, + "step": 17394 + }, + { + "epoch": 0.41, + "grad_norm": 1.9498892988392476, + "learning_rate": 1.3341448966167954e-05, + "loss": 0.9456, + "step": 17395 + }, + { + "epoch": 0.41, + "grad_norm": 1.949136699971156, + "learning_rate": 1.3340729773670416e-05, + "loss": 0.9503, + "step": 17396 + }, + { + "epoch": 0.41, + "grad_norm": 1.1291830859219067, + "learning_rate": 1.3340010561722093e-05, + "loss": 1.0247, + "step": 17397 + }, + { + "epoch": 0.41, + "grad_norm": 2.14184956923653, + "learning_rate": 1.3339291330327172e-05, + "loss": 0.9971, + "step": 17398 + }, + { + "epoch": 0.41, + "grad_norm": 2.04226984086617, + "learning_rate": 1.3338572079489835e-05, + "loss": 1.0518, + "step": 17399 + }, + { + "epoch": 0.41, + "grad_norm": 1.9299701384022911, + "learning_rate": 1.3337852809214278e-05, + "loss": 0.9544, + "step": 17400 + }, + { + "epoch": 0.41, + "grad_norm": 2.2004372539485995, + "learning_rate": 1.3337133519504685e-05, + "loss": 1.0525, + "step": 17401 + }, + { + "epoch": 0.41, + "grad_norm": 1.9105230985482842, + "learning_rate": 1.3336414210365245e-05, + "loss": 0.8631, + "step": 17402 + }, + { + "epoch": 0.41, + "grad_norm": 1.9131696706701102, + "learning_rate": 1.3335694881800145e-05, + "loss": 1.0395, + "step": 17403 + }, + { + "epoch": 0.41, + "grad_norm": 1.913215085057098, + "learning_rate": 1.3334975533813572e-05, + "loss": 1.0596, + "step": 17404 + }, + { + "epoch": 0.41, + "grad_norm": 2.1174069080647895, + "learning_rate": 1.3334256166409717e-05, + "loss": 0.7898, + "step": 17405 + }, + { + "epoch": 0.41, + "grad_norm": 2.0021456485146856, + "learning_rate": 1.3333536779592766e-05, + "loss": 0.9709, + "step": 17406 + }, + { + "epoch": 0.41, + "grad_norm": 2.635378707844747, + "learning_rate": 1.333281737336691e-05, + "loss": 1.0191, + "step": 17407 + }, + { + "epoch": 0.41, + "grad_norm": 1.1534049738286865, + "learning_rate": 1.3332097947736333e-05, + "loss": 0.9154, + "step": 17408 + }, + { + "epoch": 0.41, + "grad_norm": 2.1807511196320486, + "learning_rate": 1.333137850270523e-05, + "loss": 1.0292, + "step": 17409 + }, + { + "epoch": 0.41, + "grad_norm": 2.0221639231579043, + "learning_rate": 1.3330659038277785e-05, + "loss": 1.0352, + "step": 17410 + }, + { + "epoch": 0.41, + "grad_norm": 1.1026012016267632, + "learning_rate": 1.3329939554458188e-05, + "loss": 0.892, + "step": 17411 + }, + { + "epoch": 0.41, + "grad_norm": 1.87579456918032, + "learning_rate": 1.332922005125063e-05, + "loss": 0.9293, + "step": 17412 + }, + { + "epoch": 0.41, + "grad_norm": 2.063842581315014, + "learning_rate": 1.3328500528659301e-05, + "loss": 0.9245, + "step": 17413 + }, + { + "epoch": 0.41, + "grad_norm": 2.2648006742120748, + "learning_rate": 1.3327780986688384e-05, + "loss": 0.9434, + "step": 17414 + }, + { + "epoch": 0.41, + "grad_norm": 1.9481201377441741, + "learning_rate": 1.3327061425342073e-05, + "loss": 1.0339, + "step": 17415 + }, + { + "epoch": 0.41, + "grad_norm": 2.4324533559729407, + "learning_rate": 1.332634184462456e-05, + "loss": 1.0163, + "step": 17416 + }, + { + "epoch": 0.41, + "grad_norm": 2.0235891078245953, + "learning_rate": 1.3325622244540028e-05, + "loss": 1.0186, + "step": 17417 + }, + { + "epoch": 0.41, + "grad_norm": 1.8760799990568866, + "learning_rate": 1.3324902625092672e-05, + "loss": 0.9773, + "step": 17418 + }, + { + "epoch": 0.41, + "grad_norm": 1.9397692647190616, + "learning_rate": 1.3324182986286683e-05, + "loss": 0.9591, + "step": 17419 + }, + { + "epoch": 0.41, + "grad_norm": 2.007241145583771, + "learning_rate": 1.3323463328126245e-05, + "loss": 1.0017, + "step": 17420 + }, + { + "epoch": 0.41, + "grad_norm": 2.0277919798628945, + "learning_rate": 1.3322743650615552e-05, + "loss": 0.8922, + "step": 17421 + }, + { + "epoch": 0.41, + "grad_norm": 2.107495213510559, + "learning_rate": 1.3322023953758794e-05, + "loss": 1.1292, + "step": 17422 + }, + { + "epoch": 0.41, + "grad_norm": 1.9791054731937616, + "learning_rate": 1.332130423756016e-05, + "loss": 1.0858, + "step": 17423 + }, + { + "epoch": 0.41, + "grad_norm": 2.1951008456879664, + "learning_rate": 1.3320584502023841e-05, + "loss": 1.1666, + "step": 17424 + }, + { + "epoch": 0.41, + "grad_norm": 2.0262549756475194, + "learning_rate": 1.3319864747154028e-05, + "loss": 1.008, + "step": 17425 + }, + { + "epoch": 0.41, + "grad_norm": 2.2072938106710454, + "learning_rate": 1.3319144972954915e-05, + "loss": 1.0063, + "step": 17426 + }, + { + "epoch": 0.41, + "grad_norm": 2.296954003077432, + "learning_rate": 1.3318425179430683e-05, + "loss": 0.9929, + "step": 17427 + }, + { + "epoch": 0.41, + "grad_norm": 2.068438691895881, + "learning_rate": 1.3317705366585534e-05, + "loss": 0.9498, + "step": 17428 + }, + { + "epoch": 0.41, + "grad_norm": 2.0810756706905584, + "learning_rate": 1.331698553442365e-05, + "loss": 0.8612, + "step": 17429 + }, + { + "epoch": 0.41, + "grad_norm": 1.974199017582064, + "learning_rate": 1.331626568294923e-05, + "loss": 0.8234, + "step": 17430 + }, + { + "epoch": 0.41, + "grad_norm": 2.246293395955097, + "learning_rate": 1.3315545812166457e-05, + "loss": 0.9889, + "step": 17431 + }, + { + "epoch": 0.41, + "grad_norm": 2.431466459352705, + "learning_rate": 1.3314825922079531e-05, + "loss": 1.0745, + "step": 17432 + }, + { + "epoch": 0.41, + "grad_norm": 2.052197581604331, + "learning_rate": 1.3314106012692633e-05, + "loss": 1.0495, + "step": 17433 + }, + { + "epoch": 0.41, + "grad_norm": 2.003838316284586, + "learning_rate": 1.331338608400997e-05, + "loss": 0.9562, + "step": 17434 + }, + { + "epoch": 0.41, + "grad_norm": 1.8479071973069754, + "learning_rate": 1.3312666136035716e-05, + "loss": 0.9949, + "step": 17435 + }, + { + "epoch": 0.41, + "grad_norm": 1.9232866095133838, + "learning_rate": 1.3311946168774072e-05, + "loss": 1.0327, + "step": 17436 + }, + { + "epoch": 0.41, + "grad_norm": 2.132351589322032, + "learning_rate": 1.3311226182229233e-05, + "loss": 1.1099, + "step": 17437 + }, + { + "epoch": 0.41, + "grad_norm": 1.9015876058640828, + "learning_rate": 1.3310506176405383e-05, + "loss": 1.1537, + "step": 17438 + }, + { + "epoch": 0.41, + "grad_norm": 2.2343046707267216, + "learning_rate": 1.3309786151306721e-05, + "loss": 1.1399, + "step": 17439 + }, + { + "epoch": 0.41, + "grad_norm": 2.174857930202718, + "learning_rate": 1.3309066106937432e-05, + "loss": 1.0581, + "step": 17440 + }, + { + "epoch": 0.41, + "grad_norm": 3.4185119401640907, + "learning_rate": 1.3308346043301717e-05, + "loss": 1.0834, + "step": 17441 + }, + { + "epoch": 0.41, + "grad_norm": 2.1718859345303207, + "learning_rate": 1.3307625960403763e-05, + "loss": 0.8799, + "step": 17442 + }, + { + "epoch": 0.41, + "grad_norm": 2.1581110447016187, + "learning_rate": 1.330690585824776e-05, + "loss": 0.9795, + "step": 17443 + }, + { + "epoch": 0.41, + "grad_norm": 2.1994917731432313, + "learning_rate": 1.3306185736837908e-05, + "loss": 1.215, + "step": 17444 + }, + { + "epoch": 0.41, + "grad_norm": 1.9493007498131947, + "learning_rate": 1.3305465596178395e-05, + "loss": 0.9986, + "step": 17445 + }, + { + "epoch": 0.41, + "grad_norm": 1.9351073705805033, + "learning_rate": 1.3304745436273415e-05, + "loss": 0.9777, + "step": 17446 + }, + { + "epoch": 0.41, + "grad_norm": 1.9473070979125908, + "learning_rate": 1.330402525712716e-05, + "loss": 0.9615, + "step": 17447 + }, + { + "epoch": 0.41, + "grad_norm": 2.011893175373321, + "learning_rate": 1.3303305058743823e-05, + "loss": 1.0587, + "step": 17448 + }, + { + "epoch": 0.41, + "grad_norm": 1.9812901330601915, + "learning_rate": 1.33025848411276e-05, + "loss": 1.0262, + "step": 17449 + }, + { + "epoch": 0.41, + "grad_norm": 1.9412770885545987, + "learning_rate": 1.330186460428268e-05, + "loss": 1.0276, + "step": 17450 + }, + { + "epoch": 0.41, + "grad_norm": 1.1922621521443943, + "learning_rate": 1.3301144348213262e-05, + "loss": 0.9361, + "step": 17451 + }, + { + "epoch": 0.41, + "grad_norm": 1.8069491698367837, + "learning_rate": 1.3300424072923534e-05, + "loss": 1.0055, + "step": 17452 + }, + { + "epoch": 0.41, + "grad_norm": 2.281629571727076, + "learning_rate": 1.3299703778417694e-05, + "loss": 0.953, + "step": 17453 + }, + { + "epoch": 0.41, + "grad_norm": 2.1143959594396393, + "learning_rate": 1.3298983464699933e-05, + "loss": 1.0295, + "step": 17454 + }, + { + "epoch": 0.41, + "grad_norm": 2.0794030062143074, + "learning_rate": 1.3298263131774445e-05, + "loss": 0.9875, + "step": 17455 + }, + { + "epoch": 0.41, + "grad_norm": 1.908899121632794, + "learning_rate": 1.3297542779645428e-05, + "loss": 0.9949, + "step": 17456 + }, + { + "epoch": 0.41, + "grad_norm": 2.084478114682358, + "learning_rate": 1.3296822408317067e-05, + "loss": 1.0271, + "step": 17457 + }, + { + "epoch": 0.41, + "grad_norm": 2.0520630665698514, + "learning_rate": 1.3296102017793569e-05, + "loss": 1.0425, + "step": 17458 + }, + { + "epoch": 0.41, + "grad_norm": 2.0815984303358284, + "learning_rate": 1.3295381608079116e-05, + "loss": 1.031, + "step": 17459 + }, + { + "epoch": 0.41, + "grad_norm": 2.017838720680339, + "learning_rate": 1.3294661179177912e-05, + "loss": 1.0331, + "step": 17460 + }, + { + "epoch": 0.41, + "grad_norm": 2.157043336732441, + "learning_rate": 1.3293940731094143e-05, + "loss": 0.8787, + "step": 17461 + }, + { + "epoch": 0.41, + "grad_norm": 2.654545045297203, + "learning_rate": 1.3293220263832015e-05, + "loss": 1.1156, + "step": 17462 + }, + { + "epoch": 0.41, + "grad_norm": 1.9078558083019133, + "learning_rate": 1.3292499777395711e-05, + "loss": 0.961, + "step": 17463 + }, + { + "epoch": 0.41, + "grad_norm": 1.0772578087941644, + "learning_rate": 1.3291779271789433e-05, + "loss": 0.9737, + "step": 17464 + }, + { + "epoch": 0.41, + "grad_norm": 1.897230317623905, + "learning_rate": 1.3291058747017373e-05, + "loss": 0.9486, + "step": 17465 + }, + { + "epoch": 0.41, + "grad_norm": 1.983410890400202, + "learning_rate": 1.3290338203083728e-05, + "loss": 0.9607, + "step": 17466 + }, + { + "epoch": 0.41, + "grad_norm": 2.1885646189975745, + "learning_rate": 1.328961763999269e-05, + "loss": 1.0031, + "step": 17467 + }, + { + "epoch": 0.41, + "grad_norm": 2.2784275006417007, + "learning_rate": 1.3288897057748458e-05, + "loss": 1.0409, + "step": 17468 + }, + { + "epoch": 0.41, + "grad_norm": 1.9937679829835624, + "learning_rate": 1.3288176456355226e-05, + "loss": 1.1484, + "step": 17469 + }, + { + "epoch": 0.41, + "grad_norm": 2.953586615298439, + "learning_rate": 1.3287455835817189e-05, + "loss": 1.0169, + "step": 17470 + }, + { + "epoch": 0.41, + "grad_norm": 1.741676706530068, + "learning_rate": 1.3286735196138544e-05, + "loss": 0.9266, + "step": 17471 + }, + { + "epoch": 0.41, + "grad_norm": 1.95690561377159, + "learning_rate": 1.3286014537323486e-05, + "loss": 0.9272, + "step": 17472 + }, + { + "epoch": 0.41, + "grad_norm": 2.0147898964583133, + "learning_rate": 1.3285293859376212e-05, + "loss": 0.997, + "step": 17473 + }, + { + "epoch": 0.41, + "grad_norm": 1.9354840933898416, + "learning_rate": 1.3284573162300916e-05, + "loss": 1.0809, + "step": 17474 + }, + { + "epoch": 0.41, + "grad_norm": 2.432702437828546, + "learning_rate": 1.3283852446101794e-05, + "loss": 1.0242, + "step": 17475 + }, + { + "epoch": 0.41, + "grad_norm": 1.7999252246243482, + "learning_rate": 1.3283131710783046e-05, + "loss": 1.0854, + "step": 17476 + }, + { + "epoch": 0.41, + "grad_norm": 2.0863592455035125, + "learning_rate": 1.3282410956348865e-05, + "loss": 1.0038, + "step": 17477 + }, + { + "epoch": 0.41, + "grad_norm": 2.7580141601303056, + "learning_rate": 1.3281690182803448e-05, + "loss": 0.9847, + "step": 17478 + }, + { + "epoch": 0.41, + "grad_norm": 1.8944964828447237, + "learning_rate": 1.3280969390150993e-05, + "loss": 1.1363, + "step": 17479 + }, + { + "epoch": 0.41, + "grad_norm": 1.9878211593362485, + "learning_rate": 1.328024857839569e-05, + "loss": 0.9207, + "step": 17480 + }, + { + "epoch": 0.41, + "grad_norm": 2.132267120748608, + "learning_rate": 1.327952774754175e-05, + "loss": 0.9612, + "step": 17481 + }, + { + "epoch": 0.41, + "grad_norm": 1.8957140179392271, + "learning_rate": 1.3278806897593352e-05, + "loss": 0.9861, + "step": 17482 + }, + { + "epoch": 0.41, + "grad_norm": 1.9331876023562127, + "learning_rate": 1.3278086028554711e-05, + "loss": 1.1107, + "step": 17483 + }, + { + "epoch": 0.41, + "grad_norm": 2.0520451544004326, + "learning_rate": 1.327736514043001e-05, + "loss": 1.1423, + "step": 17484 + }, + { + "epoch": 0.41, + "grad_norm": 2.046082853185922, + "learning_rate": 1.327664423322345e-05, + "loss": 1.0235, + "step": 17485 + }, + { + "epoch": 0.41, + "grad_norm": 1.9535559135962985, + "learning_rate": 1.3275923306939235e-05, + "loss": 1.0889, + "step": 17486 + }, + { + "epoch": 0.41, + "grad_norm": 1.9272423546761812, + "learning_rate": 1.3275202361581552e-05, + "loss": 0.8346, + "step": 17487 + }, + { + "epoch": 0.41, + "grad_norm": 2.065865978814327, + "learning_rate": 1.327448139715461e-05, + "loss": 0.9747, + "step": 17488 + }, + { + "epoch": 0.41, + "grad_norm": 2.14987756172052, + "learning_rate": 1.3273760413662596e-05, + "loss": 0.9598, + "step": 17489 + }, + { + "epoch": 0.41, + "grad_norm": 1.9379923157999568, + "learning_rate": 1.3273039411109714e-05, + "loss": 0.9106, + "step": 17490 + }, + { + "epoch": 0.41, + "grad_norm": 1.8927980012050638, + "learning_rate": 1.3272318389500158e-05, + "loss": 1.0051, + "step": 17491 + }, + { + "epoch": 0.41, + "grad_norm": 2.0112281387410067, + "learning_rate": 1.3271597348838133e-05, + "loss": 1.109, + "step": 17492 + }, + { + "epoch": 0.41, + "grad_norm": 2.4007326528229544, + "learning_rate": 1.327087628912783e-05, + "loss": 1.1185, + "step": 17493 + }, + { + "epoch": 0.41, + "grad_norm": 1.8007019711035654, + "learning_rate": 1.327015521037345e-05, + "loss": 0.9805, + "step": 17494 + }, + { + "epoch": 0.41, + "grad_norm": 2.270245766643455, + "learning_rate": 1.326943411257919e-05, + "loss": 0.9638, + "step": 17495 + }, + { + "epoch": 0.41, + "grad_norm": 1.837128351418926, + "learning_rate": 1.3268712995749251e-05, + "loss": 1.0396, + "step": 17496 + }, + { + "epoch": 0.41, + "grad_norm": 1.902744601570614, + "learning_rate": 1.3267991859887827e-05, + "loss": 0.9688, + "step": 17497 + }, + { + "epoch": 0.41, + "grad_norm": 1.9903976317434273, + "learning_rate": 1.3267270704999122e-05, + "loss": 1.0575, + "step": 17498 + }, + { + "epoch": 0.41, + "grad_norm": 1.92224241176613, + "learning_rate": 1.3266549531087333e-05, + "loss": 1.02, + "step": 17499 + }, + { + "epoch": 0.41, + "grad_norm": 2.813008880937103, + "learning_rate": 1.3265828338156658e-05, + "loss": 1.043, + "step": 17500 + }, + { + "epoch": 0.41, + "grad_norm": 2.1681481651221266, + "learning_rate": 1.3265107126211296e-05, + "loss": 1.0622, + "step": 17501 + }, + { + "epoch": 0.41, + "grad_norm": 2.353941108599493, + "learning_rate": 1.326438589525545e-05, + "loss": 0.9878, + "step": 17502 + }, + { + "epoch": 0.41, + "grad_norm": 1.9931798331082065, + "learning_rate": 1.326366464529331e-05, + "loss": 0.9052, + "step": 17503 + }, + { + "epoch": 0.41, + "grad_norm": 1.1698135007756905, + "learning_rate": 1.3262943376329085e-05, + "loss": 0.9284, + "step": 17504 + }, + { + "epoch": 0.41, + "grad_norm": 2.1093493711715996, + "learning_rate": 1.326222208836697e-05, + "loss": 0.9553, + "step": 17505 + }, + { + "epoch": 0.41, + "grad_norm": 2.2921549365851677, + "learning_rate": 1.3261500781411165e-05, + "loss": 0.9926, + "step": 17506 + }, + { + "epoch": 0.41, + "grad_norm": 2.383377366846957, + "learning_rate": 1.3260779455465871e-05, + "loss": 0.9883, + "step": 17507 + }, + { + "epoch": 0.41, + "grad_norm": 1.9780618313353842, + "learning_rate": 1.3260058110535283e-05, + "loss": 0.9554, + "step": 17508 + }, + { + "epoch": 0.41, + "grad_norm": 1.176518433112591, + "learning_rate": 1.325933674662361e-05, + "loss": 0.9737, + "step": 17509 + }, + { + "epoch": 0.41, + "grad_norm": 1.9329772406486907, + "learning_rate": 1.3258615363735041e-05, + "loss": 0.9492, + "step": 17510 + }, + { + "epoch": 0.41, + "grad_norm": 2.5691147266449663, + "learning_rate": 1.3257893961873787e-05, + "loss": 1.0288, + "step": 17511 + }, + { + "epoch": 0.41, + "grad_norm": 2.506386762351466, + "learning_rate": 1.3257172541044039e-05, + "loss": 0.9072, + "step": 17512 + }, + { + "epoch": 0.41, + "grad_norm": 2.14060410683853, + "learning_rate": 1.3256451101250002e-05, + "loss": 0.8956, + "step": 17513 + }, + { + "epoch": 0.41, + "grad_norm": 2.1596316566967952, + "learning_rate": 1.3255729642495878e-05, + "loss": 1.1004, + "step": 17514 + }, + { + "epoch": 0.41, + "grad_norm": 1.6960812333188537, + "learning_rate": 1.3255008164785863e-05, + "loss": 1.0575, + "step": 17515 + }, + { + "epoch": 0.41, + "grad_norm": 1.9563984401620842, + "learning_rate": 1.325428666812416e-05, + "loss": 0.9728, + "step": 17516 + }, + { + "epoch": 0.41, + "grad_norm": 2.1971610082516686, + "learning_rate": 1.3253565152514971e-05, + "loss": 1.101, + "step": 17517 + }, + { + "epoch": 0.41, + "grad_norm": 1.9079385063209766, + "learning_rate": 1.3252843617962494e-05, + "loss": 1.1205, + "step": 17518 + }, + { + "epoch": 0.41, + "grad_norm": 1.9117548866514207, + "learning_rate": 1.3252122064470934e-05, + "loss": 1.1122, + "step": 17519 + }, + { + "epoch": 0.41, + "grad_norm": 2.0119497186280277, + "learning_rate": 1.3251400492044487e-05, + "loss": 0.912, + "step": 17520 + }, + { + "epoch": 0.41, + "grad_norm": 2.1613932358703565, + "learning_rate": 1.3250678900687356e-05, + "loss": 1.0435, + "step": 17521 + }, + { + "epoch": 0.41, + "grad_norm": 3.6353661786913154, + "learning_rate": 1.3249957290403748e-05, + "loss": 0.9878, + "step": 17522 + }, + { + "epoch": 0.41, + "grad_norm": 1.968815382193936, + "learning_rate": 1.3249235661197857e-05, + "loss": 1.102, + "step": 17523 + }, + { + "epoch": 0.41, + "grad_norm": 2.0790586528722055, + "learning_rate": 1.3248514013073886e-05, + "loss": 1.0457, + "step": 17524 + }, + { + "epoch": 0.41, + "grad_norm": 1.9303253880862647, + "learning_rate": 1.3247792346036039e-05, + "loss": 1.044, + "step": 17525 + }, + { + "epoch": 0.41, + "grad_norm": 2.247636387614271, + "learning_rate": 1.3247070660088519e-05, + "loss": 1.0254, + "step": 17526 + }, + { + "epoch": 0.41, + "grad_norm": 2.3004970859785465, + "learning_rate": 1.3246348955235521e-05, + "loss": 0.9928, + "step": 17527 + }, + { + "epoch": 0.41, + "grad_norm": 1.833724989870221, + "learning_rate": 1.3245627231481258e-05, + "loss": 0.9392, + "step": 17528 + }, + { + "epoch": 0.41, + "grad_norm": 1.7822070033816062, + "learning_rate": 1.324490548882992e-05, + "loss": 0.9068, + "step": 17529 + }, + { + "epoch": 0.41, + "grad_norm": 2.0036941217593855, + "learning_rate": 1.3244183727285717e-05, + "loss": 0.9441, + "step": 17530 + }, + { + "epoch": 0.41, + "grad_norm": 2.0585452737746626, + "learning_rate": 1.3243461946852849e-05, + "loss": 1.1026, + "step": 17531 + }, + { + "epoch": 0.41, + "grad_norm": 2.0090814198284566, + "learning_rate": 1.324274014753552e-05, + "loss": 1.0372, + "step": 17532 + }, + { + "epoch": 0.41, + "grad_norm": 2.310049368227718, + "learning_rate": 1.3242018329337928e-05, + "loss": 0.8298, + "step": 17533 + }, + { + "epoch": 0.41, + "grad_norm": 2.2493428569946494, + "learning_rate": 1.3241296492264282e-05, + "loss": 0.9915, + "step": 17534 + }, + { + "epoch": 0.41, + "grad_norm": 1.154036817769752, + "learning_rate": 1.324057463631878e-05, + "loss": 0.9514, + "step": 17535 + }, + { + "epoch": 0.41, + "grad_norm": 2.037798113964289, + "learning_rate": 1.3239852761505627e-05, + "loss": 1.0556, + "step": 17536 + }, + { + "epoch": 0.41, + "grad_norm": 1.9998104131446008, + "learning_rate": 1.3239130867829028e-05, + "loss": 1.0461, + "step": 17537 + }, + { + "epoch": 0.41, + "grad_norm": 2.0387101052731267, + "learning_rate": 1.3238408955293179e-05, + "loss": 1.0066, + "step": 17538 + }, + { + "epoch": 0.41, + "grad_norm": 2.116666693476046, + "learning_rate": 1.3237687023902292e-05, + "loss": 0.9583, + "step": 17539 + }, + { + "epoch": 0.41, + "grad_norm": 2.071368887121674, + "learning_rate": 1.3236965073660562e-05, + "loss": 1.0686, + "step": 17540 + }, + { + "epoch": 0.41, + "grad_norm": 1.8731391811150255, + "learning_rate": 1.3236243104572202e-05, + "loss": 1.0662, + "step": 17541 + }, + { + "epoch": 0.41, + "grad_norm": 2.0388477210496787, + "learning_rate": 1.3235521116641405e-05, + "loss": 0.9466, + "step": 17542 + }, + { + "epoch": 0.41, + "grad_norm": 2.1127772383548398, + "learning_rate": 1.3234799109872382e-05, + "loss": 1.0358, + "step": 17543 + }, + { + "epoch": 0.41, + "grad_norm": 2.2441415164238574, + "learning_rate": 1.3234077084269334e-05, + "loss": 1.072, + "step": 17544 + }, + { + "epoch": 0.41, + "grad_norm": 2.551566896630816, + "learning_rate": 1.3233355039836466e-05, + "loss": 0.9855, + "step": 17545 + }, + { + "epoch": 0.41, + "grad_norm": 2.1908338729127212, + "learning_rate": 1.323263297657798e-05, + "loss": 1.14, + "step": 17546 + }, + { + "epoch": 0.41, + "grad_norm": 2.0415283834399203, + "learning_rate": 1.3231910894498082e-05, + "loss": 0.8968, + "step": 17547 + }, + { + "epoch": 0.41, + "grad_norm": 2.009002987449797, + "learning_rate": 1.3231188793600976e-05, + "loss": 1.1365, + "step": 17548 + }, + { + "epoch": 0.41, + "grad_norm": 1.1417915624213697, + "learning_rate": 1.3230466673890865e-05, + "loss": 0.9633, + "step": 17549 + }, + { + "epoch": 0.41, + "grad_norm": 3.877936199665966, + "learning_rate": 1.3229744535371953e-05, + "loss": 0.9897, + "step": 17550 + }, + { + "epoch": 0.41, + "grad_norm": 2.243161572752664, + "learning_rate": 1.3229022378048448e-05, + "loss": 1.009, + "step": 17551 + }, + { + "epoch": 0.41, + "grad_norm": 2.040934999712394, + "learning_rate": 1.322830020192455e-05, + "loss": 1.1579, + "step": 17552 + }, + { + "epoch": 0.41, + "grad_norm": 2.021161550047081, + "learning_rate": 1.3227578007004471e-05, + "loss": 0.9693, + "step": 17553 + }, + { + "epoch": 0.41, + "grad_norm": 2.9478962712065377, + "learning_rate": 1.3226855793292406e-05, + "loss": 0.8936, + "step": 17554 + }, + { + "epoch": 0.41, + "grad_norm": 2.5055391914702687, + "learning_rate": 1.322613356079257e-05, + "loss": 1.0853, + "step": 17555 + }, + { + "epoch": 0.41, + "grad_norm": 2.0213301362521685, + "learning_rate": 1.3225411309509158e-05, + "loss": 0.9668, + "step": 17556 + }, + { + "epoch": 0.41, + "grad_norm": 1.7524772061379206, + "learning_rate": 1.322468903944638e-05, + "loss": 1.171, + "step": 17557 + }, + { + "epoch": 0.41, + "grad_norm": 1.940747416938226, + "learning_rate": 1.3223966750608447e-05, + "loss": 0.9565, + "step": 17558 + }, + { + "epoch": 0.41, + "grad_norm": 1.9992192962221076, + "learning_rate": 1.3223244442999554e-05, + "loss": 0.8871, + "step": 17559 + }, + { + "epoch": 0.41, + "grad_norm": 2.090384606201673, + "learning_rate": 1.3222522116623915e-05, + "loss": 0.9411, + "step": 17560 + }, + { + "epoch": 0.41, + "grad_norm": 1.8785460808699135, + "learning_rate": 1.3221799771485728e-05, + "loss": 0.9266, + "step": 17561 + }, + { + "epoch": 0.41, + "grad_norm": 2.194804597051915, + "learning_rate": 1.3221077407589207e-05, + "loss": 1.0054, + "step": 17562 + }, + { + "epoch": 0.41, + "grad_norm": 2.0573894996433357, + "learning_rate": 1.3220355024938549e-05, + "loss": 1.0533, + "step": 17563 + }, + { + "epoch": 0.41, + "grad_norm": 1.9309619328635292, + "learning_rate": 1.3219632623537968e-05, + "loss": 1.0305, + "step": 17564 + }, + { + "epoch": 0.41, + "grad_norm": 2.038536747342493, + "learning_rate": 1.3218910203391666e-05, + "loss": 1.0043, + "step": 17565 + }, + { + "epoch": 0.41, + "grad_norm": 1.9102236619074957, + "learning_rate": 1.321818776450385e-05, + "loss": 1.1067, + "step": 17566 + }, + { + "epoch": 0.41, + "grad_norm": 1.636272734816441, + "learning_rate": 1.3217465306878725e-05, + "loss": 0.9359, + "step": 17567 + }, + { + "epoch": 0.41, + "grad_norm": 1.102030945955592, + "learning_rate": 1.3216742830520497e-05, + "loss": 0.9447, + "step": 17568 + }, + { + "epoch": 0.41, + "grad_norm": 2.090440510169627, + "learning_rate": 1.3216020335433375e-05, + "loss": 0.8975, + "step": 17569 + }, + { + "epoch": 0.41, + "grad_norm": 2.2841021007605513, + "learning_rate": 1.3215297821621565e-05, + "loss": 1.031, + "step": 17570 + }, + { + "epoch": 0.41, + "grad_norm": 1.8305176162170247, + "learning_rate": 1.3214575289089272e-05, + "loss": 1.0773, + "step": 17571 + }, + { + "epoch": 0.41, + "grad_norm": 2.226302412385716, + "learning_rate": 1.3213852737840705e-05, + "loss": 1.1368, + "step": 17572 + }, + { + "epoch": 0.41, + "grad_norm": 2.6499936852230186, + "learning_rate": 1.3213130167880071e-05, + "loss": 1.0304, + "step": 17573 + }, + { + "epoch": 0.41, + "grad_norm": 1.1162573626601557, + "learning_rate": 1.3212407579211575e-05, + "loss": 0.9926, + "step": 17574 + }, + { + "epoch": 0.41, + "grad_norm": 1.9113025548480884, + "learning_rate": 1.3211684971839422e-05, + "loss": 1.0406, + "step": 17575 + }, + { + "epoch": 0.41, + "grad_norm": 2.190144440904416, + "learning_rate": 1.3210962345767826e-05, + "loss": 1.0298, + "step": 17576 + }, + { + "epoch": 0.41, + "grad_norm": 1.9496012262983475, + "learning_rate": 1.3210239701000989e-05, + "loss": 1.069, + "step": 17577 + }, + { + "epoch": 0.41, + "grad_norm": 1.845625251055635, + "learning_rate": 1.320951703754312e-05, + "loss": 1.045, + "step": 17578 + }, + { + "epoch": 0.41, + "grad_norm": 1.0575258994045043, + "learning_rate": 1.3208794355398431e-05, + "loss": 0.9226, + "step": 17579 + }, + { + "epoch": 0.41, + "grad_norm": 2.0573352302343135, + "learning_rate": 1.320807165457112e-05, + "loss": 1.1715, + "step": 17580 + }, + { + "epoch": 0.41, + "grad_norm": 2.089277416103221, + "learning_rate": 1.3207348935065403e-05, + "loss": 1.0296, + "step": 17581 + }, + { + "epoch": 0.41, + "grad_norm": 1.2067957433476209, + "learning_rate": 1.3206626196885481e-05, + "loss": 1.0259, + "step": 17582 + }, + { + "epoch": 0.41, + "grad_norm": 2.022464939727765, + "learning_rate": 1.320590344003557e-05, + "loss": 1.1094, + "step": 17583 + }, + { + "epoch": 0.41, + "grad_norm": 2.000324368715795, + "learning_rate": 1.3205180664519872e-05, + "loss": 1.0352, + "step": 17584 + }, + { + "epoch": 0.41, + "grad_norm": 1.0981079286706157, + "learning_rate": 1.3204457870342599e-05, + "loss": 0.9904, + "step": 17585 + }, + { + "epoch": 0.41, + "grad_norm": 2.6445806592357215, + "learning_rate": 1.320373505750796e-05, + "loss": 0.9614, + "step": 17586 + }, + { + "epoch": 0.41, + "grad_norm": 2.234519310897598, + "learning_rate": 1.3203012226020156e-05, + "loss": 1.0742, + "step": 17587 + }, + { + "epoch": 0.41, + "grad_norm": 2.207690510776234, + "learning_rate": 1.3202289375883405e-05, + "loss": 1.0687, + "step": 17588 + }, + { + "epoch": 0.41, + "grad_norm": 2.0815875250870564, + "learning_rate": 1.3201566507101908e-05, + "loss": 1.0256, + "step": 17589 + }, + { + "epoch": 0.41, + "grad_norm": 2.2092387611977, + "learning_rate": 1.3200843619679882e-05, + "loss": 1.1438, + "step": 17590 + }, + { + "epoch": 0.41, + "grad_norm": 1.0298505524236807, + "learning_rate": 1.3200120713621525e-05, + "loss": 0.9867, + "step": 17591 + }, + { + "epoch": 0.41, + "grad_norm": 2.158882824746762, + "learning_rate": 1.3199397788931057e-05, + "loss": 0.9252, + "step": 17592 + }, + { + "epoch": 0.41, + "grad_norm": 2.709045872715649, + "learning_rate": 1.319867484561268e-05, + "loss": 1.1573, + "step": 17593 + }, + { + "epoch": 0.41, + "grad_norm": 1.804741871389412, + "learning_rate": 1.3197951883670606e-05, + "loss": 1.0549, + "step": 17594 + }, + { + "epoch": 0.41, + "grad_norm": 2.1486550340196025, + "learning_rate": 1.3197228903109044e-05, + "loss": 1.0672, + "step": 17595 + }, + { + "epoch": 0.41, + "grad_norm": 1.803464245952549, + "learning_rate": 1.3196505903932203e-05, + "loss": 1.0144, + "step": 17596 + }, + { + "epoch": 0.41, + "grad_norm": 2.1087655100841394, + "learning_rate": 1.3195782886144291e-05, + "loss": 1.0278, + "step": 17597 + }, + { + "epoch": 0.41, + "grad_norm": 2.347754861086414, + "learning_rate": 1.3195059849749519e-05, + "loss": 1.1531, + "step": 17598 + }, + { + "epoch": 0.41, + "grad_norm": 1.1495397856720029, + "learning_rate": 1.3194336794752099e-05, + "loss": 0.9864, + "step": 17599 + }, + { + "epoch": 0.41, + "grad_norm": 2.440962511744752, + "learning_rate": 1.3193613721156237e-05, + "loss": 0.9299, + "step": 17600 + }, + { + "epoch": 0.41, + "grad_norm": 2.558775040158838, + "learning_rate": 1.3192890628966146e-05, + "loss": 1.1117, + "step": 17601 + }, + { + "epoch": 0.41, + "grad_norm": 2.00950615195398, + "learning_rate": 1.3192167518186036e-05, + "loss": 1.1039, + "step": 17602 + }, + { + "epoch": 0.41, + "grad_norm": 2.771502033040287, + "learning_rate": 1.3191444388820114e-05, + "loss": 0.9437, + "step": 17603 + }, + { + "epoch": 0.41, + "grad_norm": 2.5016435514283866, + "learning_rate": 1.3190721240872595e-05, + "loss": 1.1188, + "step": 17604 + }, + { + "epoch": 0.41, + "grad_norm": 1.9552970590013374, + "learning_rate": 1.3189998074347684e-05, + "loss": 0.9957, + "step": 17605 + }, + { + "epoch": 0.41, + "grad_norm": 2.153445219236233, + "learning_rate": 1.3189274889249598e-05, + "loss": 1.0175, + "step": 17606 + }, + { + "epoch": 0.41, + "grad_norm": 1.9348056022850748, + "learning_rate": 1.3188551685582542e-05, + "loss": 0.9684, + "step": 17607 + }, + { + "epoch": 0.41, + "grad_norm": 2.004239339837691, + "learning_rate": 1.3187828463350725e-05, + "loss": 0.9722, + "step": 17608 + }, + { + "epoch": 0.41, + "grad_norm": 1.102301669390569, + "learning_rate": 1.3187105222558368e-05, + "loss": 1.0076, + "step": 17609 + }, + { + "epoch": 0.41, + "grad_norm": 2.276997560562263, + "learning_rate": 1.3186381963209666e-05, + "loss": 1.0267, + "step": 17610 + }, + { + "epoch": 0.41, + "grad_norm": 1.936645370517056, + "learning_rate": 1.3185658685308848e-05, + "loss": 1.0197, + "step": 17611 + }, + { + "epoch": 0.41, + "grad_norm": 1.0694848009246836, + "learning_rate": 1.3184935388860111e-05, + "loss": 0.9963, + "step": 17612 + }, + { + "epoch": 0.41, + "grad_norm": 1.9166607143858856, + "learning_rate": 1.3184212073867678e-05, + "loss": 0.9446, + "step": 17613 + }, + { + "epoch": 0.41, + "grad_norm": 1.75780314734408, + "learning_rate": 1.3183488740335748e-05, + "loss": 1.0198, + "step": 17614 + }, + { + "epoch": 0.41, + "grad_norm": 2.019902760056152, + "learning_rate": 1.3182765388268544e-05, + "loss": 1.0671, + "step": 17615 + }, + { + "epoch": 0.42, + "grad_norm": 2.248631317493038, + "learning_rate": 1.318204201767027e-05, + "loss": 1.0963, + "step": 17616 + }, + { + "epoch": 0.42, + "grad_norm": 2.2712930072261726, + "learning_rate": 1.3181318628545138e-05, + "loss": 0.9896, + "step": 17617 + }, + { + "epoch": 0.42, + "grad_norm": 1.845080576713446, + "learning_rate": 1.3180595220897365e-05, + "loss": 1.0116, + "step": 17618 + }, + { + "epoch": 0.42, + "grad_norm": 2.8876410426967087, + "learning_rate": 1.3179871794731158e-05, + "loss": 1.0428, + "step": 17619 + }, + { + "epoch": 0.42, + "grad_norm": 2.195173543175582, + "learning_rate": 1.3179148350050733e-05, + "loss": 1.0842, + "step": 17620 + }, + { + "epoch": 0.42, + "grad_norm": 2.157239781145357, + "learning_rate": 1.3178424886860292e-05, + "loss": 0.9361, + "step": 17621 + }, + { + "epoch": 0.42, + "grad_norm": 1.8573625227423596, + "learning_rate": 1.3177701405164065e-05, + "loss": 0.924, + "step": 17622 + }, + { + "epoch": 0.42, + "grad_norm": 2.2499221471742588, + "learning_rate": 1.3176977904966248e-05, + "loss": 0.9061, + "step": 17623 + }, + { + "epoch": 0.42, + "grad_norm": 2.613730268363437, + "learning_rate": 1.3176254386271063e-05, + "loss": 1.1256, + "step": 17624 + }, + { + "epoch": 0.42, + "grad_norm": 2.068077975106259, + "learning_rate": 1.3175530849082718e-05, + "loss": 1.0465, + "step": 17625 + }, + { + "epoch": 0.42, + "grad_norm": 1.9945430550373453, + "learning_rate": 1.3174807293405427e-05, + "loss": 0.9926, + "step": 17626 + }, + { + "epoch": 0.42, + "grad_norm": 2.642258916107674, + "learning_rate": 1.3174083719243402e-05, + "loss": 1.0233, + "step": 17627 + }, + { + "epoch": 0.42, + "grad_norm": 1.0526670575293124, + "learning_rate": 1.3173360126600858e-05, + "loss": 0.9829, + "step": 17628 + }, + { + "epoch": 0.42, + "grad_norm": 1.9266530482502184, + "learning_rate": 1.3172636515482007e-05, + "loss": 1.0826, + "step": 17629 + }, + { + "epoch": 0.42, + "grad_norm": 2.0099163576984806, + "learning_rate": 1.3171912885891063e-05, + "loss": 1.0019, + "step": 17630 + }, + { + "epoch": 0.42, + "grad_norm": 2.310305961123699, + "learning_rate": 1.3171189237832234e-05, + "loss": 1.0684, + "step": 17631 + }, + { + "epoch": 0.42, + "grad_norm": 2.065658322095636, + "learning_rate": 1.3170465571309738e-05, + "loss": 1.1743, + "step": 17632 + }, + { + "epoch": 0.42, + "grad_norm": 1.8797824860095615, + "learning_rate": 1.316974188632779e-05, + "loss": 1.0089, + "step": 17633 + }, + { + "epoch": 0.42, + "grad_norm": 1.9560195227645973, + "learning_rate": 1.31690181828906e-05, + "loss": 1.0389, + "step": 17634 + }, + { + "epoch": 0.42, + "grad_norm": 1.961435590633547, + "learning_rate": 1.3168294461002385e-05, + "loss": 0.9751, + "step": 17635 + }, + { + "epoch": 0.42, + "grad_norm": 1.878205143662377, + "learning_rate": 1.3167570720667353e-05, + "loss": 0.8224, + "step": 17636 + }, + { + "epoch": 0.42, + "grad_norm": 1.9247923620304512, + "learning_rate": 1.3166846961889725e-05, + "loss": 1.0761, + "step": 17637 + }, + { + "epoch": 0.42, + "grad_norm": 2.06115835696419, + "learning_rate": 1.316612318467371e-05, + "loss": 1.0316, + "step": 17638 + }, + { + "epoch": 0.42, + "grad_norm": 2.3872285363236587, + "learning_rate": 1.3165399389023523e-05, + "loss": 1.0815, + "step": 17639 + }, + { + "epoch": 0.42, + "grad_norm": 2.0878389237705113, + "learning_rate": 1.3164675574943377e-05, + "loss": 1.0172, + "step": 17640 + }, + { + "epoch": 0.42, + "grad_norm": 1.120757890237753, + "learning_rate": 1.3163951742437493e-05, + "loss": 1.0084, + "step": 17641 + }, + { + "epoch": 0.42, + "grad_norm": 2.498556875123675, + "learning_rate": 1.3163227891510073e-05, + "loss": 0.9359, + "step": 17642 + }, + { + "epoch": 0.42, + "grad_norm": 2.094619731413015, + "learning_rate": 1.3162504022165345e-05, + "loss": 1.0594, + "step": 17643 + }, + { + "epoch": 0.42, + "grad_norm": 2.082446109778832, + "learning_rate": 1.3161780134407515e-05, + "loss": 1.1392, + "step": 17644 + }, + { + "epoch": 0.42, + "grad_norm": 1.961352571076004, + "learning_rate": 1.3161056228240798e-05, + "loss": 0.8921, + "step": 17645 + }, + { + "epoch": 0.42, + "grad_norm": 2.5416203531710826, + "learning_rate": 1.3160332303669414e-05, + "loss": 1.1104, + "step": 17646 + }, + { + "epoch": 0.42, + "grad_norm": 2.064752244324484, + "learning_rate": 1.3159608360697573e-05, + "loss": 1.0252, + "step": 17647 + }, + { + "epoch": 0.42, + "grad_norm": 2.0008944302682368, + "learning_rate": 1.3158884399329492e-05, + "loss": 1.0217, + "step": 17648 + }, + { + "epoch": 0.42, + "grad_norm": 2.0073601349030303, + "learning_rate": 1.3158160419569388e-05, + "loss": 0.942, + "step": 17649 + }, + { + "epoch": 0.42, + "grad_norm": 1.9713362250095323, + "learning_rate": 1.3157436421421472e-05, + "loss": 1.0263, + "step": 17650 + }, + { + "epoch": 0.42, + "grad_norm": 3.846765579330957, + "learning_rate": 1.3156712404889962e-05, + "loss": 1.071, + "step": 17651 + }, + { + "epoch": 0.42, + "grad_norm": 1.9103840184348073, + "learning_rate": 1.3155988369979071e-05, + "loss": 1.0017, + "step": 17652 + }, + { + "epoch": 0.42, + "grad_norm": 1.9426898385206024, + "learning_rate": 1.315526431669302e-05, + "loss": 1.0516, + "step": 17653 + }, + { + "epoch": 0.42, + "grad_norm": 1.8662461821952359, + "learning_rate": 1.315454024503602e-05, + "loss": 1.0663, + "step": 17654 + }, + { + "epoch": 0.42, + "grad_norm": 2.2031925105258585, + "learning_rate": 1.3153816155012287e-05, + "loss": 0.9277, + "step": 17655 + }, + { + "epoch": 0.42, + "grad_norm": 2.0231264577911685, + "learning_rate": 1.3153092046626039e-05, + "loss": 0.9103, + "step": 17656 + }, + { + "epoch": 0.42, + "grad_norm": 2.068049418654381, + "learning_rate": 1.315236791988149e-05, + "loss": 0.9949, + "step": 17657 + }, + { + "epoch": 0.42, + "grad_norm": 2.3377593111344774, + "learning_rate": 1.315164377478286e-05, + "loss": 1.035, + "step": 17658 + }, + { + "epoch": 0.42, + "grad_norm": 2.020055491332767, + "learning_rate": 1.3150919611334358e-05, + "loss": 1.0373, + "step": 17659 + }, + { + "epoch": 0.42, + "grad_norm": 2.0167090514310337, + "learning_rate": 1.3150195429540207e-05, + "loss": 0.9743, + "step": 17660 + }, + { + "epoch": 0.42, + "grad_norm": 1.9270716689845246, + "learning_rate": 1.3149471229404617e-05, + "loss": 1.0453, + "step": 17661 + }, + { + "epoch": 0.42, + "grad_norm": 2.087925281830283, + "learning_rate": 1.3148747010931814e-05, + "loss": 0.8513, + "step": 17662 + }, + { + "epoch": 0.42, + "grad_norm": 2.1137328329753644, + "learning_rate": 1.3148022774126003e-05, + "loss": 1.0188, + "step": 17663 + }, + { + "epoch": 0.42, + "grad_norm": 1.9803320733616718, + "learning_rate": 1.3147298518991413e-05, + "loss": 0.9653, + "step": 17664 + }, + { + "epoch": 0.42, + "grad_norm": 1.877641359341742, + "learning_rate": 1.3146574245532254e-05, + "loss": 1.055, + "step": 17665 + }, + { + "epoch": 0.42, + "grad_norm": 1.9132317440731843, + "learning_rate": 1.3145849953752739e-05, + "loss": 0.9385, + "step": 17666 + }, + { + "epoch": 0.42, + "grad_norm": 2.446986864363246, + "learning_rate": 1.3145125643657093e-05, + "loss": 1.0033, + "step": 17667 + }, + { + "epoch": 0.42, + "grad_norm": 1.9264654386276492, + "learning_rate": 1.3144401315249529e-05, + "loss": 1.0859, + "step": 17668 + }, + { + "epoch": 0.42, + "grad_norm": 1.0837862959300713, + "learning_rate": 1.3143676968534264e-05, + "loss": 0.9568, + "step": 17669 + }, + { + "epoch": 0.42, + "grad_norm": 1.1217828914524748, + "learning_rate": 1.3142952603515515e-05, + "loss": 0.9304, + "step": 17670 + }, + { + "epoch": 0.42, + "grad_norm": 2.124833052555704, + "learning_rate": 1.3142228220197503e-05, + "loss": 1.0172, + "step": 17671 + }, + { + "epoch": 0.42, + "grad_norm": 1.9271095935244804, + "learning_rate": 1.3141503818584444e-05, + "loss": 0.9441, + "step": 17672 + }, + { + "epoch": 0.42, + "grad_norm": 2.044378350615963, + "learning_rate": 1.3140779398680557e-05, + "loss": 1.0278, + "step": 17673 + }, + { + "epoch": 0.42, + "grad_norm": 2.074601179902934, + "learning_rate": 1.3140054960490053e-05, + "loss": 1.0882, + "step": 17674 + }, + { + "epoch": 0.42, + "grad_norm": 1.8781932550633265, + "learning_rate": 1.3139330504017159e-05, + "loss": 0.9541, + "step": 17675 + }, + { + "epoch": 0.42, + "grad_norm": 2.008071533310227, + "learning_rate": 1.3138606029266087e-05, + "loss": 1.0419, + "step": 17676 + }, + { + "epoch": 0.42, + "grad_norm": 1.1114911251502635, + "learning_rate": 1.3137881536241058e-05, + "loss": 0.8972, + "step": 17677 + }, + { + "epoch": 0.42, + "grad_norm": 2.2110762317766617, + "learning_rate": 1.3137157024946288e-05, + "loss": 1.0621, + "step": 17678 + }, + { + "epoch": 0.42, + "grad_norm": 1.9057681644095148, + "learning_rate": 1.3136432495385997e-05, + "loss": 0.9132, + "step": 17679 + }, + { + "epoch": 0.42, + "grad_norm": 2.086693643426408, + "learning_rate": 1.3135707947564402e-05, + "loss": 1.0477, + "step": 17680 + }, + { + "epoch": 0.42, + "grad_norm": 2.132184968908529, + "learning_rate": 1.3134983381485724e-05, + "loss": 1.0216, + "step": 17681 + }, + { + "epoch": 0.42, + "grad_norm": 1.915478386242394, + "learning_rate": 1.3134258797154183e-05, + "loss": 1.0747, + "step": 17682 + }, + { + "epoch": 0.42, + "grad_norm": 2.3200202324435817, + "learning_rate": 1.3133534194573992e-05, + "loss": 1.0294, + "step": 17683 + }, + { + "epoch": 0.42, + "grad_norm": 2.3428036313935685, + "learning_rate": 1.3132809573749371e-05, + "loss": 1.0612, + "step": 17684 + }, + { + "epoch": 0.42, + "grad_norm": 1.975098439936724, + "learning_rate": 1.3132084934684544e-05, + "loss": 0.92, + "step": 17685 + }, + { + "epoch": 0.42, + "grad_norm": 2.390026144133234, + "learning_rate": 1.3131360277383725e-05, + "loss": 1.0839, + "step": 17686 + }, + { + "epoch": 0.42, + "grad_norm": 2.052893214155361, + "learning_rate": 1.3130635601851136e-05, + "loss": 1.0519, + "step": 17687 + }, + { + "epoch": 0.42, + "grad_norm": 2.027796311205102, + "learning_rate": 1.3129910908090997e-05, + "loss": 1.065, + "step": 17688 + }, + { + "epoch": 0.42, + "grad_norm": 2.203639979108836, + "learning_rate": 1.3129186196107522e-05, + "loss": 0.8491, + "step": 17689 + }, + { + "epoch": 0.42, + "grad_norm": 1.9566072161769181, + "learning_rate": 1.3128461465904938e-05, + "loss": 1.0949, + "step": 17690 + }, + { + "epoch": 0.42, + "grad_norm": 1.795197270851616, + "learning_rate": 1.3127736717487459e-05, + "loss": 1.0666, + "step": 17691 + }, + { + "epoch": 0.42, + "grad_norm": 2.107363009536638, + "learning_rate": 1.3127011950859307e-05, + "loss": 0.9302, + "step": 17692 + }, + { + "epoch": 0.42, + "grad_norm": 1.1895685695817937, + "learning_rate": 1.31262871660247e-05, + "loss": 1.0071, + "step": 17693 + }, + { + "epoch": 0.42, + "grad_norm": 1.2291112386115604, + "learning_rate": 1.3125562362987863e-05, + "loss": 0.9471, + "step": 17694 + }, + { + "epoch": 0.42, + "grad_norm": 2.0964813990391478, + "learning_rate": 1.3124837541753008e-05, + "loss": 0.8896, + "step": 17695 + }, + { + "epoch": 0.42, + "grad_norm": 2.0366378423147995, + "learning_rate": 1.3124112702324363e-05, + "loss": 1.0437, + "step": 17696 + }, + { + "epoch": 0.42, + "grad_norm": 1.8389906587288347, + "learning_rate": 1.3123387844706144e-05, + "loss": 0.954, + "step": 17697 + }, + { + "epoch": 0.42, + "grad_norm": 1.1047082509640607, + "learning_rate": 1.312266296890257e-05, + "loss": 0.9892, + "step": 17698 + }, + { + "epoch": 0.42, + "grad_norm": 1.0616880594537714, + "learning_rate": 1.3121938074917866e-05, + "loss": 0.9798, + "step": 17699 + }, + { + "epoch": 0.42, + "grad_norm": 1.8757902973327052, + "learning_rate": 1.3121213162756247e-05, + "loss": 1.117, + "step": 17700 + }, + { + "epoch": 0.42, + "grad_norm": 2.3347532865247094, + "learning_rate": 1.3120488232421939e-05, + "loss": 1.0326, + "step": 17701 + }, + { + "epoch": 0.42, + "grad_norm": 1.2110353015044102, + "learning_rate": 1.311976328391916e-05, + "loss": 1.0349, + "step": 17702 + }, + { + "epoch": 0.42, + "grad_norm": 1.7238704466254644, + "learning_rate": 1.3119038317252133e-05, + "loss": 0.9925, + "step": 17703 + }, + { + "epoch": 0.42, + "grad_norm": 1.890555453954818, + "learning_rate": 1.3118313332425077e-05, + "loss": 0.9766, + "step": 17704 + }, + { + "epoch": 0.42, + "grad_norm": 2.3155757280000238, + "learning_rate": 1.3117588329442213e-05, + "loss": 0.9412, + "step": 17705 + }, + { + "epoch": 0.42, + "grad_norm": 1.957383882019313, + "learning_rate": 1.3116863308307762e-05, + "loss": 1.1711, + "step": 17706 + }, + { + "epoch": 0.42, + "grad_norm": 2.005473673056659, + "learning_rate": 1.3116138269025947e-05, + "loss": 1.0159, + "step": 17707 + }, + { + "epoch": 0.42, + "grad_norm": 1.8067562205475234, + "learning_rate": 1.3115413211600987e-05, + "loss": 1.0225, + "step": 17708 + }, + { + "epoch": 0.42, + "grad_norm": 1.1144414032932375, + "learning_rate": 1.3114688136037109e-05, + "loss": 1.0481, + "step": 17709 + }, + { + "epoch": 0.42, + "grad_norm": 1.862006429103637, + "learning_rate": 1.3113963042338527e-05, + "loss": 1.0621, + "step": 17710 + }, + { + "epoch": 0.42, + "grad_norm": 2.189864530245308, + "learning_rate": 1.3113237930509469e-05, + "loss": 0.989, + "step": 17711 + }, + { + "epoch": 0.42, + "grad_norm": 1.8237236033033373, + "learning_rate": 1.3112512800554149e-05, + "loss": 1.0187, + "step": 17712 + }, + { + "epoch": 0.42, + "grad_norm": 2.2557722248235854, + "learning_rate": 1.31117876524768e-05, + "loss": 1.0109, + "step": 17713 + }, + { + "epoch": 0.42, + "grad_norm": 1.9125670050043815, + "learning_rate": 1.3111062486281632e-05, + "loss": 1.025, + "step": 17714 + }, + { + "epoch": 0.42, + "grad_norm": 1.8813558263227539, + "learning_rate": 1.3110337301972877e-05, + "loss": 0.9596, + "step": 17715 + }, + { + "epoch": 0.42, + "grad_norm": 1.8120025939321036, + "learning_rate": 1.3109612099554754e-05, + "loss": 1.0748, + "step": 17716 + }, + { + "epoch": 0.42, + "grad_norm": 1.8895099643370155, + "learning_rate": 1.3108886879031485e-05, + "loss": 1.0805, + "step": 17717 + }, + { + "epoch": 0.42, + "grad_norm": 1.0969130996079262, + "learning_rate": 1.310816164040729e-05, + "loss": 0.9954, + "step": 17718 + }, + { + "epoch": 0.42, + "grad_norm": 2.2985715352195846, + "learning_rate": 1.3107436383686394e-05, + "loss": 1.0485, + "step": 17719 + }, + { + "epoch": 0.42, + "grad_norm": 2.0565758208809055, + "learning_rate": 1.3106711108873022e-05, + "loss": 1.1538, + "step": 17720 + }, + { + "epoch": 0.42, + "grad_norm": 1.9272239095195367, + "learning_rate": 1.3105985815971392e-05, + "loss": 0.9504, + "step": 17721 + }, + { + "epoch": 0.42, + "grad_norm": 2.05331517338824, + "learning_rate": 1.3105260504985732e-05, + "loss": 0.9984, + "step": 17722 + }, + { + "epoch": 0.42, + "grad_norm": 1.1192022988167325, + "learning_rate": 1.3104535175920261e-05, + "loss": 0.972, + "step": 17723 + }, + { + "epoch": 0.42, + "grad_norm": 2.1930895301376716, + "learning_rate": 1.3103809828779202e-05, + "loss": 1.0574, + "step": 17724 + }, + { + "epoch": 0.42, + "grad_norm": 1.8940440719657334, + "learning_rate": 1.310308446356678e-05, + "loss": 1.001, + "step": 17725 + }, + { + "epoch": 0.42, + "grad_norm": 1.8339114754201915, + "learning_rate": 1.3102359080287219e-05, + "loss": 0.9126, + "step": 17726 + }, + { + "epoch": 0.42, + "grad_norm": 2.1860005822471384, + "learning_rate": 1.3101633678944741e-05, + "loss": 1.079, + "step": 17727 + }, + { + "epoch": 0.42, + "grad_norm": 2.311165339403738, + "learning_rate": 1.3100908259543567e-05, + "loss": 1.1566, + "step": 17728 + }, + { + "epoch": 0.42, + "grad_norm": 1.9939074767636353, + "learning_rate": 1.3100182822087928e-05, + "loss": 0.9276, + "step": 17729 + }, + { + "epoch": 0.42, + "grad_norm": 2.0233391379784096, + "learning_rate": 1.309945736658204e-05, + "loss": 1.0093, + "step": 17730 + }, + { + "epoch": 0.42, + "grad_norm": 1.0942521015674986, + "learning_rate": 1.309873189303013e-05, + "loss": 0.9396, + "step": 17731 + }, + { + "epoch": 0.42, + "grad_norm": 1.998448835676127, + "learning_rate": 1.3098006401436422e-05, + "loss": 1.104, + "step": 17732 + }, + { + "epoch": 0.42, + "grad_norm": 2.4358604111946516, + "learning_rate": 1.3097280891805141e-05, + "loss": 0.9855, + "step": 17733 + }, + { + "epoch": 0.42, + "grad_norm": 2.1026926279131466, + "learning_rate": 1.309655536414051e-05, + "loss": 0.8595, + "step": 17734 + }, + { + "epoch": 0.42, + "grad_norm": 2.212898502689356, + "learning_rate": 1.3095829818446753e-05, + "loss": 1.0516, + "step": 17735 + }, + { + "epoch": 0.42, + "grad_norm": 2.106819020865839, + "learning_rate": 1.3095104254728095e-05, + "loss": 1.0298, + "step": 17736 + }, + { + "epoch": 0.42, + "grad_norm": 2.0009299859750795, + "learning_rate": 1.3094378672988759e-05, + "loss": 1.0058, + "step": 17737 + }, + { + "epoch": 0.42, + "grad_norm": 2.688646422447479, + "learning_rate": 1.3093653073232972e-05, + "loss": 0.9787, + "step": 17738 + }, + { + "epoch": 0.42, + "grad_norm": 1.9649527015444805, + "learning_rate": 1.3092927455464959e-05, + "loss": 1.0333, + "step": 17739 + }, + { + "epoch": 0.42, + "grad_norm": 2.0448596114500597, + "learning_rate": 1.3092201819688937e-05, + "loss": 1.1196, + "step": 17740 + }, + { + "epoch": 0.42, + "grad_norm": 2.0166608182869754, + "learning_rate": 1.3091476165909142e-05, + "loss": 0.9391, + "step": 17741 + }, + { + "epoch": 0.42, + "grad_norm": 1.8508038913337042, + "learning_rate": 1.3090750494129792e-05, + "loss": 0.9765, + "step": 17742 + }, + { + "epoch": 0.42, + "grad_norm": 2.8980398253117174, + "learning_rate": 1.3090024804355117e-05, + "loss": 1.0128, + "step": 17743 + }, + { + "epoch": 0.42, + "grad_norm": 2.0214769339654866, + "learning_rate": 1.3089299096589336e-05, + "loss": 1.1113, + "step": 17744 + }, + { + "epoch": 0.42, + "grad_norm": 1.6950149680191209, + "learning_rate": 1.308857337083668e-05, + "loss": 1.1083, + "step": 17745 + }, + { + "epoch": 0.42, + "grad_norm": 2.066762758682511, + "learning_rate": 1.3087847627101372e-05, + "loss": 0.9911, + "step": 17746 + }, + { + "epoch": 0.42, + "grad_norm": 1.9525097395276831, + "learning_rate": 1.3087121865387636e-05, + "loss": 0.9181, + "step": 17747 + }, + { + "epoch": 0.42, + "grad_norm": 2.3446948783335286, + "learning_rate": 1.30863960856997e-05, + "loss": 1.0175, + "step": 17748 + }, + { + "epoch": 0.42, + "grad_norm": 2.2819914165661155, + "learning_rate": 1.3085670288041787e-05, + "loss": 1.0846, + "step": 17749 + }, + { + "epoch": 0.42, + "grad_norm": 2.6533550486398654, + "learning_rate": 1.3084944472418126e-05, + "loss": 1.0738, + "step": 17750 + }, + { + "epoch": 0.42, + "grad_norm": 2.608733384100918, + "learning_rate": 1.3084218638832943e-05, + "loss": 1.082, + "step": 17751 + }, + { + "epoch": 0.42, + "grad_norm": 1.8324071864710607, + "learning_rate": 1.3083492787290462e-05, + "loss": 1.08, + "step": 17752 + }, + { + "epoch": 0.42, + "grad_norm": 1.9325656586428195, + "learning_rate": 1.308276691779491e-05, + "loss": 1.0212, + "step": 17753 + }, + { + "epoch": 0.42, + "grad_norm": 2.143973934736292, + "learning_rate": 1.3082041030350512e-05, + "loss": 1.0885, + "step": 17754 + }, + { + "epoch": 0.42, + "grad_norm": 1.9906041776961911, + "learning_rate": 1.3081315124961496e-05, + "loss": 0.9703, + "step": 17755 + }, + { + "epoch": 0.42, + "grad_norm": 1.0964413443968244, + "learning_rate": 1.3080589201632086e-05, + "loss": 1.0162, + "step": 17756 + }, + { + "epoch": 0.42, + "grad_norm": 2.246424492243667, + "learning_rate": 1.3079863260366512e-05, + "loss": 1.0805, + "step": 17757 + }, + { + "epoch": 0.42, + "grad_norm": 2.270454160199967, + "learning_rate": 1.3079137301168998e-05, + "loss": 0.9132, + "step": 17758 + }, + { + "epoch": 0.42, + "grad_norm": 2.0850099156945294, + "learning_rate": 1.3078411324043772e-05, + "loss": 1.0407, + "step": 17759 + }, + { + "epoch": 0.42, + "grad_norm": 1.9768615356743806, + "learning_rate": 1.3077685328995065e-05, + "loss": 0.9692, + "step": 17760 + }, + { + "epoch": 0.42, + "grad_norm": 1.9515589443721097, + "learning_rate": 1.3076959316027093e-05, + "loss": 1.0172, + "step": 17761 + }, + { + "epoch": 0.42, + "grad_norm": 2.0573720546098495, + "learning_rate": 1.3076233285144096e-05, + "loss": 1.0579, + "step": 17762 + }, + { + "epoch": 0.42, + "grad_norm": 2.0022565605901392, + "learning_rate": 1.307550723635029e-05, + "loss": 1.0784, + "step": 17763 + }, + { + "epoch": 0.42, + "grad_norm": 1.9885835031995067, + "learning_rate": 1.3074781169649908e-05, + "loss": 0.9522, + "step": 17764 + }, + { + "epoch": 0.42, + "grad_norm": 1.8757282800962616, + "learning_rate": 1.3074055085047174e-05, + "loss": 1.0804, + "step": 17765 + }, + { + "epoch": 0.42, + "grad_norm": 2.2821464170495602, + "learning_rate": 1.3073328982546325e-05, + "loss": 0.9764, + "step": 17766 + }, + { + "epoch": 0.42, + "grad_norm": 1.9808177545258232, + "learning_rate": 1.3072602862151577e-05, + "loss": 0.9407, + "step": 17767 + }, + { + "epoch": 0.42, + "grad_norm": 2.0339568373623886, + "learning_rate": 1.307187672386716e-05, + "loss": 1.0357, + "step": 17768 + }, + { + "epoch": 0.42, + "grad_norm": 2.235241531961296, + "learning_rate": 1.3071150567697313e-05, + "loss": 0.9247, + "step": 17769 + }, + { + "epoch": 0.42, + "grad_norm": 1.767063213926696, + "learning_rate": 1.3070424393646246e-05, + "loss": 1.0671, + "step": 17770 + }, + { + "epoch": 0.42, + "grad_norm": 2.146475500018175, + "learning_rate": 1.3069698201718202e-05, + "loss": 1.0826, + "step": 17771 + }, + { + "epoch": 0.42, + "grad_norm": 1.902980023659691, + "learning_rate": 1.3068971991917398e-05, + "loss": 0.8717, + "step": 17772 + }, + { + "epoch": 0.42, + "grad_norm": 1.9593371496654266, + "learning_rate": 1.3068245764248073e-05, + "loss": 0.9329, + "step": 17773 + }, + { + "epoch": 0.42, + "grad_norm": 2.0610797308354014, + "learning_rate": 1.3067519518714447e-05, + "loss": 1.0687, + "step": 17774 + }, + { + "epoch": 0.42, + "grad_norm": 1.862184593802284, + "learning_rate": 1.306679325532075e-05, + "loss": 1.1395, + "step": 17775 + }, + { + "epoch": 0.42, + "grad_norm": 2.0498633564862545, + "learning_rate": 1.3066066974071213e-05, + "loss": 1.0563, + "step": 17776 + }, + { + "epoch": 0.42, + "grad_norm": 2.01085851202219, + "learning_rate": 1.3065340674970064e-05, + "loss": 0.9101, + "step": 17777 + }, + { + "epoch": 0.42, + "grad_norm": 2.2551362185211983, + "learning_rate": 1.3064614358021532e-05, + "loss": 0.9451, + "step": 17778 + }, + { + "epoch": 0.42, + "grad_norm": 1.8137316213087376, + "learning_rate": 1.3063888023229842e-05, + "loss": 0.8782, + "step": 17779 + }, + { + "epoch": 0.42, + "grad_norm": 2.3931117687856776, + "learning_rate": 1.3063161670599227e-05, + "loss": 1.168, + "step": 17780 + }, + { + "epoch": 0.42, + "grad_norm": 2.1518746594394953, + "learning_rate": 1.3062435300133917e-05, + "loss": 1.1842, + "step": 17781 + }, + { + "epoch": 0.42, + "grad_norm": 1.0981116711867935, + "learning_rate": 1.3061708911838138e-05, + "loss": 0.9452, + "step": 17782 + }, + { + "epoch": 0.42, + "grad_norm": 2.2531687304205255, + "learning_rate": 1.306098250571612e-05, + "loss": 0.9397, + "step": 17783 + }, + { + "epoch": 0.42, + "grad_norm": 1.9459380973695852, + "learning_rate": 1.3060256081772092e-05, + "loss": 0.9948, + "step": 17784 + }, + { + "epoch": 0.42, + "grad_norm": 1.2238699391375425, + "learning_rate": 1.3059529640010284e-05, + "loss": 0.9846, + "step": 17785 + }, + { + "epoch": 0.42, + "grad_norm": 2.2341285668453996, + "learning_rate": 1.3058803180434927e-05, + "loss": 0.9645, + "step": 17786 + }, + { + "epoch": 0.42, + "grad_norm": 2.099218262018247, + "learning_rate": 1.3058076703050249e-05, + "loss": 0.9695, + "step": 17787 + }, + { + "epoch": 0.42, + "grad_norm": 1.872393863076861, + "learning_rate": 1.3057350207860484e-05, + "loss": 1.1177, + "step": 17788 + }, + { + "epoch": 0.42, + "grad_norm": 1.8892226129975789, + "learning_rate": 1.305662369486985e-05, + "loss": 1.0365, + "step": 17789 + }, + { + "epoch": 0.42, + "grad_norm": 2.140731719238732, + "learning_rate": 1.3055897164082594e-05, + "loss": 0.8987, + "step": 17790 + }, + { + "epoch": 0.42, + "grad_norm": 1.9359592077696544, + "learning_rate": 1.305517061550293e-05, + "loss": 1.1339, + "step": 17791 + }, + { + "epoch": 0.42, + "grad_norm": 1.8807052252536727, + "learning_rate": 1.3054444049135099e-05, + "loss": 1.0777, + "step": 17792 + }, + { + "epoch": 0.42, + "grad_norm": 1.11431844000572, + "learning_rate": 1.3053717464983327e-05, + "loss": 1.0184, + "step": 17793 + }, + { + "epoch": 0.42, + "grad_norm": 2.0505893301284, + "learning_rate": 1.3052990863051846e-05, + "loss": 1.0766, + "step": 17794 + }, + { + "epoch": 0.42, + "grad_norm": 1.799045973710024, + "learning_rate": 1.3052264243344885e-05, + "loss": 1.0422, + "step": 17795 + }, + { + "epoch": 0.42, + "grad_norm": 2.143251036293501, + "learning_rate": 1.3051537605866675e-05, + "loss": 0.9998, + "step": 17796 + }, + { + "epoch": 0.42, + "grad_norm": 1.8572803031205944, + "learning_rate": 1.305081095062145e-05, + "loss": 1.0475, + "step": 17797 + }, + { + "epoch": 0.42, + "grad_norm": 2.1139019057579005, + "learning_rate": 1.3050084277613433e-05, + "loss": 1.0714, + "step": 17798 + }, + { + "epoch": 0.42, + "grad_norm": 2.229131410843958, + "learning_rate": 1.3049357586846862e-05, + "loss": 1.0475, + "step": 17799 + }, + { + "epoch": 0.42, + "grad_norm": 1.0195305622626503, + "learning_rate": 1.3048630878325966e-05, + "loss": 0.9624, + "step": 17800 + }, + { + "epoch": 0.42, + "grad_norm": 2.142700443497085, + "learning_rate": 1.3047904152054974e-05, + "loss": 0.975, + "step": 17801 + }, + { + "epoch": 0.42, + "grad_norm": 2.557025737076145, + "learning_rate": 1.3047177408038119e-05, + "loss": 1.0425, + "step": 17802 + }, + { + "epoch": 0.42, + "grad_norm": 3.0148129292429084, + "learning_rate": 1.3046450646279633e-05, + "loss": 0.9438, + "step": 17803 + }, + { + "epoch": 0.42, + "grad_norm": 2.084858194428346, + "learning_rate": 1.3045723866783749e-05, + "loss": 0.8859, + "step": 17804 + }, + { + "epoch": 0.42, + "grad_norm": 2.4871671394849484, + "learning_rate": 1.3044997069554695e-05, + "loss": 0.989, + "step": 17805 + }, + { + "epoch": 0.42, + "grad_norm": 1.850076888358732, + "learning_rate": 1.3044270254596706e-05, + "loss": 0.9384, + "step": 17806 + }, + { + "epoch": 0.42, + "grad_norm": 1.1396496590115, + "learning_rate": 1.3043543421914007e-05, + "loss": 0.9514, + "step": 17807 + }, + { + "epoch": 0.42, + "grad_norm": 1.983280347293108, + "learning_rate": 1.3042816571510836e-05, + "loss": 1.0506, + "step": 17808 + }, + { + "epoch": 0.42, + "grad_norm": 2.058795358211046, + "learning_rate": 1.3042089703391425e-05, + "loss": 1.1031, + "step": 17809 + }, + { + "epoch": 0.42, + "grad_norm": 1.8714631584901062, + "learning_rate": 1.3041362817560007e-05, + "loss": 0.9196, + "step": 17810 + }, + { + "epoch": 0.42, + "grad_norm": 2.027596434805435, + "learning_rate": 1.304063591402081e-05, + "loss": 0.9713, + "step": 17811 + }, + { + "epoch": 0.42, + "grad_norm": 1.854505159358798, + "learning_rate": 1.3039908992778065e-05, + "loss": 1.0462, + "step": 17812 + }, + { + "epoch": 0.42, + "grad_norm": 1.874963158628003, + "learning_rate": 1.3039182053836013e-05, + "loss": 0.9059, + "step": 17813 + }, + { + "epoch": 0.42, + "grad_norm": 1.9938750564229046, + "learning_rate": 1.3038455097198875e-05, + "loss": 1.0602, + "step": 17814 + }, + { + "epoch": 0.42, + "grad_norm": 1.9629963000637476, + "learning_rate": 1.3037728122870891e-05, + "loss": 0.9371, + "step": 17815 + }, + { + "epoch": 0.42, + "grad_norm": 2.2841913584622615, + "learning_rate": 1.3037001130856296e-05, + "loss": 1.0441, + "step": 17816 + }, + { + "epoch": 0.42, + "grad_norm": 2.2539628318654414, + "learning_rate": 1.3036274121159313e-05, + "loss": 1.0483, + "step": 17817 + }, + { + "epoch": 0.42, + "grad_norm": 1.1278661990772523, + "learning_rate": 1.3035547093784187e-05, + "loss": 0.9703, + "step": 17818 + }, + { + "epoch": 0.42, + "grad_norm": 2.0812459592173203, + "learning_rate": 1.3034820048735139e-05, + "loss": 0.8916, + "step": 17819 + }, + { + "epoch": 0.42, + "grad_norm": 1.934203674410252, + "learning_rate": 1.3034092986016413e-05, + "loss": 0.9981, + "step": 17820 + }, + { + "epoch": 0.42, + "grad_norm": 2.12575947354507, + "learning_rate": 1.3033365905632234e-05, + "loss": 1.0948, + "step": 17821 + }, + { + "epoch": 0.42, + "grad_norm": 2.021013642179841, + "learning_rate": 1.3032638807586842e-05, + "loss": 1.0145, + "step": 17822 + }, + { + "epoch": 0.42, + "grad_norm": 1.9469797907822601, + "learning_rate": 1.303191169188446e-05, + "loss": 0.9913, + "step": 17823 + }, + { + "epoch": 0.42, + "grad_norm": 2.1406945665177695, + "learning_rate": 1.3031184558529335e-05, + "loss": 0.9803, + "step": 17824 + }, + { + "epoch": 0.42, + "grad_norm": 2.025906981983683, + "learning_rate": 1.303045740752569e-05, + "loss": 1.0075, + "step": 17825 + }, + { + "epoch": 0.42, + "grad_norm": 2.1880008771210733, + "learning_rate": 1.3029730238877764e-05, + "loss": 1.0535, + "step": 17826 + }, + { + "epoch": 0.42, + "grad_norm": 2.116725598625171, + "learning_rate": 1.3029003052589789e-05, + "loss": 1.0895, + "step": 17827 + }, + { + "epoch": 0.42, + "grad_norm": 1.8342051503919614, + "learning_rate": 1.3028275848665998e-05, + "loss": 1.0988, + "step": 17828 + }, + { + "epoch": 0.42, + "grad_norm": 2.253991359280601, + "learning_rate": 1.3027548627110629e-05, + "loss": 1.1331, + "step": 17829 + }, + { + "epoch": 0.42, + "grad_norm": 1.939254701131445, + "learning_rate": 1.3026821387927911e-05, + "loss": 1.1066, + "step": 17830 + }, + { + "epoch": 0.42, + "grad_norm": 1.9859122399583373, + "learning_rate": 1.302609413112208e-05, + "loss": 0.9671, + "step": 17831 + }, + { + "epoch": 0.42, + "grad_norm": 2.0179140272005256, + "learning_rate": 1.3025366856697372e-05, + "loss": 1.0052, + "step": 17832 + }, + { + "epoch": 0.42, + "grad_norm": 2.0920809404648204, + "learning_rate": 1.3024639564658022e-05, + "loss": 1.1103, + "step": 17833 + }, + { + "epoch": 0.42, + "grad_norm": 2.2742652443091, + "learning_rate": 1.3023912255008261e-05, + "loss": 1.065, + "step": 17834 + }, + { + "epoch": 0.42, + "grad_norm": 2.153828515094928, + "learning_rate": 1.3023184927752326e-05, + "loss": 1.1544, + "step": 17835 + }, + { + "epoch": 0.42, + "grad_norm": 2.332720570088729, + "learning_rate": 1.3022457582894452e-05, + "loss": 1.0585, + "step": 17836 + }, + { + "epoch": 0.42, + "grad_norm": 1.8727730212976883, + "learning_rate": 1.302173022043887e-05, + "loss": 1.0212, + "step": 17837 + }, + { + "epoch": 0.42, + "grad_norm": 2.053428313414346, + "learning_rate": 1.302100284038982e-05, + "loss": 0.953, + "step": 17838 + }, + { + "epoch": 0.42, + "grad_norm": 2.130659900549844, + "learning_rate": 1.3020275442751538e-05, + "loss": 0.9814, + "step": 17839 + }, + { + "epoch": 0.42, + "grad_norm": 1.7580188879735994, + "learning_rate": 1.3019548027528253e-05, + "loss": 1.0053, + "step": 17840 + }, + { + "epoch": 0.42, + "grad_norm": 1.9618361397915292, + "learning_rate": 1.3018820594724204e-05, + "loss": 1.0002, + "step": 17841 + }, + { + "epoch": 0.42, + "grad_norm": 1.8476931827102074, + "learning_rate": 1.3018093144343624e-05, + "loss": 0.9562, + "step": 17842 + }, + { + "epoch": 0.42, + "grad_norm": 2.005656651970668, + "learning_rate": 1.3017365676390755e-05, + "loss": 0.8826, + "step": 17843 + }, + { + "epoch": 0.42, + "grad_norm": 1.0897194615671928, + "learning_rate": 1.3016638190869822e-05, + "loss": 0.9339, + "step": 17844 + }, + { + "epoch": 0.42, + "grad_norm": 1.862468079136502, + "learning_rate": 1.3015910687785072e-05, + "loss": 1.0236, + "step": 17845 + }, + { + "epoch": 0.42, + "grad_norm": 1.9870053670448597, + "learning_rate": 1.3015183167140731e-05, + "loss": 1.0089, + "step": 17846 + }, + { + "epoch": 0.42, + "grad_norm": 1.9941400737883117, + "learning_rate": 1.3014455628941041e-05, + "loss": 1.0748, + "step": 17847 + }, + { + "epoch": 0.42, + "grad_norm": 2.101654907159475, + "learning_rate": 1.3013728073190235e-05, + "loss": 1.0973, + "step": 17848 + }, + { + "epoch": 0.42, + "grad_norm": 1.0097925790826874, + "learning_rate": 1.3013000499892549e-05, + "loss": 0.91, + "step": 17849 + }, + { + "epoch": 0.42, + "grad_norm": 1.9638115269395482, + "learning_rate": 1.3012272909052221e-05, + "loss": 1.0631, + "step": 17850 + }, + { + "epoch": 0.42, + "grad_norm": 2.2602451754467516, + "learning_rate": 1.3011545300673486e-05, + "loss": 0.9311, + "step": 17851 + }, + { + "epoch": 0.42, + "grad_norm": 2.1011474836183335, + "learning_rate": 1.3010817674760583e-05, + "loss": 1.0007, + "step": 17852 + }, + { + "epoch": 0.42, + "grad_norm": 2.0464516114354407, + "learning_rate": 1.3010090031317746e-05, + "loss": 1.1136, + "step": 17853 + }, + { + "epoch": 0.42, + "grad_norm": 2.3055088892967945, + "learning_rate": 1.300936237034921e-05, + "loss": 1.1064, + "step": 17854 + }, + { + "epoch": 0.42, + "grad_norm": 2.2874547329512023, + "learning_rate": 1.3008634691859214e-05, + "loss": 0.8971, + "step": 17855 + }, + { + "epoch": 0.42, + "grad_norm": 1.9259589603634413, + "learning_rate": 1.3007906995851994e-05, + "loss": 1.0097, + "step": 17856 + }, + { + "epoch": 0.42, + "grad_norm": 2.0741429160151177, + "learning_rate": 1.3007179282331788e-05, + "loss": 0.9756, + "step": 17857 + }, + { + "epoch": 0.42, + "grad_norm": 2.254122596729678, + "learning_rate": 1.3006451551302833e-05, + "loss": 0.9305, + "step": 17858 + }, + { + "epoch": 0.42, + "grad_norm": 2.003731003171015, + "learning_rate": 1.3005723802769363e-05, + "loss": 0.9979, + "step": 17859 + }, + { + "epoch": 0.42, + "grad_norm": 1.85496442729624, + "learning_rate": 1.300499603673562e-05, + "loss": 1.0874, + "step": 17860 + }, + { + "epoch": 0.42, + "grad_norm": 2.368713602525407, + "learning_rate": 1.3004268253205837e-05, + "loss": 0.9723, + "step": 17861 + }, + { + "epoch": 0.42, + "grad_norm": 2.208775025978456, + "learning_rate": 1.3003540452184254e-05, + "loss": 0.9818, + "step": 17862 + }, + { + "epoch": 0.42, + "grad_norm": 2.3067831464564295, + "learning_rate": 1.3002812633675105e-05, + "loss": 1.1436, + "step": 17863 + }, + { + "epoch": 0.42, + "grad_norm": 2.052238257574931, + "learning_rate": 1.3002084797682633e-05, + "loss": 1.095, + "step": 17864 + }, + { + "epoch": 0.42, + "grad_norm": 2.441513565279804, + "learning_rate": 1.300135694421107e-05, + "loss": 0.8965, + "step": 17865 + }, + { + "epoch": 0.42, + "grad_norm": 2.009440751343168, + "learning_rate": 1.3000629073264658e-05, + "loss": 1.1449, + "step": 17866 + }, + { + "epoch": 0.42, + "grad_norm": 2.009416502378252, + "learning_rate": 1.2999901184847636e-05, + "loss": 1.1094, + "step": 17867 + }, + { + "epoch": 0.42, + "grad_norm": 2.242814925293912, + "learning_rate": 1.2999173278964237e-05, + "loss": 0.9632, + "step": 17868 + }, + { + "epoch": 0.42, + "grad_norm": 2.7763811328972445, + "learning_rate": 1.2998445355618704e-05, + "loss": 0.9866, + "step": 17869 + }, + { + "epoch": 0.42, + "grad_norm": 1.9555609450579525, + "learning_rate": 1.2997717414815269e-05, + "loss": 1.1409, + "step": 17870 + }, + { + "epoch": 0.42, + "grad_norm": 2.979435072769982, + "learning_rate": 1.299698945655818e-05, + "loss": 0.9159, + "step": 17871 + }, + { + "epoch": 0.42, + "grad_norm": 2.1893695850241373, + "learning_rate": 1.2996261480851664e-05, + "loss": 1.0341, + "step": 17872 + }, + { + "epoch": 0.42, + "grad_norm": 1.9434891974274966, + "learning_rate": 1.299553348769997e-05, + "loss": 0.9737, + "step": 17873 + }, + { + "epoch": 0.42, + "grad_norm": 1.8882101938501348, + "learning_rate": 1.2994805477107323e-05, + "loss": 0.9263, + "step": 17874 + }, + { + "epoch": 0.42, + "grad_norm": 2.071795606768666, + "learning_rate": 1.299407744907798e-05, + "loss": 1.0877, + "step": 17875 + }, + { + "epoch": 0.42, + "grad_norm": 1.0604258813168626, + "learning_rate": 1.2993349403616167e-05, + "loss": 0.9569, + "step": 17876 + }, + { + "epoch": 0.42, + "grad_norm": 2.043205758509778, + "learning_rate": 1.2992621340726124e-05, + "loss": 0.9878, + "step": 17877 + }, + { + "epoch": 0.42, + "grad_norm": 1.113394109752472, + "learning_rate": 1.2991893260412093e-05, + "loss": 0.9845, + "step": 17878 + }, + { + "epoch": 0.42, + "grad_norm": 1.8845858704378131, + "learning_rate": 1.2991165162678315e-05, + "loss": 0.9965, + "step": 17879 + }, + { + "epoch": 0.42, + "grad_norm": 1.0669144063359712, + "learning_rate": 1.2990437047529022e-05, + "loss": 0.9484, + "step": 17880 + }, + { + "epoch": 0.42, + "grad_norm": 1.8730154257342784, + "learning_rate": 1.298970891496846e-05, + "loss": 1.0297, + "step": 17881 + }, + { + "epoch": 0.42, + "grad_norm": 2.2635668477394644, + "learning_rate": 1.2988980765000867e-05, + "loss": 1.0055, + "step": 17882 + }, + { + "epoch": 0.42, + "grad_norm": 2.0028537162050633, + "learning_rate": 1.298825259763048e-05, + "loss": 1.1381, + "step": 17883 + }, + { + "epoch": 0.42, + "grad_norm": 2.0668519242335965, + "learning_rate": 1.298752441286154e-05, + "loss": 1.0148, + "step": 17884 + }, + { + "epoch": 0.42, + "grad_norm": 2.0043108886663026, + "learning_rate": 1.2986796210698288e-05, + "loss": 1.1146, + "step": 17885 + }, + { + "epoch": 0.42, + "grad_norm": 1.999142781945419, + "learning_rate": 1.2986067991144963e-05, + "loss": 0.974, + "step": 17886 + }, + { + "epoch": 0.42, + "grad_norm": 2.7805492353433334, + "learning_rate": 1.2985339754205806e-05, + "loss": 1.0155, + "step": 17887 + }, + { + "epoch": 0.42, + "grad_norm": 1.1718611295197099, + "learning_rate": 1.2984611499885053e-05, + "loss": 1.0181, + "step": 17888 + }, + { + "epoch": 0.42, + "grad_norm": 2.176901691057785, + "learning_rate": 1.2983883228186949e-05, + "loss": 0.9722, + "step": 17889 + }, + { + "epoch": 0.42, + "grad_norm": 1.8961811473938464, + "learning_rate": 1.2983154939115734e-05, + "loss": 1.0833, + "step": 17890 + }, + { + "epoch": 0.42, + "grad_norm": 2.0802530452447594, + "learning_rate": 1.2982426632675641e-05, + "loss": 1.047, + "step": 17891 + }, + { + "epoch": 0.42, + "grad_norm": 2.1653579892981196, + "learning_rate": 1.2981698308870923e-05, + "loss": 1.0578, + "step": 17892 + }, + { + "epoch": 0.42, + "grad_norm": 1.9502403116669733, + "learning_rate": 1.2980969967705806e-05, + "loss": 1.0359, + "step": 17893 + }, + { + "epoch": 0.42, + "grad_norm": 2.2050690918949427, + "learning_rate": 1.2980241609184546e-05, + "loss": 1.0239, + "step": 17894 + }, + { + "epoch": 0.42, + "grad_norm": 1.8992300953314343, + "learning_rate": 1.2979513233311367e-05, + "loss": 0.889, + "step": 17895 + }, + { + "epoch": 0.42, + "grad_norm": 1.8040679019221506, + "learning_rate": 1.2978784840090526e-05, + "loss": 1.0498, + "step": 17896 + }, + { + "epoch": 0.42, + "grad_norm": 2.140472009205427, + "learning_rate": 1.2978056429526254e-05, + "loss": 1.0542, + "step": 17897 + }, + { + "epoch": 0.42, + "grad_norm": 2.2154023794313753, + "learning_rate": 1.2977328001622793e-05, + "loss": 0.9202, + "step": 17898 + }, + { + "epoch": 0.42, + "grad_norm": 2.2858606035488758, + "learning_rate": 1.2976599556384385e-05, + "loss": 1.0212, + "step": 17899 + }, + { + "epoch": 0.42, + "grad_norm": 1.9707610846730605, + "learning_rate": 1.2975871093815275e-05, + "loss": 0.9926, + "step": 17900 + }, + { + "epoch": 0.42, + "grad_norm": 2.197291803468948, + "learning_rate": 1.2975142613919702e-05, + "loss": 1.0633, + "step": 17901 + }, + { + "epoch": 0.42, + "grad_norm": 2.3088836591769932, + "learning_rate": 1.2974414116701903e-05, + "loss": 1.052, + "step": 17902 + }, + { + "epoch": 0.42, + "grad_norm": 2.8613891166234104, + "learning_rate": 1.2973685602166129e-05, + "loss": 1.1051, + "step": 17903 + }, + { + "epoch": 0.42, + "grad_norm": 1.9206719145543252, + "learning_rate": 1.297295707031661e-05, + "loss": 0.8816, + "step": 17904 + }, + { + "epoch": 0.42, + "grad_norm": 2.132379643442993, + "learning_rate": 1.2972228521157598e-05, + "loss": 0.9743, + "step": 17905 + }, + { + "epoch": 0.42, + "grad_norm": 2.4224117402425933, + "learning_rate": 1.2971499954693327e-05, + "loss": 1.0173, + "step": 17906 + }, + { + "epoch": 0.42, + "grad_norm": 2.1489298063542575, + "learning_rate": 1.2970771370928045e-05, + "loss": 1.0723, + "step": 17907 + }, + { + "epoch": 0.42, + "grad_norm": 1.0499723674242227, + "learning_rate": 1.2970042769865992e-05, + "loss": 0.991, + "step": 17908 + }, + { + "epoch": 0.42, + "grad_norm": 2.3957860325747395, + "learning_rate": 1.2969314151511407e-05, + "loss": 0.964, + "step": 17909 + }, + { + "epoch": 0.42, + "grad_norm": 1.9102969749174117, + "learning_rate": 1.2968585515868537e-05, + "loss": 1.0413, + "step": 17910 + }, + { + "epoch": 0.42, + "grad_norm": 2.1221877398370173, + "learning_rate": 1.2967856862941623e-05, + "loss": 1.012, + "step": 17911 + }, + { + "epoch": 0.42, + "grad_norm": 2.1210650148781194, + "learning_rate": 1.2967128192734903e-05, + "loss": 0.9775, + "step": 17912 + }, + { + "epoch": 0.42, + "grad_norm": 1.9285830751118769, + "learning_rate": 1.2966399505252628e-05, + "loss": 1.1304, + "step": 17913 + }, + { + "epoch": 0.42, + "grad_norm": 2.8590498012430126, + "learning_rate": 1.2965670800499032e-05, + "loss": 1.0264, + "step": 17914 + }, + { + "epoch": 0.42, + "grad_norm": 2.243116303854299, + "learning_rate": 1.2964942078478365e-05, + "loss": 1.0078, + "step": 17915 + }, + { + "epoch": 0.42, + "grad_norm": 2.0867976605588563, + "learning_rate": 1.2964213339194865e-05, + "loss": 1.0182, + "step": 17916 + }, + { + "epoch": 0.42, + "grad_norm": 1.9845688691273349, + "learning_rate": 1.2963484582652775e-05, + "loss": 1.1104, + "step": 17917 + }, + { + "epoch": 0.42, + "grad_norm": 2.0880292437404164, + "learning_rate": 1.2962755808856341e-05, + "loss": 1.0358, + "step": 17918 + }, + { + "epoch": 0.42, + "grad_norm": 2.12849080533682, + "learning_rate": 1.2962027017809804e-05, + "loss": 1.0359, + "step": 17919 + }, + { + "epoch": 0.42, + "grad_norm": 2.32589874237497, + "learning_rate": 1.296129820951741e-05, + "loss": 1.0547, + "step": 17920 + }, + { + "epoch": 0.42, + "grad_norm": 1.956642899104484, + "learning_rate": 1.2960569383983397e-05, + "loss": 1.0623, + "step": 17921 + }, + { + "epoch": 0.42, + "grad_norm": 1.7535561448315915, + "learning_rate": 1.2959840541212017e-05, + "loss": 1.0259, + "step": 17922 + }, + { + "epoch": 0.42, + "grad_norm": 2.049223795155435, + "learning_rate": 1.29591116812075e-05, + "loss": 1.0298, + "step": 17923 + }, + { + "epoch": 0.42, + "grad_norm": 2.1525322564165172, + "learning_rate": 1.2958382803974106e-05, + "loss": 1.0887, + "step": 17924 + }, + { + "epoch": 0.42, + "grad_norm": 1.944780234304698, + "learning_rate": 1.2957653909516065e-05, + "loss": 1.111, + "step": 17925 + }, + { + "epoch": 0.42, + "grad_norm": 2.6610149773485663, + "learning_rate": 1.2956924997837633e-05, + "loss": 0.9127, + "step": 17926 + }, + { + "epoch": 0.42, + "grad_norm": 1.856669779616239, + "learning_rate": 1.2956196068943043e-05, + "loss": 0.7881, + "step": 17927 + }, + { + "epoch": 0.42, + "grad_norm": 3.4344625208672936, + "learning_rate": 1.2955467122836545e-05, + "loss": 0.9111, + "step": 17928 + }, + { + "epoch": 0.42, + "grad_norm": 1.9358080824591455, + "learning_rate": 1.295473815952238e-05, + "loss": 1.1588, + "step": 17929 + }, + { + "epoch": 0.42, + "grad_norm": 2.222835297131605, + "learning_rate": 1.2954009179004794e-05, + "loss": 1.0389, + "step": 17930 + }, + { + "epoch": 0.42, + "grad_norm": 2.333988126747755, + "learning_rate": 1.2953280181288033e-05, + "loss": 0.9566, + "step": 17931 + }, + { + "epoch": 0.42, + "grad_norm": 2.0187506069661856, + "learning_rate": 1.2952551166376339e-05, + "loss": 1.1153, + "step": 17932 + }, + { + "epoch": 0.42, + "grad_norm": 2.578106314101922, + "learning_rate": 1.2951822134273957e-05, + "loss": 1.0652, + "step": 17933 + }, + { + "epoch": 0.42, + "grad_norm": 1.173053135035049, + "learning_rate": 1.2951093084985131e-05, + "loss": 0.9538, + "step": 17934 + }, + { + "epoch": 0.42, + "grad_norm": 1.9948447794691977, + "learning_rate": 1.295036401851411e-05, + "loss": 0.9891, + "step": 17935 + }, + { + "epoch": 0.42, + "grad_norm": 2.060251276680235, + "learning_rate": 1.2949634934865134e-05, + "loss": 1.1584, + "step": 17936 + }, + { + "epoch": 0.42, + "grad_norm": 2.013762465662951, + "learning_rate": 1.294890583404245e-05, + "loss": 1.0742, + "step": 17937 + }, + { + "epoch": 0.42, + "grad_norm": 1.9687909540555502, + "learning_rate": 1.29481767160503e-05, + "loss": 1.1041, + "step": 17938 + }, + { + "epoch": 0.42, + "grad_norm": 1.8420731965749495, + "learning_rate": 1.2947447580892933e-05, + "loss": 1.0624, + "step": 17939 + }, + { + "epoch": 0.42, + "grad_norm": 2.1124935557868287, + "learning_rate": 1.2946718428574595e-05, + "loss": 1.0256, + "step": 17940 + }, + { + "epoch": 0.42, + "grad_norm": 2.0556151529233873, + "learning_rate": 1.294598925909953e-05, + "loss": 1.0247, + "step": 17941 + }, + { + "epoch": 0.42, + "grad_norm": 1.8172502534693944, + "learning_rate": 1.294526007247198e-05, + "loss": 0.98, + "step": 17942 + }, + { + "epoch": 0.42, + "grad_norm": 2.0177963067662135, + "learning_rate": 1.2944530868696194e-05, + "loss": 1.0819, + "step": 17943 + }, + { + "epoch": 0.42, + "grad_norm": 1.6809035371093703, + "learning_rate": 1.2943801647776418e-05, + "loss": 0.9474, + "step": 17944 + }, + { + "epoch": 0.42, + "grad_norm": 1.9485493861152825, + "learning_rate": 1.2943072409716897e-05, + "loss": 0.961, + "step": 17945 + }, + { + "epoch": 0.42, + "grad_norm": 2.1458026625218274, + "learning_rate": 1.2942343154521875e-05, + "loss": 0.9463, + "step": 17946 + }, + { + "epoch": 0.42, + "grad_norm": 2.469431320053914, + "learning_rate": 1.2941613882195601e-05, + "loss": 1.1027, + "step": 17947 + }, + { + "epoch": 0.42, + "grad_norm": 2.6268356688734715, + "learning_rate": 1.294088459274232e-05, + "loss": 1.0226, + "step": 17948 + }, + { + "epoch": 0.42, + "grad_norm": 2.0622843982794237, + "learning_rate": 1.2940155286166279e-05, + "loss": 1.0867, + "step": 17949 + }, + { + "epoch": 0.42, + "grad_norm": 1.9115765433643739, + "learning_rate": 1.2939425962471719e-05, + "loss": 1.0891, + "step": 17950 + }, + { + "epoch": 0.42, + "grad_norm": 1.8749093592609993, + "learning_rate": 1.293869662166289e-05, + "loss": 1.1371, + "step": 17951 + }, + { + "epoch": 0.42, + "grad_norm": 1.9062863341810459, + "learning_rate": 1.2937967263744044e-05, + "loss": 1.0714, + "step": 17952 + }, + { + "epoch": 0.42, + "grad_norm": 2.1215786557450564, + "learning_rate": 1.2937237888719418e-05, + "loss": 1.0114, + "step": 17953 + }, + { + "epoch": 0.42, + "grad_norm": 1.8939886562927215, + "learning_rate": 1.2936508496593268e-05, + "loss": 1.0797, + "step": 17954 + }, + { + "epoch": 0.42, + "grad_norm": 2.0750662801681754, + "learning_rate": 1.2935779087369832e-05, + "loss": 1.0197, + "step": 17955 + }, + { + "epoch": 0.42, + "grad_norm": 1.9675991327241773, + "learning_rate": 1.2935049661053363e-05, + "loss": 1.0042, + "step": 17956 + }, + { + "epoch": 0.42, + "grad_norm": 2.0667728734772197, + "learning_rate": 1.2934320217648101e-05, + "loss": 1.1686, + "step": 17957 + }, + { + "epoch": 0.42, + "grad_norm": 2.2815256405916786, + "learning_rate": 1.2933590757158303e-05, + "loss": 1.0043, + "step": 17958 + }, + { + "epoch": 0.42, + "grad_norm": 2.575886889385061, + "learning_rate": 1.293286127958821e-05, + "loss": 1.0974, + "step": 17959 + }, + { + "epoch": 0.42, + "grad_norm": 2.179231698565334, + "learning_rate": 1.2932131784942069e-05, + "loss": 1.0935, + "step": 17960 + }, + { + "epoch": 0.42, + "grad_norm": 2.0528222784563153, + "learning_rate": 1.2931402273224128e-05, + "loss": 1.0253, + "step": 17961 + }, + { + "epoch": 0.42, + "grad_norm": 2.3588120397955454, + "learning_rate": 1.2930672744438634e-05, + "loss": 1.0588, + "step": 17962 + }, + { + "epoch": 0.42, + "grad_norm": 2.181673379240669, + "learning_rate": 1.2929943198589836e-05, + "loss": 1.0129, + "step": 17963 + }, + { + "epoch": 0.42, + "grad_norm": 2.030113217562672, + "learning_rate": 1.2929213635681983e-05, + "loss": 0.9917, + "step": 17964 + }, + { + "epoch": 0.42, + "grad_norm": 2.2646308592278195, + "learning_rate": 1.292848405571932e-05, + "loss": 1.0914, + "step": 17965 + }, + { + "epoch": 0.42, + "grad_norm": 1.7830661071330494, + "learning_rate": 1.2927754458706094e-05, + "loss": 0.9462, + "step": 17966 + }, + { + "epoch": 0.42, + "grad_norm": 2.4227002215562123, + "learning_rate": 1.2927024844646556e-05, + "loss": 1.0921, + "step": 17967 + }, + { + "epoch": 0.42, + "grad_norm": 2.0308623313056837, + "learning_rate": 1.2926295213544953e-05, + "loss": 0.9658, + "step": 17968 + }, + { + "epoch": 0.42, + "grad_norm": 2.1997260532128258, + "learning_rate": 1.2925565565405536e-05, + "loss": 1.0921, + "step": 17969 + }, + { + "epoch": 0.42, + "grad_norm": 2.555986155998172, + "learning_rate": 1.2924835900232542e-05, + "loss": 0.9247, + "step": 17970 + }, + { + "epoch": 0.42, + "grad_norm": 2.170090586067776, + "learning_rate": 1.2924106218030235e-05, + "loss": 0.9482, + "step": 17971 + }, + { + "epoch": 0.42, + "grad_norm": 2.1037057534533714, + "learning_rate": 1.292337651880285e-05, + "loss": 0.9085, + "step": 17972 + }, + { + "epoch": 0.42, + "grad_norm": 2.011584608341677, + "learning_rate": 1.2922646802554649e-05, + "loss": 1.083, + "step": 17973 + }, + { + "epoch": 0.42, + "grad_norm": 1.8717523589879634, + "learning_rate": 1.2921917069289864e-05, + "loss": 0.9402, + "step": 17974 + }, + { + "epoch": 0.42, + "grad_norm": 1.9271253138906528, + "learning_rate": 1.2921187319012761e-05, + "loss": 0.9945, + "step": 17975 + }, + { + "epoch": 0.42, + "grad_norm": 2.144637384572052, + "learning_rate": 1.2920457551727575e-05, + "loss": 0.9957, + "step": 17976 + }, + { + "epoch": 0.42, + "grad_norm": 1.9390726999705037, + "learning_rate": 1.2919727767438562e-05, + "loss": 0.9769, + "step": 17977 + }, + { + "epoch": 0.42, + "grad_norm": 2.2154389170227664, + "learning_rate": 1.2918997966149969e-05, + "loss": 1.1983, + "step": 17978 + }, + { + "epoch": 0.42, + "grad_norm": 2.2744631182571244, + "learning_rate": 1.2918268147866047e-05, + "loss": 0.8888, + "step": 17979 + }, + { + "epoch": 0.42, + "grad_norm": 3.072441688135015, + "learning_rate": 1.2917538312591043e-05, + "loss": 0.9888, + "step": 17980 + }, + { + "epoch": 0.42, + "grad_norm": 2.295592390274621, + "learning_rate": 1.2916808460329206e-05, + "loss": 1.1236, + "step": 17981 + }, + { + "epoch": 0.42, + "grad_norm": 2.322169411485255, + "learning_rate": 1.2916078591084788e-05, + "loss": 0.8996, + "step": 17982 + }, + { + "epoch": 0.42, + "grad_norm": 1.955230058779102, + "learning_rate": 1.291534870486204e-05, + "loss": 1.0858, + "step": 17983 + }, + { + "epoch": 0.42, + "grad_norm": 2.2275014905089714, + "learning_rate": 1.2914618801665205e-05, + "loss": 1.0926, + "step": 17984 + }, + { + "epoch": 0.42, + "grad_norm": 2.0615516340594393, + "learning_rate": 1.2913888881498537e-05, + "loss": 1.0791, + "step": 17985 + }, + { + "epoch": 0.42, + "grad_norm": 1.1095605169574643, + "learning_rate": 1.2913158944366286e-05, + "loss": 1.0009, + "step": 17986 + }, + { + "epoch": 0.42, + "grad_norm": 1.1615315032362703, + "learning_rate": 1.2912428990272704e-05, + "loss": 0.9708, + "step": 17987 + }, + { + "epoch": 0.42, + "grad_norm": 1.8492164499089645, + "learning_rate": 1.2911699019222035e-05, + "loss": 1.0178, + "step": 17988 + }, + { + "epoch": 0.42, + "grad_norm": 2.017756638264514, + "learning_rate": 1.2910969031218532e-05, + "loss": 0.9989, + "step": 17989 + }, + { + "epoch": 0.42, + "grad_norm": 2.05559247938822, + "learning_rate": 1.2910239026266446e-05, + "loss": 1.022, + "step": 17990 + }, + { + "epoch": 0.42, + "grad_norm": 1.9544510552664143, + "learning_rate": 1.290950900437003e-05, + "loss": 0.9747, + "step": 17991 + }, + { + "epoch": 0.42, + "grad_norm": 2.1258637249833994, + "learning_rate": 1.2908778965533533e-05, + "loss": 1.0781, + "step": 17992 + }, + { + "epoch": 0.42, + "grad_norm": 1.0224667336510052, + "learning_rate": 1.2908048909761197e-05, + "loss": 0.9786, + "step": 17993 + }, + { + "epoch": 0.42, + "grad_norm": 2.125876479537597, + "learning_rate": 1.2907318837057286e-05, + "loss": 0.9836, + "step": 17994 + }, + { + "epoch": 0.42, + "grad_norm": 2.1098959283966825, + "learning_rate": 1.2906588747426038e-05, + "loss": 1.0148, + "step": 17995 + }, + { + "epoch": 0.42, + "grad_norm": 2.1792035567565042, + "learning_rate": 1.2905858640871717e-05, + "loss": 1.0003, + "step": 17996 + }, + { + "epoch": 0.42, + "grad_norm": 1.1330710790100953, + "learning_rate": 1.2905128517398562e-05, + "loss": 1.0617, + "step": 17997 + }, + { + "epoch": 0.42, + "grad_norm": 2.2887858824396705, + "learning_rate": 1.2904398377010832e-05, + "loss": 0.9381, + "step": 17998 + }, + { + "epoch": 0.42, + "grad_norm": 2.219271947481557, + "learning_rate": 1.2903668219712773e-05, + "loss": 0.9934, + "step": 17999 + }, + { + "epoch": 0.42, + "grad_norm": 2.762866102486929, + "learning_rate": 1.2902938045508638e-05, + "loss": 1.0192, + "step": 18000 + }, + { + "epoch": 0.42, + "grad_norm": 2.8489652251075093, + "learning_rate": 1.2902207854402682e-05, + "loss": 0.8694, + "step": 18001 + }, + { + "epoch": 0.42, + "grad_norm": 2.1207482251253165, + "learning_rate": 1.2901477646399147e-05, + "loss": 0.9472, + "step": 18002 + }, + { + "epoch": 0.42, + "grad_norm": 2.0045447795047306, + "learning_rate": 1.2900747421502297e-05, + "loss": 1.0006, + "step": 18003 + }, + { + "epoch": 0.42, + "grad_norm": 1.838582626772753, + "learning_rate": 1.2900017179716373e-05, + "loss": 0.9481, + "step": 18004 + }, + { + "epoch": 0.42, + "grad_norm": 1.8940629955589112, + "learning_rate": 1.2899286921045633e-05, + "loss": 1.0196, + "step": 18005 + }, + { + "epoch": 0.42, + "grad_norm": 2.0224402928323273, + "learning_rate": 1.2898556645494327e-05, + "loss": 0.9949, + "step": 18006 + }, + { + "epoch": 0.42, + "grad_norm": 1.1464874288454925, + "learning_rate": 1.2897826353066705e-05, + "loss": 0.9388, + "step": 18007 + }, + { + "epoch": 0.42, + "grad_norm": 1.080038607475429, + "learning_rate": 1.289709604376702e-05, + "loss": 0.9076, + "step": 18008 + }, + { + "epoch": 0.42, + "grad_norm": 1.8683904816213133, + "learning_rate": 1.2896365717599526e-05, + "loss": 1.0701, + "step": 18009 + }, + { + "epoch": 0.42, + "grad_norm": 2.0594050808923408, + "learning_rate": 1.2895635374568474e-05, + "loss": 1.0049, + "step": 18010 + }, + { + "epoch": 0.42, + "grad_norm": 2.1836893430950615, + "learning_rate": 1.2894905014678117e-05, + "loss": 1.1007, + "step": 18011 + }, + { + "epoch": 0.42, + "grad_norm": 2.02260086911136, + "learning_rate": 1.2894174637932703e-05, + "loss": 0.9593, + "step": 18012 + }, + { + "epoch": 0.42, + "grad_norm": 1.9097439844193185, + "learning_rate": 1.289344424433649e-05, + "loss": 1.0057, + "step": 18013 + }, + { + "epoch": 0.42, + "grad_norm": 1.8889721081478699, + "learning_rate": 1.2892713833893731e-05, + "loss": 1.0723, + "step": 18014 + }, + { + "epoch": 0.42, + "grad_norm": 2.4002615649134116, + "learning_rate": 1.2891983406608673e-05, + "loss": 0.8641, + "step": 18015 + }, + { + "epoch": 0.42, + "grad_norm": 1.9231077640344596, + "learning_rate": 1.2891252962485573e-05, + "loss": 1.1715, + "step": 18016 + }, + { + "epoch": 0.42, + "grad_norm": 1.981766819427693, + "learning_rate": 1.2890522501528682e-05, + "loss": 0.9823, + "step": 18017 + }, + { + "epoch": 0.42, + "grad_norm": 2.243281185968577, + "learning_rate": 1.2889792023742253e-05, + "loss": 1.073, + "step": 18018 + }, + { + "epoch": 0.42, + "grad_norm": 1.8988338042584125, + "learning_rate": 1.2889061529130542e-05, + "loss": 1.1285, + "step": 18019 + }, + { + "epoch": 0.42, + "grad_norm": 2.0463469853477045, + "learning_rate": 1.2888331017697804e-05, + "loss": 0.9894, + "step": 18020 + }, + { + "epoch": 0.42, + "grad_norm": 2.192237473663922, + "learning_rate": 1.2887600489448282e-05, + "loss": 1.0369, + "step": 18021 + }, + { + "epoch": 0.42, + "grad_norm": 2.072099879145847, + "learning_rate": 1.2886869944386237e-05, + "loss": 0.9795, + "step": 18022 + }, + { + "epoch": 0.42, + "grad_norm": 2.5659669493039132, + "learning_rate": 1.2886139382515924e-05, + "loss": 0.8871, + "step": 18023 + }, + { + "epoch": 0.42, + "grad_norm": 1.0678792058377562, + "learning_rate": 1.2885408803841593e-05, + "loss": 0.9768, + "step": 18024 + }, + { + "epoch": 0.42, + "grad_norm": 1.9554415997553258, + "learning_rate": 1.2884678208367493e-05, + "loss": 0.9589, + "step": 18025 + }, + { + "epoch": 0.42, + "grad_norm": 2.8652112449727825, + "learning_rate": 1.288394759609789e-05, + "loss": 1.1147, + "step": 18026 + }, + { + "epoch": 0.42, + "grad_norm": 2.2305684536071535, + "learning_rate": 1.2883216967037029e-05, + "loss": 0.9291, + "step": 18027 + }, + { + "epoch": 0.42, + "grad_norm": 2.212606910482994, + "learning_rate": 1.2882486321189165e-05, + "loss": 1.0324, + "step": 18028 + }, + { + "epoch": 0.42, + "grad_norm": 2.2491654983936464, + "learning_rate": 1.2881755658558555e-05, + "loss": 0.933, + "step": 18029 + }, + { + "epoch": 0.42, + "grad_norm": 1.9510257766366765, + "learning_rate": 1.2881024979149448e-05, + "loss": 1.0229, + "step": 18030 + }, + { + "epoch": 0.42, + "grad_norm": 2.0014270865924235, + "learning_rate": 1.2880294282966104e-05, + "loss": 0.993, + "step": 18031 + }, + { + "epoch": 0.42, + "grad_norm": 1.01518080401417, + "learning_rate": 1.2879563570012773e-05, + "loss": 0.9742, + "step": 18032 + }, + { + "epoch": 0.42, + "grad_norm": 2.1061833693633116, + "learning_rate": 1.2878832840293713e-05, + "loss": 1.0248, + "step": 18033 + }, + { + "epoch": 0.42, + "grad_norm": 1.7774246065119117, + "learning_rate": 1.2878102093813176e-05, + "loss": 1.0293, + "step": 18034 + }, + { + "epoch": 0.42, + "grad_norm": 2.268229057425698, + "learning_rate": 1.287737133057542e-05, + "loss": 1.0706, + "step": 18035 + }, + { + "epoch": 0.42, + "grad_norm": 1.1510578323127372, + "learning_rate": 1.2876640550584693e-05, + "loss": 0.9883, + "step": 18036 + }, + { + "epoch": 0.42, + "grad_norm": 2.2549865593674534, + "learning_rate": 1.2875909753845255e-05, + "loss": 0.9466, + "step": 18037 + }, + { + "epoch": 0.42, + "grad_norm": 6.839114083440803, + "learning_rate": 1.2875178940361361e-05, + "loss": 1.0702, + "step": 18038 + }, + { + "epoch": 0.42, + "grad_norm": 2.322127343771783, + "learning_rate": 1.2874448110137263e-05, + "loss": 0.9747, + "step": 18039 + }, + { + "epoch": 0.43, + "grad_norm": 2.2867002522356077, + "learning_rate": 1.2873717263177221e-05, + "loss": 1.0938, + "step": 18040 + }, + { + "epoch": 0.43, + "grad_norm": 2.4995157164847046, + "learning_rate": 1.2872986399485487e-05, + "loss": 1.0597, + "step": 18041 + }, + { + "epoch": 0.43, + "grad_norm": 1.9723932027117195, + "learning_rate": 1.2872255519066314e-05, + "loss": 0.9323, + "step": 18042 + }, + { + "epoch": 0.43, + "grad_norm": 1.9505152055538633, + "learning_rate": 1.2871524621923963e-05, + "loss": 0.9174, + "step": 18043 + }, + { + "epoch": 0.43, + "grad_norm": 2.2742842062482658, + "learning_rate": 1.2870793708062684e-05, + "loss": 1.0707, + "step": 18044 + }, + { + "epoch": 0.43, + "grad_norm": 2.221711750847142, + "learning_rate": 1.2870062777486738e-05, + "loss": 1.126, + "step": 18045 + }, + { + "epoch": 0.43, + "grad_norm": 1.8834645377143409, + "learning_rate": 1.2869331830200374e-05, + "loss": 0.9907, + "step": 18046 + }, + { + "epoch": 0.43, + "grad_norm": 2.257611671470387, + "learning_rate": 1.2868600866207856e-05, + "loss": 1.0945, + "step": 18047 + }, + { + "epoch": 0.43, + "grad_norm": 1.9561854114430333, + "learning_rate": 1.2867869885513432e-05, + "loss": 0.9512, + "step": 18048 + }, + { + "epoch": 0.43, + "grad_norm": 3.581134579458853, + "learning_rate": 1.2867138888121362e-05, + "loss": 1.109, + "step": 18049 + }, + { + "epoch": 0.43, + "grad_norm": 4.474151858487634, + "learning_rate": 1.2866407874035904e-05, + "loss": 1.0597, + "step": 18050 + }, + { + "epoch": 0.43, + "grad_norm": 2.1150941412610313, + "learning_rate": 1.2865676843261308e-05, + "loss": 1.0997, + "step": 18051 + }, + { + "epoch": 0.43, + "grad_norm": 2.052314942525858, + "learning_rate": 1.2864945795801837e-05, + "loss": 1.0901, + "step": 18052 + }, + { + "epoch": 0.43, + "grad_norm": 2.113496006962648, + "learning_rate": 1.2864214731661743e-05, + "loss": 1.1157, + "step": 18053 + }, + { + "epoch": 0.43, + "grad_norm": 2.08420591944242, + "learning_rate": 1.2863483650845288e-05, + "loss": 0.9352, + "step": 18054 + }, + { + "epoch": 0.43, + "grad_norm": 2.0954188314459983, + "learning_rate": 1.2862752553356721e-05, + "loss": 0.9085, + "step": 18055 + }, + { + "epoch": 0.43, + "grad_norm": 2.034549732299624, + "learning_rate": 1.2862021439200306e-05, + "loss": 0.9937, + "step": 18056 + }, + { + "epoch": 0.43, + "grad_norm": 1.1107569460276556, + "learning_rate": 1.2861290308380289e-05, + "loss": 1.016, + "step": 18057 + }, + { + "epoch": 0.43, + "grad_norm": 1.883964551535245, + "learning_rate": 1.286055916090094e-05, + "loss": 0.969, + "step": 18058 + }, + { + "epoch": 0.43, + "grad_norm": 1.8886819831687118, + "learning_rate": 1.2859827996766508e-05, + "loss": 0.9453, + "step": 18059 + }, + { + "epoch": 0.43, + "grad_norm": 1.9349548324417607, + "learning_rate": 1.2859096815981254e-05, + "loss": 0.9129, + "step": 18060 + }, + { + "epoch": 0.43, + "grad_norm": 2.205443218892003, + "learning_rate": 1.285836561854943e-05, + "loss": 1.0656, + "step": 18061 + }, + { + "epoch": 0.43, + "grad_norm": 2.12989902479236, + "learning_rate": 1.2857634404475297e-05, + "loss": 1.0085, + "step": 18062 + }, + { + "epoch": 0.43, + "grad_norm": 2.398247775571408, + "learning_rate": 1.2856903173763114e-05, + "loss": 1.0454, + "step": 18063 + }, + { + "epoch": 0.43, + "grad_norm": 2.238261556351489, + "learning_rate": 1.2856171926417134e-05, + "loss": 0.9911, + "step": 18064 + }, + { + "epoch": 0.43, + "grad_norm": 2.179695546825828, + "learning_rate": 1.2855440662441617e-05, + "loss": 0.938, + "step": 18065 + }, + { + "epoch": 0.43, + "grad_norm": 1.859392550229344, + "learning_rate": 1.285470938184082e-05, + "loss": 1.0511, + "step": 18066 + }, + { + "epoch": 0.43, + "grad_norm": 2.286882547005739, + "learning_rate": 1.2853978084619004e-05, + "loss": 1.1397, + "step": 18067 + }, + { + "epoch": 0.43, + "grad_norm": 2.2012277309833306, + "learning_rate": 1.285324677078042e-05, + "loss": 0.966, + "step": 18068 + }, + { + "epoch": 0.43, + "grad_norm": 2.962357634761406, + "learning_rate": 1.2852515440329331e-05, + "loss": 1.0521, + "step": 18069 + }, + { + "epoch": 0.43, + "grad_norm": 1.1365163391843487, + "learning_rate": 1.2851784093269997e-05, + "loss": 1.0283, + "step": 18070 + }, + { + "epoch": 0.43, + "grad_norm": 2.054721706832227, + "learning_rate": 1.2851052729606673e-05, + "loss": 0.8879, + "step": 18071 + }, + { + "epoch": 0.43, + "grad_norm": 1.8532318831629842, + "learning_rate": 1.2850321349343613e-05, + "loss": 1.0378, + "step": 18072 + }, + { + "epoch": 0.43, + "grad_norm": 2.2921975274011315, + "learning_rate": 1.2849589952485084e-05, + "loss": 1.1331, + "step": 18073 + }, + { + "epoch": 0.43, + "grad_norm": 2.037071804983025, + "learning_rate": 1.2848858539035339e-05, + "loss": 0.8535, + "step": 18074 + }, + { + "epoch": 0.43, + "grad_norm": 2.180593737441157, + "learning_rate": 1.2848127108998638e-05, + "loss": 0.9999, + "step": 18075 + }, + { + "epoch": 0.43, + "grad_norm": 2.0975606347559013, + "learning_rate": 1.2847395662379236e-05, + "loss": 0.9395, + "step": 18076 + }, + { + "epoch": 0.43, + "grad_norm": 1.9411758689359127, + "learning_rate": 1.28466641991814e-05, + "loss": 1.1405, + "step": 18077 + }, + { + "epoch": 0.43, + "grad_norm": 2.2804682494230266, + "learning_rate": 1.284593271940938e-05, + "loss": 1.0019, + "step": 18078 + }, + { + "epoch": 0.43, + "grad_norm": 1.1061120099412716, + "learning_rate": 1.2845201223067441e-05, + "loss": 0.9736, + "step": 18079 + }, + { + "epoch": 0.43, + "grad_norm": 2.047965660745912, + "learning_rate": 1.2844469710159837e-05, + "loss": 1.0214, + "step": 18080 + }, + { + "epoch": 0.43, + "grad_norm": 2.0516506349321877, + "learning_rate": 1.2843738180690833e-05, + "loss": 0.9757, + "step": 18081 + }, + { + "epoch": 0.43, + "grad_norm": 2.0571570800818604, + "learning_rate": 1.2843006634664683e-05, + "loss": 0.991, + "step": 18082 + }, + { + "epoch": 0.43, + "grad_norm": 2.3685250948962526, + "learning_rate": 1.2842275072085649e-05, + "loss": 0.9975, + "step": 18083 + }, + { + "epoch": 0.43, + "grad_norm": 2.1355025916238173, + "learning_rate": 1.2841543492957988e-05, + "loss": 0.9957, + "step": 18084 + }, + { + "epoch": 0.43, + "grad_norm": 1.952805566653703, + "learning_rate": 1.2840811897285963e-05, + "loss": 1.2147, + "step": 18085 + }, + { + "epoch": 0.43, + "grad_norm": 2.060478598245404, + "learning_rate": 1.284008028507383e-05, + "loss": 1.0203, + "step": 18086 + }, + { + "epoch": 0.43, + "grad_norm": 1.9140492586607196, + "learning_rate": 1.2839348656325851e-05, + "loss": 1.0513, + "step": 18087 + }, + { + "epoch": 0.43, + "grad_norm": 1.9108880326803788, + "learning_rate": 1.2838617011046288e-05, + "loss": 0.9252, + "step": 18088 + }, + { + "epoch": 0.43, + "grad_norm": 2.017488749865766, + "learning_rate": 1.2837885349239396e-05, + "loss": 0.9973, + "step": 18089 + }, + { + "epoch": 0.43, + "grad_norm": 1.0873930376774046, + "learning_rate": 1.2837153670909437e-05, + "loss": 0.9915, + "step": 18090 + }, + { + "epoch": 0.43, + "grad_norm": 2.2297592746661166, + "learning_rate": 1.2836421976060672e-05, + "loss": 1.0052, + "step": 18091 + }, + { + "epoch": 0.43, + "grad_norm": 1.9371650650487109, + "learning_rate": 1.283569026469736e-05, + "loss": 1.0433, + "step": 18092 + }, + { + "epoch": 0.43, + "grad_norm": 2.333519128492976, + "learning_rate": 1.2834958536823758e-05, + "loss": 1.062, + "step": 18093 + }, + { + "epoch": 0.43, + "grad_norm": 1.8844812381230904, + "learning_rate": 1.2834226792444136e-05, + "loss": 1.0523, + "step": 18094 + }, + { + "epoch": 0.43, + "grad_norm": 2.133505941895496, + "learning_rate": 1.2833495031562744e-05, + "loss": 0.9701, + "step": 18095 + }, + { + "epoch": 0.43, + "grad_norm": 2.0424796165302204, + "learning_rate": 1.283276325418385e-05, + "loss": 1.0904, + "step": 18096 + }, + { + "epoch": 0.43, + "grad_norm": 1.8606625774430194, + "learning_rate": 1.2832031460311707e-05, + "loss": 0.905, + "step": 18097 + }, + { + "epoch": 0.43, + "grad_norm": 1.9517911086744322, + "learning_rate": 1.2831299649950582e-05, + "loss": 1.0624, + "step": 18098 + }, + { + "epoch": 0.43, + "grad_norm": 1.9537870107707442, + "learning_rate": 1.2830567823104736e-05, + "loss": 1.0922, + "step": 18099 + }, + { + "epoch": 0.43, + "grad_norm": 2.360150600325254, + "learning_rate": 1.2829835979778424e-05, + "loss": 1.0038, + "step": 18100 + }, + { + "epoch": 0.43, + "grad_norm": 2.0807240316095275, + "learning_rate": 1.2829104119975917e-05, + "loss": 0.8953, + "step": 18101 + }, + { + "epoch": 0.43, + "grad_norm": 1.8740096001017292, + "learning_rate": 1.2828372243701463e-05, + "loss": 0.9442, + "step": 18102 + }, + { + "epoch": 0.43, + "grad_norm": 2.0023195338468587, + "learning_rate": 1.2827640350959338e-05, + "loss": 1.0575, + "step": 18103 + }, + { + "epoch": 0.43, + "grad_norm": 2.40168075581385, + "learning_rate": 1.2826908441753789e-05, + "loss": 1.01, + "step": 18104 + }, + { + "epoch": 0.43, + "grad_norm": 2.06741131002474, + "learning_rate": 1.2826176516089088e-05, + "loss": 1.0786, + "step": 18105 + }, + { + "epoch": 0.43, + "grad_norm": 2.157860157759368, + "learning_rate": 1.2825444573969494e-05, + "loss": 1.0724, + "step": 18106 + }, + { + "epoch": 0.43, + "grad_norm": 1.9071371130273744, + "learning_rate": 1.2824712615399264e-05, + "loss": 1.0202, + "step": 18107 + }, + { + "epoch": 0.43, + "grad_norm": 1.9741379425533765, + "learning_rate": 1.2823980640382663e-05, + "loss": 1.0271, + "step": 18108 + }, + { + "epoch": 0.43, + "grad_norm": 2.08122221271531, + "learning_rate": 1.2823248648923954e-05, + "loss": 1.0059, + "step": 18109 + }, + { + "epoch": 0.43, + "grad_norm": 2.0747098116946963, + "learning_rate": 1.2822516641027396e-05, + "loss": 1.099, + "step": 18110 + }, + { + "epoch": 0.43, + "grad_norm": 2.6991656214678876, + "learning_rate": 1.2821784616697253e-05, + "loss": 0.9978, + "step": 18111 + }, + { + "epoch": 0.43, + "grad_norm": 2.150489694461832, + "learning_rate": 1.2821052575937786e-05, + "loss": 1.0223, + "step": 18112 + }, + { + "epoch": 0.43, + "grad_norm": 2.0790157192886767, + "learning_rate": 1.282032051875326e-05, + "loss": 0.9851, + "step": 18113 + }, + { + "epoch": 0.43, + "grad_norm": 2.0105101373097023, + "learning_rate": 1.2819588445147932e-05, + "loss": 1.0506, + "step": 18114 + }, + { + "epoch": 0.43, + "grad_norm": 2.1166852598599015, + "learning_rate": 1.2818856355126071e-05, + "loss": 0.934, + "step": 18115 + }, + { + "epoch": 0.43, + "grad_norm": 2.76511364177294, + "learning_rate": 1.2818124248691932e-05, + "loss": 0.8847, + "step": 18116 + }, + { + "epoch": 0.43, + "grad_norm": 2.095005335853284, + "learning_rate": 1.2817392125849785e-05, + "loss": 1.0566, + "step": 18117 + }, + { + "epoch": 0.43, + "grad_norm": 2.0369330617454047, + "learning_rate": 1.2816659986603887e-05, + "loss": 1.1176, + "step": 18118 + }, + { + "epoch": 0.43, + "grad_norm": 2.074845176105295, + "learning_rate": 1.2815927830958504e-05, + "loss": 1.0112, + "step": 18119 + }, + { + "epoch": 0.43, + "grad_norm": 2.26895605722845, + "learning_rate": 1.2815195658917896e-05, + "loss": 0.9477, + "step": 18120 + }, + { + "epoch": 0.43, + "grad_norm": 1.1041620207513003, + "learning_rate": 1.2814463470486329e-05, + "loss": 0.9713, + "step": 18121 + }, + { + "epoch": 0.43, + "grad_norm": 2.9615753573187575, + "learning_rate": 1.2813731265668066e-05, + "loss": 1.0027, + "step": 18122 + }, + { + "epoch": 0.43, + "grad_norm": 1.8830980517427232, + "learning_rate": 1.2812999044467364e-05, + "loss": 1.0137, + "step": 18123 + }, + { + "epoch": 0.43, + "grad_norm": 2.102748521890625, + "learning_rate": 1.2812266806888496e-05, + "loss": 1.0377, + "step": 18124 + }, + { + "epoch": 0.43, + "grad_norm": 2.076077788478331, + "learning_rate": 1.2811534552935716e-05, + "loss": 1.0374, + "step": 18125 + }, + { + "epoch": 0.43, + "grad_norm": 2.1048702834657202, + "learning_rate": 1.2810802282613297e-05, + "loss": 1.1032, + "step": 18126 + }, + { + "epoch": 0.43, + "grad_norm": 2.162114106214459, + "learning_rate": 1.2810069995925491e-05, + "loss": 1.0439, + "step": 18127 + }, + { + "epoch": 0.43, + "grad_norm": 2.3346371984171395, + "learning_rate": 1.2809337692876573e-05, + "loss": 0.9436, + "step": 18128 + }, + { + "epoch": 0.43, + "grad_norm": 2.4373481466105544, + "learning_rate": 1.2808605373470799e-05, + "loss": 1.0386, + "step": 18129 + }, + { + "epoch": 0.43, + "grad_norm": 2.2134111792540296, + "learning_rate": 1.2807873037712437e-05, + "loss": 0.962, + "step": 18130 + }, + { + "epoch": 0.43, + "grad_norm": 2.0571433021997856, + "learning_rate": 1.2807140685605748e-05, + "loss": 1.0497, + "step": 18131 + }, + { + "epoch": 0.43, + "grad_norm": 2.0142200071516827, + "learning_rate": 1.2806408317154997e-05, + "loss": 1.0492, + "step": 18132 + }, + { + "epoch": 0.43, + "grad_norm": 2.158841640967076, + "learning_rate": 1.2805675932364449e-05, + "loss": 0.9427, + "step": 18133 + }, + { + "epoch": 0.43, + "grad_norm": 2.027324531717664, + "learning_rate": 1.2804943531238366e-05, + "loss": 1.0201, + "step": 18134 + }, + { + "epoch": 0.43, + "grad_norm": 1.8271999988360639, + "learning_rate": 1.2804211113781016e-05, + "loss": 1.0024, + "step": 18135 + }, + { + "epoch": 0.43, + "grad_norm": 2.109125116015674, + "learning_rate": 1.280347867999666e-05, + "loss": 1.0815, + "step": 18136 + }, + { + "epoch": 0.43, + "grad_norm": 1.8675552501930193, + "learning_rate": 1.2802746229889563e-05, + "loss": 1.0821, + "step": 18137 + }, + { + "epoch": 0.43, + "grad_norm": 2.1602663292136817, + "learning_rate": 1.280201376346399e-05, + "loss": 0.9219, + "step": 18138 + }, + { + "epoch": 0.43, + "grad_norm": 2.5255722771976465, + "learning_rate": 1.2801281280724207e-05, + "loss": 1.114, + "step": 18139 + }, + { + "epoch": 0.43, + "grad_norm": 1.8717666953367054, + "learning_rate": 1.2800548781674476e-05, + "loss": 0.9833, + "step": 18140 + }, + { + "epoch": 0.43, + "grad_norm": 1.9612129868526633, + "learning_rate": 1.2799816266319062e-05, + "loss": 1.1563, + "step": 18141 + }, + { + "epoch": 0.43, + "grad_norm": 1.1609307662560324, + "learning_rate": 1.2799083734662237e-05, + "loss": 0.9877, + "step": 18142 + }, + { + "epoch": 0.43, + "grad_norm": 1.7893479182806418, + "learning_rate": 1.2798351186708258e-05, + "loss": 0.9918, + "step": 18143 + }, + { + "epoch": 0.43, + "grad_norm": 2.282367223494669, + "learning_rate": 1.2797618622461387e-05, + "loss": 1.0892, + "step": 18144 + }, + { + "epoch": 0.43, + "grad_norm": 1.8896702049908987, + "learning_rate": 1.2796886041925903e-05, + "loss": 1.1451, + "step": 18145 + }, + { + "epoch": 0.43, + "grad_norm": 2.221951333248657, + "learning_rate": 1.2796153445106057e-05, + "loss": 1.0543, + "step": 18146 + }, + { + "epoch": 0.43, + "grad_norm": 1.9133526578162017, + "learning_rate": 1.279542083200612e-05, + "loss": 1.0335, + "step": 18147 + }, + { + "epoch": 0.43, + "grad_norm": 1.9085959320118346, + "learning_rate": 1.2794688202630361e-05, + "loss": 1.0042, + "step": 18148 + }, + { + "epoch": 0.43, + "grad_norm": 2.085526728356365, + "learning_rate": 1.2793955556983042e-05, + "loss": 0.8906, + "step": 18149 + }, + { + "epoch": 0.43, + "grad_norm": 2.064130370787891, + "learning_rate": 1.2793222895068431e-05, + "loss": 0.916, + "step": 18150 + }, + { + "epoch": 0.43, + "grad_norm": 2.4570695888487717, + "learning_rate": 1.2792490216890786e-05, + "loss": 1.0793, + "step": 18151 + }, + { + "epoch": 0.43, + "grad_norm": 2.2137647221414696, + "learning_rate": 1.2791757522454387e-05, + "loss": 1.1956, + "step": 18152 + }, + { + "epoch": 0.43, + "grad_norm": 1.8459719662678016, + "learning_rate": 1.2791024811763484e-05, + "loss": 0.9067, + "step": 18153 + }, + { + "epoch": 0.43, + "grad_norm": 2.2558763230913295, + "learning_rate": 1.2790292084822355e-05, + "loss": 1.1041, + "step": 18154 + }, + { + "epoch": 0.43, + "grad_norm": 1.9116004023156812, + "learning_rate": 1.278955934163526e-05, + "loss": 0.8919, + "step": 18155 + }, + { + "epoch": 0.43, + "grad_norm": 1.909853265251548, + "learning_rate": 1.2788826582206473e-05, + "loss": 0.9405, + "step": 18156 + }, + { + "epoch": 0.43, + "grad_norm": 2.3780829288211303, + "learning_rate": 1.2788093806540249e-05, + "loss": 1.0056, + "step": 18157 + }, + { + "epoch": 0.43, + "grad_norm": 2.3955960325534433, + "learning_rate": 1.2787361014640862e-05, + "loss": 1.0723, + "step": 18158 + }, + { + "epoch": 0.43, + "grad_norm": 1.9315377635081423, + "learning_rate": 1.2786628206512577e-05, + "loss": 0.9476, + "step": 18159 + }, + { + "epoch": 0.43, + "grad_norm": 2.1867338727129657, + "learning_rate": 1.2785895382159658e-05, + "loss": 1.031, + "step": 18160 + }, + { + "epoch": 0.43, + "grad_norm": 2.0613089812568837, + "learning_rate": 1.2785162541586378e-05, + "loss": 1.1055, + "step": 18161 + }, + { + "epoch": 0.43, + "grad_norm": 2.776593885129656, + "learning_rate": 1.2784429684796997e-05, + "loss": 0.8999, + "step": 18162 + }, + { + "epoch": 0.43, + "grad_norm": 2.0978572856124296, + "learning_rate": 1.2783696811795787e-05, + "loss": 0.8985, + "step": 18163 + }, + { + "epoch": 0.43, + "grad_norm": 2.1526924502485527, + "learning_rate": 1.278296392258701e-05, + "loss": 0.9849, + "step": 18164 + }, + { + "epoch": 0.43, + "grad_norm": 1.8875471078888675, + "learning_rate": 1.2782231017174937e-05, + "loss": 1.1233, + "step": 18165 + }, + { + "epoch": 0.43, + "grad_norm": 2.812224850341656, + "learning_rate": 1.2781498095563835e-05, + "loss": 1.0247, + "step": 18166 + }, + { + "epoch": 0.43, + "grad_norm": 3.0279031667364222, + "learning_rate": 1.2780765157757971e-05, + "loss": 0.9361, + "step": 18167 + }, + { + "epoch": 0.43, + "grad_norm": 2.3309531960349337, + "learning_rate": 1.278003220376161e-05, + "loss": 0.9906, + "step": 18168 + }, + { + "epoch": 0.43, + "grad_norm": 1.680476699661328, + "learning_rate": 1.2779299233579024e-05, + "loss": 0.9552, + "step": 18169 + }, + { + "epoch": 0.43, + "grad_norm": 2.039647960790637, + "learning_rate": 1.2778566247214474e-05, + "loss": 0.8667, + "step": 18170 + }, + { + "epoch": 0.43, + "grad_norm": 1.8501913442622495, + "learning_rate": 1.2777833244672235e-05, + "loss": 0.987, + "step": 18171 + }, + { + "epoch": 0.43, + "grad_norm": 2.0305228362268193, + "learning_rate": 1.277710022595657e-05, + "loss": 1.0064, + "step": 18172 + }, + { + "epoch": 0.43, + "grad_norm": 2.1783616594664905, + "learning_rate": 1.2776367191071752e-05, + "loss": 1.0205, + "step": 18173 + }, + { + "epoch": 0.43, + "grad_norm": 2.3066098594107696, + "learning_rate": 1.2775634140022039e-05, + "loss": 1.1042, + "step": 18174 + }, + { + "epoch": 0.43, + "grad_norm": 1.9405488438754457, + "learning_rate": 1.2774901072811709e-05, + "loss": 1.093, + "step": 18175 + }, + { + "epoch": 0.43, + "grad_norm": 1.0205578744965806, + "learning_rate": 1.2774167989445025e-05, + "loss": 0.9584, + "step": 18176 + }, + { + "epoch": 0.43, + "grad_norm": 1.1121712134409583, + "learning_rate": 1.2773434889926259e-05, + "loss": 0.9763, + "step": 18177 + }, + { + "epoch": 0.43, + "grad_norm": 2.0349759822402453, + "learning_rate": 1.2772701774259674e-05, + "loss": 1.0314, + "step": 18178 + }, + { + "epoch": 0.43, + "grad_norm": 2.4286897774952956, + "learning_rate": 1.2771968642449544e-05, + "loss": 1.089, + "step": 18179 + }, + { + "epoch": 0.43, + "grad_norm": 1.9297323222929679, + "learning_rate": 1.2771235494500133e-05, + "loss": 1.0484, + "step": 18180 + }, + { + "epoch": 0.43, + "grad_norm": 1.9545907955181654, + "learning_rate": 1.2770502330415712e-05, + "loss": 1.0604, + "step": 18181 + }, + { + "epoch": 0.43, + "grad_norm": 1.9407723846057239, + "learning_rate": 1.276976915020055e-05, + "loss": 1.0202, + "step": 18182 + }, + { + "epoch": 0.43, + "grad_norm": 2.303490506855244, + "learning_rate": 1.2769035953858913e-05, + "loss": 1.0021, + "step": 18183 + }, + { + "epoch": 0.43, + "grad_norm": 1.9064050321556134, + "learning_rate": 1.2768302741395076e-05, + "loss": 0.971, + "step": 18184 + }, + { + "epoch": 0.43, + "grad_norm": 1.9482804206213402, + "learning_rate": 1.27675695128133e-05, + "loss": 1.0719, + "step": 18185 + }, + { + "epoch": 0.43, + "grad_norm": 1.0578242896758092, + "learning_rate": 1.2766836268117862e-05, + "loss": 0.9547, + "step": 18186 + }, + { + "epoch": 0.43, + "grad_norm": 2.282642989430193, + "learning_rate": 1.2766103007313024e-05, + "loss": 1.1697, + "step": 18187 + }, + { + "epoch": 0.43, + "grad_norm": 1.9918834947524466, + "learning_rate": 1.2765369730403061e-05, + "loss": 0.9739, + "step": 18188 + }, + { + "epoch": 0.43, + "grad_norm": 1.9754246974980183, + "learning_rate": 1.2764636437392239e-05, + "loss": 1.0972, + "step": 18189 + }, + { + "epoch": 0.43, + "grad_norm": 2.0918458381678255, + "learning_rate": 1.2763903128284827e-05, + "loss": 1.0221, + "step": 18190 + }, + { + "epoch": 0.43, + "grad_norm": 1.9420166477206002, + "learning_rate": 1.2763169803085095e-05, + "loss": 0.8507, + "step": 18191 + }, + { + "epoch": 0.43, + "grad_norm": 2.1378562971407113, + "learning_rate": 1.2762436461797318e-05, + "loss": 0.9914, + "step": 18192 + }, + { + "epoch": 0.43, + "grad_norm": 2.843532596192881, + "learning_rate": 1.2761703104425757e-05, + "loss": 0.9471, + "step": 18193 + }, + { + "epoch": 0.43, + "grad_norm": 1.9784943154892396, + "learning_rate": 1.2760969730974692e-05, + "loss": 0.935, + "step": 18194 + }, + { + "epoch": 0.43, + "grad_norm": 2.123675756718432, + "learning_rate": 1.276023634144838e-05, + "loss": 1.0973, + "step": 18195 + }, + { + "epoch": 0.43, + "grad_norm": 1.8171180065076231, + "learning_rate": 1.2759502935851102e-05, + "loss": 0.9313, + "step": 18196 + }, + { + "epoch": 0.43, + "grad_norm": 3.525868649547711, + "learning_rate": 1.2758769514187123e-05, + "loss": 0.9972, + "step": 18197 + }, + { + "epoch": 0.43, + "grad_norm": 2.1064282399141367, + "learning_rate": 1.2758036076460716e-05, + "loss": 1.1113, + "step": 18198 + }, + { + "epoch": 0.43, + "grad_norm": 1.8627721315370893, + "learning_rate": 1.2757302622676149e-05, + "loss": 1.0973, + "step": 18199 + }, + { + "epoch": 0.43, + "grad_norm": 2.355174915765783, + "learning_rate": 1.2756569152837693e-05, + "loss": 0.9746, + "step": 18200 + }, + { + "epoch": 0.43, + "grad_norm": 2.0329362505625115, + "learning_rate": 1.275583566694962e-05, + "loss": 1.0864, + "step": 18201 + }, + { + "epoch": 0.43, + "grad_norm": 2.2787999407790123, + "learning_rate": 1.2755102165016195e-05, + "loss": 0.9231, + "step": 18202 + }, + { + "epoch": 0.43, + "grad_norm": 1.73346149303028, + "learning_rate": 1.2754368647041699e-05, + "loss": 0.9429, + "step": 18203 + }, + { + "epoch": 0.43, + "grad_norm": 2.1962547635622545, + "learning_rate": 1.275363511303039e-05, + "loss": 1.1272, + "step": 18204 + }, + { + "epoch": 0.43, + "grad_norm": 1.9277437559148847, + "learning_rate": 1.2752901562986553e-05, + "loss": 1.0282, + "step": 18205 + }, + { + "epoch": 0.43, + "grad_norm": 1.1489074468872318, + "learning_rate": 1.2752167996914446e-05, + "loss": 0.9621, + "step": 18206 + }, + { + "epoch": 0.43, + "grad_norm": 2.133988891340515, + "learning_rate": 1.2751434414818349e-05, + "loss": 0.9511, + "step": 18207 + }, + { + "epoch": 0.43, + "grad_norm": 2.0186519419435474, + "learning_rate": 1.2750700816702528e-05, + "loss": 0.9159, + "step": 18208 + }, + { + "epoch": 0.43, + "grad_norm": 2.2620681745912865, + "learning_rate": 1.2749967202571256e-05, + "loss": 1.0022, + "step": 18209 + }, + { + "epoch": 0.43, + "grad_norm": 1.8738804399113145, + "learning_rate": 1.2749233572428805e-05, + "loss": 1.0735, + "step": 18210 + }, + { + "epoch": 0.43, + "grad_norm": 1.0695674327772424, + "learning_rate": 1.2748499926279445e-05, + "loss": 1.0064, + "step": 18211 + }, + { + "epoch": 0.43, + "grad_norm": 2.5398796992924475, + "learning_rate": 1.2747766264127449e-05, + "loss": 0.9458, + "step": 18212 + }, + { + "epoch": 0.43, + "grad_norm": 1.9896069958400135, + "learning_rate": 1.2747032585977089e-05, + "loss": 0.8845, + "step": 18213 + }, + { + "epoch": 0.43, + "grad_norm": 2.0202857324023857, + "learning_rate": 1.2746298891832631e-05, + "loss": 1.0629, + "step": 18214 + }, + { + "epoch": 0.43, + "grad_norm": 1.977017909949913, + "learning_rate": 1.2745565181698357e-05, + "loss": 1.1439, + "step": 18215 + }, + { + "epoch": 0.43, + "grad_norm": 2.1212744616090373, + "learning_rate": 1.2744831455578528e-05, + "loss": 1.1607, + "step": 18216 + }, + { + "epoch": 0.43, + "grad_norm": 2.1540989514680167, + "learning_rate": 1.2744097713477426e-05, + "loss": 1.0031, + "step": 18217 + }, + { + "epoch": 0.43, + "grad_norm": 1.8307037125861718, + "learning_rate": 1.2743363955399315e-05, + "loss": 1.02, + "step": 18218 + }, + { + "epoch": 0.43, + "grad_norm": 1.945949946768189, + "learning_rate": 1.2742630181348473e-05, + "loss": 1.0078, + "step": 18219 + }, + { + "epoch": 0.43, + "grad_norm": 2.1536583169146155, + "learning_rate": 1.2741896391329167e-05, + "loss": 1.016, + "step": 18220 + }, + { + "epoch": 0.43, + "grad_norm": 2.262817772987037, + "learning_rate": 1.2741162585345674e-05, + "loss": 1.0795, + "step": 18221 + }, + { + "epoch": 0.43, + "grad_norm": 1.955097359758929, + "learning_rate": 1.2740428763402266e-05, + "loss": 1.0287, + "step": 18222 + }, + { + "epoch": 0.43, + "grad_norm": 1.7783135311085392, + "learning_rate": 1.2739694925503209e-05, + "loss": 0.9717, + "step": 18223 + }, + { + "epoch": 0.43, + "grad_norm": 1.0693251246728683, + "learning_rate": 1.2738961071652787e-05, + "loss": 0.9873, + "step": 18224 + }, + { + "epoch": 0.43, + "grad_norm": 1.927319910763482, + "learning_rate": 1.2738227201855263e-05, + "loss": 1.0511, + "step": 18225 + }, + { + "epoch": 0.43, + "grad_norm": 2.014587738659323, + "learning_rate": 1.2737493316114916e-05, + "loss": 1.006, + "step": 18226 + }, + { + "epoch": 0.43, + "grad_norm": 1.8688601886481353, + "learning_rate": 1.273675941443601e-05, + "loss": 0.9818, + "step": 18227 + }, + { + "epoch": 0.43, + "grad_norm": 1.8954290387649153, + "learning_rate": 1.2736025496822833e-05, + "loss": 1.0137, + "step": 18228 + }, + { + "epoch": 0.43, + "grad_norm": 1.9798003566350157, + "learning_rate": 1.2735291563279644e-05, + "loss": 1.0548, + "step": 18229 + }, + { + "epoch": 0.43, + "grad_norm": 2.1290609738794335, + "learning_rate": 1.2734557613810723e-05, + "loss": 0.955, + "step": 18230 + }, + { + "epoch": 0.43, + "grad_norm": 1.975495823710891, + "learning_rate": 1.2733823648420341e-05, + "loss": 0.9754, + "step": 18231 + }, + { + "epoch": 0.43, + "grad_norm": 3.0137775885310267, + "learning_rate": 1.273308966711277e-05, + "loss": 0.9526, + "step": 18232 + }, + { + "epoch": 0.43, + "grad_norm": 1.931819749685653, + "learning_rate": 1.2732355669892289e-05, + "loss": 0.9764, + "step": 18233 + }, + { + "epoch": 0.43, + "grad_norm": 2.7664144981595817, + "learning_rate": 1.2731621656763165e-05, + "loss": 0.9472, + "step": 18234 + }, + { + "epoch": 0.43, + "grad_norm": 1.8553215136585348, + "learning_rate": 1.273088762772968e-05, + "loss": 1.0772, + "step": 18235 + }, + { + "epoch": 0.43, + "grad_norm": 2.1827063673477007, + "learning_rate": 1.2730153582796097e-05, + "loss": 0.9854, + "step": 18236 + }, + { + "epoch": 0.43, + "grad_norm": 1.8394195698889344, + "learning_rate": 1.2729419521966702e-05, + "loss": 1.0166, + "step": 18237 + }, + { + "epoch": 0.43, + "grad_norm": 1.1529618902634995, + "learning_rate": 1.2728685445245757e-05, + "loss": 0.8802, + "step": 18238 + }, + { + "epoch": 0.43, + "grad_norm": 1.1102614755202187, + "learning_rate": 1.2727951352637544e-05, + "loss": 1.0103, + "step": 18239 + }, + { + "epoch": 0.43, + "grad_norm": 2.167428831150077, + "learning_rate": 1.2727217244146334e-05, + "loss": 0.9667, + "step": 18240 + }, + { + "epoch": 0.43, + "grad_norm": 2.590940483191415, + "learning_rate": 1.2726483119776401e-05, + "loss": 0.9661, + "step": 18241 + }, + { + "epoch": 0.43, + "grad_norm": 2.040160596317131, + "learning_rate": 1.272574897953202e-05, + "loss": 1.0468, + "step": 18242 + }, + { + "epoch": 0.43, + "grad_norm": 2.055081546312826, + "learning_rate": 1.2725014823417468e-05, + "loss": 1.0296, + "step": 18243 + }, + { + "epoch": 0.43, + "grad_norm": 2.8040256872491334, + "learning_rate": 1.2724280651437014e-05, + "loss": 1.0445, + "step": 18244 + }, + { + "epoch": 0.43, + "grad_norm": 2.089818643258703, + "learning_rate": 1.2723546463594936e-05, + "loss": 0.9702, + "step": 18245 + }, + { + "epoch": 0.43, + "grad_norm": 2.172554739705481, + "learning_rate": 1.272281225989551e-05, + "loss": 1.044, + "step": 18246 + }, + { + "epoch": 0.43, + "grad_norm": 2.0594196816171633, + "learning_rate": 1.2722078040343007e-05, + "loss": 1.0678, + "step": 18247 + }, + { + "epoch": 0.43, + "grad_norm": 1.9727243835419703, + "learning_rate": 1.2721343804941704e-05, + "loss": 0.9873, + "step": 18248 + }, + { + "epoch": 0.43, + "grad_norm": 1.9056483823172785, + "learning_rate": 1.2720609553695877e-05, + "loss": 1.0412, + "step": 18249 + }, + { + "epoch": 0.43, + "grad_norm": 1.8725774393538368, + "learning_rate": 1.27198752866098e-05, + "loss": 1.0584, + "step": 18250 + }, + { + "epoch": 0.43, + "grad_norm": 2.2986413453208145, + "learning_rate": 1.2719141003687746e-05, + "loss": 1.039, + "step": 18251 + }, + { + "epoch": 0.43, + "grad_norm": 2.115681280950428, + "learning_rate": 1.2718406704933997e-05, + "loss": 1.0694, + "step": 18252 + }, + { + "epoch": 0.43, + "grad_norm": 1.9630440991218425, + "learning_rate": 1.2717672390352818e-05, + "loss": 0.9585, + "step": 18253 + }, + { + "epoch": 0.43, + "grad_norm": 1.9302367910151426, + "learning_rate": 1.2716938059948493e-05, + "loss": 0.8958, + "step": 18254 + }, + { + "epoch": 0.43, + "grad_norm": 2.1514765273995877, + "learning_rate": 1.2716203713725293e-05, + "loss": 1.0076, + "step": 18255 + }, + { + "epoch": 0.43, + "grad_norm": 2.02833899194145, + "learning_rate": 1.2715469351687498e-05, + "loss": 0.9748, + "step": 18256 + }, + { + "epoch": 0.43, + "grad_norm": 2.191767808935274, + "learning_rate": 1.2714734973839375e-05, + "loss": 1.0461, + "step": 18257 + }, + { + "epoch": 0.43, + "grad_norm": 2.3222496897011955, + "learning_rate": 1.2714000580185211e-05, + "loss": 0.9387, + "step": 18258 + }, + { + "epoch": 0.43, + "grad_norm": 2.1796784987186877, + "learning_rate": 1.2713266170729275e-05, + "loss": 1.0962, + "step": 18259 + }, + { + "epoch": 0.43, + "grad_norm": 2.6331407077298663, + "learning_rate": 1.2712531745475842e-05, + "loss": 0.9686, + "step": 18260 + }, + { + "epoch": 0.43, + "grad_norm": 1.0559464510566925, + "learning_rate": 1.2711797304429192e-05, + "loss": 0.975, + "step": 18261 + }, + { + "epoch": 0.43, + "grad_norm": 1.9140605273285256, + "learning_rate": 1.27110628475936e-05, + "loss": 0.987, + "step": 18262 + }, + { + "epoch": 0.43, + "grad_norm": 2.0641797702646283, + "learning_rate": 1.2710328374973342e-05, + "loss": 0.9736, + "step": 18263 + }, + { + "epoch": 0.43, + "grad_norm": 1.9796843744795605, + "learning_rate": 1.2709593886572693e-05, + "loss": 0.9933, + "step": 18264 + }, + { + "epoch": 0.43, + "grad_norm": 1.7559178575200638, + "learning_rate": 1.270885938239593e-05, + "loss": 1.0551, + "step": 18265 + }, + { + "epoch": 0.43, + "grad_norm": 1.8714163165321571, + "learning_rate": 1.2708124862447331e-05, + "loss": 1.0078, + "step": 18266 + }, + { + "epoch": 0.43, + "grad_norm": 1.8442618543023772, + "learning_rate": 1.2707390326731173e-05, + "loss": 1.0068, + "step": 18267 + }, + { + "epoch": 0.43, + "grad_norm": 1.8688179405745484, + "learning_rate": 1.2706655775251728e-05, + "loss": 0.8058, + "step": 18268 + }, + { + "epoch": 0.43, + "grad_norm": 2.1348550457998634, + "learning_rate": 1.270592120801328e-05, + "loss": 1.0395, + "step": 18269 + }, + { + "epoch": 0.43, + "grad_norm": 2.0644123290803065, + "learning_rate": 1.2705186625020098e-05, + "loss": 1.0946, + "step": 18270 + }, + { + "epoch": 0.43, + "grad_norm": 1.0962024528864385, + "learning_rate": 1.2704452026276468e-05, + "loss": 0.9453, + "step": 18271 + }, + { + "epoch": 0.43, + "grad_norm": 2.0841810092112136, + "learning_rate": 1.2703717411786658e-05, + "loss": 1.0474, + "step": 18272 + }, + { + "epoch": 0.43, + "grad_norm": 1.9050482729805482, + "learning_rate": 1.2702982781554953e-05, + "loss": 0.9331, + "step": 18273 + }, + { + "epoch": 0.43, + "grad_norm": 1.937767950315187, + "learning_rate": 1.2702248135585622e-05, + "loss": 1.1077, + "step": 18274 + }, + { + "epoch": 0.43, + "grad_norm": 2.67035529826998, + "learning_rate": 1.2701513473882951e-05, + "loss": 0.9438, + "step": 18275 + }, + { + "epoch": 0.43, + "grad_norm": 1.0317977231681053, + "learning_rate": 1.270077879645121e-05, + "loss": 0.9544, + "step": 18276 + }, + { + "epoch": 0.43, + "grad_norm": 2.224632890085522, + "learning_rate": 1.2700044103294683e-05, + "loss": 0.9671, + "step": 18277 + }, + { + "epoch": 0.43, + "grad_norm": 1.1035474186089718, + "learning_rate": 1.2699309394417641e-05, + "loss": 0.9922, + "step": 18278 + }, + { + "epoch": 0.43, + "grad_norm": 1.8341911885365352, + "learning_rate": 1.2698574669824366e-05, + "loss": 0.9837, + "step": 18279 + }, + { + "epoch": 0.43, + "grad_norm": 2.1559461340862325, + "learning_rate": 1.2697839929519136e-05, + "loss": 0.9398, + "step": 18280 + }, + { + "epoch": 0.43, + "grad_norm": 1.9469831106535982, + "learning_rate": 1.2697105173506225e-05, + "loss": 0.9882, + "step": 18281 + }, + { + "epoch": 0.43, + "grad_norm": 2.1329844705340855, + "learning_rate": 1.269637040178992e-05, + "loss": 1.0292, + "step": 18282 + }, + { + "epoch": 0.43, + "grad_norm": 1.8084334087152512, + "learning_rate": 1.2695635614374484e-05, + "loss": 0.9827, + "step": 18283 + }, + { + "epoch": 0.43, + "grad_norm": 1.821115175593286, + "learning_rate": 1.2694900811264212e-05, + "loss": 0.9688, + "step": 18284 + }, + { + "epoch": 0.43, + "grad_norm": 1.9493961448207173, + "learning_rate": 1.269416599246337e-05, + "loss": 1.0229, + "step": 18285 + }, + { + "epoch": 0.43, + "grad_norm": 2.166168400881925, + "learning_rate": 1.2693431157976243e-05, + "loss": 1.0093, + "step": 18286 + }, + { + "epoch": 0.43, + "grad_norm": 1.7586859860046242, + "learning_rate": 1.2692696307807106e-05, + "loss": 1.0962, + "step": 18287 + }, + { + "epoch": 0.43, + "grad_norm": 1.9917963700519228, + "learning_rate": 1.2691961441960238e-05, + "loss": 1.0347, + "step": 18288 + }, + { + "epoch": 0.43, + "grad_norm": 2.159291021145781, + "learning_rate": 1.269122656043992e-05, + "loss": 1.003, + "step": 18289 + }, + { + "epoch": 0.43, + "grad_norm": 2.0466494215267197, + "learning_rate": 1.2690491663250428e-05, + "loss": 1.0886, + "step": 18290 + }, + { + "epoch": 0.43, + "grad_norm": 1.8152737210630951, + "learning_rate": 1.268975675039604e-05, + "loss": 1.0633, + "step": 18291 + }, + { + "epoch": 0.43, + "grad_norm": 1.9861285917997085, + "learning_rate": 1.268902182188104e-05, + "loss": 1.1357, + "step": 18292 + }, + { + "epoch": 0.43, + "grad_norm": 2.466205042011558, + "learning_rate": 1.26882868777097e-05, + "loss": 1.0176, + "step": 18293 + }, + { + "epoch": 0.43, + "grad_norm": 1.9587945655991166, + "learning_rate": 1.2687551917886304e-05, + "loss": 1.0015, + "step": 18294 + }, + { + "epoch": 0.43, + "grad_norm": 2.192472384158706, + "learning_rate": 1.2686816942415133e-05, + "loss": 0.9081, + "step": 18295 + }, + { + "epoch": 0.43, + "grad_norm": 2.0661881073446686, + "learning_rate": 1.268608195130046e-05, + "loss": 1.0163, + "step": 18296 + }, + { + "epoch": 0.43, + "grad_norm": 1.1651552536877539, + "learning_rate": 1.2685346944546566e-05, + "loss": 0.9445, + "step": 18297 + }, + { + "epoch": 0.43, + "grad_norm": 1.088957850996788, + "learning_rate": 1.2684611922157735e-05, + "loss": 0.9592, + "step": 18298 + }, + { + "epoch": 0.43, + "grad_norm": 1.8938977191230708, + "learning_rate": 1.2683876884138242e-05, + "loss": 1.2012, + "step": 18299 + }, + { + "epoch": 0.43, + "grad_norm": 1.9918940608510347, + "learning_rate": 1.2683141830492368e-05, + "loss": 0.9561, + "step": 18300 + }, + { + "epoch": 0.43, + "grad_norm": 3.27340840327298, + "learning_rate": 1.2682406761224394e-05, + "loss": 1.0227, + "step": 18301 + }, + { + "epoch": 0.43, + "grad_norm": 1.952034741564353, + "learning_rate": 1.2681671676338597e-05, + "loss": 1.1541, + "step": 18302 + }, + { + "epoch": 0.43, + "grad_norm": 2.2723068987134316, + "learning_rate": 1.268093657583926e-05, + "loss": 0.9603, + "step": 18303 + }, + { + "epoch": 0.43, + "grad_norm": 2.00538602994786, + "learning_rate": 1.2680201459730661e-05, + "loss": 0.983, + "step": 18304 + }, + { + "epoch": 0.43, + "grad_norm": 2.22533735001748, + "learning_rate": 1.267946632801708e-05, + "loss": 1.0063, + "step": 18305 + }, + { + "epoch": 0.43, + "grad_norm": 2.1417699816980993, + "learning_rate": 1.2678731180702796e-05, + "loss": 0.9319, + "step": 18306 + }, + { + "epoch": 0.43, + "grad_norm": 2.2488740093076847, + "learning_rate": 1.2677996017792095e-05, + "loss": 1.0483, + "step": 18307 + }, + { + "epoch": 0.43, + "grad_norm": 2.1322853423545474, + "learning_rate": 1.2677260839289249e-05, + "loss": 0.9896, + "step": 18308 + }, + { + "epoch": 0.43, + "grad_norm": 1.8520501080458223, + "learning_rate": 1.267652564519855e-05, + "loss": 0.9696, + "step": 18309 + }, + { + "epoch": 0.43, + "grad_norm": 2.161466681049449, + "learning_rate": 1.2675790435524264e-05, + "loss": 1.088, + "step": 18310 + }, + { + "epoch": 0.43, + "grad_norm": 2.0671276986779374, + "learning_rate": 1.2675055210270681e-05, + "loss": 1.1877, + "step": 18311 + }, + { + "epoch": 0.43, + "grad_norm": 1.1609298226436695, + "learning_rate": 1.2674319969442082e-05, + "loss": 0.9182, + "step": 18312 + }, + { + "epoch": 0.43, + "grad_norm": 1.0421804897687033, + "learning_rate": 1.2673584713042743e-05, + "loss": 0.9694, + "step": 18313 + }, + { + "epoch": 0.43, + "grad_norm": 1.0840930914273839, + "learning_rate": 1.2672849441076948e-05, + "loss": 0.952, + "step": 18314 + }, + { + "epoch": 0.43, + "grad_norm": 1.980726418107505, + "learning_rate": 1.2672114153548977e-05, + "loss": 0.9667, + "step": 18315 + }, + { + "epoch": 0.43, + "grad_norm": 2.2054840532357036, + "learning_rate": 1.2671378850463112e-05, + "loss": 0.9766, + "step": 18316 + }, + { + "epoch": 0.43, + "grad_norm": 2.175791568962776, + "learning_rate": 1.2670643531823634e-05, + "loss": 0.9121, + "step": 18317 + }, + { + "epoch": 0.43, + "grad_norm": 1.8716802856760226, + "learning_rate": 1.2669908197634822e-05, + "loss": 0.9265, + "step": 18318 + }, + { + "epoch": 0.43, + "grad_norm": 3.092115557721229, + "learning_rate": 1.2669172847900958e-05, + "loss": 0.9852, + "step": 18319 + }, + { + "epoch": 0.43, + "grad_norm": 1.9131211800417949, + "learning_rate": 1.266843748262633e-05, + "loss": 1.1215, + "step": 18320 + }, + { + "epoch": 0.43, + "grad_norm": 2.250776813740465, + "learning_rate": 1.266770210181521e-05, + "loss": 0.9923, + "step": 18321 + }, + { + "epoch": 0.43, + "grad_norm": 1.8182998879512424, + "learning_rate": 1.2666966705471886e-05, + "loss": 0.9433, + "step": 18322 + }, + { + "epoch": 0.43, + "grad_norm": 2.2461911035254496, + "learning_rate": 1.2666231293600634e-05, + "loss": 0.8802, + "step": 18323 + }, + { + "epoch": 0.43, + "grad_norm": 2.1673432277920894, + "learning_rate": 1.2665495866205745e-05, + "loss": 1.0117, + "step": 18324 + }, + { + "epoch": 0.43, + "grad_norm": 1.9142095239283987, + "learning_rate": 1.2664760423291492e-05, + "loss": 1.0402, + "step": 18325 + }, + { + "epoch": 0.43, + "grad_norm": 1.8623093801018213, + "learning_rate": 1.266402496486216e-05, + "loss": 0.9016, + "step": 18326 + }, + { + "epoch": 0.43, + "grad_norm": 2.1565889025458933, + "learning_rate": 1.2663289490922029e-05, + "loss": 1.1072, + "step": 18327 + }, + { + "epoch": 0.43, + "grad_norm": 1.1233001516675782, + "learning_rate": 1.266255400147539e-05, + "loss": 0.9999, + "step": 18328 + }, + { + "epoch": 0.43, + "grad_norm": 2.444783951058404, + "learning_rate": 1.266181849652651e-05, + "loss": 1.0836, + "step": 18329 + }, + { + "epoch": 0.43, + "grad_norm": 1.8983768563337162, + "learning_rate": 1.2661082976079684e-05, + "loss": 1.0376, + "step": 18330 + }, + { + "epoch": 0.43, + "grad_norm": 1.9838780409519625, + "learning_rate": 1.2660347440139194e-05, + "loss": 0.8525, + "step": 18331 + }, + { + "epoch": 0.43, + "grad_norm": 1.9552396550126099, + "learning_rate": 1.2659611888709313e-05, + "loss": 1.0, + "step": 18332 + }, + { + "epoch": 0.43, + "grad_norm": 1.9150333218437416, + "learning_rate": 1.2658876321794333e-05, + "loss": 1.0451, + "step": 18333 + }, + { + "epoch": 0.43, + "grad_norm": 1.117739131614509, + "learning_rate": 1.2658140739398532e-05, + "loss": 0.9889, + "step": 18334 + }, + { + "epoch": 0.43, + "grad_norm": 2.0509802815948777, + "learning_rate": 1.2657405141526196e-05, + "loss": 1.0237, + "step": 18335 + }, + { + "epoch": 0.43, + "grad_norm": 3.3802105395960904, + "learning_rate": 1.2656669528181603e-05, + "loss": 1.127, + "step": 18336 + }, + { + "epoch": 0.43, + "grad_norm": 2.4086272698800446, + "learning_rate": 1.2655933899369042e-05, + "loss": 1.089, + "step": 18337 + }, + { + "epoch": 0.43, + "grad_norm": 1.9192241659617333, + "learning_rate": 1.2655198255092791e-05, + "loss": 1.1259, + "step": 18338 + }, + { + "epoch": 0.43, + "grad_norm": 1.7283012016700974, + "learning_rate": 1.2654462595357136e-05, + "loss": 1.0833, + "step": 18339 + }, + { + "epoch": 0.43, + "grad_norm": 2.0678205346566365, + "learning_rate": 1.2653726920166358e-05, + "loss": 1.0951, + "step": 18340 + }, + { + "epoch": 0.43, + "grad_norm": 2.11743615971473, + "learning_rate": 1.2652991229524742e-05, + "loss": 1.0827, + "step": 18341 + }, + { + "epoch": 0.43, + "grad_norm": 2.4284968207959574, + "learning_rate": 1.2652255523436572e-05, + "loss": 1.1063, + "step": 18342 + }, + { + "epoch": 0.43, + "grad_norm": 2.3147411399612556, + "learning_rate": 1.2651519801906131e-05, + "loss": 0.9589, + "step": 18343 + }, + { + "epoch": 0.43, + "grad_norm": 2.2169584704834504, + "learning_rate": 1.2650784064937702e-05, + "loss": 1.1177, + "step": 18344 + }, + { + "epoch": 0.43, + "grad_norm": 1.9011367795151117, + "learning_rate": 1.265004831253557e-05, + "loss": 1.1345, + "step": 18345 + }, + { + "epoch": 0.43, + "grad_norm": 2.114089158381548, + "learning_rate": 1.2649312544704016e-05, + "loss": 0.8821, + "step": 18346 + }, + { + "epoch": 0.43, + "grad_norm": 1.8971464713256068, + "learning_rate": 1.2648576761447324e-05, + "loss": 0.9889, + "step": 18347 + }, + { + "epoch": 0.43, + "grad_norm": 2.093882497600166, + "learning_rate": 1.2647840962769781e-05, + "loss": 0.9193, + "step": 18348 + }, + { + "epoch": 0.43, + "grad_norm": 2.0429253826589444, + "learning_rate": 1.2647105148675671e-05, + "loss": 0.9761, + "step": 18349 + }, + { + "epoch": 0.43, + "grad_norm": 2.3553890832888333, + "learning_rate": 1.2646369319169275e-05, + "loss": 0.9708, + "step": 18350 + }, + { + "epoch": 0.43, + "grad_norm": 2.199338577056789, + "learning_rate": 1.2645633474254882e-05, + "loss": 1.1259, + "step": 18351 + }, + { + "epoch": 0.43, + "grad_norm": 3.3952470261817282, + "learning_rate": 1.2644897613936773e-05, + "loss": 0.9936, + "step": 18352 + }, + { + "epoch": 0.43, + "grad_norm": 2.184677849421939, + "learning_rate": 1.2644161738219229e-05, + "loss": 0.9939, + "step": 18353 + }, + { + "epoch": 0.43, + "grad_norm": 1.9363931160203758, + "learning_rate": 1.2643425847106543e-05, + "loss": 1.0456, + "step": 18354 + }, + { + "epoch": 0.43, + "grad_norm": 2.1699125739730043, + "learning_rate": 1.264268994060299e-05, + "loss": 1.0065, + "step": 18355 + }, + { + "epoch": 0.43, + "grad_norm": 2.0163955653691605, + "learning_rate": 1.2641954018712863e-05, + "loss": 1.0939, + "step": 18356 + }, + { + "epoch": 0.43, + "grad_norm": 1.8947470268040536, + "learning_rate": 1.264121808144044e-05, + "loss": 1.0723, + "step": 18357 + }, + { + "epoch": 0.43, + "grad_norm": 2.595770025754575, + "learning_rate": 1.2640482128790012e-05, + "loss": 0.9897, + "step": 18358 + }, + { + "epoch": 0.43, + "grad_norm": 1.8417494074741276, + "learning_rate": 1.2639746160765858e-05, + "loss": 1.0968, + "step": 18359 + }, + { + "epoch": 0.43, + "grad_norm": 1.9555503188529975, + "learning_rate": 1.2639010177372271e-05, + "loss": 1.0434, + "step": 18360 + }, + { + "epoch": 0.43, + "grad_norm": 2.165288167516179, + "learning_rate": 1.2638274178613527e-05, + "loss": 0.9442, + "step": 18361 + }, + { + "epoch": 0.43, + "grad_norm": 2.1414423016104975, + "learning_rate": 1.2637538164493916e-05, + "loss": 1.0445, + "step": 18362 + }, + { + "epoch": 0.43, + "grad_norm": 1.1497200963194876, + "learning_rate": 1.2636802135017726e-05, + "loss": 0.9227, + "step": 18363 + }, + { + "epoch": 0.43, + "grad_norm": 2.0057174885827105, + "learning_rate": 1.2636066090189237e-05, + "loss": 1.0952, + "step": 18364 + }, + { + "epoch": 0.43, + "grad_norm": 2.2769970782331286, + "learning_rate": 1.2635330030012735e-05, + "loss": 1.0504, + "step": 18365 + }, + { + "epoch": 0.43, + "grad_norm": 2.0143343892309655, + "learning_rate": 1.2634593954492509e-05, + "loss": 0.8287, + "step": 18366 + }, + { + "epoch": 0.43, + "grad_norm": 2.0023281463152935, + "learning_rate": 1.2633857863632843e-05, + "loss": 1.0742, + "step": 18367 + }, + { + "epoch": 0.43, + "grad_norm": 2.055145246338264, + "learning_rate": 1.2633121757438022e-05, + "loss": 1.0907, + "step": 18368 + }, + { + "epoch": 0.43, + "grad_norm": 2.1912315378904683, + "learning_rate": 1.2632385635912332e-05, + "loss": 1.047, + "step": 18369 + }, + { + "epoch": 0.43, + "grad_norm": 2.226444817662157, + "learning_rate": 1.2631649499060061e-05, + "loss": 0.9516, + "step": 18370 + }, + { + "epoch": 0.43, + "grad_norm": 1.8638253787005647, + "learning_rate": 1.2630913346885491e-05, + "loss": 0.9235, + "step": 18371 + }, + { + "epoch": 0.43, + "grad_norm": 2.1065653338789567, + "learning_rate": 1.2630177179392913e-05, + "loss": 0.9547, + "step": 18372 + }, + { + "epoch": 0.43, + "grad_norm": 2.1471505192280835, + "learning_rate": 1.262944099658661e-05, + "loss": 1.1431, + "step": 18373 + }, + { + "epoch": 0.43, + "grad_norm": 2.3784064678524146, + "learning_rate": 1.262870479847087e-05, + "loss": 1.1005, + "step": 18374 + }, + { + "epoch": 0.43, + "grad_norm": 2.0445375934230205, + "learning_rate": 1.2627968585049979e-05, + "loss": 1.1073, + "step": 18375 + }, + { + "epoch": 0.43, + "grad_norm": 1.147293883906097, + "learning_rate": 1.262723235632822e-05, + "loss": 0.9375, + "step": 18376 + }, + { + "epoch": 0.43, + "grad_norm": 2.907804977788015, + "learning_rate": 1.2626496112309885e-05, + "loss": 0.9657, + "step": 18377 + }, + { + "epoch": 0.43, + "grad_norm": 2.2044894901090824, + "learning_rate": 1.2625759852999256e-05, + "loss": 1.1368, + "step": 18378 + }, + { + "epoch": 0.43, + "grad_norm": 1.8915996710285818, + "learning_rate": 1.2625023578400622e-05, + "loss": 0.8572, + "step": 18379 + }, + { + "epoch": 0.43, + "grad_norm": 2.201485361050085, + "learning_rate": 1.2624287288518272e-05, + "loss": 1.0488, + "step": 18380 + }, + { + "epoch": 0.43, + "grad_norm": 2.54569687496287, + "learning_rate": 1.2623550983356487e-05, + "loss": 1.084, + "step": 18381 + }, + { + "epoch": 0.43, + "grad_norm": 2.0432297452590866, + "learning_rate": 1.2622814662919562e-05, + "loss": 0.8986, + "step": 18382 + }, + { + "epoch": 0.43, + "grad_norm": 2.1237728343759614, + "learning_rate": 1.2622078327211776e-05, + "loss": 1.0022, + "step": 18383 + }, + { + "epoch": 0.43, + "grad_norm": 2.015588858930987, + "learning_rate": 1.2621341976237423e-05, + "loss": 1.1148, + "step": 18384 + }, + { + "epoch": 0.43, + "grad_norm": 2.0009775934360987, + "learning_rate": 1.2620605610000782e-05, + "loss": 1.0785, + "step": 18385 + }, + { + "epoch": 0.43, + "grad_norm": 1.7713337355002738, + "learning_rate": 1.2619869228506151e-05, + "loss": 1.0054, + "step": 18386 + }, + { + "epoch": 0.43, + "grad_norm": 2.1534168048786473, + "learning_rate": 1.2619132831757808e-05, + "loss": 1.025, + "step": 18387 + }, + { + "epoch": 0.43, + "grad_norm": 2.0680282173340614, + "learning_rate": 1.2618396419760049e-05, + "loss": 1.0236, + "step": 18388 + }, + { + "epoch": 0.43, + "grad_norm": 2.5978267971319897, + "learning_rate": 1.2617659992517153e-05, + "loss": 1.1218, + "step": 18389 + }, + { + "epoch": 0.43, + "grad_norm": 2.0788627219006575, + "learning_rate": 1.2616923550033413e-05, + "loss": 1.0787, + "step": 18390 + }, + { + "epoch": 0.43, + "grad_norm": 1.7425737666430652, + "learning_rate": 1.2616187092313114e-05, + "loss": 0.9059, + "step": 18391 + }, + { + "epoch": 0.43, + "grad_norm": 2.2941914761680393, + "learning_rate": 1.2615450619360547e-05, + "loss": 1.0129, + "step": 18392 + }, + { + "epoch": 0.43, + "grad_norm": 1.9092691773712192, + "learning_rate": 1.2614714131179999e-05, + "loss": 1.065, + "step": 18393 + }, + { + "epoch": 0.43, + "grad_norm": 1.8888215515568891, + "learning_rate": 1.2613977627775757e-05, + "loss": 1.0089, + "step": 18394 + }, + { + "epoch": 0.43, + "grad_norm": 1.9830316485459605, + "learning_rate": 1.261324110915211e-05, + "loss": 0.9693, + "step": 18395 + }, + { + "epoch": 0.43, + "grad_norm": 2.347889646616833, + "learning_rate": 1.2612504575313344e-05, + "loss": 0.9255, + "step": 18396 + }, + { + "epoch": 0.43, + "grad_norm": 1.1386314476109305, + "learning_rate": 1.2611768026263752e-05, + "loss": 1.0049, + "step": 18397 + }, + { + "epoch": 0.43, + "grad_norm": 1.830971379991679, + "learning_rate": 1.2611031462007619e-05, + "loss": 0.9447, + "step": 18398 + }, + { + "epoch": 0.43, + "grad_norm": 2.0679056896571963, + "learning_rate": 1.2610294882549232e-05, + "loss": 0.8463, + "step": 18399 + }, + { + "epoch": 0.43, + "grad_norm": 2.283660724065523, + "learning_rate": 1.2609558287892884e-05, + "loss": 1.0571, + "step": 18400 + }, + { + "epoch": 0.43, + "grad_norm": 2.491495423501937, + "learning_rate": 1.2608821678042862e-05, + "loss": 1.046, + "step": 18401 + }, + { + "epoch": 0.43, + "grad_norm": 2.105537018991724, + "learning_rate": 1.2608085053003452e-05, + "loss": 1.0326, + "step": 18402 + }, + { + "epoch": 0.43, + "grad_norm": 2.12498473425336, + "learning_rate": 1.2607348412778949e-05, + "loss": 0.9572, + "step": 18403 + }, + { + "epoch": 0.43, + "grad_norm": 1.1355841931439938, + "learning_rate": 1.2606611757373633e-05, + "loss": 0.9821, + "step": 18404 + }, + { + "epoch": 0.43, + "grad_norm": 1.887611126292893, + "learning_rate": 1.2605875086791804e-05, + "loss": 1.1122, + "step": 18405 + }, + { + "epoch": 0.43, + "grad_norm": 1.08885932650102, + "learning_rate": 1.2605138401037738e-05, + "loss": 0.9309, + "step": 18406 + }, + { + "epoch": 0.43, + "grad_norm": 1.9523574794463612, + "learning_rate": 1.2604401700115738e-05, + "loss": 0.9764, + "step": 18407 + }, + { + "epoch": 0.43, + "grad_norm": 1.937920657076382, + "learning_rate": 1.2603664984030082e-05, + "loss": 1.0942, + "step": 18408 + }, + { + "epoch": 0.43, + "grad_norm": 2.31824890442351, + "learning_rate": 1.2602928252785068e-05, + "loss": 0.9933, + "step": 18409 + }, + { + "epoch": 0.43, + "grad_norm": 2.144535271409786, + "learning_rate": 1.260219150638498e-05, + "loss": 1.0239, + "step": 18410 + }, + { + "epoch": 0.43, + "grad_norm": 1.8161110880438973, + "learning_rate": 1.2601454744834108e-05, + "loss": 1.0667, + "step": 18411 + }, + { + "epoch": 0.43, + "grad_norm": 1.8081437041700246, + "learning_rate": 1.2600717968136742e-05, + "loss": 0.9068, + "step": 18412 + }, + { + "epoch": 0.43, + "grad_norm": 1.9965659543446903, + "learning_rate": 1.2599981176297176e-05, + "loss": 1.0151, + "step": 18413 + }, + { + "epoch": 0.43, + "grad_norm": 2.550079347985707, + "learning_rate": 1.2599244369319694e-05, + "loss": 1.0192, + "step": 18414 + }, + { + "epoch": 0.43, + "grad_norm": 1.1376159532848642, + "learning_rate": 1.2598507547208589e-05, + "loss": 0.9701, + "step": 18415 + }, + { + "epoch": 0.43, + "grad_norm": 1.9768888540746192, + "learning_rate": 1.259777070996815e-05, + "loss": 0.882, + "step": 18416 + }, + { + "epoch": 0.43, + "grad_norm": 2.119255154180263, + "learning_rate": 1.259703385760267e-05, + "loss": 0.9886, + "step": 18417 + }, + { + "epoch": 0.43, + "grad_norm": 2.0592335001202446, + "learning_rate": 1.2596296990116435e-05, + "loss": 0.8922, + "step": 18418 + }, + { + "epoch": 0.43, + "grad_norm": 2.254433620713832, + "learning_rate": 1.2595560107513735e-05, + "loss": 1.0377, + "step": 18419 + }, + { + "epoch": 0.43, + "grad_norm": 1.1386834169771245, + "learning_rate": 1.2594823209798865e-05, + "loss": 0.9551, + "step": 18420 + }, + { + "epoch": 0.43, + "grad_norm": 2.0571573044061546, + "learning_rate": 1.2594086296976111e-05, + "loss": 0.9491, + "step": 18421 + }, + { + "epoch": 0.43, + "grad_norm": 1.9229936474584466, + "learning_rate": 1.2593349369049767e-05, + "loss": 1.0087, + "step": 18422 + }, + { + "epoch": 0.43, + "grad_norm": 2.177413908297254, + "learning_rate": 1.259261242602412e-05, + "loss": 0.9752, + "step": 18423 + }, + { + "epoch": 0.43, + "grad_norm": 1.7889498324815762, + "learning_rate": 1.2591875467903465e-05, + "loss": 1.0112, + "step": 18424 + }, + { + "epoch": 0.43, + "grad_norm": 2.100935053931762, + "learning_rate": 1.2591138494692087e-05, + "loss": 0.8552, + "step": 18425 + }, + { + "epoch": 0.43, + "grad_norm": 2.266694753280967, + "learning_rate": 1.2590401506394285e-05, + "loss": 1.0045, + "step": 18426 + }, + { + "epoch": 0.43, + "grad_norm": 1.1197194524677678, + "learning_rate": 1.258966450301434e-05, + "loss": 0.9352, + "step": 18427 + }, + { + "epoch": 0.43, + "grad_norm": 2.1896100961009446, + "learning_rate": 1.2588927484556552e-05, + "loss": 1.0376, + "step": 18428 + }, + { + "epoch": 0.43, + "grad_norm": 1.8348508023849726, + "learning_rate": 1.2588190451025209e-05, + "loss": 0.9616, + "step": 18429 + }, + { + "epoch": 0.43, + "grad_norm": 1.9807113951671875, + "learning_rate": 1.2587453402424598e-05, + "loss": 1.1288, + "step": 18430 + }, + { + "epoch": 0.43, + "grad_norm": 3.6109904337155516, + "learning_rate": 1.2586716338759018e-05, + "loss": 1.1646, + "step": 18431 + }, + { + "epoch": 0.43, + "grad_norm": 2.1555796381089265, + "learning_rate": 1.2585979260032754e-05, + "loss": 1.1741, + "step": 18432 + }, + { + "epoch": 0.43, + "grad_norm": 2.1770688836269945, + "learning_rate": 1.2585242166250105e-05, + "loss": 0.9718, + "step": 18433 + }, + { + "epoch": 0.43, + "grad_norm": 2.189200145082059, + "learning_rate": 1.2584505057415352e-05, + "loss": 1.032, + "step": 18434 + }, + { + "epoch": 0.43, + "grad_norm": 2.147554733484006, + "learning_rate": 1.2583767933532795e-05, + "loss": 0.833, + "step": 18435 + }, + { + "epoch": 0.43, + "grad_norm": 1.987924248191203, + "learning_rate": 1.258303079460672e-05, + "loss": 1.0233, + "step": 18436 + }, + { + "epoch": 0.43, + "grad_norm": 2.0369200259596987, + "learning_rate": 1.2582293640641427e-05, + "loss": 1.0871, + "step": 18437 + }, + { + "epoch": 0.43, + "grad_norm": 1.9812577773661295, + "learning_rate": 1.2581556471641199e-05, + "loss": 1.0585, + "step": 18438 + }, + { + "epoch": 0.43, + "grad_norm": 1.9272779194982934, + "learning_rate": 1.2580819287610332e-05, + "loss": 0.9981, + "step": 18439 + }, + { + "epoch": 0.43, + "grad_norm": 1.9387244861134973, + "learning_rate": 1.2580082088553121e-05, + "loss": 0.9193, + "step": 18440 + }, + { + "epoch": 0.43, + "grad_norm": 2.69180030059292, + "learning_rate": 1.2579344874473854e-05, + "loss": 0.9689, + "step": 18441 + }, + { + "epoch": 0.43, + "grad_norm": 2.0694529089422984, + "learning_rate": 1.2578607645376823e-05, + "loss": 0.9246, + "step": 18442 + }, + { + "epoch": 0.43, + "grad_norm": 1.923555305786193, + "learning_rate": 1.2577870401266324e-05, + "loss": 0.9731, + "step": 18443 + }, + { + "epoch": 0.43, + "grad_norm": 1.9538257863959554, + "learning_rate": 1.2577133142146646e-05, + "loss": 1.0097, + "step": 18444 + }, + { + "epoch": 0.43, + "grad_norm": 1.0292881376232226, + "learning_rate": 1.2576395868022082e-05, + "loss": 0.8895, + "step": 18445 + }, + { + "epoch": 0.43, + "grad_norm": 1.9710274996256731, + "learning_rate": 1.2575658578896928e-05, + "loss": 1.0413, + "step": 18446 + }, + { + "epoch": 0.43, + "grad_norm": 1.9287450513357884, + "learning_rate": 1.2574921274775472e-05, + "loss": 1.0315, + "step": 18447 + }, + { + "epoch": 0.43, + "grad_norm": 1.9876329102915868, + "learning_rate": 1.257418395566201e-05, + "loss": 1.06, + "step": 18448 + }, + { + "epoch": 0.43, + "grad_norm": 1.8688213206823312, + "learning_rate": 1.2573446621560833e-05, + "loss": 1.0685, + "step": 18449 + }, + { + "epoch": 0.43, + "grad_norm": 2.0390136217784938, + "learning_rate": 1.2572709272476237e-05, + "loss": 1.0246, + "step": 18450 + }, + { + "epoch": 0.43, + "grad_norm": 1.9510450781374902, + "learning_rate": 1.2571971908412512e-05, + "loss": 0.944, + "step": 18451 + }, + { + "epoch": 0.43, + "grad_norm": 1.1477059660703028, + "learning_rate": 1.2571234529373953e-05, + "loss": 0.9604, + "step": 18452 + }, + { + "epoch": 0.43, + "grad_norm": 2.06512466296498, + "learning_rate": 1.2570497135364851e-05, + "loss": 0.9478, + "step": 18453 + }, + { + "epoch": 0.43, + "grad_norm": 2.196876882391031, + "learning_rate": 1.2569759726389505e-05, + "loss": 0.9318, + "step": 18454 + }, + { + "epoch": 0.43, + "grad_norm": 1.2107922111151812, + "learning_rate": 1.2569022302452199e-05, + "loss": 0.9575, + "step": 18455 + }, + { + "epoch": 0.43, + "grad_norm": 1.0664498373527858, + "learning_rate": 1.2568284863557236e-05, + "loss": 0.9351, + "step": 18456 + }, + { + "epoch": 0.43, + "grad_norm": 2.330014719184314, + "learning_rate": 1.25675474097089e-05, + "loss": 0.9765, + "step": 18457 + }, + { + "epoch": 0.43, + "grad_norm": 1.1504003550882225, + "learning_rate": 1.2566809940911497e-05, + "loss": 0.9654, + "step": 18458 + }, + { + "epoch": 0.43, + "grad_norm": 2.4762349243008437, + "learning_rate": 1.2566072457169308e-05, + "loss": 1.1126, + "step": 18459 + }, + { + "epoch": 0.43, + "grad_norm": 1.98970324721747, + "learning_rate": 1.2565334958486636e-05, + "loss": 0.9791, + "step": 18460 + }, + { + "epoch": 0.43, + "grad_norm": 1.1719774741827178, + "learning_rate": 1.2564597444867772e-05, + "loss": 0.9287, + "step": 18461 + }, + { + "epoch": 0.43, + "grad_norm": 2.027627391533761, + "learning_rate": 1.256385991631701e-05, + "loss": 1.0212, + "step": 18462 + }, + { + "epoch": 0.43, + "grad_norm": 2.0865429852680593, + "learning_rate": 1.2563122372838642e-05, + "loss": 0.9857, + "step": 18463 + }, + { + "epoch": 0.43, + "grad_norm": 2.3303724354478446, + "learning_rate": 1.2562384814436961e-05, + "loss": 0.8886, + "step": 18464 + }, + { + "epoch": 0.44, + "grad_norm": 1.9154380807442777, + "learning_rate": 1.256164724111627e-05, + "loss": 0.8002, + "step": 18465 + }, + { + "epoch": 0.44, + "grad_norm": 1.8952376952203767, + "learning_rate": 1.2560909652880854e-05, + "loss": 0.9829, + "step": 18466 + }, + { + "epoch": 0.44, + "grad_norm": 2.0112694449189354, + "learning_rate": 1.2560172049735016e-05, + "loss": 0.9485, + "step": 18467 + }, + { + "epoch": 0.44, + "grad_norm": 1.0449745578274998, + "learning_rate": 1.2559434431683042e-05, + "loss": 0.9375, + "step": 18468 + }, + { + "epoch": 0.44, + "grad_norm": 1.8833001040718815, + "learning_rate": 1.2558696798729231e-05, + "loss": 0.9697, + "step": 18469 + }, + { + "epoch": 0.44, + "grad_norm": 2.0938064924868316, + "learning_rate": 1.255795915087788e-05, + "loss": 1.0184, + "step": 18470 + }, + { + "epoch": 0.44, + "grad_norm": 2.7155808996022053, + "learning_rate": 1.2557221488133276e-05, + "loss": 0.9867, + "step": 18471 + }, + { + "epoch": 0.44, + "grad_norm": 1.942607022703747, + "learning_rate": 1.255648381049972e-05, + "loss": 0.9375, + "step": 18472 + }, + { + "epoch": 0.44, + "grad_norm": 2.248316744749226, + "learning_rate": 1.2555746117981506e-05, + "loss": 0.8739, + "step": 18473 + }, + { + "epoch": 0.44, + "grad_norm": 2.0073133921627035, + "learning_rate": 1.2555008410582932e-05, + "loss": 1.0691, + "step": 18474 + }, + { + "epoch": 0.44, + "grad_norm": 2.404542605081669, + "learning_rate": 1.255427068830829e-05, + "loss": 0.9838, + "step": 18475 + }, + { + "epoch": 0.44, + "grad_norm": 1.8270597362446679, + "learning_rate": 1.255353295116187e-05, + "loss": 1.0535, + "step": 18476 + }, + { + "epoch": 0.44, + "grad_norm": 1.9384352399159501, + "learning_rate": 1.2552795199147979e-05, + "loss": 1.0098, + "step": 18477 + }, + { + "epoch": 0.44, + "grad_norm": 2.0336812410638654, + "learning_rate": 1.2552057432270903e-05, + "loss": 1.0052, + "step": 18478 + }, + { + "epoch": 0.44, + "grad_norm": 2.0151862613137563, + "learning_rate": 1.255131965053494e-05, + "loss": 1.0762, + "step": 18479 + }, + { + "epoch": 0.44, + "grad_norm": 1.9396795242007159, + "learning_rate": 1.2550581853944388e-05, + "loss": 0.88, + "step": 18480 + }, + { + "epoch": 0.44, + "grad_norm": 1.125341190854885, + "learning_rate": 1.254984404250354e-05, + "loss": 0.9477, + "step": 18481 + }, + { + "epoch": 0.44, + "grad_norm": 2.108756494685749, + "learning_rate": 1.2549106216216693e-05, + "loss": 1.1124, + "step": 18482 + }, + { + "epoch": 0.44, + "grad_norm": 2.076721965108555, + "learning_rate": 1.2548368375088143e-05, + "loss": 1.0918, + "step": 18483 + }, + { + "epoch": 0.44, + "grad_norm": 2.3232589119168514, + "learning_rate": 1.2547630519122187e-05, + "loss": 1.0307, + "step": 18484 + }, + { + "epoch": 0.44, + "grad_norm": 2.074497568187013, + "learning_rate": 1.2546892648323114e-05, + "loss": 1.071, + "step": 18485 + }, + { + "epoch": 0.44, + "grad_norm": 1.8187006272023711, + "learning_rate": 1.2546154762695233e-05, + "loss": 1.0712, + "step": 18486 + }, + { + "epoch": 0.44, + "grad_norm": 1.930200044183685, + "learning_rate": 1.2545416862242827e-05, + "loss": 1.092, + "step": 18487 + }, + { + "epoch": 0.44, + "grad_norm": 1.8926351251275908, + "learning_rate": 1.2544678946970201e-05, + "loss": 0.8832, + "step": 18488 + }, + { + "epoch": 0.44, + "grad_norm": 2.4150031736813933, + "learning_rate": 1.2543941016881647e-05, + "loss": 1.0971, + "step": 18489 + }, + { + "epoch": 0.44, + "grad_norm": 2.185670688290768, + "learning_rate": 1.2543203071981467e-05, + "loss": 0.9606, + "step": 18490 + }, + { + "epoch": 0.44, + "grad_norm": 1.8285944448756346, + "learning_rate": 1.254246511227395e-05, + "loss": 0.9157, + "step": 18491 + }, + { + "epoch": 0.44, + "grad_norm": 1.9801571690726498, + "learning_rate": 1.2541727137763395e-05, + "loss": 0.9377, + "step": 18492 + }, + { + "epoch": 0.44, + "grad_norm": 1.8366094414730856, + "learning_rate": 1.2540989148454102e-05, + "loss": 0.9849, + "step": 18493 + }, + { + "epoch": 0.44, + "grad_norm": 1.8113849387088106, + "learning_rate": 1.2540251144350366e-05, + "loss": 0.8758, + "step": 18494 + }, + { + "epoch": 0.44, + "grad_norm": 2.0172111939444823, + "learning_rate": 1.2539513125456484e-05, + "loss": 1.1399, + "step": 18495 + }, + { + "epoch": 0.44, + "grad_norm": 2.0670472740195924, + "learning_rate": 1.253877509177675e-05, + "loss": 0.9942, + "step": 18496 + }, + { + "epoch": 0.44, + "grad_norm": 2.370221032774284, + "learning_rate": 1.2538037043315465e-05, + "loss": 0.9649, + "step": 18497 + }, + { + "epoch": 0.44, + "grad_norm": 2.3561066683749576, + "learning_rate": 1.2537298980076926e-05, + "loss": 1.0275, + "step": 18498 + }, + { + "epoch": 0.44, + "grad_norm": 1.9592156490604589, + "learning_rate": 1.2536560902065427e-05, + "loss": 1.0113, + "step": 18499 + }, + { + "epoch": 0.44, + "grad_norm": 1.9703332125075546, + "learning_rate": 1.253582280928527e-05, + "loss": 0.949, + "step": 18500 + }, + { + "epoch": 0.44, + "grad_norm": 1.949271119323, + "learning_rate": 1.2535084701740748e-05, + "loss": 1.0873, + "step": 18501 + }, + { + "epoch": 0.44, + "grad_norm": 2.3652221880582274, + "learning_rate": 1.2534346579436158e-05, + "loss": 0.8581, + "step": 18502 + }, + { + "epoch": 0.44, + "grad_norm": 1.7973256700503524, + "learning_rate": 1.2533608442375803e-05, + "loss": 0.9352, + "step": 18503 + }, + { + "epoch": 0.44, + "grad_norm": 2.302133492196915, + "learning_rate": 1.2532870290563976e-05, + "loss": 1.0732, + "step": 18504 + }, + { + "epoch": 0.44, + "grad_norm": 1.822852326908174, + "learning_rate": 1.253213212400498e-05, + "loss": 1.0925, + "step": 18505 + }, + { + "epoch": 0.44, + "grad_norm": 1.8560803677756228, + "learning_rate": 1.2531393942703104e-05, + "loss": 1.0055, + "step": 18506 + }, + { + "epoch": 0.44, + "grad_norm": 2.281564432635989, + "learning_rate": 1.2530655746662657e-05, + "loss": 1.0173, + "step": 18507 + }, + { + "epoch": 0.44, + "grad_norm": 2.026080127513836, + "learning_rate": 1.2529917535887923e-05, + "loss": 1.1552, + "step": 18508 + }, + { + "epoch": 0.44, + "grad_norm": 1.9605372010104642, + "learning_rate": 1.2529179310383215e-05, + "loss": 1.0966, + "step": 18509 + }, + { + "epoch": 0.44, + "grad_norm": 1.7711039520299228, + "learning_rate": 1.252844107015282e-05, + "loss": 0.9888, + "step": 18510 + }, + { + "epoch": 0.44, + "grad_norm": 1.2243615718037524, + "learning_rate": 1.2527702815201043e-05, + "loss": 0.9649, + "step": 18511 + }, + { + "epoch": 0.44, + "grad_norm": 2.011075099501525, + "learning_rate": 1.2526964545532182e-05, + "loss": 1.0782, + "step": 18512 + }, + { + "epoch": 0.44, + "grad_norm": 1.1324013508561548, + "learning_rate": 1.2526226261150527e-05, + "loss": 0.9809, + "step": 18513 + }, + { + "epoch": 0.44, + "grad_norm": 1.963711822070054, + "learning_rate": 1.2525487962060389e-05, + "loss": 1.07, + "step": 18514 + }, + { + "epoch": 0.44, + "grad_norm": 1.2257282788040598, + "learning_rate": 1.2524749648266054e-05, + "loss": 1.0413, + "step": 18515 + }, + { + "epoch": 0.44, + "grad_norm": 2.4800767985001064, + "learning_rate": 1.2524011319771833e-05, + "loss": 0.9002, + "step": 18516 + }, + { + "epoch": 0.44, + "grad_norm": 2.044088286312606, + "learning_rate": 1.2523272976582018e-05, + "loss": 1.0498, + "step": 18517 + }, + { + "epoch": 0.44, + "grad_norm": 2.0803781338381753, + "learning_rate": 1.252253461870091e-05, + "loss": 1.1125, + "step": 18518 + }, + { + "epoch": 0.44, + "grad_norm": 1.9387366583393038, + "learning_rate": 1.2521796246132804e-05, + "loss": 0.9429, + "step": 18519 + }, + { + "epoch": 0.44, + "grad_norm": 1.7646690271397631, + "learning_rate": 1.2521057858882002e-05, + "loss": 1.0122, + "step": 18520 + }, + { + "epoch": 0.44, + "grad_norm": 1.089434905299906, + "learning_rate": 1.2520319456952806e-05, + "loss": 0.972, + "step": 18521 + }, + { + "epoch": 0.44, + "grad_norm": 2.0604468573346044, + "learning_rate": 1.2519581040349508e-05, + "loss": 1.0238, + "step": 18522 + }, + { + "epoch": 0.44, + "grad_norm": 2.1428618605085825, + "learning_rate": 1.2518842609076414e-05, + "loss": 0.8926, + "step": 18523 + }, + { + "epoch": 0.44, + "grad_norm": 2.181732193101803, + "learning_rate": 1.2518104163137821e-05, + "loss": 1.0413, + "step": 18524 + }, + { + "epoch": 0.44, + "grad_norm": 2.126497872377491, + "learning_rate": 1.2517365702538026e-05, + "loss": 1.0675, + "step": 18525 + }, + { + "epoch": 0.44, + "grad_norm": 1.0817326459043068, + "learning_rate": 1.2516627227281336e-05, + "loss": 0.9366, + "step": 18526 + }, + { + "epoch": 0.44, + "grad_norm": 1.9592735205142775, + "learning_rate": 1.2515888737372042e-05, + "loss": 1.0299, + "step": 18527 + }, + { + "epoch": 0.44, + "grad_norm": 2.044542143809639, + "learning_rate": 1.2515150232814446e-05, + "loss": 1.107, + "step": 18528 + }, + { + "epoch": 0.44, + "grad_norm": 2.2718329960586536, + "learning_rate": 1.251441171361285e-05, + "loss": 1.0567, + "step": 18529 + }, + { + "epoch": 0.44, + "grad_norm": 2.1438339660818033, + "learning_rate": 1.2513673179771555e-05, + "loss": 0.9125, + "step": 18530 + }, + { + "epoch": 0.44, + "grad_norm": 1.9944022463087188, + "learning_rate": 1.2512934631294858e-05, + "loss": 0.9788, + "step": 18531 + }, + { + "epoch": 0.44, + "grad_norm": 2.1023038690731246, + "learning_rate": 1.251219606818706e-05, + "loss": 0.8433, + "step": 18532 + }, + { + "epoch": 0.44, + "grad_norm": 2.4562362545958325, + "learning_rate": 1.251145749045246e-05, + "loss": 1.0011, + "step": 18533 + }, + { + "epoch": 0.44, + "grad_norm": 1.7758528555678688, + "learning_rate": 1.2510718898095359e-05, + "loss": 0.905, + "step": 18534 + }, + { + "epoch": 0.44, + "grad_norm": 2.240864546920333, + "learning_rate": 1.2509980291120061e-05, + "loss": 1.0079, + "step": 18535 + }, + { + "epoch": 0.44, + "grad_norm": 2.1047299410085136, + "learning_rate": 1.2509241669530858e-05, + "loss": 0.9913, + "step": 18536 + }, + { + "epoch": 0.44, + "grad_norm": 2.2435371346516404, + "learning_rate": 1.250850303333206e-05, + "loss": 0.9774, + "step": 18537 + }, + { + "epoch": 0.44, + "grad_norm": 1.1569509540181757, + "learning_rate": 1.250776438252796e-05, + "loss": 1.0406, + "step": 18538 + }, + { + "epoch": 0.44, + "grad_norm": 2.2693192566415266, + "learning_rate": 1.2507025717122865e-05, + "loss": 0.9955, + "step": 18539 + }, + { + "epoch": 0.44, + "grad_norm": 1.8897103767076209, + "learning_rate": 1.2506287037121072e-05, + "loss": 0.8493, + "step": 18540 + }, + { + "epoch": 0.44, + "grad_norm": 2.1708946306287067, + "learning_rate": 1.250554834252688e-05, + "loss": 0.9436, + "step": 18541 + }, + { + "epoch": 0.44, + "grad_norm": 1.912517848332807, + "learning_rate": 1.2504809633344594e-05, + "loss": 1.0217, + "step": 18542 + }, + { + "epoch": 0.44, + "grad_norm": 2.065534387915287, + "learning_rate": 1.2504070909578515e-05, + "loss": 0.9607, + "step": 18543 + }, + { + "epoch": 0.44, + "grad_norm": 2.028907095816586, + "learning_rate": 1.2503332171232938e-05, + "loss": 1.0627, + "step": 18544 + }, + { + "epoch": 0.44, + "grad_norm": 1.9576245529765461, + "learning_rate": 1.250259341831217e-05, + "loss": 0.906, + "step": 18545 + }, + { + "epoch": 0.44, + "grad_norm": 2.1100279088979663, + "learning_rate": 1.2501854650820513e-05, + "loss": 1.0984, + "step": 18546 + }, + { + "epoch": 0.44, + "grad_norm": 1.968571236371767, + "learning_rate": 1.2501115868762264e-05, + "loss": 0.9934, + "step": 18547 + }, + { + "epoch": 0.44, + "grad_norm": 3.6355612731199467, + "learning_rate": 1.2500377072141727e-05, + "loss": 1.0982, + "step": 18548 + }, + { + "epoch": 0.44, + "grad_norm": 1.0867527229766751, + "learning_rate": 1.2499638260963204e-05, + "loss": 1.0354, + "step": 18549 + }, + { + "epoch": 0.44, + "grad_norm": 2.2453830601182947, + "learning_rate": 1.2498899435230996e-05, + "loss": 0.9481, + "step": 18550 + }, + { + "epoch": 0.44, + "grad_norm": 2.059098949344779, + "learning_rate": 1.2498160594949402e-05, + "loss": 1.0397, + "step": 18551 + }, + { + "epoch": 0.44, + "grad_norm": 2.14061795554603, + "learning_rate": 1.2497421740122728e-05, + "loss": 1.0647, + "step": 18552 + }, + { + "epoch": 0.44, + "grad_norm": 2.02295902559655, + "learning_rate": 1.2496682870755275e-05, + "loss": 1.0807, + "step": 18553 + }, + { + "epoch": 0.44, + "grad_norm": 1.9362981277327416, + "learning_rate": 1.2495943986851342e-05, + "loss": 0.9783, + "step": 18554 + }, + { + "epoch": 0.44, + "grad_norm": 1.7291024068107297, + "learning_rate": 1.2495205088415233e-05, + "loss": 1.0227, + "step": 18555 + }, + { + "epoch": 0.44, + "grad_norm": 2.090914796728203, + "learning_rate": 1.2494466175451253e-05, + "loss": 1.107, + "step": 18556 + }, + { + "epoch": 0.44, + "grad_norm": 2.086030139410782, + "learning_rate": 1.2493727247963695e-05, + "loss": 0.8865, + "step": 18557 + }, + { + "epoch": 0.44, + "grad_norm": 2.13239564316314, + "learning_rate": 1.2492988305956875e-05, + "loss": 1.0087, + "step": 18558 + }, + { + "epoch": 0.44, + "grad_norm": 2.1370045120620835, + "learning_rate": 1.249224934943508e-05, + "loss": 1.0414, + "step": 18559 + }, + { + "epoch": 0.44, + "grad_norm": 1.9643956606856834, + "learning_rate": 1.2491510378402628e-05, + "loss": 1.0114, + "step": 18560 + }, + { + "epoch": 0.44, + "grad_norm": 2.0884874744659987, + "learning_rate": 1.2490771392863805e-05, + "loss": 0.9873, + "step": 18561 + }, + { + "epoch": 0.44, + "grad_norm": 1.875109600114826, + "learning_rate": 1.2490032392822929e-05, + "loss": 1.1123, + "step": 18562 + }, + { + "epoch": 0.44, + "grad_norm": 1.9096400826558335, + "learning_rate": 1.2489293378284298e-05, + "loss": 0.9315, + "step": 18563 + }, + { + "epoch": 0.44, + "grad_norm": 1.777207223657999, + "learning_rate": 1.2488554349252205e-05, + "loss": 0.9969, + "step": 18564 + }, + { + "epoch": 0.44, + "grad_norm": 1.9134097866984712, + "learning_rate": 1.2487815305730967e-05, + "loss": 1.0368, + "step": 18565 + }, + { + "epoch": 0.44, + "grad_norm": 1.0798153058454125, + "learning_rate": 1.2487076247724876e-05, + "loss": 1.0252, + "step": 18566 + }, + { + "epoch": 0.44, + "grad_norm": 2.316950650032945, + "learning_rate": 1.2486337175238246e-05, + "loss": 0.9784, + "step": 18567 + }, + { + "epoch": 0.44, + "grad_norm": 2.0350212888828225, + "learning_rate": 1.2485598088275367e-05, + "loss": 1.085, + "step": 18568 + }, + { + "epoch": 0.44, + "grad_norm": 1.859678855554234, + "learning_rate": 1.2484858986840553e-05, + "loss": 0.9643, + "step": 18569 + }, + { + "epoch": 0.44, + "grad_norm": 1.9324485514644203, + "learning_rate": 1.2484119870938102e-05, + "loss": 1.0781, + "step": 18570 + }, + { + "epoch": 0.44, + "grad_norm": 2.108850013893086, + "learning_rate": 1.2483380740572319e-05, + "loss": 0.9722, + "step": 18571 + }, + { + "epoch": 0.44, + "grad_norm": 2.1704609662749963, + "learning_rate": 1.2482641595747508e-05, + "loss": 0.9694, + "step": 18572 + }, + { + "epoch": 0.44, + "grad_norm": 1.11989525383553, + "learning_rate": 1.2481902436467972e-05, + "loss": 0.9852, + "step": 18573 + }, + { + "epoch": 0.44, + "grad_norm": 1.9921917570526788, + "learning_rate": 1.2481163262738011e-05, + "loss": 0.9431, + "step": 18574 + }, + { + "epoch": 0.44, + "grad_norm": 2.0449902330411756, + "learning_rate": 1.2480424074561934e-05, + "loss": 0.9898, + "step": 18575 + }, + { + "epoch": 0.44, + "grad_norm": 2.0602633689132905, + "learning_rate": 1.2479684871944044e-05, + "loss": 1.0701, + "step": 18576 + }, + { + "epoch": 0.44, + "grad_norm": 1.8643968958900472, + "learning_rate": 1.247894565488864e-05, + "loss": 1.1281, + "step": 18577 + }, + { + "epoch": 0.44, + "grad_norm": 1.845929566941656, + "learning_rate": 1.2478206423400034e-05, + "loss": 1.1031, + "step": 18578 + }, + { + "epoch": 0.44, + "grad_norm": 2.1819190284827794, + "learning_rate": 1.2477467177482524e-05, + "loss": 0.994, + "step": 18579 + }, + { + "epoch": 0.44, + "grad_norm": 1.7608858062906572, + "learning_rate": 1.2476727917140415e-05, + "loss": 0.9047, + "step": 18580 + }, + { + "epoch": 0.44, + "grad_norm": 1.8925455990379065, + "learning_rate": 1.2475988642378011e-05, + "loss": 0.9771, + "step": 18581 + }, + { + "epoch": 0.44, + "grad_norm": 1.8650982477534854, + "learning_rate": 1.247524935319962e-05, + "loss": 1.0031, + "step": 18582 + }, + { + "epoch": 0.44, + "grad_norm": 2.091085098708048, + "learning_rate": 1.247451004960954e-05, + "loss": 1.0511, + "step": 18583 + }, + { + "epoch": 0.44, + "grad_norm": 1.936356291528559, + "learning_rate": 1.2473770731612084e-05, + "loss": 0.9935, + "step": 18584 + }, + { + "epoch": 0.44, + "grad_norm": 2.025954660225364, + "learning_rate": 1.2473031399211548e-05, + "loss": 0.9446, + "step": 18585 + }, + { + "epoch": 0.44, + "grad_norm": 2.306751326085668, + "learning_rate": 1.2472292052412243e-05, + "loss": 1.0298, + "step": 18586 + }, + { + "epoch": 0.44, + "grad_norm": 1.7637800774117207, + "learning_rate": 1.2471552691218465e-05, + "loss": 0.9548, + "step": 18587 + }, + { + "epoch": 0.44, + "grad_norm": 1.961652873878487, + "learning_rate": 1.247081331563453e-05, + "loss": 1.065, + "step": 18588 + }, + { + "epoch": 0.44, + "grad_norm": 2.2711090969365, + "learning_rate": 1.2470073925664736e-05, + "loss": 0.9698, + "step": 18589 + }, + { + "epoch": 0.44, + "grad_norm": 2.0289382068027235, + "learning_rate": 1.2469334521313392e-05, + "loss": 0.8971, + "step": 18590 + }, + { + "epoch": 0.44, + "grad_norm": 2.233606576707602, + "learning_rate": 1.2468595102584798e-05, + "loss": 1.0577, + "step": 18591 + }, + { + "epoch": 0.44, + "grad_norm": 1.1710149804053143, + "learning_rate": 1.2467855669483262e-05, + "loss": 1.0215, + "step": 18592 + }, + { + "epoch": 0.44, + "grad_norm": 1.9719005100184053, + "learning_rate": 1.2467116222013089e-05, + "loss": 0.9582, + "step": 18593 + }, + { + "epoch": 0.44, + "grad_norm": 2.006627842052253, + "learning_rate": 1.2466376760178584e-05, + "loss": 1.103, + "step": 18594 + }, + { + "epoch": 0.44, + "grad_norm": 1.6435669664359445, + "learning_rate": 1.2465637283984056e-05, + "loss": 0.898, + "step": 18595 + }, + { + "epoch": 0.44, + "grad_norm": 2.253385892517826, + "learning_rate": 1.2464897793433803e-05, + "loss": 0.9412, + "step": 18596 + }, + { + "epoch": 0.44, + "grad_norm": 2.311973240339815, + "learning_rate": 1.2464158288532136e-05, + "loss": 1.0072, + "step": 18597 + }, + { + "epoch": 0.44, + "grad_norm": 2.2492698470468278, + "learning_rate": 1.246341876928336e-05, + "loss": 0.9811, + "step": 18598 + }, + { + "epoch": 0.44, + "grad_norm": 1.937025339152263, + "learning_rate": 1.2462679235691779e-05, + "loss": 0.9029, + "step": 18599 + }, + { + "epoch": 0.44, + "grad_norm": 2.0079028820315914, + "learning_rate": 1.24619396877617e-05, + "loss": 0.8823, + "step": 18600 + }, + { + "epoch": 0.44, + "grad_norm": 1.8673312845592873, + "learning_rate": 1.2461200125497428e-05, + "loss": 0.9968, + "step": 18601 + }, + { + "epoch": 0.44, + "grad_norm": 1.7516744928586228, + "learning_rate": 1.2460460548903273e-05, + "loss": 0.9899, + "step": 18602 + }, + { + "epoch": 0.44, + "grad_norm": 2.0197886240952276, + "learning_rate": 1.2459720957983535e-05, + "loss": 0.9271, + "step": 18603 + }, + { + "epoch": 0.44, + "grad_norm": 2.395873690062051, + "learning_rate": 1.2458981352742522e-05, + "loss": 1.0498, + "step": 18604 + }, + { + "epoch": 0.44, + "grad_norm": 5.582622072151782, + "learning_rate": 1.2458241733184542e-05, + "loss": 1.0163, + "step": 18605 + }, + { + "epoch": 0.44, + "grad_norm": 1.9966960995327157, + "learning_rate": 1.2457502099313901e-05, + "loss": 1.0101, + "step": 18606 + }, + { + "epoch": 0.44, + "grad_norm": 1.850579059205571, + "learning_rate": 1.2456762451134905e-05, + "loss": 0.9869, + "step": 18607 + }, + { + "epoch": 0.44, + "grad_norm": 2.0878563374950407, + "learning_rate": 1.2456022788651858e-05, + "loss": 0.8879, + "step": 18608 + }, + { + "epoch": 0.44, + "grad_norm": 1.9132070577279934, + "learning_rate": 1.2455283111869072e-05, + "loss": 0.8916, + "step": 18609 + }, + { + "epoch": 0.44, + "grad_norm": 2.2041741484591646, + "learning_rate": 1.2454543420790847e-05, + "loss": 1.0463, + "step": 18610 + }, + { + "epoch": 0.44, + "grad_norm": 2.3815955463127683, + "learning_rate": 1.2453803715421496e-05, + "loss": 0.8993, + "step": 18611 + }, + { + "epoch": 0.44, + "grad_norm": 2.046082916341236, + "learning_rate": 1.245306399576532e-05, + "loss": 1.0017, + "step": 18612 + }, + { + "epoch": 0.44, + "grad_norm": 2.0321235816248224, + "learning_rate": 1.2452324261826632e-05, + "loss": 1.1016, + "step": 18613 + }, + { + "epoch": 0.44, + "grad_norm": 2.084185492482724, + "learning_rate": 1.2451584513609736e-05, + "loss": 1.0016, + "step": 18614 + }, + { + "epoch": 0.44, + "grad_norm": 2.1047075699077276, + "learning_rate": 1.2450844751118934e-05, + "loss": 1.0946, + "step": 18615 + }, + { + "epoch": 0.44, + "grad_norm": 1.1980736414970006, + "learning_rate": 1.2450104974358544e-05, + "loss": 0.9743, + "step": 18616 + }, + { + "epoch": 0.44, + "grad_norm": 2.085894532645165, + "learning_rate": 1.2449365183332862e-05, + "loss": 1.0676, + "step": 18617 + }, + { + "epoch": 0.44, + "grad_norm": 2.020326671251926, + "learning_rate": 1.2448625378046205e-05, + "loss": 1.0274, + "step": 18618 + }, + { + "epoch": 0.44, + "grad_norm": 1.9384925134180309, + "learning_rate": 1.244788555850287e-05, + "loss": 1.1096, + "step": 18619 + }, + { + "epoch": 0.44, + "grad_norm": 2.2600924330059873, + "learning_rate": 1.2447145724707178e-05, + "loss": 0.8404, + "step": 18620 + }, + { + "epoch": 0.44, + "grad_norm": 1.922139654506747, + "learning_rate": 1.2446405876663424e-05, + "loss": 1.0499, + "step": 18621 + }, + { + "epoch": 0.44, + "grad_norm": 2.266015637465498, + "learning_rate": 1.2445666014375918e-05, + "loss": 1.0002, + "step": 18622 + }, + { + "epoch": 0.44, + "grad_norm": 1.910453604301353, + "learning_rate": 1.2444926137848974e-05, + "loss": 1.0051, + "step": 18623 + }, + { + "epoch": 0.44, + "grad_norm": 2.013044299627454, + "learning_rate": 1.2444186247086895e-05, + "loss": 0.8334, + "step": 18624 + }, + { + "epoch": 0.44, + "grad_norm": 2.1790802441504953, + "learning_rate": 1.244344634209399e-05, + "loss": 1.0228, + "step": 18625 + }, + { + "epoch": 0.44, + "grad_norm": 2.2246084744934724, + "learning_rate": 1.2442706422874567e-05, + "loss": 1.0472, + "step": 18626 + }, + { + "epoch": 0.44, + "grad_norm": 1.8821771475808113, + "learning_rate": 1.2441966489432936e-05, + "loss": 1.0821, + "step": 18627 + }, + { + "epoch": 0.44, + "grad_norm": 2.0928987835101696, + "learning_rate": 1.24412265417734e-05, + "loss": 1.0373, + "step": 18628 + }, + { + "epoch": 0.44, + "grad_norm": 1.8337591969181655, + "learning_rate": 1.2440486579900268e-05, + "loss": 1.0231, + "step": 18629 + }, + { + "epoch": 0.44, + "grad_norm": 1.1846434135477022, + "learning_rate": 1.2439746603817855e-05, + "loss": 0.9809, + "step": 18630 + }, + { + "epoch": 0.44, + "grad_norm": 1.8155746988152492, + "learning_rate": 1.2439006613530462e-05, + "loss": 0.9781, + "step": 18631 + }, + { + "epoch": 0.44, + "grad_norm": 1.9349629538313693, + "learning_rate": 1.2438266609042403e-05, + "loss": 1.0406, + "step": 18632 + }, + { + "epoch": 0.44, + "grad_norm": 2.0230427451523387, + "learning_rate": 1.2437526590357982e-05, + "loss": 1.038, + "step": 18633 + }, + { + "epoch": 0.44, + "grad_norm": 1.8826885980234693, + "learning_rate": 1.243678655748151e-05, + "loss": 1.0503, + "step": 18634 + }, + { + "epoch": 0.44, + "grad_norm": 2.1876923283364276, + "learning_rate": 1.2436046510417297e-05, + "loss": 1.056, + "step": 18635 + }, + { + "epoch": 0.44, + "grad_norm": 1.2177119621936476, + "learning_rate": 1.2435306449169646e-05, + "loss": 0.9882, + "step": 18636 + }, + { + "epoch": 0.44, + "grad_norm": 2.2421845217387593, + "learning_rate": 1.2434566373742873e-05, + "loss": 1.0018, + "step": 18637 + }, + { + "epoch": 0.44, + "grad_norm": 2.0206059743273017, + "learning_rate": 1.2433826284141282e-05, + "loss": 1.0498, + "step": 18638 + }, + { + "epoch": 0.44, + "grad_norm": 1.9947160381957227, + "learning_rate": 1.2433086180369184e-05, + "loss": 1.0697, + "step": 18639 + }, + { + "epoch": 0.44, + "grad_norm": 1.9399306234352436, + "learning_rate": 1.2432346062430889e-05, + "loss": 0.916, + "step": 18640 + }, + { + "epoch": 0.44, + "grad_norm": 1.8794675741226172, + "learning_rate": 1.2431605930330706e-05, + "loss": 0.9227, + "step": 18641 + }, + { + "epoch": 0.44, + "grad_norm": 2.1274578014177172, + "learning_rate": 1.243086578407294e-05, + "loss": 0.9932, + "step": 18642 + }, + { + "epoch": 0.44, + "grad_norm": 1.8248954961570385, + "learning_rate": 1.2430125623661909e-05, + "loss": 1.0893, + "step": 18643 + }, + { + "epoch": 0.44, + "grad_norm": 2.0591910048232327, + "learning_rate": 1.2429385449101912e-05, + "loss": 1.0063, + "step": 18644 + }, + { + "epoch": 0.44, + "grad_norm": 2.0372332770018042, + "learning_rate": 1.2428645260397267e-05, + "loss": 1.0052, + "step": 18645 + }, + { + "epoch": 0.44, + "grad_norm": 2.0911550717112144, + "learning_rate": 1.2427905057552279e-05, + "loss": 1.0201, + "step": 18646 + }, + { + "epoch": 0.44, + "grad_norm": 2.165149849671227, + "learning_rate": 1.242716484057126e-05, + "loss": 1.0274, + "step": 18647 + }, + { + "epoch": 0.44, + "grad_norm": 2.198389666527665, + "learning_rate": 1.2426424609458518e-05, + "loss": 1.1213, + "step": 18648 + }, + { + "epoch": 0.44, + "grad_norm": 1.964563979532942, + "learning_rate": 1.2425684364218365e-05, + "loss": 1.0425, + "step": 18649 + }, + { + "epoch": 0.44, + "grad_norm": 2.397894142309125, + "learning_rate": 1.2424944104855107e-05, + "loss": 0.9764, + "step": 18650 + }, + { + "epoch": 0.44, + "grad_norm": 2.1116456109367405, + "learning_rate": 1.2424203831373058e-05, + "loss": 1.118, + "step": 18651 + }, + { + "epoch": 0.44, + "grad_norm": 1.1526028467107725, + "learning_rate": 1.2423463543776529e-05, + "loss": 0.9754, + "step": 18652 + }, + { + "epoch": 0.44, + "grad_norm": 2.627195468700267, + "learning_rate": 1.2422723242069826e-05, + "loss": 0.9316, + "step": 18653 + }, + { + "epoch": 0.44, + "grad_norm": 2.04433228401389, + "learning_rate": 1.2421982926257262e-05, + "loss": 0.9438, + "step": 18654 + }, + { + "epoch": 0.44, + "grad_norm": 2.0023712814728922, + "learning_rate": 1.2421242596343144e-05, + "loss": 1.0382, + "step": 18655 + }, + { + "epoch": 0.44, + "grad_norm": 1.926797738437384, + "learning_rate": 1.242050225233179e-05, + "loss": 0.9211, + "step": 18656 + }, + { + "epoch": 0.44, + "grad_norm": 2.041642319709357, + "learning_rate": 1.24197618942275e-05, + "loss": 0.9523, + "step": 18657 + }, + { + "epoch": 0.44, + "grad_norm": 1.7179605126373176, + "learning_rate": 1.2419021522034596e-05, + "loss": 0.9783, + "step": 18658 + }, + { + "epoch": 0.44, + "grad_norm": 1.9005489510069598, + "learning_rate": 1.2418281135757378e-05, + "loss": 1.0657, + "step": 18659 + }, + { + "epoch": 0.44, + "grad_norm": 2.0314145531621106, + "learning_rate": 1.2417540735400161e-05, + "loss": 1.0566, + "step": 18660 + }, + { + "epoch": 0.44, + "grad_norm": 2.1605399437284354, + "learning_rate": 1.241680032096726e-05, + "loss": 0.9762, + "step": 18661 + }, + { + "epoch": 0.44, + "grad_norm": 1.9817332875863658, + "learning_rate": 1.2416059892462979e-05, + "loss": 1.0694, + "step": 18662 + }, + { + "epoch": 0.44, + "grad_norm": 2.117482412480132, + "learning_rate": 1.2415319449891638e-05, + "loss": 1.0646, + "step": 18663 + }, + { + "epoch": 0.44, + "grad_norm": 1.9140358459553362, + "learning_rate": 1.2414578993257535e-05, + "loss": 1.0233, + "step": 18664 + }, + { + "epoch": 0.44, + "grad_norm": 1.8672580997435426, + "learning_rate": 1.2413838522564997e-05, + "loss": 0.9843, + "step": 18665 + }, + { + "epoch": 0.44, + "grad_norm": 2.1654727023605616, + "learning_rate": 1.2413098037818321e-05, + "loss": 1.0261, + "step": 18666 + }, + { + "epoch": 0.44, + "grad_norm": 1.0695846745162174, + "learning_rate": 1.2412357539021826e-05, + "loss": 0.9795, + "step": 18667 + }, + { + "epoch": 0.44, + "grad_norm": 1.815484277864791, + "learning_rate": 1.2411617026179821e-05, + "loss": 1.1272, + "step": 18668 + }, + { + "epoch": 0.44, + "grad_norm": 1.1861464924025809, + "learning_rate": 1.241087649929662e-05, + "loss": 0.9949, + "step": 18669 + }, + { + "epoch": 0.44, + "grad_norm": 1.736126581685192, + "learning_rate": 1.241013595837653e-05, + "loss": 0.9974, + "step": 18670 + }, + { + "epoch": 0.44, + "grad_norm": 2.0192803859079893, + "learning_rate": 1.240939540342387e-05, + "loss": 1.1082, + "step": 18671 + }, + { + "epoch": 0.44, + "grad_norm": 2.026955455049788, + "learning_rate": 1.2408654834442942e-05, + "loss": 1.1706, + "step": 18672 + }, + { + "epoch": 0.44, + "grad_norm": 2.3095782416935435, + "learning_rate": 1.2407914251438065e-05, + "loss": 0.922, + "step": 18673 + }, + { + "epoch": 0.44, + "grad_norm": 2.056223772058862, + "learning_rate": 1.2407173654413549e-05, + "loss": 0.8788, + "step": 18674 + }, + { + "epoch": 0.44, + "grad_norm": 2.7266365441027856, + "learning_rate": 1.2406433043373707e-05, + "loss": 1.1219, + "step": 18675 + }, + { + "epoch": 0.44, + "grad_norm": 1.9880455029326933, + "learning_rate": 1.2405692418322847e-05, + "loss": 1.026, + "step": 18676 + }, + { + "epoch": 0.44, + "grad_norm": 2.1887099737041598, + "learning_rate": 1.2404951779265287e-05, + "loss": 0.9578, + "step": 18677 + }, + { + "epoch": 0.44, + "grad_norm": 2.642235283187618, + "learning_rate": 1.2404211126205335e-05, + "loss": 1.0211, + "step": 18678 + }, + { + "epoch": 0.44, + "grad_norm": 2.060199438369952, + "learning_rate": 1.2403470459147305e-05, + "loss": 1.0966, + "step": 18679 + }, + { + "epoch": 0.44, + "grad_norm": 1.9346744156950955, + "learning_rate": 1.2402729778095507e-05, + "loss": 0.967, + "step": 18680 + }, + { + "epoch": 0.44, + "grad_norm": 1.9399334821916074, + "learning_rate": 1.2401989083054258e-05, + "loss": 1.0761, + "step": 18681 + }, + { + "epoch": 0.44, + "grad_norm": 1.968550593087758, + "learning_rate": 1.2401248374027868e-05, + "loss": 1.0382, + "step": 18682 + }, + { + "epoch": 0.44, + "grad_norm": 1.9077108900372417, + "learning_rate": 1.240050765102065e-05, + "loss": 0.8584, + "step": 18683 + }, + { + "epoch": 0.44, + "grad_norm": 1.9807462732940706, + "learning_rate": 1.2399766914036914e-05, + "loss": 1.0615, + "step": 18684 + }, + { + "epoch": 0.44, + "grad_norm": 1.9348976279831636, + "learning_rate": 1.2399026163080978e-05, + "loss": 1.1115, + "step": 18685 + }, + { + "epoch": 0.44, + "grad_norm": 1.8809464140882977, + "learning_rate": 1.2398285398157153e-05, + "loss": 0.8388, + "step": 18686 + }, + { + "epoch": 0.44, + "grad_norm": 2.057212270251362, + "learning_rate": 1.2397544619269748e-05, + "loss": 0.9808, + "step": 18687 + }, + { + "epoch": 0.44, + "grad_norm": 1.9846796070432056, + "learning_rate": 1.2396803826423083e-05, + "loss": 1.0321, + "step": 18688 + }, + { + "epoch": 0.44, + "grad_norm": 1.9896654064692096, + "learning_rate": 1.2396063019621463e-05, + "loss": 1.0638, + "step": 18689 + }, + { + "epoch": 0.44, + "grad_norm": 2.0643205995460168, + "learning_rate": 1.2395322198869209e-05, + "loss": 1.0674, + "step": 18690 + }, + { + "epoch": 0.44, + "grad_norm": 2.2357097206939214, + "learning_rate": 1.2394581364170626e-05, + "loss": 0.9898, + "step": 18691 + }, + { + "epoch": 0.44, + "grad_norm": 2.909275354283089, + "learning_rate": 1.2393840515530038e-05, + "loss": 0.919, + "step": 18692 + }, + { + "epoch": 0.44, + "grad_norm": 2.2263739486173026, + "learning_rate": 1.239309965295175e-05, + "loss": 0.9627, + "step": 18693 + }, + { + "epoch": 0.44, + "grad_norm": 3.114899300997695, + "learning_rate": 1.2392358776440076e-05, + "loss": 0.995, + "step": 18694 + }, + { + "epoch": 0.44, + "grad_norm": 1.9509695163434972, + "learning_rate": 1.2391617885999336e-05, + "loss": 1.0528, + "step": 18695 + }, + { + "epoch": 0.44, + "grad_norm": 2.087487855018618, + "learning_rate": 1.2390876981633838e-05, + "loss": 1.0422, + "step": 18696 + }, + { + "epoch": 0.44, + "grad_norm": 1.8189375831467838, + "learning_rate": 1.2390136063347895e-05, + "loss": 0.9769, + "step": 18697 + }, + { + "epoch": 0.44, + "grad_norm": 2.204152295002415, + "learning_rate": 1.2389395131145823e-05, + "loss": 0.9768, + "step": 18698 + }, + { + "epoch": 0.44, + "grad_norm": 1.1033775031104647, + "learning_rate": 1.2388654185031941e-05, + "loss": 1.0049, + "step": 18699 + }, + { + "epoch": 0.44, + "grad_norm": 1.0746389318283722, + "learning_rate": 1.2387913225010554e-05, + "loss": 0.9908, + "step": 18700 + }, + { + "epoch": 0.44, + "grad_norm": 1.9858287973239737, + "learning_rate": 1.2387172251085982e-05, + "loss": 1.0488, + "step": 18701 + }, + { + "epoch": 0.44, + "grad_norm": 2.094938032050887, + "learning_rate": 1.2386431263262538e-05, + "loss": 0.961, + "step": 18702 + }, + { + "epoch": 0.44, + "grad_norm": 1.083793523672371, + "learning_rate": 1.2385690261544531e-05, + "loss": 0.9799, + "step": 18703 + }, + { + "epoch": 0.44, + "grad_norm": 2.0001664347380377, + "learning_rate": 1.2384949245936284e-05, + "loss": 0.9258, + "step": 18704 + }, + { + "epoch": 0.44, + "grad_norm": 2.655880680249851, + "learning_rate": 1.2384208216442106e-05, + "loss": 0.992, + "step": 18705 + }, + { + "epoch": 0.44, + "grad_norm": 1.7449780305557592, + "learning_rate": 1.2383467173066315e-05, + "loss": 0.9215, + "step": 18706 + }, + { + "epoch": 0.44, + "grad_norm": 1.9014605698579845, + "learning_rate": 1.2382726115813224e-05, + "loss": 0.9756, + "step": 18707 + }, + { + "epoch": 0.44, + "grad_norm": 1.0739799596068325, + "learning_rate": 1.2381985044687144e-05, + "loss": 0.9, + "step": 18708 + }, + { + "epoch": 0.44, + "grad_norm": 1.9089344021971457, + "learning_rate": 1.2381243959692394e-05, + "loss": 0.9952, + "step": 18709 + }, + { + "epoch": 0.44, + "grad_norm": 1.0924519244026931, + "learning_rate": 1.2380502860833287e-05, + "loss": 0.997, + "step": 18710 + }, + { + "epoch": 0.44, + "grad_norm": 1.9138088750427822, + "learning_rate": 1.237976174811414e-05, + "loss": 1.1059, + "step": 18711 + }, + { + "epoch": 0.44, + "grad_norm": 1.9174164787341355, + "learning_rate": 1.2379020621539266e-05, + "loss": 0.9988, + "step": 18712 + }, + { + "epoch": 0.44, + "grad_norm": 1.9575984234325916, + "learning_rate": 1.2378279481112983e-05, + "loss": 1.0872, + "step": 18713 + }, + { + "epoch": 0.44, + "grad_norm": 2.012987705922395, + "learning_rate": 1.2377538326839603e-05, + "loss": 0.9304, + "step": 18714 + }, + { + "epoch": 0.44, + "grad_norm": 2.2958984436861054, + "learning_rate": 1.2376797158723439e-05, + "loss": 0.999, + "step": 18715 + }, + { + "epoch": 0.44, + "grad_norm": 1.8420809082302314, + "learning_rate": 1.2376055976768815e-05, + "loss": 0.8479, + "step": 18716 + }, + { + "epoch": 0.44, + "grad_norm": 1.7622811249545853, + "learning_rate": 1.2375314780980034e-05, + "loss": 0.91, + "step": 18717 + }, + { + "epoch": 0.44, + "grad_norm": 1.1082652915269229, + "learning_rate": 1.2374573571361426e-05, + "loss": 0.9753, + "step": 18718 + }, + { + "epoch": 0.44, + "grad_norm": 2.3291158434292085, + "learning_rate": 1.2373832347917292e-05, + "loss": 1.0216, + "step": 18719 + }, + { + "epoch": 0.44, + "grad_norm": 2.470261196429983, + "learning_rate": 1.2373091110651958e-05, + "loss": 1.0252, + "step": 18720 + }, + { + "epoch": 0.44, + "grad_norm": 2.178595948743487, + "learning_rate": 1.2372349859569737e-05, + "loss": 0.9227, + "step": 18721 + }, + { + "epoch": 0.44, + "grad_norm": 1.8508066208714955, + "learning_rate": 1.2371608594674943e-05, + "loss": 0.9675, + "step": 18722 + }, + { + "epoch": 0.44, + "grad_norm": 1.1163313989226196, + "learning_rate": 1.237086731597189e-05, + "loss": 1.0204, + "step": 18723 + }, + { + "epoch": 0.44, + "grad_norm": 2.086236382263428, + "learning_rate": 1.2370126023464901e-05, + "loss": 1.0635, + "step": 18724 + }, + { + "epoch": 0.44, + "grad_norm": 1.9797107937391925, + "learning_rate": 1.2369384717158285e-05, + "loss": 1.1784, + "step": 18725 + }, + { + "epoch": 0.44, + "grad_norm": 2.483719707271226, + "learning_rate": 1.2368643397056364e-05, + "loss": 0.9508, + "step": 18726 + }, + { + "epoch": 0.44, + "grad_norm": 2.3445746212413616, + "learning_rate": 1.2367902063163448e-05, + "loss": 0.9113, + "step": 18727 + }, + { + "epoch": 0.44, + "grad_norm": 2.0464796543749517, + "learning_rate": 1.236716071548386e-05, + "loss": 0.967, + "step": 18728 + }, + { + "epoch": 0.44, + "grad_norm": 2.0122025404595125, + "learning_rate": 1.236641935402191e-05, + "loss": 0.986, + "step": 18729 + }, + { + "epoch": 0.44, + "grad_norm": 1.15246182701535, + "learning_rate": 1.2365677978781919e-05, + "loss": 0.8998, + "step": 18730 + }, + { + "epoch": 0.44, + "grad_norm": 1.739604563714986, + "learning_rate": 1.2364936589768202e-05, + "loss": 0.8942, + "step": 18731 + }, + { + "epoch": 0.44, + "grad_norm": 2.2477057686619184, + "learning_rate": 1.2364195186985073e-05, + "loss": 1.1016, + "step": 18732 + }, + { + "epoch": 0.44, + "grad_norm": 1.9004047729949602, + "learning_rate": 1.2363453770436853e-05, + "loss": 1.0842, + "step": 18733 + }, + { + "epoch": 0.44, + "grad_norm": 2.257709035000089, + "learning_rate": 1.2362712340127856e-05, + "loss": 1.074, + "step": 18734 + }, + { + "epoch": 0.44, + "grad_norm": 2.040797941677796, + "learning_rate": 1.23619708960624e-05, + "loss": 1.0863, + "step": 18735 + }, + { + "epoch": 0.44, + "grad_norm": 2.1623334570379997, + "learning_rate": 1.2361229438244803e-05, + "loss": 1.0136, + "step": 18736 + }, + { + "epoch": 0.44, + "grad_norm": 1.974826372563475, + "learning_rate": 1.2360487966679383e-05, + "loss": 1.0643, + "step": 18737 + }, + { + "epoch": 0.44, + "grad_norm": 2.6336114066436527, + "learning_rate": 1.2359746481370447e-05, + "loss": 1.1488, + "step": 18738 + }, + { + "epoch": 0.44, + "grad_norm": 2.0687997934901494, + "learning_rate": 1.2359004982322328e-05, + "loss": 1.0924, + "step": 18739 + }, + { + "epoch": 0.44, + "grad_norm": 1.682862253436486, + "learning_rate": 1.2358263469539328e-05, + "loss": 1.0269, + "step": 18740 + }, + { + "epoch": 0.44, + "grad_norm": 2.175710717246664, + "learning_rate": 1.2357521943025779e-05, + "loss": 1.1267, + "step": 18741 + }, + { + "epoch": 0.44, + "grad_norm": 2.0032008690551875, + "learning_rate": 1.2356780402785985e-05, + "loss": 0.9785, + "step": 18742 + }, + { + "epoch": 0.44, + "grad_norm": 1.9975942072005017, + "learning_rate": 1.2356038848824272e-05, + "loss": 1.0283, + "step": 18743 + }, + { + "epoch": 0.44, + "grad_norm": 1.9015169504714002, + "learning_rate": 1.2355297281144953e-05, + "loss": 1.0302, + "step": 18744 + }, + { + "epoch": 0.44, + "grad_norm": 2.2542168589786193, + "learning_rate": 1.2354555699752351e-05, + "loss": 1.0644, + "step": 18745 + }, + { + "epoch": 0.44, + "grad_norm": 1.0665818982825501, + "learning_rate": 1.2353814104650777e-05, + "loss": 0.9539, + "step": 18746 + }, + { + "epoch": 0.44, + "grad_norm": 1.8557187336745764, + "learning_rate": 1.2353072495844551e-05, + "loss": 0.9313, + "step": 18747 + }, + { + "epoch": 0.44, + "grad_norm": 2.07903511119003, + "learning_rate": 1.2352330873337995e-05, + "loss": 1.0551, + "step": 18748 + }, + { + "epoch": 0.44, + "grad_norm": 1.888490927217773, + "learning_rate": 1.2351589237135422e-05, + "loss": 0.9966, + "step": 18749 + }, + { + "epoch": 0.44, + "grad_norm": 3.1678248530349578, + "learning_rate": 1.2350847587241155e-05, + "loss": 1.0419, + "step": 18750 + }, + { + "epoch": 0.44, + "grad_norm": 2.0090034535739125, + "learning_rate": 1.2350105923659506e-05, + "loss": 1.0696, + "step": 18751 + }, + { + "epoch": 0.44, + "grad_norm": 1.9851999121716797, + "learning_rate": 1.2349364246394797e-05, + "loss": 1.069, + "step": 18752 + }, + { + "epoch": 0.44, + "grad_norm": 1.8869183310885014, + "learning_rate": 1.2348622555451346e-05, + "loss": 1.0693, + "step": 18753 + }, + { + "epoch": 0.44, + "grad_norm": 1.8750861888189354, + "learning_rate": 1.2347880850833472e-05, + "loss": 0.9662, + "step": 18754 + }, + { + "epoch": 0.44, + "grad_norm": 2.1333036555862184, + "learning_rate": 1.234713913254549e-05, + "loss": 1.0397, + "step": 18755 + }, + { + "epoch": 0.44, + "grad_norm": 1.157109959276848, + "learning_rate": 1.2346397400591721e-05, + "loss": 1.043, + "step": 18756 + }, + { + "epoch": 0.44, + "grad_norm": 2.391676572978134, + "learning_rate": 1.2345655654976485e-05, + "loss": 0.9991, + "step": 18757 + }, + { + "epoch": 0.44, + "grad_norm": 1.9690061909400864, + "learning_rate": 1.2344913895704099e-05, + "loss": 0.961, + "step": 18758 + }, + { + "epoch": 0.44, + "grad_norm": 2.1542661504508556, + "learning_rate": 1.2344172122778882e-05, + "loss": 0.9958, + "step": 18759 + }, + { + "epoch": 0.44, + "grad_norm": 1.8075326163520944, + "learning_rate": 1.234343033620515e-05, + "loss": 0.9466, + "step": 18760 + }, + { + "epoch": 0.44, + "grad_norm": 1.9101803480424426, + "learning_rate": 1.2342688535987228e-05, + "loss": 1.0817, + "step": 18761 + }, + { + "epoch": 0.44, + "grad_norm": 1.849036874797341, + "learning_rate": 1.234194672212943e-05, + "loss": 1.1215, + "step": 18762 + }, + { + "epoch": 0.44, + "grad_norm": 2.1235194437003098, + "learning_rate": 1.234120489463608e-05, + "loss": 1.0996, + "step": 18763 + }, + { + "epoch": 0.44, + "grad_norm": 2.262166318170965, + "learning_rate": 1.234046305351149e-05, + "loss": 1.0253, + "step": 18764 + }, + { + "epoch": 0.44, + "grad_norm": 2.011060079824803, + "learning_rate": 1.2339721198759987e-05, + "loss": 1.0, + "step": 18765 + }, + { + "epoch": 0.44, + "grad_norm": 2.9071612804055302, + "learning_rate": 1.2338979330385881e-05, + "loss": 0.974, + "step": 18766 + }, + { + "epoch": 0.44, + "grad_norm": 2.184043287061071, + "learning_rate": 1.23382374483935e-05, + "loss": 1.0241, + "step": 18767 + }, + { + "epoch": 0.44, + "grad_norm": 2.0589896509960175, + "learning_rate": 1.2337495552787157e-05, + "loss": 1.1428, + "step": 18768 + }, + { + "epoch": 0.44, + "grad_norm": 1.0813816788792778, + "learning_rate": 1.2336753643571181e-05, + "loss": 0.9212, + "step": 18769 + }, + { + "epoch": 0.44, + "grad_norm": 2.08841335491419, + "learning_rate": 1.2336011720749881e-05, + "loss": 1.1569, + "step": 18770 + }, + { + "epoch": 0.44, + "grad_norm": 1.1009934564901183, + "learning_rate": 1.2335269784327583e-05, + "loss": 1.0197, + "step": 18771 + }, + { + "epoch": 0.44, + "grad_norm": 3.36544534032477, + "learning_rate": 1.2334527834308606e-05, + "loss": 0.9816, + "step": 18772 + }, + { + "epoch": 0.44, + "grad_norm": 1.8443654395648124, + "learning_rate": 1.2333785870697265e-05, + "loss": 1.0701, + "step": 18773 + }, + { + "epoch": 0.44, + "grad_norm": 2.0338907255198375, + "learning_rate": 1.2333043893497886e-05, + "loss": 0.956, + "step": 18774 + }, + { + "epoch": 0.44, + "grad_norm": 2.0024140546093907, + "learning_rate": 1.2332301902714789e-05, + "loss": 0.9234, + "step": 18775 + }, + { + "epoch": 0.44, + "grad_norm": 1.9776466409243971, + "learning_rate": 1.233155989835229e-05, + "loss": 0.7905, + "step": 18776 + }, + { + "epoch": 0.44, + "grad_norm": 1.8048970587962496, + "learning_rate": 1.233081788041471e-05, + "loss": 1.0608, + "step": 18777 + }, + { + "epoch": 0.44, + "grad_norm": 2.3179600789764874, + "learning_rate": 1.2330075848906371e-05, + "loss": 0.9178, + "step": 18778 + }, + { + "epoch": 0.44, + "grad_norm": 2.7708660252331017, + "learning_rate": 1.2329333803831593e-05, + "loss": 1.0377, + "step": 18779 + }, + { + "epoch": 0.44, + "grad_norm": 1.9091672920107587, + "learning_rate": 1.2328591745194696e-05, + "loss": 1.0388, + "step": 18780 + }, + { + "epoch": 0.44, + "grad_norm": 1.7962878916277134, + "learning_rate": 1.2327849673000001e-05, + "loss": 1.0384, + "step": 18781 + }, + { + "epoch": 0.44, + "grad_norm": 1.142332677887318, + "learning_rate": 1.2327107587251827e-05, + "loss": 1.0289, + "step": 18782 + }, + { + "epoch": 0.44, + "grad_norm": 2.086478600702742, + "learning_rate": 1.2326365487954497e-05, + "loss": 0.9827, + "step": 18783 + }, + { + "epoch": 0.44, + "grad_norm": 1.3052664535115288, + "learning_rate": 1.232562337511233e-05, + "loss": 0.9611, + "step": 18784 + }, + { + "epoch": 0.44, + "grad_norm": 1.0902485708759901, + "learning_rate": 1.232488124872965e-05, + "loss": 0.9075, + "step": 18785 + }, + { + "epoch": 0.44, + "grad_norm": 2.398500018050553, + "learning_rate": 1.2324139108810775e-05, + "loss": 1.0568, + "step": 18786 + }, + { + "epoch": 0.44, + "grad_norm": 1.0889210510188798, + "learning_rate": 1.2323396955360023e-05, + "loss": 1.0257, + "step": 18787 + }, + { + "epoch": 0.44, + "grad_norm": 1.9554706248807052, + "learning_rate": 1.232265478838172e-05, + "loss": 1.0612, + "step": 18788 + }, + { + "epoch": 0.44, + "grad_norm": 1.8360814918831052, + "learning_rate": 1.2321912607880184e-05, + "loss": 0.9605, + "step": 18789 + }, + { + "epoch": 0.44, + "grad_norm": 3.5496546037999064, + "learning_rate": 1.2321170413859742e-05, + "loss": 1.1158, + "step": 18790 + }, + { + "epoch": 0.44, + "grad_norm": 1.966124137866349, + "learning_rate": 1.2320428206324703e-05, + "loss": 1.1535, + "step": 18791 + }, + { + "epoch": 0.44, + "grad_norm": 2.1023105815232896, + "learning_rate": 1.2319685985279406e-05, + "loss": 1.0961, + "step": 18792 + }, + { + "epoch": 0.44, + "grad_norm": 2.3955025209235474, + "learning_rate": 1.2318943750728157e-05, + "loss": 1.2067, + "step": 18793 + }, + { + "epoch": 0.44, + "grad_norm": 2.217441260163013, + "learning_rate": 1.2318201502675285e-05, + "loss": 1.0008, + "step": 18794 + }, + { + "epoch": 0.44, + "grad_norm": 2.381559888507676, + "learning_rate": 1.231745924112511e-05, + "loss": 1.1269, + "step": 18795 + }, + { + "epoch": 0.44, + "grad_norm": 1.9693150280134015, + "learning_rate": 1.231671696608195e-05, + "loss": 1.042, + "step": 18796 + }, + { + "epoch": 0.44, + "grad_norm": 1.9858883124187012, + "learning_rate": 1.2315974677550136e-05, + "loss": 1.1611, + "step": 18797 + }, + { + "epoch": 0.44, + "grad_norm": 2.556368419000247, + "learning_rate": 1.231523237553398e-05, + "loss": 0.9314, + "step": 18798 + }, + { + "epoch": 0.44, + "grad_norm": 2.311121591418088, + "learning_rate": 1.2314490060037812e-05, + "loss": 1.0646, + "step": 18799 + }, + { + "epoch": 0.44, + "grad_norm": 2.0329958532248966, + "learning_rate": 1.2313747731065947e-05, + "loss": 1.0648, + "step": 18800 + }, + { + "epoch": 0.44, + "grad_norm": 2.099901308789362, + "learning_rate": 1.2313005388622712e-05, + "loss": 0.936, + "step": 18801 + }, + { + "epoch": 0.44, + "grad_norm": 2.0275699360906296, + "learning_rate": 1.2312263032712425e-05, + "loss": 1.0959, + "step": 18802 + }, + { + "epoch": 0.44, + "grad_norm": 1.9875990978753597, + "learning_rate": 1.2311520663339411e-05, + "loss": 1.0994, + "step": 18803 + }, + { + "epoch": 0.44, + "grad_norm": 1.9185475717395315, + "learning_rate": 1.2310778280507994e-05, + "loss": 0.8873, + "step": 18804 + }, + { + "epoch": 0.44, + "grad_norm": 2.299110572540773, + "learning_rate": 1.2310035884222493e-05, + "loss": 1.0754, + "step": 18805 + }, + { + "epoch": 0.44, + "grad_norm": 1.9214233787558284, + "learning_rate": 1.2309293474487232e-05, + "loss": 0.9893, + "step": 18806 + }, + { + "epoch": 0.44, + "grad_norm": 1.9589616858821903, + "learning_rate": 1.2308551051306533e-05, + "loss": 1.0575, + "step": 18807 + }, + { + "epoch": 0.44, + "grad_norm": 2.1706505951223964, + "learning_rate": 1.230780861468472e-05, + "loss": 1.0554, + "step": 18808 + }, + { + "epoch": 0.44, + "grad_norm": 2.353780688697417, + "learning_rate": 1.2307066164626113e-05, + "loss": 0.9765, + "step": 18809 + }, + { + "epoch": 0.44, + "grad_norm": 1.0942866209588304, + "learning_rate": 1.2306323701135036e-05, + "loss": 0.9408, + "step": 18810 + }, + { + "epoch": 0.44, + "grad_norm": 2.333577516539711, + "learning_rate": 1.2305581224215812e-05, + "loss": 1.0773, + "step": 18811 + }, + { + "epoch": 0.44, + "grad_norm": 2.0646330183032773, + "learning_rate": 1.2304838733872768e-05, + "loss": 0.9397, + "step": 18812 + }, + { + "epoch": 0.44, + "grad_norm": 2.3326782937547494, + "learning_rate": 1.2304096230110219e-05, + "loss": 1.0225, + "step": 18813 + }, + { + "epoch": 0.44, + "grad_norm": 2.480654646213462, + "learning_rate": 1.2303353712932495e-05, + "loss": 1.1001, + "step": 18814 + }, + { + "epoch": 0.44, + "grad_norm": 1.9205106142278319, + "learning_rate": 1.2302611182343915e-05, + "loss": 1.0453, + "step": 18815 + }, + { + "epoch": 0.44, + "grad_norm": 2.217861346388462, + "learning_rate": 1.2301868638348805e-05, + "loss": 1.1887, + "step": 18816 + }, + { + "epoch": 0.44, + "grad_norm": 2.2357083950736434, + "learning_rate": 1.2301126080951484e-05, + "loss": 0.9238, + "step": 18817 + }, + { + "epoch": 0.44, + "grad_norm": 1.1039546947570538, + "learning_rate": 1.2300383510156284e-05, + "loss": 0.9486, + "step": 18818 + }, + { + "epoch": 0.44, + "grad_norm": 1.9808804732534167, + "learning_rate": 1.2299640925967518e-05, + "loss": 1.0675, + "step": 18819 + }, + { + "epoch": 0.44, + "grad_norm": 1.9023862375399638, + "learning_rate": 1.2298898328389517e-05, + "loss": 0.9493, + "step": 18820 + }, + { + "epoch": 0.44, + "grad_norm": 1.093400414406367, + "learning_rate": 1.2298155717426598e-05, + "loss": 0.9446, + "step": 18821 + }, + { + "epoch": 0.44, + "grad_norm": 1.9598742749446938, + "learning_rate": 1.2297413093083094e-05, + "loss": 0.9328, + "step": 18822 + }, + { + "epoch": 0.44, + "grad_norm": 4.079900087605517, + "learning_rate": 1.2296670455363322e-05, + "loss": 1.096, + "step": 18823 + }, + { + "epoch": 0.44, + "grad_norm": 1.9131512163425195, + "learning_rate": 1.2295927804271606e-05, + "loss": 1.073, + "step": 18824 + }, + { + "epoch": 0.44, + "grad_norm": 1.8924449775645724, + "learning_rate": 1.2295185139812274e-05, + "loss": 1.0829, + "step": 18825 + }, + { + "epoch": 0.44, + "grad_norm": 1.831605537843823, + "learning_rate": 1.2294442461989644e-05, + "loss": 1.1042, + "step": 18826 + }, + { + "epoch": 0.44, + "grad_norm": 2.764256429604512, + "learning_rate": 1.2293699770808047e-05, + "loss": 1.0172, + "step": 18827 + }, + { + "epoch": 0.44, + "grad_norm": 1.9686464673211046, + "learning_rate": 1.2292957066271802e-05, + "loss": 0.9954, + "step": 18828 + }, + { + "epoch": 0.44, + "grad_norm": 1.9674627568786398, + "learning_rate": 1.2292214348385235e-05, + "loss": 0.9652, + "step": 18829 + }, + { + "epoch": 0.44, + "grad_norm": 2.1116788322090856, + "learning_rate": 1.229147161715267e-05, + "loss": 1.0742, + "step": 18830 + }, + { + "epoch": 0.44, + "grad_norm": 1.0755240894185403, + "learning_rate": 1.2290728872578432e-05, + "loss": 0.9283, + "step": 18831 + }, + { + "epoch": 0.44, + "grad_norm": 2.332492092738594, + "learning_rate": 1.2289986114666847e-05, + "loss": 0.9517, + "step": 18832 + }, + { + "epoch": 0.44, + "grad_norm": 2.7323743284000193, + "learning_rate": 1.2289243343422234e-05, + "loss": 1.0954, + "step": 18833 + }, + { + "epoch": 0.44, + "grad_norm": 1.939133061024086, + "learning_rate": 1.2288500558848926e-05, + "loss": 1.0613, + "step": 18834 + }, + { + "epoch": 0.44, + "grad_norm": 2.229109787802552, + "learning_rate": 1.228775776095124e-05, + "loss": 1.0407, + "step": 18835 + }, + { + "epoch": 0.44, + "grad_norm": 2.1760807501103376, + "learning_rate": 1.2287014949733506e-05, + "loss": 0.9728, + "step": 18836 + }, + { + "epoch": 0.44, + "grad_norm": 2.0869464503583415, + "learning_rate": 1.2286272125200048e-05, + "loss": 1.0152, + "step": 18837 + }, + { + "epoch": 0.44, + "grad_norm": 2.670225491384102, + "learning_rate": 1.2285529287355188e-05, + "loss": 0.9611, + "step": 18838 + }, + { + "epoch": 0.44, + "grad_norm": 1.8376064258160822, + "learning_rate": 1.2284786436203253e-05, + "loss": 0.9424, + "step": 18839 + }, + { + "epoch": 0.44, + "grad_norm": 2.12581118195384, + "learning_rate": 1.2284043571748565e-05, + "loss": 1.0322, + "step": 18840 + }, + { + "epoch": 0.44, + "grad_norm": 2.29494662915738, + "learning_rate": 1.2283300693995459e-05, + "loss": 1.1112, + "step": 18841 + }, + { + "epoch": 0.44, + "grad_norm": 2.5634678667011817, + "learning_rate": 1.2282557802948247e-05, + "loss": 1.1442, + "step": 18842 + }, + { + "epoch": 0.44, + "grad_norm": 2.2322175102607744, + "learning_rate": 1.2281814898611267e-05, + "loss": 1.0155, + "step": 18843 + }, + { + "epoch": 0.44, + "grad_norm": 2.0954550480063947, + "learning_rate": 1.2281071980988832e-05, + "loss": 1.1932, + "step": 18844 + }, + { + "epoch": 0.44, + "grad_norm": 1.9557685615405367, + "learning_rate": 1.2280329050085276e-05, + "loss": 1.0605, + "step": 18845 + }, + { + "epoch": 0.44, + "grad_norm": 1.0541373654799087, + "learning_rate": 1.2279586105904924e-05, + "loss": 0.8946, + "step": 18846 + }, + { + "epoch": 0.44, + "grad_norm": 1.9751581307714496, + "learning_rate": 1.2278843148452095e-05, + "loss": 1.2126, + "step": 18847 + }, + { + "epoch": 0.44, + "grad_norm": 1.1703167710191311, + "learning_rate": 1.2278100177731128e-05, + "loss": 1.0781, + "step": 18848 + }, + { + "epoch": 0.44, + "grad_norm": 2.1816393298125627, + "learning_rate": 1.2277357193746333e-05, + "loss": 0.7832, + "step": 18849 + }, + { + "epoch": 0.44, + "grad_norm": 1.9817110153055344, + "learning_rate": 1.2276614196502047e-05, + "loss": 1.0955, + "step": 18850 + }, + { + "epoch": 0.44, + "grad_norm": 1.8595104351025447, + "learning_rate": 1.2275871186002592e-05, + "loss": 0.9613, + "step": 18851 + }, + { + "epoch": 0.44, + "grad_norm": 2.0045163680949294, + "learning_rate": 1.2275128162252297e-05, + "loss": 0.986, + "step": 18852 + }, + { + "epoch": 0.44, + "grad_norm": 2.08241937447911, + "learning_rate": 1.2274385125255481e-05, + "loss": 1.0825, + "step": 18853 + }, + { + "epoch": 0.44, + "grad_norm": 1.8352187570267895, + "learning_rate": 1.2273642075016476e-05, + "loss": 0.9675, + "step": 18854 + }, + { + "epoch": 0.44, + "grad_norm": 2.2135499372806966, + "learning_rate": 1.2272899011539607e-05, + "loss": 1.0265, + "step": 18855 + }, + { + "epoch": 0.44, + "grad_norm": 2.278640580262078, + "learning_rate": 1.2272155934829203e-05, + "loss": 1.1396, + "step": 18856 + }, + { + "epoch": 0.44, + "grad_norm": 1.9072225182915705, + "learning_rate": 1.2271412844889585e-05, + "loss": 1.0475, + "step": 18857 + }, + { + "epoch": 0.44, + "grad_norm": 1.8462571524864158, + "learning_rate": 1.2270669741725083e-05, + "loss": 1.1017, + "step": 18858 + }, + { + "epoch": 0.44, + "grad_norm": 2.3876723799355077, + "learning_rate": 1.2269926625340022e-05, + "loss": 1.0431, + "step": 18859 + }, + { + "epoch": 0.44, + "grad_norm": 1.6766590961792003, + "learning_rate": 1.226918349573873e-05, + "loss": 0.9289, + "step": 18860 + }, + { + "epoch": 0.44, + "grad_norm": 2.5268603709076585, + "learning_rate": 1.2268440352925534e-05, + "loss": 0.8441, + "step": 18861 + }, + { + "epoch": 0.44, + "grad_norm": 1.8280532331980501, + "learning_rate": 1.226769719690476e-05, + "loss": 0.9914, + "step": 18862 + }, + { + "epoch": 0.44, + "grad_norm": 1.1958878513878788, + "learning_rate": 1.2266954027680736e-05, + "loss": 0.8931, + "step": 18863 + }, + { + "epoch": 0.44, + "grad_norm": 2.406007408720279, + "learning_rate": 1.2266210845257787e-05, + "loss": 1.1122, + "step": 18864 + }, + { + "epoch": 0.44, + "grad_norm": 2.031148835961899, + "learning_rate": 1.226546764964024e-05, + "loss": 1.0489, + "step": 18865 + }, + { + "epoch": 0.44, + "grad_norm": 2.1647276765199353, + "learning_rate": 1.2264724440832423e-05, + "loss": 0.9303, + "step": 18866 + }, + { + "epoch": 0.44, + "grad_norm": 2.1301140824724283, + "learning_rate": 1.2263981218838668e-05, + "loss": 1.1121, + "step": 18867 + }, + { + "epoch": 0.44, + "grad_norm": 1.8392738331004552, + "learning_rate": 1.226323798366329e-05, + "loss": 1.0695, + "step": 18868 + }, + { + "epoch": 0.44, + "grad_norm": 1.9609166444969781, + "learning_rate": 1.226249473531063e-05, + "loss": 1.048, + "step": 18869 + }, + { + "epoch": 0.44, + "grad_norm": 1.875230691918927, + "learning_rate": 1.2261751473785004e-05, + "loss": 1.0859, + "step": 18870 + }, + { + "epoch": 0.44, + "grad_norm": 1.8142586533019562, + "learning_rate": 1.2261008199090751e-05, + "loss": 1.1028, + "step": 18871 + }, + { + "epoch": 0.44, + "grad_norm": 1.0574798723795509, + "learning_rate": 1.2260264911232187e-05, + "loss": 0.9664, + "step": 18872 + }, + { + "epoch": 0.44, + "grad_norm": 2.154530612607053, + "learning_rate": 1.225952161021365e-05, + "loss": 0.9467, + "step": 18873 + }, + { + "epoch": 0.44, + "grad_norm": 4.350489774384027, + "learning_rate": 1.225877829603946e-05, + "loss": 1.0272, + "step": 18874 + }, + { + "epoch": 0.44, + "grad_norm": 1.202839360761041, + "learning_rate": 1.2258034968713948e-05, + "loss": 0.9828, + "step": 18875 + }, + { + "epoch": 0.44, + "grad_norm": 1.1952256398259236, + "learning_rate": 1.2257291628241441e-05, + "loss": 0.9412, + "step": 18876 + }, + { + "epoch": 0.44, + "grad_norm": 2.0634564379781755, + "learning_rate": 1.2256548274626267e-05, + "loss": 0.992, + "step": 18877 + }, + { + "epoch": 0.44, + "grad_norm": 1.9275560259688627, + "learning_rate": 1.2255804907872755e-05, + "loss": 0.9822, + "step": 18878 + }, + { + "epoch": 0.44, + "grad_norm": 2.374309803702449, + "learning_rate": 1.2255061527985235e-05, + "loss": 1.0161, + "step": 18879 + }, + { + "epoch": 0.44, + "grad_norm": 2.051310909537656, + "learning_rate": 1.2254318134968029e-05, + "loss": 0.9407, + "step": 18880 + }, + { + "epoch": 0.44, + "grad_norm": 1.9865970899510672, + "learning_rate": 1.2253574728825472e-05, + "loss": 0.9891, + "step": 18881 + }, + { + "epoch": 0.44, + "grad_norm": 1.8266614623576216, + "learning_rate": 1.2252831309561888e-05, + "loss": 1.0466, + "step": 18882 + }, + { + "epoch": 0.44, + "grad_norm": 1.057731211037908, + "learning_rate": 1.2252087877181606e-05, + "loss": 0.8779, + "step": 18883 + }, + { + "epoch": 0.44, + "grad_norm": 1.8998929639686215, + "learning_rate": 1.2251344431688959e-05, + "loss": 1.1021, + "step": 18884 + }, + { + "epoch": 0.44, + "grad_norm": 1.8948943613552052, + "learning_rate": 1.225060097308827e-05, + "loss": 1.11, + "step": 18885 + }, + { + "epoch": 0.44, + "grad_norm": 1.8794799278840169, + "learning_rate": 1.2249857501383869e-05, + "loss": 1.13, + "step": 18886 + }, + { + "epoch": 0.44, + "grad_norm": 1.1009068386170635, + "learning_rate": 1.2249114016580083e-05, + "loss": 0.9587, + "step": 18887 + }, + { + "epoch": 0.44, + "grad_norm": 2.558885599191502, + "learning_rate": 1.224837051868125e-05, + "loss": 1.0263, + "step": 18888 + }, + { + "epoch": 0.45, + "grad_norm": 2.517655921870435, + "learning_rate": 1.2247627007691684e-05, + "loss": 1.0166, + "step": 18889 + }, + { + "epoch": 0.45, + "grad_norm": 2.0978848466638036, + "learning_rate": 1.2246883483615731e-05, + "loss": 1.1091, + "step": 18890 + }, + { + "epoch": 0.45, + "grad_norm": 1.9612150545572982, + "learning_rate": 1.2246139946457703e-05, + "loss": 1.0664, + "step": 18891 + }, + { + "epoch": 0.45, + "grad_norm": 1.8044468245216887, + "learning_rate": 1.2245396396221938e-05, + "loss": 1.1213, + "step": 18892 + }, + { + "epoch": 0.45, + "grad_norm": 2.281314063070833, + "learning_rate": 1.2244652832912766e-05, + "loss": 0.995, + "step": 18893 + }, + { + "epoch": 0.45, + "grad_norm": 2.3715478606685028, + "learning_rate": 1.2243909256534517e-05, + "loss": 1.0007, + "step": 18894 + }, + { + "epoch": 0.45, + "grad_norm": 2.5302244830944494, + "learning_rate": 1.2243165667091517e-05, + "loss": 0.9949, + "step": 18895 + }, + { + "epoch": 0.45, + "grad_norm": 1.7422431882240674, + "learning_rate": 1.224242206458809e-05, + "loss": 0.973, + "step": 18896 + }, + { + "epoch": 0.45, + "grad_norm": 1.0506729367954266, + "learning_rate": 1.224167844902858e-05, + "loss": 0.9739, + "step": 18897 + }, + { + "epoch": 0.45, + "grad_norm": 2.250447809422809, + "learning_rate": 1.2240934820417303e-05, + "loss": 0.959, + "step": 18898 + }, + { + "epoch": 0.45, + "grad_norm": 1.9924695287485554, + "learning_rate": 1.2240191178758598e-05, + "loss": 1.0858, + "step": 18899 + }, + { + "epoch": 0.45, + "grad_norm": 1.965083362058961, + "learning_rate": 1.2239447524056786e-05, + "loss": 0.981, + "step": 18900 + }, + { + "epoch": 0.45, + "grad_norm": 1.9376854336739289, + "learning_rate": 1.2238703856316208e-05, + "loss": 0.9883, + "step": 18901 + }, + { + "epoch": 0.45, + "grad_norm": 2.0557661104809988, + "learning_rate": 1.2237960175541184e-05, + "loss": 0.9787, + "step": 18902 + }, + { + "epoch": 0.45, + "grad_norm": 1.8555028828191853, + "learning_rate": 1.2237216481736047e-05, + "loss": 1.2141, + "step": 18903 + }, + { + "epoch": 0.45, + "grad_norm": 1.985744251368525, + "learning_rate": 1.2236472774905128e-05, + "loss": 1.0415, + "step": 18904 + }, + { + "epoch": 0.45, + "grad_norm": 1.95206916809392, + "learning_rate": 1.2235729055052757e-05, + "loss": 1.0581, + "step": 18905 + }, + { + "epoch": 0.45, + "grad_norm": 1.9580556940945102, + "learning_rate": 1.2234985322183262e-05, + "loss": 1.0688, + "step": 18906 + }, + { + "epoch": 0.45, + "grad_norm": 2.5090067229098962, + "learning_rate": 1.2234241576300974e-05, + "loss": 1.0994, + "step": 18907 + }, + { + "epoch": 0.45, + "grad_norm": 1.9841132209569292, + "learning_rate": 1.2233497817410226e-05, + "loss": 1.066, + "step": 18908 + }, + { + "epoch": 0.45, + "grad_norm": 2.4956394193993656, + "learning_rate": 1.2232754045515348e-05, + "loss": 0.9285, + "step": 18909 + }, + { + "epoch": 0.45, + "grad_norm": 1.9828236001386383, + "learning_rate": 1.2232010260620665e-05, + "loss": 0.9476, + "step": 18910 + }, + { + "epoch": 0.45, + "grad_norm": 1.8720589257084532, + "learning_rate": 1.2231266462730514e-05, + "loss": 1.03, + "step": 18911 + }, + { + "epoch": 0.45, + "grad_norm": 2.0182721151760736, + "learning_rate": 1.2230522651849224e-05, + "loss": 1.046, + "step": 18912 + }, + { + "epoch": 0.45, + "grad_norm": 1.895418460482725, + "learning_rate": 1.2229778827981122e-05, + "loss": 0.934, + "step": 18913 + }, + { + "epoch": 0.45, + "grad_norm": 2.1444583925611025, + "learning_rate": 1.2229034991130544e-05, + "loss": 1.0005, + "step": 18914 + }, + { + "epoch": 0.45, + "grad_norm": 2.0312716181457486, + "learning_rate": 1.2228291141301818e-05, + "loss": 1.0643, + "step": 18915 + }, + { + "epoch": 0.45, + "grad_norm": 2.0599396265077177, + "learning_rate": 1.2227547278499276e-05, + "loss": 0.9813, + "step": 18916 + }, + { + "epoch": 0.45, + "grad_norm": 1.8615918247883851, + "learning_rate": 1.222680340272725e-05, + "loss": 1.0416, + "step": 18917 + }, + { + "epoch": 0.45, + "grad_norm": 1.955269568961225, + "learning_rate": 1.222605951399007e-05, + "loss": 0.9433, + "step": 18918 + }, + { + "epoch": 0.45, + "grad_norm": 2.2068126555444536, + "learning_rate": 1.2225315612292061e-05, + "loss": 1.0648, + "step": 18919 + }, + { + "epoch": 0.45, + "grad_norm": 1.7906936084453788, + "learning_rate": 1.2224571697637566e-05, + "loss": 1.0112, + "step": 18920 + }, + { + "epoch": 0.45, + "grad_norm": 2.4612098951811423, + "learning_rate": 1.2223827770030907e-05, + "loss": 1.066, + "step": 18921 + }, + { + "epoch": 0.45, + "grad_norm": 2.2102262697559683, + "learning_rate": 1.2223083829476421e-05, + "loss": 1.0401, + "step": 18922 + }, + { + "epoch": 0.45, + "grad_norm": 3.3096937136580054, + "learning_rate": 1.2222339875978435e-05, + "loss": 0.9555, + "step": 18923 + }, + { + "epoch": 0.45, + "grad_norm": 1.9876702301098297, + "learning_rate": 1.2221595909541283e-05, + "loss": 1.0965, + "step": 18924 + }, + { + "epoch": 0.45, + "grad_norm": 2.000854407706069, + "learning_rate": 1.22208519301693e-05, + "loss": 1.061, + "step": 18925 + }, + { + "epoch": 0.45, + "grad_norm": 1.0398745393243038, + "learning_rate": 1.222010793786681e-05, + "loss": 0.9399, + "step": 18926 + }, + { + "epoch": 0.45, + "grad_norm": 1.9429453592686101, + "learning_rate": 1.2219363932638148e-05, + "loss": 0.9439, + "step": 18927 + }, + { + "epoch": 0.45, + "grad_norm": 1.0657895119353042, + "learning_rate": 1.2218619914487648e-05, + "loss": 0.9428, + "step": 18928 + }, + { + "epoch": 0.45, + "grad_norm": 2.5773592718483185, + "learning_rate": 1.2217875883419642e-05, + "loss": 0.9756, + "step": 18929 + }, + { + "epoch": 0.45, + "grad_norm": 1.1339960806676421, + "learning_rate": 1.2217131839438457e-05, + "loss": 0.9923, + "step": 18930 + }, + { + "epoch": 0.45, + "grad_norm": 1.9737806413470078, + "learning_rate": 1.221638778254843e-05, + "loss": 1.0318, + "step": 18931 + }, + { + "epoch": 0.45, + "grad_norm": 1.9250254066605126, + "learning_rate": 1.2215643712753893e-05, + "loss": 1.0051, + "step": 18932 + }, + { + "epoch": 0.45, + "grad_norm": 2.1576115391844577, + "learning_rate": 1.2214899630059176e-05, + "loss": 0.934, + "step": 18933 + }, + { + "epoch": 0.45, + "grad_norm": 2.1193981521776033, + "learning_rate": 1.2214155534468612e-05, + "loss": 1.0647, + "step": 18934 + }, + { + "epoch": 0.45, + "grad_norm": 2.1720537806330205, + "learning_rate": 1.2213411425986533e-05, + "loss": 1.1902, + "step": 18935 + }, + { + "epoch": 0.45, + "grad_norm": 1.144737139490101, + "learning_rate": 1.2212667304617272e-05, + "loss": 1.0167, + "step": 18936 + }, + { + "epoch": 0.45, + "grad_norm": 1.0836259541473783, + "learning_rate": 1.2211923170365163e-05, + "loss": 1.059, + "step": 18937 + }, + { + "epoch": 0.45, + "grad_norm": 2.03469569165838, + "learning_rate": 1.2211179023234534e-05, + "loss": 0.8844, + "step": 18938 + }, + { + "epoch": 0.45, + "grad_norm": 2.0672418926315586, + "learning_rate": 1.2210434863229725e-05, + "loss": 1.1729, + "step": 18939 + }, + { + "epoch": 0.45, + "grad_norm": 2.064610504586447, + "learning_rate": 1.2209690690355061e-05, + "loss": 1.0694, + "step": 18940 + }, + { + "epoch": 0.45, + "grad_norm": 1.9118897578108895, + "learning_rate": 1.2208946504614878e-05, + "loss": 1.0082, + "step": 18941 + }, + { + "epoch": 0.45, + "grad_norm": 1.796394312226364, + "learning_rate": 1.220820230601351e-05, + "loss": 1.1621, + "step": 18942 + }, + { + "epoch": 0.45, + "grad_norm": 2.0244277707743117, + "learning_rate": 1.2207458094555289e-05, + "loss": 1.1052, + "step": 18943 + }, + { + "epoch": 0.45, + "grad_norm": 1.9845211443274147, + "learning_rate": 1.2206713870244549e-05, + "loss": 1.0152, + "step": 18944 + }, + { + "epoch": 0.45, + "grad_norm": 2.7320008479269764, + "learning_rate": 1.2205969633085621e-05, + "loss": 1.0416, + "step": 18945 + }, + { + "epoch": 0.45, + "grad_norm": 1.9032111821050663, + "learning_rate": 1.2205225383082844e-05, + "loss": 0.9493, + "step": 18946 + }, + { + "epoch": 0.45, + "grad_norm": 1.8355301321349426, + "learning_rate": 1.220448112024054e-05, + "loss": 1.0941, + "step": 18947 + }, + { + "epoch": 0.45, + "grad_norm": 2.5797577270723693, + "learning_rate": 1.2203736844563055e-05, + "loss": 1.0894, + "step": 18948 + }, + { + "epoch": 0.45, + "grad_norm": 1.8455668906235976, + "learning_rate": 1.2202992556054709e-05, + "loss": 0.9881, + "step": 18949 + }, + { + "epoch": 0.45, + "grad_norm": 1.7998819519602804, + "learning_rate": 1.2202248254719849e-05, + "loss": 1.0579, + "step": 18950 + }, + { + "epoch": 0.45, + "grad_norm": 1.1215063658298503, + "learning_rate": 1.2201503940562798e-05, + "loss": 0.922, + "step": 18951 + }, + { + "epoch": 0.45, + "grad_norm": 2.087731076338911, + "learning_rate": 1.22007596135879e-05, + "loss": 1.0337, + "step": 18952 + }, + { + "epoch": 0.45, + "grad_norm": 2.7664540334308683, + "learning_rate": 1.220001527379948e-05, + "loss": 1.1099, + "step": 18953 + }, + { + "epoch": 0.45, + "grad_norm": 2.3690950962450077, + "learning_rate": 1.2199270921201873e-05, + "loss": 1.039, + "step": 18954 + }, + { + "epoch": 0.45, + "grad_norm": 1.8833169000572265, + "learning_rate": 1.2198526555799415e-05, + "loss": 1.0345, + "step": 18955 + }, + { + "epoch": 0.45, + "grad_norm": 2.146475478215476, + "learning_rate": 1.219778217759644e-05, + "loss": 0.9786, + "step": 18956 + }, + { + "epoch": 0.45, + "grad_norm": 2.0423822186673366, + "learning_rate": 1.2197037786597278e-05, + "loss": 0.9446, + "step": 18957 + }, + { + "epoch": 0.45, + "grad_norm": 2.4434411671172556, + "learning_rate": 1.219629338280627e-05, + "loss": 0.9959, + "step": 18958 + }, + { + "epoch": 0.45, + "grad_norm": 1.8399882278844464, + "learning_rate": 1.2195548966227749e-05, + "loss": 1.0026, + "step": 18959 + }, + { + "epoch": 0.45, + "grad_norm": 2.2523124877278113, + "learning_rate": 1.2194804536866043e-05, + "loss": 1.0703, + "step": 18960 + }, + { + "epoch": 0.45, + "grad_norm": 2.4328002439678964, + "learning_rate": 1.219406009472549e-05, + "loss": 1.1005, + "step": 18961 + }, + { + "epoch": 0.45, + "grad_norm": 2.495588029473234, + "learning_rate": 1.2193315639810426e-05, + "loss": 1.1108, + "step": 18962 + }, + { + "epoch": 0.45, + "grad_norm": 2.3098662808769372, + "learning_rate": 1.2192571172125182e-05, + "loss": 1.0236, + "step": 18963 + }, + { + "epoch": 0.45, + "grad_norm": 3.0731213075849544, + "learning_rate": 1.2191826691674096e-05, + "loss": 0.8894, + "step": 18964 + }, + { + "epoch": 0.45, + "grad_norm": 1.9809119778536293, + "learning_rate": 1.2191082198461502e-05, + "loss": 1.1185, + "step": 18965 + }, + { + "epoch": 0.45, + "grad_norm": 2.0908238042546583, + "learning_rate": 1.2190337692491732e-05, + "loss": 1.0085, + "step": 18966 + }, + { + "epoch": 0.45, + "grad_norm": 2.021075288871115, + "learning_rate": 1.2189593173769126e-05, + "loss": 1.0143, + "step": 18967 + }, + { + "epoch": 0.45, + "grad_norm": 2.614396819482715, + "learning_rate": 1.218884864229801e-05, + "loss": 1.0019, + "step": 18968 + }, + { + "epoch": 0.45, + "grad_norm": 2.172078034375466, + "learning_rate": 1.218810409808273e-05, + "loss": 1.0646, + "step": 18969 + }, + { + "epoch": 0.45, + "grad_norm": 2.032344677543377, + "learning_rate": 1.2187359541127607e-05, + "loss": 0.9681, + "step": 18970 + }, + { + "epoch": 0.45, + "grad_norm": 2.2634095307565443, + "learning_rate": 1.2186614971436992e-05, + "loss": 1.0187, + "step": 18971 + }, + { + "epoch": 0.45, + "grad_norm": 1.9581268016352662, + "learning_rate": 1.2185870389015208e-05, + "loss": 0.95, + "step": 18972 + }, + { + "epoch": 0.45, + "grad_norm": 3.3405620802844394, + "learning_rate": 1.21851257938666e-05, + "loss": 1.0974, + "step": 18973 + }, + { + "epoch": 0.45, + "grad_norm": 1.9715093552204066, + "learning_rate": 1.2184381185995494e-05, + "loss": 1.0667, + "step": 18974 + }, + { + "epoch": 0.45, + "grad_norm": 1.1031705766281361, + "learning_rate": 1.2183636565406227e-05, + "loss": 1.0376, + "step": 18975 + }, + { + "epoch": 0.45, + "grad_norm": 1.9764398437941129, + "learning_rate": 1.2182891932103138e-05, + "loss": 0.9422, + "step": 18976 + }, + { + "epoch": 0.45, + "grad_norm": 1.1812683113983566, + "learning_rate": 1.2182147286090562e-05, + "loss": 0.9554, + "step": 18977 + }, + { + "epoch": 0.45, + "grad_norm": 1.1554333728484023, + "learning_rate": 1.2181402627372834e-05, + "loss": 1.0367, + "step": 18978 + }, + { + "epoch": 0.45, + "grad_norm": 2.2953062447822337, + "learning_rate": 1.2180657955954287e-05, + "loss": 1.0032, + "step": 18979 + }, + { + "epoch": 0.45, + "grad_norm": 2.121958547199672, + "learning_rate": 1.2179913271839264e-05, + "loss": 0.9946, + "step": 18980 + }, + { + "epoch": 0.45, + "grad_norm": 1.9106139807692197, + "learning_rate": 1.2179168575032088e-05, + "loss": 1.0887, + "step": 18981 + }, + { + "epoch": 0.45, + "grad_norm": 1.0336894761501305, + "learning_rate": 1.217842386553711e-05, + "loss": 0.8853, + "step": 18982 + }, + { + "epoch": 0.45, + "grad_norm": 2.3195526417333316, + "learning_rate": 1.2177679143358655e-05, + "loss": 0.9921, + "step": 18983 + }, + { + "epoch": 0.45, + "grad_norm": 2.4265698164380067, + "learning_rate": 1.2176934408501065e-05, + "loss": 0.9893, + "step": 18984 + }, + { + "epoch": 0.45, + "grad_norm": 2.701546909920639, + "learning_rate": 1.2176189660968671e-05, + "loss": 1.0225, + "step": 18985 + }, + { + "epoch": 0.45, + "grad_norm": 2.055557685419261, + "learning_rate": 1.2175444900765811e-05, + "loss": 0.9641, + "step": 18986 + }, + { + "epoch": 0.45, + "grad_norm": 2.3072840156084773, + "learning_rate": 1.2174700127896826e-05, + "loss": 0.9865, + "step": 18987 + }, + { + "epoch": 0.45, + "grad_norm": 1.8674615987747851, + "learning_rate": 1.2173955342366046e-05, + "loss": 1.1305, + "step": 18988 + }, + { + "epoch": 0.45, + "grad_norm": 4.205916781098807, + "learning_rate": 1.2173210544177809e-05, + "loss": 1.0447, + "step": 18989 + }, + { + "epoch": 0.45, + "grad_norm": 1.9825500235456452, + "learning_rate": 1.2172465733336453e-05, + "loss": 0.8912, + "step": 18990 + }, + { + "epoch": 0.45, + "grad_norm": 3.6352759254921065, + "learning_rate": 1.2171720909846314e-05, + "loss": 0.9169, + "step": 18991 + }, + { + "epoch": 0.45, + "grad_norm": 1.9325217800049226, + "learning_rate": 1.217097607371173e-05, + "loss": 1.0124, + "step": 18992 + }, + { + "epoch": 0.45, + "grad_norm": 2.332290841834783, + "learning_rate": 1.2170231224937032e-05, + "loss": 1.1247, + "step": 18993 + }, + { + "epoch": 0.45, + "grad_norm": 2.1032600766723304, + "learning_rate": 1.2169486363526564e-05, + "loss": 1.0755, + "step": 18994 + }, + { + "epoch": 0.45, + "grad_norm": 2.181750861779158, + "learning_rate": 1.2168741489484656e-05, + "loss": 0.9652, + "step": 18995 + }, + { + "epoch": 0.45, + "grad_norm": 2.169581892764733, + "learning_rate": 1.2167996602815652e-05, + "loss": 0.9965, + "step": 18996 + }, + { + "epoch": 0.45, + "grad_norm": 1.9706444436193729, + "learning_rate": 1.2167251703523886e-05, + "loss": 0.9743, + "step": 18997 + }, + { + "epoch": 0.45, + "grad_norm": 1.0841987558965074, + "learning_rate": 1.216650679161369e-05, + "loss": 1.0011, + "step": 18998 + }, + { + "epoch": 0.45, + "grad_norm": 2.243697678260518, + "learning_rate": 1.2165761867089411e-05, + "loss": 1.0427, + "step": 18999 + }, + { + "epoch": 0.45, + "grad_norm": 2.0611954769915113, + "learning_rate": 1.2165016929955376e-05, + "loss": 1.0221, + "step": 19000 + }, + { + "epoch": 0.45, + "grad_norm": 2.1099124003128504, + "learning_rate": 1.2164271980215931e-05, + "loss": 0.9027, + "step": 19001 + }, + { + "epoch": 0.45, + "grad_norm": 2.2445581551347367, + "learning_rate": 1.2163527017875406e-05, + "loss": 1.0195, + "step": 19002 + }, + { + "epoch": 0.45, + "grad_norm": 2.1270733560474824, + "learning_rate": 1.2162782042938144e-05, + "loss": 0.9044, + "step": 19003 + }, + { + "epoch": 0.45, + "grad_norm": 2.039644586329709, + "learning_rate": 1.216203705540848e-05, + "loss": 0.803, + "step": 19004 + }, + { + "epoch": 0.45, + "grad_norm": 2.116181030224939, + "learning_rate": 1.2161292055290751e-05, + "loss": 1.0169, + "step": 19005 + }, + { + "epoch": 0.45, + "grad_norm": 1.9075324520797794, + "learning_rate": 1.2160547042589294e-05, + "loss": 1.0643, + "step": 19006 + }, + { + "epoch": 0.45, + "grad_norm": 1.985775916761512, + "learning_rate": 1.2159802017308451e-05, + "loss": 1.0934, + "step": 19007 + }, + { + "epoch": 0.45, + "grad_norm": 2.100119016805231, + "learning_rate": 1.2159056979452555e-05, + "loss": 1.1256, + "step": 19008 + }, + { + "epoch": 0.45, + "grad_norm": 1.5627290790239086, + "learning_rate": 1.2158311929025945e-05, + "loss": 1.048, + "step": 19009 + }, + { + "epoch": 0.45, + "grad_norm": 3.041419757144987, + "learning_rate": 1.215756686603296e-05, + "loss": 1.0308, + "step": 19010 + }, + { + "epoch": 0.45, + "grad_norm": 2.6784482606193047, + "learning_rate": 1.2156821790477939e-05, + "loss": 0.9304, + "step": 19011 + }, + { + "epoch": 0.45, + "grad_norm": 2.1902724022291453, + "learning_rate": 1.2156076702365219e-05, + "loss": 0.9601, + "step": 19012 + }, + { + "epoch": 0.45, + "grad_norm": 2.1905106094483546, + "learning_rate": 1.2155331601699136e-05, + "loss": 0.9477, + "step": 19013 + }, + { + "epoch": 0.45, + "grad_norm": 2.1937338705926916, + "learning_rate": 1.2154586488484033e-05, + "loss": 1.0906, + "step": 19014 + }, + { + "epoch": 0.45, + "grad_norm": 1.910757996114441, + "learning_rate": 1.2153841362724244e-05, + "loss": 1.0019, + "step": 19015 + }, + { + "epoch": 0.45, + "grad_norm": 2.195527363275165, + "learning_rate": 1.2153096224424106e-05, + "loss": 1.0942, + "step": 19016 + }, + { + "epoch": 0.45, + "grad_norm": 2.052985667092511, + "learning_rate": 1.2152351073587963e-05, + "loss": 1.0136, + "step": 19017 + }, + { + "epoch": 0.45, + "grad_norm": 1.9414575794743332, + "learning_rate": 1.2151605910220153e-05, + "loss": 1.0323, + "step": 19018 + }, + { + "epoch": 0.45, + "grad_norm": 1.903801818100837, + "learning_rate": 1.2150860734325005e-05, + "loss": 1.0494, + "step": 19019 + }, + { + "epoch": 0.45, + "grad_norm": 1.9240883645143023, + "learning_rate": 1.2150115545906871e-05, + "loss": 0.911, + "step": 19020 + }, + { + "epoch": 0.45, + "grad_norm": 1.075177439944736, + "learning_rate": 1.2149370344970079e-05, + "loss": 0.9248, + "step": 19021 + }, + { + "epoch": 0.45, + "grad_norm": 2.0441306350731065, + "learning_rate": 1.2148625131518976e-05, + "loss": 1.0095, + "step": 19022 + }, + { + "epoch": 0.45, + "grad_norm": 1.8855554683575229, + "learning_rate": 1.2147879905557896e-05, + "loss": 0.9328, + "step": 19023 + }, + { + "epoch": 0.45, + "grad_norm": 2.007875226994705, + "learning_rate": 1.2147134667091181e-05, + "loss": 0.9915, + "step": 19024 + }, + { + "epoch": 0.45, + "grad_norm": 1.8621292881814153, + "learning_rate": 1.2146389416123166e-05, + "loss": 0.9839, + "step": 19025 + }, + { + "epoch": 0.45, + "grad_norm": 2.0241144971215053, + "learning_rate": 1.2145644152658191e-05, + "loss": 0.9619, + "step": 19026 + }, + { + "epoch": 0.45, + "grad_norm": 1.0358176426177508, + "learning_rate": 1.2144898876700599e-05, + "loss": 0.9897, + "step": 19027 + }, + { + "epoch": 0.45, + "grad_norm": 1.8906191967892843, + "learning_rate": 1.2144153588254723e-05, + "loss": 1.0899, + "step": 19028 + }, + { + "epoch": 0.45, + "grad_norm": 1.9442823843092771, + "learning_rate": 1.2143408287324912e-05, + "loss": 1.1029, + "step": 19029 + }, + { + "epoch": 0.45, + "grad_norm": 2.048980737662493, + "learning_rate": 1.2142662973915494e-05, + "loss": 1.0492, + "step": 19030 + }, + { + "epoch": 0.45, + "grad_norm": 2.035886417661682, + "learning_rate": 1.214191764803082e-05, + "loss": 1.0637, + "step": 19031 + }, + { + "epoch": 0.45, + "grad_norm": 1.8308001025917207, + "learning_rate": 1.2141172309675218e-05, + "loss": 1.1632, + "step": 19032 + }, + { + "epoch": 0.45, + "grad_norm": 2.1505320775986743, + "learning_rate": 1.2140426958853034e-05, + "loss": 0.966, + "step": 19033 + }, + { + "epoch": 0.45, + "grad_norm": 1.9837584482243633, + "learning_rate": 1.2139681595568606e-05, + "loss": 0.9505, + "step": 19034 + }, + { + "epoch": 0.45, + "grad_norm": 2.2108466690802646, + "learning_rate": 1.2138936219826273e-05, + "loss": 1.0564, + "step": 19035 + }, + { + "epoch": 0.45, + "grad_norm": 2.0718441558411227, + "learning_rate": 1.2138190831630381e-05, + "loss": 1.1378, + "step": 19036 + }, + { + "epoch": 0.45, + "grad_norm": 2.0429814139533446, + "learning_rate": 1.2137445430985263e-05, + "loss": 1.0379, + "step": 19037 + }, + { + "epoch": 0.45, + "grad_norm": 2.0230957092038597, + "learning_rate": 1.213670001789526e-05, + "loss": 1.073, + "step": 19038 + }, + { + "epoch": 0.45, + "grad_norm": 1.9335022628940473, + "learning_rate": 1.2135954592364712e-05, + "loss": 1.0349, + "step": 19039 + }, + { + "epoch": 0.45, + "grad_norm": 2.0020262776942523, + "learning_rate": 1.2135209154397962e-05, + "loss": 0.987, + "step": 19040 + }, + { + "epoch": 0.45, + "grad_norm": 1.0727422107619826, + "learning_rate": 1.2134463703999346e-05, + "loss": 0.9405, + "step": 19041 + }, + { + "epoch": 0.45, + "grad_norm": 1.9127898890226624, + "learning_rate": 1.2133718241173209e-05, + "loss": 0.8741, + "step": 19042 + }, + { + "epoch": 0.45, + "grad_norm": 2.362869413112423, + "learning_rate": 1.213297276592389e-05, + "loss": 1.0034, + "step": 19043 + }, + { + "epoch": 0.45, + "grad_norm": 1.979693398320083, + "learning_rate": 1.2132227278255727e-05, + "loss": 1.1233, + "step": 19044 + }, + { + "epoch": 0.45, + "grad_norm": 2.065249637214021, + "learning_rate": 1.213148177817306e-05, + "loss": 0.9755, + "step": 19045 + }, + { + "epoch": 0.45, + "grad_norm": 2.478874183660256, + "learning_rate": 1.2130736265680231e-05, + "loss": 1.0254, + "step": 19046 + }, + { + "epoch": 0.45, + "grad_norm": 2.2527622278271653, + "learning_rate": 1.2129990740781583e-05, + "loss": 1.047, + "step": 19047 + }, + { + "epoch": 0.45, + "grad_norm": 1.9316643354495828, + "learning_rate": 1.2129245203481457e-05, + "loss": 1.01, + "step": 19048 + }, + { + "epoch": 0.45, + "grad_norm": 1.9877919172716652, + "learning_rate": 1.2128499653784184e-05, + "loss": 1.0016, + "step": 19049 + }, + { + "epoch": 0.45, + "grad_norm": 2.1607614345498454, + "learning_rate": 1.2127754091694119e-05, + "loss": 0.9222, + "step": 19050 + }, + { + "epoch": 0.45, + "grad_norm": 2.2529196626326904, + "learning_rate": 1.2127008517215593e-05, + "loss": 0.9461, + "step": 19051 + }, + { + "epoch": 0.45, + "grad_norm": 1.7075268123074403, + "learning_rate": 1.2126262930352953e-05, + "loss": 0.9417, + "step": 19052 + }, + { + "epoch": 0.45, + "grad_norm": 2.3200728038964846, + "learning_rate": 1.2125517331110532e-05, + "loss": 1.1359, + "step": 19053 + }, + { + "epoch": 0.45, + "grad_norm": 2.0070864026891417, + "learning_rate": 1.2124771719492681e-05, + "loss": 1.011, + "step": 19054 + }, + { + "epoch": 0.45, + "grad_norm": 2.258388885738308, + "learning_rate": 1.2124026095503735e-05, + "loss": 1.0659, + "step": 19055 + }, + { + "epoch": 0.45, + "grad_norm": 1.8767661151350996, + "learning_rate": 1.2123280459148034e-05, + "loss": 0.9761, + "step": 19056 + }, + { + "epoch": 0.45, + "grad_norm": 2.215297004094044, + "learning_rate": 1.2122534810429922e-05, + "loss": 1.0776, + "step": 19057 + }, + { + "epoch": 0.45, + "grad_norm": 1.049129118343181, + "learning_rate": 1.2121789149353745e-05, + "loss": 1.0002, + "step": 19058 + }, + { + "epoch": 0.45, + "grad_norm": 1.1815404268217407, + "learning_rate": 1.2121043475923834e-05, + "loss": 0.897, + "step": 19059 + }, + { + "epoch": 0.45, + "grad_norm": 2.3329756148916085, + "learning_rate": 1.2120297790144539e-05, + "loss": 1.0704, + "step": 19060 + }, + { + "epoch": 0.45, + "grad_norm": 2.5472079871883797, + "learning_rate": 1.21195520920202e-05, + "loss": 1.0659, + "step": 19061 + }, + { + "epoch": 0.45, + "grad_norm": 2.60181460746683, + "learning_rate": 1.2118806381555157e-05, + "loss": 1.0399, + "step": 19062 + }, + { + "epoch": 0.45, + "grad_norm": 1.8189400602076653, + "learning_rate": 1.2118060658753751e-05, + "loss": 1.0878, + "step": 19063 + }, + { + "epoch": 0.45, + "grad_norm": 2.1290084161435505, + "learning_rate": 1.2117314923620326e-05, + "loss": 1.0279, + "step": 19064 + }, + { + "epoch": 0.45, + "grad_norm": 2.2347169003670087, + "learning_rate": 1.2116569176159225e-05, + "loss": 0.9469, + "step": 19065 + }, + { + "epoch": 0.45, + "grad_norm": 2.203925281830917, + "learning_rate": 1.2115823416374788e-05, + "loss": 0.8889, + "step": 19066 + }, + { + "epoch": 0.45, + "grad_norm": 1.8952704131229978, + "learning_rate": 1.2115077644271356e-05, + "loss": 1.0771, + "step": 19067 + }, + { + "epoch": 0.45, + "grad_norm": 1.7724412528171853, + "learning_rate": 1.2114331859853272e-05, + "loss": 1.1436, + "step": 19068 + }, + { + "epoch": 0.45, + "grad_norm": 2.0064765455224443, + "learning_rate": 1.211358606312488e-05, + "loss": 1.0507, + "step": 19069 + }, + { + "epoch": 0.45, + "grad_norm": 1.9603941983926108, + "learning_rate": 1.2112840254090519e-05, + "loss": 1.0289, + "step": 19070 + }, + { + "epoch": 0.45, + "grad_norm": 1.9685740484505616, + "learning_rate": 1.2112094432754536e-05, + "loss": 1.0674, + "step": 19071 + }, + { + "epoch": 0.45, + "grad_norm": 1.8869623134107245, + "learning_rate": 1.2111348599121266e-05, + "loss": 1.0627, + "step": 19072 + }, + { + "epoch": 0.45, + "grad_norm": 1.1306869973768994, + "learning_rate": 1.211060275319506e-05, + "loss": 1.0679, + "step": 19073 + }, + { + "epoch": 0.45, + "grad_norm": 2.159151214947162, + "learning_rate": 1.2109856894980254e-05, + "loss": 1.0046, + "step": 19074 + }, + { + "epoch": 0.45, + "grad_norm": 2.4942714719198333, + "learning_rate": 1.2109111024481194e-05, + "loss": 1.0228, + "step": 19075 + }, + { + "epoch": 0.45, + "grad_norm": 2.0528283566698398, + "learning_rate": 1.2108365141702224e-05, + "loss": 1.0152, + "step": 19076 + }, + { + "epoch": 0.45, + "grad_norm": 2.5712899608480098, + "learning_rate": 1.2107619246647679e-05, + "loss": 1.0575, + "step": 19077 + }, + { + "epoch": 0.45, + "grad_norm": 2.2840625959012537, + "learning_rate": 1.2106873339321914e-05, + "loss": 1.0762, + "step": 19078 + }, + { + "epoch": 0.45, + "grad_norm": 1.0801994995363253, + "learning_rate": 1.210612741972926e-05, + "loss": 0.9522, + "step": 19079 + }, + { + "epoch": 0.45, + "grad_norm": 1.7015177571386155, + "learning_rate": 1.210538148787407e-05, + "loss": 0.9314, + "step": 19080 + }, + { + "epoch": 0.45, + "grad_norm": 2.005652154407372, + "learning_rate": 1.210463554376068e-05, + "loss": 1.0589, + "step": 19081 + }, + { + "epoch": 0.45, + "grad_norm": 1.9487186915773689, + "learning_rate": 1.2103889587393437e-05, + "loss": 0.9951, + "step": 19082 + }, + { + "epoch": 0.45, + "grad_norm": 1.9870128942246077, + "learning_rate": 1.2103143618776682e-05, + "loss": 1.0528, + "step": 19083 + }, + { + "epoch": 0.45, + "grad_norm": 1.9686252541109037, + "learning_rate": 1.2102397637914758e-05, + "loss": 1.0607, + "step": 19084 + }, + { + "epoch": 0.45, + "grad_norm": 2.061198414376059, + "learning_rate": 1.210165164481201e-05, + "loss": 0.9098, + "step": 19085 + }, + { + "epoch": 0.45, + "grad_norm": 1.9961406371923522, + "learning_rate": 1.210090563947278e-05, + "loss": 0.9307, + "step": 19086 + }, + { + "epoch": 0.45, + "grad_norm": 2.0931235035984286, + "learning_rate": 1.2100159621901413e-05, + "loss": 1.0414, + "step": 19087 + }, + { + "epoch": 0.45, + "grad_norm": 1.187676569946506, + "learning_rate": 1.2099413592102255e-05, + "loss": 0.9981, + "step": 19088 + }, + { + "epoch": 0.45, + "grad_norm": 1.0711103544993263, + "learning_rate": 1.2098667550079642e-05, + "loss": 1.0136, + "step": 19089 + }, + { + "epoch": 0.45, + "grad_norm": 1.8349707731599503, + "learning_rate": 1.2097921495837923e-05, + "loss": 1.0752, + "step": 19090 + }, + { + "epoch": 0.45, + "grad_norm": 2.0783207592104658, + "learning_rate": 1.209717542938144e-05, + "loss": 1.0851, + "step": 19091 + }, + { + "epoch": 0.45, + "grad_norm": 2.247752796275908, + "learning_rate": 1.2096429350714538e-05, + "loss": 0.9654, + "step": 19092 + }, + { + "epoch": 0.45, + "grad_norm": 2.065920091581505, + "learning_rate": 1.2095683259841562e-05, + "loss": 1.0055, + "step": 19093 + }, + { + "epoch": 0.45, + "grad_norm": 1.0380287655951226, + "learning_rate": 1.2094937156766855e-05, + "loss": 0.9263, + "step": 19094 + }, + { + "epoch": 0.45, + "grad_norm": 1.9269296598605503, + "learning_rate": 1.2094191041494759e-05, + "loss": 1.0454, + "step": 19095 + }, + { + "epoch": 0.45, + "grad_norm": 1.9865360992747476, + "learning_rate": 1.2093444914029622e-05, + "loss": 1.1075, + "step": 19096 + }, + { + "epoch": 0.45, + "grad_norm": 2.0773956580346002, + "learning_rate": 1.2092698774375786e-05, + "loss": 1.0329, + "step": 19097 + }, + { + "epoch": 0.45, + "grad_norm": 2.0550357480194528, + "learning_rate": 1.209195262253759e-05, + "loss": 1.0313, + "step": 19098 + }, + { + "epoch": 0.45, + "grad_norm": 1.979792870831243, + "learning_rate": 1.2091206458519389e-05, + "loss": 1.0458, + "step": 19099 + }, + { + "epoch": 0.45, + "grad_norm": 2.234407016947412, + "learning_rate": 1.2090460282325518e-05, + "loss": 0.9966, + "step": 19100 + }, + { + "epoch": 0.45, + "grad_norm": 2.039719922885944, + "learning_rate": 1.208971409396033e-05, + "loss": 1.0422, + "step": 19101 + }, + { + "epoch": 0.45, + "grad_norm": 1.8354714790074145, + "learning_rate": 1.2088967893428161e-05, + "loss": 0.8256, + "step": 19102 + }, + { + "epoch": 0.45, + "grad_norm": 2.1168870430081967, + "learning_rate": 1.2088221680733362e-05, + "loss": 1.024, + "step": 19103 + }, + { + "epoch": 0.45, + "grad_norm": 1.8834329741184497, + "learning_rate": 1.2087475455880272e-05, + "loss": 1.0174, + "step": 19104 + }, + { + "epoch": 0.45, + "grad_norm": 2.4804905317882042, + "learning_rate": 1.2086729218873242e-05, + "loss": 1.1399, + "step": 19105 + }, + { + "epoch": 0.45, + "grad_norm": 2.024885351051318, + "learning_rate": 1.2085982969716615e-05, + "loss": 0.963, + "step": 19106 + }, + { + "epoch": 0.45, + "grad_norm": 1.8507336534915158, + "learning_rate": 1.208523670841473e-05, + "loss": 1.1125, + "step": 19107 + }, + { + "epoch": 0.45, + "grad_norm": 1.9417065242134337, + "learning_rate": 1.2084490434971942e-05, + "loss": 1.073, + "step": 19108 + }, + { + "epoch": 0.45, + "grad_norm": 2.525050400528073, + "learning_rate": 1.2083744149392586e-05, + "loss": 1.0442, + "step": 19109 + }, + { + "epoch": 0.45, + "grad_norm": 2.5091551856558487, + "learning_rate": 1.2082997851681015e-05, + "loss": 0.9929, + "step": 19110 + }, + { + "epoch": 0.45, + "grad_norm": 1.8291692569123554, + "learning_rate": 1.2082251541841573e-05, + "loss": 0.9786, + "step": 19111 + }, + { + "epoch": 0.45, + "grad_norm": 1.9452434042428193, + "learning_rate": 1.2081505219878597e-05, + "loss": 0.9308, + "step": 19112 + }, + { + "epoch": 0.45, + "grad_norm": 2.1528802583163302, + "learning_rate": 1.2080758885796444e-05, + "loss": 0.9617, + "step": 19113 + }, + { + "epoch": 0.45, + "grad_norm": 2.185224262116091, + "learning_rate": 1.208001253959945e-05, + "loss": 0.969, + "step": 19114 + }, + { + "epoch": 0.45, + "grad_norm": 1.7318105475797463, + "learning_rate": 1.207926618129197e-05, + "loss": 0.9782, + "step": 19115 + }, + { + "epoch": 0.45, + "grad_norm": 1.8640393917380291, + "learning_rate": 1.207851981087834e-05, + "loss": 0.9624, + "step": 19116 + }, + { + "epoch": 0.45, + "grad_norm": 2.0190027257144547, + "learning_rate": 1.2077773428362909e-05, + "loss": 0.9917, + "step": 19117 + }, + { + "epoch": 0.45, + "grad_norm": 2.279602995901436, + "learning_rate": 1.2077027033750025e-05, + "loss": 1.0968, + "step": 19118 + }, + { + "epoch": 0.45, + "grad_norm": 2.03296868097607, + "learning_rate": 1.207628062704403e-05, + "loss": 1.0663, + "step": 19119 + }, + { + "epoch": 0.45, + "grad_norm": 1.793405465107304, + "learning_rate": 1.2075534208249275e-05, + "loss": 1.0048, + "step": 19120 + }, + { + "epoch": 0.45, + "grad_norm": 1.9243548149122411, + "learning_rate": 1.2074787777370101e-05, + "loss": 1.0159, + "step": 19121 + }, + { + "epoch": 0.45, + "grad_norm": 2.0637001883636654, + "learning_rate": 1.2074041334410858e-05, + "loss": 1.0481, + "step": 19122 + }, + { + "epoch": 0.45, + "grad_norm": 8.172450702438061, + "learning_rate": 1.2073294879375883e-05, + "loss": 1.0017, + "step": 19123 + }, + { + "epoch": 0.45, + "grad_norm": 1.9261599104600013, + "learning_rate": 1.2072548412269536e-05, + "loss": 1.0642, + "step": 19124 + }, + { + "epoch": 0.45, + "grad_norm": 1.9353442442075466, + "learning_rate": 1.2071801933096153e-05, + "loss": 1.1097, + "step": 19125 + }, + { + "epoch": 0.45, + "grad_norm": 1.9305338391141806, + "learning_rate": 1.2071055441860082e-05, + "loss": 1.0846, + "step": 19126 + }, + { + "epoch": 0.45, + "grad_norm": 2.236475934220475, + "learning_rate": 1.2070308938565674e-05, + "loss": 1.0111, + "step": 19127 + }, + { + "epoch": 0.45, + "grad_norm": 1.81458156724434, + "learning_rate": 1.2069562423217268e-05, + "loss": 0.998, + "step": 19128 + }, + { + "epoch": 0.45, + "grad_norm": 1.937542719475798, + "learning_rate": 1.2068815895819219e-05, + "loss": 1.0731, + "step": 19129 + }, + { + "epoch": 0.45, + "grad_norm": 2.015114446045059, + "learning_rate": 1.2068069356375864e-05, + "loss": 0.9924, + "step": 19130 + }, + { + "epoch": 0.45, + "grad_norm": 1.135529819922417, + "learning_rate": 1.2067322804891556e-05, + "loss": 0.9902, + "step": 19131 + }, + { + "epoch": 0.45, + "grad_norm": 2.099353334096147, + "learning_rate": 1.206657624137064e-05, + "loss": 1.1589, + "step": 19132 + }, + { + "epoch": 0.45, + "grad_norm": 2.397219109721068, + "learning_rate": 1.2065829665817465e-05, + "loss": 1.0058, + "step": 19133 + }, + { + "epoch": 0.45, + "grad_norm": 1.9144669539719394, + "learning_rate": 1.2065083078236375e-05, + "loss": 1.1046, + "step": 19134 + }, + { + "epoch": 0.45, + "grad_norm": 1.8002370530020282, + "learning_rate": 1.2064336478631717e-05, + "loss": 1.125, + "step": 19135 + }, + { + "epoch": 0.45, + "grad_norm": 1.8396278996814102, + "learning_rate": 1.2063589867007838e-05, + "loss": 1.0561, + "step": 19136 + }, + { + "epoch": 0.45, + "grad_norm": 4.323197335142399, + "learning_rate": 1.2062843243369086e-05, + "loss": 1.0133, + "step": 19137 + }, + { + "epoch": 0.45, + "grad_norm": 2.173582447762296, + "learning_rate": 1.2062096607719809e-05, + "loss": 0.9701, + "step": 19138 + }, + { + "epoch": 0.45, + "grad_norm": 2.028274098371681, + "learning_rate": 1.206134996006435e-05, + "loss": 1.0148, + "step": 19139 + }, + { + "epoch": 0.45, + "grad_norm": 2.497677432073485, + "learning_rate": 1.206060330040706e-05, + "loss": 0.9314, + "step": 19140 + }, + { + "epoch": 0.45, + "grad_norm": 1.8902072965994554, + "learning_rate": 1.2059856628752285e-05, + "loss": 1.0985, + "step": 19141 + }, + { + "epoch": 0.45, + "grad_norm": 2.395364912553222, + "learning_rate": 1.2059109945104373e-05, + "loss": 0.961, + "step": 19142 + }, + { + "epoch": 0.45, + "grad_norm": 2.0909935760511393, + "learning_rate": 1.2058363249467671e-05, + "loss": 0.9412, + "step": 19143 + }, + { + "epoch": 0.45, + "grad_norm": 2.202783876446911, + "learning_rate": 1.2057616541846526e-05, + "loss": 1.0482, + "step": 19144 + }, + { + "epoch": 0.45, + "grad_norm": 1.9409514137001638, + "learning_rate": 1.2056869822245285e-05, + "loss": 0.9087, + "step": 19145 + }, + { + "epoch": 0.45, + "grad_norm": 1.9588431218747782, + "learning_rate": 1.20561230906683e-05, + "loss": 1.0285, + "step": 19146 + }, + { + "epoch": 0.45, + "grad_norm": 2.3191569597555044, + "learning_rate": 1.2055376347119912e-05, + "loss": 0.9769, + "step": 19147 + }, + { + "epoch": 0.45, + "grad_norm": 1.9854180808728799, + "learning_rate": 1.2054629591604474e-05, + "loss": 0.9983, + "step": 19148 + }, + { + "epoch": 0.45, + "grad_norm": 1.0535279051825595, + "learning_rate": 1.2053882824126331e-05, + "loss": 0.9731, + "step": 19149 + }, + { + "epoch": 0.45, + "grad_norm": 2.2208577907839215, + "learning_rate": 1.2053136044689834e-05, + "loss": 1.0716, + "step": 19150 + }, + { + "epoch": 0.45, + "grad_norm": 2.2612044391192403, + "learning_rate": 1.2052389253299326e-05, + "loss": 0.9945, + "step": 19151 + }, + { + "epoch": 0.45, + "grad_norm": 1.933821016258973, + "learning_rate": 1.2051642449959162e-05, + "loss": 0.9727, + "step": 19152 + }, + { + "epoch": 0.45, + "grad_norm": 2.247543478489581, + "learning_rate": 1.2050895634673681e-05, + "loss": 0.9753, + "step": 19153 + }, + { + "epoch": 0.45, + "grad_norm": 1.7852049216807744, + "learning_rate": 1.2050148807447242e-05, + "loss": 1.0691, + "step": 19154 + }, + { + "epoch": 0.45, + "grad_norm": 1.0363213723042117, + "learning_rate": 1.2049401968284182e-05, + "loss": 0.9341, + "step": 19155 + }, + { + "epoch": 0.45, + "grad_norm": 2.3123446431785104, + "learning_rate": 1.2048655117188859e-05, + "loss": 1.0567, + "step": 19156 + }, + { + "epoch": 0.45, + "grad_norm": 1.1260686574735235, + "learning_rate": 1.2047908254165614e-05, + "loss": 0.994, + "step": 19157 + }, + { + "epoch": 0.45, + "grad_norm": 2.0623684160657176, + "learning_rate": 1.20471613792188e-05, + "loss": 0.9041, + "step": 19158 + }, + { + "epoch": 0.45, + "grad_norm": 1.9509497892966003, + "learning_rate": 1.2046414492352766e-05, + "loss": 1.1042, + "step": 19159 + }, + { + "epoch": 0.45, + "grad_norm": 2.269873985708672, + "learning_rate": 1.2045667593571858e-05, + "loss": 1.1188, + "step": 19160 + }, + { + "epoch": 0.45, + "grad_norm": 2.0363419181838456, + "learning_rate": 1.2044920682880425e-05, + "loss": 1.0171, + "step": 19161 + }, + { + "epoch": 0.45, + "grad_norm": 1.993134104822701, + "learning_rate": 1.2044173760282818e-05, + "loss": 0.9845, + "step": 19162 + }, + { + "epoch": 0.45, + "grad_norm": 2.1302132284623885, + "learning_rate": 1.2043426825783383e-05, + "loss": 0.9571, + "step": 19163 + }, + { + "epoch": 0.45, + "grad_norm": 1.8815332555567181, + "learning_rate": 1.204267987938647e-05, + "loss": 0.9776, + "step": 19164 + }, + { + "epoch": 0.45, + "grad_norm": 2.3120162254861576, + "learning_rate": 1.2041932921096427e-05, + "loss": 0.86, + "step": 19165 + }, + { + "epoch": 0.45, + "grad_norm": 1.9127475132860243, + "learning_rate": 1.2041185950917604e-05, + "loss": 0.9182, + "step": 19166 + }, + { + "epoch": 0.45, + "grad_norm": 2.1257801222517827, + "learning_rate": 1.2040438968854354e-05, + "loss": 1.0238, + "step": 19167 + }, + { + "epoch": 0.45, + "grad_norm": 2.0442521739318713, + "learning_rate": 1.2039691974911019e-05, + "loss": 1.0365, + "step": 19168 + }, + { + "epoch": 0.45, + "grad_norm": 2.0158984822109094, + "learning_rate": 1.2038944969091953e-05, + "loss": 1.0708, + "step": 19169 + }, + { + "epoch": 0.45, + "grad_norm": 1.965118268631633, + "learning_rate": 1.20381979514015e-05, + "loss": 1.0477, + "step": 19170 + }, + { + "epoch": 0.45, + "grad_norm": 2.11475416946895, + "learning_rate": 1.203745092184402e-05, + "loss": 1.0096, + "step": 19171 + }, + { + "epoch": 0.45, + "grad_norm": 1.9343697317799669, + "learning_rate": 1.203670388042385e-05, + "loss": 1.0188, + "step": 19172 + }, + { + "epoch": 0.45, + "grad_norm": 1.1034463036757407, + "learning_rate": 1.203595682714535e-05, + "loss": 0.9781, + "step": 19173 + }, + { + "epoch": 0.45, + "grad_norm": 2.064109082909179, + "learning_rate": 1.2035209762012858e-05, + "loss": 1.0466, + "step": 19174 + }, + { + "epoch": 0.45, + "grad_norm": 1.157587150335608, + "learning_rate": 1.2034462685030734e-05, + "loss": 1.025, + "step": 19175 + }, + { + "epoch": 0.45, + "grad_norm": 1.9313342252313401, + "learning_rate": 1.2033715596203326e-05, + "loss": 1.0836, + "step": 19176 + }, + { + "epoch": 0.45, + "grad_norm": 2.051627743949922, + "learning_rate": 1.203296849553498e-05, + "loss": 1.0994, + "step": 19177 + }, + { + "epoch": 0.45, + "grad_norm": 2.0206256756484944, + "learning_rate": 1.203222138303005e-05, + "loss": 0.9579, + "step": 19178 + }, + { + "epoch": 0.45, + "grad_norm": 2.1309251957586883, + "learning_rate": 1.2031474258692878e-05, + "loss": 1.0091, + "step": 19179 + }, + { + "epoch": 0.45, + "grad_norm": 2.1984446853641937, + "learning_rate": 1.2030727122527828e-05, + "loss": 1.0748, + "step": 19180 + }, + { + "epoch": 0.45, + "grad_norm": 2.026997200337261, + "learning_rate": 1.2029979974539233e-05, + "loss": 1.0229, + "step": 19181 + }, + { + "epoch": 0.45, + "grad_norm": 3.0658632858467856, + "learning_rate": 1.202923281473146e-05, + "loss": 1.043, + "step": 19182 + }, + { + "epoch": 0.45, + "grad_norm": 1.0176026129476032, + "learning_rate": 1.2028485643108848e-05, + "loss": 0.9484, + "step": 19183 + }, + { + "epoch": 0.45, + "grad_norm": 5.7826920032331985, + "learning_rate": 1.202773845967575e-05, + "loss": 1.119, + "step": 19184 + }, + { + "epoch": 0.45, + "grad_norm": 2.0234041785429353, + "learning_rate": 1.2026991264436516e-05, + "loss": 1.0295, + "step": 19185 + }, + { + "epoch": 0.45, + "grad_norm": 1.967870833043035, + "learning_rate": 1.2026244057395497e-05, + "loss": 1.0081, + "step": 19186 + }, + { + "epoch": 0.45, + "grad_norm": 1.8481008013637548, + "learning_rate": 1.2025496838557045e-05, + "loss": 1.1072, + "step": 19187 + }, + { + "epoch": 0.45, + "grad_norm": 1.9272458659022844, + "learning_rate": 1.202474960792551e-05, + "loss": 0.9238, + "step": 19188 + }, + { + "epoch": 0.45, + "grad_norm": 1.8941776701499955, + "learning_rate": 1.2024002365505238e-05, + "loss": 0.9565, + "step": 19189 + }, + { + "epoch": 0.45, + "grad_norm": 1.0438168735954498, + "learning_rate": 1.2023255111300587e-05, + "loss": 0.9435, + "step": 19190 + }, + { + "epoch": 0.45, + "grad_norm": 1.9993625636472698, + "learning_rate": 1.2022507845315903e-05, + "loss": 1.0135, + "step": 19191 + }, + { + "epoch": 0.45, + "grad_norm": 1.910163855244995, + "learning_rate": 1.2021760567555539e-05, + "loss": 0.9051, + "step": 19192 + }, + { + "epoch": 0.45, + "grad_norm": 1.0722067959903316, + "learning_rate": 1.2021013278023843e-05, + "loss": 0.9974, + "step": 19193 + }, + { + "epoch": 0.45, + "grad_norm": 2.625076560606246, + "learning_rate": 1.2020265976725167e-05, + "loss": 1.0363, + "step": 19194 + }, + { + "epoch": 0.45, + "grad_norm": 2.414940255789695, + "learning_rate": 1.2019518663663865e-05, + "loss": 0.9913, + "step": 19195 + }, + { + "epoch": 0.45, + "grad_norm": 2.045535207326117, + "learning_rate": 1.2018771338844287e-05, + "loss": 1.0513, + "step": 19196 + }, + { + "epoch": 0.45, + "grad_norm": 2.094052924476002, + "learning_rate": 1.201802400227078e-05, + "loss": 1.0655, + "step": 19197 + }, + { + "epoch": 0.45, + "grad_norm": 2.1540655433838602, + "learning_rate": 1.2017276653947701e-05, + "loss": 1.0683, + "step": 19198 + }, + { + "epoch": 0.45, + "grad_norm": 1.9376856941789586, + "learning_rate": 1.20165292938794e-05, + "loss": 1.0308, + "step": 19199 + }, + { + "epoch": 0.45, + "grad_norm": 2.1678448597631714, + "learning_rate": 1.2015781922070223e-05, + "loss": 1.1243, + "step": 19200 + }, + { + "epoch": 0.45, + "grad_norm": 2.613691509709989, + "learning_rate": 1.2015034538524528e-05, + "loss": 1.0272, + "step": 19201 + }, + { + "epoch": 0.45, + "grad_norm": 2.116661126103109, + "learning_rate": 1.2014287143246662e-05, + "loss": 0.9161, + "step": 19202 + }, + { + "epoch": 0.45, + "grad_norm": 1.0532138822549972, + "learning_rate": 1.201353973624098e-05, + "loss": 0.91, + "step": 19203 + }, + { + "epoch": 0.45, + "grad_norm": 2.092103750294576, + "learning_rate": 1.201279231751183e-05, + "loss": 1.1063, + "step": 19204 + }, + { + "epoch": 0.45, + "grad_norm": 2.0204512385744406, + "learning_rate": 1.201204488706357e-05, + "loss": 0.9857, + "step": 19205 + }, + { + "epoch": 0.45, + "grad_norm": 1.9687739408168934, + "learning_rate": 1.2011297444900545e-05, + "loss": 0.9615, + "step": 19206 + }, + { + "epoch": 0.45, + "grad_norm": 2.3842140265906173, + "learning_rate": 1.2010549991027109e-05, + "loss": 1.0056, + "step": 19207 + }, + { + "epoch": 0.45, + "grad_norm": 2.026640502238094, + "learning_rate": 1.2009802525447616e-05, + "loss": 1.12, + "step": 19208 + }, + { + "epoch": 0.45, + "grad_norm": 1.6836787182183983, + "learning_rate": 1.2009055048166415e-05, + "loss": 1.0276, + "step": 19209 + }, + { + "epoch": 0.45, + "grad_norm": 2.00371297372238, + "learning_rate": 1.2008307559187862e-05, + "loss": 0.9854, + "step": 19210 + }, + { + "epoch": 0.45, + "grad_norm": 1.922698452928882, + "learning_rate": 1.2007560058516301e-05, + "loss": 1.0423, + "step": 19211 + }, + { + "epoch": 0.45, + "grad_norm": 2.0813151236424767, + "learning_rate": 1.2006812546156097e-05, + "loss": 1.1356, + "step": 19212 + }, + { + "epoch": 0.45, + "grad_norm": 1.869061427870089, + "learning_rate": 1.2006065022111591e-05, + "loss": 1.0902, + "step": 19213 + }, + { + "epoch": 0.45, + "grad_norm": 2.028840192716274, + "learning_rate": 1.200531748638714e-05, + "loss": 0.9726, + "step": 19214 + }, + { + "epoch": 0.45, + "grad_norm": 1.866320186465908, + "learning_rate": 1.2004569938987093e-05, + "loss": 1.0004, + "step": 19215 + }, + { + "epoch": 0.45, + "grad_norm": 2.0809502471366774, + "learning_rate": 1.200382237991581e-05, + "loss": 0.9081, + "step": 19216 + }, + { + "epoch": 0.45, + "grad_norm": 2.145240069418288, + "learning_rate": 1.2003074809177635e-05, + "loss": 1.0231, + "step": 19217 + }, + { + "epoch": 0.45, + "grad_norm": 1.9903823841490504, + "learning_rate": 1.2002327226776926e-05, + "loss": 0.9098, + "step": 19218 + }, + { + "epoch": 0.45, + "grad_norm": 1.9085439654092258, + "learning_rate": 1.2001579632718034e-05, + "loss": 1.0354, + "step": 19219 + }, + { + "epoch": 0.45, + "grad_norm": 2.146716258277463, + "learning_rate": 1.2000832027005314e-05, + "loss": 0.9816, + "step": 19220 + }, + { + "epoch": 0.45, + "grad_norm": 2.1185400162244488, + "learning_rate": 1.200008440964311e-05, + "loss": 1.0356, + "step": 19221 + }, + { + "epoch": 0.45, + "grad_norm": 1.2053104494321945, + "learning_rate": 1.1999336780635788e-05, + "loss": 0.9116, + "step": 19222 + }, + { + "epoch": 0.45, + "grad_norm": 6.832829695536055, + "learning_rate": 1.199858913998769e-05, + "loss": 0.8862, + "step": 19223 + }, + { + "epoch": 0.45, + "grad_norm": 2.132814441953376, + "learning_rate": 1.1997841487703175e-05, + "loss": 0.9615, + "step": 19224 + }, + { + "epoch": 0.45, + "grad_norm": 1.978772269250052, + "learning_rate": 1.1997093823786594e-05, + "loss": 1.1556, + "step": 19225 + }, + { + "epoch": 0.45, + "grad_norm": 1.1253970384332277, + "learning_rate": 1.1996346148242299e-05, + "loss": 0.9627, + "step": 19226 + }, + { + "epoch": 0.45, + "grad_norm": 3.306623505853906, + "learning_rate": 1.1995598461074647e-05, + "loss": 1.0027, + "step": 19227 + }, + { + "epoch": 0.45, + "grad_norm": 1.998040995339816, + "learning_rate": 1.1994850762287988e-05, + "loss": 0.9429, + "step": 19228 + }, + { + "epoch": 0.45, + "grad_norm": 2.599202248859455, + "learning_rate": 1.199410305188668e-05, + "loss": 1.1039, + "step": 19229 + }, + { + "epoch": 0.45, + "grad_norm": 1.0773435281531318, + "learning_rate": 1.1993355329875067e-05, + "loss": 0.903, + "step": 19230 + }, + { + "epoch": 0.45, + "grad_norm": 2.0224166989150643, + "learning_rate": 1.1992607596257513e-05, + "loss": 1.0574, + "step": 19231 + }, + { + "epoch": 0.45, + "grad_norm": 3.1011267466744594, + "learning_rate": 1.1991859851038362e-05, + "loss": 0.8858, + "step": 19232 + }, + { + "epoch": 0.45, + "grad_norm": 2.0527058888862526, + "learning_rate": 1.1991112094221975e-05, + "loss": 0.9793, + "step": 19233 + }, + { + "epoch": 0.45, + "grad_norm": 1.8537084344272838, + "learning_rate": 1.19903643258127e-05, + "loss": 0.9456, + "step": 19234 + }, + { + "epoch": 0.45, + "grad_norm": 1.1223209516591468, + "learning_rate": 1.1989616545814899e-05, + "loss": 0.9989, + "step": 19235 + }, + { + "epoch": 0.45, + "grad_norm": 1.9163321879777842, + "learning_rate": 1.198886875423292e-05, + "loss": 1.0755, + "step": 19236 + }, + { + "epoch": 0.45, + "grad_norm": 2.195269962977085, + "learning_rate": 1.1988120951071113e-05, + "loss": 1.0365, + "step": 19237 + }, + { + "epoch": 0.45, + "grad_norm": 1.1045416286307108, + "learning_rate": 1.198737313633384e-05, + "loss": 1.0212, + "step": 19238 + }, + { + "epoch": 0.45, + "grad_norm": 1.9253404235602762, + "learning_rate": 1.1986625310025451e-05, + "loss": 1.0486, + "step": 19239 + }, + { + "epoch": 0.45, + "grad_norm": 2.7007923470723783, + "learning_rate": 1.1985877472150299e-05, + "loss": 1.0101, + "step": 19240 + }, + { + "epoch": 0.45, + "grad_norm": 2.605477223050096, + "learning_rate": 1.1985129622712738e-05, + "loss": 0.982, + "step": 19241 + }, + { + "epoch": 0.45, + "grad_norm": 3.789416491548965, + "learning_rate": 1.1984381761717127e-05, + "loss": 1.0711, + "step": 19242 + }, + { + "epoch": 0.45, + "grad_norm": 2.1981216270080184, + "learning_rate": 1.1983633889167815e-05, + "loss": 1.0526, + "step": 19243 + }, + { + "epoch": 0.45, + "grad_norm": 3.543902544915634, + "learning_rate": 1.198288600506916e-05, + "loss": 0.8779, + "step": 19244 + }, + { + "epoch": 0.45, + "grad_norm": 2.6966843565699232, + "learning_rate": 1.1982138109425516e-05, + "loss": 1.0358, + "step": 19245 + }, + { + "epoch": 0.45, + "grad_norm": 2.026462019918887, + "learning_rate": 1.1981390202241235e-05, + "loss": 0.9506, + "step": 19246 + }, + { + "epoch": 0.45, + "grad_norm": 1.8931025420319798, + "learning_rate": 1.1980642283520672e-05, + "loss": 1.0319, + "step": 19247 + }, + { + "epoch": 0.45, + "grad_norm": 1.0469942618785264, + "learning_rate": 1.1979894353268185e-05, + "loss": 0.9618, + "step": 19248 + }, + { + "epoch": 0.45, + "grad_norm": 2.0718082610652506, + "learning_rate": 1.1979146411488122e-05, + "loss": 1.0076, + "step": 19249 + }, + { + "epoch": 0.45, + "grad_norm": 2.2013063205677237, + "learning_rate": 1.1978398458184848e-05, + "loss": 1.1143, + "step": 19250 + }, + { + "epoch": 0.45, + "grad_norm": 1.97745712875833, + "learning_rate": 1.1977650493362708e-05, + "loss": 0.9976, + "step": 19251 + }, + { + "epoch": 0.45, + "grad_norm": 1.1369635655894865, + "learning_rate": 1.1976902517026063e-05, + "loss": 0.904, + "step": 19252 + }, + { + "epoch": 0.45, + "grad_norm": 1.9975166186428892, + "learning_rate": 1.1976154529179263e-05, + "loss": 0.9689, + "step": 19253 + }, + { + "epoch": 0.45, + "grad_norm": 1.9667360612223765, + "learning_rate": 1.1975406529826668e-05, + "loss": 0.9065, + "step": 19254 + }, + { + "epoch": 0.45, + "grad_norm": 1.8561024364245216, + "learning_rate": 1.1974658518972625e-05, + "loss": 1.0878, + "step": 19255 + }, + { + "epoch": 0.45, + "grad_norm": 1.8888725602652698, + "learning_rate": 1.1973910496621503e-05, + "loss": 0.9116, + "step": 19256 + }, + { + "epoch": 0.45, + "grad_norm": 1.917066636294088, + "learning_rate": 1.1973162462777645e-05, + "loss": 1.1946, + "step": 19257 + }, + { + "epoch": 0.45, + "grad_norm": 3.4737792637055844, + "learning_rate": 1.197241441744541e-05, + "loss": 0.9615, + "step": 19258 + }, + { + "epoch": 0.45, + "grad_norm": 1.93730925130071, + "learning_rate": 1.1971666360629157e-05, + "loss": 1.1454, + "step": 19259 + }, + { + "epoch": 0.45, + "grad_norm": 1.086217968278151, + "learning_rate": 1.1970918292333234e-05, + "loss": 0.9889, + "step": 19260 + }, + { + "epoch": 0.45, + "grad_norm": 2.075456111564235, + "learning_rate": 1.1970170212562004e-05, + "loss": 1.0155, + "step": 19261 + }, + { + "epoch": 0.45, + "grad_norm": 2.138191779783703, + "learning_rate": 1.1969422121319815e-05, + "loss": 1.0096, + "step": 19262 + }, + { + "epoch": 0.45, + "grad_norm": 1.880969176548555, + "learning_rate": 1.1968674018611033e-05, + "loss": 1.1248, + "step": 19263 + }, + { + "epoch": 0.45, + "grad_norm": 1.9445547707485977, + "learning_rate": 1.1967925904440005e-05, + "loss": 1.0798, + "step": 19264 + }, + { + "epoch": 0.45, + "grad_norm": 2.04852934075604, + "learning_rate": 1.1967177778811087e-05, + "loss": 1.0797, + "step": 19265 + }, + { + "epoch": 0.45, + "grad_norm": 2.332737301109011, + "learning_rate": 1.1966429641728641e-05, + "loss": 1.0073, + "step": 19266 + }, + { + "epoch": 0.45, + "grad_norm": 1.1081360210131117, + "learning_rate": 1.1965681493197017e-05, + "loss": 0.9769, + "step": 19267 + }, + { + "epoch": 0.45, + "grad_norm": 2.4096148689897094, + "learning_rate": 1.1964933333220573e-05, + "loss": 0.9548, + "step": 19268 + }, + { + "epoch": 0.45, + "grad_norm": 1.8775997271211935, + "learning_rate": 1.1964185161803666e-05, + "loss": 1.0107, + "step": 19269 + }, + { + "epoch": 0.45, + "grad_norm": 2.5051170516592163, + "learning_rate": 1.196343697895065e-05, + "loss": 1.0532, + "step": 19270 + }, + { + "epoch": 0.45, + "grad_norm": 1.98126177151572, + "learning_rate": 1.1962688784665883e-05, + "loss": 1.0378, + "step": 19271 + }, + { + "epoch": 0.45, + "grad_norm": 3.5403023986211837, + "learning_rate": 1.1961940578953719e-05, + "loss": 0.9815, + "step": 19272 + }, + { + "epoch": 0.45, + "grad_norm": 2.3377558116397217, + "learning_rate": 1.1961192361818517e-05, + "loss": 1.0495, + "step": 19273 + }, + { + "epoch": 0.45, + "grad_norm": 1.9885556976837637, + "learning_rate": 1.1960444133264632e-05, + "loss": 1.1179, + "step": 19274 + }, + { + "epoch": 0.45, + "grad_norm": 1.9130321617318558, + "learning_rate": 1.1959695893296421e-05, + "loss": 1.0088, + "step": 19275 + }, + { + "epoch": 0.45, + "grad_norm": 2.168925809321684, + "learning_rate": 1.1958947641918239e-05, + "loss": 0.9714, + "step": 19276 + }, + { + "epoch": 0.45, + "grad_norm": 2.2911651922065017, + "learning_rate": 1.1958199379134445e-05, + "loss": 0.985, + "step": 19277 + }, + { + "epoch": 0.45, + "grad_norm": 1.09991975493809, + "learning_rate": 1.1957451104949396e-05, + "loss": 0.9792, + "step": 19278 + }, + { + "epoch": 0.45, + "grad_norm": 1.1478035191204532, + "learning_rate": 1.1956702819367442e-05, + "loss": 0.9038, + "step": 19279 + }, + { + "epoch": 0.45, + "grad_norm": 1.131418228829474, + "learning_rate": 1.195595452239295e-05, + "loss": 1.0166, + "step": 19280 + }, + { + "epoch": 0.45, + "grad_norm": 1.9621634000913708, + "learning_rate": 1.1955206214030266e-05, + "loss": 1.0014, + "step": 19281 + }, + { + "epoch": 0.45, + "grad_norm": 1.9003033396199067, + "learning_rate": 1.1954457894283758e-05, + "loss": 0.9614, + "step": 19282 + }, + { + "epoch": 0.45, + "grad_norm": 2.131106252424895, + "learning_rate": 1.1953709563157771e-05, + "loss": 0.9189, + "step": 19283 + }, + { + "epoch": 0.45, + "grad_norm": 2.4798244318606386, + "learning_rate": 1.1952961220656676e-05, + "loss": 1.0144, + "step": 19284 + }, + { + "epoch": 0.45, + "grad_norm": 2.409085939336062, + "learning_rate": 1.1952212866784819e-05, + "loss": 1.1016, + "step": 19285 + }, + { + "epoch": 0.45, + "grad_norm": 1.8843877475320792, + "learning_rate": 1.1951464501546559e-05, + "loss": 0.9486, + "step": 19286 + }, + { + "epoch": 0.45, + "grad_norm": 2.1154634137283197, + "learning_rate": 1.1950716124946256e-05, + "loss": 1.0304, + "step": 19287 + }, + { + "epoch": 0.45, + "grad_norm": 1.8358305056572035, + "learning_rate": 1.1949967736988265e-05, + "loss": 1.0986, + "step": 19288 + }, + { + "epoch": 0.45, + "grad_norm": 2.3246822728435315, + "learning_rate": 1.1949219337676948e-05, + "loss": 1.0827, + "step": 19289 + }, + { + "epoch": 0.45, + "grad_norm": 1.933703062430689, + "learning_rate": 1.1948470927016656e-05, + "loss": 1.1118, + "step": 19290 + }, + { + "epoch": 0.45, + "grad_norm": 2.4000970159994575, + "learning_rate": 1.1947722505011749e-05, + "loss": 1.0499, + "step": 19291 + }, + { + "epoch": 0.45, + "grad_norm": 2.039163154231549, + "learning_rate": 1.1946974071666587e-05, + "loss": 0.9513, + "step": 19292 + }, + { + "epoch": 0.45, + "grad_norm": 2.0661544202417863, + "learning_rate": 1.1946225626985523e-05, + "loss": 1.0895, + "step": 19293 + }, + { + "epoch": 0.45, + "grad_norm": 1.8959300904423342, + "learning_rate": 1.194547717097292e-05, + "loss": 0.993, + "step": 19294 + }, + { + "epoch": 0.45, + "grad_norm": 2.0162535320916635, + "learning_rate": 1.194472870363313e-05, + "loss": 1.0138, + "step": 19295 + }, + { + "epoch": 0.45, + "grad_norm": 1.9272895193293653, + "learning_rate": 1.1943980224970516e-05, + "loss": 1.0018, + "step": 19296 + }, + { + "epoch": 0.45, + "grad_norm": 2.0398520862872145, + "learning_rate": 1.1943231734989434e-05, + "loss": 0.9399, + "step": 19297 + }, + { + "epoch": 0.45, + "grad_norm": 2.05025551204848, + "learning_rate": 1.194248323369424e-05, + "loss": 0.992, + "step": 19298 + }, + { + "epoch": 0.45, + "grad_norm": 2.099254429778042, + "learning_rate": 1.1941734721089293e-05, + "loss": 0.9406, + "step": 19299 + }, + { + "epoch": 0.45, + "grad_norm": 1.700292525615375, + "learning_rate": 1.1940986197178955e-05, + "loss": 1.113, + "step": 19300 + }, + { + "epoch": 0.45, + "grad_norm": 1.145896253779606, + "learning_rate": 1.1940237661967579e-05, + "loss": 0.9669, + "step": 19301 + }, + { + "epoch": 0.45, + "grad_norm": 1.1909131960677404, + "learning_rate": 1.1939489115459523e-05, + "loss": 0.9767, + "step": 19302 + }, + { + "epoch": 0.45, + "grad_norm": 1.9806981455285702, + "learning_rate": 1.1938740557659153e-05, + "loss": 1.0983, + "step": 19303 + }, + { + "epoch": 0.45, + "grad_norm": 2.2596876206984544, + "learning_rate": 1.1937991988570815e-05, + "loss": 0.9986, + "step": 19304 + }, + { + "epoch": 0.45, + "grad_norm": 2.4860603096696425, + "learning_rate": 1.193724340819888e-05, + "loss": 1.0011, + "step": 19305 + }, + { + "epoch": 0.45, + "grad_norm": 2.048888325261361, + "learning_rate": 1.1936494816547695e-05, + "loss": 0.9701, + "step": 19306 + }, + { + "epoch": 0.45, + "grad_norm": 2.1299339760097027, + "learning_rate": 1.1935746213621629e-05, + "loss": 1.2358, + "step": 19307 + }, + { + "epoch": 0.45, + "grad_norm": 1.9443586289845842, + "learning_rate": 1.1934997599425034e-05, + "loss": 1.1361, + "step": 19308 + }, + { + "epoch": 0.45, + "grad_norm": 2.2400335249498515, + "learning_rate": 1.1934248973962266e-05, + "loss": 0.9759, + "step": 19309 + }, + { + "epoch": 0.45, + "grad_norm": 1.8153276794445459, + "learning_rate": 1.1933500337237693e-05, + "loss": 1.0279, + "step": 19310 + }, + { + "epoch": 0.45, + "grad_norm": 2.0083266852934396, + "learning_rate": 1.1932751689255666e-05, + "loss": 1.0342, + "step": 19311 + }, + { + "epoch": 0.45, + "grad_norm": 2.4625234445982134, + "learning_rate": 1.1932003030020552e-05, + "loss": 1.0392, + "step": 19312 + }, + { + "epoch": 0.45, + "grad_norm": 2.6509645672693085, + "learning_rate": 1.1931254359536698e-05, + "loss": 1.015, + "step": 19313 + }, + { + "epoch": 0.46, + "grad_norm": 2.1660818472741044, + "learning_rate": 1.1930505677808476e-05, + "loss": 1.1124, + "step": 19314 + }, + { + "epoch": 0.46, + "grad_norm": 2.5717873177480732, + "learning_rate": 1.1929756984840236e-05, + "loss": 1.0155, + "step": 19315 + }, + { + "epoch": 0.46, + "grad_norm": 2.3664998800274706, + "learning_rate": 1.1929008280636338e-05, + "loss": 1.0005, + "step": 19316 + }, + { + "epoch": 0.46, + "grad_norm": 1.895150177301977, + "learning_rate": 1.1928259565201145e-05, + "loss": 1.0085, + "step": 19317 + }, + { + "epoch": 0.46, + "grad_norm": 1.9375491551827977, + "learning_rate": 1.1927510838539014e-05, + "loss": 0.9945, + "step": 19318 + }, + { + "epoch": 0.46, + "grad_norm": 2.35326659910176, + "learning_rate": 1.1926762100654306e-05, + "loss": 0.9893, + "step": 19319 + }, + { + "epoch": 0.46, + "grad_norm": 1.8536598030909806, + "learning_rate": 1.1926013351551375e-05, + "loss": 0.9911, + "step": 19320 + }, + { + "epoch": 0.46, + "grad_norm": 1.176723360016335, + "learning_rate": 1.1925264591234588e-05, + "loss": 1.0327, + "step": 19321 + }, + { + "epoch": 0.46, + "grad_norm": 2.6395365280797662, + "learning_rate": 1.19245158197083e-05, + "loss": 1.0098, + "step": 19322 + }, + { + "epoch": 0.46, + "grad_norm": 2.1526627282589734, + "learning_rate": 1.1923767036976872e-05, + "loss": 1.1064, + "step": 19323 + }, + { + "epoch": 0.46, + "grad_norm": 2.146061382559089, + "learning_rate": 1.1923018243044664e-05, + "loss": 1.0927, + "step": 19324 + }, + { + "epoch": 0.46, + "grad_norm": 1.9901251903342103, + "learning_rate": 1.1922269437916035e-05, + "loss": 0.9294, + "step": 19325 + }, + { + "epoch": 0.46, + "grad_norm": 1.9685992038164015, + "learning_rate": 1.1921520621595342e-05, + "loss": 1.2, + "step": 19326 + }, + { + "epoch": 0.46, + "grad_norm": 1.7811176397661548, + "learning_rate": 1.1920771794086949e-05, + "loss": 1.03, + "step": 19327 + }, + { + "epoch": 0.46, + "grad_norm": 2.088183417389596, + "learning_rate": 1.1920022955395217e-05, + "loss": 1.2524, + "step": 19328 + }, + { + "epoch": 0.46, + "grad_norm": 1.9618298731826889, + "learning_rate": 1.1919274105524501e-05, + "loss": 1.0104, + "step": 19329 + }, + { + "epoch": 0.46, + "grad_norm": 2.8181094520937164, + "learning_rate": 1.1918525244479161e-05, + "loss": 0.9705, + "step": 19330 + }, + { + "epoch": 0.46, + "grad_norm": 2.3359089972437994, + "learning_rate": 1.1917776372263568e-05, + "loss": 1.0378, + "step": 19331 + }, + { + "epoch": 0.46, + "grad_norm": 1.7737530086832392, + "learning_rate": 1.1917027488882066e-05, + "loss": 0.9736, + "step": 19332 + }, + { + "epoch": 0.46, + "grad_norm": 1.998273418770819, + "learning_rate": 1.1916278594339026e-05, + "loss": 0.8444, + "step": 19333 + }, + { + "epoch": 0.46, + "grad_norm": 1.8539662666885397, + "learning_rate": 1.1915529688638803e-05, + "loss": 1.178, + "step": 19334 + }, + { + "epoch": 0.46, + "grad_norm": 2.1352720946805004, + "learning_rate": 1.1914780771785764e-05, + "loss": 0.904, + "step": 19335 + }, + { + "epoch": 0.46, + "grad_norm": 2.0532055542241556, + "learning_rate": 1.1914031843784262e-05, + "loss": 1.005, + "step": 19336 + }, + { + "epoch": 0.46, + "grad_norm": 2.1353854361969913, + "learning_rate": 1.1913282904638661e-05, + "loss": 1.0866, + "step": 19337 + }, + { + "epoch": 0.46, + "grad_norm": 2.5267556526445833, + "learning_rate": 1.1912533954353323e-05, + "loss": 1.0346, + "step": 19338 + }, + { + "epoch": 0.46, + "grad_norm": 1.9326513484000145, + "learning_rate": 1.1911784992932605e-05, + "loss": 1.0232, + "step": 19339 + }, + { + "epoch": 0.46, + "grad_norm": 1.9381279337604997, + "learning_rate": 1.1911036020380868e-05, + "loss": 1.1816, + "step": 19340 + }, + { + "epoch": 0.46, + "grad_norm": 2.0016228299862826, + "learning_rate": 1.1910287036702478e-05, + "loss": 1.0239, + "step": 19341 + }, + { + "epoch": 0.46, + "grad_norm": 1.9635543834951579, + "learning_rate": 1.1909538041901789e-05, + "loss": 1.0717, + "step": 19342 + }, + { + "epoch": 0.46, + "grad_norm": 1.1414031644824503, + "learning_rate": 1.1908789035983166e-05, + "loss": 0.9239, + "step": 19343 + }, + { + "epoch": 0.46, + "grad_norm": 2.0213569543365577, + "learning_rate": 1.1908040018950968e-05, + "loss": 1.0685, + "step": 19344 + }, + { + "epoch": 0.46, + "grad_norm": 1.9226393040026888, + "learning_rate": 1.190729099080956e-05, + "loss": 0.9948, + "step": 19345 + }, + { + "epoch": 0.46, + "grad_norm": 1.9686829185599959, + "learning_rate": 1.1906541951563295e-05, + "loss": 1.0882, + "step": 19346 + }, + { + "epoch": 0.46, + "grad_norm": 1.7495049692886648, + "learning_rate": 1.1905792901216544e-05, + "loss": 0.9011, + "step": 19347 + }, + { + "epoch": 0.46, + "grad_norm": 1.9614946607527888, + "learning_rate": 1.190504383977366e-05, + "loss": 0.9756, + "step": 19348 + }, + { + "epoch": 0.46, + "grad_norm": 2.5461458611620347, + "learning_rate": 1.1904294767239009e-05, + "loss": 0.9976, + "step": 19349 + }, + { + "epoch": 0.46, + "grad_norm": 1.727013151755427, + "learning_rate": 1.1903545683616953e-05, + "loss": 0.893, + "step": 19350 + }, + { + "epoch": 0.46, + "grad_norm": 2.018267014927938, + "learning_rate": 1.1902796588911847e-05, + "loss": 1.0909, + "step": 19351 + }, + { + "epoch": 0.46, + "grad_norm": 1.1615097860257624, + "learning_rate": 1.1902047483128063e-05, + "loss": 0.9603, + "step": 19352 + }, + { + "epoch": 0.46, + "grad_norm": 1.941854503972984, + "learning_rate": 1.190129836626995e-05, + "loss": 1.0969, + "step": 19353 + }, + { + "epoch": 0.46, + "grad_norm": 1.9495514650711583, + "learning_rate": 1.1900549238341881e-05, + "loss": 0.9875, + "step": 19354 + }, + { + "epoch": 0.46, + "grad_norm": 2.0035942827748126, + "learning_rate": 1.1899800099348209e-05, + "loss": 1.1372, + "step": 19355 + }, + { + "epoch": 0.46, + "grad_norm": 1.807278610811819, + "learning_rate": 1.1899050949293303e-05, + "loss": 1.0393, + "step": 19356 + }, + { + "epoch": 0.46, + "grad_norm": 2.6002398060728544, + "learning_rate": 1.1898301788181515e-05, + "loss": 1.072, + "step": 19357 + }, + { + "epoch": 0.46, + "grad_norm": 2.0344704830130786, + "learning_rate": 1.1897552616017216e-05, + "loss": 1.0257, + "step": 19358 + }, + { + "epoch": 0.46, + "grad_norm": 1.8660258080183871, + "learning_rate": 1.1896803432804768e-05, + "loss": 1.0354, + "step": 19359 + }, + { + "epoch": 0.46, + "grad_norm": 1.9448199494784553, + "learning_rate": 1.1896054238548527e-05, + "loss": 1.0403, + "step": 19360 + }, + { + "epoch": 0.46, + "grad_norm": 1.9318288993639747, + "learning_rate": 1.189530503325286e-05, + "loss": 1.1082, + "step": 19361 + }, + { + "epoch": 0.46, + "grad_norm": 1.9250691722431066, + "learning_rate": 1.1894555816922122e-05, + "loss": 0.9053, + "step": 19362 + }, + { + "epoch": 0.46, + "grad_norm": 2.1225774072344676, + "learning_rate": 1.1893806589560689e-05, + "loss": 0.9633, + "step": 19363 + }, + { + "epoch": 0.46, + "grad_norm": 1.9570560857059245, + "learning_rate": 1.1893057351172905e-05, + "loss": 1.0741, + "step": 19364 + }, + { + "epoch": 0.46, + "grad_norm": 2.4449662206268648, + "learning_rate": 1.1892308101763148e-05, + "loss": 1.1089, + "step": 19365 + }, + { + "epoch": 0.46, + "grad_norm": 1.815247688307229, + "learning_rate": 1.1891558841335773e-05, + "loss": 1.1283, + "step": 19366 + }, + { + "epoch": 0.46, + "grad_norm": 2.594898327754783, + "learning_rate": 1.189080956989514e-05, + "loss": 1.042, + "step": 19367 + }, + { + "epoch": 0.46, + "grad_norm": 2.031807391323591, + "learning_rate": 1.189006028744562e-05, + "loss": 1.076, + "step": 19368 + }, + { + "epoch": 0.46, + "grad_norm": 2.3159247486647425, + "learning_rate": 1.1889310993991567e-05, + "loss": 0.8874, + "step": 19369 + }, + { + "epoch": 0.46, + "grad_norm": 1.8737270861754343, + "learning_rate": 1.188856168953735e-05, + "loss": 1.0252, + "step": 19370 + }, + { + "epoch": 0.46, + "grad_norm": 2.120258663834753, + "learning_rate": 1.1887812374087328e-05, + "loss": 1.0002, + "step": 19371 + }, + { + "epoch": 0.46, + "grad_norm": 3.002995721432131, + "learning_rate": 1.1887063047645863e-05, + "loss": 0.9862, + "step": 19372 + }, + { + "epoch": 0.46, + "grad_norm": 1.0906974140328463, + "learning_rate": 1.1886313710217322e-05, + "loss": 0.9418, + "step": 19373 + }, + { + "epoch": 0.46, + "grad_norm": 2.083042572923431, + "learning_rate": 1.1885564361806063e-05, + "loss": 0.9593, + "step": 19374 + }, + { + "epoch": 0.46, + "grad_norm": 1.9807946061212678, + "learning_rate": 1.1884815002416452e-05, + "loss": 1.0443, + "step": 19375 + }, + { + "epoch": 0.46, + "grad_norm": 2.5527512931773875, + "learning_rate": 1.1884065632052852e-05, + "loss": 1.157, + "step": 19376 + }, + { + "epoch": 0.46, + "grad_norm": 1.9409525366769806, + "learning_rate": 1.1883316250719625e-05, + "loss": 1.0521, + "step": 19377 + }, + { + "epoch": 0.46, + "grad_norm": 1.83230826534537, + "learning_rate": 1.1882566858421137e-05, + "loss": 1.1884, + "step": 19378 + }, + { + "epoch": 0.46, + "grad_norm": 2.468433610720745, + "learning_rate": 1.1881817455161747e-05, + "loss": 1.0753, + "step": 19379 + }, + { + "epoch": 0.46, + "grad_norm": 1.9249250150231527, + "learning_rate": 1.188106804094582e-05, + "loss": 1.1628, + "step": 19380 + }, + { + "epoch": 0.46, + "grad_norm": 2.1890399222971832, + "learning_rate": 1.1880318615777718e-05, + "loss": 1.0415, + "step": 19381 + }, + { + "epoch": 0.46, + "grad_norm": 2.0801946270468976, + "learning_rate": 1.1879569179661808e-05, + "loss": 1.0302, + "step": 19382 + }, + { + "epoch": 0.46, + "grad_norm": 1.8081092591056684, + "learning_rate": 1.1878819732602448e-05, + "loss": 0.8504, + "step": 19383 + }, + { + "epoch": 0.46, + "grad_norm": 1.175545001671938, + "learning_rate": 1.1878070274604008e-05, + "loss": 0.9277, + "step": 19384 + }, + { + "epoch": 0.46, + "grad_norm": 2.2448341362108026, + "learning_rate": 1.1877320805670845e-05, + "loss": 0.9722, + "step": 19385 + }, + { + "epoch": 0.46, + "grad_norm": 2.0284118119742742, + "learning_rate": 1.1876571325807328e-05, + "loss": 1.0331, + "step": 19386 + }, + { + "epoch": 0.46, + "grad_norm": 1.8921220221705757, + "learning_rate": 1.1875821835017819e-05, + "loss": 0.9458, + "step": 19387 + }, + { + "epoch": 0.46, + "grad_norm": 2.8334532706064866, + "learning_rate": 1.1875072333306681e-05, + "loss": 1.025, + "step": 19388 + }, + { + "epoch": 0.46, + "grad_norm": 2.132615334987491, + "learning_rate": 1.1874322820678277e-05, + "loss": 0.9659, + "step": 19389 + }, + { + "epoch": 0.46, + "grad_norm": 1.0931906314975741, + "learning_rate": 1.1873573297136973e-05, + "loss": 0.9767, + "step": 19390 + }, + { + "epoch": 0.46, + "grad_norm": 2.1652546571171847, + "learning_rate": 1.1872823762687133e-05, + "loss": 1.0196, + "step": 19391 + }, + { + "epoch": 0.46, + "grad_norm": 2.1375310754355774, + "learning_rate": 1.1872074217333118e-05, + "loss": 1.0562, + "step": 19392 + }, + { + "epoch": 0.46, + "grad_norm": 1.9108261541928089, + "learning_rate": 1.1871324661079295e-05, + "loss": 1.0286, + "step": 19393 + }, + { + "epoch": 0.46, + "grad_norm": 2.5102414004896056, + "learning_rate": 1.1870575093930027e-05, + "loss": 0.9791, + "step": 19394 + }, + { + "epoch": 0.46, + "grad_norm": 2.2594311070103914, + "learning_rate": 1.1869825515889682e-05, + "loss": 1.0113, + "step": 19395 + }, + { + "epoch": 0.46, + "grad_norm": 2.347127018972662, + "learning_rate": 1.1869075926962616e-05, + "loss": 1.0576, + "step": 19396 + }, + { + "epoch": 0.46, + "grad_norm": 1.8124268405661008, + "learning_rate": 1.1868326327153202e-05, + "loss": 1.0308, + "step": 19397 + }, + { + "epoch": 0.46, + "grad_norm": 2.14432283326344, + "learning_rate": 1.1867576716465799e-05, + "loss": 0.9221, + "step": 19398 + }, + { + "epoch": 0.46, + "grad_norm": 2.234896409362399, + "learning_rate": 1.1866827094904773e-05, + "loss": 1.0417, + "step": 19399 + }, + { + "epoch": 0.46, + "grad_norm": 1.8002541639782592, + "learning_rate": 1.186607746247449e-05, + "loss": 1.0165, + "step": 19400 + }, + { + "epoch": 0.46, + "grad_norm": 1.9077059371140834, + "learning_rate": 1.1865327819179313e-05, + "loss": 0.9893, + "step": 19401 + }, + { + "epoch": 0.46, + "grad_norm": 2.0167214311807413, + "learning_rate": 1.1864578165023604e-05, + "loss": 1.0741, + "step": 19402 + }, + { + "epoch": 0.46, + "grad_norm": 2.22299249909654, + "learning_rate": 1.1863828500011736e-05, + "loss": 0.9918, + "step": 19403 + }, + { + "epoch": 0.46, + "grad_norm": 2.096996983961854, + "learning_rate": 1.1863078824148062e-05, + "loss": 0.9287, + "step": 19404 + }, + { + "epoch": 0.46, + "grad_norm": 2.1650031181004925, + "learning_rate": 1.1862329137436958e-05, + "loss": 0.998, + "step": 19405 + }, + { + "epoch": 0.46, + "grad_norm": 2.193160607697625, + "learning_rate": 1.1861579439882779e-05, + "loss": 1.1052, + "step": 19406 + }, + { + "epoch": 0.46, + "grad_norm": 1.946835316932682, + "learning_rate": 1.18608297314899e-05, + "loss": 1.0297, + "step": 19407 + }, + { + "epoch": 0.46, + "grad_norm": 2.090201724000053, + "learning_rate": 1.1860080012262683e-05, + "loss": 1.1358, + "step": 19408 + }, + { + "epoch": 0.46, + "grad_norm": 2.1345445836098023, + "learning_rate": 1.1859330282205483e-05, + "loss": 1.1459, + "step": 19409 + }, + { + "epoch": 0.46, + "grad_norm": 2.336032909937461, + "learning_rate": 1.1858580541322682e-05, + "loss": 1.1078, + "step": 19410 + }, + { + "epoch": 0.46, + "grad_norm": 1.127542172575837, + "learning_rate": 1.1857830789618632e-05, + "loss": 0.972, + "step": 19411 + }, + { + "epoch": 0.46, + "grad_norm": 2.0929615160527653, + "learning_rate": 1.1857081027097706e-05, + "loss": 0.9081, + "step": 19412 + }, + { + "epoch": 0.46, + "grad_norm": 2.466260024796919, + "learning_rate": 1.1856331253764261e-05, + "loss": 0.8956, + "step": 19413 + }, + { + "epoch": 0.46, + "grad_norm": 2.0479912296395706, + "learning_rate": 1.1855581469622673e-05, + "loss": 1.0177, + "step": 19414 + }, + { + "epoch": 0.46, + "grad_norm": 1.0720825512389525, + "learning_rate": 1.1854831674677298e-05, + "loss": 1.0183, + "step": 19415 + }, + { + "epoch": 0.46, + "grad_norm": 1.0352548657983112, + "learning_rate": 1.185408186893251e-05, + "loss": 1.0351, + "step": 19416 + }, + { + "epoch": 0.46, + "grad_norm": 2.0473935836689305, + "learning_rate": 1.185333205239267e-05, + "loss": 1.1768, + "step": 19417 + }, + { + "epoch": 0.46, + "grad_norm": 2.0464205244162414, + "learning_rate": 1.1852582225062141e-05, + "loss": 1.0215, + "step": 19418 + }, + { + "epoch": 0.46, + "grad_norm": 2.411651623492677, + "learning_rate": 1.1851832386945294e-05, + "loss": 0.9032, + "step": 19419 + }, + { + "epoch": 0.46, + "grad_norm": 1.0366317112397345, + "learning_rate": 1.1851082538046492e-05, + "loss": 0.9198, + "step": 19420 + }, + { + "epoch": 0.46, + "grad_norm": 1.9451310980910639, + "learning_rate": 1.18503326783701e-05, + "loss": 0.8383, + "step": 19421 + }, + { + "epoch": 0.46, + "grad_norm": 1.1228994379496946, + "learning_rate": 1.1849582807920488e-05, + "loss": 0.9399, + "step": 19422 + }, + { + "epoch": 0.46, + "grad_norm": 2.545662984287171, + "learning_rate": 1.1848832926702017e-05, + "loss": 1.0185, + "step": 19423 + }, + { + "epoch": 0.46, + "grad_norm": 2.387280969807004, + "learning_rate": 1.1848083034719055e-05, + "loss": 1.1646, + "step": 19424 + }, + { + "epoch": 0.46, + "grad_norm": 2.486065387890411, + "learning_rate": 1.1847333131975971e-05, + "loss": 0.9485, + "step": 19425 + }, + { + "epoch": 0.46, + "grad_norm": 2.0404975025854863, + "learning_rate": 1.1846583218477129e-05, + "loss": 0.8444, + "step": 19426 + }, + { + "epoch": 0.46, + "grad_norm": 2.0636276249286105, + "learning_rate": 1.1845833294226892e-05, + "loss": 1.0631, + "step": 19427 + }, + { + "epoch": 0.46, + "grad_norm": 1.1366731650512971, + "learning_rate": 1.184508335922963e-05, + "loss": 1.0215, + "step": 19428 + }, + { + "epoch": 0.46, + "grad_norm": 2.0033324186415435, + "learning_rate": 1.1844333413489709e-05, + "loss": 1.0978, + "step": 19429 + }, + { + "epoch": 0.46, + "grad_norm": 1.884280824695109, + "learning_rate": 1.1843583457011497e-05, + "loss": 1.0918, + "step": 19430 + }, + { + "epoch": 0.46, + "grad_norm": 2.924278129725866, + "learning_rate": 1.1842833489799356e-05, + "loss": 0.9729, + "step": 19431 + }, + { + "epoch": 0.46, + "grad_norm": 1.934350311974128, + "learning_rate": 1.1842083511857654e-05, + "loss": 1.0578, + "step": 19432 + }, + { + "epoch": 0.46, + "grad_norm": 4.837908646792933, + "learning_rate": 1.184133352319076e-05, + "loss": 0.9268, + "step": 19433 + }, + { + "epoch": 0.46, + "grad_norm": 2.274525950129628, + "learning_rate": 1.184058352380304e-05, + "loss": 1.0042, + "step": 19434 + }, + { + "epoch": 0.46, + "grad_norm": 3.0205693466832675, + "learning_rate": 1.183983351369886e-05, + "loss": 1.106, + "step": 19435 + }, + { + "epoch": 0.46, + "grad_norm": 2.474319655568364, + "learning_rate": 1.1839083492882583e-05, + "loss": 1.0558, + "step": 19436 + }, + { + "epoch": 0.46, + "grad_norm": 2.722004218201304, + "learning_rate": 1.1838333461358587e-05, + "loss": 0.9271, + "step": 19437 + }, + { + "epoch": 0.46, + "grad_norm": 2.010771857946401, + "learning_rate": 1.1837583419131227e-05, + "loss": 1.2241, + "step": 19438 + }, + { + "epoch": 0.46, + "grad_norm": 1.9025092664891157, + "learning_rate": 1.1836833366204874e-05, + "loss": 1.009, + "step": 19439 + }, + { + "epoch": 0.46, + "grad_norm": 2.2029181818319614, + "learning_rate": 1.1836083302583896e-05, + "loss": 0.9041, + "step": 19440 + }, + { + "epoch": 0.46, + "grad_norm": 2.0196865738465304, + "learning_rate": 1.183533322827266e-05, + "loss": 0.9215, + "step": 19441 + }, + { + "epoch": 0.46, + "grad_norm": 1.0965873370686423, + "learning_rate": 1.1834583143275534e-05, + "loss": 0.9347, + "step": 19442 + }, + { + "epoch": 0.46, + "grad_norm": 1.1589417779858238, + "learning_rate": 1.1833833047596884e-05, + "loss": 0.9675, + "step": 19443 + }, + { + "epoch": 0.46, + "grad_norm": 2.0171097229481494, + "learning_rate": 1.1833082941241076e-05, + "loss": 0.9656, + "step": 19444 + }, + { + "epoch": 0.46, + "grad_norm": 2.0822820090851657, + "learning_rate": 1.1832332824212478e-05, + "loss": 1.1283, + "step": 19445 + }, + { + "epoch": 0.46, + "grad_norm": 2.109366827525096, + "learning_rate": 1.183158269651546e-05, + "loss": 1.0046, + "step": 19446 + }, + { + "epoch": 0.46, + "grad_norm": 2.5208419852077104, + "learning_rate": 1.1830832558154388e-05, + "loss": 1.0032, + "step": 19447 + }, + { + "epoch": 0.46, + "grad_norm": 2.1571238242034267, + "learning_rate": 1.1830082409133627e-05, + "loss": 1.1253, + "step": 19448 + }, + { + "epoch": 0.46, + "grad_norm": 1.8472069305230443, + "learning_rate": 1.182933224945755e-05, + "loss": 0.974, + "step": 19449 + }, + { + "epoch": 0.46, + "grad_norm": 4.976240746018671, + "learning_rate": 1.1828582079130519e-05, + "loss": 1.0418, + "step": 19450 + }, + { + "epoch": 0.46, + "grad_norm": 2.1295801778078833, + "learning_rate": 1.1827831898156905e-05, + "loss": 0.9674, + "step": 19451 + }, + { + "epoch": 0.46, + "grad_norm": 1.9544753163309914, + "learning_rate": 1.1827081706541076e-05, + "loss": 1.1069, + "step": 19452 + }, + { + "epoch": 0.46, + "grad_norm": 1.8961951568577424, + "learning_rate": 1.1826331504287396e-05, + "loss": 1.0306, + "step": 19453 + }, + { + "epoch": 0.46, + "grad_norm": 2.271179901961103, + "learning_rate": 1.182558129140024e-05, + "loss": 1.0441, + "step": 19454 + }, + { + "epoch": 0.46, + "grad_norm": 2.6185735197720956, + "learning_rate": 1.1824831067883969e-05, + "loss": 1.0347, + "step": 19455 + }, + { + "epoch": 0.46, + "grad_norm": 2.052487589449975, + "learning_rate": 1.1824080833742952e-05, + "loss": 0.9152, + "step": 19456 + }, + { + "epoch": 0.46, + "grad_norm": 1.8379641567214948, + "learning_rate": 1.1823330588981562e-05, + "loss": 0.9648, + "step": 19457 + }, + { + "epoch": 0.46, + "grad_norm": 2.3015593097006537, + "learning_rate": 1.1822580333604163e-05, + "loss": 1.1039, + "step": 19458 + }, + { + "epoch": 0.46, + "grad_norm": 2.030502590579583, + "learning_rate": 1.1821830067615128e-05, + "loss": 0.9969, + "step": 19459 + }, + { + "epoch": 0.46, + "grad_norm": 2.4554234933596755, + "learning_rate": 1.1821079791018815e-05, + "loss": 0.8788, + "step": 19460 + }, + { + "epoch": 0.46, + "grad_norm": 1.9083468756782904, + "learning_rate": 1.1820329503819604e-05, + "loss": 0.9001, + "step": 19461 + }, + { + "epoch": 0.46, + "grad_norm": 1.8534733062240318, + "learning_rate": 1.1819579206021854e-05, + "loss": 0.9792, + "step": 19462 + }, + { + "epoch": 0.46, + "grad_norm": 2.0923309814126405, + "learning_rate": 1.1818828897629941e-05, + "loss": 1.0124, + "step": 19463 + }, + { + "epoch": 0.46, + "grad_norm": 2.5821209418903748, + "learning_rate": 1.181807857864823e-05, + "loss": 1.0413, + "step": 19464 + }, + { + "epoch": 0.46, + "grad_norm": 2.0176486924340904, + "learning_rate": 1.1817328249081088e-05, + "loss": 1.1127, + "step": 19465 + }, + { + "epoch": 0.46, + "grad_norm": 1.8350104188388805, + "learning_rate": 1.1816577908932887e-05, + "loss": 1.2022, + "step": 19466 + }, + { + "epoch": 0.46, + "grad_norm": 2.085032460618569, + "learning_rate": 1.1815827558207996e-05, + "loss": 1.0471, + "step": 19467 + }, + { + "epoch": 0.46, + "grad_norm": 1.949408937769392, + "learning_rate": 1.181507719691078e-05, + "loss": 1.1338, + "step": 19468 + }, + { + "epoch": 0.46, + "grad_norm": 1.949059796684651, + "learning_rate": 1.181432682504561e-05, + "loss": 1.0715, + "step": 19469 + }, + { + "epoch": 0.46, + "grad_norm": 2.51195715059093, + "learning_rate": 1.1813576442616854e-05, + "loss": 1.0081, + "step": 19470 + }, + { + "epoch": 0.46, + "grad_norm": 1.9100855261341654, + "learning_rate": 1.1812826049628883e-05, + "loss": 1.0458, + "step": 19471 + }, + { + "epoch": 0.46, + "grad_norm": 1.755559393607794, + "learning_rate": 1.1812075646086066e-05, + "loss": 1.0401, + "step": 19472 + }, + { + "epoch": 0.46, + "grad_norm": 1.934436639185396, + "learning_rate": 1.1811325231992768e-05, + "loss": 0.9608, + "step": 19473 + }, + { + "epoch": 0.46, + "grad_norm": 1.9214241143553674, + "learning_rate": 1.1810574807353362e-05, + "loss": 1.0426, + "step": 19474 + }, + { + "epoch": 0.46, + "grad_norm": 1.8742538325928835, + "learning_rate": 1.1809824372172217e-05, + "loss": 0.9941, + "step": 19475 + }, + { + "epoch": 0.46, + "grad_norm": 2.2156614685453655, + "learning_rate": 1.18090739264537e-05, + "loss": 1.1296, + "step": 19476 + }, + { + "epoch": 0.46, + "grad_norm": 1.921660264451404, + "learning_rate": 1.1808323470202185e-05, + "loss": 1.008, + "step": 19477 + }, + { + "epoch": 0.46, + "grad_norm": 2.417427510833557, + "learning_rate": 1.1807573003422035e-05, + "loss": 1.0112, + "step": 19478 + }, + { + "epoch": 0.46, + "grad_norm": 2.0692591930855615, + "learning_rate": 1.1806822526117626e-05, + "loss": 1.0073, + "step": 19479 + }, + { + "epoch": 0.46, + "grad_norm": 2.0579178342213016, + "learning_rate": 1.1806072038293322e-05, + "loss": 1.0226, + "step": 19480 + }, + { + "epoch": 0.46, + "grad_norm": 2.232959404758253, + "learning_rate": 1.1805321539953493e-05, + "loss": 1.036, + "step": 19481 + }, + { + "epoch": 0.46, + "grad_norm": 2.8346770107751813, + "learning_rate": 1.1804571031102516e-05, + "loss": 1.067, + "step": 19482 + }, + { + "epoch": 0.46, + "grad_norm": 2.1702104815621337, + "learning_rate": 1.1803820511744749e-05, + "loss": 0.9768, + "step": 19483 + }, + { + "epoch": 0.46, + "grad_norm": 1.9847806764378058, + "learning_rate": 1.1803069981884573e-05, + "loss": 0.8468, + "step": 19484 + }, + { + "epoch": 0.46, + "grad_norm": 2.643088302437039, + "learning_rate": 1.1802319441526348e-05, + "loss": 1.0478, + "step": 19485 + }, + { + "epoch": 0.46, + "grad_norm": 2.4883613551074673, + "learning_rate": 1.1801568890674453e-05, + "loss": 0.9423, + "step": 19486 + }, + { + "epoch": 0.46, + "grad_norm": 2.1982791138957607, + "learning_rate": 1.180081832933325e-05, + "loss": 1.077, + "step": 19487 + }, + { + "epoch": 0.46, + "grad_norm": 2.1100076422094616, + "learning_rate": 1.1800067757507115e-05, + "loss": 1.0738, + "step": 19488 + }, + { + "epoch": 0.46, + "grad_norm": 2.1279875041500222, + "learning_rate": 1.1799317175200416e-05, + "loss": 1.0396, + "step": 19489 + }, + { + "epoch": 0.46, + "grad_norm": 1.8765318162755609, + "learning_rate": 1.1798566582417521e-05, + "loss": 1.0563, + "step": 19490 + }, + { + "epoch": 0.46, + "grad_norm": 1.9640123334064044, + "learning_rate": 1.1797815979162801e-05, + "loss": 0.9609, + "step": 19491 + }, + { + "epoch": 0.46, + "grad_norm": 1.9382034169371807, + "learning_rate": 1.179706536544063e-05, + "loss": 1.0224, + "step": 19492 + }, + { + "epoch": 0.46, + "grad_norm": 2.035061141302236, + "learning_rate": 1.1796314741255372e-05, + "loss": 0.8973, + "step": 19493 + }, + { + "epoch": 0.46, + "grad_norm": 2.1076920508393173, + "learning_rate": 1.1795564106611401e-05, + "loss": 1.0024, + "step": 19494 + }, + { + "epoch": 0.46, + "grad_norm": 1.9774608331677672, + "learning_rate": 1.1794813461513092e-05, + "loss": 1.0588, + "step": 19495 + }, + { + "epoch": 0.46, + "grad_norm": 2.2017738842096306, + "learning_rate": 1.1794062805964808e-05, + "loss": 0.8958, + "step": 19496 + }, + { + "epoch": 0.46, + "grad_norm": 2.056426026327863, + "learning_rate": 1.1793312139970923e-05, + "loss": 1.0861, + "step": 19497 + }, + { + "epoch": 0.46, + "grad_norm": 2.3563750585975023, + "learning_rate": 1.1792561463535805e-05, + "loss": 1.1572, + "step": 19498 + }, + { + "epoch": 0.46, + "grad_norm": 1.8924648901336916, + "learning_rate": 1.1791810776663829e-05, + "loss": 0.8349, + "step": 19499 + }, + { + "epoch": 0.46, + "grad_norm": 1.969211065519056, + "learning_rate": 1.1791060079359362e-05, + "loss": 1.0221, + "step": 19500 + }, + { + "epoch": 0.46, + "grad_norm": 2.2601801134393127, + "learning_rate": 1.1790309371626775e-05, + "loss": 1.0419, + "step": 19501 + }, + { + "epoch": 0.46, + "grad_norm": 2.0474725928807693, + "learning_rate": 1.1789558653470443e-05, + "loss": 1.0008, + "step": 19502 + }, + { + "epoch": 0.46, + "grad_norm": 1.9290674037719813, + "learning_rate": 1.1788807924894734e-05, + "loss": 0.8903, + "step": 19503 + }, + { + "epoch": 0.46, + "grad_norm": 2.248967837765733, + "learning_rate": 1.1788057185904015e-05, + "loss": 1.0986, + "step": 19504 + }, + { + "epoch": 0.46, + "grad_norm": 2.078392040893227, + "learning_rate": 1.1787306436502664e-05, + "loss": 0.9116, + "step": 19505 + }, + { + "epoch": 0.46, + "grad_norm": 2.307347908020309, + "learning_rate": 1.1786555676695049e-05, + "loss": 0.934, + "step": 19506 + }, + { + "epoch": 0.46, + "grad_norm": 1.9488339780765287, + "learning_rate": 1.178580490648554e-05, + "loss": 1.0766, + "step": 19507 + }, + { + "epoch": 0.46, + "grad_norm": 1.1177895941309026, + "learning_rate": 1.178505412587851e-05, + "loss": 0.9683, + "step": 19508 + }, + { + "epoch": 0.46, + "grad_norm": 1.1636710020627388, + "learning_rate": 1.1784303334878331e-05, + "loss": 0.9418, + "step": 19509 + }, + { + "epoch": 0.46, + "grad_norm": 2.4843401457520398, + "learning_rate": 1.1783552533489372e-05, + "loss": 1.0101, + "step": 19510 + }, + { + "epoch": 0.46, + "grad_norm": 1.9456838922080404, + "learning_rate": 1.1782801721716004e-05, + "loss": 0.9451, + "step": 19511 + }, + { + "epoch": 0.46, + "grad_norm": 2.119632779820648, + "learning_rate": 1.1782050899562605e-05, + "loss": 0.9526, + "step": 19512 + }, + { + "epoch": 0.46, + "grad_norm": 1.083050671744982, + "learning_rate": 1.1781300067033534e-05, + "loss": 0.9873, + "step": 19513 + }, + { + "epoch": 0.46, + "grad_norm": 2.0398388928069204, + "learning_rate": 1.1780549224133177e-05, + "loss": 0.9777, + "step": 19514 + }, + { + "epoch": 0.46, + "grad_norm": 1.9889804438845797, + "learning_rate": 1.1779798370865893e-05, + "loss": 1.138, + "step": 19515 + }, + { + "epoch": 0.46, + "grad_norm": 2.0149093611284146, + "learning_rate": 1.1779047507236064e-05, + "loss": 1.0561, + "step": 19516 + }, + { + "epoch": 0.46, + "grad_norm": 1.7645765215628075, + "learning_rate": 1.1778296633248056e-05, + "loss": 0.9854, + "step": 19517 + }, + { + "epoch": 0.46, + "grad_norm": 2.1020859269007452, + "learning_rate": 1.177754574890624e-05, + "loss": 0.9646, + "step": 19518 + }, + { + "epoch": 0.46, + "grad_norm": 2.336361510290673, + "learning_rate": 1.1776794854214993e-05, + "loss": 1.0717, + "step": 19519 + }, + { + "epoch": 0.46, + "grad_norm": 2.2278510117335757, + "learning_rate": 1.1776043949178681e-05, + "loss": 1.1198, + "step": 19520 + }, + { + "epoch": 0.46, + "grad_norm": 2.8936255633338064, + "learning_rate": 1.177529303380168e-05, + "loss": 1.0556, + "step": 19521 + }, + { + "epoch": 0.46, + "grad_norm": 2.0867061708916306, + "learning_rate": 1.177454210808836e-05, + "loss": 1.0775, + "step": 19522 + }, + { + "epoch": 0.46, + "grad_norm": 2.1605434416705895, + "learning_rate": 1.1773791172043094e-05, + "loss": 1.1198, + "step": 19523 + }, + { + "epoch": 0.46, + "grad_norm": 2.016198899707003, + "learning_rate": 1.1773040225670257e-05, + "loss": 1.0756, + "step": 19524 + }, + { + "epoch": 0.46, + "grad_norm": 1.1487225287089333, + "learning_rate": 1.1772289268974215e-05, + "loss": 0.971, + "step": 19525 + }, + { + "epoch": 0.46, + "grad_norm": 2.4885406965220844, + "learning_rate": 1.1771538301959346e-05, + "loss": 1.1214, + "step": 19526 + }, + { + "epoch": 0.46, + "grad_norm": 1.9564640910111974, + "learning_rate": 1.1770787324630019e-05, + "loss": 0.9756, + "step": 19527 + }, + { + "epoch": 0.46, + "grad_norm": 2.127449505790332, + "learning_rate": 1.1770036336990609e-05, + "loss": 1.1205, + "step": 19528 + }, + { + "epoch": 0.46, + "grad_norm": 1.9048040869944831, + "learning_rate": 1.1769285339045484e-05, + "loss": 0.9157, + "step": 19529 + }, + { + "epoch": 0.46, + "grad_norm": 1.9156326555012373, + "learning_rate": 1.1768534330799021e-05, + "loss": 1.1038, + "step": 19530 + }, + { + "epoch": 0.46, + "grad_norm": 2.1224746845474507, + "learning_rate": 1.1767783312255595e-05, + "loss": 1.0155, + "step": 19531 + }, + { + "epoch": 0.46, + "grad_norm": 1.1344835939148534, + "learning_rate": 1.176703228341957e-05, + "loss": 0.9877, + "step": 19532 + }, + { + "epoch": 0.46, + "grad_norm": 2.5240700740087774, + "learning_rate": 1.1766281244295328e-05, + "loss": 1.0754, + "step": 19533 + }, + { + "epoch": 0.46, + "grad_norm": 2.178945926144825, + "learning_rate": 1.1765530194887231e-05, + "loss": 1.0564, + "step": 19534 + }, + { + "epoch": 0.46, + "grad_norm": 2.0439122585666776, + "learning_rate": 1.1764779135199664e-05, + "loss": 1.0598, + "step": 19535 + }, + { + "epoch": 0.46, + "grad_norm": 1.1202326002585115, + "learning_rate": 1.1764028065236991e-05, + "loss": 1.0052, + "step": 19536 + }, + { + "epoch": 0.46, + "grad_norm": 1.7514458123670762, + "learning_rate": 1.1763276985003592e-05, + "loss": 0.9577, + "step": 19537 + }, + { + "epoch": 0.46, + "grad_norm": 2.2649882484404467, + "learning_rate": 1.1762525894503831e-05, + "loss": 0.9902, + "step": 19538 + }, + { + "epoch": 0.46, + "grad_norm": 2.018476580777481, + "learning_rate": 1.1761774793742092e-05, + "loss": 1.0328, + "step": 19539 + }, + { + "epoch": 0.46, + "grad_norm": 1.9455859096736645, + "learning_rate": 1.176102368272274e-05, + "loss": 1.0562, + "step": 19540 + }, + { + "epoch": 0.46, + "grad_norm": 2.1470591187391355, + "learning_rate": 1.1760272561450146e-05, + "loss": 1.052, + "step": 19541 + }, + { + "epoch": 0.46, + "grad_norm": 2.085346153952893, + "learning_rate": 1.1759521429928693e-05, + "loss": 1.13, + "step": 19542 + }, + { + "epoch": 0.46, + "grad_norm": 2.1489977821296327, + "learning_rate": 1.1758770288162746e-05, + "loss": 0.9087, + "step": 19543 + }, + { + "epoch": 0.46, + "grad_norm": 2.0423939757485714, + "learning_rate": 1.1758019136156684e-05, + "loss": 1.0134, + "step": 19544 + }, + { + "epoch": 0.46, + "grad_norm": 2.0522999890284797, + "learning_rate": 1.1757267973914876e-05, + "loss": 0.9163, + "step": 19545 + }, + { + "epoch": 0.46, + "grad_norm": 2.051759391213944, + "learning_rate": 1.17565168014417e-05, + "loss": 0.9065, + "step": 19546 + }, + { + "epoch": 0.46, + "grad_norm": 2.1213424211493868, + "learning_rate": 1.1755765618741526e-05, + "loss": 1.0171, + "step": 19547 + }, + { + "epoch": 0.46, + "grad_norm": 2.879320249218734, + "learning_rate": 1.175501442581873e-05, + "loss": 0.9977, + "step": 19548 + }, + { + "epoch": 0.46, + "grad_norm": 2.646553702211787, + "learning_rate": 1.1754263222677681e-05, + "loss": 1.0203, + "step": 19549 + }, + { + "epoch": 0.46, + "grad_norm": 1.8097305504825083, + "learning_rate": 1.1753512009322758e-05, + "loss": 1.0231, + "step": 19550 + }, + { + "epoch": 0.46, + "grad_norm": 1.863013822504736, + "learning_rate": 1.1752760785758332e-05, + "loss": 0.9241, + "step": 19551 + }, + { + "epoch": 0.46, + "grad_norm": 2.150199173721047, + "learning_rate": 1.1752009551988781e-05, + "loss": 0.8788, + "step": 19552 + }, + { + "epoch": 0.46, + "grad_norm": 1.8641293840772641, + "learning_rate": 1.1751258308018473e-05, + "loss": 1.0422, + "step": 19553 + }, + { + "epoch": 0.46, + "grad_norm": 2.2278871114875827, + "learning_rate": 1.1750507053851784e-05, + "loss": 1.1321, + "step": 19554 + }, + { + "epoch": 0.46, + "grad_norm": 2.1178894868582256, + "learning_rate": 1.1749755789493092e-05, + "loss": 0.9869, + "step": 19555 + }, + { + "epoch": 0.46, + "grad_norm": 2.0162297579202777, + "learning_rate": 1.1749004514946765e-05, + "loss": 1.0593, + "step": 19556 + }, + { + "epoch": 0.46, + "grad_norm": 2.3946463961126088, + "learning_rate": 1.174825323021718e-05, + "loss": 1.0016, + "step": 19557 + }, + { + "epoch": 0.46, + "grad_norm": 2.201352498169475, + "learning_rate": 1.1747501935308712e-05, + "loss": 1.1495, + "step": 19558 + }, + { + "epoch": 0.46, + "grad_norm": 1.9819768392600732, + "learning_rate": 1.1746750630225734e-05, + "loss": 1.0083, + "step": 19559 + }, + { + "epoch": 0.46, + "grad_norm": 2.448107069385408, + "learning_rate": 1.1745999314972623e-05, + "loss": 0.9833, + "step": 19560 + }, + { + "epoch": 0.46, + "grad_norm": 2.113252063574354, + "learning_rate": 1.174524798955375e-05, + "loss": 1.103, + "step": 19561 + }, + { + "epoch": 0.46, + "grad_norm": 1.8414582325986713, + "learning_rate": 1.1744496653973488e-05, + "loss": 0.9428, + "step": 19562 + }, + { + "epoch": 0.46, + "grad_norm": 2.46315116434483, + "learning_rate": 1.1743745308236218e-05, + "loss": 1.0674, + "step": 19563 + }, + { + "epoch": 0.46, + "grad_norm": 1.117655058952405, + "learning_rate": 1.1742993952346307e-05, + "loss": 0.9716, + "step": 19564 + }, + { + "epoch": 0.46, + "grad_norm": 2.0372611945452963, + "learning_rate": 1.174224258630814e-05, + "loss": 1.0344, + "step": 19565 + }, + { + "epoch": 0.46, + "grad_norm": 2.2501612171308065, + "learning_rate": 1.1741491210126077e-05, + "loss": 1.0539, + "step": 19566 + }, + { + "epoch": 0.46, + "grad_norm": 1.9114506956744444, + "learning_rate": 1.1740739823804506e-05, + "loss": 0.85, + "step": 19567 + }, + { + "epoch": 0.46, + "grad_norm": 1.9197641858614307, + "learning_rate": 1.1739988427347795e-05, + "loss": 0.9933, + "step": 19568 + }, + { + "epoch": 0.46, + "grad_norm": 2.1272980537954504, + "learning_rate": 1.1739237020760319e-05, + "loss": 1.0576, + "step": 19569 + }, + { + "epoch": 0.46, + "grad_norm": 1.9280322795480715, + "learning_rate": 1.1738485604046457e-05, + "loss": 1.1349, + "step": 19570 + }, + { + "epoch": 0.46, + "grad_norm": 2.1006316389958517, + "learning_rate": 1.1737734177210581e-05, + "loss": 1.0267, + "step": 19571 + }, + { + "epoch": 0.46, + "grad_norm": 2.536685099542296, + "learning_rate": 1.1736982740257065e-05, + "loss": 0.9457, + "step": 19572 + }, + { + "epoch": 0.46, + "grad_norm": 2.406338381421393, + "learning_rate": 1.1736231293190286e-05, + "loss": 0.9621, + "step": 19573 + }, + { + "epoch": 0.46, + "grad_norm": 1.9544927283219875, + "learning_rate": 1.1735479836014621e-05, + "loss": 1.0544, + "step": 19574 + }, + { + "epoch": 0.46, + "grad_norm": 2.284864818092974, + "learning_rate": 1.1734728368734441e-05, + "loss": 1.0612, + "step": 19575 + }, + { + "epoch": 0.46, + "grad_norm": 1.1078709042279984, + "learning_rate": 1.1733976891354122e-05, + "loss": 0.9322, + "step": 19576 + }, + { + "epoch": 0.46, + "grad_norm": 2.188805395052591, + "learning_rate": 1.1733225403878042e-05, + "loss": 1.097, + "step": 19577 + }, + { + "epoch": 0.46, + "grad_norm": 1.960071751370441, + "learning_rate": 1.1732473906310576e-05, + "loss": 0.9436, + "step": 19578 + }, + { + "epoch": 0.46, + "grad_norm": 2.2084230998760543, + "learning_rate": 1.1731722398656096e-05, + "loss": 1.0256, + "step": 19579 + }, + { + "epoch": 0.46, + "grad_norm": 1.8406375327547833, + "learning_rate": 1.1730970880918981e-05, + "loss": 0.9524, + "step": 19580 + }, + { + "epoch": 0.46, + "grad_norm": 1.1340417534143477, + "learning_rate": 1.1730219353103604e-05, + "loss": 0.9521, + "step": 19581 + }, + { + "epoch": 0.46, + "grad_norm": 2.060921088358323, + "learning_rate": 1.1729467815214348e-05, + "loss": 1.0499, + "step": 19582 + }, + { + "epoch": 0.46, + "grad_norm": 1.8494151557504588, + "learning_rate": 1.1728716267255575e-05, + "loss": 0.9811, + "step": 19583 + }, + { + "epoch": 0.46, + "grad_norm": 1.9299124286667735, + "learning_rate": 1.1727964709231676e-05, + "loss": 1.0599, + "step": 19584 + }, + { + "epoch": 0.46, + "grad_norm": 2.1797694112655885, + "learning_rate": 1.1727213141147013e-05, + "loss": 0.9223, + "step": 19585 + }, + { + "epoch": 0.46, + "grad_norm": 2.811356832716271, + "learning_rate": 1.1726461563005973e-05, + "loss": 0.9382, + "step": 19586 + }, + { + "epoch": 0.46, + "grad_norm": 1.0590901949388916, + "learning_rate": 1.1725709974812922e-05, + "loss": 0.9177, + "step": 19587 + }, + { + "epoch": 0.46, + "grad_norm": 1.9230728939864128, + "learning_rate": 1.1724958376572246e-05, + "loss": 1.0972, + "step": 19588 + }, + { + "epoch": 0.46, + "grad_norm": 2.4680605149824246, + "learning_rate": 1.1724206768288316e-05, + "loss": 1.1009, + "step": 19589 + }, + { + "epoch": 0.46, + "grad_norm": 2.241237638663513, + "learning_rate": 1.1723455149965504e-05, + "loss": 0.9951, + "step": 19590 + }, + { + "epoch": 0.46, + "grad_norm": 1.8378747157483581, + "learning_rate": 1.1722703521608197e-05, + "loss": 1.0745, + "step": 19591 + }, + { + "epoch": 0.46, + "grad_norm": 1.8127378002531098, + "learning_rate": 1.1721951883220756e-05, + "loss": 1.0928, + "step": 19592 + }, + { + "epoch": 0.46, + "grad_norm": 1.9541190796338153, + "learning_rate": 1.1721200234807574e-05, + "loss": 0.9151, + "step": 19593 + }, + { + "epoch": 0.46, + "grad_norm": 1.9895720349573758, + "learning_rate": 1.1720448576373013e-05, + "loss": 1.0312, + "step": 19594 + }, + { + "epoch": 0.46, + "grad_norm": 2.161602065687874, + "learning_rate": 1.1719696907921463e-05, + "loss": 0.9447, + "step": 19595 + }, + { + "epoch": 0.46, + "grad_norm": 2.0331159304098603, + "learning_rate": 1.171894522945729e-05, + "loss": 0.9404, + "step": 19596 + }, + { + "epoch": 0.46, + "grad_norm": 2.276969213096756, + "learning_rate": 1.1718193540984873e-05, + "loss": 0.8424, + "step": 19597 + }, + { + "epoch": 0.46, + "grad_norm": 2.2152231119027967, + "learning_rate": 1.1717441842508588e-05, + "loss": 1.08, + "step": 19598 + }, + { + "epoch": 0.46, + "grad_norm": 2.1634333401526282, + "learning_rate": 1.1716690134032813e-05, + "loss": 1.1005, + "step": 19599 + }, + { + "epoch": 0.46, + "grad_norm": 2.0461539540249407, + "learning_rate": 1.1715938415561928e-05, + "loss": 0.9806, + "step": 19600 + }, + { + "epoch": 0.46, + "grad_norm": 1.9263920201131368, + "learning_rate": 1.1715186687100302e-05, + "loss": 1.0671, + "step": 19601 + }, + { + "epoch": 0.46, + "grad_norm": 1.8107340188974221, + "learning_rate": 1.1714434948652317e-05, + "loss": 0.9281, + "step": 19602 + }, + { + "epoch": 0.46, + "grad_norm": 2.4232026510394684, + "learning_rate": 1.171368320022235e-05, + "loss": 1.1543, + "step": 19603 + }, + { + "epoch": 0.46, + "grad_norm": 2.0574341880673606, + "learning_rate": 1.1712931441814776e-05, + "loss": 1.0504, + "step": 19604 + }, + { + "epoch": 0.46, + "grad_norm": 2.1338434188431425, + "learning_rate": 1.1712179673433974e-05, + "loss": 1.072, + "step": 19605 + }, + { + "epoch": 0.46, + "grad_norm": 2.2719102785757115, + "learning_rate": 1.171142789508432e-05, + "loss": 0.9809, + "step": 19606 + }, + { + "epoch": 0.46, + "grad_norm": 2.06190743127302, + "learning_rate": 1.1710676106770191e-05, + "loss": 1.0291, + "step": 19607 + }, + { + "epoch": 0.46, + "grad_norm": 2.1035071599058583, + "learning_rate": 1.1709924308495964e-05, + "loss": 1.1285, + "step": 19608 + }, + { + "epoch": 0.46, + "grad_norm": 3.516074864758115, + "learning_rate": 1.1709172500266015e-05, + "loss": 1.0673, + "step": 19609 + }, + { + "epoch": 0.46, + "grad_norm": 4.020371092003305, + "learning_rate": 1.1708420682084722e-05, + "loss": 1.0016, + "step": 19610 + }, + { + "epoch": 0.46, + "grad_norm": 1.8888798083798928, + "learning_rate": 1.1707668853956465e-05, + "loss": 0.8628, + "step": 19611 + }, + { + "epoch": 0.46, + "grad_norm": 2.062323217239178, + "learning_rate": 1.1706917015885621e-05, + "loss": 0.9577, + "step": 19612 + }, + { + "epoch": 0.46, + "grad_norm": 2.1438132547219597, + "learning_rate": 1.1706165167876562e-05, + "loss": 1.1871, + "step": 19613 + }, + { + "epoch": 0.46, + "grad_norm": 1.7545545250505754, + "learning_rate": 1.170541330993367e-05, + "loss": 1.1125, + "step": 19614 + }, + { + "epoch": 0.46, + "grad_norm": 2.08376026713793, + "learning_rate": 1.1704661442061323e-05, + "loss": 1.0864, + "step": 19615 + }, + { + "epoch": 0.46, + "grad_norm": 2.340120429149576, + "learning_rate": 1.1703909564263898e-05, + "loss": 0.8261, + "step": 19616 + }, + { + "epoch": 0.46, + "grad_norm": 1.9528650632978417, + "learning_rate": 1.1703157676545768e-05, + "loss": 0.9017, + "step": 19617 + }, + { + "epoch": 0.46, + "grad_norm": 1.9484964694170397, + "learning_rate": 1.170240577891132e-05, + "loss": 0.7979, + "step": 19618 + }, + { + "epoch": 0.46, + "grad_norm": 2.0036802374715883, + "learning_rate": 1.1701653871364924e-05, + "loss": 1.0504, + "step": 19619 + }, + { + "epoch": 0.46, + "grad_norm": 1.9169835925091736, + "learning_rate": 1.1700901953910961e-05, + "loss": 0.9222, + "step": 19620 + }, + { + "epoch": 0.46, + "grad_norm": 1.923801343992855, + "learning_rate": 1.1700150026553807e-05, + "loss": 1.0883, + "step": 19621 + }, + { + "epoch": 0.46, + "grad_norm": 2.0006460784472346, + "learning_rate": 1.1699398089297842e-05, + "loss": 1.0638, + "step": 19622 + }, + { + "epoch": 0.46, + "grad_norm": 1.8923506678806856, + "learning_rate": 1.1698646142147446e-05, + "loss": 0.9445, + "step": 19623 + }, + { + "epoch": 0.46, + "grad_norm": 2.7166894245975604, + "learning_rate": 1.1697894185106991e-05, + "loss": 1.1631, + "step": 19624 + }, + { + "epoch": 0.46, + "grad_norm": 1.157708104860924, + "learning_rate": 1.1697142218180859e-05, + "loss": 1.0492, + "step": 19625 + }, + { + "epoch": 0.46, + "grad_norm": 2.021176246296475, + "learning_rate": 1.1696390241373427e-05, + "loss": 0.9363, + "step": 19626 + }, + { + "epoch": 0.46, + "grad_norm": 2.1881493862612698, + "learning_rate": 1.1695638254689074e-05, + "loss": 1.0439, + "step": 19627 + }, + { + "epoch": 0.46, + "grad_norm": 2.2052811105993495, + "learning_rate": 1.1694886258132182e-05, + "loss": 0.9546, + "step": 19628 + }, + { + "epoch": 0.46, + "grad_norm": 2.282379344111889, + "learning_rate": 1.1694134251707121e-05, + "loss": 1.0348, + "step": 19629 + }, + { + "epoch": 0.46, + "grad_norm": 2.1970987547982417, + "learning_rate": 1.1693382235418276e-05, + "loss": 1.0646, + "step": 19630 + }, + { + "epoch": 0.46, + "grad_norm": 1.0554232602813394, + "learning_rate": 1.1692630209270024e-05, + "loss": 1.0238, + "step": 19631 + }, + { + "epoch": 0.46, + "grad_norm": 2.1403075489888033, + "learning_rate": 1.1691878173266743e-05, + "loss": 1.0247, + "step": 19632 + }, + { + "epoch": 0.46, + "grad_norm": 2.7055753251813224, + "learning_rate": 1.1691126127412814e-05, + "loss": 1.0016, + "step": 19633 + }, + { + "epoch": 0.46, + "grad_norm": 2.102515991029428, + "learning_rate": 1.1690374071712608e-05, + "loss": 1.0641, + "step": 19634 + }, + { + "epoch": 0.46, + "grad_norm": 2.1026950011826884, + "learning_rate": 1.168962200617051e-05, + "loss": 1.0861, + "step": 19635 + }, + { + "epoch": 0.46, + "grad_norm": 2.065275343297769, + "learning_rate": 1.1688869930790898e-05, + "loss": 1.1518, + "step": 19636 + }, + { + "epoch": 0.46, + "grad_norm": 1.9330298854512535, + "learning_rate": 1.1688117845578155e-05, + "loss": 0.9766, + "step": 19637 + }, + { + "epoch": 0.46, + "grad_norm": 1.8452627370614503, + "learning_rate": 1.168736575053665e-05, + "loss": 1.0405, + "step": 19638 + }, + { + "epoch": 0.46, + "grad_norm": 1.938078207660162, + "learning_rate": 1.168661364567077e-05, + "loss": 1.0456, + "step": 19639 + }, + { + "epoch": 0.46, + "grad_norm": 2.1775969811578295, + "learning_rate": 1.1685861530984893e-05, + "loss": 0.9803, + "step": 19640 + }, + { + "epoch": 0.46, + "grad_norm": 2.0648697914862137, + "learning_rate": 1.1685109406483392e-05, + "loss": 1.0082, + "step": 19641 + }, + { + "epoch": 0.46, + "grad_norm": 3.5053281062413237, + "learning_rate": 1.1684357272170655e-05, + "loss": 1.0086, + "step": 19642 + }, + { + "epoch": 0.46, + "grad_norm": 2.304223190226577, + "learning_rate": 1.1683605128051053e-05, + "loss": 0.9992, + "step": 19643 + }, + { + "epoch": 0.46, + "grad_norm": 1.884257153251878, + "learning_rate": 1.1682852974128974e-05, + "loss": 0.928, + "step": 19644 + }, + { + "epoch": 0.46, + "grad_norm": 1.8487741269721754, + "learning_rate": 1.1682100810408785e-05, + "loss": 1.0519, + "step": 19645 + }, + { + "epoch": 0.46, + "grad_norm": 2.5594192361691315, + "learning_rate": 1.1681348636894879e-05, + "loss": 0.9914, + "step": 19646 + }, + { + "epoch": 0.46, + "grad_norm": 1.9060488011098944, + "learning_rate": 1.1680596453591626e-05, + "loss": 1.026, + "step": 19647 + }, + { + "epoch": 0.46, + "grad_norm": 1.879778452862318, + "learning_rate": 1.1679844260503407e-05, + "loss": 1.0513, + "step": 19648 + }, + { + "epoch": 0.46, + "grad_norm": 2.514342256442338, + "learning_rate": 1.1679092057634605e-05, + "loss": 1.0377, + "step": 19649 + }, + { + "epoch": 0.46, + "grad_norm": 2.0075286289887044, + "learning_rate": 1.1678339844989597e-05, + "loss": 1.0791, + "step": 19650 + }, + { + "epoch": 0.46, + "grad_norm": 2.118455224725949, + "learning_rate": 1.1677587622572763e-05, + "loss": 0.9202, + "step": 19651 + }, + { + "epoch": 0.46, + "grad_norm": 1.9801351273996115, + "learning_rate": 1.1676835390388484e-05, + "loss": 1.02, + "step": 19652 + }, + { + "epoch": 0.46, + "grad_norm": 1.9258826897185857, + "learning_rate": 1.1676083148441135e-05, + "loss": 0.9127, + "step": 19653 + }, + { + "epoch": 0.46, + "grad_norm": 1.8846645947046694, + "learning_rate": 1.1675330896735103e-05, + "loss": 0.8506, + "step": 19654 + }, + { + "epoch": 0.46, + "grad_norm": 2.1258349433896, + "learning_rate": 1.1674578635274763e-05, + "loss": 1.1328, + "step": 19655 + }, + { + "epoch": 0.46, + "grad_norm": 2.2202097930666187, + "learning_rate": 1.1673826364064496e-05, + "loss": 1.0102, + "step": 19656 + }, + { + "epoch": 0.46, + "grad_norm": 2.2508720971565177, + "learning_rate": 1.167307408310868e-05, + "loss": 1.0896, + "step": 19657 + }, + { + "epoch": 0.46, + "grad_norm": 2.5195851436661725, + "learning_rate": 1.1672321792411698e-05, + "loss": 0.9456, + "step": 19658 + }, + { + "epoch": 0.46, + "grad_norm": 1.0741452073802296, + "learning_rate": 1.167156949197793e-05, + "loss": 0.9261, + "step": 19659 + }, + { + "epoch": 0.46, + "grad_norm": 1.1083522112697481, + "learning_rate": 1.1670817181811756e-05, + "loss": 0.8871, + "step": 19660 + }, + { + "epoch": 0.46, + "grad_norm": 1.973460037114271, + "learning_rate": 1.1670064861917554e-05, + "loss": 0.8904, + "step": 19661 + }, + { + "epoch": 0.46, + "grad_norm": 1.1463343771562071, + "learning_rate": 1.1669312532299705e-05, + "loss": 0.984, + "step": 19662 + }, + { + "epoch": 0.46, + "grad_norm": 1.9951574149489857, + "learning_rate": 1.1668560192962595e-05, + "loss": 1.0157, + "step": 19663 + }, + { + "epoch": 0.46, + "grad_norm": 2.053198189652007, + "learning_rate": 1.1667807843910592e-05, + "loss": 0.9552, + "step": 19664 + }, + { + "epoch": 0.46, + "grad_norm": 2.134168133737548, + "learning_rate": 1.166705548514809e-05, + "loss": 0.9814, + "step": 19665 + }, + { + "epoch": 0.46, + "grad_norm": 1.0952383299072563, + "learning_rate": 1.1666303116679456e-05, + "loss": 0.9969, + "step": 19666 + }, + { + "epoch": 0.46, + "grad_norm": 1.989329848076053, + "learning_rate": 1.1665550738509084e-05, + "loss": 1.0926, + "step": 19667 + }, + { + "epoch": 0.46, + "grad_norm": 2.5128782116261505, + "learning_rate": 1.1664798350641344e-05, + "loss": 0.9369, + "step": 19668 + }, + { + "epoch": 0.46, + "grad_norm": 2.378643437918859, + "learning_rate": 1.1664045953080625e-05, + "loss": 1.0343, + "step": 19669 + }, + { + "epoch": 0.46, + "grad_norm": 2.04324201457585, + "learning_rate": 1.1663293545831302e-05, + "loss": 1.0771, + "step": 19670 + }, + { + "epoch": 0.46, + "grad_norm": 2.399446691369612, + "learning_rate": 1.1662541128897758e-05, + "loss": 0.9142, + "step": 19671 + }, + { + "epoch": 0.46, + "grad_norm": 2.0824888449596015, + "learning_rate": 1.166178870228437e-05, + "loss": 1.1165, + "step": 19672 + }, + { + "epoch": 0.46, + "grad_norm": 2.0690533883685, + "learning_rate": 1.1661036265995528e-05, + "loss": 0.8464, + "step": 19673 + }, + { + "epoch": 0.46, + "grad_norm": 1.9638628007592254, + "learning_rate": 1.1660283820035601e-05, + "loss": 1.1136, + "step": 19674 + }, + { + "epoch": 0.46, + "grad_norm": 2.168833726299464, + "learning_rate": 1.1659531364408979e-05, + "loss": 1.0958, + "step": 19675 + }, + { + "epoch": 0.46, + "grad_norm": 2.0904597522074493, + "learning_rate": 1.1658778899120038e-05, + "loss": 0.9235, + "step": 19676 + }, + { + "epoch": 0.46, + "grad_norm": 1.9641240787210774, + "learning_rate": 1.1658026424173163e-05, + "loss": 0.9299, + "step": 19677 + }, + { + "epoch": 0.46, + "grad_norm": 1.7561183734182004, + "learning_rate": 1.1657273939572734e-05, + "loss": 1.0257, + "step": 19678 + }, + { + "epoch": 0.46, + "grad_norm": 1.8259049291878682, + "learning_rate": 1.1656521445323129e-05, + "loss": 0.9891, + "step": 19679 + }, + { + "epoch": 0.46, + "grad_norm": 1.8528904684626197, + "learning_rate": 1.1655768941428734e-05, + "loss": 0.9232, + "step": 19680 + }, + { + "epoch": 0.46, + "grad_norm": 1.8993232520246488, + "learning_rate": 1.1655016427893926e-05, + "loss": 1.0508, + "step": 19681 + }, + { + "epoch": 0.46, + "grad_norm": 2.3923211079365885, + "learning_rate": 1.1654263904723089e-05, + "loss": 0.884, + "step": 19682 + }, + { + "epoch": 0.46, + "grad_norm": 1.8546690383209619, + "learning_rate": 1.1653511371920603e-05, + "loss": 0.9348, + "step": 19683 + }, + { + "epoch": 0.46, + "grad_norm": 2.107998074263685, + "learning_rate": 1.1652758829490854e-05, + "loss": 1.0078, + "step": 19684 + }, + { + "epoch": 0.46, + "grad_norm": 2.4298518627772245, + "learning_rate": 1.1652006277438214e-05, + "loss": 1.0224, + "step": 19685 + }, + { + "epoch": 0.46, + "grad_norm": 4.345100043754641, + "learning_rate": 1.1651253715767077e-05, + "loss": 0.8546, + "step": 19686 + }, + { + "epoch": 0.46, + "grad_norm": 2.565602481422471, + "learning_rate": 1.1650501144481812e-05, + "loss": 1.0611, + "step": 19687 + }, + { + "epoch": 0.46, + "grad_norm": 1.9562784901081511, + "learning_rate": 1.164974856358681e-05, + "loss": 0.9665, + "step": 19688 + }, + { + "epoch": 0.46, + "grad_norm": 1.1077917816688956, + "learning_rate": 1.1648995973086448e-05, + "loss": 0.9095, + "step": 19689 + }, + { + "epoch": 0.46, + "grad_norm": 2.3571464011995533, + "learning_rate": 1.1648243372985111e-05, + "loss": 0.9154, + "step": 19690 + }, + { + "epoch": 0.46, + "grad_norm": 2.496661241010823, + "learning_rate": 1.1647490763287178e-05, + "loss": 0.9783, + "step": 19691 + }, + { + "epoch": 0.46, + "grad_norm": 2.079458457135425, + "learning_rate": 1.1646738143997031e-05, + "loss": 0.9327, + "step": 19692 + }, + { + "epoch": 0.46, + "grad_norm": 1.120538159553421, + "learning_rate": 1.1645985515119059e-05, + "loss": 0.9811, + "step": 19693 + }, + { + "epoch": 0.46, + "grad_norm": 1.944423142378687, + "learning_rate": 1.164523287665763e-05, + "loss": 1.0225, + "step": 19694 + }, + { + "epoch": 0.46, + "grad_norm": 2.1273287409940447, + "learning_rate": 1.164448022861714e-05, + "loss": 1.0345, + "step": 19695 + }, + { + "epoch": 0.46, + "grad_norm": 1.9984320436416967, + "learning_rate": 1.1643727571001964e-05, + "loss": 0.9059, + "step": 19696 + }, + { + "epoch": 0.46, + "grad_norm": 2.4267800516226306, + "learning_rate": 1.1642974903816488e-05, + "loss": 1.0299, + "step": 19697 + }, + { + "epoch": 0.46, + "grad_norm": 1.0560308828364249, + "learning_rate": 1.164222222706509e-05, + "loss": 0.9608, + "step": 19698 + }, + { + "epoch": 0.46, + "grad_norm": 2.319624537933313, + "learning_rate": 1.1641469540752153e-05, + "loss": 1.0128, + "step": 19699 + }, + { + "epoch": 0.46, + "grad_norm": 1.9946473144723031, + "learning_rate": 1.164071684488206e-05, + "loss": 1.1025, + "step": 19700 + }, + { + "epoch": 0.46, + "grad_norm": 2.3197779696010854, + "learning_rate": 1.1639964139459196e-05, + "loss": 0.9459, + "step": 19701 + }, + { + "epoch": 0.46, + "grad_norm": 1.09408818342782, + "learning_rate": 1.1639211424487941e-05, + "loss": 0.9861, + "step": 19702 + }, + { + "epoch": 0.46, + "grad_norm": 2.193046842661842, + "learning_rate": 1.1638458699972678e-05, + "loss": 1.0691, + "step": 19703 + }, + { + "epoch": 0.46, + "grad_norm": 1.9025388060525434, + "learning_rate": 1.1637705965917791e-05, + "loss": 1.0628, + "step": 19704 + }, + { + "epoch": 0.46, + "grad_norm": 1.1431911083543487, + "learning_rate": 1.163695322232766e-05, + "loss": 0.9044, + "step": 19705 + }, + { + "epoch": 0.46, + "grad_norm": 2.019612027435248, + "learning_rate": 1.1636200469206671e-05, + "loss": 0.9891, + "step": 19706 + }, + { + "epoch": 0.46, + "grad_norm": 1.977647014426484, + "learning_rate": 1.1635447706559205e-05, + "loss": 0.9406, + "step": 19707 + }, + { + "epoch": 0.46, + "grad_norm": 2.3922654819578804, + "learning_rate": 1.1634694934389641e-05, + "loss": 0.9847, + "step": 19708 + }, + { + "epoch": 0.46, + "grad_norm": 2.0573022745810623, + "learning_rate": 1.1633942152702367e-05, + "loss": 1.0589, + "step": 19709 + }, + { + "epoch": 0.46, + "grad_norm": 3.0009610226191885, + "learning_rate": 1.1633189361501767e-05, + "loss": 0.9264, + "step": 19710 + }, + { + "epoch": 0.46, + "grad_norm": 1.7878598372208059, + "learning_rate": 1.1632436560792217e-05, + "loss": 1.041, + "step": 19711 + }, + { + "epoch": 0.46, + "grad_norm": 1.9748872795807482, + "learning_rate": 1.1631683750578112e-05, + "loss": 0.9937, + "step": 19712 + }, + { + "epoch": 0.46, + "grad_norm": 1.6737204658122142, + "learning_rate": 1.163093093086382e-05, + "loss": 0.9346, + "step": 19713 + }, + { + "epoch": 0.46, + "grad_norm": 1.1069284566432784, + "learning_rate": 1.1630178101653736e-05, + "loss": 1.0114, + "step": 19714 + }, + { + "epoch": 0.46, + "grad_norm": 2.0574716458584503, + "learning_rate": 1.1629425262952235e-05, + "loss": 1.0072, + "step": 19715 + }, + { + "epoch": 0.46, + "grad_norm": 1.1562214373430748, + "learning_rate": 1.1628672414763709e-05, + "loss": 0.9595, + "step": 19716 + }, + { + "epoch": 0.46, + "grad_norm": 1.9534308910443057, + "learning_rate": 1.1627919557092532e-05, + "loss": 0.9997, + "step": 19717 + }, + { + "epoch": 0.46, + "grad_norm": 2.050505702489811, + "learning_rate": 1.1627166689943098e-05, + "loss": 0.9489, + "step": 19718 + }, + { + "epoch": 0.46, + "grad_norm": 2.0559839056464893, + "learning_rate": 1.1626413813319779e-05, + "loss": 1.0405, + "step": 19719 + }, + { + "epoch": 0.46, + "grad_norm": 1.94371273168525, + "learning_rate": 1.1625660927226966e-05, + "loss": 1.0542, + "step": 19720 + }, + { + "epoch": 0.46, + "grad_norm": 1.890413902121531, + "learning_rate": 1.162490803166904e-05, + "loss": 1.0122, + "step": 19721 + }, + { + "epoch": 0.46, + "grad_norm": 2.1763201865578217, + "learning_rate": 1.1624155126650385e-05, + "loss": 0.8964, + "step": 19722 + }, + { + "epoch": 0.46, + "grad_norm": 2.1482670985725005, + "learning_rate": 1.1623402212175384e-05, + "loss": 0.9045, + "step": 19723 + }, + { + "epoch": 0.46, + "grad_norm": 1.964281743162156, + "learning_rate": 1.1622649288248422e-05, + "loss": 0.9441, + "step": 19724 + }, + { + "epoch": 0.46, + "grad_norm": 1.6751752510172704, + "learning_rate": 1.1621896354873883e-05, + "loss": 0.9725, + "step": 19725 + }, + { + "epoch": 0.46, + "grad_norm": 2.314022086643497, + "learning_rate": 1.1621143412056146e-05, + "loss": 0.9376, + "step": 19726 + }, + { + "epoch": 0.46, + "grad_norm": 2.6396516715406393, + "learning_rate": 1.1620390459799605e-05, + "loss": 0.9128, + "step": 19727 + }, + { + "epoch": 0.46, + "grad_norm": 1.9877018154480592, + "learning_rate": 1.1619637498108634e-05, + "loss": 0.9608, + "step": 19728 + }, + { + "epoch": 0.46, + "grad_norm": 2.5803110636947015, + "learning_rate": 1.1618884526987619e-05, + "loss": 0.9999, + "step": 19729 + }, + { + "epoch": 0.46, + "grad_norm": 1.6958187268727543, + "learning_rate": 1.1618131546440949e-05, + "loss": 1.0019, + "step": 19730 + }, + { + "epoch": 0.46, + "grad_norm": 1.0162154624756838, + "learning_rate": 1.1617378556473004e-05, + "loss": 0.9284, + "step": 19731 + }, + { + "epoch": 0.46, + "grad_norm": 2.1256862642817604, + "learning_rate": 1.1616625557088169e-05, + "loss": 1.0443, + "step": 19732 + }, + { + "epoch": 0.46, + "grad_norm": 2.2995018441758606, + "learning_rate": 1.1615872548290827e-05, + "loss": 0.9732, + "step": 19733 + }, + { + "epoch": 0.46, + "grad_norm": 2.1596087739366996, + "learning_rate": 1.1615119530085365e-05, + "loss": 0.9828, + "step": 19734 + }, + { + "epoch": 0.46, + "grad_norm": 2.0576971702867644, + "learning_rate": 1.1614366502476169e-05, + "loss": 1.0075, + "step": 19735 + }, + { + "epoch": 0.46, + "grad_norm": 2.091776002362152, + "learning_rate": 1.1613613465467612e-05, + "loss": 1.0644, + "step": 19736 + }, + { + "epoch": 0.46, + "grad_norm": 1.8718618887333067, + "learning_rate": 1.1612860419064092e-05, + "loss": 0.993, + "step": 19737 + }, + { + "epoch": 0.47, + "grad_norm": 1.8587606911264194, + "learning_rate": 1.1612107363269987e-05, + "loss": 0.9825, + "step": 19738 + }, + { + "epoch": 0.47, + "grad_norm": 1.9223050662061054, + "learning_rate": 1.1611354298089683e-05, + "loss": 0.9739, + "step": 19739 + }, + { + "epoch": 0.47, + "grad_norm": 1.8981519103354516, + "learning_rate": 1.1610601223527563e-05, + "loss": 1.0505, + "step": 19740 + }, + { + "epoch": 0.47, + "grad_norm": 1.1507654168499077, + "learning_rate": 1.1609848139588014e-05, + "loss": 0.9194, + "step": 19741 + }, + { + "epoch": 0.47, + "grad_norm": 2.121949895302803, + "learning_rate": 1.160909504627542e-05, + "loss": 1.1419, + "step": 19742 + }, + { + "epoch": 0.47, + "grad_norm": 2.335867739994526, + "learning_rate": 1.1608341943594162e-05, + "loss": 0.9848, + "step": 19743 + }, + { + "epoch": 0.47, + "grad_norm": 1.8638728390235928, + "learning_rate": 1.160758883154863e-05, + "loss": 1.046, + "step": 19744 + }, + { + "epoch": 0.47, + "grad_norm": 2.249000717119547, + "learning_rate": 1.1606835710143207e-05, + "loss": 0.9512, + "step": 19745 + }, + { + "epoch": 0.47, + "grad_norm": 2.6939178348407435, + "learning_rate": 1.160608257938228e-05, + "loss": 1.2113, + "step": 19746 + }, + { + "epoch": 0.47, + "grad_norm": 2.6471103486560392, + "learning_rate": 1.1605329439270227e-05, + "loss": 0.9887, + "step": 19747 + }, + { + "epoch": 0.47, + "grad_norm": 2.3098661991294973, + "learning_rate": 1.1604576289811442e-05, + "loss": 1.0693, + "step": 19748 + }, + { + "epoch": 0.47, + "grad_norm": 2.103460487601193, + "learning_rate": 1.1603823131010303e-05, + "loss": 0.8348, + "step": 19749 + }, + { + "epoch": 0.47, + "grad_norm": 2.406643389945755, + "learning_rate": 1.16030699628712e-05, + "loss": 0.9826, + "step": 19750 + }, + { + "epoch": 0.47, + "grad_norm": 2.2346583816915473, + "learning_rate": 1.1602316785398515e-05, + "loss": 0.9923, + "step": 19751 + }, + { + "epoch": 0.47, + "grad_norm": 2.7675965432366527, + "learning_rate": 1.1601563598596635e-05, + "loss": 0.9516, + "step": 19752 + }, + { + "epoch": 0.47, + "grad_norm": 2.526505725355549, + "learning_rate": 1.1600810402469944e-05, + "loss": 1.0394, + "step": 19753 + }, + { + "epoch": 0.47, + "grad_norm": 1.8660609013067746, + "learning_rate": 1.1600057197022826e-05, + "loss": 1.0431, + "step": 19754 + }, + { + "epoch": 0.47, + "grad_norm": 2.172207355101454, + "learning_rate": 1.1599303982259673e-05, + "loss": 0.99, + "step": 19755 + }, + { + "epoch": 0.47, + "grad_norm": 2.0085241020610143, + "learning_rate": 1.1598550758184864e-05, + "loss": 1.0586, + "step": 19756 + }, + { + "epoch": 0.47, + "grad_norm": 2.053076331541218, + "learning_rate": 1.1597797524802786e-05, + "loss": 1.0498, + "step": 19757 + }, + { + "epoch": 0.47, + "grad_norm": 2.087861256653823, + "learning_rate": 1.1597044282117824e-05, + "loss": 1.0577, + "step": 19758 + }, + { + "epoch": 0.47, + "grad_norm": 2.266045492301245, + "learning_rate": 1.1596291030134366e-05, + "loss": 1.1015, + "step": 19759 + }, + { + "epoch": 0.47, + "grad_norm": 1.8284075261545039, + "learning_rate": 1.1595537768856794e-05, + "loss": 1.0876, + "step": 19760 + }, + { + "epoch": 0.47, + "grad_norm": 1.7981123499413725, + "learning_rate": 1.15947844982895e-05, + "loss": 1.0623, + "step": 19761 + }, + { + "epoch": 0.47, + "grad_norm": 1.8762532895273276, + "learning_rate": 1.1594031218436864e-05, + "loss": 1.1119, + "step": 19762 + }, + { + "epoch": 0.47, + "grad_norm": 1.0031523393542232, + "learning_rate": 1.1593277929303275e-05, + "loss": 0.8875, + "step": 19763 + }, + { + "epoch": 0.47, + "grad_norm": 2.3203625230602882, + "learning_rate": 1.1592524630893111e-05, + "loss": 1.0143, + "step": 19764 + }, + { + "epoch": 0.47, + "grad_norm": 1.8626913497925053, + "learning_rate": 1.1591771323210771e-05, + "loss": 0.9536, + "step": 19765 + }, + { + "epoch": 0.47, + "grad_norm": 2.1569821846167927, + "learning_rate": 1.1591018006260631e-05, + "loss": 0.9037, + "step": 19766 + }, + { + "epoch": 0.47, + "grad_norm": 2.0794295264629628, + "learning_rate": 1.1590264680047084e-05, + "loss": 1.105, + "step": 19767 + }, + { + "epoch": 0.47, + "grad_norm": 2.132188203901124, + "learning_rate": 1.1589511344574509e-05, + "loss": 0.9976, + "step": 19768 + }, + { + "epoch": 0.47, + "grad_norm": 2.13866535288108, + "learning_rate": 1.1588757999847297e-05, + "loss": 1.0295, + "step": 19769 + }, + { + "epoch": 0.47, + "grad_norm": 1.9126194534393903, + "learning_rate": 1.1588004645869835e-05, + "loss": 0.9447, + "step": 19770 + }, + { + "epoch": 0.47, + "grad_norm": 1.1554937870003241, + "learning_rate": 1.1587251282646505e-05, + "loss": 0.9754, + "step": 19771 + }, + { + "epoch": 0.47, + "grad_norm": 1.9552523460799596, + "learning_rate": 1.1586497910181695e-05, + "loss": 1.0933, + "step": 19772 + }, + { + "epoch": 0.47, + "grad_norm": 1.8957016701793923, + "learning_rate": 1.1585744528479794e-05, + "loss": 1.0757, + "step": 19773 + }, + { + "epoch": 0.47, + "grad_norm": 2.971527037161341, + "learning_rate": 1.1584991137545185e-05, + "loss": 1.0891, + "step": 19774 + }, + { + "epoch": 0.47, + "grad_norm": 2.0866124756241815, + "learning_rate": 1.1584237737382253e-05, + "loss": 0.9795, + "step": 19775 + }, + { + "epoch": 0.47, + "grad_norm": 1.9155115461901404, + "learning_rate": 1.158348432799539e-05, + "loss": 0.9395, + "step": 19776 + }, + { + "epoch": 0.47, + "grad_norm": 2.0631424463668244, + "learning_rate": 1.1582730909388982e-05, + "loss": 0.9513, + "step": 19777 + }, + { + "epoch": 0.47, + "grad_norm": 2.325581459405284, + "learning_rate": 1.1581977481567413e-05, + "loss": 0.9962, + "step": 19778 + }, + { + "epoch": 0.47, + "grad_norm": 2.1572788965241587, + "learning_rate": 1.1581224044535067e-05, + "loss": 1.126, + "step": 19779 + }, + { + "epoch": 0.47, + "grad_norm": 2.0486931726850424, + "learning_rate": 1.1580470598296337e-05, + "loss": 1.0295, + "step": 19780 + }, + { + "epoch": 0.47, + "grad_norm": 1.8556758766261978, + "learning_rate": 1.1579717142855605e-05, + "loss": 0.9906, + "step": 19781 + }, + { + "epoch": 0.47, + "grad_norm": 2.3504036559122023, + "learning_rate": 1.157896367821726e-05, + "loss": 1.0987, + "step": 19782 + }, + { + "epoch": 0.47, + "grad_norm": 1.8854724303642558, + "learning_rate": 1.1578210204385689e-05, + "loss": 0.976, + "step": 19783 + }, + { + "epoch": 0.47, + "grad_norm": 1.1390712827432528, + "learning_rate": 1.157745672136528e-05, + "loss": 0.9564, + "step": 19784 + }, + { + "epoch": 0.47, + "grad_norm": 1.106643593336784, + "learning_rate": 1.1576703229160416e-05, + "loss": 0.9295, + "step": 19785 + }, + { + "epoch": 0.47, + "grad_norm": 1.9212312830589624, + "learning_rate": 1.1575949727775487e-05, + "loss": 1.0024, + "step": 19786 + }, + { + "epoch": 0.47, + "grad_norm": 1.9685465704270553, + "learning_rate": 1.1575196217214879e-05, + "loss": 0.9226, + "step": 19787 + }, + { + "epoch": 0.47, + "grad_norm": 2.0228272560574942, + "learning_rate": 1.1574442697482982e-05, + "loss": 0.9168, + "step": 19788 + }, + { + "epoch": 0.47, + "grad_norm": 2.2571816149743094, + "learning_rate": 1.1573689168584182e-05, + "loss": 1.0319, + "step": 19789 + }, + { + "epoch": 0.47, + "grad_norm": 1.9869450367904034, + "learning_rate": 1.1572935630522864e-05, + "loss": 1.0538, + "step": 19790 + }, + { + "epoch": 0.47, + "grad_norm": 2.4727364552704154, + "learning_rate": 1.1572182083303417e-05, + "loss": 0.9684, + "step": 19791 + }, + { + "epoch": 0.47, + "grad_norm": 2.0350877406867967, + "learning_rate": 1.1571428526930227e-05, + "loss": 1.1531, + "step": 19792 + }, + { + "epoch": 0.47, + "grad_norm": 2.717500254676598, + "learning_rate": 1.1570674961407684e-05, + "loss": 0.9602, + "step": 19793 + }, + { + "epoch": 0.47, + "grad_norm": 1.9688153527473808, + "learning_rate": 1.156992138674017e-05, + "loss": 1.0117, + "step": 19794 + }, + { + "epoch": 0.47, + "grad_norm": 1.8780952318408204, + "learning_rate": 1.1569167802932084e-05, + "loss": 1.0148, + "step": 19795 + }, + { + "epoch": 0.47, + "grad_norm": 2.198767146042462, + "learning_rate": 1.15684142099878e-05, + "loss": 1.0635, + "step": 19796 + }, + { + "epoch": 0.47, + "grad_norm": 1.2125571801526207, + "learning_rate": 1.1567660607911718e-05, + "loss": 1.0117, + "step": 19797 + }, + { + "epoch": 0.47, + "grad_norm": 2.0453506836949664, + "learning_rate": 1.1566906996708212e-05, + "loss": 0.9486, + "step": 19798 + }, + { + "epoch": 0.47, + "grad_norm": 1.7237360021728105, + "learning_rate": 1.1566153376381683e-05, + "loss": 0.9963, + "step": 19799 + }, + { + "epoch": 0.47, + "grad_norm": 1.1008550537873403, + "learning_rate": 1.1565399746936513e-05, + "loss": 0.8815, + "step": 19800 + }, + { + "epoch": 0.47, + "grad_norm": 1.0636736198583294, + "learning_rate": 1.1564646108377088e-05, + "loss": 0.8891, + "step": 19801 + }, + { + "epoch": 0.47, + "grad_norm": 1.826311094923717, + "learning_rate": 1.1563892460707799e-05, + "loss": 0.8838, + "step": 19802 + }, + { + "epoch": 0.47, + "grad_norm": 2.2466551948702302, + "learning_rate": 1.1563138803933031e-05, + "loss": 1.1046, + "step": 19803 + }, + { + "epoch": 0.47, + "grad_norm": 2.0058249788784965, + "learning_rate": 1.1562385138057177e-05, + "loss": 0.9667, + "step": 19804 + }, + { + "epoch": 0.47, + "grad_norm": 2.578619592088613, + "learning_rate": 1.1561631463084621e-05, + "loss": 1.0426, + "step": 19805 + }, + { + "epoch": 0.47, + "grad_norm": 2.436540686999249, + "learning_rate": 1.1560877779019752e-05, + "loss": 1.0813, + "step": 19806 + }, + { + "epoch": 0.47, + "grad_norm": 1.061974856076707, + "learning_rate": 1.1560124085866957e-05, + "loss": 0.9222, + "step": 19807 + }, + { + "epoch": 0.47, + "grad_norm": 2.15084763892804, + "learning_rate": 1.1559370383630628e-05, + "loss": 1.0546, + "step": 19808 + }, + { + "epoch": 0.47, + "grad_norm": 1.8372626723562284, + "learning_rate": 1.155861667231515e-05, + "loss": 1.1068, + "step": 19809 + }, + { + "epoch": 0.47, + "grad_norm": 2.1931256849559513, + "learning_rate": 1.1557862951924912e-05, + "loss": 1.0922, + "step": 19810 + }, + { + "epoch": 0.47, + "grad_norm": 2.39142268346401, + "learning_rate": 1.1557109222464302e-05, + "loss": 0.9654, + "step": 19811 + }, + { + "epoch": 0.47, + "grad_norm": 2.1697849636938473, + "learning_rate": 1.155635548393771e-05, + "loss": 0.9502, + "step": 19812 + }, + { + "epoch": 0.47, + "grad_norm": 1.8808304150279003, + "learning_rate": 1.1555601736349525e-05, + "loss": 1.1349, + "step": 19813 + }, + { + "epoch": 0.47, + "grad_norm": 2.167794781548092, + "learning_rate": 1.1554847979704133e-05, + "loss": 1.1719, + "step": 19814 + }, + { + "epoch": 0.47, + "grad_norm": 2.0420833566838166, + "learning_rate": 1.1554094214005922e-05, + "loss": 0.9386, + "step": 19815 + }, + { + "epoch": 0.47, + "grad_norm": 2.042200154419075, + "learning_rate": 1.1553340439259287e-05, + "loss": 0.9362, + "step": 19816 + }, + { + "epoch": 0.47, + "grad_norm": 2.2392044204505357, + "learning_rate": 1.1552586655468608e-05, + "loss": 0.9719, + "step": 19817 + }, + { + "epoch": 0.47, + "grad_norm": 2.288044929340426, + "learning_rate": 1.1551832862638282e-05, + "loss": 1.1386, + "step": 19818 + }, + { + "epoch": 0.47, + "grad_norm": 2.0025417412738373, + "learning_rate": 1.1551079060772687e-05, + "loss": 0.9533, + "step": 19819 + }, + { + "epoch": 0.47, + "grad_norm": 1.9621123928794535, + "learning_rate": 1.1550325249876226e-05, + "loss": 0.9937, + "step": 19820 + }, + { + "epoch": 0.47, + "grad_norm": 1.202822556272135, + "learning_rate": 1.1549571429953276e-05, + "loss": 0.9833, + "step": 19821 + }, + { + "epoch": 0.47, + "grad_norm": 2.225873063138831, + "learning_rate": 1.1548817601008233e-05, + "loss": 1.0587, + "step": 19822 + }, + { + "epoch": 0.47, + "grad_norm": 1.9265050974886997, + "learning_rate": 1.1548063763045483e-05, + "loss": 1.0233, + "step": 19823 + }, + { + "epoch": 0.47, + "grad_norm": 1.9135403164106228, + "learning_rate": 1.154730991606941e-05, + "loss": 1.0496, + "step": 19824 + }, + { + "epoch": 0.47, + "grad_norm": 1.1342037216446348, + "learning_rate": 1.1546556060084418e-05, + "loss": 1.0016, + "step": 19825 + }, + { + "epoch": 0.47, + "grad_norm": 1.7735869325922549, + "learning_rate": 1.1545802195094879e-05, + "loss": 0.9335, + "step": 19826 + }, + { + "epoch": 0.47, + "grad_norm": 1.1014315084683552, + "learning_rate": 1.1545048321105196e-05, + "loss": 0.9776, + "step": 19827 + }, + { + "epoch": 0.47, + "grad_norm": 1.849904360040338, + "learning_rate": 1.1544294438119751e-05, + "loss": 1.0331, + "step": 19828 + }, + { + "epoch": 0.47, + "grad_norm": 2.2214695359491237, + "learning_rate": 1.1543540546142934e-05, + "loss": 0.874, + "step": 19829 + }, + { + "epoch": 0.47, + "grad_norm": 2.0692153433539673, + "learning_rate": 1.1542786645179139e-05, + "loss": 0.9471, + "step": 19830 + }, + { + "epoch": 0.47, + "grad_norm": 2.060709080142276, + "learning_rate": 1.1542032735232746e-05, + "loss": 0.9865, + "step": 19831 + }, + { + "epoch": 0.47, + "grad_norm": 1.082436036790271, + "learning_rate": 1.1541278816308152e-05, + "loss": 1.0115, + "step": 19832 + }, + { + "epoch": 0.47, + "grad_norm": 1.8274670883843347, + "learning_rate": 1.1540524888409746e-05, + "loss": 1.0429, + "step": 19833 + }, + { + "epoch": 0.47, + "grad_norm": 1.7634195593351714, + "learning_rate": 1.1539770951541916e-05, + "loss": 0.9894, + "step": 19834 + }, + { + "epoch": 0.47, + "grad_norm": 2.027103830324171, + "learning_rate": 1.1539017005709052e-05, + "loss": 0.9352, + "step": 19835 + }, + { + "epoch": 0.47, + "grad_norm": 1.8349919786005078, + "learning_rate": 1.1538263050915543e-05, + "loss": 0.9091, + "step": 19836 + }, + { + "epoch": 0.47, + "grad_norm": 2.0619286717595324, + "learning_rate": 1.1537509087165781e-05, + "loss": 0.9365, + "step": 19837 + }, + { + "epoch": 0.47, + "grad_norm": 3.1044160842417323, + "learning_rate": 1.1536755114464153e-05, + "loss": 1.0436, + "step": 19838 + }, + { + "epoch": 0.47, + "grad_norm": 2.9521192064330286, + "learning_rate": 1.1536001132815051e-05, + "loss": 0.8673, + "step": 19839 + }, + { + "epoch": 0.47, + "grad_norm": 1.9757268596571156, + "learning_rate": 1.1535247142222862e-05, + "loss": 0.8789, + "step": 19840 + }, + { + "epoch": 0.47, + "grad_norm": 1.9189888773285164, + "learning_rate": 1.153449314269198e-05, + "loss": 1.1128, + "step": 19841 + }, + { + "epoch": 0.47, + "grad_norm": 1.9921363753883354, + "learning_rate": 1.1533739134226797e-05, + "loss": 1.2041, + "step": 19842 + }, + { + "epoch": 0.47, + "grad_norm": 2.255009347054973, + "learning_rate": 1.1532985116831692e-05, + "loss": 1.021, + "step": 19843 + }, + { + "epoch": 0.47, + "grad_norm": 2.100791840368767, + "learning_rate": 1.1532231090511069e-05, + "loss": 1.1344, + "step": 19844 + }, + { + "epoch": 0.47, + "grad_norm": 2.1627056174978843, + "learning_rate": 1.1531477055269305e-05, + "loss": 0.9687, + "step": 19845 + }, + { + "epoch": 0.47, + "grad_norm": 2.5808662987355344, + "learning_rate": 1.15307230111108e-05, + "loss": 1.0309, + "step": 19846 + }, + { + "epoch": 0.47, + "grad_norm": 2.1385942454815488, + "learning_rate": 1.152996895803994e-05, + "loss": 1.0179, + "step": 19847 + }, + { + "epoch": 0.47, + "grad_norm": 1.8400165144935068, + "learning_rate": 1.1529214896061118e-05, + "loss": 0.9668, + "step": 19848 + }, + { + "epoch": 0.47, + "grad_norm": 1.8591971211513196, + "learning_rate": 1.152846082517872e-05, + "loss": 0.9834, + "step": 19849 + }, + { + "epoch": 0.47, + "grad_norm": 1.8750772133714093, + "learning_rate": 1.1527706745397143e-05, + "loss": 0.9814, + "step": 19850 + }, + { + "epoch": 0.47, + "grad_norm": 2.2335070448948833, + "learning_rate": 1.1526952656720772e-05, + "loss": 1.1128, + "step": 19851 + }, + { + "epoch": 0.47, + "grad_norm": 2.5954591214023237, + "learning_rate": 1.1526198559153998e-05, + "loss": 1.0252, + "step": 19852 + }, + { + "epoch": 0.47, + "grad_norm": 2.330612316834267, + "learning_rate": 1.1525444452701213e-05, + "loss": 1.064, + "step": 19853 + }, + { + "epoch": 0.47, + "grad_norm": 1.799139051482864, + "learning_rate": 1.1524690337366808e-05, + "loss": 0.9777, + "step": 19854 + }, + { + "epoch": 0.47, + "grad_norm": 1.9394456732646161, + "learning_rate": 1.1523936213155174e-05, + "loss": 0.9435, + "step": 19855 + }, + { + "epoch": 0.47, + "grad_norm": 2.0813638405140877, + "learning_rate": 1.1523182080070701e-05, + "loss": 0.9868, + "step": 19856 + }, + { + "epoch": 0.47, + "grad_norm": 2.018581715242756, + "learning_rate": 1.1522427938117777e-05, + "loss": 1.0109, + "step": 19857 + }, + { + "epoch": 0.47, + "grad_norm": 2.2803490688475274, + "learning_rate": 1.1521673787300798e-05, + "loss": 1.0925, + "step": 19858 + }, + { + "epoch": 0.47, + "grad_norm": 2.189690244267665, + "learning_rate": 1.1520919627624152e-05, + "loss": 0.9465, + "step": 19859 + }, + { + "epoch": 0.47, + "grad_norm": 2.3013050934864347, + "learning_rate": 1.152016545909223e-05, + "loss": 1.0469, + "step": 19860 + }, + { + "epoch": 0.47, + "grad_norm": 2.156571264538673, + "learning_rate": 1.1519411281709423e-05, + "loss": 0.9704, + "step": 19861 + }, + { + "epoch": 0.47, + "grad_norm": 1.923637442712017, + "learning_rate": 1.1518657095480123e-05, + "loss": 0.9811, + "step": 19862 + }, + { + "epoch": 0.47, + "grad_norm": 2.229508759096142, + "learning_rate": 1.1517902900408722e-05, + "loss": 0.9725, + "step": 19863 + }, + { + "epoch": 0.47, + "grad_norm": 1.891345001786572, + "learning_rate": 1.1517148696499607e-05, + "loss": 1.0495, + "step": 19864 + }, + { + "epoch": 0.47, + "grad_norm": 2.0509640641699414, + "learning_rate": 1.1516394483757176e-05, + "loss": 1.1195, + "step": 19865 + }, + { + "epoch": 0.47, + "grad_norm": 1.171850144003479, + "learning_rate": 1.1515640262185813e-05, + "loss": 0.9962, + "step": 19866 + }, + { + "epoch": 0.47, + "grad_norm": 2.7954425143572115, + "learning_rate": 1.1514886031789913e-05, + "loss": 0.9599, + "step": 19867 + }, + { + "epoch": 0.47, + "grad_norm": 2.0382631612213937, + "learning_rate": 1.1514131792573865e-05, + "loss": 0.9639, + "step": 19868 + }, + { + "epoch": 0.47, + "grad_norm": 1.9771693776578203, + "learning_rate": 1.1513377544542069e-05, + "loss": 0.8728, + "step": 19869 + }, + { + "epoch": 0.47, + "grad_norm": 2.075280631162742, + "learning_rate": 1.1512623287698901e-05, + "loss": 1.173, + "step": 19870 + }, + { + "epoch": 0.47, + "grad_norm": 1.1036400463212959, + "learning_rate": 1.1511869022048767e-05, + "loss": 0.9947, + "step": 19871 + }, + { + "epoch": 0.47, + "grad_norm": 2.0153902516876903, + "learning_rate": 1.1511114747596052e-05, + "loss": 0.8944, + "step": 19872 + }, + { + "epoch": 0.47, + "grad_norm": 2.1086729158359208, + "learning_rate": 1.1510360464345147e-05, + "loss": 1.0139, + "step": 19873 + }, + { + "epoch": 0.47, + "grad_norm": 2.0579569775465703, + "learning_rate": 1.1509606172300447e-05, + "loss": 1.1556, + "step": 19874 + }, + { + "epoch": 0.47, + "grad_norm": 1.1553660707910285, + "learning_rate": 1.150885187146634e-05, + "loss": 1.0572, + "step": 19875 + }, + { + "epoch": 0.47, + "grad_norm": 2.022409380753426, + "learning_rate": 1.1508097561847223e-05, + "loss": 1.0577, + "step": 19876 + }, + { + "epoch": 0.47, + "grad_norm": 2.529674540864103, + "learning_rate": 1.150734324344748e-05, + "loss": 1.0788, + "step": 19877 + }, + { + "epoch": 0.47, + "grad_norm": 2.0992515564018914, + "learning_rate": 1.1506588916271512e-05, + "loss": 1.0003, + "step": 19878 + }, + { + "epoch": 0.47, + "grad_norm": 2.191946023640855, + "learning_rate": 1.1505834580323704e-05, + "loss": 0.8366, + "step": 19879 + }, + { + "epoch": 0.47, + "grad_norm": 1.9924170046328251, + "learning_rate": 1.150508023560845e-05, + "loss": 1.172, + "step": 19880 + }, + { + "epoch": 0.47, + "grad_norm": 2.0851478892627178, + "learning_rate": 1.1504325882130143e-05, + "loss": 0.9474, + "step": 19881 + }, + { + "epoch": 0.47, + "grad_norm": 2.16262088059267, + "learning_rate": 1.1503571519893175e-05, + "loss": 0.9772, + "step": 19882 + }, + { + "epoch": 0.47, + "grad_norm": 1.1478871016771868, + "learning_rate": 1.1502817148901937e-05, + "loss": 0.9512, + "step": 19883 + }, + { + "epoch": 0.47, + "grad_norm": 2.0161404169011403, + "learning_rate": 1.1502062769160821e-05, + "loss": 1.0669, + "step": 19884 + }, + { + "epoch": 0.47, + "grad_norm": 1.7600477515883972, + "learning_rate": 1.1501308380674221e-05, + "loss": 1.1014, + "step": 19885 + }, + { + "epoch": 0.47, + "grad_norm": 2.02316605017943, + "learning_rate": 1.1500553983446527e-05, + "loss": 0.8894, + "step": 19886 + }, + { + "epoch": 0.47, + "grad_norm": 2.0911231514285324, + "learning_rate": 1.1499799577482134e-05, + "loss": 0.8836, + "step": 19887 + }, + { + "epoch": 0.47, + "grad_norm": 2.140323491511762, + "learning_rate": 1.1499045162785434e-05, + "loss": 1.0372, + "step": 19888 + }, + { + "epoch": 0.47, + "grad_norm": 1.9424249264259277, + "learning_rate": 1.1498290739360815e-05, + "loss": 1.0316, + "step": 19889 + }, + { + "epoch": 0.47, + "grad_norm": 2.0392315321090133, + "learning_rate": 1.1497536307212675e-05, + "loss": 1.0532, + "step": 19890 + }, + { + "epoch": 0.47, + "grad_norm": 2.182388312438634, + "learning_rate": 1.1496781866345406e-05, + "loss": 1.1158, + "step": 19891 + }, + { + "epoch": 0.47, + "grad_norm": 2.0033221023725813, + "learning_rate": 1.1496027416763399e-05, + "loss": 0.9576, + "step": 19892 + }, + { + "epoch": 0.47, + "grad_norm": 2.15315782762207, + "learning_rate": 1.1495272958471046e-05, + "loss": 0.9549, + "step": 19893 + }, + { + "epoch": 0.47, + "grad_norm": 1.1655981753879907, + "learning_rate": 1.1494518491472738e-05, + "loss": 0.9558, + "step": 19894 + }, + { + "epoch": 0.47, + "grad_norm": 2.0895301684966396, + "learning_rate": 1.1493764015772875e-05, + "loss": 1.0882, + "step": 19895 + }, + { + "epoch": 0.47, + "grad_norm": 2.1885020867569582, + "learning_rate": 1.1493009531375838e-05, + "loss": 1.0701, + "step": 19896 + }, + { + "epoch": 0.47, + "grad_norm": 2.0679029885616784, + "learning_rate": 1.1492255038286034e-05, + "loss": 0.9497, + "step": 19897 + }, + { + "epoch": 0.47, + "grad_norm": 2.0196361704721033, + "learning_rate": 1.1491500536507845e-05, + "loss": 0.9474, + "step": 19898 + }, + { + "epoch": 0.47, + "grad_norm": 2.475580750785375, + "learning_rate": 1.149074602604567e-05, + "loss": 0.9966, + "step": 19899 + }, + { + "epoch": 0.47, + "grad_norm": 2.203215080856232, + "learning_rate": 1.1489991506903896e-05, + "loss": 1.008, + "step": 19900 + }, + { + "epoch": 0.47, + "grad_norm": 2.2349517394077183, + "learning_rate": 1.1489236979086922e-05, + "loss": 1.089, + "step": 19901 + }, + { + "epoch": 0.47, + "grad_norm": 2.1158632562528346, + "learning_rate": 1.1488482442599138e-05, + "loss": 1.0754, + "step": 19902 + }, + { + "epoch": 0.47, + "grad_norm": 2.31572535773714, + "learning_rate": 1.148772789744494e-05, + "loss": 0.9573, + "step": 19903 + }, + { + "epoch": 0.47, + "grad_norm": 1.8850090408274396, + "learning_rate": 1.1486973343628719e-05, + "loss": 1.0106, + "step": 19904 + }, + { + "epoch": 0.47, + "grad_norm": 2.0630554368865686, + "learning_rate": 1.1486218781154866e-05, + "loss": 0.998, + "step": 19905 + }, + { + "epoch": 0.47, + "grad_norm": 1.9518590465021783, + "learning_rate": 1.1485464210027778e-05, + "loss": 0.9191, + "step": 19906 + }, + { + "epoch": 0.47, + "grad_norm": 1.9077407662090387, + "learning_rate": 1.1484709630251847e-05, + "loss": 1.1163, + "step": 19907 + }, + { + "epoch": 0.47, + "grad_norm": 2.045196185527567, + "learning_rate": 1.1483955041831465e-05, + "loss": 0.9081, + "step": 19908 + }, + { + "epoch": 0.47, + "grad_norm": 1.859232761473283, + "learning_rate": 1.148320044477103e-05, + "loss": 0.8829, + "step": 19909 + }, + { + "epoch": 0.47, + "grad_norm": 1.8556683449303795, + "learning_rate": 1.1482445839074929e-05, + "loss": 1.1039, + "step": 19910 + }, + { + "epoch": 0.47, + "grad_norm": 1.069030026003546, + "learning_rate": 1.1481691224747563e-05, + "loss": 0.9939, + "step": 19911 + }, + { + "epoch": 0.47, + "grad_norm": 1.9438171265740831, + "learning_rate": 1.1480936601793318e-05, + "loss": 1.2633, + "step": 19912 + }, + { + "epoch": 0.47, + "grad_norm": 1.1269574152866462, + "learning_rate": 1.1480181970216593e-05, + "loss": 0.959, + "step": 19913 + }, + { + "epoch": 0.47, + "grad_norm": 2.0248126325035622, + "learning_rate": 1.1479427330021778e-05, + "loss": 0.9122, + "step": 19914 + }, + { + "epoch": 0.47, + "grad_norm": 1.0490316705393088, + "learning_rate": 1.147867268121327e-05, + "loss": 0.9585, + "step": 19915 + }, + { + "epoch": 0.47, + "grad_norm": 1.9139206891538447, + "learning_rate": 1.1477918023795464e-05, + "loss": 0.9732, + "step": 19916 + }, + { + "epoch": 0.47, + "grad_norm": 2.0282234914690607, + "learning_rate": 1.1477163357772747e-05, + "loss": 1.0021, + "step": 19917 + }, + { + "epoch": 0.47, + "grad_norm": 2.475558009074752, + "learning_rate": 1.1476408683149522e-05, + "loss": 1.0151, + "step": 19918 + }, + { + "epoch": 0.47, + "grad_norm": 1.9924009375326932, + "learning_rate": 1.1475653999930174e-05, + "loss": 1.0989, + "step": 19919 + }, + { + "epoch": 0.47, + "grad_norm": 2.0804066479740895, + "learning_rate": 1.1474899308119105e-05, + "loss": 1.1791, + "step": 19920 + }, + { + "epoch": 0.47, + "grad_norm": 1.0719925893478675, + "learning_rate": 1.1474144607720701e-05, + "loss": 1.0176, + "step": 19921 + }, + { + "epoch": 0.47, + "grad_norm": 2.5103832494257987, + "learning_rate": 1.1473389898739363e-05, + "loss": 0.9713, + "step": 19922 + }, + { + "epoch": 0.47, + "grad_norm": 2.4371349998679377, + "learning_rate": 1.1472635181179483e-05, + "loss": 0.9396, + "step": 19923 + }, + { + "epoch": 0.47, + "grad_norm": 1.9480266080101911, + "learning_rate": 1.1471880455045452e-05, + "loss": 1.0487, + "step": 19924 + }, + { + "epoch": 0.47, + "grad_norm": 1.9815386869540226, + "learning_rate": 1.1471125720341671e-05, + "loss": 0.8933, + "step": 19925 + }, + { + "epoch": 0.47, + "grad_norm": 1.9701146850747788, + "learning_rate": 1.1470370977072524e-05, + "loss": 1.0157, + "step": 19926 + }, + { + "epoch": 0.47, + "grad_norm": 2.6143302176701204, + "learning_rate": 1.1469616225242418e-05, + "loss": 1.1244, + "step": 19927 + }, + { + "epoch": 0.47, + "grad_norm": 2.0790202476319384, + "learning_rate": 1.1468861464855738e-05, + "loss": 0.9805, + "step": 19928 + }, + { + "epoch": 0.47, + "grad_norm": 2.110287936326958, + "learning_rate": 1.1468106695916885e-05, + "loss": 0.9681, + "step": 19929 + }, + { + "epoch": 0.47, + "grad_norm": 3.640588922383246, + "learning_rate": 1.1467351918430248e-05, + "loss": 0.955, + "step": 19930 + }, + { + "epoch": 0.47, + "grad_norm": 2.139497909012753, + "learning_rate": 1.1466597132400221e-05, + "loss": 0.8981, + "step": 19931 + }, + { + "epoch": 0.47, + "grad_norm": 2.2113280599539693, + "learning_rate": 1.1465842337831203e-05, + "loss": 1.0787, + "step": 19932 + }, + { + "epoch": 0.47, + "grad_norm": 1.834290759938212, + "learning_rate": 1.1465087534727587e-05, + "loss": 1.0081, + "step": 19933 + }, + { + "epoch": 0.47, + "grad_norm": 2.2351777814106715, + "learning_rate": 1.1464332723093769e-05, + "loss": 0.9764, + "step": 19934 + }, + { + "epoch": 0.47, + "grad_norm": 2.1365915180273194, + "learning_rate": 1.146357790293414e-05, + "loss": 0.8672, + "step": 19935 + }, + { + "epoch": 0.47, + "grad_norm": 2.929424387243057, + "learning_rate": 1.1462823074253098e-05, + "loss": 1.1437, + "step": 19936 + }, + { + "epoch": 0.47, + "grad_norm": 3.7488237335492975, + "learning_rate": 1.1462068237055037e-05, + "loss": 0.9662, + "step": 19937 + }, + { + "epoch": 0.47, + "grad_norm": 1.8967080372949476, + "learning_rate": 1.146131339134435e-05, + "loss": 0.9883, + "step": 19938 + }, + { + "epoch": 0.47, + "grad_norm": 1.867708998989333, + "learning_rate": 1.1460558537125437e-05, + "loss": 0.9817, + "step": 19939 + }, + { + "epoch": 0.47, + "grad_norm": 1.1780588484337198, + "learning_rate": 1.1459803674402687e-05, + "loss": 0.9776, + "step": 19940 + }, + { + "epoch": 0.47, + "grad_norm": 2.0721227325157345, + "learning_rate": 1.14590488031805e-05, + "loss": 1.0339, + "step": 19941 + }, + { + "epoch": 0.47, + "grad_norm": 1.9277615499368594, + "learning_rate": 1.1458293923463269e-05, + "loss": 1.1062, + "step": 19942 + }, + { + "epoch": 0.47, + "grad_norm": 2.2352444558296116, + "learning_rate": 1.1457539035255387e-05, + "loss": 1.1798, + "step": 19943 + }, + { + "epoch": 0.47, + "grad_norm": 1.0635278175375003, + "learning_rate": 1.1456784138561254e-05, + "loss": 0.9586, + "step": 19944 + }, + { + "epoch": 0.47, + "grad_norm": 1.8864274017205378, + "learning_rate": 1.1456029233385259e-05, + "loss": 1.1305, + "step": 19945 + }, + { + "epoch": 0.47, + "grad_norm": 1.133586534524645, + "learning_rate": 1.1455274319731806e-05, + "loss": 0.8644, + "step": 19946 + }, + { + "epoch": 0.47, + "grad_norm": 2.1651085739501057, + "learning_rate": 1.1454519397605279e-05, + "loss": 0.9762, + "step": 19947 + }, + { + "epoch": 0.47, + "grad_norm": 2.2381670993054175, + "learning_rate": 1.1453764467010084e-05, + "loss": 1.0428, + "step": 19948 + }, + { + "epoch": 0.47, + "grad_norm": 1.878355964765756, + "learning_rate": 1.145300952795061e-05, + "loss": 0.9093, + "step": 19949 + }, + { + "epoch": 0.47, + "grad_norm": 1.845761692351619, + "learning_rate": 1.1452254580431256e-05, + "loss": 1.011, + "step": 19950 + }, + { + "epoch": 0.47, + "grad_norm": 1.8576149926590269, + "learning_rate": 1.1451499624456415e-05, + "loss": 0.9431, + "step": 19951 + }, + { + "epoch": 0.47, + "grad_norm": 1.8660138470638454, + "learning_rate": 1.1450744660030484e-05, + "loss": 0.9465, + "step": 19952 + }, + { + "epoch": 0.47, + "grad_norm": 1.9421527878819542, + "learning_rate": 1.1449989687157859e-05, + "loss": 1.0582, + "step": 19953 + }, + { + "epoch": 0.47, + "grad_norm": 1.7601480977324238, + "learning_rate": 1.1449234705842933e-05, + "loss": 1.0895, + "step": 19954 + }, + { + "epoch": 0.47, + "grad_norm": 3.325432765425495, + "learning_rate": 1.1448479716090103e-05, + "loss": 1.0254, + "step": 19955 + }, + { + "epoch": 0.47, + "grad_norm": 1.7837087874214432, + "learning_rate": 1.144772471790377e-05, + "loss": 1.0056, + "step": 19956 + }, + { + "epoch": 0.47, + "grad_norm": 2.0853053462891773, + "learning_rate": 1.1446969711288322e-05, + "loss": 0.9464, + "step": 19957 + }, + { + "epoch": 0.47, + "grad_norm": 4.046452415927737, + "learning_rate": 1.144621469624816e-05, + "loss": 0.9778, + "step": 19958 + }, + { + "epoch": 0.47, + "grad_norm": 1.0873247876064482, + "learning_rate": 1.1445459672787676e-05, + "loss": 0.9985, + "step": 19959 + }, + { + "epoch": 0.47, + "grad_norm": 2.6293906791821726, + "learning_rate": 1.144470464091127e-05, + "loss": 1.0946, + "step": 19960 + }, + { + "epoch": 0.47, + "grad_norm": 1.968128347456088, + "learning_rate": 1.1443949600623334e-05, + "loss": 0.9683, + "step": 19961 + }, + { + "epoch": 0.47, + "grad_norm": 2.0310624316847674, + "learning_rate": 1.1443194551928267e-05, + "loss": 1.0827, + "step": 19962 + }, + { + "epoch": 0.47, + "grad_norm": 2.530613055086468, + "learning_rate": 1.1442439494830467e-05, + "loss": 1.0382, + "step": 19963 + }, + { + "epoch": 0.47, + "grad_norm": 2.143010651212346, + "learning_rate": 1.1441684429334323e-05, + "loss": 1.0717, + "step": 19964 + }, + { + "epoch": 0.47, + "grad_norm": 2.4585920641861208, + "learning_rate": 1.144092935544424e-05, + "loss": 1.1017, + "step": 19965 + }, + { + "epoch": 0.47, + "grad_norm": 1.8680779259760398, + "learning_rate": 1.1440174273164608e-05, + "loss": 1.0497, + "step": 19966 + }, + { + "epoch": 0.47, + "grad_norm": 1.1328572356911117, + "learning_rate": 1.143941918249983e-05, + "loss": 0.9068, + "step": 19967 + }, + { + "epoch": 0.47, + "grad_norm": 2.0207329986870537, + "learning_rate": 1.143866408345429e-05, + "loss": 0.988, + "step": 19968 + }, + { + "epoch": 0.47, + "grad_norm": 3.0783440366289097, + "learning_rate": 1.1437908976032395e-05, + "loss": 1.1214, + "step": 19969 + }, + { + "epoch": 0.47, + "grad_norm": 1.810671940064691, + "learning_rate": 1.1437153860238541e-05, + "loss": 0.9866, + "step": 19970 + }, + { + "epoch": 0.47, + "grad_norm": 1.9189601180546476, + "learning_rate": 1.143639873607712e-05, + "loss": 0.8942, + "step": 19971 + }, + { + "epoch": 0.47, + "grad_norm": 2.3500568957809436, + "learning_rate": 1.1435643603552532e-05, + "loss": 1.0453, + "step": 19972 + }, + { + "epoch": 0.47, + "grad_norm": 2.046023448889404, + "learning_rate": 1.143488846266917e-05, + "loss": 0.9427, + "step": 19973 + }, + { + "epoch": 0.47, + "grad_norm": 2.1697337720241947, + "learning_rate": 1.143413331343144e-05, + "loss": 0.9573, + "step": 19974 + }, + { + "epoch": 0.47, + "grad_norm": 2.0411088833196414, + "learning_rate": 1.1433378155843725e-05, + "loss": 0.9537, + "step": 19975 + }, + { + "epoch": 0.47, + "grad_norm": 2.2013445580611997, + "learning_rate": 1.1432622989910434e-05, + "loss": 1.0604, + "step": 19976 + }, + { + "epoch": 0.47, + "grad_norm": 1.0993752628405258, + "learning_rate": 1.1431867815635951e-05, + "loss": 0.9668, + "step": 19977 + }, + { + "epoch": 0.47, + "grad_norm": 2.31468384675175, + "learning_rate": 1.1431112633024689e-05, + "loss": 0.9667, + "step": 19978 + }, + { + "epoch": 0.47, + "grad_norm": 2.351769740376425, + "learning_rate": 1.1430357442081027e-05, + "loss": 1.1589, + "step": 19979 + }, + { + "epoch": 0.47, + "grad_norm": 1.9424779152710756, + "learning_rate": 1.1429602242809379e-05, + "loss": 1.0048, + "step": 19980 + }, + { + "epoch": 0.47, + "grad_norm": 2.216886766749563, + "learning_rate": 1.142884703521413e-05, + "loss": 0.9837, + "step": 19981 + }, + { + "epoch": 0.47, + "grad_norm": 2.5038216635361694, + "learning_rate": 1.1428091819299683e-05, + "loss": 0.9749, + "step": 19982 + }, + { + "epoch": 0.47, + "grad_norm": 2.2497339128571534, + "learning_rate": 1.142733659507043e-05, + "loss": 1.0564, + "step": 19983 + }, + { + "epoch": 0.47, + "grad_norm": 3.434120340525492, + "learning_rate": 1.1426581362530775e-05, + "loss": 0.9934, + "step": 19984 + }, + { + "epoch": 0.47, + "grad_norm": 2.689562281268228, + "learning_rate": 1.142582612168511e-05, + "loss": 0.9425, + "step": 19985 + }, + { + "epoch": 0.47, + "grad_norm": 1.889957532011651, + "learning_rate": 1.1425070872537836e-05, + "loss": 1.0779, + "step": 19986 + }, + { + "epoch": 0.47, + "grad_norm": 1.9770012247178022, + "learning_rate": 1.1424315615093343e-05, + "loss": 1.06, + "step": 19987 + }, + { + "epoch": 0.47, + "grad_norm": 1.9243709632217378, + "learning_rate": 1.142356034935604e-05, + "loss": 1.1067, + "step": 19988 + }, + { + "epoch": 0.47, + "grad_norm": 2.3920393693375503, + "learning_rate": 1.1422805075330315e-05, + "loss": 1.0861, + "step": 19989 + }, + { + "epoch": 0.47, + "grad_norm": 1.8085251837909586, + "learning_rate": 1.1422049793020566e-05, + "loss": 1.1245, + "step": 19990 + }, + { + "epoch": 0.47, + "grad_norm": 1.9718209110279852, + "learning_rate": 1.1421294502431197e-05, + "loss": 1.0767, + "step": 19991 + }, + { + "epoch": 0.47, + "grad_norm": 2.126472184620744, + "learning_rate": 1.1420539203566598e-05, + "loss": 1.1584, + "step": 19992 + }, + { + "epoch": 0.47, + "grad_norm": 2.7548186110206494, + "learning_rate": 1.1419783896431174e-05, + "loss": 0.9839, + "step": 19993 + }, + { + "epoch": 0.47, + "grad_norm": 2.0504876771335443, + "learning_rate": 1.1419028581029315e-05, + "loss": 1.0308, + "step": 19994 + }, + { + "epoch": 0.47, + "grad_norm": 1.0994583790801709, + "learning_rate": 1.1418273257365426e-05, + "loss": 0.9331, + "step": 19995 + }, + { + "epoch": 0.47, + "grad_norm": 1.0355672041932973, + "learning_rate": 1.1417517925443897e-05, + "loss": 1.0019, + "step": 19996 + }, + { + "epoch": 0.47, + "grad_norm": 2.0311444388066615, + "learning_rate": 1.1416762585269134e-05, + "loss": 1.0202, + "step": 19997 + }, + { + "epoch": 0.47, + "grad_norm": 2.1659501685441565, + "learning_rate": 1.1416007236845526e-05, + "loss": 0.9671, + "step": 19998 + }, + { + "epoch": 0.47, + "grad_norm": 2.1086784623416173, + "learning_rate": 1.141525188017748e-05, + "loss": 1.1144, + "step": 19999 + }, + { + "epoch": 0.47, + "grad_norm": 2.316462464036458, + "learning_rate": 1.1414496515269384e-05, + "loss": 1.1451, + "step": 20000 + }, + { + "epoch": 0.47, + "grad_norm": 2.092087413719834, + "learning_rate": 1.141374114212565e-05, + "loss": 1.0101, + "step": 20001 + }, + { + "epoch": 0.47, + "grad_norm": 2.3742020281688916, + "learning_rate": 1.141298576075066e-05, + "loss": 0.9616, + "step": 20002 + }, + { + "epoch": 0.47, + "grad_norm": 2.17061065395692, + "learning_rate": 1.1412230371148822e-05, + "loss": 0.951, + "step": 20003 + }, + { + "epoch": 0.47, + "grad_norm": 1.9083206107402286, + "learning_rate": 1.1411474973324533e-05, + "loss": 0.9378, + "step": 20004 + }, + { + "epoch": 0.47, + "grad_norm": 3.2840775759779977, + "learning_rate": 1.1410719567282188e-05, + "loss": 1.0036, + "step": 20005 + }, + { + "epoch": 0.47, + "grad_norm": 1.9933680951392891, + "learning_rate": 1.1409964153026188e-05, + "loss": 1.0777, + "step": 20006 + }, + { + "epoch": 0.47, + "grad_norm": 2.526441986143813, + "learning_rate": 1.1409208730560927e-05, + "loss": 0.8637, + "step": 20007 + }, + { + "epoch": 0.47, + "grad_norm": 1.83974910279275, + "learning_rate": 1.1408453299890811e-05, + "loss": 1.1156, + "step": 20008 + }, + { + "epoch": 0.47, + "grad_norm": 2.0192498463778557, + "learning_rate": 1.1407697861020232e-05, + "loss": 0.9194, + "step": 20009 + }, + { + "epoch": 0.47, + "grad_norm": 1.7652821991353929, + "learning_rate": 1.1406942413953593e-05, + "loss": 1.0009, + "step": 20010 + }, + { + "epoch": 0.47, + "grad_norm": 1.856402776432504, + "learning_rate": 1.1406186958695287e-05, + "loss": 1.0005, + "step": 20011 + }, + { + "epoch": 0.47, + "grad_norm": 2.162714159642325, + "learning_rate": 1.1405431495249714e-05, + "loss": 0.9922, + "step": 20012 + }, + { + "epoch": 0.47, + "grad_norm": 2.180190886629923, + "learning_rate": 1.1404676023621278e-05, + "loss": 1.1642, + "step": 20013 + }, + { + "epoch": 0.47, + "grad_norm": 2.0536925559461547, + "learning_rate": 1.140392054381437e-05, + "loss": 0.9647, + "step": 20014 + }, + { + "epoch": 0.47, + "grad_norm": 2.090148617664637, + "learning_rate": 1.1403165055833394e-05, + "loss": 1.0131, + "step": 20015 + }, + { + "epoch": 0.47, + "grad_norm": 1.8767286154230294, + "learning_rate": 1.1402409559682748e-05, + "loss": 1.1552, + "step": 20016 + }, + { + "epoch": 0.47, + "grad_norm": 2.0344590952860777, + "learning_rate": 1.1401654055366826e-05, + "loss": 0.9931, + "step": 20017 + }, + { + "epoch": 0.47, + "grad_norm": 1.9095518632364124, + "learning_rate": 1.1400898542890034e-05, + "loss": 1.0954, + "step": 20018 + }, + { + "epoch": 0.47, + "grad_norm": 1.0840859217376766, + "learning_rate": 1.1400143022256764e-05, + "loss": 0.9679, + "step": 20019 + }, + { + "epoch": 0.47, + "grad_norm": 1.9104600734955135, + "learning_rate": 1.139938749347142e-05, + "loss": 1.0131, + "step": 20020 + }, + { + "epoch": 0.47, + "grad_norm": 1.965898153855063, + "learning_rate": 1.1398631956538398e-05, + "loss": 0.9407, + "step": 20021 + }, + { + "epoch": 0.47, + "grad_norm": 2.2516481236622194, + "learning_rate": 1.1397876411462096e-05, + "loss": 1.1007, + "step": 20022 + }, + { + "epoch": 0.47, + "grad_norm": 2.2961703085352942, + "learning_rate": 1.1397120858246921e-05, + "loss": 1.0305, + "step": 20023 + }, + { + "epoch": 0.47, + "grad_norm": 2.0057725330412093, + "learning_rate": 1.139636529689726e-05, + "loss": 1.0726, + "step": 20024 + }, + { + "epoch": 0.47, + "grad_norm": 1.7381915857597585, + "learning_rate": 1.1395609727417522e-05, + "loss": 1.1023, + "step": 20025 + }, + { + "epoch": 0.47, + "grad_norm": 1.8165748805683888, + "learning_rate": 1.13948541498121e-05, + "loss": 0.9916, + "step": 20026 + }, + { + "epoch": 0.47, + "grad_norm": 1.8945768562829837, + "learning_rate": 1.1394098564085399e-05, + "loss": 1.0495, + "step": 20027 + }, + { + "epoch": 0.47, + "grad_norm": 1.954812353702194, + "learning_rate": 1.1393342970241809e-05, + "loss": 1.1183, + "step": 20028 + }, + { + "epoch": 0.47, + "grad_norm": 2.1675480988657583, + "learning_rate": 1.139258736828574e-05, + "loss": 0.915, + "step": 20029 + }, + { + "epoch": 0.47, + "grad_norm": 1.8530772933711308, + "learning_rate": 1.1391831758221585e-05, + "loss": 0.9326, + "step": 20030 + }, + { + "epoch": 0.47, + "grad_norm": 1.0641377545090764, + "learning_rate": 1.1391076140053745e-05, + "loss": 0.9132, + "step": 20031 + }, + { + "epoch": 0.47, + "grad_norm": 2.0494816755258642, + "learning_rate": 1.139032051378662e-05, + "loss": 0.963, + "step": 20032 + }, + { + "epoch": 0.47, + "grad_norm": 1.0822705557672376, + "learning_rate": 1.1389564879424605e-05, + "loss": 0.9569, + "step": 20033 + }, + { + "epoch": 0.47, + "grad_norm": 2.3049470233070584, + "learning_rate": 1.1388809236972107e-05, + "loss": 1.0658, + "step": 20034 + }, + { + "epoch": 0.47, + "grad_norm": 1.8570710123694554, + "learning_rate": 1.138805358643352e-05, + "loss": 0.9551, + "step": 20035 + }, + { + "epoch": 0.47, + "grad_norm": 2.348930353719985, + "learning_rate": 1.1387297927813248e-05, + "loss": 1.0067, + "step": 20036 + }, + { + "epoch": 0.47, + "grad_norm": 2.0149684029930373, + "learning_rate": 1.1386542261115685e-05, + "loss": 0.9343, + "step": 20037 + }, + { + "epoch": 0.47, + "grad_norm": 2.093034371535933, + "learning_rate": 1.1385786586345238e-05, + "loss": 0.9902, + "step": 20038 + }, + { + "epoch": 0.47, + "grad_norm": 2.1018568236335082, + "learning_rate": 1.13850309035063e-05, + "loss": 1.0898, + "step": 20039 + }, + { + "epoch": 0.47, + "grad_norm": 2.1165552662317193, + "learning_rate": 1.1384275212603273e-05, + "loss": 0.9145, + "step": 20040 + }, + { + "epoch": 0.47, + "grad_norm": 2.1317530741294752, + "learning_rate": 1.138351951364056e-05, + "loss": 0.9021, + "step": 20041 + }, + { + "epoch": 0.47, + "grad_norm": 1.8947662135303662, + "learning_rate": 1.1382763806622556e-05, + "loss": 1.1061, + "step": 20042 + }, + { + "epoch": 0.47, + "grad_norm": 2.325789672431393, + "learning_rate": 1.1382008091553664e-05, + "loss": 1.0292, + "step": 20043 + }, + { + "epoch": 0.47, + "grad_norm": 2.846714546127593, + "learning_rate": 1.1381252368438283e-05, + "loss": 0.9569, + "step": 20044 + }, + { + "epoch": 0.47, + "grad_norm": 2.6517598603421844, + "learning_rate": 1.1380496637280816e-05, + "loss": 0.9137, + "step": 20045 + }, + { + "epoch": 0.47, + "grad_norm": 1.9733397704180269, + "learning_rate": 1.1379740898085657e-05, + "loss": 0.917, + "step": 20046 + }, + { + "epoch": 0.47, + "grad_norm": 1.982617505995817, + "learning_rate": 1.1378985150857211e-05, + "loss": 1.0203, + "step": 20047 + }, + { + "epoch": 0.47, + "grad_norm": 1.8652015558321378, + "learning_rate": 1.137822939559988e-05, + "loss": 1.1353, + "step": 20048 + }, + { + "epoch": 0.47, + "grad_norm": 1.8913840904684769, + "learning_rate": 1.1377473632318057e-05, + "loss": 0.9946, + "step": 20049 + }, + { + "epoch": 0.47, + "grad_norm": 1.9606460208852012, + "learning_rate": 1.1376717861016149e-05, + "loss": 1.1162, + "step": 20050 + }, + { + "epoch": 0.47, + "grad_norm": 1.868584398390512, + "learning_rate": 1.137596208169855e-05, + "loss": 1.0634, + "step": 20051 + }, + { + "epoch": 0.47, + "grad_norm": 2.2942478046260986, + "learning_rate": 1.1375206294369669e-05, + "loss": 0.9957, + "step": 20052 + }, + { + "epoch": 0.47, + "grad_norm": 2.1308286666405643, + "learning_rate": 1.1374450499033898e-05, + "loss": 1.0475, + "step": 20053 + }, + { + "epoch": 0.47, + "grad_norm": 2.0044615672769646, + "learning_rate": 1.1373694695695644e-05, + "loss": 1.1613, + "step": 20054 + }, + { + "epoch": 0.47, + "grad_norm": 1.8780534958110773, + "learning_rate": 1.1372938884359302e-05, + "loss": 0.9237, + "step": 20055 + }, + { + "epoch": 0.47, + "grad_norm": 2.077050152804133, + "learning_rate": 1.1372183065029273e-05, + "loss": 1.0388, + "step": 20056 + }, + { + "epoch": 0.47, + "grad_norm": 1.8094484680763427, + "learning_rate": 1.1371427237709964e-05, + "loss": 1.0706, + "step": 20057 + }, + { + "epoch": 0.47, + "grad_norm": 2.107028636247409, + "learning_rate": 1.1370671402405768e-05, + "loss": 1.0273, + "step": 20058 + }, + { + "epoch": 0.47, + "grad_norm": 2.026045086435584, + "learning_rate": 1.1369915559121094e-05, + "loss": 1.0152, + "step": 20059 + }, + { + "epoch": 0.47, + "grad_norm": 2.2502016006782837, + "learning_rate": 1.1369159707860336e-05, + "loss": 1.0374, + "step": 20060 + }, + { + "epoch": 0.47, + "grad_norm": 2.033937807028072, + "learning_rate": 1.1368403848627896e-05, + "loss": 1.0257, + "step": 20061 + }, + { + "epoch": 0.47, + "grad_norm": 2.0045011806806667, + "learning_rate": 1.1367647981428173e-05, + "loss": 0.9822, + "step": 20062 + }, + { + "epoch": 0.47, + "grad_norm": 2.1157471843339173, + "learning_rate": 1.1366892106265575e-05, + "loss": 1.1526, + "step": 20063 + }, + { + "epoch": 0.47, + "grad_norm": 2.05197211861693, + "learning_rate": 1.1366136223144495e-05, + "loss": 0.9901, + "step": 20064 + }, + { + "epoch": 0.47, + "grad_norm": 2.043877736062151, + "learning_rate": 1.136538033206934e-05, + "loss": 1.1546, + "step": 20065 + }, + { + "epoch": 0.47, + "grad_norm": 1.830343814378018, + "learning_rate": 1.1364624433044507e-05, + "loss": 1.0762, + "step": 20066 + }, + { + "epoch": 0.47, + "grad_norm": 2.326850106209965, + "learning_rate": 1.1363868526074397e-05, + "loss": 0.9778, + "step": 20067 + }, + { + "epoch": 0.47, + "grad_norm": 2.0875908552901206, + "learning_rate": 1.1363112611163415e-05, + "loss": 0.9239, + "step": 20068 + }, + { + "epoch": 0.47, + "grad_norm": 1.1213907871564537, + "learning_rate": 1.1362356688315959e-05, + "loss": 0.9722, + "step": 20069 + }, + { + "epoch": 0.47, + "grad_norm": 1.9283289114539064, + "learning_rate": 1.136160075753643e-05, + "loss": 1.0785, + "step": 20070 + }, + { + "epoch": 0.47, + "grad_norm": 1.1028229427799192, + "learning_rate": 1.1360844818829234e-05, + "loss": 0.9495, + "step": 20071 + }, + { + "epoch": 0.47, + "grad_norm": 1.8413903862604066, + "learning_rate": 1.1360088872198764e-05, + "loss": 1.0539, + "step": 20072 + }, + { + "epoch": 0.47, + "grad_norm": 4.163335088980627, + "learning_rate": 1.1359332917649427e-05, + "loss": 0.8725, + "step": 20073 + }, + { + "epoch": 0.47, + "grad_norm": 2.076314028613027, + "learning_rate": 1.135857695518563e-05, + "loss": 0.9933, + "step": 20074 + }, + { + "epoch": 0.47, + "grad_norm": 2.398127363000356, + "learning_rate": 1.1357820984811758e-05, + "loss": 0.9534, + "step": 20075 + }, + { + "epoch": 0.47, + "grad_norm": 2.2088638480635114, + "learning_rate": 1.135706500653223e-05, + "loss": 0.9303, + "step": 20076 + }, + { + "epoch": 0.47, + "grad_norm": 2.0590096130327833, + "learning_rate": 1.1356309020351435e-05, + "loss": 1.1299, + "step": 20077 + }, + { + "epoch": 0.47, + "grad_norm": 1.9153460025520912, + "learning_rate": 1.1355553026273783e-05, + "loss": 1.0495, + "step": 20078 + }, + { + "epoch": 0.47, + "grad_norm": 1.976408148688013, + "learning_rate": 1.1354797024303668e-05, + "loss": 1.1148, + "step": 20079 + }, + { + "epoch": 0.47, + "grad_norm": 2.2322046484185702, + "learning_rate": 1.1354041014445499e-05, + "loss": 1.1109, + "step": 20080 + }, + { + "epoch": 0.47, + "grad_norm": 2.821079305936937, + "learning_rate": 1.1353284996703672e-05, + "loss": 1.0256, + "step": 20081 + }, + { + "epoch": 0.47, + "grad_norm": 1.8583096536500425, + "learning_rate": 1.1352528971082594e-05, + "loss": 1.0677, + "step": 20082 + }, + { + "epoch": 0.47, + "grad_norm": 1.8868470343323152, + "learning_rate": 1.1351772937586663e-05, + "loss": 1.0547, + "step": 20083 + }, + { + "epoch": 0.47, + "grad_norm": 2.1346155754764813, + "learning_rate": 1.1351016896220282e-05, + "loss": 1.1511, + "step": 20084 + }, + { + "epoch": 0.47, + "grad_norm": 1.9789200718429378, + "learning_rate": 1.1350260846987852e-05, + "loss": 0.93, + "step": 20085 + }, + { + "epoch": 0.47, + "grad_norm": 1.8536341921811228, + "learning_rate": 1.1349504789893776e-05, + "loss": 1.015, + "step": 20086 + }, + { + "epoch": 0.47, + "grad_norm": 2.144714880465868, + "learning_rate": 1.1348748724942458e-05, + "loss": 0.8426, + "step": 20087 + }, + { + "epoch": 0.47, + "grad_norm": 1.9552244291597685, + "learning_rate": 1.1347992652138298e-05, + "loss": 1.1301, + "step": 20088 + }, + { + "epoch": 0.47, + "grad_norm": 1.9578347719456282, + "learning_rate": 1.1347236571485696e-05, + "loss": 0.8398, + "step": 20089 + }, + { + "epoch": 0.47, + "grad_norm": 1.8841631005942954, + "learning_rate": 1.1346480482989055e-05, + "loss": 0.8354, + "step": 20090 + }, + { + "epoch": 0.47, + "grad_norm": 1.0966155629165724, + "learning_rate": 1.1345724386652781e-05, + "loss": 0.9675, + "step": 20091 + }, + { + "epoch": 0.47, + "grad_norm": 2.1807823987192707, + "learning_rate": 1.1344968282481273e-05, + "loss": 1.0754, + "step": 20092 + }, + { + "epoch": 0.47, + "grad_norm": 2.2527970385865053, + "learning_rate": 1.1344212170478932e-05, + "loss": 1.0593, + "step": 20093 + }, + { + "epoch": 0.47, + "grad_norm": 2.074002365546265, + "learning_rate": 1.1343456050650163e-05, + "loss": 1.0646, + "step": 20094 + }, + { + "epoch": 0.47, + "grad_norm": 1.8978094862112804, + "learning_rate": 1.134269992299937e-05, + "loss": 1.1339, + "step": 20095 + }, + { + "epoch": 0.47, + "grad_norm": 2.066301328570204, + "learning_rate": 1.1341943787530951e-05, + "loss": 0.9836, + "step": 20096 + }, + { + "epoch": 0.47, + "grad_norm": 1.8265410909364763, + "learning_rate": 1.1341187644249311e-05, + "loss": 1.0948, + "step": 20097 + }, + { + "epoch": 0.47, + "grad_norm": 1.8182323245735512, + "learning_rate": 1.134043149315885e-05, + "loss": 0.951, + "step": 20098 + }, + { + "epoch": 0.47, + "grad_norm": 2.5986288462905693, + "learning_rate": 1.1339675334263976e-05, + "loss": 0.9756, + "step": 20099 + }, + { + "epoch": 0.47, + "grad_norm": 2.416115148636949, + "learning_rate": 1.1338919167569085e-05, + "loss": 0.9799, + "step": 20100 + }, + { + "epoch": 0.47, + "grad_norm": 2.1612683013266767, + "learning_rate": 1.1338162993078585e-05, + "loss": 1.1207, + "step": 20101 + }, + { + "epoch": 0.47, + "grad_norm": 1.9701384263658486, + "learning_rate": 1.1337406810796872e-05, + "loss": 1.0372, + "step": 20102 + }, + { + "epoch": 0.47, + "grad_norm": 2.1748095274881227, + "learning_rate": 1.133665062072836e-05, + "loss": 0.9447, + "step": 20103 + }, + { + "epoch": 0.47, + "grad_norm": 1.8836737446543121, + "learning_rate": 1.1335894422877441e-05, + "loss": 1.092, + "step": 20104 + }, + { + "epoch": 0.47, + "grad_norm": 6.638833652941185, + "learning_rate": 1.1335138217248518e-05, + "loss": 1.0274, + "step": 20105 + }, + { + "epoch": 0.47, + "grad_norm": 1.8076766195991132, + "learning_rate": 1.1334382003846006e-05, + "loss": 0.9815, + "step": 20106 + }, + { + "epoch": 0.47, + "grad_norm": 2.1171700294986144, + "learning_rate": 1.133362578267429e-05, + "loss": 1.0694, + "step": 20107 + }, + { + "epoch": 0.47, + "grad_norm": 1.5191614052103368, + "learning_rate": 1.133286955373779e-05, + "loss": 0.9596, + "step": 20108 + }, + { + "epoch": 0.47, + "grad_norm": 2.693848240433796, + "learning_rate": 1.1332113317040898e-05, + "loss": 1.0896, + "step": 20109 + }, + { + "epoch": 0.47, + "grad_norm": 1.8702923044081972, + "learning_rate": 1.1331357072588024e-05, + "loss": 1.0606, + "step": 20110 + }, + { + "epoch": 0.47, + "grad_norm": 1.9788001275814513, + "learning_rate": 1.1330600820383564e-05, + "loss": 1.0124, + "step": 20111 + }, + { + "epoch": 0.47, + "grad_norm": 1.0709852267642983, + "learning_rate": 1.1329844560431926e-05, + "loss": 0.9572, + "step": 20112 + }, + { + "epoch": 0.47, + "grad_norm": 1.9734801463888192, + "learning_rate": 1.132908829273751e-05, + "loss": 1.1168, + "step": 20113 + }, + { + "epoch": 0.47, + "grad_norm": 2.10250812959352, + "learning_rate": 1.1328332017304724e-05, + "loss": 1.108, + "step": 20114 + }, + { + "epoch": 0.47, + "grad_norm": 1.9408063533167135, + "learning_rate": 1.1327575734137968e-05, + "loss": 1.0164, + "step": 20115 + }, + { + "epoch": 0.47, + "grad_norm": 2.0672914344484674, + "learning_rate": 1.1326819443241645e-05, + "loss": 1.0258, + "step": 20116 + }, + { + "epoch": 0.47, + "grad_norm": 1.78662991039387, + "learning_rate": 1.1326063144620158e-05, + "loss": 1.0277, + "step": 20117 + }, + { + "epoch": 0.47, + "grad_norm": 2.0697228865173702, + "learning_rate": 1.1325306838277914e-05, + "loss": 0.9193, + "step": 20118 + }, + { + "epoch": 0.47, + "grad_norm": 1.998609225291356, + "learning_rate": 1.1324550524219313e-05, + "loss": 1.0697, + "step": 20119 + }, + { + "epoch": 0.47, + "grad_norm": 2.492903711974029, + "learning_rate": 1.1323794202448759e-05, + "loss": 1.0868, + "step": 20120 + }, + { + "epoch": 0.47, + "grad_norm": 1.1560848466903093, + "learning_rate": 1.1323037872970657e-05, + "loss": 0.9381, + "step": 20121 + }, + { + "epoch": 0.47, + "grad_norm": 2.9661202174681365, + "learning_rate": 1.1322281535789408e-05, + "loss": 0.9273, + "step": 20122 + }, + { + "epoch": 0.47, + "grad_norm": 2.1055527469718665, + "learning_rate": 1.1321525190909419e-05, + "loss": 1.12, + "step": 20123 + }, + { + "epoch": 0.47, + "grad_norm": 2.179377776212848, + "learning_rate": 1.1320768838335092e-05, + "loss": 0.9787, + "step": 20124 + }, + { + "epoch": 0.47, + "grad_norm": 1.9165381128453598, + "learning_rate": 1.1320012478070831e-05, + "loss": 1.0979, + "step": 20125 + }, + { + "epoch": 0.47, + "grad_norm": 2.102156240252423, + "learning_rate": 1.1319256110121036e-05, + "loss": 0.8656, + "step": 20126 + }, + { + "epoch": 0.47, + "grad_norm": 1.8671807876694708, + "learning_rate": 1.131849973449012e-05, + "loss": 1.0533, + "step": 20127 + }, + { + "epoch": 0.47, + "grad_norm": 1.9393955551824817, + "learning_rate": 1.1317743351182475e-05, + "loss": 0.8949, + "step": 20128 + }, + { + "epoch": 0.47, + "grad_norm": 1.056856125927999, + "learning_rate": 1.1316986960202516e-05, + "loss": 0.9279, + "step": 20129 + }, + { + "epoch": 0.47, + "grad_norm": 2.0431384231319565, + "learning_rate": 1.1316230561554637e-05, + "loss": 1.0418, + "step": 20130 + }, + { + "epoch": 0.47, + "grad_norm": 2.0429199084812995, + "learning_rate": 1.131547415524325e-05, + "loss": 1.0755, + "step": 20131 + }, + { + "epoch": 0.47, + "grad_norm": 2.0067806752571244, + "learning_rate": 1.1314717741272757e-05, + "loss": 1.1078, + "step": 20132 + }, + { + "epoch": 0.47, + "grad_norm": 2.210153183959386, + "learning_rate": 1.1313961319647561e-05, + "loss": 1.1046, + "step": 20133 + }, + { + "epoch": 0.47, + "grad_norm": 1.9922818410498129, + "learning_rate": 1.1313204890372065e-05, + "loss": 0.9632, + "step": 20134 + }, + { + "epoch": 0.47, + "grad_norm": 1.9903806974447769, + "learning_rate": 1.1312448453450672e-05, + "loss": 0.9516, + "step": 20135 + }, + { + "epoch": 0.47, + "grad_norm": 1.1427276927099712, + "learning_rate": 1.1311692008887793e-05, + "loss": 0.9286, + "step": 20136 + }, + { + "epoch": 0.47, + "grad_norm": 1.1461957996031025, + "learning_rate": 1.1310935556687825e-05, + "loss": 1.0317, + "step": 20137 + }, + { + "epoch": 0.47, + "grad_norm": 1.693385343451514, + "learning_rate": 1.1310179096855176e-05, + "loss": 1.0213, + "step": 20138 + }, + { + "epoch": 0.47, + "grad_norm": 1.7972849077725392, + "learning_rate": 1.1309422629394251e-05, + "loss": 0.9936, + "step": 20139 + }, + { + "epoch": 0.47, + "grad_norm": 2.115678464549937, + "learning_rate": 1.1308666154309451e-05, + "loss": 1.0382, + "step": 20140 + }, + { + "epoch": 0.47, + "grad_norm": 1.9786100373872972, + "learning_rate": 1.1307909671605182e-05, + "loss": 0.9541, + "step": 20141 + }, + { + "epoch": 0.47, + "grad_norm": 1.995329959893066, + "learning_rate": 1.1307153181285848e-05, + "loss": 1.0871, + "step": 20142 + }, + { + "epoch": 0.47, + "grad_norm": 2.3181647027770333, + "learning_rate": 1.1306396683355858e-05, + "loss": 1.0217, + "step": 20143 + }, + { + "epoch": 0.47, + "grad_norm": 1.7858883654997662, + "learning_rate": 1.130564017781961e-05, + "loss": 1.0077, + "step": 20144 + }, + { + "epoch": 0.47, + "grad_norm": 2.3788282052949663, + "learning_rate": 1.1304883664681512e-05, + "loss": 0.8065, + "step": 20145 + }, + { + "epoch": 0.47, + "grad_norm": 2.0376450478618477, + "learning_rate": 1.1304127143945971e-05, + "loss": 0.9171, + "step": 20146 + }, + { + "epoch": 0.47, + "grad_norm": 2.0300331557254965, + "learning_rate": 1.1303370615617384e-05, + "loss": 0.9514, + "step": 20147 + }, + { + "epoch": 0.47, + "grad_norm": 2.034424800568145, + "learning_rate": 1.1302614079700165e-05, + "loss": 0.955, + "step": 20148 + }, + { + "epoch": 0.47, + "grad_norm": 1.800306900566457, + "learning_rate": 1.130185753619871e-05, + "loss": 0.9133, + "step": 20149 + }, + { + "epoch": 0.47, + "grad_norm": 27.624738056521576, + "learning_rate": 1.1301100985117433e-05, + "loss": 0.9287, + "step": 20150 + }, + { + "epoch": 0.47, + "grad_norm": 3.726639771122712, + "learning_rate": 1.130034442646073e-05, + "loss": 0.9871, + "step": 20151 + }, + { + "epoch": 0.47, + "grad_norm": 1.9936789851986059, + "learning_rate": 1.1299587860233014e-05, + "loss": 1.0404, + "step": 20152 + }, + { + "epoch": 0.47, + "grad_norm": 2.126301230082756, + "learning_rate": 1.1298831286438683e-05, + "loss": 1.1198, + "step": 20153 + }, + { + "epoch": 0.47, + "grad_norm": 1.7902449351197374, + "learning_rate": 1.1298074705082145e-05, + "loss": 1.0838, + "step": 20154 + }, + { + "epoch": 0.47, + "grad_norm": 2.1974284031230855, + "learning_rate": 1.1297318116167808e-05, + "loss": 0.9271, + "step": 20155 + }, + { + "epoch": 0.47, + "grad_norm": 2.240311916650621, + "learning_rate": 1.129656151970007e-05, + "loss": 0.986, + "step": 20156 + }, + { + "epoch": 0.47, + "grad_norm": 1.999397089982946, + "learning_rate": 1.1295804915683343e-05, + "loss": 1.1278, + "step": 20157 + }, + { + "epoch": 0.47, + "grad_norm": 2.3360869155936204, + "learning_rate": 1.1295048304122029e-05, + "loss": 1.0711, + "step": 20158 + }, + { + "epoch": 0.47, + "grad_norm": 2.0261969174236127, + "learning_rate": 1.1294291685020535e-05, + "loss": 0.9878, + "step": 20159 + }, + { + "epoch": 0.47, + "grad_norm": 2.041074130711868, + "learning_rate": 1.1293535058383262e-05, + "loss": 1.0678, + "step": 20160 + }, + { + "epoch": 0.47, + "grad_norm": 1.9578120885328727, + "learning_rate": 1.1292778424214623e-05, + "loss": 1.0724, + "step": 20161 + }, + { + "epoch": 0.48, + "grad_norm": 1.9292091750986542, + "learning_rate": 1.1292021782519015e-05, + "loss": 1.01, + "step": 20162 + }, + { + "epoch": 0.48, + "grad_norm": 1.9277877991045853, + "learning_rate": 1.1291265133300849e-05, + "loss": 1.1286, + "step": 20163 + }, + { + "epoch": 0.48, + "grad_norm": 1.821282350269559, + "learning_rate": 1.1290508476564528e-05, + "loss": 0.9893, + "step": 20164 + }, + { + "epoch": 0.48, + "grad_norm": 1.971773745282857, + "learning_rate": 1.1289751812314458e-05, + "loss": 0.957, + "step": 20165 + }, + { + "epoch": 0.48, + "grad_norm": 1.070933792400141, + "learning_rate": 1.1288995140555044e-05, + "loss": 0.9111, + "step": 20166 + }, + { + "epoch": 0.48, + "grad_norm": 2.2358099721340583, + "learning_rate": 1.1288238461290693e-05, + "loss": 1.1091, + "step": 20167 + }, + { + "epoch": 0.48, + "grad_norm": 2.1968189147811037, + "learning_rate": 1.128748177452581e-05, + "loss": 1.0345, + "step": 20168 + }, + { + "epoch": 0.48, + "grad_norm": 1.1119848998184023, + "learning_rate": 1.12867250802648e-05, + "loss": 1.0384, + "step": 20169 + }, + { + "epoch": 0.48, + "grad_norm": 1.0832901342445278, + "learning_rate": 1.1285968378512069e-05, + "loss": 0.9265, + "step": 20170 + }, + { + "epoch": 0.48, + "grad_norm": 2.204674405052989, + "learning_rate": 1.1285211669272023e-05, + "loss": 1.0327, + "step": 20171 + }, + { + "epoch": 0.48, + "grad_norm": 1.076066235564527, + "learning_rate": 1.1284454952549067e-05, + "loss": 0.9941, + "step": 20172 + }, + { + "epoch": 0.48, + "grad_norm": 2.142017298057576, + "learning_rate": 1.1283698228347609e-05, + "loss": 0.9414, + "step": 20173 + }, + { + "epoch": 0.48, + "grad_norm": 3.1525482212756293, + "learning_rate": 1.1282941496672053e-05, + "loss": 1.0419, + "step": 20174 + }, + { + "epoch": 0.48, + "grad_norm": 2.585625751617487, + "learning_rate": 1.1282184757526806e-05, + "loss": 1.0102, + "step": 20175 + }, + { + "epoch": 0.48, + "grad_norm": 2.0355656661862978, + "learning_rate": 1.1281428010916276e-05, + "loss": 1.0397, + "step": 20176 + }, + { + "epoch": 0.48, + "grad_norm": 2.0866369091067836, + "learning_rate": 1.128067125684486e-05, + "loss": 1.0028, + "step": 20177 + }, + { + "epoch": 0.48, + "grad_norm": 2.137421958632206, + "learning_rate": 1.1279914495316975e-05, + "loss": 1.0634, + "step": 20178 + }, + { + "epoch": 0.48, + "grad_norm": 2.3781243388543793, + "learning_rate": 1.127915772633702e-05, + "loss": 0.9095, + "step": 20179 + }, + { + "epoch": 0.48, + "grad_norm": 1.8932466628993605, + "learning_rate": 1.1278400949909407e-05, + "loss": 0.9847, + "step": 20180 + }, + { + "epoch": 0.48, + "grad_norm": 1.9625364379030434, + "learning_rate": 1.1277644166038535e-05, + "loss": 1.0275, + "step": 20181 + }, + { + "epoch": 0.48, + "grad_norm": 2.05955727777928, + "learning_rate": 1.127688737472882e-05, + "loss": 1.1013, + "step": 20182 + }, + { + "epoch": 0.48, + "grad_norm": 2.6475175412999694, + "learning_rate": 1.1276130575984658e-05, + "loss": 0.9083, + "step": 20183 + }, + { + "epoch": 0.48, + "grad_norm": 2.231627930841778, + "learning_rate": 1.1275373769810461e-05, + "loss": 0.9283, + "step": 20184 + }, + { + "epoch": 0.48, + "grad_norm": 1.814451792524524, + "learning_rate": 1.1274616956210633e-05, + "loss": 1.0255, + "step": 20185 + }, + { + "epoch": 0.48, + "grad_norm": 3.1379104145611554, + "learning_rate": 1.1273860135189583e-05, + "loss": 0.9959, + "step": 20186 + }, + { + "epoch": 0.48, + "grad_norm": 2.267797853320179, + "learning_rate": 1.1273103306751713e-05, + "loss": 0.7978, + "step": 20187 + }, + { + "epoch": 0.48, + "grad_norm": 1.1327133760164034, + "learning_rate": 1.1272346470901436e-05, + "loss": 0.892, + "step": 20188 + }, + { + "epoch": 0.48, + "grad_norm": 2.1016597218453437, + "learning_rate": 1.1271589627643154e-05, + "loss": 0.9782, + "step": 20189 + }, + { + "epoch": 0.48, + "grad_norm": 1.858122920963165, + "learning_rate": 1.1270832776981272e-05, + "loss": 1.0519, + "step": 20190 + }, + { + "epoch": 0.48, + "grad_norm": 2.2460470097281098, + "learning_rate": 1.1270075918920202e-05, + "loss": 0.9894, + "step": 20191 + }, + { + "epoch": 0.48, + "grad_norm": 2.018014080163921, + "learning_rate": 1.126931905346435e-05, + "loss": 0.849, + "step": 20192 + }, + { + "epoch": 0.48, + "grad_norm": 1.8843408476524501, + "learning_rate": 1.1268562180618116e-05, + "loss": 0.9587, + "step": 20193 + }, + { + "epoch": 0.48, + "grad_norm": 2.519225585771094, + "learning_rate": 1.1267805300385914e-05, + "loss": 1.0177, + "step": 20194 + }, + { + "epoch": 0.48, + "grad_norm": 1.154567334033849, + "learning_rate": 1.1267048412772147e-05, + "loss": 1.0014, + "step": 20195 + }, + { + "epoch": 0.48, + "grad_norm": 2.215928261579041, + "learning_rate": 1.1266291517781225e-05, + "loss": 1.0028, + "step": 20196 + }, + { + "epoch": 0.48, + "grad_norm": 1.0898648538990185, + "learning_rate": 1.1265534615417553e-05, + "loss": 0.9467, + "step": 20197 + }, + { + "epoch": 0.48, + "grad_norm": 1.986953964789403, + "learning_rate": 1.1264777705685535e-05, + "loss": 1.0717, + "step": 20198 + }, + { + "epoch": 0.48, + "grad_norm": 2.3533200972208155, + "learning_rate": 1.1264020788589582e-05, + "loss": 0.999, + "step": 20199 + }, + { + "epoch": 0.48, + "grad_norm": 3.931914563556117, + "learning_rate": 1.12632638641341e-05, + "loss": 1.0041, + "step": 20200 + }, + { + "epoch": 0.48, + "grad_norm": 2.156912008365685, + "learning_rate": 1.1262506932323496e-05, + "loss": 1.0095, + "step": 20201 + }, + { + "epoch": 0.48, + "grad_norm": 2.2036271600943405, + "learning_rate": 1.1261749993162175e-05, + "loss": 1.1003, + "step": 20202 + }, + { + "epoch": 0.48, + "grad_norm": 2.0961222206239176, + "learning_rate": 1.1260993046654546e-05, + "loss": 0.9979, + "step": 20203 + }, + { + "epoch": 0.48, + "grad_norm": 1.9134687505120445, + "learning_rate": 1.1260236092805022e-05, + "loss": 0.9907, + "step": 20204 + }, + { + "epoch": 0.48, + "grad_norm": 1.1262047392521868, + "learning_rate": 1.1259479131617999e-05, + "loss": 1.0423, + "step": 20205 + }, + { + "epoch": 0.48, + "grad_norm": 2.7333119703424527, + "learning_rate": 1.1258722163097893e-05, + "loss": 1.1238, + "step": 20206 + }, + { + "epoch": 0.48, + "grad_norm": 1.92978585529252, + "learning_rate": 1.1257965187249102e-05, + "loss": 0.9488, + "step": 20207 + }, + { + "epoch": 0.48, + "grad_norm": 2.5159249254656224, + "learning_rate": 1.1257208204076046e-05, + "loss": 1.2487, + "step": 20208 + }, + { + "epoch": 0.48, + "grad_norm": 1.1340410624417556, + "learning_rate": 1.1256451213583121e-05, + "loss": 0.9661, + "step": 20209 + }, + { + "epoch": 0.48, + "grad_norm": 2.320497323632084, + "learning_rate": 1.1255694215774743e-05, + "loss": 0.9213, + "step": 20210 + }, + { + "epoch": 0.48, + "grad_norm": 1.0384259311993822, + "learning_rate": 1.1254937210655316e-05, + "loss": 0.9366, + "step": 20211 + }, + { + "epoch": 0.48, + "grad_norm": 2.151084903460691, + "learning_rate": 1.1254180198229245e-05, + "loss": 1.0738, + "step": 20212 + }, + { + "epoch": 0.48, + "grad_norm": 1.9045469401081223, + "learning_rate": 1.1253423178500943e-05, + "loss": 1.0894, + "step": 20213 + }, + { + "epoch": 0.48, + "grad_norm": 2.0610477745471507, + "learning_rate": 1.125266615147481e-05, + "loss": 0.9253, + "step": 20214 + }, + { + "epoch": 0.48, + "grad_norm": 2.2198675158507513, + "learning_rate": 1.125190911715526e-05, + "loss": 0.9518, + "step": 20215 + }, + { + "epoch": 0.48, + "grad_norm": 1.7953182056338826, + "learning_rate": 1.1251152075546698e-05, + "loss": 1.0205, + "step": 20216 + }, + { + "epoch": 0.48, + "grad_norm": 2.0392453184330552, + "learning_rate": 1.1250395026653536e-05, + "loss": 1.0462, + "step": 20217 + }, + { + "epoch": 0.48, + "grad_norm": 2.4751723218587736, + "learning_rate": 1.1249637970480174e-05, + "loss": 1.1016, + "step": 20218 + }, + { + "epoch": 0.48, + "grad_norm": 2.0144084526530484, + "learning_rate": 1.1248880907031025e-05, + "loss": 1.1128, + "step": 20219 + }, + { + "epoch": 0.48, + "grad_norm": 2.404331460939512, + "learning_rate": 1.1248123836310499e-05, + "loss": 1.2295, + "step": 20220 + }, + { + "epoch": 0.48, + "grad_norm": 2.039068172024917, + "learning_rate": 1.1247366758322997e-05, + "loss": 0.9703, + "step": 20221 + }, + { + "epoch": 0.48, + "grad_norm": 1.9850348195536354, + "learning_rate": 1.1246609673072932e-05, + "loss": 1.003, + "step": 20222 + }, + { + "epoch": 0.48, + "grad_norm": 2.3037533273067448, + "learning_rate": 1.124585258056471e-05, + "loss": 1.0473, + "step": 20223 + }, + { + "epoch": 0.48, + "grad_norm": 2.0015951102823486, + "learning_rate": 1.1245095480802743e-05, + "loss": 1.1054, + "step": 20224 + }, + { + "epoch": 0.48, + "grad_norm": 1.101690112493634, + "learning_rate": 1.1244338373791432e-05, + "loss": 0.9919, + "step": 20225 + }, + { + "epoch": 0.48, + "grad_norm": 3.270833421324531, + "learning_rate": 1.1243581259535193e-05, + "loss": 1.0064, + "step": 20226 + }, + { + "epoch": 0.48, + "grad_norm": 2.037513041893344, + "learning_rate": 1.1242824138038431e-05, + "loss": 0.9784, + "step": 20227 + }, + { + "epoch": 0.48, + "grad_norm": 1.920986261163738, + "learning_rate": 1.1242067009305546e-05, + "loss": 1.0172, + "step": 20228 + }, + { + "epoch": 0.48, + "grad_norm": 2.1300600454539755, + "learning_rate": 1.1241309873340962e-05, + "loss": 1.0405, + "step": 20229 + }, + { + "epoch": 0.48, + "grad_norm": 2.0654538183628834, + "learning_rate": 1.1240552730149073e-05, + "loss": 0.8665, + "step": 20230 + }, + { + "epoch": 0.48, + "grad_norm": 2.4075431219028385, + "learning_rate": 1.1239795579734296e-05, + "loss": 0.9698, + "step": 20231 + }, + { + "epoch": 0.48, + "grad_norm": 2.7939392419447113, + "learning_rate": 1.1239038422101033e-05, + "loss": 0.9483, + "step": 20232 + }, + { + "epoch": 0.48, + "grad_norm": 2.28475639378409, + "learning_rate": 1.1238281257253703e-05, + "loss": 1.0328, + "step": 20233 + }, + { + "epoch": 0.48, + "grad_norm": 2.262050083186511, + "learning_rate": 1.1237524085196703e-05, + "loss": 1.0627, + "step": 20234 + }, + { + "epoch": 0.48, + "grad_norm": 2.3058805485243927, + "learning_rate": 1.1236766905934447e-05, + "loss": 1.0363, + "step": 20235 + }, + { + "epoch": 0.48, + "grad_norm": 2.183760915330376, + "learning_rate": 1.123600971947134e-05, + "loss": 1.0589, + "step": 20236 + }, + { + "epoch": 0.48, + "grad_norm": 2.8573082308421505, + "learning_rate": 1.1235252525811797e-05, + "loss": 0.9318, + "step": 20237 + }, + { + "epoch": 0.48, + "grad_norm": 1.066398507044702, + "learning_rate": 1.1234495324960221e-05, + "loss": 0.9141, + "step": 20238 + }, + { + "epoch": 0.48, + "grad_norm": 1.9019525838597708, + "learning_rate": 1.123373811692102e-05, + "loss": 0.9712, + "step": 20239 + }, + { + "epoch": 0.48, + "grad_norm": 2.009239376840375, + "learning_rate": 1.123298090169861e-05, + "loss": 1.1468, + "step": 20240 + }, + { + "epoch": 0.48, + "grad_norm": 2.2625892764650923, + "learning_rate": 1.1232223679297391e-05, + "loss": 0.9582, + "step": 20241 + }, + { + "epoch": 0.48, + "grad_norm": 1.998163882192102, + "learning_rate": 1.1231466449721777e-05, + "loss": 1.0169, + "step": 20242 + }, + { + "epoch": 0.48, + "grad_norm": 1.9761040620268207, + "learning_rate": 1.1230709212976176e-05, + "loss": 0.9598, + "step": 20243 + }, + { + "epoch": 0.48, + "grad_norm": 1.1056660578860829, + "learning_rate": 1.1229951969064996e-05, + "loss": 0.8841, + "step": 20244 + }, + { + "epoch": 0.48, + "grad_norm": 2.459692022894581, + "learning_rate": 1.1229194717992644e-05, + "loss": 1.1327, + "step": 20245 + }, + { + "epoch": 0.48, + "grad_norm": 2.3611674774746683, + "learning_rate": 1.1228437459763535e-05, + "loss": 0.9862, + "step": 20246 + }, + { + "epoch": 0.48, + "grad_norm": 1.858592372939276, + "learning_rate": 1.1227680194382072e-05, + "loss": 1.0365, + "step": 20247 + }, + { + "epoch": 0.48, + "grad_norm": 1.9153669921266427, + "learning_rate": 1.1226922921852668e-05, + "loss": 0.9865, + "step": 20248 + }, + { + "epoch": 0.48, + "grad_norm": 2.0111492767578816, + "learning_rate": 1.1226165642179727e-05, + "loss": 1.0676, + "step": 20249 + }, + { + "epoch": 0.48, + "grad_norm": 1.9812284197434902, + "learning_rate": 1.1225408355367665e-05, + "loss": 1.0804, + "step": 20250 + }, + { + "epoch": 0.48, + "grad_norm": 1.957569349929579, + "learning_rate": 1.1224651061420884e-05, + "loss": 1.0514, + "step": 20251 + }, + { + "epoch": 0.48, + "grad_norm": 2.114094639135241, + "learning_rate": 1.1223893760343798e-05, + "loss": 0.992, + "step": 20252 + }, + { + "epoch": 0.48, + "grad_norm": 3.452035069121647, + "learning_rate": 1.1223136452140815e-05, + "loss": 0.9824, + "step": 20253 + }, + { + "epoch": 0.48, + "grad_norm": 1.0620265389647057, + "learning_rate": 1.1222379136816347e-05, + "loss": 0.9576, + "step": 20254 + }, + { + "epoch": 0.48, + "grad_norm": 2.131750568615038, + "learning_rate": 1.12216218143748e-05, + "loss": 0.9381, + "step": 20255 + }, + { + "epoch": 0.48, + "grad_norm": 1.9205581134357652, + "learning_rate": 1.122086448482058e-05, + "loss": 0.9971, + "step": 20256 + }, + { + "epoch": 0.48, + "grad_norm": 1.9061515423096682, + "learning_rate": 1.1220107148158104e-05, + "loss": 0.9298, + "step": 20257 + }, + { + "epoch": 0.48, + "grad_norm": 2.142609723487528, + "learning_rate": 1.1219349804391774e-05, + "loss": 0.9103, + "step": 20258 + }, + { + "epoch": 0.48, + "grad_norm": 2.450554317402013, + "learning_rate": 1.121859245352601e-05, + "loss": 1.1835, + "step": 20259 + }, + { + "epoch": 0.48, + "grad_norm": 1.985098248822861, + "learning_rate": 1.1217835095565208e-05, + "loss": 1.0659, + "step": 20260 + }, + { + "epoch": 0.48, + "grad_norm": 1.9040885299025705, + "learning_rate": 1.1217077730513792e-05, + "loss": 1.0939, + "step": 20261 + }, + { + "epoch": 0.48, + "grad_norm": 2.338903690642606, + "learning_rate": 1.1216320358376158e-05, + "loss": 0.9442, + "step": 20262 + }, + { + "epoch": 0.48, + "grad_norm": 1.1114227855053085, + "learning_rate": 1.1215562979156725e-05, + "loss": 0.948, + "step": 20263 + }, + { + "epoch": 0.48, + "grad_norm": 1.1242154915969849, + "learning_rate": 1.1214805592859898e-05, + "loss": 0.9296, + "step": 20264 + }, + { + "epoch": 0.48, + "grad_norm": 2.234430414334629, + "learning_rate": 1.1214048199490089e-05, + "loss": 1.1371, + "step": 20265 + }, + { + "epoch": 0.48, + "grad_norm": 1.9246596288236621, + "learning_rate": 1.1213290799051706e-05, + "loss": 1.1532, + "step": 20266 + }, + { + "epoch": 0.48, + "grad_norm": 1.9080666633699226, + "learning_rate": 1.121253339154916e-05, + "loss": 0.921, + "step": 20267 + }, + { + "epoch": 0.48, + "grad_norm": 1.091313459078293, + "learning_rate": 1.1211775976986861e-05, + "loss": 0.9378, + "step": 20268 + }, + { + "epoch": 0.48, + "grad_norm": 1.8479757961361174, + "learning_rate": 1.121101855536922e-05, + "loss": 0.9706, + "step": 20269 + }, + { + "epoch": 0.48, + "grad_norm": 1.7373575723317187, + "learning_rate": 1.1210261126700643e-05, + "loss": 0.9548, + "step": 20270 + }, + { + "epoch": 0.48, + "grad_norm": 1.8941669725920922, + "learning_rate": 1.1209503690985544e-05, + "loss": 1.1106, + "step": 20271 + }, + { + "epoch": 0.48, + "grad_norm": 1.7222658234460082, + "learning_rate": 1.1208746248228332e-05, + "loss": 0.8546, + "step": 20272 + }, + { + "epoch": 0.48, + "grad_norm": 2.362049882258507, + "learning_rate": 1.1207988798433416e-05, + "loss": 1.0133, + "step": 20273 + }, + { + "epoch": 0.48, + "grad_norm": 1.9819103353789258, + "learning_rate": 1.1207231341605208e-05, + "loss": 1.1745, + "step": 20274 + }, + { + "epoch": 0.48, + "grad_norm": 2.666744925364339, + "learning_rate": 1.1206473877748115e-05, + "loss": 1.0075, + "step": 20275 + }, + { + "epoch": 0.48, + "grad_norm": 1.9821050285013389, + "learning_rate": 1.120571640686655e-05, + "loss": 1.0473, + "step": 20276 + }, + { + "epoch": 0.48, + "grad_norm": 1.8264262433422849, + "learning_rate": 1.1204958928964922e-05, + "loss": 1.0608, + "step": 20277 + }, + { + "epoch": 0.48, + "grad_norm": 2.3003519283371245, + "learning_rate": 1.1204201444047644e-05, + "loss": 1.0128, + "step": 20278 + }, + { + "epoch": 0.48, + "grad_norm": 2.2775001905634347, + "learning_rate": 1.120344395211912e-05, + "loss": 1.0359, + "step": 20279 + }, + { + "epoch": 0.48, + "grad_norm": 1.8434825708067346, + "learning_rate": 1.1202686453183769e-05, + "loss": 1.0065, + "step": 20280 + }, + { + "epoch": 0.48, + "grad_norm": 2.1028749458551728, + "learning_rate": 1.120192894724599e-05, + "loss": 0.9873, + "step": 20281 + }, + { + "epoch": 0.48, + "grad_norm": 1.8993119089144048, + "learning_rate": 1.1201171434310207e-05, + "loss": 0.9161, + "step": 20282 + }, + { + "epoch": 0.48, + "grad_norm": 1.968956922970469, + "learning_rate": 1.1200413914380819e-05, + "loss": 1.0513, + "step": 20283 + }, + { + "epoch": 0.48, + "grad_norm": 2.3453371380243127, + "learning_rate": 1.1199656387462244e-05, + "loss": 0.9348, + "step": 20284 + }, + { + "epoch": 0.48, + "grad_norm": 1.8078741686182744, + "learning_rate": 1.1198898853558889e-05, + "loss": 1.0707, + "step": 20285 + }, + { + "epoch": 0.48, + "grad_norm": 2.4892288393764144, + "learning_rate": 1.1198141312675164e-05, + "loss": 1.0467, + "step": 20286 + }, + { + "epoch": 0.48, + "grad_norm": 1.9670208470320354, + "learning_rate": 1.1197383764815482e-05, + "loss": 0.8881, + "step": 20287 + }, + { + "epoch": 0.48, + "grad_norm": 2.041703456896706, + "learning_rate": 1.119662620998425e-05, + "loss": 1.0019, + "step": 20288 + }, + { + "epoch": 0.48, + "grad_norm": 2.051269938621671, + "learning_rate": 1.1195868648185886e-05, + "loss": 1.0068, + "step": 20289 + }, + { + "epoch": 0.48, + "grad_norm": 2.140384798186629, + "learning_rate": 1.1195111079424791e-05, + "loss": 1.0752, + "step": 20290 + }, + { + "epoch": 0.48, + "grad_norm": 2.5032887075730805, + "learning_rate": 1.1194353503705387e-05, + "loss": 1.0186, + "step": 20291 + }, + { + "epoch": 0.48, + "grad_norm": 2.066310817999548, + "learning_rate": 1.1193595921032074e-05, + "loss": 0.9973, + "step": 20292 + }, + { + "epoch": 0.48, + "grad_norm": 2.1277715140356763, + "learning_rate": 1.119283833140927e-05, + "loss": 0.9673, + "step": 20293 + }, + { + "epoch": 0.48, + "grad_norm": 1.9931111645302455, + "learning_rate": 1.1192080734841385e-05, + "loss": 0.8864, + "step": 20294 + }, + { + "epoch": 0.48, + "grad_norm": 1.03477338062201, + "learning_rate": 1.1191323131332825e-05, + "loss": 0.9018, + "step": 20295 + }, + { + "epoch": 0.48, + "grad_norm": 1.0764769895713302, + "learning_rate": 1.1190565520888005e-05, + "loss": 0.9409, + "step": 20296 + }, + { + "epoch": 0.48, + "grad_norm": 1.9389736080596016, + "learning_rate": 1.1189807903511336e-05, + "loss": 1.0466, + "step": 20297 + }, + { + "epoch": 0.48, + "grad_norm": 2.1323138601009126, + "learning_rate": 1.1189050279207229e-05, + "loss": 1.1474, + "step": 20298 + }, + { + "epoch": 0.48, + "grad_norm": 1.118873207654227, + "learning_rate": 1.1188292647980099e-05, + "loss": 0.8931, + "step": 20299 + }, + { + "epoch": 0.48, + "grad_norm": 2.5372520534035425, + "learning_rate": 1.1187535009834347e-05, + "loss": 1.0069, + "step": 20300 + }, + { + "epoch": 0.48, + "grad_norm": 2.0380146669772445, + "learning_rate": 1.1186777364774393e-05, + "loss": 1.1572, + "step": 20301 + }, + { + "epoch": 0.48, + "grad_norm": 1.0816064564087426, + "learning_rate": 1.1186019712804644e-05, + "loss": 0.8955, + "step": 20302 + }, + { + "epoch": 0.48, + "grad_norm": 2.669707683287102, + "learning_rate": 1.1185262053929513e-05, + "loss": 1.0593, + "step": 20303 + }, + { + "epoch": 0.48, + "grad_norm": 1.8773724672287608, + "learning_rate": 1.1184504388153413e-05, + "loss": 0.9716, + "step": 20304 + }, + { + "epoch": 0.48, + "grad_norm": 2.5013212128398297, + "learning_rate": 1.118374671548075e-05, + "loss": 1.0333, + "step": 20305 + }, + { + "epoch": 0.48, + "grad_norm": 1.8200000281910318, + "learning_rate": 1.1182989035915943e-05, + "loss": 0.9693, + "step": 20306 + }, + { + "epoch": 0.48, + "grad_norm": 1.8429349284981997, + "learning_rate": 1.1182231349463394e-05, + "loss": 1.0146, + "step": 20307 + }, + { + "epoch": 0.48, + "grad_norm": 1.1597449562216735, + "learning_rate": 1.1181473656127526e-05, + "loss": 0.9688, + "step": 20308 + }, + { + "epoch": 0.48, + "grad_norm": 2.217410167961994, + "learning_rate": 1.118071595591274e-05, + "loss": 1.0347, + "step": 20309 + }, + { + "epoch": 0.48, + "grad_norm": 2.0319136660987422, + "learning_rate": 1.1179958248823455e-05, + "loss": 0.9875, + "step": 20310 + }, + { + "epoch": 0.48, + "grad_norm": 2.1174585658916634, + "learning_rate": 1.1179200534864077e-05, + "loss": 1.0413, + "step": 20311 + }, + { + "epoch": 0.48, + "grad_norm": 2.0497100301852655, + "learning_rate": 1.1178442814039021e-05, + "loss": 0.9748, + "step": 20312 + }, + { + "epoch": 0.48, + "grad_norm": 2.1594618936809677, + "learning_rate": 1.1177685086352699e-05, + "loss": 0.9249, + "step": 20313 + }, + { + "epoch": 0.48, + "grad_norm": 1.9198070313628193, + "learning_rate": 1.1176927351809519e-05, + "loss": 1.0138, + "step": 20314 + }, + { + "epoch": 0.48, + "grad_norm": 2.1457589788796048, + "learning_rate": 1.1176169610413895e-05, + "loss": 1.0939, + "step": 20315 + }, + { + "epoch": 0.48, + "grad_norm": 2.2711273339775087, + "learning_rate": 1.1175411862170242e-05, + "loss": 0.8384, + "step": 20316 + }, + { + "epoch": 0.48, + "grad_norm": 1.910233570164389, + "learning_rate": 1.117465410708297e-05, + "loss": 1.033, + "step": 20317 + }, + { + "epoch": 0.48, + "grad_norm": 2.0629534958835465, + "learning_rate": 1.1173896345156486e-05, + "loss": 1.0404, + "step": 20318 + }, + { + "epoch": 0.48, + "grad_norm": 1.9974059607481358, + "learning_rate": 1.1173138576395208e-05, + "loss": 1.008, + "step": 20319 + }, + { + "epoch": 0.48, + "grad_norm": 1.823770141042392, + "learning_rate": 1.1172380800803544e-05, + "loss": 1.0938, + "step": 20320 + }, + { + "epoch": 0.48, + "grad_norm": 1.9496669120897872, + "learning_rate": 1.117162301838591e-05, + "loss": 0.9663, + "step": 20321 + }, + { + "epoch": 0.48, + "grad_norm": 2.183155981335659, + "learning_rate": 1.1170865229146715e-05, + "loss": 0.9887, + "step": 20322 + }, + { + "epoch": 0.48, + "grad_norm": 2.385933041404463, + "learning_rate": 1.1170107433090371e-05, + "loss": 0.9532, + "step": 20323 + }, + { + "epoch": 0.48, + "grad_norm": 2.032017708434985, + "learning_rate": 1.1169349630221294e-05, + "loss": 1.051, + "step": 20324 + }, + { + "epoch": 0.48, + "grad_norm": 2.2818893080186125, + "learning_rate": 1.116859182054389e-05, + "loss": 1.1131, + "step": 20325 + }, + { + "epoch": 0.48, + "grad_norm": 2.0025685878550075, + "learning_rate": 1.1167834004062576e-05, + "loss": 1.0982, + "step": 20326 + }, + { + "epoch": 0.48, + "grad_norm": 1.816175096939484, + "learning_rate": 1.1167076180781764e-05, + "loss": 0.9703, + "step": 20327 + }, + { + "epoch": 0.48, + "grad_norm": 1.9017005110502991, + "learning_rate": 1.116631835070586e-05, + "loss": 1.061, + "step": 20328 + }, + { + "epoch": 0.48, + "grad_norm": 1.8806957039224932, + "learning_rate": 1.1165560513839289e-05, + "loss": 0.9513, + "step": 20329 + }, + { + "epoch": 0.48, + "grad_norm": 1.0944463491824035, + "learning_rate": 1.1164802670186448e-05, + "loss": 0.9066, + "step": 20330 + }, + { + "epoch": 0.48, + "grad_norm": 2.030873256829612, + "learning_rate": 1.1164044819751763e-05, + "loss": 1.0419, + "step": 20331 + }, + { + "epoch": 0.48, + "grad_norm": 2.092260040023398, + "learning_rate": 1.1163286962539637e-05, + "loss": 0.9157, + "step": 20332 + }, + { + "epoch": 0.48, + "grad_norm": 1.7813038152535643, + "learning_rate": 1.1162529098554489e-05, + "loss": 1.12, + "step": 20333 + }, + { + "epoch": 0.48, + "grad_norm": 2.102653306135385, + "learning_rate": 1.1161771227800727e-05, + "loss": 1.0336, + "step": 20334 + }, + { + "epoch": 0.48, + "grad_norm": 2.2137135263886156, + "learning_rate": 1.1161013350282763e-05, + "loss": 1.0783, + "step": 20335 + }, + { + "epoch": 0.48, + "grad_norm": 2.1225086551378936, + "learning_rate": 1.1160255466005014e-05, + "loss": 1.0326, + "step": 20336 + }, + { + "epoch": 0.48, + "grad_norm": 2.8935790615329777, + "learning_rate": 1.1159497574971888e-05, + "loss": 0.9776, + "step": 20337 + }, + { + "epoch": 0.48, + "grad_norm": 2.2119279165234076, + "learning_rate": 1.1158739677187802e-05, + "loss": 1.1423, + "step": 20338 + }, + { + "epoch": 0.48, + "grad_norm": 1.15427276784635, + "learning_rate": 1.1157981772657165e-05, + "loss": 0.9379, + "step": 20339 + }, + { + "epoch": 0.48, + "grad_norm": 2.1148606855362067, + "learning_rate": 1.1157223861384394e-05, + "loss": 1.0269, + "step": 20340 + }, + { + "epoch": 0.48, + "grad_norm": 2.4596917288854057, + "learning_rate": 1.1156465943373897e-05, + "loss": 1.126, + "step": 20341 + }, + { + "epoch": 0.48, + "grad_norm": 3.039691866469039, + "learning_rate": 1.1155708018630094e-05, + "loss": 0.9585, + "step": 20342 + }, + { + "epoch": 0.48, + "grad_norm": 1.9759517647760587, + "learning_rate": 1.1154950087157389e-05, + "loss": 0.9538, + "step": 20343 + }, + { + "epoch": 0.48, + "grad_norm": 1.7551893707884392, + "learning_rate": 1.1154192148960199e-05, + "loss": 1.0, + "step": 20344 + }, + { + "epoch": 0.48, + "grad_norm": 2.1148602966338403, + "learning_rate": 1.1153434204042936e-05, + "loss": 1.0068, + "step": 20345 + }, + { + "epoch": 0.48, + "grad_norm": 2.114614230237152, + "learning_rate": 1.1152676252410015e-05, + "loss": 0.9278, + "step": 20346 + }, + { + "epoch": 0.48, + "grad_norm": 2.112690220461952, + "learning_rate": 1.115191829406585e-05, + "loss": 1.0564, + "step": 20347 + }, + { + "epoch": 0.48, + "grad_norm": 1.1464901924798034, + "learning_rate": 1.1151160329014849e-05, + "loss": 0.993, + "step": 20348 + }, + { + "epoch": 0.48, + "grad_norm": 1.1077249560482678, + "learning_rate": 1.1150402357261429e-05, + "loss": 0.9334, + "step": 20349 + }, + { + "epoch": 0.48, + "grad_norm": 2.215772387535644, + "learning_rate": 1.1149644378810003e-05, + "loss": 0.9565, + "step": 20350 + }, + { + "epoch": 0.48, + "grad_norm": 1.0994040420717346, + "learning_rate": 1.1148886393664984e-05, + "loss": 0.9528, + "step": 20351 + }, + { + "epoch": 0.48, + "grad_norm": 2.0094090273732426, + "learning_rate": 1.1148128401830783e-05, + "loss": 1.0964, + "step": 20352 + }, + { + "epoch": 0.48, + "grad_norm": 2.2713925032679567, + "learning_rate": 1.1147370403311815e-05, + "loss": 1.1412, + "step": 20353 + }, + { + "epoch": 0.48, + "grad_norm": 1.7708538966995155, + "learning_rate": 1.1146612398112492e-05, + "loss": 1.0077, + "step": 20354 + }, + { + "epoch": 0.48, + "grad_norm": 2.633759189552283, + "learning_rate": 1.1145854386237231e-05, + "loss": 0.9587, + "step": 20355 + }, + { + "epoch": 0.48, + "grad_norm": 17.455052480084415, + "learning_rate": 1.1145096367690445e-05, + "loss": 0.9916, + "step": 20356 + }, + { + "epoch": 0.48, + "grad_norm": 2.0871284281719333, + "learning_rate": 1.1144338342476544e-05, + "loss": 0.9017, + "step": 20357 + }, + { + "epoch": 0.48, + "grad_norm": 2.0086867743265366, + "learning_rate": 1.114358031059994e-05, + "loss": 0.9464, + "step": 20358 + }, + { + "epoch": 0.48, + "grad_norm": 1.9893724777052026, + "learning_rate": 1.1142822272065053e-05, + "loss": 1.0341, + "step": 20359 + }, + { + "epoch": 0.48, + "grad_norm": 2.0857463361867814, + "learning_rate": 1.1142064226876289e-05, + "loss": 1.0362, + "step": 20360 + }, + { + "epoch": 0.48, + "grad_norm": 2.804846866836344, + "learning_rate": 1.1141306175038069e-05, + "loss": 0.9296, + "step": 20361 + }, + { + "epoch": 0.48, + "grad_norm": 1.9594887899860567, + "learning_rate": 1.1140548116554799e-05, + "loss": 1.126, + "step": 20362 + }, + { + "epoch": 0.48, + "grad_norm": 1.123997326633786, + "learning_rate": 1.1139790051430901e-05, + "loss": 0.9208, + "step": 20363 + }, + { + "epoch": 0.48, + "grad_norm": 2.1179272581737445, + "learning_rate": 1.1139031979670785e-05, + "loss": 1.0367, + "step": 20364 + }, + { + "epoch": 0.48, + "grad_norm": 1.891176545943749, + "learning_rate": 1.1138273901278863e-05, + "loss": 1.0962, + "step": 20365 + }, + { + "epoch": 0.48, + "grad_norm": 1.8855905558302024, + "learning_rate": 1.1137515816259548e-05, + "loss": 1.0032, + "step": 20366 + }, + { + "epoch": 0.48, + "grad_norm": 2.165614523740161, + "learning_rate": 1.1136757724617256e-05, + "loss": 0.9764, + "step": 20367 + }, + { + "epoch": 0.48, + "grad_norm": 2.2947270649545697, + "learning_rate": 1.1135999626356403e-05, + "loss": 1.0111, + "step": 20368 + }, + { + "epoch": 0.48, + "grad_norm": 2.8704642747602542, + "learning_rate": 1.1135241521481397e-05, + "loss": 0.9566, + "step": 20369 + }, + { + "epoch": 0.48, + "grad_norm": 1.845214199975561, + "learning_rate": 1.1134483409996657e-05, + "loss": 1.0305, + "step": 20370 + }, + { + "epoch": 0.48, + "grad_norm": 2.2064228455678703, + "learning_rate": 1.1133725291906596e-05, + "loss": 0.9843, + "step": 20371 + }, + { + "epoch": 0.48, + "grad_norm": 1.0760737121661155, + "learning_rate": 1.1132967167215627e-05, + "loss": 0.9341, + "step": 20372 + }, + { + "epoch": 0.48, + "grad_norm": 1.8419448144236168, + "learning_rate": 1.1132209035928166e-05, + "loss": 0.9675, + "step": 20373 + }, + { + "epoch": 0.48, + "grad_norm": 2.725266764791259, + "learning_rate": 1.1131450898048624e-05, + "loss": 0.9935, + "step": 20374 + }, + { + "epoch": 0.48, + "grad_norm": 2.1917290194866332, + "learning_rate": 1.1130692753581416e-05, + "loss": 0.9618, + "step": 20375 + }, + { + "epoch": 0.48, + "grad_norm": 1.942138842829056, + "learning_rate": 1.1129934602530959e-05, + "loss": 1.0133, + "step": 20376 + }, + { + "epoch": 0.48, + "grad_norm": 1.0371496458550487, + "learning_rate": 1.1129176444901663e-05, + "loss": 0.865, + "step": 20377 + }, + { + "epoch": 0.48, + "grad_norm": 1.9412822139288433, + "learning_rate": 1.1128418280697948e-05, + "loss": 0.9648, + "step": 20378 + }, + { + "epoch": 0.48, + "grad_norm": 1.8444572180030165, + "learning_rate": 1.1127660109924218e-05, + "loss": 0.9814, + "step": 20379 + }, + { + "epoch": 0.48, + "grad_norm": 1.0497426961737748, + "learning_rate": 1.11269019325849e-05, + "loss": 0.9486, + "step": 20380 + }, + { + "epoch": 0.48, + "grad_norm": 2.0300470718986503, + "learning_rate": 1.1126143748684397e-05, + "loss": 0.9947, + "step": 20381 + }, + { + "epoch": 0.48, + "grad_norm": 1.904193389293716, + "learning_rate": 1.1125385558227136e-05, + "loss": 1.0384, + "step": 20382 + }, + { + "epoch": 0.48, + "grad_norm": 1.9566575492331426, + "learning_rate": 1.1124627361217514e-05, + "loss": 0.9364, + "step": 20383 + }, + { + "epoch": 0.48, + "grad_norm": 1.9248119802399721, + "learning_rate": 1.1123869157659962e-05, + "loss": 0.9307, + "step": 20384 + }, + { + "epoch": 0.48, + "grad_norm": 2.314148994172907, + "learning_rate": 1.1123110947558888e-05, + "loss": 0.9548, + "step": 20385 + }, + { + "epoch": 0.48, + "grad_norm": 2.0488711234190893, + "learning_rate": 1.1122352730918704e-05, + "loss": 1.0971, + "step": 20386 + }, + { + "epoch": 0.48, + "grad_norm": 1.9870276504045143, + "learning_rate": 1.112159450774383e-05, + "loss": 0.8531, + "step": 20387 + }, + { + "epoch": 0.48, + "grad_norm": 2.0269681914819557, + "learning_rate": 1.1120836278038674e-05, + "loss": 1.105, + "step": 20388 + }, + { + "epoch": 0.48, + "grad_norm": 2.3696838373914955, + "learning_rate": 1.1120078041807658e-05, + "loss": 1.029, + "step": 20389 + }, + { + "epoch": 0.48, + "grad_norm": 1.9065331871981714, + "learning_rate": 1.1119319799055188e-05, + "loss": 0.8155, + "step": 20390 + }, + { + "epoch": 0.48, + "grad_norm": 1.8752127883042375, + "learning_rate": 1.111856154978569e-05, + "loss": 1.0855, + "step": 20391 + }, + { + "epoch": 0.48, + "grad_norm": 2.1304602774812493, + "learning_rate": 1.1117803294003568e-05, + "loss": 1.0638, + "step": 20392 + }, + { + "epoch": 0.48, + "grad_norm": 2.0434126536929695, + "learning_rate": 1.1117045031713242e-05, + "loss": 0.9943, + "step": 20393 + }, + { + "epoch": 0.48, + "grad_norm": 2.406075096924404, + "learning_rate": 1.1116286762919129e-05, + "loss": 0.9433, + "step": 20394 + }, + { + "epoch": 0.48, + "grad_norm": 1.9042725885505365, + "learning_rate": 1.1115528487625638e-05, + "loss": 1.0755, + "step": 20395 + }, + { + "epoch": 0.48, + "grad_norm": 2.014861617964983, + "learning_rate": 1.1114770205837189e-05, + "loss": 1.1369, + "step": 20396 + }, + { + "epoch": 0.48, + "grad_norm": 1.0514195760798537, + "learning_rate": 1.1114011917558194e-05, + "loss": 0.9264, + "step": 20397 + }, + { + "epoch": 0.48, + "grad_norm": 2.138084083936854, + "learning_rate": 1.1113253622793067e-05, + "loss": 1.0892, + "step": 20398 + }, + { + "epoch": 0.48, + "grad_norm": 2.103810588093963, + "learning_rate": 1.1112495321546228e-05, + "loss": 1.0362, + "step": 20399 + }, + { + "epoch": 0.48, + "grad_norm": 2.216679841577881, + "learning_rate": 1.1111737013822089e-05, + "loss": 0.9498, + "step": 20400 + }, + { + "epoch": 0.48, + "grad_norm": 1.8203018229349472, + "learning_rate": 1.1110978699625061e-05, + "loss": 1.0725, + "step": 20401 + }, + { + "epoch": 0.48, + "grad_norm": 2.0503776876989295, + "learning_rate": 1.1110220378959568e-05, + "loss": 1.0714, + "step": 20402 + }, + { + "epoch": 0.48, + "grad_norm": 2.3774312507426596, + "learning_rate": 1.1109462051830018e-05, + "loss": 1.0223, + "step": 20403 + }, + { + "epoch": 0.48, + "grad_norm": 2.1573964437810664, + "learning_rate": 1.110870371824083e-05, + "loss": 1.0499, + "step": 20404 + }, + { + "epoch": 0.48, + "grad_norm": 1.9567435430482736, + "learning_rate": 1.1107945378196417e-05, + "loss": 1.1013, + "step": 20405 + }, + { + "epoch": 0.48, + "grad_norm": 1.9773417372035795, + "learning_rate": 1.1107187031701194e-05, + "loss": 1.1688, + "step": 20406 + }, + { + "epoch": 0.48, + "grad_norm": 2.372293340681883, + "learning_rate": 1.110642867875958e-05, + "loss": 0.9331, + "step": 20407 + }, + { + "epoch": 0.48, + "grad_norm": 1.9327614129499493, + "learning_rate": 1.110567031937599e-05, + "loss": 0.8799, + "step": 20408 + }, + { + "epoch": 0.48, + "grad_norm": 1.986934182474758, + "learning_rate": 1.1104911953554831e-05, + "loss": 1.0463, + "step": 20409 + }, + { + "epoch": 0.48, + "grad_norm": 2.000409125530422, + "learning_rate": 1.1104153581300532e-05, + "loss": 1.0146, + "step": 20410 + }, + { + "epoch": 0.48, + "grad_norm": 2.216590859066052, + "learning_rate": 1.1103395202617495e-05, + "loss": 1.0152, + "step": 20411 + }, + { + "epoch": 0.48, + "grad_norm": 2.4024014434050778, + "learning_rate": 1.1102636817510147e-05, + "loss": 1.0596, + "step": 20412 + }, + { + "epoch": 0.48, + "grad_norm": 1.9345673999639286, + "learning_rate": 1.1101878425982892e-05, + "loss": 0.981, + "step": 20413 + }, + { + "epoch": 0.48, + "grad_norm": 1.9944304799446575, + "learning_rate": 1.1101120028040158e-05, + "loss": 1.0857, + "step": 20414 + }, + { + "epoch": 0.48, + "grad_norm": 1.9310572251140874, + "learning_rate": 1.1100361623686352e-05, + "loss": 0.918, + "step": 20415 + }, + { + "epoch": 0.48, + "grad_norm": 2.2374605726452454, + "learning_rate": 1.1099603212925893e-05, + "loss": 0.9461, + "step": 20416 + }, + { + "epoch": 0.48, + "grad_norm": 2.1121455592048557, + "learning_rate": 1.1098844795763195e-05, + "loss": 1.1802, + "step": 20417 + }, + { + "epoch": 0.48, + "grad_norm": 2.115193637172062, + "learning_rate": 1.1098086372202674e-05, + "loss": 0.9938, + "step": 20418 + }, + { + "epoch": 0.48, + "grad_norm": 2.2773021638788173, + "learning_rate": 1.109732794224875e-05, + "loss": 1.0848, + "step": 20419 + }, + { + "epoch": 0.48, + "grad_norm": 1.968868309477183, + "learning_rate": 1.1096569505905832e-05, + "loss": 0.9003, + "step": 20420 + }, + { + "epoch": 0.48, + "grad_norm": 2.0602053063689763, + "learning_rate": 1.109581106317834e-05, + "loss": 0.9816, + "step": 20421 + }, + { + "epoch": 0.48, + "grad_norm": 2.0506956809220975, + "learning_rate": 1.109505261407069e-05, + "loss": 1.0292, + "step": 20422 + }, + { + "epoch": 0.48, + "grad_norm": 1.8013635512213644, + "learning_rate": 1.1094294158587295e-05, + "loss": 1.004, + "step": 20423 + }, + { + "epoch": 0.48, + "grad_norm": 2.011475056406169, + "learning_rate": 1.1093535696732574e-05, + "loss": 1.2011, + "step": 20424 + }, + { + "epoch": 0.48, + "grad_norm": 1.7583028770281741, + "learning_rate": 1.1092777228510943e-05, + "loss": 1.0817, + "step": 20425 + }, + { + "epoch": 0.48, + "grad_norm": 1.9919857217478234, + "learning_rate": 1.1092018753926817e-05, + "loss": 0.9962, + "step": 20426 + }, + { + "epoch": 0.48, + "grad_norm": 1.9978250235179276, + "learning_rate": 1.109126027298461e-05, + "loss": 0.9798, + "step": 20427 + }, + { + "epoch": 0.48, + "grad_norm": 2.0585055594762993, + "learning_rate": 1.109050178568874e-05, + "loss": 1.0177, + "step": 20428 + }, + { + "epoch": 0.48, + "grad_norm": 2.3643829002599426, + "learning_rate": 1.1089743292043629e-05, + "loss": 0.8835, + "step": 20429 + }, + { + "epoch": 0.48, + "grad_norm": 2.0900703731774612, + "learning_rate": 1.1088984792053681e-05, + "loss": 0.9836, + "step": 20430 + }, + { + "epoch": 0.48, + "grad_norm": 1.077713109373841, + "learning_rate": 1.1088226285723323e-05, + "loss": 1.0421, + "step": 20431 + }, + { + "epoch": 0.48, + "grad_norm": 1.870014238947689, + "learning_rate": 1.1087467773056962e-05, + "loss": 1.0708, + "step": 20432 + }, + { + "epoch": 0.48, + "grad_norm": 1.113765786681848, + "learning_rate": 1.1086709254059025e-05, + "loss": 0.9993, + "step": 20433 + }, + { + "epoch": 0.48, + "grad_norm": 1.9790252801572468, + "learning_rate": 1.1085950728733919e-05, + "loss": 1.0197, + "step": 20434 + }, + { + "epoch": 0.48, + "grad_norm": 1.8544373833109962, + "learning_rate": 1.1085192197086063e-05, + "loss": 1.1213, + "step": 20435 + }, + { + "epoch": 0.48, + "grad_norm": 1.870326231062152, + "learning_rate": 1.108443365911988e-05, + "loss": 0.9788, + "step": 20436 + }, + { + "epoch": 0.48, + "grad_norm": 2.028846917266709, + "learning_rate": 1.1083675114839775e-05, + "loss": 1.0322, + "step": 20437 + }, + { + "epoch": 0.48, + "grad_norm": 1.820090063875219, + "learning_rate": 1.1082916564250175e-05, + "loss": 1.1116, + "step": 20438 + }, + { + "epoch": 0.48, + "grad_norm": 1.8822605801374486, + "learning_rate": 1.1082158007355487e-05, + "loss": 1.0133, + "step": 20439 + }, + { + "epoch": 0.48, + "grad_norm": 1.9848995590179166, + "learning_rate": 1.108139944416014e-05, + "loss": 0.8254, + "step": 20440 + }, + { + "epoch": 0.48, + "grad_norm": 1.1266690473411527, + "learning_rate": 1.1080640874668535e-05, + "loss": 1.0086, + "step": 20441 + }, + { + "epoch": 0.48, + "grad_norm": 2.124975056515008, + "learning_rate": 1.1079882298885102e-05, + "loss": 1.0113, + "step": 20442 + }, + { + "epoch": 0.48, + "grad_norm": 2.1240405895373713, + "learning_rate": 1.107912371681425e-05, + "loss": 1.0557, + "step": 20443 + }, + { + "epoch": 0.48, + "grad_norm": 2.4480152981710366, + "learning_rate": 1.1078365128460398e-05, + "loss": 0.9593, + "step": 20444 + }, + { + "epoch": 0.48, + "grad_norm": 2.258268481564867, + "learning_rate": 1.1077606533827965e-05, + "loss": 1.1116, + "step": 20445 + }, + { + "epoch": 0.48, + "grad_norm": 1.8745289385909645, + "learning_rate": 1.1076847932921364e-05, + "loss": 1.0409, + "step": 20446 + }, + { + "epoch": 0.48, + "grad_norm": 1.9017621726901566, + "learning_rate": 1.1076089325745013e-05, + "loss": 1.1139, + "step": 20447 + }, + { + "epoch": 0.48, + "grad_norm": 2.2531548219326196, + "learning_rate": 1.1075330712303329e-05, + "loss": 1.0084, + "step": 20448 + }, + { + "epoch": 0.48, + "grad_norm": 1.9721834731572774, + "learning_rate": 1.107457209260073e-05, + "loss": 0.8605, + "step": 20449 + }, + { + "epoch": 0.48, + "grad_norm": 1.9875035412578028, + "learning_rate": 1.1073813466641633e-05, + "loss": 0.9767, + "step": 20450 + }, + { + "epoch": 0.48, + "grad_norm": 2.150924509368229, + "learning_rate": 1.1073054834430452e-05, + "loss": 1.105, + "step": 20451 + }, + { + "epoch": 0.48, + "grad_norm": 2.104814539138649, + "learning_rate": 1.1072296195971606e-05, + "loss": 1.0407, + "step": 20452 + }, + { + "epoch": 0.48, + "grad_norm": 1.9063837334749427, + "learning_rate": 1.1071537551269513e-05, + "loss": 0.9477, + "step": 20453 + }, + { + "epoch": 0.48, + "grad_norm": 2.309009750771209, + "learning_rate": 1.107077890032859e-05, + "loss": 1.0121, + "step": 20454 + }, + { + "epoch": 0.48, + "grad_norm": 2.300833089872705, + "learning_rate": 1.1070020243153253e-05, + "loss": 1.1049, + "step": 20455 + }, + { + "epoch": 0.48, + "grad_norm": 2.017962906973366, + "learning_rate": 1.1069261579747919e-05, + "loss": 1.0129, + "step": 20456 + }, + { + "epoch": 0.48, + "grad_norm": 1.1281875893488877, + "learning_rate": 1.1068502910117006e-05, + "loss": 0.9506, + "step": 20457 + }, + { + "epoch": 0.48, + "grad_norm": 2.472053607218308, + "learning_rate": 1.1067744234264928e-05, + "loss": 0.9129, + "step": 20458 + }, + { + "epoch": 0.48, + "grad_norm": 2.0889281945684863, + "learning_rate": 1.106698555219611e-05, + "loss": 1.0357, + "step": 20459 + }, + { + "epoch": 0.48, + "grad_norm": 1.9133936178573157, + "learning_rate": 1.106622686391496e-05, + "loss": 1.127, + "step": 20460 + }, + { + "epoch": 0.48, + "grad_norm": 1.932358333124374, + "learning_rate": 1.1065468169425902e-05, + "loss": 0.8647, + "step": 20461 + }, + { + "epoch": 0.48, + "grad_norm": 2.2260292474999086, + "learning_rate": 1.1064709468733348e-05, + "loss": 1.0534, + "step": 20462 + }, + { + "epoch": 0.48, + "grad_norm": 2.61770855125831, + "learning_rate": 1.1063950761841722e-05, + "loss": 1.0839, + "step": 20463 + }, + { + "epoch": 0.48, + "grad_norm": 2.442771616993231, + "learning_rate": 1.1063192048755436e-05, + "loss": 0.9424, + "step": 20464 + }, + { + "epoch": 0.48, + "grad_norm": 2.3216417284046447, + "learning_rate": 1.106243332947891e-05, + "loss": 0.9883, + "step": 20465 + }, + { + "epoch": 0.48, + "grad_norm": 2.017407538007835, + "learning_rate": 1.106167460401656e-05, + "loss": 1.0818, + "step": 20466 + }, + { + "epoch": 0.48, + "grad_norm": 1.9847172333826695, + "learning_rate": 1.1060915872372804e-05, + "loss": 1.1795, + "step": 20467 + }, + { + "epoch": 0.48, + "grad_norm": 1.9723565986457952, + "learning_rate": 1.1060157134552061e-05, + "loss": 0.924, + "step": 20468 + }, + { + "epoch": 0.48, + "grad_norm": 1.0845574483652967, + "learning_rate": 1.105939839055875e-05, + "loss": 0.9713, + "step": 20469 + }, + { + "epoch": 0.48, + "grad_norm": 2.0056967573029896, + "learning_rate": 1.1058639640397282e-05, + "loss": 0.9853, + "step": 20470 + }, + { + "epoch": 0.48, + "grad_norm": 2.061721779284704, + "learning_rate": 1.1057880884072078e-05, + "loss": 1.0451, + "step": 20471 + }, + { + "epoch": 0.48, + "grad_norm": 2.002494258143603, + "learning_rate": 1.1057122121587561e-05, + "loss": 0.967, + "step": 20472 + }, + { + "epoch": 0.48, + "grad_norm": 2.1068602907241405, + "learning_rate": 1.1056363352948141e-05, + "loss": 1.062, + "step": 20473 + }, + { + "epoch": 0.48, + "grad_norm": 1.1016372589139098, + "learning_rate": 1.105560457815824e-05, + "loss": 1.0067, + "step": 20474 + }, + { + "epoch": 0.48, + "grad_norm": 2.0701011671311433, + "learning_rate": 1.1054845797222276e-05, + "loss": 0.9381, + "step": 20475 + }, + { + "epoch": 0.48, + "grad_norm": 2.120640642307902, + "learning_rate": 1.1054087010144662e-05, + "loss": 1.0057, + "step": 20476 + }, + { + "epoch": 0.48, + "grad_norm": 2.1493134050401608, + "learning_rate": 1.1053328216929823e-05, + "loss": 0.9286, + "step": 20477 + }, + { + "epoch": 0.48, + "grad_norm": 2.249202015864373, + "learning_rate": 1.1052569417582174e-05, + "loss": 1.0254, + "step": 20478 + }, + { + "epoch": 0.48, + "grad_norm": 1.9779812172507272, + "learning_rate": 1.1051810612106134e-05, + "loss": 1.1187, + "step": 20479 + }, + { + "epoch": 0.48, + "grad_norm": 2.044522151550518, + "learning_rate": 1.1051051800506118e-05, + "loss": 0.9591, + "step": 20480 + }, + { + "epoch": 0.48, + "grad_norm": 2.1671787608628135, + "learning_rate": 1.1050292982786543e-05, + "loss": 0.9856, + "step": 20481 + }, + { + "epoch": 0.48, + "grad_norm": 2.0304251834771945, + "learning_rate": 1.1049534158951834e-05, + "loss": 1.1047, + "step": 20482 + }, + { + "epoch": 0.48, + "grad_norm": 2.1758606912682485, + "learning_rate": 1.10487753290064e-05, + "loss": 1.1066, + "step": 20483 + }, + { + "epoch": 0.48, + "grad_norm": 2.138844240549875, + "learning_rate": 1.1048016492954668e-05, + "loss": 1.0327, + "step": 20484 + }, + { + "epoch": 0.48, + "grad_norm": 1.8862647533139978, + "learning_rate": 1.104725765080105e-05, + "loss": 1.0122, + "step": 20485 + }, + { + "epoch": 0.48, + "grad_norm": 2.1475773329197163, + "learning_rate": 1.1046498802549968e-05, + "loss": 1.069, + "step": 20486 + }, + { + "epoch": 0.48, + "grad_norm": 1.8722739253207346, + "learning_rate": 1.104573994820584e-05, + "loss": 1.1257, + "step": 20487 + }, + { + "epoch": 0.48, + "grad_norm": 1.953855742010872, + "learning_rate": 1.104498108777308e-05, + "loss": 1.0354, + "step": 20488 + }, + { + "epoch": 0.48, + "grad_norm": 2.0815399101891665, + "learning_rate": 1.104422222125611e-05, + "loss": 1.0944, + "step": 20489 + }, + { + "epoch": 0.48, + "grad_norm": 2.1443520769242763, + "learning_rate": 1.1043463348659346e-05, + "loss": 0.9765, + "step": 20490 + }, + { + "epoch": 0.48, + "grad_norm": 1.762113840350762, + "learning_rate": 1.1042704469987213e-05, + "loss": 1.0247, + "step": 20491 + }, + { + "epoch": 0.48, + "grad_norm": 2.2405829003658844, + "learning_rate": 1.1041945585244118e-05, + "loss": 1.0028, + "step": 20492 + }, + { + "epoch": 0.48, + "grad_norm": 2.1185051903631407, + "learning_rate": 1.1041186694434492e-05, + "loss": 0.9946, + "step": 20493 + }, + { + "epoch": 0.48, + "grad_norm": 2.337278248142579, + "learning_rate": 1.1040427797562744e-05, + "loss": 0.9555, + "step": 20494 + }, + { + "epoch": 0.48, + "grad_norm": 2.0183061346123745, + "learning_rate": 1.1039668894633298e-05, + "loss": 0.9759, + "step": 20495 + }, + { + "epoch": 0.48, + "grad_norm": 2.0397400498785783, + "learning_rate": 1.1038909985650568e-05, + "loss": 1.0752, + "step": 20496 + }, + { + "epoch": 0.48, + "grad_norm": 2.139192807319102, + "learning_rate": 1.1038151070618976e-05, + "loss": 1.0142, + "step": 20497 + }, + { + "epoch": 0.48, + "grad_norm": 1.0429164707282654, + "learning_rate": 1.103739214954294e-05, + "loss": 0.9515, + "step": 20498 + }, + { + "epoch": 0.48, + "grad_norm": 1.9525622869739534, + "learning_rate": 1.103663322242688e-05, + "loss": 0.8841, + "step": 20499 + }, + { + "epoch": 0.48, + "grad_norm": 2.2475908567887997, + "learning_rate": 1.1035874289275212e-05, + "loss": 1.0305, + "step": 20500 + }, + { + "epoch": 0.48, + "grad_norm": 2.6085059437650235, + "learning_rate": 1.1035115350092356e-05, + "loss": 0.9323, + "step": 20501 + }, + { + "epoch": 0.48, + "grad_norm": 1.066484768478084, + "learning_rate": 1.1034356404882728e-05, + "loss": 0.8933, + "step": 20502 + }, + { + "epoch": 0.48, + "grad_norm": 1.7516550350853384, + "learning_rate": 1.1033597453650752e-05, + "loss": 1.0715, + "step": 20503 + }, + { + "epoch": 0.48, + "grad_norm": 1.9268803366611298, + "learning_rate": 1.1032838496400846e-05, + "loss": 1.0134, + "step": 20504 + }, + { + "epoch": 0.48, + "grad_norm": 2.132751938255674, + "learning_rate": 1.1032079533137426e-05, + "loss": 1.0268, + "step": 20505 + }, + { + "epoch": 0.48, + "grad_norm": 6.824019288722037, + "learning_rate": 1.1031320563864912e-05, + "loss": 0.9435, + "step": 20506 + }, + { + "epoch": 0.48, + "grad_norm": 4.084586005225649, + "learning_rate": 1.1030561588587722e-05, + "loss": 1.0101, + "step": 20507 + }, + { + "epoch": 0.48, + "grad_norm": 2.575825317786545, + "learning_rate": 1.102980260731028e-05, + "loss": 0.9981, + "step": 20508 + }, + { + "epoch": 0.48, + "grad_norm": 1.1831342320324771, + "learning_rate": 1.1029043620036996e-05, + "loss": 0.9526, + "step": 20509 + }, + { + "epoch": 0.48, + "grad_norm": 2.1736462838663546, + "learning_rate": 1.1028284626772299e-05, + "loss": 0.9887, + "step": 20510 + }, + { + "epoch": 0.48, + "grad_norm": 1.8837210509821158, + "learning_rate": 1.1027525627520599e-05, + "loss": 0.976, + "step": 20511 + }, + { + "epoch": 0.48, + "grad_norm": 2.2688411595093863, + "learning_rate": 1.1026766622286322e-05, + "loss": 0.9742, + "step": 20512 + }, + { + "epoch": 0.48, + "grad_norm": 1.9427270144391298, + "learning_rate": 1.1026007611073883e-05, + "loss": 1.1083, + "step": 20513 + }, + { + "epoch": 0.48, + "grad_norm": 2.062850885273922, + "learning_rate": 1.1025248593887707e-05, + "loss": 1.0945, + "step": 20514 + }, + { + "epoch": 0.48, + "grad_norm": 2.2586576864346393, + "learning_rate": 1.1024489570732207e-05, + "loss": 1.0647, + "step": 20515 + }, + { + "epoch": 0.48, + "grad_norm": 2.1383928857274883, + "learning_rate": 1.10237305416118e-05, + "loss": 0.9427, + "step": 20516 + }, + { + "epoch": 0.48, + "grad_norm": 2.0563539631720533, + "learning_rate": 1.1022971506530915e-05, + "loss": 0.9445, + "step": 20517 + }, + { + "epoch": 0.48, + "grad_norm": 1.9765083146301268, + "learning_rate": 1.1022212465493965e-05, + "loss": 0.9748, + "step": 20518 + }, + { + "epoch": 0.48, + "grad_norm": 1.9878467713235892, + "learning_rate": 1.102145341850537e-05, + "loss": 0.9881, + "step": 20519 + }, + { + "epoch": 0.48, + "grad_norm": 1.920644549054985, + "learning_rate": 1.1020694365569548e-05, + "loss": 1.0785, + "step": 20520 + }, + { + "epoch": 0.48, + "grad_norm": 5.65045586791283, + "learning_rate": 1.1019935306690921e-05, + "loss": 0.7785, + "step": 20521 + }, + { + "epoch": 0.48, + "grad_norm": 2.5274439597272607, + "learning_rate": 1.101917624187391e-05, + "loss": 1.0729, + "step": 20522 + }, + { + "epoch": 0.48, + "grad_norm": 2.4386159807726133, + "learning_rate": 1.101841717112293e-05, + "loss": 1.0545, + "step": 20523 + }, + { + "epoch": 0.48, + "grad_norm": 2.2381043743937443, + "learning_rate": 1.1017658094442403e-05, + "loss": 0.9845, + "step": 20524 + }, + { + "epoch": 0.48, + "grad_norm": 1.9373718951688146, + "learning_rate": 1.1016899011836749e-05, + "loss": 1.0986, + "step": 20525 + }, + { + "epoch": 0.48, + "grad_norm": 1.9802585991482131, + "learning_rate": 1.1016139923310388e-05, + "loss": 1.0484, + "step": 20526 + }, + { + "epoch": 0.48, + "grad_norm": 1.9159211542305348, + "learning_rate": 1.1015380828867736e-05, + "loss": 0.9767, + "step": 20527 + }, + { + "epoch": 0.48, + "grad_norm": 2.4868863346110768, + "learning_rate": 1.1014621728513218e-05, + "loss": 1.1191, + "step": 20528 + }, + { + "epoch": 0.48, + "grad_norm": 2.225413502402661, + "learning_rate": 1.101386262225125e-05, + "loss": 1.102, + "step": 20529 + }, + { + "epoch": 0.48, + "grad_norm": 2.0606397021433067, + "learning_rate": 1.1013103510086252e-05, + "loss": 0.9349, + "step": 20530 + }, + { + "epoch": 0.48, + "grad_norm": 2.0998157975834597, + "learning_rate": 1.1012344392022647e-05, + "loss": 1.0562, + "step": 20531 + }, + { + "epoch": 0.48, + "grad_norm": 1.2157504128658116, + "learning_rate": 1.1011585268064849e-05, + "loss": 1.0036, + "step": 20532 + }, + { + "epoch": 0.48, + "grad_norm": 2.328681437770161, + "learning_rate": 1.1010826138217284e-05, + "loss": 1.0743, + "step": 20533 + }, + { + "epoch": 0.48, + "grad_norm": 1.8065406239386774, + "learning_rate": 1.1010067002484368e-05, + "loss": 0.9448, + "step": 20534 + }, + { + "epoch": 0.48, + "grad_norm": 2.1130688802092545, + "learning_rate": 1.1009307860870522e-05, + "loss": 1.0315, + "step": 20535 + }, + { + "epoch": 0.48, + "grad_norm": 1.9311410763745807, + "learning_rate": 1.1008548713380167e-05, + "loss": 0.8924, + "step": 20536 + }, + { + "epoch": 0.48, + "grad_norm": 1.886574829527921, + "learning_rate": 1.1007789560017722e-05, + "loss": 1.0718, + "step": 20537 + }, + { + "epoch": 0.48, + "grad_norm": 1.7588118205379946, + "learning_rate": 1.1007030400787608e-05, + "loss": 1.0741, + "step": 20538 + }, + { + "epoch": 0.48, + "grad_norm": 1.7319103258359, + "learning_rate": 1.1006271235694242e-05, + "loss": 0.9068, + "step": 20539 + }, + { + "epoch": 0.48, + "grad_norm": 2.7324867682218947, + "learning_rate": 1.1005512064742048e-05, + "loss": 1.1111, + "step": 20540 + }, + { + "epoch": 0.48, + "grad_norm": 2.3224421964207953, + "learning_rate": 1.1004752887935442e-05, + "loss": 1.0395, + "step": 20541 + }, + { + "epoch": 0.48, + "grad_norm": 2.011393919547628, + "learning_rate": 1.100399370527885e-05, + "loss": 0.9831, + "step": 20542 + }, + { + "epoch": 0.48, + "grad_norm": 2.183357433052126, + "learning_rate": 1.1003234516776685e-05, + "loss": 1.1239, + "step": 20543 + }, + { + "epoch": 0.48, + "grad_norm": 2.10135971576888, + "learning_rate": 1.1002475322433374e-05, + "loss": 1.1302, + "step": 20544 + }, + { + "epoch": 0.48, + "grad_norm": 1.8990833107269576, + "learning_rate": 1.1001716122253335e-05, + "loss": 1.0535, + "step": 20545 + }, + { + "epoch": 0.48, + "grad_norm": 2.103488648882733, + "learning_rate": 1.1000956916240985e-05, + "loss": 0.9357, + "step": 20546 + }, + { + "epoch": 0.48, + "grad_norm": 2.344228684994285, + "learning_rate": 1.1000197704400748e-05, + "loss": 0.9923, + "step": 20547 + }, + { + "epoch": 0.48, + "grad_norm": 1.7941225312952838, + "learning_rate": 1.0999438486737043e-05, + "loss": 0.9316, + "step": 20548 + }, + { + "epoch": 0.48, + "grad_norm": 1.875900121041925, + "learning_rate": 1.099867926325429e-05, + "loss": 0.9738, + "step": 20549 + }, + { + "epoch": 0.48, + "grad_norm": 2.323770654282374, + "learning_rate": 1.0997920033956912e-05, + "loss": 1.0805, + "step": 20550 + }, + { + "epoch": 0.48, + "grad_norm": 1.918270682150975, + "learning_rate": 1.0997160798849327e-05, + "loss": 0.8973, + "step": 20551 + }, + { + "epoch": 0.48, + "grad_norm": 1.729846931426512, + "learning_rate": 1.0996401557935957e-05, + "loss": 0.9053, + "step": 20552 + }, + { + "epoch": 0.48, + "grad_norm": 2.185776344244965, + "learning_rate": 1.099564231122122e-05, + "loss": 1.0924, + "step": 20553 + }, + { + "epoch": 0.48, + "grad_norm": 2.054129594244516, + "learning_rate": 1.0994883058709538e-05, + "loss": 0.9367, + "step": 20554 + }, + { + "epoch": 0.48, + "grad_norm": 2.213220117777863, + "learning_rate": 1.0994123800405332e-05, + "loss": 1.1353, + "step": 20555 + }, + { + "epoch": 0.48, + "grad_norm": 3.644083532811453, + "learning_rate": 1.0993364536313022e-05, + "loss": 1.0389, + "step": 20556 + }, + { + "epoch": 0.48, + "grad_norm": 1.8548063457298753, + "learning_rate": 1.0992605266437032e-05, + "loss": 0.9844, + "step": 20557 + }, + { + "epoch": 0.48, + "grad_norm": 2.0571423674332476, + "learning_rate": 1.0991845990781777e-05, + "loss": 0.9097, + "step": 20558 + }, + { + "epoch": 0.48, + "grad_norm": 2.274316982709176, + "learning_rate": 1.0991086709351686e-05, + "loss": 0.9544, + "step": 20559 + }, + { + "epoch": 0.48, + "grad_norm": 2.1151305113289682, + "learning_rate": 1.0990327422151168e-05, + "loss": 0.8725, + "step": 20560 + }, + { + "epoch": 0.48, + "grad_norm": 2.021222786863211, + "learning_rate": 1.0989568129184654e-05, + "loss": 0.8479, + "step": 20561 + }, + { + "epoch": 0.48, + "grad_norm": 1.7442301991364468, + "learning_rate": 1.0988808830456557e-05, + "loss": 0.9422, + "step": 20562 + }, + { + "epoch": 0.48, + "grad_norm": 1.934415686817207, + "learning_rate": 1.0988049525971307e-05, + "loss": 0.9438, + "step": 20563 + }, + { + "epoch": 0.48, + "grad_norm": 1.8371191711968617, + "learning_rate": 1.0987290215733315e-05, + "loss": 0.9318, + "step": 20564 + }, + { + "epoch": 0.48, + "grad_norm": 1.9518051631628828, + "learning_rate": 1.098653089974701e-05, + "loss": 1.1255, + "step": 20565 + }, + { + "epoch": 0.48, + "grad_norm": 2.1216085261653848, + "learning_rate": 1.0985771578016808e-05, + "loss": 0.9584, + "step": 20566 + }, + { + "epoch": 0.48, + "grad_norm": 1.8187436524680642, + "learning_rate": 1.0985012250547133e-05, + "loss": 0.9348, + "step": 20567 + }, + { + "epoch": 0.48, + "grad_norm": 2.0977374975490513, + "learning_rate": 1.0984252917342406e-05, + "loss": 0.914, + "step": 20568 + }, + { + "epoch": 0.48, + "grad_norm": 1.99162312617351, + "learning_rate": 1.0983493578407044e-05, + "loss": 1.1304, + "step": 20569 + }, + { + "epoch": 0.48, + "grad_norm": 1.8101405605489056, + "learning_rate": 1.0982734233745473e-05, + "loss": 0.9927, + "step": 20570 + }, + { + "epoch": 0.48, + "grad_norm": 2.10961376777091, + "learning_rate": 1.098197488336211e-05, + "loss": 0.9856, + "step": 20571 + }, + { + "epoch": 0.48, + "grad_norm": 2.0400736924761573, + "learning_rate": 1.0981215527261383e-05, + "loss": 0.9273, + "step": 20572 + }, + { + "epoch": 0.48, + "grad_norm": 2.1248395800958098, + "learning_rate": 1.0980456165447704e-05, + "loss": 0.9755, + "step": 20573 + }, + { + "epoch": 0.48, + "grad_norm": 1.099705421943527, + "learning_rate": 1.09796967979255e-05, + "loss": 0.966, + "step": 20574 + }, + { + "epoch": 0.48, + "grad_norm": 1.1302442889429063, + "learning_rate": 1.0978937424699191e-05, + "loss": 0.9401, + "step": 20575 + }, + { + "epoch": 0.48, + "grad_norm": 1.8223105467357894, + "learning_rate": 1.0978178045773197e-05, + "loss": 0.9883, + "step": 20576 + }, + { + "epoch": 0.48, + "grad_norm": 1.928575221369029, + "learning_rate": 1.0977418661151943e-05, + "loss": 0.9592, + "step": 20577 + }, + { + "epoch": 0.48, + "grad_norm": 2.0299307778056828, + "learning_rate": 1.097665927083985e-05, + "loss": 0.9761, + "step": 20578 + }, + { + "epoch": 0.48, + "grad_norm": 1.911849135466384, + "learning_rate": 1.0975899874841334e-05, + "loss": 1.0084, + "step": 20579 + }, + { + "epoch": 0.48, + "grad_norm": 2.2103761995236377, + "learning_rate": 1.0975140473160824e-05, + "loss": 1.0962, + "step": 20580 + }, + { + "epoch": 0.48, + "grad_norm": 2.0726830667515297, + "learning_rate": 1.0974381065802731e-05, + "loss": 1.0758, + "step": 20581 + }, + { + "epoch": 0.48, + "grad_norm": 1.8979768383017552, + "learning_rate": 1.0973621652771487e-05, + "loss": 0.9172, + "step": 20582 + }, + { + "epoch": 0.48, + "grad_norm": 2.0240809833209212, + "learning_rate": 1.0972862234071508e-05, + "loss": 1.0678, + "step": 20583 + }, + { + "epoch": 0.48, + "grad_norm": 2.3749970373139915, + "learning_rate": 1.0972102809707219e-05, + "loss": 1.0886, + "step": 20584 + }, + { + "epoch": 0.48, + "grad_norm": 2.3584257927388155, + "learning_rate": 1.0971343379683037e-05, + "loss": 0.9768, + "step": 20585 + }, + { + "epoch": 0.48, + "grad_norm": 2.120436691269152, + "learning_rate": 1.0970583944003388e-05, + "loss": 1.0301, + "step": 20586 + }, + { + "epoch": 0.49, + "grad_norm": 1.7558791899517763, + "learning_rate": 1.0969824502672694e-05, + "loss": 0.943, + "step": 20587 + }, + { + "epoch": 0.49, + "grad_norm": 2.2976542472019954, + "learning_rate": 1.0969065055695371e-05, + "loss": 0.9856, + "step": 20588 + }, + { + "epoch": 0.49, + "grad_norm": 2.0566625344299476, + "learning_rate": 1.0968305603075849e-05, + "loss": 1.0162, + "step": 20589 + }, + { + "epoch": 0.49, + "grad_norm": 2.177753088853598, + "learning_rate": 1.096754614481854e-05, + "loss": 1.0129, + "step": 20590 + }, + { + "epoch": 0.49, + "grad_norm": 1.9727722145096076, + "learning_rate": 1.0966786680927875e-05, + "loss": 0.9615, + "step": 20591 + }, + { + "epoch": 0.49, + "grad_norm": 2.2079635406996654, + "learning_rate": 1.0966027211408268e-05, + "loss": 1.0125, + "step": 20592 + }, + { + "epoch": 0.49, + "grad_norm": 1.871451226066562, + "learning_rate": 1.096526773626415e-05, + "loss": 1.1102, + "step": 20593 + }, + { + "epoch": 0.49, + "grad_norm": 2.019407129716158, + "learning_rate": 1.0964508255499933e-05, + "loss": 0.9133, + "step": 20594 + }, + { + "epoch": 0.49, + "grad_norm": 2.345113624316041, + "learning_rate": 1.0963748769120046e-05, + "loss": 0.9453, + "step": 20595 + }, + { + "epoch": 0.49, + "grad_norm": 1.8380021166986311, + "learning_rate": 1.0962989277128909e-05, + "loss": 0.8995, + "step": 20596 + }, + { + "epoch": 0.49, + "grad_norm": 1.1203497212520102, + "learning_rate": 1.0962229779530941e-05, + "loss": 0.9528, + "step": 20597 + }, + { + "epoch": 0.49, + "grad_norm": 2.3257976423327533, + "learning_rate": 1.0961470276330568e-05, + "loss": 1.0478, + "step": 20598 + }, + { + "epoch": 0.49, + "grad_norm": 1.950066320857748, + "learning_rate": 1.0960710767532209e-05, + "loss": 1.0146, + "step": 20599 + }, + { + "epoch": 0.49, + "grad_norm": 1.9932685679336324, + "learning_rate": 1.095995125314029e-05, + "loss": 1.1324, + "step": 20600 + }, + { + "epoch": 0.49, + "grad_norm": 2.123227928753445, + "learning_rate": 1.095919173315923e-05, + "loss": 1.0837, + "step": 20601 + }, + { + "epoch": 0.49, + "grad_norm": 1.9530469386475262, + "learning_rate": 1.0958432207593453e-05, + "loss": 0.9673, + "step": 20602 + }, + { + "epoch": 0.49, + "grad_norm": 2.028838666516277, + "learning_rate": 1.095767267644738e-05, + "loss": 1.0052, + "step": 20603 + }, + { + "epoch": 0.49, + "grad_norm": 2.314572339367748, + "learning_rate": 1.095691313972543e-05, + "loss": 1.167, + "step": 20604 + }, + { + "epoch": 0.49, + "grad_norm": 2.0621655014111626, + "learning_rate": 1.0956153597432033e-05, + "loss": 0.8698, + "step": 20605 + }, + { + "epoch": 0.49, + "grad_norm": 1.9984878430287027, + "learning_rate": 1.0955394049571606e-05, + "loss": 0.8697, + "step": 20606 + }, + { + "epoch": 0.49, + "grad_norm": 1.0992961104041774, + "learning_rate": 1.0954634496148573e-05, + "loss": 0.9182, + "step": 20607 + }, + { + "epoch": 0.49, + "grad_norm": 1.8661071998283096, + "learning_rate": 1.0953874937167355e-05, + "loss": 0.9953, + "step": 20608 + }, + { + "epoch": 0.49, + "grad_norm": 2.1267234602282117, + "learning_rate": 1.0953115372632375e-05, + "loss": 1.0355, + "step": 20609 + }, + { + "epoch": 0.49, + "grad_norm": 1.1268550950615128, + "learning_rate": 1.0952355802548058e-05, + "loss": 0.9259, + "step": 20610 + }, + { + "epoch": 0.49, + "grad_norm": 2.73847256283284, + "learning_rate": 1.095159622691882e-05, + "loss": 1.0321, + "step": 20611 + }, + { + "epoch": 0.49, + "grad_norm": 2.1242270196713013, + "learning_rate": 1.095083664574909e-05, + "loss": 1.0786, + "step": 20612 + }, + { + "epoch": 0.49, + "grad_norm": 1.9275032976153577, + "learning_rate": 1.0950077059043285e-05, + "loss": 1.0728, + "step": 20613 + }, + { + "epoch": 0.49, + "grad_norm": 2.1765339594230855, + "learning_rate": 1.0949317466805834e-05, + "loss": 1.0561, + "step": 20614 + }, + { + "epoch": 0.49, + "grad_norm": 2.2781247723106546, + "learning_rate": 1.0948557869041151e-05, + "loss": 0.9855, + "step": 20615 + }, + { + "epoch": 0.49, + "grad_norm": 2.0577373070703935, + "learning_rate": 1.094779826575367e-05, + "loss": 1.0366, + "step": 20616 + }, + { + "epoch": 0.49, + "grad_norm": 1.9524134986246013, + "learning_rate": 1.0947038656947806e-05, + "loss": 0.9639, + "step": 20617 + }, + { + "epoch": 0.49, + "grad_norm": 2.429416419052275, + "learning_rate": 1.0946279042627981e-05, + "loss": 1.1086, + "step": 20618 + }, + { + "epoch": 0.49, + "grad_norm": 2.3121432280072134, + "learning_rate": 1.0945519422798622e-05, + "loss": 0.9497, + "step": 20619 + }, + { + "epoch": 0.49, + "grad_norm": 1.8739028257476518, + "learning_rate": 1.0944759797464146e-05, + "loss": 0.9879, + "step": 20620 + }, + { + "epoch": 0.49, + "grad_norm": 1.1335530850759223, + "learning_rate": 1.0944000166628986e-05, + "loss": 0.9721, + "step": 20621 + }, + { + "epoch": 0.49, + "grad_norm": 2.0365483802156232, + "learning_rate": 1.094324053029755e-05, + "loss": 1.08, + "step": 20622 + }, + { + "epoch": 0.49, + "grad_norm": 2.1832067455040587, + "learning_rate": 1.0942480888474276e-05, + "loss": 1.0255, + "step": 20623 + }, + { + "epoch": 0.49, + "grad_norm": 2.3496654032926294, + "learning_rate": 1.0941721241163577e-05, + "loss": 1.033, + "step": 20624 + }, + { + "epoch": 0.49, + "grad_norm": 2.0482156516971894, + "learning_rate": 1.0940961588369879e-05, + "loss": 0.9618, + "step": 20625 + }, + { + "epoch": 0.49, + "grad_norm": 2.1595878138255213, + "learning_rate": 1.0940201930097603e-05, + "loss": 1.1476, + "step": 20626 + }, + { + "epoch": 0.49, + "grad_norm": 2.0642215573284943, + "learning_rate": 1.0939442266351174e-05, + "loss": 1.0618, + "step": 20627 + }, + { + "epoch": 0.49, + "grad_norm": 1.132366259934812, + "learning_rate": 1.0938682597135017e-05, + "loss": 0.9203, + "step": 20628 + }, + { + "epoch": 0.49, + "grad_norm": 2.012451064336446, + "learning_rate": 1.0937922922453553e-05, + "loss": 1.0088, + "step": 20629 + }, + { + "epoch": 0.49, + "grad_norm": 1.9996832288467534, + "learning_rate": 1.0937163242311201e-05, + "loss": 0.9914, + "step": 20630 + }, + { + "epoch": 0.49, + "grad_norm": 1.1208253355976945, + "learning_rate": 1.0936403556712392e-05, + "loss": 1.0341, + "step": 20631 + }, + { + "epoch": 0.49, + "grad_norm": 2.201447152548649, + "learning_rate": 1.0935643865661543e-05, + "loss": 0.989, + "step": 20632 + }, + { + "epoch": 0.49, + "grad_norm": 2.0089727220646063, + "learning_rate": 1.0934884169163078e-05, + "loss": 1.0144, + "step": 20633 + }, + { + "epoch": 0.49, + "grad_norm": 2.1065051097278773, + "learning_rate": 1.0934124467221424e-05, + "loss": 1.0576, + "step": 20634 + }, + { + "epoch": 0.49, + "grad_norm": 1.0916244746262633, + "learning_rate": 1.0933364759841001e-05, + "loss": 0.9883, + "step": 20635 + }, + { + "epoch": 0.49, + "grad_norm": 1.0224006566975217, + "learning_rate": 1.0932605047026232e-05, + "loss": 0.9229, + "step": 20636 + }, + { + "epoch": 0.49, + "grad_norm": 1.8017860800637193, + "learning_rate": 1.0931845328781541e-05, + "loss": 1.136, + "step": 20637 + }, + { + "epoch": 0.49, + "grad_norm": 1.9331713676301123, + "learning_rate": 1.0931085605111354e-05, + "loss": 1.0113, + "step": 20638 + }, + { + "epoch": 0.49, + "grad_norm": 1.812093908013065, + "learning_rate": 1.0930325876020086e-05, + "loss": 0.9638, + "step": 20639 + }, + { + "epoch": 0.49, + "grad_norm": 1.9905817287471468, + "learning_rate": 1.0929566141512173e-05, + "loss": 1.0384, + "step": 20640 + }, + { + "epoch": 0.49, + "grad_norm": 2.2349241251869496, + "learning_rate": 1.0928806401592026e-05, + "loss": 1.0429, + "step": 20641 + }, + { + "epoch": 0.49, + "grad_norm": 2.272209134267755, + "learning_rate": 1.092804665626408e-05, + "loss": 1.1251, + "step": 20642 + }, + { + "epoch": 0.49, + "grad_norm": 2.24307619175173, + "learning_rate": 1.0927286905532749e-05, + "loss": 1.0141, + "step": 20643 + }, + { + "epoch": 0.49, + "grad_norm": 1.738259370336634, + "learning_rate": 1.0926527149402461e-05, + "loss": 1.0029, + "step": 20644 + }, + { + "epoch": 0.49, + "grad_norm": 2.148653062605031, + "learning_rate": 1.0925767387877638e-05, + "loss": 1.095, + "step": 20645 + }, + { + "epoch": 0.49, + "grad_norm": 2.2914985288454885, + "learning_rate": 1.0925007620962703e-05, + "loss": 0.9322, + "step": 20646 + }, + { + "epoch": 0.49, + "grad_norm": 2.0420296628852674, + "learning_rate": 1.0924247848662084e-05, + "loss": 0.9724, + "step": 20647 + }, + { + "epoch": 0.49, + "grad_norm": 1.9405296465042394, + "learning_rate": 1.0923488070980197e-05, + "loss": 1.0839, + "step": 20648 + }, + { + "epoch": 0.49, + "grad_norm": 2.112795240419955, + "learning_rate": 1.0922728287921474e-05, + "loss": 1.089, + "step": 20649 + }, + { + "epoch": 0.49, + "grad_norm": 1.1026477464505042, + "learning_rate": 1.0921968499490332e-05, + "loss": 0.8737, + "step": 20650 + }, + { + "epoch": 0.49, + "grad_norm": 2.8093971713552253, + "learning_rate": 1.09212087056912e-05, + "loss": 1.0163, + "step": 20651 + }, + { + "epoch": 0.49, + "grad_norm": 2.035413157832665, + "learning_rate": 1.0920448906528498e-05, + "loss": 1.092, + "step": 20652 + }, + { + "epoch": 0.49, + "grad_norm": 1.9330104427294696, + "learning_rate": 1.091968910200665e-05, + "loss": 1.1463, + "step": 20653 + }, + { + "epoch": 0.49, + "grad_norm": 1.9422923027909058, + "learning_rate": 1.0918929292130082e-05, + "loss": 1.0021, + "step": 20654 + }, + { + "epoch": 0.49, + "grad_norm": 1.9744547948268725, + "learning_rate": 1.0918169476903217e-05, + "loss": 1.0198, + "step": 20655 + }, + { + "epoch": 0.49, + "grad_norm": 2.046560651450255, + "learning_rate": 1.0917409656330477e-05, + "loss": 1.0158, + "step": 20656 + }, + { + "epoch": 0.49, + "grad_norm": 2.118484618679463, + "learning_rate": 1.091664983041629e-05, + "loss": 1.0419, + "step": 20657 + }, + { + "epoch": 0.49, + "grad_norm": 3.666499079262816, + "learning_rate": 1.0915889999165077e-05, + "loss": 1.0717, + "step": 20658 + }, + { + "epoch": 0.49, + "grad_norm": 2.2680267755517844, + "learning_rate": 1.091513016258126e-05, + "loss": 0.9738, + "step": 20659 + }, + { + "epoch": 0.49, + "grad_norm": 2.1247585778326936, + "learning_rate": 1.0914370320669267e-05, + "loss": 1.1171, + "step": 20660 + }, + { + "epoch": 0.49, + "grad_norm": 2.004499484949062, + "learning_rate": 1.0913610473433525e-05, + "loss": 0.9884, + "step": 20661 + }, + { + "epoch": 0.49, + "grad_norm": 1.9489931645418619, + "learning_rate": 1.0912850620878445e-05, + "loss": 1.0795, + "step": 20662 + }, + { + "epoch": 0.49, + "grad_norm": 2.378090956974674, + "learning_rate": 1.0912090763008468e-05, + "loss": 0.9666, + "step": 20663 + }, + { + "epoch": 0.49, + "grad_norm": 2.0463577328528944, + "learning_rate": 1.0911330899828003e-05, + "loss": 0.9551, + "step": 20664 + }, + { + "epoch": 0.49, + "grad_norm": 1.8984402035470245, + "learning_rate": 1.0910571031341485e-05, + "loss": 1.0554, + "step": 20665 + }, + { + "epoch": 0.49, + "grad_norm": 2.1269358045531326, + "learning_rate": 1.0909811157553331e-05, + "loss": 0.8245, + "step": 20666 + }, + { + "epoch": 0.49, + "grad_norm": 2.3804862800996647, + "learning_rate": 1.090905127846797e-05, + "loss": 1.027, + "step": 20667 + }, + { + "epoch": 0.49, + "grad_norm": 1.9591121601004984, + "learning_rate": 1.0908291394089828e-05, + "loss": 0.9925, + "step": 20668 + }, + { + "epoch": 0.49, + "grad_norm": 2.021547803885717, + "learning_rate": 1.0907531504423321e-05, + "loss": 1.0623, + "step": 20669 + }, + { + "epoch": 0.49, + "grad_norm": 2.0915014181190412, + "learning_rate": 1.0906771609472882e-05, + "loss": 0.9563, + "step": 20670 + }, + { + "epoch": 0.49, + "grad_norm": 2.006473288985755, + "learning_rate": 1.0906011709242927e-05, + "loss": 1.0335, + "step": 20671 + }, + { + "epoch": 0.49, + "grad_norm": 1.867769450163781, + "learning_rate": 1.0905251803737891e-05, + "loss": 0.9004, + "step": 20672 + }, + { + "epoch": 0.49, + "grad_norm": 2.094483519672907, + "learning_rate": 1.0904491892962186e-05, + "loss": 1.0662, + "step": 20673 + }, + { + "epoch": 0.49, + "grad_norm": 1.096683916927039, + "learning_rate": 1.0903731976920249e-05, + "loss": 0.9267, + "step": 20674 + }, + { + "epoch": 0.49, + "grad_norm": 1.0647927084885047, + "learning_rate": 1.0902972055616494e-05, + "loss": 0.9281, + "step": 20675 + }, + { + "epoch": 0.49, + "grad_norm": 2.192257752706687, + "learning_rate": 1.0902212129055353e-05, + "loss": 1.0007, + "step": 20676 + }, + { + "epoch": 0.49, + "grad_norm": 2.076555961185261, + "learning_rate": 1.0901452197241245e-05, + "loss": 1.0112, + "step": 20677 + }, + { + "epoch": 0.49, + "grad_norm": 1.1835003058998108, + "learning_rate": 1.0900692260178596e-05, + "loss": 0.9806, + "step": 20678 + }, + { + "epoch": 0.49, + "grad_norm": 2.2906206564720586, + "learning_rate": 1.0899932317871832e-05, + "loss": 0.9708, + "step": 20679 + }, + { + "epoch": 0.49, + "grad_norm": 2.298223334496324, + "learning_rate": 1.0899172370325379e-05, + "loss": 1.0135, + "step": 20680 + }, + { + "epoch": 0.49, + "grad_norm": 2.3556119713567827, + "learning_rate": 1.0898412417543658e-05, + "loss": 1.1044, + "step": 20681 + }, + { + "epoch": 0.49, + "grad_norm": 2.1948341370465623, + "learning_rate": 1.0897652459531098e-05, + "loss": 1.0193, + "step": 20682 + }, + { + "epoch": 0.49, + "grad_norm": 2.3462743277092226, + "learning_rate": 1.089689249629212e-05, + "loss": 1.066, + "step": 20683 + }, + { + "epoch": 0.49, + "grad_norm": 2.202510833102662, + "learning_rate": 1.0896132527831149e-05, + "loss": 0.9831, + "step": 20684 + }, + { + "epoch": 0.49, + "grad_norm": 2.1106331841233468, + "learning_rate": 1.0895372554152611e-05, + "loss": 1.1261, + "step": 20685 + }, + { + "epoch": 0.49, + "grad_norm": 1.9120200045442288, + "learning_rate": 1.089461257526093e-05, + "loss": 1.0777, + "step": 20686 + }, + { + "epoch": 0.49, + "grad_norm": 1.8861099084160864, + "learning_rate": 1.0893852591160533e-05, + "loss": 1.0844, + "step": 20687 + }, + { + "epoch": 0.49, + "grad_norm": 1.8060594150916147, + "learning_rate": 1.0893092601855841e-05, + "loss": 1.0207, + "step": 20688 + }, + { + "epoch": 0.49, + "grad_norm": 2.2281414958295076, + "learning_rate": 1.0892332607351283e-05, + "loss": 0.9764, + "step": 20689 + }, + { + "epoch": 0.49, + "grad_norm": 2.0190055067030332, + "learning_rate": 1.0891572607651281e-05, + "loss": 0.9265, + "step": 20690 + }, + { + "epoch": 0.49, + "grad_norm": 1.9018889470147062, + "learning_rate": 1.0890812602760263e-05, + "loss": 1.0124, + "step": 20691 + }, + { + "epoch": 0.49, + "grad_norm": 5.024690199234333, + "learning_rate": 1.089005259268265e-05, + "loss": 0.8397, + "step": 20692 + }, + { + "epoch": 0.49, + "grad_norm": 2.0350093175305117, + "learning_rate": 1.0889292577422869e-05, + "loss": 1.0168, + "step": 20693 + }, + { + "epoch": 0.49, + "grad_norm": 1.8155430094822933, + "learning_rate": 1.0888532556985346e-05, + "loss": 1.1645, + "step": 20694 + }, + { + "epoch": 0.49, + "grad_norm": 1.2123757869133303, + "learning_rate": 1.0887772531374506e-05, + "loss": 0.9654, + "step": 20695 + }, + { + "epoch": 0.49, + "grad_norm": 1.922845156549, + "learning_rate": 1.0887012500594771e-05, + "loss": 1.1591, + "step": 20696 + }, + { + "epoch": 0.49, + "grad_norm": 2.0888270300241563, + "learning_rate": 1.0886252464650569e-05, + "loss": 0.9252, + "step": 20697 + }, + { + "epoch": 0.49, + "grad_norm": 2.6171936789482397, + "learning_rate": 1.0885492423546324e-05, + "loss": 0.9777, + "step": 20698 + }, + { + "epoch": 0.49, + "grad_norm": 1.92422323018287, + "learning_rate": 1.0884732377286464e-05, + "loss": 1.0947, + "step": 20699 + }, + { + "epoch": 0.49, + "grad_norm": 2.3047552016990633, + "learning_rate": 1.0883972325875411e-05, + "loss": 1.1396, + "step": 20700 + }, + { + "epoch": 0.49, + "grad_norm": 2.418604073712997, + "learning_rate": 1.088321226931759e-05, + "loss": 1.0691, + "step": 20701 + }, + { + "epoch": 0.49, + "grad_norm": 2.055808120831027, + "learning_rate": 1.0882452207617428e-05, + "loss": 0.9238, + "step": 20702 + }, + { + "epoch": 0.49, + "grad_norm": 1.943834282005696, + "learning_rate": 1.0881692140779352e-05, + "loss": 1.1161, + "step": 20703 + }, + { + "epoch": 0.49, + "grad_norm": 1.9627146140225786, + "learning_rate": 1.0880932068807783e-05, + "loss": 0.9826, + "step": 20704 + }, + { + "epoch": 0.49, + "grad_norm": 1.1590146501239584, + "learning_rate": 1.0880171991707149e-05, + "loss": 1.0118, + "step": 20705 + }, + { + "epoch": 0.49, + "grad_norm": 1.171484437470595, + "learning_rate": 1.0879411909481874e-05, + "loss": 0.9322, + "step": 20706 + }, + { + "epoch": 0.49, + "grad_norm": 2.1498422254724985, + "learning_rate": 1.0878651822136386e-05, + "loss": 1.1438, + "step": 20707 + }, + { + "epoch": 0.49, + "grad_norm": 2.1271774392751177, + "learning_rate": 1.087789172967511e-05, + "loss": 1.023, + "step": 20708 + }, + { + "epoch": 0.49, + "grad_norm": 1.856743779914389, + "learning_rate": 1.0877131632102468e-05, + "loss": 0.9741, + "step": 20709 + }, + { + "epoch": 0.49, + "grad_norm": 2.2421958052272863, + "learning_rate": 1.087637152942289e-05, + "loss": 1.0574, + "step": 20710 + }, + { + "epoch": 0.49, + "grad_norm": 2.0783877530984713, + "learning_rate": 1.0875611421640797e-05, + "loss": 0.9835, + "step": 20711 + }, + { + "epoch": 0.49, + "grad_norm": 1.8087397386692181, + "learning_rate": 1.0874851308760622e-05, + "loss": 0.8535, + "step": 20712 + }, + { + "epoch": 0.49, + "grad_norm": 1.8828958265006501, + "learning_rate": 1.087409119078678e-05, + "loss": 1.0243, + "step": 20713 + }, + { + "epoch": 0.49, + "grad_norm": 2.241803358771212, + "learning_rate": 1.0873331067723704e-05, + "loss": 1.09, + "step": 20714 + }, + { + "epoch": 0.49, + "grad_norm": 2.1112963967846206, + "learning_rate": 1.0872570939575818e-05, + "loss": 1.1242, + "step": 20715 + }, + { + "epoch": 0.49, + "grad_norm": 2.090182813652915, + "learning_rate": 1.087181080634755e-05, + "loss": 0.9113, + "step": 20716 + }, + { + "epoch": 0.49, + "grad_norm": 1.9646302517453984, + "learning_rate": 1.087105066804332e-05, + "loss": 1.0942, + "step": 20717 + }, + { + "epoch": 0.49, + "grad_norm": 1.908180525615639, + "learning_rate": 1.087029052466756e-05, + "loss": 0.9546, + "step": 20718 + }, + { + "epoch": 0.49, + "grad_norm": 2.2372042069543636, + "learning_rate": 1.0869530376224693e-05, + "loss": 1.0722, + "step": 20719 + }, + { + "epoch": 0.49, + "grad_norm": 1.8351199709000843, + "learning_rate": 1.0868770222719143e-05, + "loss": 0.9712, + "step": 20720 + }, + { + "epoch": 0.49, + "grad_norm": 1.9803977175653147, + "learning_rate": 1.0868010064155338e-05, + "loss": 1.0151, + "step": 20721 + }, + { + "epoch": 0.49, + "grad_norm": 1.929345340572809, + "learning_rate": 1.0867249900537703e-05, + "loss": 1.0591, + "step": 20722 + }, + { + "epoch": 0.49, + "grad_norm": 1.9868746385925753, + "learning_rate": 1.0866489731870668e-05, + "loss": 1.0211, + "step": 20723 + }, + { + "epoch": 0.49, + "grad_norm": 2.3827557973290494, + "learning_rate": 1.0865729558158652e-05, + "loss": 1.1171, + "step": 20724 + }, + { + "epoch": 0.49, + "grad_norm": 2.862637360999254, + "learning_rate": 1.0864969379406088e-05, + "loss": 1.1537, + "step": 20725 + }, + { + "epoch": 0.49, + "grad_norm": 2.1734095134594122, + "learning_rate": 1.0864209195617395e-05, + "loss": 0.931, + "step": 20726 + }, + { + "epoch": 0.49, + "grad_norm": 2.2313404875618494, + "learning_rate": 1.0863449006797005e-05, + "loss": 0.9011, + "step": 20727 + }, + { + "epoch": 0.49, + "grad_norm": 1.9402210252272623, + "learning_rate": 1.0862688812949338e-05, + "loss": 0.9623, + "step": 20728 + }, + { + "epoch": 0.49, + "grad_norm": 2.167869381947101, + "learning_rate": 1.0861928614078827e-05, + "loss": 1.0904, + "step": 20729 + }, + { + "epoch": 0.49, + "grad_norm": 2.1038528852072775, + "learning_rate": 1.0861168410189894e-05, + "loss": 1.1025, + "step": 20730 + }, + { + "epoch": 0.49, + "grad_norm": 1.1319367106341043, + "learning_rate": 1.0860408201286964e-05, + "loss": 1.0382, + "step": 20731 + }, + { + "epoch": 0.49, + "grad_norm": 2.199007120354817, + "learning_rate": 1.0859647987374467e-05, + "loss": 0.9326, + "step": 20732 + }, + { + "epoch": 0.49, + "grad_norm": 1.9912097730020484, + "learning_rate": 1.0858887768456824e-05, + "loss": 1.0224, + "step": 20733 + }, + { + "epoch": 0.49, + "grad_norm": 2.512129643372495, + "learning_rate": 1.0858127544538468e-05, + "loss": 0.9427, + "step": 20734 + }, + { + "epoch": 0.49, + "grad_norm": 1.1132489135923087, + "learning_rate": 1.085736731562382e-05, + "loss": 0.9828, + "step": 20735 + }, + { + "epoch": 0.49, + "grad_norm": 2.077880176081975, + "learning_rate": 1.085660708171731e-05, + "loss": 1.0352, + "step": 20736 + }, + { + "epoch": 0.49, + "grad_norm": 1.0799587725534752, + "learning_rate": 1.085584684282336e-05, + "loss": 0.962, + "step": 20737 + }, + { + "epoch": 0.49, + "grad_norm": 2.057837512604521, + "learning_rate": 1.0855086598946401e-05, + "loss": 1.0626, + "step": 20738 + }, + { + "epoch": 0.49, + "grad_norm": 1.7545634037676896, + "learning_rate": 1.0854326350090855e-05, + "loss": 1.0697, + "step": 20739 + }, + { + "epoch": 0.49, + "grad_norm": 1.9829665685358688, + "learning_rate": 1.0853566096261155e-05, + "loss": 0.9477, + "step": 20740 + }, + { + "epoch": 0.49, + "grad_norm": 2.0013418495452973, + "learning_rate": 1.0852805837461718e-05, + "loss": 1.0278, + "step": 20741 + }, + { + "epoch": 0.49, + "grad_norm": 1.1536631131241477, + "learning_rate": 1.0852045573696979e-05, + "loss": 1.0064, + "step": 20742 + }, + { + "epoch": 0.49, + "grad_norm": 2.068786796663229, + "learning_rate": 1.0851285304971354e-05, + "loss": 0.9199, + "step": 20743 + }, + { + "epoch": 0.49, + "grad_norm": 3.299624888251091, + "learning_rate": 1.0850525031289286e-05, + "loss": 1.1013, + "step": 20744 + }, + { + "epoch": 0.49, + "grad_norm": 2.4544867206869885, + "learning_rate": 1.0849764752655185e-05, + "loss": 0.8142, + "step": 20745 + }, + { + "epoch": 0.49, + "grad_norm": 1.9351005849245586, + "learning_rate": 1.0849004469073487e-05, + "loss": 1.0759, + "step": 20746 + }, + { + "epoch": 0.49, + "grad_norm": 1.129955641980089, + "learning_rate": 1.0848244180548618e-05, + "loss": 0.9352, + "step": 20747 + }, + { + "epoch": 0.49, + "grad_norm": 2.0151671948495813, + "learning_rate": 1.0847483887085e-05, + "loss": 0.935, + "step": 20748 + }, + { + "epoch": 0.49, + "grad_norm": 2.0738329496047316, + "learning_rate": 1.0846723588687065e-05, + "loss": 0.9794, + "step": 20749 + }, + { + "epoch": 0.49, + "grad_norm": 1.9951253712038608, + "learning_rate": 1.0845963285359233e-05, + "loss": 0.9934, + "step": 20750 + }, + { + "epoch": 0.49, + "grad_norm": 2.077953205926624, + "learning_rate": 1.084520297710594e-05, + "loss": 1.0633, + "step": 20751 + }, + { + "epoch": 0.49, + "grad_norm": 2.37993295735692, + "learning_rate": 1.0844442663931604e-05, + "loss": 1.0836, + "step": 20752 + }, + { + "epoch": 0.49, + "grad_norm": 1.850185914324305, + "learning_rate": 1.0843682345840655e-05, + "loss": 0.9265, + "step": 20753 + }, + { + "epoch": 0.49, + "grad_norm": 2.3090849291758464, + "learning_rate": 1.0842922022837524e-05, + "loss": 0.986, + "step": 20754 + }, + { + "epoch": 0.49, + "grad_norm": 1.1149560951517605, + "learning_rate": 1.0842161694926632e-05, + "loss": 0.9329, + "step": 20755 + }, + { + "epoch": 0.49, + "grad_norm": 1.8999498960499488, + "learning_rate": 1.0841401362112408e-05, + "loss": 0.9503, + "step": 20756 + }, + { + "epoch": 0.49, + "grad_norm": 2.458776037383346, + "learning_rate": 1.0840641024399278e-05, + "loss": 1.0088, + "step": 20757 + }, + { + "epoch": 0.49, + "grad_norm": 1.094273548119638, + "learning_rate": 1.083988068179167e-05, + "loss": 1.0205, + "step": 20758 + }, + { + "epoch": 0.49, + "grad_norm": 2.020179405183704, + "learning_rate": 1.0839120334294013e-05, + "loss": 1.0218, + "step": 20759 + }, + { + "epoch": 0.49, + "grad_norm": 2.098053124946638, + "learning_rate": 1.083835998191073e-05, + "loss": 0.9711, + "step": 20760 + }, + { + "epoch": 0.49, + "grad_norm": 1.0968434102466047, + "learning_rate": 1.0837599624646251e-05, + "loss": 0.9616, + "step": 20761 + }, + { + "epoch": 0.49, + "grad_norm": 2.213181353624159, + "learning_rate": 1.0836839262504998e-05, + "loss": 1.0126, + "step": 20762 + }, + { + "epoch": 0.49, + "grad_norm": 2.0026912885604746, + "learning_rate": 1.0836078895491408e-05, + "loss": 1.0824, + "step": 20763 + }, + { + "epoch": 0.49, + "grad_norm": 2.0663005114716895, + "learning_rate": 1.0835318523609897e-05, + "loss": 0.9856, + "step": 20764 + }, + { + "epoch": 0.49, + "grad_norm": 1.087491842851109, + "learning_rate": 1.08345581468649e-05, + "loss": 0.9209, + "step": 20765 + }, + { + "epoch": 0.49, + "grad_norm": 2.0509369071288694, + "learning_rate": 1.0833797765260838e-05, + "loss": 0.9853, + "step": 20766 + }, + { + "epoch": 0.49, + "grad_norm": 1.8069172682182104, + "learning_rate": 1.0833037378802145e-05, + "loss": 1.1327, + "step": 20767 + }, + { + "epoch": 0.49, + "grad_norm": 1.785673940594759, + "learning_rate": 1.0832276987493243e-05, + "loss": 1.0752, + "step": 20768 + }, + { + "epoch": 0.49, + "grad_norm": 1.9751357630105584, + "learning_rate": 1.083151659133856e-05, + "loss": 1.0087, + "step": 20769 + }, + { + "epoch": 0.49, + "grad_norm": 1.812222525176861, + "learning_rate": 1.0830756190342528e-05, + "loss": 0.9745, + "step": 20770 + }, + { + "epoch": 0.49, + "grad_norm": 2.0659171313827573, + "learning_rate": 1.0829995784509565e-05, + "loss": 1.0824, + "step": 20771 + }, + { + "epoch": 0.49, + "grad_norm": 1.906118844505643, + "learning_rate": 1.0829235373844109e-05, + "loss": 0.9599, + "step": 20772 + }, + { + "epoch": 0.49, + "grad_norm": 1.8329720282783422, + "learning_rate": 1.0828474958350577e-05, + "loss": 1.0002, + "step": 20773 + }, + { + "epoch": 0.49, + "grad_norm": 1.9596068396681359, + "learning_rate": 1.0827714538033407e-05, + "loss": 1.161, + "step": 20774 + }, + { + "epoch": 0.49, + "grad_norm": 1.9513380823641335, + "learning_rate": 1.0826954112897014e-05, + "loss": 1.0703, + "step": 20775 + }, + { + "epoch": 0.49, + "grad_norm": 1.943717590461246, + "learning_rate": 1.0826193682945839e-05, + "loss": 0.9344, + "step": 20776 + }, + { + "epoch": 0.49, + "grad_norm": 1.1251650807504683, + "learning_rate": 1.0825433248184301e-05, + "loss": 0.9928, + "step": 20777 + }, + { + "epoch": 0.49, + "grad_norm": 2.0803003768319996, + "learning_rate": 1.0824672808616827e-05, + "loss": 1.0745, + "step": 20778 + }, + { + "epoch": 0.49, + "grad_norm": 2.202201943054124, + "learning_rate": 1.0823912364247848e-05, + "loss": 1.0305, + "step": 20779 + }, + { + "epoch": 0.49, + "grad_norm": 2.1514649669818082, + "learning_rate": 1.0823151915081791e-05, + "loss": 1.0347, + "step": 20780 + }, + { + "epoch": 0.49, + "grad_norm": 1.067928711654115, + "learning_rate": 1.0822391461123083e-05, + "loss": 0.9467, + "step": 20781 + }, + { + "epoch": 0.49, + "grad_norm": 2.020037721865979, + "learning_rate": 1.0821631002376152e-05, + "loss": 1.0469, + "step": 20782 + }, + { + "epoch": 0.49, + "grad_norm": 1.966367620319165, + "learning_rate": 1.0820870538845423e-05, + "loss": 0.9924, + "step": 20783 + }, + { + "epoch": 0.49, + "grad_norm": 1.9040998849094453, + "learning_rate": 1.0820110070535326e-05, + "loss": 0.864, + "step": 20784 + }, + { + "epoch": 0.49, + "grad_norm": 2.1112500591164083, + "learning_rate": 1.081934959745029e-05, + "loss": 0.9067, + "step": 20785 + }, + { + "epoch": 0.49, + "grad_norm": 2.0751271853203335, + "learning_rate": 1.081858911959474e-05, + "loss": 1.0401, + "step": 20786 + }, + { + "epoch": 0.49, + "grad_norm": 1.9186519994379607, + "learning_rate": 1.0817828636973106e-05, + "loss": 0.9152, + "step": 20787 + }, + { + "epoch": 0.49, + "grad_norm": 1.9530731638943852, + "learning_rate": 1.0817068149589816e-05, + "loss": 0.9625, + "step": 20788 + }, + { + "epoch": 0.49, + "grad_norm": 2.1001872133611283, + "learning_rate": 1.0816307657449294e-05, + "loss": 1.0726, + "step": 20789 + }, + { + "epoch": 0.49, + "grad_norm": 1.91949279391093, + "learning_rate": 1.0815547160555973e-05, + "loss": 1.0215, + "step": 20790 + }, + { + "epoch": 0.49, + "grad_norm": 2.114922674668322, + "learning_rate": 1.0814786658914279e-05, + "loss": 0.9805, + "step": 20791 + }, + { + "epoch": 0.49, + "grad_norm": 2.2321807649982586, + "learning_rate": 1.0814026152528634e-05, + "loss": 1.0124, + "step": 20792 + }, + { + "epoch": 0.49, + "grad_norm": 1.9183008919171738, + "learning_rate": 1.0813265641403476e-05, + "loss": 0.8903, + "step": 20793 + }, + { + "epoch": 0.49, + "grad_norm": 2.197332309770631, + "learning_rate": 1.0812505125543223e-05, + "loss": 1.1859, + "step": 20794 + }, + { + "epoch": 0.49, + "grad_norm": 2.243297253674705, + "learning_rate": 1.0811744604952312e-05, + "loss": 1.0038, + "step": 20795 + }, + { + "epoch": 0.49, + "grad_norm": 2.0045543562793373, + "learning_rate": 1.0810984079635164e-05, + "loss": 1.0124, + "step": 20796 + }, + { + "epoch": 0.49, + "grad_norm": 1.1811297463017287, + "learning_rate": 1.0810223549596215e-05, + "loss": 0.996, + "step": 20797 + }, + { + "epoch": 0.49, + "grad_norm": 2.05995933802651, + "learning_rate": 1.0809463014839884e-05, + "loss": 0.9954, + "step": 20798 + }, + { + "epoch": 0.49, + "grad_norm": 2.0004838339267574, + "learning_rate": 1.0808702475370603e-05, + "loss": 1.0216, + "step": 20799 + }, + { + "epoch": 0.49, + "grad_norm": 1.889221111299208, + "learning_rate": 1.0807941931192802e-05, + "loss": 0.9675, + "step": 20800 + }, + { + "epoch": 0.49, + "grad_norm": 2.288411081419417, + "learning_rate": 1.0807181382310904e-05, + "loss": 0.9328, + "step": 20801 + }, + { + "epoch": 0.49, + "grad_norm": 2.010035023182492, + "learning_rate": 1.0806420828729345e-05, + "loss": 0.9471, + "step": 20802 + }, + { + "epoch": 0.49, + "grad_norm": 1.8356526840615617, + "learning_rate": 1.0805660270452542e-05, + "loss": 1.0457, + "step": 20803 + }, + { + "epoch": 0.49, + "grad_norm": 2.0376152312743225, + "learning_rate": 1.0804899707484938e-05, + "loss": 0.9483, + "step": 20804 + }, + { + "epoch": 0.49, + "grad_norm": 2.168435763413899, + "learning_rate": 1.0804139139830947e-05, + "loss": 1.0575, + "step": 20805 + }, + { + "epoch": 0.49, + "grad_norm": 1.0849382915760115, + "learning_rate": 1.0803378567495006e-05, + "loss": 0.9036, + "step": 20806 + }, + { + "epoch": 0.49, + "grad_norm": 1.9844201836196202, + "learning_rate": 1.0802617990481542e-05, + "loss": 1.0801, + "step": 20807 + }, + { + "epoch": 0.49, + "grad_norm": 1.7948559993084894, + "learning_rate": 1.0801857408794978e-05, + "loss": 1.0152, + "step": 20808 + }, + { + "epoch": 0.49, + "grad_norm": 2.2719080225287267, + "learning_rate": 1.080109682243975e-05, + "loss": 0.9419, + "step": 20809 + }, + { + "epoch": 0.49, + "grad_norm": 2.0827488016123774, + "learning_rate": 1.0800336231420278e-05, + "loss": 0.9642, + "step": 20810 + }, + { + "epoch": 0.49, + "grad_norm": 1.986853023467249, + "learning_rate": 1.0799575635741e-05, + "loss": 0.9625, + "step": 20811 + }, + { + "epoch": 0.49, + "grad_norm": 2.140147887241466, + "learning_rate": 1.079881503540634e-05, + "loss": 1.0795, + "step": 20812 + }, + { + "epoch": 0.49, + "grad_norm": 2.8408344073960583, + "learning_rate": 1.079805443042072e-05, + "loss": 0.9674, + "step": 20813 + }, + { + "epoch": 0.49, + "grad_norm": 1.2699542098459549, + "learning_rate": 1.0797293820788578e-05, + "loss": 1.0598, + "step": 20814 + }, + { + "epoch": 0.49, + "grad_norm": 1.8491534827593719, + "learning_rate": 1.079653320651434e-05, + "loss": 1.0123, + "step": 20815 + }, + { + "epoch": 0.49, + "grad_norm": 1.9057400752536011, + "learning_rate": 1.0795772587602432e-05, + "loss": 1.0142, + "step": 20816 + }, + { + "epoch": 0.49, + "grad_norm": 1.087046091033028, + "learning_rate": 1.0795011964057285e-05, + "loss": 0.9784, + "step": 20817 + }, + { + "epoch": 0.49, + "grad_norm": 2.0554561726202722, + "learning_rate": 1.0794251335883326e-05, + "loss": 1.0585, + "step": 20818 + }, + { + "epoch": 0.49, + "grad_norm": 2.101736674062009, + "learning_rate": 1.0793490703084985e-05, + "loss": 1.0315, + "step": 20819 + }, + { + "epoch": 0.49, + "grad_norm": 2.141503957167136, + "learning_rate": 1.0792730065666688e-05, + "loss": 1.0722, + "step": 20820 + }, + { + "epoch": 0.49, + "grad_norm": 2.9711459101513893, + "learning_rate": 1.0791969423632869e-05, + "loss": 0.8721, + "step": 20821 + }, + { + "epoch": 0.49, + "grad_norm": 2.1751461510396077, + "learning_rate": 1.0791208776987949e-05, + "loss": 0.9979, + "step": 20822 + }, + { + "epoch": 0.49, + "grad_norm": 3.61741207921028, + "learning_rate": 1.0790448125736366e-05, + "loss": 0.8822, + "step": 20823 + }, + { + "epoch": 0.49, + "grad_norm": 2.2302772066859684, + "learning_rate": 1.0789687469882538e-05, + "loss": 1.0535, + "step": 20824 + }, + { + "epoch": 0.49, + "grad_norm": 1.9355913331187993, + "learning_rate": 1.0788926809430904e-05, + "loss": 0.9965, + "step": 20825 + }, + { + "epoch": 0.49, + "grad_norm": 2.3058824823498982, + "learning_rate": 1.0788166144385888e-05, + "loss": 0.9065, + "step": 20826 + }, + { + "epoch": 0.49, + "grad_norm": 2.4199790469652096, + "learning_rate": 1.0787405474751917e-05, + "loss": 0.9607, + "step": 20827 + }, + { + "epoch": 0.49, + "grad_norm": 1.9651618449645467, + "learning_rate": 1.0786644800533423e-05, + "loss": 1.0259, + "step": 20828 + }, + { + "epoch": 0.49, + "grad_norm": 1.086637161672103, + "learning_rate": 1.0785884121734836e-05, + "loss": 0.9353, + "step": 20829 + }, + { + "epoch": 0.49, + "grad_norm": 1.0830034245358278, + "learning_rate": 1.0785123438360578e-05, + "loss": 1.0214, + "step": 20830 + }, + { + "epoch": 0.49, + "grad_norm": 2.1768315534008287, + "learning_rate": 1.0784362750415088e-05, + "loss": 1.1197, + "step": 20831 + }, + { + "epoch": 0.49, + "grad_norm": 1.9214039830858465, + "learning_rate": 1.0783602057902786e-05, + "loss": 1.0798, + "step": 20832 + }, + { + "epoch": 0.49, + "grad_norm": 1.9094664917400903, + "learning_rate": 1.0782841360828107e-05, + "loss": 1.0397, + "step": 20833 + }, + { + "epoch": 0.49, + "grad_norm": 2.707302834358546, + "learning_rate": 1.0782080659195475e-05, + "loss": 0.9247, + "step": 20834 + }, + { + "epoch": 0.49, + "grad_norm": 2.3582896910448556, + "learning_rate": 1.0781319953009324e-05, + "loss": 1.1146, + "step": 20835 + }, + { + "epoch": 0.49, + "grad_norm": 1.9912226524822239, + "learning_rate": 1.078055924227408e-05, + "loss": 1.0317, + "step": 20836 + }, + { + "epoch": 0.49, + "grad_norm": 3.539259931457809, + "learning_rate": 1.0779798526994173e-05, + "loss": 1.0294, + "step": 20837 + }, + { + "epoch": 0.49, + "grad_norm": 2.1939658039346392, + "learning_rate": 1.0779037807174032e-05, + "loss": 1.0099, + "step": 20838 + }, + { + "epoch": 0.49, + "grad_norm": 3.982335681482976, + "learning_rate": 1.077827708281809e-05, + "loss": 0.8533, + "step": 20839 + }, + { + "epoch": 0.49, + "grad_norm": 2.087363497947052, + "learning_rate": 1.0777516353930767e-05, + "loss": 0.8913, + "step": 20840 + }, + { + "epoch": 0.49, + "grad_norm": 1.9983308315097181, + "learning_rate": 1.0776755620516501e-05, + "loss": 0.965, + "step": 20841 + }, + { + "epoch": 0.49, + "grad_norm": 2.016447716379513, + "learning_rate": 1.0775994882579718e-05, + "loss": 0.9691, + "step": 20842 + }, + { + "epoch": 0.49, + "grad_norm": 2.0290631085257864, + "learning_rate": 1.0775234140124845e-05, + "loss": 0.9954, + "step": 20843 + }, + { + "epoch": 0.49, + "grad_norm": 2.1239931164692933, + "learning_rate": 1.0774473393156315e-05, + "loss": 0.9643, + "step": 20844 + }, + { + "epoch": 0.49, + "grad_norm": 2.8503421134050018, + "learning_rate": 1.0773712641678553e-05, + "loss": 0.9781, + "step": 20845 + }, + { + "epoch": 0.49, + "grad_norm": 2.1047862174441576, + "learning_rate": 1.0772951885695994e-05, + "loss": 1.0928, + "step": 20846 + }, + { + "epoch": 0.49, + "grad_norm": 1.047855284422942, + "learning_rate": 1.0772191125213063e-05, + "loss": 0.9741, + "step": 20847 + }, + { + "epoch": 0.49, + "grad_norm": 1.8658126844525529, + "learning_rate": 1.0771430360234194e-05, + "loss": 1.0096, + "step": 20848 + }, + { + "epoch": 0.49, + "grad_norm": 2.248410397329224, + "learning_rate": 1.0770669590763809e-05, + "loss": 1.1055, + "step": 20849 + }, + { + "epoch": 0.49, + "grad_norm": 2.0553360723802467, + "learning_rate": 1.0769908816806343e-05, + "loss": 1.1408, + "step": 20850 + }, + { + "epoch": 0.49, + "grad_norm": 1.978790494879071, + "learning_rate": 1.0769148038366223e-05, + "loss": 1.0482, + "step": 20851 + }, + { + "epoch": 0.49, + "grad_norm": 1.0844225862263974, + "learning_rate": 1.076838725544788e-05, + "loss": 0.9317, + "step": 20852 + }, + { + "epoch": 0.49, + "grad_norm": 1.9484895285056842, + "learning_rate": 1.0767626468055747e-05, + "loss": 1.1187, + "step": 20853 + }, + { + "epoch": 0.49, + "grad_norm": 2.460835805562871, + "learning_rate": 1.0766865676194245e-05, + "loss": 0.8645, + "step": 20854 + }, + { + "epoch": 0.49, + "grad_norm": 1.8932524101559558, + "learning_rate": 1.0766104879867811e-05, + "loss": 0.7501, + "step": 20855 + }, + { + "epoch": 0.49, + "grad_norm": 2.0054061657625417, + "learning_rate": 1.076534407908087e-05, + "loss": 1.0948, + "step": 20856 + }, + { + "epoch": 0.49, + "grad_norm": 2.0882528560562816, + "learning_rate": 1.0764583273837855e-05, + "loss": 1.1248, + "step": 20857 + }, + { + "epoch": 0.49, + "grad_norm": 1.9035689591299036, + "learning_rate": 1.0763822464143191e-05, + "loss": 1.0825, + "step": 20858 + }, + { + "epoch": 0.49, + "grad_norm": 1.862844655431016, + "learning_rate": 1.0763061650001314e-05, + "loss": 0.9533, + "step": 20859 + }, + { + "epoch": 0.49, + "grad_norm": 1.8765329845651892, + "learning_rate": 1.076230083141665e-05, + "loss": 0.9981, + "step": 20860 + }, + { + "epoch": 0.49, + "grad_norm": 2.292249330318595, + "learning_rate": 1.076154000839363e-05, + "loss": 0.9348, + "step": 20861 + }, + { + "epoch": 0.49, + "grad_norm": 1.8150480138756426, + "learning_rate": 1.076077918093668e-05, + "loss": 0.9923, + "step": 20862 + }, + { + "epoch": 0.49, + "grad_norm": 2.373499157420113, + "learning_rate": 1.0760018349050236e-05, + "loss": 0.9173, + "step": 20863 + }, + { + "epoch": 0.49, + "grad_norm": 1.8200774704918115, + "learning_rate": 1.0759257512738723e-05, + "loss": 1.0226, + "step": 20864 + }, + { + "epoch": 0.49, + "grad_norm": 1.9075123037686545, + "learning_rate": 1.0758496672006571e-05, + "loss": 0.8803, + "step": 20865 + }, + { + "epoch": 0.49, + "grad_norm": 2.2816329320917665, + "learning_rate": 1.0757735826858213e-05, + "loss": 0.9976, + "step": 20866 + }, + { + "epoch": 0.49, + "grad_norm": 2.104488304426678, + "learning_rate": 1.0756974977298078e-05, + "loss": 0.9257, + "step": 20867 + }, + { + "epoch": 0.49, + "grad_norm": 1.0908743333545665, + "learning_rate": 1.0756214123330593e-05, + "loss": 1.01, + "step": 20868 + }, + { + "epoch": 0.49, + "grad_norm": 2.513436888520151, + "learning_rate": 1.0755453264960192e-05, + "loss": 1.2102, + "step": 20869 + }, + { + "epoch": 0.49, + "grad_norm": 2.6922202506141306, + "learning_rate": 1.0754692402191303e-05, + "loss": 0.8805, + "step": 20870 + }, + { + "epoch": 0.49, + "grad_norm": 1.8587087585369988, + "learning_rate": 1.0753931535028353e-05, + "loss": 1.1101, + "step": 20871 + }, + { + "epoch": 0.49, + "grad_norm": 2.035685437611071, + "learning_rate": 1.0753170663475778e-05, + "loss": 0.9998, + "step": 20872 + }, + { + "epoch": 0.49, + "grad_norm": 2.251599831878841, + "learning_rate": 1.0752409787538e-05, + "loss": 1.0321, + "step": 20873 + }, + { + "epoch": 0.49, + "grad_norm": 2.2833767500938014, + "learning_rate": 1.0751648907219461e-05, + "loss": 1.0226, + "step": 20874 + }, + { + "epoch": 0.49, + "grad_norm": 2.1713184399996663, + "learning_rate": 1.075088802252458e-05, + "loss": 0.96, + "step": 20875 + }, + { + "epoch": 0.49, + "grad_norm": 1.940592770953851, + "learning_rate": 1.0750127133457792e-05, + "loss": 1.0066, + "step": 20876 + }, + { + "epoch": 0.49, + "grad_norm": 2.2146780616019655, + "learning_rate": 1.0749366240023526e-05, + "loss": 0.9817, + "step": 20877 + }, + { + "epoch": 0.49, + "grad_norm": 1.8965173442461642, + "learning_rate": 1.0748605342226213e-05, + "loss": 1.0349, + "step": 20878 + }, + { + "epoch": 0.49, + "grad_norm": 1.8489876616641325, + "learning_rate": 1.0747844440070284e-05, + "loss": 1.1031, + "step": 20879 + }, + { + "epoch": 0.49, + "grad_norm": 2.4802040720135436, + "learning_rate": 1.0747083533560166e-05, + "loss": 1.0859, + "step": 20880 + }, + { + "epoch": 0.49, + "grad_norm": 2.0351628075299995, + "learning_rate": 1.0746322622700289e-05, + "loss": 1.0366, + "step": 20881 + }, + { + "epoch": 0.49, + "grad_norm": 1.7763782146855025, + "learning_rate": 1.0745561707495088e-05, + "loss": 0.8956, + "step": 20882 + }, + { + "epoch": 0.49, + "grad_norm": 1.1949642660145074, + "learning_rate": 1.074480078794899e-05, + "loss": 1.0056, + "step": 20883 + }, + { + "epoch": 0.49, + "grad_norm": 1.789781910632151, + "learning_rate": 1.0744039864066427e-05, + "loss": 0.9652, + "step": 20884 + }, + { + "epoch": 0.49, + "grad_norm": 2.0479750001412445, + "learning_rate": 1.0743278935851828e-05, + "loss": 1.1027, + "step": 20885 + }, + { + "epoch": 0.49, + "grad_norm": 1.0635944653362306, + "learning_rate": 1.0742518003309624e-05, + "loss": 0.9144, + "step": 20886 + }, + { + "epoch": 0.49, + "grad_norm": 4.582111022983264, + "learning_rate": 1.0741757066444244e-05, + "loss": 1.015, + "step": 20887 + }, + { + "epoch": 0.49, + "grad_norm": 3.276084993847712, + "learning_rate": 1.0740996125260122e-05, + "loss": 0.9561, + "step": 20888 + }, + { + "epoch": 0.49, + "grad_norm": 2.167017371618837, + "learning_rate": 1.0740235179761682e-05, + "loss": 1.0738, + "step": 20889 + }, + { + "epoch": 0.49, + "grad_norm": 2.082781176157595, + "learning_rate": 1.0739474229953362e-05, + "loss": 0.9869, + "step": 20890 + }, + { + "epoch": 0.49, + "grad_norm": 1.930898996269221, + "learning_rate": 1.073871327583959e-05, + "loss": 0.9376, + "step": 20891 + }, + { + "epoch": 0.49, + "grad_norm": 2.1226548896359776, + "learning_rate": 1.0737952317424789e-05, + "loss": 0.9993, + "step": 20892 + }, + { + "epoch": 0.49, + "grad_norm": 2.1175964060009944, + "learning_rate": 1.0737191354713403e-05, + "loss": 0.9522, + "step": 20893 + }, + { + "epoch": 0.49, + "grad_norm": 1.822029340638397, + "learning_rate": 1.0736430387709851e-05, + "loss": 0.9661, + "step": 20894 + }, + { + "epoch": 0.49, + "grad_norm": 2.432414933808578, + "learning_rate": 1.0735669416418571e-05, + "loss": 0.9872, + "step": 20895 + }, + { + "epoch": 0.49, + "grad_norm": 2.0827677857163063, + "learning_rate": 1.0734908440843987e-05, + "loss": 1.0557, + "step": 20896 + }, + { + "epoch": 0.49, + "grad_norm": 1.8547510008051438, + "learning_rate": 1.0734147460990539e-05, + "loss": 0.9603, + "step": 20897 + }, + { + "epoch": 0.49, + "grad_norm": 1.8178083205060274, + "learning_rate": 1.0733386476862649e-05, + "loss": 0.9966, + "step": 20898 + }, + { + "epoch": 0.49, + "grad_norm": 1.8944000159360972, + "learning_rate": 1.0732625488464749e-05, + "loss": 0.9076, + "step": 20899 + }, + { + "epoch": 0.49, + "grad_norm": 2.2231636038690374, + "learning_rate": 1.0731864495801272e-05, + "loss": 1.1517, + "step": 20900 + }, + { + "epoch": 0.49, + "grad_norm": 1.9766146698670901, + "learning_rate": 1.0731103498876647e-05, + "loss": 0.9285, + "step": 20901 + }, + { + "epoch": 0.49, + "grad_norm": 2.498239335970414, + "learning_rate": 1.0730342497695312e-05, + "loss": 1.0472, + "step": 20902 + }, + { + "epoch": 0.49, + "grad_norm": 2.042386748548233, + "learning_rate": 1.0729581492261684e-05, + "loss": 0.9921, + "step": 20903 + }, + { + "epoch": 0.49, + "grad_norm": 2.334886148851259, + "learning_rate": 1.072882048258021e-05, + "loss": 0.9567, + "step": 20904 + }, + { + "epoch": 0.49, + "grad_norm": 2.119314894893854, + "learning_rate": 1.0728059468655303e-05, + "loss": 0.9274, + "step": 20905 + }, + { + "epoch": 0.49, + "grad_norm": 1.8567554286593746, + "learning_rate": 1.0727298450491413e-05, + "loss": 1.0388, + "step": 20906 + }, + { + "epoch": 0.49, + "grad_norm": 2.112633875745471, + "learning_rate": 1.0726537428092956e-05, + "loss": 0.9777, + "step": 20907 + }, + { + "epoch": 0.49, + "grad_norm": 2.3579938466475205, + "learning_rate": 1.0725776401464366e-05, + "loss": 0.9388, + "step": 20908 + }, + { + "epoch": 0.49, + "grad_norm": 1.818589575654302, + "learning_rate": 1.0725015370610078e-05, + "loss": 1.0269, + "step": 20909 + }, + { + "epoch": 0.49, + "grad_norm": 1.954400999965531, + "learning_rate": 1.0724254335534524e-05, + "loss": 0.9527, + "step": 20910 + }, + { + "epoch": 0.49, + "grad_norm": 1.9500007843382225, + "learning_rate": 1.0723493296242129e-05, + "loss": 1.0544, + "step": 20911 + }, + { + "epoch": 0.49, + "grad_norm": 2.1357069118335503, + "learning_rate": 1.0722732252737327e-05, + "loss": 1.0715, + "step": 20912 + }, + { + "epoch": 0.49, + "grad_norm": 2.1154385028583587, + "learning_rate": 1.072197120502455e-05, + "loss": 1.1029, + "step": 20913 + }, + { + "epoch": 0.49, + "grad_norm": 2.1337406422470124, + "learning_rate": 1.0721210153108226e-05, + "loss": 1.0215, + "step": 20914 + }, + { + "epoch": 0.49, + "grad_norm": 2.451984488055117, + "learning_rate": 1.072044909699279e-05, + "loss": 1.0095, + "step": 20915 + }, + { + "epoch": 0.49, + "grad_norm": 2.0677859180702907, + "learning_rate": 1.0719688036682672e-05, + "loss": 1.03, + "step": 20916 + }, + { + "epoch": 0.49, + "grad_norm": 2.053666438565235, + "learning_rate": 1.07189269721823e-05, + "loss": 0.9927, + "step": 20917 + }, + { + "epoch": 0.49, + "grad_norm": 2.505895698076722, + "learning_rate": 1.0718165903496109e-05, + "loss": 1.0101, + "step": 20918 + }, + { + "epoch": 0.49, + "grad_norm": 3.081358434800146, + "learning_rate": 1.0717404830628529e-05, + "loss": 0.9958, + "step": 20919 + }, + { + "epoch": 0.49, + "grad_norm": 2.177313257992134, + "learning_rate": 1.071664375358399e-05, + "loss": 1.0459, + "step": 20920 + }, + { + "epoch": 0.49, + "grad_norm": 2.2020015150526913, + "learning_rate": 1.0715882672366926e-05, + "loss": 1.1432, + "step": 20921 + }, + { + "epoch": 0.49, + "grad_norm": 2.063352260645942, + "learning_rate": 1.0715121586981763e-05, + "loss": 1.1658, + "step": 20922 + }, + { + "epoch": 0.49, + "grad_norm": 1.993263779715245, + "learning_rate": 1.071436049743294e-05, + "loss": 1.0018, + "step": 20923 + }, + { + "epoch": 0.49, + "grad_norm": 2.3074315594156234, + "learning_rate": 1.0713599403724879e-05, + "loss": 0.941, + "step": 20924 + }, + { + "epoch": 0.49, + "grad_norm": 2.1032628498132624, + "learning_rate": 1.0712838305862023e-05, + "loss": 0.9663, + "step": 20925 + }, + { + "epoch": 0.49, + "grad_norm": 2.0734789921298296, + "learning_rate": 1.071207720384879e-05, + "loss": 1.01, + "step": 20926 + }, + { + "epoch": 0.49, + "grad_norm": 1.8197248831951356, + "learning_rate": 1.0711316097689623e-05, + "loss": 0.9809, + "step": 20927 + }, + { + "epoch": 0.49, + "grad_norm": 1.072149556206262, + "learning_rate": 1.0710554987388946e-05, + "loss": 1.0146, + "step": 20928 + }, + { + "epoch": 0.49, + "grad_norm": 1.1956385007428554, + "learning_rate": 1.0709793872951193e-05, + "loss": 1.0377, + "step": 20929 + }, + { + "epoch": 0.49, + "grad_norm": 2.0282456331918244, + "learning_rate": 1.0709032754380797e-05, + "loss": 1.0434, + "step": 20930 + }, + { + "epoch": 0.49, + "grad_norm": 1.8817298194958894, + "learning_rate": 1.0708271631682186e-05, + "loss": 0.8865, + "step": 20931 + }, + { + "epoch": 0.49, + "grad_norm": 2.279180884934167, + "learning_rate": 1.0707510504859795e-05, + "loss": 1.1061, + "step": 20932 + }, + { + "epoch": 0.49, + "grad_norm": 2.179138724658214, + "learning_rate": 1.070674937391805e-05, + "loss": 0.9539, + "step": 20933 + }, + { + "epoch": 0.49, + "grad_norm": 0.993928321032375, + "learning_rate": 1.0705988238861391e-05, + "loss": 0.9484, + "step": 20934 + }, + { + "epoch": 0.49, + "grad_norm": 2.003258793416171, + "learning_rate": 1.0705227099694242e-05, + "loss": 0.9858, + "step": 20935 + }, + { + "epoch": 0.49, + "grad_norm": 2.1072700208684845, + "learning_rate": 1.070446595642104e-05, + "loss": 1.0924, + "step": 20936 + }, + { + "epoch": 0.49, + "grad_norm": 1.9162230579249326, + "learning_rate": 1.0703704809046211e-05, + "loss": 1.017, + "step": 20937 + }, + { + "epoch": 0.49, + "grad_norm": 1.8172328101480606, + "learning_rate": 1.0702943657574193e-05, + "loss": 1.0515, + "step": 20938 + }, + { + "epoch": 0.49, + "grad_norm": 1.8243473669287544, + "learning_rate": 1.0702182502009412e-05, + "loss": 0.8792, + "step": 20939 + }, + { + "epoch": 0.49, + "grad_norm": 1.144384029875018, + "learning_rate": 1.0701421342356304e-05, + "loss": 0.9728, + "step": 20940 + }, + { + "epoch": 0.49, + "grad_norm": 1.0752591525388395, + "learning_rate": 1.0700660178619297e-05, + "loss": 0.892, + "step": 20941 + }, + { + "epoch": 0.49, + "grad_norm": 1.9376232130630922, + "learning_rate": 1.069989901080283e-05, + "loss": 1.0282, + "step": 20942 + }, + { + "epoch": 0.49, + "grad_norm": 1.6885471356824202, + "learning_rate": 1.0699137838911323e-05, + "loss": 0.9172, + "step": 20943 + }, + { + "epoch": 0.49, + "grad_norm": 1.9230314581574826, + "learning_rate": 1.0698376662949217e-05, + "loss": 0.985, + "step": 20944 + }, + { + "epoch": 0.49, + "grad_norm": 6.222419745425261, + "learning_rate": 1.0697615482920937e-05, + "loss": 0.8943, + "step": 20945 + }, + { + "epoch": 0.49, + "grad_norm": 1.957276644723573, + "learning_rate": 1.0696854298830925e-05, + "loss": 1.0382, + "step": 20946 + }, + { + "epoch": 0.49, + "grad_norm": 1.9595975249675706, + "learning_rate": 1.0696093110683603e-05, + "loss": 1.1093, + "step": 20947 + }, + { + "epoch": 0.49, + "grad_norm": 3.1831640631515117, + "learning_rate": 1.069533191848341e-05, + "loss": 1.0838, + "step": 20948 + }, + { + "epoch": 0.49, + "grad_norm": 2.1198504444645243, + "learning_rate": 1.0694570722234769e-05, + "loss": 1.0188, + "step": 20949 + }, + { + "epoch": 0.49, + "grad_norm": 2.089435935560384, + "learning_rate": 1.0693809521942118e-05, + "loss": 1.0023, + "step": 20950 + }, + { + "epoch": 0.49, + "grad_norm": 2.1590902284833837, + "learning_rate": 1.0693048317609893e-05, + "loss": 1.041, + "step": 20951 + }, + { + "epoch": 0.49, + "grad_norm": 2.2549304081522377, + "learning_rate": 1.0692287109242516e-05, + "loss": 0.9507, + "step": 20952 + }, + { + "epoch": 0.49, + "grad_norm": 2.622240994401825, + "learning_rate": 1.069152589684443e-05, + "loss": 0.9307, + "step": 20953 + }, + { + "epoch": 0.49, + "grad_norm": 1.9119275893612482, + "learning_rate": 1.0690764680420056e-05, + "loss": 1.0145, + "step": 20954 + }, + { + "epoch": 0.49, + "grad_norm": 2.095163081998153, + "learning_rate": 1.0690003459973836e-05, + "loss": 1.0717, + "step": 20955 + }, + { + "epoch": 0.49, + "grad_norm": 1.87574063576962, + "learning_rate": 1.0689242235510195e-05, + "loss": 1.0079, + "step": 20956 + }, + { + "epoch": 0.49, + "grad_norm": 2.2008023024756627, + "learning_rate": 1.0688481007033566e-05, + "loss": 1.1252, + "step": 20957 + }, + { + "epoch": 0.49, + "grad_norm": 1.849047758279029, + "learning_rate": 1.0687719774548384e-05, + "loss": 0.8884, + "step": 20958 + }, + { + "epoch": 0.49, + "grad_norm": 1.885562262527628, + "learning_rate": 1.068695853805908e-05, + "loss": 1.1105, + "step": 20959 + }, + { + "epoch": 0.49, + "grad_norm": 1.74546948803607, + "learning_rate": 1.0686197297570087e-05, + "loss": 1.0374, + "step": 20960 + }, + { + "epoch": 0.49, + "grad_norm": 2.3475334469386104, + "learning_rate": 1.0685436053085833e-05, + "loss": 0.91, + "step": 20961 + }, + { + "epoch": 0.49, + "grad_norm": 1.9674005532570176, + "learning_rate": 1.0684674804610756e-05, + "loss": 1.0378, + "step": 20962 + }, + { + "epoch": 0.49, + "grad_norm": 2.1098689259857717, + "learning_rate": 1.0683913552149286e-05, + "loss": 0.9995, + "step": 20963 + }, + { + "epoch": 0.49, + "grad_norm": 5.9848700531741255, + "learning_rate": 1.0683152295705853e-05, + "loss": 1.0286, + "step": 20964 + }, + { + "epoch": 0.49, + "grad_norm": 2.1478045954255993, + "learning_rate": 1.0682391035284892e-05, + "loss": 0.9263, + "step": 20965 + }, + { + "epoch": 0.49, + "grad_norm": 2.1129989337392576, + "learning_rate": 1.0681629770890832e-05, + "loss": 1.1138, + "step": 20966 + }, + { + "epoch": 0.49, + "grad_norm": 2.521954045222429, + "learning_rate": 1.068086850252811e-05, + "loss": 0.9573, + "step": 20967 + }, + { + "epoch": 0.49, + "grad_norm": 2.410253530158935, + "learning_rate": 1.0680107230201155e-05, + "loss": 0.9969, + "step": 20968 + }, + { + "epoch": 0.49, + "grad_norm": 1.9434979638566394, + "learning_rate": 1.06793459539144e-05, + "loss": 1.0186, + "step": 20969 + }, + { + "epoch": 0.49, + "grad_norm": 1.9720386991425527, + "learning_rate": 1.0678584673672279e-05, + "loss": 0.9554, + "step": 20970 + }, + { + "epoch": 0.49, + "grad_norm": 2.057858400798729, + "learning_rate": 1.0677823389479223e-05, + "loss": 0.8887, + "step": 20971 + }, + { + "epoch": 0.49, + "grad_norm": 1.934616018462558, + "learning_rate": 1.0677062101339666e-05, + "loss": 1.0129, + "step": 20972 + }, + { + "epoch": 0.49, + "grad_norm": 2.3206775111064446, + "learning_rate": 1.0676300809258035e-05, + "loss": 0.9366, + "step": 20973 + }, + { + "epoch": 0.49, + "grad_norm": 2.2268829387002156, + "learning_rate": 1.0675539513238774e-05, + "loss": 0.9112, + "step": 20974 + }, + { + "epoch": 0.49, + "grad_norm": 2.041620244815005, + "learning_rate": 1.06747782132863e-05, + "loss": 1.0417, + "step": 20975 + }, + { + "epoch": 0.49, + "grad_norm": 1.965370367872801, + "learning_rate": 1.0674016909405059e-05, + "loss": 0.8647, + "step": 20976 + }, + { + "epoch": 0.49, + "grad_norm": 1.9512558656233732, + "learning_rate": 1.0673255601599472e-05, + "loss": 1.1339, + "step": 20977 + }, + { + "epoch": 0.49, + "grad_norm": 2.108905967802084, + "learning_rate": 1.0672494289873985e-05, + "loss": 0.9675, + "step": 20978 + }, + { + "epoch": 0.49, + "grad_norm": 1.9547538776046816, + "learning_rate": 1.067173297423302e-05, + "loss": 1.0655, + "step": 20979 + }, + { + "epoch": 0.49, + "grad_norm": 2.4944798493518126, + "learning_rate": 1.0670971654681014e-05, + "loss": 1.0269, + "step": 20980 + }, + { + "epoch": 0.49, + "grad_norm": 2.365841670088715, + "learning_rate": 1.0670210331222397e-05, + "loss": 1.1206, + "step": 20981 + }, + { + "epoch": 0.49, + "grad_norm": 2.1694751999956714, + "learning_rate": 1.0669449003861603e-05, + "loss": 0.9818, + "step": 20982 + }, + { + "epoch": 0.49, + "grad_norm": 1.1994916713281298, + "learning_rate": 1.0668687672603066e-05, + "loss": 0.9685, + "step": 20983 + }, + { + "epoch": 0.49, + "grad_norm": 2.2168348769152972, + "learning_rate": 1.0667926337451217e-05, + "loss": 0.9683, + "step": 20984 + }, + { + "epoch": 0.49, + "grad_norm": 1.0998109137130727, + "learning_rate": 1.0667164998410491e-05, + "loss": 1.0104, + "step": 20985 + }, + { + "epoch": 0.49, + "grad_norm": 2.1109529128974946, + "learning_rate": 1.0666403655485318e-05, + "loss": 0.9687, + "step": 20986 + }, + { + "epoch": 0.49, + "grad_norm": 1.9163325961504742, + "learning_rate": 1.066564230868013e-05, + "loss": 0.9949, + "step": 20987 + }, + { + "epoch": 0.49, + "grad_norm": 1.9736009154108711, + "learning_rate": 1.0664880957999365e-05, + "loss": 1.0037, + "step": 20988 + }, + { + "epoch": 0.49, + "grad_norm": 2.786775976799367, + "learning_rate": 1.0664119603447451e-05, + "loss": 1.026, + "step": 20989 + }, + { + "epoch": 0.49, + "grad_norm": 1.8978689170972176, + "learning_rate": 1.0663358245028822e-05, + "loss": 1.0146, + "step": 20990 + }, + { + "epoch": 0.49, + "grad_norm": 2.2467103646628153, + "learning_rate": 1.0662596882747913e-05, + "loss": 0.8545, + "step": 20991 + }, + { + "epoch": 0.49, + "grad_norm": 1.9587847892874566, + "learning_rate": 1.0661835516609153e-05, + "loss": 1.0539, + "step": 20992 + }, + { + "epoch": 0.49, + "grad_norm": 2.4498128781633755, + "learning_rate": 1.066107414661698e-05, + "loss": 0.9864, + "step": 20993 + }, + { + "epoch": 0.49, + "grad_norm": 2.25784314678852, + "learning_rate": 1.066031277277582e-05, + "loss": 0.9691, + "step": 20994 + }, + { + "epoch": 0.49, + "grad_norm": 1.1328916760431809, + "learning_rate": 1.0659551395090115e-05, + "loss": 0.9252, + "step": 20995 + }, + { + "epoch": 0.49, + "grad_norm": 2.3314084476590082, + "learning_rate": 1.0658790013564289e-05, + "loss": 0.9549, + "step": 20996 + }, + { + "epoch": 0.49, + "grad_norm": 2.031181167160698, + "learning_rate": 1.0658028628202782e-05, + "loss": 0.9721, + "step": 20997 + }, + { + "epoch": 0.49, + "grad_norm": 1.957929995648105, + "learning_rate": 1.0657267239010021e-05, + "loss": 0.96, + "step": 20998 + }, + { + "epoch": 0.49, + "grad_norm": 2.1224757825116933, + "learning_rate": 1.0656505845990442e-05, + "loss": 1.0314, + "step": 20999 + }, + { + "epoch": 0.49, + "grad_norm": 3.8561112322436717, + "learning_rate": 1.0655744449148482e-05, + "loss": 1.0242, + "step": 21000 + }, + { + "epoch": 0.49, + "grad_norm": 1.0570396980369556, + "learning_rate": 1.0654983048488566e-05, + "loss": 0.8867, + "step": 21001 + }, + { + "epoch": 0.49, + "grad_norm": 2.4852916261569415, + "learning_rate": 1.0654221644015135e-05, + "loss": 1.0288, + "step": 21002 + }, + { + "epoch": 0.49, + "grad_norm": 1.9312251595707801, + "learning_rate": 1.0653460235732613e-05, + "loss": 1.0759, + "step": 21003 + }, + { + "epoch": 0.49, + "grad_norm": 3.3711925123446127, + "learning_rate": 1.0652698823645444e-05, + "loss": 0.9352, + "step": 21004 + }, + { + "epoch": 0.49, + "grad_norm": 2.3697314251021604, + "learning_rate": 1.0651937407758052e-05, + "loss": 0.9455, + "step": 21005 + }, + { + "epoch": 0.49, + "grad_norm": 1.917023214457376, + "learning_rate": 1.0651175988074877e-05, + "loss": 1.1411, + "step": 21006 + }, + { + "epoch": 0.49, + "grad_norm": 2.0320768313766138, + "learning_rate": 1.0650414564600349e-05, + "loss": 1.0871, + "step": 21007 + }, + { + "epoch": 0.49, + "grad_norm": 2.1291018798606727, + "learning_rate": 1.06496531373389e-05, + "loss": 0.9708, + "step": 21008 + }, + { + "epoch": 0.49, + "grad_norm": 1.0250149238404227, + "learning_rate": 1.0648891706294967e-05, + "loss": 1.0115, + "step": 21009 + }, + { + "epoch": 0.49, + "grad_norm": 1.0728751835878179, + "learning_rate": 1.0648130271472978e-05, + "loss": 0.999, + "step": 21010 + }, + { + "epoch": 0.5, + "grad_norm": 2.054501472054703, + "learning_rate": 1.0647368832877371e-05, + "loss": 0.9901, + "step": 21011 + }, + { + "epoch": 0.5, + "grad_norm": 1.1031087049153345, + "learning_rate": 1.0646607390512578e-05, + "loss": 0.9754, + "step": 21012 + }, + { + "epoch": 0.5, + "grad_norm": 2.072576108570089, + "learning_rate": 1.064584594438303e-05, + "loss": 0.863, + "step": 21013 + }, + { + "epoch": 0.5, + "grad_norm": 1.7763761833549285, + "learning_rate": 1.0645084494493166e-05, + "loss": 1.0145, + "step": 21014 + }, + { + "epoch": 0.5, + "grad_norm": 2.022546961836884, + "learning_rate": 1.0644323040847414e-05, + "loss": 1.1317, + "step": 21015 + }, + { + "epoch": 0.5, + "grad_norm": 1.995411852737547, + "learning_rate": 1.0643561583450208e-05, + "loss": 0.9725, + "step": 21016 + }, + { + "epoch": 0.5, + "grad_norm": 1.9650979358740044, + "learning_rate": 1.0642800122305983e-05, + "loss": 1.0228, + "step": 21017 + }, + { + "epoch": 0.5, + "grad_norm": 1.9708376990683942, + "learning_rate": 1.0642038657419174e-05, + "loss": 0.9576, + "step": 21018 + }, + { + "epoch": 0.5, + "grad_norm": 1.1434829939464355, + "learning_rate": 1.064127718879421e-05, + "loss": 1.0227, + "step": 21019 + }, + { + "epoch": 0.5, + "grad_norm": 1.9054360872642664, + "learning_rate": 1.0640515716435529e-05, + "loss": 1.0149, + "step": 21020 + }, + { + "epoch": 0.5, + "grad_norm": 2.399955285624443, + "learning_rate": 1.0639754240347564e-05, + "loss": 1.0586, + "step": 21021 + }, + { + "epoch": 0.5, + "grad_norm": 2.2375394223687186, + "learning_rate": 1.0638992760534743e-05, + "loss": 1.0301, + "step": 21022 + }, + { + "epoch": 0.5, + "grad_norm": 2.1870044946391527, + "learning_rate": 1.063823127700151e-05, + "loss": 0.9867, + "step": 21023 + }, + { + "epoch": 0.5, + "grad_norm": 1.8980090509949326, + "learning_rate": 1.0637469789752286e-05, + "loss": 1.1413, + "step": 21024 + }, + { + "epoch": 0.5, + "grad_norm": 2.102714072448271, + "learning_rate": 1.0636708298791515e-05, + "loss": 1.1574, + "step": 21025 + }, + { + "epoch": 0.5, + "grad_norm": 2.1581956303906527, + "learning_rate": 1.0635946804123625e-05, + "loss": 1.1906, + "step": 21026 + }, + { + "epoch": 0.5, + "grad_norm": 1.75924519771023, + "learning_rate": 1.0635185305753054e-05, + "loss": 0.9939, + "step": 21027 + }, + { + "epoch": 0.5, + "grad_norm": 2.0308308967947815, + "learning_rate": 1.0634423803684231e-05, + "loss": 0.9865, + "step": 21028 + }, + { + "epoch": 0.5, + "grad_norm": 2.160693322832319, + "learning_rate": 1.0633662297921595e-05, + "loss": 1.0288, + "step": 21029 + }, + { + "epoch": 0.5, + "grad_norm": 1.746085583544522, + "learning_rate": 1.0632900788469572e-05, + "loss": 1.0318, + "step": 21030 + }, + { + "epoch": 0.5, + "grad_norm": 2.265232237631317, + "learning_rate": 1.06321392753326e-05, + "loss": 1.1264, + "step": 21031 + }, + { + "epoch": 0.5, + "grad_norm": 1.902934486933602, + "learning_rate": 1.0631377758515116e-05, + "loss": 0.9778, + "step": 21032 + }, + { + "epoch": 0.5, + "grad_norm": 1.0747946849037295, + "learning_rate": 1.0630616238021549e-05, + "loss": 0.893, + "step": 21033 + }, + { + "epoch": 0.5, + "grad_norm": 2.2345442416045036, + "learning_rate": 1.0629854713856336e-05, + "loss": 0.9867, + "step": 21034 + }, + { + "epoch": 0.5, + "grad_norm": 1.9954683812337621, + "learning_rate": 1.0629093186023906e-05, + "loss": 1.0987, + "step": 21035 + }, + { + "epoch": 0.5, + "grad_norm": 2.1125811198879076, + "learning_rate": 1.06283316545287e-05, + "loss": 0.9626, + "step": 21036 + }, + { + "epoch": 0.5, + "grad_norm": 1.9897465958623897, + "learning_rate": 1.0627570119375148e-05, + "loss": 1.0473, + "step": 21037 + }, + { + "epoch": 0.5, + "grad_norm": 2.214697528251799, + "learning_rate": 1.0626808580567683e-05, + "loss": 1.0641, + "step": 21038 + }, + { + "epoch": 0.5, + "grad_norm": 2.0942106591998146, + "learning_rate": 1.0626047038110741e-05, + "loss": 0.9821, + "step": 21039 + }, + { + "epoch": 0.5, + "grad_norm": 2.448578321925535, + "learning_rate": 1.0625285492008752e-05, + "loss": 0.932, + "step": 21040 + }, + { + "epoch": 0.5, + "grad_norm": 1.1364489140663447, + "learning_rate": 1.0624523942266157e-05, + "loss": 0.9433, + "step": 21041 + }, + { + "epoch": 0.5, + "grad_norm": 2.0901730261298956, + "learning_rate": 1.0623762388887382e-05, + "loss": 1.0102, + "step": 21042 + }, + { + "epoch": 0.5, + "grad_norm": 2.0506209265393203, + "learning_rate": 1.0623000831876868e-05, + "loss": 1.1196, + "step": 21043 + }, + { + "epoch": 0.5, + "grad_norm": 2.8137416046094437, + "learning_rate": 1.0622239271239048e-05, + "loss": 0.9732, + "step": 21044 + }, + { + "epoch": 0.5, + "grad_norm": 2.2550993849855523, + "learning_rate": 1.0621477706978347e-05, + "loss": 0.9249, + "step": 21045 + }, + { + "epoch": 0.5, + "grad_norm": 2.0814387894719992, + "learning_rate": 1.062071613909921e-05, + "loss": 0.9299, + "step": 21046 + }, + { + "epoch": 0.5, + "grad_norm": 2.1847607339266806, + "learning_rate": 1.0619954567606066e-05, + "loss": 1.1717, + "step": 21047 + }, + { + "epoch": 0.5, + "grad_norm": 2.0836549714650254, + "learning_rate": 1.0619192992503351e-05, + "loss": 0.965, + "step": 21048 + }, + { + "epoch": 0.5, + "grad_norm": 1.9567086731771102, + "learning_rate": 1.0618431413795498e-05, + "loss": 1.0562, + "step": 21049 + }, + { + "epoch": 0.5, + "grad_norm": 1.9928980561164071, + "learning_rate": 1.0617669831486944e-05, + "loss": 1.0365, + "step": 21050 + }, + { + "epoch": 0.5, + "grad_norm": 2.0290301430346136, + "learning_rate": 1.0616908245582121e-05, + "loss": 1.0869, + "step": 21051 + }, + { + "epoch": 0.5, + "grad_norm": 1.8636304796229417, + "learning_rate": 1.0616146656085457e-05, + "loss": 1.0367, + "step": 21052 + }, + { + "epoch": 0.5, + "grad_norm": 2.0440494242688274, + "learning_rate": 1.0615385063001398e-05, + "loss": 1.0647, + "step": 21053 + }, + { + "epoch": 0.5, + "grad_norm": 1.1331910426816594, + "learning_rate": 1.0614623466334368e-05, + "loss": 0.9619, + "step": 21054 + }, + { + "epoch": 0.5, + "grad_norm": 2.0481049690578104, + "learning_rate": 1.0613861866088809e-05, + "loss": 1.0919, + "step": 21055 + }, + { + "epoch": 0.5, + "grad_norm": 2.35203870853765, + "learning_rate": 1.0613100262269148e-05, + "loss": 0.9194, + "step": 21056 + }, + { + "epoch": 0.5, + "grad_norm": 1.0846082599591984, + "learning_rate": 1.0612338654879827e-05, + "loss": 0.9482, + "step": 21057 + }, + { + "epoch": 0.5, + "grad_norm": 2.0122769414674835, + "learning_rate": 1.0611577043925276e-05, + "loss": 0.9276, + "step": 21058 + }, + { + "epoch": 0.5, + "grad_norm": 1.9024989574680702, + "learning_rate": 1.0610815429409928e-05, + "loss": 0.968, + "step": 21059 + }, + { + "epoch": 0.5, + "grad_norm": 2.421903596120873, + "learning_rate": 1.0610053811338222e-05, + "loss": 0.9667, + "step": 21060 + }, + { + "epoch": 0.5, + "grad_norm": 2.980722600761881, + "learning_rate": 1.0609292189714586e-05, + "loss": 0.9702, + "step": 21061 + }, + { + "epoch": 0.5, + "grad_norm": 1.8224369021731683, + "learning_rate": 1.060853056454346e-05, + "loss": 1.1896, + "step": 21062 + }, + { + "epoch": 0.5, + "grad_norm": 2.120423325657307, + "learning_rate": 1.0607768935829277e-05, + "loss": 0.8784, + "step": 21063 + }, + { + "epoch": 0.5, + "grad_norm": 2.0766125441922605, + "learning_rate": 1.0607007303576469e-05, + "loss": 1.0842, + "step": 21064 + }, + { + "epoch": 0.5, + "grad_norm": 1.8183639480981981, + "learning_rate": 1.0606245667789473e-05, + "loss": 1.0768, + "step": 21065 + }, + { + "epoch": 0.5, + "grad_norm": 1.9256076794975625, + "learning_rate": 1.0605484028472723e-05, + "loss": 1.0248, + "step": 21066 + }, + { + "epoch": 0.5, + "grad_norm": 2.2363953653163438, + "learning_rate": 1.0604722385630652e-05, + "loss": 1.1209, + "step": 21067 + }, + { + "epoch": 0.5, + "grad_norm": 1.0904601571053607, + "learning_rate": 1.0603960739267698e-05, + "loss": 1.0073, + "step": 21068 + }, + { + "epoch": 0.5, + "grad_norm": 1.9602364803530345, + "learning_rate": 1.0603199089388293e-05, + "loss": 0.9908, + "step": 21069 + }, + { + "epoch": 0.5, + "grad_norm": 1.8777720013328072, + "learning_rate": 1.060243743599687e-05, + "loss": 0.9775, + "step": 21070 + }, + { + "epoch": 0.5, + "grad_norm": 1.0633199418154067, + "learning_rate": 1.0601675779097868e-05, + "loss": 0.9017, + "step": 21071 + }, + { + "epoch": 0.5, + "grad_norm": 1.8774090344945644, + "learning_rate": 1.060091411869572e-05, + "loss": 0.9774, + "step": 21072 + }, + { + "epoch": 0.5, + "grad_norm": 1.924663896240382, + "learning_rate": 1.0600152454794853e-05, + "loss": 0.9731, + "step": 21073 + }, + { + "epoch": 0.5, + "grad_norm": 2.0677664789987324, + "learning_rate": 1.0599390787399717e-05, + "loss": 0.9319, + "step": 21074 + }, + { + "epoch": 0.5, + "grad_norm": 1.0836077229060284, + "learning_rate": 1.0598629116514733e-05, + "loss": 0.9027, + "step": 21075 + }, + { + "epoch": 0.5, + "grad_norm": 2.2372862272591614, + "learning_rate": 1.0597867442144343e-05, + "loss": 0.9094, + "step": 21076 + }, + { + "epoch": 0.5, + "grad_norm": 2.4216923337160554, + "learning_rate": 1.0597105764292976e-05, + "loss": 1.0886, + "step": 21077 + }, + { + "epoch": 0.5, + "grad_norm": 1.7566457066802819, + "learning_rate": 1.0596344082965076e-05, + "loss": 0.9937, + "step": 21078 + }, + { + "epoch": 0.5, + "grad_norm": 2.12296542947568, + "learning_rate": 1.0595582398165067e-05, + "loss": 1.0379, + "step": 21079 + }, + { + "epoch": 0.5, + "grad_norm": 1.07502920139165, + "learning_rate": 1.059482070989739e-05, + "loss": 0.909, + "step": 21080 + }, + { + "epoch": 0.5, + "grad_norm": 1.0276719534982637, + "learning_rate": 1.059405901816648e-05, + "loss": 0.944, + "step": 21081 + }, + { + "epoch": 0.5, + "grad_norm": 2.1765188710251704, + "learning_rate": 1.059329732297677e-05, + "loss": 0.9378, + "step": 21082 + }, + { + "epoch": 0.5, + "grad_norm": 1.8108148942277915, + "learning_rate": 1.0592535624332694e-05, + "loss": 1.0658, + "step": 21083 + }, + { + "epoch": 0.5, + "grad_norm": 2.089655310945467, + "learning_rate": 1.0591773922238685e-05, + "loss": 1.1506, + "step": 21084 + }, + { + "epoch": 0.5, + "grad_norm": 1.8839057327081388, + "learning_rate": 1.0591012216699189e-05, + "loss": 1.0324, + "step": 21085 + }, + { + "epoch": 0.5, + "grad_norm": 1.0698795943029162, + "learning_rate": 1.0590250507718624e-05, + "loss": 0.9697, + "step": 21086 + }, + { + "epoch": 0.5, + "grad_norm": 1.9583579661188906, + "learning_rate": 1.058948879530144e-05, + "loss": 1.161, + "step": 21087 + }, + { + "epoch": 0.5, + "grad_norm": 2.107256564618327, + "learning_rate": 1.0588727079452063e-05, + "loss": 0.92, + "step": 21088 + }, + { + "epoch": 0.5, + "grad_norm": 2.9866074822173077, + "learning_rate": 1.058796536017493e-05, + "loss": 0.9834, + "step": 21089 + }, + { + "epoch": 0.5, + "grad_norm": 2.123389456558004, + "learning_rate": 1.058720363747448e-05, + "loss": 1.0442, + "step": 21090 + }, + { + "epoch": 0.5, + "grad_norm": 1.8567938021016765, + "learning_rate": 1.0586441911355138e-05, + "loss": 1.0074, + "step": 21091 + }, + { + "epoch": 0.5, + "grad_norm": 2.185037893058709, + "learning_rate": 1.0585680181821351e-05, + "loss": 1.0937, + "step": 21092 + }, + { + "epoch": 0.5, + "grad_norm": 1.7755802381669716, + "learning_rate": 1.0584918448877547e-05, + "loss": 0.8618, + "step": 21093 + }, + { + "epoch": 0.5, + "grad_norm": 2.2440094127177255, + "learning_rate": 1.0584156712528161e-05, + "loss": 1.042, + "step": 21094 + }, + { + "epoch": 0.5, + "grad_norm": 1.9723374557083981, + "learning_rate": 1.0583394972777632e-05, + "loss": 0.9481, + "step": 21095 + }, + { + "epoch": 0.5, + "grad_norm": 5.088684762950983, + "learning_rate": 1.0582633229630393e-05, + "loss": 1.023, + "step": 21096 + }, + { + "epoch": 0.5, + "grad_norm": 1.7953406558565903, + "learning_rate": 1.0581871483090878e-05, + "loss": 1.0265, + "step": 21097 + }, + { + "epoch": 0.5, + "grad_norm": 2.600168871553518, + "learning_rate": 1.0581109733163522e-05, + "loss": 1.0872, + "step": 21098 + }, + { + "epoch": 0.5, + "grad_norm": 1.9790601395289182, + "learning_rate": 1.0580347979852763e-05, + "loss": 1.0351, + "step": 21099 + }, + { + "epoch": 0.5, + "grad_norm": 2.0249823038647374, + "learning_rate": 1.0579586223163032e-05, + "loss": 0.9686, + "step": 21100 + }, + { + "epoch": 0.5, + "grad_norm": 1.148327526597752, + "learning_rate": 1.0578824463098769e-05, + "loss": 0.8875, + "step": 21101 + }, + { + "epoch": 0.5, + "grad_norm": 3.816387544955094, + "learning_rate": 1.0578062699664406e-05, + "loss": 0.9509, + "step": 21102 + }, + { + "epoch": 0.5, + "grad_norm": 1.9887645177218385, + "learning_rate": 1.0577300932864378e-05, + "loss": 1.0514, + "step": 21103 + }, + { + "epoch": 0.5, + "grad_norm": 2.061734401736344, + "learning_rate": 1.0576539162703123e-05, + "loss": 0.9166, + "step": 21104 + }, + { + "epoch": 0.5, + "grad_norm": 2.1602094421252964, + "learning_rate": 1.0575777389185071e-05, + "loss": 1.0851, + "step": 21105 + }, + { + "epoch": 0.5, + "grad_norm": 2.0825888870380265, + "learning_rate": 1.0575015612314665e-05, + "loss": 0.9517, + "step": 21106 + }, + { + "epoch": 0.5, + "grad_norm": 1.9589358653701998, + "learning_rate": 1.0574253832096332e-05, + "loss": 0.9216, + "step": 21107 + }, + { + "epoch": 0.5, + "grad_norm": 2.3206327890406797, + "learning_rate": 1.0573492048534515e-05, + "loss": 0.9248, + "step": 21108 + }, + { + "epoch": 0.5, + "grad_norm": 1.9236302562076542, + "learning_rate": 1.0572730261633644e-05, + "loss": 1.1751, + "step": 21109 + }, + { + "epoch": 0.5, + "grad_norm": 2.473220494337937, + "learning_rate": 1.0571968471398157e-05, + "loss": 0.9342, + "step": 21110 + }, + { + "epoch": 0.5, + "grad_norm": 2.031971406064601, + "learning_rate": 1.0571206677832488e-05, + "loss": 1.0333, + "step": 21111 + }, + { + "epoch": 0.5, + "grad_norm": 2.079503648077314, + "learning_rate": 1.0570444880941074e-05, + "loss": 0.9803, + "step": 21112 + }, + { + "epoch": 0.5, + "grad_norm": 1.934149475505957, + "learning_rate": 1.0569683080728345e-05, + "loss": 1.0594, + "step": 21113 + }, + { + "epoch": 0.5, + "grad_norm": 1.999942893511235, + "learning_rate": 1.0568921277198744e-05, + "loss": 0.998, + "step": 21114 + }, + { + "epoch": 0.5, + "grad_norm": 2.0246182090698843, + "learning_rate": 1.0568159470356702e-05, + "loss": 0.8019, + "step": 21115 + }, + { + "epoch": 0.5, + "grad_norm": 2.053979961925126, + "learning_rate": 1.0567397660206658e-05, + "loss": 1.0745, + "step": 21116 + }, + { + "epoch": 0.5, + "grad_norm": 1.0887426252707628, + "learning_rate": 1.0566635846753044e-05, + "loss": 0.9931, + "step": 21117 + }, + { + "epoch": 0.5, + "grad_norm": 1.1949808499771282, + "learning_rate": 1.0565874030000296e-05, + "loss": 0.9931, + "step": 21118 + }, + { + "epoch": 0.5, + "grad_norm": 2.5865411716479163, + "learning_rate": 1.056511220995285e-05, + "loss": 0.9849, + "step": 21119 + }, + { + "epoch": 0.5, + "grad_norm": 2.927322506245312, + "learning_rate": 1.0564350386615143e-05, + "loss": 1.044, + "step": 21120 + }, + { + "epoch": 0.5, + "grad_norm": 2.0489533900195394, + "learning_rate": 1.0563588559991608e-05, + "loss": 0.949, + "step": 21121 + }, + { + "epoch": 0.5, + "grad_norm": 2.003267818572818, + "learning_rate": 1.0562826730086684e-05, + "loss": 1.1079, + "step": 21122 + }, + { + "epoch": 0.5, + "grad_norm": 2.2613772262481815, + "learning_rate": 1.0562064896904804e-05, + "loss": 1.0022, + "step": 21123 + }, + { + "epoch": 0.5, + "grad_norm": 1.8770669237495279, + "learning_rate": 1.0561303060450404e-05, + "loss": 1.0397, + "step": 21124 + }, + { + "epoch": 0.5, + "grad_norm": 2.1602160556966368, + "learning_rate": 1.056054122072792e-05, + "loss": 0.8913, + "step": 21125 + }, + { + "epoch": 0.5, + "grad_norm": 1.8458069754453654, + "learning_rate": 1.0559779377741786e-05, + "loss": 1.0168, + "step": 21126 + }, + { + "epoch": 0.5, + "grad_norm": 2.133249569481412, + "learning_rate": 1.055901753149644e-05, + "loss": 1.0465, + "step": 21127 + }, + { + "epoch": 0.5, + "grad_norm": 1.9798631873244283, + "learning_rate": 1.0558255681996316e-05, + "loss": 1.0299, + "step": 21128 + }, + { + "epoch": 0.5, + "grad_norm": 2.001005358695898, + "learning_rate": 1.0557493829245854e-05, + "loss": 1.0103, + "step": 21129 + }, + { + "epoch": 0.5, + "grad_norm": 1.7792430286838679, + "learning_rate": 1.0556731973249486e-05, + "loss": 1.0077, + "step": 21130 + }, + { + "epoch": 0.5, + "grad_norm": 2.270882287164466, + "learning_rate": 1.0555970114011644e-05, + "loss": 1.0463, + "step": 21131 + }, + { + "epoch": 0.5, + "grad_norm": 2.364030912801884, + "learning_rate": 1.055520825153677e-05, + "loss": 1.0657, + "step": 21132 + }, + { + "epoch": 0.5, + "grad_norm": 1.9524749079264516, + "learning_rate": 1.0554446385829297e-05, + "loss": 1.2674, + "step": 21133 + }, + { + "epoch": 0.5, + "grad_norm": 2.066624522737323, + "learning_rate": 1.0553684516893665e-05, + "loss": 1.0448, + "step": 21134 + }, + { + "epoch": 0.5, + "grad_norm": 1.0418436653788432, + "learning_rate": 1.0552922644734302e-05, + "loss": 0.9461, + "step": 21135 + }, + { + "epoch": 0.5, + "grad_norm": 1.8327475646553497, + "learning_rate": 1.0552160769355653e-05, + "loss": 0.9072, + "step": 21136 + }, + { + "epoch": 0.5, + "grad_norm": 1.09414945613834, + "learning_rate": 1.0551398890762148e-05, + "loss": 0.9786, + "step": 21137 + }, + { + "epoch": 0.5, + "grad_norm": 2.128417762765964, + "learning_rate": 1.0550637008958222e-05, + "loss": 1.1641, + "step": 21138 + }, + { + "epoch": 0.5, + "grad_norm": 2.0718353333832207, + "learning_rate": 1.0549875123948314e-05, + "loss": 1.0084, + "step": 21139 + }, + { + "epoch": 0.5, + "grad_norm": 1.9353742346266682, + "learning_rate": 1.0549113235736858e-05, + "loss": 1.0489, + "step": 21140 + }, + { + "epoch": 0.5, + "grad_norm": 1.9686059849437818, + "learning_rate": 1.0548351344328293e-05, + "loss": 0.9943, + "step": 21141 + }, + { + "epoch": 0.5, + "grad_norm": 1.8448567300342547, + "learning_rate": 1.0547589449727051e-05, + "loss": 0.9747, + "step": 21142 + }, + { + "epoch": 0.5, + "grad_norm": 1.8782349610876408, + "learning_rate": 1.054682755193757e-05, + "loss": 0.9899, + "step": 21143 + }, + { + "epoch": 0.5, + "grad_norm": 2.2696075654371426, + "learning_rate": 1.0546065650964288e-05, + "loss": 1.112, + "step": 21144 + }, + { + "epoch": 0.5, + "grad_norm": 2.0827151802525266, + "learning_rate": 1.0545303746811637e-05, + "loss": 1.0452, + "step": 21145 + }, + { + "epoch": 0.5, + "grad_norm": 1.701294047064216, + "learning_rate": 1.0544541839484055e-05, + "loss": 0.9123, + "step": 21146 + }, + { + "epoch": 0.5, + "grad_norm": 1.1090207419961107, + "learning_rate": 1.0543779928985978e-05, + "loss": 0.9231, + "step": 21147 + }, + { + "epoch": 0.5, + "grad_norm": 2.1292666884199005, + "learning_rate": 1.0543018015321842e-05, + "loss": 1.0222, + "step": 21148 + }, + { + "epoch": 0.5, + "grad_norm": 2.080611394694822, + "learning_rate": 1.0542256098496085e-05, + "loss": 1.0538, + "step": 21149 + }, + { + "epoch": 0.5, + "grad_norm": 2.0632974587074298, + "learning_rate": 1.0541494178513138e-05, + "loss": 1.1173, + "step": 21150 + }, + { + "epoch": 0.5, + "grad_norm": 2.3343000460875127, + "learning_rate": 1.0540732255377442e-05, + "loss": 0.9542, + "step": 21151 + }, + { + "epoch": 0.5, + "grad_norm": 1.9813778876597103, + "learning_rate": 1.0539970329093432e-05, + "loss": 0.9202, + "step": 21152 + }, + { + "epoch": 0.5, + "grad_norm": 2.1238801982418187, + "learning_rate": 1.0539208399665546e-05, + "loss": 0.9283, + "step": 21153 + }, + { + "epoch": 0.5, + "grad_norm": 2.5627155872006964, + "learning_rate": 1.0538446467098214e-05, + "loss": 1.1083, + "step": 21154 + }, + { + "epoch": 0.5, + "grad_norm": 2.518718203409438, + "learning_rate": 1.053768453139588e-05, + "loss": 1.0193, + "step": 21155 + }, + { + "epoch": 0.5, + "grad_norm": 2.1368433688905015, + "learning_rate": 1.0536922592562973e-05, + "loss": 1.0582, + "step": 21156 + }, + { + "epoch": 0.5, + "grad_norm": 1.054855778634152, + "learning_rate": 1.0536160650603935e-05, + "loss": 0.9188, + "step": 21157 + }, + { + "epoch": 0.5, + "grad_norm": 2.4409607582859265, + "learning_rate": 1.0535398705523197e-05, + "loss": 0.9963, + "step": 21158 + }, + { + "epoch": 0.5, + "grad_norm": 2.1803035510685844, + "learning_rate": 1.0534636757325203e-05, + "loss": 0.9251, + "step": 21159 + }, + { + "epoch": 0.5, + "grad_norm": 1.9649408540539608, + "learning_rate": 1.0533874806014382e-05, + "loss": 1.0155, + "step": 21160 + }, + { + "epoch": 0.5, + "grad_norm": 2.284014664996818, + "learning_rate": 1.053311285159517e-05, + "loss": 1.1951, + "step": 21161 + }, + { + "epoch": 0.5, + "grad_norm": 1.7912339659298162, + "learning_rate": 1.053235089407201e-05, + "loss": 1.0534, + "step": 21162 + }, + { + "epoch": 0.5, + "grad_norm": 2.0832195355459056, + "learning_rate": 1.0531588933449332e-05, + "loss": 1.0842, + "step": 21163 + }, + { + "epoch": 0.5, + "grad_norm": 2.125488469279163, + "learning_rate": 1.0530826969731577e-05, + "loss": 0.9487, + "step": 21164 + }, + { + "epoch": 0.5, + "grad_norm": 1.0663854641513022, + "learning_rate": 1.053006500292318e-05, + "loss": 0.9826, + "step": 21165 + }, + { + "epoch": 0.5, + "grad_norm": 1.9521118293495563, + "learning_rate": 1.0529303033028575e-05, + "loss": 1.1137, + "step": 21166 + }, + { + "epoch": 0.5, + "grad_norm": 1.929759513072567, + "learning_rate": 1.05285410600522e-05, + "loss": 0.9799, + "step": 21167 + }, + { + "epoch": 0.5, + "grad_norm": 2.065376106823005, + "learning_rate": 1.0527779083998492e-05, + "loss": 1.0568, + "step": 21168 + }, + { + "epoch": 0.5, + "grad_norm": 1.9265773603377856, + "learning_rate": 1.0527017104871885e-05, + "loss": 0.9302, + "step": 21169 + }, + { + "epoch": 0.5, + "grad_norm": 2.5313673570773654, + "learning_rate": 1.0526255122676823e-05, + "loss": 0.9759, + "step": 21170 + }, + { + "epoch": 0.5, + "grad_norm": 2.560531920419199, + "learning_rate": 1.0525493137417733e-05, + "loss": 1.0162, + "step": 21171 + }, + { + "epoch": 0.5, + "grad_norm": 2.3969510616169774, + "learning_rate": 1.0524731149099057e-05, + "loss": 0.9852, + "step": 21172 + }, + { + "epoch": 0.5, + "grad_norm": 1.8476285855508612, + "learning_rate": 1.0523969157725229e-05, + "loss": 1.086, + "step": 21173 + }, + { + "epoch": 0.5, + "grad_norm": 2.144538872432538, + "learning_rate": 1.0523207163300688e-05, + "loss": 0.9559, + "step": 21174 + }, + { + "epoch": 0.5, + "grad_norm": 1.9700734831098818, + "learning_rate": 1.0522445165829867e-05, + "loss": 1.0395, + "step": 21175 + }, + { + "epoch": 0.5, + "grad_norm": 1.9251749504757356, + "learning_rate": 1.0521683165317209e-05, + "loss": 0.9642, + "step": 21176 + }, + { + "epoch": 0.5, + "grad_norm": 1.8774526704244432, + "learning_rate": 1.0520921161767142e-05, + "loss": 1.0409, + "step": 21177 + }, + { + "epoch": 0.5, + "grad_norm": 1.9105023282553695, + "learning_rate": 1.0520159155184111e-05, + "loss": 1.0253, + "step": 21178 + }, + { + "epoch": 0.5, + "grad_norm": 1.7588611864804127, + "learning_rate": 1.0519397145572542e-05, + "loss": 0.9775, + "step": 21179 + }, + { + "epoch": 0.5, + "grad_norm": 2.2152671606354093, + "learning_rate": 1.0518635132936888e-05, + "loss": 0.9933, + "step": 21180 + }, + { + "epoch": 0.5, + "grad_norm": 2.665307805092022, + "learning_rate": 1.051787311728157e-05, + "loss": 1.1589, + "step": 21181 + }, + { + "epoch": 0.5, + "grad_norm": 3.344529214591957, + "learning_rate": 1.0517111098611028e-05, + "loss": 0.9572, + "step": 21182 + }, + { + "epoch": 0.5, + "grad_norm": 2.000067314746459, + "learning_rate": 1.0516349076929708e-05, + "loss": 0.9555, + "step": 21183 + }, + { + "epoch": 0.5, + "grad_norm": 2.1272026857893502, + "learning_rate": 1.0515587052242035e-05, + "loss": 1.022, + "step": 21184 + }, + { + "epoch": 0.5, + "grad_norm": 1.7873032135145004, + "learning_rate": 1.0514825024552456e-05, + "loss": 1.0074, + "step": 21185 + }, + { + "epoch": 0.5, + "grad_norm": 1.96538321955581, + "learning_rate": 1.0514062993865396e-05, + "loss": 0.9959, + "step": 21186 + }, + { + "epoch": 0.5, + "grad_norm": 1.104706015408081, + "learning_rate": 1.0513300960185307e-05, + "loss": 0.9561, + "step": 21187 + }, + { + "epoch": 0.5, + "grad_norm": 2.003966765112782, + "learning_rate": 1.0512538923516613e-05, + "loss": 1.0293, + "step": 21188 + }, + { + "epoch": 0.5, + "grad_norm": 3.6936559345616304, + "learning_rate": 1.0511776883863753e-05, + "loss": 0.9882, + "step": 21189 + }, + { + "epoch": 0.5, + "grad_norm": 1.9141665916335664, + "learning_rate": 1.0511014841231168e-05, + "loss": 0.9754, + "step": 21190 + }, + { + "epoch": 0.5, + "grad_norm": 2.102352077819993, + "learning_rate": 1.0510252795623294e-05, + "loss": 0.886, + "step": 21191 + }, + { + "epoch": 0.5, + "grad_norm": 1.9956629429589516, + "learning_rate": 1.0509490747044566e-05, + "loss": 1.0465, + "step": 21192 + }, + { + "epoch": 0.5, + "grad_norm": 2.112179616002112, + "learning_rate": 1.050872869549942e-05, + "loss": 1.0681, + "step": 21193 + }, + { + "epoch": 0.5, + "grad_norm": 1.9320439843493915, + "learning_rate": 1.0507966640992295e-05, + "loss": 0.9482, + "step": 21194 + }, + { + "epoch": 0.5, + "grad_norm": 2.040660474532404, + "learning_rate": 1.0507204583527629e-05, + "loss": 0.9511, + "step": 21195 + }, + { + "epoch": 0.5, + "grad_norm": 1.8733456439730005, + "learning_rate": 1.0506442523109857e-05, + "loss": 1.0639, + "step": 21196 + }, + { + "epoch": 0.5, + "grad_norm": 2.0546855172947285, + "learning_rate": 1.0505680459743414e-05, + "loss": 1.0528, + "step": 21197 + }, + { + "epoch": 0.5, + "grad_norm": 2.0946815621493053, + "learning_rate": 1.050491839343274e-05, + "loss": 0.9714, + "step": 21198 + }, + { + "epoch": 0.5, + "grad_norm": 2.1261011781692054, + "learning_rate": 1.0504156324182274e-05, + "loss": 1.0527, + "step": 21199 + }, + { + "epoch": 0.5, + "grad_norm": 2.1132010552214493, + "learning_rate": 1.0503394251996446e-05, + "loss": 1.0976, + "step": 21200 + }, + { + "epoch": 0.5, + "grad_norm": 2.1665842001846976, + "learning_rate": 1.0502632176879701e-05, + "loss": 1.0089, + "step": 21201 + }, + { + "epoch": 0.5, + "grad_norm": 2.1232857793467304, + "learning_rate": 1.0501870098836473e-05, + "loss": 1.0302, + "step": 21202 + }, + { + "epoch": 0.5, + "grad_norm": 1.9512177233803178, + "learning_rate": 1.0501108017871193e-05, + "loss": 1.0267, + "step": 21203 + }, + { + "epoch": 0.5, + "grad_norm": 1.9620841604333183, + "learning_rate": 1.050034593398831e-05, + "loss": 1.0653, + "step": 21204 + }, + { + "epoch": 0.5, + "grad_norm": 2.2888812576806776, + "learning_rate": 1.0499583847192249e-05, + "loss": 0.9414, + "step": 21205 + }, + { + "epoch": 0.5, + "grad_norm": 1.1756108485670724, + "learning_rate": 1.0498821757487456e-05, + "loss": 1.0342, + "step": 21206 + }, + { + "epoch": 0.5, + "grad_norm": 1.995993500751756, + "learning_rate": 1.0498059664878361e-05, + "loss": 0.9469, + "step": 21207 + }, + { + "epoch": 0.5, + "grad_norm": 2.35550652999732, + "learning_rate": 1.0497297569369412e-05, + "loss": 0.956, + "step": 21208 + }, + { + "epoch": 0.5, + "grad_norm": 1.8492703658343224, + "learning_rate": 1.0496535470965033e-05, + "loss": 1.217, + "step": 21209 + }, + { + "epoch": 0.5, + "grad_norm": 2.027794609923235, + "learning_rate": 1.049577336966967e-05, + "loss": 0.9323, + "step": 21210 + }, + { + "epoch": 0.5, + "grad_norm": 2.2654931153354565, + "learning_rate": 1.0495011265487758e-05, + "loss": 0.8496, + "step": 21211 + }, + { + "epoch": 0.5, + "grad_norm": 1.9772484623530289, + "learning_rate": 1.0494249158423733e-05, + "loss": 1.0825, + "step": 21212 + }, + { + "epoch": 0.5, + "grad_norm": 2.0578913482248606, + "learning_rate": 1.0493487048482031e-05, + "loss": 1.1215, + "step": 21213 + }, + { + "epoch": 0.5, + "grad_norm": 1.0887449952916841, + "learning_rate": 1.049272493566709e-05, + "loss": 0.927, + "step": 21214 + }, + { + "epoch": 0.5, + "grad_norm": 1.9041192679844434, + "learning_rate": 1.0491962819983351e-05, + "loss": 0.8842, + "step": 21215 + }, + { + "epoch": 0.5, + "grad_norm": 3.3109220496833642, + "learning_rate": 1.0491200701435252e-05, + "loss": 0.9572, + "step": 21216 + }, + { + "epoch": 0.5, + "grad_norm": 3.014806907005602, + "learning_rate": 1.0490438580027223e-05, + "loss": 1.0441, + "step": 21217 + }, + { + "epoch": 0.5, + "grad_norm": 1.8491744854569496, + "learning_rate": 1.0489676455763704e-05, + "loss": 1.0201, + "step": 21218 + }, + { + "epoch": 0.5, + "grad_norm": 1.8508339711273998, + "learning_rate": 1.0488914328649136e-05, + "loss": 1.0988, + "step": 21219 + }, + { + "epoch": 0.5, + "grad_norm": 1.960255065748109, + "learning_rate": 1.0488152198687955e-05, + "loss": 1.0282, + "step": 21220 + }, + { + "epoch": 0.5, + "grad_norm": 1.976435908873786, + "learning_rate": 1.0487390065884596e-05, + "loss": 0.9124, + "step": 21221 + }, + { + "epoch": 0.5, + "grad_norm": 1.7718687704390348, + "learning_rate": 1.0486627930243497e-05, + "loss": 0.9844, + "step": 21222 + }, + { + "epoch": 0.5, + "grad_norm": 2.0911115491855705, + "learning_rate": 1.0485865791769096e-05, + "loss": 1.0685, + "step": 21223 + }, + { + "epoch": 0.5, + "grad_norm": 1.8284342107706975, + "learning_rate": 1.0485103650465832e-05, + "loss": 0.9496, + "step": 21224 + }, + { + "epoch": 0.5, + "grad_norm": 3.6286531074623753, + "learning_rate": 1.0484341506338143e-05, + "loss": 1.0061, + "step": 21225 + }, + { + "epoch": 0.5, + "grad_norm": 1.9043178935602731, + "learning_rate": 1.0483579359390459e-05, + "loss": 0.9927, + "step": 21226 + }, + { + "epoch": 0.5, + "grad_norm": 2.21319653603165, + "learning_rate": 1.0482817209627228e-05, + "loss": 0.9238, + "step": 21227 + }, + { + "epoch": 0.5, + "grad_norm": 2.018290444227021, + "learning_rate": 1.0482055057052878e-05, + "loss": 0.9776, + "step": 21228 + }, + { + "epoch": 0.5, + "grad_norm": 2.0620578340219566, + "learning_rate": 1.0481292901671853e-05, + "loss": 0.9625, + "step": 21229 + }, + { + "epoch": 0.5, + "grad_norm": 2.2458672775243707, + "learning_rate": 1.0480530743488586e-05, + "loss": 1.0492, + "step": 21230 + }, + { + "epoch": 0.5, + "grad_norm": 2.0302053101115294, + "learning_rate": 1.0479768582507519e-05, + "loss": 0.8819, + "step": 21231 + }, + { + "epoch": 0.5, + "grad_norm": 2.127672319048297, + "learning_rate": 1.0479006418733091e-05, + "loss": 0.9663, + "step": 21232 + }, + { + "epoch": 0.5, + "grad_norm": 2.0542332227271363, + "learning_rate": 1.047824425216973e-05, + "loss": 1.1006, + "step": 21233 + }, + { + "epoch": 0.5, + "grad_norm": 2.3412795571703517, + "learning_rate": 1.0477482082821884e-05, + "loss": 0.97, + "step": 21234 + }, + { + "epoch": 0.5, + "grad_norm": 1.7675258317214373, + "learning_rate": 1.0476719910693981e-05, + "loss": 1.0937, + "step": 21235 + }, + { + "epoch": 0.5, + "grad_norm": 1.865699025393569, + "learning_rate": 1.0475957735790469e-05, + "loss": 0.8582, + "step": 21236 + }, + { + "epoch": 0.5, + "grad_norm": 2.283325243442334, + "learning_rate": 1.0475195558115776e-05, + "loss": 1.0425, + "step": 21237 + }, + { + "epoch": 0.5, + "grad_norm": 2.809381222718023, + "learning_rate": 1.0474433377674349e-05, + "loss": 1.059, + "step": 21238 + }, + { + "epoch": 0.5, + "grad_norm": 1.8988017092933023, + "learning_rate": 1.0473671194470618e-05, + "loss": 1.1563, + "step": 21239 + }, + { + "epoch": 0.5, + "grad_norm": 2.0215467552797226, + "learning_rate": 1.0472909008509023e-05, + "loss": 0.9248, + "step": 21240 + }, + { + "epoch": 0.5, + "grad_norm": 2.7095350786622787, + "learning_rate": 1.0472146819794003e-05, + "loss": 0.9661, + "step": 21241 + }, + { + "epoch": 0.5, + "grad_norm": 1.9659139023559176, + "learning_rate": 1.0471384628329993e-05, + "loss": 1.0013, + "step": 21242 + }, + { + "epoch": 0.5, + "grad_norm": 2.147393304055558, + "learning_rate": 1.0470622434121434e-05, + "loss": 1.016, + "step": 21243 + }, + { + "epoch": 0.5, + "grad_norm": 1.8794335102083777, + "learning_rate": 1.0469860237172759e-05, + "loss": 1.0403, + "step": 21244 + }, + { + "epoch": 0.5, + "grad_norm": 2.3575870208960796, + "learning_rate": 1.0469098037488415e-05, + "loss": 0.9896, + "step": 21245 + }, + { + "epoch": 0.5, + "grad_norm": 1.9741893817121494, + "learning_rate": 1.0468335835072829e-05, + "loss": 1.0883, + "step": 21246 + }, + { + "epoch": 0.5, + "grad_norm": 1.8089746748386792, + "learning_rate": 1.0467573629930445e-05, + "loss": 0.9615, + "step": 21247 + }, + { + "epoch": 0.5, + "grad_norm": 2.4301819638378066, + "learning_rate": 1.0466811422065699e-05, + "loss": 0.9845, + "step": 21248 + }, + { + "epoch": 0.5, + "grad_norm": 2.2111727667082186, + "learning_rate": 1.046604921148303e-05, + "loss": 0.8665, + "step": 21249 + }, + { + "epoch": 0.5, + "grad_norm": 2.0204724091626587, + "learning_rate": 1.0465286998186872e-05, + "loss": 0.9501, + "step": 21250 + }, + { + "epoch": 0.5, + "grad_norm": 1.8600229240859123, + "learning_rate": 1.0464524782181669e-05, + "loss": 1.0411, + "step": 21251 + }, + { + "epoch": 0.5, + "grad_norm": 2.110972792748998, + "learning_rate": 1.0463762563471854e-05, + "loss": 1.0954, + "step": 21252 + }, + { + "epoch": 0.5, + "grad_norm": 2.008761834786244, + "learning_rate": 1.0463000342061868e-05, + "loss": 1.0017, + "step": 21253 + }, + { + "epoch": 0.5, + "grad_norm": 1.9435351711581752, + "learning_rate": 1.0462238117956145e-05, + "loss": 1.0047, + "step": 21254 + }, + { + "epoch": 0.5, + "grad_norm": 1.9709587704027285, + "learning_rate": 1.046147589115913e-05, + "loss": 1.0371, + "step": 21255 + }, + { + "epoch": 0.5, + "grad_norm": 2.5092789793222607, + "learning_rate": 1.046071366167525e-05, + "loss": 1.1051, + "step": 21256 + }, + { + "epoch": 0.5, + "grad_norm": 1.1654353235305694, + "learning_rate": 1.0459951429508955e-05, + "loss": 0.923, + "step": 21257 + }, + { + "epoch": 0.5, + "grad_norm": 2.353934774430131, + "learning_rate": 1.0459189194664671e-05, + "loss": 0.8921, + "step": 21258 + }, + { + "epoch": 0.5, + "grad_norm": 1.9531219387030454, + "learning_rate": 1.045842695714685e-05, + "loss": 1.0297, + "step": 21259 + }, + { + "epoch": 0.5, + "grad_norm": 1.7065976688039182, + "learning_rate": 1.0457664716959916e-05, + "loss": 1.0275, + "step": 21260 + }, + { + "epoch": 0.5, + "grad_norm": 1.8792796094856943, + "learning_rate": 1.0456902474108315e-05, + "loss": 0.8668, + "step": 21261 + }, + { + "epoch": 0.5, + "grad_norm": 1.9831003098146722, + "learning_rate": 1.0456140228596484e-05, + "loss": 0.9577, + "step": 21262 + }, + { + "epoch": 0.5, + "grad_norm": 2.0148232631947143, + "learning_rate": 1.0455377980428858e-05, + "loss": 0.992, + "step": 21263 + }, + { + "epoch": 0.5, + "grad_norm": 2.525902875338188, + "learning_rate": 1.045461572960988e-05, + "loss": 0.9891, + "step": 21264 + }, + { + "epoch": 0.5, + "grad_norm": 2.250664287594736, + "learning_rate": 1.0453853476143982e-05, + "loss": 1.0622, + "step": 21265 + }, + { + "epoch": 0.5, + "grad_norm": 2.8856906777349765, + "learning_rate": 1.0453091220035607e-05, + "loss": 1.0016, + "step": 21266 + }, + { + "epoch": 0.5, + "grad_norm": 1.8543240737522468, + "learning_rate": 1.0452328961289192e-05, + "loss": 0.8585, + "step": 21267 + }, + { + "epoch": 0.5, + "grad_norm": 1.86350750036859, + "learning_rate": 1.0451566699909174e-05, + "loss": 0.9372, + "step": 21268 + }, + { + "epoch": 0.5, + "grad_norm": 2.091293890124496, + "learning_rate": 1.0450804435899993e-05, + "loss": 0.9879, + "step": 21269 + }, + { + "epoch": 0.5, + "grad_norm": 1.953824602356932, + "learning_rate": 1.0450042169266083e-05, + "loss": 1.0313, + "step": 21270 + }, + { + "epoch": 0.5, + "grad_norm": 2.069490272276309, + "learning_rate": 1.0449279900011888e-05, + "loss": 1.131, + "step": 21271 + }, + { + "epoch": 0.5, + "grad_norm": 1.8928859641669846, + "learning_rate": 1.044851762814184e-05, + "loss": 1.014, + "step": 21272 + }, + { + "epoch": 0.5, + "grad_norm": 2.8129640395092155, + "learning_rate": 1.0447755353660384e-05, + "loss": 1.0338, + "step": 21273 + }, + { + "epoch": 0.5, + "grad_norm": 1.8892797643965011, + "learning_rate": 1.0446993076571952e-05, + "loss": 1.0794, + "step": 21274 + }, + { + "epoch": 0.5, + "grad_norm": 1.8598819248760259, + "learning_rate": 1.0446230796880987e-05, + "loss": 1.0872, + "step": 21275 + }, + { + "epoch": 0.5, + "grad_norm": 2.0385460818513277, + "learning_rate": 1.0445468514591925e-05, + "loss": 1.0143, + "step": 21276 + }, + { + "epoch": 0.5, + "grad_norm": 1.9161041179300033, + "learning_rate": 1.0444706229709201e-05, + "loss": 1.0766, + "step": 21277 + }, + { + "epoch": 0.5, + "grad_norm": 2.2326968293647282, + "learning_rate": 1.044394394223726e-05, + "loss": 1.089, + "step": 21278 + }, + { + "epoch": 0.5, + "grad_norm": 2.06588297575146, + "learning_rate": 1.0443181652180534e-05, + "loss": 0.9299, + "step": 21279 + }, + { + "epoch": 0.5, + "grad_norm": 1.1547202404559287, + "learning_rate": 1.0442419359543465e-05, + "loss": 0.9577, + "step": 21280 + }, + { + "epoch": 0.5, + "grad_norm": 2.0673027658259877, + "learning_rate": 1.044165706433049e-05, + "loss": 1.0159, + "step": 21281 + }, + { + "epoch": 0.5, + "grad_norm": 1.9076946562349903, + "learning_rate": 1.0440894766546051e-05, + "loss": 1.056, + "step": 21282 + }, + { + "epoch": 0.5, + "grad_norm": 1.9418571181766209, + "learning_rate": 1.0440132466194582e-05, + "loss": 0.9964, + "step": 21283 + }, + { + "epoch": 0.5, + "grad_norm": 1.872251770031242, + "learning_rate": 1.0439370163280518e-05, + "loss": 0.9762, + "step": 21284 + }, + { + "epoch": 0.5, + "grad_norm": 2.1519838947901433, + "learning_rate": 1.0438607857808307e-05, + "loss": 0.9106, + "step": 21285 + }, + { + "epoch": 0.5, + "grad_norm": 2.0246304939413498, + "learning_rate": 1.0437845549782377e-05, + "loss": 1.0515, + "step": 21286 + }, + { + "epoch": 0.5, + "grad_norm": 2.4135916767341743, + "learning_rate": 1.0437083239207177e-05, + "loss": 1.0335, + "step": 21287 + }, + { + "epoch": 0.5, + "grad_norm": 2.1639808202987227, + "learning_rate": 1.0436320926087135e-05, + "loss": 0.9105, + "step": 21288 + }, + { + "epoch": 0.5, + "grad_norm": 1.9158303697903283, + "learning_rate": 1.0435558610426701e-05, + "loss": 0.9444, + "step": 21289 + }, + { + "epoch": 0.5, + "grad_norm": 1.8237549450775339, + "learning_rate": 1.0434796292230303e-05, + "loss": 1.1521, + "step": 21290 + }, + { + "epoch": 0.5, + "grad_norm": 1.857838310100712, + "learning_rate": 1.0434033971502381e-05, + "loss": 0.9063, + "step": 21291 + }, + { + "epoch": 0.5, + "grad_norm": 1.0918656887929081, + "learning_rate": 1.043327164824738e-05, + "loss": 1.0067, + "step": 21292 + }, + { + "epoch": 0.5, + "grad_norm": 2.127582301985049, + "learning_rate": 1.0432509322469732e-05, + "loss": 0.9798, + "step": 21293 + }, + { + "epoch": 0.5, + "grad_norm": 1.1197743305792955, + "learning_rate": 1.043174699417388e-05, + "loss": 1.027, + "step": 21294 + }, + { + "epoch": 0.5, + "grad_norm": 1.0752020699455864, + "learning_rate": 1.0430984663364256e-05, + "loss": 1.009, + "step": 21295 + }, + { + "epoch": 0.5, + "grad_norm": 1.9511707751478162, + "learning_rate": 1.0430222330045306e-05, + "loss": 1.0872, + "step": 21296 + }, + { + "epoch": 0.5, + "grad_norm": 1.9741483048887798, + "learning_rate": 1.0429459994221465e-05, + "loss": 0.9584, + "step": 21297 + }, + { + "epoch": 0.5, + "grad_norm": 2.322477267188422, + "learning_rate": 1.0428697655897172e-05, + "loss": 1.0106, + "step": 21298 + }, + { + "epoch": 0.5, + "grad_norm": 1.9040223371239091, + "learning_rate": 1.0427935315076864e-05, + "loss": 1.0699, + "step": 21299 + }, + { + "epoch": 0.5, + "grad_norm": 2.524679481933859, + "learning_rate": 1.0427172971764984e-05, + "loss": 0.9989, + "step": 21300 + }, + { + "epoch": 0.5, + "grad_norm": 1.9887279141579213, + "learning_rate": 1.0426410625965965e-05, + "loss": 0.9307, + "step": 21301 + }, + { + "epoch": 0.5, + "grad_norm": 1.8497230123275932, + "learning_rate": 1.0425648277684251e-05, + "loss": 1.0437, + "step": 21302 + }, + { + "epoch": 0.5, + "grad_norm": 2.1735552321306972, + "learning_rate": 1.0424885926924275e-05, + "loss": 0.9913, + "step": 21303 + }, + { + "epoch": 0.5, + "grad_norm": 2.238398661361317, + "learning_rate": 1.042412357369048e-05, + "loss": 0.9652, + "step": 21304 + }, + { + "epoch": 0.5, + "grad_norm": 2.0718920209803646, + "learning_rate": 1.0423361217987303e-05, + "loss": 1.0788, + "step": 21305 + }, + { + "epoch": 0.5, + "grad_norm": 2.1394905824727144, + "learning_rate": 1.0422598859819186e-05, + "loss": 1.0599, + "step": 21306 + }, + { + "epoch": 0.5, + "grad_norm": 1.9119313415801045, + "learning_rate": 1.0421836499190559e-05, + "loss": 1.0584, + "step": 21307 + }, + { + "epoch": 0.5, + "grad_norm": 1.8432436424278489, + "learning_rate": 1.0421074136105872e-05, + "loss": 0.9131, + "step": 21308 + }, + { + "epoch": 0.5, + "grad_norm": 2.0236023186559384, + "learning_rate": 1.0420311770569553e-05, + "loss": 1.0276, + "step": 21309 + }, + { + "epoch": 0.5, + "grad_norm": 1.0442723908441498, + "learning_rate": 1.0419549402586049e-05, + "loss": 0.9298, + "step": 21310 + }, + { + "epoch": 0.5, + "grad_norm": 1.7196347871079005, + "learning_rate": 1.0418787032159797e-05, + "loss": 0.9131, + "step": 21311 + }, + { + "epoch": 0.5, + "grad_norm": 2.0224211708241064, + "learning_rate": 1.041802465929523e-05, + "loss": 0.9138, + "step": 21312 + }, + { + "epoch": 0.5, + "grad_norm": 2.079262696195241, + "learning_rate": 1.0417262283996793e-05, + "loss": 0.9323, + "step": 21313 + }, + { + "epoch": 0.5, + "grad_norm": 1.824948180078204, + "learning_rate": 1.0416499906268922e-05, + "loss": 0.8412, + "step": 21314 + }, + { + "epoch": 0.5, + "grad_norm": 2.0910335407098875, + "learning_rate": 1.0415737526116059e-05, + "loss": 0.9318, + "step": 21315 + }, + { + "epoch": 0.5, + "grad_norm": 1.8175111461533788, + "learning_rate": 1.0414975143542636e-05, + "loss": 0.9798, + "step": 21316 + }, + { + "epoch": 0.5, + "grad_norm": 2.2952025019749627, + "learning_rate": 1.0414212758553102e-05, + "loss": 0.9112, + "step": 21317 + }, + { + "epoch": 0.5, + "grad_norm": 2.1614521174695747, + "learning_rate": 1.0413450371151888e-05, + "loss": 1.0625, + "step": 21318 + }, + { + "epoch": 0.5, + "grad_norm": 1.1053871506518018, + "learning_rate": 1.0412687981343435e-05, + "loss": 0.9216, + "step": 21319 + }, + { + "epoch": 0.5, + "grad_norm": 3.159112141873206, + "learning_rate": 1.0411925589132182e-05, + "loss": 0.8918, + "step": 21320 + }, + { + "epoch": 0.5, + "grad_norm": 2.209135430244867, + "learning_rate": 1.0411163194522566e-05, + "loss": 1.0304, + "step": 21321 + }, + { + "epoch": 0.5, + "grad_norm": 1.0615444225142632, + "learning_rate": 1.041040079751903e-05, + "loss": 0.9787, + "step": 21322 + }, + { + "epoch": 0.5, + "grad_norm": 1.916229116335898, + "learning_rate": 1.0409638398126009e-05, + "loss": 0.926, + "step": 21323 + }, + { + "epoch": 0.5, + "grad_norm": 2.0679612697972054, + "learning_rate": 1.0408875996347945e-05, + "loss": 0.9575, + "step": 21324 + }, + { + "epoch": 0.5, + "grad_norm": 2.1611668445180383, + "learning_rate": 1.0408113592189275e-05, + "loss": 1.0403, + "step": 21325 + }, + { + "epoch": 0.5, + "grad_norm": 2.485237418272232, + "learning_rate": 1.0407351185654437e-05, + "loss": 1.0619, + "step": 21326 + }, + { + "epoch": 0.5, + "grad_norm": 1.1493572504252538, + "learning_rate": 1.0406588776747875e-05, + "loss": 0.9506, + "step": 21327 + }, + { + "epoch": 0.5, + "grad_norm": 2.1773960317907193, + "learning_rate": 1.040582636547402e-05, + "loss": 0.9757, + "step": 21328 + }, + { + "epoch": 0.5, + "grad_norm": 1.974705039729325, + "learning_rate": 1.0405063951837318e-05, + "loss": 1.0282, + "step": 21329 + }, + { + "epoch": 0.5, + "grad_norm": 1.876404320561963, + "learning_rate": 1.0404301535842205e-05, + "loss": 1.0483, + "step": 21330 + }, + { + "epoch": 0.5, + "grad_norm": 2.1395795647615397, + "learning_rate": 1.040353911749312e-05, + "loss": 0.961, + "step": 21331 + }, + { + "epoch": 0.5, + "grad_norm": 5.3519796778797915, + "learning_rate": 1.0402776696794506e-05, + "loss": 0.9777, + "step": 21332 + }, + { + "epoch": 0.5, + "grad_norm": 1.8654472069870685, + "learning_rate": 1.040201427375079e-05, + "loss": 1.0695, + "step": 21333 + }, + { + "epoch": 0.5, + "grad_norm": 1.8911769681637713, + "learning_rate": 1.0401251848366429e-05, + "loss": 1.0368, + "step": 21334 + }, + { + "epoch": 0.5, + "grad_norm": 2.0996344996213256, + "learning_rate": 1.0400489420645848e-05, + "loss": 0.9762, + "step": 21335 + }, + { + "epoch": 0.5, + "grad_norm": 1.9325702720212785, + "learning_rate": 1.0399726990593493e-05, + "loss": 1.0119, + "step": 21336 + }, + { + "epoch": 0.5, + "grad_norm": 2.6427131395958168, + "learning_rate": 1.0398964558213796e-05, + "loss": 0.9532, + "step": 21337 + }, + { + "epoch": 0.5, + "grad_norm": 1.169390704298257, + "learning_rate": 1.0398202123511207e-05, + "loss": 0.9742, + "step": 21338 + }, + { + "epoch": 0.5, + "grad_norm": 2.14814697338669, + "learning_rate": 1.0397439686490154e-05, + "loss": 1.0054, + "step": 21339 + }, + { + "epoch": 0.5, + "grad_norm": 1.9890125618104282, + "learning_rate": 1.0396677247155087e-05, + "loss": 1.0762, + "step": 21340 + }, + { + "epoch": 0.5, + "grad_norm": 3.6556269245204596, + "learning_rate": 1.0395914805510437e-05, + "loss": 0.9341, + "step": 21341 + }, + { + "epoch": 0.5, + "grad_norm": 1.0685636132947731, + "learning_rate": 1.0395152361560645e-05, + "loss": 0.9782, + "step": 21342 + }, + { + "epoch": 0.5, + "grad_norm": 1.7911138783193865, + "learning_rate": 1.039438991531015e-05, + "loss": 0.9669, + "step": 21343 + }, + { + "epoch": 0.5, + "grad_norm": 1.9275356958982008, + "learning_rate": 1.0393627466763395e-05, + "loss": 1.0968, + "step": 21344 + }, + { + "epoch": 0.5, + "grad_norm": 1.9744376687793284, + "learning_rate": 1.0392865015924813e-05, + "loss": 0.8599, + "step": 21345 + }, + { + "epoch": 0.5, + "grad_norm": 1.805263732669254, + "learning_rate": 1.0392102562798847e-05, + "loss": 0.9413, + "step": 21346 + }, + { + "epoch": 0.5, + "grad_norm": 1.9467950509282403, + "learning_rate": 1.0391340107389938e-05, + "loss": 1.0048, + "step": 21347 + }, + { + "epoch": 0.5, + "grad_norm": 1.1449265172833998, + "learning_rate": 1.039057764970252e-05, + "loss": 0.9387, + "step": 21348 + }, + { + "epoch": 0.5, + "grad_norm": 2.0571036934863782, + "learning_rate": 1.0389815189741038e-05, + "loss": 0.9927, + "step": 21349 + }, + { + "epoch": 0.5, + "grad_norm": 1.898959750168125, + "learning_rate": 1.0389052727509928e-05, + "loss": 1.0337, + "step": 21350 + }, + { + "epoch": 0.5, + "grad_norm": 2.0191223703198684, + "learning_rate": 1.0388290263013628e-05, + "loss": 1.1258, + "step": 21351 + }, + { + "epoch": 0.5, + "grad_norm": 2.403397612741068, + "learning_rate": 1.038752779625658e-05, + "loss": 1.0121, + "step": 21352 + }, + { + "epoch": 0.5, + "grad_norm": 1.8224710007471367, + "learning_rate": 1.0386765327243222e-05, + "loss": 0.9711, + "step": 21353 + }, + { + "epoch": 0.5, + "grad_norm": 2.200985163120842, + "learning_rate": 1.0386002855977995e-05, + "loss": 1.075, + "step": 21354 + }, + { + "epoch": 0.5, + "grad_norm": 2.0610112894009807, + "learning_rate": 1.038524038246534e-05, + "loss": 1.0032, + "step": 21355 + }, + { + "epoch": 0.5, + "grad_norm": 1.1474084829962006, + "learning_rate": 1.0384477906709687e-05, + "loss": 0.9681, + "step": 21356 + }, + { + "epoch": 0.5, + "grad_norm": 1.0760808638376085, + "learning_rate": 1.0383715428715489e-05, + "loss": 0.8855, + "step": 21357 + }, + { + "epoch": 0.5, + "grad_norm": 1.8695649391134639, + "learning_rate": 1.038295294848717e-05, + "loss": 0.9707, + "step": 21358 + }, + { + "epoch": 0.5, + "grad_norm": 2.0711004953956262, + "learning_rate": 1.0382190466029186e-05, + "loss": 1.0687, + "step": 21359 + }, + { + "epoch": 0.5, + "grad_norm": 1.0446469486725782, + "learning_rate": 1.0381427981345961e-05, + "loss": 0.9418, + "step": 21360 + }, + { + "epoch": 0.5, + "grad_norm": 1.9727417975704131, + "learning_rate": 1.0380665494441949e-05, + "loss": 1.0477, + "step": 21361 + }, + { + "epoch": 0.5, + "grad_norm": 1.9030764992684894, + "learning_rate": 1.0379903005321576e-05, + "loss": 0.9434, + "step": 21362 + }, + { + "epoch": 0.5, + "grad_norm": 1.9098304662573704, + "learning_rate": 1.037914051398929e-05, + "loss": 1.027, + "step": 21363 + }, + { + "epoch": 0.5, + "grad_norm": 1.9882297053663178, + "learning_rate": 1.0378378020449527e-05, + "loss": 1.0193, + "step": 21364 + }, + { + "epoch": 0.5, + "grad_norm": 1.8521719618098562, + "learning_rate": 1.0377615524706726e-05, + "loss": 1.0217, + "step": 21365 + }, + { + "epoch": 0.5, + "grad_norm": 2.193456995788142, + "learning_rate": 1.0376853026765332e-05, + "loss": 1.189, + "step": 21366 + }, + { + "epoch": 0.5, + "grad_norm": 2.008020884302761, + "learning_rate": 1.0376090526629773e-05, + "loss": 1.0079, + "step": 21367 + }, + { + "epoch": 0.5, + "grad_norm": 2.076151925571589, + "learning_rate": 1.0375328024304505e-05, + "loss": 1.0283, + "step": 21368 + }, + { + "epoch": 0.5, + "grad_norm": 2.068779444172103, + "learning_rate": 1.0374565519793953e-05, + "loss": 0.9313, + "step": 21369 + }, + { + "epoch": 0.5, + "grad_norm": 1.0841521479102232, + "learning_rate": 1.0373803013102563e-05, + "loss": 0.9295, + "step": 21370 + }, + { + "epoch": 0.5, + "grad_norm": 1.9358814210934177, + "learning_rate": 1.0373040504234775e-05, + "loss": 0.9141, + "step": 21371 + }, + { + "epoch": 0.5, + "grad_norm": 2.0225029031091863, + "learning_rate": 1.0372277993195026e-05, + "loss": 1.1033, + "step": 21372 + }, + { + "epoch": 0.5, + "grad_norm": 2.4313770239402, + "learning_rate": 1.0371515479987755e-05, + "loss": 1.0044, + "step": 21373 + }, + { + "epoch": 0.5, + "grad_norm": 2.1088161591395385, + "learning_rate": 1.0370752964617405e-05, + "loss": 0.9622, + "step": 21374 + }, + { + "epoch": 0.5, + "grad_norm": 2.1283555358044604, + "learning_rate": 1.0369990447088415e-05, + "loss": 1.1131, + "step": 21375 + }, + { + "epoch": 0.5, + "grad_norm": 1.0729756649143651, + "learning_rate": 1.0369227927405223e-05, + "loss": 0.9947, + "step": 21376 + }, + { + "epoch": 0.5, + "grad_norm": 2.020464469749925, + "learning_rate": 1.0368465405572265e-05, + "loss": 0.9908, + "step": 21377 + }, + { + "epoch": 0.5, + "grad_norm": 1.9242345424822773, + "learning_rate": 1.0367702881593988e-05, + "loss": 1.1316, + "step": 21378 + }, + { + "epoch": 0.5, + "grad_norm": 1.1222208066776636, + "learning_rate": 1.0366940355474829e-05, + "loss": 1.0219, + "step": 21379 + }, + { + "epoch": 0.5, + "grad_norm": 1.1381081096226855, + "learning_rate": 1.0366177827219227e-05, + "loss": 0.963, + "step": 21380 + }, + { + "epoch": 0.5, + "grad_norm": 1.1273885166645723, + "learning_rate": 1.036541529683162e-05, + "loss": 0.8688, + "step": 21381 + }, + { + "epoch": 0.5, + "grad_norm": 4.779826523265414, + "learning_rate": 1.036465276431645e-05, + "loss": 1.0498, + "step": 21382 + }, + { + "epoch": 0.5, + "grad_norm": 1.9896669570421006, + "learning_rate": 1.036389022967816e-05, + "loss": 1.0002, + "step": 21383 + }, + { + "epoch": 0.5, + "grad_norm": 2.24733729208427, + "learning_rate": 1.036312769292118e-05, + "loss": 1.0498, + "step": 21384 + }, + { + "epoch": 0.5, + "grad_norm": 2.158708663727782, + "learning_rate": 1.0362365154049959e-05, + "loss": 1.0062, + "step": 21385 + }, + { + "epoch": 0.5, + "grad_norm": 1.7416501423311725, + "learning_rate": 1.036160261306893e-05, + "loss": 1.0562, + "step": 21386 + }, + { + "epoch": 0.5, + "grad_norm": 3.834792037743508, + "learning_rate": 1.036084006998254e-05, + "loss": 1.1098, + "step": 21387 + }, + { + "epoch": 0.5, + "grad_norm": 1.086016121898096, + "learning_rate": 1.0360077524795222e-05, + "loss": 0.9114, + "step": 21388 + }, + { + "epoch": 0.5, + "grad_norm": 1.9342562831244987, + "learning_rate": 1.0359314977511423e-05, + "loss": 1.137, + "step": 21389 + }, + { + "epoch": 0.5, + "grad_norm": 2.655106994895118, + "learning_rate": 1.0358552428135576e-05, + "loss": 1.1281, + "step": 21390 + }, + { + "epoch": 0.5, + "grad_norm": 1.0938482456488359, + "learning_rate": 1.0357789876672122e-05, + "loss": 1.0466, + "step": 21391 + }, + { + "epoch": 0.5, + "grad_norm": 1.1273626309906783, + "learning_rate": 1.0357027323125503e-05, + "loss": 0.8635, + "step": 21392 + }, + { + "epoch": 0.5, + "grad_norm": 1.0495484831250455, + "learning_rate": 1.0356264767500158e-05, + "loss": 0.9727, + "step": 21393 + }, + { + "epoch": 0.5, + "grad_norm": 1.9596259513206051, + "learning_rate": 1.0355502209800525e-05, + "loss": 1.1123, + "step": 21394 + }, + { + "epoch": 0.5, + "grad_norm": 1.9471064366668704, + "learning_rate": 1.0354739650031047e-05, + "loss": 1.0169, + "step": 21395 + }, + { + "epoch": 0.5, + "grad_norm": 1.9306738300394783, + "learning_rate": 1.0353977088196164e-05, + "loss": 0.9741, + "step": 21396 + }, + { + "epoch": 0.5, + "grad_norm": 2.1587596763656345, + "learning_rate": 1.0353214524300312e-05, + "loss": 1.07, + "step": 21397 + }, + { + "epoch": 0.5, + "grad_norm": 2.160910100726123, + "learning_rate": 1.0352451958347936e-05, + "loss": 1.0628, + "step": 21398 + }, + { + "epoch": 0.5, + "grad_norm": 1.1005503936238994, + "learning_rate": 1.035168939034347e-05, + "loss": 0.9295, + "step": 21399 + }, + { + "epoch": 0.5, + "grad_norm": 1.9310814781786354, + "learning_rate": 1.0350926820291359e-05, + "loss": 1.125, + "step": 21400 + }, + { + "epoch": 0.5, + "grad_norm": 2.026617260404837, + "learning_rate": 1.035016424819604e-05, + "loss": 1.0419, + "step": 21401 + }, + { + "epoch": 0.5, + "grad_norm": 1.8921136498953237, + "learning_rate": 1.0349401674061952e-05, + "loss": 1.1078, + "step": 21402 + }, + { + "epoch": 0.5, + "grad_norm": 1.8794289182161819, + "learning_rate": 1.034863909789354e-05, + "loss": 0.9345, + "step": 21403 + }, + { + "epoch": 0.5, + "grad_norm": 1.9251782600562497, + "learning_rate": 1.034787651969524e-05, + "loss": 1.1031, + "step": 21404 + }, + { + "epoch": 0.5, + "grad_norm": 2.0785306849612466, + "learning_rate": 1.0347113939471491e-05, + "loss": 0.9016, + "step": 21405 + }, + { + "epoch": 0.5, + "grad_norm": 1.8552757423805017, + "learning_rate": 1.0346351357226737e-05, + "loss": 0.9087, + "step": 21406 + }, + { + "epoch": 0.5, + "grad_norm": 2.039992361141347, + "learning_rate": 1.0345588772965413e-05, + "loss": 0.9092, + "step": 21407 + }, + { + "epoch": 0.5, + "grad_norm": 2.1543307301722123, + "learning_rate": 1.0344826186691968e-05, + "loss": 1.0556, + "step": 21408 + }, + { + "epoch": 0.5, + "grad_norm": 1.9540300568266291, + "learning_rate": 1.0344063598410829e-05, + "loss": 0.9591, + "step": 21409 + }, + { + "epoch": 0.5, + "grad_norm": 1.8268424540060353, + "learning_rate": 1.0343301008126447e-05, + "loss": 1.0839, + "step": 21410 + }, + { + "epoch": 0.5, + "grad_norm": 2.1764463388593698, + "learning_rate": 1.0342538415843255e-05, + "loss": 1.159, + "step": 21411 + }, + { + "epoch": 0.5, + "grad_norm": 2.154357038699253, + "learning_rate": 1.0341775821565699e-05, + "loss": 0.9162, + "step": 21412 + }, + { + "epoch": 0.5, + "grad_norm": 1.767630557078196, + "learning_rate": 1.0341013225298213e-05, + "loss": 1.0103, + "step": 21413 + }, + { + "epoch": 0.5, + "grad_norm": 2.2546233446708297, + "learning_rate": 1.0340250627045238e-05, + "loss": 0.9643, + "step": 21414 + }, + { + "epoch": 0.5, + "grad_norm": 2.701282015328561, + "learning_rate": 1.0339488026811222e-05, + "loss": 0.9224, + "step": 21415 + }, + { + "epoch": 0.5, + "grad_norm": 2.4911031372585772, + "learning_rate": 1.0338725424600596e-05, + "loss": 1.0307, + "step": 21416 + }, + { + "epoch": 0.5, + "grad_norm": 2.2311297938613643, + "learning_rate": 1.0337962820417804e-05, + "loss": 1.1502, + "step": 21417 + }, + { + "epoch": 0.5, + "grad_norm": 1.8637337252700557, + "learning_rate": 1.0337200214267282e-05, + "loss": 0.9782, + "step": 21418 + }, + { + "epoch": 0.5, + "grad_norm": 2.325248081312639, + "learning_rate": 1.033643760615348e-05, + "loss": 0.9688, + "step": 21419 + }, + { + "epoch": 0.5, + "grad_norm": 2.1513723303222143, + "learning_rate": 1.033567499608083e-05, + "loss": 0.8752, + "step": 21420 + }, + { + "epoch": 0.5, + "grad_norm": 2.694302641478622, + "learning_rate": 1.033491238405377e-05, + "loss": 0.9403, + "step": 21421 + }, + { + "epoch": 0.5, + "grad_norm": 1.8186915881128982, + "learning_rate": 1.0334149770076747e-05, + "loss": 0.9965, + "step": 21422 + }, + { + "epoch": 0.5, + "grad_norm": 1.9679867444842714, + "learning_rate": 1.0333387154154197e-05, + "loss": 0.9988, + "step": 21423 + }, + { + "epoch": 0.5, + "grad_norm": 1.0986854084988344, + "learning_rate": 1.0332624536290563e-05, + "loss": 1.0164, + "step": 21424 + }, + { + "epoch": 0.5, + "grad_norm": 1.1308054313039924, + "learning_rate": 1.0331861916490283e-05, + "loss": 1.0342, + "step": 21425 + }, + { + "epoch": 0.5, + "grad_norm": 2.08480255052438, + "learning_rate": 1.0331099294757798e-05, + "loss": 1.0625, + "step": 21426 + }, + { + "epoch": 0.5, + "grad_norm": 1.9414415262747624, + "learning_rate": 1.0330336671097547e-05, + "loss": 1.077, + "step": 21427 + }, + { + "epoch": 0.5, + "grad_norm": 2.288088172405973, + "learning_rate": 1.0329574045513973e-05, + "loss": 1.1974, + "step": 21428 + }, + { + "epoch": 0.5, + "grad_norm": 1.862560696763438, + "learning_rate": 1.0328811418011514e-05, + "loss": 0.9524, + "step": 21429 + }, + { + "epoch": 0.5, + "grad_norm": 2.517265923240085, + "learning_rate": 1.0328048788594609e-05, + "loss": 1.0181, + "step": 21430 + }, + { + "epoch": 0.5, + "grad_norm": 1.9203083579116322, + "learning_rate": 1.0327286157267701e-05, + "loss": 0.9521, + "step": 21431 + }, + { + "epoch": 0.5, + "grad_norm": 1.9423038499939405, + "learning_rate": 1.032652352403523e-05, + "loss": 1.0156, + "step": 21432 + }, + { + "epoch": 0.5, + "grad_norm": 1.9944038432228346, + "learning_rate": 1.0325760888901638e-05, + "loss": 0.8924, + "step": 21433 + }, + { + "epoch": 0.5, + "grad_norm": 1.1108384470878456, + "learning_rate": 1.032499825187136e-05, + "loss": 0.9684, + "step": 21434 + }, + { + "epoch": 0.5, + "grad_norm": 1.9468099033110058, + "learning_rate": 1.0324235612948839e-05, + "loss": 1.0527, + "step": 21435 + }, + { + "epoch": 0.51, + "grad_norm": 1.7792929102071477, + "learning_rate": 1.0323472972138518e-05, + "loss": 0.9965, + "step": 21436 + }, + { + "epoch": 0.51, + "grad_norm": 2.169403027339113, + "learning_rate": 1.0322710329444831e-05, + "loss": 1.1256, + "step": 21437 + }, + { + "epoch": 0.51, + "grad_norm": 1.9237141209918642, + "learning_rate": 1.0321947684872227e-05, + "loss": 1.1082, + "step": 21438 + }, + { + "epoch": 0.51, + "grad_norm": 1.9186289657575801, + "learning_rate": 1.0321185038425137e-05, + "loss": 1.0237, + "step": 21439 + }, + { + "epoch": 0.51, + "grad_norm": 2.0018348604253675, + "learning_rate": 1.032042239010801e-05, + "loss": 1.0263, + "step": 21440 + }, + { + "epoch": 0.51, + "grad_norm": 2.246464708109001, + "learning_rate": 1.031965973992528e-05, + "loss": 1.0012, + "step": 21441 + }, + { + "epoch": 0.51, + "grad_norm": 2.0522522904681204, + "learning_rate": 1.0318897087881392e-05, + "loss": 1.0325, + "step": 21442 + }, + { + "epoch": 0.51, + "grad_norm": 2.3795113421355434, + "learning_rate": 1.0318134433980781e-05, + "loss": 0.9965, + "step": 21443 + }, + { + "epoch": 0.51, + "grad_norm": 2.048692767642205, + "learning_rate": 1.0317371778227893e-05, + "loss": 1.0576, + "step": 21444 + }, + { + "epoch": 0.51, + "grad_norm": 2.0854626651572468, + "learning_rate": 1.0316609120627165e-05, + "loss": 1.0399, + "step": 21445 + }, + { + "epoch": 0.51, + "grad_norm": 2.341511816308217, + "learning_rate": 1.0315846461183039e-05, + "loss": 0.9336, + "step": 21446 + }, + { + "epoch": 0.51, + "grad_norm": 2.021395831944747, + "learning_rate": 1.0315083799899957e-05, + "loss": 0.9355, + "step": 21447 + }, + { + "epoch": 0.51, + "grad_norm": 1.9132388909609277, + "learning_rate": 1.0314321136782354e-05, + "loss": 0.9251, + "step": 21448 + }, + { + "epoch": 0.51, + "grad_norm": 1.9958487896232207, + "learning_rate": 1.0313558471834674e-05, + "loss": 1.1454, + "step": 21449 + }, + { + "epoch": 0.51, + "grad_norm": 1.832278244656774, + "learning_rate": 1.031279580506136e-05, + "loss": 0.9881, + "step": 21450 + }, + { + "epoch": 0.51, + "grad_norm": 2.167587585921822, + "learning_rate": 1.0312033136466848e-05, + "loss": 1.0973, + "step": 21451 + }, + { + "epoch": 0.51, + "grad_norm": 2.017879603656488, + "learning_rate": 1.0311270466055578e-05, + "loss": 0.9878, + "step": 21452 + }, + { + "epoch": 0.51, + "grad_norm": 1.1188826879540974, + "learning_rate": 1.0310507793831997e-05, + "loss": 0.9349, + "step": 21453 + }, + { + "epoch": 0.51, + "grad_norm": 2.246027258531501, + "learning_rate": 1.0309745119800539e-05, + "loss": 1.021, + "step": 21454 + }, + { + "epoch": 0.51, + "grad_norm": 2.1647306659446617, + "learning_rate": 1.030898244396565e-05, + "loss": 1.0255, + "step": 21455 + }, + { + "epoch": 0.51, + "grad_norm": 1.083207963600969, + "learning_rate": 1.0308219766331762e-05, + "loss": 0.9881, + "step": 21456 + }, + { + "epoch": 0.51, + "grad_norm": 2.070420948035705, + "learning_rate": 1.0307457086903325e-05, + "loss": 0.936, + "step": 21457 + }, + { + "epoch": 0.51, + "grad_norm": 1.0283913238030638, + "learning_rate": 1.0306694405684771e-05, + "loss": 0.9599, + "step": 21458 + }, + { + "epoch": 0.51, + "grad_norm": 2.079402579160887, + "learning_rate": 1.030593172268055e-05, + "loss": 1.0421, + "step": 21459 + }, + { + "epoch": 0.51, + "grad_norm": 1.8754248389235055, + "learning_rate": 1.0305169037895095e-05, + "loss": 1.0603, + "step": 21460 + }, + { + "epoch": 0.51, + "grad_norm": 1.95975805013199, + "learning_rate": 1.0304406351332852e-05, + "loss": 1.1396, + "step": 21461 + }, + { + "epoch": 0.51, + "grad_norm": 2.07225110472959, + "learning_rate": 1.0303643662998255e-05, + "loss": 0.8882, + "step": 21462 + }, + { + "epoch": 0.51, + "grad_norm": 3.579908916194132, + "learning_rate": 1.0302880972895751e-05, + "loss": 0.9935, + "step": 21463 + }, + { + "epoch": 0.51, + "grad_norm": 2.1201241027257134, + "learning_rate": 1.030211828102978e-05, + "loss": 0.9981, + "step": 21464 + }, + { + "epoch": 0.51, + "grad_norm": 1.9688626425195315, + "learning_rate": 1.0301355587404775e-05, + "loss": 1.0307, + "step": 21465 + }, + { + "epoch": 0.51, + "grad_norm": 2.3390617063741734, + "learning_rate": 1.0300592892025187e-05, + "loss": 1.0265, + "step": 21466 + }, + { + "epoch": 0.51, + "grad_norm": 2.4833692250519355, + "learning_rate": 1.029983019489545e-05, + "loss": 1.0952, + "step": 21467 + }, + { + "epoch": 0.51, + "grad_norm": 2.1963347741488635, + "learning_rate": 1.029906749602001e-05, + "loss": 1.0072, + "step": 21468 + }, + { + "epoch": 0.51, + "grad_norm": 1.8813881185973844, + "learning_rate": 1.0298304795403298e-05, + "loss": 0.9389, + "step": 21469 + }, + { + "epoch": 0.51, + "grad_norm": 1.0943816221758207, + "learning_rate": 1.0297542093049767e-05, + "loss": 0.9546, + "step": 21470 + }, + { + "epoch": 0.51, + "grad_norm": 2.1429150507921535, + "learning_rate": 1.0296779388963849e-05, + "loss": 0.9799, + "step": 21471 + }, + { + "epoch": 0.51, + "grad_norm": 1.077394306104542, + "learning_rate": 1.0296016683149988e-05, + "loss": 0.9897, + "step": 21472 + }, + { + "epoch": 0.51, + "grad_norm": 2.0595235069757147, + "learning_rate": 1.0295253975612624e-05, + "loss": 0.9797, + "step": 21473 + }, + { + "epoch": 0.51, + "grad_norm": 1.8455966244478643, + "learning_rate": 1.02944912663562e-05, + "loss": 1.0866, + "step": 21474 + }, + { + "epoch": 0.51, + "grad_norm": 1.9774218052373906, + "learning_rate": 1.0293728555385149e-05, + "loss": 0.9019, + "step": 21475 + }, + { + "epoch": 0.51, + "grad_norm": 1.9586466646007992, + "learning_rate": 1.0292965842703922e-05, + "loss": 1.0211, + "step": 21476 + }, + { + "epoch": 0.51, + "grad_norm": 1.9921670955893942, + "learning_rate": 1.0292203128316953e-05, + "loss": 0.9926, + "step": 21477 + }, + { + "epoch": 0.51, + "grad_norm": 2.141343493683245, + "learning_rate": 1.0291440412228688e-05, + "loss": 1.1083, + "step": 21478 + }, + { + "epoch": 0.51, + "grad_norm": 1.088694473754233, + "learning_rate": 1.029067769444356e-05, + "loss": 0.9792, + "step": 21479 + }, + { + "epoch": 0.51, + "grad_norm": 1.9552414059694516, + "learning_rate": 1.0289914974966018e-05, + "loss": 1.0149, + "step": 21480 + }, + { + "epoch": 0.51, + "grad_norm": 2.3167267083385226, + "learning_rate": 1.0289152253800499e-05, + "loss": 1.1227, + "step": 21481 + }, + { + "epoch": 0.51, + "grad_norm": 2.0139576658709113, + "learning_rate": 1.0288389530951444e-05, + "loss": 0.9288, + "step": 21482 + }, + { + "epoch": 0.51, + "grad_norm": 7.971359634128497, + "learning_rate": 1.0287626806423292e-05, + "loss": 0.9691, + "step": 21483 + }, + { + "epoch": 0.51, + "grad_norm": 1.8308457561820233, + "learning_rate": 1.0286864080220486e-05, + "loss": 0.9278, + "step": 21484 + }, + { + "epoch": 0.51, + "grad_norm": 2.747926050043275, + "learning_rate": 1.028610135234747e-05, + "loss": 0.8877, + "step": 21485 + }, + { + "epoch": 0.51, + "grad_norm": 1.6973294686731764, + "learning_rate": 1.0285338622808676e-05, + "loss": 0.9096, + "step": 21486 + }, + { + "epoch": 0.51, + "grad_norm": 1.8376965009157618, + "learning_rate": 1.0284575891608556e-05, + "loss": 1.1407, + "step": 21487 + }, + { + "epoch": 0.51, + "grad_norm": 1.9133439552711193, + "learning_rate": 1.0283813158751539e-05, + "loss": 1.0649, + "step": 21488 + }, + { + "epoch": 0.51, + "grad_norm": 1.0733396518003582, + "learning_rate": 1.0283050424242076e-05, + "loss": 0.898, + "step": 21489 + }, + { + "epoch": 0.51, + "grad_norm": 2.1578660056536725, + "learning_rate": 1.02822876880846e-05, + "loss": 1.0822, + "step": 21490 + }, + { + "epoch": 0.51, + "grad_norm": 2.0160727804037344, + "learning_rate": 1.0281524950283562e-05, + "loss": 1.0397, + "step": 21491 + }, + { + "epoch": 0.51, + "grad_norm": 1.9156034036618084, + "learning_rate": 1.0280762210843391e-05, + "loss": 1.0207, + "step": 21492 + }, + { + "epoch": 0.51, + "grad_norm": 2.2435820179358292, + "learning_rate": 1.0279999469768535e-05, + "loss": 0.869, + "step": 21493 + }, + { + "epoch": 0.51, + "grad_norm": 1.858805877374606, + "learning_rate": 1.0279236727063436e-05, + "loss": 1.0375, + "step": 21494 + }, + { + "epoch": 0.51, + "grad_norm": 1.1061222708756442, + "learning_rate": 1.0278473982732529e-05, + "loss": 0.966, + "step": 21495 + }, + { + "epoch": 0.51, + "grad_norm": 2.0942848262764193, + "learning_rate": 1.0277711236780261e-05, + "loss": 1.0853, + "step": 21496 + }, + { + "epoch": 0.51, + "grad_norm": 1.1279598300175973, + "learning_rate": 1.0276948489211068e-05, + "loss": 1.0506, + "step": 21497 + }, + { + "epoch": 0.51, + "grad_norm": 2.0348409141392834, + "learning_rate": 1.0276185740029393e-05, + "loss": 0.9249, + "step": 21498 + }, + { + "epoch": 0.51, + "grad_norm": 2.251821159654963, + "learning_rate": 1.027542298923968e-05, + "loss": 1.0732, + "step": 21499 + }, + { + "epoch": 0.51, + "grad_norm": 2.0910053802670765, + "learning_rate": 1.0274660236846364e-05, + "loss": 1.0453, + "step": 21500 + }, + { + "epoch": 0.51, + "grad_norm": 2.2330507120130445, + "learning_rate": 1.0273897482853891e-05, + "loss": 0.9684, + "step": 21501 + }, + { + "epoch": 0.51, + "grad_norm": 2.2822555334380903, + "learning_rate": 1.02731347272667e-05, + "loss": 0.8863, + "step": 21502 + }, + { + "epoch": 0.51, + "grad_norm": 2.3711477389190803, + "learning_rate": 1.0272371970089232e-05, + "loss": 0.9869, + "step": 21503 + }, + { + "epoch": 0.51, + "grad_norm": 1.9272975243064747, + "learning_rate": 1.0271609211325928e-05, + "loss": 0.971, + "step": 21504 + }, + { + "epoch": 0.51, + "grad_norm": 1.8415138578603716, + "learning_rate": 1.0270846450981228e-05, + "loss": 1.0865, + "step": 21505 + }, + { + "epoch": 0.51, + "grad_norm": 1.9805489441334285, + "learning_rate": 1.0270083689059578e-05, + "loss": 1.0634, + "step": 21506 + }, + { + "epoch": 0.51, + "grad_norm": 1.9428826288847694, + "learning_rate": 1.0269320925565409e-05, + "loss": 0.9007, + "step": 21507 + }, + { + "epoch": 0.51, + "grad_norm": 2.041501778102186, + "learning_rate": 1.0268558160503174e-05, + "loss": 0.9072, + "step": 21508 + }, + { + "epoch": 0.51, + "grad_norm": 2.0002870711062872, + "learning_rate": 1.0267795393877305e-05, + "loss": 0.9618, + "step": 21509 + }, + { + "epoch": 0.51, + "grad_norm": 2.074131212502127, + "learning_rate": 1.026703262569225e-05, + "loss": 1.0954, + "step": 21510 + }, + { + "epoch": 0.51, + "grad_norm": 2.085198314072027, + "learning_rate": 1.0266269855952442e-05, + "loss": 1.0295, + "step": 21511 + }, + { + "epoch": 0.51, + "grad_norm": 2.269620998868261, + "learning_rate": 1.026550708466233e-05, + "loss": 1.0117, + "step": 21512 + }, + { + "epoch": 0.51, + "grad_norm": 2.1683797466057815, + "learning_rate": 1.026474431182635e-05, + "loss": 1.0202, + "step": 21513 + }, + { + "epoch": 0.51, + "grad_norm": 2.1075359356482455, + "learning_rate": 1.0263981537448944e-05, + "loss": 1.0532, + "step": 21514 + }, + { + "epoch": 0.51, + "grad_norm": 1.1227490014525443, + "learning_rate": 1.0263218761534557e-05, + "loss": 0.8826, + "step": 21515 + }, + { + "epoch": 0.51, + "grad_norm": 1.765068510724194, + "learning_rate": 1.0262455984087626e-05, + "loss": 1.0015, + "step": 21516 + }, + { + "epoch": 0.51, + "grad_norm": 2.12069832622451, + "learning_rate": 1.0261693205112592e-05, + "loss": 1.0348, + "step": 21517 + }, + { + "epoch": 0.51, + "grad_norm": 2.069273946953314, + "learning_rate": 1.0260930424613895e-05, + "loss": 0.7678, + "step": 21518 + }, + { + "epoch": 0.51, + "grad_norm": 1.980670746893663, + "learning_rate": 1.0260167642595984e-05, + "loss": 1.0038, + "step": 21519 + }, + { + "epoch": 0.51, + "grad_norm": 1.937869628302497, + "learning_rate": 1.0259404859063288e-05, + "loss": 1.0919, + "step": 21520 + }, + { + "epoch": 0.51, + "grad_norm": 1.971670899647915, + "learning_rate": 1.0258642074020261e-05, + "loss": 1.1748, + "step": 21521 + }, + { + "epoch": 0.51, + "grad_norm": 2.105467128952089, + "learning_rate": 1.0257879287471337e-05, + "loss": 1.0139, + "step": 21522 + }, + { + "epoch": 0.51, + "grad_norm": 2.993513902184562, + "learning_rate": 1.0257116499420955e-05, + "loss": 1.0083, + "step": 21523 + }, + { + "epoch": 0.51, + "grad_norm": 2.7676199632745067, + "learning_rate": 1.0256353709873563e-05, + "loss": 0.9563, + "step": 21524 + }, + { + "epoch": 0.51, + "grad_norm": 2.7699555011055, + "learning_rate": 1.0255590918833596e-05, + "loss": 0.9587, + "step": 21525 + }, + { + "epoch": 0.51, + "grad_norm": 2.2211961454758953, + "learning_rate": 1.0254828126305498e-05, + "loss": 1.1135, + "step": 21526 + }, + { + "epoch": 0.51, + "grad_norm": 1.8929546369696801, + "learning_rate": 1.025406533229371e-05, + "loss": 0.9307, + "step": 21527 + }, + { + "epoch": 0.51, + "grad_norm": 2.052934237111654, + "learning_rate": 1.0253302536802674e-05, + "loss": 0.9813, + "step": 21528 + }, + { + "epoch": 0.51, + "grad_norm": 2.035187101519267, + "learning_rate": 1.025253973983683e-05, + "loss": 1.1154, + "step": 21529 + }, + { + "epoch": 0.51, + "grad_norm": 1.9171700375666385, + "learning_rate": 1.025177694140062e-05, + "loss": 0.9232, + "step": 21530 + }, + { + "epoch": 0.51, + "grad_norm": 2.3864960474383103, + "learning_rate": 1.0251014141498486e-05, + "loss": 1.0335, + "step": 21531 + }, + { + "epoch": 0.51, + "grad_norm": 2.206210825985331, + "learning_rate": 1.0250251340134865e-05, + "loss": 1.0059, + "step": 21532 + }, + { + "epoch": 0.51, + "grad_norm": 2.856207948016851, + "learning_rate": 1.0249488537314203e-05, + "loss": 1.1961, + "step": 21533 + }, + { + "epoch": 0.51, + "grad_norm": 1.8938044124657158, + "learning_rate": 1.0248725733040941e-05, + "loss": 0.9713, + "step": 21534 + }, + { + "epoch": 0.51, + "grad_norm": 1.8696600671553978, + "learning_rate": 1.0247962927319518e-05, + "loss": 1.0304, + "step": 21535 + }, + { + "epoch": 0.51, + "grad_norm": 2.0432368346156085, + "learning_rate": 1.0247200120154377e-05, + "loss": 0.9055, + "step": 21536 + }, + { + "epoch": 0.51, + "grad_norm": 2.185669786761478, + "learning_rate": 1.0246437311549957e-05, + "loss": 0.9992, + "step": 21537 + }, + { + "epoch": 0.51, + "grad_norm": 1.8275107786343883, + "learning_rate": 1.0245674501510703e-05, + "loss": 1.0245, + "step": 21538 + }, + { + "epoch": 0.51, + "grad_norm": 2.1685126365570664, + "learning_rate": 1.0244911690041051e-05, + "loss": 1.0354, + "step": 21539 + }, + { + "epoch": 0.51, + "grad_norm": 1.9303916167054387, + "learning_rate": 1.024414887714545e-05, + "loss": 1.1034, + "step": 21540 + }, + { + "epoch": 0.51, + "grad_norm": 1.9884573912541768, + "learning_rate": 1.0243386062828333e-05, + "loss": 0.8492, + "step": 21541 + }, + { + "epoch": 0.51, + "grad_norm": 1.9878183826845022, + "learning_rate": 1.024262324709415e-05, + "loss": 1.1508, + "step": 21542 + }, + { + "epoch": 0.51, + "grad_norm": 1.1432496204715257, + "learning_rate": 1.0241860429947334e-05, + "loss": 1.0277, + "step": 21543 + }, + { + "epoch": 0.51, + "grad_norm": 2.2396280269220044, + "learning_rate": 1.024109761139233e-05, + "loss": 0.9668, + "step": 21544 + }, + { + "epoch": 0.51, + "grad_norm": 1.8904552644602852, + "learning_rate": 1.0240334791433581e-05, + "loss": 0.9778, + "step": 21545 + }, + { + "epoch": 0.51, + "grad_norm": 2.0199566731759533, + "learning_rate": 1.0239571970075526e-05, + "loss": 1.0668, + "step": 21546 + }, + { + "epoch": 0.51, + "grad_norm": 1.9006097749673119, + "learning_rate": 1.0238809147322606e-05, + "loss": 1.0098, + "step": 21547 + }, + { + "epoch": 0.51, + "grad_norm": 2.0331809763430315, + "learning_rate": 1.0238046323179265e-05, + "loss": 1.1295, + "step": 21548 + }, + { + "epoch": 0.51, + "grad_norm": 2.0297124761104843, + "learning_rate": 1.0237283497649942e-05, + "loss": 1.0593, + "step": 21549 + }, + { + "epoch": 0.51, + "grad_norm": 2.1761029046803464, + "learning_rate": 1.0236520670739081e-05, + "loss": 0.9089, + "step": 21550 + }, + { + "epoch": 0.51, + "grad_norm": 2.0339569689476864, + "learning_rate": 1.023575784245112e-05, + "loss": 0.8918, + "step": 21551 + }, + { + "epoch": 0.51, + "grad_norm": 2.0133596843359403, + "learning_rate": 1.0234995012790503e-05, + "loss": 0.9762, + "step": 21552 + }, + { + "epoch": 0.51, + "grad_norm": 1.8446296186302886, + "learning_rate": 1.023423218176167e-05, + "loss": 0.9516, + "step": 21553 + }, + { + "epoch": 0.51, + "grad_norm": 2.0057123395197145, + "learning_rate": 1.0233469349369062e-05, + "loss": 0.9726, + "step": 21554 + }, + { + "epoch": 0.51, + "grad_norm": 2.110436927257339, + "learning_rate": 1.0232706515617123e-05, + "loss": 0.9517, + "step": 21555 + }, + { + "epoch": 0.51, + "grad_norm": 1.8925640644027526, + "learning_rate": 1.0231943680510295e-05, + "loss": 1.0323, + "step": 21556 + }, + { + "epoch": 0.51, + "grad_norm": 1.0756834461317684, + "learning_rate": 1.0231180844053017e-05, + "loss": 1.0063, + "step": 21557 + }, + { + "epoch": 0.51, + "grad_norm": 2.0456521404931975, + "learning_rate": 1.0230418006249728e-05, + "loss": 1.1687, + "step": 21558 + }, + { + "epoch": 0.51, + "grad_norm": 2.1430705525781204, + "learning_rate": 1.0229655167104876e-05, + "loss": 1.0574, + "step": 21559 + }, + { + "epoch": 0.51, + "grad_norm": 2.4903873804247603, + "learning_rate": 1.0228892326622892e-05, + "loss": 0.9537, + "step": 21560 + }, + { + "epoch": 0.51, + "grad_norm": 2.1772163447469115, + "learning_rate": 1.022812948480823e-05, + "loss": 0.8727, + "step": 21561 + }, + { + "epoch": 0.51, + "grad_norm": 1.9450758105173962, + "learning_rate": 1.0227366641665327e-05, + "loss": 1.0508, + "step": 21562 + }, + { + "epoch": 0.51, + "grad_norm": 2.263502728338301, + "learning_rate": 1.022660379719862e-05, + "loss": 0.9147, + "step": 21563 + }, + { + "epoch": 0.51, + "grad_norm": 2.4597769132135214, + "learning_rate": 1.0225840951412557e-05, + "loss": 0.9841, + "step": 21564 + }, + { + "epoch": 0.51, + "grad_norm": 2.012322945624421, + "learning_rate": 1.0225078104311572e-05, + "loss": 0.9647, + "step": 21565 + }, + { + "epoch": 0.51, + "grad_norm": 1.9910508613450872, + "learning_rate": 1.0224315255900116e-05, + "loss": 1.1203, + "step": 21566 + }, + { + "epoch": 0.51, + "grad_norm": 2.098567655429665, + "learning_rate": 1.0223552406182622e-05, + "loss": 0.7878, + "step": 21567 + }, + { + "epoch": 0.51, + "grad_norm": 2.017662916682049, + "learning_rate": 1.0222789555163538e-05, + "loss": 1.0954, + "step": 21568 + }, + { + "epoch": 0.51, + "grad_norm": 2.2544203207425917, + "learning_rate": 1.02220267028473e-05, + "loss": 1.0118, + "step": 21569 + }, + { + "epoch": 0.51, + "grad_norm": 2.0901002160577695, + "learning_rate": 1.0221263849238357e-05, + "loss": 1.1537, + "step": 21570 + }, + { + "epoch": 0.51, + "grad_norm": 1.0434764220757047, + "learning_rate": 1.022050099434114e-05, + "loss": 1.0012, + "step": 21571 + }, + { + "epoch": 0.51, + "grad_norm": 1.8756543332773572, + "learning_rate": 1.02197381381601e-05, + "loss": 1.1111, + "step": 21572 + }, + { + "epoch": 0.51, + "grad_norm": 1.946520600181821, + "learning_rate": 1.0218975280699675e-05, + "loss": 0.9878, + "step": 21573 + }, + { + "epoch": 0.51, + "grad_norm": 2.0103656362718976, + "learning_rate": 1.0218212421964304e-05, + "loss": 0.9489, + "step": 21574 + }, + { + "epoch": 0.51, + "grad_norm": 1.8331404807155351, + "learning_rate": 1.0217449561958432e-05, + "loss": 1.1044, + "step": 21575 + }, + { + "epoch": 0.51, + "grad_norm": 1.9402549974110888, + "learning_rate": 1.02166867006865e-05, + "loss": 0.9797, + "step": 21576 + }, + { + "epoch": 0.51, + "grad_norm": 2.266809331892808, + "learning_rate": 1.021592383815295e-05, + "loss": 0.9895, + "step": 21577 + }, + { + "epoch": 0.51, + "grad_norm": 4.480228312078574, + "learning_rate": 1.0215160974362224e-05, + "loss": 1.0844, + "step": 21578 + }, + { + "epoch": 0.51, + "grad_norm": 1.1234557577781419, + "learning_rate": 1.0214398109318761e-05, + "loss": 0.9071, + "step": 21579 + }, + { + "epoch": 0.51, + "grad_norm": 1.9295078627857742, + "learning_rate": 1.0213635243027004e-05, + "loss": 0.9336, + "step": 21580 + }, + { + "epoch": 0.51, + "grad_norm": 2.128761756219324, + "learning_rate": 1.0212872375491396e-05, + "loss": 1.004, + "step": 21581 + }, + { + "epoch": 0.51, + "grad_norm": 2.0562043980595357, + "learning_rate": 1.021210950671638e-05, + "loss": 1.0258, + "step": 21582 + }, + { + "epoch": 0.51, + "grad_norm": 2.142832581624658, + "learning_rate": 1.021134663670639e-05, + "loss": 1.0404, + "step": 21583 + }, + { + "epoch": 0.51, + "grad_norm": 1.0915457489175802, + "learning_rate": 1.0210583765465875e-05, + "loss": 0.9224, + "step": 21584 + }, + { + "epoch": 0.51, + "grad_norm": 2.627046835257806, + "learning_rate": 1.0209820892999275e-05, + "loss": 0.9994, + "step": 21585 + }, + { + "epoch": 0.51, + "grad_norm": 1.1184788669936743, + "learning_rate": 1.0209058019311034e-05, + "loss": 0.9385, + "step": 21586 + }, + { + "epoch": 0.51, + "grad_norm": 2.04122885840443, + "learning_rate": 1.0208295144405592e-05, + "loss": 1.0285, + "step": 21587 + }, + { + "epoch": 0.51, + "grad_norm": 1.9938220806843154, + "learning_rate": 1.0207532268287384e-05, + "loss": 1.1101, + "step": 21588 + }, + { + "epoch": 0.51, + "grad_norm": 2.653016120955571, + "learning_rate": 1.020676939096086e-05, + "loss": 1.204, + "step": 21589 + }, + { + "epoch": 0.51, + "grad_norm": 1.8190433107021382, + "learning_rate": 1.0206006512430457e-05, + "loss": 0.9388, + "step": 21590 + }, + { + "epoch": 0.51, + "grad_norm": 2.005951088375193, + "learning_rate": 1.0205243632700622e-05, + "loss": 1.1216, + "step": 21591 + }, + { + "epoch": 0.51, + "grad_norm": 2.015948962289491, + "learning_rate": 1.0204480751775792e-05, + "loss": 1.1072, + "step": 21592 + }, + { + "epoch": 0.51, + "grad_norm": 2.379516303126462, + "learning_rate": 1.0203717869660413e-05, + "loss": 1.0092, + "step": 21593 + }, + { + "epoch": 0.51, + "grad_norm": 1.9376979873811022, + "learning_rate": 1.0202954986358921e-05, + "loss": 1.0015, + "step": 21594 + }, + { + "epoch": 0.51, + "grad_norm": 2.176651004025133, + "learning_rate": 1.0202192101875763e-05, + "loss": 1.0425, + "step": 21595 + }, + { + "epoch": 0.51, + "grad_norm": 2.0368449350594022, + "learning_rate": 1.0201429216215378e-05, + "loss": 1.1362, + "step": 21596 + }, + { + "epoch": 0.51, + "grad_norm": 2.0139633605068137, + "learning_rate": 1.0200666329382205e-05, + "loss": 0.9125, + "step": 21597 + }, + { + "epoch": 0.51, + "grad_norm": 2.2232609351181285, + "learning_rate": 1.0199903441380692e-05, + "loss": 1.1196, + "step": 21598 + }, + { + "epoch": 0.51, + "grad_norm": 1.9400194590699613, + "learning_rate": 1.0199140552215276e-05, + "loss": 0.9341, + "step": 21599 + }, + { + "epoch": 0.51, + "grad_norm": 1.9979655300755315, + "learning_rate": 1.0198377661890404e-05, + "loss": 0.8689, + "step": 21600 + }, + { + "epoch": 0.51, + "grad_norm": 1.9203732633938777, + "learning_rate": 1.0197614770410513e-05, + "loss": 1.1249, + "step": 21601 + }, + { + "epoch": 0.51, + "grad_norm": 1.0417810851204983, + "learning_rate": 1.0196851877780045e-05, + "loss": 0.9537, + "step": 21602 + }, + { + "epoch": 0.51, + "grad_norm": 1.9146479619907084, + "learning_rate": 1.0196088984003444e-05, + "loss": 1.0082, + "step": 21603 + }, + { + "epoch": 0.51, + "grad_norm": 1.1282405859317368, + "learning_rate": 1.0195326089085152e-05, + "loss": 0.9276, + "step": 21604 + }, + { + "epoch": 0.51, + "grad_norm": 3.1439476220889726, + "learning_rate": 1.0194563193029608e-05, + "loss": 1.0189, + "step": 21605 + }, + { + "epoch": 0.51, + "grad_norm": 1.9348801843566887, + "learning_rate": 1.0193800295841257e-05, + "loss": 0.9385, + "step": 21606 + }, + { + "epoch": 0.51, + "grad_norm": 2.020907277507919, + "learning_rate": 1.0193037397524538e-05, + "loss": 0.9196, + "step": 21607 + }, + { + "epoch": 0.51, + "grad_norm": 5.938210566174289, + "learning_rate": 1.0192274498083895e-05, + "loss": 0.9733, + "step": 21608 + }, + { + "epoch": 0.51, + "grad_norm": 2.07473810960005, + "learning_rate": 1.0191511597523767e-05, + "loss": 0.9598, + "step": 21609 + }, + { + "epoch": 0.51, + "grad_norm": 2.028242156764147, + "learning_rate": 1.01907486958486e-05, + "loss": 0.8781, + "step": 21610 + }, + { + "epoch": 0.51, + "grad_norm": 2.860533162258905, + "learning_rate": 1.0189985793062833e-05, + "loss": 1.0486, + "step": 21611 + }, + { + "epoch": 0.51, + "grad_norm": 2.346417344931349, + "learning_rate": 1.0189222889170906e-05, + "loss": 0.9195, + "step": 21612 + }, + { + "epoch": 0.51, + "grad_norm": 1.8990919325419795, + "learning_rate": 1.0188459984177268e-05, + "loss": 1.1009, + "step": 21613 + }, + { + "epoch": 0.51, + "grad_norm": 2.478910463659713, + "learning_rate": 1.0187697078086354e-05, + "loss": 0.9155, + "step": 21614 + }, + { + "epoch": 0.51, + "grad_norm": 1.964448890934635, + "learning_rate": 1.0186934170902611e-05, + "loss": 0.922, + "step": 21615 + }, + { + "epoch": 0.51, + "grad_norm": 1.9552240560185163, + "learning_rate": 1.0186171262630473e-05, + "loss": 1.0574, + "step": 21616 + }, + { + "epoch": 0.51, + "grad_norm": 2.8534541712151484, + "learning_rate": 1.018540835327439e-05, + "loss": 0.9463, + "step": 21617 + }, + { + "epoch": 0.51, + "grad_norm": 2.098490136374735, + "learning_rate": 1.0184645442838802e-05, + "loss": 0.9884, + "step": 21618 + }, + { + "epoch": 0.51, + "grad_norm": 1.980110123289811, + "learning_rate": 1.0183882531328148e-05, + "loss": 0.9794, + "step": 21619 + }, + { + "epoch": 0.51, + "grad_norm": 1.9207433032840797, + "learning_rate": 1.0183119618746871e-05, + "loss": 0.9573, + "step": 21620 + }, + { + "epoch": 0.51, + "grad_norm": 2.025382917995559, + "learning_rate": 1.0182356705099416e-05, + "loss": 1.0014, + "step": 21621 + }, + { + "epoch": 0.51, + "grad_norm": 2.215159592571133, + "learning_rate": 1.0181593790390221e-05, + "loss": 1.0309, + "step": 21622 + }, + { + "epoch": 0.51, + "grad_norm": 2.160651116549699, + "learning_rate": 1.0180830874623729e-05, + "loss": 1.0775, + "step": 21623 + }, + { + "epoch": 0.51, + "grad_norm": 3.630377454950933, + "learning_rate": 1.0180067957804383e-05, + "loss": 1.0233, + "step": 21624 + }, + { + "epoch": 0.51, + "grad_norm": 2.0810523142772017, + "learning_rate": 1.0179305039936624e-05, + "loss": 1.1633, + "step": 21625 + }, + { + "epoch": 0.51, + "grad_norm": 2.0623671207105216, + "learning_rate": 1.0178542121024895e-05, + "loss": 0.975, + "step": 21626 + }, + { + "epoch": 0.51, + "grad_norm": 2.070757179592771, + "learning_rate": 1.0177779201073636e-05, + "loss": 0.8803, + "step": 21627 + }, + { + "epoch": 0.51, + "grad_norm": 3.373181816250184, + "learning_rate": 1.017701628008729e-05, + "loss": 0.9907, + "step": 21628 + }, + { + "epoch": 0.51, + "grad_norm": 2.1363732777408426, + "learning_rate": 1.0176253358070301e-05, + "loss": 1.0497, + "step": 21629 + }, + { + "epoch": 0.51, + "grad_norm": 2.081515320959597, + "learning_rate": 1.017549043502711e-05, + "loss": 0.9957, + "step": 21630 + }, + { + "epoch": 0.51, + "grad_norm": 3.1540945775085034, + "learning_rate": 1.0174727510962154e-05, + "loss": 0.9583, + "step": 21631 + }, + { + "epoch": 0.51, + "grad_norm": 1.917064177870984, + "learning_rate": 1.0173964585879884e-05, + "loss": 0.9926, + "step": 21632 + }, + { + "epoch": 0.51, + "grad_norm": 1.9244740102665099, + "learning_rate": 1.0173201659784732e-05, + "loss": 1.0034, + "step": 21633 + }, + { + "epoch": 0.51, + "grad_norm": 2.1275213862180093, + "learning_rate": 1.017243873268115e-05, + "loss": 0.9654, + "step": 21634 + }, + { + "epoch": 0.51, + "grad_norm": 2.033277611691852, + "learning_rate": 1.017167580457357e-05, + "loss": 1.1009, + "step": 21635 + }, + { + "epoch": 0.51, + "grad_norm": 2.132941745808645, + "learning_rate": 1.0170912875466445e-05, + "loss": 0.9901, + "step": 21636 + }, + { + "epoch": 0.51, + "grad_norm": 2.8927048945804295, + "learning_rate": 1.0170149945364204e-05, + "loss": 1.1726, + "step": 21637 + }, + { + "epoch": 0.51, + "grad_norm": 1.0892041556308205, + "learning_rate": 1.0169387014271303e-05, + "loss": 0.9565, + "step": 21638 + }, + { + "epoch": 0.51, + "grad_norm": 2.1888969167482735, + "learning_rate": 1.0168624082192173e-05, + "loss": 0.9072, + "step": 21639 + }, + { + "epoch": 0.51, + "grad_norm": 2.3119628295845254, + "learning_rate": 1.0167861149131265e-05, + "loss": 1.0059, + "step": 21640 + }, + { + "epoch": 0.51, + "grad_norm": 2.0424238629035605, + "learning_rate": 1.016709821509301e-05, + "loss": 0.9579, + "step": 21641 + }, + { + "epoch": 0.51, + "grad_norm": 1.9517379945164757, + "learning_rate": 1.0166335280081861e-05, + "loss": 1.1474, + "step": 21642 + }, + { + "epoch": 0.51, + "grad_norm": 2.131922876268974, + "learning_rate": 1.0165572344102252e-05, + "loss": 1.1179, + "step": 21643 + }, + { + "epoch": 0.51, + "grad_norm": 2.0968470100227456, + "learning_rate": 1.0164809407158632e-05, + "loss": 1.0239, + "step": 21644 + }, + { + "epoch": 0.51, + "grad_norm": 2.0863661990774287, + "learning_rate": 1.0164046469255437e-05, + "loss": 1.0393, + "step": 21645 + }, + { + "epoch": 0.51, + "grad_norm": 1.8452091171527072, + "learning_rate": 1.0163283530397112e-05, + "loss": 0.9236, + "step": 21646 + }, + { + "epoch": 0.51, + "grad_norm": 1.8600114486454793, + "learning_rate": 1.01625205905881e-05, + "loss": 1.0084, + "step": 21647 + }, + { + "epoch": 0.51, + "grad_norm": 1.900313035014686, + "learning_rate": 1.0161757649832836e-05, + "loss": 1.0455, + "step": 21648 + }, + { + "epoch": 0.51, + "grad_norm": 1.1569949442412595, + "learning_rate": 1.0160994708135775e-05, + "loss": 0.9327, + "step": 21649 + }, + { + "epoch": 0.51, + "grad_norm": 1.7722325687571105, + "learning_rate": 1.0160231765501345e-05, + "loss": 0.907, + "step": 21650 + }, + { + "epoch": 0.51, + "grad_norm": 1.6471983808361803, + "learning_rate": 1.0159468821934001e-05, + "loss": 0.9477, + "step": 21651 + }, + { + "epoch": 0.51, + "grad_norm": 2.1553107450935287, + "learning_rate": 1.0158705877438178e-05, + "loss": 1.0397, + "step": 21652 + }, + { + "epoch": 0.51, + "grad_norm": 2.24455241096198, + "learning_rate": 1.0157942932018317e-05, + "loss": 0.9916, + "step": 21653 + }, + { + "epoch": 0.51, + "grad_norm": 2.018091314001947, + "learning_rate": 1.0157179985678862e-05, + "loss": 0.9927, + "step": 21654 + }, + { + "epoch": 0.51, + "grad_norm": 2.3451166535471653, + "learning_rate": 1.0156417038424259e-05, + "loss": 1.0698, + "step": 21655 + }, + { + "epoch": 0.51, + "grad_norm": 1.841928879071892, + "learning_rate": 1.0155654090258942e-05, + "loss": 1.0327, + "step": 21656 + }, + { + "epoch": 0.51, + "grad_norm": 1.985908033927467, + "learning_rate": 1.015489114118736e-05, + "loss": 0.9546, + "step": 21657 + }, + { + "epoch": 0.51, + "grad_norm": 2.199135112770704, + "learning_rate": 1.0154128191213952e-05, + "loss": 0.9471, + "step": 21658 + }, + { + "epoch": 0.51, + "grad_norm": 2.1498605295005477, + "learning_rate": 1.015336524034316e-05, + "loss": 1.0944, + "step": 21659 + }, + { + "epoch": 0.51, + "grad_norm": 1.8472751466088662, + "learning_rate": 1.015260228857943e-05, + "loss": 1.0001, + "step": 21660 + }, + { + "epoch": 0.51, + "grad_norm": 1.8798479193384452, + "learning_rate": 1.0151839335927199e-05, + "loss": 1.0559, + "step": 21661 + }, + { + "epoch": 0.51, + "grad_norm": 1.9414011196763317, + "learning_rate": 1.015107638239091e-05, + "loss": 0.9193, + "step": 21662 + }, + { + "epoch": 0.51, + "grad_norm": 2.1916465362985775, + "learning_rate": 1.015031342797501e-05, + "loss": 1.1079, + "step": 21663 + }, + { + "epoch": 0.51, + "grad_norm": 3.578813842095899, + "learning_rate": 1.0149550472683933e-05, + "loss": 1.0573, + "step": 21664 + }, + { + "epoch": 0.51, + "grad_norm": 1.8547784522584212, + "learning_rate": 1.0148787516522129e-05, + "loss": 1.0883, + "step": 21665 + }, + { + "epoch": 0.51, + "grad_norm": 1.765226493098429, + "learning_rate": 1.0148024559494038e-05, + "loss": 0.9686, + "step": 21666 + }, + { + "epoch": 0.51, + "grad_norm": 1.9203978317945765, + "learning_rate": 1.0147261601604099e-05, + "loss": 1.0413, + "step": 21667 + }, + { + "epoch": 0.51, + "grad_norm": 2.0655678049084933, + "learning_rate": 1.014649864285676e-05, + "loss": 1.0367, + "step": 21668 + }, + { + "epoch": 0.51, + "grad_norm": 2.0244451177481237, + "learning_rate": 1.0145735683256453e-05, + "loss": 1.0135, + "step": 21669 + }, + { + "epoch": 0.51, + "grad_norm": 2.156082559840479, + "learning_rate": 1.0144972722807633e-05, + "loss": 0.9296, + "step": 21670 + }, + { + "epoch": 0.51, + "grad_norm": 2.4980510551974655, + "learning_rate": 1.0144209761514729e-05, + "loss": 1.0093, + "step": 21671 + }, + { + "epoch": 0.51, + "grad_norm": 1.905303164129967, + "learning_rate": 1.0143446799382197e-05, + "loss": 0.889, + "step": 21672 + }, + { + "epoch": 0.51, + "grad_norm": 1.948747820658651, + "learning_rate": 1.0142683836414472e-05, + "loss": 1.0286, + "step": 21673 + }, + { + "epoch": 0.51, + "grad_norm": 2.260392950686137, + "learning_rate": 1.0141920872615992e-05, + "loss": 0.9869, + "step": 21674 + }, + { + "epoch": 0.51, + "grad_norm": 2.0181842231809495, + "learning_rate": 1.0141157907991206e-05, + "loss": 0.9975, + "step": 21675 + }, + { + "epoch": 0.51, + "grad_norm": 2.7005191188897704, + "learning_rate": 1.0140394942544554e-05, + "loss": 0.9722, + "step": 21676 + }, + { + "epoch": 0.51, + "grad_norm": 2.0012636719916994, + "learning_rate": 1.0139631976280477e-05, + "loss": 0.9886, + "step": 21677 + }, + { + "epoch": 0.51, + "grad_norm": 2.131217516370315, + "learning_rate": 1.013886900920342e-05, + "loss": 0.9432, + "step": 21678 + }, + { + "epoch": 0.51, + "grad_norm": 1.1277390302390886, + "learning_rate": 1.0138106041317823e-05, + "loss": 0.9624, + "step": 21679 + }, + { + "epoch": 0.51, + "grad_norm": 1.9226351605790941, + "learning_rate": 1.013734307262813e-05, + "loss": 0.8501, + "step": 21680 + }, + { + "epoch": 0.51, + "grad_norm": 2.0136450752783728, + "learning_rate": 1.013658010313878e-05, + "loss": 0.9326, + "step": 21681 + }, + { + "epoch": 0.51, + "grad_norm": 2.160303396892185, + "learning_rate": 1.0135817132854218e-05, + "loss": 0.972, + "step": 21682 + }, + { + "epoch": 0.51, + "grad_norm": 2.0088587662134256, + "learning_rate": 1.0135054161778886e-05, + "loss": 0.9329, + "step": 21683 + }, + { + "epoch": 0.51, + "grad_norm": 2.636465727136477, + "learning_rate": 1.0134291189917227e-05, + "loss": 0.9984, + "step": 21684 + }, + { + "epoch": 0.51, + "grad_norm": 1.9487948405836915, + "learning_rate": 1.013352821727368e-05, + "loss": 1.0355, + "step": 21685 + }, + { + "epoch": 0.51, + "grad_norm": 1.8769400074045914, + "learning_rate": 1.0132765243852691e-05, + "loss": 1.0719, + "step": 21686 + }, + { + "epoch": 0.51, + "grad_norm": 1.9439594560768068, + "learning_rate": 1.0132002269658703e-05, + "loss": 1.0239, + "step": 21687 + }, + { + "epoch": 0.51, + "grad_norm": 1.8878514451741195, + "learning_rate": 1.0131239294696151e-05, + "loss": 1.0579, + "step": 21688 + }, + { + "epoch": 0.51, + "grad_norm": 2.036868663050182, + "learning_rate": 1.0130476318969487e-05, + "loss": 1.0219, + "step": 21689 + }, + { + "epoch": 0.51, + "grad_norm": 1.1284532432126797, + "learning_rate": 1.0129713342483145e-05, + "loss": 0.9521, + "step": 21690 + }, + { + "epoch": 0.51, + "grad_norm": 1.9955370475701193, + "learning_rate": 1.0128950365241575e-05, + "loss": 1.0621, + "step": 21691 + }, + { + "epoch": 0.51, + "grad_norm": 2.3359500489866676, + "learning_rate": 1.0128187387249208e-05, + "loss": 0.9434, + "step": 21692 + }, + { + "epoch": 0.51, + "grad_norm": 1.8376151163691088, + "learning_rate": 1.01274244085105e-05, + "loss": 1.0021, + "step": 21693 + }, + { + "epoch": 0.51, + "grad_norm": 2.090954916000775, + "learning_rate": 1.0126661429029885e-05, + "loss": 0.9896, + "step": 21694 + }, + { + "epoch": 0.51, + "grad_norm": 1.8920840654592574, + "learning_rate": 1.0125898448811803e-05, + "loss": 0.8962, + "step": 21695 + }, + { + "epoch": 0.51, + "grad_norm": 2.0400984600462864, + "learning_rate": 1.0125135467860706e-05, + "loss": 1.0704, + "step": 21696 + }, + { + "epoch": 0.51, + "grad_norm": 2.3322743580848933, + "learning_rate": 1.0124372486181027e-05, + "loss": 0.9454, + "step": 21697 + }, + { + "epoch": 0.51, + "grad_norm": 1.9369233633290055, + "learning_rate": 1.0123609503777216e-05, + "loss": 1.0002, + "step": 21698 + }, + { + "epoch": 0.51, + "grad_norm": 2.0161923341965418, + "learning_rate": 1.0122846520653707e-05, + "loss": 0.9924, + "step": 21699 + }, + { + "epoch": 0.51, + "grad_norm": 2.828521850836174, + "learning_rate": 1.0122083536814951e-05, + "loss": 1.0098, + "step": 21700 + }, + { + "epoch": 0.51, + "grad_norm": 2.101257591481341, + "learning_rate": 1.0121320552265384e-05, + "loss": 1.0827, + "step": 21701 + }, + { + "epoch": 0.51, + "grad_norm": 2.2685555103159287, + "learning_rate": 1.0120557567009449e-05, + "loss": 1.0881, + "step": 21702 + }, + { + "epoch": 0.51, + "grad_norm": 1.7943224357378524, + "learning_rate": 1.011979458105159e-05, + "loss": 1.0951, + "step": 21703 + }, + { + "epoch": 0.51, + "grad_norm": 2.2830150771957602, + "learning_rate": 1.0119031594396249e-05, + "loss": 0.9916, + "step": 21704 + }, + { + "epoch": 0.51, + "grad_norm": 2.0254084958536027, + "learning_rate": 1.011826860704787e-05, + "loss": 0.9404, + "step": 21705 + }, + { + "epoch": 0.51, + "grad_norm": 2.3001210819156546, + "learning_rate": 1.0117505619010891e-05, + "loss": 1.0397, + "step": 21706 + }, + { + "epoch": 0.51, + "grad_norm": 1.7885954026652982, + "learning_rate": 1.011674263028976e-05, + "loss": 1.0133, + "step": 21707 + }, + { + "epoch": 0.51, + "grad_norm": 1.9216116458493027, + "learning_rate": 1.0115979640888916e-05, + "loss": 0.9369, + "step": 21708 + }, + { + "epoch": 0.51, + "grad_norm": 1.30341958619407, + "learning_rate": 1.01152166508128e-05, + "loss": 1.0118, + "step": 21709 + }, + { + "epoch": 0.51, + "grad_norm": 1.9016408762495391, + "learning_rate": 1.0114453660065856e-05, + "loss": 1.0197, + "step": 21710 + }, + { + "epoch": 0.51, + "grad_norm": 1.8644224086590369, + "learning_rate": 1.0113690668652529e-05, + "loss": 0.916, + "step": 21711 + }, + { + "epoch": 0.51, + "grad_norm": 1.9320234822338072, + "learning_rate": 1.0112927676577258e-05, + "loss": 1.1138, + "step": 21712 + }, + { + "epoch": 0.51, + "grad_norm": 2.0009995033788486, + "learning_rate": 1.0112164683844487e-05, + "loss": 1.0374, + "step": 21713 + }, + { + "epoch": 0.51, + "grad_norm": 1.0415549602192369, + "learning_rate": 1.0111401690458655e-05, + "loss": 1.0435, + "step": 21714 + }, + { + "epoch": 0.51, + "grad_norm": 2.1095008021166537, + "learning_rate": 1.0110638696424211e-05, + "loss": 1.0147, + "step": 21715 + }, + { + "epoch": 0.51, + "grad_norm": 2.1079094173013906, + "learning_rate": 1.010987570174559e-05, + "loss": 0.9196, + "step": 21716 + }, + { + "epoch": 0.51, + "grad_norm": 2.586589825224044, + "learning_rate": 1.0109112706427242e-05, + "loss": 0.8573, + "step": 21717 + }, + { + "epoch": 0.51, + "grad_norm": 1.9116914288600395, + "learning_rate": 1.01083497104736e-05, + "loss": 1.0952, + "step": 21718 + }, + { + "epoch": 0.51, + "grad_norm": 2.074050037860251, + "learning_rate": 1.0107586713889117e-05, + "loss": 1.0786, + "step": 21719 + }, + { + "epoch": 0.51, + "grad_norm": 1.9446614791394183, + "learning_rate": 1.0106823716678225e-05, + "loss": 1.0459, + "step": 21720 + }, + { + "epoch": 0.51, + "grad_norm": 2.3754793273780557, + "learning_rate": 1.0106060718845377e-05, + "loss": 1.0127, + "step": 21721 + }, + { + "epoch": 0.51, + "grad_norm": 2.0639179018863003, + "learning_rate": 1.0105297720395003e-05, + "loss": 0.9247, + "step": 21722 + }, + { + "epoch": 0.51, + "grad_norm": 1.8651366683199573, + "learning_rate": 1.0104534721331558e-05, + "loss": 0.9355, + "step": 21723 + }, + { + "epoch": 0.51, + "grad_norm": 2.358101961305689, + "learning_rate": 1.0103771721659476e-05, + "loss": 0.9041, + "step": 21724 + }, + { + "epoch": 0.51, + "grad_norm": 2.055337482389085, + "learning_rate": 1.0103008721383202e-05, + "loss": 0.9154, + "step": 21725 + }, + { + "epoch": 0.51, + "grad_norm": 3.0866251303495864, + "learning_rate": 1.010224572050718e-05, + "loss": 1.0247, + "step": 21726 + }, + { + "epoch": 0.51, + "grad_norm": 1.9974531874614032, + "learning_rate": 1.0101482719035851e-05, + "loss": 1.0662, + "step": 21727 + }, + { + "epoch": 0.51, + "grad_norm": 1.981955870108211, + "learning_rate": 1.0100719716973657e-05, + "loss": 1.0163, + "step": 21728 + }, + { + "epoch": 0.51, + "grad_norm": 2.1077772846520677, + "learning_rate": 1.009995671432504e-05, + "loss": 0.9552, + "step": 21729 + }, + { + "epoch": 0.51, + "grad_norm": 2.0839480807621658, + "learning_rate": 1.0099193711094446e-05, + "loss": 0.9956, + "step": 21730 + }, + { + "epoch": 0.51, + "grad_norm": 2.213537670183314, + "learning_rate": 1.0098430707286313e-05, + "loss": 0.9855, + "step": 21731 + }, + { + "epoch": 0.51, + "grad_norm": 1.1032370135431777, + "learning_rate": 1.0097667702905085e-05, + "loss": 0.9584, + "step": 21732 + }, + { + "epoch": 0.51, + "grad_norm": 1.1032406179436993, + "learning_rate": 1.0096904697955205e-05, + "loss": 0.9799, + "step": 21733 + }, + { + "epoch": 0.51, + "grad_norm": 1.9369172910315473, + "learning_rate": 1.0096141692441117e-05, + "loss": 0.967, + "step": 21734 + }, + { + "epoch": 0.51, + "grad_norm": 2.1443497875442836, + "learning_rate": 1.0095378686367258e-05, + "loss": 1.0231, + "step": 21735 + }, + { + "epoch": 0.51, + "grad_norm": 2.832323929198377, + "learning_rate": 1.0094615679738078e-05, + "loss": 0.8643, + "step": 21736 + }, + { + "epoch": 0.51, + "grad_norm": 1.98177381484915, + "learning_rate": 1.0093852672558013e-05, + "loss": 0.9645, + "step": 21737 + }, + { + "epoch": 0.51, + "grad_norm": 2.1554760090168905, + "learning_rate": 1.009308966483151e-05, + "loss": 0.9449, + "step": 21738 + }, + { + "epoch": 0.51, + "grad_norm": 2.6253564333963113, + "learning_rate": 1.0092326656563006e-05, + "loss": 0.9363, + "step": 21739 + }, + { + "epoch": 0.51, + "grad_norm": 1.8236113063789836, + "learning_rate": 1.0091563647756953e-05, + "loss": 1.049, + "step": 21740 + }, + { + "epoch": 0.51, + "grad_norm": 2.0566787600717595, + "learning_rate": 1.0090800638417779e-05, + "loss": 0.977, + "step": 21741 + }, + { + "epoch": 0.51, + "grad_norm": 1.769630709988717, + "learning_rate": 1.0090037628549943e-05, + "loss": 1.0211, + "step": 21742 + }, + { + "epoch": 0.51, + "grad_norm": 2.02031827743354, + "learning_rate": 1.0089274618157874e-05, + "loss": 1.0041, + "step": 21743 + }, + { + "epoch": 0.51, + "grad_norm": 1.951695905228694, + "learning_rate": 1.0088511607246021e-05, + "loss": 0.9944, + "step": 21744 + }, + { + "epoch": 0.51, + "grad_norm": 1.9588043897046046, + "learning_rate": 1.008774859581883e-05, + "loss": 0.9717, + "step": 21745 + }, + { + "epoch": 0.51, + "grad_norm": 1.8311995757953257, + "learning_rate": 1.0086985583880731e-05, + "loss": 1.0588, + "step": 21746 + }, + { + "epoch": 0.51, + "grad_norm": 2.2128569065060373, + "learning_rate": 1.0086222571436182e-05, + "loss": 0.9228, + "step": 21747 + }, + { + "epoch": 0.51, + "grad_norm": 1.0484342690235415, + "learning_rate": 1.0085459558489612e-05, + "loss": 0.9842, + "step": 21748 + }, + { + "epoch": 0.51, + "grad_norm": 1.959775720515755, + "learning_rate": 1.0084696545045474e-05, + "loss": 1.1327, + "step": 21749 + }, + { + "epoch": 0.51, + "grad_norm": 1.657016162450102, + "learning_rate": 1.0083933531108202e-05, + "loss": 0.8726, + "step": 21750 + }, + { + "epoch": 0.51, + "grad_norm": 1.0817056311828206, + "learning_rate": 1.0083170516682244e-05, + "loss": 0.9329, + "step": 21751 + }, + { + "epoch": 0.51, + "grad_norm": 1.9373192025052475, + "learning_rate": 1.0082407501772043e-05, + "loss": 1.0608, + "step": 21752 + }, + { + "epoch": 0.51, + "grad_norm": 1.871254769001467, + "learning_rate": 1.0081644486382036e-05, + "loss": 1.1267, + "step": 21753 + }, + { + "epoch": 0.51, + "grad_norm": 2.07271916323279, + "learning_rate": 1.008088147051667e-05, + "loss": 1.2028, + "step": 21754 + }, + { + "epoch": 0.51, + "grad_norm": 1.8881473643830045, + "learning_rate": 1.0080118454180387e-05, + "loss": 1.0856, + "step": 21755 + }, + { + "epoch": 0.51, + "grad_norm": 1.9595535438909064, + "learning_rate": 1.007935543737763e-05, + "loss": 0.9909, + "step": 21756 + }, + { + "epoch": 0.51, + "grad_norm": 1.0744802414633627, + "learning_rate": 1.0078592420112838e-05, + "loss": 0.9637, + "step": 21757 + }, + { + "epoch": 0.51, + "grad_norm": 1.064036464403275, + "learning_rate": 1.0077829402390458e-05, + "loss": 0.955, + "step": 21758 + }, + { + "epoch": 0.51, + "grad_norm": 2.0944873918362528, + "learning_rate": 1.007706638421493e-05, + "loss": 0.9471, + "step": 21759 + }, + { + "epoch": 0.51, + "grad_norm": 1.1201695799526639, + "learning_rate": 1.0076303365590696e-05, + "loss": 0.967, + "step": 21760 + }, + { + "epoch": 0.51, + "grad_norm": 2.214272519821809, + "learning_rate": 1.0075540346522202e-05, + "loss": 1.0629, + "step": 21761 + }, + { + "epoch": 0.51, + "grad_norm": 1.9679898492070869, + "learning_rate": 1.0074777327013887e-05, + "loss": 0.9439, + "step": 21762 + }, + { + "epoch": 0.51, + "grad_norm": 2.1817627384794287, + "learning_rate": 1.0074014307070194e-05, + "loss": 0.9191, + "step": 21763 + }, + { + "epoch": 0.51, + "grad_norm": 2.010733226693165, + "learning_rate": 1.0073251286695567e-05, + "loss": 1.0834, + "step": 21764 + }, + { + "epoch": 0.51, + "grad_norm": 1.930078925026622, + "learning_rate": 1.007248826589445e-05, + "loss": 1.1534, + "step": 21765 + }, + { + "epoch": 0.51, + "grad_norm": 1.9481812093508797, + "learning_rate": 1.0071725244671281e-05, + "loss": 1.0717, + "step": 21766 + }, + { + "epoch": 0.51, + "grad_norm": 2.1802770968067064, + "learning_rate": 1.0070962223030502e-05, + "loss": 1.0636, + "step": 21767 + }, + { + "epoch": 0.51, + "grad_norm": 2.0968682186482344, + "learning_rate": 1.0070199200976564e-05, + "loss": 1.1407, + "step": 21768 + }, + { + "epoch": 0.51, + "grad_norm": 2.397208861328069, + "learning_rate": 1.00694361785139e-05, + "loss": 0.9937, + "step": 21769 + }, + { + "epoch": 0.51, + "grad_norm": 2.672145546512126, + "learning_rate": 1.006867315564696e-05, + "loss": 1.014, + "step": 21770 + }, + { + "epoch": 0.51, + "grad_norm": 2.342787585050792, + "learning_rate": 1.0067910132380179e-05, + "loss": 0.9897, + "step": 21771 + }, + { + "epoch": 0.51, + "grad_norm": 2.7200160368170123, + "learning_rate": 1.006714710871801e-05, + "loss": 1.1439, + "step": 21772 + }, + { + "epoch": 0.51, + "grad_norm": 1.9112734811071272, + "learning_rate": 1.0066384084664882e-05, + "loss": 0.9541, + "step": 21773 + }, + { + "epoch": 0.51, + "grad_norm": 2.196202224157083, + "learning_rate": 1.006562106022525e-05, + "loss": 1.0233, + "step": 21774 + }, + { + "epoch": 0.51, + "grad_norm": 1.9141461976275453, + "learning_rate": 1.0064858035403549e-05, + "loss": 0.8576, + "step": 21775 + }, + { + "epoch": 0.51, + "grad_norm": 1.8709015040696577, + "learning_rate": 1.0064095010204225e-05, + "loss": 1.0162, + "step": 21776 + }, + { + "epoch": 0.51, + "grad_norm": 1.070599105366832, + "learning_rate": 1.0063331984631719e-05, + "loss": 0.9819, + "step": 21777 + }, + { + "epoch": 0.51, + "grad_norm": 1.9578239065594931, + "learning_rate": 1.0062568958690475e-05, + "loss": 0.9733, + "step": 21778 + }, + { + "epoch": 0.51, + "grad_norm": 2.0527935172112444, + "learning_rate": 1.0061805932384933e-05, + "loss": 0.9555, + "step": 21779 + }, + { + "epoch": 0.51, + "grad_norm": 1.0696887247587783, + "learning_rate": 1.0061042905719538e-05, + "loss": 0.8899, + "step": 21780 + }, + { + "epoch": 0.51, + "grad_norm": 2.492572581371419, + "learning_rate": 1.0060279878698734e-05, + "loss": 1.0118, + "step": 21781 + }, + { + "epoch": 0.51, + "grad_norm": 2.166178499075572, + "learning_rate": 1.0059516851326958e-05, + "loss": 1.0517, + "step": 21782 + }, + { + "epoch": 0.51, + "grad_norm": 2.0369453732187917, + "learning_rate": 1.0058753823608658e-05, + "loss": 1.0246, + "step": 21783 + }, + { + "epoch": 0.51, + "grad_norm": 2.055520484875684, + "learning_rate": 1.0057990795548274e-05, + "loss": 1.1236, + "step": 21784 + }, + { + "epoch": 0.51, + "grad_norm": 1.8461774139080882, + "learning_rate": 1.0057227767150249e-05, + "loss": 0.9107, + "step": 21785 + }, + { + "epoch": 0.51, + "grad_norm": 1.9351097494106893, + "learning_rate": 1.0056464738419028e-05, + "loss": 1.0575, + "step": 21786 + }, + { + "epoch": 0.51, + "grad_norm": 1.9772128576687402, + "learning_rate": 1.005570170935905e-05, + "loss": 0.9545, + "step": 21787 + }, + { + "epoch": 0.51, + "grad_norm": 2.605762417643462, + "learning_rate": 1.0054938679974757e-05, + "loss": 0.9252, + "step": 21788 + }, + { + "epoch": 0.51, + "grad_norm": 1.9435158665346721, + "learning_rate": 1.0054175650270596e-05, + "loss": 1.0238, + "step": 21789 + }, + { + "epoch": 0.51, + "grad_norm": 1.8129850127999503, + "learning_rate": 1.0053412620251006e-05, + "loss": 1.1273, + "step": 21790 + }, + { + "epoch": 0.51, + "grad_norm": 2.1623331404715884, + "learning_rate": 1.0052649589920433e-05, + "loss": 0.9856, + "step": 21791 + }, + { + "epoch": 0.51, + "grad_norm": 1.9849824116558805, + "learning_rate": 1.0051886559283315e-05, + "loss": 1.1287, + "step": 21792 + }, + { + "epoch": 0.51, + "grad_norm": 2.1267772604919912, + "learning_rate": 1.0051123528344096e-05, + "loss": 0.9424, + "step": 21793 + }, + { + "epoch": 0.51, + "grad_norm": 2.1301083462839765, + "learning_rate": 1.0050360497107221e-05, + "loss": 1.0154, + "step": 21794 + }, + { + "epoch": 0.51, + "grad_norm": 1.944460509192832, + "learning_rate": 1.0049597465577133e-05, + "loss": 1.1947, + "step": 21795 + }, + { + "epoch": 0.51, + "grad_norm": 1.9817275480354655, + "learning_rate": 1.0048834433758275e-05, + "loss": 0.9829, + "step": 21796 + }, + { + "epoch": 0.51, + "grad_norm": 2.240025051064641, + "learning_rate": 1.004807140165508e-05, + "loss": 0.8974, + "step": 21797 + }, + { + "epoch": 0.51, + "grad_norm": 2.016003895603318, + "learning_rate": 1.0047308369272004e-05, + "loss": 0.9634, + "step": 21798 + }, + { + "epoch": 0.51, + "grad_norm": 1.9260928049957105, + "learning_rate": 1.0046545336613481e-05, + "loss": 1.0897, + "step": 21799 + }, + { + "epoch": 0.51, + "grad_norm": 2.1869345901773096, + "learning_rate": 1.0045782303683958e-05, + "loss": 0.9462, + "step": 21800 + }, + { + "epoch": 0.51, + "grad_norm": 2.12668858280672, + "learning_rate": 1.0045019270487873e-05, + "loss": 1.0341, + "step": 21801 + }, + { + "epoch": 0.51, + "grad_norm": 1.9803459111280208, + "learning_rate": 1.0044256237029674e-05, + "loss": 0.8728, + "step": 21802 + }, + { + "epoch": 0.51, + "grad_norm": 2.1502003005166817, + "learning_rate": 1.0043493203313803e-05, + "loss": 0.9363, + "step": 21803 + }, + { + "epoch": 0.51, + "grad_norm": 2.040424893530502, + "learning_rate": 1.0042730169344698e-05, + "loss": 0.9661, + "step": 21804 + }, + { + "epoch": 0.51, + "grad_norm": 2.122614805821184, + "learning_rate": 1.0041967135126805e-05, + "loss": 1.0157, + "step": 21805 + }, + { + "epoch": 0.51, + "grad_norm": 1.9470408140317175, + "learning_rate": 1.0041204100664566e-05, + "loss": 0.9751, + "step": 21806 + }, + { + "epoch": 0.51, + "grad_norm": 2.271058348237319, + "learning_rate": 1.0040441065962423e-05, + "loss": 0.9931, + "step": 21807 + }, + { + "epoch": 0.51, + "grad_norm": 1.9525561661880757, + "learning_rate": 1.0039678031024819e-05, + "loss": 1.0025, + "step": 21808 + }, + { + "epoch": 0.51, + "grad_norm": 1.900489696731303, + "learning_rate": 1.0038914995856199e-05, + "loss": 1.0109, + "step": 21809 + }, + { + "epoch": 0.51, + "grad_norm": 2.3028451098808396, + "learning_rate": 1.0038151960461002e-05, + "loss": 1.0305, + "step": 21810 + }, + { + "epoch": 0.51, + "grad_norm": 1.9665850434825156, + "learning_rate": 1.0037388924843674e-05, + "loss": 1.0136, + "step": 21811 + }, + { + "epoch": 0.51, + "grad_norm": 1.102605264537033, + "learning_rate": 1.0036625889008656e-05, + "loss": 0.9162, + "step": 21812 + }, + { + "epoch": 0.51, + "grad_norm": 2.830690775431157, + "learning_rate": 1.0035862852960388e-05, + "loss": 0.9683, + "step": 21813 + }, + { + "epoch": 0.51, + "grad_norm": 1.1361948510994286, + "learning_rate": 1.0035099816703315e-05, + "loss": 0.9771, + "step": 21814 + }, + { + "epoch": 0.51, + "grad_norm": 1.9085438069391962, + "learning_rate": 1.0034336780241882e-05, + "loss": 0.9956, + "step": 21815 + }, + { + "epoch": 0.51, + "grad_norm": 2.092638013646135, + "learning_rate": 1.0033573743580527e-05, + "loss": 1.0116, + "step": 21816 + }, + { + "epoch": 0.51, + "grad_norm": 2.0845949583214853, + "learning_rate": 1.0032810706723699e-05, + "loss": 1.0652, + "step": 21817 + }, + { + "epoch": 0.51, + "grad_norm": 2.0309582968479223, + "learning_rate": 1.0032047669675832e-05, + "loss": 1.0309, + "step": 21818 + }, + { + "epoch": 0.51, + "grad_norm": 1.7856183140107627, + "learning_rate": 1.0031284632441378e-05, + "loss": 1.0079, + "step": 21819 + }, + { + "epoch": 0.51, + "grad_norm": 2.0994231556216283, + "learning_rate": 1.003052159502477e-05, + "loss": 1.1368, + "step": 21820 + }, + { + "epoch": 0.51, + "grad_norm": 1.1175553224428019, + "learning_rate": 1.002975855743046e-05, + "loss": 0.9556, + "step": 21821 + }, + { + "epoch": 0.51, + "grad_norm": 1.89712651879461, + "learning_rate": 1.0028995519662883e-05, + "loss": 0.8845, + "step": 21822 + }, + { + "epoch": 0.51, + "grad_norm": 1.8264764893801024, + "learning_rate": 1.0028232481726488e-05, + "loss": 0.9773, + "step": 21823 + }, + { + "epoch": 0.51, + "grad_norm": 2.182710948206042, + "learning_rate": 1.0027469443625712e-05, + "loss": 1.1225, + "step": 21824 + }, + { + "epoch": 0.51, + "grad_norm": 1.9447473102076578, + "learning_rate": 1.0026706405365001e-05, + "loss": 1.1128, + "step": 21825 + }, + { + "epoch": 0.51, + "grad_norm": 2.0441324749318857, + "learning_rate": 1.0025943366948795e-05, + "loss": 1.019, + "step": 21826 + }, + { + "epoch": 0.51, + "grad_norm": 2.6506610236879267, + "learning_rate": 1.002518032838154e-05, + "loss": 1.0141, + "step": 21827 + }, + { + "epoch": 0.51, + "grad_norm": 2.1845020105310833, + "learning_rate": 1.0024417289667678e-05, + "loss": 0.9884, + "step": 21828 + }, + { + "epoch": 0.51, + "grad_norm": 2.6503919877805697, + "learning_rate": 1.0023654250811652e-05, + "loss": 0.8655, + "step": 21829 + }, + { + "epoch": 0.51, + "grad_norm": 1.7987083437632896, + "learning_rate": 1.00228912118179e-05, + "loss": 0.9841, + "step": 21830 + }, + { + "epoch": 0.51, + "grad_norm": 1.9715527961988035, + "learning_rate": 1.0022128172690868e-05, + "loss": 0.8239, + "step": 21831 + }, + { + "epoch": 0.51, + "grad_norm": 1.9114868187546705, + "learning_rate": 1.0021365133435004e-05, + "loss": 1.059, + "step": 21832 + }, + { + "epoch": 0.51, + "grad_norm": 2.3904305755725166, + "learning_rate": 1.002060209405474e-05, + "loss": 0.9125, + "step": 21833 + }, + { + "epoch": 0.51, + "grad_norm": 1.1052415750413547, + "learning_rate": 1.0019839054554528e-05, + "loss": 0.9848, + "step": 21834 + }, + { + "epoch": 0.51, + "grad_norm": 1.8785489825963468, + "learning_rate": 1.0019076014938805e-05, + "loss": 0.9798, + "step": 21835 + }, + { + "epoch": 0.51, + "grad_norm": 1.9203241474130694, + "learning_rate": 1.0018312975212015e-05, + "loss": 1.0266, + "step": 21836 + }, + { + "epoch": 0.51, + "grad_norm": 1.9508749642394125, + "learning_rate": 1.0017549935378601e-05, + "loss": 1.0519, + "step": 21837 + }, + { + "epoch": 0.51, + "grad_norm": 2.1979767095929867, + "learning_rate": 1.0016786895443005e-05, + "loss": 1.1004, + "step": 21838 + }, + { + "epoch": 0.51, + "grad_norm": 1.8231084397321615, + "learning_rate": 1.0016023855409671e-05, + "loss": 0.8927, + "step": 21839 + }, + { + "epoch": 0.51, + "grad_norm": 1.9244196984935344, + "learning_rate": 1.0015260815283043e-05, + "loss": 1.0421, + "step": 21840 + }, + { + "epoch": 0.51, + "grad_norm": 1.9957532451126043, + "learning_rate": 1.001449777506756e-05, + "loss": 1.0224, + "step": 21841 + }, + { + "epoch": 0.51, + "grad_norm": 4.846969306835044, + "learning_rate": 1.0013734734767666e-05, + "loss": 1.0198, + "step": 21842 + }, + { + "epoch": 0.51, + "grad_norm": 2.134564232400903, + "learning_rate": 1.0012971694387803e-05, + "loss": 1.0931, + "step": 21843 + }, + { + "epoch": 0.51, + "grad_norm": 2.621199292866425, + "learning_rate": 1.0012208653932416e-05, + "loss": 0.9982, + "step": 21844 + }, + { + "epoch": 0.51, + "grad_norm": 2.126484731324985, + "learning_rate": 1.0011445613405948e-05, + "loss": 0.9896, + "step": 21845 + }, + { + "epoch": 0.51, + "grad_norm": 2.524017032985477, + "learning_rate": 1.0010682572812837e-05, + "loss": 0.9454, + "step": 21846 + }, + { + "epoch": 0.51, + "grad_norm": 2.1395076945796965, + "learning_rate": 1.0009919532157532e-05, + "loss": 1.0164, + "step": 21847 + }, + { + "epoch": 0.51, + "grad_norm": 1.948015398793737, + "learning_rate": 1.000915649144447e-05, + "loss": 1.0397, + "step": 21848 + }, + { + "epoch": 0.51, + "grad_norm": 2.0986727173332125, + "learning_rate": 1.0008393450678098e-05, + "loss": 1.025, + "step": 21849 + }, + { + "epoch": 0.51, + "grad_norm": 1.0873971844967436, + "learning_rate": 1.0007630409862854e-05, + "loss": 0.9392, + "step": 21850 + }, + { + "epoch": 0.51, + "grad_norm": 1.9789459144487231, + "learning_rate": 1.0006867369003185e-05, + "loss": 0.9869, + "step": 21851 + }, + { + "epoch": 0.51, + "grad_norm": 2.191938662514931, + "learning_rate": 1.000610432810353e-05, + "loss": 0.9292, + "step": 21852 + }, + { + "epoch": 0.51, + "grad_norm": 1.038516071572269, + "learning_rate": 1.0005341287168338e-05, + "loss": 0.9253, + "step": 21853 + }, + { + "epoch": 0.51, + "grad_norm": 1.0983140036756234, + "learning_rate": 1.0004578246202046e-05, + "loss": 1.0009, + "step": 21854 + }, + { + "epoch": 0.51, + "grad_norm": 1.9600284418200016, + "learning_rate": 1.0003815205209095e-05, + "loss": 0.9352, + "step": 21855 + }, + { + "epoch": 0.51, + "grad_norm": 1.8387387389610765, + "learning_rate": 1.0003052164193931e-05, + "loss": 0.883, + "step": 21856 + }, + { + "epoch": 0.51, + "grad_norm": 2.171581152378705, + "learning_rate": 1.0002289123161e-05, + "loss": 1.0743, + "step": 21857 + }, + { + "epoch": 0.51, + "grad_norm": 2.6385863203442317, + "learning_rate": 1.0001526082114738e-05, + "loss": 1.0423, + "step": 21858 + }, + { + "epoch": 0.51, + "grad_norm": 1.9197857756728804, + "learning_rate": 1.0000763041059591e-05, + "loss": 1.0309, + "step": 21859 + }, + { + "epoch": 0.52, + "grad_norm": 0.9952976103723284, + "learning_rate": 1e-05, + "loss": 0.9162, + "step": 21860 + }, + { + "epoch": 0.52, + "grad_norm": 2.2240035443694746, + "learning_rate": 9.999236958940412e-06, + "loss": 1.076, + "step": 21861 + }, + { + "epoch": 0.52, + "grad_norm": 1.8328155192903575, + "learning_rate": 9.998473917885268e-06, + "loss": 1.0775, + "step": 21862 + }, + { + "epoch": 0.52, + "grad_norm": 1.99636199160342, + "learning_rate": 9.997710876839005e-06, + "loss": 1.1165, + "step": 21863 + }, + { + "epoch": 0.52, + "grad_norm": 2.185984729100442, + "learning_rate": 9.99694783580607e-06, + "loss": 0.9749, + "step": 21864 + }, + { + "epoch": 0.52, + "grad_norm": 2.256273037740094, + "learning_rate": 9.996184794790907e-06, + "loss": 0.9667, + "step": 21865 + }, + { + "epoch": 0.52, + "grad_norm": 2.12503609173029, + "learning_rate": 9.99542175379796e-06, + "loss": 0.9912, + "step": 21866 + }, + { + "epoch": 0.52, + "grad_norm": 2.0229769391654258, + "learning_rate": 9.994658712831666e-06, + "loss": 0.9441, + "step": 21867 + }, + { + "epoch": 0.52, + "grad_norm": 2.0180238579206518, + "learning_rate": 9.99389567189647e-06, + "loss": 1.0276, + "step": 21868 + }, + { + "epoch": 0.52, + "grad_norm": 1.896257137737961, + "learning_rate": 9.99313263099682e-06, + "loss": 1.0626, + "step": 21869 + }, + { + "epoch": 0.52, + "grad_norm": 1.84440056280265, + "learning_rate": 9.99236959013715e-06, + "loss": 1.0701, + "step": 21870 + }, + { + "epoch": 0.52, + "grad_norm": 4.641968324796498, + "learning_rate": 9.991606549321904e-06, + "loss": 1.0386, + "step": 21871 + }, + { + "epoch": 0.52, + "grad_norm": 2.224306852066279, + "learning_rate": 9.99084350855553e-06, + "loss": 0.8335, + "step": 21872 + }, + { + "epoch": 0.52, + "grad_norm": 2.195373383179232, + "learning_rate": 9.990080467842472e-06, + "loss": 0.9544, + "step": 21873 + }, + { + "epoch": 0.52, + "grad_norm": 2.276267043254772, + "learning_rate": 9.989317427187165e-06, + "loss": 1.1383, + "step": 21874 + }, + { + "epoch": 0.52, + "grad_norm": 2.222942092458256, + "learning_rate": 9.988554386594052e-06, + "loss": 1.0653, + "step": 21875 + }, + { + "epoch": 0.52, + "grad_norm": 5.858587457620352, + "learning_rate": 9.987791346067586e-06, + "loss": 0.9399, + "step": 21876 + }, + { + "epoch": 0.52, + "grad_norm": 2.0128250951883526, + "learning_rate": 9.987028305612199e-06, + "loss": 1.0701, + "step": 21877 + }, + { + "epoch": 0.52, + "grad_norm": 2.4265070535255, + "learning_rate": 9.986265265232337e-06, + "loss": 1.1406, + "step": 21878 + }, + { + "epoch": 0.52, + "grad_norm": 2.0779444742824023, + "learning_rate": 9.985502224932442e-06, + "loss": 0.9947, + "step": 21879 + }, + { + "epoch": 0.52, + "grad_norm": 1.893232011795639, + "learning_rate": 9.98473918471696e-06, + "loss": 1.0756, + "step": 21880 + }, + { + "epoch": 0.52, + "grad_norm": 1.8849078447169016, + "learning_rate": 9.983976144590332e-06, + "loss": 1.0378, + "step": 21881 + }, + { + "epoch": 0.52, + "grad_norm": 2.8363105128621053, + "learning_rate": 9.983213104556995e-06, + "loss": 1.0979, + "step": 21882 + }, + { + "epoch": 0.52, + "grad_norm": 2.0042859948308474, + "learning_rate": 9.982450064621405e-06, + "loss": 0.8552, + "step": 21883 + }, + { + "epoch": 0.52, + "grad_norm": 2.0439886698200027, + "learning_rate": 9.98168702478799e-06, + "loss": 0.9666, + "step": 21884 + }, + { + "epoch": 0.52, + "grad_norm": 2.0212133260429384, + "learning_rate": 9.980923985061199e-06, + "loss": 1.0148, + "step": 21885 + }, + { + "epoch": 0.52, + "grad_norm": 2.0634263285720555, + "learning_rate": 9.980160945445474e-06, + "loss": 0.8876, + "step": 21886 + }, + { + "epoch": 0.52, + "grad_norm": 2.111769418193687, + "learning_rate": 9.979397905945264e-06, + "loss": 0.9541, + "step": 21887 + }, + { + "epoch": 0.52, + "grad_norm": 1.9882374798017055, + "learning_rate": 9.978634866565e-06, + "loss": 0.956, + "step": 21888 + }, + { + "epoch": 0.52, + "grad_norm": 2.076567704627107, + "learning_rate": 9.977871827309132e-06, + "loss": 0.9326, + "step": 21889 + }, + { + "epoch": 0.52, + "grad_norm": 1.970962292878413, + "learning_rate": 9.977108788182104e-06, + "loss": 1.07, + "step": 21890 + }, + { + "epoch": 0.52, + "grad_norm": 2.2064271201313774, + "learning_rate": 9.976345749188353e-06, + "loss": 1.0412, + "step": 21891 + }, + { + "epoch": 0.52, + "grad_norm": 2.020852217100073, + "learning_rate": 9.975582710332325e-06, + "loss": 0.9724, + "step": 21892 + }, + { + "epoch": 0.52, + "grad_norm": 3.7545020089869117, + "learning_rate": 9.974819671618462e-06, + "loss": 0.9098, + "step": 21893 + }, + { + "epoch": 0.52, + "grad_norm": 1.9204839194685412, + "learning_rate": 9.97405663305121e-06, + "loss": 0.9928, + "step": 21894 + }, + { + "epoch": 0.52, + "grad_norm": 2.1845769802165687, + "learning_rate": 9.973293594635002e-06, + "loss": 0.9645, + "step": 21895 + }, + { + "epoch": 0.52, + "grad_norm": 1.9563466585977758, + "learning_rate": 9.972530556374291e-06, + "loss": 0.9622, + "step": 21896 + }, + { + "epoch": 0.52, + "grad_norm": 2.056212213627949, + "learning_rate": 9.971767518273517e-06, + "loss": 0.9812, + "step": 21897 + }, + { + "epoch": 0.52, + "grad_norm": 1.97437616309628, + "learning_rate": 9.97100448033712e-06, + "loss": 1.0843, + "step": 21898 + }, + { + "epoch": 0.52, + "grad_norm": 1.8494626518753123, + "learning_rate": 9.970241442569543e-06, + "loss": 0.9715, + "step": 21899 + }, + { + "epoch": 0.52, + "grad_norm": 1.902440738176082, + "learning_rate": 9.96947840497523e-06, + "loss": 1.0156, + "step": 21900 + }, + { + "epoch": 0.52, + "grad_norm": 2.347626038383119, + "learning_rate": 9.968715367558627e-06, + "loss": 1.1207, + "step": 21901 + }, + { + "epoch": 0.52, + "grad_norm": 2.081693790861122, + "learning_rate": 9.967952330324171e-06, + "loss": 1.0828, + "step": 21902 + }, + { + "epoch": 0.52, + "grad_norm": 2.192586587088964, + "learning_rate": 9.967189293276303e-06, + "loss": 1.0556, + "step": 21903 + }, + { + "epoch": 0.52, + "grad_norm": 2.048208740821936, + "learning_rate": 9.966426256419475e-06, + "loss": 0.9899, + "step": 21904 + }, + { + "epoch": 0.52, + "grad_norm": 2.2106160306228704, + "learning_rate": 9.965663219758121e-06, + "loss": 0.9325, + "step": 21905 + }, + { + "epoch": 0.52, + "grad_norm": 1.994351411377809, + "learning_rate": 9.964900183296688e-06, + "loss": 0.9873, + "step": 21906 + }, + { + "epoch": 0.52, + "grad_norm": 2.082781196447178, + "learning_rate": 9.964137147039613e-06, + "loss": 1.0906, + "step": 21907 + }, + { + "epoch": 0.52, + "grad_norm": 1.8686850862398072, + "learning_rate": 9.96337411099135e-06, + "loss": 0.9291, + "step": 21908 + }, + { + "epoch": 0.52, + "grad_norm": 2.193464766218235, + "learning_rate": 9.962611075156329e-06, + "loss": 1.0005, + "step": 21909 + }, + { + "epoch": 0.52, + "grad_norm": 2.2302959706854106, + "learning_rate": 9.961848039539e-06, + "loss": 1.0352, + "step": 21910 + }, + { + "epoch": 0.52, + "grad_norm": 2.2358173021203656, + "learning_rate": 9.961085004143806e-06, + "loss": 0.8576, + "step": 21911 + }, + { + "epoch": 0.52, + "grad_norm": 1.9779072166884883, + "learning_rate": 9.960321968975182e-06, + "loss": 0.9129, + "step": 21912 + }, + { + "epoch": 0.52, + "grad_norm": 1.1527570740915467, + "learning_rate": 9.959558934037578e-06, + "loss": 0.9632, + "step": 21913 + }, + { + "epoch": 0.52, + "grad_norm": 1.1266994448017214, + "learning_rate": 9.958795899335438e-06, + "loss": 1.0067, + "step": 21914 + }, + { + "epoch": 0.52, + "grad_norm": 1.1281631676013986, + "learning_rate": 9.9580328648732e-06, + "loss": 1.0244, + "step": 21915 + }, + { + "epoch": 0.52, + "grad_norm": 1.9777844622313332, + "learning_rate": 9.957269830655305e-06, + "loss": 0.9642, + "step": 21916 + }, + { + "epoch": 0.52, + "grad_norm": 1.9922062012785724, + "learning_rate": 9.9565067966862e-06, + "loss": 1.0613, + "step": 21917 + }, + { + "epoch": 0.52, + "grad_norm": 2.1584059247257352, + "learning_rate": 9.955743762970329e-06, + "loss": 0.8927, + "step": 21918 + }, + { + "epoch": 0.52, + "grad_norm": 2.1004719555849207, + "learning_rate": 9.95498072951213e-06, + "loss": 1.0003, + "step": 21919 + }, + { + "epoch": 0.52, + "grad_norm": 1.997664118720704, + "learning_rate": 9.954217696316043e-06, + "loss": 0.9599, + "step": 21920 + }, + { + "epoch": 0.52, + "grad_norm": 2.331839485187326, + "learning_rate": 9.95345466338652e-06, + "loss": 0.9668, + "step": 21921 + }, + { + "epoch": 0.52, + "grad_norm": 2.291691120335128, + "learning_rate": 9.952691630728001e-06, + "loss": 0.9484, + "step": 21922 + }, + { + "epoch": 0.52, + "grad_norm": 2.24213631439434, + "learning_rate": 9.951928598344923e-06, + "loss": 1.1054, + "step": 21923 + }, + { + "epoch": 0.52, + "grad_norm": 1.7926700757260003, + "learning_rate": 9.95116556624173e-06, + "loss": 0.9825, + "step": 21924 + }, + { + "epoch": 0.52, + "grad_norm": 2.0545078887370303, + "learning_rate": 9.950402534422869e-06, + "loss": 0.9174, + "step": 21925 + }, + { + "epoch": 0.52, + "grad_norm": 2.114786119507257, + "learning_rate": 9.94963950289278e-06, + "loss": 0.948, + "step": 21926 + }, + { + "epoch": 0.52, + "grad_norm": 2.02258301011481, + "learning_rate": 9.948876471655906e-06, + "loss": 1.1151, + "step": 21927 + }, + { + "epoch": 0.52, + "grad_norm": 2.2720841001884042, + "learning_rate": 9.948113440716687e-06, + "loss": 0.9684, + "step": 21928 + }, + { + "epoch": 0.52, + "grad_norm": 1.9932238515743748, + "learning_rate": 9.947350410079572e-06, + "loss": 0.9694, + "step": 21929 + }, + { + "epoch": 0.52, + "grad_norm": 2.011340771797135, + "learning_rate": 9.946587379748997e-06, + "loss": 1.0162, + "step": 21930 + }, + { + "epoch": 0.52, + "grad_norm": 1.1631506626398702, + "learning_rate": 9.945824349729405e-06, + "loss": 0.9809, + "step": 21931 + }, + { + "epoch": 0.52, + "grad_norm": 3.2257488689884184, + "learning_rate": 9.945061320025243e-06, + "loss": 1.2221, + "step": 21932 + }, + { + "epoch": 0.52, + "grad_norm": 1.8238637371216055, + "learning_rate": 9.944298290640955e-06, + "loss": 1.0141, + "step": 21933 + }, + { + "epoch": 0.52, + "grad_norm": 2.327446529975968, + "learning_rate": 9.943535261580977e-06, + "loss": 1.0114, + "step": 21934 + }, + { + "epoch": 0.52, + "grad_norm": 1.1116140784198327, + "learning_rate": 9.942772232849751e-06, + "loss": 1.0269, + "step": 21935 + }, + { + "epoch": 0.52, + "grad_norm": 1.9769603824463424, + "learning_rate": 9.942009204451731e-06, + "loss": 0.9427, + "step": 21936 + }, + { + "epoch": 0.52, + "grad_norm": 2.136762568662619, + "learning_rate": 9.941246176391345e-06, + "loss": 1.0224, + "step": 21937 + }, + { + "epoch": 0.52, + "grad_norm": 1.949437635887582, + "learning_rate": 9.940483148673045e-06, + "loss": 0.8136, + "step": 21938 + }, + { + "epoch": 0.52, + "grad_norm": 2.227779411573972, + "learning_rate": 9.939720121301268e-06, + "loss": 1.1517, + "step": 21939 + }, + { + "epoch": 0.52, + "grad_norm": 1.5674324460058875, + "learning_rate": 9.938957094280464e-06, + "loss": 0.9185, + "step": 21940 + }, + { + "epoch": 0.52, + "grad_norm": 1.1206933828537418, + "learning_rate": 9.93819406761507e-06, + "loss": 0.9824, + "step": 21941 + }, + { + "epoch": 0.52, + "grad_norm": 2.0747441851619253, + "learning_rate": 9.937431041309528e-06, + "loss": 0.9108, + "step": 21942 + }, + { + "epoch": 0.52, + "grad_norm": 2.100821817881666, + "learning_rate": 9.936668015368286e-06, + "loss": 1.0101, + "step": 21943 + }, + { + "epoch": 0.52, + "grad_norm": 2.0422474311896175, + "learning_rate": 9.935904989795778e-06, + "loss": 1.0311, + "step": 21944 + }, + { + "epoch": 0.52, + "grad_norm": 1.8505924623570622, + "learning_rate": 9.935141964596453e-06, + "loss": 1.0311, + "step": 21945 + }, + { + "epoch": 0.52, + "grad_norm": 1.045036297397993, + "learning_rate": 9.934378939774753e-06, + "loss": 0.9258, + "step": 21946 + }, + { + "epoch": 0.52, + "grad_norm": 1.9028582193042334, + "learning_rate": 9.933615915335121e-06, + "loss": 0.9631, + "step": 21947 + }, + { + "epoch": 0.52, + "grad_norm": 2.2888495768134782, + "learning_rate": 9.932852891281994e-06, + "loss": 1.0348, + "step": 21948 + }, + { + "epoch": 0.52, + "grad_norm": 1.8159708229885594, + "learning_rate": 9.932089867619821e-06, + "loss": 0.9417, + "step": 21949 + }, + { + "epoch": 0.52, + "grad_norm": 2.178137022373325, + "learning_rate": 9.931326844353044e-06, + "loss": 0.906, + "step": 21950 + }, + { + "epoch": 0.52, + "grad_norm": 2.0703718235405435, + "learning_rate": 9.930563821486101e-06, + "loss": 1.0217, + "step": 21951 + }, + { + "epoch": 0.52, + "grad_norm": 1.9531857076985373, + "learning_rate": 9.929800799023438e-06, + "loss": 1.0472, + "step": 21952 + }, + { + "epoch": 0.52, + "grad_norm": 1.9809937671163407, + "learning_rate": 9.929037776969498e-06, + "loss": 0.9687, + "step": 21953 + }, + { + "epoch": 0.52, + "grad_norm": 2.225711796249507, + "learning_rate": 9.928274755328724e-06, + "loss": 1.0703, + "step": 21954 + }, + { + "epoch": 0.52, + "grad_norm": 1.9365105269646625, + "learning_rate": 9.927511734105555e-06, + "loss": 1.0318, + "step": 21955 + }, + { + "epoch": 0.52, + "grad_norm": 2.1325209253142132, + "learning_rate": 9.926748713304433e-06, + "loss": 0.958, + "step": 21956 + }, + { + "epoch": 0.52, + "grad_norm": 2.847522093954558, + "learning_rate": 9.925985692929809e-06, + "loss": 0.9708, + "step": 21957 + }, + { + "epoch": 0.52, + "grad_norm": 2.0371721749258507, + "learning_rate": 9.925222672986116e-06, + "loss": 0.8704, + "step": 21958 + }, + { + "epoch": 0.52, + "grad_norm": 1.8693158142109552, + "learning_rate": 9.924459653477802e-06, + "loss": 1.0169, + "step": 21959 + }, + { + "epoch": 0.52, + "grad_norm": 2.3131655837898437, + "learning_rate": 9.923696634409304e-06, + "loss": 1.0204, + "step": 21960 + }, + { + "epoch": 0.52, + "grad_norm": 1.9831172728801243, + "learning_rate": 9.922933615785073e-06, + "loss": 0.8409, + "step": 21961 + }, + { + "epoch": 0.52, + "grad_norm": 2.8694599395417253, + "learning_rate": 9.922170597609545e-06, + "loss": 1.0845, + "step": 21962 + }, + { + "epoch": 0.52, + "grad_norm": 2.0112774401924245, + "learning_rate": 9.921407579887164e-06, + "loss": 1.1674, + "step": 21963 + }, + { + "epoch": 0.52, + "grad_norm": 2.297693267629538, + "learning_rate": 9.920644562622375e-06, + "loss": 1.0362, + "step": 21964 + }, + { + "epoch": 0.52, + "grad_norm": 2.0907635718944513, + "learning_rate": 9.919881545819614e-06, + "loss": 1.0503, + "step": 21965 + }, + { + "epoch": 0.52, + "grad_norm": 1.099899315383225, + "learning_rate": 9.919118529483333e-06, + "loss": 0.9554, + "step": 21966 + }, + { + "epoch": 0.52, + "grad_norm": 1.079488638353009, + "learning_rate": 9.918355513617964e-06, + "loss": 0.9708, + "step": 21967 + }, + { + "epoch": 0.52, + "grad_norm": 2.175692460488818, + "learning_rate": 9.917592498227964e-06, + "loss": 1.0272, + "step": 21968 + }, + { + "epoch": 0.52, + "grad_norm": 1.1425442883176276, + "learning_rate": 9.916829483317758e-06, + "loss": 0.9682, + "step": 21969 + }, + { + "epoch": 0.52, + "grad_norm": 1.0685407671547245, + "learning_rate": 9.9160664688918e-06, + "loss": 0.9214, + "step": 21970 + }, + { + "epoch": 0.52, + "grad_norm": 1.12081596811769, + "learning_rate": 9.915303454954533e-06, + "loss": 0.949, + "step": 21971 + }, + { + "epoch": 0.52, + "grad_norm": 1.1302318803447302, + "learning_rate": 9.914540441510392e-06, + "loss": 0.919, + "step": 21972 + }, + { + "epoch": 0.52, + "grad_norm": 1.8808172089761206, + "learning_rate": 9.913777428563823e-06, + "loss": 0.9105, + "step": 21973 + }, + { + "epoch": 0.52, + "grad_norm": 1.121647764554065, + "learning_rate": 9.91301441611927e-06, + "loss": 1.048, + "step": 21974 + }, + { + "epoch": 0.52, + "grad_norm": 2.006880500338133, + "learning_rate": 9.912251404181177e-06, + "loss": 1.0987, + "step": 21975 + }, + { + "epoch": 0.52, + "grad_norm": 2.131976641429654, + "learning_rate": 9.911488392753982e-06, + "loss": 1.0512, + "step": 21976 + }, + { + "epoch": 0.52, + "grad_norm": 1.901346843441168, + "learning_rate": 9.910725381842127e-06, + "loss": 1.0687, + "step": 21977 + }, + { + "epoch": 0.52, + "grad_norm": 2.3864917907681735, + "learning_rate": 9.909962371450062e-06, + "loss": 0.962, + "step": 21978 + }, + { + "epoch": 0.52, + "grad_norm": 1.8300206454999854, + "learning_rate": 9.909199361582223e-06, + "loss": 1.0315, + "step": 21979 + }, + { + "epoch": 0.52, + "grad_norm": 2.0696859631345674, + "learning_rate": 9.908436352243052e-06, + "loss": 0.8246, + "step": 21980 + }, + { + "epoch": 0.52, + "grad_norm": 3.056688953130338, + "learning_rate": 9.907673343436994e-06, + "loss": 0.985, + "step": 21981 + }, + { + "epoch": 0.52, + "grad_norm": 1.7317507995286816, + "learning_rate": 9.906910335168495e-06, + "loss": 0.9281, + "step": 21982 + }, + { + "epoch": 0.52, + "grad_norm": 1.8277864140082956, + "learning_rate": 9.90614732744199e-06, + "loss": 0.9991, + "step": 21983 + }, + { + "epoch": 0.52, + "grad_norm": 2.1261379554573985, + "learning_rate": 9.905384320261924e-06, + "loss": 0.8207, + "step": 21984 + }, + { + "epoch": 0.52, + "grad_norm": 2.0880621019153174, + "learning_rate": 9.904621313632747e-06, + "loss": 0.9791, + "step": 21985 + }, + { + "epoch": 0.52, + "grad_norm": 2.1307235278053254, + "learning_rate": 9.903858307558887e-06, + "loss": 1.062, + "step": 21986 + }, + { + "epoch": 0.52, + "grad_norm": 2.162346316755547, + "learning_rate": 9.903095302044798e-06, + "loss": 1.2002, + "step": 21987 + }, + { + "epoch": 0.52, + "grad_norm": 2.167106410949043, + "learning_rate": 9.902332297094915e-06, + "loss": 1.1068, + "step": 21988 + }, + { + "epoch": 0.52, + "grad_norm": 2.0843175161036616, + "learning_rate": 9.901569292713689e-06, + "loss": 0.9356, + "step": 21989 + }, + { + "epoch": 0.52, + "grad_norm": 2.208027134127059, + "learning_rate": 9.900806288905557e-06, + "loss": 0.9417, + "step": 21990 + }, + { + "epoch": 0.52, + "grad_norm": 1.8562765975392836, + "learning_rate": 9.900043285674961e-06, + "loss": 1.0114, + "step": 21991 + }, + { + "epoch": 0.52, + "grad_norm": 2.059976881352433, + "learning_rate": 9.899280283026346e-06, + "loss": 1.1357, + "step": 21992 + }, + { + "epoch": 0.52, + "grad_norm": 2.0489314680973743, + "learning_rate": 9.898517280964152e-06, + "loss": 0.9838, + "step": 21993 + }, + { + "epoch": 0.52, + "grad_norm": 1.8388642424014388, + "learning_rate": 9.897754279492822e-06, + "loss": 1.0058, + "step": 21994 + }, + { + "epoch": 0.52, + "grad_norm": 1.860784416892232, + "learning_rate": 9.8969912786168e-06, + "loss": 1.0684, + "step": 21995 + }, + { + "epoch": 0.52, + "grad_norm": 2.0030492574162513, + "learning_rate": 9.896228278340529e-06, + "loss": 0.9692, + "step": 21996 + }, + { + "epoch": 0.52, + "grad_norm": 1.8833516023557126, + "learning_rate": 9.895465278668445e-06, + "loss": 0.9886, + "step": 21997 + }, + { + "epoch": 0.52, + "grad_norm": 2.241746746893828, + "learning_rate": 9.894702279604997e-06, + "loss": 1.0416, + "step": 21998 + }, + { + "epoch": 0.52, + "grad_norm": 1.8777130725337488, + "learning_rate": 9.89393928115463e-06, + "loss": 0.98, + "step": 21999 + }, + { + "epoch": 0.52, + "grad_norm": 2.4534536574522785, + "learning_rate": 9.893176283321779e-06, + "loss": 0.9713, + "step": 22000 + }, + { + "epoch": 0.52, + "grad_norm": 1.938457114412774, + "learning_rate": 9.892413286110886e-06, + "loss": 1.2159, + "step": 22001 + }, + { + "epoch": 0.52, + "grad_norm": 1.9612025635319215, + "learning_rate": 9.8916502895264e-06, + "loss": 0.9715, + "step": 22002 + }, + { + "epoch": 0.52, + "grad_norm": 2.064183782855676, + "learning_rate": 9.890887293572763e-06, + "loss": 1.0909, + "step": 22003 + }, + { + "epoch": 0.52, + "grad_norm": 1.1019652850664363, + "learning_rate": 9.890124298254412e-06, + "loss": 0.8762, + "step": 22004 + }, + { + "epoch": 0.52, + "grad_norm": 2.006068852391975, + "learning_rate": 9.88936130357579e-06, + "loss": 1.0408, + "step": 22005 + }, + { + "epoch": 0.52, + "grad_norm": 1.8988552341564426, + "learning_rate": 9.888598309541347e-06, + "loss": 1.0193, + "step": 22006 + }, + { + "epoch": 0.52, + "grad_norm": 2.0156998832971977, + "learning_rate": 9.887835316155516e-06, + "loss": 0.9921, + "step": 22007 + }, + { + "epoch": 0.52, + "grad_norm": 1.1110835413805138, + "learning_rate": 9.887072323422743e-06, + "loss": 0.9647, + "step": 22008 + }, + { + "epoch": 0.52, + "grad_norm": 1.1325239558066458, + "learning_rate": 9.886309331347471e-06, + "loss": 0.9566, + "step": 22009 + }, + { + "epoch": 0.52, + "grad_norm": 2.08655044416243, + "learning_rate": 9.885546339934145e-06, + "loss": 1.0505, + "step": 22010 + }, + { + "epoch": 0.52, + "grad_norm": 1.8290216713101186, + "learning_rate": 9.884783349187202e-06, + "loss": 0.9173, + "step": 22011 + }, + { + "epoch": 0.52, + "grad_norm": 1.8265154142501727, + "learning_rate": 9.884020359111087e-06, + "loss": 0.8879, + "step": 22012 + }, + { + "epoch": 0.52, + "grad_norm": 1.101868758826795, + "learning_rate": 9.883257369710244e-06, + "loss": 0.9155, + "step": 22013 + }, + { + "epoch": 0.52, + "grad_norm": 1.075665076053552, + "learning_rate": 9.88249438098911e-06, + "loss": 0.9517, + "step": 22014 + }, + { + "epoch": 0.52, + "grad_norm": 1.0658272269047224, + "learning_rate": 9.881731392952134e-06, + "loss": 0.9385, + "step": 22015 + }, + { + "epoch": 0.52, + "grad_norm": 1.8517089846918604, + "learning_rate": 9.880968405603751e-06, + "loss": 0.98, + "step": 22016 + }, + { + "epoch": 0.52, + "grad_norm": 2.3296778213928677, + "learning_rate": 9.880205418948414e-06, + "loss": 1.0106, + "step": 22017 + }, + { + "epoch": 0.52, + "grad_norm": 2.037246660158243, + "learning_rate": 9.879442432990553e-06, + "loss": 0.9326, + "step": 22018 + }, + { + "epoch": 0.52, + "grad_norm": 2.209720424016501, + "learning_rate": 9.87867944773462e-06, + "loss": 0.9769, + "step": 22019 + }, + { + "epoch": 0.52, + "grad_norm": 1.7113568773016987, + "learning_rate": 9.877916463185054e-06, + "loss": 0.8734, + "step": 22020 + }, + { + "epoch": 0.52, + "grad_norm": 1.9685448569705115, + "learning_rate": 9.877153479346296e-06, + "loss": 1.0883, + "step": 22021 + }, + { + "epoch": 0.52, + "grad_norm": 1.1422965173432722, + "learning_rate": 9.876390496222787e-06, + "loss": 0.9647, + "step": 22022 + }, + { + "epoch": 0.52, + "grad_norm": 1.9274151945093068, + "learning_rate": 9.875627513818973e-06, + "loss": 1.0217, + "step": 22023 + }, + { + "epoch": 0.52, + "grad_norm": 2.034892679498152, + "learning_rate": 9.874864532139299e-06, + "loss": 0.9518, + "step": 22024 + }, + { + "epoch": 0.52, + "grad_norm": 2.5276481234723502, + "learning_rate": 9.874101551188198e-06, + "loss": 0.9843, + "step": 22025 + }, + { + "epoch": 0.52, + "grad_norm": 1.846405450658112, + "learning_rate": 9.873338570970119e-06, + "loss": 1.138, + "step": 22026 + }, + { + "epoch": 0.52, + "grad_norm": 1.8631870791541953, + "learning_rate": 9.872575591489505e-06, + "loss": 0.998, + "step": 22027 + }, + { + "epoch": 0.52, + "grad_norm": 1.8341574481848295, + "learning_rate": 9.871812612750794e-06, + "loss": 1.0323, + "step": 22028 + }, + { + "epoch": 0.52, + "grad_norm": 1.9579532713202519, + "learning_rate": 9.87104963475843e-06, + "loss": 1.0805, + "step": 22029 + }, + { + "epoch": 0.52, + "grad_norm": 2.236438067932046, + "learning_rate": 9.870286657516856e-06, + "loss": 0.9763, + "step": 22030 + }, + { + "epoch": 0.52, + "grad_norm": 1.9770556457720656, + "learning_rate": 9.869523681030518e-06, + "loss": 1.1154, + "step": 22031 + }, + { + "epoch": 0.52, + "grad_norm": 2.1090977798969397, + "learning_rate": 9.868760705303852e-06, + "loss": 1.0448, + "step": 22032 + }, + { + "epoch": 0.52, + "grad_norm": 2.4580237389421, + "learning_rate": 9.8679977303413e-06, + "loss": 1.0666, + "step": 22033 + }, + { + "epoch": 0.52, + "grad_norm": 1.1786947138632973, + "learning_rate": 9.867234756147314e-06, + "loss": 0.9052, + "step": 22034 + }, + { + "epoch": 0.52, + "grad_norm": 2.4881361804171123, + "learning_rate": 9.866471782726323e-06, + "loss": 0.9526, + "step": 22035 + }, + { + "epoch": 0.52, + "grad_norm": 1.9440822498087593, + "learning_rate": 9.865708810082776e-06, + "loss": 1.0896, + "step": 22036 + }, + { + "epoch": 0.52, + "grad_norm": 1.8392004019089645, + "learning_rate": 9.864945838221114e-06, + "loss": 1.013, + "step": 22037 + }, + { + "epoch": 0.52, + "grad_norm": 2.1011699891387767, + "learning_rate": 9.864182867145784e-06, + "loss": 0.9535, + "step": 22038 + }, + { + "epoch": 0.52, + "grad_norm": 2.04944715680379, + "learning_rate": 9.863419896861222e-06, + "loss": 0.9926, + "step": 22039 + }, + { + "epoch": 0.52, + "grad_norm": 2.0267799028385833, + "learning_rate": 9.862656927371873e-06, + "loss": 1.0935, + "step": 22040 + }, + { + "epoch": 0.52, + "grad_norm": 1.9236678017943147, + "learning_rate": 9.861893958682183e-06, + "loss": 0.9615, + "step": 22041 + }, + { + "epoch": 0.52, + "grad_norm": 2.035103999851899, + "learning_rate": 9.861130990796583e-06, + "loss": 1.1576, + "step": 22042 + }, + { + "epoch": 0.52, + "grad_norm": 2.1561144188180785, + "learning_rate": 9.860368023719525e-06, + "loss": 1.0891, + "step": 22043 + }, + { + "epoch": 0.52, + "grad_norm": 1.119529176713111, + "learning_rate": 9.85960505745545e-06, + "loss": 0.9546, + "step": 22044 + }, + { + "epoch": 0.52, + "grad_norm": 1.9902249613044183, + "learning_rate": 9.858842092008799e-06, + "loss": 0.9922, + "step": 22045 + }, + { + "epoch": 0.52, + "grad_norm": 2.015513612128461, + "learning_rate": 9.85807912738401e-06, + "loss": 0.9778, + "step": 22046 + }, + { + "epoch": 0.52, + "grad_norm": 1.9822948421569082, + "learning_rate": 9.857316163585533e-06, + "loss": 0.8743, + "step": 22047 + }, + { + "epoch": 0.52, + "grad_norm": 2.1816592449419887, + "learning_rate": 9.856553200617805e-06, + "loss": 1.019, + "step": 22048 + }, + { + "epoch": 0.52, + "grad_norm": 2.077313201014092, + "learning_rate": 9.855790238485273e-06, + "loss": 0.9417, + "step": 22049 + }, + { + "epoch": 0.52, + "grad_norm": 1.8387620577917871, + "learning_rate": 9.85502727719237e-06, + "loss": 0.9692, + "step": 22050 + }, + { + "epoch": 0.52, + "grad_norm": 1.9355107787882362, + "learning_rate": 9.854264316743549e-06, + "loss": 1.0286, + "step": 22051 + }, + { + "epoch": 0.52, + "grad_norm": 2.2264481215524774, + "learning_rate": 9.853501357143247e-06, + "loss": 0.9635, + "step": 22052 + }, + { + "epoch": 0.52, + "grad_norm": 1.8698884885155103, + "learning_rate": 9.852738398395904e-06, + "loss": 1.0707, + "step": 22053 + }, + { + "epoch": 0.52, + "grad_norm": 2.1251996805316007, + "learning_rate": 9.851975440505964e-06, + "loss": 0.9905, + "step": 22054 + }, + { + "epoch": 0.52, + "grad_norm": 3.6328238861997657, + "learning_rate": 9.851212483477871e-06, + "loss": 1.0538, + "step": 22055 + }, + { + "epoch": 0.52, + "grad_norm": 2.1397146839766417, + "learning_rate": 9.850449527316069e-06, + "loss": 1.1244, + "step": 22056 + }, + { + "epoch": 0.52, + "grad_norm": 1.8456766165111382, + "learning_rate": 9.849686572024995e-06, + "loss": 1.0035, + "step": 22057 + }, + { + "epoch": 0.52, + "grad_norm": 2.191528774068706, + "learning_rate": 9.84892361760909e-06, + "loss": 1.0377, + "step": 22058 + }, + { + "epoch": 0.52, + "grad_norm": 1.959399864214502, + "learning_rate": 9.848160664072806e-06, + "loss": 1.0219, + "step": 22059 + }, + { + "epoch": 0.52, + "grad_norm": 1.8610054935596148, + "learning_rate": 9.847397711420574e-06, + "loss": 1.0217, + "step": 22060 + }, + { + "epoch": 0.52, + "grad_norm": 1.0673993607588073, + "learning_rate": 9.846634759656843e-06, + "loss": 0.9472, + "step": 22061 + }, + { + "epoch": 0.52, + "grad_norm": 2.346147723788626, + "learning_rate": 9.84587180878605e-06, + "loss": 1.0596, + "step": 22062 + }, + { + "epoch": 0.52, + "grad_norm": 2.137245713647126, + "learning_rate": 9.845108858812645e-06, + "loss": 1.1133, + "step": 22063 + }, + { + "epoch": 0.52, + "grad_norm": 2.1856203403517314, + "learning_rate": 9.844345909741061e-06, + "loss": 0.9688, + "step": 22064 + }, + { + "epoch": 0.52, + "grad_norm": 1.9307001247520847, + "learning_rate": 9.843582961575743e-06, + "loss": 1.0673, + "step": 22065 + }, + { + "epoch": 0.52, + "grad_norm": 1.7476063892599876, + "learning_rate": 9.842820014321143e-06, + "loss": 0.9549, + "step": 22066 + }, + { + "epoch": 0.52, + "grad_norm": 2.167067833633493, + "learning_rate": 9.842057067981686e-06, + "loss": 1.0032, + "step": 22067 + }, + { + "epoch": 0.52, + "grad_norm": 2.0174227284949384, + "learning_rate": 9.841294122561826e-06, + "loss": 0.9086, + "step": 22068 + }, + { + "epoch": 0.52, + "grad_norm": 2.2084838664222217, + "learning_rate": 9.840531178066e-06, + "loss": 1.0252, + "step": 22069 + }, + { + "epoch": 0.52, + "grad_norm": 1.941996957594576, + "learning_rate": 9.839768234498658e-06, + "loss": 0.956, + "step": 22070 + }, + { + "epoch": 0.52, + "grad_norm": 2.8604376906272413, + "learning_rate": 9.839005291864229e-06, + "loss": 0.9966, + "step": 22071 + }, + { + "epoch": 0.52, + "grad_norm": 1.8832948404949825, + "learning_rate": 9.838242350167164e-06, + "loss": 1.0071, + "step": 22072 + }, + { + "epoch": 0.52, + "grad_norm": 2.0357953604986068, + "learning_rate": 9.837479409411907e-06, + "loss": 0.9669, + "step": 22073 + }, + { + "epoch": 0.52, + "grad_norm": 2.0521644886464143, + "learning_rate": 9.836716469602891e-06, + "loss": 1.0444, + "step": 22074 + }, + { + "epoch": 0.52, + "grad_norm": 2.6307713663628145, + "learning_rate": 9.835953530744567e-06, + "loss": 1.0657, + "step": 22075 + }, + { + "epoch": 0.52, + "grad_norm": 1.9734944422941765, + "learning_rate": 9.835190592841371e-06, + "loss": 1.124, + "step": 22076 + }, + { + "epoch": 0.52, + "grad_norm": 1.0842344766636343, + "learning_rate": 9.834427655897751e-06, + "loss": 0.8402, + "step": 22077 + }, + { + "epoch": 0.52, + "grad_norm": 2.0286088049807134, + "learning_rate": 9.83366471991814e-06, + "loss": 0.9715, + "step": 22078 + }, + { + "epoch": 0.52, + "grad_norm": 1.060768881041186, + "learning_rate": 9.83290178490699e-06, + "loss": 0.9304, + "step": 22079 + }, + { + "epoch": 0.52, + "grad_norm": 1.9282679637001532, + "learning_rate": 9.83213885086874e-06, + "loss": 0.9803, + "step": 22080 + }, + { + "epoch": 0.52, + "grad_norm": 2.023672777612795, + "learning_rate": 9.831375917807829e-06, + "loss": 0.8569, + "step": 22081 + }, + { + "epoch": 0.52, + "grad_norm": 2.010121120031196, + "learning_rate": 9.830612985728699e-06, + "loss": 0.9939, + "step": 22082 + }, + { + "epoch": 0.52, + "grad_norm": 1.929999237155555, + "learning_rate": 9.829850054635794e-06, + "loss": 0.9421, + "step": 22083 + }, + { + "epoch": 0.52, + "grad_norm": 2.075767351294394, + "learning_rate": 9.829087124533559e-06, + "loss": 0.9628, + "step": 22084 + }, + { + "epoch": 0.52, + "grad_norm": 2.5557877203105632, + "learning_rate": 9.828324195426431e-06, + "loss": 1.0803, + "step": 22085 + }, + { + "epoch": 0.52, + "grad_norm": 1.0799370004364226, + "learning_rate": 9.827561267318852e-06, + "loss": 0.947, + "step": 22086 + }, + { + "epoch": 0.52, + "grad_norm": 1.0835653426254892, + "learning_rate": 9.82679834021527e-06, + "loss": 1.0058, + "step": 22087 + }, + { + "epoch": 0.52, + "grad_norm": 1.8732362988029119, + "learning_rate": 9.826035414120122e-06, + "loss": 1.041, + "step": 22088 + }, + { + "epoch": 0.52, + "grad_norm": 2.600453761255321, + "learning_rate": 9.825272489037847e-06, + "loss": 0.9205, + "step": 22089 + }, + { + "epoch": 0.52, + "grad_norm": 2.2613087100827944, + "learning_rate": 9.824509564972892e-06, + "loss": 0.8523, + "step": 22090 + }, + { + "epoch": 0.52, + "grad_norm": 2.185717642613666, + "learning_rate": 9.823746641929702e-06, + "loss": 0.9761, + "step": 22091 + }, + { + "epoch": 0.52, + "grad_norm": 1.0822235380478824, + "learning_rate": 9.822983719912712e-06, + "loss": 0.9936, + "step": 22092 + }, + { + "epoch": 0.52, + "grad_norm": 2.177857101835463, + "learning_rate": 9.822220798926367e-06, + "loss": 0.9859, + "step": 22093 + }, + { + "epoch": 0.52, + "grad_norm": 1.8922488903756445, + "learning_rate": 9.82145787897511e-06, + "loss": 0.9375, + "step": 22094 + }, + { + "epoch": 0.52, + "grad_norm": 2.203273356038515, + "learning_rate": 9.82069496006338e-06, + "loss": 1.0463, + "step": 22095 + }, + { + "epoch": 0.52, + "grad_norm": 2.058183310293528, + "learning_rate": 9.81993204219562e-06, + "loss": 1.0125, + "step": 22096 + }, + { + "epoch": 0.52, + "grad_norm": 2.5718474802253195, + "learning_rate": 9.819169125376275e-06, + "loss": 0.9274, + "step": 22097 + }, + { + "epoch": 0.52, + "grad_norm": 1.8564020101141467, + "learning_rate": 9.818406209609784e-06, + "loss": 1.0605, + "step": 22098 + }, + { + "epoch": 0.52, + "grad_norm": 2.2908738154192405, + "learning_rate": 9.817643294900587e-06, + "loss": 0.9657, + "step": 22099 + }, + { + "epoch": 0.52, + "grad_norm": 2.09795923793087, + "learning_rate": 9.81688038125313e-06, + "loss": 0.9848, + "step": 22100 + }, + { + "epoch": 0.52, + "grad_norm": 2.0558781168074525, + "learning_rate": 9.816117468671858e-06, + "loss": 1.0807, + "step": 22101 + }, + { + "epoch": 0.52, + "grad_norm": 1.9206793252985397, + "learning_rate": 9.815354557161203e-06, + "loss": 1.1372, + "step": 22102 + }, + { + "epoch": 0.52, + "grad_norm": 2.11372871083035, + "learning_rate": 9.814591646725611e-06, + "loss": 0.9218, + "step": 22103 + }, + { + "epoch": 0.52, + "grad_norm": 1.1151557120491187, + "learning_rate": 9.813828737369527e-06, + "loss": 0.9318, + "step": 22104 + }, + { + "epoch": 0.52, + "grad_norm": 2.445909217936978, + "learning_rate": 9.813065829097394e-06, + "loss": 0.9151, + "step": 22105 + }, + { + "epoch": 0.52, + "grad_norm": 2.6153596573123328, + "learning_rate": 9.81230292191365e-06, + "loss": 0.9258, + "step": 22106 + }, + { + "epoch": 0.52, + "grad_norm": 1.8342078430024291, + "learning_rate": 9.811540015822733e-06, + "loss": 1.0545, + "step": 22107 + }, + { + "epoch": 0.52, + "grad_norm": 2.381215581502816, + "learning_rate": 9.810777110829096e-06, + "loss": 1.0353, + "step": 22108 + }, + { + "epoch": 0.52, + "grad_norm": 2.11792610717222, + "learning_rate": 9.81001420693717e-06, + "loss": 1.0322, + "step": 22109 + }, + { + "epoch": 0.52, + "grad_norm": 2.081020583350384, + "learning_rate": 9.809251304151404e-06, + "loss": 0.9872, + "step": 22110 + }, + { + "epoch": 0.52, + "grad_norm": 1.8761030942084576, + "learning_rate": 9.808488402476234e-06, + "loss": 1.0109, + "step": 22111 + }, + { + "epoch": 0.52, + "grad_norm": 2.34197293344345, + "learning_rate": 9.80772550191611e-06, + "loss": 1.1122, + "step": 22112 + }, + { + "epoch": 0.52, + "grad_norm": 2.2990238571053183, + "learning_rate": 9.806962602475466e-06, + "loss": 1.0112, + "step": 22113 + }, + { + "epoch": 0.52, + "grad_norm": 2.7880352057402686, + "learning_rate": 9.806199704158745e-06, + "loss": 1.0756, + "step": 22114 + }, + { + "epoch": 0.52, + "grad_norm": 2.093022408563826, + "learning_rate": 9.805436806970397e-06, + "loss": 1.0393, + "step": 22115 + }, + { + "epoch": 0.52, + "grad_norm": 1.9313883805373178, + "learning_rate": 9.804673910914851e-06, + "loss": 1.0749, + "step": 22116 + }, + { + "epoch": 0.52, + "grad_norm": 2.3927917573395043, + "learning_rate": 9.803911015996557e-06, + "loss": 1.1503, + "step": 22117 + }, + { + "epoch": 0.52, + "grad_norm": 1.823269509528103, + "learning_rate": 9.803148122219955e-06, + "loss": 0.9979, + "step": 22118 + }, + { + "epoch": 0.52, + "grad_norm": 2.1856568022696643, + "learning_rate": 9.802385229589493e-06, + "loss": 0.9237, + "step": 22119 + }, + { + "epoch": 0.52, + "grad_norm": 1.964864757621771, + "learning_rate": 9.801622338109599e-06, + "loss": 0.9106, + "step": 22120 + }, + { + "epoch": 0.52, + "grad_norm": 2.1854797916327064, + "learning_rate": 9.800859447784726e-06, + "loss": 0.9635, + "step": 22121 + }, + { + "epoch": 0.52, + "grad_norm": 2.1710448054856775, + "learning_rate": 9.800096558619313e-06, + "loss": 0.9385, + "step": 22122 + }, + { + "epoch": 0.52, + "grad_norm": 2.0395195058237956, + "learning_rate": 9.799333670617797e-06, + "loss": 1.009, + "step": 22123 + }, + { + "epoch": 0.52, + "grad_norm": 2.1315620807179405, + "learning_rate": 9.798570783784627e-06, + "loss": 0.9134, + "step": 22124 + }, + { + "epoch": 0.52, + "grad_norm": 2.3465761141694697, + "learning_rate": 9.79780789812424e-06, + "loss": 1.0538, + "step": 22125 + }, + { + "epoch": 0.52, + "grad_norm": 1.9702574509642543, + "learning_rate": 9.797045013641084e-06, + "loss": 1.0161, + "step": 22126 + }, + { + "epoch": 0.52, + "grad_norm": 2.116575067282501, + "learning_rate": 9.79628213033959e-06, + "loss": 1.0021, + "step": 22127 + }, + { + "epoch": 0.52, + "grad_norm": 2.1355656030614165, + "learning_rate": 9.79551924822421e-06, + "loss": 1.0595, + "step": 22128 + }, + { + "epoch": 0.52, + "grad_norm": 2.4602168493546688, + "learning_rate": 9.794756367299383e-06, + "loss": 1.0628, + "step": 22129 + }, + { + "epoch": 0.52, + "grad_norm": 2.184948551934065, + "learning_rate": 9.793993487569544e-06, + "loss": 1.0463, + "step": 22130 + }, + { + "epoch": 0.52, + "grad_norm": 2.0390273496276436, + "learning_rate": 9.793230609039142e-06, + "loss": 1.0517, + "step": 22131 + }, + { + "epoch": 0.52, + "grad_norm": 1.8538279167297194, + "learning_rate": 9.792467731712618e-06, + "loss": 0.9092, + "step": 22132 + }, + { + "epoch": 0.52, + "grad_norm": 1.9110425509665947, + "learning_rate": 9.791704855594415e-06, + "loss": 0.9817, + "step": 22133 + }, + { + "epoch": 0.52, + "grad_norm": 1.9038358245158398, + "learning_rate": 9.79094198068897e-06, + "loss": 1.0595, + "step": 22134 + }, + { + "epoch": 0.52, + "grad_norm": 1.1537843072880227, + "learning_rate": 9.790179107000723e-06, + "loss": 0.9347, + "step": 22135 + }, + { + "epoch": 0.52, + "grad_norm": 1.9817975185029848, + "learning_rate": 9.789416234534126e-06, + "loss": 0.8079, + "step": 22136 + }, + { + "epoch": 0.52, + "grad_norm": 2.193953174814525, + "learning_rate": 9.788653363293612e-06, + "loss": 0.9671, + "step": 22137 + }, + { + "epoch": 0.52, + "grad_norm": 1.0639689138176194, + "learning_rate": 9.787890493283626e-06, + "loss": 0.9857, + "step": 22138 + }, + { + "epoch": 0.52, + "grad_norm": 1.084254483558579, + "learning_rate": 9.787127624508604e-06, + "loss": 0.9417, + "step": 22139 + }, + { + "epoch": 0.52, + "grad_norm": 1.089552273043666, + "learning_rate": 9.786364756972997e-06, + "loss": 1.0175, + "step": 22140 + }, + { + "epoch": 0.52, + "grad_norm": 1.9338197897511946, + "learning_rate": 9.785601890681242e-06, + "loss": 1.0051, + "step": 22141 + }, + { + "epoch": 0.52, + "grad_norm": 1.9596780281513972, + "learning_rate": 9.78483902563778e-06, + "loss": 1.0248, + "step": 22142 + }, + { + "epoch": 0.52, + "grad_norm": 1.1722970794486025, + "learning_rate": 9.784076161847055e-06, + "loss": 0.9593, + "step": 22143 + }, + { + "epoch": 0.52, + "grad_norm": 2.0841111269303783, + "learning_rate": 9.783313299313502e-06, + "loss": 1.064, + "step": 22144 + }, + { + "epoch": 0.52, + "grad_norm": 1.8699483366328704, + "learning_rate": 9.78255043804157e-06, + "loss": 0.9532, + "step": 22145 + }, + { + "epoch": 0.52, + "grad_norm": 4.286568382307137, + "learning_rate": 9.781787578035698e-06, + "loss": 0.9398, + "step": 22146 + }, + { + "epoch": 0.52, + "grad_norm": 3.2787162688231226, + "learning_rate": 9.781024719300332e-06, + "loss": 0.9629, + "step": 22147 + }, + { + "epoch": 0.52, + "grad_norm": 2.369657917580069, + "learning_rate": 9.780261861839903e-06, + "loss": 1.1541, + "step": 22148 + }, + { + "epoch": 0.52, + "grad_norm": 2.1316696744129215, + "learning_rate": 9.779499005658862e-06, + "loss": 1.0251, + "step": 22149 + }, + { + "epoch": 0.52, + "grad_norm": 1.1178078658790658, + "learning_rate": 9.77873615076165e-06, + "loss": 0.9555, + "step": 22150 + }, + { + "epoch": 0.52, + "grad_norm": 2.1707416210516017, + "learning_rate": 9.777973297152703e-06, + "loss": 1.0004, + "step": 22151 + }, + { + "epoch": 0.52, + "grad_norm": 1.0595227242915264, + "learning_rate": 9.777210444836463e-06, + "loss": 0.8781, + "step": 22152 + }, + { + "epoch": 0.52, + "grad_norm": 1.9646338023963108, + "learning_rate": 9.77644759381738e-06, + "loss": 1.005, + "step": 22153 + }, + { + "epoch": 0.52, + "grad_norm": 2.0535814090927342, + "learning_rate": 9.775684744099889e-06, + "loss": 1.0561, + "step": 22154 + }, + { + "epoch": 0.52, + "grad_norm": 2.070012991423781, + "learning_rate": 9.77492189568843e-06, + "loss": 1.0085, + "step": 22155 + }, + { + "epoch": 0.52, + "grad_norm": 1.9704921757664366, + "learning_rate": 9.774159048587445e-06, + "loss": 1.0416, + "step": 22156 + }, + { + "epoch": 0.52, + "grad_norm": 2.127115802568839, + "learning_rate": 9.773396202801383e-06, + "loss": 1.0196, + "step": 22157 + }, + { + "epoch": 0.52, + "grad_norm": 2.239364952574122, + "learning_rate": 9.772633358334678e-06, + "loss": 0.9917, + "step": 22158 + }, + { + "epoch": 0.52, + "grad_norm": 2.0372528847857665, + "learning_rate": 9.771870515191773e-06, + "loss": 0.9989, + "step": 22159 + }, + { + "epoch": 0.52, + "grad_norm": 2.0394440930805287, + "learning_rate": 9.771107673377106e-06, + "loss": 0.9942, + "step": 22160 + }, + { + "epoch": 0.52, + "grad_norm": 2.1080049739098574, + "learning_rate": 9.77034483289513e-06, + "loss": 0.9954, + "step": 22161 + }, + { + "epoch": 0.52, + "grad_norm": 2.0226030309025482, + "learning_rate": 9.769581993750277e-06, + "loss": 0.9439, + "step": 22162 + }, + { + "epoch": 0.52, + "grad_norm": 2.117612197408257, + "learning_rate": 9.768819155946986e-06, + "loss": 0.8951, + "step": 22163 + }, + { + "epoch": 0.52, + "grad_norm": 2.272534754463617, + "learning_rate": 9.768056319489711e-06, + "loss": 1.0603, + "step": 22164 + }, + { + "epoch": 0.52, + "grad_norm": 2.085648377400191, + "learning_rate": 9.767293484382879e-06, + "loss": 0.9423, + "step": 22165 + }, + { + "epoch": 0.52, + "grad_norm": 1.1366212513614875, + "learning_rate": 9.76653065063094e-06, + "loss": 0.9672, + "step": 22166 + }, + { + "epoch": 0.52, + "grad_norm": 2.4585022040767557, + "learning_rate": 9.76576781823833e-06, + "loss": 1.0365, + "step": 22167 + }, + { + "epoch": 0.52, + "grad_norm": 1.9527653696843896, + "learning_rate": 9.765004987209504e-06, + "loss": 0.9828, + "step": 22168 + }, + { + "epoch": 0.52, + "grad_norm": 1.969803556839126, + "learning_rate": 9.764242157548884e-06, + "loss": 0.9039, + "step": 22169 + }, + { + "epoch": 0.52, + "grad_norm": 1.9884581696755916, + "learning_rate": 9.763479329260924e-06, + "loss": 0.9677, + "step": 22170 + }, + { + "epoch": 0.52, + "grad_norm": 1.9029951286310183, + "learning_rate": 9.762716502350058e-06, + "loss": 1.0312, + "step": 22171 + }, + { + "epoch": 0.52, + "grad_norm": 1.1972090192795999, + "learning_rate": 9.761953676820738e-06, + "loss": 0.9963, + "step": 22172 + }, + { + "epoch": 0.52, + "grad_norm": 1.8272771323616128, + "learning_rate": 9.761190852677397e-06, + "loss": 0.9597, + "step": 22173 + }, + { + "epoch": 0.52, + "grad_norm": 2.1547680182459414, + "learning_rate": 9.760428029924479e-06, + "loss": 1.1893, + "step": 22174 + }, + { + "epoch": 0.52, + "grad_norm": 1.8354963936899378, + "learning_rate": 9.759665208566425e-06, + "loss": 1.0375, + "step": 22175 + }, + { + "epoch": 0.52, + "grad_norm": 1.1342574283881197, + "learning_rate": 9.758902388607672e-06, + "loss": 1.0025, + "step": 22176 + }, + { + "epoch": 0.52, + "grad_norm": 1.8761065112044153, + "learning_rate": 9.75813957005267e-06, + "loss": 1.0823, + "step": 22177 + }, + { + "epoch": 0.52, + "grad_norm": 2.0110410849806573, + "learning_rate": 9.757376752905854e-06, + "loss": 0.9809, + "step": 22178 + }, + { + "epoch": 0.52, + "grad_norm": 1.8653014197208586, + "learning_rate": 9.756613937171669e-06, + "loss": 1.0953, + "step": 22179 + }, + { + "epoch": 0.52, + "grad_norm": 2.3216132107991996, + "learning_rate": 9.755851122854553e-06, + "loss": 0.9067, + "step": 22180 + }, + { + "epoch": 0.52, + "grad_norm": 1.923987647162149, + "learning_rate": 9.755088309958949e-06, + "loss": 0.8961, + "step": 22181 + }, + { + "epoch": 0.52, + "grad_norm": 1.978601977477628, + "learning_rate": 9.7543254984893e-06, + "loss": 0.908, + "step": 22182 + }, + { + "epoch": 0.52, + "grad_norm": 1.1264381781713706, + "learning_rate": 9.753562688450047e-06, + "loss": 1.0116, + "step": 22183 + }, + { + "epoch": 0.52, + "grad_norm": 2.0318500611356054, + "learning_rate": 9.752799879845624e-06, + "loss": 1.0181, + "step": 22184 + }, + { + "epoch": 0.52, + "grad_norm": 1.9175763207620613, + "learning_rate": 9.752037072680484e-06, + "loss": 1.0294, + "step": 22185 + }, + { + "epoch": 0.52, + "grad_norm": 2.0174478512367124, + "learning_rate": 9.751274266959064e-06, + "loss": 1.0483, + "step": 22186 + }, + { + "epoch": 0.52, + "grad_norm": 1.8259542748656552, + "learning_rate": 9.7505114626858e-06, + "loss": 0.8939, + "step": 22187 + }, + { + "epoch": 0.52, + "grad_norm": 2.475752072054768, + "learning_rate": 9.749748659865136e-06, + "loss": 0.9826, + "step": 22188 + }, + { + "epoch": 0.52, + "grad_norm": 1.968633010345827, + "learning_rate": 9.74898585850152e-06, + "loss": 1.093, + "step": 22189 + }, + { + "epoch": 0.52, + "grad_norm": 1.9671975502358872, + "learning_rate": 9.748223058599383e-06, + "loss": 1.0046, + "step": 22190 + }, + { + "epoch": 0.52, + "grad_norm": 2.048418279744245, + "learning_rate": 9.747460260163173e-06, + "loss": 1.2304, + "step": 22191 + }, + { + "epoch": 0.52, + "grad_norm": 2.17254016080667, + "learning_rate": 9.746697463197327e-06, + "loss": 0.9605, + "step": 22192 + }, + { + "epoch": 0.52, + "grad_norm": 1.985864680631791, + "learning_rate": 9.745934667706291e-06, + "loss": 1.0694, + "step": 22193 + }, + { + "epoch": 0.52, + "grad_norm": 2.0876822765772025, + "learning_rate": 9.745171873694504e-06, + "loss": 1.0292, + "step": 22194 + }, + { + "epoch": 0.52, + "grad_norm": 2.0095804328299356, + "learning_rate": 9.744409081166406e-06, + "loss": 0.9296, + "step": 22195 + }, + { + "epoch": 0.52, + "grad_norm": 2.00110966778232, + "learning_rate": 9.743646290126444e-06, + "loss": 0.9633, + "step": 22196 + }, + { + "epoch": 0.52, + "grad_norm": 1.9207445751137409, + "learning_rate": 9.742883500579046e-06, + "loss": 0.893, + "step": 22197 + }, + { + "epoch": 0.52, + "grad_norm": 1.8251900540169474, + "learning_rate": 9.742120712528665e-06, + "loss": 0.9162, + "step": 22198 + }, + { + "epoch": 0.52, + "grad_norm": 1.925903146950133, + "learning_rate": 9.741357925979737e-06, + "loss": 1.1442, + "step": 22199 + }, + { + "epoch": 0.52, + "grad_norm": 1.845370904917671, + "learning_rate": 9.740595140936714e-06, + "loss": 0.9136, + "step": 22200 + }, + { + "epoch": 0.52, + "grad_norm": 2.0424601050913673, + "learning_rate": 9.73983235740402e-06, + "loss": 1.0361, + "step": 22201 + }, + { + "epoch": 0.52, + "grad_norm": 1.8897318183999823, + "learning_rate": 9.739069575386105e-06, + "loss": 1.0645, + "step": 22202 + }, + { + "epoch": 0.52, + "grad_norm": 2.322773819153011, + "learning_rate": 9.738306794887413e-06, + "loss": 0.9704, + "step": 22203 + }, + { + "epoch": 0.52, + "grad_norm": 2.092534149693038, + "learning_rate": 9.73754401591238e-06, + "loss": 1.0612, + "step": 22204 + }, + { + "epoch": 0.52, + "grad_norm": 1.101431892985108, + "learning_rate": 9.736781238465444e-06, + "loss": 0.9288, + "step": 22205 + }, + { + "epoch": 0.52, + "grad_norm": 2.0044885612158754, + "learning_rate": 9.736018462551057e-06, + "loss": 1.0166, + "step": 22206 + }, + { + "epoch": 0.52, + "grad_norm": 2.168079803388977, + "learning_rate": 9.735255688173654e-06, + "loss": 1.0255, + "step": 22207 + }, + { + "epoch": 0.52, + "grad_norm": 1.9233508014705767, + "learning_rate": 9.734492915337675e-06, + "loss": 1.0753, + "step": 22208 + }, + { + "epoch": 0.52, + "grad_norm": 1.9893861000898774, + "learning_rate": 9.733730144047558e-06, + "loss": 0.9756, + "step": 22209 + }, + { + "epoch": 0.52, + "grad_norm": 1.8569017716499006, + "learning_rate": 9.732967374307755e-06, + "loss": 0.8963, + "step": 22210 + }, + { + "epoch": 0.52, + "grad_norm": 1.8393021128475027, + "learning_rate": 9.732204606122698e-06, + "loss": 0.9233, + "step": 22211 + }, + { + "epoch": 0.52, + "grad_norm": 2.1176794677466564, + "learning_rate": 9.731441839496828e-06, + "loss": 0.9679, + "step": 22212 + }, + { + "epoch": 0.52, + "grad_norm": 1.9229242454337074, + "learning_rate": 9.73067907443459e-06, + "loss": 0.9987, + "step": 22213 + }, + { + "epoch": 0.52, + "grad_norm": 1.8883130509088213, + "learning_rate": 9.729916310940427e-06, + "loss": 1.0081, + "step": 22214 + }, + { + "epoch": 0.52, + "grad_norm": 2.0955385965012683, + "learning_rate": 9.729153549018775e-06, + "loss": 1.0929, + "step": 22215 + }, + { + "epoch": 0.52, + "grad_norm": 1.93168376224997, + "learning_rate": 9.728390788674074e-06, + "loss": 0.9616, + "step": 22216 + }, + { + "epoch": 0.52, + "grad_norm": 1.8778851770877905, + "learning_rate": 9.727628029910774e-06, + "loss": 0.9331, + "step": 22217 + }, + { + "epoch": 0.52, + "grad_norm": 2.212635034466081, + "learning_rate": 9.726865272733304e-06, + "loss": 0.9727, + "step": 22218 + }, + { + "epoch": 0.52, + "grad_norm": 2.267635285647283, + "learning_rate": 9.726102517146112e-06, + "loss": 0.9998, + "step": 22219 + }, + { + "epoch": 0.52, + "grad_norm": 1.896970713429088, + "learning_rate": 9.725339763153636e-06, + "loss": 0.8884, + "step": 22220 + }, + { + "epoch": 0.52, + "grad_norm": 2.4267418816732436, + "learning_rate": 9.724577010760324e-06, + "loss": 1.0319, + "step": 22221 + }, + { + "epoch": 0.52, + "grad_norm": 1.8131086427427183, + "learning_rate": 9.723814259970608e-06, + "loss": 0.9314, + "step": 22222 + }, + { + "epoch": 0.52, + "grad_norm": 1.8136698615198341, + "learning_rate": 9.723051510788934e-06, + "loss": 1.1693, + "step": 22223 + }, + { + "epoch": 0.52, + "grad_norm": 2.1973809660202495, + "learning_rate": 9.722288763219744e-06, + "loss": 1.0416, + "step": 22224 + }, + { + "epoch": 0.52, + "grad_norm": 2.0635620456375503, + "learning_rate": 9.721526017267473e-06, + "loss": 1.0243, + "step": 22225 + }, + { + "epoch": 0.52, + "grad_norm": 2.2880322899433176, + "learning_rate": 9.720763272936568e-06, + "loss": 0.9249, + "step": 22226 + }, + { + "epoch": 0.52, + "grad_norm": 2.281352908976437, + "learning_rate": 9.720000530231466e-06, + "loss": 1.0708, + "step": 22227 + }, + { + "epoch": 0.52, + "grad_norm": 2.038440944860829, + "learning_rate": 9.719237789156612e-06, + "loss": 0.9446, + "step": 22228 + }, + { + "epoch": 0.52, + "grad_norm": 2.2127218868132315, + "learning_rate": 9.718475049716441e-06, + "loss": 1.0384, + "step": 22229 + }, + { + "epoch": 0.52, + "grad_norm": 1.8955484172917514, + "learning_rate": 9.7177123119154e-06, + "loss": 1.0188, + "step": 22230 + }, + { + "epoch": 0.52, + "grad_norm": 2.00427387759554, + "learning_rate": 9.71694957575793e-06, + "loss": 0.9153, + "step": 22231 + }, + { + "epoch": 0.52, + "grad_norm": 1.9873972308264758, + "learning_rate": 9.716186841248465e-06, + "loss": 0.9191, + "step": 22232 + }, + { + "epoch": 0.52, + "grad_norm": 2.977246034306063, + "learning_rate": 9.715424108391448e-06, + "loss": 0.9585, + "step": 22233 + }, + { + "epoch": 0.52, + "grad_norm": 2.2075951934421, + "learning_rate": 9.714661377191324e-06, + "loss": 0.964, + "step": 22234 + }, + { + "epoch": 0.52, + "grad_norm": 1.0977466932466498, + "learning_rate": 9.713898647652535e-06, + "loss": 0.9589, + "step": 22235 + }, + { + "epoch": 0.52, + "grad_norm": 1.0975747410158099, + "learning_rate": 9.713135919779516e-06, + "loss": 0.9794, + "step": 22236 + }, + { + "epoch": 0.52, + "grad_norm": 1.8641745509762373, + "learning_rate": 9.712373193576708e-06, + "loss": 1.0327, + "step": 22237 + }, + { + "epoch": 0.52, + "grad_norm": 1.8825225838117337, + "learning_rate": 9.71161046904856e-06, + "loss": 0.9374, + "step": 22238 + }, + { + "epoch": 0.52, + "grad_norm": 2.0740266082783263, + "learning_rate": 9.710847746199505e-06, + "loss": 0.8936, + "step": 22239 + }, + { + "epoch": 0.52, + "grad_norm": 3.7546735230246036, + "learning_rate": 9.710085025033984e-06, + "loss": 0.8992, + "step": 22240 + }, + { + "epoch": 0.52, + "grad_norm": 1.9467404870102891, + "learning_rate": 9.709322305556438e-06, + "loss": 1.0985, + "step": 22241 + }, + { + "epoch": 0.52, + "grad_norm": 2.0086387000558945, + "learning_rate": 9.708559587771316e-06, + "loss": 0.9856, + "step": 22242 + }, + { + "epoch": 0.52, + "grad_norm": 2.2230584160170523, + "learning_rate": 9.707796871683049e-06, + "loss": 1.0627, + "step": 22243 + }, + { + "epoch": 0.52, + "grad_norm": 2.0335133140271573, + "learning_rate": 9.70703415729608e-06, + "loss": 1.0165, + "step": 22244 + }, + { + "epoch": 0.52, + "grad_norm": 2.073354618071437, + "learning_rate": 9.706271444614854e-06, + "loss": 1.0355, + "step": 22245 + }, + { + "epoch": 0.52, + "grad_norm": 1.0997682553182837, + "learning_rate": 9.705508733643805e-06, + "loss": 0.984, + "step": 22246 + }, + { + "epoch": 0.52, + "grad_norm": 2.070316348769551, + "learning_rate": 9.704746024387377e-06, + "loss": 1.0232, + "step": 22247 + }, + { + "epoch": 0.52, + "grad_norm": 2.1376042415246115, + "learning_rate": 9.703983316850012e-06, + "loss": 1.0909, + "step": 22248 + }, + { + "epoch": 0.52, + "grad_norm": 2.3187907284841356, + "learning_rate": 9.703220611036156e-06, + "loss": 1.0337, + "step": 22249 + }, + { + "epoch": 0.52, + "grad_norm": 2.1270795015923625, + "learning_rate": 9.702457906950235e-06, + "loss": 0.9518, + "step": 22250 + }, + { + "epoch": 0.52, + "grad_norm": 2.0540264855861636, + "learning_rate": 9.701695204596702e-06, + "loss": 1.0741, + "step": 22251 + }, + { + "epoch": 0.52, + "grad_norm": 2.0635311633557554, + "learning_rate": 9.700932503979996e-06, + "loss": 0.9407, + "step": 22252 + }, + { + "epoch": 0.52, + "grad_norm": 1.9658956699216177, + "learning_rate": 9.700169805104553e-06, + "loss": 0.9966, + "step": 22253 + }, + { + "epoch": 0.52, + "grad_norm": 1.094250335099277, + "learning_rate": 9.699407107974815e-06, + "loss": 0.8989, + "step": 22254 + }, + { + "epoch": 0.52, + "grad_norm": 2.065447294777927, + "learning_rate": 9.698644412595225e-06, + "loss": 0.9163, + "step": 22255 + }, + { + "epoch": 0.52, + "grad_norm": 2.020445440706943, + "learning_rate": 9.697881718970226e-06, + "loss": 1.0723, + "step": 22256 + }, + { + "epoch": 0.52, + "grad_norm": 1.9215304971627958, + "learning_rate": 9.697119027104252e-06, + "loss": 1.0215, + "step": 22257 + }, + { + "epoch": 0.52, + "grad_norm": 1.8639862474473854, + "learning_rate": 9.696356337001745e-06, + "loss": 0.956, + "step": 22258 + }, + { + "epoch": 0.52, + "grad_norm": 2.3868792243068224, + "learning_rate": 9.695593648667152e-06, + "loss": 1.0481, + "step": 22259 + }, + { + "epoch": 0.52, + "grad_norm": 1.9290201522936943, + "learning_rate": 9.694830962104907e-06, + "loss": 0.9875, + "step": 22260 + }, + { + "epoch": 0.52, + "grad_norm": 2.5609017983345157, + "learning_rate": 9.694068277319451e-06, + "loss": 0.9385, + "step": 22261 + }, + { + "epoch": 0.52, + "grad_norm": 2.0147243074969223, + "learning_rate": 9.693305594315229e-06, + "loss": 1.0718, + "step": 22262 + }, + { + "epoch": 0.52, + "grad_norm": 1.9821143005098154, + "learning_rate": 9.69254291309668e-06, + "loss": 0.9722, + "step": 22263 + }, + { + "epoch": 0.52, + "grad_norm": 2.0177589429645466, + "learning_rate": 9.691780233668241e-06, + "loss": 0.961, + "step": 22264 + }, + { + "epoch": 0.52, + "grad_norm": 1.7462526765525224, + "learning_rate": 9.691017556034354e-06, + "loss": 1.0046, + "step": 22265 + }, + { + "epoch": 0.52, + "grad_norm": 1.9160982633262398, + "learning_rate": 9.690254880199466e-06, + "loss": 1.1069, + "step": 22266 + }, + { + "epoch": 0.52, + "grad_norm": 1.9502822362814496, + "learning_rate": 9.689492206168006e-06, + "loss": 1.0149, + "step": 22267 + }, + { + "epoch": 0.52, + "grad_norm": 2.026071855096224, + "learning_rate": 9.688729533944423e-06, + "loss": 1.0545, + "step": 22268 + }, + { + "epoch": 0.52, + "grad_norm": 2.340531990085494, + "learning_rate": 9.687966863533154e-06, + "loss": 0.9656, + "step": 22269 + }, + { + "epoch": 0.52, + "grad_norm": 2.153640767124764, + "learning_rate": 9.687204194938645e-06, + "loss": 0.9321, + "step": 22270 + }, + { + "epoch": 0.52, + "grad_norm": 2.0200953297620603, + "learning_rate": 9.686441528165327e-06, + "loss": 1.0397, + "step": 22271 + }, + { + "epoch": 0.52, + "grad_norm": 2.329599990035065, + "learning_rate": 9.685678863217649e-06, + "loss": 1.0079, + "step": 22272 + }, + { + "epoch": 0.52, + "grad_norm": 2.235263437964686, + "learning_rate": 9.68491620010005e-06, + "loss": 1.0749, + "step": 22273 + }, + { + "epoch": 0.52, + "grad_norm": 2.4232249134695105, + "learning_rate": 9.684153538816963e-06, + "loss": 1.0623, + "step": 22274 + }, + { + "epoch": 0.52, + "grad_norm": 2.156397487729203, + "learning_rate": 9.683390879372837e-06, + "loss": 0.9597, + "step": 22275 + }, + { + "epoch": 0.52, + "grad_norm": 2.134693319687668, + "learning_rate": 9.682628221772109e-06, + "loss": 1.0165, + "step": 22276 + }, + { + "epoch": 0.52, + "grad_norm": 1.9141902340263368, + "learning_rate": 9.681865566019224e-06, + "loss": 1.1036, + "step": 22277 + }, + { + "epoch": 0.52, + "grad_norm": 2.8081021255814202, + "learning_rate": 9.68110291211861e-06, + "loss": 1.047, + "step": 22278 + }, + { + "epoch": 0.52, + "grad_norm": 1.9422999500874711, + "learning_rate": 9.680340260074721e-06, + "loss": 1.0462, + "step": 22279 + }, + { + "epoch": 0.52, + "grad_norm": 1.8880624049709918, + "learning_rate": 9.679577609891994e-06, + "loss": 1.0879, + "step": 22280 + }, + { + "epoch": 0.52, + "grad_norm": 1.8588782438801956, + "learning_rate": 9.678814961574866e-06, + "loss": 1.0569, + "step": 22281 + }, + { + "epoch": 0.52, + "grad_norm": 2.685133511548552, + "learning_rate": 9.678052315127776e-06, + "loss": 1.0147, + "step": 22282 + }, + { + "epoch": 0.52, + "grad_norm": 2.0813355887290546, + "learning_rate": 9.67728967055517e-06, + "loss": 1.0479, + "step": 22283 + }, + { + "epoch": 0.52, + "grad_norm": 2.3938799328367697, + "learning_rate": 9.676527027861487e-06, + "loss": 1.0023, + "step": 22284 + }, + { + "epoch": 0.53, + "grad_norm": 1.9312476929733584, + "learning_rate": 9.675764387051164e-06, + "loss": 0.9626, + "step": 22285 + }, + { + "epoch": 0.53, + "grad_norm": 2.0833060382629522, + "learning_rate": 9.675001748128641e-06, + "loss": 0.9553, + "step": 22286 + }, + { + "epoch": 0.53, + "grad_norm": 2.0676246314351596, + "learning_rate": 9.674239111098367e-06, + "loss": 0.9534, + "step": 22287 + }, + { + "epoch": 0.53, + "grad_norm": 1.907876456813351, + "learning_rate": 9.673476475964773e-06, + "loss": 1.0022, + "step": 22288 + }, + { + "epoch": 0.53, + "grad_norm": 1.9228786126472266, + "learning_rate": 9.6727138427323e-06, + "loss": 0.9091, + "step": 22289 + }, + { + "epoch": 0.53, + "grad_norm": 1.8802823049711241, + "learning_rate": 9.671951211405391e-06, + "loss": 0.9186, + "step": 22290 + }, + { + "epoch": 0.53, + "grad_norm": 1.7216334414505354, + "learning_rate": 9.67118858198849e-06, + "loss": 0.8724, + "step": 22291 + }, + { + "epoch": 0.53, + "grad_norm": 2.006611529795243, + "learning_rate": 9.67042595448603e-06, + "loss": 0.9867, + "step": 22292 + }, + { + "epoch": 0.53, + "grad_norm": 1.0780509715130633, + "learning_rate": 9.669663328902455e-06, + "loss": 1.0085, + "step": 22293 + }, + { + "epoch": 0.53, + "grad_norm": 2.1712838465506232, + "learning_rate": 9.668900705242208e-06, + "loss": 1.1244, + "step": 22294 + }, + { + "epoch": 0.53, + "grad_norm": 1.928009765625827, + "learning_rate": 9.66813808350972e-06, + "loss": 0.9372, + "step": 22295 + }, + { + "epoch": 0.53, + "grad_norm": 2.1061013734683813, + "learning_rate": 9.66737546370944e-06, + "loss": 0.9997, + "step": 22296 + }, + { + "epoch": 0.53, + "grad_norm": 1.8175958652540596, + "learning_rate": 9.666612845845803e-06, + "loss": 1.0668, + "step": 22297 + }, + { + "epoch": 0.53, + "grad_norm": 2.1419742443038934, + "learning_rate": 9.665850229923258e-06, + "loss": 1.0809, + "step": 22298 + }, + { + "epoch": 0.53, + "grad_norm": 1.8769718411372864, + "learning_rate": 9.665087615946231e-06, + "loss": 0.9162, + "step": 22299 + }, + { + "epoch": 0.53, + "grad_norm": 1.9847482093074351, + "learning_rate": 9.664325003919176e-06, + "loss": 1.0659, + "step": 22300 + }, + { + "epoch": 0.53, + "grad_norm": 1.9866224052076273, + "learning_rate": 9.66356239384652e-06, + "loss": 0.9828, + "step": 22301 + }, + { + "epoch": 0.53, + "grad_norm": 2.0383858683941525, + "learning_rate": 9.66279978573272e-06, + "loss": 1.0245, + "step": 22302 + }, + { + "epoch": 0.53, + "grad_norm": 1.904512031578153, + "learning_rate": 9.662037179582199e-06, + "loss": 1.0385, + "step": 22303 + }, + { + "epoch": 0.53, + "grad_norm": 1.887548458704143, + "learning_rate": 9.661274575399406e-06, + "loss": 1.0868, + "step": 22304 + }, + { + "epoch": 0.53, + "grad_norm": 2.038264841828441, + "learning_rate": 9.660511973188783e-06, + "loss": 0.92, + "step": 22305 + }, + { + "epoch": 0.53, + "grad_norm": 2.0644292677788347, + "learning_rate": 9.659749372954764e-06, + "loss": 1.0699, + "step": 22306 + }, + { + "epoch": 0.53, + "grad_norm": 1.1387774962796118, + "learning_rate": 9.65898677470179e-06, + "loss": 1.0401, + "step": 22307 + }, + { + "epoch": 0.53, + "grad_norm": 2.037622542929765, + "learning_rate": 9.658224178434305e-06, + "loss": 0.9928, + "step": 22308 + }, + { + "epoch": 0.53, + "grad_norm": 2.0933691859882937, + "learning_rate": 9.657461584156748e-06, + "loss": 0.9662, + "step": 22309 + }, + { + "epoch": 0.53, + "grad_norm": 1.8981172756144506, + "learning_rate": 9.656698991873555e-06, + "loss": 0.9925, + "step": 22310 + }, + { + "epoch": 0.53, + "grad_norm": 2.306490077478641, + "learning_rate": 9.655936401589173e-06, + "loss": 0.9708, + "step": 22311 + }, + { + "epoch": 0.53, + "grad_norm": 1.837109104083372, + "learning_rate": 9.655173813308037e-06, + "loss": 0.961, + "step": 22312 + }, + { + "epoch": 0.53, + "grad_norm": 2.013937410002123, + "learning_rate": 9.654411227034588e-06, + "loss": 1.0431, + "step": 22313 + }, + { + "epoch": 0.53, + "grad_norm": 2.2058696212204643, + "learning_rate": 9.653648642773264e-06, + "loss": 1.0258, + "step": 22314 + }, + { + "epoch": 0.53, + "grad_norm": 1.8896995909615362, + "learning_rate": 9.652886060528509e-06, + "loss": 0.9664, + "step": 22315 + }, + { + "epoch": 0.53, + "grad_norm": 1.7839919039343115, + "learning_rate": 9.652123480304764e-06, + "loss": 0.9436, + "step": 22316 + }, + { + "epoch": 0.53, + "grad_norm": 1.9096235302041642, + "learning_rate": 9.651360902106463e-06, + "loss": 0.9603, + "step": 22317 + }, + { + "epoch": 0.53, + "grad_norm": 2.3039534482781403, + "learning_rate": 9.650598325938048e-06, + "loss": 1.1385, + "step": 22318 + }, + { + "epoch": 0.53, + "grad_norm": 2.0032538214965836, + "learning_rate": 9.649835751803965e-06, + "loss": 1.0544, + "step": 22319 + }, + { + "epoch": 0.53, + "grad_norm": 2.1079669918883646, + "learning_rate": 9.649073179708646e-06, + "loss": 1.007, + "step": 22320 + }, + { + "epoch": 0.53, + "grad_norm": 1.8680367106079028, + "learning_rate": 9.648310609656534e-06, + "loss": 1.1162, + "step": 22321 + }, + { + "epoch": 0.53, + "grad_norm": 1.9848860712373544, + "learning_rate": 9.647548041652066e-06, + "loss": 1.0047, + "step": 22322 + }, + { + "epoch": 0.53, + "grad_norm": 1.7723903820612774, + "learning_rate": 9.646785475699691e-06, + "loss": 0.9354, + "step": 22323 + }, + { + "epoch": 0.53, + "grad_norm": 2.259606292921861, + "learning_rate": 9.64602291180384e-06, + "loss": 1.0448, + "step": 22324 + }, + { + "epoch": 0.53, + "grad_norm": 2.644455321572863, + "learning_rate": 9.645260349968955e-06, + "loss": 1.1205, + "step": 22325 + }, + { + "epoch": 0.53, + "grad_norm": 2.0221868434733836, + "learning_rate": 9.644497790199478e-06, + "loss": 1.0661, + "step": 22326 + }, + { + "epoch": 0.53, + "grad_norm": 2.061935741597499, + "learning_rate": 9.643735232499846e-06, + "loss": 0.9672, + "step": 22327 + }, + { + "epoch": 0.53, + "grad_norm": 2.0118774150109973, + "learning_rate": 9.6429726768745e-06, + "loss": 0.9707, + "step": 22328 + }, + { + "epoch": 0.53, + "grad_norm": 1.9931639184676047, + "learning_rate": 9.642210123327882e-06, + "loss": 1.0882, + "step": 22329 + }, + { + "epoch": 0.53, + "grad_norm": 1.8764132477215452, + "learning_rate": 9.641447571864429e-06, + "loss": 0.842, + "step": 22330 + }, + { + "epoch": 0.53, + "grad_norm": 2.0795961707055426, + "learning_rate": 9.64068502248858e-06, + "loss": 1.0713, + "step": 22331 + }, + { + "epoch": 0.53, + "grad_norm": 1.972365944458218, + "learning_rate": 9.639922475204778e-06, + "loss": 0.9445, + "step": 22332 + }, + { + "epoch": 0.53, + "grad_norm": 1.9623835952633912, + "learning_rate": 9.639159930017465e-06, + "loss": 1.076, + "step": 22333 + }, + { + "epoch": 0.53, + "grad_norm": 1.985534956671341, + "learning_rate": 9.638397386931072e-06, + "loss": 1.0043, + "step": 22334 + }, + { + "epoch": 0.53, + "grad_norm": 2.033875502340085, + "learning_rate": 9.637634845950043e-06, + "loss": 0.9503, + "step": 22335 + }, + { + "epoch": 0.53, + "grad_norm": 1.1132551636693029, + "learning_rate": 9.636872307078821e-06, + "loss": 0.9427, + "step": 22336 + }, + { + "epoch": 0.53, + "grad_norm": 1.964576335767705, + "learning_rate": 9.636109770321846e-06, + "loss": 0.9848, + "step": 22337 + }, + { + "epoch": 0.53, + "grad_norm": 2.1121197671142506, + "learning_rate": 9.635347235683552e-06, + "loss": 1.1305, + "step": 22338 + }, + { + "epoch": 0.53, + "grad_norm": 2.155559995313609, + "learning_rate": 9.634584703168381e-06, + "loss": 0.9463, + "step": 22339 + }, + { + "epoch": 0.53, + "grad_norm": 1.9610137964980017, + "learning_rate": 9.633822172780778e-06, + "loss": 0.9103, + "step": 22340 + }, + { + "epoch": 0.53, + "grad_norm": 1.8592879748064282, + "learning_rate": 9.633059644525175e-06, + "loss": 0.9967, + "step": 22341 + }, + { + "epoch": 0.53, + "grad_norm": 2.1684061347015087, + "learning_rate": 9.632297118406015e-06, + "loss": 1.0501, + "step": 22342 + }, + { + "epoch": 0.53, + "grad_norm": 2.0724360731173976, + "learning_rate": 9.631534594427735e-06, + "loss": 1.1314, + "step": 22343 + }, + { + "epoch": 0.53, + "grad_norm": 2.040478348628722, + "learning_rate": 9.630772072594784e-06, + "loss": 0.9747, + "step": 22344 + }, + { + "epoch": 0.53, + "grad_norm": 1.9397851858025055, + "learning_rate": 9.63000955291159e-06, + "loss": 1.0919, + "step": 22345 + }, + { + "epoch": 0.53, + "grad_norm": 2.0678918174844663, + "learning_rate": 9.629247035382597e-06, + "loss": 0.9303, + "step": 22346 + }, + { + "epoch": 0.53, + "grad_norm": 2.003279612745048, + "learning_rate": 9.62848452001225e-06, + "loss": 1.0823, + "step": 22347 + }, + { + "epoch": 0.53, + "grad_norm": 1.828367906667563, + "learning_rate": 9.627722006804977e-06, + "loss": 0.9175, + "step": 22348 + }, + { + "epoch": 0.53, + "grad_norm": 2.0513726263919394, + "learning_rate": 9.626959495765229e-06, + "loss": 0.9978, + "step": 22349 + }, + { + "epoch": 0.53, + "grad_norm": 1.9013378362564963, + "learning_rate": 9.626196986897437e-06, + "loss": 0.9409, + "step": 22350 + }, + { + "epoch": 0.53, + "grad_norm": 2.0392065262261885, + "learning_rate": 9.62543448020605e-06, + "loss": 0.9969, + "step": 22351 + }, + { + "epoch": 0.53, + "grad_norm": 2.0487613995535234, + "learning_rate": 9.624671975695497e-06, + "loss": 0.9986, + "step": 22352 + }, + { + "epoch": 0.53, + "grad_norm": 2.455451383350641, + "learning_rate": 9.623909473370225e-06, + "loss": 1.0155, + "step": 22353 + }, + { + "epoch": 0.53, + "grad_norm": 1.9303863487976318, + "learning_rate": 9.623146973234673e-06, + "loss": 0.9292, + "step": 22354 + }, + { + "epoch": 0.53, + "grad_norm": 2.1518175216667514, + "learning_rate": 9.622384475293278e-06, + "loss": 1.1531, + "step": 22355 + }, + { + "epoch": 0.53, + "grad_norm": 3.7912146770550628, + "learning_rate": 9.621621979550475e-06, + "loss": 0.9299, + "step": 22356 + }, + { + "epoch": 0.53, + "grad_norm": 1.9328835404659088, + "learning_rate": 9.620859486010713e-06, + "loss": 1.1121, + "step": 22357 + }, + { + "epoch": 0.53, + "grad_norm": 2.233089925693799, + "learning_rate": 9.620096994678429e-06, + "loss": 1.0148, + "step": 22358 + }, + { + "epoch": 0.53, + "grad_norm": 2.0150043131165245, + "learning_rate": 9.619334505558055e-06, + "loss": 1.0563, + "step": 22359 + }, + { + "epoch": 0.53, + "grad_norm": 1.7985529707275805, + "learning_rate": 9.618572018654039e-06, + "loss": 1.0065, + "step": 22360 + }, + { + "epoch": 0.53, + "grad_norm": 1.8490959471238497, + "learning_rate": 9.61780953397082e-06, + "loss": 0.9299, + "step": 22361 + }, + { + "epoch": 0.53, + "grad_norm": 1.8535765363032541, + "learning_rate": 9.617047051512832e-06, + "loss": 0.9139, + "step": 22362 + }, + { + "epoch": 0.53, + "grad_norm": 2.0945531941187077, + "learning_rate": 9.616284571284515e-06, + "loss": 1.0859, + "step": 22363 + }, + { + "epoch": 0.53, + "grad_norm": 1.911780191307363, + "learning_rate": 9.615522093290313e-06, + "loss": 1.0028, + "step": 22364 + }, + { + "epoch": 0.53, + "grad_norm": 1.0969259710777026, + "learning_rate": 9.614759617534665e-06, + "loss": 0.9381, + "step": 22365 + }, + { + "epoch": 0.53, + "grad_norm": 2.2308133222623767, + "learning_rate": 9.613997144022006e-06, + "loss": 0.9208, + "step": 22366 + }, + { + "epoch": 0.53, + "grad_norm": 1.7945667347204237, + "learning_rate": 9.613234672756777e-06, + "loss": 1.0727, + "step": 22367 + }, + { + "epoch": 0.53, + "grad_norm": 2.2948540821613603, + "learning_rate": 9.612472203743423e-06, + "loss": 0.9807, + "step": 22368 + }, + { + "epoch": 0.53, + "grad_norm": 1.7849996666870802, + "learning_rate": 9.611709736986375e-06, + "loss": 0.9917, + "step": 22369 + }, + { + "epoch": 0.53, + "grad_norm": 1.9840715871673935, + "learning_rate": 9.610947272490077e-06, + "loss": 1.1087, + "step": 22370 + }, + { + "epoch": 0.53, + "grad_norm": 1.0834369705610545, + "learning_rate": 9.610184810258964e-06, + "loss": 0.924, + "step": 22371 + }, + { + "epoch": 0.53, + "grad_norm": 1.9881126099493167, + "learning_rate": 9.609422350297482e-06, + "loss": 1.0975, + "step": 22372 + }, + { + "epoch": 0.53, + "grad_norm": 1.8566812812279234, + "learning_rate": 9.608659892610067e-06, + "loss": 1.0391, + "step": 22373 + }, + { + "epoch": 0.53, + "grad_norm": 1.9258210506204894, + "learning_rate": 9.607897437201155e-06, + "loss": 0.9279, + "step": 22374 + }, + { + "epoch": 0.53, + "grad_norm": 2.0105438866453196, + "learning_rate": 9.607134984075193e-06, + "loss": 1.0224, + "step": 22375 + }, + { + "epoch": 0.53, + "grad_norm": 1.922934678336418, + "learning_rate": 9.60637253323661e-06, + "loss": 0.9669, + "step": 22376 + }, + { + "epoch": 0.53, + "grad_norm": 2.1786676028091403, + "learning_rate": 9.605610084689853e-06, + "loss": 0.9347, + "step": 22377 + }, + { + "epoch": 0.53, + "grad_norm": 1.9170075402347062, + "learning_rate": 9.604847638439358e-06, + "loss": 0.97, + "step": 22378 + }, + { + "epoch": 0.53, + "grad_norm": 1.0927713085627677, + "learning_rate": 9.604085194489568e-06, + "loss": 0.9596, + "step": 22379 + }, + { + "epoch": 0.53, + "grad_norm": 1.8607857136872443, + "learning_rate": 9.603322752844917e-06, + "loss": 0.9109, + "step": 22380 + }, + { + "epoch": 0.53, + "grad_norm": 2.125803861472375, + "learning_rate": 9.602560313509846e-06, + "loss": 1.0144, + "step": 22381 + }, + { + "epoch": 0.53, + "grad_norm": 2.055840053296316, + "learning_rate": 9.601797876488798e-06, + "loss": 0.9756, + "step": 22382 + }, + { + "epoch": 0.53, + "grad_norm": 2.090390816002081, + "learning_rate": 9.601035441786206e-06, + "loss": 0.9245, + "step": 22383 + }, + { + "epoch": 0.53, + "grad_norm": 2.5933599240216556, + "learning_rate": 9.60027300940651e-06, + "loss": 1.1504, + "step": 22384 + }, + { + "epoch": 0.53, + "grad_norm": 2.08328337099719, + "learning_rate": 9.599510579354153e-06, + "loss": 0.9679, + "step": 22385 + }, + { + "epoch": 0.53, + "grad_norm": 2.232201619392318, + "learning_rate": 9.598748151633576e-06, + "loss": 0.9913, + "step": 22386 + }, + { + "epoch": 0.53, + "grad_norm": 1.0766044455233912, + "learning_rate": 9.597985726249211e-06, + "loss": 0.9927, + "step": 22387 + }, + { + "epoch": 0.53, + "grad_norm": 2.5927041846435857, + "learning_rate": 9.5972233032055e-06, + "loss": 0.9822, + "step": 22388 + }, + { + "epoch": 0.53, + "grad_norm": 1.7912207506008961, + "learning_rate": 9.596460882506885e-06, + "loss": 0.9053, + "step": 22389 + }, + { + "epoch": 0.53, + "grad_norm": 2.0268280177461873, + "learning_rate": 9.595698464157799e-06, + "loss": 1.1121, + "step": 22390 + }, + { + "epoch": 0.53, + "grad_norm": 2.1322316412137114, + "learning_rate": 9.594936048162685e-06, + "loss": 0.927, + "step": 22391 + }, + { + "epoch": 0.53, + "grad_norm": 2.029561762897012, + "learning_rate": 9.59417363452598e-06, + "loss": 0.9431, + "step": 22392 + }, + { + "epoch": 0.53, + "grad_norm": 2.186583132693512, + "learning_rate": 9.59341122325213e-06, + "loss": 1.0565, + "step": 22393 + }, + { + "epoch": 0.53, + "grad_norm": 2.1662221737853025, + "learning_rate": 9.592648814345565e-06, + "loss": 0.9551, + "step": 22394 + }, + { + "epoch": 0.53, + "grad_norm": 2.1237202817148324, + "learning_rate": 9.591886407810727e-06, + "loss": 1.0504, + "step": 22395 + }, + { + "epoch": 0.53, + "grad_norm": 1.8517616032006459, + "learning_rate": 9.591124003652062e-06, + "loss": 0.9783, + "step": 22396 + }, + { + "epoch": 0.53, + "grad_norm": 2.028559643084004, + "learning_rate": 9.590361601873994e-06, + "loss": 1.0447, + "step": 22397 + }, + { + "epoch": 0.53, + "grad_norm": 1.89767564189788, + "learning_rate": 9.589599202480974e-06, + "loss": 0.9257, + "step": 22398 + }, + { + "epoch": 0.53, + "grad_norm": 2.25845113212537, + "learning_rate": 9.588836805477435e-06, + "loss": 1.0679, + "step": 22399 + }, + { + "epoch": 0.53, + "grad_norm": 1.8765526632995888, + "learning_rate": 9.588074410867825e-06, + "loss": 1.0063, + "step": 22400 + }, + { + "epoch": 0.53, + "grad_norm": 1.944031880940617, + "learning_rate": 9.587312018656568e-06, + "loss": 0.8998, + "step": 22401 + }, + { + "epoch": 0.53, + "grad_norm": 2.215069140801948, + "learning_rate": 9.586549628848115e-06, + "loss": 1.1094, + "step": 22402 + }, + { + "epoch": 0.53, + "grad_norm": 1.7905274098076143, + "learning_rate": 9.585787241446903e-06, + "loss": 0.9866, + "step": 22403 + }, + { + "epoch": 0.53, + "grad_norm": 2.344564007824517, + "learning_rate": 9.585024856457367e-06, + "loss": 1.0439, + "step": 22404 + }, + { + "epoch": 0.53, + "grad_norm": 2.2129961171303334, + "learning_rate": 9.584262473883945e-06, + "loss": 1.0564, + "step": 22405 + }, + { + "epoch": 0.53, + "grad_norm": 2.1343102640236933, + "learning_rate": 9.58350009373108e-06, + "loss": 1.0407, + "step": 22406 + }, + { + "epoch": 0.53, + "grad_norm": 2.1410616233405233, + "learning_rate": 9.582737716003212e-06, + "loss": 0.9572, + "step": 22407 + }, + { + "epoch": 0.53, + "grad_norm": 1.1050077794029611, + "learning_rate": 9.581975340704772e-06, + "loss": 1.0041, + "step": 22408 + }, + { + "epoch": 0.53, + "grad_norm": 2.063146126594218, + "learning_rate": 9.581212967840208e-06, + "loss": 0.984, + "step": 22409 + }, + { + "epoch": 0.53, + "grad_norm": 2.239097715927432, + "learning_rate": 9.580450597413955e-06, + "loss": 1.1081, + "step": 22410 + }, + { + "epoch": 0.53, + "grad_norm": 1.9292179008763017, + "learning_rate": 9.57968822943045e-06, + "loss": 1.006, + "step": 22411 + }, + { + "epoch": 0.53, + "grad_norm": 2.0087240839830787, + "learning_rate": 9.578925863894132e-06, + "loss": 0.9509, + "step": 22412 + }, + { + "epoch": 0.53, + "grad_norm": 2.080457626454315, + "learning_rate": 9.578163500809443e-06, + "loss": 1.077, + "step": 22413 + }, + { + "epoch": 0.53, + "grad_norm": 2.2895624296128374, + "learning_rate": 9.577401140180819e-06, + "loss": 1.0897, + "step": 22414 + }, + { + "epoch": 0.53, + "grad_norm": 2.1179760754531403, + "learning_rate": 9.5766387820127e-06, + "loss": 0.9657, + "step": 22415 + }, + { + "epoch": 0.53, + "grad_norm": 2.151804618935777, + "learning_rate": 9.575876426309521e-06, + "loss": 0.9945, + "step": 22416 + }, + { + "epoch": 0.53, + "grad_norm": 2.0376748878368924, + "learning_rate": 9.575114073075729e-06, + "loss": 0.9569, + "step": 22417 + }, + { + "epoch": 0.53, + "grad_norm": 1.9643592389957394, + "learning_rate": 9.574351722315754e-06, + "loss": 1.0038, + "step": 22418 + }, + { + "epoch": 0.53, + "grad_norm": 1.8487923614869108, + "learning_rate": 9.573589374034038e-06, + "loss": 1.0763, + "step": 22419 + }, + { + "epoch": 0.53, + "grad_norm": 1.7114290569344814, + "learning_rate": 9.572827028235018e-06, + "loss": 1.0042, + "step": 22420 + }, + { + "epoch": 0.53, + "grad_norm": 2.2290813660297895, + "learning_rate": 9.572064684923138e-06, + "loss": 0.9646, + "step": 22421 + }, + { + "epoch": 0.53, + "grad_norm": 2.1219028601253545, + "learning_rate": 9.57130234410283e-06, + "loss": 1.0055, + "step": 22422 + }, + { + "epoch": 0.53, + "grad_norm": 1.9728016407897513, + "learning_rate": 9.570540005778539e-06, + "loss": 1.0176, + "step": 22423 + }, + { + "epoch": 0.53, + "grad_norm": 1.9181786318420773, + "learning_rate": 9.569777669954694e-06, + "loss": 1.0551, + "step": 22424 + }, + { + "epoch": 0.53, + "grad_norm": 1.0672810288715164, + "learning_rate": 9.569015336635746e-06, + "loss": 0.9364, + "step": 22425 + }, + { + "epoch": 0.53, + "grad_norm": 2.556124305389149, + "learning_rate": 9.568253005826124e-06, + "loss": 0.9004, + "step": 22426 + }, + { + "epoch": 0.53, + "grad_norm": 1.9354288419346368, + "learning_rate": 9.56749067753027e-06, + "loss": 0.8953, + "step": 22427 + }, + { + "epoch": 0.53, + "grad_norm": 1.930713114479538, + "learning_rate": 9.566728351752626e-06, + "loss": 0.9633, + "step": 22428 + }, + { + "epoch": 0.53, + "grad_norm": 3.5146644635112034, + "learning_rate": 9.56596602849762e-06, + "loss": 1.1786, + "step": 22429 + }, + { + "epoch": 0.53, + "grad_norm": 2.0180023924132917, + "learning_rate": 9.5652037077697e-06, + "loss": 1.0512, + "step": 22430 + }, + { + "epoch": 0.53, + "grad_norm": 1.860565904724822, + "learning_rate": 9.5644413895733e-06, + "loss": 0.9225, + "step": 22431 + }, + { + "epoch": 0.53, + "grad_norm": 1.852324531924878, + "learning_rate": 9.563679073912866e-06, + "loss": 1.0208, + "step": 22432 + }, + { + "epoch": 0.53, + "grad_norm": 2.262582393840746, + "learning_rate": 9.562916760792825e-06, + "loss": 1.0224, + "step": 22433 + }, + { + "epoch": 0.53, + "grad_norm": 3.105691102752656, + "learning_rate": 9.562154450217623e-06, + "loss": 0.9338, + "step": 22434 + }, + { + "epoch": 0.53, + "grad_norm": 2.2493311977467374, + "learning_rate": 9.561392142191698e-06, + "loss": 0.9405, + "step": 22435 + }, + { + "epoch": 0.53, + "grad_norm": 2.311368347810498, + "learning_rate": 9.560629836719485e-06, + "loss": 1.0821, + "step": 22436 + }, + { + "epoch": 0.53, + "grad_norm": 2.4532208008267724, + "learning_rate": 9.559867533805421e-06, + "loss": 0.9505, + "step": 22437 + }, + { + "epoch": 0.53, + "grad_norm": 2.004241509352202, + "learning_rate": 9.55910523345395e-06, + "loss": 1.1089, + "step": 22438 + }, + { + "epoch": 0.53, + "grad_norm": 1.917086096006909, + "learning_rate": 9.558342935669512e-06, + "loss": 1.0972, + "step": 22439 + }, + { + "epoch": 0.53, + "grad_norm": 2.098807032668368, + "learning_rate": 9.557580640456538e-06, + "loss": 0.9533, + "step": 22440 + }, + { + "epoch": 0.53, + "grad_norm": 1.8994212284051024, + "learning_rate": 9.556818347819467e-06, + "loss": 0.9583, + "step": 22441 + }, + { + "epoch": 0.53, + "grad_norm": 1.9442000835598579, + "learning_rate": 9.556056057762746e-06, + "loss": 1.1295, + "step": 22442 + }, + { + "epoch": 0.53, + "grad_norm": 2.6101590733853977, + "learning_rate": 9.555293770290802e-06, + "loss": 1.0445, + "step": 22443 + }, + { + "epoch": 0.53, + "grad_norm": 2.8265345794781194, + "learning_rate": 9.554531485408078e-06, + "loss": 1.101, + "step": 22444 + }, + { + "epoch": 0.53, + "grad_norm": 2.0904728946258717, + "learning_rate": 9.553769203119014e-06, + "loss": 1.0633, + "step": 22445 + }, + { + "epoch": 0.53, + "grad_norm": 2.1931130843867015, + "learning_rate": 9.55300692342805e-06, + "loss": 0.9599, + "step": 22446 + }, + { + "epoch": 0.53, + "grad_norm": 1.0776102115065693, + "learning_rate": 9.552244646339619e-06, + "loss": 0.9769, + "step": 22447 + }, + { + "epoch": 0.53, + "grad_norm": 1.8199426063880553, + "learning_rate": 9.55148237185816e-06, + "loss": 1.0372, + "step": 22448 + }, + { + "epoch": 0.53, + "grad_norm": 1.9630263910265933, + "learning_rate": 9.550720099988117e-06, + "loss": 0.9793, + "step": 22449 + }, + { + "epoch": 0.53, + "grad_norm": 2.045706948887626, + "learning_rate": 9.54995783073392e-06, + "loss": 1.0914, + "step": 22450 + }, + { + "epoch": 0.53, + "grad_norm": 8.06429709784546, + "learning_rate": 9.54919556410001e-06, + "loss": 0.9862, + "step": 22451 + }, + { + "epoch": 0.53, + "grad_norm": 2.127767663488341, + "learning_rate": 9.548433300090826e-06, + "loss": 0.9635, + "step": 22452 + }, + { + "epoch": 0.53, + "grad_norm": 2.1543005545974667, + "learning_rate": 9.547671038710813e-06, + "loss": 1.028, + "step": 22453 + }, + { + "epoch": 0.53, + "grad_norm": 1.9550284969831533, + "learning_rate": 9.546908779964394e-06, + "loss": 0.9461, + "step": 22454 + }, + { + "epoch": 0.53, + "grad_norm": 2.6298545714773267, + "learning_rate": 9.546146523856021e-06, + "loss": 1.0245, + "step": 22455 + }, + { + "epoch": 0.53, + "grad_norm": 1.7770714929890081, + "learning_rate": 9.545384270390125e-06, + "loss": 1.0092, + "step": 22456 + }, + { + "epoch": 0.53, + "grad_norm": 1.8005994282945443, + "learning_rate": 9.544622019571144e-06, + "loss": 1.0463, + "step": 22457 + }, + { + "epoch": 0.53, + "grad_norm": 2.0669460121248244, + "learning_rate": 9.543859771403518e-06, + "loss": 0.9415, + "step": 22458 + }, + { + "epoch": 0.53, + "grad_norm": 2.0448550063761863, + "learning_rate": 9.543097525891686e-06, + "loss": 1.0957, + "step": 22459 + }, + { + "epoch": 0.53, + "grad_norm": 2.187180127871184, + "learning_rate": 9.542335283040089e-06, + "loss": 0.974, + "step": 22460 + }, + { + "epoch": 0.53, + "grad_norm": 2.1911726450334754, + "learning_rate": 9.541573042853153e-06, + "loss": 1.0739, + "step": 22461 + }, + { + "epoch": 0.53, + "grad_norm": 2.09749208683478, + "learning_rate": 9.540810805335327e-06, + "loss": 1.132, + "step": 22462 + }, + { + "epoch": 0.53, + "grad_norm": 1.0056814444903595, + "learning_rate": 9.540048570491049e-06, + "loss": 0.953, + "step": 22463 + }, + { + "epoch": 0.53, + "grad_norm": 1.0679227773301956, + "learning_rate": 9.539286338324753e-06, + "loss": 0.965, + "step": 22464 + }, + { + "epoch": 0.53, + "grad_norm": 1.051588243152385, + "learning_rate": 9.538524108840873e-06, + "loss": 0.9632, + "step": 22465 + }, + { + "epoch": 0.53, + "grad_norm": 1.9216560158084328, + "learning_rate": 9.537761882043855e-06, + "loss": 0.9938, + "step": 22466 + }, + { + "epoch": 0.53, + "grad_norm": 2.2431442727364788, + "learning_rate": 9.536999657938135e-06, + "loss": 1.0512, + "step": 22467 + }, + { + "epoch": 0.53, + "grad_norm": 1.0346426665923982, + "learning_rate": 9.536237436528148e-06, + "loss": 0.9341, + "step": 22468 + }, + { + "epoch": 0.53, + "grad_norm": 1.9772581607384252, + "learning_rate": 9.535475217818331e-06, + "loss": 0.9918, + "step": 22469 + }, + { + "epoch": 0.53, + "grad_norm": 2.152268421824578, + "learning_rate": 9.534713001813131e-06, + "loss": 1.0496, + "step": 22470 + }, + { + "epoch": 0.53, + "grad_norm": 2.1202624038762954, + "learning_rate": 9.533950788516974e-06, + "loss": 0.964, + "step": 22471 + }, + { + "epoch": 0.53, + "grad_norm": 2.353076109417455, + "learning_rate": 9.533188577934304e-06, + "loss": 0.8331, + "step": 22472 + }, + { + "epoch": 0.53, + "grad_norm": 2.098553593907958, + "learning_rate": 9.532426370069557e-06, + "loss": 1.0422, + "step": 22473 + }, + { + "epoch": 0.53, + "grad_norm": 1.864403810624169, + "learning_rate": 9.531664164927175e-06, + "loss": 1.0047, + "step": 22474 + }, + { + "epoch": 0.53, + "grad_norm": 2.183224633086683, + "learning_rate": 9.53090196251159e-06, + "loss": 1.0557, + "step": 22475 + }, + { + "epoch": 0.53, + "grad_norm": 2.1930948596438085, + "learning_rate": 9.530139762827243e-06, + "loss": 0.9948, + "step": 22476 + }, + { + "epoch": 0.53, + "grad_norm": 1.9773253915287563, + "learning_rate": 9.529377565878571e-06, + "loss": 1.0359, + "step": 22477 + }, + { + "epoch": 0.53, + "grad_norm": 2.5842421089395073, + "learning_rate": 9.52861537167001e-06, + "loss": 1.1491, + "step": 22478 + }, + { + "epoch": 0.53, + "grad_norm": 3.0710190237348134, + "learning_rate": 9.527853180206e-06, + "loss": 0.9706, + "step": 22479 + }, + { + "epoch": 0.53, + "grad_norm": 2.8455506510884017, + "learning_rate": 9.527090991490977e-06, + "loss": 1.146, + "step": 22480 + }, + { + "epoch": 0.53, + "grad_norm": 2.0564302537698067, + "learning_rate": 9.526328805529387e-06, + "loss": 0.9325, + "step": 22481 + }, + { + "epoch": 0.53, + "grad_norm": 2.100396979561619, + "learning_rate": 9.525566622325653e-06, + "loss": 0.9313, + "step": 22482 + }, + { + "epoch": 0.53, + "grad_norm": 2.2724148465307303, + "learning_rate": 9.524804441884224e-06, + "loss": 0.9477, + "step": 22483 + }, + { + "epoch": 0.53, + "grad_norm": 2.169855327190914, + "learning_rate": 9.524042264209535e-06, + "loss": 0.9897, + "step": 22484 + }, + { + "epoch": 0.53, + "grad_norm": 2.766760959874162, + "learning_rate": 9.523280089306022e-06, + "loss": 0.9203, + "step": 22485 + }, + { + "epoch": 0.53, + "grad_norm": 1.9157553242244532, + "learning_rate": 9.52251791717812e-06, + "loss": 0.9802, + "step": 22486 + }, + { + "epoch": 0.53, + "grad_norm": 2.530147962539833, + "learning_rate": 9.52175574783027e-06, + "loss": 1.042, + "step": 22487 + }, + { + "epoch": 0.53, + "grad_norm": 2.0415035871420177, + "learning_rate": 9.520993581266916e-06, + "loss": 1.0332, + "step": 22488 + }, + { + "epoch": 0.53, + "grad_norm": 1.8852389339236821, + "learning_rate": 9.520231417492484e-06, + "loss": 0.9006, + "step": 22489 + }, + { + "epoch": 0.53, + "grad_norm": 1.9242535163177623, + "learning_rate": 9.519469256511415e-06, + "loss": 1.0482, + "step": 22490 + }, + { + "epoch": 0.53, + "grad_norm": 2.098920926179978, + "learning_rate": 9.518707098328152e-06, + "loss": 0.971, + "step": 22491 + }, + { + "epoch": 0.53, + "grad_norm": 1.0585389351817367, + "learning_rate": 9.517944942947126e-06, + "loss": 0.9156, + "step": 22492 + }, + { + "epoch": 0.53, + "grad_norm": 2.1496932276526635, + "learning_rate": 9.517182790372776e-06, + "loss": 1.1015, + "step": 22493 + }, + { + "epoch": 0.53, + "grad_norm": 2.134232542405127, + "learning_rate": 9.516420640609541e-06, + "loss": 1.1222, + "step": 22494 + }, + { + "epoch": 0.53, + "grad_norm": 2.178413926465961, + "learning_rate": 9.515658493661863e-06, + "loss": 0.9803, + "step": 22495 + }, + { + "epoch": 0.53, + "grad_norm": 1.8352209795001215, + "learning_rate": 9.51489634953417e-06, + "loss": 1.1796, + "step": 22496 + }, + { + "epoch": 0.53, + "grad_norm": 1.126975348119601, + "learning_rate": 9.514134208230904e-06, + "loss": 0.961, + "step": 22497 + }, + { + "epoch": 0.53, + "grad_norm": 1.954374380097872, + "learning_rate": 9.513372069756508e-06, + "loss": 1.1337, + "step": 22498 + }, + { + "epoch": 0.53, + "grad_norm": 1.8917640889219687, + "learning_rate": 9.512609934115407e-06, + "loss": 1.0501, + "step": 22499 + }, + { + "epoch": 0.53, + "grad_norm": 1.8694717222721522, + "learning_rate": 9.511847801312048e-06, + "loss": 1.0272, + "step": 22500 + }, + { + "epoch": 0.53, + "grad_norm": 2.1824823597092915, + "learning_rate": 9.511085671350864e-06, + "loss": 1.0289, + "step": 22501 + }, + { + "epoch": 0.53, + "grad_norm": 1.9995168297507857, + "learning_rate": 9.510323544236298e-06, + "loss": 1.0731, + "step": 22502 + }, + { + "epoch": 0.53, + "grad_norm": 2.1458451928647015, + "learning_rate": 9.50956141997278e-06, + "loss": 0.9551, + "step": 22503 + }, + { + "epoch": 0.53, + "grad_norm": 1.9851782791633603, + "learning_rate": 9.508799298564753e-06, + "loss": 1.0153, + "step": 22504 + }, + { + "epoch": 0.53, + "grad_norm": 1.9779533071276332, + "learning_rate": 9.508037180016652e-06, + "loss": 1.1401, + "step": 22505 + }, + { + "epoch": 0.53, + "grad_norm": 2.0814601604098173, + "learning_rate": 9.50727506433291e-06, + "loss": 0.9928, + "step": 22506 + }, + { + "epoch": 0.53, + "grad_norm": 2.1665336405020676, + "learning_rate": 9.506512951517972e-06, + "loss": 0.954, + "step": 22507 + }, + { + "epoch": 0.53, + "grad_norm": 1.1776409019160565, + "learning_rate": 9.50575084157627e-06, + "loss": 1.0395, + "step": 22508 + }, + { + "epoch": 0.53, + "grad_norm": 2.3990433724480504, + "learning_rate": 9.504988734512249e-06, + "loss": 1.0432, + "step": 22509 + }, + { + "epoch": 0.53, + "grad_norm": 1.8866447786921419, + "learning_rate": 9.504226630330333e-06, + "loss": 0.9675, + "step": 22510 + }, + { + "epoch": 0.53, + "grad_norm": 2.1205265768639, + "learning_rate": 9.503464529034967e-06, + "loss": 0.9792, + "step": 22511 + }, + { + "epoch": 0.53, + "grad_norm": 2.363149150198712, + "learning_rate": 9.502702430630593e-06, + "loss": 0.993, + "step": 22512 + }, + { + "epoch": 0.53, + "grad_norm": 1.9001207292348599, + "learning_rate": 9.50194033512164e-06, + "loss": 1.0049, + "step": 22513 + }, + { + "epoch": 0.53, + "grad_norm": 1.9802798451717756, + "learning_rate": 9.501178242512546e-06, + "loss": 1.0686, + "step": 22514 + }, + { + "epoch": 0.53, + "grad_norm": 1.942046657327038, + "learning_rate": 9.500416152807752e-06, + "loss": 0.8814, + "step": 22515 + }, + { + "epoch": 0.53, + "grad_norm": 1.7956140400220402, + "learning_rate": 9.499654066011696e-06, + "loss": 1.1123, + "step": 22516 + }, + { + "epoch": 0.53, + "grad_norm": 1.1273751717470732, + "learning_rate": 9.498891982128809e-06, + "loss": 1.0048, + "step": 22517 + }, + { + "epoch": 0.53, + "grad_norm": 2.07733652005833, + "learning_rate": 9.49812990116353e-06, + "loss": 0.9996, + "step": 22518 + }, + { + "epoch": 0.53, + "grad_norm": 1.8306738073304727, + "learning_rate": 9.497367823120302e-06, + "loss": 1.0156, + "step": 22519 + }, + { + "epoch": 0.53, + "grad_norm": 1.0012519948274379, + "learning_rate": 9.496605748003556e-06, + "loss": 0.9458, + "step": 22520 + }, + { + "epoch": 0.53, + "grad_norm": 2.055633399376776, + "learning_rate": 9.49584367581773e-06, + "loss": 0.9363, + "step": 22521 + }, + { + "epoch": 0.53, + "grad_norm": 1.0726305350672043, + "learning_rate": 9.495081606567261e-06, + "loss": 0.8425, + "step": 22522 + }, + { + "epoch": 0.53, + "grad_norm": 1.9328389871782259, + "learning_rate": 9.49431954025659e-06, + "loss": 0.9154, + "step": 22523 + }, + { + "epoch": 0.53, + "grad_norm": 1.8149273126874457, + "learning_rate": 9.493557476890146e-06, + "loss": 1.0733, + "step": 22524 + }, + { + "epoch": 0.53, + "grad_norm": 1.8130086231272, + "learning_rate": 9.492795416472374e-06, + "loss": 1.0164, + "step": 22525 + }, + { + "epoch": 0.53, + "grad_norm": 1.9983286398489806, + "learning_rate": 9.49203335900771e-06, + "loss": 1.087, + "step": 22526 + }, + { + "epoch": 0.53, + "grad_norm": 1.7436465592022663, + "learning_rate": 9.491271304500581e-06, + "loss": 1.0187, + "step": 22527 + }, + { + "epoch": 0.53, + "grad_norm": 2.012805069855607, + "learning_rate": 9.490509252955437e-06, + "loss": 0.9496, + "step": 22528 + }, + { + "epoch": 0.53, + "grad_norm": 2.1122394789569827, + "learning_rate": 9.489747204376707e-06, + "loss": 1.081, + "step": 22529 + }, + { + "epoch": 0.53, + "grad_norm": 1.8700872415132022, + "learning_rate": 9.488985158768836e-06, + "loss": 1.0042, + "step": 22530 + }, + { + "epoch": 0.53, + "grad_norm": 2.170964111011722, + "learning_rate": 9.488223116136248e-06, + "loss": 0.9052, + "step": 22531 + }, + { + "epoch": 0.53, + "grad_norm": 2.0766615841822733, + "learning_rate": 9.48746107648339e-06, + "loss": 0.9926, + "step": 22532 + }, + { + "epoch": 0.53, + "grad_norm": 1.8588215135381327, + "learning_rate": 9.486699039814698e-06, + "loss": 0.9338, + "step": 22533 + }, + { + "epoch": 0.53, + "grad_norm": 1.1905697080318363, + "learning_rate": 9.485937006134605e-06, + "loss": 0.9727, + "step": 22534 + }, + { + "epoch": 0.53, + "grad_norm": 1.9941294226229023, + "learning_rate": 9.485174975447546e-06, + "loss": 0.9419, + "step": 22535 + }, + { + "epoch": 0.53, + "grad_norm": 1.1083308009168422, + "learning_rate": 9.484412947757965e-06, + "loss": 0.8908, + "step": 22536 + }, + { + "epoch": 0.53, + "grad_norm": 1.9969891943231148, + "learning_rate": 9.483650923070297e-06, + "loss": 1.1049, + "step": 22537 + }, + { + "epoch": 0.53, + "grad_norm": 1.083790080811955, + "learning_rate": 9.482888901388973e-06, + "loss": 0.9385, + "step": 22538 + }, + { + "epoch": 0.53, + "grad_norm": 1.781867561239506, + "learning_rate": 9.482126882718433e-06, + "loss": 1.02, + "step": 22539 + }, + { + "epoch": 0.53, + "grad_norm": 2.263008734152546, + "learning_rate": 9.481364867063118e-06, + "loss": 0.961, + "step": 22540 + }, + { + "epoch": 0.53, + "grad_norm": 1.954966302178495, + "learning_rate": 9.48060285442746e-06, + "loss": 1.1002, + "step": 22541 + }, + { + "epoch": 0.53, + "grad_norm": 2.160208917121479, + "learning_rate": 9.479840844815892e-06, + "loss": 1.0277, + "step": 22542 + }, + { + "epoch": 0.53, + "grad_norm": 1.9891511319558002, + "learning_rate": 9.47907883823286e-06, + "loss": 0.8955, + "step": 22543 + }, + { + "epoch": 0.53, + "grad_norm": 2.0412238053648175, + "learning_rate": 9.478316834682798e-06, + "loss": 1.0741, + "step": 22544 + }, + { + "epoch": 0.53, + "grad_norm": 1.9059868733801852, + "learning_rate": 9.477554834170137e-06, + "loss": 0.9618, + "step": 22545 + }, + { + "epoch": 0.53, + "grad_norm": 3.7995075579265123, + "learning_rate": 9.476792836699315e-06, + "loss": 0.9693, + "step": 22546 + }, + { + "epoch": 0.53, + "grad_norm": 1.765649897291179, + "learning_rate": 9.476030842274773e-06, + "loss": 0.8951, + "step": 22547 + }, + { + "epoch": 0.53, + "grad_norm": 2.105308343588397, + "learning_rate": 9.475268850900947e-06, + "loss": 0.9059, + "step": 22548 + }, + { + "epoch": 0.53, + "grad_norm": 1.0739541833369877, + "learning_rate": 9.47450686258227e-06, + "loss": 0.947, + "step": 22549 + }, + { + "epoch": 0.53, + "grad_norm": 1.0354001149341565, + "learning_rate": 9.473744877323178e-06, + "loss": 0.8854, + "step": 22550 + }, + { + "epoch": 0.53, + "grad_norm": 2.0633165380492815, + "learning_rate": 9.472982895128116e-06, + "loss": 0.9047, + "step": 22551 + }, + { + "epoch": 0.53, + "grad_norm": 1.9557468949534864, + "learning_rate": 9.472220916001511e-06, + "loss": 0.8642, + "step": 22552 + }, + { + "epoch": 0.53, + "grad_norm": 2.032585710015297, + "learning_rate": 9.471458939947804e-06, + "loss": 0.9834, + "step": 22553 + }, + { + "epoch": 0.53, + "grad_norm": 2.052104062496192, + "learning_rate": 9.470696966971427e-06, + "loss": 1.0469, + "step": 22554 + }, + { + "epoch": 0.53, + "grad_norm": 1.864949269374793, + "learning_rate": 9.469934997076825e-06, + "loss": 1.0601, + "step": 22555 + }, + { + "epoch": 0.53, + "grad_norm": 1.8246822515631824, + "learning_rate": 9.469173030268424e-06, + "loss": 0.928, + "step": 22556 + }, + { + "epoch": 0.53, + "grad_norm": 1.0548174758136337, + "learning_rate": 9.46841106655067e-06, + "loss": 0.9213, + "step": 22557 + }, + { + "epoch": 0.53, + "grad_norm": 1.8889963714850027, + "learning_rate": 9.467649105927995e-06, + "loss": 0.8206, + "step": 22558 + }, + { + "epoch": 0.53, + "grad_norm": 1.779123143187084, + "learning_rate": 9.466887148404831e-06, + "loss": 0.985, + "step": 22559 + }, + { + "epoch": 0.53, + "grad_norm": 2.0223559916909823, + "learning_rate": 9.466125193985623e-06, + "loss": 0.9554, + "step": 22560 + }, + { + "epoch": 0.53, + "grad_norm": 1.8826699116341998, + "learning_rate": 9.465363242674802e-06, + "loss": 1.03, + "step": 22561 + }, + { + "epoch": 0.53, + "grad_norm": 1.749012744307511, + "learning_rate": 9.464601294476807e-06, + "loss": 0.9422, + "step": 22562 + }, + { + "epoch": 0.53, + "grad_norm": 2.152697789239178, + "learning_rate": 9.463839349396069e-06, + "loss": 1.0739, + "step": 22563 + }, + { + "epoch": 0.53, + "grad_norm": 2.2605460132281157, + "learning_rate": 9.463077407437028e-06, + "loss": 0.9049, + "step": 22564 + }, + { + "epoch": 0.53, + "grad_norm": 2.0792823511774152, + "learning_rate": 9.462315468604126e-06, + "loss": 0.9632, + "step": 22565 + }, + { + "epoch": 0.53, + "grad_norm": 2.136492368066124, + "learning_rate": 9.461553532901789e-06, + "loss": 0.9788, + "step": 22566 + }, + { + "epoch": 0.53, + "grad_norm": 1.8532552741649237, + "learning_rate": 9.460791600334457e-06, + "loss": 0.9195, + "step": 22567 + }, + { + "epoch": 0.53, + "grad_norm": 2.0831141545875256, + "learning_rate": 9.460029670906568e-06, + "loss": 1.004, + "step": 22568 + }, + { + "epoch": 0.53, + "grad_norm": 1.7226031438757687, + "learning_rate": 9.45926774462256e-06, + "loss": 0.9125, + "step": 22569 + }, + { + "epoch": 0.53, + "grad_norm": 1.870226809772262, + "learning_rate": 9.458505821486864e-06, + "loss": 1.0814, + "step": 22570 + }, + { + "epoch": 0.53, + "grad_norm": 2.0205002901435973, + "learning_rate": 9.457743901503917e-06, + "loss": 1.1257, + "step": 22571 + }, + { + "epoch": 0.53, + "grad_norm": 2.234018226567332, + "learning_rate": 9.45698198467816e-06, + "loss": 1.0183, + "step": 22572 + }, + { + "epoch": 0.53, + "grad_norm": 1.8325807161354937, + "learning_rate": 9.456220071014025e-06, + "loss": 0.9986, + "step": 22573 + }, + { + "epoch": 0.53, + "grad_norm": 1.8698383990642788, + "learning_rate": 9.455458160515948e-06, + "loss": 0.9398, + "step": 22574 + }, + { + "epoch": 0.53, + "grad_norm": 3.0553403440895943, + "learning_rate": 9.454696253188365e-06, + "loss": 0.9775, + "step": 22575 + }, + { + "epoch": 0.53, + "grad_norm": 1.8573621889205816, + "learning_rate": 9.453934349035715e-06, + "loss": 1.0414, + "step": 22576 + }, + { + "epoch": 0.53, + "grad_norm": 1.854122518116408, + "learning_rate": 9.453172448062433e-06, + "loss": 0.9443, + "step": 22577 + }, + { + "epoch": 0.53, + "grad_norm": 2.228104314993963, + "learning_rate": 9.452410550272949e-06, + "loss": 1.0166, + "step": 22578 + }, + { + "epoch": 0.53, + "grad_norm": 1.9253926176394418, + "learning_rate": 9.451648655671712e-06, + "loss": 1.1248, + "step": 22579 + }, + { + "epoch": 0.53, + "grad_norm": 1.126962312548522, + "learning_rate": 9.450886764263144e-06, + "loss": 0.9642, + "step": 22580 + }, + { + "epoch": 0.53, + "grad_norm": 1.133051056037461, + "learning_rate": 9.450124876051688e-06, + "loss": 1.0528, + "step": 22581 + }, + { + "epoch": 0.53, + "grad_norm": 1.932716112173529, + "learning_rate": 9.449362991041778e-06, + "loss": 0.9961, + "step": 22582 + }, + { + "epoch": 0.53, + "grad_norm": 2.1307956069098752, + "learning_rate": 9.448601109237859e-06, + "loss": 0.9532, + "step": 22583 + }, + { + "epoch": 0.53, + "grad_norm": 1.7993240576908776, + "learning_rate": 9.447839230644349e-06, + "loss": 0.9359, + "step": 22584 + }, + { + "epoch": 0.53, + "grad_norm": 2.3457492282681915, + "learning_rate": 9.447077355265698e-06, + "loss": 0.994, + "step": 22585 + }, + { + "epoch": 0.53, + "grad_norm": 2.0193942614289506, + "learning_rate": 9.44631548310634e-06, + "loss": 0.9703, + "step": 22586 + }, + { + "epoch": 0.53, + "grad_norm": 2.024293943012114, + "learning_rate": 9.445553614170705e-06, + "loss": 1.0607, + "step": 22587 + }, + { + "epoch": 0.53, + "grad_norm": 1.928064864148633, + "learning_rate": 9.444791748463232e-06, + "loss": 1.1608, + "step": 22588 + }, + { + "epoch": 0.53, + "grad_norm": 1.8406753751122529, + "learning_rate": 9.444029885988359e-06, + "loss": 1.0184, + "step": 22589 + }, + { + "epoch": 0.53, + "grad_norm": 1.1015711890336188, + "learning_rate": 9.443268026750521e-06, + "loss": 0.9532, + "step": 22590 + }, + { + "epoch": 0.53, + "grad_norm": 2.3493977512431137, + "learning_rate": 9.442506170754149e-06, + "loss": 0.9131, + "step": 22591 + }, + { + "epoch": 0.53, + "grad_norm": 2.1600464653168054, + "learning_rate": 9.441744318003685e-06, + "loss": 0.9853, + "step": 22592 + }, + { + "epoch": 0.53, + "grad_norm": 2.2105058482040594, + "learning_rate": 9.440982468503563e-06, + "loss": 1.0008, + "step": 22593 + }, + { + "epoch": 0.53, + "grad_norm": 2.1210723087946515, + "learning_rate": 9.440220622258217e-06, + "loss": 0.9318, + "step": 22594 + }, + { + "epoch": 0.53, + "grad_norm": 2.1794098682379275, + "learning_rate": 9.439458779272082e-06, + "loss": 0.9126, + "step": 22595 + }, + { + "epoch": 0.53, + "grad_norm": 2.6515738437441687, + "learning_rate": 9.438696939549598e-06, + "loss": 1.0375, + "step": 22596 + }, + { + "epoch": 0.53, + "grad_norm": 1.8482914396324879, + "learning_rate": 9.437935103095201e-06, + "loss": 0.9817, + "step": 22597 + }, + { + "epoch": 0.53, + "grad_norm": 1.0741641981188355, + "learning_rate": 9.43717326991332e-06, + "loss": 0.922, + "step": 22598 + }, + { + "epoch": 0.53, + "grad_norm": 1.8934392710078114, + "learning_rate": 9.436411440008392e-06, + "loss": 1.0045, + "step": 22599 + }, + { + "epoch": 0.53, + "grad_norm": 2.4685292933742677, + "learning_rate": 9.435649613384859e-06, + "loss": 0.9699, + "step": 22600 + }, + { + "epoch": 0.53, + "grad_norm": 2.273385565736437, + "learning_rate": 9.434887790047152e-06, + "loss": 1.0372, + "step": 22601 + }, + { + "epoch": 0.53, + "grad_norm": 1.8682718001634975, + "learning_rate": 9.434125969999707e-06, + "loss": 0.895, + "step": 22602 + }, + { + "epoch": 0.53, + "grad_norm": 1.9292650946519254, + "learning_rate": 9.433364153246958e-06, + "loss": 0.9659, + "step": 22603 + }, + { + "epoch": 0.53, + "grad_norm": 2.113811748265459, + "learning_rate": 9.432602339793346e-06, + "loss": 1.0286, + "step": 22604 + }, + { + "epoch": 0.53, + "grad_norm": 2.042606017665401, + "learning_rate": 9.4318405296433e-06, + "loss": 0.9749, + "step": 22605 + }, + { + "epoch": 0.53, + "grad_norm": 2.2114791726801024, + "learning_rate": 9.43107872280126e-06, + "loss": 0.8802, + "step": 22606 + }, + { + "epoch": 0.53, + "grad_norm": 2.2508827796067576, + "learning_rate": 9.43031691927166e-06, + "loss": 1.014, + "step": 22607 + }, + { + "epoch": 0.53, + "grad_norm": 1.9317236490506866, + "learning_rate": 9.429555119058931e-06, + "loss": 1.0382, + "step": 22608 + }, + { + "epoch": 0.53, + "grad_norm": 2.170967590334872, + "learning_rate": 9.428793322167516e-06, + "loss": 0.9804, + "step": 22609 + }, + { + "epoch": 0.53, + "grad_norm": 2.0601832245213356, + "learning_rate": 9.428031528601846e-06, + "loss": 1.1313, + "step": 22610 + }, + { + "epoch": 0.53, + "grad_norm": 1.8908872949505695, + "learning_rate": 9.427269738366361e-06, + "loss": 1.0714, + "step": 22611 + }, + { + "epoch": 0.53, + "grad_norm": 2.0693118111768825, + "learning_rate": 9.426507951465487e-06, + "loss": 0.9623, + "step": 22612 + }, + { + "epoch": 0.53, + "grad_norm": 1.829632401639509, + "learning_rate": 9.42574616790367e-06, + "loss": 1.0199, + "step": 22613 + }, + { + "epoch": 0.53, + "grad_norm": 1.9612577420740067, + "learning_rate": 9.42498438768534e-06, + "loss": 1.0284, + "step": 22614 + }, + { + "epoch": 0.53, + "grad_norm": 2.0307691513309822, + "learning_rate": 9.424222610814932e-06, + "loss": 0.9099, + "step": 22615 + }, + { + "epoch": 0.53, + "grad_norm": 1.9308398386956487, + "learning_rate": 9.423460837296879e-06, + "loss": 1.0722, + "step": 22616 + }, + { + "epoch": 0.53, + "grad_norm": 1.9474490559022, + "learning_rate": 9.422699067135624e-06, + "loss": 1.0763, + "step": 22617 + }, + { + "epoch": 0.53, + "grad_norm": 2.006969324897565, + "learning_rate": 9.4219373003356e-06, + "loss": 0.9391, + "step": 22618 + }, + { + "epoch": 0.53, + "grad_norm": 1.9621325028755874, + "learning_rate": 9.421175536901234e-06, + "loss": 1.0869, + "step": 22619 + }, + { + "epoch": 0.53, + "grad_norm": 2.115299241030342, + "learning_rate": 9.420413776836968e-06, + "loss": 1.0251, + "step": 22620 + }, + { + "epoch": 0.53, + "grad_norm": 1.9592728686437269, + "learning_rate": 9.41965202014724e-06, + "loss": 1.1185, + "step": 22621 + }, + { + "epoch": 0.53, + "grad_norm": 1.8490420533112784, + "learning_rate": 9.41889026683648e-06, + "loss": 1.0947, + "step": 22622 + }, + { + "epoch": 0.53, + "grad_norm": 1.123454183647269, + "learning_rate": 9.418128516909126e-06, + "loss": 1.0621, + "step": 22623 + }, + { + "epoch": 0.53, + "grad_norm": 1.1485172257435925, + "learning_rate": 9.417366770369609e-06, + "loss": 0.9805, + "step": 22624 + }, + { + "epoch": 0.53, + "grad_norm": 1.8737755505154712, + "learning_rate": 9.41660502722237e-06, + "loss": 0.936, + "step": 22625 + }, + { + "epoch": 0.53, + "grad_norm": 2.402533040303903, + "learning_rate": 9.41584328747184e-06, + "loss": 1.0963, + "step": 22626 + }, + { + "epoch": 0.53, + "grad_norm": 1.7736333555618329, + "learning_rate": 9.415081551122453e-06, + "loss": 1.0099, + "step": 22627 + }, + { + "epoch": 0.53, + "grad_norm": 2.2857851520187937, + "learning_rate": 9.414319818178654e-06, + "loss": 1.1661, + "step": 22628 + }, + { + "epoch": 0.53, + "grad_norm": 1.1066854075230175, + "learning_rate": 9.413558088644864e-06, + "loss": 0.8996, + "step": 22629 + }, + { + "epoch": 0.53, + "grad_norm": 2.336357538777244, + "learning_rate": 9.412796362525526e-06, + "loss": 0.9471, + "step": 22630 + }, + { + "epoch": 0.53, + "grad_norm": 1.116883880407353, + "learning_rate": 9.41203463982507e-06, + "loss": 0.9432, + "step": 22631 + }, + { + "epoch": 0.53, + "grad_norm": 2.0309998447637763, + "learning_rate": 9.411272920547942e-06, + "loss": 1.0445, + "step": 22632 + }, + { + "epoch": 0.53, + "grad_norm": 1.8611302303742057, + "learning_rate": 9.410511204698562e-06, + "loss": 1.058, + "step": 22633 + }, + { + "epoch": 0.53, + "grad_norm": 2.3417064455013823, + "learning_rate": 9.409749492281378e-06, + "loss": 1.0594, + "step": 22634 + }, + { + "epoch": 0.53, + "grad_norm": 2.0454853263286688, + "learning_rate": 9.408987783300818e-06, + "loss": 0.974, + "step": 22635 + }, + { + "epoch": 0.53, + "grad_norm": 2.1877113031613984, + "learning_rate": 9.408226077761316e-06, + "loss": 1.0335, + "step": 22636 + }, + { + "epoch": 0.53, + "grad_norm": 1.9404190831837742, + "learning_rate": 9.407464375667309e-06, + "loss": 0.9743, + "step": 22637 + }, + { + "epoch": 0.53, + "grad_norm": 2.0117979308397014, + "learning_rate": 9.406702677023233e-06, + "loss": 1.0399, + "step": 22638 + }, + { + "epoch": 0.53, + "grad_norm": 2.3730919730646005, + "learning_rate": 9.405940981833525e-06, + "loss": 1.1099, + "step": 22639 + }, + { + "epoch": 0.53, + "grad_norm": 2.1188080426695897, + "learning_rate": 9.405179290102611e-06, + "loss": 0.9856, + "step": 22640 + }, + { + "epoch": 0.53, + "grad_norm": 2.057398913048754, + "learning_rate": 9.404417601834935e-06, + "loss": 1.0959, + "step": 22641 + }, + { + "epoch": 0.53, + "grad_norm": 2.468786989403467, + "learning_rate": 9.40365591703493e-06, + "loss": 1.0949, + "step": 22642 + }, + { + "epoch": 0.53, + "grad_norm": 1.933362313469807, + "learning_rate": 9.402894235707026e-06, + "loss": 0.9416, + "step": 22643 + }, + { + "epoch": 0.53, + "grad_norm": 2.007903493705328, + "learning_rate": 9.402132557855659e-06, + "loss": 1.0679, + "step": 22644 + }, + { + "epoch": 0.53, + "grad_norm": 2.320286874276399, + "learning_rate": 9.401370883485268e-06, + "loss": 0.9431, + "step": 22645 + }, + { + "epoch": 0.53, + "grad_norm": 1.9110654104887634, + "learning_rate": 9.400609212600288e-06, + "loss": 0.9767, + "step": 22646 + }, + { + "epoch": 0.53, + "grad_norm": 1.8736501474636331, + "learning_rate": 9.399847545205148e-06, + "loss": 1.1313, + "step": 22647 + }, + { + "epoch": 0.53, + "grad_norm": 2.1431469945602544, + "learning_rate": 9.399085881304283e-06, + "loss": 1.0415, + "step": 22648 + }, + { + "epoch": 0.53, + "grad_norm": 1.7764401186263115, + "learning_rate": 9.398324220902135e-06, + "loss": 0.9531, + "step": 22649 + }, + { + "epoch": 0.53, + "grad_norm": 1.1442931455660923, + "learning_rate": 9.397562564003131e-06, + "loss": 0.9771, + "step": 22650 + }, + { + "epoch": 0.53, + "grad_norm": 2.4968781235052293, + "learning_rate": 9.39680091061171e-06, + "loss": 0.9728, + "step": 22651 + }, + { + "epoch": 0.53, + "grad_norm": 2.009074026437747, + "learning_rate": 9.396039260732302e-06, + "loss": 0.9917, + "step": 22652 + }, + { + "epoch": 0.53, + "grad_norm": 1.981480144454968, + "learning_rate": 9.39527761436935e-06, + "loss": 0.9264, + "step": 22653 + }, + { + "epoch": 0.53, + "grad_norm": 2.1757088365352795, + "learning_rate": 9.39451597152728e-06, + "loss": 1.0943, + "step": 22654 + }, + { + "epoch": 0.53, + "grad_norm": 2.0763899955711453, + "learning_rate": 9.393754332210529e-06, + "loss": 0.9934, + "step": 22655 + }, + { + "epoch": 0.53, + "grad_norm": 2.0818313802174817, + "learning_rate": 9.392992696423536e-06, + "loss": 0.8843, + "step": 22656 + }, + { + "epoch": 0.53, + "grad_norm": 1.7514465762850877, + "learning_rate": 9.392231064170728e-06, + "loss": 1.0133, + "step": 22657 + }, + { + "epoch": 0.53, + "grad_norm": 1.9418664812244781, + "learning_rate": 9.391469435456543e-06, + "loss": 1.0089, + "step": 22658 + }, + { + "epoch": 0.53, + "grad_norm": 1.1036581763886144, + "learning_rate": 9.390707810285416e-06, + "loss": 0.954, + "step": 22659 + }, + { + "epoch": 0.53, + "grad_norm": 2.128714048498148, + "learning_rate": 9.389946188661785e-06, + "loss": 1.0915, + "step": 22660 + }, + { + "epoch": 0.53, + "grad_norm": 2.1079394845135044, + "learning_rate": 9.389184570590075e-06, + "loss": 1.065, + "step": 22661 + }, + { + "epoch": 0.53, + "grad_norm": 2.0043350040131926, + "learning_rate": 9.388422956074727e-06, + "loss": 0.9064, + "step": 22662 + }, + { + "epoch": 0.53, + "grad_norm": 1.9974616270091161, + "learning_rate": 9.387661345120178e-06, + "loss": 0.9608, + "step": 22663 + }, + { + "epoch": 0.53, + "grad_norm": 2.1913231569999057, + "learning_rate": 9.386899737730854e-06, + "loss": 0.9365, + "step": 22664 + }, + { + "epoch": 0.53, + "grad_norm": 1.9136756984194006, + "learning_rate": 9.386138133911193e-06, + "loss": 1.0041, + "step": 22665 + }, + { + "epoch": 0.53, + "grad_norm": 2.4337040889947263, + "learning_rate": 9.385376533665633e-06, + "loss": 0.8685, + "step": 22666 + }, + { + "epoch": 0.53, + "grad_norm": 2.046522845826651, + "learning_rate": 9.384614936998607e-06, + "loss": 1.0076, + "step": 22667 + }, + { + "epoch": 0.53, + "grad_norm": 2.051673238659815, + "learning_rate": 9.383853343914546e-06, + "loss": 1.0469, + "step": 22668 + }, + { + "epoch": 0.53, + "grad_norm": 2.193628803875869, + "learning_rate": 9.383091754417884e-06, + "loss": 1.1128, + "step": 22669 + }, + { + "epoch": 0.53, + "grad_norm": 2.1481691947208255, + "learning_rate": 9.382330168513058e-06, + "loss": 1.0751, + "step": 22670 + }, + { + "epoch": 0.53, + "grad_norm": 1.991979417732856, + "learning_rate": 9.381568586204504e-06, + "loss": 1.0546, + "step": 22671 + }, + { + "epoch": 0.53, + "grad_norm": 1.1221601949868516, + "learning_rate": 9.380807007496652e-06, + "loss": 0.8786, + "step": 22672 + }, + { + "epoch": 0.53, + "grad_norm": 2.121214516768774, + "learning_rate": 9.380045432393934e-06, + "loss": 0.9804, + "step": 22673 + }, + { + "epoch": 0.53, + "grad_norm": 2.0993134570621206, + "learning_rate": 9.379283860900793e-06, + "loss": 0.846, + "step": 22674 + }, + { + "epoch": 0.53, + "grad_norm": 2.0980473888272653, + "learning_rate": 9.378522293021656e-06, + "loss": 1.0597, + "step": 22675 + }, + { + "epoch": 0.53, + "grad_norm": 2.1261052231588757, + "learning_rate": 9.377760728760957e-06, + "loss": 1.0606, + "step": 22676 + }, + { + "epoch": 0.53, + "grad_norm": 2.243842148561347, + "learning_rate": 9.376999168123134e-06, + "loss": 1.0474, + "step": 22677 + }, + { + "epoch": 0.53, + "grad_norm": 2.1342902498961203, + "learning_rate": 9.37623761111262e-06, + "loss": 0.9927, + "step": 22678 + }, + { + "epoch": 0.53, + "grad_norm": 1.0471313820050232, + "learning_rate": 9.375476057733847e-06, + "loss": 0.9961, + "step": 22679 + }, + { + "epoch": 0.53, + "grad_norm": 2.289627883167743, + "learning_rate": 9.374714507991248e-06, + "loss": 1.0581, + "step": 22680 + }, + { + "epoch": 0.53, + "grad_norm": 2.1202213903312614, + "learning_rate": 9.373952961889265e-06, + "loss": 0.9624, + "step": 22681 + }, + { + "epoch": 0.53, + "grad_norm": 2.732226201666265, + "learning_rate": 9.373191419432319e-06, + "loss": 1.053, + "step": 22682 + }, + { + "epoch": 0.53, + "grad_norm": 2.266471754812755, + "learning_rate": 9.372429880624856e-06, + "loss": 0.9951, + "step": 22683 + }, + { + "epoch": 0.53, + "grad_norm": 1.7474019663088876, + "learning_rate": 9.3716683454713e-06, + "loss": 0.9607, + "step": 22684 + }, + { + "epoch": 0.53, + "grad_norm": 1.9312508108255388, + "learning_rate": 9.370906813976097e-06, + "loss": 0.9811, + "step": 22685 + }, + { + "epoch": 0.53, + "grad_norm": 1.7867089052141762, + "learning_rate": 9.370145286143667e-06, + "loss": 0.9742, + "step": 22686 + }, + { + "epoch": 0.53, + "grad_norm": 1.96737700212674, + "learning_rate": 9.369383761978453e-06, + "loss": 0.9307, + "step": 22687 + }, + { + "epoch": 0.53, + "grad_norm": 1.7825521613258712, + "learning_rate": 9.36862224148489e-06, + "loss": 0.9243, + "step": 22688 + }, + { + "epoch": 0.53, + "grad_norm": 1.8793251461983225, + "learning_rate": 9.367860724667401e-06, + "loss": 1.0456, + "step": 22689 + }, + { + "epoch": 0.53, + "grad_norm": 1.8222623178016604, + "learning_rate": 9.367099211530432e-06, + "loss": 1.0863, + "step": 22690 + }, + { + "epoch": 0.53, + "grad_norm": 2.2666651119221624, + "learning_rate": 9.36633770207841e-06, + "loss": 1.0174, + "step": 22691 + }, + { + "epoch": 0.53, + "grad_norm": 1.993874510042264, + "learning_rate": 9.365576196315772e-06, + "loss": 0.9772, + "step": 22692 + }, + { + "epoch": 0.53, + "grad_norm": 2.2485587956514412, + "learning_rate": 9.364814694246947e-06, + "loss": 1.024, + "step": 22693 + }, + { + "epoch": 0.53, + "grad_norm": 2.0358514583373633, + "learning_rate": 9.364053195876375e-06, + "loss": 1.0273, + "step": 22694 + }, + { + "epoch": 0.53, + "grad_norm": 1.9476205706532084, + "learning_rate": 9.363291701208488e-06, + "loss": 1.0495, + "step": 22695 + }, + { + "epoch": 0.53, + "grad_norm": 1.9747509863132384, + "learning_rate": 9.362530210247715e-06, + "loss": 0.9877, + "step": 22696 + }, + { + "epoch": 0.53, + "grad_norm": 1.9703667155539022, + "learning_rate": 9.361768722998494e-06, + "loss": 1.0794, + "step": 22697 + }, + { + "epoch": 0.53, + "grad_norm": 2.1594808499800924, + "learning_rate": 9.361007239465257e-06, + "loss": 1.0615, + "step": 22698 + }, + { + "epoch": 0.53, + "grad_norm": 2.097628564958248, + "learning_rate": 9.36024575965244e-06, + "loss": 1.079, + "step": 22699 + }, + { + "epoch": 0.53, + "grad_norm": 1.949469799108333, + "learning_rate": 9.359484283564474e-06, + "loss": 1.0665, + "step": 22700 + }, + { + "epoch": 0.53, + "grad_norm": 1.9729670551806842, + "learning_rate": 9.35872281120579e-06, + "loss": 0.8791, + "step": 22701 + }, + { + "epoch": 0.53, + "grad_norm": 2.130337340450156, + "learning_rate": 9.35796134258083e-06, + "loss": 1.0536, + "step": 22702 + }, + { + "epoch": 0.53, + "grad_norm": 1.9615858586381405, + "learning_rate": 9.35719987769402e-06, + "loss": 1.0866, + "step": 22703 + }, + { + "epoch": 0.53, + "grad_norm": 2.572921994816671, + "learning_rate": 9.356438416549795e-06, + "loss": 1.1416, + "step": 22704 + }, + { + "epoch": 0.53, + "grad_norm": 1.8457057949077127, + "learning_rate": 9.355676959152588e-06, + "loss": 1.1014, + "step": 22705 + }, + { + "epoch": 0.53, + "grad_norm": 2.1916692731251457, + "learning_rate": 9.354915505506839e-06, + "loss": 1.0485, + "step": 22706 + }, + { + "epoch": 0.53, + "grad_norm": 2.1140466868770593, + "learning_rate": 9.354154055616971e-06, + "loss": 1.0481, + "step": 22707 + }, + { + "epoch": 0.53, + "grad_norm": 2.3038203875231846, + "learning_rate": 9.353392609487427e-06, + "loss": 0.9703, + "step": 22708 + }, + { + "epoch": 0.54, + "grad_norm": 2.0438040444391157, + "learning_rate": 9.352631167122635e-06, + "loss": 1.1207, + "step": 22709 + }, + { + "epoch": 0.54, + "grad_norm": 2.1688732903222236, + "learning_rate": 9.351869728527024e-06, + "loss": 0.9334, + "step": 22710 + }, + { + "epoch": 0.54, + "grad_norm": 1.9745831694170526, + "learning_rate": 9.351108293705037e-06, + "loss": 0.9289, + "step": 22711 + }, + { + "epoch": 0.54, + "grad_norm": 1.8718139267537466, + "learning_rate": 9.3503468626611e-06, + "loss": 1.0818, + "step": 22712 + }, + { + "epoch": 0.54, + "grad_norm": 2.0581854659899617, + "learning_rate": 9.349585435399656e-06, + "loss": 1.1034, + "step": 22713 + }, + { + "epoch": 0.54, + "grad_norm": 1.7808147894642026, + "learning_rate": 9.348824011925124e-06, + "loss": 1.0332, + "step": 22714 + }, + { + "epoch": 0.54, + "grad_norm": 2.0487144908493815, + "learning_rate": 9.34806259224195e-06, + "loss": 0.9405, + "step": 22715 + }, + { + "epoch": 0.54, + "grad_norm": 2.2611900258839115, + "learning_rate": 9.34730117635456e-06, + "loss": 1.0452, + "step": 22716 + }, + { + "epoch": 0.54, + "grad_norm": 1.8317427967934314, + "learning_rate": 9.346539764267389e-06, + "loss": 0.8952, + "step": 22717 + }, + { + "epoch": 0.54, + "grad_norm": 1.8861006548404515, + "learning_rate": 9.345778355984869e-06, + "loss": 0.9757, + "step": 22718 + }, + { + "epoch": 0.54, + "grad_norm": 2.056403398983365, + "learning_rate": 9.345016951511436e-06, + "loss": 0.9987, + "step": 22719 + }, + { + "epoch": 0.54, + "grad_norm": 1.0879749099793867, + "learning_rate": 9.344255550851523e-06, + "loss": 0.9689, + "step": 22720 + }, + { + "epoch": 0.54, + "grad_norm": 2.1586223880103446, + "learning_rate": 9.34349415400956e-06, + "loss": 0.9523, + "step": 22721 + }, + { + "epoch": 0.54, + "grad_norm": 2.040822192088169, + "learning_rate": 9.34273276098998e-06, + "loss": 1.0952, + "step": 22722 + }, + { + "epoch": 0.54, + "grad_norm": 1.9935975387228766, + "learning_rate": 9.341971371797223e-06, + "loss": 1.1127, + "step": 22723 + }, + { + "epoch": 0.54, + "grad_norm": 1.7974695485429097, + "learning_rate": 9.341209986435715e-06, + "loss": 1.0047, + "step": 22724 + }, + { + "epoch": 0.54, + "grad_norm": 1.049087633384799, + "learning_rate": 9.340448604909888e-06, + "loss": 0.8868, + "step": 22725 + }, + { + "epoch": 0.54, + "grad_norm": 1.9203111796184436, + "learning_rate": 9.339687227224178e-06, + "loss": 1.0637, + "step": 22726 + }, + { + "epoch": 0.54, + "grad_norm": 2.035136064746904, + "learning_rate": 9.338925853383024e-06, + "loss": 0.9581, + "step": 22727 + }, + { + "epoch": 0.54, + "grad_norm": 1.1364550746586877, + "learning_rate": 9.338164483390849e-06, + "loss": 0.9191, + "step": 22728 + }, + { + "epoch": 0.54, + "grad_norm": 1.9096389633946942, + "learning_rate": 9.337403117252089e-06, + "loss": 0.8473, + "step": 22729 + }, + { + "epoch": 0.54, + "grad_norm": 2.2675995985882293, + "learning_rate": 9.336641754971183e-06, + "loss": 1.0252, + "step": 22730 + }, + { + "epoch": 0.54, + "grad_norm": 2.1455819824023328, + "learning_rate": 9.335880396552552e-06, + "loss": 0.9864, + "step": 22731 + }, + { + "epoch": 0.54, + "grad_norm": 2.699318432916706, + "learning_rate": 9.335119042000637e-06, + "loss": 0.9438, + "step": 22732 + }, + { + "epoch": 0.54, + "grad_norm": 3.5312763259020317, + "learning_rate": 9.33435769131987e-06, + "loss": 0.9418, + "step": 22733 + }, + { + "epoch": 0.54, + "grad_norm": 1.05795769007174, + "learning_rate": 9.333596344514689e-06, + "loss": 1.0309, + "step": 22734 + }, + { + "epoch": 0.54, + "grad_norm": 2.1758184326006442, + "learning_rate": 9.332835001589514e-06, + "loss": 1.0553, + "step": 22735 + }, + { + "epoch": 0.54, + "grad_norm": 1.1258116916328609, + "learning_rate": 9.332073662548785e-06, + "loss": 0.9818, + "step": 22736 + }, + { + "epoch": 0.54, + "grad_norm": 2.015060347562692, + "learning_rate": 9.33131232739694e-06, + "loss": 1.1002, + "step": 22737 + }, + { + "epoch": 0.54, + "grad_norm": 1.1788643746589123, + "learning_rate": 9.3305509961384e-06, + "loss": 1.0131, + "step": 22738 + }, + { + "epoch": 0.54, + "grad_norm": 1.8725248671954775, + "learning_rate": 9.329789668777606e-06, + "loss": 1.0694, + "step": 22739 + }, + { + "epoch": 0.54, + "grad_norm": 1.9054501859439799, + "learning_rate": 9.32902834531899e-06, + "loss": 1.0484, + "step": 22740 + }, + { + "epoch": 0.54, + "grad_norm": 1.6938320072866444, + "learning_rate": 9.328267025766984e-06, + "loss": 1.0756, + "step": 22741 + }, + { + "epoch": 0.54, + "grad_norm": 1.6868264716966717, + "learning_rate": 9.327505710126018e-06, + "loss": 0.9492, + "step": 22742 + }, + { + "epoch": 0.54, + "grad_norm": 1.9460260945228822, + "learning_rate": 9.326744398400527e-06, + "loss": 1.0202, + "step": 22743 + }, + { + "epoch": 0.54, + "grad_norm": 2.0737667440412784, + "learning_rate": 9.325983090594946e-06, + "loss": 0.873, + "step": 22744 + }, + { + "epoch": 0.54, + "grad_norm": 2.301242574818801, + "learning_rate": 9.325221786713704e-06, + "loss": 1.005, + "step": 22745 + }, + { + "epoch": 0.54, + "grad_norm": 1.959276162159382, + "learning_rate": 9.324460486761231e-06, + "loss": 0.8552, + "step": 22746 + }, + { + "epoch": 0.54, + "grad_norm": 2.097844898653505, + "learning_rate": 9.323699190741964e-06, + "loss": 0.9918, + "step": 22747 + }, + { + "epoch": 0.54, + "grad_norm": 1.8457654683871723, + "learning_rate": 9.322937898660339e-06, + "loss": 0.9136, + "step": 22748 + }, + { + "epoch": 0.54, + "grad_norm": 2.3576712522856416, + "learning_rate": 9.32217661052078e-06, + "loss": 1.0122, + "step": 22749 + }, + { + "epoch": 0.54, + "grad_norm": 2.257755207144157, + "learning_rate": 9.321415326327721e-06, + "loss": 0.8524, + "step": 22750 + }, + { + "epoch": 0.54, + "grad_norm": 1.3034177003892922, + "learning_rate": 9.320654046085602e-06, + "loss": 0.9525, + "step": 22751 + }, + { + "epoch": 0.54, + "grad_norm": 2.4531783830664153, + "learning_rate": 9.319892769798848e-06, + "loss": 1.0144, + "step": 22752 + }, + { + "epoch": 0.54, + "grad_norm": 2.8202599770867454, + "learning_rate": 9.319131497471894e-06, + "loss": 1.027, + "step": 22753 + }, + { + "epoch": 0.54, + "grad_norm": 2.020018495854958, + "learning_rate": 9.318370229109168e-06, + "loss": 0.9533, + "step": 22754 + }, + { + "epoch": 0.54, + "grad_norm": 1.9311765285260372, + "learning_rate": 9.317608964715113e-06, + "loss": 1.0302, + "step": 22755 + }, + { + "epoch": 0.54, + "grad_norm": 1.8670890311538617, + "learning_rate": 9.31684770429415e-06, + "loss": 1.0129, + "step": 22756 + }, + { + "epoch": 0.54, + "grad_norm": 2.300387948327832, + "learning_rate": 9.316086447850718e-06, + "loss": 0.967, + "step": 22757 + }, + { + "epoch": 0.54, + "grad_norm": 2.0065040164223884, + "learning_rate": 9.31532519538925e-06, + "loss": 0.8987, + "step": 22758 + }, + { + "epoch": 0.54, + "grad_norm": 2.0828742295410763, + "learning_rate": 9.314563946914169e-06, + "loss": 1.118, + "step": 22759 + }, + { + "epoch": 0.54, + "grad_norm": 1.8839704340708245, + "learning_rate": 9.313802702429918e-06, + "loss": 1.0329, + "step": 22760 + }, + { + "epoch": 0.54, + "grad_norm": 1.8473101002077723, + "learning_rate": 9.31304146194092e-06, + "loss": 0.9372, + "step": 22761 + }, + { + "epoch": 0.54, + "grad_norm": 1.0654605774461832, + "learning_rate": 9.31228022545162e-06, + "loss": 0.9381, + "step": 22762 + }, + { + "epoch": 0.54, + "grad_norm": 1.9024937222257912, + "learning_rate": 9.311518992966436e-06, + "loss": 0.9588, + "step": 22763 + }, + { + "epoch": 0.54, + "grad_norm": 2.345708874235643, + "learning_rate": 9.310757764489808e-06, + "loss": 1.0443, + "step": 22764 + }, + { + "epoch": 0.54, + "grad_norm": 2.689512626139347, + "learning_rate": 9.309996540026169e-06, + "loss": 1.0139, + "step": 22765 + }, + { + "epoch": 0.54, + "grad_norm": 1.957095627013791, + "learning_rate": 9.309235319579947e-06, + "loss": 0.9682, + "step": 22766 + }, + { + "epoch": 0.54, + "grad_norm": 2.401446699409991, + "learning_rate": 9.308474103155573e-06, + "loss": 0.9536, + "step": 22767 + }, + { + "epoch": 0.54, + "grad_norm": 1.955915329911588, + "learning_rate": 9.307712890757483e-06, + "loss": 0.9544, + "step": 22768 + }, + { + "epoch": 0.54, + "grad_norm": 1.0200395234049264, + "learning_rate": 9.306951682390112e-06, + "loss": 0.9637, + "step": 22769 + }, + { + "epoch": 0.54, + "grad_norm": 2.022474051054949, + "learning_rate": 9.306190478057884e-06, + "loss": 1.0083, + "step": 22770 + }, + { + "epoch": 0.54, + "grad_norm": 1.9895486383815568, + "learning_rate": 9.305429277765233e-06, + "loss": 1.0317, + "step": 22771 + }, + { + "epoch": 0.54, + "grad_norm": 2.1029472028024663, + "learning_rate": 9.304668081516598e-06, + "loss": 1.0102, + "step": 22772 + }, + { + "epoch": 0.54, + "grad_norm": 1.9937982352834198, + "learning_rate": 9.3039068893164e-06, + "loss": 0.8989, + "step": 22773 + }, + { + "epoch": 0.54, + "grad_norm": 2.043274471600692, + "learning_rate": 9.303145701169077e-06, + "loss": 1.147, + "step": 22774 + }, + { + "epoch": 0.54, + "grad_norm": 1.8636663089497973, + "learning_rate": 9.302384517079061e-06, + "loss": 0.9521, + "step": 22775 + }, + { + "epoch": 0.54, + "grad_norm": 2.079904629743093, + "learning_rate": 9.301623337050788e-06, + "loss": 0.8581, + "step": 22776 + }, + { + "epoch": 0.54, + "grad_norm": 2.153835509098977, + "learning_rate": 9.30086216108868e-06, + "loss": 1.0166, + "step": 22777 + }, + { + "epoch": 0.54, + "grad_norm": 1.90600209878927, + "learning_rate": 9.300100989197174e-06, + "loss": 0.9922, + "step": 22778 + }, + { + "epoch": 0.54, + "grad_norm": 2.3838664805707395, + "learning_rate": 9.299339821380708e-06, + "loss": 1.1251, + "step": 22779 + }, + { + "epoch": 0.54, + "grad_norm": 1.8354449818839542, + "learning_rate": 9.2985786576437e-06, + "loss": 0.8003, + "step": 22780 + }, + { + "epoch": 0.54, + "grad_norm": 2.075551782396828, + "learning_rate": 9.29781749799059e-06, + "loss": 0.9498, + "step": 22781 + }, + { + "epoch": 0.54, + "grad_norm": 2.3643161333376117, + "learning_rate": 9.297056342425807e-06, + "loss": 1.0496, + "step": 22782 + }, + { + "epoch": 0.54, + "grad_norm": 1.998797347995, + "learning_rate": 9.296295190953792e-06, + "loss": 1.0312, + "step": 22783 + }, + { + "epoch": 0.54, + "grad_norm": 1.8850419998747825, + "learning_rate": 9.295534043578963e-06, + "loss": 0.8586, + "step": 22784 + }, + { + "epoch": 0.54, + "grad_norm": 2.0698949207155204, + "learning_rate": 9.29477290030576e-06, + "loss": 1.0675, + "step": 22785 + }, + { + "epoch": 0.54, + "grad_norm": 2.566123245332567, + "learning_rate": 9.294011761138614e-06, + "loss": 1.0606, + "step": 22786 + }, + { + "epoch": 0.54, + "grad_norm": 2.026059688711005, + "learning_rate": 9.293250626081951e-06, + "loss": 0.9854, + "step": 22787 + }, + { + "epoch": 0.54, + "grad_norm": 1.9136172726823328, + "learning_rate": 9.292489495140209e-06, + "loss": 1.0593, + "step": 22788 + }, + { + "epoch": 0.54, + "grad_norm": 2.0883744959754567, + "learning_rate": 9.291728368317816e-06, + "loss": 1.0479, + "step": 22789 + }, + { + "epoch": 0.54, + "grad_norm": 2.052907535692775, + "learning_rate": 9.290967245619208e-06, + "loss": 1.1357, + "step": 22790 + }, + { + "epoch": 0.54, + "grad_norm": 1.0986614611387178, + "learning_rate": 9.290206127048809e-06, + "loss": 0.9659, + "step": 22791 + }, + { + "epoch": 0.54, + "grad_norm": 2.001822539117821, + "learning_rate": 9.289445012611056e-06, + "loss": 0.9213, + "step": 22792 + }, + { + "epoch": 0.54, + "grad_norm": 2.6202440584353357, + "learning_rate": 9.28868390231038e-06, + "loss": 0.9711, + "step": 22793 + }, + { + "epoch": 0.54, + "grad_norm": 2.073509478824382, + "learning_rate": 9.287922796151214e-06, + "loss": 1.0221, + "step": 22794 + }, + { + "epoch": 0.54, + "grad_norm": 2.073003942777529, + "learning_rate": 9.28716169413798e-06, + "loss": 0.9369, + "step": 22795 + }, + { + "epoch": 0.54, + "grad_norm": 2.0080483003344574, + "learning_rate": 9.28640059627512e-06, + "loss": 1.0226, + "step": 22796 + }, + { + "epoch": 0.54, + "grad_norm": 2.2723510956006128, + "learning_rate": 9.285639502567065e-06, + "loss": 1.035, + "step": 22797 + }, + { + "epoch": 0.54, + "grad_norm": 1.943486581765271, + "learning_rate": 9.284878413018238e-06, + "loss": 1.101, + "step": 22798 + }, + { + "epoch": 0.54, + "grad_norm": 1.9528409672680653, + "learning_rate": 9.284117327633077e-06, + "loss": 1.182, + "step": 22799 + }, + { + "epoch": 0.54, + "grad_norm": 2.199443516310891, + "learning_rate": 9.28335624641601e-06, + "loss": 0.9337, + "step": 22800 + }, + { + "epoch": 0.54, + "grad_norm": 2.6046480807831025, + "learning_rate": 9.282595169371474e-06, + "loss": 1.0611, + "step": 22801 + }, + { + "epoch": 0.54, + "grad_norm": 1.7872977569182251, + "learning_rate": 9.281834096503895e-06, + "loss": 0.9421, + "step": 22802 + }, + { + "epoch": 0.54, + "grad_norm": 1.8353454397372817, + "learning_rate": 9.281073027817702e-06, + "loss": 0.9424, + "step": 22803 + }, + { + "epoch": 0.54, + "grad_norm": 2.060207090000943, + "learning_rate": 9.280311963317332e-06, + "loss": 0.9415, + "step": 22804 + }, + { + "epoch": 0.54, + "grad_norm": 1.9905409240507375, + "learning_rate": 9.279550903007212e-06, + "loss": 0.9182, + "step": 22805 + }, + { + "epoch": 0.54, + "grad_norm": 2.1642686993306444, + "learning_rate": 9.278789846891777e-06, + "loss": 1.0204, + "step": 22806 + }, + { + "epoch": 0.54, + "grad_norm": 2.263414901344804, + "learning_rate": 9.278028794975452e-06, + "loss": 0.9078, + "step": 22807 + }, + { + "epoch": 0.54, + "grad_norm": 1.8670005223395167, + "learning_rate": 9.277267747262676e-06, + "loss": 1.0294, + "step": 22808 + }, + { + "epoch": 0.54, + "grad_norm": 1.9415766603555409, + "learning_rate": 9.276506703757875e-06, + "loss": 1.0496, + "step": 22809 + }, + { + "epoch": 0.54, + "grad_norm": 2.1069821339979864, + "learning_rate": 9.275745664465478e-06, + "loss": 1.0655, + "step": 22810 + }, + { + "epoch": 0.54, + "grad_norm": 2.125636993155369, + "learning_rate": 9.274984629389927e-06, + "loss": 1.1053, + "step": 22811 + }, + { + "epoch": 0.54, + "grad_norm": 2.039342435008656, + "learning_rate": 9.274223598535635e-06, + "loss": 0.9375, + "step": 22812 + }, + { + "epoch": 0.54, + "grad_norm": 1.9014987953888698, + "learning_rate": 9.273462571907049e-06, + "loss": 0.9853, + "step": 22813 + }, + { + "epoch": 0.54, + "grad_norm": 2.187597821860492, + "learning_rate": 9.272701549508589e-06, + "loss": 1.014, + "step": 22814 + }, + { + "epoch": 0.54, + "grad_norm": 2.107430938421917, + "learning_rate": 9.271940531344699e-06, + "loss": 1.0762, + "step": 22815 + }, + { + "epoch": 0.54, + "grad_norm": 1.959322454095961, + "learning_rate": 9.271179517419794e-06, + "loss": 0.9759, + "step": 22816 + }, + { + "epoch": 0.54, + "grad_norm": 1.7469865022278321, + "learning_rate": 9.270418507738316e-06, + "loss": 1.0971, + "step": 22817 + }, + { + "epoch": 0.54, + "grad_norm": 2.881198096163382, + "learning_rate": 9.269657502304694e-06, + "loss": 0.9997, + "step": 22818 + }, + { + "epoch": 0.54, + "grad_norm": 1.7836541877620842, + "learning_rate": 9.268896501123354e-06, + "loss": 0.9499, + "step": 22819 + }, + { + "epoch": 0.54, + "grad_norm": 1.8633327791898102, + "learning_rate": 9.26813550419873e-06, + "loss": 1.0004, + "step": 22820 + }, + { + "epoch": 0.54, + "grad_norm": 1.9538853033331771, + "learning_rate": 9.267374511535254e-06, + "loss": 1.0272, + "step": 22821 + }, + { + "epoch": 0.54, + "grad_norm": 2.3086525901284785, + "learning_rate": 9.266613523137358e-06, + "loss": 1.0676, + "step": 22822 + }, + { + "epoch": 0.54, + "grad_norm": 2.1045626899566257, + "learning_rate": 9.265852539009466e-06, + "loss": 1.0707, + "step": 22823 + }, + { + "epoch": 0.54, + "grad_norm": 1.9028611168095775, + "learning_rate": 9.265091559156014e-06, + "loss": 0.9615, + "step": 22824 + }, + { + "epoch": 0.54, + "grad_norm": 1.1717988368953935, + "learning_rate": 9.264330583581434e-06, + "loss": 0.9471, + "step": 22825 + }, + { + "epoch": 0.54, + "grad_norm": 2.0855969236259275, + "learning_rate": 9.263569612290154e-06, + "loss": 0.9748, + "step": 22826 + }, + { + "epoch": 0.54, + "grad_norm": 1.948289109324689, + "learning_rate": 9.2628086452866e-06, + "loss": 0.9944, + "step": 22827 + }, + { + "epoch": 0.54, + "grad_norm": 1.9243856112462727, + "learning_rate": 9.262047682575211e-06, + "loss": 0.9344, + "step": 22828 + }, + { + "epoch": 0.54, + "grad_norm": 1.923848182433277, + "learning_rate": 9.261286724160417e-06, + "loss": 1.0362, + "step": 22829 + }, + { + "epoch": 0.54, + "grad_norm": 1.212599251808216, + "learning_rate": 9.260525770046642e-06, + "loss": 0.9232, + "step": 22830 + }, + { + "epoch": 0.54, + "grad_norm": 1.867593771187374, + "learning_rate": 9.259764820238318e-06, + "loss": 1.0756, + "step": 22831 + }, + { + "epoch": 0.54, + "grad_norm": 2.182042005028769, + "learning_rate": 9.259003874739885e-06, + "loss": 1.078, + "step": 22832 + }, + { + "epoch": 0.54, + "grad_norm": 2.374489193455897, + "learning_rate": 9.258242933555759e-06, + "loss": 0.9468, + "step": 22833 + }, + { + "epoch": 0.54, + "grad_norm": 1.8434581629175777, + "learning_rate": 9.25748199669038e-06, + "loss": 1.0729, + "step": 22834 + }, + { + "epoch": 0.54, + "grad_norm": 1.1165549295962436, + "learning_rate": 9.256721064148173e-06, + "loss": 0.9264, + "step": 22835 + }, + { + "epoch": 0.54, + "grad_norm": 1.9535215443746674, + "learning_rate": 9.255960135933577e-06, + "loss": 0.9805, + "step": 22836 + }, + { + "epoch": 0.54, + "grad_norm": 1.9190532683079171, + "learning_rate": 9.255199212051012e-06, + "loss": 0.9733, + "step": 22837 + }, + { + "epoch": 0.54, + "grad_norm": 1.8145604637592463, + "learning_rate": 9.254438292504915e-06, + "loss": 1.0364, + "step": 22838 + }, + { + "epoch": 0.54, + "grad_norm": 1.0644144048997881, + "learning_rate": 9.253677377299716e-06, + "loss": 0.9094, + "step": 22839 + }, + { + "epoch": 0.54, + "grad_norm": 2.0662664875925674, + "learning_rate": 9.252916466439839e-06, + "loss": 0.9773, + "step": 22840 + }, + { + "epoch": 0.54, + "grad_norm": 2.0076507929286453, + "learning_rate": 9.252155559929722e-06, + "loss": 0.9966, + "step": 22841 + }, + { + "epoch": 0.54, + "grad_norm": 1.8027917153707755, + "learning_rate": 9.25139465777379e-06, + "loss": 0.9775, + "step": 22842 + }, + { + "epoch": 0.54, + "grad_norm": 2.6432695391778838, + "learning_rate": 9.25063375997648e-06, + "loss": 1.0923, + "step": 22843 + }, + { + "epoch": 0.54, + "grad_norm": 1.8747970814599786, + "learning_rate": 9.249872866542211e-06, + "loss": 1.0045, + "step": 22844 + }, + { + "epoch": 0.54, + "grad_norm": 1.9097031629338603, + "learning_rate": 9.249111977475422e-06, + "loss": 0.9202, + "step": 22845 + }, + { + "epoch": 0.54, + "grad_norm": 1.8412848248258888, + "learning_rate": 9.248351092780544e-06, + "loss": 1.0336, + "step": 22846 + }, + { + "epoch": 0.54, + "grad_norm": 2.33349510275865, + "learning_rate": 9.247590212462001e-06, + "loss": 1.1255, + "step": 22847 + }, + { + "epoch": 0.54, + "grad_norm": 2.653632135499385, + "learning_rate": 9.246829336524225e-06, + "loss": 1.0652, + "step": 22848 + }, + { + "epoch": 0.54, + "grad_norm": 1.1919286851744202, + "learning_rate": 9.246068464971649e-06, + "loss": 1.0041, + "step": 22849 + }, + { + "epoch": 0.54, + "grad_norm": 2.180613450388927, + "learning_rate": 9.245307597808702e-06, + "loss": 0.9057, + "step": 22850 + }, + { + "epoch": 0.54, + "grad_norm": 2.0460748209520143, + "learning_rate": 9.244546735039811e-06, + "loss": 0.9531, + "step": 22851 + }, + { + "epoch": 0.54, + "grad_norm": 2.2448688695286876, + "learning_rate": 9.243785876669407e-06, + "loss": 0.9431, + "step": 22852 + }, + { + "epoch": 0.54, + "grad_norm": 1.075024196848333, + "learning_rate": 9.243025022701926e-06, + "loss": 1.0232, + "step": 22853 + }, + { + "epoch": 0.54, + "grad_norm": 1.069500107769832, + "learning_rate": 9.242264173141789e-06, + "loss": 0.9469, + "step": 22854 + }, + { + "epoch": 0.54, + "grad_norm": 1.933821770981666, + "learning_rate": 9.24150332799343e-06, + "loss": 1.1115, + "step": 22855 + }, + { + "epoch": 0.54, + "grad_norm": 2.0493624909581554, + "learning_rate": 9.240742487261279e-06, + "loss": 1.0192, + "step": 22856 + }, + { + "epoch": 0.54, + "grad_norm": 1.9442247230689922, + "learning_rate": 9.239981650949768e-06, + "loss": 1.0028, + "step": 22857 + }, + { + "epoch": 0.54, + "grad_norm": 1.8688576172263562, + "learning_rate": 9.239220819063322e-06, + "loss": 0.8646, + "step": 22858 + }, + { + "epoch": 0.54, + "grad_norm": 1.8923539432668526, + "learning_rate": 9.238459991606371e-06, + "loss": 1.0203, + "step": 22859 + }, + { + "epoch": 0.54, + "grad_norm": 1.123372901675323, + "learning_rate": 9.237699168583355e-06, + "loss": 0.9927, + "step": 22860 + }, + { + "epoch": 0.54, + "grad_norm": 2.0264927294345063, + "learning_rate": 9.236938349998688e-06, + "loss": 0.9819, + "step": 22861 + }, + { + "epoch": 0.54, + "grad_norm": 2.28785148268878, + "learning_rate": 9.23617753585681e-06, + "loss": 1.0891, + "step": 22862 + }, + { + "epoch": 0.54, + "grad_norm": 2.361788860339434, + "learning_rate": 9.235416726162147e-06, + "loss": 1.0724, + "step": 22863 + }, + { + "epoch": 0.54, + "grad_norm": 1.940829041081245, + "learning_rate": 9.234655920919135e-06, + "loss": 1.0709, + "step": 22864 + }, + { + "epoch": 0.54, + "grad_norm": 1.9969150433304093, + "learning_rate": 9.233895120132192e-06, + "loss": 0.9426, + "step": 22865 + }, + { + "epoch": 0.54, + "grad_norm": 4.227937336488004, + "learning_rate": 9.233134323805757e-06, + "loss": 1.0641, + "step": 22866 + }, + { + "epoch": 0.54, + "grad_norm": 2.141226931962571, + "learning_rate": 9.23237353194426e-06, + "loss": 1.0068, + "step": 22867 + }, + { + "epoch": 0.54, + "grad_norm": 2.2359171241905478, + "learning_rate": 9.231612744552123e-06, + "loss": 0.9697, + "step": 22868 + }, + { + "epoch": 0.54, + "grad_norm": 2.21596372125796, + "learning_rate": 9.230851961633779e-06, + "loss": 0.9832, + "step": 22869 + }, + { + "epoch": 0.54, + "grad_norm": 1.9713855995724565, + "learning_rate": 9.230091183193659e-06, + "loss": 0.8425, + "step": 22870 + }, + { + "epoch": 0.54, + "grad_norm": 1.9825209891154802, + "learning_rate": 9.229330409236196e-06, + "loss": 0.9934, + "step": 22871 + }, + { + "epoch": 0.54, + "grad_norm": 1.994985671687143, + "learning_rate": 9.22856963976581e-06, + "loss": 1.0054, + "step": 22872 + }, + { + "epoch": 0.54, + "grad_norm": 2.3239645530434214, + "learning_rate": 9.227808874786937e-06, + "loss": 1.1257, + "step": 22873 + }, + { + "epoch": 0.54, + "grad_norm": 1.9764930862274142, + "learning_rate": 9.22704811430401e-06, + "loss": 1.011, + "step": 22874 + }, + { + "epoch": 0.54, + "grad_norm": 1.999643805150234, + "learning_rate": 9.22628735832145e-06, + "loss": 1.0541, + "step": 22875 + }, + { + "epoch": 0.54, + "grad_norm": 2.030064100814737, + "learning_rate": 9.225526606843687e-06, + "loss": 1.1139, + "step": 22876 + }, + { + "epoch": 0.54, + "grad_norm": 2.046439457443392, + "learning_rate": 9.224765859875157e-06, + "loss": 1.0966, + "step": 22877 + }, + { + "epoch": 0.54, + "grad_norm": 2.4524116099536717, + "learning_rate": 9.224005117420287e-06, + "loss": 0.9277, + "step": 22878 + }, + { + "epoch": 0.54, + "grad_norm": 2.0766897053682603, + "learning_rate": 9.223244379483502e-06, + "loss": 0.8999, + "step": 22879 + }, + { + "epoch": 0.54, + "grad_norm": 2.4424298002412614, + "learning_rate": 9.222483646069233e-06, + "loss": 1.0295, + "step": 22880 + }, + { + "epoch": 0.54, + "grad_norm": 1.7711712710088976, + "learning_rate": 9.221722917181917e-06, + "loss": 0.9566, + "step": 22881 + }, + { + "epoch": 0.54, + "grad_norm": 2.054003809023552, + "learning_rate": 9.22096219282597e-06, + "loss": 0.9476, + "step": 22882 + }, + { + "epoch": 0.54, + "grad_norm": 2.208335493520345, + "learning_rate": 9.220201473005828e-06, + "loss": 1.0528, + "step": 22883 + }, + { + "epoch": 0.54, + "grad_norm": 1.945918511116093, + "learning_rate": 9.21944075772592e-06, + "loss": 1.0368, + "step": 22884 + }, + { + "epoch": 0.54, + "grad_norm": 2.1342628937133687, + "learning_rate": 9.218680046990678e-06, + "loss": 1.0229, + "step": 22885 + }, + { + "epoch": 0.54, + "grad_norm": 2.2310250366476083, + "learning_rate": 9.217919340804527e-06, + "loss": 1.0812, + "step": 22886 + }, + { + "epoch": 0.54, + "grad_norm": 1.0905526288670038, + "learning_rate": 9.217158639171898e-06, + "loss": 0.9203, + "step": 22887 + }, + { + "epoch": 0.54, + "grad_norm": 2.274130227177106, + "learning_rate": 9.216397942097219e-06, + "loss": 0.9789, + "step": 22888 + }, + { + "epoch": 0.54, + "grad_norm": 2.0721537134429955, + "learning_rate": 9.215637249584915e-06, + "loss": 0.944, + "step": 22889 + }, + { + "epoch": 0.54, + "grad_norm": 2.0176192907904866, + "learning_rate": 9.214876561639424e-06, + "loss": 1.0913, + "step": 22890 + }, + { + "epoch": 0.54, + "grad_norm": 1.9879394511011388, + "learning_rate": 9.214115878265168e-06, + "loss": 0.942, + "step": 22891 + }, + { + "epoch": 0.54, + "grad_norm": 2.1363209606077724, + "learning_rate": 9.213355199466582e-06, + "loss": 0.9, + "step": 22892 + }, + { + "epoch": 0.54, + "grad_norm": 2.231013949476551, + "learning_rate": 9.212594525248086e-06, + "loss": 1.041, + "step": 22893 + }, + { + "epoch": 0.54, + "grad_norm": 2.070145513102494, + "learning_rate": 9.211833855614115e-06, + "loss": 0.9614, + "step": 22894 + }, + { + "epoch": 0.54, + "grad_norm": 2.315391238238054, + "learning_rate": 9.2110731905691e-06, + "loss": 0.9972, + "step": 22895 + }, + { + "epoch": 0.54, + "grad_norm": 2.03080438244382, + "learning_rate": 9.210312530117464e-06, + "loss": 1.0135, + "step": 22896 + }, + { + "epoch": 0.54, + "grad_norm": 1.9563307363815965, + "learning_rate": 9.209551874263637e-06, + "loss": 0.9468, + "step": 22897 + }, + { + "epoch": 0.54, + "grad_norm": 1.9662017972654522, + "learning_rate": 9.208791223012051e-06, + "loss": 0.98, + "step": 22898 + }, + { + "epoch": 0.54, + "grad_norm": 1.8246472749324387, + "learning_rate": 9.208030576367136e-06, + "loss": 1.0082, + "step": 22899 + }, + { + "epoch": 0.54, + "grad_norm": 1.90513131120168, + "learning_rate": 9.207269934333315e-06, + "loss": 0.968, + "step": 22900 + }, + { + "epoch": 0.54, + "grad_norm": 2.0143562899953253, + "learning_rate": 9.206509296915016e-06, + "loss": 1.0053, + "step": 22901 + }, + { + "epoch": 0.54, + "grad_norm": 2.0910744998064774, + "learning_rate": 9.205748664116677e-06, + "loss": 0.9056, + "step": 22902 + }, + { + "epoch": 0.54, + "grad_norm": 1.1557382849979623, + "learning_rate": 9.20498803594272e-06, + "loss": 1.0266, + "step": 22903 + }, + { + "epoch": 0.54, + "grad_norm": 2.03035499691964, + "learning_rate": 9.20422741239757e-06, + "loss": 0.9438, + "step": 22904 + }, + { + "epoch": 0.54, + "grad_norm": 2.008792058782981, + "learning_rate": 9.203466793485662e-06, + "loss": 1.0165, + "step": 22905 + }, + { + "epoch": 0.54, + "grad_norm": 1.0605070028045884, + "learning_rate": 9.202706179211425e-06, + "loss": 0.987, + "step": 22906 + }, + { + "epoch": 0.54, + "grad_norm": 2.162915751821255, + "learning_rate": 9.201945569579283e-06, + "loss": 0.9939, + "step": 22907 + }, + { + "epoch": 0.54, + "grad_norm": 2.156131828178022, + "learning_rate": 9.201184964593664e-06, + "loss": 0.941, + "step": 22908 + }, + { + "epoch": 0.54, + "grad_norm": 2.365534774014027, + "learning_rate": 9.200424364259006e-06, + "loss": 1.0205, + "step": 22909 + }, + { + "epoch": 0.54, + "grad_norm": 2.035810617356603, + "learning_rate": 9.199663768579724e-06, + "loss": 0.9789, + "step": 22910 + }, + { + "epoch": 0.54, + "grad_norm": 1.8918948754794835, + "learning_rate": 9.198903177560255e-06, + "loss": 0.9396, + "step": 22911 + }, + { + "epoch": 0.54, + "grad_norm": 1.9267000531927556, + "learning_rate": 9.198142591205022e-06, + "loss": 1.0257, + "step": 22912 + }, + { + "epoch": 0.54, + "grad_norm": 2.0856250965321044, + "learning_rate": 9.197382009518465e-06, + "loss": 0.9882, + "step": 22913 + }, + { + "epoch": 0.54, + "grad_norm": 1.9112185561068245, + "learning_rate": 9.196621432504996e-06, + "loss": 1.1315, + "step": 22914 + }, + { + "epoch": 0.54, + "grad_norm": 1.0614553914900524, + "learning_rate": 9.195860860169054e-06, + "loss": 0.9551, + "step": 22915 + }, + { + "epoch": 0.54, + "grad_norm": 1.9317276626953832, + "learning_rate": 9.195100292515064e-06, + "loss": 0.958, + "step": 22916 + }, + { + "epoch": 0.54, + "grad_norm": 1.7992666392903762, + "learning_rate": 9.194339729547461e-06, + "loss": 0.8908, + "step": 22917 + }, + { + "epoch": 0.54, + "grad_norm": 1.991614332328373, + "learning_rate": 9.193579171270659e-06, + "loss": 1.1531, + "step": 22918 + }, + { + "epoch": 0.54, + "grad_norm": 2.0067177488279158, + "learning_rate": 9.192818617689097e-06, + "loss": 1.0414, + "step": 22919 + }, + { + "epoch": 0.54, + "grad_norm": 2.3693909379189866, + "learning_rate": 9.192058068807203e-06, + "loss": 0.9982, + "step": 22920 + }, + { + "epoch": 0.54, + "grad_norm": 2.0841083353999603, + "learning_rate": 9.191297524629398e-06, + "loss": 0.9959, + "step": 22921 + }, + { + "epoch": 0.54, + "grad_norm": 2.0814646297427197, + "learning_rate": 9.190536985160119e-06, + "loss": 0.987, + "step": 22922 + }, + { + "epoch": 0.54, + "grad_norm": 1.9985442450275006, + "learning_rate": 9.189776450403789e-06, + "loss": 0.9926, + "step": 22923 + }, + { + "epoch": 0.54, + "grad_norm": 2.0068009162315934, + "learning_rate": 9.189015920364839e-06, + "loss": 0.8215, + "step": 22924 + }, + { + "epoch": 0.54, + "grad_norm": 1.7660368199364398, + "learning_rate": 9.18825539504769e-06, + "loss": 0.8905, + "step": 22925 + }, + { + "epoch": 0.54, + "grad_norm": 2.2080603140511537, + "learning_rate": 9.187494874456777e-06, + "loss": 0.9306, + "step": 22926 + }, + { + "epoch": 0.54, + "grad_norm": 1.146907667127132, + "learning_rate": 9.186734358596529e-06, + "loss": 0.8376, + "step": 22927 + }, + { + "epoch": 0.54, + "grad_norm": 2.1625186842656623, + "learning_rate": 9.185973847471369e-06, + "loss": 1.167, + "step": 22928 + }, + { + "epoch": 0.54, + "grad_norm": 2.093878316305636, + "learning_rate": 9.185213341085725e-06, + "loss": 1.0513, + "step": 22929 + }, + { + "epoch": 0.54, + "grad_norm": 2.138581910810429, + "learning_rate": 9.184452839444029e-06, + "loss": 1.0737, + "step": 22930 + }, + { + "epoch": 0.54, + "grad_norm": 1.9015661015203162, + "learning_rate": 9.183692342550708e-06, + "loss": 0.9658, + "step": 22931 + }, + { + "epoch": 0.54, + "grad_norm": 2.1872666197200323, + "learning_rate": 9.182931850410187e-06, + "loss": 0.8747, + "step": 22932 + }, + { + "epoch": 0.54, + "grad_norm": 2.073779509305601, + "learning_rate": 9.182171363026894e-06, + "loss": 1.074, + "step": 22933 + }, + { + "epoch": 0.54, + "grad_norm": 1.9170104858867933, + "learning_rate": 9.181410880405262e-06, + "loss": 0.9713, + "step": 22934 + }, + { + "epoch": 0.54, + "grad_norm": 1.9765150741590647, + "learning_rate": 9.180650402549713e-06, + "loss": 1.0231, + "step": 22935 + }, + { + "epoch": 0.54, + "grad_norm": 2.273968471853753, + "learning_rate": 9.179889929464676e-06, + "loss": 1.098, + "step": 22936 + }, + { + "epoch": 0.54, + "grad_norm": 1.1542909419525307, + "learning_rate": 9.179129461154578e-06, + "loss": 0.9757, + "step": 22937 + }, + { + "epoch": 0.54, + "grad_norm": 2.0663678882522833, + "learning_rate": 9.178368997623853e-06, + "loss": 0.9727, + "step": 22938 + }, + { + "epoch": 0.54, + "grad_norm": 1.8588702481930208, + "learning_rate": 9.17760853887692e-06, + "loss": 0.9886, + "step": 22939 + }, + { + "epoch": 0.54, + "grad_norm": 1.9479848328412455, + "learning_rate": 9.176848084918212e-06, + "loss": 1.0005, + "step": 22940 + }, + { + "epoch": 0.54, + "grad_norm": 2.935867697255588, + "learning_rate": 9.176087635752157e-06, + "loss": 1.0139, + "step": 22941 + }, + { + "epoch": 0.54, + "grad_norm": 1.0977182029315165, + "learning_rate": 9.175327191383175e-06, + "loss": 0.8973, + "step": 22942 + }, + { + "epoch": 0.54, + "grad_norm": 1.8528651339885773, + "learning_rate": 9.174566751815704e-06, + "loss": 0.9701, + "step": 22943 + }, + { + "epoch": 0.54, + "grad_norm": 2.1901037802404835, + "learning_rate": 9.173806317054161e-06, + "loss": 1.0735, + "step": 22944 + }, + { + "epoch": 0.54, + "grad_norm": 2.074966477623677, + "learning_rate": 9.173045887102988e-06, + "loss": 0.941, + "step": 22945 + }, + { + "epoch": 0.54, + "grad_norm": 2.0741428382231595, + "learning_rate": 9.172285461966597e-06, + "loss": 0.9891, + "step": 22946 + }, + { + "epoch": 0.54, + "grad_norm": 2.0622948691064567, + "learning_rate": 9.171525041649423e-06, + "loss": 1.0481, + "step": 22947 + }, + { + "epoch": 0.54, + "grad_norm": 1.8268698064965656, + "learning_rate": 9.170764626155896e-06, + "loss": 0.9156, + "step": 22948 + }, + { + "epoch": 0.54, + "grad_norm": 2.0042122124890738, + "learning_rate": 9.170004215490438e-06, + "loss": 1.1111, + "step": 22949 + }, + { + "epoch": 0.54, + "grad_norm": 2.2112532682295947, + "learning_rate": 9.169243809657476e-06, + "loss": 0.9922, + "step": 22950 + }, + { + "epoch": 0.54, + "grad_norm": 2.4444898935207653, + "learning_rate": 9.16848340866144e-06, + "loss": 0.8886, + "step": 22951 + }, + { + "epoch": 0.54, + "grad_norm": 1.8990652166069062, + "learning_rate": 9.16772301250676e-06, + "loss": 1.11, + "step": 22952 + }, + { + "epoch": 0.54, + "grad_norm": 2.2026434626927407, + "learning_rate": 9.16696262119786e-06, + "loss": 1.0116, + "step": 22953 + }, + { + "epoch": 0.54, + "grad_norm": 1.996216608305212, + "learning_rate": 9.166202234739161e-06, + "loss": 1.0304, + "step": 22954 + }, + { + "epoch": 0.54, + "grad_norm": 1.718387741823576, + "learning_rate": 9.165441853135104e-06, + "loss": 1.0235, + "step": 22955 + }, + { + "epoch": 0.54, + "grad_norm": 2.091089827388594, + "learning_rate": 9.164681476390107e-06, + "loss": 1.0586, + "step": 22956 + }, + { + "epoch": 0.54, + "grad_norm": 1.919978592135281, + "learning_rate": 9.163921104508593e-06, + "loss": 0.9641, + "step": 22957 + }, + { + "epoch": 0.54, + "grad_norm": 1.9145928899103222, + "learning_rate": 9.163160737495e-06, + "loss": 1.0403, + "step": 22958 + }, + { + "epoch": 0.54, + "grad_norm": 2.0462636973230857, + "learning_rate": 9.162400375353752e-06, + "loss": 1.0213, + "step": 22959 + }, + { + "epoch": 0.54, + "grad_norm": 2.146533087766872, + "learning_rate": 9.161640018089273e-06, + "loss": 0.9653, + "step": 22960 + }, + { + "epoch": 0.54, + "grad_norm": 1.8761757557981908, + "learning_rate": 9.160879665705988e-06, + "loss": 1.1268, + "step": 22961 + }, + { + "epoch": 0.54, + "grad_norm": 2.0202859127053596, + "learning_rate": 9.160119318208334e-06, + "loss": 1.0057, + "step": 22962 + }, + { + "epoch": 0.54, + "grad_norm": 2.243338320147342, + "learning_rate": 9.159358975600723e-06, + "loss": 0.9315, + "step": 22963 + }, + { + "epoch": 0.54, + "grad_norm": 2.114832100388719, + "learning_rate": 9.158598637887594e-06, + "loss": 0.9866, + "step": 22964 + }, + { + "epoch": 0.54, + "grad_norm": 2.0119757755442333, + "learning_rate": 9.15783830507337e-06, + "loss": 1.0513, + "step": 22965 + }, + { + "epoch": 0.54, + "grad_norm": 1.9263931628175524, + "learning_rate": 9.157077977162481e-06, + "loss": 0.9309, + "step": 22966 + }, + { + "epoch": 0.54, + "grad_norm": 2.078309560948164, + "learning_rate": 9.156317654159346e-06, + "loss": 1.0182, + "step": 22967 + }, + { + "epoch": 0.54, + "grad_norm": 1.8481099673507368, + "learning_rate": 9.155557336068399e-06, + "loss": 0.9519, + "step": 22968 + }, + { + "epoch": 0.54, + "grad_norm": 2.2849265729990744, + "learning_rate": 9.154797022894067e-06, + "loss": 1.0169, + "step": 22969 + }, + { + "epoch": 0.54, + "grad_norm": 1.030618424531967, + "learning_rate": 9.154036714640768e-06, + "loss": 1.0307, + "step": 22970 + }, + { + "epoch": 0.54, + "grad_norm": 2.513395695716045, + "learning_rate": 9.15327641131294e-06, + "loss": 0.922, + "step": 22971 + }, + { + "epoch": 0.54, + "grad_norm": 2.1285043127565646, + "learning_rate": 9.152516112915002e-06, + "loss": 1.0469, + "step": 22972 + }, + { + "epoch": 0.54, + "grad_norm": 1.8652560874414208, + "learning_rate": 9.151755819451387e-06, + "loss": 0.9617, + "step": 22973 + }, + { + "epoch": 0.54, + "grad_norm": 1.975218322425241, + "learning_rate": 9.150995530926515e-06, + "loss": 0.9558, + "step": 22974 + }, + { + "epoch": 0.54, + "grad_norm": 1.965907818580117, + "learning_rate": 9.150235247344817e-06, + "loss": 0.9683, + "step": 22975 + }, + { + "epoch": 0.54, + "grad_norm": 1.985340527078344, + "learning_rate": 9.14947496871072e-06, + "loss": 0.8936, + "step": 22976 + }, + { + "epoch": 0.54, + "grad_norm": 2.1736350687111075, + "learning_rate": 9.148714695028647e-06, + "loss": 1.0065, + "step": 22977 + }, + { + "epoch": 0.54, + "grad_norm": 2.06188598956797, + "learning_rate": 9.147954426303024e-06, + "loss": 0.8991, + "step": 22978 + }, + { + "epoch": 0.54, + "grad_norm": 1.085958423816004, + "learning_rate": 9.147194162538284e-06, + "loss": 0.9724, + "step": 22979 + }, + { + "epoch": 0.54, + "grad_norm": 2.7536977251050963, + "learning_rate": 9.14643390373885e-06, + "loss": 0.9399, + "step": 22980 + }, + { + "epoch": 0.54, + "grad_norm": 2.0527077272973973, + "learning_rate": 9.145673649909147e-06, + "loss": 1.0904, + "step": 22981 + }, + { + "epoch": 0.54, + "grad_norm": 1.1071458262016816, + "learning_rate": 9.144913401053599e-06, + "loss": 0.9163, + "step": 22982 + }, + { + "epoch": 0.54, + "grad_norm": 1.9482440003169141, + "learning_rate": 9.144153157176641e-06, + "loss": 1.1247, + "step": 22983 + }, + { + "epoch": 0.54, + "grad_norm": 2.0733895388710706, + "learning_rate": 9.143392918282693e-06, + "loss": 0.9399, + "step": 22984 + }, + { + "epoch": 0.54, + "grad_norm": 1.113766693323175, + "learning_rate": 9.14263268437618e-06, + "loss": 0.9738, + "step": 22985 + }, + { + "epoch": 0.54, + "grad_norm": 1.8638594474270007, + "learning_rate": 9.141872455461532e-06, + "loss": 0.9404, + "step": 22986 + }, + { + "epoch": 0.54, + "grad_norm": 2.047679194075683, + "learning_rate": 9.141112231543177e-06, + "loss": 1.1737, + "step": 22987 + }, + { + "epoch": 0.54, + "grad_norm": 2.573312562092378, + "learning_rate": 9.140352012625538e-06, + "loss": 1.0915, + "step": 22988 + }, + { + "epoch": 0.54, + "grad_norm": 2.110829468793607, + "learning_rate": 9.139591798713038e-06, + "loss": 1.0172, + "step": 22989 + }, + { + "epoch": 0.54, + "grad_norm": 1.9236803364143529, + "learning_rate": 9.138831589810111e-06, + "loss": 1.0353, + "step": 22990 + }, + { + "epoch": 0.54, + "grad_norm": 1.0399962252986963, + "learning_rate": 9.138071385921176e-06, + "loss": 0.9589, + "step": 22991 + }, + { + "epoch": 0.54, + "grad_norm": 1.8313672144252344, + "learning_rate": 9.137311187050663e-06, + "loss": 1.0942, + "step": 22992 + }, + { + "epoch": 0.54, + "grad_norm": 1.9960983002437231, + "learning_rate": 9.136550993202997e-06, + "loss": 0.9624, + "step": 22993 + }, + { + "epoch": 0.54, + "grad_norm": 1.1424667380520541, + "learning_rate": 9.13579080438261e-06, + "loss": 0.9, + "step": 22994 + }, + { + "epoch": 0.54, + "grad_norm": 2.144575546053409, + "learning_rate": 9.135030620593915e-06, + "loss": 0.9145, + "step": 22995 + }, + { + "epoch": 0.54, + "grad_norm": 2.0059952455411665, + "learning_rate": 9.13427044184135e-06, + "loss": 1.1007, + "step": 22996 + }, + { + "epoch": 0.54, + "grad_norm": 1.9614065462308363, + "learning_rate": 9.133510268129337e-06, + "loss": 0.9204, + "step": 22997 + }, + { + "epoch": 0.54, + "grad_norm": 1.0670669390172662, + "learning_rate": 9.1327500994623e-06, + "loss": 0.8817, + "step": 22998 + }, + { + "epoch": 0.54, + "grad_norm": 2.5748243980670393, + "learning_rate": 9.131989935844663e-06, + "loss": 1.0341, + "step": 22999 + }, + { + "epoch": 0.54, + "grad_norm": 1.832760291517562, + "learning_rate": 9.131229777280859e-06, + "loss": 0.7898, + "step": 23000 + }, + { + "epoch": 0.54, + "grad_norm": 2.004270244350656, + "learning_rate": 9.130469623775312e-06, + "loss": 1.0166, + "step": 23001 + }, + { + "epoch": 0.54, + "grad_norm": 2.0408750984192, + "learning_rate": 9.129709475332445e-06, + "loss": 0.9455, + "step": 23002 + }, + { + "epoch": 0.54, + "grad_norm": 1.1191274203608297, + "learning_rate": 9.12894933195668e-06, + "loss": 0.9258, + "step": 23003 + }, + { + "epoch": 0.54, + "grad_norm": 3.9850547234106997, + "learning_rate": 9.128189193652455e-06, + "loss": 0.9903, + "step": 23004 + }, + { + "epoch": 0.54, + "grad_norm": 1.9739273045994041, + "learning_rate": 9.127429060424184e-06, + "loss": 0.9194, + "step": 23005 + }, + { + "epoch": 0.54, + "grad_norm": 1.8094567163131678, + "learning_rate": 9.126668932276298e-06, + "loss": 1.0367, + "step": 23006 + }, + { + "epoch": 0.54, + "grad_norm": 1.0606756897188476, + "learning_rate": 9.125908809213221e-06, + "loss": 0.9148, + "step": 23007 + }, + { + "epoch": 0.54, + "grad_norm": 1.0515194590037076, + "learning_rate": 9.125148691239383e-06, + "loss": 0.9145, + "step": 23008 + }, + { + "epoch": 0.54, + "grad_norm": 2.269590099129003, + "learning_rate": 9.124388578359206e-06, + "loss": 1.0866, + "step": 23009 + }, + { + "epoch": 0.54, + "grad_norm": 1.9248202400459804, + "learning_rate": 9.123628470577111e-06, + "loss": 1.0686, + "step": 23010 + }, + { + "epoch": 0.54, + "grad_norm": 2.4023334342068265, + "learning_rate": 9.122868367897537e-06, + "loss": 0.9644, + "step": 23011 + }, + { + "epoch": 0.54, + "grad_norm": 2.390777602790835, + "learning_rate": 9.122108270324894e-06, + "loss": 0.8597, + "step": 23012 + }, + { + "epoch": 0.54, + "grad_norm": 2.3000315795225768, + "learning_rate": 9.121348177863615e-06, + "loss": 0.9763, + "step": 23013 + }, + { + "epoch": 0.54, + "grad_norm": 2.6572881766469196, + "learning_rate": 9.120588090518126e-06, + "loss": 1.0502, + "step": 23014 + }, + { + "epoch": 0.54, + "grad_norm": 2.499252395866974, + "learning_rate": 9.119828008292856e-06, + "loss": 0.9091, + "step": 23015 + }, + { + "epoch": 0.54, + "grad_norm": 2.009488143500641, + "learning_rate": 9.11906793119222e-06, + "loss": 1.0328, + "step": 23016 + }, + { + "epoch": 0.54, + "grad_norm": 2.265570454105292, + "learning_rate": 9.118307859220651e-06, + "loss": 1.1275, + "step": 23017 + }, + { + "epoch": 0.54, + "grad_norm": 1.057642612785683, + "learning_rate": 9.117547792382577e-06, + "loss": 0.9668, + "step": 23018 + }, + { + "epoch": 0.54, + "grad_norm": 3.719942337699589, + "learning_rate": 9.116787730682413e-06, + "loss": 1.0755, + "step": 23019 + }, + { + "epoch": 0.54, + "grad_norm": 2.261891657288708, + "learning_rate": 9.116027674124592e-06, + "loss": 1.0348, + "step": 23020 + }, + { + "epoch": 0.54, + "grad_norm": 2.2777287198114102, + "learning_rate": 9.115267622713539e-06, + "loss": 0.8002, + "step": 23021 + }, + { + "epoch": 0.54, + "grad_norm": 2.1029477551662765, + "learning_rate": 9.11450757645368e-06, + "loss": 1.1114, + "step": 23022 + }, + { + "epoch": 0.54, + "grad_norm": 2.0305712383087835, + "learning_rate": 9.113747535349433e-06, + "loss": 1.0471, + "step": 23023 + }, + { + "epoch": 0.54, + "grad_norm": 1.836667373206193, + "learning_rate": 9.112987499405232e-06, + "loss": 0.9441, + "step": 23024 + }, + { + "epoch": 0.54, + "grad_norm": 2.0713737517466213, + "learning_rate": 9.112227468625499e-06, + "loss": 0.9424, + "step": 23025 + }, + { + "epoch": 0.54, + "grad_norm": 1.8435357973822497, + "learning_rate": 9.11146744301466e-06, + "loss": 0.9317, + "step": 23026 + }, + { + "epoch": 0.54, + "grad_norm": 1.909415346001682, + "learning_rate": 9.110707422577133e-06, + "loss": 0.9452, + "step": 23027 + }, + { + "epoch": 0.54, + "grad_norm": 1.8555829828635229, + "learning_rate": 9.109947407317352e-06, + "loss": 0.9988, + "step": 23028 + }, + { + "epoch": 0.54, + "grad_norm": 1.9135595310384355, + "learning_rate": 9.109187397239742e-06, + "loss": 1.0123, + "step": 23029 + }, + { + "epoch": 0.54, + "grad_norm": 1.0789349111191542, + "learning_rate": 9.108427392348722e-06, + "loss": 1.0268, + "step": 23030 + }, + { + "epoch": 0.54, + "grad_norm": 2.3047984890560045, + "learning_rate": 9.107667392648718e-06, + "loss": 1.1523, + "step": 23031 + }, + { + "epoch": 0.54, + "grad_norm": 1.0533001330276646, + "learning_rate": 9.10690739814416e-06, + "loss": 0.974, + "step": 23032 + }, + { + "epoch": 0.54, + "grad_norm": 1.9328662286941793, + "learning_rate": 9.10614740883947e-06, + "loss": 1.0198, + "step": 23033 + }, + { + "epoch": 0.54, + "grad_norm": 2.0010449651222104, + "learning_rate": 9.105387424739073e-06, + "loss": 1.0286, + "step": 23034 + }, + { + "epoch": 0.54, + "grad_norm": 2.0154246394244555, + "learning_rate": 9.10462744584739e-06, + "loss": 0.9737, + "step": 23035 + }, + { + "epoch": 0.54, + "grad_norm": 1.888875898322483, + "learning_rate": 9.103867472168854e-06, + "loss": 0.9358, + "step": 23036 + }, + { + "epoch": 0.54, + "grad_norm": 2.286817781229498, + "learning_rate": 9.103107503707885e-06, + "loss": 1.0046, + "step": 23037 + }, + { + "epoch": 0.54, + "grad_norm": 1.9998838707151605, + "learning_rate": 9.102347540468906e-06, + "loss": 1.0696, + "step": 23038 + }, + { + "epoch": 0.54, + "grad_norm": 1.0657427126457382, + "learning_rate": 9.101587582456342e-06, + "loss": 1.0315, + "step": 23039 + }, + { + "epoch": 0.54, + "grad_norm": 1.8343281481479852, + "learning_rate": 9.100827629674623e-06, + "loss": 1.1301, + "step": 23040 + }, + { + "epoch": 0.54, + "grad_norm": 2.0979206979186693, + "learning_rate": 9.10006768212817e-06, + "loss": 1.0866, + "step": 23041 + }, + { + "epoch": 0.54, + "grad_norm": 2.202226547116235, + "learning_rate": 9.099307739821404e-06, + "loss": 0.8653, + "step": 23042 + }, + { + "epoch": 0.54, + "grad_norm": 1.8068044860735577, + "learning_rate": 9.098547802758762e-06, + "loss": 0.9013, + "step": 23043 + }, + { + "epoch": 0.54, + "grad_norm": 1.9521823921728543, + "learning_rate": 9.09778787094465e-06, + "loss": 0.9748, + "step": 23044 + }, + { + "epoch": 0.54, + "grad_norm": 2.0172848749744583, + "learning_rate": 9.097027944383508e-06, + "loss": 1.0142, + "step": 23045 + }, + { + "epoch": 0.54, + "grad_norm": 2.498634978162754, + "learning_rate": 9.096268023079753e-06, + "loss": 1.0379, + "step": 23046 + }, + { + "epoch": 0.54, + "grad_norm": 2.0398165021949892, + "learning_rate": 9.095508107037817e-06, + "loss": 1.0281, + "step": 23047 + }, + { + "epoch": 0.54, + "grad_norm": 2.0990673671909192, + "learning_rate": 9.094748196262114e-06, + "loss": 1.0757, + "step": 23048 + }, + { + "epoch": 0.54, + "grad_norm": 2.0426596599528475, + "learning_rate": 9.093988290757075e-06, + "loss": 0.9446, + "step": 23049 + }, + { + "epoch": 0.54, + "grad_norm": 1.8928954078214064, + "learning_rate": 9.093228390527123e-06, + "loss": 0.9636, + "step": 23050 + }, + { + "epoch": 0.54, + "grad_norm": 1.8621554838771297, + "learning_rate": 9.092468495576682e-06, + "loss": 0.954, + "step": 23051 + }, + { + "epoch": 0.54, + "grad_norm": 2.2756816799630664, + "learning_rate": 9.091708605910175e-06, + "loss": 1.0191, + "step": 23052 + }, + { + "epoch": 0.54, + "grad_norm": 2.207257621895301, + "learning_rate": 9.09094872153203e-06, + "loss": 1.07, + "step": 23053 + }, + { + "epoch": 0.54, + "grad_norm": 1.9612643923563402, + "learning_rate": 9.090188842446672e-06, + "loss": 1.0962, + "step": 23054 + }, + { + "epoch": 0.54, + "grad_norm": 2.879611546403619, + "learning_rate": 9.089428968658518e-06, + "loss": 1.0324, + "step": 23055 + }, + { + "epoch": 0.54, + "grad_norm": 2.4418499450907625, + "learning_rate": 9.088669100171999e-06, + "loss": 0.9727, + "step": 23056 + }, + { + "epoch": 0.54, + "grad_norm": 1.799162884791969, + "learning_rate": 9.08790923699154e-06, + "loss": 1.0365, + "step": 23057 + }, + { + "epoch": 0.54, + "grad_norm": 2.0665651374803238, + "learning_rate": 9.087149379121556e-06, + "loss": 1.0573, + "step": 23058 + }, + { + "epoch": 0.54, + "grad_norm": 2.016756210562765, + "learning_rate": 9.08638952656648e-06, + "loss": 0.8988, + "step": 23059 + }, + { + "epoch": 0.54, + "grad_norm": 1.9565181805578133, + "learning_rate": 9.085629679330733e-06, + "loss": 0.9541, + "step": 23060 + }, + { + "epoch": 0.54, + "grad_norm": 2.207123367636308, + "learning_rate": 9.084869837418742e-06, + "loss": 0.9678, + "step": 23061 + }, + { + "epoch": 0.54, + "grad_norm": 1.1012941500303675, + "learning_rate": 9.084110000834928e-06, + "loss": 1.0003, + "step": 23062 + }, + { + "epoch": 0.54, + "grad_norm": 1.9391216528317017, + "learning_rate": 9.083350169583711e-06, + "loss": 0.9038, + "step": 23063 + }, + { + "epoch": 0.54, + "grad_norm": 2.4276302087813844, + "learning_rate": 9.082590343669526e-06, + "loss": 1.0127, + "step": 23064 + }, + { + "epoch": 0.54, + "grad_norm": 1.9904094044442944, + "learning_rate": 9.081830523096786e-06, + "loss": 0.8447, + "step": 23065 + }, + { + "epoch": 0.54, + "grad_norm": 1.8650180271762906, + "learning_rate": 9.081070707869921e-06, + "loss": 1.063, + "step": 23066 + }, + { + "epoch": 0.54, + "grad_norm": 2.157256086159546, + "learning_rate": 9.08031089799335e-06, + "loss": 1.163, + "step": 23067 + }, + { + "epoch": 0.54, + "grad_norm": 1.9763818152797987, + "learning_rate": 9.079551093471507e-06, + "loss": 0.9828, + "step": 23068 + }, + { + "epoch": 0.54, + "grad_norm": 2.509806320123966, + "learning_rate": 9.078791294308803e-06, + "loss": 0.9355, + "step": 23069 + }, + { + "epoch": 0.54, + "grad_norm": 2.2097618240474133, + "learning_rate": 9.07803150050967e-06, + "loss": 0.8558, + "step": 23070 + }, + { + "epoch": 0.54, + "grad_norm": 1.8187228835915572, + "learning_rate": 9.077271712078531e-06, + "loss": 0.9308, + "step": 23071 + }, + { + "epoch": 0.54, + "grad_norm": 1.1184667414992542, + "learning_rate": 9.076511929019804e-06, + "loss": 0.9997, + "step": 23072 + }, + { + "epoch": 0.54, + "grad_norm": 2.186598241087856, + "learning_rate": 9.07575215133792e-06, + "loss": 1.0141, + "step": 23073 + }, + { + "epoch": 0.54, + "grad_norm": 2.276477123937753, + "learning_rate": 9.074992379037298e-06, + "loss": 1.037, + "step": 23074 + }, + { + "epoch": 0.54, + "grad_norm": 6.864205004057083, + "learning_rate": 9.074232612122367e-06, + "loss": 0.9366, + "step": 23075 + }, + { + "epoch": 0.54, + "grad_norm": 1.6708113941276075, + "learning_rate": 9.073472850597542e-06, + "loss": 0.9597, + "step": 23076 + }, + { + "epoch": 0.54, + "grad_norm": 1.8175876348020426, + "learning_rate": 9.072713094467253e-06, + "loss": 1.0076, + "step": 23077 + }, + { + "epoch": 0.54, + "grad_norm": 1.8493092043103332, + "learning_rate": 9.071953343735925e-06, + "loss": 1.0881, + "step": 23078 + }, + { + "epoch": 0.54, + "grad_norm": 1.9920781173316036, + "learning_rate": 9.071193598407975e-06, + "loss": 0.9521, + "step": 23079 + }, + { + "epoch": 0.54, + "grad_norm": 2.1402690952257184, + "learning_rate": 9.070433858487829e-06, + "loss": 1.0264, + "step": 23080 + }, + { + "epoch": 0.54, + "grad_norm": 1.978970779025913, + "learning_rate": 9.069674123979912e-06, + "loss": 1.0051, + "step": 23081 + }, + { + "epoch": 0.54, + "grad_norm": 1.0969228305128644, + "learning_rate": 9.068914394888651e-06, + "loss": 0.9124, + "step": 23082 + }, + { + "epoch": 0.54, + "grad_norm": 2.492852913188761, + "learning_rate": 9.068154671218462e-06, + "loss": 0.9408, + "step": 23083 + }, + { + "epoch": 0.54, + "grad_norm": 2.152273031118354, + "learning_rate": 9.06739495297377e-06, + "loss": 1.0547, + "step": 23084 + }, + { + "epoch": 0.54, + "grad_norm": 2.015255413217806, + "learning_rate": 9.066635240159002e-06, + "loss": 0.9732, + "step": 23085 + }, + { + "epoch": 0.54, + "grad_norm": 2.0446806751059268, + "learning_rate": 9.065875532778578e-06, + "loss": 1.0872, + "step": 23086 + }, + { + "epoch": 0.54, + "grad_norm": 1.9981519228904467, + "learning_rate": 9.065115830836923e-06, + "loss": 1.1127, + "step": 23087 + }, + { + "epoch": 0.54, + "grad_norm": 2.197767165853186, + "learning_rate": 9.064356134338459e-06, + "loss": 0.8908, + "step": 23088 + }, + { + "epoch": 0.54, + "grad_norm": 1.9569345258677902, + "learning_rate": 9.063596443287612e-06, + "loss": 0.9456, + "step": 23089 + }, + { + "epoch": 0.54, + "grad_norm": 1.0634367499023423, + "learning_rate": 9.0628367576888e-06, + "loss": 0.9138, + "step": 23090 + }, + { + "epoch": 0.54, + "grad_norm": 2.20032490828202, + "learning_rate": 9.062077077546449e-06, + "loss": 1.0487, + "step": 23091 + }, + { + "epoch": 0.54, + "grad_norm": 2.2934172477422807, + "learning_rate": 9.061317402864988e-06, + "loss": 0.9618, + "step": 23092 + }, + { + "epoch": 0.54, + "grad_norm": 1.9110314992181254, + "learning_rate": 9.060557733648827e-06, + "loss": 0.9055, + "step": 23093 + }, + { + "epoch": 0.54, + "grad_norm": 1.9623668733415682, + "learning_rate": 9.059798069902399e-06, + "loss": 1.027, + "step": 23094 + }, + { + "epoch": 0.54, + "grad_norm": 1.8032837261381354, + "learning_rate": 9.059038411630123e-06, + "loss": 0.9735, + "step": 23095 + }, + { + "epoch": 0.54, + "grad_norm": 2.0050523572901366, + "learning_rate": 9.058278758836428e-06, + "loss": 0.9909, + "step": 23096 + }, + { + "epoch": 0.54, + "grad_norm": 1.9342311134992136, + "learning_rate": 9.057519111525728e-06, + "loss": 0.9806, + "step": 23097 + }, + { + "epoch": 0.54, + "grad_norm": 1.9647034309158313, + "learning_rate": 9.05675946970245e-06, + "loss": 1.0053, + "step": 23098 + }, + { + "epoch": 0.54, + "grad_norm": 2.079036602333153, + "learning_rate": 9.05599983337102e-06, + "loss": 1.0803, + "step": 23099 + }, + { + "epoch": 0.54, + "grad_norm": 2.0799161097959575, + "learning_rate": 9.055240202535856e-06, + "loss": 0.8725, + "step": 23100 + }, + { + "epoch": 0.54, + "grad_norm": 1.0778188103283803, + "learning_rate": 9.054480577201381e-06, + "loss": 1.0127, + "step": 23101 + }, + { + "epoch": 0.54, + "grad_norm": 1.8864794139028263, + "learning_rate": 9.05372095737202e-06, + "loss": 0.891, + "step": 23102 + }, + { + "epoch": 0.54, + "grad_norm": 2.122485504554948, + "learning_rate": 9.052961343052199e-06, + "loss": 1.1303, + "step": 23103 + }, + { + "epoch": 0.54, + "grad_norm": 2.0373288310242925, + "learning_rate": 9.052201734246332e-06, + "loss": 1.0532, + "step": 23104 + }, + { + "epoch": 0.54, + "grad_norm": 2.1451277326195117, + "learning_rate": 9.051442130958848e-06, + "loss": 1.0342, + "step": 23105 + }, + { + "epoch": 0.54, + "grad_norm": 1.8824938244980751, + "learning_rate": 9.050682533194171e-06, + "loss": 0.9481, + "step": 23106 + }, + { + "epoch": 0.54, + "grad_norm": 2.0140370162543926, + "learning_rate": 9.049922940956719e-06, + "loss": 1.0111, + "step": 23107 + }, + { + "epoch": 0.54, + "grad_norm": 1.9918218848093194, + "learning_rate": 9.049163354250914e-06, + "loss": 1.0491, + "step": 23108 + }, + { + "epoch": 0.54, + "grad_norm": 1.9503016836834945, + "learning_rate": 9.048403773081182e-06, + "loss": 1.0162, + "step": 23109 + }, + { + "epoch": 0.54, + "grad_norm": 2.2188828041462156, + "learning_rate": 9.047644197451949e-06, + "loss": 0.9733, + "step": 23110 + }, + { + "epoch": 0.54, + "grad_norm": 2.1159668721390155, + "learning_rate": 9.046884627367628e-06, + "loss": 0.9786, + "step": 23111 + }, + { + "epoch": 0.54, + "grad_norm": 2.038813219508189, + "learning_rate": 9.046125062832646e-06, + "loss": 1.0135, + "step": 23112 + }, + { + "epoch": 0.54, + "grad_norm": 1.068818214907626, + "learning_rate": 9.045365503851432e-06, + "loss": 0.968, + "step": 23113 + }, + { + "epoch": 0.54, + "grad_norm": 1.9962856905315947, + "learning_rate": 9.044605950428397e-06, + "loss": 0.9757, + "step": 23114 + }, + { + "epoch": 0.54, + "grad_norm": 1.7277799077775402, + "learning_rate": 9.043846402567969e-06, + "loss": 0.9087, + "step": 23115 + }, + { + "epoch": 0.54, + "grad_norm": 1.8699113716384734, + "learning_rate": 9.043086860274567e-06, + "loss": 0.9372, + "step": 23116 + }, + { + "epoch": 0.54, + "grad_norm": 1.785001241065318, + "learning_rate": 9.042327323552624e-06, + "loss": 0.9876, + "step": 23117 + }, + { + "epoch": 0.54, + "grad_norm": 1.1007989437821422, + "learning_rate": 9.04156779240655e-06, + "loss": 1.0445, + "step": 23118 + }, + { + "epoch": 0.54, + "grad_norm": 2.048957618788566, + "learning_rate": 9.040808266840773e-06, + "loss": 1.0231, + "step": 23119 + }, + { + "epoch": 0.54, + "grad_norm": 1.885578921550814, + "learning_rate": 9.040048746859715e-06, + "loss": 1.0598, + "step": 23120 + }, + { + "epoch": 0.54, + "grad_norm": 2.0020250570353593, + "learning_rate": 9.039289232467792e-06, + "loss": 0.9311, + "step": 23121 + }, + { + "epoch": 0.54, + "grad_norm": 2.1250143393609497, + "learning_rate": 9.038529723669435e-06, + "loss": 0.9981, + "step": 23122 + }, + { + "epoch": 0.54, + "grad_norm": 2.4895733040938643, + "learning_rate": 9.037770220469062e-06, + "loss": 0.9308, + "step": 23123 + }, + { + "epoch": 0.54, + "grad_norm": 1.8278485072765447, + "learning_rate": 9.037010722871098e-06, + "loss": 1.001, + "step": 23124 + }, + { + "epoch": 0.54, + "grad_norm": 2.6254651628377075, + "learning_rate": 9.036251230879956e-06, + "loss": 0.9432, + "step": 23125 + }, + { + "epoch": 0.54, + "grad_norm": 1.8774006346800787, + "learning_rate": 9.035491744500069e-06, + "loss": 1.0306, + "step": 23126 + }, + { + "epoch": 0.54, + "grad_norm": 1.994669384586488, + "learning_rate": 9.034732263735856e-06, + "loss": 0.8944, + "step": 23127 + }, + { + "epoch": 0.54, + "grad_norm": 2.177813935750976, + "learning_rate": 9.033972788591735e-06, + "loss": 1.1, + "step": 23128 + }, + { + "epoch": 0.54, + "grad_norm": 2.1591929235186393, + "learning_rate": 9.033213319072126e-06, + "loss": 0.9654, + "step": 23129 + }, + { + "epoch": 0.54, + "grad_norm": 2.1115826287722634, + "learning_rate": 9.032453855181462e-06, + "loss": 1.0176, + "step": 23130 + }, + { + "epoch": 0.54, + "grad_norm": 2.222795554888296, + "learning_rate": 9.031694396924156e-06, + "loss": 1.0083, + "step": 23131 + }, + { + "epoch": 0.54, + "grad_norm": 2.245579233325276, + "learning_rate": 9.030934944304632e-06, + "loss": 0.9929, + "step": 23132 + }, + { + "epoch": 0.54, + "grad_norm": 2.598725537708675, + "learning_rate": 9.030175497327308e-06, + "loss": 1.0674, + "step": 23133 + }, + { + "epoch": 0.55, + "grad_norm": 2.1658667853145426, + "learning_rate": 9.029416055996614e-06, + "loss": 0.9723, + "step": 23134 + }, + { + "epoch": 0.55, + "grad_norm": 2.417061830442063, + "learning_rate": 9.028656620316964e-06, + "loss": 0.9594, + "step": 23135 + }, + { + "epoch": 0.55, + "grad_norm": 1.9071298672149533, + "learning_rate": 9.027897190292784e-06, + "loss": 1.0167, + "step": 23136 + }, + { + "epoch": 0.55, + "grad_norm": 1.1865728651552039, + "learning_rate": 9.027137765928492e-06, + "loss": 1.0516, + "step": 23137 + }, + { + "epoch": 0.55, + "grad_norm": 1.9511910980893858, + "learning_rate": 9.026378347228516e-06, + "loss": 0.933, + "step": 23138 + }, + { + "epoch": 0.55, + "grad_norm": 2.1246170953403265, + "learning_rate": 9.025618934197272e-06, + "loss": 0.9925, + "step": 23139 + }, + { + "epoch": 0.55, + "grad_norm": 2.716776814513529, + "learning_rate": 9.02485952683918e-06, + "loss": 0.8202, + "step": 23140 + }, + { + "epoch": 0.55, + "grad_norm": 1.8809828062273521, + "learning_rate": 9.024100125158671e-06, + "loss": 0.9765, + "step": 23141 + }, + { + "epoch": 0.55, + "grad_norm": 1.9002421149291986, + "learning_rate": 9.023340729160154e-06, + "loss": 0.9816, + "step": 23142 + }, + { + "epoch": 0.55, + "grad_norm": 2.039925877529669, + "learning_rate": 9.022581338848058e-06, + "loss": 0.9954, + "step": 23143 + }, + { + "epoch": 0.55, + "grad_norm": 1.950620365908492, + "learning_rate": 9.021821954226801e-06, + "loss": 1.0988, + "step": 23144 + }, + { + "epoch": 0.55, + "grad_norm": 2.252841097702567, + "learning_rate": 9.021062575300814e-06, + "loss": 1.1476, + "step": 23145 + }, + { + "epoch": 0.55, + "grad_norm": 2.3015122397950356, + "learning_rate": 9.020303202074503e-06, + "loss": 0.9175, + "step": 23146 + }, + { + "epoch": 0.55, + "grad_norm": 2.1970864993832766, + "learning_rate": 9.0195438345523e-06, + "loss": 0.9529, + "step": 23147 + }, + { + "epoch": 0.55, + "grad_norm": 8.12392265553021, + "learning_rate": 9.018784472738624e-06, + "loss": 0.9488, + "step": 23148 + }, + { + "epoch": 0.55, + "grad_norm": 1.9190873585487505, + "learning_rate": 9.018025116637895e-06, + "loss": 0.9862, + "step": 23149 + }, + { + "epoch": 0.55, + "grad_norm": 1.9414510930425959, + "learning_rate": 9.01726576625453e-06, + "loss": 0.913, + "step": 23150 + }, + { + "epoch": 0.55, + "grad_norm": 1.879865136930425, + "learning_rate": 9.016506421592959e-06, + "loss": 0.893, + "step": 23151 + }, + { + "epoch": 0.55, + "grad_norm": 1.8978109342486407, + "learning_rate": 9.0157470826576e-06, + "loss": 0.9801, + "step": 23152 + }, + { + "epoch": 0.55, + "grad_norm": 1.1195658841968403, + "learning_rate": 9.01498774945287e-06, + "loss": 0.932, + "step": 23153 + }, + { + "epoch": 0.55, + "grad_norm": 1.903297039817441, + "learning_rate": 9.014228421983194e-06, + "loss": 0.8974, + "step": 23154 + }, + { + "epoch": 0.55, + "grad_norm": 1.1084358711055269, + "learning_rate": 9.013469100252995e-06, + "loss": 0.86, + "step": 23155 + }, + { + "epoch": 0.55, + "grad_norm": 2.259976646190555, + "learning_rate": 9.012709784266688e-06, + "loss": 0.9877, + "step": 23156 + }, + { + "epoch": 0.55, + "grad_norm": 2.2438607257649665, + "learning_rate": 9.011950474028697e-06, + "loss": 0.9692, + "step": 23157 + }, + { + "epoch": 0.55, + "grad_norm": 1.1371151362667622, + "learning_rate": 9.011191169543444e-06, + "loss": 0.9441, + "step": 23158 + }, + { + "epoch": 0.55, + "grad_norm": 2.2945510796023565, + "learning_rate": 9.010431870815351e-06, + "loss": 1.0462, + "step": 23159 + }, + { + "epoch": 0.55, + "grad_norm": 1.098843161049173, + "learning_rate": 9.009672577848836e-06, + "loss": 0.9931, + "step": 23160 + }, + { + "epoch": 0.55, + "grad_norm": 1.9171261555382355, + "learning_rate": 9.008913290648317e-06, + "loss": 1.0286, + "step": 23161 + }, + { + "epoch": 0.55, + "grad_norm": 2.0675654390341567, + "learning_rate": 9.008154009218228e-06, + "loss": 1.0415, + "step": 23162 + }, + { + "epoch": 0.55, + "grad_norm": 1.9177618965248828, + "learning_rate": 9.00739473356297e-06, + "loss": 0.9448, + "step": 23163 + }, + { + "epoch": 0.55, + "grad_norm": 2.0331716173169387, + "learning_rate": 9.00663546368698e-06, + "loss": 0.9421, + "step": 23164 + }, + { + "epoch": 0.55, + "grad_norm": 1.9772919556971942, + "learning_rate": 9.005876199594668e-06, + "loss": 0.9753, + "step": 23165 + }, + { + "epoch": 0.55, + "grad_norm": 2.261218764637884, + "learning_rate": 9.005116941290465e-06, + "loss": 1.0153, + "step": 23166 + }, + { + "epoch": 0.55, + "grad_norm": 1.8640805133159464, + "learning_rate": 9.004357688778784e-06, + "loss": 0.994, + "step": 23167 + }, + { + "epoch": 0.55, + "grad_norm": 1.0418122498409743, + "learning_rate": 9.003598442064048e-06, + "loss": 0.9587, + "step": 23168 + }, + { + "epoch": 0.55, + "grad_norm": 1.0742646038499213, + "learning_rate": 9.002839201150673e-06, + "loss": 0.9367, + "step": 23169 + }, + { + "epoch": 0.55, + "grad_norm": 1.8872457629222879, + "learning_rate": 9.00207996604309e-06, + "loss": 0.8724, + "step": 23170 + }, + { + "epoch": 0.55, + "grad_norm": 1.9725223741922313, + "learning_rate": 9.001320736745711e-06, + "loss": 0.9192, + "step": 23171 + }, + { + "epoch": 0.55, + "grad_norm": 2.025976084171089, + "learning_rate": 9.00056151326296e-06, + "loss": 0.9809, + "step": 23172 + }, + { + "epoch": 0.55, + "grad_norm": 1.793490094467683, + "learning_rate": 8.999802295599257e-06, + "loss": 0.9524, + "step": 23173 + }, + { + "epoch": 0.55, + "grad_norm": 2.03097846068537, + "learning_rate": 8.999043083759016e-06, + "loss": 0.9622, + "step": 23174 + }, + { + "epoch": 0.55, + "grad_norm": 1.9749679233419588, + "learning_rate": 8.998283877746669e-06, + "loss": 0.9805, + "step": 23175 + }, + { + "epoch": 0.55, + "grad_norm": 2.146421784871614, + "learning_rate": 8.997524677566627e-06, + "loss": 0.9691, + "step": 23176 + }, + { + "epoch": 0.55, + "grad_norm": 1.997449020186257, + "learning_rate": 8.996765483223317e-06, + "loss": 1.1201, + "step": 23177 + }, + { + "epoch": 0.55, + "grad_norm": 1.8798186038557847, + "learning_rate": 8.996006294721153e-06, + "loss": 0.9682, + "step": 23178 + }, + { + "epoch": 0.55, + "grad_norm": 2.1787015714020876, + "learning_rate": 8.99524711206456e-06, + "loss": 0.8457, + "step": 23179 + }, + { + "epoch": 0.55, + "grad_norm": 2.5035138919094893, + "learning_rate": 8.994487935257957e-06, + "loss": 1.0663, + "step": 23180 + }, + { + "epoch": 0.55, + "grad_norm": 2.0900959403069383, + "learning_rate": 8.993728764305762e-06, + "loss": 1.0465, + "step": 23181 + }, + { + "epoch": 0.55, + "grad_norm": 2.1520781213571514, + "learning_rate": 8.992969599212395e-06, + "loss": 0.9095, + "step": 23182 + }, + { + "epoch": 0.55, + "grad_norm": 1.9802929395311142, + "learning_rate": 8.99221043998228e-06, + "loss": 0.9521, + "step": 23183 + }, + { + "epoch": 0.55, + "grad_norm": 1.7646532872729217, + "learning_rate": 8.991451286619837e-06, + "loss": 0.9405, + "step": 23184 + }, + { + "epoch": 0.55, + "grad_norm": 1.7801896198873637, + "learning_rate": 8.990692139129481e-06, + "loss": 0.9867, + "step": 23185 + }, + { + "epoch": 0.55, + "grad_norm": 2.040972029412026, + "learning_rate": 8.989932997515634e-06, + "loss": 1.043, + "step": 23186 + }, + { + "epoch": 0.55, + "grad_norm": 1.1021065765457416, + "learning_rate": 8.98917386178272e-06, + "loss": 0.9612, + "step": 23187 + }, + { + "epoch": 0.55, + "grad_norm": 1.9152886493330694, + "learning_rate": 8.988414731935153e-06, + "loss": 0.9941, + "step": 23188 + }, + { + "epoch": 0.55, + "grad_norm": 2.100251075801113, + "learning_rate": 8.987655607977357e-06, + "loss": 1.0548, + "step": 23189 + }, + { + "epoch": 0.55, + "grad_norm": 1.0883866656828167, + "learning_rate": 8.98689648991375e-06, + "loss": 0.8531, + "step": 23190 + }, + { + "epoch": 0.55, + "grad_norm": 1.7536410565860479, + "learning_rate": 8.986137377748754e-06, + "loss": 0.9791, + "step": 23191 + }, + { + "epoch": 0.55, + "grad_norm": 1.9904173038353095, + "learning_rate": 8.985378271486786e-06, + "loss": 0.9696, + "step": 23192 + }, + { + "epoch": 0.55, + "grad_norm": 3.207111550100015, + "learning_rate": 8.984619171132264e-06, + "loss": 1.0036, + "step": 23193 + }, + { + "epoch": 0.55, + "grad_norm": 2.087722768673509, + "learning_rate": 8.983860076689617e-06, + "loss": 0.8886, + "step": 23194 + }, + { + "epoch": 0.55, + "grad_norm": 2.0254178458042724, + "learning_rate": 8.983100988163253e-06, + "loss": 0.9669, + "step": 23195 + }, + { + "epoch": 0.55, + "grad_norm": 1.74479052514097, + "learning_rate": 8.982341905557599e-06, + "loss": 0.8786, + "step": 23196 + }, + { + "epoch": 0.55, + "grad_norm": 2.075909188618524, + "learning_rate": 8.981582828877072e-06, + "loss": 1.0622, + "step": 23197 + }, + { + "epoch": 0.55, + "grad_norm": 1.9441496105596738, + "learning_rate": 8.980823758126096e-06, + "loss": 0.8606, + "step": 23198 + }, + { + "epoch": 0.55, + "grad_norm": 2.002754517394981, + "learning_rate": 8.98006469330908e-06, + "loss": 1.0849, + "step": 23199 + }, + { + "epoch": 0.55, + "grad_norm": 1.8932262782509799, + "learning_rate": 8.979305634430453e-06, + "loss": 1.0654, + "step": 23200 + }, + { + "epoch": 0.55, + "grad_norm": 2.0516608162865766, + "learning_rate": 8.978546581494637e-06, + "loss": 1.0809, + "step": 23201 + }, + { + "epoch": 0.55, + "grad_norm": 1.0482096616769416, + "learning_rate": 8.977787534506038e-06, + "loss": 0.893, + "step": 23202 + }, + { + "epoch": 0.55, + "grad_norm": 1.8330952134987888, + "learning_rate": 8.977028493469088e-06, + "loss": 0.9939, + "step": 23203 + }, + { + "epoch": 0.55, + "grad_norm": 1.1323333773174735, + "learning_rate": 8.976269458388201e-06, + "loss": 0.9883, + "step": 23204 + }, + { + "epoch": 0.55, + "grad_norm": 3.2281807975664383, + "learning_rate": 8.9755104292678e-06, + "loss": 0.9532, + "step": 23205 + }, + { + "epoch": 0.55, + "grad_norm": 2.127756458190721, + "learning_rate": 8.974751406112296e-06, + "loss": 0.9768, + "step": 23206 + }, + { + "epoch": 0.55, + "grad_norm": 2.4207119453440917, + "learning_rate": 8.973992388926117e-06, + "loss": 0.9805, + "step": 23207 + }, + { + "epoch": 0.55, + "grad_norm": 1.1077876262821473, + "learning_rate": 8.973233377713681e-06, + "loss": 0.886, + "step": 23208 + }, + { + "epoch": 0.55, + "grad_norm": 1.9880266130740187, + "learning_rate": 8.972474372479404e-06, + "loss": 0.9881, + "step": 23209 + }, + { + "epoch": 0.55, + "grad_norm": 2.2179323393126404, + "learning_rate": 8.971715373227704e-06, + "loss": 1.0535, + "step": 23210 + }, + { + "epoch": 0.55, + "grad_norm": 2.0876645531406086, + "learning_rate": 8.970956379963006e-06, + "loss": 0.9398, + "step": 23211 + }, + { + "epoch": 0.55, + "grad_norm": 1.9056028608485929, + "learning_rate": 8.970197392689725e-06, + "loss": 1.0109, + "step": 23212 + }, + { + "epoch": 0.55, + "grad_norm": 2.2857245213463893, + "learning_rate": 8.96943841141228e-06, + "loss": 1.0206, + "step": 23213 + }, + { + "epoch": 0.55, + "grad_norm": 5.946847566426012, + "learning_rate": 8.96867943613509e-06, + "loss": 1.1196, + "step": 23214 + }, + { + "epoch": 0.55, + "grad_norm": 2.0309860976200436, + "learning_rate": 8.967920466862577e-06, + "loss": 1.1, + "step": 23215 + }, + { + "epoch": 0.55, + "grad_norm": 2.1715967125269207, + "learning_rate": 8.967161503599157e-06, + "loss": 1.0282, + "step": 23216 + }, + { + "epoch": 0.55, + "grad_norm": 2.212918164489265, + "learning_rate": 8.96640254634925e-06, + "loss": 0.9167, + "step": 23217 + }, + { + "epoch": 0.55, + "grad_norm": 2.1170509865560727, + "learning_rate": 8.96564359511727e-06, + "loss": 0.9589, + "step": 23218 + }, + { + "epoch": 0.55, + "grad_norm": 1.9064312133242307, + "learning_rate": 8.964884649907648e-06, + "loss": 0.8955, + "step": 23219 + }, + { + "epoch": 0.55, + "grad_norm": 2.2146641774732196, + "learning_rate": 8.964125710724791e-06, + "loss": 0.9393, + "step": 23220 + }, + { + "epoch": 0.55, + "grad_norm": 2.035023363341054, + "learning_rate": 8.963366777573123e-06, + "loss": 1.0845, + "step": 23221 + }, + { + "epoch": 0.55, + "grad_norm": 2.044253421198541, + "learning_rate": 8.962607850457063e-06, + "loss": 0.9741, + "step": 23222 + }, + { + "epoch": 0.55, + "grad_norm": 2.3498213773238867, + "learning_rate": 8.961848929381027e-06, + "loss": 1.0023, + "step": 23223 + }, + { + "epoch": 0.55, + "grad_norm": 1.9517407175195527, + "learning_rate": 8.961090014349433e-06, + "loss": 1.0646, + "step": 23224 + }, + { + "epoch": 0.55, + "grad_norm": 2.1885701983565005, + "learning_rate": 8.960331105366705e-06, + "loss": 0.8945, + "step": 23225 + }, + { + "epoch": 0.55, + "grad_norm": 2.0469944985920825, + "learning_rate": 8.95957220243726e-06, + "loss": 0.869, + "step": 23226 + }, + { + "epoch": 0.55, + "grad_norm": 1.110129152726654, + "learning_rate": 8.95881330556551e-06, + "loss": 0.8725, + "step": 23227 + }, + { + "epoch": 0.55, + "grad_norm": 1.90994955550599, + "learning_rate": 8.958054414755882e-06, + "loss": 1.0056, + "step": 23228 + }, + { + "epoch": 0.55, + "grad_norm": 2.216676767389233, + "learning_rate": 8.957295530012792e-06, + "loss": 1.0606, + "step": 23229 + }, + { + "epoch": 0.55, + "grad_norm": 2.0779247173978117, + "learning_rate": 8.956536651340655e-06, + "loss": 0.9352, + "step": 23230 + }, + { + "epoch": 0.55, + "grad_norm": 2.9881482379893307, + "learning_rate": 8.955777778743891e-06, + "loss": 0.9844, + "step": 23231 + }, + { + "epoch": 0.55, + "grad_norm": 2.429859484265516, + "learning_rate": 8.95501891222692e-06, + "loss": 0.9065, + "step": 23232 + }, + { + "epoch": 0.55, + "grad_norm": 1.7748977313447443, + "learning_rate": 8.954260051794165e-06, + "loss": 0.8781, + "step": 23233 + }, + { + "epoch": 0.55, + "grad_norm": 1.900977928895419, + "learning_rate": 8.953501197450035e-06, + "loss": 1.0152, + "step": 23234 + }, + { + "epoch": 0.55, + "grad_norm": 1.9038083836432333, + "learning_rate": 8.952742349198951e-06, + "loss": 0.9791, + "step": 23235 + }, + { + "epoch": 0.55, + "grad_norm": 1.8542833640054555, + "learning_rate": 8.951983507045335e-06, + "loss": 1.004, + "step": 23236 + }, + { + "epoch": 0.55, + "grad_norm": 2.7271091831445826, + "learning_rate": 8.951224670993603e-06, + "loss": 0.937, + "step": 23237 + }, + { + "epoch": 0.55, + "grad_norm": 2.0014189482580425, + "learning_rate": 8.950465841048168e-06, + "loss": 0.8011, + "step": 23238 + }, + { + "epoch": 0.55, + "grad_norm": 1.7034397258242762, + "learning_rate": 8.949707017213457e-06, + "loss": 1.0144, + "step": 23239 + }, + { + "epoch": 0.55, + "grad_norm": 1.8325458017016796, + "learning_rate": 8.948948199493887e-06, + "loss": 1.0743, + "step": 23240 + }, + { + "epoch": 0.55, + "grad_norm": 1.8837392780900049, + "learning_rate": 8.948189387893871e-06, + "loss": 0.9906, + "step": 23241 + }, + { + "epoch": 0.55, + "grad_norm": 1.9665410074018144, + "learning_rate": 8.947430582417827e-06, + "loss": 1.0894, + "step": 23242 + }, + { + "epoch": 0.55, + "grad_norm": 2.1553411299208642, + "learning_rate": 8.946671783070182e-06, + "loss": 1.0989, + "step": 23243 + }, + { + "epoch": 0.55, + "grad_norm": 1.927429349975329, + "learning_rate": 8.94591298985534e-06, + "loss": 1.0379, + "step": 23244 + }, + { + "epoch": 0.55, + "grad_norm": 1.9265221428917565, + "learning_rate": 8.945154202777729e-06, + "loss": 0.9738, + "step": 23245 + }, + { + "epoch": 0.55, + "grad_norm": 2.1344031060361144, + "learning_rate": 8.94439542184176e-06, + "loss": 0.9157, + "step": 23246 + }, + { + "epoch": 0.55, + "grad_norm": 1.0109929212609918, + "learning_rate": 8.943636647051864e-06, + "loss": 0.9452, + "step": 23247 + }, + { + "epoch": 0.55, + "grad_norm": 1.787842031288991, + "learning_rate": 8.942877878412444e-06, + "loss": 0.931, + "step": 23248 + }, + { + "epoch": 0.55, + "grad_norm": 2.0101890820481567, + "learning_rate": 8.942119115927925e-06, + "loss": 1.04, + "step": 23249 + }, + { + "epoch": 0.55, + "grad_norm": 2.389803193720289, + "learning_rate": 8.941360359602724e-06, + "loss": 0.9647, + "step": 23250 + }, + { + "epoch": 0.55, + "grad_norm": 1.0897458534127649, + "learning_rate": 8.940601609441256e-06, + "loss": 0.8897, + "step": 23251 + }, + { + "epoch": 0.55, + "grad_norm": 1.7798716912601849, + "learning_rate": 8.93984286544794e-06, + "loss": 1.0025, + "step": 23252 + }, + { + "epoch": 0.55, + "grad_norm": 1.7880763530734756, + "learning_rate": 8.939084127627198e-06, + "loss": 0.9621, + "step": 23253 + }, + { + "epoch": 0.55, + "grad_norm": 2.112486691797148, + "learning_rate": 8.938325395983445e-06, + "loss": 0.9638, + "step": 23254 + }, + { + "epoch": 0.55, + "grad_norm": 1.9569664723607656, + "learning_rate": 8.937566670521092e-06, + "loss": 0.8971, + "step": 23255 + }, + { + "epoch": 0.55, + "grad_norm": 1.8330571865190617, + "learning_rate": 8.936807951244566e-06, + "loss": 0.9749, + "step": 23256 + }, + { + "epoch": 0.55, + "grad_norm": 2.0712170777228858, + "learning_rate": 8.936049238158283e-06, + "loss": 1.0463, + "step": 23257 + }, + { + "epoch": 0.55, + "grad_norm": 1.8983829298143275, + "learning_rate": 8.935290531266656e-06, + "loss": 0.9817, + "step": 23258 + }, + { + "epoch": 0.55, + "grad_norm": 1.817628058528628, + "learning_rate": 8.934531830574101e-06, + "loss": 0.9641, + "step": 23259 + }, + { + "epoch": 0.55, + "grad_norm": 2.218081789551631, + "learning_rate": 8.933773136085042e-06, + "loss": 1.0883, + "step": 23260 + }, + { + "epoch": 0.55, + "grad_norm": 1.8546010296064295, + "learning_rate": 8.933014447803896e-06, + "loss": 1.0238, + "step": 23261 + }, + { + "epoch": 0.55, + "grad_norm": 2.0521968078458475, + "learning_rate": 8.932255765735075e-06, + "loss": 0.9585, + "step": 23262 + }, + { + "epoch": 0.55, + "grad_norm": 1.8858656639090423, + "learning_rate": 8.931497089882997e-06, + "loss": 1.1265, + "step": 23263 + }, + { + "epoch": 0.55, + "grad_norm": 2.3511824067992655, + "learning_rate": 8.930738420252085e-06, + "loss": 0.9368, + "step": 23264 + }, + { + "epoch": 0.55, + "grad_norm": 2.037116867627238, + "learning_rate": 8.92997975684675e-06, + "loss": 0.9969, + "step": 23265 + }, + { + "epoch": 0.55, + "grad_norm": 1.147833918561953, + "learning_rate": 8.929221099671414e-06, + "loss": 0.9725, + "step": 23266 + }, + { + "epoch": 0.55, + "grad_norm": 2.1930435098973233, + "learning_rate": 8.928462448730487e-06, + "loss": 1.0043, + "step": 23267 + }, + { + "epoch": 0.55, + "grad_norm": 2.0753557890383165, + "learning_rate": 8.927703804028396e-06, + "loss": 1.1188, + "step": 23268 + }, + { + "epoch": 0.55, + "grad_norm": 2.476822158582496, + "learning_rate": 8.926945165569551e-06, + "loss": 1.0566, + "step": 23269 + }, + { + "epoch": 0.55, + "grad_norm": 2.9608502062955258, + "learning_rate": 8.92618653335837e-06, + "loss": 1.0364, + "step": 23270 + }, + { + "epoch": 0.55, + "grad_norm": 1.8887940892912447, + "learning_rate": 8.925427907399276e-06, + "loss": 1.0195, + "step": 23271 + }, + { + "epoch": 0.55, + "grad_norm": 2.025288583301108, + "learning_rate": 8.924669287696673e-06, + "loss": 0.8135, + "step": 23272 + }, + { + "epoch": 0.55, + "grad_norm": 2.0365777237425804, + "learning_rate": 8.92391067425499e-06, + "loss": 1.0041, + "step": 23273 + }, + { + "epoch": 0.55, + "grad_norm": 1.8775051141478478, + "learning_rate": 8.92315206707864e-06, + "loss": 0.8933, + "step": 23274 + }, + { + "epoch": 0.55, + "grad_norm": 1.9373309578187883, + "learning_rate": 8.922393466172042e-06, + "loss": 0.8935, + "step": 23275 + }, + { + "epoch": 0.55, + "grad_norm": 3.4701345245091204, + "learning_rate": 8.921634871539603e-06, + "loss": 1.0779, + "step": 23276 + }, + { + "epoch": 0.55, + "grad_norm": 2.2919320751607986, + "learning_rate": 8.920876283185753e-06, + "loss": 1.0234, + "step": 23277 + }, + { + "epoch": 0.55, + "grad_norm": 2.1372498671823363, + "learning_rate": 8.920117701114903e-06, + "loss": 0.9673, + "step": 23278 + }, + { + "epoch": 0.55, + "grad_norm": 1.082544981623326, + "learning_rate": 8.919359125331468e-06, + "loss": 0.9804, + "step": 23279 + }, + { + "epoch": 0.55, + "grad_norm": 1.8997149783167129, + "learning_rate": 8.918600555839865e-06, + "loss": 0.8923, + "step": 23280 + }, + { + "epoch": 0.55, + "grad_norm": 2.2495743690222243, + "learning_rate": 8.917841992644513e-06, + "loss": 0.9242, + "step": 23281 + }, + { + "epoch": 0.55, + "grad_norm": 1.9830591563342048, + "learning_rate": 8.91708343574983e-06, + "loss": 0.9644, + "step": 23282 + }, + { + "epoch": 0.55, + "grad_norm": 2.088315204204466, + "learning_rate": 8.916324885160228e-06, + "loss": 0.9524, + "step": 23283 + }, + { + "epoch": 0.55, + "grad_norm": 2.0095703889311234, + "learning_rate": 8.915566340880122e-06, + "loss": 0.96, + "step": 23284 + }, + { + "epoch": 0.55, + "grad_norm": 2.449771485204041, + "learning_rate": 8.914807802913939e-06, + "loss": 0.9709, + "step": 23285 + }, + { + "epoch": 0.55, + "grad_norm": 2.4495101092629494, + "learning_rate": 8.914049271266085e-06, + "loss": 0.9791, + "step": 23286 + }, + { + "epoch": 0.55, + "grad_norm": 1.7199228489231044, + "learning_rate": 8.913290745940979e-06, + "loss": 0.9379, + "step": 23287 + }, + { + "epoch": 0.55, + "grad_norm": 2.2403143527164557, + "learning_rate": 8.912532226943038e-06, + "loss": 1.1816, + "step": 23288 + }, + { + "epoch": 0.55, + "grad_norm": 2.3443100830266297, + "learning_rate": 8.911773714276682e-06, + "loss": 1.0467, + "step": 23289 + }, + { + "epoch": 0.55, + "grad_norm": 1.9166616503070126, + "learning_rate": 8.911015207946324e-06, + "loss": 0.9189, + "step": 23290 + }, + { + "epoch": 0.55, + "grad_norm": 2.289882835455958, + "learning_rate": 8.910256707956375e-06, + "loss": 0.9504, + "step": 23291 + }, + { + "epoch": 0.55, + "grad_norm": 2.053534448017467, + "learning_rate": 8.90949821431126e-06, + "loss": 1.103, + "step": 23292 + }, + { + "epoch": 0.55, + "grad_norm": 2.03268210872432, + "learning_rate": 8.908739727015394e-06, + "loss": 0.9649, + "step": 23293 + }, + { + "epoch": 0.55, + "grad_norm": 1.8861385702517541, + "learning_rate": 8.907981246073188e-06, + "loss": 1.0494, + "step": 23294 + }, + { + "epoch": 0.55, + "grad_norm": 2.1599359155289206, + "learning_rate": 8.907222771489059e-06, + "loss": 1.0461, + "step": 23295 + }, + { + "epoch": 0.55, + "grad_norm": 1.8359488904756915, + "learning_rate": 8.90646430326743e-06, + "loss": 1.0784, + "step": 23296 + }, + { + "epoch": 0.55, + "grad_norm": 1.9246508785090717, + "learning_rate": 8.905705841412708e-06, + "loss": 0.9926, + "step": 23297 + }, + { + "epoch": 0.55, + "grad_norm": 2.0188804646980008, + "learning_rate": 8.904947385929314e-06, + "loss": 0.9253, + "step": 23298 + }, + { + "epoch": 0.55, + "grad_norm": 2.065722132330939, + "learning_rate": 8.90418893682166e-06, + "loss": 1.0271, + "step": 23299 + }, + { + "epoch": 0.55, + "grad_norm": 1.8968234165503646, + "learning_rate": 8.903430494094171e-06, + "loss": 1.0963, + "step": 23300 + }, + { + "epoch": 0.55, + "grad_norm": 2.1350877145635048, + "learning_rate": 8.902672057751254e-06, + "loss": 0.9891, + "step": 23301 + }, + { + "epoch": 0.55, + "grad_norm": 1.7153562280200851, + "learning_rate": 8.901913627797328e-06, + "loss": 0.9953, + "step": 23302 + }, + { + "epoch": 0.55, + "grad_norm": 2.843869201305681, + "learning_rate": 8.90115520423681e-06, + "loss": 0.9827, + "step": 23303 + }, + { + "epoch": 0.55, + "grad_norm": 2.743697234813493, + "learning_rate": 8.90039678707411e-06, + "loss": 0.9235, + "step": 23304 + }, + { + "epoch": 0.55, + "grad_norm": 2.616495340010286, + "learning_rate": 8.899638376313651e-06, + "loss": 0.8773, + "step": 23305 + }, + { + "epoch": 0.55, + "grad_norm": 2.704891827947238, + "learning_rate": 8.898879971959844e-06, + "loss": 0.9938, + "step": 23306 + }, + { + "epoch": 0.55, + "grad_norm": 1.9282593208028456, + "learning_rate": 8.89812157401711e-06, + "loss": 1.0698, + "step": 23307 + }, + { + "epoch": 0.55, + "grad_norm": 3.7834246439619648, + "learning_rate": 8.897363182489856e-06, + "loss": 0.9458, + "step": 23308 + }, + { + "epoch": 0.55, + "grad_norm": 1.9991861121354082, + "learning_rate": 8.896604797382507e-06, + "loss": 1.0199, + "step": 23309 + }, + { + "epoch": 0.55, + "grad_norm": 2.786694031083874, + "learning_rate": 8.895846418699475e-06, + "loss": 1.0497, + "step": 23310 + }, + { + "epoch": 0.55, + "grad_norm": 2.1489760581074475, + "learning_rate": 8.89508804644517e-06, + "loss": 0.9475, + "step": 23311 + }, + { + "epoch": 0.55, + "grad_norm": 1.8742890285314426, + "learning_rate": 8.894329680624012e-06, + "loss": 1.1062, + "step": 23312 + }, + { + "epoch": 0.55, + "grad_norm": 1.0607900320998922, + "learning_rate": 8.89357132124042e-06, + "loss": 0.9599, + "step": 23313 + }, + { + "epoch": 0.55, + "grad_norm": 1.7458910071647702, + "learning_rate": 8.892812968298807e-06, + "loss": 1.1444, + "step": 23314 + }, + { + "epoch": 0.55, + "grad_norm": 1.1329082929459255, + "learning_rate": 8.892054621803586e-06, + "loss": 1.0553, + "step": 23315 + }, + { + "epoch": 0.55, + "grad_norm": 2.5070078249011094, + "learning_rate": 8.89129628175917e-06, + "loss": 1.0271, + "step": 23316 + }, + { + "epoch": 0.55, + "grad_norm": 1.8616513962108687, + "learning_rate": 8.890537948169984e-06, + "loss": 1.0623, + "step": 23317 + }, + { + "epoch": 0.55, + "grad_norm": 2.3480092913201527, + "learning_rate": 8.889779621040435e-06, + "loss": 0.9505, + "step": 23318 + }, + { + "epoch": 0.55, + "grad_norm": 1.9878978654718722, + "learning_rate": 8.88902130037494e-06, + "loss": 1.0593, + "step": 23319 + }, + { + "epoch": 0.55, + "grad_norm": 2.0129894680659457, + "learning_rate": 8.888262986177913e-06, + "loss": 1.0459, + "step": 23320 + }, + { + "epoch": 0.55, + "grad_norm": 2.1369310607858187, + "learning_rate": 8.887504678453775e-06, + "loss": 0.9787, + "step": 23321 + }, + { + "epoch": 0.55, + "grad_norm": 2.0367399516893387, + "learning_rate": 8.886746377206934e-06, + "loss": 0.9933, + "step": 23322 + }, + { + "epoch": 0.55, + "grad_norm": 1.949919281077019, + "learning_rate": 8.88598808244181e-06, + "loss": 0.9606, + "step": 23323 + }, + { + "epoch": 0.55, + "grad_norm": 2.298051579687162, + "learning_rate": 8.885229794162818e-06, + "loss": 0.9914, + "step": 23324 + }, + { + "epoch": 0.55, + "grad_norm": 1.935882798077081, + "learning_rate": 8.884471512374363e-06, + "loss": 0.9639, + "step": 23325 + }, + { + "epoch": 0.55, + "grad_norm": 1.7856831841096632, + "learning_rate": 8.883713237080875e-06, + "loss": 0.9761, + "step": 23326 + }, + { + "epoch": 0.55, + "grad_norm": 2.0130473134029288, + "learning_rate": 8.882954968286756e-06, + "loss": 1.1383, + "step": 23327 + }, + { + "epoch": 0.55, + "grad_norm": 2.1694644598017576, + "learning_rate": 8.882196705996436e-06, + "loss": 0.9246, + "step": 23328 + }, + { + "epoch": 0.55, + "grad_norm": 1.8948168461888235, + "learning_rate": 8.881438450214313e-06, + "loss": 0.9252, + "step": 23329 + }, + { + "epoch": 0.55, + "grad_norm": 2.0299608748942815, + "learning_rate": 8.880680200944812e-06, + "loss": 0.9947, + "step": 23330 + }, + { + "epoch": 0.55, + "grad_norm": 1.912617577153721, + "learning_rate": 8.879921958192347e-06, + "loss": 0.9856, + "step": 23331 + }, + { + "epoch": 0.55, + "grad_norm": 1.8919412713323964, + "learning_rate": 8.87916372196133e-06, + "loss": 0.9852, + "step": 23332 + }, + { + "epoch": 0.55, + "grad_norm": 2.1027734618502207, + "learning_rate": 8.878405492256174e-06, + "loss": 1.0858, + "step": 23333 + }, + { + "epoch": 0.55, + "grad_norm": 2.418483811578134, + "learning_rate": 8.877647269081298e-06, + "loss": 1.0055, + "step": 23334 + }, + { + "epoch": 0.55, + "grad_norm": 2.123660411285242, + "learning_rate": 8.876889052441117e-06, + "loss": 0.9968, + "step": 23335 + }, + { + "epoch": 0.55, + "grad_norm": 2.172014728946667, + "learning_rate": 8.876130842340039e-06, + "loss": 1.1165, + "step": 23336 + }, + { + "epoch": 0.55, + "grad_norm": 2.113008905219231, + "learning_rate": 8.875372638782485e-06, + "loss": 1.0799, + "step": 23337 + }, + { + "epoch": 0.55, + "grad_norm": 1.8772215388298068, + "learning_rate": 8.874614441772871e-06, + "loss": 0.9086, + "step": 23338 + }, + { + "epoch": 0.55, + "grad_norm": 1.983897140484006, + "learning_rate": 8.873856251315605e-06, + "loss": 1.0072, + "step": 23339 + }, + { + "epoch": 0.55, + "grad_norm": 2.0365080377349165, + "learning_rate": 8.873098067415103e-06, + "loss": 0.9284, + "step": 23340 + }, + { + "epoch": 0.55, + "grad_norm": 1.9423434084256062, + "learning_rate": 8.872339890075782e-06, + "loss": 1.0548, + "step": 23341 + }, + { + "epoch": 0.55, + "grad_norm": 2.2222636309539503, + "learning_rate": 8.871581719302057e-06, + "loss": 0.9583, + "step": 23342 + }, + { + "epoch": 0.55, + "grad_norm": 1.9362791761475957, + "learning_rate": 8.870823555098339e-06, + "loss": 0.9832, + "step": 23343 + }, + { + "epoch": 0.55, + "grad_norm": 2.0831805062367024, + "learning_rate": 8.870065397469043e-06, + "loss": 1.0564, + "step": 23344 + }, + { + "epoch": 0.55, + "grad_norm": 2.217541438733788, + "learning_rate": 8.869307246418589e-06, + "loss": 1.0396, + "step": 23345 + }, + { + "epoch": 0.55, + "grad_norm": 2.5030506798184464, + "learning_rate": 8.86854910195138e-06, + "loss": 0.9826, + "step": 23346 + }, + { + "epoch": 0.55, + "grad_norm": 2.5941324249208915, + "learning_rate": 8.867790964071838e-06, + "loss": 0.9276, + "step": 23347 + }, + { + "epoch": 0.55, + "grad_norm": 1.1152022365684287, + "learning_rate": 8.867032832784372e-06, + "loss": 0.9221, + "step": 23348 + }, + { + "epoch": 0.55, + "grad_norm": 1.781536286074975, + "learning_rate": 8.866274708093407e-06, + "loss": 1.0532, + "step": 23349 + }, + { + "epoch": 0.55, + "grad_norm": 1.8694342100014583, + "learning_rate": 8.865516590003344e-06, + "loss": 1.0991, + "step": 23350 + }, + { + "epoch": 0.55, + "grad_norm": 1.987531194502065, + "learning_rate": 8.864758478518606e-06, + "loss": 0.8742, + "step": 23351 + }, + { + "epoch": 0.55, + "grad_norm": 1.9466300407631085, + "learning_rate": 8.864000373643604e-06, + "loss": 1.0337, + "step": 23352 + }, + { + "epoch": 0.55, + "grad_norm": 2.0378034372582285, + "learning_rate": 8.863242275382747e-06, + "loss": 1.0114, + "step": 23353 + }, + { + "epoch": 0.55, + "grad_norm": 1.1015130083273272, + "learning_rate": 8.862484183740457e-06, + "loss": 0.9668, + "step": 23354 + }, + { + "epoch": 0.55, + "grad_norm": 2.03122691705316, + "learning_rate": 8.861726098721142e-06, + "loss": 0.9726, + "step": 23355 + }, + { + "epoch": 0.55, + "grad_norm": 1.2038387428842887, + "learning_rate": 8.860968020329222e-06, + "loss": 0.9553, + "step": 23356 + }, + { + "epoch": 0.55, + "grad_norm": 1.7730817469487063, + "learning_rate": 8.8602099485691e-06, + "loss": 0.9339, + "step": 23357 + }, + { + "epoch": 0.55, + "grad_norm": 1.9745465884381177, + "learning_rate": 8.859451883445201e-06, + "loss": 1.0189, + "step": 23358 + }, + { + "epoch": 0.55, + "grad_norm": 2.945279153384012, + "learning_rate": 8.858693824961936e-06, + "loss": 1.0064, + "step": 23359 + }, + { + "epoch": 0.55, + "grad_norm": 1.8734678084706289, + "learning_rate": 8.857935773123715e-06, + "loss": 0.9182, + "step": 23360 + }, + { + "epoch": 0.55, + "grad_norm": 2.131885933497293, + "learning_rate": 8.85717772793495e-06, + "loss": 0.9769, + "step": 23361 + }, + { + "epoch": 0.55, + "grad_norm": 1.8066838821181317, + "learning_rate": 8.856419689400062e-06, + "loss": 1.0281, + "step": 23362 + }, + { + "epoch": 0.55, + "grad_norm": 1.969973566155097, + "learning_rate": 8.855661657523461e-06, + "loss": 1.0189, + "step": 23363 + }, + { + "epoch": 0.55, + "grad_norm": 1.9429813720657816, + "learning_rate": 8.85490363230956e-06, + "loss": 0.9456, + "step": 23364 + }, + { + "epoch": 0.55, + "grad_norm": 2.6887609376099086, + "learning_rate": 8.854145613762767e-06, + "loss": 0.9921, + "step": 23365 + }, + { + "epoch": 0.55, + "grad_norm": 1.9353276743556518, + "learning_rate": 8.85338760188751e-06, + "loss": 1.0414, + "step": 23366 + }, + { + "epoch": 0.55, + "grad_norm": 1.05458813018158, + "learning_rate": 8.852629596688189e-06, + "loss": 0.9439, + "step": 23367 + }, + { + "epoch": 0.55, + "grad_norm": 1.7354768918171766, + "learning_rate": 8.85187159816922e-06, + "loss": 1.0639, + "step": 23368 + }, + { + "epoch": 0.55, + "grad_norm": 2.1415217064425565, + "learning_rate": 8.851113606335018e-06, + "loss": 1.1105, + "step": 23369 + }, + { + "epoch": 0.55, + "grad_norm": 2.01666585124715, + "learning_rate": 8.85035562119e-06, + "loss": 1.0447, + "step": 23370 + }, + { + "epoch": 0.55, + "grad_norm": 1.137904930859651, + "learning_rate": 8.849597642738573e-06, + "loss": 0.958, + "step": 23371 + }, + { + "epoch": 0.55, + "grad_norm": 2.0766608195390144, + "learning_rate": 8.848839670985151e-06, + "loss": 1.0666, + "step": 23372 + }, + { + "epoch": 0.55, + "grad_norm": 2.0209748206409537, + "learning_rate": 8.848081705934155e-06, + "loss": 1.0373, + "step": 23373 + }, + { + "epoch": 0.55, + "grad_norm": 1.1379928346163246, + "learning_rate": 8.847323747589987e-06, + "loss": 0.9162, + "step": 23374 + }, + { + "epoch": 0.55, + "grad_norm": 2.012916436188122, + "learning_rate": 8.846565795957066e-06, + "loss": 1.056, + "step": 23375 + }, + { + "epoch": 0.55, + "grad_norm": 3.0552426965761588, + "learning_rate": 8.845807851039801e-06, + "loss": 0.8515, + "step": 23376 + }, + { + "epoch": 0.55, + "grad_norm": 2.307325625333415, + "learning_rate": 8.845049912842616e-06, + "loss": 0.8679, + "step": 23377 + }, + { + "epoch": 0.55, + "grad_norm": 2.1856860944452787, + "learning_rate": 8.844291981369911e-06, + "loss": 0.9949, + "step": 23378 + }, + { + "epoch": 0.55, + "grad_norm": 2.0245776266791604, + "learning_rate": 8.843534056626103e-06, + "loss": 0.9326, + "step": 23379 + }, + { + "epoch": 0.55, + "grad_norm": 2.697818691589106, + "learning_rate": 8.842776138615609e-06, + "loss": 0.8876, + "step": 23380 + }, + { + "epoch": 0.55, + "grad_norm": 2.3023127328769033, + "learning_rate": 8.842018227342839e-06, + "loss": 0.9376, + "step": 23381 + }, + { + "epoch": 0.55, + "grad_norm": 1.9342995248418224, + "learning_rate": 8.8412603228122e-06, + "loss": 0.9852, + "step": 23382 + }, + { + "epoch": 0.55, + "grad_norm": 2.1293419356040424, + "learning_rate": 8.840502425028114e-06, + "loss": 0.9456, + "step": 23383 + }, + { + "epoch": 0.55, + "grad_norm": 2.166474541823423, + "learning_rate": 8.839744533994991e-06, + "loss": 0.9936, + "step": 23384 + }, + { + "epoch": 0.55, + "grad_norm": 2.259776662290188, + "learning_rate": 8.83898664971724e-06, + "loss": 1.0989, + "step": 23385 + }, + { + "epoch": 0.55, + "grad_norm": 1.9355530444017048, + "learning_rate": 8.838228772199278e-06, + "loss": 0.9145, + "step": 23386 + }, + { + "epoch": 0.55, + "grad_norm": 2.5837012058847497, + "learning_rate": 8.837470901445518e-06, + "loss": 0.9569, + "step": 23387 + }, + { + "epoch": 0.55, + "grad_norm": 2.22238702067391, + "learning_rate": 8.836713037460368e-06, + "loss": 0.9741, + "step": 23388 + }, + { + "epoch": 0.55, + "grad_norm": 2.1822968568997165, + "learning_rate": 8.83595518024824e-06, + "loss": 1.0427, + "step": 23389 + }, + { + "epoch": 0.55, + "grad_norm": 1.945359763036646, + "learning_rate": 8.835197329813552e-06, + "loss": 0.9643, + "step": 23390 + }, + { + "epoch": 0.55, + "grad_norm": 1.8958681852051733, + "learning_rate": 8.834439486160716e-06, + "loss": 1.015, + "step": 23391 + }, + { + "epoch": 0.55, + "grad_norm": 1.0699344762747365, + "learning_rate": 8.833681649294142e-06, + "loss": 0.9888, + "step": 23392 + }, + { + "epoch": 0.55, + "grad_norm": 2.5004840562885677, + "learning_rate": 8.832923819218238e-06, + "loss": 0.9733, + "step": 23393 + }, + { + "epoch": 0.55, + "grad_norm": 2.121935683692049, + "learning_rate": 8.83216599593743e-06, + "loss": 0.9697, + "step": 23394 + }, + { + "epoch": 0.55, + "grad_norm": 1.1396935937380315, + "learning_rate": 8.831408179456112e-06, + "loss": 0.946, + "step": 23395 + }, + { + "epoch": 0.55, + "grad_norm": 2.2638127356783295, + "learning_rate": 8.830650369778711e-06, + "loss": 1.0568, + "step": 23396 + }, + { + "epoch": 0.55, + "grad_norm": 1.802226826118572, + "learning_rate": 8.829892566909629e-06, + "loss": 1.0114, + "step": 23397 + }, + { + "epoch": 0.55, + "grad_norm": 1.0827498423264084, + "learning_rate": 8.829134770853289e-06, + "loss": 0.8942, + "step": 23398 + }, + { + "epoch": 0.55, + "grad_norm": 2.0685573173378726, + "learning_rate": 8.828376981614093e-06, + "loss": 1.0939, + "step": 23399 + }, + { + "epoch": 0.55, + "grad_norm": 2.0352680055261616, + "learning_rate": 8.827619199196457e-06, + "loss": 1.0317, + "step": 23400 + }, + { + "epoch": 0.55, + "grad_norm": 3.8617221749274333, + "learning_rate": 8.826861423604797e-06, + "loss": 0.9313, + "step": 23401 + }, + { + "epoch": 0.55, + "grad_norm": 2.498045840050645, + "learning_rate": 8.826103654843516e-06, + "loss": 0.9927, + "step": 23402 + }, + { + "epoch": 0.55, + "grad_norm": 1.933063734027531, + "learning_rate": 8.825345892917035e-06, + "loss": 0.8667, + "step": 23403 + }, + { + "epoch": 0.55, + "grad_norm": 1.1580690737338075, + "learning_rate": 8.824588137829761e-06, + "loss": 1.0178, + "step": 23404 + }, + { + "epoch": 0.55, + "grad_norm": 1.8330521395922124, + "learning_rate": 8.823830389586108e-06, + "loss": 0.9271, + "step": 23405 + }, + { + "epoch": 0.55, + "grad_norm": 2.519512997882735, + "learning_rate": 8.823072648190484e-06, + "loss": 1.0728, + "step": 23406 + }, + { + "epoch": 0.55, + "grad_norm": 2.1314444824719794, + "learning_rate": 8.822314913647305e-06, + "loss": 1.1236, + "step": 23407 + }, + { + "epoch": 0.55, + "grad_norm": 3.8483276245578937, + "learning_rate": 8.821557185960984e-06, + "loss": 0.8899, + "step": 23408 + }, + { + "epoch": 0.55, + "grad_norm": 2.140270568292198, + "learning_rate": 8.820799465135928e-06, + "loss": 0.8788, + "step": 23409 + }, + { + "epoch": 0.55, + "grad_norm": 1.8678377350543622, + "learning_rate": 8.820041751176549e-06, + "loss": 1.0036, + "step": 23410 + }, + { + "epoch": 0.55, + "grad_norm": 1.8584366380781498, + "learning_rate": 8.81928404408726e-06, + "loss": 0.957, + "step": 23411 + }, + { + "epoch": 0.55, + "grad_norm": 1.9753783895299568, + "learning_rate": 8.81852634387248e-06, + "loss": 0.9686, + "step": 23412 + }, + { + "epoch": 0.55, + "grad_norm": 3.253753336594059, + "learning_rate": 8.817768650536607e-06, + "loss": 1.0202, + "step": 23413 + }, + { + "epoch": 0.55, + "grad_norm": 1.7970578661675132, + "learning_rate": 8.817010964084059e-06, + "loss": 1.0429, + "step": 23414 + }, + { + "epoch": 0.55, + "grad_norm": 2.1025968819201384, + "learning_rate": 8.81625328451925e-06, + "loss": 0.9882, + "step": 23415 + }, + { + "epoch": 0.55, + "grad_norm": 2.2449416912924085, + "learning_rate": 8.815495611846592e-06, + "loss": 0.9102, + "step": 23416 + }, + { + "epoch": 0.55, + "grad_norm": 1.1458052894676238, + "learning_rate": 8.814737946070489e-06, + "loss": 0.9755, + "step": 23417 + }, + { + "epoch": 0.55, + "grad_norm": 1.927277113346158, + "learning_rate": 8.813980287195358e-06, + "loss": 0.9666, + "step": 23418 + }, + { + "epoch": 0.55, + "grad_norm": 2.1782231097850486, + "learning_rate": 8.81322263522561e-06, + "loss": 1.0583, + "step": 23419 + }, + { + "epoch": 0.55, + "grad_norm": 1.7738875813653934, + "learning_rate": 8.812464990165656e-06, + "loss": 0.9561, + "step": 23420 + }, + { + "epoch": 0.55, + "grad_norm": 2.534462975781281, + "learning_rate": 8.811707352019905e-06, + "loss": 1.0307, + "step": 23421 + }, + { + "epoch": 0.55, + "grad_norm": 1.9821639274861487, + "learning_rate": 8.810949720792771e-06, + "loss": 1.0227, + "step": 23422 + }, + { + "epoch": 0.55, + "grad_norm": 2.0962044660826513, + "learning_rate": 8.810192096488666e-06, + "loss": 1.0837, + "step": 23423 + }, + { + "epoch": 0.55, + "grad_norm": 1.904168193455225, + "learning_rate": 8.809434479111998e-06, + "loss": 1.0172, + "step": 23424 + }, + { + "epoch": 0.55, + "grad_norm": 2.101345995693247, + "learning_rate": 8.808676868667177e-06, + "loss": 0.9973, + "step": 23425 + }, + { + "epoch": 0.55, + "grad_norm": 1.987290570609638, + "learning_rate": 8.807919265158622e-06, + "loss": 1.0887, + "step": 23426 + }, + { + "epoch": 0.55, + "grad_norm": 2.1603942142035737, + "learning_rate": 8.807161668590733e-06, + "loss": 0.9607, + "step": 23427 + }, + { + "epoch": 0.55, + "grad_norm": 1.9344064637361493, + "learning_rate": 8.806404078967928e-06, + "loss": 1.0803, + "step": 23428 + }, + { + "epoch": 0.55, + "grad_norm": 1.9585239154729297, + "learning_rate": 8.805646496294613e-06, + "loss": 0.9031, + "step": 23429 + }, + { + "epoch": 0.55, + "grad_norm": 1.8317405561815463, + "learning_rate": 8.80488892057521e-06, + "loss": 0.9896, + "step": 23430 + }, + { + "epoch": 0.55, + "grad_norm": 1.8139721304232441, + "learning_rate": 8.804131351814117e-06, + "loss": 0.9726, + "step": 23431 + }, + { + "epoch": 0.55, + "grad_norm": 1.8835056913348593, + "learning_rate": 8.80337379001575e-06, + "loss": 1.006, + "step": 23432 + }, + { + "epoch": 0.55, + "grad_norm": 2.2547362994716065, + "learning_rate": 8.802616235184524e-06, + "loss": 0.9877, + "step": 23433 + }, + { + "epoch": 0.55, + "grad_norm": 1.9154571912474552, + "learning_rate": 8.801858687324839e-06, + "loss": 0.9775, + "step": 23434 + }, + { + "epoch": 0.55, + "grad_norm": 2.7440623482641238, + "learning_rate": 8.801101146441114e-06, + "loss": 1.1278, + "step": 23435 + }, + { + "epoch": 0.55, + "grad_norm": 2.556589944066126, + "learning_rate": 8.800343612537758e-06, + "loss": 0.9343, + "step": 23436 + }, + { + "epoch": 0.55, + "grad_norm": 2.0714591168725582, + "learning_rate": 8.799586085619184e-06, + "loss": 0.9479, + "step": 23437 + }, + { + "epoch": 0.55, + "grad_norm": 3.1975795587575813, + "learning_rate": 8.798828565689796e-06, + "loss": 1.1039, + "step": 23438 + }, + { + "epoch": 0.55, + "grad_norm": 1.9028242493789203, + "learning_rate": 8.79807105275401e-06, + "loss": 0.9423, + "step": 23439 + }, + { + "epoch": 0.55, + "grad_norm": 1.95804815387549, + "learning_rate": 8.797313546816238e-06, + "loss": 1.0481, + "step": 23440 + }, + { + "epoch": 0.55, + "grad_norm": 1.9502588740125826, + "learning_rate": 8.796556047880883e-06, + "loss": 1.0925, + "step": 23441 + }, + { + "epoch": 0.55, + "grad_norm": 2.0986584065617575, + "learning_rate": 8.79579855595236e-06, + "loss": 0.9648, + "step": 23442 + }, + { + "epoch": 0.55, + "grad_norm": 1.8833968413360147, + "learning_rate": 8.795041071035078e-06, + "loss": 0.9895, + "step": 23443 + }, + { + "epoch": 0.55, + "grad_norm": 1.986492424825874, + "learning_rate": 8.794283593133454e-06, + "loss": 0.8939, + "step": 23444 + }, + { + "epoch": 0.55, + "grad_norm": 1.057342441655834, + "learning_rate": 8.793526122251888e-06, + "loss": 0.939, + "step": 23445 + }, + { + "epoch": 0.55, + "grad_norm": 2.3492978687017243, + "learning_rate": 8.792768658394794e-06, + "loss": 0.9926, + "step": 23446 + }, + { + "epoch": 0.55, + "grad_norm": 1.902945606445931, + "learning_rate": 8.792011201566587e-06, + "loss": 0.9974, + "step": 23447 + }, + { + "epoch": 0.55, + "grad_norm": 1.9637528803075492, + "learning_rate": 8.791253751771671e-06, + "loss": 1.2341, + "step": 23448 + }, + { + "epoch": 0.55, + "grad_norm": 2.1199804657955714, + "learning_rate": 8.790496309014457e-06, + "loss": 0.9983, + "step": 23449 + }, + { + "epoch": 0.55, + "grad_norm": 1.9853228637555633, + "learning_rate": 8.789738873299356e-06, + "loss": 1.1056, + "step": 23450 + }, + { + "epoch": 0.55, + "grad_norm": 2.222497076224645, + "learning_rate": 8.788981444630784e-06, + "loss": 1.0673, + "step": 23451 + }, + { + "epoch": 0.55, + "grad_norm": 2.0171613401228092, + "learning_rate": 8.78822402301314e-06, + "loss": 1.1057, + "step": 23452 + }, + { + "epoch": 0.55, + "grad_norm": 3.065637267404256, + "learning_rate": 8.787466608450843e-06, + "loss": 0.9057, + "step": 23453 + }, + { + "epoch": 0.55, + "grad_norm": 1.1538052748536731, + "learning_rate": 8.786709200948299e-06, + "loss": 0.9181, + "step": 23454 + }, + { + "epoch": 0.55, + "grad_norm": 1.0541216422863748, + "learning_rate": 8.785951800509915e-06, + "loss": 0.9157, + "step": 23455 + }, + { + "epoch": 0.55, + "grad_norm": 2.1143404788095133, + "learning_rate": 8.785194407140105e-06, + "loss": 0.9505, + "step": 23456 + }, + { + "epoch": 0.55, + "grad_norm": 1.8813742564098275, + "learning_rate": 8.784437020843278e-06, + "loss": 0.8757, + "step": 23457 + }, + { + "epoch": 0.55, + "grad_norm": 1.8487904719330404, + "learning_rate": 8.783679641623845e-06, + "loss": 1.003, + "step": 23458 + }, + { + "epoch": 0.55, + "grad_norm": 2.153905245857157, + "learning_rate": 8.782922269486211e-06, + "loss": 0.9679, + "step": 23459 + }, + { + "epoch": 0.55, + "grad_norm": 1.9227776043367448, + "learning_rate": 8.782164904434791e-06, + "loss": 0.976, + "step": 23460 + }, + { + "epoch": 0.55, + "grad_norm": 1.8160610063145939, + "learning_rate": 8.781407546473995e-06, + "loss": 0.8959, + "step": 23461 + }, + { + "epoch": 0.55, + "grad_norm": 2.1461811317895343, + "learning_rate": 8.780650195608228e-06, + "loss": 0.9888, + "step": 23462 + }, + { + "epoch": 0.55, + "grad_norm": 1.9059786050078533, + "learning_rate": 8.779892851841897e-06, + "loss": 1.087, + "step": 23463 + }, + { + "epoch": 0.55, + "grad_norm": 1.0962341920450727, + "learning_rate": 8.77913551517942e-06, + "loss": 1.0034, + "step": 23464 + }, + { + "epoch": 0.55, + "grad_norm": 1.1831221525210074, + "learning_rate": 8.778378185625206e-06, + "loss": 1.0162, + "step": 23465 + }, + { + "epoch": 0.55, + "grad_norm": 2.309662236406612, + "learning_rate": 8.777620863183658e-06, + "loss": 0.8611, + "step": 23466 + }, + { + "epoch": 0.55, + "grad_norm": 1.9453343597011195, + "learning_rate": 8.776863547859185e-06, + "loss": 1.0573, + "step": 23467 + }, + { + "epoch": 0.55, + "grad_norm": 2.082336192652773, + "learning_rate": 8.776106239656206e-06, + "loss": 0.9519, + "step": 23468 + }, + { + "epoch": 0.55, + "grad_norm": 1.04207322811929, + "learning_rate": 8.77534893857912e-06, + "loss": 0.9269, + "step": 23469 + }, + { + "epoch": 0.55, + "grad_norm": 1.0960136284284185, + "learning_rate": 8.774591644632338e-06, + "loss": 0.9206, + "step": 23470 + }, + { + "epoch": 0.55, + "grad_norm": 1.6536194989859045, + "learning_rate": 8.773834357820275e-06, + "loss": 1.039, + "step": 23471 + }, + { + "epoch": 0.55, + "grad_norm": 1.0674794274201957, + "learning_rate": 8.773077078147337e-06, + "loss": 0.9983, + "step": 23472 + }, + { + "epoch": 0.55, + "grad_norm": 1.9413429330041287, + "learning_rate": 8.772319805617931e-06, + "loss": 0.9841, + "step": 23473 + }, + { + "epoch": 0.55, + "grad_norm": 2.0174353469170883, + "learning_rate": 8.771562540236467e-06, + "loss": 0.95, + "step": 23474 + }, + { + "epoch": 0.55, + "grad_norm": 2.070037965408562, + "learning_rate": 8.77080528200736e-06, + "loss": 1.0483, + "step": 23475 + }, + { + "epoch": 0.55, + "grad_norm": 2.1646083731188965, + "learning_rate": 8.770048030935007e-06, + "loss": 0.9383, + "step": 23476 + }, + { + "epoch": 0.55, + "grad_norm": 2.1706803784397968, + "learning_rate": 8.769290787023828e-06, + "loss": 1.0253, + "step": 23477 + }, + { + "epoch": 0.55, + "grad_norm": 2.110453368087509, + "learning_rate": 8.768533550278223e-06, + "loss": 1.0519, + "step": 23478 + }, + { + "epoch": 0.55, + "grad_norm": 2.2766663396391618, + "learning_rate": 8.767776320702614e-06, + "loss": 0.8769, + "step": 23479 + }, + { + "epoch": 0.55, + "grad_norm": 2.2278821450301085, + "learning_rate": 8.767019098301394e-06, + "loss": 0.9679, + "step": 23480 + }, + { + "epoch": 0.55, + "grad_norm": 2.427423702090946, + "learning_rate": 8.766261883078981e-06, + "loss": 1.0066, + "step": 23481 + }, + { + "epoch": 0.55, + "grad_norm": 2.189348965718597, + "learning_rate": 8.765504675039784e-06, + "loss": 1.0424, + "step": 23482 + }, + { + "epoch": 0.55, + "grad_norm": 2.265226234497715, + "learning_rate": 8.764747474188205e-06, + "loss": 1.1237, + "step": 23483 + }, + { + "epoch": 0.55, + "grad_norm": 1.8793155178329983, + "learning_rate": 8.763990280528661e-06, + "loss": 1.0658, + "step": 23484 + }, + { + "epoch": 0.55, + "grad_norm": 2.010006233215935, + "learning_rate": 8.763233094065556e-06, + "loss": 0.9109, + "step": 23485 + }, + { + "epoch": 0.55, + "grad_norm": 2.605619141761411, + "learning_rate": 8.7624759148033e-06, + "loss": 0.9146, + "step": 23486 + }, + { + "epoch": 0.55, + "grad_norm": 2.0984823422387766, + "learning_rate": 8.7617187427463e-06, + "loss": 0.9465, + "step": 23487 + }, + { + "epoch": 0.55, + "grad_norm": 1.7790734374416242, + "learning_rate": 8.760961577898965e-06, + "loss": 1.0224, + "step": 23488 + }, + { + "epoch": 0.55, + "grad_norm": 2.1148055646695103, + "learning_rate": 8.760204420265707e-06, + "loss": 1.0821, + "step": 23489 + }, + { + "epoch": 0.55, + "grad_norm": 2.5741760738059565, + "learning_rate": 8.75944726985093e-06, + "loss": 1.0315, + "step": 23490 + }, + { + "epoch": 0.55, + "grad_norm": 1.1693337308258498, + "learning_rate": 8.758690126659041e-06, + "loss": 0.9664, + "step": 23491 + }, + { + "epoch": 0.55, + "grad_norm": 1.1196521800492403, + "learning_rate": 8.757932990694454e-06, + "loss": 0.9572, + "step": 23492 + }, + { + "epoch": 0.55, + "grad_norm": 2.1354488270313023, + "learning_rate": 8.757175861961576e-06, + "loss": 1.0678, + "step": 23493 + }, + { + "epoch": 0.55, + "grad_norm": 2.233981382417517, + "learning_rate": 8.75641874046481e-06, + "loss": 1.0161, + "step": 23494 + }, + { + "epoch": 0.55, + "grad_norm": 2.2609479688993224, + "learning_rate": 8.755661626208568e-06, + "loss": 0.8683, + "step": 23495 + }, + { + "epoch": 0.55, + "grad_norm": 1.9068999764071652, + "learning_rate": 8.75490451919726e-06, + "loss": 0.9681, + "step": 23496 + }, + { + "epoch": 0.55, + "grad_norm": 1.936863488722507, + "learning_rate": 8.754147419435291e-06, + "loss": 0.9714, + "step": 23497 + }, + { + "epoch": 0.55, + "grad_norm": 1.9030839926612184, + "learning_rate": 8.75339032692707e-06, + "loss": 1.0348, + "step": 23498 + }, + { + "epoch": 0.55, + "grad_norm": 1.9487463083772856, + "learning_rate": 8.752633241677003e-06, + "loss": 0.9515, + "step": 23499 + }, + { + "epoch": 0.55, + "grad_norm": 1.94839474181149, + "learning_rate": 8.751876163689506e-06, + "loss": 1.0007, + "step": 23500 + }, + { + "epoch": 0.55, + "grad_norm": 2.5690186850934666, + "learning_rate": 8.751119092968978e-06, + "loss": 1.0118, + "step": 23501 + }, + { + "epoch": 0.55, + "grad_norm": 1.9685699107571868, + "learning_rate": 8.75036202951983e-06, + "loss": 0.9479, + "step": 23502 + }, + { + "epoch": 0.55, + "grad_norm": 2.1864931358752484, + "learning_rate": 8.74960497334647e-06, + "loss": 1.0278, + "step": 23503 + }, + { + "epoch": 0.55, + "grad_norm": 2.134120816349342, + "learning_rate": 8.748847924453304e-06, + "loss": 0.9368, + "step": 23504 + }, + { + "epoch": 0.55, + "grad_norm": 1.9917624291229925, + "learning_rate": 8.748090882844742e-06, + "loss": 1.0425, + "step": 23505 + }, + { + "epoch": 0.55, + "grad_norm": 1.1337222663661146, + "learning_rate": 8.747333848525193e-06, + "loss": 0.9209, + "step": 23506 + }, + { + "epoch": 0.55, + "grad_norm": 2.181962002273606, + "learning_rate": 8.746576821499062e-06, + "loss": 1.0363, + "step": 23507 + }, + { + "epoch": 0.55, + "grad_norm": 2.543644715907432, + "learning_rate": 8.745819801770756e-06, + "loss": 1.1669, + "step": 23508 + }, + { + "epoch": 0.55, + "grad_norm": 2.456465364575123, + "learning_rate": 8.745062789344686e-06, + "loss": 0.8828, + "step": 23509 + }, + { + "epoch": 0.55, + "grad_norm": 2.0052291645388722, + "learning_rate": 8.74430578422526e-06, + "loss": 0.9457, + "step": 23510 + }, + { + "epoch": 0.55, + "grad_norm": 2.129089174920476, + "learning_rate": 8.74354878641688e-06, + "loss": 1.0288, + "step": 23511 + }, + { + "epoch": 0.55, + "grad_norm": 2.496649438495289, + "learning_rate": 8.742791795923956e-06, + "loss": 1.1031, + "step": 23512 + }, + { + "epoch": 0.55, + "grad_norm": 1.9214659367338967, + "learning_rate": 8.742034812750896e-06, + "loss": 1.0336, + "step": 23513 + }, + { + "epoch": 0.55, + "grad_norm": 1.910834515151297, + "learning_rate": 8.741277836902114e-06, + "loss": 0.8735, + "step": 23514 + }, + { + "epoch": 0.55, + "grad_norm": 1.1160403898980986, + "learning_rate": 8.740520868382005e-06, + "loss": 0.9708, + "step": 23515 + }, + { + "epoch": 0.55, + "grad_norm": 1.9547970110118382, + "learning_rate": 8.739763907194981e-06, + "loss": 0.93, + "step": 23516 + }, + { + "epoch": 0.55, + "grad_norm": 5.282359728903886, + "learning_rate": 8.739006953345455e-06, + "loss": 0.9731, + "step": 23517 + }, + { + "epoch": 0.55, + "grad_norm": 1.8290841212003723, + "learning_rate": 8.738250006837828e-06, + "loss": 0.8905, + "step": 23518 + }, + { + "epoch": 0.55, + "grad_norm": 1.98325273239442, + "learning_rate": 8.737493067676508e-06, + "loss": 1.0053, + "step": 23519 + }, + { + "epoch": 0.55, + "grad_norm": 2.990605628546047, + "learning_rate": 8.736736135865901e-06, + "loss": 0.9979, + "step": 23520 + }, + { + "epoch": 0.55, + "grad_norm": 2.307610555862054, + "learning_rate": 8.735979211410423e-06, + "loss": 0.9705, + "step": 23521 + }, + { + "epoch": 0.55, + "grad_norm": 2.169355737605757, + "learning_rate": 8.73522229431447e-06, + "loss": 0.9551, + "step": 23522 + }, + { + "epoch": 0.55, + "grad_norm": 1.9820493147204987, + "learning_rate": 8.73446538458245e-06, + "loss": 0.9899, + "step": 23523 + }, + { + "epoch": 0.55, + "grad_norm": 1.90114494268012, + "learning_rate": 8.733708482218782e-06, + "loss": 1.0026, + "step": 23524 + }, + { + "epoch": 0.55, + "grad_norm": 1.888122876002185, + "learning_rate": 8.732951587227856e-06, + "loss": 1.103, + "step": 23525 + }, + { + "epoch": 0.55, + "grad_norm": 1.8935276412458646, + "learning_rate": 8.732194699614089e-06, + "loss": 0.9531, + "step": 23526 + }, + { + "epoch": 0.55, + "grad_norm": 2.2505063096365325, + "learning_rate": 8.731437819381884e-06, + "loss": 1.0131, + "step": 23527 + }, + { + "epoch": 0.55, + "grad_norm": 2.1351799999002106, + "learning_rate": 8.730680946535657e-06, + "loss": 1.0133, + "step": 23528 + }, + { + "epoch": 0.55, + "grad_norm": 2.2348035628749745, + "learning_rate": 8.7299240810798e-06, + "loss": 0.9678, + "step": 23529 + }, + { + "epoch": 0.55, + "grad_norm": 1.9157056596632727, + "learning_rate": 8.72916722301873e-06, + "loss": 1.1605, + "step": 23530 + }, + { + "epoch": 0.55, + "grad_norm": 1.8928118077755203, + "learning_rate": 8.72841037235685e-06, + "loss": 1.0717, + "step": 23531 + }, + { + "epoch": 0.55, + "grad_norm": 1.853656534068442, + "learning_rate": 8.727653529098567e-06, + "loss": 0.9022, + "step": 23532 + }, + { + "epoch": 0.55, + "grad_norm": 1.9847089850593678, + "learning_rate": 8.726896693248288e-06, + "loss": 1.0333, + "step": 23533 + }, + { + "epoch": 0.55, + "grad_norm": 2.3257097641839772, + "learning_rate": 8.726139864810422e-06, + "loss": 1.0007, + "step": 23534 + }, + { + "epoch": 0.55, + "grad_norm": 1.1026077791817996, + "learning_rate": 8.725383043789372e-06, + "loss": 0.9123, + "step": 23535 + }, + { + "epoch": 0.55, + "grad_norm": 2.0340123876943417, + "learning_rate": 8.724626230189544e-06, + "loss": 1.0587, + "step": 23536 + }, + { + "epoch": 0.55, + "grad_norm": 2.455429215352655, + "learning_rate": 8.723869424015345e-06, + "loss": 0.9413, + "step": 23537 + }, + { + "epoch": 0.55, + "grad_norm": 3.7093782789519953, + "learning_rate": 8.723112625271184e-06, + "loss": 1.0123, + "step": 23538 + }, + { + "epoch": 0.55, + "grad_norm": 2.150991551613632, + "learning_rate": 8.722355833961467e-06, + "loss": 1.1605, + "step": 23539 + }, + { + "epoch": 0.55, + "grad_norm": 1.8119812862356066, + "learning_rate": 8.721599050090596e-06, + "loss": 1.0425, + "step": 23540 + }, + { + "epoch": 0.55, + "grad_norm": 2.2335317165979367, + "learning_rate": 8.72084227366298e-06, + "loss": 1.0366, + "step": 23541 + }, + { + "epoch": 0.55, + "grad_norm": 1.7615524640923756, + "learning_rate": 8.720085504683028e-06, + "loss": 1.0213, + "step": 23542 + }, + { + "epoch": 0.55, + "grad_norm": 2.01628108595442, + "learning_rate": 8.719328743155144e-06, + "loss": 1.0417, + "step": 23543 + }, + { + "epoch": 0.55, + "grad_norm": 1.9783002778798633, + "learning_rate": 8.718571989083729e-06, + "loss": 1.0262, + "step": 23544 + }, + { + "epoch": 0.55, + "grad_norm": 1.1084708969445825, + "learning_rate": 8.717815242473196e-06, + "loss": 0.9673, + "step": 23545 + }, + { + "epoch": 0.55, + "grad_norm": 1.9762474769524379, + "learning_rate": 8.71705850332795e-06, + "loss": 1.0014, + "step": 23546 + }, + { + "epoch": 0.55, + "grad_norm": 1.9918415207421862, + "learning_rate": 8.716301771652393e-06, + "loss": 0.916, + "step": 23547 + }, + { + "epoch": 0.55, + "grad_norm": 1.0832357781547894, + "learning_rate": 8.715545047450933e-06, + "loss": 0.9251, + "step": 23548 + }, + { + "epoch": 0.55, + "grad_norm": 1.925603077066701, + "learning_rate": 8.71478833072798e-06, + "loss": 0.8049, + "step": 23549 + }, + { + "epoch": 0.55, + "grad_norm": 2.180592300850346, + "learning_rate": 8.714031621487934e-06, + "loss": 1.0738, + "step": 23550 + }, + { + "epoch": 0.55, + "grad_norm": 1.1241042019474077, + "learning_rate": 8.713274919735204e-06, + "loss": 0.9998, + "step": 23551 + }, + { + "epoch": 0.55, + "grad_norm": 2.210467182071266, + "learning_rate": 8.712518225474191e-06, + "loss": 1.0235, + "step": 23552 + }, + { + "epoch": 0.55, + "grad_norm": 1.9204689665993195, + "learning_rate": 8.71176153870931e-06, + "loss": 1.0093, + "step": 23553 + }, + { + "epoch": 0.55, + "grad_norm": 2.1430113427473167, + "learning_rate": 8.711004859444957e-06, + "loss": 1.0384, + "step": 23554 + }, + { + "epoch": 0.55, + "grad_norm": 2.615237862069709, + "learning_rate": 8.710248187685545e-06, + "loss": 1.0359, + "step": 23555 + }, + { + "epoch": 0.55, + "grad_norm": 2.2946025790624685, + "learning_rate": 8.709491523435477e-06, + "loss": 0.8644, + "step": 23556 + }, + { + "epoch": 0.55, + "grad_norm": 2.53379787093472, + "learning_rate": 8.708734866699153e-06, + "loss": 0.9189, + "step": 23557 + }, + { + "epoch": 0.56, + "grad_norm": 1.8919624496765077, + "learning_rate": 8.707978217480986e-06, + "loss": 0.963, + "step": 23558 + }, + { + "epoch": 0.56, + "grad_norm": 2.104083095471445, + "learning_rate": 8.707221575785379e-06, + "loss": 1.0249, + "step": 23559 + }, + { + "epoch": 0.56, + "grad_norm": 1.9533560536926804, + "learning_rate": 8.706464941616741e-06, + "loss": 0.9596, + "step": 23560 + }, + { + "epoch": 0.56, + "grad_norm": 1.888883312522067, + "learning_rate": 8.705708314979468e-06, + "loss": 1.0033, + "step": 23561 + }, + { + "epoch": 0.56, + "grad_norm": 1.885420316432262, + "learning_rate": 8.704951695877973e-06, + "loss": 0.9329, + "step": 23562 + }, + { + "epoch": 0.56, + "grad_norm": 1.2069650866451382, + "learning_rate": 8.70419508431666e-06, + "loss": 0.9624, + "step": 23563 + }, + { + "epoch": 0.56, + "grad_norm": 2.1926977124194327, + "learning_rate": 8.703438480299932e-06, + "loss": 1.131, + "step": 23564 + }, + { + "epoch": 0.56, + "grad_norm": 1.9963051473827766, + "learning_rate": 8.702681883832196e-06, + "loss": 1.0615, + "step": 23565 + }, + { + "epoch": 0.56, + "grad_norm": 1.1121287525963939, + "learning_rate": 8.701925294917857e-06, + "loss": 0.8861, + "step": 23566 + }, + { + "epoch": 0.56, + "grad_norm": 2.043841669099481, + "learning_rate": 8.70116871356132e-06, + "loss": 0.9198, + "step": 23567 + }, + { + "epoch": 0.56, + "grad_norm": 2.284174890327635, + "learning_rate": 8.700412139766989e-06, + "loss": 0.9928, + "step": 23568 + }, + { + "epoch": 0.56, + "grad_norm": 2.176906916447761, + "learning_rate": 8.69965557353927e-06, + "loss": 0.9731, + "step": 23569 + }, + { + "epoch": 0.56, + "grad_norm": 2.29935430806163, + "learning_rate": 8.698899014882572e-06, + "loss": 0.8449, + "step": 23570 + }, + { + "epoch": 0.56, + "grad_norm": 1.9448380487741852, + "learning_rate": 8.698142463801292e-06, + "loss": 1.0958, + "step": 23571 + }, + { + "epoch": 0.56, + "grad_norm": 2.2715152576836712, + "learning_rate": 8.697385920299838e-06, + "loss": 0.9941, + "step": 23572 + }, + { + "epoch": 0.56, + "grad_norm": 1.960660353071528, + "learning_rate": 8.696629384382616e-06, + "loss": 0.8597, + "step": 23573 + }, + { + "epoch": 0.56, + "grad_norm": 2.3579462959638726, + "learning_rate": 8.695872856054034e-06, + "loss": 1.0569, + "step": 23574 + }, + { + "epoch": 0.56, + "grad_norm": 2.21350040645105, + "learning_rate": 8.695116335318492e-06, + "loss": 0.9763, + "step": 23575 + }, + { + "epoch": 0.56, + "grad_norm": 2.3564234081235145, + "learning_rate": 8.694359822180391e-06, + "loss": 0.9911, + "step": 23576 + }, + { + "epoch": 0.56, + "grad_norm": 1.9643975922920347, + "learning_rate": 8.693603316644147e-06, + "loss": 0.9338, + "step": 23577 + }, + { + "epoch": 0.56, + "grad_norm": 2.051757966036566, + "learning_rate": 8.692846818714154e-06, + "loss": 1.1762, + "step": 23578 + }, + { + "epoch": 0.56, + "grad_norm": 1.1025346165863898, + "learning_rate": 8.692090328394821e-06, + "loss": 1.0531, + "step": 23579 + }, + { + "epoch": 0.56, + "grad_norm": 1.9013840992248427, + "learning_rate": 8.69133384569055e-06, + "loss": 0.9449, + "step": 23580 + }, + { + "epoch": 0.56, + "grad_norm": 2.207627273356984, + "learning_rate": 8.690577370605752e-06, + "loss": 1.077, + "step": 23581 + }, + { + "epoch": 0.56, + "grad_norm": 2.413634700391636, + "learning_rate": 8.689820903144827e-06, + "loss": 1.0834, + "step": 23582 + }, + { + "epoch": 0.56, + "grad_norm": 2.47497279081175, + "learning_rate": 8.689064443312177e-06, + "loss": 0.9556, + "step": 23583 + }, + { + "epoch": 0.56, + "grad_norm": 2.045905640914909, + "learning_rate": 8.688307991112212e-06, + "loss": 1.0774, + "step": 23584 + }, + { + "epoch": 0.56, + "grad_norm": 2.3992220317933963, + "learning_rate": 8.68755154654933e-06, + "loss": 0.9347, + "step": 23585 + }, + { + "epoch": 0.56, + "grad_norm": 2.0428868124843236, + "learning_rate": 8.68679510962794e-06, + "loss": 1.1327, + "step": 23586 + }, + { + "epoch": 0.56, + "grad_norm": 1.8769138515166397, + "learning_rate": 8.686038680352442e-06, + "loss": 0.9866, + "step": 23587 + }, + { + "epoch": 0.56, + "grad_norm": 1.8379060158556386, + "learning_rate": 8.685282258727248e-06, + "loss": 0.9382, + "step": 23588 + }, + { + "epoch": 0.56, + "grad_norm": 3.3868706304696015, + "learning_rate": 8.684525844756751e-06, + "loss": 0.979, + "step": 23589 + }, + { + "epoch": 0.56, + "grad_norm": 2.154437082933947, + "learning_rate": 8.683769438445365e-06, + "loss": 1.0274, + "step": 23590 + }, + { + "epoch": 0.56, + "grad_norm": 1.9169236190093528, + "learning_rate": 8.68301303979749e-06, + "loss": 1.095, + "step": 23591 + }, + { + "epoch": 0.56, + "grad_norm": 2.0757472022465544, + "learning_rate": 8.682256648817528e-06, + "loss": 0.9573, + "step": 23592 + }, + { + "epoch": 0.56, + "grad_norm": 2.0517572508942363, + "learning_rate": 8.681500265509884e-06, + "loss": 1.0731, + "step": 23593 + }, + { + "epoch": 0.56, + "grad_norm": 1.9364789024893527, + "learning_rate": 8.680743889878964e-06, + "loss": 1.0986, + "step": 23594 + }, + { + "epoch": 0.56, + "grad_norm": 2.143841645512947, + "learning_rate": 8.679987521929174e-06, + "loss": 1.01, + "step": 23595 + }, + { + "epoch": 0.56, + "grad_norm": 1.8683117802755962, + "learning_rate": 8.679231161664911e-06, + "loss": 0.9478, + "step": 23596 + }, + { + "epoch": 0.56, + "grad_norm": 2.086499583635023, + "learning_rate": 8.678474809090583e-06, + "loss": 1.0181, + "step": 23597 + }, + { + "epoch": 0.56, + "grad_norm": 8.05068327848416, + "learning_rate": 8.677718464210594e-06, + "loss": 0.9173, + "step": 23598 + }, + { + "epoch": 0.56, + "grad_norm": 1.8712248509873666, + "learning_rate": 8.676962127029348e-06, + "loss": 0.9865, + "step": 23599 + }, + { + "epoch": 0.56, + "grad_norm": 1.1276491615171857, + "learning_rate": 8.676205797551243e-06, + "loss": 0.9627, + "step": 23600 + }, + { + "epoch": 0.56, + "grad_norm": 1.091447315957273, + "learning_rate": 8.675449475780689e-06, + "loss": 0.9411, + "step": 23601 + }, + { + "epoch": 0.56, + "grad_norm": 2.17353972885152, + "learning_rate": 8.674693161722089e-06, + "loss": 1.0049, + "step": 23602 + }, + { + "epoch": 0.56, + "grad_norm": 1.1295685528918484, + "learning_rate": 8.673936855379844e-06, + "loss": 0.9052, + "step": 23603 + }, + { + "epoch": 0.56, + "grad_norm": 2.232516189144852, + "learning_rate": 8.673180556758358e-06, + "loss": 0.9654, + "step": 23604 + }, + { + "epoch": 0.56, + "grad_norm": 2.14163695964749, + "learning_rate": 8.672424265862037e-06, + "loss": 0.9551, + "step": 23605 + }, + { + "epoch": 0.56, + "grad_norm": 1.7410613837925037, + "learning_rate": 8.671667982695279e-06, + "loss": 0.9438, + "step": 23606 + }, + { + "epoch": 0.56, + "grad_norm": 2.218586740649271, + "learning_rate": 8.670911707262491e-06, + "loss": 0.9932, + "step": 23607 + }, + { + "epoch": 0.56, + "grad_norm": 2.147059903097481, + "learning_rate": 8.670155439568076e-06, + "loss": 1.0274, + "step": 23608 + }, + { + "epoch": 0.56, + "grad_norm": 2.073451263538184, + "learning_rate": 8.669399179616441e-06, + "loss": 1.0286, + "step": 23609 + }, + { + "epoch": 0.56, + "grad_norm": 1.9167118766344549, + "learning_rate": 8.66864292741198e-06, + "loss": 0.9789, + "step": 23610 + }, + { + "epoch": 0.56, + "grad_norm": 2.0243838055415617, + "learning_rate": 8.667886682959103e-06, + "loss": 1.0689, + "step": 23611 + }, + { + "epoch": 0.56, + "grad_norm": 1.8214166907164393, + "learning_rate": 8.667130446262214e-06, + "loss": 1.0113, + "step": 23612 + }, + { + "epoch": 0.56, + "grad_norm": 2.149529674420026, + "learning_rate": 8.666374217325713e-06, + "loss": 1.0206, + "step": 23613 + }, + { + "epoch": 0.56, + "grad_norm": 1.9862903403361496, + "learning_rate": 8.665617996154e-06, + "loss": 0.9137, + "step": 23614 + }, + { + "epoch": 0.56, + "grad_norm": 1.989596653340516, + "learning_rate": 8.664861782751482e-06, + "loss": 1.1156, + "step": 23615 + }, + { + "epoch": 0.56, + "grad_norm": 1.9891927988929592, + "learning_rate": 8.664105577122566e-06, + "loss": 1.0614, + "step": 23616 + }, + { + "epoch": 0.56, + "grad_norm": 2.1451314620805855, + "learning_rate": 8.663349379271645e-06, + "loss": 0.9827, + "step": 23617 + }, + { + "epoch": 0.56, + "grad_norm": 1.9943626978481372, + "learning_rate": 8.662593189203128e-06, + "loss": 0.9209, + "step": 23618 + }, + { + "epoch": 0.56, + "grad_norm": 1.9091745885063822, + "learning_rate": 8.661837006921422e-06, + "loss": 1.0061, + "step": 23619 + }, + { + "epoch": 0.56, + "grad_norm": 1.9940369471182524, + "learning_rate": 8.661080832430919e-06, + "loss": 0.9283, + "step": 23620 + }, + { + "epoch": 0.56, + "grad_norm": 2.034782561610495, + "learning_rate": 8.660324665736026e-06, + "loss": 1.1124, + "step": 23621 + }, + { + "epoch": 0.56, + "grad_norm": 1.9754086765890588, + "learning_rate": 8.659568506841151e-06, + "loss": 1.1086, + "step": 23622 + }, + { + "epoch": 0.56, + "grad_norm": 2.19753918979298, + "learning_rate": 8.658812355750692e-06, + "loss": 0.9871, + "step": 23623 + }, + { + "epoch": 0.56, + "grad_norm": 1.8277300295652765, + "learning_rate": 8.658056212469052e-06, + "loss": 0.8935, + "step": 23624 + }, + { + "epoch": 0.56, + "grad_norm": 1.9065343630545248, + "learning_rate": 8.657300077000632e-06, + "loss": 1.0057, + "step": 23625 + }, + { + "epoch": 0.56, + "grad_norm": 1.9340009295528033, + "learning_rate": 8.65654394934984e-06, + "loss": 1.0096, + "step": 23626 + }, + { + "epoch": 0.56, + "grad_norm": 1.995461188500953, + "learning_rate": 8.65578782952107e-06, + "loss": 1.1769, + "step": 23627 + }, + { + "epoch": 0.56, + "grad_norm": 2.163517494185528, + "learning_rate": 8.655031717518732e-06, + "loss": 1.0071, + "step": 23628 + }, + { + "epoch": 0.56, + "grad_norm": 1.1325265234653807, + "learning_rate": 8.65427561334722e-06, + "loss": 0.9544, + "step": 23629 + }, + { + "epoch": 0.56, + "grad_norm": 1.765076981387078, + "learning_rate": 8.653519517010947e-06, + "loss": 1.009, + "step": 23630 + }, + { + "epoch": 0.56, + "grad_norm": 2.7520412037726327, + "learning_rate": 8.652763428514309e-06, + "loss": 1.0347, + "step": 23631 + }, + { + "epoch": 0.56, + "grad_norm": 1.9005555767891609, + "learning_rate": 8.652007347861707e-06, + "loss": 0.9258, + "step": 23632 + }, + { + "epoch": 0.56, + "grad_norm": 2.2942047735670674, + "learning_rate": 8.651251275057547e-06, + "loss": 1.0816, + "step": 23633 + }, + { + "epoch": 0.56, + "grad_norm": 2.4230216962913502, + "learning_rate": 8.650495210106225e-06, + "loss": 0.9654, + "step": 23634 + }, + { + "epoch": 0.56, + "grad_norm": 1.9056581789889449, + "learning_rate": 8.64973915301215e-06, + "loss": 0.913, + "step": 23635 + }, + { + "epoch": 0.56, + "grad_norm": 2.1529705942197803, + "learning_rate": 8.648983103779722e-06, + "loss": 1.0158, + "step": 23636 + }, + { + "epoch": 0.56, + "grad_norm": 2.1599420923985075, + "learning_rate": 8.648227062413342e-06, + "loss": 0.8801, + "step": 23637 + }, + { + "epoch": 0.56, + "grad_norm": 2.0654632884081314, + "learning_rate": 8.647471028917409e-06, + "loss": 1.0211, + "step": 23638 + }, + { + "epoch": 0.56, + "grad_norm": 2.2675643120820586, + "learning_rate": 8.64671500329633e-06, + "loss": 1.0029, + "step": 23639 + }, + { + "epoch": 0.56, + "grad_norm": 2.1143494793179527, + "learning_rate": 8.645958985554506e-06, + "loss": 0.9339, + "step": 23640 + }, + { + "epoch": 0.56, + "grad_norm": 1.9765043701276803, + "learning_rate": 8.645202975696336e-06, + "loss": 0.9311, + "step": 23641 + }, + { + "epoch": 0.56, + "grad_norm": 1.7757322938801383, + "learning_rate": 8.644446973726222e-06, + "loss": 1.0084, + "step": 23642 + }, + { + "epoch": 0.56, + "grad_norm": 2.027757233949401, + "learning_rate": 8.643690979648567e-06, + "loss": 1.1459, + "step": 23643 + }, + { + "epoch": 0.56, + "grad_norm": 2.024744671813156, + "learning_rate": 8.642934993467776e-06, + "loss": 0.9606, + "step": 23644 + }, + { + "epoch": 0.56, + "grad_norm": 2.5305084088035623, + "learning_rate": 8.642179015188245e-06, + "loss": 1.0525, + "step": 23645 + }, + { + "epoch": 0.56, + "grad_norm": 2.0593229249703633, + "learning_rate": 8.641423044814375e-06, + "loss": 0.9696, + "step": 23646 + }, + { + "epoch": 0.56, + "grad_norm": 1.108113909533993, + "learning_rate": 8.640667082350574e-06, + "loss": 0.9712, + "step": 23647 + }, + { + "epoch": 0.56, + "grad_norm": 2.0224683148204874, + "learning_rate": 8.639911127801238e-06, + "loss": 1.1156, + "step": 23648 + }, + { + "epoch": 0.56, + "grad_norm": 2.1761805630866364, + "learning_rate": 8.639155181170771e-06, + "loss": 0.9874, + "step": 23649 + }, + { + "epoch": 0.56, + "grad_norm": 2.1973491717455587, + "learning_rate": 8.63839924246357e-06, + "loss": 1.0037, + "step": 23650 + }, + { + "epoch": 0.56, + "grad_norm": 3.180687938087576, + "learning_rate": 8.637643311684045e-06, + "loss": 0.982, + "step": 23651 + }, + { + "epoch": 0.56, + "grad_norm": 2.106714048362514, + "learning_rate": 8.636887388836588e-06, + "loss": 1.0437, + "step": 23652 + }, + { + "epoch": 0.56, + "grad_norm": 1.9481951207137906, + "learning_rate": 8.636131473925605e-06, + "loss": 1.0343, + "step": 23653 + }, + { + "epoch": 0.56, + "grad_norm": 3.3084892628856313, + "learning_rate": 8.6353755669555e-06, + "loss": 0.8652, + "step": 23654 + }, + { + "epoch": 0.56, + "grad_norm": 2.6241068478946774, + "learning_rate": 8.634619667930664e-06, + "loss": 1.0101, + "step": 23655 + }, + { + "epoch": 0.56, + "grad_norm": 1.9329571534553163, + "learning_rate": 8.633863776855508e-06, + "loss": 1.0124, + "step": 23656 + }, + { + "epoch": 0.56, + "grad_norm": 1.0758259729242905, + "learning_rate": 8.633107893734427e-06, + "loss": 0.9334, + "step": 23657 + }, + { + "epoch": 0.56, + "grad_norm": 1.8770741511513986, + "learning_rate": 8.63235201857183e-06, + "loss": 1.0272, + "step": 23658 + }, + { + "epoch": 0.56, + "grad_norm": 1.7431040445648904, + "learning_rate": 8.631596151372107e-06, + "loss": 0.9521, + "step": 23659 + }, + { + "epoch": 0.56, + "grad_norm": 1.8362624845851678, + "learning_rate": 8.630840292139668e-06, + "loss": 1.0023, + "step": 23660 + }, + { + "epoch": 0.56, + "grad_norm": 1.8325497379298588, + "learning_rate": 8.630084440878906e-06, + "loss": 1.0112, + "step": 23661 + }, + { + "epoch": 0.56, + "grad_norm": 1.910902157357354, + "learning_rate": 8.629328597594233e-06, + "loss": 1.0021, + "step": 23662 + }, + { + "epoch": 0.56, + "grad_norm": 2.03599846623101, + "learning_rate": 8.628572762290038e-06, + "loss": 1.0888, + "step": 23663 + }, + { + "epoch": 0.56, + "grad_norm": 1.8731757231310358, + "learning_rate": 8.627816934970728e-06, + "loss": 0.9087, + "step": 23664 + }, + { + "epoch": 0.56, + "grad_norm": 1.685682763761205, + "learning_rate": 8.627061115640703e-06, + "loss": 1.0065, + "step": 23665 + }, + { + "epoch": 0.56, + "grad_norm": 2.010428817844894, + "learning_rate": 8.62630530430436e-06, + "loss": 1.1592, + "step": 23666 + }, + { + "epoch": 0.56, + "grad_norm": 1.9072453836190153, + "learning_rate": 8.625549500966103e-06, + "loss": 0.9431, + "step": 23667 + }, + { + "epoch": 0.56, + "grad_norm": 2.0642545457049484, + "learning_rate": 8.624793705630334e-06, + "loss": 1.1666, + "step": 23668 + }, + { + "epoch": 0.56, + "grad_norm": 2.1171564433329126, + "learning_rate": 8.624037918301453e-06, + "loss": 0.9391, + "step": 23669 + }, + { + "epoch": 0.56, + "grad_norm": 1.0816139208681979, + "learning_rate": 8.623282138983854e-06, + "loss": 0.9276, + "step": 23670 + }, + { + "epoch": 0.56, + "grad_norm": 1.906100107112698, + "learning_rate": 8.622526367681944e-06, + "loss": 0.8816, + "step": 23671 + }, + { + "epoch": 0.56, + "grad_norm": 2.119550027831777, + "learning_rate": 8.621770604400126e-06, + "loss": 1.0658, + "step": 23672 + }, + { + "epoch": 0.56, + "grad_norm": 1.8308571230163584, + "learning_rate": 8.621014849142792e-06, + "loss": 1.0909, + "step": 23673 + }, + { + "epoch": 0.56, + "grad_norm": 2.5650219466212687, + "learning_rate": 8.620259101914344e-06, + "loss": 1.0488, + "step": 23674 + }, + { + "epoch": 0.56, + "grad_norm": 2.084031415549641, + "learning_rate": 8.619503362719186e-06, + "loss": 0.9945, + "step": 23675 + }, + { + "epoch": 0.56, + "grad_norm": 2.092867655219715, + "learning_rate": 8.618747631561719e-06, + "loss": 1.0951, + "step": 23676 + }, + { + "epoch": 0.56, + "grad_norm": 1.9164396117292712, + "learning_rate": 8.617991908446339e-06, + "loss": 0.9753, + "step": 23677 + }, + { + "epoch": 0.56, + "grad_norm": 2.0342243518060164, + "learning_rate": 8.617236193377445e-06, + "loss": 1.0112, + "step": 23678 + }, + { + "epoch": 0.56, + "grad_norm": 1.9978142115321436, + "learning_rate": 8.616480486359444e-06, + "loss": 1.1283, + "step": 23679 + }, + { + "epoch": 0.56, + "grad_norm": 2.3230292779704373, + "learning_rate": 8.61572478739673e-06, + "loss": 0.9881, + "step": 23680 + }, + { + "epoch": 0.56, + "grad_norm": 2.1148297860432668, + "learning_rate": 8.614969096493704e-06, + "loss": 1.1515, + "step": 23681 + }, + { + "epoch": 0.56, + "grad_norm": 1.9139921980106516, + "learning_rate": 8.614213413654764e-06, + "loss": 0.9736, + "step": 23682 + }, + { + "epoch": 0.56, + "grad_norm": 1.0747030035824763, + "learning_rate": 8.613457738884316e-06, + "loss": 1.0013, + "step": 23683 + }, + { + "epoch": 0.56, + "grad_norm": 1.9495877849961345, + "learning_rate": 8.612702072186757e-06, + "loss": 0.8713, + "step": 23684 + }, + { + "epoch": 0.56, + "grad_norm": 1.1310961020333954, + "learning_rate": 8.611946413566481e-06, + "loss": 0.8952, + "step": 23685 + }, + { + "epoch": 0.56, + "grad_norm": 1.871672028461009, + "learning_rate": 8.611190763027898e-06, + "loss": 1.0661, + "step": 23686 + }, + { + "epoch": 0.56, + "grad_norm": 1.8961076984817844, + "learning_rate": 8.610435120575396e-06, + "loss": 0.9491, + "step": 23687 + }, + { + "epoch": 0.56, + "grad_norm": 1.9206407678534223, + "learning_rate": 8.609679486213384e-06, + "loss": 0.9089, + "step": 23688 + }, + { + "epoch": 0.56, + "grad_norm": 1.8636875597407319, + "learning_rate": 8.608923859946259e-06, + "loss": 1.0814, + "step": 23689 + }, + { + "epoch": 0.56, + "grad_norm": 1.8851228680552043, + "learning_rate": 8.60816824177842e-06, + "loss": 0.9213, + "step": 23690 + }, + { + "epoch": 0.56, + "grad_norm": 1.769258056745297, + "learning_rate": 8.607412631714263e-06, + "loss": 1.1102, + "step": 23691 + }, + { + "epoch": 0.56, + "grad_norm": 2.2334610776369295, + "learning_rate": 8.606657029758193e-06, + "loss": 0.9974, + "step": 23692 + }, + { + "epoch": 0.56, + "grad_norm": 2.207729649178786, + "learning_rate": 8.605901435914608e-06, + "loss": 1.0463, + "step": 23693 + }, + { + "epoch": 0.56, + "grad_norm": 1.9375924110726128, + "learning_rate": 8.605145850187904e-06, + "loss": 0.9806, + "step": 23694 + }, + { + "epoch": 0.56, + "grad_norm": 2.1437281377650033, + "learning_rate": 8.60439027258248e-06, + "loss": 1.0634, + "step": 23695 + }, + { + "epoch": 0.56, + "grad_norm": 2.364013153918398, + "learning_rate": 8.60363470310274e-06, + "loss": 0.9747, + "step": 23696 + }, + { + "epoch": 0.56, + "grad_norm": 1.0643026597743588, + "learning_rate": 8.602879141753084e-06, + "loss": 0.9609, + "step": 23697 + }, + { + "epoch": 0.56, + "grad_norm": 1.9232377590302012, + "learning_rate": 8.602123588537906e-06, + "loss": 0.9012, + "step": 23698 + }, + { + "epoch": 0.56, + "grad_norm": 2.2257287030713786, + "learning_rate": 8.601368043461604e-06, + "loss": 1.0478, + "step": 23699 + }, + { + "epoch": 0.56, + "grad_norm": 1.7033925600233912, + "learning_rate": 8.600612506528584e-06, + "loss": 1.01, + "step": 23700 + }, + { + "epoch": 0.56, + "grad_norm": 2.537089476377042, + "learning_rate": 8.59985697774324e-06, + "loss": 0.8793, + "step": 23701 + }, + { + "epoch": 0.56, + "grad_norm": 1.88503397256708, + "learning_rate": 8.599101457109971e-06, + "loss": 1.1421, + "step": 23702 + }, + { + "epoch": 0.56, + "grad_norm": 2.339695854659473, + "learning_rate": 8.598345944633174e-06, + "loss": 1.0843, + "step": 23703 + }, + { + "epoch": 0.56, + "grad_norm": 1.9897986526802511, + "learning_rate": 8.597590440317257e-06, + "loss": 0.9975, + "step": 23704 + }, + { + "epoch": 0.56, + "grad_norm": 2.4112149792773487, + "learning_rate": 8.596834944166608e-06, + "loss": 0.9886, + "step": 23705 + }, + { + "epoch": 0.56, + "grad_norm": 1.9841082511684947, + "learning_rate": 8.59607945618563e-06, + "loss": 1.0638, + "step": 23706 + }, + { + "epoch": 0.56, + "grad_norm": 1.9161058460195548, + "learning_rate": 8.595323976378727e-06, + "loss": 1.0391, + "step": 23707 + }, + { + "epoch": 0.56, + "grad_norm": 1.8148230553084257, + "learning_rate": 8.594568504750287e-06, + "loss": 0.8349, + "step": 23708 + }, + { + "epoch": 0.56, + "grad_norm": 1.9430052019990078, + "learning_rate": 8.593813041304716e-06, + "loss": 0.9676, + "step": 23709 + }, + { + "epoch": 0.56, + "grad_norm": 2.0535928967310566, + "learning_rate": 8.593057586046409e-06, + "loss": 1.0055, + "step": 23710 + }, + { + "epoch": 0.56, + "grad_norm": 1.8908167602149875, + "learning_rate": 8.592302138979773e-06, + "loss": 1.0004, + "step": 23711 + }, + { + "epoch": 0.56, + "grad_norm": 2.314153114724585, + "learning_rate": 8.59154670010919e-06, + "loss": 1.027, + "step": 23712 + }, + { + "epoch": 0.56, + "grad_norm": 1.9498622921076092, + "learning_rate": 8.590791269439074e-06, + "loss": 0.8825, + "step": 23713 + }, + { + "epoch": 0.56, + "grad_norm": 1.9815431058862092, + "learning_rate": 8.590035846973817e-06, + "loss": 0.9422, + "step": 23714 + }, + { + "epoch": 0.56, + "grad_norm": 2.269265439164483, + "learning_rate": 8.589280432717814e-06, + "loss": 1.0386, + "step": 23715 + }, + { + "epoch": 0.56, + "grad_norm": 1.9272974751381071, + "learning_rate": 8.58852502667547e-06, + "loss": 1.0249, + "step": 23716 + }, + { + "epoch": 0.56, + "grad_norm": 1.1352690444470215, + "learning_rate": 8.58776962885118e-06, + "loss": 0.9953, + "step": 23717 + }, + { + "epoch": 0.56, + "grad_norm": 2.1720322240723418, + "learning_rate": 8.587014239249344e-06, + "loss": 1.0284, + "step": 23718 + }, + { + "epoch": 0.56, + "grad_norm": 2.011865934987457, + "learning_rate": 8.586258857874355e-06, + "loss": 1.0609, + "step": 23719 + }, + { + "epoch": 0.56, + "grad_norm": 1.811403171320668, + "learning_rate": 8.585503484730616e-06, + "loss": 0.9242, + "step": 23720 + }, + { + "epoch": 0.56, + "grad_norm": 1.9949650209293999, + "learning_rate": 8.584748119822525e-06, + "loss": 0.9718, + "step": 23721 + }, + { + "epoch": 0.56, + "grad_norm": 1.9842772526630474, + "learning_rate": 8.583992763154476e-06, + "loss": 0.977, + "step": 23722 + }, + { + "epoch": 0.56, + "grad_norm": 1.7292297360426794, + "learning_rate": 8.583237414730869e-06, + "loss": 1.0525, + "step": 23723 + }, + { + "epoch": 0.56, + "grad_norm": 2.1662551119300937, + "learning_rate": 8.582482074556103e-06, + "loss": 1.0013, + "step": 23724 + }, + { + "epoch": 0.56, + "grad_norm": 1.1436864736655097, + "learning_rate": 8.581726742634579e-06, + "loss": 0.968, + "step": 23725 + }, + { + "epoch": 0.56, + "grad_norm": 2.135559521622487, + "learning_rate": 8.580971418970687e-06, + "loss": 0.9348, + "step": 23726 + }, + { + "epoch": 0.56, + "grad_norm": 0.9931128310743658, + "learning_rate": 8.580216103568828e-06, + "loss": 0.9383, + "step": 23727 + }, + { + "epoch": 0.56, + "grad_norm": 2.102922288596961, + "learning_rate": 8.579460796433403e-06, + "loss": 1.0056, + "step": 23728 + }, + { + "epoch": 0.56, + "grad_norm": 2.2372870809835517, + "learning_rate": 8.578705497568807e-06, + "loss": 0.9745, + "step": 23729 + }, + { + "epoch": 0.56, + "grad_norm": 2.0217435209972496, + "learning_rate": 8.577950206979436e-06, + "loss": 0.9113, + "step": 23730 + }, + { + "epoch": 0.56, + "grad_norm": 2.1338628317522623, + "learning_rate": 8.577194924669687e-06, + "loss": 1.0968, + "step": 23731 + }, + { + "epoch": 0.56, + "grad_norm": 1.9458671323383496, + "learning_rate": 8.576439650643964e-06, + "loss": 1.063, + "step": 23732 + }, + { + "epoch": 0.56, + "grad_norm": 1.134417329513535, + "learning_rate": 8.575684384906658e-06, + "loss": 0.9478, + "step": 23733 + }, + { + "epoch": 0.56, + "grad_norm": 2.039423427847798, + "learning_rate": 8.574929127462168e-06, + "loss": 1.032, + "step": 23734 + }, + { + "epoch": 0.56, + "grad_norm": 1.8685766043513574, + "learning_rate": 8.574173878314895e-06, + "loss": 1.0629, + "step": 23735 + }, + { + "epoch": 0.56, + "grad_norm": 2.0598238981753147, + "learning_rate": 8.573418637469228e-06, + "loss": 1.0342, + "step": 23736 + }, + { + "epoch": 0.56, + "grad_norm": 1.9638774426731707, + "learning_rate": 8.572663404929573e-06, + "loss": 1.1505, + "step": 23737 + }, + { + "epoch": 0.56, + "grad_norm": 1.749662446830836, + "learning_rate": 8.57190818070032e-06, + "loss": 1.0234, + "step": 23738 + }, + { + "epoch": 0.56, + "grad_norm": 1.1331183939262928, + "learning_rate": 8.571152964785876e-06, + "loss": 0.9829, + "step": 23739 + }, + { + "epoch": 0.56, + "grad_norm": 1.8257754863365243, + "learning_rate": 8.570397757190624e-06, + "loss": 1.0989, + "step": 23740 + }, + { + "epoch": 0.56, + "grad_norm": 1.9089837437760964, + "learning_rate": 8.569642557918974e-06, + "loss": 1.0305, + "step": 23741 + }, + { + "epoch": 0.56, + "grad_norm": 2.119558901312916, + "learning_rate": 8.568887366975318e-06, + "loss": 0.9771, + "step": 23742 + }, + { + "epoch": 0.56, + "grad_norm": 1.8809906912083822, + "learning_rate": 8.56813218436405e-06, + "loss": 1.0101, + "step": 23743 + }, + { + "epoch": 0.56, + "grad_norm": 2.5279684006040983, + "learning_rate": 8.567377010089571e-06, + "loss": 1.0692, + "step": 23744 + }, + { + "epoch": 0.56, + "grad_norm": 1.8938385176252501, + "learning_rate": 8.566621844156275e-06, + "loss": 0.9403, + "step": 23745 + }, + { + "epoch": 0.56, + "grad_norm": 2.244147890426123, + "learning_rate": 8.565866686568565e-06, + "loss": 1.0167, + "step": 23746 + }, + { + "epoch": 0.56, + "grad_norm": 1.949561236450175, + "learning_rate": 8.565111537330831e-06, + "loss": 1.0732, + "step": 23747 + }, + { + "epoch": 0.56, + "grad_norm": 2.0572805176190885, + "learning_rate": 8.56435639644747e-06, + "loss": 1.0585, + "step": 23748 + }, + { + "epoch": 0.56, + "grad_norm": 2.0110257921633927, + "learning_rate": 8.563601263922883e-06, + "loss": 1.1141, + "step": 23749 + }, + { + "epoch": 0.56, + "grad_norm": 2.2275051409535482, + "learning_rate": 8.562846139761462e-06, + "loss": 1.0923, + "step": 23750 + }, + { + "epoch": 0.56, + "grad_norm": 2.0012138963895154, + "learning_rate": 8.562091023967606e-06, + "loss": 1.0117, + "step": 23751 + }, + { + "epoch": 0.56, + "grad_norm": 2.139665965383234, + "learning_rate": 8.56133591654571e-06, + "loss": 0.8837, + "step": 23752 + }, + { + "epoch": 0.56, + "grad_norm": 2.282899520665833, + "learning_rate": 8.560580817500177e-06, + "loss": 0.9405, + "step": 23753 + }, + { + "epoch": 0.56, + "grad_norm": 2.017358986576009, + "learning_rate": 8.559825726835395e-06, + "loss": 0.9525, + "step": 23754 + }, + { + "epoch": 0.56, + "grad_norm": 6.719984767909494, + "learning_rate": 8.559070644555761e-06, + "loss": 1.0904, + "step": 23755 + }, + { + "epoch": 0.56, + "grad_norm": 1.9829739348067843, + "learning_rate": 8.55831557066568e-06, + "loss": 0.9968, + "step": 23756 + }, + { + "epoch": 0.56, + "grad_norm": 2.228518652976584, + "learning_rate": 8.557560505169538e-06, + "loss": 0.9672, + "step": 23757 + }, + { + "epoch": 0.56, + "grad_norm": 1.918041368569393, + "learning_rate": 8.556805448071736e-06, + "loss": 1.1018, + "step": 23758 + }, + { + "epoch": 0.56, + "grad_norm": 2.0450243721535752, + "learning_rate": 8.556050399376667e-06, + "loss": 0.8723, + "step": 23759 + }, + { + "epoch": 0.56, + "grad_norm": 2.922392164532182, + "learning_rate": 8.555295359088736e-06, + "loss": 0.9156, + "step": 23760 + }, + { + "epoch": 0.56, + "grad_norm": 1.8353974567694231, + "learning_rate": 8.554540327212328e-06, + "loss": 1.0749, + "step": 23761 + }, + { + "epoch": 0.56, + "grad_norm": 2.093611328649036, + "learning_rate": 8.553785303751844e-06, + "loss": 1.0675, + "step": 23762 + }, + { + "epoch": 0.56, + "grad_norm": 2.2111095760202106, + "learning_rate": 8.553030288711683e-06, + "loss": 0.9941, + "step": 23763 + }, + { + "epoch": 0.56, + "grad_norm": 2.176137501124677, + "learning_rate": 8.552275282096233e-06, + "loss": 1.1487, + "step": 23764 + }, + { + "epoch": 0.56, + "grad_norm": 1.8760718642139556, + "learning_rate": 8.551520283909898e-06, + "loss": 1.0671, + "step": 23765 + }, + { + "epoch": 0.56, + "grad_norm": 1.9148411382253088, + "learning_rate": 8.550765294157069e-06, + "loss": 1.0107, + "step": 23766 + }, + { + "epoch": 0.56, + "grad_norm": 2.25668779061895, + "learning_rate": 8.550010312842148e-06, + "loss": 0.973, + "step": 23767 + }, + { + "epoch": 0.56, + "grad_norm": 4.201867456243071, + "learning_rate": 8.54925533996952e-06, + "loss": 0.9062, + "step": 23768 + }, + { + "epoch": 0.56, + "grad_norm": 1.1610733243375182, + "learning_rate": 8.548500375543587e-06, + "loss": 0.9673, + "step": 23769 + }, + { + "epoch": 0.56, + "grad_norm": 1.8498598986995465, + "learning_rate": 8.547745419568749e-06, + "loss": 1.0127, + "step": 23770 + }, + { + "epoch": 0.56, + "grad_norm": 2.1393233982319946, + "learning_rate": 8.546990472049394e-06, + "loss": 0.9647, + "step": 23771 + }, + { + "epoch": 0.56, + "grad_norm": 2.095997045833384, + "learning_rate": 8.546235532989919e-06, + "loss": 1.0483, + "step": 23772 + }, + { + "epoch": 0.56, + "grad_norm": 2.0633436483804504, + "learning_rate": 8.545480602394721e-06, + "loss": 1.0076, + "step": 23773 + }, + { + "epoch": 0.56, + "grad_norm": 1.9695042757557513, + "learning_rate": 8.544725680268199e-06, + "loss": 1.139, + "step": 23774 + }, + { + "epoch": 0.56, + "grad_norm": 2.1467361186506873, + "learning_rate": 8.543970766614743e-06, + "loss": 1.1566, + "step": 23775 + }, + { + "epoch": 0.56, + "grad_norm": 1.9569831232379808, + "learning_rate": 8.543215861438749e-06, + "loss": 1.0583, + "step": 23776 + }, + { + "epoch": 0.56, + "grad_norm": 1.8037651871731164, + "learning_rate": 8.542460964744615e-06, + "loss": 0.8712, + "step": 23777 + }, + { + "epoch": 0.56, + "grad_norm": 2.780789231155042, + "learning_rate": 8.541706076536735e-06, + "loss": 0.957, + "step": 23778 + }, + { + "epoch": 0.56, + "grad_norm": 1.2070887617110029, + "learning_rate": 8.540951196819502e-06, + "loss": 1.0138, + "step": 23779 + }, + { + "epoch": 0.56, + "grad_norm": 1.8336567156922727, + "learning_rate": 8.540196325597312e-06, + "loss": 1.0146, + "step": 23780 + }, + { + "epoch": 0.56, + "grad_norm": 1.147965517905963, + "learning_rate": 8.539441462874566e-06, + "loss": 0.9418, + "step": 23781 + }, + { + "epoch": 0.56, + "grad_norm": 2.03778584637583, + "learning_rate": 8.538686608655652e-06, + "loss": 1.005, + "step": 23782 + }, + { + "epoch": 0.56, + "grad_norm": 2.1357199789173715, + "learning_rate": 8.537931762944966e-06, + "loss": 0.9136, + "step": 23783 + }, + { + "epoch": 0.56, + "grad_norm": 2.582091891558452, + "learning_rate": 8.537176925746902e-06, + "loss": 0.9009, + "step": 23784 + }, + { + "epoch": 0.56, + "grad_norm": 1.9780634127209595, + "learning_rate": 8.536422097065863e-06, + "loss": 0.9023, + "step": 23785 + }, + { + "epoch": 0.56, + "grad_norm": 2.1568622199407805, + "learning_rate": 8.535667276906234e-06, + "loss": 0.9997, + "step": 23786 + }, + { + "epoch": 0.56, + "grad_norm": 2.5145589589430393, + "learning_rate": 8.534912465272414e-06, + "loss": 1.0461, + "step": 23787 + }, + { + "epoch": 0.56, + "grad_norm": 1.9733274371406127, + "learning_rate": 8.5341576621688e-06, + "loss": 1.0617, + "step": 23788 + }, + { + "epoch": 0.56, + "grad_norm": 1.991291005361516, + "learning_rate": 8.53340286759978e-06, + "loss": 1.053, + "step": 23789 + }, + { + "epoch": 0.56, + "grad_norm": 2.023428059085524, + "learning_rate": 8.532648081569756e-06, + "loss": 0.9455, + "step": 23790 + }, + { + "epoch": 0.56, + "grad_norm": 1.808252212617721, + "learning_rate": 8.531893304083117e-06, + "loss": 0.8794, + "step": 23791 + }, + { + "epoch": 0.56, + "grad_norm": 1.8935200530421552, + "learning_rate": 8.531138535144265e-06, + "loss": 0.9465, + "step": 23792 + }, + { + "epoch": 0.56, + "grad_norm": 1.9087383222291041, + "learning_rate": 8.530383774757583e-06, + "loss": 1.0072, + "step": 23793 + }, + { + "epoch": 0.56, + "grad_norm": 1.721233200620941, + "learning_rate": 8.529629022927474e-06, + "loss": 0.9528, + "step": 23794 + }, + { + "epoch": 0.56, + "grad_norm": 1.8037687159683031, + "learning_rate": 8.528874279658335e-06, + "loss": 1.0414, + "step": 23795 + }, + { + "epoch": 0.56, + "grad_norm": 2.0050517312764615, + "learning_rate": 8.528119544954551e-06, + "loss": 0.9626, + "step": 23796 + }, + { + "epoch": 0.56, + "grad_norm": 2.048712030727197, + "learning_rate": 8.52736481882052e-06, + "loss": 1.0599, + "step": 23797 + }, + { + "epoch": 0.56, + "grad_norm": 2.0840241804674133, + "learning_rate": 8.526610101260639e-06, + "loss": 1.0424, + "step": 23798 + }, + { + "epoch": 0.56, + "grad_norm": 2.1385154777980144, + "learning_rate": 8.525855392279302e-06, + "loss": 1.0047, + "step": 23799 + }, + { + "epoch": 0.56, + "grad_norm": 1.1706345680809906, + "learning_rate": 8.525100691880898e-06, + "loss": 0.9998, + "step": 23800 + }, + { + "epoch": 0.56, + "grad_norm": 1.7413631241390586, + "learning_rate": 8.524346000069828e-06, + "loss": 1.0349, + "step": 23801 + }, + { + "epoch": 0.56, + "grad_norm": 1.7560519181431116, + "learning_rate": 8.523591316850483e-06, + "loss": 1.013, + "step": 23802 + }, + { + "epoch": 0.56, + "grad_norm": 1.7757123393633656, + "learning_rate": 8.522836642227254e-06, + "loss": 1.0624, + "step": 23803 + }, + { + "epoch": 0.56, + "grad_norm": 2.099811950603084, + "learning_rate": 8.522081976204539e-06, + "loss": 0.9054, + "step": 23804 + }, + { + "epoch": 0.56, + "grad_norm": 1.0535725332750192, + "learning_rate": 8.52132731878673e-06, + "loss": 0.881, + "step": 23805 + }, + { + "epoch": 0.56, + "grad_norm": 1.8554539021098697, + "learning_rate": 8.520572669978224e-06, + "loss": 0.9807, + "step": 23806 + }, + { + "epoch": 0.56, + "grad_norm": 1.9331545458598343, + "learning_rate": 8.519818029783412e-06, + "loss": 0.8886, + "step": 23807 + }, + { + "epoch": 0.56, + "grad_norm": 2.1793544849877517, + "learning_rate": 8.519063398206684e-06, + "loss": 0.9027, + "step": 23808 + }, + { + "epoch": 0.56, + "grad_norm": 2.8970625075977927, + "learning_rate": 8.518308775252444e-06, + "loss": 0.9267, + "step": 23809 + }, + { + "epoch": 0.56, + "grad_norm": 1.0982903199932879, + "learning_rate": 8.517554160925073e-06, + "loss": 0.9349, + "step": 23810 + }, + { + "epoch": 0.56, + "grad_norm": 1.042352332273872, + "learning_rate": 8.516799555228974e-06, + "loss": 0.9949, + "step": 23811 + }, + { + "epoch": 0.56, + "grad_norm": 1.0552419611271313, + "learning_rate": 8.516044958168535e-06, + "loss": 0.9791, + "step": 23812 + }, + { + "epoch": 0.56, + "grad_norm": 2.085985265878741, + "learning_rate": 8.515290369748158e-06, + "loss": 1.1275, + "step": 23813 + }, + { + "epoch": 0.56, + "grad_norm": 1.9053287075263567, + "learning_rate": 8.514535789972226e-06, + "loss": 0.9427, + "step": 23814 + }, + { + "epoch": 0.56, + "grad_norm": 1.7998148880476483, + "learning_rate": 8.513781218845137e-06, + "loss": 0.9806, + "step": 23815 + }, + { + "epoch": 0.56, + "grad_norm": 2.3754349069391676, + "learning_rate": 8.513026656371288e-06, + "loss": 1.0554, + "step": 23816 + }, + { + "epoch": 0.56, + "grad_norm": 2.1120836253146047, + "learning_rate": 8.512272102555063e-06, + "loss": 1.0031, + "step": 23817 + }, + { + "epoch": 0.56, + "grad_norm": 1.9146720079998365, + "learning_rate": 8.511517557400864e-06, + "loss": 1.0723, + "step": 23818 + }, + { + "epoch": 0.56, + "grad_norm": 2.37418328201768, + "learning_rate": 8.51076302091308e-06, + "loss": 0.9074, + "step": 23819 + }, + { + "epoch": 0.56, + "grad_norm": 1.2403289071799022, + "learning_rate": 8.510008493096109e-06, + "loss": 0.9709, + "step": 23820 + }, + { + "epoch": 0.56, + "grad_norm": 1.77447167737594, + "learning_rate": 8.509253973954334e-06, + "loss": 0.9238, + "step": 23821 + }, + { + "epoch": 0.56, + "grad_norm": 2.066232416276782, + "learning_rate": 8.508499463492157e-06, + "loss": 1.0685, + "step": 23822 + }, + { + "epoch": 0.56, + "grad_norm": 2.34064414109793, + "learning_rate": 8.507744961713972e-06, + "loss": 0.9698, + "step": 23823 + }, + { + "epoch": 0.56, + "grad_norm": 1.9439128633862988, + "learning_rate": 8.506990468624164e-06, + "loss": 0.9491, + "step": 23824 + }, + { + "epoch": 0.56, + "grad_norm": 1.1074729329277595, + "learning_rate": 8.506235984227129e-06, + "loss": 0.9417, + "step": 23825 + }, + { + "epoch": 0.56, + "grad_norm": 2.0536942124195896, + "learning_rate": 8.505481508527263e-06, + "loss": 1.0383, + "step": 23826 + }, + { + "epoch": 0.56, + "grad_norm": 2.179602336651187, + "learning_rate": 8.504727041528959e-06, + "loss": 1.0304, + "step": 23827 + }, + { + "epoch": 0.56, + "grad_norm": 2.170548757444347, + "learning_rate": 8.503972583236605e-06, + "loss": 0.9964, + "step": 23828 + }, + { + "epoch": 0.56, + "grad_norm": 1.9668380980755262, + "learning_rate": 8.503218133654595e-06, + "loss": 1.1781, + "step": 23829 + }, + { + "epoch": 0.56, + "grad_norm": 1.8716947038071625, + "learning_rate": 8.502463692787327e-06, + "loss": 1.0354, + "step": 23830 + }, + { + "epoch": 0.56, + "grad_norm": 1.9135328611054663, + "learning_rate": 8.501709260639187e-06, + "loss": 1.0717, + "step": 23831 + }, + { + "epoch": 0.56, + "grad_norm": 1.8678611476388276, + "learning_rate": 8.50095483721457e-06, + "loss": 1.1067, + "step": 23832 + }, + { + "epoch": 0.56, + "grad_norm": 2.159762866207658, + "learning_rate": 8.500200422517866e-06, + "loss": 0.8832, + "step": 23833 + }, + { + "epoch": 0.56, + "grad_norm": 2.143522213720285, + "learning_rate": 8.499446016553475e-06, + "loss": 1.1385, + "step": 23834 + }, + { + "epoch": 0.56, + "grad_norm": 1.9672850744692894, + "learning_rate": 8.498691619325782e-06, + "loss": 1.0038, + "step": 23835 + }, + { + "epoch": 0.56, + "grad_norm": 2.0480624381626713, + "learning_rate": 8.497937230839182e-06, + "loss": 1.0998, + "step": 23836 + }, + { + "epoch": 0.56, + "grad_norm": 1.7910289059316085, + "learning_rate": 8.497182851098068e-06, + "loss": 1.1066, + "step": 23837 + }, + { + "epoch": 0.56, + "grad_norm": 2.0005190852519132, + "learning_rate": 8.496428480106828e-06, + "loss": 0.9059, + "step": 23838 + }, + { + "epoch": 0.56, + "grad_norm": 2.1435815291364313, + "learning_rate": 8.495674117869859e-06, + "loss": 1.0518, + "step": 23839 + }, + { + "epoch": 0.56, + "grad_norm": 1.9857115501268299, + "learning_rate": 8.494919764391548e-06, + "loss": 0.9664, + "step": 23840 + }, + { + "epoch": 0.56, + "grad_norm": 2.086844691246819, + "learning_rate": 8.4941654196763e-06, + "loss": 1.07, + "step": 23841 + }, + { + "epoch": 0.56, + "grad_norm": 2.25818413850757, + "learning_rate": 8.493411083728491e-06, + "loss": 0.9713, + "step": 23842 + }, + { + "epoch": 0.56, + "grad_norm": 2.222256001560431, + "learning_rate": 8.49265675655252e-06, + "loss": 0.9654, + "step": 23843 + }, + { + "epoch": 0.56, + "grad_norm": 1.0800146654225236, + "learning_rate": 8.49190243815278e-06, + "loss": 0.8919, + "step": 23844 + }, + { + "epoch": 0.56, + "grad_norm": 1.9642483322650257, + "learning_rate": 8.491148128533662e-06, + "loss": 0.9961, + "step": 23845 + }, + { + "epoch": 0.56, + "grad_norm": 1.9401610015165192, + "learning_rate": 8.490393827699554e-06, + "loss": 1.0883, + "step": 23846 + }, + { + "epoch": 0.56, + "grad_norm": 2.3493403197862994, + "learning_rate": 8.489639535654855e-06, + "loss": 0.9871, + "step": 23847 + }, + { + "epoch": 0.56, + "grad_norm": 2.220907832409213, + "learning_rate": 8.488885252403953e-06, + "loss": 0.924, + "step": 23848 + }, + { + "epoch": 0.56, + "grad_norm": 1.986706460660861, + "learning_rate": 8.488130977951235e-06, + "loss": 0.9866, + "step": 23849 + }, + { + "epoch": 0.56, + "grad_norm": 2.1505292928734776, + "learning_rate": 8.487376712301099e-06, + "loss": 1.0067, + "step": 23850 + }, + { + "epoch": 0.56, + "grad_norm": 1.917077380350634, + "learning_rate": 8.486622455457938e-06, + "loss": 0.8759, + "step": 23851 + }, + { + "epoch": 0.56, + "grad_norm": 1.8752955084629914, + "learning_rate": 8.485868207426136e-06, + "loss": 1.1635, + "step": 23852 + }, + { + "epoch": 0.56, + "grad_norm": 2.3380959523414866, + "learning_rate": 8.485113968210089e-06, + "loss": 1.1027, + "step": 23853 + }, + { + "epoch": 0.56, + "grad_norm": 1.9708497364575388, + "learning_rate": 8.484359737814189e-06, + "loss": 0.9204, + "step": 23854 + }, + { + "epoch": 0.56, + "grad_norm": 1.8044079303037506, + "learning_rate": 8.483605516242829e-06, + "loss": 0.9586, + "step": 23855 + }, + { + "epoch": 0.56, + "grad_norm": 2.1982589939861206, + "learning_rate": 8.482851303500395e-06, + "loss": 1.1108, + "step": 23856 + }, + { + "epoch": 0.56, + "grad_norm": 2.106269190627266, + "learning_rate": 8.48209709959128e-06, + "loss": 1.1095, + "step": 23857 + }, + { + "epoch": 0.56, + "grad_norm": 2.046227969736534, + "learning_rate": 8.481342904519882e-06, + "loss": 0.9304, + "step": 23858 + }, + { + "epoch": 0.56, + "grad_norm": 1.7842538769387513, + "learning_rate": 8.480588718290579e-06, + "loss": 1.085, + "step": 23859 + }, + { + "epoch": 0.56, + "grad_norm": 2.3724977771456377, + "learning_rate": 8.479834540907774e-06, + "loss": 0.9714, + "step": 23860 + }, + { + "epoch": 0.56, + "grad_norm": 1.1436980835181119, + "learning_rate": 8.479080372375848e-06, + "loss": 0.9708, + "step": 23861 + }, + { + "epoch": 0.56, + "grad_norm": 2.0603076867176164, + "learning_rate": 8.478326212699205e-06, + "loss": 1.0594, + "step": 23862 + }, + { + "epoch": 0.56, + "grad_norm": 1.9737370463857111, + "learning_rate": 8.477572061882225e-06, + "loss": 1.0516, + "step": 23863 + }, + { + "epoch": 0.56, + "grad_norm": 1.113887279345344, + "learning_rate": 8.476817919929304e-06, + "loss": 0.9409, + "step": 23864 + }, + { + "epoch": 0.56, + "grad_norm": 2.189038024213457, + "learning_rate": 8.47606378684483e-06, + "loss": 1.1248, + "step": 23865 + }, + { + "epoch": 0.56, + "grad_norm": 1.17926871241082, + "learning_rate": 8.475309662633195e-06, + "loss": 0.9447, + "step": 23866 + }, + { + "epoch": 0.56, + "grad_norm": 2.0765186071421318, + "learning_rate": 8.474555547298789e-06, + "loss": 1.0793, + "step": 23867 + }, + { + "epoch": 0.56, + "grad_norm": 1.9086335696252488, + "learning_rate": 8.473801440846005e-06, + "loss": 1.0497, + "step": 23868 + }, + { + "epoch": 0.56, + "grad_norm": 1.78951216274814, + "learning_rate": 8.473047343279233e-06, + "loss": 0.948, + "step": 23869 + }, + { + "epoch": 0.56, + "grad_norm": 2.0060538840215676, + "learning_rate": 8.472293254602858e-06, + "loss": 1.0013, + "step": 23870 + }, + { + "epoch": 0.56, + "grad_norm": 1.9672544197186668, + "learning_rate": 8.47153917482128e-06, + "loss": 0.971, + "step": 23871 + }, + { + "epoch": 0.56, + "grad_norm": 1.88921522459311, + "learning_rate": 8.470785103938887e-06, + "loss": 0.9808, + "step": 23872 + }, + { + "epoch": 0.56, + "grad_norm": 3.041394871563796, + "learning_rate": 8.470031041960062e-06, + "loss": 0.8994, + "step": 23873 + }, + { + "epoch": 0.56, + "grad_norm": 1.9300407199689016, + "learning_rate": 8.469276988889202e-06, + "loss": 1.0164, + "step": 23874 + }, + { + "epoch": 0.56, + "grad_norm": 2.346772893522268, + "learning_rate": 8.468522944730697e-06, + "loss": 1.0172, + "step": 23875 + }, + { + "epoch": 0.56, + "grad_norm": 2.014873805687205, + "learning_rate": 8.467768909488938e-06, + "loss": 1.0355, + "step": 23876 + }, + { + "epoch": 0.56, + "grad_norm": 2.0849405332366455, + "learning_rate": 8.46701488316831e-06, + "loss": 1.091, + "step": 23877 + }, + { + "epoch": 0.56, + "grad_norm": 1.9903546219542199, + "learning_rate": 8.466260865773207e-06, + "loss": 0.9884, + "step": 23878 + }, + { + "epoch": 0.56, + "grad_norm": 1.9453147871174448, + "learning_rate": 8.465506857308021e-06, + "loss": 0.9214, + "step": 23879 + }, + { + "epoch": 0.56, + "grad_norm": 2.1306876340401995, + "learning_rate": 8.46475285777714e-06, + "loss": 1.0548, + "step": 23880 + }, + { + "epoch": 0.56, + "grad_norm": 1.9390680275568533, + "learning_rate": 8.463998867184952e-06, + "loss": 0.9734, + "step": 23881 + }, + { + "epoch": 0.56, + "grad_norm": 1.9732654016964841, + "learning_rate": 8.463244885535849e-06, + "loss": 1.0755, + "step": 23882 + }, + { + "epoch": 0.56, + "grad_norm": 2.2513731272028292, + "learning_rate": 8.462490912834222e-06, + "loss": 1.105, + "step": 23883 + }, + { + "epoch": 0.56, + "grad_norm": 1.9670567836794697, + "learning_rate": 8.461736949084458e-06, + "loss": 1.1106, + "step": 23884 + }, + { + "epoch": 0.56, + "grad_norm": 1.8384220731208902, + "learning_rate": 8.460982994290951e-06, + "loss": 1.0407, + "step": 23885 + }, + { + "epoch": 0.56, + "grad_norm": 1.9965335368235197, + "learning_rate": 8.46022904845809e-06, + "loss": 1.0793, + "step": 23886 + }, + { + "epoch": 0.56, + "grad_norm": 1.7203382881027514, + "learning_rate": 8.459475111590256e-06, + "loss": 0.8995, + "step": 23887 + }, + { + "epoch": 0.56, + "grad_norm": 1.1664164812980053, + "learning_rate": 8.45872118369185e-06, + "loss": 0.9526, + "step": 23888 + }, + { + "epoch": 0.56, + "grad_norm": 1.9855683218244053, + "learning_rate": 8.457967264767253e-06, + "loss": 0.9769, + "step": 23889 + }, + { + "epoch": 0.56, + "grad_norm": 1.9823202948078278, + "learning_rate": 8.457213354820868e-06, + "loss": 1.0374, + "step": 23890 + }, + { + "epoch": 0.56, + "grad_norm": 1.826219146927558, + "learning_rate": 8.456459453857067e-06, + "loss": 0.9689, + "step": 23891 + }, + { + "epoch": 0.56, + "grad_norm": 2.047763979091263, + "learning_rate": 8.45570556188025e-06, + "loss": 1.0016, + "step": 23892 + }, + { + "epoch": 0.56, + "grad_norm": 1.966297309716175, + "learning_rate": 8.454951678894807e-06, + "loss": 0.9635, + "step": 23893 + }, + { + "epoch": 0.56, + "grad_norm": 2.204782033487134, + "learning_rate": 8.454197804905123e-06, + "loss": 0.9844, + "step": 23894 + }, + { + "epoch": 0.56, + "grad_norm": 1.9957954321224622, + "learning_rate": 8.453443939915586e-06, + "loss": 0.8729, + "step": 23895 + }, + { + "epoch": 0.56, + "grad_norm": 1.965347439841766, + "learning_rate": 8.45269008393059e-06, + "loss": 0.9664, + "step": 23896 + }, + { + "epoch": 0.56, + "grad_norm": 1.9515798890789482, + "learning_rate": 8.451936236954524e-06, + "loss": 0.8688, + "step": 23897 + }, + { + "epoch": 0.56, + "grad_norm": 2.163168098426413, + "learning_rate": 8.45118239899177e-06, + "loss": 0.9827, + "step": 23898 + }, + { + "epoch": 0.56, + "grad_norm": 2.235759731589463, + "learning_rate": 8.450428570046725e-06, + "loss": 1.0538, + "step": 23899 + }, + { + "epoch": 0.56, + "grad_norm": 2.2228158490623864, + "learning_rate": 8.44967475012378e-06, + "loss": 1.0249, + "step": 23900 + }, + { + "epoch": 0.56, + "grad_norm": 1.9421904120035969, + "learning_rate": 8.448920939227316e-06, + "loss": 1.0743, + "step": 23901 + }, + { + "epoch": 0.56, + "grad_norm": 1.8651823567082269, + "learning_rate": 8.448167137361723e-06, + "loss": 0.9796, + "step": 23902 + }, + { + "epoch": 0.56, + "grad_norm": 1.1081152035225164, + "learning_rate": 8.447413344531394e-06, + "loss": 0.9348, + "step": 23903 + }, + { + "epoch": 0.56, + "grad_norm": 2.046709420699598, + "learning_rate": 8.446659560740718e-06, + "loss": 1.0154, + "step": 23904 + }, + { + "epoch": 0.56, + "grad_norm": 1.9275122728229377, + "learning_rate": 8.44590578599408e-06, + "loss": 0.8299, + "step": 23905 + }, + { + "epoch": 0.56, + "grad_norm": 1.9255578703914034, + "learning_rate": 8.445152020295869e-06, + "loss": 1.0691, + "step": 23906 + }, + { + "epoch": 0.56, + "grad_norm": 1.8853382929292783, + "learning_rate": 8.444398263650477e-06, + "loss": 1.0081, + "step": 23907 + }, + { + "epoch": 0.56, + "grad_norm": 2.029321406838481, + "learning_rate": 8.443644516062293e-06, + "loss": 0.9492, + "step": 23908 + }, + { + "epoch": 0.56, + "grad_norm": 2.1426466439804694, + "learning_rate": 8.4428907775357e-06, + "loss": 1.0227, + "step": 23909 + }, + { + "epoch": 0.56, + "grad_norm": 1.1320053551631482, + "learning_rate": 8.44213704807509e-06, + "loss": 0.9649, + "step": 23910 + }, + { + "epoch": 0.56, + "grad_norm": 1.9262091813345672, + "learning_rate": 8.441383327684854e-06, + "loss": 1.0361, + "step": 23911 + }, + { + "epoch": 0.56, + "grad_norm": 1.9146159084852392, + "learning_rate": 8.440629616369375e-06, + "loss": 1.0729, + "step": 23912 + }, + { + "epoch": 0.56, + "grad_norm": 2.024778856063202, + "learning_rate": 8.439875914133044e-06, + "loss": 1.1293, + "step": 23913 + }, + { + "epoch": 0.56, + "grad_norm": 2.006099837992559, + "learning_rate": 8.439122220980249e-06, + "loss": 0.9807, + "step": 23914 + }, + { + "epoch": 0.56, + "grad_norm": 1.8931049214356925, + "learning_rate": 8.438368536915382e-06, + "loss": 0.9657, + "step": 23915 + }, + { + "epoch": 0.56, + "grad_norm": 2.1095641677747707, + "learning_rate": 8.437614861942826e-06, + "loss": 1.013, + "step": 23916 + }, + { + "epoch": 0.56, + "grad_norm": 2.118689672550494, + "learning_rate": 8.43686119606697e-06, + "loss": 1.0302, + "step": 23917 + }, + { + "epoch": 0.56, + "grad_norm": 2.0414500819047037, + "learning_rate": 8.436107539292206e-06, + "loss": 1.0806, + "step": 23918 + }, + { + "epoch": 0.56, + "grad_norm": 1.1332217201857517, + "learning_rate": 8.435353891622915e-06, + "loss": 0.9522, + "step": 23919 + }, + { + "epoch": 0.56, + "grad_norm": 1.8509723644238167, + "learning_rate": 8.434600253063489e-06, + "loss": 0.9086, + "step": 23920 + }, + { + "epoch": 0.56, + "grad_norm": 1.9278377250531071, + "learning_rate": 8.433846623618319e-06, + "loss": 0.9554, + "step": 23921 + }, + { + "epoch": 0.56, + "grad_norm": 2.1867599382828167, + "learning_rate": 8.43309300329179e-06, + "loss": 0.9076, + "step": 23922 + }, + { + "epoch": 0.56, + "grad_norm": 2.130117018821308, + "learning_rate": 8.432339392088286e-06, + "loss": 1.0784, + "step": 23923 + }, + { + "epoch": 0.56, + "grad_norm": 1.8458264468637675, + "learning_rate": 8.4315857900122e-06, + "loss": 1.0083, + "step": 23924 + }, + { + "epoch": 0.56, + "grad_norm": 2.212433834984175, + "learning_rate": 8.43083219706792e-06, + "loss": 1.0148, + "step": 23925 + }, + { + "epoch": 0.56, + "grad_norm": 2.215603744293234, + "learning_rate": 8.430078613259832e-06, + "loss": 0.9322, + "step": 23926 + }, + { + "epoch": 0.56, + "grad_norm": 2.0598889899526984, + "learning_rate": 8.42932503859232e-06, + "loss": 1.1262, + "step": 23927 + }, + { + "epoch": 0.56, + "grad_norm": 2.0923282801347005, + "learning_rate": 8.428571473069775e-06, + "loss": 1.0639, + "step": 23928 + }, + { + "epoch": 0.56, + "grad_norm": 2.697297221275978, + "learning_rate": 8.427817916696588e-06, + "loss": 0.8152, + "step": 23929 + }, + { + "epoch": 0.56, + "grad_norm": 2.240514218165531, + "learning_rate": 8.42706436947714e-06, + "loss": 0.9659, + "step": 23930 + }, + { + "epoch": 0.56, + "grad_norm": 1.0789924003225142, + "learning_rate": 8.42631083141582e-06, + "loss": 0.9513, + "step": 23931 + }, + { + "epoch": 0.56, + "grad_norm": 1.8547414120796144, + "learning_rate": 8.42555730251702e-06, + "loss": 0.9772, + "step": 23932 + }, + { + "epoch": 0.56, + "grad_norm": 2.293865047821141, + "learning_rate": 8.424803782785123e-06, + "loss": 1.0598, + "step": 23933 + }, + { + "epoch": 0.56, + "grad_norm": 1.966460289016536, + "learning_rate": 8.424050272224515e-06, + "loss": 0.9529, + "step": 23934 + }, + { + "epoch": 0.56, + "grad_norm": 3.2824990954009956, + "learning_rate": 8.423296770839584e-06, + "loss": 0.8758, + "step": 23935 + }, + { + "epoch": 0.56, + "grad_norm": 1.9682543770291907, + "learning_rate": 8.422543278634726e-06, + "loss": 1.1135, + "step": 23936 + }, + { + "epoch": 0.56, + "grad_norm": 2.186395596148425, + "learning_rate": 8.421789795614315e-06, + "loss": 0.9916, + "step": 23937 + }, + { + "epoch": 0.56, + "grad_norm": 2.101793734964705, + "learning_rate": 8.421036321782741e-06, + "loss": 0.9619, + "step": 23938 + }, + { + "epoch": 0.56, + "grad_norm": 1.8675392107803024, + "learning_rate": 8.4202828571444e-06, + "loss": 0.9977, + "step": 23939 + }, + { + "epoch": 0.56, + "grad_norm": 2.856909170227283, + "learning_rate": 8.419529401703667e-06, + "loss": 1.142, + "step": 23940 + }, + { + "epoch": 0.56, + "grad_norm": 1.9291068043246054, + "learning_rate": 8.418775955464935e-06, + "loss": 1.1265, + "step": 23941 + }, + { + "epoch": 0.56, + "grad_norm": 2.046655862485538, + "learning_rate": 8.418022518432588e-06, + "loss": 1.1546, + "step": 23942 + }, + { + "epoch": 0.56, + "grad_norm": 1.0778870471158954, + "learning_rate": 8.417269090611023e-06, + "loss": 0.9988, + "step": 23943 + }, + { + "epoch": 0.56, + "grad_norm": 1.0542970937019849, + "learning_rate": 8.416515672004611e-06, + "loss": 1.0119, + "step": 23944 + }, + { + "epoch": 0.56, + "grad_norm": 3.2327262356970548, + "learning_rate": 8.415762262617748e-06, + "loss": 1.0455, + "step": 23945 + }, + { + "epoch": 0.56, + "grad_norm": 1.8721175155499272, + "learning_rate": 8.41500886245482e-06, + "loss": 0.9778, + "step": 23946 + }, + { + "epoch": 0.56, + "grad_norm": 1.0667588535038564, + "learning_rate": 8.41425547152021e-06, + "loss": 0.9403, + "step": 23947 + }, + { + "epoch": 0.56, + "grad_norm": 2.026990075260716, + "learning_rate": 8.413502089818307e-06, + "loss": 0.9, + "step": 23948 + }, + { + "epoch": 0.56, + "grad_norm": 2.0810850871571938, + "learning_rate": 8.412748717353499e-06, + "loss": 1.0577, + "step": 23949 + }, + { + "epoch": 0.56, + "grad_norm": 2.0739767181224105, + "learning_rate": 8.41199535413017e-06, + "loss": 1.0236, + "step": 23950 + }, + { + "epoch": 0.56, + "grad_norm": 1.9255423695098353, + "learning_rate": 8.411242000152705e-06, + "loss": 0.9698, + "step": 23951 + }, + { + "epoch": 0.56, + "grad_norm": 1.9429436830917597, + "learning_rate": 8.410488655425493e-06, + "loss": 0.9196, + "step": 23952 + }, + { + "epoch": 0.56, + "grad_norm": 1.0507474584978005, + "learning_rate": 8.40973531995292e-06, + "loss": 0.9624, + "step": 23953 + }, + { + "epoch": 0.56, + "grad_norm": 1.8304039934475602, + "learning_rate": 8.40898199373937e-06, + "loss": 0.8698, + "step": 23954 + }, + { + "epoch": 0.56, + "grad_norm": 1.8708071202313468, + "learning_rate": 8.40822867678923e-06, + "loss": 0.9973, + "step": 23955 + }, + { + "epoch": 0.56, + "grad_norm": 2.3787920918873606, + "learning_rate": 8.407475369106889e-06, + "loss": 1.0492, + "step": 23956 + }, + { + "epoch": 0.56, + "grad_norm": 1.1829652341039154, + "learning_rate": 8.406722070696731e-06, + "loss": 0.9839, + "step": 23957 + }, + { + "epoch": 0.56, + "grad_norm": 2.013462776368528, + "learning_rate": 8.405968781563141e-06, + "loss": 0.9774, + "step": 23958 + }, + { + "epoch": 0.56, + "grad_norm": 2.366479980965829, + "learning_rate": 8.405215501710501e-06, + "loss": 1.0005, + "step": 23959 + }, + { + "epoch": 0.56, + "grad_norm": 2.3026568245629804, + "learning_rate": 8.404462231143208e-06, + "loss": 1.0205, + "step": 23960 + }, + { + "epoch": 0.56, + "grad_norm": 1.9730407877078395, + "learning_rate": 8.403708969865638e-06, + "loss": 0.8901, + "step": 23961 + }, + { + "epoch": 0.56, + "grad_norm": 2.1123757793510403, + "learning_rate": 8.402955717882179e-06, + "loss": 1.0349, + "step": 23962 + }, + { + "epoch": 0.56, + "grad_norm": 2.0191006329338412, + "learning_rate": 8.402202475197216e-06, + "loss": 0.9414, + "step": 23963 + }, + { + "epoch": 0.56, + "grad_norm": 2.386708965745809, + "learning_rate": 8.40144924181514e-06, + "loss": 0.9759, + "step": 23964 + }, + { + "epoch": 0.56, + "grad_norm": 2.079984264987322, + "learning_rate": 8.400696017740329e-06, + "loss": 1.1081, + "step": 23965 + }, + { + "epoch": 0.56, + "grad_norm": 1.074320051011768, + "learning_rate": 8.399942802977175e-06, + "loss": 0.8789, + "step": 23966 + }, + { + "epoch": 0.56, + "grad_norm": 1.8988276498818877, + "learning_rate": 8.39918959753006e-06, + "loss": 1.1001, + "step": 23967 + }, + { + "epoch": 0.56, + "grad_norm": 2.0908587419144355, + "learning_rate": 8.398436401403368e-06, + "loss": 1.0948, + "step": 23968 + }, + { + "epoch": 0.56, + "grad_norm": 2.0494101061883225, + "learning_rate": 8.397683214601487e-06, + "loss": 0.9753, + "step": 23969 + }, + { + "epoch": 0.56, + "grad_norm": 2.10241199772555, + "learning_rate": 8.396930037128803e-06, + "loss": 1.0309, + "step": 23970 + }, + { + "epoch": 0.56, + "grad_norm": 1.0515595344831814, + "learning_rate": 8.3961768689897e-06, + "loss": 0.9405, + "step": 23971 + }, + { + "epoch": 0.56, + "grad_norm": 2.508618717175537, + "learning_rate": 8.395423710188561e-06, + "loss": 1.0289, + "step": 23972 + }, + { + "epoch": 0.56, + "grad_norm": 2.1083911113337863, + "learning_rate": 8.394670560729774e-06, + "loss": 0.9719, + "step": 23973 + }, + { + "epoch": 0.56, + "grad_norm": 1.0685976220888522, + "learning_rate": 8.393917420617726e-06, + "loss": 0.9544, + "step": 23974 + }, + { + "epoch": 0.56, + "grad_norm": 1.831355397263955, + "learning_rate": 8.393164289856797e-06, + "loss": 1.0154, + "step": 23975 + }, + { + "epoch": 0.56, + "grad_norm": 2.1137731206885033, + "learning_rate": 8.392411168451371e-06, + "loss": 1.0183, + "step": 23976 + }, + { + "epoch": 0.56, + "grad_norm": 1.1275426479248234, + "learning_rate": 8.39165805640584e-06, + "loss": 0.9051, + "step": 23977 + }, + { + "epoch": 0.56, + "grad_norm": 1.0917242997210517, + "learning_rate": 8.390904953724586e-06, + "loss": 0.9845, + "step": 23978 + }, + { + "epoch": 0.56, + "grad_norm": 2.1057018935427783, + "learning_rate": 8.390151860411991e-06, + "loss": 0.9863, + "step": 23979 + }, + { + "epoch": 0.56, + "grad_norm": 2.058461223458784, + "learning_rate": 8.389398776472439e-06, + "loss": 1.0562, + "step": 23980 + }, + { + "epoch": 0.56, + "grad_norm": 1.1160382173058654, + "learning_rate": 8.38864570191032e-06, + "loss": 1.0345, + "step": 23981 + }, + { + "epoch": 0.56, + "grad_norm": 2.0953093698892995, + "learning_rate": 8.387892636730016e-06, + "loss": 1.1119, + "step": 23982 + }, + { + "epoch": 0.57, + "grad_norm": 1.9448041828664464, + "learning_rate": 8.387139580935912e-06, + "loss": 0.9605, + "step": 23983 + }, + { + "epoch": 0.57, + "grad_norm": 1.7415552641248366, + "learning_rate": 8.386386534532388e-06, + "loss": 0.9994, + "step": 23984 + }, + { + "epoch": 0.57, + "grad_norm": 1.918340361888248, + "learning_rate": 8.385633497523838e-06, + "loss": 0.9605, + "step": 23985 + }, + { + "epoch": 0.57, + "grad_norm": 1.9658176349379093, + "learning_rate": 8.384880469914638e-06, + "loss": 0.9564, + "step": 23986 + }, + { + "epoch": 0.57, + "grad_norm": 2.2782646017870563, + "learning_rate": 8.384127451709173e-06, + "loss": 0.9876, + "step": 23987 + }, + { + "epoch": 0.57, + "grad_norm": 1.9702769286242507, + "learning_rate": 8.383374442911836e-06, + "loss": 0.9192, + "step": 23988 + }, + { + "epoch": 0.57, + "grad_norm": 1.982718001173709, + "learning_rate": 8.382621443527e-06, + "loss": 1.0488, + "step": 23989 + }, + { + "epoch": 0.57, + "grad_norm": 2.001595766864864, + "learning_rate": 8.381868453559053e-06, + "loss": 0.9911, + "step": 23990 + }, + { + "epoch": 0.57, + "grad_norm": 2.020988548556886, + "learning_rate": 8.381115473012381e-06, + "loss": 0.9151, + "step": 23991 + }, + { + "epoch": 0.57, + "grad_norm": 2.165792941457056, + "learning_rate": 8.380362501891373e-06, + "loss": 1.219, + "step": 23992 + }, + { + "epoch": 0.57, + "grad_norm": 1.9709405927877073, + "learning_rate": 8.3796095402004e-06, + "loss": 1.0975, + "step": 23993 + }, + { + "epoch": 0.57, + "grad_norm": 2.2108140881925933, + "learning_rate": 8.378856587943856e-06, + "loss": 1.1211, + "step": 23994 + }, + { + "epoch": 0.57, + "grad_norm": 2.3691553029299905, + "learning_rate": 8.378103645126124e-06, + "loss": 1.0165, + "step": 23995 + }, + { + "epoch": 0.57, + "grad_norm": 2.0358677677275745, + "learning_rate": 8.377350711751583e-06, + "loss": 1.091, + "step": 23996 + }, + { + "epoch": 0.57, + "grad_norm": 1.9083658039103846, + "learning_rate": 8.376597787824618e-06, + "loss": 0.9588, + "step": 23997 + }, + { + "epoch": 0.57, + "grad_norm": 1.9085913993260404, + "learning_rate": 8.37584487334962e-06, + "loss": 1.0009, + "step": 23998 + }, + { + "epoch": 0.57, + "grad_norm": 1.9029591490739595, + "learning_rate": 8.375091968330966e-06, + "loss": 1.0816, + "step": 23999 + }, + { + "epoch": 0.57, + "grad_norm": 2.151740704283439, + "learning_rate": 8.374339072773036e-06, + "loss": 1.0921, + "step": 24000 + }, + { + "epoch": 0.57, + "grad_norm": 1.964465903288773, + "learning_rate": 8.373586186680224e-06, + "loss": 0.974, + "step": 24001 + }, + { + "epoch": 0.57, + "grad_norm": 2.099189757066857, + "learning_rate": 8.37283331005691e-06, + "loss": 0.9355, + "step": 24002 + }, + { + "epoch": 0.57, + "grad_norm": 1.0851751638779994, + "learning_rate": 8.372080442907471e-06, + "loss": 1.0246, + "step": 24003 + }, + { + "epoch": 0.57, + "grad_norm": 1.1010870031980635, + "learning_rate": 8.371327585236294e-06, + "loss": 0.8964, + "step": 24004 + }, + { + "epoch": 0.57, + "grad_norm": 1.8497423097287258, + "learning_rate": 8.370574737047765e-06, + "loss": 0.9065, + "step": 24005 + }, + { + "epoch": 0.57, + "grad_norm": 1.7566574989687085, + "learning_rate": 8.369821898346269e-06, + "loss": 1.0303, + "step": 24006 + }, + { + "epoch": 0.57, + "grad_norm": 1.933194503603087, + "learning_rate": 8.369069069136184e-06, + "loss": 0.9323, + "step": 24007 + }, + { + "epoch": 0.57, + "grad_norm": 2.149887119887368, + "learning_rate": 8.368316249421894e-06, + "loss": 1.0446, + "step": 24008 + }, + { + "epoch": 0.57, + "grad_norm": 2.107326488450782, + "learning_rate": 8.367563439207784e-06, + "loss": 0.9499, + "step": 24009 + }, + { + "epoch": 0.57, + "grad_norm": 1.721175022317811, + "learning_rate": 8.366810638498237e-06, + "loss": 0.9143, + "step": 24010 + }, + { + "epoch": 0.57, + "grad_norm": 2.0074147214115334, + "learning_rate": 8.366057847297635e-06, + "loss": 1.0242, + "step": 24011 + }, + { + "epoch": 0.57, + "grad_norm": 2.426368444827102, + "learning_rate": 8.36530506561036e-06, + "loss": 1.0272, + "step": 24012 + }, + { + "epoch": 0.57, + "grad_norm": 1.7520034929252108, + "learning_rate": 8.3645522934408e-06, + "loss": 1.0057, + "step": 24013 + }, + { + "epoch": 0.57, + "grad_norm": 1.846437215772775, + "learning_rate": 8.363799530793332e-06, + "loss": 0.9413, + "step": 24014 + }, + { + "epoch": 0.57, + "grad_norm": 1.9265492204792816, + "learning_rate": 8.363046777672342e-06, + "loss": 1.0711, + "step": 24015 + }, + { + "epoch": 0.57, + "grad_norm": 1.9254772026539246, + "learning_rate": 8.362294034082214e-06, + "loss": 1.0195, + "step": 24016 + }, + { + "epoch": 0.57, + "grad_norm": 1.8200593614259886, + "learning_rate": 8.361541300027323e-06, + "loss": 1.0314, + "step": 24017 + }, + { + "epoch": 0.57, + "grad_norm": 2.019742599201751, + "learning_rate": 8.36078857551206e-06, + "loss": 1.0509, + "step": 24018 + }, + { + "epoch": 0.57, + "grad_norm": 1.9730605947968716, + "learning_rate": 8.360035860540806e-06, + "loss": 0.9817, + "step": 24019 + }, + { + "epoch": 0.57, + "grad_norm": 2.1898719703230154, + "learning_rate": 8.359283155117943e-06, + "loss": 0.9833, + "step": 24020 + }, + { + "epoch": 0.57, + "grad_norm": 1.9562381237241964, + "learning_rate": 8.35853045924785e-06, + "loss": 0.9689, + "step": 24021 + }, + { + "epoch": 0.57, + "grad_norm": 2.0237714093124306, + "learning_rate": 8.357777772934914e-06, + "loss": 1.0269, + "step": 24022 + }, + { + "epoch": 0.57, + "grad_norm": 2.0170902996944595, + "learning_rate": 8.357025096183518e-06, + "loss": 1.0296, + "step": 24023 + }, + { + "epoch": 0.57, + "grad_norm": 1.9836322659737966, + "learning_rate": 8.356272428998039e-06, + "loss": 0.9829, + "step": 24024 + }, + { + "epoch": 0.57, + "grad_norm": 2.226276004914208, + "learning_rate": 8.355519771382861e-06, + "loss": 0.9117, + "step": 24025 + }, + { + "epoch": 0.57, + "grad_norm": 1.0529705874725903, + "learning_rate": 8.35476712334237e-06, + "loss": 0.982, + "step": 24026 + }, + { + "epoch": 0.57, + "grad_norm": 3.3841918377386544, + "learning_rate": 8.354014484880946e-06, + "loss": 0.8995, + "step": 24027 + }, + { + "epoch": 0.57, + "grad_norm": 2.023639234521243, + "learning_rate": 8.35326185600297e-06, + "loss": 1.0439, + "step": 24028 + }, + { + "epoch": 0.57, + "grad_norm": 2.54088752914473, + "learning_rate": 8.352509236712823e-06, + "loss": 0.8959, + "step": 24029 + }, + { + "epoch": 0.57, + "grad_norm": 2.2638709721917434, + "learning_rate": 8.351756627014894e-06, + "loss": 0.9075, + "step": 24030 + }, + { + "epoch": 0.57, + "grad_norm": 2.117292985276613, + "learning_rate": 8.351004026913555e-06, + "loss": 1.1081, + "step": 24031 + }, + { + "epoch": 0.57, + "grad_norm": 2.0856362640557093, + "learning_rate": 8.350251436413194e-06, + "loss": 1.0044, + "step": 24032 + }, + { + "epoch": 0.57, + "grad_norm": 1.8525441243319212, + "learning_rate": 8.34949885551819e-06, + "loss": 0.9924, + "step": 24033 + }, + { + "epoch": 0.57, + "grad_norm": 2.103219192262206, + "learning_rate": 8.34874628423293e-06, + "loss": 1.0369, + "step": 24034 + }, + { + "epoch": 0.57, + "grad_norm": 1.1206640209533867, + "learning_rate": 8.347993722561789e-06, + "loss": 0.9577, + "step": 24035 + }, + { + "epoch": 0.57, + "grad_norm": 1.1477936960205992, + "learning_rate": 8.34724117050915e-06, + "loss": 0.9666, + "step": 24036 + }, + { + "epoch": 0.57, + "grad_norm": 2.230670750597684, + "learning_rate": 8.346488628079397e-06, + "loss": 1.0012, + "step": 24037 + }, + { + "epoch": 0.57, + "grad_norm": 1.8982804985529234, + "learning_rate": 8.345736095276914e-06, + "loss": 1.0636, + "step": 24038 + }, + { + "epoch": 0.57, + "grad_norm": 1.8659873873547719, + "learning_rate": 8.344983572106077e-06, + "loss": 1.0529, + "step": 24039 + }, + { + "epoch": 0.57, + "grad_norm": 1.967338217544953, + "learning_rate": 8.344231058571268e-06, + "loss": 1.027, + "step": 24040 + }, + { + "epoch": 0.57, + "grad_norm": 2.1347237975079203, + "learning_rate": 8.343478554676876e-06, + "loss": 0.9478, + "step": 24041 + }, + { + "epoch": 0.57, + "grad_norm": 2.020255601065842, + "learning_rate": 8.342726060427271e-06, + "loss": 0.9442, + "step": 24042 + }, + { + "epoch": 0.57, + "grad_norm": 2.0854812399447384, + "learning_rate": 8.34197357582684e-06, + "loss": 1.0544, + "step": 24043 + }, + { + "epoch": 0.57, + "grad_norm": 1.954318499688525, + "learning_rate": 8.341221100879962e-06, + "loss": 1.0125, + "step": 24044 + }, + { + "epoch": 0.57, + "grad_norm": 1.1047949376023523, + "learning_rate": 8.340468635591023e-06, + "loss": 0.9795, + "step": 24045 + }, + { + "epoch": 0.57, + "grad_norm": 2.1278956641636735, + "learning_rate": 8.3397161799644e-06, + "loss": 1.0105, + "step": 24046 + }, + { + "epoch": 0.57, + "grad_norm": 1.978594456309125, + "learning_rate": 8.338963734004477e-06, + "loss": 1.0141, + "step": 24047 + }, + { + "epoch": 0.57, + "grad_norm": 2.1745116654578203, + "learning_rate": 8.338211297715633e-06, + "loss": 1.0588, + "step": 24048 + }, + { + "epoch": 0.57, + "grad_norm": 2.1233787316736534, + "learning_rate": 8.337458871102245e-06, + "loss": 0.9516, + "step": 24049 + }, + { + "epoch": 0.57, + "grad_norm": 1.865451249027067, + "learning_rate": 8.336706454168701e-06, + "loss": 1.1094, + "step": 24050 + }, + { + "epoch": 0.57, + "grad_norm": 2.075546828398936, + "learning_rate": 8.335954046919376e-06, + "loss": 1.0183, + "step": 24051 + }, + { + "epoch": 0.57, + "grad_norm": 1.8374902823036734, + "learning_rate": 8.335201649358658e-06, + "loss": 0.9271, + "step": 24052 + }, + { + "epoch": 0.57, + "grad_norm": 2.3071921086610847, + "learning_rate": 8.334449261490917e-06, + "loss": 1.0053, + "step": 24053 + }, + { + "epoch": 0.57, + "grad_norm": 1.7534655722803734, + "learning_rate": 8.333696883320544e-06, + "loss": 0.8266, + "step": 24054 + }, + { + "epoch": 0.57, + "grad_norm": 2.030724521786092, + "learning_rate": 8.332944514851916e-06, + "loss": 1.1188, + "step": 24055 + }, + { + "epoch": 0.57, + "grad_norm": 1.0726472895394557, + "learning_rate": 8.332192156089411e-06, + "loss": 0.9467, + "step": 24056 + }, + { + "epoch": 0.57, + "grad_norm": 2.0561160856885166, + "learning_rate": 8.33143980703741e-06, + "loss": 0.9603, + "step": 24057 + }, + { + "epoch": 0.57, + "grad_norm": 2.1576024473154187, + "learning_rate": 8.330687467700295e-06, + "loss": 1.0034, + "step": 24058 + }, + { + "epoch": 0.57, + "grad_norm": 1.907526914092275, + "learning_rate": 8.32993513808245e-06, + "loss": 0.9956, + "step": 24059 + }, + { + "epoch": 0.57, + "grad_norm": 1.8423094344694173, + "learning_rate": 8.329182818188247e-06, + "loss": 1.0482, + "step": 24060 + }, + { + "epoch": 0.57, + "grad_norm": 1.878919314276587, + "learning_rate": 8.32843050802207e-06, + "loss": 1.0745, + "step": 24061 + }, + { + "epoch": 0.57, + "grad_norm": 2.2770053723985506, + "learning_rate": 8.327678207588304e-06, + "loss": 0.9936, + "step": 24062 + }, + { + "epoch": 0.57, + "grad_norm": 1.9470402851811386, + "learning_rate": 8.326925916891322e-06, + "loss": 0.9607, + "step": 24063 + }, + { + "epoch": 0.57, + "grad_norm": 1.1293855997434161, + "learning_rate": 8.326173635935509e-06, + "loss": 0.9114, + "step": 24064 + }, + { + "epoch": 0.57, + "grad_norm": 1.9272151242747777, + "learning_rate": 8.325421364725237e-06, + "loss": 0.9852, + "step": 24065 + }, + { + "epoch": 0.57, + "grad_norm": 1.1746624129531062, + "learning_rate": 8.324669103264899e-06, + "loss": 0.9613, + "step": 24066 + }, + { + "epoch": 0.57, + "grad_norm": 1.0601774457224924, + "learning_rate": 8.323916851558867e-06, + "loss": 0.9976, + "step": 24067 + }, + { + "epoch": 0.57, + "grad_norm": 2.1287426972122208, + "learning_rate": 8.32316460961152e-06, + "loss": 0.9652, + "step": 24068 + }, + { + "epoch": 0.57, + "grad_norm": 2.1275111707490835, + "learning_rate": 8.322412377427242e-06, + "loss": 0.927, + "step": 24069 + }, + { + "epoch": 0.57, + "grad_norm": 2.039671945546014, + "learning_rate": 8.321660155010407e-06, + "loss": 0.9944, + "step": 24070 + }, + { + "epoch": 0.57, + "grad_norm": 1.9604774891702141, + "learning_rate": 8.320907942365397e-06, + "loss": 1.0334, + "step": 24071 + }, + { + "epoch": 0.57, + "grad_norm": 1.8818023923750535, + "learning_rate": 8.320155739496593e-06, + "loss": 0.8819, + "step": 24072 + }, + { + "epoch": 0.57, + "grad_norm": 1.713082819962897, + "learning_rate": 8.31940354640838e-06, + "loss": 0.9498, + "step": 24073 + }, + { + "epoch": 0.57, + "grad_norm": 2.627462169275939, + "learning_rate": 8.318651363105125e-06, + "loss": 0.9529, + "step": 24074 + }, + { + "epoch": 0.57, + "grad_norm": 2.0122463004490645, + "learning_rate": 8.317899189591216e-06, + "loss": 1.1559, + "step": 24075 + }, + { + "epoch": 0.57, + "grad_norm": 1.8111843346983059, + "learning_rate": 8.317147025871033e-06, + "loss": 1.1708, + "step": 24076 + }, + { + "epoch": 0.57, + "grad_norm": 1.1370438353820442, + "learning_rate": 8.31639487194895e-06, + "loss": 1.0085, + "step": 24077 + }, + { + "epoch": 0.57, + "grad_norm": 2.1294251012564986, + "learning_rate": 8.315642727829347e-06, + "loss": 1.0261, + "step": 24078 + }, + { + "epoch": 0.57, + "grad_norm": 2.646353038954798, + "learning_rate": 8.31489059351661e-06, + "loss": 1.0591, + "step": 24079 + }, + { + "epoch": 0.57, + "grad_norm": 2.1416118438280827, + "learning_rate": 8.314138469015112e-06, + "loss": 0.8973, + "step": 24080 + }, + { + "epoch": 0.57, + "grad_norm": 1.8582844997444308, + "learning_rate": 8.313386354329234e-06, + "loss": 1.0599, + "step": 24081 + }, + { + "epoch": 0.57, + "grad_norm": 1.957044772593479, + "learning_rate": 8.31263424946335e-06, + "loss": 0.9931, + "step": 24082 + }, + { + "epoch": 0.57, + "grad_norm": 1.8862210804164727, + "learning_rate": 8.31188215442185e-06, + "loss": 1.0417, + "step": 24083 + }, + { + "epoch": 0.57, + "grad_norm": 2.1328750452060024, + "learning_rate": 8.311130069209103e-06, + "loss": 0.9077, + "step": 24084 + }, + { + "epoch": 0.57, + "grad_norm": 2.515088787318278, + "learning_rate": 8.310377993829491e-06, + "loss": 0.993, + "step": 24085 + }, + { + "epoch": 0.57, + "grad_norm": 2.0417363332279894, + "learning_rate": 8.309625928287394e-06, + "loss": 0.9494, + "step": 24086 + }, + { + "epoch": 0.57, + "grad_norm": 1.935396348352477, + "learning_rate": 8.308873872587193e-06, + "loss": 1.0275, + "step": 24087 + }, + { + "epoch": 0.57, + "grad_norm": 2.035508428407996, + "learning_rate": 8.30812182673326e-06, + "loss": 1.0151, + "step": 24088 + }, + { + "epoch": 0.57, + "grad_norm": 2.1850678183717847, + "learning_rate": 8.307369790729977e-06, + "loss": 1.0481, + "step": 24089 + }, + { + "epoch": 0.57, + "grad_norm": 2.0176086443668146, + "learning_rate": 8.306617764581728e-06, + "loss": 1.0813, + "step": 24090 + }, + { + "epoch": 0.57, + "grad_norm": 2.0771500223608386, + "learning_rate": 8.305865748292882e-06, + "loss": 0.9314, + "step": 24091 + }, + { + "epoch": 0.57, + "grad_norm": 2.102455070486108, + "learning_rate": 8.305113741867822e-06, + "loss": 0.9816, + "step": 24092 + }, + { + "epoch": 0.57, + "grad_norm": 2.0541326970191625, + "learning_rate": 8.304361745310926e-06, + "loss": 0.8182, + "step": 24093 + }, + { + "epoch": 0.57, + "grad_norm": 1.839329346635107, + "learning_rate": 8.303609758626574e-06, + "loss": 1.0762, + "step": 24094 + }, + { + "epoch": 0.57, + "grad_norm": 2.046910084653969, + "learning_rate": 8.302857781819144e-06, + "loss": 0.8945, + "step": 24095 + }, + { + "epoch": 0.57, + "grad_norm": 1.8000107756260482, + "learning_rate": 8.302105814893014e-06, + "loss": 0.978, + "step": 24096 + }, + { + "epoch": 0.57, + "grad_norm": 1.095300085651092, + "learning_rate": 8.301353857852561e-06, + "loss": 0.984, + "step": 24097 + }, + { + "epoch": 0.57, + "grad_norm": 2.0798490452996337, + "learning_rate": 8.30060191070216e-06, + "loss": 0.9584, + "step": 24098 + }, + { + "epoch": 0.57, + "grad_norm": 2.1245051949150344, + "learning_rate": 8.299849973446196e-06, + "loss": 0.9252, + "step": 24099 + }, + { + "epoch": 0.57, + "grad_norm": 2.1262245398291997, + "learning_rate": 8.299098046089042e-06, + "loss": 0.994, + "step": 24100 + }, + { + "epoch": 0.57, + "grad_norm": 1.9257263650675351, + "learning_rate": 8.298346128635081e-06, + "loss": 0.9414, + "step": 24101 + }, + { + "epoch": 0.57, + "grad_norm": 2.0097452662903916, + "learning_rate": 8.297594221088683e-06, + "loss": 1.0427, + "step": 24102 + }, + { + "epoch": 0.57, + "grad_norm": 2.093067732138537, + "learning_rate": 8.296842323454234e-06, + "loss": 1.0963, + "step": 24103 + }, + { + "epoch": 0.57, + "grad_norm": 1.139856864106638, + "learning_rate": 8.296090435736107e-06, + "loss": 0.9177, + "step": 24104 + }, + { + "epoch": 0.57, + "grad_norm": 2.6479858363020323, + "learning_rate": 8.295338557938682e-06, + "loss": 1.0276, + "step": 24105 + }, + { + "epoch": 0.57, + "grad_norm": 1.9954325918664095, + "learning_rate": 8.294586690066331e-06, + "loss": 0.8865, + "step": 24106 + }, + { + "epoch": 0.57, + "grad_norm": 2.329697721986265, + "learning_rate": 8.29383483212344e-06, + "loss": 0.9482, + "step": 24107 + }, + { + "epoch": 0.57, + "grad_norm": 2.1725893345663017, + "learning_rate": 8.293082984114384e-06, + "loss": 0.9285, + "step": 24108 + }, + { + "epoch": 0.57, + "grad_norm": 2.4657332557002594, + "learning_rate": 8.292331146043537e-06, + "loss": 0.9794, + "step": 24109 + }, + { + "epoch": 0.57, + "grad_norm": 2.2272998122337517, + "learning_rate": 8.291579317915278e-06, + "loss": 0.7199, + "step": 24110 + }, + { + "epoch": 0.57, + "grad_norm": 1.830917172735903, + "learning_rate": 8.290827499733988e-06, + "loss": 1.0254, + "step": 24111 + }, + { + "epoch": 0.57, + "grad_norm": 2.463114527115267, + "learning_rate": 8.290075691504041e-06, + "loss": 1.0257, + "step": 24112 + }, + { + "epoch": 0.57, + "grad_norm": 1.9951669183350331, + "learning_rate": 8.289323893229812e-06, + "loss": 0.9485, + "step": 24113 + }, + { + "epoch": 0.57, + "grad_norm": 2.1665108627738974, + "learning_rate": 8.28857210491568e-06, + "loss": 1.1435, + "step": 24114 + }, + { + "epoch": 0.57, + "grad_norm": 2.8025811589626066, + "learning_rate": 8.287820326566028e-06, + "loss": 1.0021, + "step": 24115 + }, + { + "epoch": 0.57, + "grad_norm": 2.015214655334729, + "learning_rate": 8.287068558185225e-06, + "loss": 1.057, + "step": 24116 + }, + { + "epoch": 0.57, + "grad_norm": 1.9382870103419119, + "learning_rate": 8.286316799777651e-06, + "loss": 0.9843, + "step": 24117 + }, + { + "epoch": 0.57, + "grad_norm": 1.0962876662064154, + "learning_rate": 8.285565051347688e-06, + "loss": 0.9546, + "step": 24118 + }, + { + "epoch": 0.57, + "grad_norm": 2.039460087762746, + "learning_rate": 8.284813312899701e-06, + "loss": 1.0287, + "step": 24119 + }, + { + "epoch": 0.57, + "grad_norm": 1.0724180845797506, + "learning_rate": 8.284061584438077e-06, + "loss": 0.9781, + "step": 24120 + }, + { + "epoch": 0.57, + "grad_norm": 2.1526120803077835, + "learning_rate": 8.283309865967187e-06, + "loss": 0.8911, + "step": 24121 + }, + { + "epoch": 0.57, + "grad_norm": 1.9801961356838893, + "learning_rate": 8.282558157491417e-06, + "loss": 0.9249, + "step": 24122 + }, + { + "epoch": 0.57, + "grad_norm": 1.9609624390422062, + "learning_rate": 8.281806459015132e-06, + "loss": 1.0441, + "step": 24123 + }, + { + "epoch": 0.57, + "grad_norm": 1.0609312464141043, + "learning_rate": 8.281054770542713e-06, + "loss": 0.9484, + "step": 24124 + }, + { + "epoch": 0.57, + "grad_norm": 1.9084214614659896, + "learning_rate": 8.280303092078542e-06, + "loss": 1.0734, + "step": 24125 + }, + { + "epoch": 0.57, + "grad_norm": 2.2309773490750353, + "learning_rate": 8.279551423626988e-06, + "loss": 1.0647, + "step": 24126 + }, + { + "epoch": 0.57, + "grad_norm": 2.0048728051077527, + "learning_rate": 8.278799765192427e-06, + "loss": 1.0929, + "step": 24127 + }, + { + "epoch": 0.57, + "grad_norm": 2.849719405337759, + "learning_rate": 8.278048116779242e-06, + "loss": 1.0058, + "step": 24128 + }, + { + "epoch": 0.57, + "grad_norm": 2.376823980559814, + "learning_rate": 8.27729647839181e-06, + "loss": 1.0178, + "step": 24129 + }, + { + "epoch": 0.57, + "grad_norm": 2.053547399311151, + "learning_rate": 8.276544850034498e-06, + "loss": 0.9113, + "step": 24130 + }, + { + "epoch": 0.57, + "grad_norm": 2.1312559912110043, + "learning_rate": 8.275793231711687e-06, + "loss": 1.0767, + "step": 24131 + }, + { + "epoch": 0.57, + "grad_norm": 1.060677164506671, + "learning_rate": 8.275041623427757e-06, + "loss": 1.0093, + "step": 24132 + }, + { + "epoch": 0.57, + "grad_norm": 2.633114641998277, + "learning_rate": 8.27429002518708e-06, + "loss": 1.0197, + "step": 24133 + }, + { + "epoch": 0.57, + "grad_norm": 1.926660840402076, + "learning_rate": 8.27353843699403e-06, + "loss": 1.1483, + "step": 24134 + }, + { + "epoch": 0.57, + "grad_norm": 1.982815110263692, + "learning_rate": 8.272786858852989e-06, + "loss": 1.0045, + "step": 24135 + }, + { + "epoch": 0.57, + "grad_norm": 1.8800083851055602, + "learning_rate": 8.27203529076833e-06, + "loss": 1.0994, + "step": 24136 + }, + { + "epoch": 0.57, + "grad_norm": 1.817440696496048, + "learning_rate": 8.271283732744427e-06, + "loss": 1.0073, + "step": 24137 + }, + { + "epoch": 0.57, + "grad_norm": 1.8529993826624966, + "learning_rate": 8.270532184785656e-06, + "loss": 0.9807, + "step": 24138 + }, + { + "epoch": 0.57, + "grad_norm": 2.2636656526573837, + "learning_rate": 8.2697806468964e-06, + "loss": 0.8613, + "step": 24139 + }, + { + "epoch": 0.57, + "grad_norm": 2.073772415175223, + "learning_rate": 8.269029119081022e-06, + "loss": 1.1092, + "step": 24140 + }, + { + "epoch": 0.57, + "grad_norm": 1.9738616007371204, + "learning_rate": 8.268277601343908e-06, + "loss": 1.1071, + "step": 24141 + }, + { + "epoch": 0.57, + "grad_norm": 2.023035805598821, + "learning_rate": 8.267526093689426e-06, + "loss": 0.9449, + "step": 24142 + }, + { + "epoch": 0.57, + "grad_norm": 1.8590579443776531, + "learning_rate": 8.266774596121962e-06, + "loss": 0.9948, + "step": 24143 + }, + { + "epoch": 0.57, + "grad_norm": 2.1839402123112124, + "learning_rate": 8.26602310864588e-06, + "loss": 0.9584, + "step": 24144 + }, + { + "epoch": 0.57, + "grad_norm": 1.1907573371090918, + "learning_rate": 8.265271631265564e-06, + "loss": 0.9709, + "step": 24145 + }, + { + "epoch": 0.57, + "grad_norm": 1.104860784569999, + "learning_rate": 8.264520163985385e-06, + "loss": 0.9352, + "step": 24146 + }, + { + "epoch": 0.57, + "grad_norm": 1.9763737416886769, + "learning_rate": 8.263768706809715e-06, + "loss": 0.8823, + "step": 24147 + }, + { + "epoch": 0.57, + "grad_norm": 2.0547598464135506, + "learning_rate": 8.263017259742938e-06, + "loss": 1.0006, + "step": 24148 + }, + { + "epoch": 0.57, + "grad_norm": 1.9755806083720462, + "learning_rate": 8.262265822789422e-06, + "loss": 0.9422, + "step": 24149 + }, + { + "epoch": 0.57, + "grad_norm": 1.949606308292636, + "learning_rate": 8.261514395953548e-06, + "loss": 1.0502, + "step": 24150 + }, + { + "epoch": 0.57, + "grad_norm": 2.0549127175183957, + "learning_rate": 8.260762979239683e-06, + "loss": 1.089, + "step": 24151 + }, + { + "epoch": 0.57, + "grad_norm": 1.8965977446179805, + "learning_rate": 8.260011572652208e-06, + "loss": 0.9542, + "step": 24152 + }, + { + "epoch": 0.57, + "grad_norm": 1.0887069269884329, + "learning_rate": 8.259260176195499e-06, + "loss": 0.9306, + "step": 24153 + }, + { + "epoch": 0.57, + "grad_norm": 2.007968425378357, + "learning_rate": 8.258508789873926e-06, + "loss": 1.0695, + "step": 24154 + }, + { + "epoch": 0.57, + "grad_norm": 1.9454949205584349, + "learning_rate": 8.257757413691865e-06, + "loss": 0.9, + "step": 24155 + }, + { + "epoch": 0.57, + "grad_norm": 2.353811911179522, + "learning_rate": 8.257006047653693e-06, + "loss": 1.0393, + "step": 24156 + }, + { + "epoch": 0.57, + "grad_norm": 1.8238622545657963, + "learning_rate": 8.256254691763787e-06, + "loss": 1.0013, + "step": 24157 + }, + { + "epoch": 0.57, + "grad_norm": 1.9714357868584171, + "learning_rate": 8.255503346026514e-06, + "loss": 1.0877, + "step": 24158 + }, + { + "epoch": 0.57, + "grad_norm": 2.0863897416242616, + "learning_rate": 8.254752010446252e-06, + "loss": 0.9762, + "step": 24159 + }, + { + "epoch": 0.57, + "grad_norm": 1.8496446763479517, + "learning_rate": 8.254000685027379e-06, + "loss": 1.0368, + "step": 24160 + }, + { + "epoch": 0.57, + "grad_norm": 1.9742676745137262, + "learning_rate": 8.253249369774268e-06, + "loss": 0.9968, + "step": 24161 + }, + { + "epoch": 0.57, + "grad_norm": 2.3695673891348084, + "learning_rate": 8.252498064691291e-06, + "loss": 0.9495, + "step": 24162 + }, + { + "epoch": 0.57, + "grad_norm": 2.098823390932635, + "learning_rate": 8.251746769782821e-06, + "loss": 0.9636, + "step": 24163 + }, + { + "epoch": 0.57, + "grad_norm": 2.070140695078516, + "learning_rate": 8.250995485053238e-06, + "loss": 0.9659, + "step": 24164 + }, + { + "epoch": 0.57, + "grad_norm": 2.5776798751775014, + "learning_rate": 8.250244210506913e-06, + "loss": 1.1014, + "step": 24165 + }, + { + "epoch": 0.57, + "grad_norm": 2.9472975048991565, + "learning_rate": 8.249492946148217e-06, + "loss": 1.0931, + "step": 24166 + }, + { + "epoch": 0.57, + "grad_norm": 2.0027492400991016, + "learning_rate": 8.248741691981528e-06, + "loss": 1.0488, + "step": 24167 + }, + { + "epoch": 0.57, + "grad_norm": 1.6809589910210445, + "learning_rate": 8.247990448011222e-06, + "loss": 0.9281, + "step": 24168 + }, + { + "epoch": 0.57, + "grad_norm": 2.2662323946308107, + "learning_rate": 8.24723921424167e-06, + "loss": 1.0557, + "step": 24169 + }, + { + "epoch": 0.57, + "grad_norm": 1.859218226475786, + "learning_rate": 8.246487990677242e-06, + "loss": 0.9668, + "step": 24170 + }, + { + "epoch": 0.57, + "grad_norm": 2.021160952983771, + "learning_rate": 8.245736777322324e-06, + "loss": 0.9511, + "step": 24171 + }, + { + "epoch": 0.57, + "grad_norm": 1.8381712206010807, + "learning_rate": 8.244985574181274e-06, + "loss": 1.1199, + "step": 24172 + }, + { + "epoch": 0.57, + "grad_norm": 2.0585799825816906, + "learning_rate": 8.244234381258477e-06, + "loss": 0.8956, + "step": 24173 + }, + { + "epoch": 0.57, + "grad_norm": 1.9252114071888722, + "learning_rate": 8.243483198558299e-06, + "loss": 0.9523, + "step": 24174 + }, + { + "epoch": 0.57, + "grad_norm": 2.2627441092700833, + "learning_rate": 8.242732026085127e-06, + "loss": 0.953, + "step": 24175 + }, + { + "epoch": 0.57, + "grad_norm": 1.9025820923301664, + "learning_rate": 8.24198086384332e-06, + "loss": 0.8996, + "step": 24176 + }, + { + "epoch": 0.57, + "grad_norm": 1.9399650523321426, + "learning_rate": 8.241229711837256e-06, + "loss": 0.9593, + "step": 24177 + }, + { + "epoch": 0.57, + "grad_norm": 2.267692507176217, + "learning_rate": 8.240478570071312e-06, + "loss": 0.9282, + "step": 24178 + }, + { + "epoch": 0.57, + "grad_norm": 1.9926145296934767, + "learning_rate": 8.239727438549859e-06, + "loss": 0.8942, + "step": 24179 + }, + { + "epoch": 0.57, + "grad_norm": 2.1920887480120523, + "learning_rate": 8.238976317277266e-06, + "loss": 0.8632, + "step": 24180 + }, + { + "epoch": 0.57, + "grad_norm": 2.0841882435255976, + "learning_rate": 8.238225206257913e-06, + "loss": 0.9558, + "step": 24181 + }, + { + "epoch": 0.57, + "grad_norm": 2.1562639072736807, + "learning_rate": 8.237474105496172e-06, + "loss": 1.1516, + "step": 24182 + }, + { + "epoch": 0.57, + "grad_norm": 2.0100387850079433, + "learning_rate": 8.23672301499641e-06, + "loss": 0.9325, + "step": 24183 + }, + { + "epoch": 0.57, + "grad_norm": 2.1430062883983587, + "learning_rate": 8.23597193476301e-06, + "loss": 0.9565, + "step": 24184 + }, + { + "epoch": 0.57, + "grad_norm": 2.2639836778734783, + "learning_rate": 8.23522086480034e-06, + "loss": 0.9945, + "step": 24185 + }, + { + "epoch": 0.57, + "grad_norm": 1.9866911247332941, + "learning_rate": 8.23446980511277e-06, + "loss": 1.1611, + "step": 24186 + }, + { + "epoch": 0.57, + "grad_norm": 2.502560122767583, + "learning_rate": 8.233718755704676e-06, + "loss": 1.0224, + "step": 24187 + }, + { + "epoch": 0.57, + "grad_norm": 2.0111545558605015, + "learning_rate": 8.232967716580431e-06, + "loss": 1.0001, + "step": 24188 + }, + { + "epoch": 0.57, + "grad_norm": 1.924868459482168, + "learning_rate": 8.23221668774441e-06, + "loss": 0.9548, + "step": 24189 + }, + { + "epoch": 0.57, + "grad_norm": 2.319346216279392, + "learning_rate": 8.23146566920098e-06, + "loss": 1.042, + "step": 24190 + }, + { + "epoch": 0.57, + "grad_norm": 2.013941467319479, + "learning_rate": 8.230714660954516e-06, + "loss": 0.9706, + "step": 24191 + }, + { + "epoch": 0.57, + "grad_norm": 2.0547762233376243, + "learning_rate": 8.229963663009394e-06, + "loss": 1.0482, + "step": 24192 + }, + { + "epoch": 0.57, + "grad_norm": 2.033410225019163, + "learning_rate": 8.229212675369984e-06, + "loss": 0.9361, + "step": 24193 + }, + { + "epoch": 0.57, + "grad_norm": 2.6153605414128136, + "learning_rate": 8.228461698040657e-06, + "loss": 0.9517, + "step": 24194 + }, + { + "epoch": 0.57, + "grad_norm": 2.0774963443549788, + "learning_rate": 8.227710731025785e-06, + "loss": 1.0061, + "step": 24195 + }, + { + "epoch": 0.57, + "grad_norm": 2.0468311954635845, + "learning_rate": 8.226959774329747e-06, + "loss": 0.9886, + "step": 24196 + }, + { + "epoch": 0.57, + "grad_norm": 2.002620921754173, + "learning_rate": 8.226208827956908e-06, + "loss": 0.9143, + "step": 24197 + }, + { + "epoch": 0.57, + "grad_norm": 1.0824205982896964, + "learning_rate": 8.225457891911643e-06, + "loss": 0.9929, + "step": 24198 + }, + { + "epoch": 0.57, + "grad_norm": 1.725278225557912, + "learning_rate": 8.224706966198325e-06, + "loss": 0.9106, + "step": 24199 + }, + { + "epoch": 0.57, + "grad_norm": 1.8758079027536874, + "learning_rate": 8.223956050821322e-06, + "loss": 0.9872, + "step": 24200 + }, + { + "epoch": 0.57, + "grad_norm": 1.8684775187466178, + "learning_rate": 8.223205145785012e-06, + "loss": 0.9778, + "step": 24201 + }, + { + "epoch": 0.57, + "grad_norm": 1.8365679711595984, + "learning_rate": 8.222454251093763e-06, + "loss": 1.0117, + "step": 24202 + }, + { + "epoch": 0.57, + "grad_norm": 2.2309954624911006, + "learning_rate": 8.22170336675195e-06, + "loss": 1.0773, + "step": 24203 + }, + { + "epoch": 0.57, + "grad_norm": 1.8947359877742116, + "learning_rate": 8.220952492763938e-06, + "loss": 0.9401, + "step": 24204 + }, + { + "epoch": 0.57, + "grad_norm": 1.9843465050792515, + "learning_rate": 8.220201629134109e-06, + "loss": 0.9757, + "step": 24205 + }, + { + "epoch": 0.57, + "grad_norm": 2.0885205638898308, + "learning_rate": 8.219450775866828e-06, + "loss": 0.9874, + "step": 24206 + }, + { + "epoch": 0.57, + "grad_norm": 2.003956643060233, + "learning_rate": 8.218699932966468e-06, + "loss": 0.9799, + "step": 24207 + }, + { + "epoch": 0.57, + "grad_norm": 1.806084189169384, + "learning_rate": 8.2179491004374e-06, + "loss": 0.9197, + "step": 24208 + }, + { + "epoch": 0.57, + "grad_norm": 1.9191466888294666, + "learning_rate": 8.217198278283997e-06, + "loss": 0.868, + "step": 24209 + }, + { + "epoch": 0.57, + "grad_norm": 1.8791257837584352, + "learning_rate": 8.216447466510633e-06, + "loss": 1.0091, + "step": 24210 + }, + { + "epoch": 0.57, + "grad_norm": 2.1240514428934256, + "learning_rate": 8.215696665121674e-06, + "loss": 0.8801, + "step": 24211 + }, + { + "epoch": 0.57, + "grad_norm": 1.8899199333867294, + "learning_rate": 8.214945874121491e-06, + "loss": 0.9689, + "step": 24212 + }, + { + "epoch": 0.57, + "grad_norm": 1.8375748736569806, + "learning_rate": 8.214195093514463e-06, + "loss": 0.9667, + "step": 24213 + }, + { + "epoch": 0.57, + "grad_norm": 1.8681876535765665, + "learning_rate": 8.213444323304955e-06, + "loss": 0.9608, + "step": 24214 + }, + { + "epoch": 0.57, + "grad_norm": 1.985219199538546, + "learning_rate": 8.21269356349734e-06, + "loss": 0.9789, + "step": 24215 + }, + { + "epoch": 0.57, + "grad_norm": 2.044234251066567, + "learning_rate": 8.211942814095986e-06, + "loss": 0.8963, + "step": 24216 + }, + { + "epoch": 0.57, + "grad_norm": 1.8610173265008707, + "learning_rate": 8.211192075105271e-06, + "loss": 0.968, + "step": 24217 + }, + { + "epoch": 0.57, + "grad_norm": 1.9136884808038863, + "learning_rate": 8.21044134652956e-06, + "loss": 0.916, + "step": 24218 + }, + { + "epoch": 0.57, + "grad_norm": 2.43863564695334, + "learning_rate": 8.209690628373225e-06, + "loss": 0.852, + "step": 24219 + }, + { + "epoch": 0.57, + "grad_norm": 2.1493087324187194, + "learning_rate": 8.208939920640644e-06, + "loss": 1.0518, + "step": 24220 + }, + { + "epoch": 0.57, + "grad_norm": 2.129222057863284, + "learning_rate": 8.208189223336175e-06, + "loss": 0.9831, + "step": 24221 + }, + { + "epoch": 0.57, + "grad_norm": 2.0088811498841737, + "learning_rate": 8.207438536464197e-06, + "loss": 0.986, + "step": 24222 + }, + { + "epoch": 0.57, + "grad_norm": 1.8022404795682987, + "learning_rate": 8.206687860029079e-06, + "loss": 0.9054, + "step": 24223 + }, + { + "epoch": 0.57, + "grad_norm": 2.2261356988304555, + "learning_rate": 8.205937194035197e-06, + "loss": 0.9913, + "step": 24224 + }, + { + "epoch": 0.57, + "grad_norm": 1.9747270309456442, + "learning_rate": 8.205186538486911e-06, + "loss": 0.9211, + "step": 24225 + }, + { + "epoch": 0.57, + "grad_norm": 1.1170129633410635, + "learning_rate": 8.204435893388599e-06, + "loss": 0.9301, + "step": 24226 + }, + { + "epoch": 0.57, + "grad_norm": 2.886393078870014, + "learning_rate": 8.203685258744633e-06, + "loss": 1.1463, + "step": 24227 + }, + { + "epoch": 0.57, + "grad_norm": 1.9412783222245646, + "learning_rate": 8.202934634559375e-06, + "loss": 0.8859, + "step": 24228 + }, + { + "epoch": 0.57, + "grad_norm": 2.0270949486209475, + "learning_rate": 8.202184020837202e-06, + "loss": 0.9531, + "step": 24229 + }, + { + "epoch": 0.57, + "grad_norm": 2.854603880463131, + "learning_rate": 8.201433417582482e-06, + "loss": 1.0487, + "step": 24230 + }, + { + "epoch": 0.57, + "grad_norm": 2.098492898177837, + "learning_rate": 8.200682824799591e-06, + "loss": 0.9719, + "step": 24231 + }, + { + "epoch": 0.57, + "grad_norm": 2.2679269398739135, + "learning_rate": 8.199932242492887e-06, + "loss": 0.8734, + "step": 24232 + }, + { + "epoch": 0.57, + "grad_norm": 2.0224614419089204, + "learning_rate": 8.199181670666751e-06, + "loss": 0.9007, + "step": 24233 + }, + { + "epoch": 0.57, + "grad_norm": 1.8365230988406143, + "learning_rate": 8.198431109325552e-06, + "loss": 1.0545, + "step": 24234 + }, + { + "epoch": 0.57, + "grad_norm": 2.1178717498357886, + "learning_rate": 8.197680558473655e-06, + "loss": 0.9725, + "step": 24235 + }, + { + "epoch": 0.57, + "grad_norm": 1.8885929761568856, + "learning_rate": 8.196930018115429e-06, + "loss": 1.0009, + "step": 24236 + }, + { + "epoch": 0.57, + "grad_norm": 1.958391128153375, + "learning_rate": 8.196179488255251e-06, + "loss": 1.0926, + "step": 24237 + }, + { + "epoch": 0.57, + "grad_norm": 2.002305202344713, + "learning_rate": 8.19542896889749e-06, + "loss": 0.8312, + "step": 24238 + }, + { + "epoch": 0.57, + "grad_norm": 2.175731527012449, + "learning_rate": 8.194678460046509e-06, + "loss": 1.0004, + "step": 24239 + }, + { + "epoch": 0.57, + "grad_norm": 2.0333455958364683, + "learning_rate": 8.19392796170668e-06, + "loss": 0.9933, + "step": 24240 + }, + { + "epoch": 0.57, + "grad_norm": 2.088399832239169, + "learning_rate": 8.19317747388238e-06, + "loss": 1.0287, + "step": 24241 + }, + { + "epoch": 0.57, + "grad_norm": 1.9268480210665173, + "learning_rate": 8.192426996577967e-06, + "loss": 0.9475, + "step": 24242 + }, + { + "epoch": 0.57, + "grad_norm": 1.9717437047423634, + "learning_rate": 8.191676529797818e-06, + "loss": 1.0256, + "step": 24243 + }, + { + "epoch": 0.57, + "grad_norm": 1.982060067393416, + "learning_rate": 8.1909260735463e-06, + "loss": 1.0149, + "step": 24244 + }, + { + "epoch": 0.57, + "grad_norm": 2.30542018021438, + "learning_rate": 8.190175627827785e-06, + "loss": 0.9445, + "step": 24245 + }, + { + "epoch": 0.57, + "grad_norm": 2.0097241904822636, + "learning_rate": 8.18942519264664e-06, + "loss": 1.1237, + "step": 24246 + }, + { + "epoch": 0.57, + "grad_norm": 2.5267120687365034, + "learning_rate": 8.188674768007234e-06, + "loss": 0.9838, + "step": 24247 + }, + { + "epoch": 0.57, + "grad_norm": 2.027530712651223, + "learning_rate": 8.18792435391394e-06, + "loss": 0.8073, + "step": 24248 + }, + { + "epoch": 0.57, + "grad_norm": 1.8818034255609841, + "learning_rate": 8.18717395037112e-06, + "loss": 1.0167, + "step": 24249 + }, + { + "epoch": 0.57, + "grad_norm": 2.621357222865314, + "learning_rate": 8.18642355738315e-06, + "loss": 1.14, + "step": 24250 + }, + { + "epoch": 0.57, + "grad_norm": 5.270108129438211, + "learning_rate": 8.185673174954394e-06, + "loss": 0.984, + "step": 24251 + }, + { + "epoch": 0.57, + "grad_norm": 1.7660255717518654, + "learning_rate": 8.184922803089226e-06, + "loss": 1.1206, + "step": 24252 + }, + { + "epoch": 0.57, + "grad_norm": 1.1063663431547965, + "learning_rate": 8.184172441792008e-06, + "loss": 0.9834, + "step": 24253 + }, + { + "epoch": 0.57, + "grad_norm": 2.0171721190481895, + "learning_rate": 8.183422091067115e-06, + "loss": 1.0757, + "step": 24254 + }, + { + "epoch": 0.57, + "grad_norm": 2.0971664156147507, + "learning_rate": 8.182671750918915e-06, + "loss": 1.0259, + "step": 24255 + }, + { + "epoch": 0.57, + "grad_norm": 2.0609106818385494, + "learning_rate": 8.181921421351775e-06, + "loss": 0.9401, + "step": 24256 + }, + { + "epoch": 0.57, + "grad_norm": 2.884508222742385, + "learning_rate": 8.181171102370062e-06, + "loss": 0.99, + "step": 24257 + }, + { + "epoch": 0.57, + "grad_norm": 1.9088914187724704, + "learning_rate": 8.180420793978148e-06, + "loss": 1.0677, + "step": 24258 + }, + { + "epoch": 0.57, + "grad_norm": 1.8543625076154888, + "learning_rate": 8.179670496180401e-06, + "loss": 0.8949, + "step": 24259 + }, + { + "epoch": 0.57, + "grad_norm": 2.143504583385801, + "learning_rate": 8.178920208981189e-06, + "loss": 1.0577, + "step": 24260 + }, + { + "epoch": 0.57, + "grad_norm": 1.7636009713745202, + "learning_rate": 8.178169932384877e-06, + "loss": 0.9607, + "step": 24261 + }, + { + "epoch": 0.57, + "grad_norm": 2.5615509430211705, + "learning_rate": 8.177419666395839e-06, + "loss": 1.0637, + "step": 24262 + }, + { + "epoch": 0.57, + "grad_norm": 2.803857137923411, + "learning_rate": 8.176669411018441e-06, + "loss": 0.9351, + "step": 24263 + }, + { + "epoch": 0.57, + "grad_norm": 1.1379173946145487, + "learning_rate": 8.17591916625705e-06, + "loss": 0.9556, + "step": 24264 + }, + { + "epoch": 0.57, + "grad_norm": 2.0010324676759748, + "learning_rate": 8.175168932116033e-06, + "loss": 1.0464, + "step": 24265 + }, + { + "epoch": 0.57, + "grad_norm": 2.0858531628571955, + "learning_rate": 8.174418708599764e-06, + "loss": 0.9313, + "step": 24266 + }, + { + "epoch": 0.57, + "grad_norm": 1.8740860934515278, + "learning_rate": 8.173668495712605e-06, + "loss": 1.0766, + "step": 24267 + }, + { + "epoch": 0.57, + "grad_norm": 2.181911347760742, + "learning_rate": 8.172918293458925e-06, + "loss": 0.9295, + "step": 24268 + }, + { + "epoch": 0.57, + "grad_norm": 3.9995308077923313, + "learning_rate": 8.1721681018431e-06, + "loss": 0.9562, + "step": 24269 + }, + { + "epoch": 0.57, + "grad_norm": 2.2163405619893934, + "learning_rate": 8.171417920869483e-06, + "loss": 0.9885, + "step": 24270 + }, + { + "epoch": 0.57, + "grad_norm": 1.985675006719452, + "learning_rate": 8.170667750542453e-06, + "loss": 0.8302, + "step": 24271 + }, + { + "epoch": 0.57, + "grad_norm": 1.8827866882913404, + "learning_rate": 8.169917590866372e-06, + "loss": 1.0035, + "step": 24272 + }, + { + "epoch": 0.57, + "grad_norm": 2.0256599544473044, + "learning_rate": 8.169167441845617e-06, + "loss": 0.9784, + "step": 24273 + }, + { + "epoch": 0.57, + "grad_norm": 2.0014051901137324, + "learning_rate": 8.168417303484542e-06, + "loss": 0.9963, + "step": 24274 + }, + { + "epoch": 0.57, + "grad_norm": 1.9030213430702252, + "learning_rate": 8.167667175787524e-06, + "loss": 1.0075, + "step": 24275 + }, + { + "epoch": 0.57, + "grad_norm": 1.8009246047132674, + "learning_rate": 8.16691705875893e-06, + "loss": 0.9629, + "step": 24276 + }, + { + "epoch": 0.57, + "grad_norm": 1.9649307221536962, + "learning_rate": 8.16616695240312e-06, + "loss": 0.9426, + "step": 24277 + }, + { + "epoch": 0.57, + "grad_norm": 2.1348879678531785, + "learning_rate": 8.165416856724468e-06, + "loss": 1.0135, + "step": 24278 + }, + { + "epoch": 0.57, + "grad_norm": 2.1313100007478862, + "learning_rate": 8.164666771727343e-06, + "loss": 0.9928, + "step": 24279 + }, + { + "epoch": 0.57, + "grad_norm": 2.0382701916041013, + "learning_rate": 8.163916697416109e-06, + "loss": 1.0408, + "step": 24280 + }, + { + "epoch": 0.57, + "grad_norm": 2.0556820096208814, + "learning_rate": 8.163166633795129e-06, + "loss": 1.0026, + "step": 24281 + }, + { + "epoch": 0.57, + "grad_norm": 2.11601292111064, + "learning_rate": 8.162416580868777e-06, + "loss": 0.9123, + "step": 24282 + }, + { + "epoch": 0.57, + "grad_norm": 1.8488608901976924, + "learning_rate": 8.161666538641418e-06, + "loss": 0.9863, + "step": 24283 + }, + { + "epoch": 0.57, + "grad_norm": 1.7351301132576618, + "learning_rate": 8.160916507117418e-06, + "loss": 0.9631, + "step": 24284 + }, + { + "epoch": 0.57, + "grad_norm": 1.8624574551587485, + "learning_rate": 8.160166486301141e-06, + "loss": 0.9603, + "step": 24285 + }, + { + "epoch": 0.57, + "grad_norm": 2.018530127757021, + "learning_rate": 8.159416476196962e-06, + "loss": 0.9433, + "step": 24286 + }, + { + "epoch": 0.57, + "grad_norm": 2.2232640277452003, + "learning_rate": 8.158666476809243e-06, + "loss": 0.8984, + "step": 24287 + }, + { + "epoch": 0.57, + "grad_norm": 2.2603515720095495, + "learning_rate": 8.157916488142349e-06, + "loss": 0.9739, + "step": 24288 + }, + { + "epoch": 0.57, + "grad_norm": 2.228653747039856, + "learning_rate": 8.157166510200647e-06, + "loss": 1.0716, + "step": 24289 + }, + { + "epoch": 0.57, + "grad_norm": 2.014325081979853, + "learning_rate": 8.156416542988505e-06, + "loss": 1.0291, + "step": 24290 + }, + { + "epoch": 0.57, + "grad_norm": 1.949229536619136, + "learning_rate": 8.155666586510293e-06, + "loss": 0.9088, + "step": 24291 + }, + { + "epoch": 0.57, + "grad_norm": 2.0696722853634277, + "learning_rate": 8.154916640770373e-06, + "loss": 1.0514, + "step": 24292 + }, + { + "epoch": 0.57, + "grad_norm": 2.225570971060759, + "learning_rate": 8.15416670577311e-06, + "loss": 1.0756, + "step": 24293 + }, + { + "epoch": 0.57, + "grad_norm": 2.60754216350664, + "learning_rate": 8.153416781522876e-06, + "loss": 0.9122, + "step": 24294 + }, + { + "epoch": 0.57, + "grad_norm": 2.091262790829427, + "learning_rate": 8.152666868024032e-06, + "loss": 0.9893, + "step": 24295 + }, + { + "epoch": 0.57, + "grad_norm": 1.8574231278252813, + "learning_rate": 8.151916965280947e-06, + "loss": 1.0198, + "step": 24296 + }, + { + "epoch": 0.57, + "grad_norm": 2.0057590384307327, + "learning_rate": 8.151167073297983e-06, + "loss": 1.0164, + "step": 24297 + }, + { + "epoch": 0.57, + "grad_norm": 2.678109587340353, + "learning_rate": 8.150417192079516e-06, + "loss": 1.0297, + "step": 24298 + }, + { + "epoch": 0.57, + "grad_norm": 1.880332801578035, + "learning_rate": 8.149667321629902e-06, + "loss": 1.0858, + "step": 24299 + }, + { + "epoch": 0.57, + "grad_norm": 2.0234320504251393, + "learning_rate": 8.148917461953511e-06, + "loss": 1.036, + "step": 24300 + }, + { + "epoch": 0.57, + "grad_norm": 2.0162870171030303, + "learning_rate": 8.148167613054711e-06, + "loss": 0.9617, + "step": 24301 + }, + { + "epoch": 0.57, + "grad_norm": 2.3293466710927353, + "learning_rate": 8.14741777493786e-06, + "loss": 1.0356, + "step": 24302 + }, + { + "epoch": 0.57, + "grad_norm": 2.126328642542138, + "learning_rate": 8.146667947607335e-06, + "loss": 1.1392, + "step": 24303 + }, + { + "epoch": 0.57, + "grad_norm": 2.1378195354263725, + "learning_rate": 8.14591813106749e-06, + "loss": 1.0089, + "step": 24304 + }, + { + "epoch": 0.57, + "grad_norm": 3.0441250591976594, + "learning_rate": 8.145168325322704e-06, + "loss": 1.0482, + "step": 24305 + }, + { + "epoch": 0.57, + "grad_norm": 2.2477144475386974, + "learning_rate": 8.144418530377329e-06, + "loss": 0.9144, + "step": 24306 + }, + { + "epoch": 0.57, + "grad_norm": 1.9587299828544589, + "learning_rate": 8.14366874623574e-06, + "loss": 1.0442, + "step": 24307 + }, + { + "epoch": 0.57, + "grad_norm": 1.8435008019640782, + "learning_rate": 8.142918972902301e-06, + "loss": 0.9582, + "step": 24308 + }, + { + "epoch": 0.57, + "grad_norm": 2.251680117524138, + "learning_rate": 8.142169210381373e-06, + "loss": 0.9599, + "step": 24309 + }, + { + "epoch": 0.57, + "grad_norm": 2.213494950203562, + "learning_rate": 8.141419458677321e-06, + "loss": 0.9991, + "step": 24310 + }, + { + "epoch": 0.57, + "grad_norm": 1.8209032570524426, + "learning_rate": 8.140669717794515e-06, + "loss": 1.0287, + "step": 24311 + }, + { + "epoch": 0.57, + "grad_norm": 2.0985967405351302, + "learning_rate": 8.139919987737324e-06, + "loss": 1.0758, + "step": 24312 + }, + { + "epoch": 0.57, + "grad_norm": 2.822009392593712, + "learning_rate": 8.139170268510104e-06, + "loss": 1.0249, + "step": 24313 + }, + { + "epoch": 0.57, + "grad_norm": 1.9519954868841911, + "learning_rate": 8.13842056011722e-06, + "loss": 1.181, + "step": 24314 + }, + { + "epoch": 0.57, + "grad_norm": 1.9890217601217886, + "learning_rate": 8.137670862563047e-06, + "loss": 0.9846, + "step": 24315 + }, + { + "epoch": 0.57, + "grad_norm": 2.237868360709792, + "learning_rate": 8.136921175851942e-06, + "loss": 0.8688, + "step": 24316 + }, + { + "epoch": 0.57, + "grad_norm": 1.8859122560516484, + "learning_rate": 8.136171499988267e-06, + "loss": 0.9249, + "step": 24317 + }, + { + "epoch": 0.57, + "grad_norm": 2.1891977221191925, + "learning_rate": 8.135421834976396e-06, + "loss": 0.875, + "step": 24318 + }, + { + "epoch": 0.57, + "grad_norm": 2.1355461906400084, + "learning_rate": 8.134672180820692e-06, + "loss": 0.9859, + "step": 24319 + }, + { + "epoch": 0.57, + "grad_norm": 1.792602496720269, + "learning_rate": 8.133922537525515e-06, + "loss": 1.0104, + "step": 24320 + }, + { + "epoch": 0.57, + "grad_norm": 2.032605793103504, + "learning_rate": 8.133172905095227e-06, + "loss": 0.8688, + "step": 24321 + }, + { + "epoch": 0.57, + "grad_norm": 2.081795244222275, + "learning_rate": 8.132423283534206e-06, + "loss": 1.0636, + "step": 24322 + }, + { + "epoch": 0.57, + "grad_norm": 1.8567399442580312, + "learning_rate": 8.1316736728468e-06, + "loss": 1.1427, + "step": 24323 + }, + { + "epoch": 0.57, + "grad_norm": 1.9821587117171882, + "learning_rate": 8.130924073037385e-06, + "loss": 0.908, + "step": 24324 + }, + { + "epoch": 0.57, + "grad_norm": 2.4197874507175916, + "learning_rate": 8.13017448411032e-06, + "loss": 1.1114, + "step": 24325 + }, + { + "epoch": 0.57, + "grad_norm": 1.833559299997355, + "learning_rate": 8.129424906069974e-06, + "loss": 1.0029, + "step": 24326 + }, + { + "epoch": 0.57, + "grad_norm": 1.7936874664050948, + "learning_rate": 8.128675338920707e-06, + "loss": 1.043, + "step": 24327 + }, + { + "epoch": 0.57, + "grad_norm": 2.378098850969613, + "learning_rate": 8.127925782666884e-06, + "loss": 1.0528, + "step": 24328 + }, + { + "epoch": 0.57, + "grad_norm": 1.9429226107740643, + "learning_rate": 8.127176237312872e-06, + "loss": 1.0751, + "step": 24329 + }, + { + "epoch": 0.57, + "grad_norm": 1.9776727621776007, + "learning_rate": 8.126426702863029e-06, + "loss": 0.9213, + "step": 24330 + }, + { + "epoch": 0.57, + "grad_norm": 2.2826338241601163, + "learning_rate": 8.125677179321725e-06, + "loss": 1.0746, + "step": 24331 + }, + { + "epoch": 0.57, + "grad_norm": 3.3581675271300897, + "learning_rate": 8.124927666693324e-06, + "loss": 0.9831, + "step": 24332 + }, + { + "epoch": 0.57, + "grad_norm": 1.8460863232816362, + "learning_rate": 8.124178164982186e-06, + "loss": 0.9132, + "step": 24333 + }, + { + "epoch": 0.57, + "grad_norm": 4.5423663442129625, + "learning_rate": 8.123428674192675e-06, + "loss": 0.9346, + "step": 24334 + }, + { + "epoch": 0.57, + "grad_norm": 2.019067029509813, + "learning_rate": 8.122679194329157e-06, + "loss": 0.9564, + "step": 24335 + }, + { + "epoch": 0.57, + "grad_norm": 1.9977828992958906, + "learning_rate": 8.121929725396e-06, + "loss": 1.0966, + "step": 24336 + }, + { + "epoch": 0.57, + "grad_norm": 2.0291537317777735, + "learning_rate": 8.121180267397556e-06, + "loss": 0.8628, + "step": 24337 + }, + { + "epoch": 0.57, + "grad_norm": 2.234658544281144, + "learning_rate": 8.120430820338195e-06, + "loss": 1.0079, + "step": 24338 + }, + { + "epoch": 0.57, + "grad_norm": 1.8461578689450868, + "learning_rate": 8.119681384222284e-06, + "loss": 1.0317, + "step": 24339 + }, + { + "epoch": 0.57, + "grad_norm": 5.210706155699871, + "learning_rate": 8.118931959054185e-06, + "loss": 0.913, + "step": 24340 + }, + { + "epoch": 0.57, + "grad_norm": 2.0280444238119544, + "learning_rate": 8.118182544838258e-06, + "loss": 1.0445, + "step": 24341 + }, + { + "epoch": 0.57, + "grad_norm": 2.2976414121935385, + "learning_rate": 8.117433141578865e-06, + "loss": 0.8364, + "step": 24342 + }, + { + "epoch": 0.57, + "grad_norm": 1.0752467224154538, + "learning_rate": 8.116683749280376e-06, + "loss": 0.9373, + "step": 24343 + }, + { + "epoch": 0.57, + "grad_norm": 1.8621538953338448, + "learning_rate": 8.115934367947151e-06, + "loss": 1.0348, + "step": 24344 + }, + { + "epoch": 0.57, + "grad_norm": 1.1015322105955985, + "learning_rate": 8.11518499758355e-06, + "loss": 0.9646, + "step": 24345 + }, + { + "epoch": 0.57, + "grad_norm": 1.0430312658086582, + "learning_rate": 8.114435638193937e-06, + "loss": 0.926, + "step": 24346 + }, + { + "epoch": 0.57, + "grad_norm": 3.5887867077737132, + "learning_rate": 8.113686289782682e-06, + "loss": 1.0766, + "step": 24347 + }, + { + "epoch": 0.57, + "grad_norm": 2.033129628968858, + "learning_rate": 8.112936952354139e-06, + "loss": 0.9604, + "step": 24348 + }, + { + "epoch": 0.57, + "grad_norm": 2.2539239997181393, + "learning_rate": 8.112187625912675e-06, + "loss": 0.8812, + "step": 24349 + }, + { + "epoch": 0.57, + "grad_norm": 2.188028066410051, + "learning_rate": 8.111438310462656e-06, + "loss": 0.9701, + "step": 24350 + }, + { + "epoch": 0.57, + "grad_norm": 2.1970810408414647, + "learning_rate": 8.110689006008435e-06, + "loss": 1.109, + "step": 24351 + }, + { + "epoch": 0.57, + "grad_norm": 2.0555606727695928, + "learning_rate": 8.109939712554383e-06, + "loss": 0.9361, + "step": 24352 + }, + { + "epoch": 0.57, + "grad_norm": 3.254104708596267, + "learning_rate": 8.109190430104858e-06, + "loss": 1.0302, + "step": 24353 + }, + { + "epoch": 0.57, + "grad_norm": 1.0867475095418682, + "learning_rate": 8.108441158664234e-06, + "loss": 0.9305, + "step": 24354 + }, + { + "epoch": 0.57, + "grad_norm": 1.7830602291183992, + "learning_rate": 8.107691898236855e-06, + "loss": 1.0492, + "step": 24355 + }, + { + "epoch": 0.57, + "grad_norm": 2.1416022864530553, + "learning_rate": 8.106942648827097e-06, + "loss": 1.0121, + "step": 24356 + }, + { + "epoch": 0.57, + "grad_norm": 1.8897460293766128, + "learning_rate": 8.106193410439318e-06, + "loss": 0.938, + "step": 24357 + }, + { + "epoch": 0.57, + "grad_norm": 2.0021496615816847, + "learning_rate": 8.10544418307788e-06, + "loss": 1.0101, + "step": 24358 + }, + { + "epoch": 0.57, + "grad_norm": 2.1698601146959784, + "learning_rate": 8.104694966747143e-06, + "loss": 0.9253, + "step": 24359 + }, + { + "epoch": 0.57, + "grad_norm": 2.267559361113227, + "learning_rate": 8.103945761451474e-06, + "loss": 0.8967, + "step": 24360 + }, + { + "epoch": 0.57, + "grad_norm": 4.483986838376887, + "learning_rate": 8.103196567195237e-06, + "loss": 1.0946, + "step": 24361 + }, + { + "epoch": 0.57, + "grad_norm": 2.092479263257604, + "learning_rate": 8.102447383982785e-06, + "loss": 1.0886, + "step": 24362 + }, + { + "epoch": 0.57, + "grad_norm": 1.8394626338617375, + "learning_rate": 8.101698211818485e-06, + "loss": 0.9118, + "step": 24363 + }, + { + "epoch": 0.57, + "grad_norm": 3.0098718116830963, + "learning_rate": 8.100949050706704e-06, + "loss": 0.9524, + "step": 24364 + }, + { + "epoch": 0.57, + "grad_norm": 2.0664065732058794, + "learning_rate": 8.100199900651794e-06, + "loss": 0.9823, + "step": 24365 + }, + { + "epoch": 0.57, + "grad_norm": 1.8819040405904812, + "learning_rate": 8.099450761658122e-06, + "loss": 0.9518, + "step": 24366 + }, + { + "epoch": 0.57, + "grad_norm": 2.5909306904918745, + "learning_rate": 8.098701633730051e-06, + "loss": 1.108, + "step": 24367 + }, + { + "epoch": 0.57, + "grad_norm": 3.831627952196649, + "learning_rate": 8.097952516871944e-06, + "loss": 1.0413, + "step": 24368 + }, + { + "epoch": 0.57, + "grad_norm": 1.792351833576478, + "learning_rate": 8.097203411088155e-06, + "loss": 1.0315, + "step": 24369 + }, + { + "epoch": 0.57, + "grad_norm": 1.820788223599755, + "learning_rate": 8.096454316383049e-06, + "loss": 0.9878, + "step": 24370 + }, + { + "epoch": 0.57, + "grad_norm": 2.1578337570674897, + "learning_rate": 8.095705232760994e-06, + "loss": 1.066, + "step": 24371 + }, + { + "epoch": 0.57, + "grad_norm": 1.9698726609799861, + "learning_rate": 8.094956160226342e-06, + "loss": 0.9373, + "step": 24372 + }, + { + "epoch": 0.57, + "grad_norm": 1.8795610312229865, + "learning_rate": 8.09420709878346e-06, + "loss": 0.9266, + "step": 24373 + }, + { + "epoch": 0.57, + "grad_norm": 1.9539430416727948, + "learning_rate": 8.093458048436705e-06, + "loss": 0.9552, + "step": 24374 + }, + { + "epoch": 0.57, + "grad_norm": 2.23011370739653, + "learning_rate": 8.092709009190446e-06, + "loss": 1.1229, + "step": 24375 + }, + { + "epoch": 0.57, + "grad_norm": 2.001246418775503, + "learning_rate": 8.091959981049034e-06, + "loss": 0.8912, + "step": 24376 + }, + { + "epoch": 0.57, + "grad_norm": 7.88198091584302, + "learning_rate": 8.091210964016837e-06, + "loss": 1.0183, + "step": 24377 + }, + { + "epoch": 0.57, + "grad_norm": 2.0610388890422175, + "learning_rate": 8.090461958098216e-06, + "loss": 0.9007, + "step": 24378 + }, + { + "epoch": 0.57, + "grad_norm": 1.9615186128497175, + "learning_rate": 8.089712963297527e-06, + "loss": 1.0326, + "step": 24379 + }, + { + "epoch": 0.57, + "grad_norm": 1.1378541208316895, + "learning_rate": 8.088963979619135e-06, + "loss": 0.9864, + "step": 24380 + }, + { + "epoch": 0.57, + "grad_norm": 1.9962371635380733, + "learning_rate": 8.0882150070674e-06, + "loss": 0.9262, + "step": 24381 + }, + { + "epoch": 0.57, + "grad_norm": 1.9816768322033185, + "learning_rate": 8.087466045646684e-06, + "loss": 0.9315, + "step": 24382 + }, + { + "epoch": 0.57, + "grad_norm": 1.1139025715326956, + "learning_rate": 8.086717095361342e-06, + "loss": 0.9004, + "step": 24383 + }, + { + "epoch": 0.57, + "grad_norm": 2.182101341655183, + "learning_rate": 8.085968156215741e-06, + "loss": 0.9894, + "step": 24384 + }, + { + "epoch": 0.57, + "grad_norm": 2.0583161869532227, + "learning_rate": 8.085219228214241e-06, + "loss": 0.9372, + "step": 24385 + }, + { + "epoch": 0.57, + "grad_norm": 1.1193880619845011, + "learning_rate": 8.0844703113612e-06, + "loss": 0.9291, + "step": 24386 + }, + { + "epoch": 0.57, + "grad_norm": 2.0282032228122278, + "learning_rate": 8.083721405660978e-06, + "loss": 1.0279, + "step": 24387 + }, + { + "epoch": 0.57, + "grad_norm": 2.219707188257384, + "learning_rate": 8.082972511117936e-06, + "loss": 1.0122, + "step": 24388 + }, + { + "epoch": 0.57, + "grad_norm": 1.9026925286921101, + "learning_rate": 8.082223627736439e-06, + "loss": 0.8854, + "step": 24389 + }, + { + "epoch": 0.57, + "grad_norm": 1.1327910666863412, + "learning_rate": 8.08147475552084e-06, + "loss": 0.945, + "step": 24390 + }, + { + "epoch": 0.57, + "grad_norm": 2.086178715705495, + "learning_rate": 8.0807258944755e-06, + "loss": 1.0318, + "step": 24391 + }, + { + "epoch": 0.57, + "grad_norm": 2.018290418213267, + "learning_rate": 8.079977044604788e-06, + "loss": 0.9929, + "step": 24392 + }, + { + "epoch": 0.57, + "grad_norm": 2.259331274656485, + "learning_rate": 8.079228205913053e-06, + "loss": 0.9165, + "step": 24393 + }, + { + "epoch": 0.57, + "grad_norm": 1.9300946098685987, + "learning_rate": 8.07847937840466e-06, + "loss": 0.9376, + "step": 24394 + }, + { + "epoch": 0.57, + "grad_norm": 3.1770599830872217, + "learning_rate": 8.077730562083967e-06, + "loss": 0.9769, + "step": 24395 + }, + { + "epoch": 0.57, + "grad_norm": 2.60791039194557, + "learning_rate": 8.07698175695534e-06, + "loss": 1.0339, + "step": 24396 + }, + { + "epoch": 0.57, + "grad_norm": 2.1059278646006487, + "learning_rate": 8.07623296302313e-06, + "loss": 0.8962, + "step": 24397 + }, + { + "epoch": 0.57, + "grad_norm": 2.0771890107448567, + "learning_rate": 8.075484180291702e-06, + "loss": 0.9988, + "step": 24398 + }, + { + "epoch": 0.57, + "grad_norm": 2.0662915511424322, + "learning_rate": 8.074735408765417e-06, + "loss": 0.8939, + "step": 24399 + }, + { + "epoch": 0.57, + "grad_norm": 2.0995597698761426, + "learning_rate": 8.073986648448627e-06, + "loss": 0.9787, + "step": 24400 + }, + { + "epoch": 0.57, + "grad_norm": 1.1483628685050302, + "learning_rate": 8.073237899345699e-06, + "loss": 0.9429, + "step": 24401 + }, + { + "epoch": 0.57, + "grad_norm": 2.291554636467891, + "learning_rate": 8.072489161460986e-06, + "loss": 0.9815, + "step": 24402 + }, + { + "epoch": 0.57, + "grad_norm": 1.9032765750938772, + "learning_rate": 8.07174043479886e-06, + "loss": 1.1813, + "step": 24403 + }, + { + "epoch": 0.57, + "grad_norm": 5.0349467002162465, + "learning_rate": 8.070991719363664e-06, + "loss": 1.072, + "step": 24404 + }, + { + "epoch": 0.57, + "grad_norm": 2.770480654499008, + "learning_rate": 8.070243015159767e-06, + "loss": 1.0428, + "step": 24405 + }, + { + "epoch": 0.57, + "grad_norm": 2.1184751063279905, + "learning_rate": 8.069494322191526e-06, + "loss": 0.9063, + "step": 24406 + }, + { + "epoch": 0.58, + "grad_norm": 2.5458552615594754, + "learning_rate": 8.068745640463305e-06, + "loss": 0.9538, + "step": 24407 + }, + { + "epoch": 0.58, + "grad_norm": 1.9162752231300575, + "learning_rate": 8.067996969979451e-06, + "loss": 0.985, + "step": 24408 + }, + { + "epoch": 0.58, + "grad_norm": 1.893698795042589, + "learning_rate": 8.067248310744334e-06, + "loss": 1.0531, + "step": 24409 + }, + { + "epoch": 0.58, + "grad_norm": 2.2037763037627425, + "learning_rate": 8.066499662762312e-06, + "loss": 1.0377, + "step": 24410 + }, + { + "epoch": 0.58, + "grad_norm": 2.0708028998630574, + "learning_rate": 8.065751026037738e-06, + "loss": 1.088, + "step": 24411 + }, + { + "epoch": 0.58, + "grad_norm": 1.0986286262856306, + "learning_rate": 8.065002400574971e-06, + "loss": 0.9953, + "step": 24412 + }, + { + "epoch": 0.58, + "grad_norm": 2.2547678846913506, + "learning_rate": 8.064253786378375e-06, + "loss": 1.0235, + "step": 24413 + }, + { + "epoch": 0.58, + "grad_norm": 1.7556924527113846, + "learning_rate": 8.063505183452309e-06, + "loss": 1.0149, + "step": 24414 + }, + { + "epoch": 0.58, + "grad_norm": 1.0912006840062833, + "learning_rate": 8.062756591801124e-06, + "loss": 0.9183, + "step": 24415 + }, + { + "epoch": 0.58, + "grad_norm": 1.1774848730838028, + "learning_rate": 8.062008011429187e-06, + "loss": 0.9721, + "step": 24416 + }, + { + "epoch": 0.58, + "grad_norm": 1.1042798559780351, + "learning_rate": 8.061259442340852e-06, + "loss": 0.9255, + "step": 24417 + }, + { + "epoch": 0.58, + "grad_norm": 3.89567928535699, + "learning_rate": 8.06051088454048e-06, + "loss": 0.9937, + "step": 24418 + }, + { + "epoch": 0.58, + "grad_norm": 2.082277761782936, + "learning_rate": 8.059762338032423e-06, + "loss": 1.0112, + "step": 24419 + }, + { + "epoch": 0.58, + "grad_norm": 2.115089094503908, + "learning_rate": 8.059013802821047e-06, + "loss": 0.9969, + "step": 24420 + }, + { + "epoch": 0.58, + "grad_norm": 2.0948900947004283, + "learning_rate": 8.05826527891071e-06, + "loss": 1.1151, + "step": 24421 + }, + { + "epoch": 0.58, + "grad_norm": 2.1973039458430024, + "learning_rate": 8.057516766305765e-06, + "loss": 0.9551, + "step": 24422 + }, + { + "epoch": 0.58, + "grad_norm": 1.9850925426936556, + "learning_rate": 8.056768265010568e-06, + "loss": 0.9913, + "step": 24423 + }, + { + "epoch": 0.58, + "grad_norm": 2.021228918057015, + "learning_rate": 8.056019775029488e-06, + "loss": 0.8836, + "step": 24424 + }, + { + "epoch": 0.58, + "grad_norm": 1.9527668990629534, + "learning_rate": 8.055271296366874e-06, + "loss": 1.0201, + "step": 24425 + }, + { + "epoch": 0.58, + "grad_norm": 2.1418831597069237, + "learning_rate": 8.054522829027084e-06, + "loss": 1.044, + "step": 24426 + }, + { + "epoch": 0.58, + "grad_norm": 2.049037250093264, + "learning_rate": 8.053774373014478e-06, + "loss": 1.0675, + "step": 24427 + }, + { + "epoch": 0.58, + "grad_norm": 2.7430308048362564, + "learning_rate": 8.053025928333418e-06, + "loss": 0.9461, + "step": 24428 + }, + { + "epoch": 0.58, + "grad_norm": 2.122581756918223, + "learning_rate": 8.052277494988254e-06, + "loss": 0.9151, + "step": 24429 + }, + { + "epoch": 0.58, + "grad_norm": 2.0617102577187074, + "learning_rate": 8.051529072983349e-06, + "loss": 0.9761, + "step": 24430 + }, + { + "epoch": 0.58, + "grad_norm": 1.9269593647467818, + "learning_rate": 8.050780662323059e-06, + "loss": 0.9914, + "step": 24431 + }, + { + "epoch": 0.58, + "grad_norm": 1.8751297159574862, + "learning_rate": 8.050032263011738e-06, + "loss": 0.9346, + "step": 24432 + }, + { + "epoch": 0.58, + "grad_norm": 2.3842237320223294, + "learning_rate": 8.049283875053747e-06, + "loss": 1.0925, + "step": 24433 + }, + { + "epoch": 0.58, + "grad_norm": 1.968629183768531, + "learning_rate": 8.048535498453445e-06, + "loss": 0.9396, + "step": 24434 + }, + { + "epoch": 0.58, + "grad_norm": 1.7136356403161455, + "learning_rate": 8.047787133215188e-06, + "loss": 0.9538, + "step": 24435 + }, + { + "epoch": 0.58, + "grad_norm": 3.0034728715002945, + "learning_rate": 8.047038779343327e-06, + "loss": 0.916, + "step": 24436 + }, + { + "epoch": 0.58, + "grad_norm": 2.028119822354184, + "learning_rate": 8.046290436842229e-06, + "loss": 0.9327, + "step": 24437 + }, + { + "epoch": 0.58, + "grad_norm": 2.2674727335299756, + "learning_rate": 8.045542105716247e-06, + "loss": 1.0039, + "step": 24438 + }, + { + "epoch": 0.58, + "grad_norm": 1.9733818860629329, + "learning_rate": 8.044793785969736e-06, + "loss": 1.0749, + "step": 24439 + }, + { + "epoch": 0.58, + "grad_norm": 1.8886812843124858, + "learning_rate": 8.044045477607053e-06, + "loss": 1.0247, + "step": 24440 + }, + { + "epoch": 0.58, + "grad_norm": 2.0020595323382437, + "learning_rate": 8.04329718063256e-06, + "loss": 0.9303, + "step": 24441 + }, + { + "epoch": 0.58, + "grad_norm": 1.972303893587268, + "learning_rate": 8.042548895050609e-06, + "loss": 0.9603, + "step": 24442 + }, + { + "epoch": 0.58, + "grad_norm": 2.352328763558999, + "learning_rate": 8.041800620865558e-06, + "loss": 0.9897, + "step": 24443 + }, + { + "epoch": 0.58, + "grad_norm": 2.127942184579609, + "learning_rate": 8.041052358081761e-06, + "loss": 1.0186, + "step": 24444 + }, + { + "epoch": 0.58, + "grad_norm": 2.2907254931396586, + "learning_rate": 8.040304106703582e-06, + "loss": 0.9202, + "step": 24445 + }, + { + "epoch": 0.58, + "grad_norm": 2.188285726225509, + "learning_rate": 8.039555866735371e-06, + "loss": 0.9062, + "step": 24446 + }, + { + "epoch": 0.58, + "grad_norm": 2.1219673487247404, + "learning_rate": 8.038807638181486e-06, + "loss": 1.0995, + "step": 24447 + }, + { + "epoch": 0.58, + "grad_norm": 1.8501879824694816, + "learning_rate": 8.038059421046281e-06, + "loss": 1.0459, + "step": 24448 + }, + { + "epoch": 0.58, + "grad_norm": 1.952661414442172, + "learning_rate": 8.037311215334122e-06, + "loss": 0.9025, + "step": 24449 + }, + { + "epoch": 0.58, + "grad_norm": 2.0494922977033014, + "learning_rate": 8.036563021049354e-06, + "loss": 0.9808, + "step": 24450 + }, + { + "epoch": 0.58, + "grad_norm": 2.594330427833381, + "learning_rate": 8.035814838196335e-06, + "loss": 0.9385, + "step": 24451 + }, + { + "epoch": 0.58, + "grad_norm": 2.095414536769299, + "learning_rate": 8.035066666779432e-06, + "loss": 1.126, + "step": 24452 + }, + { + "epoch": 0.58, + "grad_norm": 1.9699982110416043, + "learning_rate": 8.034318506802987e-06, + "loss": 1.0658, + "step": 24453 + }, + { + "epoch": 0.58, + "grad_norm": 2.22214498677751, + "learning_rate": 8.033570358271362e-06, + "loss": 1.0406, + "step": 24454 + }, + { + "epoch": 0.58, + "grad_norm": 1.822307365022236, + "learning_rate": 8.032822221188913e-06, + "loss": 1.0533, + "step": 24455 + }, + { + "epoch": 0.58, + "grad_norm": 2.337987128953334, + "learning_rate": 8.03207409556e-06, + "loss": 0.9635, + "step": 24456 + }, + { + "epoch": 0.58, + "grad_norm": 2.013172163966724, + "learning_rate": 8.03132598138897e-06, + "loss": 0.8667, + "step": 24457 + }, + { + "epoch": 0.58, + "grad_norm": 1.9829519518017187, + "learning_rate": 8.030577878680185e-06, + "loss": 0.8551, + "step": 24458 + }, + { + "epoch": 0.58, + "grad_norm": 2.9219634113157227, + "learning_rate": 8.029829787438001e-06, + "loss": 0.9111, + "step": 24459 + }, + { + "epoch": 0.58, + "grad_norm": 2.065994201219286, + "learning_rate": 8.02908170766677e-06, + "loss": 1.142, + "step": 24460 + }, + { + "epoch": 0.58, + "grad_norm": 2.0939819788793668, + "learning_rate": 8.028333639370846e-06, + "loss": 1.0711, + "step": 24461 + }, + { + "epoch": 0.58, + "grad_norm": 2.414233605416846, + "learning_rate": 8.027585582554592e-06, + "loss": 1.0123, + "step": 24462 + }, + { + "epoch": 0.58, + "grad_norm": 1.86459644022517, + "learning_rate": 8.02683753722236e-06, + "loss": 0.9927, + "step": 24463 + }, + { + "epoch": 0.58, + "grad_norm": 2.4253936877321953, + "learning_rate": 8.0260895033785e-06, + "loss": 0.9669, + "step": 24464 + }, + { + "epoch": 0.58, + "grad_norm": 2.7634532432548258, + "learning_rate": 8.025341481027374e-06, + "loss": 1.0213, + "step": 24465 + }, + { + "epoch": 0.58, + "grad_norm": 5.778861980973259, + "learning_rate": 8.024593470173337e-06, + "loss": 0.8979, + "step": 24466 + }, + { + "epoch": 0.58, + "grad_norm": 1.9048144966052298, + "learning_rate": 8.023845470820742e-06, + "loss": 1.0888, + "step": 24467 + }, + { + "epoch": 0.58, + "grad_norm": 2.3051235369406284, + "learning_rate": 8.02309748297394e-06, + "loss": 0.9413, + "step": 24468 + }, + { + "epoch": 0.58, + "grad_norm": 2.4199099209438635, + "learning_rate": 8.022349506637294e-06, + "loss": 1.0024, + "step": 24469 + }, + { + "epoch": 0.58, + "grad_norm": 1.8230770030637162, + "learning_rate": 8.021601541815156e-06, + "loss": 0.9637, + "step": 24470 + }, + { + "epoch": 0.58, + "grad_norm": 2.1907957958804447, + "learning_rate": 8.02085358851188e-06, + "loss": 1.1076, + "step": 24471 + }, + { + "epoch": 0.58, + "grad_norm": 2.110775087156557, + "learning_rate": 8.020105646731815e-06, + "loss": 1.0084, + "step": 24472 + }, + { + "epoch": 0.58, + "grad_norm": 2.126408362815555, + "learning_rate": 8.01935771647933e-06, + "loss": 0.8979, + "step": 24473 + }, + { + "epoch": 0.58, + "grad_norm": 1.991505556985529, + "learning_rate": 8.018609797758768e-06, + "loss": 0.8867, + "step": 24474 + }, + { + "epoch": 0.58, + "grad_norm": 1.8729750921672152, + "learning_rate": 8.017861890574488e-06, + "loss": 1.0804, + "step": 24475 + }, + { + "epoch": 0.58, + "grad_norm": 2.1398818893840255, + "learning_rate": 8.01711399493084e-06, + "loss": 1.083, + "step": 24476 + }, + { + "epoch": 0.58, + "grad_norm": 2.37679987784057, + "learning_rate": 8.016366110832186e-06, + "loss": 1.2655, + "step": 24477 + }, + { + "epoch": 0.58, + "grad_norm": 2.052013857793911, + "learning_rate": 8.015618238282876e-06, + "loss": 1.135, + "step": 24478 + }, + { + "epoch": 0.58, + "grad_norm": 2.088209744510765, + "learning_rate": 8.014870377287263e-06, + "loss": 1.0303, + "step": 24479 + }, + { + "epoch": 0.58, + "grad_norm": 1.0960255120324847, + "learning_rate": 8.014122527849706e-06, + "loss": 0.9408, + "step": 24480 + }, + { + "epoch": 0.58, + "grad_norm": 2.268277897947265, + "learning_rate": 8.013374689974554e-06, + "loss": 0.9384, + "step": 24481 + }, + { + "epoch": 0.58, + "grad_norm": 2.1414798631544367, + "learning_rate": 8.012626863666163e-06, + "loss": 1.0715, + "step": 24482 + }, + { + "epoch": 0.58, + "grad_norm": 2.553520179647895, + "learning_rate": 8.011879048928888e-06, + "loss": 1.0063, + "step": 24483 + }, + { + "epoch": 0.58, + "grad_norm": 2.197962865301093, + "learning_rate": 8.011131245767086e-06, + "loss": 0.9791, + "step": 24484 + }, + { + "epoch": 0.58, + "grad_norm": 2.525796340818541, + "learning_rate": 8.010383454185103e-06, + "loss": 0.8988, + "step": 24485 + }, + { + "epoch": 0.58, + "grad_norm": 1.8097079160258471, + "learning_rate": 8.0096356741873e-06, + "loss": 0.9227, + "step": 24486 + }, + { + "epoch": 0.58, + "grad_norm": 1.0755818797700483, + "learning_rate": 8.00888790577803e-06, + "loss": 0.9278, + "step": 24487 + }, + { + "epoch": 0.58, + "grad_norm": 2.2253591853292645, + "learning_rate": 8.008140148961642e-06, + "loss": 0.938, + "step": 24488 + }, + { + "epoch": 0.58, + "grad_norm": 2.1298775628028266, + "learning_rate": 8.007392403742492e-06, + "loss": 1.0687, + "step": 24489 + }, + { + "epoch": 0.58, + "grad_norm": 1.9010370600521544, + "learning_rate": 8.006644670124935e-06, + "loss": 1.1015, + "step": 24490 + }, + { + "epoch": 0.58, + "grad_norm": 1.8851351719111442, + "learning_rate": 8.005896948113326e-06, + "loss": 1.1036, + "step": 24491 + }, + { + "epoch": 0.58, + "grad_norm": 1.8598484322011977, + "learning_rate": 8.005149237712014e-06, + "loss": 0.9645, + "step": 24492 + }, + { + "epoch": 0.58, + "grad_norm": 1.815410953717995, + "learning_rate": 8.004401538925353e-06, + "loss": 0.9723, + "step": 24493 + }, + { + "epoch": 0.58, + "grad_norm": 1.9315737133182027, + "learning_rate": 8.003653851757703e-06, + "loss": 0.8925, + "step": 24494 + }, + { + "epoch": 0.58, + "grad_norm": 2.602568464399374, + "learning_rate": 8.00290617621341e-06, + "loss": 0.9063, + "step": 24495 + }, + { + "epoch": 0.58, + "grad_norm": 2.286493793827989, + "learning_rate": 8.002158512296829e-06, + "loss": 0.8922, + "step": 24496 + }, + { + "epoch": 0.58, + "grad_norm": 1.9089350987614497, + "learning_rate": 8.001410860012311e-06, + "loss": 0.964, + "step": 24497 + }, + { + "epoch": 0.58, + "grad_norm": 2.3383687602205585, + "learning_rate": 8.000663219364217e-06, + "loss": 1.2129, + "step": 24498 + }, + { + "epoch": 0.58, + "grad_norm": 1.8905406462160332, + "learning_rate": 7.99991559035689e-06, + "loss": 1.0221, + "step": 24499 + }, + { + "epoch": 0.58, + "grad_norm": 1.9456740588154036, + "learning_rate": 7.99916797299469e-06, + "loss": 0.8627, + "step": 24500 + }, + { + "epoch": 0.58, + "grad_norm": 3.384942522285549, + "learning_rate": 7.99842036728197e-06, + "loss": 0.9667, + "step": 24501 + }, + { + "epoch": 0.58, + "grad_norm": 2.0799831541078007, + "learning_rate": 7.997672773223077e-06, + "loss": 0.9033, + "step": 24502 + }, + { + "epoch": 0.58, + "grad_norm": 2.097213845422882, + "learning_rate": 7.996925190822367e-06, + "loss": 1.0695, + "step": 24503 + }, + { + "epoch": 0.58, + "grad_norm": 1.9762918536486374, + "learning_rate": 7.996177620084191e-06, + "loss": 0.9916, + "step": 24504 + }, + { + "epoch": 0.58, + "grad_norm": 2.1172583498220474, + "learning_rate": 7.99543006101291e-06, + "loss": 1.0636, + "step": 24505 + }, + { + "epoch": 0.58, + "grad_norm": 2.395148536287529, + "learning_rate": 7.994682513612863e-06, + "loss": 1.0846, + "step": 24506 + }, + { + "epoch": 0.58, + "grad_norm": 2.450244314297447, + "learning_rate": 7.993934977888412e-06, + "loss": 0.8987, + "step": 24507 + }, + { + "epoch": 0.58, + "grad_norm": 2.4899734434228744, + "learning_rate": 7.993187453843908e-06, + "loss": 1.006, + "step": 24508 + }, + { + "epoch": 0.58, + "grad_norm": 1.028366849259266, + "learning_rate": 7.992439941483702e-06, + "loss": 0.8992, + "step": 24509 + }, + { + "epoch": 0.58, + "grad_norm": 1.8979326156172638, + "learning_rate": 7.991692440812143e-06, + "loss": 0.8866, + "step": 24510 + }, + { + "epoch": 0.58, + "grad_norm": 1.1077224262209748, + "learning_rate": 7.990944951833586e-06, + "loss": 1.0142, + "step": 24511 + }, + { + "epoch": 0.58, + "grad_norm": 2.5267301258383137, + "learning_rate": 7.99019747455239e-06, + "loss": 1.0234, + "step": 24512 + }, + { + "epoch": 0.58, + "grad_norm": 2.220511792295787, + "learning_rate": 7.989450008972893e-06, + "loss": 1.0404, + "step": 24513 + }, + { + "epoch": 0.58, + "grad_norm": 1.93547320755, + "learning_rate": 7.988702555099458e-06, + "loss": 1.0613, + "step": 24514 + }, + { + "epoch": 0.58, + "grad_norm": 2.1143974330780035, + "learning_rate": 7.987955112936435e-06, + "loss": 0.9724, + "step": 24515 + }, + { + "epoch": 0.58, + "grad_norm": 1.859795952226964, + "learning_rate": 7.987207682488172e-06, + "loss": 1.0732, + "step": 24516 + }, + { + "epoch": 0.58, + "grad_norm": 1.9156383086135247, + "learning_rate": 7.986460263759022e-06, + "loss": 0.8891, + "step": 24517 + }, + { + "epoch": 0.58, + "grad_norm": 1.9202473275222396, + "learning_rate": 7.98571285675334e-06, + "loss": 0.8936, + "step": 24518 + }, + { + "epoch": 0.58, + "grad_norm": 2.0084811756121415, + "learning_rate": 7.984965461475477e-06, + "loss": 0.9758, + "step": 24519 + }, + { + "epoch": 0.58, + "grad_norm": 2.2572480798361494, + "learning_rate": 7.984218077929782e-06, + "loss": 1.0426, + "step": 24520 + }, + { + "epoch": 0.58, + "grad_norm": 1.9722467285394887, + "learning_rate": 7.983470706120604e-06, + "loss": 1.0513, + "step": 24521 + }, + { + "epoch": 0.58, + "grad_norm": 2.097718666012052, + "learning_rate": 7.982723346052302e-06, + "loss": 0.9758, + "step": 24522 + }, + { + "epoch": 0.58, + "grad_norm": 1.1163472611115446, + "learning_rate": 7.981975997729222e-06, + "loss": 0.9113, + "step": 24523 + }, + { + "epoch": 0.58, + "grad_norm": 1.1135456064133435, + "learning_rate": 7.981228661155718e-06, + "loss": 0.9752, + "step": 24524 + }, + { + "epoch": 0.58, + "grad_norm": 2.479574131146822, + "learning_rate": 7.980481336336135e-06, + "loss": 0.9611, + "step": 24525 + }, + { + "epoch": 0.58, + "grad_norm": 1.9676458076589256, + "learning_rate": 7.979734023274835e-06, + "loss": 0.9065, + "step": 24526 + }, + { + "epoch": 0.58, + "grad_norm": 2.2591236450533922, + "learning_rate": 7.97898672197616e-06, + "loss": 1.0619, + "step": 24527 + }, + { + "epoch": 0.58, + "grad_norm": 1.8413751077284388, + "learning_rate": 7.978239432444466e-06, + "loss": 1.0436, + "step": 24528 + }, + { + "epoch": 0.58, + "grad_norm": 2.071127402774341, + "learning_rate": 7.977492154684097e-06, + "loss": 1.0936, + "step": 24529 + }, + { + "epoch": 0.58, + "grad_norm": 2.5856969062738346, + "learning_rate": 7.976744888699416e-06, + "loss": 0.9905, + "step": 24530 + }, + { + "epoch": 0.58, + "grad_norm": 1.9659533720943991, + "learning_rate": 7.975997634494764e-06, + "loss": 1.044, + "step": 24531 + }, + { + "epoch": 0.58, + "grad_norm": 1.1101773346765094, + "learning_rate": 7.975250392074493e-06, + "loss": 1.0263, + "step": 24532 + }, + { + "epoch": 0.58, + "grad_norm": 1.957290131005578, + "learning_rate": 7.97450316144296e-06, + "loss": 0.9327, + "step": 24533 + }, + { + "epoch": 0.58, + "grad_norm": 2.02819215754682, + "learning_rate": 7.973755942604505e-06, + "loss": 0.9407, + "step": 24534 + }, + { + "epoch": 0.58, + "grad_norm": 2.093314946536532, + "learning_rate": 7.973008735563486e-06, + "loss": 1.0988, + "step": 24535 + }, + { + "epoch": 0.58, + "grad_norm": 1.1277854560931615, + "learning_rate": 7.972261540324252e-06, + "loss": 0.9707, + "step": 24536 + }, + { + "epoch": 0.58, + "grad_norm": 3.0127081733194165, + "learning_rate": 7.971514356891157e-06, + "loss": 0.853, + "step": 24537 + }, + { + "epoch": 0.58, + "grad_norm": 1.872570013063668, + "learning_rate": 7.970767185268541e-06, + "loss": 1.0535, + "step": 24538 + }, + { + "epoch": 0.58, + "grad_norm": 2.172721152088938, + "learning_rate": 7.970020025460765e-06, + "loss": 0.9871, + "step": 24539 + }, + { + "epoch": 0.58, + "grad_norm": 1.9248617043800844, + "learning_rate": 7.969272877472178e-06, + "loss": 0.9908, + "step": 24540 + }, + { + "epoch": 0.58, + "grad_norm": 1.9571594848944964, + "learning_rate": 7.968525741307123e-06, + "loss": 0.9318, + "step": 24541 + }, + { + "epoch": 0.58, + "grad_norm": 1.1144750044490381, + "learning_rate": 7.967778616969952e-06, + "loss": 0.9526, + "step": 24542 + }, + { + "epoch": 0.58, + "grad_norm": 1.899715160196975, + "learning_rate": 7.96703150446502e-06, + "loss": 1.0953, + "step": 24543 + }, + { + "epoch": 0.58, + "grad_norm": 1.8680511797945945, + "learning_rate": 7.966284403796677e-06, + "loss": 1.0801, + "step": 24544 + }, + { + "epoch": 0.58, + "grad_norm": 1.1023740103221078, + "learning_rate": 7.965537314969267e-06, + "loss": 0.9812, + "step": 24545 + }, + { + "epoch": 0.58, + "grad_norm": 2.0973151784795285, + "learning_rate": 7.964790237987142e-06, + "loss": 1.1242, + "step": 24546 + }, + { + "epoch": 0.58, + "grad_norm": 2.025138204924414, + "learning_rate": 7.964043172854657e-06, + "loss": 0.9921, + "step": 24547 + }, + { + "epoch": 0.58, + "grad_norm": 1.9151203908093437, + "learning_rate": 7.963296119576154e-06, + "loss": 1.0604, + "step": 24548 + }, + { + "epoch": 0.58, + "grad_norm": 1.9132561596362947, + "learning_rate": 7.962549078155984e-06, + "loss": 1.096, + "step": 24549 + }, + { + "epoch": 0.58, + "grad_norm": 1.8960074700213907, + "learning_rate": 7.9618020485985e-06, + "loss": 0.9063, + "step": 24550 + }, + { + "epoch": 0.58, + "grad_norm": 3.633998827247909, + "learning_rate": 7.961055030908052e-06, + "loss": 0.9692, + "step": 24551 + }, + { + "epoch": 0.58, + "grad_norm": 1.8097807851366783, + "learning_rate": 7.960308025088985e-06, + "loss": 0.8905, + "step": 24552 + }, + { + "epoch": 0.58, + "grad_norm": 1.056889690397389, + "learning_rate": 7.959561031145648e-06, + "loss": 0.8877, + "step": 24553 + }, + { + "epoch": 0.58, + "grad_norm": 2.0484486216827684, + "learning_rate": 7.958814049082399e-06, + "loss": 1.132, + "step": 24554 + }, + { + "epoch": 0.58, + "grad_norm": 2.1366732014320315, + "learning_rate": 7.958067078903574e-06, + "loss": 0.9843, + "step": 24555 + }, + { + "epoch": 0.58, + "grad_norm": 1.9954078615331572, + "learning_rate": 7.957320120613535e-06, + "loss": 1.0105, + "step": 24556 + }, + { + "epoch": 0.58, + "grad_norm": 2.351405457210483, + "learning_rate": 7.956573174216619e-06, + "loss": 1.0399, + "step": 24557 + }, + { + "epoch": 0.58, + "grad_norm": 1.9886944533852327, + "learning_rate": 7.955826239717189e-06, + "loss": 0.9763, + "step": 24558 + }, + { + "epoch": 0.58, + "grad_norm": 1.9235507177508864, + "learning_rate": 7.955079317119578e-06, + "loss": 1.0434, + "step": 24559 + }, + { + "epoch": 0.58, + "grad_norm": 1.9895004346651917, + "learning_rate": 7.954332406428145e-06, + "loss": 1.0645, + "step": 24560 + }, + { + "epoch": 0.58, + "grad_norm": 2.186103733666505, + "learning_rate": 7.953585507647239e-06, + "loss": 0.9407, + "step": 24561 + }, + { + "epoch": 0.58, + "grad_norm": 1.7951206459383149, + "learning_rate": 7.952838620781201e-06, + "loss": 0.9487, + "step": 24562 + }, + { + "epoch": 0.58, + "grad_norm": 2.001967813507717, + "learning_rate": 7.952091745834387e-06, + "loss": 1.1358, + "step": 24563 + }, + { + "epoch": 0.58, + "grad_norm": 2.197151257028669, + "learning_rate": 7.951344882811145e-06, + "loss": 1.0074, + "step": 24564 + }, + { + "epoch": 0.58, + "grad_norm": 2.0598129975758104, + "learning_rate": 7.950598031715822e-06, + "loss": 1.0185, + "step": 24565 + }, + { + "epoch": 0.58, + "grad_norm": 1.9305212341770097, + "learning_rate": 7.949851192552762e-06, + "loss": 1.0096, + "step": 24566 + }, + { + "epoch": 0.58, + "grad_norm": 1.6987492939820683, + "learning_rate": 7.94910436532632e-06, + "loss": 1.042, + "step": 24567 + }, + { + "epoch": 0.58, + "grad_norm": 2.1615497754207977, + "learning_rate": 7.948357550040843e-06, + "loss": 0.9876, + "step": 24568 + }, + { + "epoch": 0.58, + "grad_norm": 1.8859152410745434, + "learning_rate": 7.947610746700677e-06, + "loss": 0.9795, + "step": 24569 + }, + { + "epoch": 0.58, + "grad_norm": 1.8835169471794388, + "learning_rate": 7.946863955310169e-06, + "loss": 0.9837, + "step": 24570 + }, + { + "epoch": 0.58, + "grad_norm": 1.8877628813115006, + "learning_rate": 7.94611717587367e-06, + "loss": 1.0836, + "step": 24571 + }, + { + "epoch": 0.58, + "grad_norm": 2.173028238378767, + "learning_rate": 7.94537040839553e-06, + "loss": 0.9781, + "step": 24572 + }, + { + "epoch": 0.58, + "grad_norm": 1.8495309133564404, + "learning_rate": 7.944623652880092e-06, + "loss": 1.0375, + "step": 24573 + }, + { + "epoch": 0.58, + "grad_norm": 1.853741934479667, + "learning_rate": 7.943876909331702e-06, + "loss": 1.068, + "step": 24574 + }, + { + "epoch": 0.58, + "grad_norm": 1.9913043675034767, + "learning_rate": 7.943130177754718e-06, + "loss": 1.0561, + "step": 24575 + }, + { + "epoch": 0.58, + "grad_norm": 2.2875885194821928, + "learning_rate": 7.942383458153478e-06, + "loss": 0.9624, + "step": 24576 + }, + { + "epoch": 0.58, + "grad_norm": 1.8177343645504478, + "learning_rate": 7.941636750532334e-06, + "loss": 0.9646, + "step": 24577 + }, + { + "epoch": 0.58, + "grad_norm": 2.179960620746025, + "learning_rate": 7.940890054895627e-06, + "loss": 0.9393, + "step": 24578 + }, + { + "epoch": 0.58, + "grad_norm": 1.9689563196550222, + "learning_rate": 7.940143371247718e-06, + "loss": 0.9416, + "step": 24579 + }, + { + "epoch": 0.58, + "grad_norm": 1.909821866953728, + "learning_rate": 7.939396699592942e-06, + "loss": 1.0242, + "step": 24580 + }, + { + "epoch": 0.58, + "grad_norm": 1.9757569008552776, + "learning_rate": 7.938650039935654e-06, + "loss": 0.9513, + "step": 24581 + }, + { + "epoch": 0.58, + "grad_norm": 2.755142104040022, + "learning_rate": 7.937903392280198e-06, + "loss": 0.9656, + "step": 24582 + }, + { + "epoch": 0.58, + "grad_norm": 1.9213107917475367, + "learning_rate": 7.937156756630915e-06, + "loss": 1.0069, + "step": 24583 + }, + { + "epoch": 0.58, + "grad_norm": 1.1570751814475613, + "learning_rate": 7.936410132992163e-06, + "loss": 0.9561, + "step": 24584 + }, + { + "epoch": 0.58, + "grad_norm": 2.1632470230244536, + "learning_rate": 7.935663521368283e-06, + "loss": 0.9391, + "step": 24585 + }, + { + "epoch": 0.58, + "grad_norm": 1.9716450742268983, + "learning_rate": 7.93491692176363e-06, + "loss": 0.9351, + "step": 24586 + }, + { + "epoch": 0.58, + "grad_norm": 2.6561453709699645, + "learning_rate": 7.934170334182537e-06, + "loss": 0.8996, + "step": 24587 + }, + { + "epoch": 0.58, + "grad_norm": 2.645408568134822, + "learning_rate": 7.93342375862936e-06, + "loss": 0.8743, + "step": 24588 + }, + { + "epoch": 0.58, + "grad_norm": 1.077295409176833, + "learning_rate": 7.932677195108447e-06, + "loss": 0.9934, + "step": 24589 + }, + { + "epoch": 0.58, + "grad_norm": 2.208233084526868, + "learning_rate": 7.93193064362414e-06, + "loss": 1.053, + "step": 24590 + }, + { + "epoch": 0.58, + "grad_norm": 1.89767419265305, + "learning_rate": 7.931184104180784e-06, + "loss": 0.9218, + "step": 24591 + }, + { + "epoch": 0.58, + "grad_norm": 2.2025471993429924, + "learning_rate": 7.930437576782732e-06, + "loss": 1.028, + "step": 24592 + }, + { + "epoch": 0.58, + "grad_norm": 1.8559996254541211, + "learning_rate": 7.92969106143433e-06, + "loss": 0.9493, + "step": 24593 + }, + { + "epoch": 0.58, + "grad_norm": 2.3761587921687335, + "learning_rate": 7.928944558139921e-06, + "loss": 0.9264, + "step": 24594 + }, + { + "epoch": 0.58, + "grad_norm": 1.928082482342623, + "learning_rate": 7.928198066903849e-06, + "loss": 1.087, + "step": 24595 + }, + { + "epoch": 0.58, + "grad_norm": 2.345728923296823, + "learning_rate": 7.927451587730467e-06, + "loss": 1.107, + "step": 24596 + }, + { + "epoch": 0.58, + "grad_norm": 2.4107310285574988, + "learning_rate": 7.926705120624119e-06, + "loss": 1.0429, + "step": 24597 + }, + { + "epoch": 0.58, + "grad_norm": 2.1040347906078187, + "learning_rate": 7.925958665589146e-06, + "loss": 0.9918, + "step": 24598 + }, + { + "epoch": 0.58, + "grad_norm": 2.203287223603471, + "learning_rate": 7.9252122226299e-06, + "loss": 0.9126, + "step": 24599 + }, + { + "epoch": 0.58, + "grad_norm": 2.105508726090816, + "learning_rate": 7.924465791750728e-06, + "loss": 1.0972, + "step": 24600 + }, + { + "epoch": 0.58, + "grad_norm": 2.1050975519675896, + "learning_rate": 7.923719372955973e-06, + "loss": 0.9228, + "step": 24601 + }, + { + "epoch": 0.58, + "grad_norm": 2.0710364786416884, + "learning_rate": 7.922972966249977e-06, + "loss": 0.9302, + "step": 24602 + }, + { + "epoch": 0.58, + "grad_norm": 2.040461737764075, + "learning_rate": 7.922226571637096e-06, + "loss": 1.1633, + "step": 24603 + }, + { + "epoch": 0.58, + "grad_norm": 2.190156389079995, + "learning_rate": 7.921480189121665e-06, + "loss": 0.8839, + "step": 24604 + }, + { + "epoch": 0.58, + "grad_norm": 1.847616779332227, + "learning_rate": 7.920733818708036e-06, + "loss": 1.0394, + "step": 24605 + }, + { + "epoch": 0.58, + "grad_norm": 2.3373930543221273, + "learning_rate": 7.919987460400549e-06, + "loss": 1.0357, + "step": 24606 + }, + { + "epoch": 0.58, + "grad_norm": 1.873994638078577, + "learning_rate": 7.919241114203563e-06, + "loss": 1.001, + "step": 24607 + }, + { + "epoch": 0.58, + "grad_norm": 1.9670326888007712, + "learning_rate": 7.918494780121404e-06, + "loss": 1.0109, + "step": 24608 + }, + { + "epoch": 0.58, + "grad_norm": 1.1235330661222072, + "learning_rate": 7.917748458158432e-06, + "loss": 1.0199, + "step": 24609 + }, + { + "epoch": 0.58, + "grad_norm": 1.886370446985311, + "learning_rate": 7.917002148318989e-06, + "loss": 1.0412, + "step": 24610 + }, + { + "epoch": 0.58, + "grad_norm": 1.8054154292186786, + "learning_rate": 7.916255850607415e-06, + "loss": 0.8899, + "step": 24611 + }, + { + "epoch": 0.58, + "grad_norm": 2.482086709008993, + "learning_rate": 7.915509565028061e-06, + "loss": 1.0162, + "step": 24612 + }, + { + "epoch": 0.58, + "grad_norm": 1.8485902497595408, + "learning_rate": 7.914763291585271e-06, + "loss": 1.0397, + "step": 24613 + }, + { + "epoch": 0.58, + "grad_norm": 1.974000778395616, + "learning_rate": 7.91401703028339e-06, + "loss": 1.103, + "step": 24614 + }, + { + "epoch": 0.58, + "grad_norm": 1.970257455277558, + "learning_rate": 7.91327078112676e-06, + "loss": 1.0361, + "step": 24615 + }, + { + "epoch": 0.58, + "grad_norm": 2.273798964934706, + "learning_rate": 7.912524544119728e-06, + "loss": 0.9916, + "step": 24616 + }, + { + "epoch": 0.58, + "grad_norm": 1.0709864637613773, + "learning_rate": 7.911778319266643e-06, + "loss": 0.936, + "step": 24617 + }, + { + "epoch": 0.58, + "grad_norm": 2.1004593241080407, + "learning_rate": 7.911032106571842e-06, + "loss": 1.0329, + "step": 24618 + }, + { + "epoch": 0.58, + "grad_norm": 2.712067483048652, + "learning_rate": 7.910285906039673e-06, + "loss": 1.0112, + "step": 24619 + }, + { + "epoch": 0.58, + "grad_norm": 2.290890896675152, + "learning_rate": 7.909539717674483e-06, + "loss": 0.9784, + "step": 24620 + }, + { + "epoch": 0.58, + "grad_norm": 2.068213929768826, + "learning_rate": 7.908793541480615e-06, + "loss": 1.0784, + "step": 24621 + }, + { + "epoch": 0.58, + "grad_norm": 1.9233614903282887, + "learning_rate": 7.908047377462413e-06, + "loss": 1.1147, + "step": 24622 + }, + { + "epoch": 0.58, + "grad_norm": 1.8080035549732645, + "learning_rate": 7.907301225624218e-06, + "loss": 0.9805, + "step": 24623 + }, + { + "epoch": 0.58, + "grad_norm": 1.0779717607390635, + "learning_rate": 7.906555085970382e-06, + "loss": 0.9724, + "step": 24624 + }, + { + "epoch": 0.58, + "grad_norm": 1.9924753942921873, + "learning_rate": 7.905808958505243e-06, + "loss": 1.0222, + "step": 24625 + }, + { + "epoch": 0.58, + "grad_norm": 2.443281302578866, + "learning_rate": 7.905062843233148e-06, + "loss": 0.9333, + "step": 24626 + }, + { + "epoch": 0.58, + "grad_norm": 1.891470519195672, + "learning_rate": 7.904316740158437e-06, + "loss": 1.0899, + "step": 24627 + }, + { + "epoch": 0.58, + "grad_norm": 2.124269150300535, + "learning_rate": 7.903570649285463e-06, + "loss": 1.0114, + "step": 24628 + }, + { + "epoch": 0.58, + "grad_norm": 2.1302027136329045, + "learning_rate": 7.902824570618562e-06, + "loss": 1.0766, + "step": 24629 + }, + { + "epoch": 0.58, + "grad_norm": 2.147869877825179, + "learning_rate": 7.90207850416208e-06, + "loss": 1.0397, + "step": 24630 + }, + { + "epoch": 0.58, + "grad_norm": 1.8963426267881673, + "learning_rate": 7.901332449920365e-06, + "loss": 0.9807, + "step": 24631 + }, + { + "epoch": 0.58, + "grad_norm": 2.0126753399573807, + "learning_rate": 7.90058640789775e-06, + "loss": 1.1395, + "step": 24632 + }, + { + "epoch": 0.58, + "grad_norm": 2.0186496814001393, + "learning_rate": 7.899840378098589e-06, + "loss": 1.0066, + "step": 24633 + }, + { + "epoch": 0.58, + "grad_norm": 2.179121046610857, + "learning_rate": 7.89909436052722e-06, + "loss": 1.0228, + "step": 24634 + }, + { + "epoch": 0.58, + "grad_norm": 1.1059887426768884, + "learning_rate": 7.898348355187994e-06, + "loss": 0.8779, + "step": 24635 + }, + { + "epoch": 0.58, + "grad_norm": 2.242437195706119, + "learning_rate": 7.897602362085246e-06, + "loss": 1.0094, + "step": 24636 + }, + { + "epoch": 0.58, + "grad_norm": 1.938532736706617, + "learning_rate": 7.896856381223321e-06, + "loss": 1.1568, + "step": 24637 + }, + { + "epoch": 0.58, + "grad_norm": 1.849816949147397, + "learning_rate": 7.896110412606569e-06, + "loss": 1.0083, + "step": 24638 + }, + { + "epoch": 0.58, + "grad_norm": 2.2342313630903994, + "learning_rate": 7.895364456239324e-06, + "loss": 0.9322, + "step": 24639 + }, + { + "epoch": 0.58, + "grad_norm": 2.3400036920554084, + "learning_rate": 7.894618512125932e-06, + "loss": 1.0209, + "step": 24640 + }, + { + "epoch": 0.58, + "grad_norm": 2.0053533229825353, + "learning_rate": 7.893872580270741e-06, + "loss": 1.1257, + "step": 24641 + }, + { + "epoch": 0.58, + "grad_norm": 1.1323631394257883, + "learning_rate": 7.893126660678091e-06, + "loss": 0.9688, + "step": 24642 + }, + { + "epoch": 0.58, + "grad_norm": 1.954694730935303, + "learning_rate": 7.892380753352324e-06, + "loss": 0.985, + "step": 24643 + }, + { + "epoch": 0.58, + "grad_norm": 2.004356836598761, + "learning_rate": 7.89163485829778e-06, + "loss": 0.9607, + "step": 24644 + }, + { + "epoch": 0.58, + "grad_norm": 2.2163796729678404, + "learning_rate": 7.89088897551881e-06, + "loss": 0.8971, + "step": 24645 + }, + { + "epoch": 0.58, + "grad_norm": 3.1810027964902203, + "learning_rate": 7.89014310501975e-06, + "loss": 0.9235, + "step": 24646 + }, + { + "epoch": 0.58, + "grad_norm": 3.0967505158799717, + "learning_rate": 7.889397246804943e-06, + "loss": 0.9465, + "step": 24647 + }, + { + "epoch": 0.58, + "grad_norm": 1.1731790878161992, + "learning_rate": 7.888651400878736e-06, + "loss": 0.9313, + "step": 24648 + }, + { + "epoch": 0.58, + "grad_norm": 1.9031069532233256, + "learning_rate": 7.88790556724547e-06, + "loss": 1.0128, + "step": 24649 + }, + { + "epoch": 0.58, + "grad_norm": 1.134274396305467, + "learning_rate": 7.887159745909484e-06, + "loss": 0.9852, + "step": 24650 + }, + { + "epoch": 0.58, + "grad_norm": 1.9427803649435602, + "learning_rate": 7.886413936875122e-06, + "loss": 1.007, + "step": 24651 + }, + { + "epoch": 0.58, + "grad_norm": 2.0952352500807234, + "learning_rate": 7.88566814014673e-06, + "loss": 1.0299, + "step": 24652 + }, + { + "epoch": 0.58, + "grad_norm": 2.3606963742282066, + "learning_rate": 7.884922355728649e-06, + "loss": 0.985, + "step": 24653 + }, + { + "epoch": 0.58, + "grad_norm": 2.3994767302525837, + "learning_rate": 7.884176583625215e-06, + "loss": 1.0888, + "step": 24654 + }, + { + "epoch": 0.58, + "grad_norm": 3.2344816775973726, + "learning_rate": 7.883430823840775e-06, + "loss": 0.9672, + "step": 24655 + }, + { + "epoch": 0.58, + "grad_norm": 2.076001496514739, + "learning_rate": 7.882685076379675e-06, + "loss": 0.935, + "step": 24656 + }, + { + "epoch": 0.58, + "grad_norm": 2.133862914841264, + "learning_rate": 7.88193934124625e-06, + "loss": 1.0192, + "step": 24657 + }, + { + "epoch": 0.58, + "grad_norm": 2.0257508772313866, + "learning_rate": 7.881193618444846e-06, + "loss": 0.9449, + "step": 24658 + }, + { + "epoch": 0.58, + "grad_norm": 2.3228590885121583, + "learning_rate": 7.8804479079798e-06, + "loss": 1.1247, + "step": 24659 + }, + { + "epoch": 0.58, + "grad_norm": 2.243037474335772, + "learning_rate": 7.879702209855463e-06, + "loss": 1.1179, + "step": 24660 + }, + { + "epoch": 0.58, + "grad_norm": 1.9224403331750768, + "learning_rate": 7.878956524076167e-06, + "loss": 0.9534, + "step": 24661 + }, + { + "epoch": 0.58, + "grad_norm": 1.9042041986484362, + "learning_rate": 7.87821085064626e-06, + "loss": 0.8926, + "step": 24662 + }, + { + "epoch": 0.58, + "grad_norm": 2.4924662872814953, + "learning_rate": 7.877465189570081e-06, + "loss": 0.9777, + "step": 24663 + }, + { + "epoch": 0.58, + "grad_norm": 1.8818547499557186, + "learning_rate": 7.876719540851969e-06, + "loss": 1.048, + "step": 24664 + }, + { + "epoch": 0.58, + "grad_norm": 2.066440435347166, + "learning_rate": 7.87597390449627e-06, + "loss": 1.0036, + "step": 24665 + }, + { + "epoch": 0.58, + "grad_norm": 2.2539917818687667, + "learning_rate": 7.875228280507322e-06, + "loss": 1.019, + "step": 24666 + }, + { + "epoch": 0.58, + "grad_norm": 1.1179518004681366, + "learning_rate": 7.874482668889472e-06, + "loss": 0.9825, + "step": 24667 + }, + { + "epoch": 0.58, + "grad_norm": 2.022010813283719, + "learning_rate": 7.87373706964705e-06, + "loss": 1.0386, + "step": 24668 + }, + { + "epoch": 0.58, + "grad_norm": 1.990667939925352, + "learning_rate": 7.872991482784408e-06, + "loss": 0.9349, + "step": 24669 + }, + { + "epoch": 0.58, + "grad_norm": 2.1396842264301426, + "learning_rate": 7.872245908305885e-06, + "loss": 1.0392, + "step": 24670 + }, + { + "epoch": 0.58, + "grad_norm": 1.9931412038562348, + "learning_rate": 7.871500346215817e-06, + "loss": 1.0792, + "step": 24671 + }, + { + "epoch": 0.58, + "grad_norm": 1.9953446831540442, + "learning_rate": 7.870754796518547e-06, + "loss": 1.0677, + "step": 24672 + }, + { + "epoch": 0.58, + "grad_norm": 2.5567033202124434, + "learning_rate": 7.870009259218417e-06, + "loss": 1.0439, + "step": 24673 + }, + { + "epoch": 0.58, + "grad_norm": 1.8916332570120256, + "learning_rate": 7.86926373431977e-06, + "loss": 0.9344, + "step": 24674 + }, + { + "epoch": 0.58, + "grad_norm": 1.8104360420753327, + "learning_rate": 7.868518221826943e-06, + "loss": 0.9528, + "step": 24675 + }, + { + "epoch": 0.58, + "grad_norm": 1.8496796323731695, + "learning_rate": 7.867772721744275e-06, + "loss": 1.0194, + "step": 24676 + }, + { + "epoch": 0.58, + "grad_norm": 1.9295270524667398, + "learning_rate": 7.867027234076115e-06, + "loss": 1.0251, + "step": 24677 + }, + { + "epoch": 0.58, + "grad_norm": 2.400392276664023, + "learning_rate": 7.866281758826793e-06, + "loss": 1.129, + "step": 24678 + }, + { + "epoch": 0.58, + "grad_norm": 3.830634827182077, + "learning_rate": 7.865536296000655e-06, + "loss": 1.0538, + "step": 24679 + }, + { + "epoch": 0.58, + "grad_norm": 2.420389490172086, + "learning_rate": 7.86479084560204e-06, + "loss": 1.0316, + "step": 24680 + }, + { + "epoch": 0.58, + "grad_norm": 2.051132480539618, + "learning_rate": 7.86404540763529e-06, + "loss": 0.8857, + "step": 24681 + }, + { + "epoch": 0.58, + "grad_norm": 2.50275514171166, + "learning_rate": 7.863299982104745e-06, + "loss": 0.8899, + "step": 24682 + }, + { + "epoch": 0.58, + "grad_norm": 2.4407942697739466, + "learning_rate": 7.862554569014739e-06, + "loss": 0.7536, + "step": 24683 + }, + { + "epoch": 0.58, + "grad_norm": 1.8746211229802987, + "learning_rate": 7.861809168369626e-06, + "loss": 1.1124, + "step": 24684 + }, + { + "epoch": 0.58, + "grad_norm": 2.1459086464347146, + "learning_rate": 7.861063780173728e-06, + "loss": 0.9984, + "step": 24685 + }, + { + "epoch": 0.58, + "grad_norm": 1.8918299742022784, + "learning_rate": 7.860318404431397e-06, + "loss": 1.0768, + "step": 24686 + }, + { + "epoch": 0.58, + "grad_norm": 1.9485563182699719, + "learning_rate": 7.859573041146968e-06, + "loss": 0.9415, + "step": 24687 + }, + { + "epoch": 0.58, + "grad_norm": 2.601364581244766, + "learning_rate": 7.858827690324788e-06, + "loss": 1.1836, + "step": 24688 + }, + { + "epoch": 0.58, + "grad_norm": 1.1108242324109399, + "learning_rate": 7.858082351969185e-06, + "loss": 1.0412, + "step": 24689 + }, + { + "epoch": 0.58, + "grad_norm": 2.228858309165702, + "learning_rate": 7.857337026084508e-06, + "loss": 0.9231, + "step": 24690 + }, + { + "epoch": 0.58, + "grad_norm": 3.372803245900599, + "learning_rate": 7.856591712675093e-06, + "loss": 1.0521, + "step": 24691 + }, + { + "epoch": 0.58, + "grad_norm": 1.09673374635969, + "learning_rate": 7.855846411745278e-06, + "loss": 0.9221, + "step": 24692 + }, + { + "epoch": 0.58, + "grad_norm": 1.8010654377786606, + "learning_rate": 7.855101123299403e-06, + "loss": 0.8836, + "step": 24693 + }, + { + "epoch": 0.58, + "grad_norm": 1.941019393382154, + "learning_rate": 7.85435584734181e-06, + "loss": 0.9128, + "step": 24694 + }, + { + "epoch": 0.58, + "grad_norm": 1.96073375302148, + "learning_rate": 7.853610583876839e-06, + "loss": 0.8976, + "step": 24695 + }, + { + "epoch": 0.58, + "grad_norm": 2.129688282368762, + "learning_rate": 7.852865332908822e-06, + "loss": 0.959, + "step": 24696 + }, + { + "epoch": 0.58, + "grad_norm": 2.2025116848197595, + "learning_rate": 7.852120094442105e-06, + "loss": 1.0561, + "step": 24697 + }, + { + "epoch": 0.58, + "grad_norm": 2.2823556950003843, + "learning_rate": 7.851374868481027e-06, + "loss": 1.1408, + "step": 24698 + }, + { + "epoch": 0.58, + "grad_norm": 1.9847351602961483, + "learning_rate": 7.850629655029923e-06, + "loss": 1.0836, + "step": 24699 + }, + { + "epoch": 0.58, + "grad_norm": 2.1542325747327786, + "learning_rate": 7.849884454093132e-06, + "loss": 1.1471, + "step": 24700 + }, + { + "epoch": 0.58, + "grad_norm": 1.839638499384879, + "learning_rate": 7.849139265674995e-06, + "loss": 1.0476, + "step": 24701 + }, + { + "epoch": 0.58, + "grad_norm": 2.1580741733760878, + "learning_rate": 7.848394089779854e-06, + "loss": 1.1742, + "step": 24702 + }, + { + "epoch": 0.58, + "grad_norm": 1.0848002475729261, + "learning_rate": 7.84764892641204e-06, + "loss": 0.9889, + "step": 24703 + }, + { + "epoch": 0.58, + "grad_norm": 2.12765195659182, + "learning_rate": 7.846903775575893e-06, + "loss": 0.8761, + "step": 24704 + }, + { + "epoch": 0.58, + "grad_norm": 2.4798283207332377, + "learning_rate": 7.846158637275761e-06, + "loss": 1.0057, + "step": 24705 + }, + { + "epoch": 0.58, + "grad_norm": 1.9075826533681306, + "learning_rate": 7.84541351151597e-06, + "loss": 1.014, + "step": 24706 + }, + { + "epoch": 0.58, + "grad_norm": 2.1973102328222924, + "learning_rate": 7.844668398300866e-06, + "loss": 1.0498, + "step": 24707 + }, + { + "epoch": 0.58, + "grad_norm": 1.9035057173416694, + "learning_rate": 7.843923297634781e-06, + "loss": 1.0128, + "step": 24708 + }, + { + "epoch": 0.58, + "grad_norm": 2.216669511392213, + "learning_rate": 7.843178209522063e-06, + "loss": 0.9667, + "step": 24709 + }, + { + "epoch": 0.58, + "grad_norm": 1.06415446175239, + "learning_rate": 7.842433133967041e-06, + "loss": 1.0193, + "step": 24710 + }, + { + "epoch": 0.58, + "grad_norm": 1.9746032173766315, + "learning_rate": 7.841688070974056e-06, + "loss": 1.002, + "step": 24711 + }, + { + "epoch": 0.58, + "grad_norm": 1.9243975709099925, + "learning_rate": 7.84094302054745e-06, + "loss": 0.9464, + "step": 24712 + }, + { + "epoch": 0.58, + "grad_norm": 2.438027618756677, + "learning_rate": 7.840197982691552e-06, + "loss": 1.0129, + "step": 24713 + }, + { + "epoch": 0.58, + "grad_norm": 7.248022799630503, + "learning_rate": 7.839452957410707e-06, + "loss": 0.9292, + "step": 24714 + }, + { + "epoch": 0.58, + "grad_norm": 2.140629381807047, + "learning_rate": 7.838707944709252e-06, + "loss": 1.0575, + "step": 24715 + }, + { + "epoch": 0.58, + "grad_norm": 2.1331611204172454, + "learning_rate": 7.837962944591525e-06, + "loss": 0.9193, + "step": 24716 + }, + { + "epoch": 0.58, + "grad_norm": 1.7794196204800297, + "learning_rate": 7.837217957061858e-06, + "loss": 0.9731, + "step": 24717 + }, + { + "epoch": 0.58, + "grad_norm": 2.0308501462126882, + "learning_rate": 7.836472982124596e-06, + "loss": 1.1599, + "step": 24718 + }, + { + "epoch": 0.58, + "grad_norm": 2.1458734389396055, + "learning_rate": 7.835728019784074e-06, + "loss": 1.1201, + "step": 24719 + }, + { + "epoch": 0.58, + "grad_norm": 2.3480186967098167, + "learning_rate": 7.834983070044626e-06, + "loss": 0.9206, + "step": 24720 + }, + { + "epoch": 0.58, + "grad_norm": 1.1004810913735261, + "learning_rate": 7.83423813291059e-06, + "loss": 0.9902, + "step": 24721 + }, + { + "epoch": 0.58, + "grad_norm": 2.431831539369288, + "learning_rate": 7.833493208386308e-06, + "loss": 1.0044, + "step": 24722 + }, + { + "epoch": 0.58, + "grad_norm": 1.070341374797875, + "learning_rate": 7.832748296476117e-06, + "loss": 0.9629, + "step": 24723 + }, + { + "epoch": 0.58, + "grad_norm": 1.8492751907233442, + "learning_rate": 7.832003397184351e-06, + "loss": 0.9709, + "step": 24724 + }, + { + "epoch": 0.58, + "grad_norm": 1.87689476500514, + "learning_rate": 7.831258510515342e-06, + "loss": 0.9263, + "step": 24725 + }, + { + "epoch": 0.58, + "grad_norm": 1.8203538754497555, + "learning_rate": 7.83051363647344e-06, + "loss": 0.9058, + "step": 24726 + }, + { + "epoch": 0.58, + "grad_norm": 1.9155012765151682, + "learning_rate": 7.82976877506297e-06, + "loss": 0.9766, + "step": 24727 + }, + { + "epoch": 0.58, + "grad_norm": 2.10128807747651, + "learning_rate": 7.829023926288275e-06, + "loss": 0.9709, + "step": 24728 + }, + { + "epoch": 0.58, + "grad_norm": 1.9701095025831747, + "learning_rate": 7.828279090153686e-06, + "loss": 1.0225, + "step": 24729 + }, + { + "epoch": 0.58, + "grad_norm": 1.9577460014077648, + "learning_rate": 7.827534266663548e-06, + "loss": 1.0089, + "step": 24730 + }, + { + "epoch": 0.58, + "grad_norm": 1.871490688269534, + "learning_rate": 7.826789455822194e-06, + "loss": 0.991, + "step": 24731 + }, + { + "epoch": 0.58, + "grad_norm": 2.0764466891190096, + "learning_rate": 7.826044657633956e-06, + "loss": 1.0791, + "step": 24732 + }, + { + "epoch": 0.58, + "grad_norm": 1.9953375442932924, + "learning_rate": 7.825299872103181e-06, + "loss": 1.0012, + "step": 24733 + }, + { + "epoch": 0.58, + "grad_norm": 1.8640193052275444, + "learning_rate": 7.82455509923419e-06, + "loss": 1.033, + "step": 24734 + }, + { + "epoch": 0.58, + "grad_norm": 1.9393117007710523, + "learning_rate": 7.823810339031332e-06, + "loss": 1.0201, + "step": 24735 + }, + { + "epoch": 0.58, + "grad_norm": 2.887495677650481, + "learning_rate": 7.823065591498937e-06, + "loss": 0.8599, + "step": 24736 + }, + { + "epoch": 0.58, + "grad_norm": 2.270459665485284, + "learning_rate": 7.82232085664135e-06, + "loss": 0.9487, + "step": 24737 + }, + { + "epoch": 0.58, + "grad_norm": 1.9175831491853672, + "learning_rate": 7.821576134462893e-06, + "loss": 0.9592, + "step": 24738 + }, + { + "epoch": 0.58, + "grad_norm": 2.616620639762589, + "learning_rate": 7.820831424967912e-06, + "loss": 0.9154, + "step": 24739 + }, + { + "epoch": 0.58, + "grad_norm": 2.126890444510663, + "learning_rate": 7.820086728160743e-06, + "loss": 1.047, + "step": 24740 + }, + { + "epoch": 0.58, + "grad_norm": 1.0972031556697668, + "learning_rate": 7.819342044045716e-06, + "loss": 0.8993, + "step": 24741 + }, + { + "epoch": 0.58, + "grad_norm": 1.1831443465753002, + "learning_rate": 7.81859737262717e-06, + "loss": 0.9807, + "step": 24742 + }, + { + "epoch": 0.58, + "grad_norm": 2.4002561046535673, + "learning_rate": 7.817852713909442e-06, + "loss": 1.0654, + "step": 24743 + }, + { + "epoch": 0.58, + "grad_norm": 1.1077702023242253, + "learning_rate": 7.817108067896865e-06, + "loss": 0.9, + "step": 24744 + }, + { + "epoch": 0.58, + "grad_norm": 1.8314964692908153, + "learning_rate": 7.816363434593776e-06, + "loss": 1.0886, + "step": 24745 + }, + { + "epoch": 0.58, + "grad_norm": 1.890783191251897, + "learning_rate": 7.815618814004511e-06, + "loss": 1.1696, + "step": 24746 + }, + { + "epoch": 0.58, + "grad_norm": 1.869594906106508, + "learning_rate": 7.814874206133407e-06, + "loss": 1.0838, + "step": 24747 + }, + { + "epoch": 0.58, + "grad_norm": 1.985126381618048, + "learning_rate": 7.814129610984795e-06, + "loss": 1.1454, + "step": 24748 + }, + { + "epoch": 0.58, + "grad_norm": 2.1956139685027636, + "learning_rate": 7.81338502856301e-06, + "loss": 0.9893, + "step": 24749 + }, + { + "epoch": 0.58, + "grad_norm": 1.9795846192958713, + "learning_rate": 7.812640458872393e-06, + "loss": 1.1062, + "step": 24750 + }, + { + "epoch": 0.58, + "grad_norm": 1.900242696380184, + "learning_rate": 7.811895901917277e-06, + "loss": 0.8516, + "step": 24751 + }, + { + "epoch": 0.58, + "grad_norm": 1.8899603742703077, + "learning_rate": 7.811151357701993e-06, + "loss": 1.1156, + "step": 24752 + }, + { + "epoch": 0.58, + "grad_norm": 2.0961267596076714, + "learning_rate": 7.810406826230877e-06, + "loss": 1.0652, + "step": 24753 + }, + { + "epoch": 0.58, + "grad_norm": 1.6848258659463793, + "learning_rate": 7.809662307508271e-06, + "loss": 1.0056, + "step": 24754 + }, + { + "epoch": 0.58, + "grad_norm": 2.0147863471221266, + "learning_rate": 7.808917801538502e-06, + "loss": 0.9698, + "step": 24755 + }, + { + "epoch": 0.58, + "grad_norm": 1.8584964870911287, + "learning_rate": 7.808173308325905e-06, + "loss": 0.9474, + "step": 24756 + }, + { + "epoch": 0.58, + "grad_norm": 2.079136533586413, + "learning_rate": 7.807428827874818e-06, + "loss": 1.0859, + "step": 24757 + }, + { + "epoch": 0.58, + "grad_norm": 2.034972358846022, + "learning_rate": 7.806684360189579e-06, + "loss": 0.9964, + "step": 24758 + }, + { + "epoch": 0.58, + "grad_norm": 1.9117819013227824, + "learning_rate": 7.805939905274514e-06, + "loss": 1.1001, + "step": 24759 + }, + { + "epoch": 0.58, + "grad_norm": 2.03487897821926, + "learning_rate": 7.80519546313396e-06, + "loss": 1.1134, + "step": 24760 + }, + { + "epoch": 0.58, + "grad_norm": 2.4180371426183145, + "learning_rate": 7.804451033772258e-06, + "loss": 0.9263, + "step": 24761 + }, + { + "epoch": 0.58, + "grad_norm": 2.192679652899692, + "learning_rate": 7.803706617193731e-06, + "loss": 0.8756, + "step": 24762 + }, + { + "epoch": 0.58, + "grad_norm": 1.0870803451328719, + "learning_rate": 7.802962213402723e-06, + "loss": 0.9871, + "step": 24763 + }, + { + "epoch": 0.58, + "grad_norm": 1.8130104493548629, + "learning_rate": 7.802217822403563e-06, + "loss": 1.098, + "step": 24764 + }, + { + "epoch": 0.58, + "grad_norm": 1.7914852276262663, + "learning_rate": 7.80147344420059e-06, + "loss": 1.0436, + "step": 24765 + }, + { + "epoch": 0.58, + "grad_norm": 1.9673043195899378, + "learning_rate": 7.80072907879813e-06, + "loss": 1.0325, + "step": 24766 + }, + { + "epoch": 0.58, + "grad_norm": 1.8872827261811875, + "learning_rate": 7.799984726200524e-06, + "loss": 1.157, + "step": 24767 + }, + { + "epoch": 0.58, + "grad_norm": 1.8170158601502433, + "learning_rate": 7.799240386412106e-06, + "loss": 0.9192, + "step": 24768 + }, + { + "epoch": 0.58, + "grad_norm": 2.105230040213096, + "learning_rate": 7.798496059437204e-06, + "loss": 0.9673, + "step": 24769 + }, + { + "epoch": 0.58, + "grad_norm": 1.9324325170420569, + "learning_rate": 7.797751745280153e-06, + "loss": 0.9862, + "step": 24770 + }, + { + "epoch": 0.58, + "grad_norm": 1.9323830288010495, + "learning_rate": 7.797007443945291e-06, + "loss": 0.99, + "step": 24771 + }, + { + "epoch": 0.58, + "grad_norm": 2.111792244965919, + "learning_rate": 7.796263155436952e-06, + "loss": 0.9998, + "step": 24772 + }, + { + "epoch": 0.58, + "grad_norm": 1.7861227116187062, + "learning_rate": 7.795518879759462e-06, + "loss": 1.0478, + "step": 24773 + }, + { + "epoch": 0.58, + "grad_norm": 1.8571745891401492, + "learning_rate": 7.79477461691716e-06, + "loss": 1.0441, + "step": 24774 + }, + { + "epoch": 0.58, + "grad_norm": 2.1981592999460053, + "learning_rate": 7.794030366914379e-06, + "loss": 0.9188, + "step": 24775 + }, + { + "epoch": 0.58, + "grad_norm": 1.972335929936414, + "learning_rate": 7.793286129755453e-06, + "loss": 1.0831, + "step": 24776 + }, + { + "epoch": 0.58, + "grad_norm": 2.1173770964150975, + "learning_rate": 7.792541905444713e-06, + "loss": 1.148, + "step": 24777 + }, + { + "epoch": 0.58, + "grad_norm": 1.9713845851677083, + "learning_rate": 7.79179769398649e-06, + "loss": 1.0579, + "step": 24778 + }, + { + "epoch": 0.58, + "grad_norm": 2.781035838586345, + "learning_rate": 7.791053495385124e-06, + "loss": 1.071, + "step": 24779 + }, + { + "epoch": 0.58, + "grad_norm": 1.9597150989846848, + "learning_rate": 7.790309309644942e-06, + "loss": 0.9452, + "step": 24780 + }, + { + "epoch": 0.58, + "grad_norm": 2.1745116382613765, + "learning_rate": 7.789565136770278e-06, + "loss": 1.0511, + "step": 24781 + }, + { + "epoch": 0.58, + "grad_norm": 1.9899206427792784, + "learning_rate": 7.788820976765466e-06, + "loss": 0.9558, + "step": 24782 + }, + { + "epoch": 0.58, + "grad_norm": 1.8512711171196086, + "learning_rate": 7.788076829634842e-06, + "loss": 0.9756, + "step": 24783 + }, + { + "epoch": 0.58, + "grad_norm": 1.9983059540588126, + "learning_rate": 7.78733269538273e-06, + "loss": 0.9908, + "step": 24784 + }, + { + "epoch": 0.58, + "grad_norm": 1.085320273079278, + "learning_rate": 7.786588574013467e-06, + "loss": 0.929, + "step": 24785 + }, + { + "epoch": 0.58, + "grad_norm": 1.8388384575202428, + "learning_rate": 7.785844465531393e-06, + "loss": 0.9151, + "step": 24786 + }, + { + "epoch": 0.58, + "grad_norm": 2.2558954248396503, + "learning_rate": 7.785100369940826e-06, + "loss": 1.0573, + "step": 24787 + }, + { + "epoch": 0.58, + "grad_norm": 2.0975422327614877, + "learning_rate": 7.78435628724611e-06, + "loss": 1.0836, + "step": 24788 + }, + { + "epoch": 0.58, + "grad_norm": 2.0656939735372553, + "learning_rate": 7.78361221745157e-06, + "loss": 1.0106, + "step": 24789 + }, + { + "epoch": 0.58, + "grad_norm": 1.9986995596314594, + "learning_rate": 7.782868160561548e-06, + "loss": 1.0319, + "step": 24790 + }, + { + "epoch": 0.58, + "grad_norm": 2.0296805506004345, + "learning_rate": 7.782124116580362e-06, + "loss": 0.8704, + "step": 24791 + }, + { + "epoch": 0.58, + "grad_norm": 2.776307717990268, + "learning_rate": 7.781380085512355e-06, + "loss": 1.0812, + "step": 24792 + }, + { + "epoch": 0.58, + "grad_norm": 2.043166865891237, + "learning_rate": 7.780636067361857e-06, + "loss": 0.9785, + "step": 24793 + }, + { + "epoch": 0.58, + "grad_norm": 2.059283345743274, + "learning_rate": 7.779892062133193e-06, + "loss": 0.9627, + "step": 24794 + }, + { + "epoch": 0.58, + "grad_norm": 2.0583629201640314, + "learning_rate": 7.779148069830705e-06, + "loss": 0.9007, + "step": 24795 + }, + { + "epoch": 0.58, + "grad_norm": 2.702658581149919, + "learning_rate": 7.778404090458718e-06, + "loss": 0.9892, + "step": 24796 + }, + { + "epoch": 0.58, + "grad_norm": 1.8837126981633538, + "learning_rate": 7.777660124021569e-06, + "loss": 0.9216, + "step": 24797 + }, + { + "epoch": 0.58, + "grad_norm": 2.339506967281342, + "learning_rate": 7.77691617052358e-06, + "loss": 1.0288, + "step": 24798 + }, + { + "epoch": 0.58, + "grad_norm": 1.8583188162669269, + "learning_rate": 7.776172229969095e-06, + "loss": 1.0637, + "step": 24799 + }, + { + "epoch": 0.58, + "grad_norm": 1.993019206034267, + "learning_rate": 7.775428302362437e-06, + "loss": 0.9478, + "step": 24800 + }, + { + "epoch": 0.58, + "grad_norm": 1.8363672283649166, + "learning_rate": 7.77468438770794e-06, + "loss": 0.9745, + "step": 24801 + }, + { + "epoch": 0.58, + "grad_norm": 1.9501877468441842, + "learning_rate": 7.773940486009934e-06, + "loss": 0.9779, + "step": 24802 + }, + { + "epoch": 0.58, + "grad_norm": 1.8785709011946945, + "learning_rate": 7.773196597272751e-06, + "loss": 1.0486, + "step": 24803 + }, + { + "epoch": 0.58, + "grad_norm": 1.9660459972554167, + "learning_rate": 7.772452721500726e-06, + "loss": 0.9435, + "step": 24804 + }, + { + "epoch": 0.58, + "grad_norm": 1.9992062465741458, + "learning_rate": 7.771708858698184e-06, + "loss": 0.9212, + "step": 24805 + }, + { + "epoch": 0.58, + "grad_norm": 1.7497846102837435, + "learning_rate": 7.770965008869455e-06, + "loss": 0.7826, + "step": 24806 + }, + { + "epoch": 0.58, + "grad_norm": 2.3488112467120215, + "learning_rate": 7.77022117201888e-06, + "loss": 1.0137, + "step": 24807 + }, + { + "epoch": 0.58, + "grad_norm": 4.64270605709721, + "learning_rate": 7.76947734815078e-06, + "loss": 0.9776, + "step": 24808 + }, + { + "epoch": 0.58, + "grad_norm": 2.0440171150434914, + "learning_rate": 7.768733537269488e-06, + "loss": 1.1059, + "step": 24809 + }, + { + "epoch": 0.58, + "grad_norm": 2.162543557541829, + "learning_rate": 7.767989739379336e-06, + "loss": 1.0145, + "step": 24810 + }, + { + "epoch": 0.58, + "grad_norm": 2.145409037435065, + "learning_rate": 7.767245954484657e-06, + "loss": 1.1003, + "step": 24811 + }, + { + "epoch": 0.58, + "grad_norm": 2.055047570308057, + "learning_rate": 7.766502182589776e-06, + "loss": 1.0987, + "step": 24812 + }, + { + "epoch": 0.58, + "grad_norm": 1.0788142449655775, + "learning_rate": 7.765758423699027e-06, + "loss": 0.8993, + "step": 24813 + }, + { + "epoch": 0.58, + "grad_norm": 2.2662303660466603, + "learning_rate": 7.765014677816745e-06, + "loss": 1.0736, + "step": 24814 + }, + { + "epoch": 0.58, + "grad_norm": 1.8957106770617138, + "learning_rate": 7.764270944947248e-06, + "loss": 1.0, + "step": 24815 + }, + { + "epoch": 0.58, + "grad_norm": 2.189967213992125, + "learning_rate": 7.763527225094876e-06, + "loss": 1.0909, + "step": 24816 + }, + { + "epoch": 0.58, + "grad_norm": 2.8069055136260257, + "learning_rate": 7.762783518263953e-06, + "loss": 0.9551, + "step": 24817 + }, + { + "epoch": 0.58, + "grad_norm": 1.1337718338031109, + "learning_rate": 7.76203982445882e-06, + "loss": 0.9518, + "step": 24818 + }, + { + "epoch": 0.58, + "grad_norm": 2.142283891814946, + "learning_rate": 7.761296143683793e-06, + "loss": 0.9833, + "step": 24819 + }, + { + "epoch": 0.58, + "grad_norm": 1.856755112158566, + "learning_rate": 7.760552475943212e-06, + "loss": 0.9253, + "step": 24820 + }, + { + "epoch": 0.58, + "grad_norm": 2.0097278011236424, + "learning_rate": 7.759808821241406e-06, + "loss": 0.956, + "step": 24821 + }, + { + "epoch": 0.58, + "grad_norm": 1.9506664813679233, + "learning_rate": 7.759065179582698e-06, + "loss": 0.886, + "step": 24822 + }, + { + "epoch": 0.58, + "grad_norm": 2.4012300695355355, + "learning_rate": 7.758321550971421e-06, + "loss": 0.9528, + "step": 24823 + }, + { + "epoch": 0.58, + "grad_norm": 2.323828753791345, + "learning_rate": 7.757577935411908e-06, + "loss": 1.0795, + "step": 24824 + }, + { + "epoch": 0.58, + "grad_norm": 2.061108187520513, + "learning_rate": 7.75683433290849e-06, + "loss": 1.1769, + "step": 24825 + }, + { + "epoch": 0.58, + "grad_norm": 2.0068228338328895, + "learning_rate": 7.756090743465488e-06, + "loss": 0.775, + "step": 24826 + }, + { + "epoch": 0.58, + "grad_norm": 1.8403507461593207, + "learning_rate": 7.755347167087234e-06, + "loss": 1.0413, + "step": 24827 + }, + { + "epoch": 0.58, + "grad_norm": 2.339532434181217, + "learning_rate": 7.754603603778063e-06, + "loss": 0.9351, + "step": 24828 + }, + { + "epoch": 0.58, + "grad_norm": 1.8534339890198963, + "learning_rate": 7.753860053542302e-06, + "loss": 0.9926, + "step": 24829 + }, + { + "epoch": 0.58, + "grad_norm": 1.7540853907754275, + "learning_rate": 7.753116516384274e-06, + "loss": 0.9513, + "step": 24830 + }, + { + "epoch": 0.58, + "grad_norm": 2.1576741067537433, + "learning_rate": 7.752372992308316e-06, + "loss": 0.8988, + "step": 24831 + }, + { + "epoch": 0.59, + "grad_norm": 2.274684980262893, + "learning_rate": 7.751629481318756e-06, + "loss": 1.0062, + "step": 24832 + }, + { + "epoch": 0.59, + "grad_norm": 1.9337127668184328, + "learning_rate": 7.750885983419919e-06, + "loss": 0.9553, + "step": 24833 + }, + { + "epoch": 0.59, + "grad_norm": 2.3725012638649674, + "learning_rate": 7.750142498616133e-06, + "loss": 0.9226, + "step": 24834 + }, + { + "epoch": 0.59, + "grad_norm": 2.0098096457628913, + "learning_rate": 7.749399026911736e-06, + "loss": 0.9134, + "step": 24835 + }, + { + "epoch": 0.59, + "grad_norm": 1.0494561858911007, + "learning_rate": 7.748655568311045e-06, + "loss": 0.9298, + "step": 24836 + }, + { + "epoch": 0.59, + "grad_norm": 1.8142509722531939, + "learning_rate": 7.747912122818395e-06, + "loss": 0.866, + "step": 24837 + }, + { + "epoch": 0.59, + "grad_norm": 1.080106716096625, + "learning_rate": 7.747168690438112e-06, + "loss": 0.9205, + "step": 24838 + }, + { + "epoch": 0.59, + "grad_norm": 1.1570881340270243, + "learning_rate": 7.746425271174533e-06, + "loss": 0.8953, + "step": 24839 + }, + { + "epoch": 0.59, + "grad_norm": 2.0877497083964993, + "learning_rate": 7.745681865031973e-06, + "loss": 1.0268, + "step": 24840 + }, + { + "epoch": 0.59, + "grad_norm": 2.0944406915659037, + "learning_rate": 7.74493847201477e-06, + "loss": 1.1572, + "step": 24841 + }, + { + "epoch": 0.59, + "grad_norm": 1.942194233848245, + "learning_rate": 7.74419509212725e-06, + "loss": 1.0226, + "step": 24842 + }, + { + "epoch": 0.59, + "grad_norm": 2.0701665748145377, + "learning_rate": 7.743451725373734e-06, + "loss": 0.9797, + "step": 24843 + }, + { + "epoch": 0.59, + "grad_norm": 2.1878337175548412, + "learning_rate": 7.742708371758562e-06, + "loss": 0.8699, + "step": 24844 + }, + { + "epoch": 0.59, + "grad_norm": 1.9181926649031613, + "learning_rate": 7.741965031286055e-06, + "loss": 1.009, + "step": 24845 + }, + { + "epoch": 0.59, + "grad_norm": 1.1122630570368575, + "learning_rate": 7.741221703960547e-06, + "loss": 0.8947, + "step": 24846 + }, + { + "epoch": 0.59, + "grad_norm": 1.9216153785528842, + "learning_rate": 7.740478389786354e-06, + "loss": 1.0071, + "step": 24847 + }, + { + "epoch": 0.59, + "grad_norm": 2.019422828546561, + "learning_rate": 7.739735088767815e-06, + "loss": 0.9302, + "step": 24848 + }, + { + "epoch": 0.59, + "grad_norm": 1.1249211102796401, + "learning_rate": 7.738991800909254e-06, + "loss": 0.9452, + "step": 24849 + }, + { + "epoch": 0.59, + "grad_norm": 2.7199611547234186, + "learning_rate": 7.738248526214998e-06, + "loss": 0.9656, + "step": 24850 + }, + { + "epoch": 0.59, + "grad_norm": 1.973117330458531, + "learning_rate": 7.737505264689374e-06, + "loss": 1.0327, + "step": 24851 + }, + { + "epoch": 0.59, + "grad_norm": 1.8120579103314622, + "learning_rate": 7.73676201633671e-06, + "loss": 1.0384, + "step": 24852 + }, + { + "epoch": 0.59, + "grad_norm": 1.868072875823971, + "learning_rate": 7.736018781161339e-06, + "loss": 0.932, + "step": 24853 + }, + { + "epoch": 0.59, + "grad_norm": 1.9408858046475734, + "learning_rate": 7.73527555916758e-06, + "loss": 0.9596, + "step": 24854 + }, + { + "epoch": 0.59, + "grad_norm": 1.1292990907236964, + "learning_rate": 7.734532350359762e-06, + "loss": 0.9217, + "step": 24855 + }, + { + "epoch": 0.59, + "grad_norm": 2.0393807433009417, + "learning_rate": 7.733789154742217e-06, + "loss": 1.0484, + "step": 24856 + }, + { + "epoch": 0.59, + "grad_norm": 2.1125063916163804, + "learning_rate": 7.733045972319268e-06, + "loss": 0.9837, + "step": 24857 + }, + { + "epoch": 0.59, + "grad_norm": 1.798653011814936, + "learning_rate": 7.732302803095242e-06, + "loss": 0.978, + "step": 24858 + }, + { + "epoch": 0.59, + "grad_norm": 1.798129731948513, + "learning_rate": 7.731559647074466e-06, + "loss": 0.9601, + "step": 24859 + }, + { + "epoch": 0.59, + "grad_norm": 2.84501022235145, + "learning_rate": 7.730816504261271e-06, + "loss": 1.0201, + "step": 24860 + }, + { + "epoch": 0.59, + "grad_norm": 2.0333637134966156, + "learning_rate": 7.73007337465998e-06, + "loss": 0.9865, + "step": 24861 + }, + { + "epoch": 0.59, + "grad_norm": 1.9479621154085525, + "learning_rate": 7.729330258274919e-06, + "loss": 0.9088, + "step": 24862 + }, + { + "epoch": 0.59, + "grad_norm": 1.0646989218155107, + "learning_rate": 7.72858715511042e-06, + "loss": 0.9682, + "step": 24863 + }, + { + "epoch": 0.59, + "grad_norm": 1.9729466317810385, + "learning_rate": 7.7278440651708e-06, + "loss": 0.9856, + "step": 24864 + }, + { + "epoch": 0.59, + "grad_norm": 2.473507565310547, + "learning_rate": 7.727100988460394e-06, + "loss": 1.0779, + "step": 24865 + }, + { + "epoch": 0.59, + "grad_norm": 2.368825955015729, + "learning_rate": 7.726357924983524e-06, + "loss": 1.015, + "step": 24866 + }, + { + "epoch": 0.59, + "grad_norm": 1.0897232701762265, + "learning_rate": 7.725614874744524e-06, + "loss": 0.9701, + "step": 24867 + }, + { + "epoch": 0.59, + "grad_norm": 1.9519327793822545, + "learning_rate": 7.724871837747708e-06, + "loss": 0.9599, + "step": 24868 + }, + { + "epoch": 0.59, + "grad_norm": 1.9915602621051776, + "learning_rate": 7.72412881399741e-06, + "loss": 0.8965, + "step": 24869 + }, + { + "epoch": 0.59, + "grad_norm": 2.4110779325343263, + "learning_rate": 7.723385803497956e-06, + "loss": 1.0362, + "step": 24870 + }, + { + "epoch": 0.59, + "grad_norm": 2.0226137231560513, + "learning_rate": 7.72264280625367e-06, + "loss": 0.9822, + "step": 24871 + }, + { + "epoch": 0.59, + "grad_norm": 1.7854070371866364, + "learning_rate": 7.721899822268875e-06, + "loss": 1.0186, + "step": 24872 + }, + { + "epoch": 0.59, + "grad_norm": 2.9614278806300005, + "learning_rate": 7.721156851547903e-06, + "loss": 0.9671, + "step": 24873 + }, + { + "epoch": 0.59, + "grad_norm": 1.1603617031671702, + "learning_rate": 7.72041389409508e-06, + "loss": 0.9877, + "step": 24874 + }, + { + "epoch": 0.59, + "grad_norm": 1.8116478075278004, + "learning_rate": 7.719670949914729e-06, + "loss": 1.006, + "step": 24875 + }, + { + "epoch": 0.59, + "grad_norm": 3.0355867873086075, + "learning_rate": 7.71892801901117e-06, + "loss": 0.9335, + "step": 24876 + }, + { + "epoch": 0.59, + "grad_norm": 10.316226277926507, + "learning_rate": 7.71818510138874e-06, + "loss": 1.13, + "step": 24877 + }, + { + "epoch": 0.59, + "grad_norm": 2.027063894858875, + "learning_rate": 7.717442197051756e-06, + "loss": 1.0351, + "step": 24878 + }, + { + "epoch": 0.59, + "grad_norm": 1.8490320128325315, + "learning_rate": 7.716699306004544e-06, + "loss": 1.0654, + "step": 24879 + }, + { + "epoch": 0.59, + "grad_norm": 2.0262317486966683, + "learning_rate": 7.715956428251433e-06, + "loss": 1.033, + "step": 24880 + }, + { + "epoch": 0.59, + "grad_norm": 1.1054935737921316, + "learning_rate": 7.715213563796752e-06, + "loss": 0.9384, + "step": 24881 + }, + { + "epoch": 0.59, + "grad_norm": 2.3473114746906782, + "learning_rate": 7.714470712644817e-06, + "loss": 0.9213, + "step": 24882 + }, + { + "epoch": 0.59, + "grad_norm": 2.026456682285645, + "learning_rate": 7.713727874799954e-06, + "loss": 0.9135, + "step": 24883 + }, + { + "epoch": 0.59, + "grad_norm": 2.1789210137341173, + "learning_rate": 7.712985050266499e-06, + "loss": 1.028, + "step": 24884 + }, + { + "epoch": 0.59, + "grad_norm": 2.11999337523615, + "learning_rate": 7.712242239048761e-06, + "loss": 0.9001, + "step": 24885 + }, + { + "epoch": 0.59, + "grad_norm": 2.1579706107761942, + "learning_rate": 7.711499441151077e-06, + "loss": 1.0133, + "step": 24886 + }, + { + "epoch": 0.59, + "grad_norm": 2.1108307971711953, + "learning_rate": 7.710756656577764e-06, + "loss": 0.9857, + "step": 24887 + }, + { + "epoch": 0.59, + "grad_norm": 1.1744077303979148, + "learning_rate": 7.710013885333158e-06, + "loss": 0.9467, + "step": 24888 + }, + { + "epoch": 0.59, + "grad_norm": 2.0388616577391008, + "learning_rate": 7.70927112742157e-06, + "loss": 0.9422, + "step": 24889 + }, + { + "epoch": 0.59, + "grad_norm": 2.4127037747447098, + "learning_rate": 7.708528382847333e-06, + "loss": 0.9861, + "step": 24890 + }, + { + "epoch": 0.59, + "grad_norm": 2.2707180600599632, + "learning_rate": 7.70778565161477e-06, + "loss": 1.1578, + "step": 24891 + }, + { + "epoch": 0.59, + "grad_norm": 2.087917800303604, + "learning_rate": 7.707042933728202e-06, + "loss": 1.1181, + "step": 24892 + }, + { + "epoch": 0.59, + "grad_norm": 1.7816046782038975, + "learning_rate": 7.706300229191956e-06, + "loss": 1.0114, + "step": 24893 + }, + { + "epoch": 0.59, + "grad_norm": 2.091393423842671, + "learning_rate": 7.705557538010358e-06, + "loss": 1.0113, + "step": 24894 + }, + { + "epoch": 0.59, + "grad_norm": 2.588262426101066, + "learning_rate": 7.704814860187731e-06, + "loss": 1.0016, + "step": 24895 + }, + { + "epoch": 0.59, + "grad_norm": 1.9490266102134166, + "learning_rate": 7.704072195728396e-06, + "loss": 0.8863, + "step": 24896 + }, + { + "epoch": 0.59, + "grad_norm": 2.01210524135636, + "learning_rate": 7.70332954463668e-06, + "loss": 1.046, + "step": 24897 + }, + { + "epoch": 0.59, + "grad_norm": 1.9463363574483656, + "learning_rate": 7.702586906916911e-06, + "loss": 0.919, + "step": 24898 + }, + { + "epoch": 0.59, + "grad_norm": 2.6416967678012866, + "learning_rate": 7.701844282573405e-06, + "loss": 1.0261, + "step": 24899 + }, + { + "epoch": 0.59, + "grad_norm": 1.996627172606763, + "learning_rate": 7.701101671610485e-06, + "loss": 0.9628, + "step": 24900 + }, + { + "epoch": 0.59, + "grad_norm": 1.909836370370027, + "learning_rate": 7.700359074032483e-06, + "loss": 0.9349, + "step": 24901 + }, + { + "epoch": 0.59, + "grad_norm": 2.107194500125982, + "learning_rate": 7.699616489843723e-06, + "loss": 0.9708, + "step": 24902 + }, + { + "epoch": 0.59, + "grad_norm": 1.9257667068953692, + "learning_rate": 7.698873919048518e-06, + "loss": 0.9665, + "step": 24903 + }, + { + "epoch": 0.59, + "grad_norm": 1.9940944274208343, + "learning_rate": 7.698131361651196e-06, + "loss": 0.8878, + "step": 24904 + }, + { + "epoch": 0.59, + "grad_norm": 1.8928615622449476, + "learning_rate": 7.697388817656087e-06, + "loss": 0.9551, + "step": 24905 + }, + { + "epoch": 0.59, + "grad_norm": 2.2099264497674542, + "learning_rate": 7.696646287067509e-06, + "loss": 0.9947, + "step": 24906 + }, + { + "epoch": 0.59, + "grad_norm": 2.383937451910129, + "learning_rate": 7.695903769889783e-06, + "loss": 1.0157, + "step": 24907 + }, + { + "epoch": 0.59, + "grad_norm": 1.8728783151758852, + "learning_rate": 7.695161266127234e-06, + "loss": 0.9447, + "step": 24908 + }, + { + "epoch": 0.59, + "grad_norm": 1.0177667520993823, + "learning_rate": 7.69441877578419e-06, + "loss": 0.9293, + "step": 24909 + }, + { + "epoch": 0.59, + "grad_norm": 2.1144886908331646, + "learning_rate": 7.693676298864967e-06, + "loss": 1.0029, + "step": 24910 + }, + { + "epoch": 0.59, + "grad_norm": 1.9628353819261515, + "learning_rate": 7.69293383537389e-06, + "loss": 1.1053, + "step": 24911 + }, + { + "epoch": 0.59, + "grad_norm": 2.0656356684013213, + "learning_rate": 7.692191385315282e-06, + "loss": 0.9675, + "step": 24912 + }, + { + "epoch": 0.59, + "grad_norm": 1.970745954754648, + "learning_rate": 7.691448948693469e-06, + "loss": 0.9673, + "step": 24913 + }, + { + "epoch": 0.59, + "grad_norm": 1.8830166724310842, + "learning_rate": 7.690706525512771e-06, + "loss": 0.8797, + "step": 24914 + }, + { + "epoch": 0.59, + "grad_norm": 2.105138058803779, + "learning_rate": 7.689964115777507e-06, + "loss": 1.0025, + "step": 24915 + }, + { + "epoch": 0.59, + "grad_norm": 1.9224401584709714, + "learning_rate": 7.68922171949201e-06, + "loss": 0.9832, + "step": 24916 + }, + { + "epoch": 0.59, + "grad_norm": 3.009599219780683, + "learning_rate": 7.68847933666059e-06, + "loss": 1.1108, + "step": 24917 + }, + { + "epoch": 0.59, + "grad_norm": 1.919633392919275, + "learning_rate": 7.687736967287578e-06, + "loss": 1.0836, + "step": 24918 + }, + { + "epoch": 0.59, + "grad_norm": 1.6724175030905146, + "learning_rate": 7.68699461137729e-06, + "loss": 0.9843, + "step": 24919 + }, + { + "epoch": 0.59, + "grad_norm": 1.935398333442758, + "learning_rate": 7.686252268934058e-06, + "loss": 0.8544, + "step": 24920 + }, + { + "epoch": 0.59, + "grad_norm": 2.1782736646156424, + "learning_rate": 7.685509939962191e-06, + "loss": 1.0266, + "step": 24921 + }, + { + "epoch": 0.59, + "grad_norm": 1.7623136057565154, + "learning_rate": 7.68476762446602e-06, + "loss": 0.9651, + "step": 24922 + }, + { + "epoch": 0.59, + "grad_norm": 2.017760224608792, + "learning_rate": 7.684025322449869e-06, + "loss": 1.1192, + "step": 24923 + }, + { + "epoch": 0.59, + "grad_norm": 2.117161764053581, + "learning_rate": 7.683283033918052e-06, + "loss": 0.9468, + "step": 24924 + }, + { + "epoch": 0.59, + "grad_norm": 1.8539628478178807, + "learning_rate": 7.682540758874894e-06, + "loss": 1.0629, + "step": 24925 + }, + { + "epoch": 0.59, + "grad_norm": 2.098723763278599, + "learning_rate": 7.681798497324717e-06, + "loss": 0.9341, + "step": 24926 + }, + { + "epoch": 0.59, + "grad_norm": 2.0831539061016424, + "learning_rate": 7.681056249271848e-06, + "loss": 0.9417, + "step": 24927 + }, + { + "epoch": 0.59, + "grad_norm": 2.300828300858792, + "learning_rate": 7.680314014720597e-06, + "loss": 0.9308, + "step": 24928 + }, + { + "epoch": 0.59, + "grad_norm": 1.9345896034877184, + "learning_rate": 7.679571793675295e-06, + "loss": 1.0873, + "step": 24929 + }, + { + "epoch": 0.59, + "grad_norm": 2.0118057065521544, + "learning_rate": 7.678829586140263e-06, + "loss": 1.0193, + "step": 24930 + }, + { + "epoch": 0.59, + "grad_norm": 2.035908443566721, + "learning_rate": 7.678087392119818e-06, + "loss": 0.9711, + "step": 24931 + }, + { + "epoch": 0.59, + "grad_norm": 2.0307290476321564, + "learning_rate": 7.677345211618281e-06, + "loss": 0.9381, + "step": 24932 + }, + { + "epoch": 0.59, + "grad_norm": 1.8646288893241265, + "learning_rate": 7.676603044639978e-06, + "loss": 1.0199, + "step": 24933 + }, + { + "epoch": 0.59, + "grad_norm": 3.5173661790929818, + "learning_rate": 7.67586089118923e-06, + "loss": 1.063, + "step": 24934 + }, + { + "epoch": 0.59, + "grad_norm": 2.0938995772948865, + "learning_rate": 7.675118751270355e-06, + "loss": 0.8001, + "step": 24935 + }, + { + "epoch": 0.59, + "grad_norm": 1.165221671724306, + "learning_rate": 7.674376624887669e-06, + "loss": 1.0328, + "step": 24936 + }, + { + "epoch": 0.59, + "grad_norm": 1.9867950867451007, + "learning_rate": 7.673634512045506e-06, + "loss": 1.0468, + "step": 24937 + }, + { + "epoch": 0.59, + "grad_norm": 1.8246739877434917, + "learning_rate": 7.672892412748175e-06, + "loss": 1.069, + "step": 24938 + }, + { + "epoch": 0.59, + "grad_norm": 2.180441790069479, + "learning_rate": 7.672150327000002e-06, + "loss": 1.1383, + "step": 24939 + }, + { + "epoch": 0.59, + "grad_norm": 2.2121185600830042, + "learning_rate": 7.671408254805304e-06, + "loss": 0.9344, + "step": 24940 + }, + { + "epoch": 0.59, + "grad_norm": 1.838448813427678, + "learning_rate": 7.67066619616841e-06, + "loss": 0.9791, + "step": 24941 + }, + { + "epoch": 0.59, + "grad_norm": 1.8228144409546867, + "learning_rate": 7.669924151093632e-06, + "loss": 0.9081, + "step": 24942 + }, + { + "epoch": 0.59, + "grad_norm": 1.8165300407834748, + "learning_rate": 7.669182119585292e-06, + "loss": 0.8385, + "step": 24943 + }, + { + "epoch": 0.59, + "grad_norm": 2.0933059481269765, + "learning_rate": 7.668440101647717e-06, + "loss": 1.0138, + "step": 24944 + }, + { + "epoch": 0.59, + "grad_norm": 1.8021129467504964, + "learning_rate": 7.667698097285214e-06, + "loss": 1.0328, + "step": 24945 + }, + { + "epoch": 0.59, + "grad_norm": 2.0981441870217044, + "learning_rate": 7.666956106502115e-06, + "loss": 1.1007, + "step": 24946 + }, + { + "epoch": 0.59, + "grad_norm": 1.8631409962417942, + "learning_rate": 7.666214129302737e-06, + "loss": 0.9443, + "step": 24947 + }, + { + "epoch": 0.59, + "grad_norm": 2.194810129784501, + "learning_rate": 7.6654721656914e-06, + "loss": 1.0655, + "step": 24948 + }, + { + "epoch": 0.59, + "grad_norm": 2.2704990634954743, + "learning_rate": 7.664730215672418e-06, + "loss": 1.0401, + "step": 24949 + }, + { + "epoch": 0.59, + "grad_norm": 1.8997533509895321, + "learning_rate": 7.66398827925012e-06, + "loss": 0.9644, + "step": 24950 + }, + { + "epoch": 0.59, + "grad_norm": 2.8215867127661003, + "learning_rate": 7.663246356428824e-06, + "loss": 1.0237, + "step": 24951 + }, + { + "epoch": 0.59, + "grad_norm": 2.2345387769327187, + "learning_rate": 7.662504447212845e-06, + "loss": 1.0203, + "step": 24952 + }, + { + "epoch": 0.59, + "grad_norm": 2.268950808118555, + "learning_rate": 7.661762551606501e-06, + "loss": 1.0567, + "step": 24953 + }, + { + "epoch": 0.59, + "grad_norm": 2.437953941188173, + "learning_rate": 7.66102066961412e-06, + "loss": 1.0809, + "step": 24954 + }, + { + "epoch": 0.59, + "grad_norm": 1.0816366658839174, + "learning_rate": 7.66027880124002e-06, + "loss": 0.924, + "step": 24955 + }, + { + "epoch": 0.59, + "grad_norm": 2.088204787148614, + "learning_rate": 7.659536946488514e-06, + "loss": 1.0224, + "step": 24956 + }, + { + "epoch": 0.59, + "grad_norm": 1.9115815460016636, + "learning_rate": 7.658795105363922e-06, + "loss": 1.0164, + "step": 24957 + }, + { + "epoch": 0.59, + "grad_norm": 2.019621309572579, + "learning_rate": 7.658053277870571e-06, + "loss": 1.0459, + "step": 24958 + }, + { + "epoch": 0.59, + "grad_norm": 2.773103686092789, + "learning_rate": 7.657311464012774e-06, + "loss": 0.9672, + "step": 24959 + }, + { + "epoch": 0.59, + "grad_norm": 1.7854583022204689, + "learning_rate": 7.656569663794851e-06, + "loss": 1.0, + "step": 24960 + }, + { + "epoch": 0.59, + "grad_norm": 1.8307013007885726, + "learning_rate": 7.65582787722112e-06, + "loss": 1.0579, + "step": 24961 + }, + { + "epoch": 0.59, + "grad_norm": 2.0497184960652923, + "learning_rate": 7.655086104295904e-06, + "loss": 0.9727, + "step": 24962 + }, + { + "epoch": 0.59, + "grad_norm": 2.061412644989361, + "learning_rate": 7.654344345023517e-06, + "loss": 0.9916, + "step": 24963 + }, + { + "epoch": 0.59, + "grad_norm": 2.1491975974273037, + "learning_rate": 7.653602599408279e-06, + "loss": 1.0956, + "step": 24964 + }, + { + "epoch": 0.59, + "grad_norm": 2.311294400681671, + "learning_rate": 7.652860867454515e-06, + "loss": 0.9405, + "step": 24965 + }, + { + "epoch": 0.59, + "grad_norm": 1.9998904500654484, + "learning_rate": 7.652119149166533e-06, + "loss": 1.0328, + "step": 24966 + }, + { + "epoch": 0.59, + "grad_norm": 2.2729021407598022, + "learning_rate": 7.651377444548657e-06, + "loss": 0.9719, + "step": 24967 + }, + { + "epoch": 0.59, + "grad_norm": 2.159430372078373, + "learning_rate": 7.650635753605203e-06, + "loss": 0.9405, + "step": 24968 + }, + { + "epoch": 0.59, + "grad_norm": 2.2261417384533395, + "learning_rate": 7.649894076340499e-06, + "loss": 1.005, + "step": 24969 + }, + { + "epoch": 0.59, + "grad_norm": 1.9545885408751171, + "learning_rate": 7.649152412758848e-06, + "loss": 1.1655, + "step": 24970 + }, + { + "epoch": 0.59, + "grad_norm": 2.145788709093705, + "learning_rate": 7.64841076286458e-06, + "loss": 1.032, + "step": 24971 + }, + { + "epoch": 0.59, + "grad_norm": 1.727325772660338, + "learning_rate": 7.64766912666201e-06, + "loss": 1.0652, + "step": 24972 + }, + { + "epoch": 0.59, + "grad_norm": 1.8131854780711985, + "learning_rate": 7.646927504155452e-06, + "loss": 1.0103, + "step": 24973 + }, + { + "epoch": 0.59, + "grad_norm": 2.0286541038753896, + "learning_rate": 7.646185895349227e-06, + "loss": 1.039, + "step": 24974 + }, + { + "epoch": 0.59, + "grad_norm": 1.8362505219824292, + "learning_rate": 7.645444300247654e-06, + "loss": 0.9089, + "step": 24975 + }, + { + "epoch": 0.59, + "grad_norm": 1.9759179923794261, + "learning_rate": 7.64470271885505e-06, + "loss": 0.9309, + "step": 24976 + }, + { + "epoch": 0.59, + "grad_norm": 1.1488829342107092, + "learning_rate": 7.64396115117573e-06, + "loss": 0.9299, + "step": 24977 + }, + { + "epoch": 0.59, + "grad_norm": 2.0258682674322612, + "learning_rate": 7.643219597214017e-06, + "loss": 1.0463, + "step": 24978 + }, + { + "epoch": 0.59, + "grad_norm": 2.1598621465898447, + "learning_rate": 7.642478056974228e-06, + "loss": 1.1075, + "step": 24979 + }, + { + "epoch": 0.59, + "grad_norm": 2.1119227565601246, + "learning_rate": 7.641736530460673e-06, + "loss": 1.0442, + "step": 24980 + }, + { + "epoch": 0.59, + "grad_norm": 2.0072486139886268, + "learning_rate": 7.640995017677674e-06, + "loss": 0.9154, + "step": 24981 + }, + { + "epoch": 0.59, + "grad_norm": 2.1182873424226947, + "learning_rate": 7.640253518629553e-06, + "loss": 1.0046, + "step": 24982 + }, + { + "epoch": 0.59, + "grad_norm": 2.388442257027223, + "learning_rate": 7.639512033320623e-06, + "loss": 1.0079, + "step": 24983 + }, + { + "epoch": 0.59, + "grad_norm": 1.0769236347128748, + "learning_rate": 7.6387705617552e-06, + "loss": 0.912, + "step": 24984 + }, + { + "epoch": 0.59, + "grad_norm": 2.2141975484261596, + "learning_rate": 7.6380291039376e-06, + "loss": 1.1038, + "step": 24985 + }, + { + "epoch": 0.59, + "grad_norm": 1.946462559611256, + "learning_rate": 7.637287659872147e-06, + "loss": 0.997, + "step": 24986 + }, + { + "epoch": 0.59, + "grad_norm": 2.0140329791868483, + "learning_rate": 7.63654622956315e-06, + "loss": 1.058, + "step": 24987 + }, + { + "epoch": 0.59, + "grad_norm": 2.1685055633560695, + "learning_rate": 7.63580481301493e-06, + "loss": 0.9954, + "step": 24988 + }, + { + "epoch": 0.59, + "grad_norm": 1.7987934707457611, + "learning_rate": 7.6350634102318e-06, + "loss": 1.0776, + "step": 24989 + }, + { + "epoch": 0.59, + "grad_norm": 2.057321268509511, + "learning_rate": 7.634322021218085e-06, + "loss": 0.911, + "step": 24990 + }, + { + "epoch": 0.59, + "grad_norm": 2.2937336901622776, + "learning_rate": 7.633580645978093e-06, + "loss": 1.1161, + "step": 24991 + }, + { + "epoch": 0.59, + "grad_norm": 1.0722729143301417, + "learning_rate": 7.632839284516144e-06, + "loss": 0.9586, + "step": 24992 + }, + { + "epoch": 0.59, + "grad_norm": 2.0096362591786976, + "learning_rate": 7.632097936836555e-06, + "loss": 0.9613, + "step": 24993 + }, + { + "epoch": 0.59, + "grad_norm": 1.906161250648618, + "learning_rate": 7.63135660294364e-06, + "loss": 0.9078, + "step": 24994 + }, + { + "epoch": 0.59, + "grad_norm": 1.920454498259686, + "learning_rate": 7.630615282841717e-06, + "loss": 0.9816, + "step": 24995 + }, + { + "epoch": 0.59, + "grad_norm": 3.6416640776932745, + "learning_rate": 7.629873976535102e-06, + "loss": 0.8675, + "step": 24996 + }, + { + "epoch": 0.59, + "grad_norm": 1.0543459212206168, + "learning_rate": 7.629132684028113e-06, + "loss": 0.9057, + "step": 24997 + }, + { + "epoch": 0.59, + "grad_norm": 1.9205783186111882, + "learning_rate": 7.6283914053250605e-06, + "loss": 0.9833, + "step": 24998 + }, + { + "epoch": 0.59, + "grad_norm": 2.5767834859146754, + "learning_rate": 7.627650140430267e-06, + "loss": 0.9503, + "step": 24999 + }, + { + "epoch": 0.59, + "grad_norm": 2.068142460803821, + "learning_rate": 7.626908889348046e-06, + "loss": 0.9896, + "step": 25000 + }, + { + "epoch": 0.59, + "grad_norm": 2.107547583136359, + "learning_rate": 7.626167652082712e-06, + "loss": 1.0629, + "step": 25001 + }, + { + "epoch": 0.59, + "grad_norm": 1.866349139553346, + "learning_rate": 7.625426428638579e-06, + "loss": 0.9865, + "step": 25002 + }, + { + "epoch": 0.59, + "grad_norm": 2.264533512613558, + "learning_rate": 7.624685219019966e-06, + "loss": 0.9725, + "step": 25003 + }, + { + "epoch": 0.59, + "grad_norm": 2.0641435912843216, + "learning_rate": 7.623944023231191e-06, + "loss": 1.001, + "step": 25004 + }, + { + "epoch": 0.59, + "grad_norm": 2.154296021459156, + "learning_rate": 7.623202841276564e-06, + "loss": 0.877, + "step": 25005 + }, + { + "epoch": 0.59, + "grad_norm": 2.085747852671835, + "learning_rate": 7.6224616731604e-06, + "loss": 0.9527, + "step": 25006 + }, + { + "epoch": 0.59, + "grad_norm": 2.075763600121486, + "learning_rate": 7.621720518887022e-06, + "loss": 0.9875, + "step": 25007 + }, + { + "epoch": 0.59, + "grad_norm": 1.9923716459998158, + "learning_rate": 7.620979378460736e-06, + "loss": 0.9612, + "step": 25008 + }, + { + "epoch": 0.59, + "grad_norm": 2.2046084391590157, + "learning_rate": 7.620238251885863e-06, + "loss": 1.0282, + "step": 25009 + }, + { + "epoch": 0.59, + "grad_norm": 1.123107255577144, + "learning_rate": 7.6194971391667126e-06, + "loss": 1.0363, + "step": 25010 + }, + { + "epoch": 0.59, + "grad_norm": 3.1289847593452986, + "learning_rate": 7.618756040307609e-06, + "loss": 0.9543, + "step": 25011 + }, + { + "epoch": 0.59, + "grad_norm": 1.1274969801581818, + "learning_rate": 7.61801495531286e-06, + "loss": 0.8622, + "step": 25012 + }, + { + "epoch": 0.59, + "grad_norm": 1.6515832488386102, + "learning_rate": 7.617273884186779e-06, + "loss": 0.9385, + "step": 25013 + }, + { + "epoch": 0.59, + "grad_norm": 2.086571613779699, + "learning_rate": 7.6165328269336905e-06, + "loss": 1.0011, + "step": 25014 + }, + { + "epoch": 0.59, + "grad_norm": 2.18585427230097, + "learning_rate": 7.615791783557895e-06, + "loss": 0.8999, + "step": 25015 + }, + { + "epoch": 0.59, + "grad_norm": 3.3052729982451208, + "learning_rate": 7.615050754063717e-06, + "loss": 0.9878, + "step": 25016 + }, + { + "epoch": 0.59, + "grad_norm": 1.0442280544810154, + "learning_rate": 7.614309738455468e-06, + "loss": 0.938, + "step": 25017 + }, + { + "epoch": 0.59, + "grad_norm": 2.51286554779154, + "learning_rate": 7.613568736737469e-06, + "loss": 0.9512, + "step": 25018 + }, + { + "epoch": 0.59, + "grad_norm": 2.1867381668114643, + "learning_rate": 7.612827748914021e-06, + "loss": 0.9718, + "step": 25019 + }, + { + "epoch": 0.59, + "grad_norm": 2.478163089322679, + "learning_rate": 7.612086774989449e-06, + "loss": 0.9908, + "step": 25020 + }, + { + "epoch": 0.59, + "grad_norm": 1.149291398103452, + "learning_rate": 7.611345814968065e-06, + "loss": 0.8972, + "step": 25021 + }, + { + "epoch": 0.59, + "grad_norm": 1.9328949077816897, + "learning_rate": 7.610604868854179e-06, + "loss": 1.0135, + "step": 25022 + }, + { + "epoch": 0.59, + "grad_norm": 1.9983788503033182, + "learning_rate": 7.609863936652106e-06, + "loss": 1.0457, + "step": 25023 + }, + { + "epoch": 0.59, + "grad_norm": 2.0663669175459165, + "learning_rate": 7.6091230183661655e-06, + "loss": 1.0884, + "step": 25024 + }, + { + "epoch": 0.59, + "grad_norm": 1.9312081095026012, + "learning_rate": 7.60838211400067e-06, + "loss": 0.9947, + "step": 25025 + }, + { + "epoch": 0.59, + "grad_norm": 2.2099350136733715, + "learning_rate": 7.607641223559925e-06, + "loss": 1.0543, + "step": 25026 + }, + { + "epoch": 0.59, + "grad_norm": 1.9313033572975138, + "learning_rate": 7.606900347048254e-06, + "loss": 0.9599, + "step": 25027 + }, + { + "epoch": 0.59, + "grad_norm": 1.928710395474901, + "learning_rate": 7.6061594844699645e-06, + "loss": 1.0551, + "step": 25028 + }, + { + "epoch": 0.59, + "grad_norm": 1.9276856128318962, + "learning_rate": 7.6054186358293765e-06, + "loss": 1.0158, + "step": 25029 + }, + { + "epoch": 0.59, + "grad_norm": 2.1056218067317225, + "learning_rate": 7.604677801130794e-06, + "loss": 1.0615, + "step": 25030 + }, + { + "epoch": 0.59, + "grad_norm": 2.3155884869283034, + "learning_rate": 7.603936980378539e-06, + "loss": 1.0036, + "step": 25031 + }, + { + "epoch": 0.59, + "grad_norm": 2.0793617120717522, + "learning_rate": 7.603196173576923e-06, + "loss": 1.0477, + "step": 25032 + }, + { + "epoch": 0.59, + "grad_norm": 1.8389828439696265, + "learning_rate": 7.602455380730256e-06, + "loss": 0.8085, + "step": 25033 + }, + { + "epoch": 0.59, + "grad_norm": 1.9958501351498719, + "learning_rate": 7.601714601842849e-06, + "loss": 0.9841, + "step": 25034 + }, + { + "epoch": 0.59, + "grad_norm": 1.0328168956122734, + "learning_rate": 7.600973836919022e-06, + "loss": 0.9717, + "step": 25035 + }, + { + "epoch": 0.59, + "grad_norm": 1.969776401399896, + "learning_rate": 7.600233085963088e-06, + "loss": 1.0378, + "step": 25036 + }, + { + "epoch": 0.59, + "grad_norm": 2.70714352415839, + "learning_rate": 7.5994923489793535e-06, + "loss": 0.9496, + "step": 25037 + }, + { + "epoch": 0.59, + "grad_norm": 2.137453846409648, + "learning_rate": 7.598751625972133e-06, + "loss": 0.8884, + "step": 25038 + }, + { + "epoch": 0.59, + "grad_norm": 2.140916986667074, + "learning_rate": 7.598010916945744e-06, + "loss": 1.0702, + "step": 25039 + }, + { + "epoch": 0.59, + "grad_norm": 1.8419766338996388, + "learning_rate": 7.597270221904494e-06, + "loss": 0.87, + "step": 25040 + }, + { + "epoch": 0.59, + "grad_norm": 1.1530213842757324, + "learning_rate": 7.596529540852699e-06, + "loss": 0.9707, + "step": 25041 + }, + { + "epoch": 0.59, + "grad_norm": 2.251100293894206, + "learning_rate": 7.595788873794666e-06, + "loss": 1.0749, + "step": 25042 + }, + { + "epoch": 0.59, + "grad_norm": 2.2062020552335206, + "learning_rate": 7.5950482207347165e-06, + "loss": 0.9015, + "step": 25043 + }, + { + "epoch": 0.59, + "grad_norm": 2.1377598449765554, + "learning_rate": 7.594307581677155e-06, + "loss": 0.9898, + "step": 25044 + }, + { + "epoch": 0.59, + "grad_norm": 2.0100922622626083, + "learning_rate": 7.5935669566262974e-06, + "loss": 0.9806, + "step": 25045 + }, + { + "epoch": 0.59, + "grad_norm": 2.104809430897927, + "learning_rate": 7.592826345586456e-06, + "loss": 1.0842, + "step": 25046 + }, + { + "epoch": 0.59, + "grad_norm": 1.8246717272225161, + "learning_rate": 7.592085748561937e-06, + "loss": 1.127, + "step": 25047 + }, + { + "epoch": 0.59, + "grad_norm": 1.8066571359910923, + "learning_rate": 7.5913451655570604e-06, + "loss": 0.9893, + "step": 25048 + }, + { + "epoch": 0.59, + "grad_norm": 2.0022902464211456, + "learning_rate": 7.590604596576134e-06, + "loss": 0.9754, + "step": 25049 + }, + { + "epoch": 0.59, + "grad_norm": 2.107516578079035, + "learning_rate": 7.589864041623473e-06, + "loss": 0.918, + "step": 25050 + }, + { + "epoch": 0.59, + "grad_norm": 1.7754238750315365, + "learning_rate": 7.589123500703383e-06, + "loss": 0.9476, + "step": 25051 + }, + { + "epoch": 0.59, + "grad_norm": 2.018827856798318, + "learning_rate": 7.588382973820181e-06, + "loss": 1.0542, + "step": 25052 + }, + { + "epoch": 0.59, + "grad_norm": 1.9669817434648593, + "learning_rate": 7.587642460978178e-06, + "loss": 0.9757, + "step": 25053 + }, + { + "epoch": 0.59, + "grad_norm": 2.089855682258767, + "learning_rate": 7.586901962181683e-06, + "loss": 0.9872, + "step": 25054 + }, + { + "epoch": 0.59, + "grad_norm": 1.9685533038343686, + "learning_rate": 7.586161477435006e-06, + "loss": 1.1055, + "step": 25055 + }, + { + "epoch": 0.59, + "grad_norm": 1.9561564926787902, + "learning_rate": 7.585421006742464e-06, + "loss": 0.919, + "step": 25056 + }, + { + "epoch": 0.59, + "grad_norm": 2.072786514578391, + "learning_rate": 7.584680550108367e-06, + "loss": 0.8933, + "step": 25057 + }, + { + "epoch": 0.59, + "grad_norm": 2.4000112468113235, + "learning_rate": 7.5839401075370225e-06, + "loss": 0.9602, + "step": 25058 + }, + { + "epoch": 0.59, + "grad_norm": 2.273132001931453, + "learning_rate": 7.583199679032741e-06, + "loss": 0.9664, + "step": 25059 + }, + { + "epoch": 0.59, + "grad_norm": 2.2881279302277475, + "learning_rate": 7.582459264599842e-06, + "loss": 0.9773, + "step": 25060 + }, + { + "epoch": 0.59, + "grad_norm": 1.8288040197257895, + "learning_rate": 7.581718864242626e-06, + "loss": 0.9136, + "step": 25061 + }, + { + "epoch": 0.59, + "grad_norm": 2.0241305997675627, + "learning_rate": 7.580978477965408e-06, + "loss": 0.9747, + "step": 25062 + }, + { + "epoch": 0.59, + "grad_norm": 1.061571515206616, + "learning_rate": 7.5802381057725e-06, + "loss": 0.9769, + "step": 25063 + }, + { + "epoch": 0.59, + "grad_norm": 1.9523851533562318, + "learning_rate": 7.579497747668216e-06, + "loss": 1.0615, + "step": 25064 + }, + { + "epoch": 0.59, + "grad_norm": 2.2304038675737123, + "learning_rate": 7.578757403656859e-06, + "loss": 0.8971, + "step": 25065 + }, + { + "epoch": 0.59, + "grad_norm": 1.7709675215474305, + "learning_rate": 7.578017073742741e-06, + "loss": 0.9625, + "step": 25066 + }, + { + "epoch": 0.59, + "grad_norm": 2.215932845838672, + "learning_rate": 7.577276757930181e-06, + "loss": 0.9377, + "step": 25067 + }, + { + "epoch": 0.59, + "grad_norm": 2.0015885802728093, + "learning_rate": 7.5765364562234756e-06, + "loss": 1.0903, + "step": 25068 + }, + { + "epoch": 0.59, + "grad_norm": 2.0829730071206085, + "learning_rate": 7.575796168626944e-06, + "loss": 0.8826, + "step": 25069 + }, + { + "epoch": 0.59, + "grad_norm": 1.1572627455249598, + "learning_rate": 7.575055895144894e-06, + "loss": 0.9682, + "step": 25070 + }, + { + "epoch": 0.59, + "grad_norm": 2.279430236159231, + "learning_rate": 7.574315635781642e-06, + "loss": 0.9435, + "step": 25071 + }, + { + "epoch": 0.59, + "grad_norm": 2.1336408857185267, + "learning_rate": 7.573575390541486e-06, + "loss": 1.0399, + "step": 25072 + }, + { + "epoch": 0.59, + "grad_norm": 2.188822836974198, + "learning_rate": 7.572835159428744e-06, + "loss": 0.9798, + "step": 25073 + }, + { + "epoch": 0.59, + "grad_norm": 1.858035291650804, + "learning_rate": 7.572094942447727e-06, + "loss": 1.0355, + "step": 25074 + }, + { + "epoch": 0.59, + "grad_norm": 2.2080623900667082, + "learning_rate": 7.571354739602736e-06, + "loss": 0.965, + "step": 25075 + }, + { + "epoch": 0.59, + "grad_norm": 2.135149805133918, + "learning_rate": 7.57061455089809e-06, + "loss": 1.1367, + "step": 25076 + }, + { + "epoch": 0.59, + "grad_norm": 2.186153469160211, + "learning_rate": 7.569874376338095e-06, + "loss": 0.9683, + "step": 25077 + }, + { + "epoch": 0.59, + "grad_norm": 1.968055519705544, + "learning_rate": 7.569134215927063e-06, + "loss": 1.1361, + "step": 25078 + }, + { + "epoch": 0.59, + "grad_norm": 1.1101954941999244, + "learning_rate": 7.568394069669297e-06, + "loss": 0.9214, + "step": 25079 + }, + { + "epoch": 0.59, + "grad_norm": 1.962877479789141, + "learning_rate": 7.567653937569113e-06, + "loss": 1.0797, + "step": 25080 + }, + { + "epoch": 0.59, + "grad_norm": 1.9122681315604688, + "learning_rate": 7.566913819630819e-06, + "loss": 0.9903, + "step": 25081 + }, + { + "epoch": 0.59, + "grad_norm": 2.2415682578464944, + "learning_rate": 7.566173715858722e-06, + "loss": 0.9833, + "step": 25082 + }, + { + "epoch": 0.59, + "grad_norm": 2.0073915503100217, + "learning_rate": 7.565433626257129e-06, + "loss": 0.9676, + "step": 25083 + }, + { + "epoch": 0.59, + "grad_norm": 2.736678590801133, + "learning_rate": 7.564693550830355e-06, + "loss": 0.8274, + "step": 25084 + }, + { + "epoch": 0.59, + "grad_norm": 2.6087373660069657, + "learning_rate": 7.563953489582708e-06, + "loss": 1.0093, + "step": 25085 + }, + { + "epoch": 0.59, + "grad_norm": 1.8115526694383854, + "learning_rate": 7.5632134425184936e-06, + "loss": 0.8598, + "step": 25086 + }, + { + "epoch": 0.59, + "grad_norm": 2.0872467155888965, + "learning_rate": 7.562473409642019e-06, + "loss": 1.092, + "step": 25087 + }, + { + "epoch": 0.59, + "grad_norm": 2.0694617132048974, + "learning_rate": 7.5617333909576006e-06, + "loss": 0.9599, + "step": 25088 + }, + { + "epoch": 0.59, + "grad_norm": 1.9014187637067868, + "learning_rate": 7.56099338646954e-06, + "loss": 0.8409, + "step": 25089 + }, + { + "epoch": 0.59, + "grad_norm": 1.9942253211097052, + "learning_rate": 7.560253396182148e-06, + "loss": 0.9498, + "step": 25090 + }, + { + "epoch": 0.59, + "grad_norm": 2.1417007771204943, + "learning_rate": 7.559513420099731e-06, + "loss": 0.9685, + "step": 25091 + }, + { + "epoch": 0.59, + "grad_norm": 2.1772873479541124, + "learning_rate": 7.558773458226605e-06, + "loss": 0.9006, + "step": 25092 + }, + { + "epoch": 0.59, + "grad_norm": 14.980093673588664, + "learning_rate": 7.558033510567069e-06, + "loss": 0.9368, + "step": 25093 + }, + { + "epoch": 0.59, + "grad_norm": 2.119280208877612, + "learning_rate": 7.557293577125435e-06, + "loss": 1.1194, + "step": 25094 + }, + { + "epoch": 0.59, + "grad_norm": 1.7888983311250048, + "learning_rate": 7.556553657906014e-06, + "loss": 1.0275, + "step": 25095 + }, + { + "epoch": 0.59, + "grad_norm": 1.9112226829830226, + "learning_rate": 7.555813752913107e-06, + "loss": 1.0031, + "step": 25096 + }, + { + "epoch": 0.59, + "grad_norm": 2.1790313954171516, + "learning_rate": 7.555073862151028e-06, + "loss": 0.8846, + "step": 25097 + }, + { + "epoch": 0.59, + "grad_norm": 1.9971241733287743, + "learning_rate": 7.554333985624082e-06, + "loss": 1.1147, + "step": 25098 + }, + { + "epoch": 0.59, + "grad_norm": 2.142998420458736, + "learning_rate": 7.553594123336583e-06, + "loss": 0.9573, + "step": 25099 + }, + { + "epoch": 0.59, + "grad_norm": 1.9454538369874828, + "learning_rate": 7.5528542752928265e-06, + "loss": 1.0484, + "step": 25100 + }, + { + "epoch": 0.59, + "grad_norm": 1.9431919488838336, + "learning_rate": 7.552114441497131e-06, + "loss": 1.0625, + "step": 25101 + }, + { + "epoch": 0.59, + "grad_norm": 1.0859540145419304, + "learning_rate": 7.551374621953801e-06, + "loss": 0.9484, + "step": 25102 + }, + { + "epoch": 0.59, + "grad_norm": 2.363401549751261, + "learning_rate": 7.550634816667142e-06, + "loss": 0.9614, + "step": 25103 + }, + { + "epoch": 0.59, + "grad_norm": 2.045212584957268, + "learning_rate": 7.549895025641459e-06, + "loss": 1.0335, + "step": 25104 + }, + { + "epoch": 0.59, + "grad_norm": 1.0503095546159444, + "learning_rate": 7.549155248881066e-06, + "loss": 0.991, + "step": 25105 + }, + { + "epoch": 0.59, + "grad_norm": 2.522113319935139, + "learning_rate": 7.5484154863902705e-06, + "loss": 0.9405, + "step": 25106 + }, + { + "epoch": 0.59, + "grad_norm": 2.148494295485757, + "learning_rate": 7.547675738173372e-06, + "loss": 1.0188, + "step": 25107 + }, + { + "epoch": 0.59, + "grad_norm": 1.053824558045222, + "learning_rate": 7.546936004234679e-06, + "loss": 0.9414, + "step": 25108 + }, + { + "epoch": 0.59, + "grad_norm": 1.7155957403406474, + "learning_rate": 7.546196284578509e-06, + "loss": 0.9555, + "step": 25109 + }, + { + "epoch": 0.59, + "grad_norm": 2.6541471261889824, + "learning_rate": 7.545456579209156e-06, + "loss": 0.9781, + "step": 25110 + }, + { + "epoch": 0.59, + "grad_norm": 2.0718967909929207, + "learning_rate": 7.5447168881309295e-06, + "loss": 0.9739, + "step": 25111 + }, + { + "epoch": 0.59, + "grad_norm": 2.058196577021727, + "learning_rate": 7.543977211348142e-06, + "loss": 1.137, + "step": 25112 + }, + { + "epoch": 0.59, + "grad_norm": 2.011138428968893, + "learning_rate": 7.543237548865099e-06, + "loss": 0.9014, + "step": 25113 + }, + { + "epoch": 0.59, + "grad_norm": 2.017144830081699, + "learning_rate": 7.542497900686102e-06, + "loss": 0.9026, + "step": 25114 + }, + { + "epoch": 0.59, + "grad_norm": 1.9550539056547795, + "learning_rate": 7.5417582668154585e-06, + "loss": 0.8989, + "step": 25115 + }, + { + "epoch": 0.59, + "grad_norm": 2.013429974646887, + "learning_rate": 7.5410186472574835e-06, + "loss": 0.9569, + "step": 25116 + }, + { + "epoch": 0.59, + "grad_norm": 2.0415926027635094, + "learning_rate": 7.540279042016469e-06, + "loss": 1.0687, + "step": 25117 + }, + { + "epoch": 0.59, + "grad_norm": 2.0011373051453116, + "learning_rate": 7.539539451096732e-06, + "loss": 0.8587, + "step": 25118 + }, + { + "epoch": 0.59, + "grad_norm": 2.11527274738476, + "learning_rate": 7.538799874502572e-06, + "loss": 1.0062, + "step": 25119 + }, + { + "epoch": 0.59, + "grad_norm": 2.4722920012208465, + "learning_rate": 7.538060312238304e-06, + "loss": 0.9709, + "step": 25120 + }, + { + "epoch": 0.59, + "grad_norm": 2.0705434243052876, + "learning_rate": 7.5373207643082245e-06, + "loss": 1.0604, + "step": 25121 + }, + { + "epoch": 0.59, + "grad_norm": 1.8940380551942453, + "learning_rate": 7.536581230716643e-06, + "loss": 1.0286, + "step": 25122 + }, + { + "epoch": 0.59, + "grad_norm": 2.03427233819207, + "learning_rate": 7.535841711467868e-06, + "loss": 0.9898, + "step": 25123 + }, + { + "epoch": 0.59, + "grad_norm": 1.7361720958924447, + "learning_rate": 7.535102206566199e-06, + "loss": 0.9326, + "step": 25124 + }, + { + "epoch": 0.59, + "grad_norm": 2.0848792717861975, + "learning_rate": 7.534362716015948e-06, + "loss": 1.0567, + "step": 25125 + }, + { + "epoch": 0.59, + "grad_norm": 2.0188471865262483, + "learning_rate": 7.533623239821417e-06, + "loss": 0.9796, + "step": 25126 + }, + { + "epoch": 0.59, + "grad_norm": 2.1289895926138236, + "learning_rate": 7.532883777986915e-06, + "loss": 0.9808, + "step": 25127 + }, + { + "epoch": 0.59, + "grad_norm": 1.920287168576671, + "learning_rate": 7.53214433051674e-06, + "loss": 0.9748, + "step": 25128 + }, + { + "epoch": 0.59, + "grad_norm": 1.9440163848403786, + "learning_rate": 7.531404897415204e-06, + "loss": 1.0411, + "step": 25129 + }, + { + "epoch": 0.59, + "grad_norm": 2.2825118952816714, + "learning_rate": 7.530665478686613e-06, + "loss": 1.1285, + "step": 25130 + }, + { + "epoch": 0.59, + "grad_norm": 1.1104479039601831, + "learning_rate": 7.5299260743352675e-06, + "loss": 0.9725, + "step": 25131 + }, + { + "epoch": 0.59, + "grad_norm": 1.8301498711739939, + "learning_rate": 7.529186684365471e-06, + "loss": 0.9113, + "step": 25132 + }, + { + "epoch": 0.59, + "grad_norm": 1.9514780887776029, + "learning_rate": 7.528447308781534e-06, + "loss": 0.8857, + "step": 25133 + }, + { + "epoch": 0.59, + "grad_norm": 2.4501586906004227, + "learning_rate": 7.527707947587764e-06, + "loss": 0.9753, + "step": 25134 + }, + { + "epoch": 0.59, + "grad_norm": 1.955044924752329, + "learning_rate": 7.526968600788456e-06, + "loss": 0.9174, + "step": 25135 + }, + { + "epoch": 0.59, + "grad_norm": 2.448005702650761, + "learning_rate": 7.526229268387919e-06, + "loss": 0.9853, + "step": 25136 + }, + { + "epoch": 0.59, + "grad_norm": 1.8767173644576367, + "learning_rate": 7.525489950390461e-06, + "loss": 1.0068, + "step": 25137 + }, + { + "epoch": 0.59, + "grad_norm": 1.7865361546397114, + "learning_rate": 7.524750646800384e-06, + "loss": 1.0173, + "step": 25138 + }, + { + "epoch": 0.59, + "grad_norm": 2.1501020168874954, + "learning_rate": 7.524011357621991e-06, + "loss": 0.9931, + "step": 25139 + }, + { + "epoch": 0.59, + "grad_norm": 1.983481612731817, + "learning_rate": 7.523272082859587e-06, + "loss": 1.0301, + "step": 25140 + }, + { + "epoch": 0.59, + "grad_norm": 1.744146761264024, + "learning_rate": 7.52253282251748e-06, + "loss": 0.9535, + "step": 25141 + }, + { + "epoch": 0.59, + "grad_norm": 1.1220496997898757, + "learning_rate": 7.52179357659997e-06, + "loss": 0.9788, + "step": 25142 + }, + { + "epoch": 0.59, + "grad_norm": 2.0665463175202037, + "learning_rate": 7.521054345111361e-06, + "loss": 0.9522, + "step": 25143 + }, + { + "epoch": 0.59, + "grad_norm": 2.5945382638993406, + "learning_rate": 7.520315128055961e-06, + "loss": 0.8479, + "step": 25144 + }, + { + "epoch": 0.59, + "grad_norm": 2.3889271984900944, + "learning_rate": 7.5195759254380674e-06, + "loss": 1.0399, + "step": 25145 + }, + { + "epoch": 0.59, + "grad_norm": 1.8155743122976167, + "learning_rate": 7.518836737261991e-06, + "loss": 0.8735, + "step": 25146 + }, + { + "epoch": 0.59, + "grad_norm": 2.170018274117105, + "learning_rate": 7.51809756353203e-06, + "loss": 1.0999, + "step": 25147 + }, + { + "epoch": 0.59, + "grad_norm": 1.8058445344758034, + "learning_rate": 7.517358404252497e-06, + "loss": 0.9163, + "step": 25148 + }, + { + "epoch": 0.59, + "grad_norm": 2.116950838077759, + "learning_rate": 7.516619259427682e-06, + "loss": 1.0118, + "step": 25149 + }, + { + "epoch": 0.59, + "grad_norm": 2.095590665627119, + "learning_rate": 7.5158801290619e-06, + "loss": 1.0131, + "step": 25150 + }, + { + "epoch": 0.59, + "grad_norm": 1.9587188592586287, + "learning_rate": 7.515141013159447e-06, + "loss": 1.0529, + "step": 25151 + }, + { + "epoch": 0.59, + "grad_norm": 2.1899640878461772, + "learning_rate": 7.514401911724636e-06, + "loss": 1.076, + "step": 25152 + }, + { + "epoch": 0.59, + "grad_norm": 2.0345343245393774, + "learning_rate": 7.513662824761758e-06, + "loss": 0.965, + "step": 25153 + }, + { + "epoch": 0.59, + "grad_norm": 2.025518875535654, + "learning_rate": 7.512923752275124e-06, + "loss": 0.8842, + "step": 25154 + }, + { + "epoch": 0.59, + "grad_norm": 2.4035699661214904, + "learning_rate": 7.5121846942690376e-06, + "loss": 0.914, + "step": 25155 + }, + { + "epoch": 0.59, + "grad_norm": 2.8525631016192077, + "learning_rate": 7.511445650747797e-06, + "loss": 1.008, + "step": 25156 + }, + { + "epoch": 0.59, + "grad_norm": 2.1804277326789725, + "learning_rate": 7.5107066217157065e-06, + "loss": 0.8733, + "step": 25157 + }, + { + "epoch": 0.59, + "grad_norm": 2.2237237874830678, + "learning_rate": 7.509967607177072e-06, + "loss": 1.0363, + "step": 25158 + }, + { + "epoch": 0.59, + "grad_norm": 1.1073029701752024, + "learning_rate": 7.509228607136196e-06, + "loss": 0.9402, + "step": 25159 + }, + { + "epoch": 0.59, + "grad_norm": 2.037036108315996, + "learning_rate": 7.508489621597376e-06, + "loss": 1.091, + "step": 25160 + }, + { + "epoch": 0.59, + "grad_norm": 2.0136346989334784, + "learning_rate": 7.50775065056492e-06, + "loss": 0.979, + "step": 25161 + }, + { + "epoch": 0.59, + "grad_norm": 2.0673775845719673, + "learning_rate": 7.5070116940431315e-06, + "loss": 1.1121, + "step": 25162 + }, + { + "epoch": 0.59, + "grad_norm": 1.9751614193708729, + "learning_rate": 7.506272752036308e-06, + "loss": 0.9766, + "step": 25163 + }, + { + "epoch": 0.59, + "grad_norm": 2.055961227118269, + "learning_rate": 7.5055338245487506e-06, + "loss": 0.8993, + "step": 25164 + }, + { + "epoch": 0.59, + "grad_norm": 1.966501228598087, + "learning_rate": 7.504794911584767e-06, + "loss": 1.1195, + "step": 25165 + }, + { + "epoch": 0.59, + "grad_norm": 2.322999615651181, + "learning_rate": 7.504056013148661e-06, + "loss": 0.9979, + "step": 25166 + }, + { + "epoch": 0.59, + "grad_norm": 1.929642783932605, + "learning_rate": 7.503317129244729e-06, + "loss": 0.9389, + "step": 25167 + }, + { + "epoch": 0.59, + "grad_norm": 1.9705293072367451, + "learning_rate": 7.502578259877272e-06, + "loss": 0.8559, + "step": 25168 + }, + { + "epoch": 0.59, + "grad_norm": 6.073698938151286, + "learning_rate": 7.501839405050601e-06, + "loss": 0.8318, + "step": 25169 + }, + { + "epoch": 0.59, + "grad_norm": 2.379735308812344, + "learning_rate": 7.501100564769007e-06, + "loss": 0.95, + "step": 25170 + }, + { + "epoch": 0.59, + "grad_norm": 2.0625941568772728, + "learning_rate": 7.500361739036799e-06, + "loss": 0.9365, + "step": 25171 + }, + { + "epoch": 0.59, + "grad_norm": 1.064640674561013, + "learning_rate": 7.499622927858272e-06, + "loss": 1.0152, + "step": 25172 + }, + { + "epoch": 0.59, + "grad_norm": 2.5504803100290774, + "learning_rate": 7.498884131237738e-06, + "loss": 0.9324, + "step": 25173 + }, + { + "epoch": 0.59, + "grad_norm": 1.9109966442035273, + "learning_rate": 7.49814534917949e-06, + "loss": 0.9715, + "step": 25174 + }, + { + "epoch": 0.59, + "grad_norm": 2.2358868611280878, + "learning_rate": 7.497406581687832e-06, + "loss": 1.0546, + "step": 25175 + }, + { + "epoch": 0.59, + "grad_norm": 1.9034557055832317, + "learning_rate": 7.496667828767066e-06, + "loss": 1.0425, + "step": 25176 + }, + { + "epoch": 0.59, + "grad_norm": 2.0423881685799223, + "learning_rate": 7.4959290904214895e-06, + "loss": 0.9155, + "step": 25177 + }, + { + "epoch": 0.59, + "grad_norm": 2.1070354325446052, + "learning_rate": 7.495190366655409e-06, + "loss": 0.9275, + "step": 25178 + }, + { + "epoch": 0.59, + "grad_norm": 2.068947697350629, + "learning_rate": 7.4944516574731215e-06, + "loss": 1.0135, + "step": 25179 + }, + { + "epoch": 0.59, + "grad_norm": 2.0538899809713373, + "learning_rate": 7.493712962878934e-06, + "loss": 0.9935, + "step": 25180 + }, + { + "epoch": 0.59, + "grad_norm": 1.0718560715611811, + "learning_rate": 7.492974282877137e-06, + "loss": 0.92, + "step": 25181 + }, + { + "epoch": 0.59, + "grad_norm": 2.2468677171005838, + "learning_rate": 7.492235617472041e-06, + "loss": 0.9691, + "step": 25182 + }, + { + "epoch": 0.59, + "grad_norm": 1.9828043245790683, + "learning_rate": 7.4914969666679435e-06, + "loss": 0.9483, + "step": 25183 + }, + { + "epoch": 0.59, + "grad_norm": 1.8226825174882537, + "learning_rate": 7.490758330469144e-06, + "loss": 1.0261, + "step": 25184 + }, + { + "epoch": 0.59, + "grad_norm": 1.0447857995425551, + "learning_rate": 7.490019708879942e-06, + "loss": 0.897, + "step": 25185 + }, + { + "epoch": 0.59, + "grad_norm": 1.9787199777089384, + "learning_rate": 7.489281101904642e-06, + "loss": 0.8706, + "step": 25186 + }, + { + "epoch": 0.59, + "grad_norm": 2.2385704461604985, + "learning_rate": 7.488542509547544e-06, + "loss": 0.9947, + "step": 25187 + }, + { + "epoch": 0.59, + "grad_norm": 2.194648671157749, + "learning_rate": 7.487803931812945e-06, + "loss": 0.9753, + "step": 25188 + }, + { + "epoch": 0.59, + "grad_norm": 2.597482980061943, + "learning_rate": 7.487065368705144e-06, + "loss": 0.9261, + "step": 25189 + }, + { + "epoch": 0.59, + "grad_norm": 2.1839384201903917, + "learning_rate": 7.486326820228449e-06, + "loss": 0.9565, + "step": 25190 + }, + { + "epoch": 0.59, + "grad_norm": 2.208244859548028, + "learning_rate": 7.485588286387152e-06, + "loss": 1.0442, + "step": 25191 + }, + { + "epoch": 0.59, + "grad_norm": 1.9024783174742537, + "learning_rate": 7.484849767185556e-06, + "loss": 1.0707, + "step": 25192 + }, + { + "epoch": 0.59, + "grad_norm": 2.173484256761309, + "learning_rate": 7.48411126262796e-06, + "loss": 1.0419, + "step": 25193 + }, + { + "epoch": 0.59, + "grad_norm": 2.0046388042813676, + "learning_rate": 7.483372772718668e-06, + "loss": 1.0374, + "step": 25194 + }, + { + "epoch": 0.59, + "grad_norm": 2.5145929260978415, + "learning_rate": 7.482634297461975e-06, + "loss": 0.9528, + "step": 25195 + }, + { + "epoch": 0.59, + "grad_norm": 2.025137696066028, + "learning_rate": 7.48189583686218e-06, + "loss": 0.9525, + "step": 25196 + }, + { + "epoch": 0.59, + "grad_norm": 2.333359098074892, + "learning_rate": 7.48115739092359e-06, + "loss": 0.9156, + "step": 25197 + }, + { + "epoch": 0.59, + "grad_norm": 1.96146283115914, + "learning_rate": 7.480418959650493e-06, + "loss": 0.9638, + "step": 25198 + }, + { + "epoch": 0.59, + "grad_norm": 1.8788541063649096, + "learning_rate": 7.4796805430471975e-06, + "loss": 1.0771, + "step": 25199 + }, + { + "epoch": 0.59, + "grad_norm": 2.23472074625467, + "learning_rate": 7.478942141117998e-06, + "loss": 1.0868, + "step": 25200 + }, + { + "epoch": 0.59, + "grad_norm": 2.0961458956311803, + "learning_rate": 7.478203753867202e-06, + "loss": 1.0629, + "step": 25201 + }, + { + "epoch": 0.59, + "grad_norm": 1.0654955192606983, + "learning_rate": 7.477465381299094e-06, + "loss": 0.9223, + "step": 25202 + }, + { + "epoch": 0.59, + "grad_norm": 1.9077753376455795, + "learning_rate": 7.476727023417985e-06, + "loss": 0.9173, + "step": 25203 + }, + { + "epoch": 0.59, + "grad_norm": 2.2028322852883266, + "learning_rate": 7.47598868022817e-06, + "loss": 0.9626, + "step": 25204 + }, + { + "epoch": 0.59, + "grad_norm": 2.0789383387539586, + "learning_rate": 7.475250351733947e-06, + "loss": 1.027, + "step": 25205 + }, + { + "epoch": 0.59, + "grad_norm": 1.8527791184940672, + "learning_rate": 7.474512037939614e-06, + "loss": 1.0359, + "step": 25206 + }, + { + "epoch": 0.59, + "grad_norm": 2.1677633019027445, + "learning_rate": 7.473773738849475e-06, + "loss": 1.0477, + "step": 25207 + }, + { + "epoch": 0.59, + "grad_norm": 2.4636859025763007, + "learning_rate": 7.473035454467825e-06, + "loss": 0.8989, + "step": 25208 + }, + { + "epoch": 0.59, + "grad_norm": 2.061101738801609, + "learning_rate": 7.472297184798959e-06, + "loss": 0.9265, + "step": 25209 + }, + { + "epoch": 0.59, + "grad_norm": 1.9934201486046232, + "learning_rate": 7.471558929847181e-06, + "loss": 1.1695, + "step": 25210 + }, + { + "epoch": 0.59, + "grad_norm": 2.5111733077202745, + "learning_rate": 7.4708206896167904e-06, + "loss": 1.0236, + "step": 25211 + }, + { + "epoch": 0.59, + "grad_norm": 1.8982364028392522, + "learning_rate": 7.470082464112079e-06, + "loss": 1.0965, + "step": 25212 + }, + { + "epoch": 0.59, + "grad_norm": 2.017362848426459, + "learning_rate": 7.469344253337347e-06, + "loss": 1.1816, + "step": 25213 + }, + { + "epoch": 0.59, + "grad_norm": 2.121282044899511, + "learning_rate": 7.468606057296896e-06, + "loss": 0.9804, + "step": 25214 + }, + { + "epoch": 0.59, + "grad_norm": 1.9670199537149822, + "learning_rate": 7.467867875995025e-06, + "loss": 0.9118, + "step": 25215 + }, + { + "epoch": 0.59, + "grad_norm": 2.201310358624616, + "learning_rate": 7.467129709436025e-06, + "loss": 1.1092, + "step": 25216 + }, + { + "epoch": 0.59, + "grad_norm": 2.0854329443716133, + "learning_rate": 7.466391557624197e-06, + "loss": 1.1177, + "step": 25217 + }, + { + "epoch": 0.59, + "grad_norm": 1.8624361328881156, + "learning_rate": 7.465653420563846e-06, + "loss": 0.9184, + "step": 25218 + }, + { + "epoch": 0.59, + "grad_norm": 2.029761393171433, + "learning_rate": 7.464915298259256e-06, + "loss": 0.9833, + "step": 25219 + }, + { + "epoch": 0.59, + "grad_norm": 2.0017317453540344, + "learning_rate": 7.464177190714734e-06, + "loss": 0.9151, + "step": 25220 + }, + { + "epoch": 0.59, + "grad_norm": 2.3301762737465337, + "learning_rate": 7.463439097934572e-06, + "loss": 0.9798, + "step": 25221 + }, + { + "epoch": 0.59, + "grad_norm": 1.0735765820029128, + "learning_rate": 7.462701019923078e-06, + "loss": 0.9647, + "step": 25222 + }, + { + "epoch": 0.59, + "grad_norm": 1.8980191751271607, + "learning_rate": 7.461962956684538e-06, + "loss": 1.0328, + "step": 25223 + }, + { + "epoch": 0.59, + "grad_norm": 1.9435467744692445, + "learning_rate": 7.4612249082232524e-06, + "loss": 1.0345, + "step": 25224 + }, + { + "epoch": 0.59, + "grad_norm": 1.829897576487412, + "learning_rate": 7.460486874543522e-06, + "loss": 0.9636, + "step": 25225 + }, + { + "epoch": 0.59, + "grad_norm": 2.183152030692725, + "learning_rate": 7.459748855649637e-06, + "loss": 0.8931, + "step": 25226 + }, + { + "epoch": 0.59, + "grad_norm": 2.1460942680439326, + "learning_rate": 7.459010851545901e-06, + "loss": 1.0237, + "step": 25227 + }, + { + "epoch": 0.59, + "grad_norm": 2.011815928454528, + "learning_rate": 7.458272862236606e-06, + "loss": 0.8868, + "step": 25228 + }, + { + "epoch": 0.59, + "grad_norm": 1.9027867544297141, + "learning_rate": 7.457534887726056e-06, + "loss": 1.1499, + "step": 25229 + }, + { + "epoch": 0.59, + "grad_norm": 2.174639739374812, + "learning_rate": 7.456796928018537e-06, + "loss": 1.0652, + "step": 25230 + }, + { + "epoch": 0.59, + "grad_norm": 2.0664135684008165, + "learning_rate": 7.456058983118354e-06, + "loss": 0.9005, + "step": 25231 + }, + { + "epoch": 0.59, + "grad_norm": 1.9078721550814457, + "learning_rate": 7.4553210530298025e-06, + "loss": 0.8707, + "step": 25232 + }, + { + "epoch": 0.59, + "grad_norm": 2.3176028687802845, + "learning_rate": 7.454583137757176e-06, + "loss": 0.9689, + "step": 25233 + }, + { + "epoch": 0.59, + "grad_norm": 1.9490518648413202, + "learning_rate": 7.4538452373047694e-06, + "loss": 0.9745, + "step": 25234 + }, + { + "epoch": 0.59, + "grad_norm": 2.0111261445161195, + "learning_rate": 7.453107351676885e-06, + "loss": 0.9625, + "step": 25235 + }, + { + "epoch": 0.59, + "grad_norm": 1.9723426419667662, + "learning_rate": 7.452369480877818e-06, + "loss": 0.9541, + "step": 25236 + }, + { + "epoch": 0.59, + "grad_norm": 2.3603058446288645, + "learning_rate": 7.45163162491186e-06, + "loss": 0.9835, + "step": 25237 + }, + { + "epoch": 0.59, + "grad_norm": 1.891874426238356, + "learning_rate": 7.450893783783308e-06, + "loss": 1.0392, + "step": 25238 + }, + { + "epoch": 0.59, + "grad_norm": 2.1144506795982854, + "learning_rate": 7.4501559574964634e-06, + "loss": 0.9513, + "step": 25239 + }, + { + "epoch": 0.59, + "grad_norm": 2.985261088414456, + "learning_rate": 7.449418146055616e-06, + "loss": 1.0078, + "step": 25240 + }, + { + "epoch": 0.59, + "grad_norm": 1.9728709330682723, + "learning_rate": 7.448680349465063e-06, + "loss": 1.0852, + "step": 25241 + }, + { + "epoch": 0.59, + "grad_norm": 1.113891529240018, + "learning_rate": 7.447942567729098e-06, + "loss": 0.9684, + "step": 25242 + }, + { + "epoch": 0.59, + "grad_norm": 2.139176103405785, + "learning_rate": 7.447204800852025e-06, + "loss": 1.0737, + "step": 25243 + }, + { + "epoch": 0.59, + "grad_norm": 1.1635435076705474, + "learning_rate": 7.446467048838131e-06, + "loss": 0.9872, + "step": 25244 + }, + { + "epoch": 0.59, + "grad_norm": 3.227194069691138, + "learning_rate": 7.445729311691713e-06, + "loss": 1.1039, + "step": 25245 + }, + { + "epoch": 0.59, + "grad_norm": 1.934831079056603, + "learning_rate": 7.444991589417073e-06, + "loss": 0.9977, + "step": 25246 + }, + { + "epoch": 0.59, + "grad_norm": 2.112525580353328, + "learning_rate": 7.444253882018495e-06, + "loss": 0.9862, + "step": 25247 + }, + { + "epoch": 0.59, + "grad_norm": 2.4052406812939484, + "learning_rate": 7.4435161895002815e-06, + "loss": 1.0994, + "step": 25248 + }, + { + "epoch": 0.59, + "grad_norm": 1.8949848370667683, + "learning_rate": 7.442778511866725e-06, + "loss": 1.1116, + "step": 25249 + }, + { + "epoch": 0.59, + "grad_norm": 2.1015014618523518, + "learning_rate": 7.442040849122127e-06, + "loss": 0.9369, + "step": 25250 + }, + { + "epoch": 0.59, + "grad_norm": 2.254140277089151, + "learning_rate": 7.441303201270772e-06, + "loss": 1.0249, + "step": 25251 + }, + { + "epoch": 0.59, + "grad_norm": 2.121877666810204, + "learning_rate": 7.4405655683169606e-06, + "loss": 0.9681, + "step": 25252 + }, + { + "epoch": 0.59, + "grad_norm": 1.7130629820868224, + "learning_rate": 7.43982795026499e-06, + "loss": 0.9797, + "step": 25253 + }, + { + "epoch": 0.59, + "grad_norm": 2.0261971008776998, + "learning_rate": 7.439090347119149e-06, + "loss": 1.1231, + "step": 25254 + }, + { + "epoch": 0.59, + "grad_norm": 1.9442591515500194, + "learning_rate": 7.438352758883732e-06, + "loss": 0.9292, + "step": 25255 + }, + { + "epoch": 0.6, + "grad_norm": 1.875151121983439, + "learning_rate": 7.437615185563038e-06, + "loss": 0.9368, + "step": 25256 + }, + { + "epoch": 0.6, + "grad_norm": 1.9904360615172774, + "learning_rate": 7.436877627161364e-06, + "loss": 0.922, + "step": 25257 + }, + { + "epoch": 0.6, + "grad_norm": 1.8578095254701812, + "learning_rate": 7.4361400836829945e-06, + "loss": 1.0513, + "step": 25258 + }, + { + "epoch": 0.6, + "grad_norm": 2.092930169651471, + "learning_rate": 7.43540255513223e-06, + "loss": 0.9754, + "step": 25259 + }, + { + "epoch": 0.6, + "grad_norm": 2.001422497510052, + "learning_rate": 7.434665041513367e-06, + "loss": 0.986, + "step": 25260 + }, + { + "epoch": 0.6, + "grad_norm": 2.1040304727397574, + "learning_rate": 7.433927542830695e-06, + "loss": 1.0968, + "step": 25261 + }, + { + "epoch": 0.6, + "grad_norm": 1.7767893975336444, + "learning_rate": 7.433190059088506e-06, + "loss": 0.9452, + "step": 25262 + }, + { + "epoch": 0.6, + "grad_norm": 2.040161646483003, + "learning_rate": 7.4324525902910996e-06, + "loss": 0.9625, + "step": 25263 + }, + { + "epoch": 0.6, + "grad_norm": 2.1336562588913504, + "learning_rate": 7.43171513644277e-06, + "loss": 1.1101, + "step": 25264 + }, + { + "epoch": 0.6, + "grad_norm": 2.063391858301762, + "learning_rate": 7.4309776975478046e-06, + "loss": 0.9985, + "step": 25265 + }, + { + "epoch": 0.6, + "grad_norm": 1.7647516451854866, + "learning_rate": 7.4302402736104985e-06, + "loss": 0.9747, + "step": 25266 + }, + { + "epoch": 0.6, + "grad_norm": 1.888204657208032, + "learning_rate": 7.429502864635154e-06, + "loss": 0.9323, + "step": 25267 + }, + { + "epoch": 0.6, + "grad_norm": 2.2344469850997046, + "learning_rate": 7.428765470626051e-06, + "loss": 1.0272, + "step": 25268 + }, + { + "epoch": 0.6, + "grad_norm": 2.156941255782189, + "learning_rate": 7.428028091587491e-06, + "loss": 0.9352, + "step": 25269 + }, + { + "epoch": 0.6, + "grad_norm": 2.1418090681514803, + "learning_rate": 7.427290727523764e-06, + "loss": 1.0902, + "step": 25270 + }, + { + "epoch": 0.6, + "grad_norm": 1.890933661594492, + "learning_rate": 7.42655337843917e-06, + "loss": 0.9916, + "step": 25271 + }, + { + "epoch": 0.6, + "grad_norm": 1.9777363118356786, + "learning_rate": 7.4258160443379935e-06, + "loss": 1.0201, + "step": 25272 + }, + { + "epoch": 0.6, + "grad_norm": 1.7956444273884251, + "learning_rate": 7.4250787252245316e-06, + "loss": 1.0202, + "step": 25273 + }, + { + "epoch": 0.6, + "grad_norm": 1.2447831034781454, + "learning_rate": 7.424341421103074e-06, + "loss": 0.9413, + "step": 25274 + }, + { + "epoch": 0.6, + "grad_norm": 3.06776276222162, + "learning_rate": 7.423604131977921e-06, + "loss": 0.9509, + "step": 25275 + }, + { + "epoch": 0.6, + "grad_norm": 2.162786348598623, + "learning_rate": 7.422866857853358e-06, + "loss": 1.0129, + "step": 25276 + }, + { + "epoch": 0.6, + "grad_norm": 2.1141347337230623, + "learning_rate": 7.422129598733681e-06, + "loss": 0.9367, + "step": 25277 + }, + { + "epoch": 0.6, + "grad_norm": 1.9960995636022976, + "learning_rate": 7.421392354623181e-06, + "loss": 1.0997, + "step": 25278 + }, + { + "epoch": 0.6, + "grad_norm": 1.8997580117554584, + "learning_rate": 7.4206551255261495e-06, + "loss": 0.9852, + "step": 25279 + }, + { + "epoch": 0.6, + "grad_norm": 1.587767368954679, + "learning_rate": 7.419917911446881e-06, + "loss": 0.9032, + "step": 25280 + }, + { + "epoch": 0.6, + "grad_norm": 1.9888063259265985, + "learning_rate": 7.4191807123896686e-06, + "loss": 0.9728, + "step": 25281 + }, + { + "epoch": 0.6, + "grad_norm": 1.915218510334465, + "learning_rate": 7.418443528358805e-06, + "loss": 1.0392, + "step": 25282 + }, + { + "epoch": 0.6, + "grad_norm": 2.2266288444188262, + "learning_rate": 7.417706359358576e-06, + "loss": 1.041, + "step": 25283 + }, + { + "epoch": 0.6, + "grad_norm": 1.9628185704988728, + "learning_rate": 7.416969205393279e-06, + "loss": 0.8903, + "step": 25284 + }, + { + "epoch": 0.6, + "grad_norm": 1.9388161623609894, + "learning_rate": 7.4162320664672084e-06, + "loss": 1.046, + "step": 25285 + }, + { + "epoch": 0.6, + "grad_norm": 2.036377115024386, + "learning_rate": 7.415494942584651e-06, + "loss": 1.0484, + "step": 25286 + }, + { + "epoch": 0.6, + "grad_norm": 2.127672160375158, + "learning_rate": 7.414757833749897e-06, + "loss": 1.1104, + "step": 25287 + }, + { + "epoch": 0.6, + "grad_norm": 1.945073227453143, + "learning_rate": 7.414020739967245e-06, + "loss": 1.0266, + "step": 25288 + }, + { + "epoch": 0.6, + "grad_norm": 1.7510007376844932, + "learning_rate": 7.413283661240984e-06, + "loss": 0.9849, + "step": 25289 + }, + { + "epoch": 0.6, + "grad_norm": 2.195229997879133, + "learning_rate": 7.412546597575403e-06, + "loss": 1.0601, + "step": 25290 + }, + { + "epoch": 0.6, + "grad_norm": 2.568962061628824, + "learning_rate": 7.411809548974792e-06, + "loss": 0.9387, + "step": 25291 + }, + { + "epoch": 0.6, + "grad_norm": 2.351952184464058, + "learning_rate": 7.411072515443449e-06, + "loss": 0.9863, + "step": 25292 + }, + { + "epoch": 0.6, + "grad_norm": 2.0661603821002705, + "learning_rate": 7.410335496985661e-06, + "loss": 0.9675, + "step": 25293 + }, + { + "epoch": 0.6, + "grad_norm": 1.7700989351549026, + "learning_rate": 7.409598493605716e-06, + "loss": 0.9104, + "step": 25294 + }, + { + "epoch": 0.6, + "grad_norm": 2.0847331513928102, + "learning_rate": 7.408861505307912e-06, + "loss": 1.0373, + "step": 25295 + }, + { + "epoch": 0.6, + "grad_norm": 2.209197333237328, + "learning_rate": 7.408124532096539e-06, + "loss": 1.0316, + "step": 25296 + }, + { + "epoch": 0.6, + "grad_norm": 2.391529303785468, + "learning_rate": 7.407387573975881e-06, + "loss": 0.9384, + "step": 25297 + }, + { + "epoch": 0.6, + "grad_norm": 2.046398025368594, + "learning_rate": 7.406650630950234e-06, + "loss": 1.0051, + "step": 25298 + }, + { + "epoch": 0.6, + "grad_norm": 2.114283321151849, + "learning_rate": 7.405913703023894e-06, + "loss": 0.9485, + "step": 25299 + }, + { + "epoch": 0.6, + "grad_norm": 1.107161470586061, + "learning_rate": 7.405176790201139e-06, + "loss": 0.9042, + "step": 25300 + }, + { + "epoch": 0.6, + "grad_norm": 2.035879968477514, + "learning_rate": 7.404439892486267e-06, + "loss": 0.9055, + "step": 25301 + }, + { + "epoch": 0.6, + "grad_norm": 1.958480844233834, + "learning_rate": 7.403703009883567e-06, + "loss": 1.1256, + "step": 25302 + }, + { + "epoch": 0.6, + "grad_norm": 2.1546169258988037, + "learning_rate": 7.402966142397335e-06, + "loss": 0.9529, + "step": 25303 + }, + { + "epoch": 0.6, + "grad_norm": 2.2197146694403465, + "learning_rate": 7.402229290031851e-06, + "loss": 1.0382, + "step": 25304 + }, + { + "epoch": 0.6, + "grad_norm": 2.37873147238655, + "learning_rate": 7.401492452791414e-06, + "loss": 0.8858, + "step": 25305 + }, + { + "epoch": 0.6, + "grad_norm": 3.2764510043092616, + "learning_rate": 7.4007556306803095e-06, + "loss": 0.961, + "step": 25306 + }, + { + "epoch": 0.6, + "grad_norm": 2.3079950511208662, + "learning_rate": 7.400018823702827e-06, + "loss": 1.0208, + "step": 25307 + }, + { + "epoch": 0.6, + "grad_norm": 1.955079862048876, + "learning_rate": 7.39928203186326e-06, + "loss": 0.9944, + "step": 25308 + }, + { + "epoch": 0.6, + "grad_norm": 2.0752041170270052, + "learning_rate": 7.3985452551658956e-06, + "loss": 0.9055, + "step": 25309 + }, + { + "epoch": 0.6, + "grad_norm": 2.055288165829153, + "learning_rate": 7.397808493615026e-06, + "loss": 1.0321, + "step": 25310 + }, + { + "epoch": 0.6, + "grad_norm": 1.894742576252061, + "learning_rate": 7.3970717472149354e-06, + "loss": 1.0779, + "step": 25311 + }, + { + "epoch": 0.6, + "grad_norm": 1.9231768924262784, + "learning_rate": 7.39633501596992e-06, + "loss": 1.0546, + "step": 25312 + }, + { + "epoch": 0.6, + "grad_norm": 1.9524471598030826, + "learning_rate": 7.3955982998842675e-06, + "loss": 1.017, + "step": 25313 + }, + { + "epoch": 0.6, + "grad_norm": 1.0700965903081088, + "learning_rate": 7.394861598962265e-06, + "loss": 0.9448, + "step": 25314 + }, + { + "epoch": 0.6, + "grad_norm": 1.8394503312019814, + "learning_rate": 7.394124913208201e-06, + "loss": 1.0637, + "step": 25315 + }, + { + "epoch": 0.6, + "grad_norm": 2.074311779651553, + "learning_rate": 7.393388242626368e-06, + "loss": 1.0304, + "step": 25316 + }, + { + "epoch": 0.6, + "grad_norm": 1.935560975807708, + "learning_rate": 7.3926515872210555e-06, + "loss": 1.0063, + "step": 25317 + }, + { + "epoch": 0.6, + "grad_norm": 1.9415394702052768, + "learning_rate": 7.3919149469965505e-06, + "loss": 0.9994, + "step": 25318 + }, + { + "epoch": 0.6, + "grad_norm": 2.1193785830467875, + "learning_rate": 7.39117832195714e-06, + "loss": 0.9982, + "step": 25319 + }, + { + "epoch": 0.6, + "grad_norm": 1.9999872403161565, + "learning_rate": 7.390441712107119e-06, + "loss": 0.9711, + "step": 25320 + }, + { + "epoch": 0.6, + "grad_norm": 1.9669061372820897, + "learning_rate": 7.389705117450769e-06, + "loss": 1.0589, + "step": 25321 + }, + { + "epoch": 0.6, + "grad_norm": 9.280256066822396, + "learning_rate": 7.388968537992385e-06, + "loss": 1.0465, + "step": 25322 + }, + { + "epoch": 0.6, + "grad_norm": 2.039281945414145, + "learning_rate": 7.388231973736248e-06, + "loss": 1.0597, + "step": 25323 + }, + { + "epoch": 0.6, + "grad_norm": 1.9349904048349331, + "learning_rate": 7.387495424686657e-06, + "loss": 1.0227, + "step": 25324 + }, + { + "epoch": 0.6, + "grad_norm": 1.1221522653231382, + "learning_rate": 7.3867588908478934e-06, + "loss": 0.9552, + "step": 25325 + }, + { + "epoch": 0.6, + "grad_norm": 2.0098829056780914, + "learning_rate": 7.386022372224246e-06, + "loss": 1.0722, + "step": 25326 + }, + { + "epoch": 0.6, + "grad_norm": 1.1181107729078104, + "learning_rate": 7.385285868820006e-06, + "loss": 0.9949, + "step": 25327 + }, + { + "epoch": 0.6, + "grad_norm": 2.015220366521891, + "learning_rate": 7.384549380639454e-06, + "loss": 0.9619, + "step": 25328 + }, + { + "epoch": 0.6, + "grad_norm": 2.456460605052654, + "learning_rate": 7.383812907686888e-06, + "loss": 1.0184, + "step": 25329 + }, + { + "epoch": 0.6, + "grad_norm": 2.061428542123777, + "learning_rate": 7.383076449966591e-06, + "loss": 1.0659, + "step": 25330 + }, + { + "epoch": 0.6, + "grad_norm": 2.15597742692828, + "learning_rate": 7.382340007482852e-06, + "loss": 1.14, + "step": 25331 + }, + { + "epoch": 0.6, + "grad_norm": 1.0780049737516317, + "learning_rate": 7.381603580239956e-06, + "loss": 0.9199, + "step": 25332 + }, + { + "epoch": 0.6, + "grad_norm": 2.0502453881620863, + "learning_rate": 7.380867168242193e-06, + "loss": 1.1101, + "step": 25333 + }, + { + "epoch": 0.6, + "grad_norm": 2.2608183761430336, + "learning_rate": 7.380130771493854e-06, + "loss": 0.9483, + "step": 25334 + }, + { + "epoch": 0.6, + "grad_norm": 2.158544382025416, + "learning_rate": 7.37939438999922e-06, + "loss": 1.0151, + "step": 25335 + }, + { + "epoch": 0.6, + "grad_norm": 2.1813538868976474, + "learning_rate": 7.3786580237625804e-06, + "loss": 1.0286, + "step": 25336 + }, + { + "epoch": 0.6, + "grad_norm": 2.3435734803563513, + "learning_rate": 7.377921672788225e-06, + "loss": 1.0784, + "step": 25337 + }, + { + "epoch": 0.6, + "grad_norm": 3.9640426562836555, + "learning_rate": 7.377185337080443e-06, + "loss": 0.9873, + "step": 25338 + }, + { + "epoch": 0.6, + "grad_norm": 1.8780765836783828, + "learning_rate": 7.376449016643516e-06, + "loss": 0.9109, + "step": 25339 + }, + { + "epoch": 0.6, + "grad_norm": 1.9077745976376481, + "learning_rate": 7.375712711481731e-06, + "loss": 0.994, + "step": 25340 + }, + { + "epoch": 0.6, + "grad_norm": 2.0117865024425687, + "learning_rate": 7.3749764215993804e-06, + "loss": 0.8803, + "step": 25341 + }, + { + "epoch": 0.6, + "grad_norm": 2.979720455832209, + "learning_rate": 7.374240147000748e-06, + "loss": 0.8937, + "step": 25342 + }, + { + "epoch": 0.6, + "grad_norm": 1.9838693176382367, + "learning_rate": 7.373503887690117e-06, + "loss": 1.0214, + "step": 25343 + }, + { + "epoch": 0.6, + "grad_norm": 1.898795583321942, + "learning_rate": 7.372767643671781e-06, + "loss": 1.0072, + "step": 25344 + }, + { + "epoch": 0.6, + "grad_norm": 2.09272849105439, + "learning_rate": 7.372031414950025e-06, + "loss": 0.9342, + "step": 25345 + }, + { + "epoch": 0.6, + "grad_norm": 1.7760518567739385, + "learning_rate": 7.371295201529133e-06, + "loss": 0.9578, + "step": 25346 + }, + { + "epoch": 0.6, + "grad_norm": 1.9579030186061965, + "learning_rate": 7.370559003413389e-06, + "loss": 0.931, + "step": 25347 + }, + { + "epoch": 0.6, + "grad_norm": 2.0296746392040834, + "learning_rate": 7.369822820607091e-06, + "loss": 0.8428, + "step": 25348 + }, + { + "epoch": 0.6, + "grad_norm": 1.9422794389331308, + "learning_rate": 7.36908665311451e-06, + "loss": 1.064, + "step": 25349 + }, + { + "epoch": 0.6, + "grad_norm": 2.178882892705883, + "learning_rate": 7.3683505009399415e-06, + "loss": 1.0272, + "step": 25350 + }, + { + "epoch": 0.6, + "grad_norm": 2.0012504579613086, + "learning_rate": 7.3676143640876675e-06, + "loss": 0.9374, + "step": 25351 + }, + { + "epoch": 0.6, + "grad_norm": 2.2926727939913176, + "learning_rate": 7.366878242561984e-06, + "loss": 1.0476, + "step": 25352 + }, + { + "epoch": 0.6, + "grad_norm": 2.030836853894666, + "learning_rate": 7.36614213636716e-06, + "loss": 1.0173, + "step": 25353 + }, + { + "epoch": 0.6, + "grad_norm": 2.1755661618400457, + "learning_rate": 7.365406045507493e-06, + "loss": 0.8893, + "step": 25354 + }, + { + "epoch": 0.6, + "grad_norm": 2.245511412180099, + "learning_rate": 7.364669969987269e-06, + "loss": 0.9901, + "step": 25355 + }, + { + "epoch": 0.6, + "grad_norm": 2.0936402486256673, + "learning_rate": 7.363933909810766e-06, + "loss": 0.9374, + "step": 25356 + }, + { + "epoch": 0.6, + "grad_norm": 2.280587486761312, + "learning_rate": 7.363197864982277e-06, + "loss": 0.9924, + "step": 25357 + }, + { + "epoch": 0.6, + "grad_norm": 2.223293464823837, + "learning_rate": 7.362461835506085e-06, + "loss": 0.9088, + "step": 25358 + }, + { + "epoch": 0.6, + "grad_norm": 2.122689219418119, + "learning_rate": 7.361725821386476e-06, + "loss": 1.0739, + "step": 25359 + }, + { + "epoch": 0.6, + "grad_norm": 1.8666557210403933, + "learning_rate": 7.360989822627732e-06, + "loss": 0.9994, + "step": 25360 + }, + { + "epoch": 0.6, + "grad_norm": 1.9707496886144533, + "learning_rate": 7.3602538392341415e-06, + "loss": 1.0033, + "step": 25361 + }, + { + "epoch": 0.6, + "grad_norm": 2.0414108026127464, + "learning_rate": 7.359517871209993e-06, + "loss": 0.9105, + "step": 25362 + }, + { + "epoch": 0.6, + "grad_norm": 1.0716091647528725, + "learning_rate": 7.358781918559563e-06, + "loss": 0.9521, + "step": 25363 + }, + { + "epoch": 0.6, + "grad_norm": 1.926899071719613, + "learning_rate": 7.358045981287141e-06, + "loss": 0.9864, + "step": 25364 + }, + { + "epoch": 0.6, + "grad_norm": 2.219467255873388, + "learning_rate": 7.357310059397011e-06, + "loss": 1.0303, + "step": 25365 + }, + { + "epoch": 0.6, + "grad_norm": 2.023863431501639, + "learning_rate": 7.356574152893463e-06, + "loss": 0.9949, + "step": 25366 + }, + { + "epoch": 0.6, + "grad_norm": 1.9310736784845648, + "learning_rate": 7.355838261780774e-06, + "loss": 1.0015, + "step": 25367 + }, + { + "epoch": 0.6, + "grad_norm": 2.0543129070930863, + "learning_rate": 7.355102386063229e-06, + "loss": 0.8921, + "step": 25368 + }, + { + "epoch": 0.6, + "grad_norm": 1.8782181612467843, + "learning_rate": 7.354366525745121e-06, + "loss": 0.8938, + "step": 25369 + }, + { + "epoch": 0.6, + "grad_norm": 1.9008270129693268, + "learning_rate": 7.3536306808307256e-06, + "loss": 0.9587, + "step": 25370 + }, + { + "epoch": 0.6, + "grad_norm": 1.1033127212077554, + "learning_rate": 7.35289485132433e-06, + "loss": 1.0169, + "step": 25371 + }, + { + "epoch": 0.6, + "grad_norm": 1.945394284710353, + "learning_rate": 7.352159037230218e-06, + "loss": 0.8618, + "step": 25372 + }, + { + "epoch": 0.6, + "grad_norm": 2.0814701398000763, + "learning_rate": 7.351423238552677e-06, + "loss": 0.9865, + "step": 25373 + }, + { + "epoch": 0.6, + "grad_norm": 1.1289815939440775, + "learning_rate": 7.3506874552959886e-06, + "loss": 0.9415, + "step": 25374 + }, + { + "epoch": 0.6, + "grad_norm": 1.9074041219345466, + "learning_rate": 7.349951687464434e-06, + "loss": 0.9965, + "step": 25375 + }, + { + "epoch": 0.6, + "grad_norm": 1.8946441770969167, + "learning_rate": 7.349215935062302e-06, + "loss": 0.9799, + "step": 25376 + }, + { + "epoch": 0.6, + "grad_norm": 1.9959082439687323, + "learning_rate": 7.348480198093871e-06, + "loss": 0.7774, + "step": 25377 + }, + { + "epoch": 0.6, + "grad_norm": 2.5765182337839008, + "learning_rate": 7.34774447656343e-06, + "loss": 1.065, + "step": 25378 + }, + { + "epoch": 0.6, + "grad_norm": 2.046534049131079, + "learning_rate": 7.347008770475259e-06, + "loss": 0.9904, + "step": 25379 + }, + { + "epoch": 0.6, + "grad_norm": 1.0875827989011289, + "learning_rate": 7.346273079833645e-06, + "loss": 0.9919, + "step": 25380 + }, + { + "epoch": 0.6, + "grad_norm": 2.1715628310425763, + "learning_rate": 7.3455374046428675e-06, + "loss": 1.0821, + "step": 25381 + }, + { + "epoch": 0.6, + "grad_norm": 1.8967596155377204, + "learning_rate": 7.344801744907212e-06, + "loss": 1.0397, + "step": 25382 + }, + { + "epoch": 0.6, + "grad_norm": 1.837469953085849, + "learning_rate": 7.344066100630963e-06, + "loss": 0.9723, + "step": 25383 + }, + { + "epoch": 0.6, + "grad_norm": 2.0546163033979585, + "learning_rate": 7.3433304718184004e-06, + "loss": 0.9776, + "step": 25384 + }, + { + "epoch": 0.6, + "grad_norm": 1.906893094771134, + "learning_rate": 7.342594858473807e-06, + "loss": 0.943, + "step": 25385 + }, + { + "epoch": 0.6, + "grad_norm": 1.1008456373521651, + "learning_rate": 7.34185926060147e-06, + "loss": 0.9368, + "step": 25386 + }, + { + "epoch": 0.6, + "grad_norm": 2.250156598407478, + "learning_rate": 7.34112367820567e-06, + "loss": 0.9337, + "step": 25387 + }, + { + "epoch": 0.6, + "grad_norm": 1.9253970530303584, + "learning_rate": 7.3403881112906905e-06, + "loss": 1.0637, + "step": 25388 + }, + { + "epoch": 0.6, + "grad_norm": 1.9047483983633238, + "learning_rate": 7.3396525598608105e-06, + "loss": 0.9568, + "step": 25389 + }, + { + "epoch": 0.6, + "grad_norm": 1.9217945875329236, + "learning_rate": 7.338917023920319e-06, + "loss": 0.986, + "step": 25390 + }, + { + "epoch": 0.6, + "grad_norm": 2.026742546177157, + "learning_rate": 7.338181503473493e-06, + "loss": 0.9628, + "step": 25391 + }, + { + "epoch": 0.6, + "grad_norm": 1.9957348900818905, + "learning_rate": 7.3374459985246146e-06, + "loss": 0.9471, + "step": 25392 + }, + { + "epoch": 0.6, + "grad_norm": 1.7723132834527995, + "learning_rate": 7.336710509077971e-06, + "loss": 1.0724, + "step": 25393 + }, + { + "epoch": 0.6, + "grad_norm": 2.081537785166819, + "learning_rate": 7.335975035137845e-06, + "loss": 1.0115, + "step": 25394 + }, + { + "epoch": 0.6, + "grad_norm": 1.885352382407311, + "learning_rate": 7.335239576708513e-06, + "loss": 0.9905, + "step": 25395 + }, + { + "epoch": 0.6, + "grad_norm": 1.9274816958000942, + "learning_rate": 7.334504133794258e-06, + "loss": 1.0222, + "step": 25396 + }, + { + "epoch": 0.6, + "grad_norm": 1.9297218170690864, + "learning_rate": 7.333768706399365e-06, + "loss": 0.9201, + "step": 25397 + }, + { + "epoch": 0.6, + "grad_norm": 1.9246120816175547, + "learning_rate": 7.333033294528118e-06, + "loss": 1.1572, + "step": 25398 + }, + { + "epoch": 0.6, + "grad_norm": 2.1603235267073284, + "learning_rate": 7.332297898184793e-06, + "loss": 1.0667, + "step": 25399 + }, + { + "epoch": 0.6, + "grad_norm": 1.9167898482323527, + "learning_rate": 7.331562517373672e-06, + "loss": 0.9513, + "step": 25400 + }, + { + "epoch": 0.6, + "grad_norm": 1.8987508491164355, + "learning_rate": 7.330827152099045e-06, + "loss": 1.0073, + "step": 25401 + }, + { + "epoch": 0.6, + "grad_norm": 1.794738543412639, + "learning_rate": 7.3300918023651805e-06, + "loss": 1.0725, + "step": 25402 + }, + { + "epoch": 0.6, + "grad_norm": 2.0752870528194385, + "learning_rate": 7.32935646817637e-06, + "loss": 0.9977, + "step": 25403 + }, + { + "epoch": 0.6, + "grad_norm": 1.724024003540781, + "learning_rate": 7.32862114953689e-06, + "loss": 0.862, + "step": 25404 + }, + { + "epoch": 0.6, + "grad_norm": 1.1417518249452145, + "learning_rate": 7.327885846451026e-06, + "loss": 1.0287, + "step": 25405 + }, + { + "epoch": 0.6, + "grad_norm": 1.953861435692194, + "learning_rate": 7.327150558923055e-06, + "loss": 0.9272, + "step": 25406 + }, + { + "epoch": 0.6, + "grad_norm": 1.7949521321121829, + "learning_rate": 7.32641528695726e-06, + "loss": 1.0173, + "step": 25407 + }, + { + "epoch": 0.6, + "grad_norm": 1.8788198458141654, + "learning_rate": 7.3256800305579234e-06, + "loss": 1.0332, + "step": 25408 + }, + { + "epoch": 0.6, + "grad_norm": 1.0681176140894948, + "learning_rate": 7.3249447897293205e-06, + "loss": 0.8519, + "step": 25409 + }, + { + "epoch": 0.6, + "grad_norm": 2.335268425169502, + "learning_rate": 7.324209564475737e-06, + "loss": 1.0254, + "step": 25410 + }, + { + "epoch": 0.6, + "grad_norm": 1.9519614222638193, + "learning_rate": 7.323474354801454e-06, + "loss": 0.9933, + "step": 25411 + }, + { + "epoch": 0.6, + "grad_norm": 1.7881058265901806, + "learning_rate": 7.322739160710752e-06, + "loss": 0.949, + "step": 25412 + }, + { + "epoch": 0.6, + "grad_norm": 2.272457931805087, + "learning_rate": 7.3220039822079066e-06, + "loss": 0.8922, + "step": 25413 + }, + { + "epoch": 0.6, + "grad_norm": 2.076074445193245, + "learning_rate": 7.321268819297204e-06, + "loss": 0.9858, + "step": 25414 + }, + { + "epoch": 0.6, + "grad_norm": 1.890980726708831, + "learning_rate": 7.320533671982924e-06, + "loss": 0.89, + "step": 25415 + }, + { + "epoch": 0.6, + "grad_norm": 1.1314956366987943, + "learning_rate": 7.319798540269343e-06, + "loss": 0.9083, + "step": 25416 + }, + { + "epoch": 0.6, + "grad_norm": 2.0688216056628637, + "learning_rate": 7.319063424160742e-06, + "loss": 0.9584, + "step": 25417 + }, + { + "epoch": 0.6, + "grad_norm": 1.9741348408360264, + "learning_rate": 7.318328323661403e-06, + "loss": 0.9311, + "step": 25418 + }, + { + "epoch": 0.6, + "grad_norm": 1.890346096191644, + "learning_rate": 7.31759323877561e-06, + "loss": 0.9509, + "step": 25419 + }, + { + "epoch": 0.6, + "grad_norm": 1.0567148838285538, + "learning_rate": 7.316858169507635e-06, + "loss": 0.9324, + "step": 25420 + }, + { + "epoch": 0.6, + "grad_norm": 2.059588435962861, + "learning_rate": 7.31612311586176e-06, + "loss": 0.9051, + "step": 25421 + }, + { + "epoch": 0.6, + "grad_norm": 2.037628863191893, + "learning_rate": 7.315388077842269e-06, + "loss": 0.9773, + "step": 25422 + }, + { + "epoch": 0.6, + "grad_norm": 2.0919195990887807, + "learning_rate": 7.3146530554534355e-06, + "loss": 0.9548, + "step": 25423 + }, + { + "epoch": 0.6, + "grad_norm": 2.2048946000993057, + "learning_rate": 7.313918048699544e-06, + "loss": 1.0119, + "step": 25424 + }, + { + "epoch": 0.6, + "grad_norm": 1.898602258362627, + "learning_rate": 7.31318305758487e-06, + "loss": 0.957, + "step": 25425 + }, + { + "epoch": 0.6, + "grad_norm": 2.073723122278561, + "learning_rate": 7.3124480821136974e-06, + "loss": 0.9738, + "step": 25426 + }, + { + "epoch": 0.6, + "grad_norm": 2.3967798497015513, + "learning_rate": 7.3117131222903025e-06, + "loss": 0.9234, + "step": 25427 + }, + { + "epoch": 0.6, + "grad_norm": 2.3069611529766383, + "learning_rate": 7.3109781781189635e-06, + "loss": 0.9484, + "step": 25428 + }, + { + "epoch": 0.6, + "grad_norm": 2.1521761531708656, + "learning_rate": 7.310243249603963e-06, + "loss": 1.0455, + "step": 25429 + }, + { + "epoch": 0.6, + "grad_norm": 2.322190436508759, + "learning_rate": 7.309508336749576e-06, + "loss": 1.0318, + "step": 25430 + }, + { + "epoch": 0.6, + "grad_norm": 1.0626344044652607, + "learning_rate": 7.308773439560083e-06, + "loss": 0.9412, + "step": 25431 + }, + { + "epoch": 0.6, + "grad_norm": 2.0860888821564214, + "learning_rate": 7.308038558039762e-06, + "loss": 1.0227, + "step": 25432 + }, + { + "epoch": 0.6, + "grad_norm": 2.025381918679693, + "learning_rate": 7.307303692192899e-06, + "loss": 1.0601, + "step": 25433 + }, + { + "epoch": 0.6, + "grad_norm": 1.0769855970820126, + "learning_rate": 7.306568842023759e-06, + "loss": 0.9408, + "step": 25434 + }, + { + "epoch": 0.6, + "grad_norm": 2.1488919398471635, + "learning_rate": 7.305834007536632e-06, + "loss": 0.9374, + "step": 25435 + }, + { + "epoch": 0.6, + "grad_norm": 1.034379212333603, + "learning_rate": 7.305099188735793e-06, + "loss": 0.8887, + "step": 25436 + }, + { + "epoch": 0.6, + "grad_norm": 1.9144012732820876, + "learning_rate": 7.3043643856255175e-06, + "loss": 0.9928, + "step": 25437 + }, + { + "epoch": 0.6, + "grad_norm": 1.2687123015721253, + "learning_rate": 7.303629598210084e-06, + "loss": 1.0578, + "step": 25438 + }, + { + "epoch": 0.6, + "grad_norm": 2.193093511874506, + "learning_rate": 7.302894826493775e-06, + "loss": 0.9842, + "step": 25439 + }, + { + "epoch": 0.6, + "grad_norm": 2.0225080966286395, + "learning_rate": 7.302160070480869e-06, + "loss": 0.9366, + "step": 25440 + }, + { + "epoch": 0.6, + "grad_norm": 1.8696272021385751, + "learning_rate": 7.301425330175635e-06, + "loss": 0.9769, + "step": 25441 + }, + { + "epoch": 0.6, + "grad_norm": 2.6595753525277677, + "learning_rate": 7.300690605582361e-06, + "loss": 0.8593, + "step": 25442 + }, + { + "epoch": 0.6, + "grad_norm": 1.8405647445720676, + "learning_rate": 7.299955896705323e-06, + "loss": 0.9272, + "step": 25443 + }, + { + "epoch": 0.6, + "grad_norm": 2.0915096492700065, + "learning_rate": 7.299221203548794e-06, + "loss": 1.1193, + "step": 25444 + }, + { + "epoch": 0.6, + "grad_norm": 2.0266048246669244, + "learning_rate": 7.2984865261170525e-06, + "loss": 1.0335, + "step": 25445 + }, + { + "epoch": 0.6, + "grad_norm": 1.9517607466138251, + "learning_rate": 7.29775186441438e-06, + "loss": 0.8608, + "step": 25446 + }, + { + "epoch": 0.6, + "grad_norm": 2.260747754438232, + "learning_rate": 7.2970172184450525e-06, + "loss": 1.071, + "step": 25447 + }, + { + "epoch": 0.6, + "grad_norm": 2.3561409495516115, + "learning_rate": 7.296282588213345e-06, + "loss": 1.1507, + "step": 25448 + }, + { + "epoch": 0.6, + "grad_norm": 1.9467646346288525, + "learning_rate": 7.295547973723533e-06, + "loss": 1.0454, + "step": 25449 + }, + { + "epoch": 0.6, + "grad_norm": 1.9227867677387445, + "learning_rate": 7.294813374979906e-06, + "loss": 1.0566, + "step": 25450 + }, + { + "epoch": 0.6, + "grad_norm": 1.0777398256937976, + "learning_rate": 7.294078791986723e-06, + "loss": 0.9068, + "step": 25451 + }, + { + "epoch": 0.6, + "grad_norm": 2.0955419387873198, + "learning_rate": 7.293344224748274e-06, + "loss": 1.1715, + "step": 25452 + }, + { + "epoch": 0.6, + "grad_norm": 2.0489586415484635, + "learning_rate": 7.292609673268829e-06, + "loss": 1.0338, + "step": 25453 + }, + { + "epoch": 0.6, + "grad_norm": 1.835767833118833, + "learning_rate": 7.291875137552671e-06, + "loss": 0.8634, + "step": 25454 + }, + { + "epoch": 0.6, + "grad_norm": 1.9553205854713747, + "learning_rate": 7.291140617604072e-06, + "loss": 0.9895, + "step": 25455 + }, + { + "epoch": 0.6, + "grad_norm": 2.1179371964467477, + "learning_rate": 7.2904061134273105e-06, + "loss": 1.0702, + "step": 25456 + }, + { + "epoch": 0.6, + "grad_norm": 2.15680990808362, + "learning_rate": 7.289671625026664e-06, + "loss": 1.041, + "step": 25457 + }, + { + "epoch": 0.6, + "grad_norm": 1.9967556087564498, + "learning_rate": 7.288937152406404e-06, + "loss": 0.9938, + "step": 25458 + }, + { + "epoch": 0.6, + "grad_norm": 1.9191376339776327, + "learning_rate": 7.28820269557081e-06, + "loss": 0.8891, + "step": 25459 + }, + { + "epoch": 0.6, + "grad_norm": 1.8409064186761448, + "learning_rate": 7.2874682545241604e-06, + "loss": 0.9138, + "step": 25460 + }, + { + "epoch": 0.6, + "grad_norm": 2.024821240484489, + "learning_rate": 7.286733829270731e-06, + "loss": 1.0919, + "step": 25461 + }, + { + "epoch": 0.6, + "grad_norm": 2.0216578191487136, + "learning_rate": 7.285999419814792e-06, + "loss": 1.0632, + "step": 25462 + }, + { + "epoch": 0.6, + "grad_norm": 2.013196342135527, + "learning_rate": 7.285265026160626e-06, + "loss": 0.9908, + "step": 25463 + }, + { + "epoch": 0.6, + "grad_norm": 2.444736516947791, + "learning_rate": 7.284530648312509e-06, + "loss": 0.9492, + "step": 25464 + }, + { + "epoch": 0.6, + "grad_norm": 1.9396951736621368, + "learning_rate": 7.283796286274711e-06, + "loss": 0.9559, + "step": 25465 + }, + { + "epoch": 0.6, + "grad_norm": 1.909662549611759, + "learning_rate": 7.283061940051509e-06, + "loss": 0.9098, + "step": 25466 + }, + { + "epoch": 0.6, + "grad_norm": 2.495180079009059, + "learning_rate": 7.2823276096471825e-06, + "loss": 1.1175, + "step": 25467 + }, + { + "epoch": 0.6, + "grad_norm": 2.4380993415722587, + "learning_rate": 7.281593295066008e-06, + "loss": 0.8618, + "step": 25468 + }, + { + "epoch": 0.6, + "grad_norm": 1.8561844894610813, + "learning_rate": 7.280858996312255e-06, + "loss": 1.0013, + "step": 25469 + }, + { + "epoch": 0.6, + "grad_norm": 2.1802013519598784, + "learning_rate": 7.280124713390201e-06, + "loss": 1.0613, + "step": 25470 + }, + { + "epoch": 0.6, + "grad_norm": 2.2510464763954974, + "learning_rate": 7.279390446304125e-06, + "loss": 0.9708, + "step": 25471 + }, + { + "epoch": 0.6, + "grad_norm": 2.808180704039255, + "learning_rate": 7.278656195058298e-06, + "loss": 1.0712, + "step": 25472 + }, + { + "epoch": 0.6, + "grad_norm": 1.920419317915183, + "learning_rate": 7.277921959656995e-06, + "loss": 1.0167, + "step": 25473 + }, + { + "epoch": 0.6, + "grad_norm": 2.00095037875584, + "learning_rate": 7.277187740104492e-06, + "loss": 1.012, + "step": 25474 + }, + { + "epoch": 0.6, + "grad_norm": 1.9186344232196542, + "learning_rate": 7.276453536405066e-06, + "loss": 0.9469, + "step": 25475 + }, + { + "epoch": 0.6, + "grad_norm": 2.020592096607219, + "learning_rate": 7.275719348562989e-06, + "loss": 0.9364, + "step": 25476 + }, + { + "epoch": 0.6, + "grad_norm": 1.8814277826780756, + "learning_rate": 7.274985176582536e-06, + "loss": 0.9858, + "step": 25477 + }, + { + "epoch": 0.6, + "grad_norm": 2.050497422332162, + "learning_rate": 7.274251020467984e-06, + "loss": 1.0425, + "step": 25478 + }, + { + "epoch": 0.6, + "grad_norm": 1.1305891786463025, + "learning_rate": 7.273516880223601e-06, + "loss": 0.9861, + "step": 25479 + }, + { + "epoch": 0.6, + "grad_norm": 1.116304064751412, + "learning_rate": 7.272782755853669e-06, + "loss": 0.9343, + "step": 25480 + }, + { + "epoch": 0.6, + "grad_norm": 1.9001739733512635, + "learning_rate": 7.272048647362457e-06, + "loss": 0.9646, + "step": 25481 + }, + { + "epoch": 0.6, + "grad_norm": 2.398975577945005, + "learning_rate": 7.2713145547542475e-06, + "loss": 1.0552, + "step": 25482 + }, + { + "epoch": 0.6, + "grad_norm": 1.1118441948079634, + "learning_rate": 7.270580478033302e-06, + "loss": 0.9367, + "step": 25483 + }, + { + "epoch": 0.6, + "grad_norm": 1.8670027955228061, + "learning_rate": 7.269846417203903e-06, + "loss": 1.0023, + "step": 25484 + }, + { + "epoch": 0.6, + "grad_norm": 1.9594163014387784, + "learning_rate": 7.269112372270325e-06, + "loss": 1.019, + "step": 25485 + }, + { + "epoch": 0.6, + "grad_norm": 1.8926896676536829, + "learning_rate": 7.268378343236837e-06, + "loss": 0.9285, + "step": 25486 + }, + { + "epoch": 0.6, + "grad_norm": 1.9020834136736577, + "learning_rate": 7.267644330107714e-06, + "loss": 0.943, + "step": 25487 + }, + { + "epoch": 0.6, + "grad_norm": 2.1075324284415338, + "learning_rate": 7.266910332887231e-06, + "loss": 0.9492, + "step": 25488 + }, + { + "epoch": 0.6, + "grad_norm": 2.8766150658725507, + "learning_rate": 7.266176351579666e-06, + "loss": 1.0571, + "step": 25489 + }, + { + "epoch": 0.6, + "grad_norm": 2.368898100598344, + "learning_rate": 7.265442386189281e-06, + "loss": 1.0539, + "step": 25490 + }, + { + "epoch": 0.6, + "grad_norm": 2.2401062139686383, + "learning_rate": 7.26470843672036e-06, + "loss": 0.9693, + "step": 25491 + }, + { + "epoch": 0.6, + "grad_norm": 1.8548219200832705, + "learning_rate": 7.263974503177174e-06, + "loss": 1.1049, + "step": 25492 + }, + { + "epoch": 0.6, + "grad_norm": 1.8118735776400692, + "learning_rate": 7.263240585563991e-06, + "loss": 0.9711, + "step": 25493 + }, + { + "epoch": 0.6, + "grad_norm": 1.130110095482039, + "learning_rate": 7.262506683885088e-06, + "loss": 0.9204, + "step": 25494 + }, + { + "epoch": 0.6, + "grad_norm": 1.9313710457798945, + "learning_rate": 7.26177279814474e-06, + "loss": 0.9609, + "step": 25495 + }, + { + "epoch": 0.6, + "grad_norm": 2.0724156299116534, + "learning_rate": 7.261038928347217e-06, + "loss": 1.0183, + "step": 25496 + }, + { + "epoch": 0.6, + "grad_norm": 1.9497933639124763, + "learning_rate": 7.260305074496793e-06, + "loss": 1.0477, + "step": 25497 + }, + { + "epoch": 0.6, + "grad_norm": 1.8925630003070868, + "learning_rate": 7.259571236597737e-06, + "loss": 1.0243, + "step": 25498 + }, + { + "epoch": 0.6, + "grad_norm": 1.889831660749665, + "learning_rate": 7.2588374146543315e-06, + "loss": 1.125, + "step": 25499 + }, + { + "epoch": 0.6, + "grad_norm": 2.1970192948122853, + "learning_rate": 7.258103608670835e-06, + "loss": 0.8941, + "step": 25500 + }, + { + "epoch": 0.6, + "grad_norm": 3.3285410142098573, + "learning_rate": 7.2573698186515316e-06, + "loss": 1.0267, + "step": 25501 + }, + { + "epoch": 0.6, + "grad_norm": 4.145212240375796, + "learning_rate": 7.256636044600687e-06, + "loss": 1.0247, + "step": 25502 + }, + { + "epoch": 0.6, + "grad_norm": 2.1239866084709607, + "learning_rate": 7.255902286522578e-06, + "loss": 0.9022, + "step": 25503 + }, + { + "epoch": 0.6, + "grad_norm": 1.9403684964728447, + "learning_rate": 7.255168544421473e-06, + "loss": 1.0812, + "step": 25504 + }, + { + "epoch": 0.6, + "grad_norm": 2.0795921860054944, + "learning_rate": 7.254434818301648e-06, + "loss": 1.0831, + "step": 25505 + }, + { + "epoch": 0.6, + "grad_norm": 1.9117185388772675, + "learning_rate": 7.253701108167373e-06, + "loss": 0.9632, + "step": 25506 + }, + { + "epoch": 0.6, + "grad_norm": 1.915039670035671, + "learning_rate": 7.252967414022915e-06, + "loss": 1.0248, + "step": 25507 + }, + { + "epoch": 0.6, + "grad_norm": 1.090238843933973, + "learning_rate": 7.252233735872554e-06, + "loss": 0.9841, + "step": 25508 + }, + { + "epoch": 0.6, + "grad_norm": 1.7535388727446468, + "learning_rate": 7.251500073720558e-06, + "loss": 1.012, + "step": 25509 + }, + { + "epoch": 0.6, + "grad_norm": 2.0053713948064056, + "learning_rate": 7.2507664275712e-06, + "loss": 0.9981, + "step": 25510 + }, + { + "epoch": 0.6, + "grad_norm": 2.187140834754632, + "learning_rate": 7.250032797428746e-06, + "loss": 1.1451, + "step": 25511 + }, + { + "epoch": 0.6, + "grad_norm": 2.023334479033725, + "learning_rate": 7.249299183297475e-06, + "loss": 1.0374, + "step": 25512 + }, + { + "epoch": 0.6, + "grad_norm": 1.864997932542038, + "learning_rate": 7.248565585181655e-06, + "loss": 1.0193, + "step": 25513 + }, + { + "epoch": 0.6, + "grad_norm": 1.1384393369102659, + "learning_rate": 7.2478320030855576e-06, + "loss": 0.958, + "step": 25514 + }, + { + "epoch": 0.6, + "grad_norm": 2.0052767605557564, + "learning_rate": 7.247098437013449e-06, + "loss": 0.9287, + "step": 25515 + }, + { + "epoch": 0.6, + "grad_norm": 2.1631649233796897, + "learning_rate": 7.246364886969609e-06, + "loss": 1.0718, + "step": 25516 + }, + { + "epoch": 0.6, + "grad_norm": 2.045436058281113, + "learning_rate": 7.245631352958306e-06, + "loss": 0.9458, + "step": 25517 + }, + { + "epoch": 0.6, + "grad_norm": 1.943702380422696, + "learning_rate": 7.244897834983806e-06, + "loss": 1.0487, + "step": 25518 + }, + { + "epoch": 0.6, + "grad_norm": 1.918120261814678, + "learning_rate": 7.244164333050383e-06, + "loss": 1.0685, + "step": 25519 + }, + { + "epoch": 0.6, + "grad_norm": 1.8976289431317905, + "learning_rate": 7.243430847162308e-06, + "loss": 1.0502, + "step": 25520 + }, + { + "epoch": 0.6, + "grad_norm": 1.9341214779570466, + "learning_rate": 7.242697377323854e-06, + "loss": 1.1502, + "step": 25521 + }, + { + "epoch": 0.6, + "grad_norm": 2.066305552879298, + "learning_rate": 7.241963923539287e-06, + "loss": 0.9181, + "step": 25522 + }, + { + "epoch": 0.6, + "grad_norm": 1.9614516704760465, + "learning_rate": 7.241230485812877e-06, + "loss": 0.9036, + "step": 25523 + }, + { + "epoch": 0.6, + "grad_norm": 2.0784601957633195, + "learning_rate": 7.2404970641489015e-06, + "loss": 1.0114, + "step": 25524 + }, + { + "epoch": 0.6, + "grad_norm": 1.198829714945562, + "learning_rate": 7.2397636585516215e-06, + "loss": 0.938, + "step": 25525 + }, + { + "epoch": 0.6, + "grad_norm": 1.926237595345406, + "learning_rate": 7.239030269025311e-06, + "loss": 1.0501, + "step": 25526 + }, + { + "epoch": 0.6, + "grad_norm": 1.865123651935507, + "learning_rate": 7.238296895574242e-06, + "loss": 0.9136, + "step": 25527 + }, + { + "epoch": 0.6, + "grad_norm": 3.0916213514709376, + "learning_rate": 7.237563538202685e-06, + "loss": 1.0237, + "step": 25528 + }, + { + "epoch": 0.6, + "grad_norm": 2.0904816323576685, + "learning_rate": 7.2368301969149054e-06, + "loss": 1.0113, + "step": 25529 + }, + { + "epoch": 0.6, + "grad_norm": 2.063119281243661, + "learning_rate": 7.236096871715174e-06, + "loss": 0.9976, + "step": 25530 + }, + { + "epoch": 0.6, + "grad_norm": 2.2345245260653033, + "learning_rate": 7.2353635626077664e-06, + "loss": 1.0377, + "step": 25531 + }, + { + "epoch": 0.6, + "grad_norm": 2.0149320865821947, + "learning_rate": 7.234630269596943e-06, + "loss": 0.9645, + "step": 25532 + }, + { + "epoch": 0.6, + "grad_norm": 2.0363314997063133, + "learning_rate": 7.233896992686977e-06, + "loss": 0.955, + "step": 25533 + }, + { + "epoch": 0.6, + "grad_norm": 2.020289530516569, + "learning_rate": 7.233163731882139e-06, + "loss": 0.951, + "step": 25534 + }, + { + "epoch": 0.6, + "grad_norm": 2.574976615783669, + "learning_rate": 7.2324304871867036e-06, + "loss": 1.1144, + "step": 25535 + }, + { + "epoch": 0.6, + "grad_norm": 1.87748781923187, + "learning_rate": 7.231697258604926e-06, + "loss": 0.9226, + "step": 25536 + }, + { + "epoch": 0.6, + "grad_norm": 2.089614430885418, + "learning_rate": 7.230964046141087e-06, + "loss": 1.0379, + "step": 25537 + }, + { + "epoch": 0.6, + "grad_norm": 1.9596392845620463, + "learning_rate": 7.230230849799454e-06, + "loss": 1.0288, + "step": 25538 + }, + { + "epoch": 0.6, + "grad_norm": 2.612082328290194, + "learning_rate": 7.229497669584289e-06, + "loss": 0.9405, + "step": 25539 + }, + { + "epoch": 0.6, + "grad_norm": 1.7697342952572592, + "learning_rate": 7.228764505499869e-06, + "loss": 0.986, + "step": 25540 + }, + { + "epoch": 0.6, + "grad_norm": 2.2601403956569244, + "learning_rate": 7.228031357550458e-06, + "loss": 1.1239, + "step": 25541 + }, + { + "epoch": 0.6, + "grad_norm": 1.9341237750576965, + "learning_rate": 7.227298225740329e-06, + "loss": 1.0019, + "step": 25542 + }, + { + "epoch": 0.6, + "grad_norm": 1.9902300543312366, + "learning_rate": 7.226565110073743e-06, + "loss": 1.1122, + "step": 25543 + }, + { + "epoch": 0.6, + "grad_norm": 1.9351228279312322, + "learning_rate": 7.225832010554976e-06, + "loss": 0.9237, + "step": 25544 + }, + { + "epoch": 0.6, + "grad_norm": 2.6965405762391597, + "learning_rate": 7.225098927188294e-06, + "loss": 0.9167, + "step": 25545 + }, + { + "epoch": 0.6, + "grad_norm": 2.218760110624709, + "learning_rate": 7.2243658599779645e-06, + "loss": 0.9582, + "step": 25546 + }, + { + "epoch": 0.6, + "grad_norm": 1.844964061339951, + "learning_rate": 7.223632808928252e-06, + "loss": 0.9451, + "step": 25547 + }, + { + "epoch": 0.6, + "grad_norm": 2.253409214457393, + "learning_rate": 7.22289977404343e-06, + "loss": 0.8647, + "step": 25548 + }, + { + "epoch": 0.6, + "grad_norm": 2.418339523560291, + "learning_rate": 7.222166755327768e-06, + "loss": 0.8685, + "step": 25549 + }, + { + "epoch": 0.6, + "grad_norm": 1.951405631965645, + "learning_rate": 7.221433752785527e-06, + "loss": 0.9242, + "step": 25550 + }, + { + "epoch": 0.6, + "grad_norm": 2.2546192199330557, + "learning_rate": 7.2207007664209785e-06, + "loss": 0.8568, + "step": 25551 + }, + { + "epoch": 0.6, + "grad_norm": 1.8682604216881735, + "learning_rate": 7.2199677962383915e-06, + "loss": 1.0379, + "step": 25552 + }, + { + "epoch": 0.6, + "grad_norm": 2.1065475970031677, + "learning_rate": 7.219234842242032e-06, + "loss": 0.9341, + "step": 25553 + }, + { + "epoch": 0.6, + "grad_norm": 1.9627580412881989, + "learning_rate": 7.2185019044361675e-06, + "loss": 1.0578, + "step": 25554 + }, + { + "epoch": 0.6, + "grad_norm": 3.0262002638874694, + "learning_rate": 7.217768982825062e-06, + "loss": 0.9557, + "step": 25555 + }, + { + "epoch": 0.6, + "grad_norm": 1.9243962775387813, + "learning_rate": 7.217036077412992e-06, + "loss": 0.9748, + "step": 25556 + }, + { + "epoch": 0.6, + "grad_norm": 1.9361728276518346, + "learning_rate": 7.216303188204217e-06, + "loss": 0.9524, + "step": 25557 + }, + { + "epoch": 0.6, + "grad_norm": 2.0263205306300454, + "learning_rate": 7.215570315203006e-06, + "loss": 0.9573, + "step": 25558 + }, + { + "epoch": 0.6, + "grad_norm": 1.990976241736635, + "learning_rate": 7.2148374584136274e-06, + "loss": 0.9691, + "step": 25559 + }, + { + "epoch": 0.6, + "grad_norm": 1.8982929272311744, + "learning_rate": 7.214104617840343e-06, + "loss": 0.8859, + "step": 25560 + }, + { + "epoch": 0.6, + "grad_norm": 2.142555902404085, + "learning_rate": 7.213371793487426e-06, + "loss": 1.0326, + "step": 25561 + }, + { + "epoch": 0.6, + "grad_norm": 1.9447569556532076, + "learning_rate": 7.212638985359139e-06, + "loss": 0.911, + "step": 25562 + }, + { + "epoch": 0.6, + "grad_norm": 2.0338528255362776, + "learning_rate": 7.2119061934597545e-06, + "loss": 0.9701, + "step": 25563 + }, + { + "epoch": 0.6, + "grad_norm": 1.065949207132999, + "learning_rate": 7.211173417793531e-06, + "loss": 0.9128, + "step": 25564 + }, + { + "epoch": 0.6, + "grad_norm": 1.9948689192618445, + "learning_rate": 7.210440658364739e-06, + "loss": 0.9985, + "step": 25565 + }, + { + "epoch": 0.6, + "grad_norm": 1.9126813213245093, + "learning_rate": 7.209707915177648e-06, + "loss": 1.0675, + "step": 25566 + }, + { + "epoch": 0.6, + "grad_norm": 1.9339414014300758, + "learning_rate": 7.208975188236518e-06, + "loss": 0.9734, + "step": 25567 + }, + { + "epoch": 0.6, + "grad_norm": 1.1001083440807504, + "learning_rate": 7.208242477545618e-06, + "loss": 0.9298, + "step": 25568 + }, + { + "epoch": 0.6, + "grad_norm": 2.488915200083931, + "learning_rate": 7.2075097831092135e-06, + "loss": 0.9875, + "step": 25569 + }, + { + "epoch": 0.6, + "grad_norm": 2.052023733433057, + "learning_rate": 7.206777104931575e-06, + "loss": 0.9592, + "step": 25570 + }, + { + "epoch": 0.6, + "grad_norm": 1.111282603717649, + "learning_rate": 7.2060444430169615e-06, + "loss": 0.9676, + "step": 25571 + }, + { + "epoch": 0.6, + "grad_norm": 1.8556438737342582, + "learning_rate": 7.20531179736964e-06, + "loss": 1.0412, + "step": 25572 + }, + { + "epoch": 0.6, + "grad_norm": 2.0411341025904495, + "learning_rate": 7.204579167993881e-06, + "loss": 0.9231, + "step": 25573 + }, + { + "epoch": 0.6, + "grad_norm": 2.7990399453201604, + "learning_rate": 7.203846554893947e-06, + "loss": 0.8594, + "step": 25574 + }, + { + "epoch": 0.6, + "grad_norm": 2.213418039160679, + "learning_rate": 7.203113958074101e-06, + "loss": 1.029, + "step": 25575 + }, + { + "epoch": 0.6, + "grad_norm": 1.7344615797607674, + "learning_rate": 7.202381377538612e-06, + "loss": 1.0588, + "step": 25576 + }, + { + "epoch": 0.6, + "grad_norm": 1.9685643962240704, + "learning_rate": 7.201648813291748e-06, + "loss": 0.9732, + "step": 25577 + }, + { + "epoch": 0.6, + "grad_norm": 2.170357559884144, + "learning_rate": 7.200916265337767e-06, + "loss": 0.8395, + "step": 25578 + }, + { + "epoch": 0.6, + "grad_norm": 1.786843602554721, + "learning_rate": 7.200183733680937e-06, + "loss": 1.078, + "step": 25579 + }, + { + "epoch": 0.6, + "grad_norm": 1.934923496242193, + "learning_rate": 7.1994512183255284e-06, + "loss": 1.0745, + "step": 25580 + }, + { + "epoch": 0.6, + "grad_norm": 1.9230628168796287, + "learning_rate": 7.1987187192757965e-06, + "loss": 0.9072, + "step": 25581 + }, + { + "epoch": 0.6, + "grad_norm": 2.0915165022295166, + "learning_rate": 7.1979862365360135e-06, + "loss": 1.0672, + "step": 25582 + }, + { + "epoch": 0.6, + "grad_norm": 1.1136221430688653, + "learning_rate": 7.197253770110438e-06, + "loss": 0.9605, + "step": 25583 + }, + { + "epoch": 0.6, + "grad_norm": 2.2398007730088785, + "learning_rate": 7.196521320003346e-06, + "loss": 0.8887, + "step": 25584 + }, + { + "epoch": 0.6, + "grad_norm": 1.7390413085292413, + "learning_rate": 7.195788886218987e-06, + "loss": 0.8986, + "step": 25585 + }, + { + "epoch": 0.6, + "grad_norm": 2.96444427757461, + "learning_rate": 7.195056468761637e-06, + "loss": 0.9514, + "step": 25586 + }, + { + "epoch": 0.6, + "grad_norm": 2.3083021312371894, + "learning_rate": 7.194324067635556e-06, + "loss": 0.9459, + "step": 25587 + }, + { + "epoch": 0.6, + "grad_norm": 3.021205457968882, + "learning_rate": 7.193591682845006e-06, + "loss": 0.938, + "step": 25588 + }, + { + "epoch": 0.6, + "grad_norm": 1.1566500553624648, + "learning_rate": 7.192859314394255e-06, + "loss": 0.945, + "step": 25589 + }, + { + "epoch": 0.6, + "grad_norm": 2.1815052391740055, + "learning_rate": 7.1921269622875655e-06, + "loss": 1.063, + "step": 25590 + }, + { + "epoch": 0.6, + "grad_norm": 2.463977915747436, + "learning_rate": 7.191394626529204e-06, + "loss": 0.9688, + "step": 25591 + }, + { + "epoch": 0.6, + "grad_norm": 2.088615850662507, + "learning_rate": 7.190662307123428e-06, + "loss": 0.999, + "step": 25592 + }, + { + "epoch": 0.6, + "grad_norm": 2.0265569568156017, + "learning_rate": 7.189930004074509e-06, + "loss": 0.9272, + "step": 25593 + }, + { + "epoch": 0.6, + "grad_norm": 2.1528270024452647, + "learning_rate": 7.189197717386709e-06, + "loss": 0.8255, + "step": 25594 + }, + { + "epoch": 0.6, + "grad_norm": 2.106166574338363, + "learning_rate": 7.188465447064286e-06, + "loss": 1.0428, + "step": 25595 + }, + { + "epoch": 0.6, + "grad_norm": 2.20776725044454, + "learning_rate": 7.187733193111505e-06, + "loss": 1.0153, + "step": 25596 + }, + { + "epoch": 0.6, + "grad_norm": 1.815958199051555, + "learning_rate": 7.187000955532636e-06, + "loss": 0.9339, + "step": 25597 + }, + { + "epoch": 0.6, + "grad_norm": 1.7877295280544705, + "learning_rate": 7.186268734331938e-06, + "loss": 1.0193, + "step": 25598 + }, + { + "epoch": 0.6, + "grad_norm": 1.0248228678845532, + "learning_rate": 7.1855365295136746e-06, + "loss": 1.0268, + "step": 25599 + }, + { + "epoch": 0.6, + "grad_norm": 1.949429595342856, + "learning_rate": 7.184804341082104e-06, + "loss": 0.9503, + "step": 25600 + }, + { + "epoch": 0.6, + "grad_norm": 1.9719780871020347, + "learning_rate": 7.184072169041499e-06, + "loss": 0.9079, + "step": 25601 + }, + { + "epoch": 0.6, + "grad_norm": 1.9485057242656132, + "learning_rate": 7.183340013396116e-06, + "loss": 1.1095, + "step": 25602 + }, + { + "epoch": 0.6, + "grad_norm": 2.3011095253212037, + "learning_rate": 7.182607874150218e-06, + "loss": 0.9447, + "step": 25603 + }, + { + "epoch": 0.6, + "grad_norm": 1.9997009712941884, + "learning_rate": 7.181875751308067e-06, + "loss": 1.1022, + "step": 25604 + }, + { + "epoch": 0.6, + "grad_norm": 1.928728448571012, + "learning_rate": 7.181143644873932e-06, + "loss": 1.0498, + "step": 25605 + }, + { + "epoch": 0.6, + "grad_norm": 1.918673184684969, + "learning_rate": 7.1804115548520695e-06, + "loss": 1.1114, + "step": 25606 + }, + { + "epoch": 0.6, + "grad_norm": 2.2267524096055604, + "learning_rate": 7.179679481246743e-06, + "loss": 0.8884, + "step": 25607 + }, + { + "epoch": 0.6, + "grad_norm": 2.0579297407592607, + "learning_rate": 7.178947424062217e-06, + "loss": 0.9825, + "step": 25608 + }, + { + "epoch": 0.6, + "grad_norm": 2.08629515669961, + "learning_rate": 7.178215383302749e-06, + "loss": 0.9789, + "step": 25609 + }, + { + "epoch": 0.6, + "grad_norm": 1.15363007273055, + "learning_rate": 7.177483358972607e-06, + "loss": 0.9149, + "step": 25610 + }, + { + "epoch": 0.6, + "grad_norm": 1.8295867103914296, + "learning_rate": 7.17675135107605e-06, + "loss": 1.018, + "step": 25611 + }, + { + "epoch": 0.6, + "grad_norm": 2.0858757730634947, + "learning_rate": 7.176019359617341e-06, + "loss": 1.0715, + "step": 25612 + }, + { + "epoch": 0.6, + "grad_norm": 2.108773520046624, + "learning_rate": 7.17528738460074e-06, + "loss": 0.8581, + "step": 25613 + }, + { + "epoch": 0.6, + "grad_norm": 2.133655614084355, + "learning_rate": 7.17455542603051e-06, + "loss": 0.9563, + "step": 25614 + }, + { + "epoch": 0.6, + "grad_norm": 2.004134108507113, + "learning_rate": 7.173823483910916e-06, + "loss": 0.9677, + "step": 25615 + }, + { + "epoch": 0.6, + "grad_norm": 1.9814074450635075, + "learning_rate": 7.173091558246214e-06, + "loss": 1.1346, + "step": 25616 + }, + { + "epoch": 0.6, + "grad_norm": 2.146334577280554, + "learning_rate": 7.172359649040666e-06, + "loss": 0.884, + "step": 25617 + }, + { + "epoch": 0.6, + "grad_norm": 1.9052092060732606, + "learning_rate": 7.171627756298536e-06, + "loss": 1.0385, + "step": 25618 + }, + { + "epoch": 0.6, + "grad_norm": 2.3787447815077103, + "learning_rate": 7.170895880024088e-06, + "loss": 1.0058, + "step": 25619 + }, + { + "epoch": 0.6, + "grad_norm": 2.0301992012670236, + "learning_rate": 7.170164020221577e-06, + "loss": 1.0389, + "step": 25620 + }, + { + "epoch": 0.6, + "grad_norm": 2.0263341642469492, + "learning_rate": 7.169432176895266e-06, + "loss": 1.0487, + "step": 25621 + }, + { + "epoch": 0.6, + "grad_norm": 1.9665795937094466, + "learning_rate": 7.168700350049421e-06, + "loss": 1.0228, + "step": 25622 + }, + { + "epoch": 0.6, + "grad_norm": 1.0773091690215155, + "learning_rate": 7.1679685396882966e-06, + "loss": 1.008, + "step": 25623 + }, + { + "epoch": 0.6, + "grad_norm": 2.3408272643258603, + "learning_rate": 7.167236745816154e-06, + "loss": 1.0269, + "step": 25624 + }, + { + "epoch": 0.6, + "grad_norm": 1.9188421800432267, + "learning_rate": 7.166504968437258e-06, + "loss": 1.0144, + "step": 25625 + }, + { + "epoch": 0.6, + "grad_norm": 2.1914212711689784, + "learning_rate": 7.165773207555869e-06, + "loss": 0.9285, + "step": 25626 + }, + { + "epoch": 0.6, + "grad_norm": 1.0875735208118495, + "learning_rate": 7.165041463176243e-06, + "loss": 0.9958, + "step": 25627 + }, + { + "epoch": 0.6, + "grad_norm": 2.111474577851237, + "learning_rate": 7.164309735302643e-06, + "loss": 0.9193, + "step": 25628 + }, + { + "epoch": 0.6, + "grad_norm": 1.7985840278062106, + "learning_rate": 7.163578023939335e-06, + "loss": 0.8474, + "step": 25629 + }, + { + "epoch": 0.6, + "grad_norm": 2.264329368297587, + "learning_rate": 7.162846329090566e-06, + "loss": 0.9866, + "step": 25630 + }, + { + "epoch": 0.6, + "grad_norm": 2.0572154150623416, + "learning_rate": 7.162114650760608e-06, + "loss": 0.9941, + "step": 25631 + }, + { + "epoch": 0.6, + "grad_norm": 1.9431203001416189, + "learning_rate": 7.161382988953714e-06, + "loss": 1.1024, + "step": 25632 + }, + { + "epoch": 0.6, + "grad_norm": 2.1657538675058947, + "learning_rate": 7.160651343674152e-06, + "loss": 1.1253, + "step": 25633 + }, + { + "epoch": 0.6, + "grad_norm": 2.0177062878795686, + "learning_rate": 7.159919714926171e-06, + "loss": 1.0697, + "step": 25634 + }, + { + "epoch": 0.6, + "grad_norm": 2.178202578651542, + "learning_rate": 7.15918810271404e-06, + "loss": 0.9246, + "step": 25635 + }, + { + "epoch": 0.6, + "grad_norm": 2.348845221159715, + "learning_rate": 7.158456507042016e-06, + "loss": 1.0755, + "step": 25636 + }, + { + "epoch": 0.6, + "grad_norm": 2.583079717641225, + "learning_rate": 7.157724927914355e-06, + "loss": 0.9633, + "step": 25637 + }, + { + "epoch": 0.6, + "grad_norm": 2.0902620739136144, + "learning_rate": 7.156993365335321e-06, + "loss": 0.9564, + "step": 25638 + }, + { + "epoch": 0.6, + "grad_norm": 1.7078011366175143, + "learning_rate": 7.156261819309171e-06, + "loss": 0.9698, + "step": 25639 + }, + { + "epoch": 0.6, + "grad_norm": 1.9672698601208303, + "learning_rate": 7.1555302898401665e-06, + "loss": 0.9851, + "step": 25640 + }, + { + "epoch": 0.6, + "grad_norm": 2.3987471353650505, + "learning_rate": 7.154798776932563e-06, + "loss": 0.9612, + "step": 25641 + }, + { + "epoch": 0.6, + "grad_norm": 2.248296886788112, + "learning_rate": 7.154067280590622e-06, + "loss": 0.9376, + "step": 25642 + }, + { + "epoch": 0.6, + "grad_norm": 2.7920130246973645, + "learning_rate": 7.153335800818604e-06, + "loss": 0.8933, + "step": 25643 + }, + { + "epoch": 0.6, + "grad_norm": 2.11399772180738, + "learning_rate": 7.1526043376207675e-06, + "loss": 0.9055, + "step": 25644 + }, + { + "epoch": 0.6, + "grad_norm": 2.161559257092476, + "learning_rate": 7.1518728910013655e-06, + "loss": 0.941, + "step": 25645 + }, + { + "epoch": 0.6, + "grad_norm": 2.121052070246808, + "learning_rate": 7.151141460964663e-06, + "loss": 0.9403, + "step": 25646 + }, + { + "epoch": 0.6, + "grad_norm": 1.9459663732275596, + "learning_rate": 7.15041004751492e-06, + "loss": 1.01, + "step": 25647 + }, + { + "epoch": 0.6, + "grad_norm": 1.0920261861338785, + "learning_rate": 7.1496786506563885e-06, + "loss": 0.968, + "step": 25648 + }, + { + "epoch": 0.6, + "grad_norm": 1.8908888422393364, + "learning_rate": 7.148947270393329e-06, + "loss": 0.999, + "step": 25649 + }, + { + "epoch": 0.6, + "grad_norm": 2.0623822751525815, + "learning_rate": 7.148215906730004e-06, + "loss": 1.0068, + "step": 25650 + }, + { + "epoch": 0.6, + "grad_norm": 2.2583782510614885, + "learning_rate": 7.14748455967067e-06, + "loss": 0.9839, + "step": 25651 + }, + { + "epoch": 0.6, + "grad_norm": 2.174964858526045, + "learning_rate": 7.146753229219583e-06, + "loss": 1.0272, + "step": 25652 + }, + { + "epoch": 0.6, + "grad_norm": 1.0891871916535651, + "learning_rate": 7.146021915380999e-06, + "loss": 0.937, + "step": 25653 + }, + { + "epoch": 0.6, + "grad_norm": 2.5647836841292975, + "learning_rate": 7.145290618159182e-06, + "loss": 1.0294, + "step": 25654 + }, + { + "epoch": 0.6, + "grad_norm": 2.033168503739529, + "learning_rate": 7.1445593375583865e-06, + "loss": 0.9945, + "step": 25655 + }, + { + "epoch": 0.6, + "grad_norm": 2.369386909990947, + "learning_rate": 7.14382807358287e-06, + "loss": 0.9042, + "step": 25656 + }, + { + "epoch": 0.6, + "grad_norm": 2.4833632291172334, + "learning_rate": 7.143096826236888e-06, + "loss": 0.9243, + "step": 25657 + }, + { + "epoch": 0.6, + "grad_norm": 2.850069213837676, + "learning_rate": 7.142365595524705e-06, + "loss": 1.0079, + "step": 25658 + }, + { + "epoch": 0.6, + "grad_norm": 1.9093098858856086, + "learning_rate": 7.141634381450573e-06, + "loss": 1.1628, + "step": 25659 + }, + { + "epoch": 0.6, + "grad_norm": 2.8196212091071007, + "learning_rate": 7.14090318401875e-06, + "loss": 1.023, + "step": 25660 + }, + { + "epoch": 0.6, + "grad_norm": 1.8500544909719516, + "learning_rate": 7.1401720032334965e-06, + "loss": 0.8488, + "step": 25661 + }, + { + "epoch": 0.6, + "grad_norm": 1.9548105654975814, + "learning_rate": 7.139440839099063e-06, + "loss": 0.9655, + "step": 25662 + }, + { + "epoch": 0.6, + "grad_norm": 1.9531835926327088, + "learning_rate": 7.138709691619712e-06, + "loss": 0.9373, + "step": 25663 + }, + { + "epoch": 0.6, + "grad_norm": 1.9649973772767177, + "learning_rate": 7.137978560799697e-06, + "loss": 1.1042, + "step": 25664 + }, + { + "epoch": 0.6, + "grad_norm": 2.189629709692626, + "learning_rate": 7.137247446643283e-06, + "loss": 0.9599, + "step": 25665 + }, + { + "epoch": 0.6, + "grad_norm": 1.1148286920232617, + "learning_rate": 7.136516349154714e-06, + "loss": 0.9309, + "step": 25666 + }, + { + "epoch": 0.6, + "grad_norm": 2.200189452136007, + "learning_rate": 7.1357852683382565e-06, + "loss": 0.8857, + "step": 25667 + }, + { + "epoch": 0.6, + "grad_norm": 1.9718514523974269, + "learning_rate": 7.135054204198166e-06, + "loss": 0.9549, + "step": 25668 + }, + { + "epoch": 0.6, + "grad_norm": 2.4560837938279043, + "learning_rate": 7.134323156738694e-06, + "loss": 0.964, + "step": 25669 + }, + { + "epoch": 0.6, + "grad_norm": 2.508981384676898, + "learning_rate": 7.1335921259640996e-06, + "loss": 1.1205, + "step": 25670 + }, + { + "epoch": 0.6, + "grad_norm": 2.1644241482914266, + "learning_rate": 7.132861111878639e-06, + "loss": 1.008, + "step": 25671 + }, + { + "epoch": 0.6, + "grad_norm": 1.9120280256539934, + "learning_rate": 7.132130114486573e-06, + "loss": 0.915, + "step": 25672 + }, + { + "epoch": 0.6, + "grad_norm": 2.2128693798666514, + "learning_rate": 7.131399133792147e-06, + "loss": 1.0592, + "step": 25673 + }, + { + "epoch": 0.6, + "grad_norm": 2.180853842691206, + "learning_rate": 7.130668169799627e-06, + "loss": 1.0385, + "step": 25674 + }, + { + "epoch": 0.6, + "grad_norm": 1.9740520642609976, + "learning_rate": 7.129937222513268e-06, + "loss": 0.9701, + "step": 25675 + }, + { + "epoch": 0.6, + "grad_norm": 2.797012705566301, + "learning_rate": 7.129206291937319e-06, + "loss": 1.0244, + "step": 25676 + }, + { + "epoch": 0.6, + "grad_norm": 2.087909734529663, + "learning_rate": 7.128475378076039e-06, + "loss": 0.9571, + "step": 25677 + }, + { + "epoch": 0.6, + "grad_norm": 2.1422573733963817, + "learning_rate": 7.127744480933687e-06, + "loss": 0.9828, + "step": 25678 + }, + { + "epoch": 0.6, + "grad_norm": 2.0799630864215537, + "learning_rate": 7.127013600514517e-06, + "loss": 0.9862, + "step": 25679 + }, + { + "epoch": 0.61, + "grad_norm": 1.9598371096460285, + "learning_rate": 7.126282736822782e-06, + "loss": 0.9105, + "step": 25680 + }, + { + "epoch": 0.61, + "grad_norm": 1.9245888955214365, + "learning_rate": 7.125551889862736e-06, + "loss": 1.014, + "step": 25681 + }, + { + "epoch": 0.61, + "grad_norm": 1.8796434240136188, + "learning_rate": 7.1248210596386444e-06, + "loss": 1.048, + "step": 25682 + }, + { + "epoch": 0.61, + "grad_norm": 1.8438091875987785, + "learning_rate": 7.1240902461547475e-06, + "loss": 0.9276, + "step": 25683 + }, + { + "epoch": 0.61, + "grad_norm": 2.084983141242623, + "learning_rate": 7.1233594494153104e-06, + "loss": 1.051, + "step": 25684 + }, + { + "epoch": 0.61, + "grad_norm": 2.052145986137113, + "learning_rate": 7.122628669424583e-06, + "loss": 0.9449, + "step": 25685 + }, + { + "epoch": 0.61, + "grad_norm": 2.319005814197363, + "learning_rate": 7.121897906186827e-06, + "loss": 0.929, + "step": 25686 + }, + { + "epoch": 0.61, + "grad_norm": 1.8039685381248962, + "learning_rate": 7.12116715970629e-06, + "loss": 0.9566, + "step": 25687 + }, + { + "epoch": 0.61, + "grad_norm": 2.1381971513885016, + "learning_rate": 7.120436429987228e-06, + "loss": 1.0965, + "step": 25688 + }, + { + "epoch": 0.61, + "grad_norm": 1.935692952833481, + "learning_rate": 7.1197057170339e-06, + "loss": 0.9951, + "step": 25689 + }, + { + "epoch": 0.61, + "grad_norm": 1.8560315780551768, + "learning_rate": 7.118975020850554e-06, + "loss": 1.1043, + "step": 25690 + }, + { + "epoch": 0.61, + "grad_norm": 1.8429734853517517, + "learning_rate": 7.1182443414414495e-06, + "loss": 0.9779, + "step": 25691 + }, + { + "epoch": 0.61, + "grad_norm": 2.1159936504436, + "learning_rate": 7.117513678810837e-06, + "loss": 1.0823, + "step": 25692 + }, + { + "epoch": 0.61, + "grad_norm": 2.1555871199410754, + "learning_rate": 7.116783032962976e-06, + "loss": 1.0248, + "step": 25693 + }, + { + "epoch": 0.61, + "grad_norm": 1.9811024920367108, + "learning_rate": 7.116052403902112e-06, + "loss": 0.9552, + "step": 25694 + }, + { + "epoch": 0.61, + "grad_norm": 2.185617477185978, + "learning_rate": 7.115321791632507e-06, + "loss": 1.0151, + "step": 25695 + }, + { + "epoch": 0.61, + "grad_norm": 1.8947174410909626, + "learning_rate": 7.114591196158413e-06, + "loss": 1.1149, + "step": 25696 + }, + { + "epoch": 0.61, + "grad_norm": 1.91119845408678, + "learning_rate": 7.1138606174840806e-06, + "loss": 0.9425, + "step": 25697 + }, + { + "epoch": 0.61, + "grad_norm": 2.0609704964107936, + "learning_rate": 7.113130055613764e-06, + "loss": 1.0048, + "step": 25698 + }, + { + "epoch": 0.61, + "grad_norm": 1.7875180992599395, + "learning_rate": 7.11239951055172e-06, + "loss": 0.9925, + "step": 25699 + }, + { + "epoch": 0.61, + "grad_norm": 3.0517065705765747, + "learning_rate": 7.111668982302202e-06, + "loss": 0.7759, + "step": 25700 + }, + { + "epoch": 0.61, + "grad_norm": 2.0943900810657876, + "learning_rate": 7.11093847086946e-06, + "loss": 0.9572, + "step": 25701 + }, + { + "epoch": 0.61, + "grad_norm": 2.07764439312451, + "learning_rate": 7.110207976257746e-06, + "loss": 1.0883, + "step": 25702 + }, + { + "epoch": 0.61, + "grad_norm": 2.190309241138582, + "learning_rate": 7.10947749847132e-06, + "loss": 1.0383, + "step": 25703 + }, + { + "epoch": 0.61, + "grad_norm": 2.8722124903990496, + "learning_rate": 7.108747037514431e-06, + "loss": 1.0035, + "step": 25704 + }, + { + "epoch": 0.61, + "grad_norm": 2.15700433738296, + "learning_rate": 7.10801659339133e-06, + "loss": 1.0616, + "step": 25705 + }, + { + "epoch": 0.61, + "grad_norm": 1.7901076612215523, + "learning_rate": 7.107286166106271e-06, + "loss": 0.9816, + "step": 25706 + }, + { + "epoch": 0.61, + "grad_norm": 2.0217676669106033, + "learning_rate": 7.106555755663511e-06, + "loss": 1.0782, + "step": 25707 + }, + { + "epoch": 0.61, + "grad_norm": 2.119600242991258, + "learning_rate": 7.105825362067299e-06, + "loss": 1.0565, + "step": 25708 + }, + { + "epoch": 0.61, + "grad_norm": 2.0875491713031793, + "learning_rate": 7.105094985321888e-06, + "loss": 0.9623, + "step": 25709 + }, + { + "epoch": 0.61, + "grad_norm": 2.8879624311910193, + "learning_rate": 7.10436462543153e-06, + "loss": 1.0185, + "step": 25710 + }, + { + "epoch": 0.61, + "grad_norm": 2.2366562671291623, + "learning_rate": 7.103634282400475e-06, + "loss": 0.9285, + "step": 25711 + }, + { + "epoch": 0.61, + "grad_norm": 2.113815743083735, + "learning_rate": 7.102903956232981e-06, + "loss": 0.9131, + "step": 25712 + }, + { + "epoch": 0.61, + "grad_norm": 1.8446376214540223, + "learning_rate": 7.102173646933295e-06, + "loss": 0.9997, + "step": 25713 + }, + { + "epoch": 0.61, + "grad_norm": 2.733438527643576, + "learning_rate": 7.1014433545056785e-06, + "loss": 0.8332, + "step": 25714 + }, + { + "epoch": 0.61, + "grad_norm": 1.9605149554992567, + "learning_rate": 7.100713078954369e-06, + "loss": 1.1447, + "step": 25715 + }, + { + "epoch": 0.61, + "grad_norm": 1.9904954996814983, + "learning_rate": 7.099982820283629e-06, + "loss": 1.0252, + "step": 25716 + }, + { + "epoch": 0.61, + "grad_norm": 1.8946936259416651, + "learning_rate": 7.099252578497709e-06, + "loss": 0.9148, + "step": 25717 + }, + { + "epoch": 0.61, + "grad_norm": 2.5598990033020304, + "learning_rate": 7.0985223536008555e-06, + "loss": 0.9304, + "step": 25718 + }, + { + "epoch": 0.61, + "grad_norm": 2.722072489778924, + "learning_rate": 7.097792145597322e-06, + "loss": 1.0056, + "step": 25719 + }, + { + "epoch": 0.61, + "grad_norm": 1.7056440752207664, + "learning_rate": 7.097061954491363e-06, + "loss": 1.0718, + "step": 25720 + }, + { + "epoch": 0.61, + "grad_norm": 1.0554994589118678, + "learning_rate": 7.096331780287231e-06, + "loss": 0.9111, + "step": 25721 + }, + { + "epoch": 0.61, + "grad_norm": 2.1214023406443046, + "learning_rate": 7.0956016229891715e-06, + "loss": 1.246, + "step": 25722 + }, + { + "epoch": 0.61, + "grad_norm": 2.4223464485803152, + "learning_rate": 7.09487148260144e-06, + "loss": 1.0685, + "step": 25723 + }, + { + "epoch": 0.61, + "grad_norm": 1.9124536080222787, + "learning_rate": 7.094141359128288e-06, + "loss": 0.9778, + "step": 25724 + }, + { + "epoch": 0.61, + "grad_norm": 3.1755941651577504, + "learning_rate": 7.093411252573964e-06, + "loss": 1.1126, + "step": 25725 + }, + { + "epoch": 0.61, + "grad_norm": 1.9328539536077551, + "learning_rate": 7.092681162942717e-06, + "loss": 0.9887, + "step": 25726 + }, + { + "epoch": 0.61, + "grad_norm": 1.9329440883473297, + "learning_rate": 7.0919510902388034e-06, + "loss": 1.0467, + "step": 25727 + }, + { + "epoch": 0.61, + "grad_norm": 1.9559340056482475, + "learning_rate": 7.0912210344664736e-06, + "loss": 0.9835, + "step": 25728 + }, + { + "epoch": 0.61, + "grad_norm": 2.1461801277512085, + "learning_rate": 7.090490995629973e-06, + "loss": 0.9839, + "step": 25729 + }, + { + "epoch": 0.61, + "grad_norm": 2.0820420050880024, + "learning_rate": 7.089760973733553e-06, + "loss": 0.9694, + "step": 25730 + }, + { + "epoch": 0.61, + "grad_norm": 2.1316812632145394, + "learning_rate": 7.089030968781472e-06, + "loss": 1.2031, + "step": 25731 + }, + { + "epoch": 0.61, + "grad_norm": 1.9543085848865207, + "learning_rate": 7.088300980777969e-06, + "loss": 1.0189, + "step": 25732 + }, + { + "epoch": 0.61, + "grad_norm": 2.534248442718587, + "learning_rate": 7.087571009727301e-06, + "loss": 1.0224, + "step": 25733 + }, + { + "epoch": 0.61, + "grad_norm": 1.9463272339809596, + "learning_rate": 7.086841055633713e-06, + "loss": 0.9365, + "step": 25734 + }, + { + "epoch": 0.61, + "grad_norm": 2.066570559670541, + "learning_rate": 7.086111118501465e-06, + "loss": 0.8926, + "step": 25735 + }, + { + "epoch": 0.61, + "grad_norm": 2.4188465214508375, + "learning_rate": 7.085381198334797e-06, + "loss": 0.9246, + "step": 25736 + }, + { + "epoch": 0.61, + "grad_norm": 2.1090955815200325, + "learning_rate": 7.084651295137964e-06, + "loss": 1.1334, + "step": 25737 + }, + { + "epoch": 0.61, + "grad_norm": 2.154215473079313, + "learning_rate": 7.0839214089152155e-06, + "loss": 0.9434, + "step": 25738 + }, + { + "epoch": 0.61, + "grad_norm": 2.535702535489959, + "learning_rate": 7.083191539670795e-06, + "loss": 0.9599, + "step": 25739 + }, + { + "epoch": 0.61, + "grad_norm": 2.005668294765447, + "learning_rate": 7.08246168740896e-06, + "loss": 0.982, + "step": 25740 + }, + { + "epoch": 0.61, + "grad_norm": 2.3946880486429967, + "learning_rate": 7.081731852133958e-06, + "loss": 0.9423, + "step": 25741 + }, + { + "epoch": 0.61, + "grad_norm": 2.2079799287385957, + "learning_rate": 7.081002033850036e-06, + "loss": 1.0271, + "step": 25742 + }, + { + "epoch": 0.61, + "grad_norm": 2.3494143650848502, + "learning_rate": 7.080272232561441e-06, + "loss": 1.057, + "step": 25743 + }, + { + "epoch": 0.61, + "grad_norm": 2.003306426251378, + "learning_rate": 7.079542448272428e-06, + "loss": 1.0061, + "step": 25744 + }, + { + "epoch": 0.61, + "grad_norm": 3.2103562977618925, + "learning_rate": 7.078812680987246e-06, + "loss": 1.0795, + "step": 25745 + }, + { + "epoch": 0.61, + "grad_norm": 1.1240057487115078, + "learning_rate": 7.0780829307101395e-06, + "loss": 0.9613, + "step": 25746 + }, + { + "epoch": 0.61, + "grad_norm": 2.184307551323259, + "learning_rate": 7.077353197445356e-06, + "loss": 1.0222, + "step": 25747 + }, + { + "epoch": 0.61, + "grad_norm": 2.72036705941865, + "learning_rate": 7.07662348119715e-06, + "loss": 1.0608, + "step": 25748 + }, + { + "epoch": 0.61, + "grad_norm": 2.1048566695114794, + "learning_rate": 7.075893781969769e-06, + "loss": 1.0407, + "step": 25749 + }, + { + "epoch": 0.61, + "grad_norm": 1.8968799554912892, + "learning_rate": 7.07516409976746e-06, + "loss": 0.9993, + "step": 25750 + }, + { + "epoch": 0.61, + "grad_norm": 2.435613448415486, + "learning_rate": 7.074434434594469e-06, + "loss": 1.0058, + "step": 25751 + }, + { + "epoch": 0.61, + "grad_norm": 1.0788439363308817, + "learning_rate": 7.073704786455049e-06, + "loss": 0.9011, + "step": 25752 + }, + { + "epoch": 0.61, + "grad_norm": 2.0002860692631614, + "learning_rate": 7.072975155353446e-06, + "loss": 1.022, + "step": 25753 + }, + { + "epoch": 0.61, + "grad_norm": 2.2711491142610125, + "learning_rate": 7.072245541293908e-06, + "loss": 1.0521, + "step": 25754 + }, + { + "epoch": 0.61, + "grad_norm": 1.9713515491129119, + "learning_rate": 7.0715159442806815e-06, + "loss": 0.9405, + "step": 25755 + }, + { + "epoch": 0.61, + "grad_norm": 1.9705374241509725, + "learning_rate": 7.0707863643180204e-06, + "loss": 0.9855, + "step": 25756 + }, + { + "epoch": 0.61, + "grad_norm": 1.8112383046764593, + "learning_rate": 7.070056801410165e-06, + "loss": 0.8379, + "step": 25757 + }, + { + "epoch": 0.61, + "grad_norm": 2.103357601229315, + "learning_rate": 7.069327255561369e-06, + "loss": 1.0194, + "step": 25758 + }, + { + "epoch": 0.61, + "grad_norm": 2.520827557536065, + "learning_rate": 7.068597726775879e-06, + "loss": 1.0528, + "step": 25759 + }, + { + "epoch": 0.61, + "grad_norm": 2.3083346172231214, + "learning_rate": 7.0678682150579355e-06, + "loss": 1.0074, + "step": 25760 + }, + { + "epoch": 0.61, + "grad_norm": 1.9673814857674985, + "learning_rate": 7.067138720411795e-06, + "loss": 1.0182, + "step": 25761 + }, + { + "epoch": 0.61, + "grad_norm": 1.8669620780859448, + "learning_rate": 7.0664092428416976e-06, + "loss": 0.9044, + "step": 25762 + }, + { + "epoch": 0.61, + "grad_norm": 1.104728274026248, + "learning_rate": 7.065679782351902e-06, + "loss": 1.0121, + "step": 25763 + }, + { + "epoch": 0.61, + "grad_norm": 2.253902523415359, + "learning_rate": 7.064950338946641e-06, + "loss": 0.9967, + "step": 25764 + }, + { + "epoch": 0.61, + "grad_norm": 3.565175351727365, + "learning_rate": 7.064220912630171e-06, + "loss": 1.0718, + "step": 25765 + }, + { + "epoch": 0.61, + "grad_norm": 2.4576229092284962, + "learning_rate": 7.063491503406737e-06, + "loss": 1.0046, + "step": 25766 + }, + { + "epoch": 0.61, + "grad_norm": 2.5169065320838895, + "learning_rate": 7.062762111280585e-06, + "loss": 0.8919, + "step": 25767 + }, + { + "epoch": 0.61, + "grad_norm": 2.0960898937914276, + "learning_rate": 7.062032736255959e-06, + "loss": 1.098, + "step": 25768 + }, + { + "epoch": 0.61, + "grad_norm": 1.9773095248486534, + "learning_rate": 7.06130337833711e-06, + "loss": 0.9287, + "step": 25769 + }, + { + "epoch": 0.61, + "grad_norm": 2.1934092798143556, + "learning_rate": 7.060574037528285e-06, + "loss": 1.1015, + "step": 25770 + }, + { + "epoch": 0.61, + "grad_norm": 2.027360248475573, + "learning_rate": 7.059844713833726e-06, + "loss": 0.8894, + "step": 25771 + }, + { + "epoch": 0.61, + "grad_norm": 2.1328176155133782, + "learning_rate": 7.059115407257683e-06, + "loss": 1.1275, + "step": 25772 + }, + { + "epoch": 0.61, + "grad_norm": 1.9989331378304738, + "learning_rate": 7.058386117804401e-06, + "loss": 0.9494, + "step": 25773 + }, + { + "epoch": 0.61, + "grad_norm": 1.1674937071382712, + "learning_rate": 7.057656845478129e-06, + "loss": 0.9501, + "step": 25774 + }, + { + "epoch": 0.61, + "grad_norm": 1.9381298607924737, + "learning_rate": 7.056927590283104e-06, + "loss": 0.9865, + "step": 25775 + }, + { + "epoch": 0.61, + "grad_norm": 1.9073213975106298, + "learning_rate": 7.056198352223585e-06, + "loss": 1.0029, + "step": 25776 + }, + { + "epoch": 0.61, + "grad_norm": 2.107923922737911, + "learning_rate": 7.055469131303809e-06, + "loss": 0.9846, + "step": 25777 + }, + { + "epoch": 0.61, + "grad_norm": 1.8966051733780331, + "learning_rate": 7.0547399275280235e-06, + "loss": 1.0692, + "step": 25778 + }, + { + "epoch": 0.61, + "grad_norm": 1.9090803208310057, + "learning_rate": 7.054010740900472e-06, + "loss": 0.9873, + "step": 25779 + }, + { + "epoch": 0.61, + "grad_norm": 2.3690398427529935, + "learning_rate": 7.053281571425405e-06, + "loss": 1.0299, + "step": 25780 + }, + { + "epoch": 0.61, + "grad_norm": 1.98030248885531, + "learning_rate": 7.052552419107068e-06, + "loss": 1.1032, + "step": 25781 + }, + { + "epoch": 0.61, + "grad_norm": 2.5939050414995863, + "learning_rate": 7.051823283949702e-06, + "loss": 1.0851, + "step": 25782 + }, + { + "epoch": 0.61, + "grad_norm": 1.1408479183679487, + "learning_rate": 7.051094165957553e-06, + "loss": 0.9594, + "step": 25783 + }, + { + "epoch": 0.61, + "grad_norm": 2.0001989676366976, + "learning_rate": 7.05036506513487e-06, + "loss": 1.0531, + "step": 25784 + }, + { + "epoch": 0.61, + "grad_norm": 1.942420300313258, + "learning_rate": 7.049635981485894e-06, + "loss": 0.8682, + "step": 25785 + }, + { + "epoch": 0.61, + "grad_norm": 1.930169140982599, + "learning_rate": 7.048906915014871e-06, + "loss": 0.969, + "step": 25786 + }, + { + "epoch": 0.61, + "grad_norm": 2.110967055669466, + "learning_rate": 7.048177865726044e-06, + "loss": 0.9406, + "step": 25787 + }, + { + "epoch": 0.61, + "grad_norm": 1.085215693251332, + "learning_rate": 7.0474488336236645e-06, + "loss": 0.9115, + "step": 25788 + }, + { + "epoch": 0.61, + "grad_norm": 2.144049612533225, + "learning_rate": 7.0467198187119705e-06, + "loss": 0.9727, + "step": 25789 + }, + { + "epoch": 0.61, + "grad_norm": 2.1421001316473056, + "learning_rate": 7.045990820995208e-06, + "loss": 1.0162, + "step": 25790 + }, + { + "epoch": 0.61, + "grad_norm": 2.2022221735284098, + "learning_rate": 7.0452618404776255e-06, + "loss": 1.0179, + "step": 25791 + }, + { + "epoch": 0.61, + "grad_norm": 1.7720942508092226, + "learning_rate": 7.044532877163459e-06, + "loss": 0.9421, + "step": 25792 + }, + { + "epoch": 0.61, + "grad_norm": 2.1122790596069096, + "learning_rate": 7.043803931056961e-06, + "loss": 0.8275, + "step": 25793 + }, + { + "epoch": 0.61, + "grad_norm": 2.0221993272775864, + "learning_rate": 7.04307500216237e-06, + "loss": 1.0736, + "step": 25794 + }, + { + "epoch": 0.61, + "grad_norm": 1.9419975954818969, + "learning_rate": 7.042346090483937e-06, + "loss": 0.855, + "step": 25795 + }, + { + "epoch": 0.61, + "grad_norm": 1.9239778542707264, + "learning_rate": 7.041617196025896e-06, + "loss": 1.0475, + "step": 25796 + }, + { + "epoch": 0.61, + "grad_norm": 1.9544807090805334, + "learning_rate": 7.040888318792499e-06, + "loss": 0.9892, + "step": 25797 + }, + { + "epoch": 0.61, + "grad_norm": 2.0369409124500577, + "learning_rate": 7.04015945878799e-06, + "loss": 1.0805, + "step": 25798 + }, + { + "epoch": 0.61, + "grad_norm": 2.140216923150116, + "learning_rate": 7.039430616016605e-06, + "loss": 0.9789, + "step": 25799 + }, + { + "epoch": 0.61, + "grad_norm": 1.1538676272225843, + "learning_rate": 7.038701790482592e-06, + "loss": 0.955, + "step": 25800 + }, + { + "epoch": 0.61, + "grad_norm": 1.1242103852046332, + "learning_rate": 7.037972982190197e-06, + "loss": 0.9443, + "step": 25801 + }, + { + "epoch": 0.61, + "grad_norm": 3.2820020264815626, + "learning_rate": 7.037244191143662e-06, + "loss": 1.0295, + "step": 25802 + }, + { + "epoch": 0.61, + "grad_norm": 2.0833735775289126, + "learning_rate": 7.036515417347228e-06, + "loss": 0.8408, + "step": 25803 + }, + { + "epoch": 0.61, + "grad_norm": 1.953970198502837, + "learning_rate": 7.035786660805137e-06, + "loss": 1.0971, + "step": 25804 + }, + { + "epoch": 0.61, + "grad_norm": 1.840454815948362, + "learning_rate": 7.03505792152164e-06, + "loss": 0.9831, + "step": 25805 + }, + { + "epoch": 0.61, + "grad_norm": 3.1439837848074097, + "learning_rate": 7.034329199500971e-06, + "loss": 1.07, + "step": 25806 + }, + { + "epoch": 0.61, + "grad_norm": 1.7184756307900941, + "learning_rate": 7.033600494747377e-06, + "loss": 1.1054, + "step": 25807 + }, + { + "epoch": 0.61, + "grad_norm": 2.1352035574098616, + "learning_rate": 7.032871807265097e-06, + "loss": 1.0965, + "step": 25808 + }, + { + "epoch": 0.61, + "grad_norm": 2.115538946456341, + "learning_rate": 7.032143137058382e-06, + "loss": 1.0493, + "step": 25809 + }, + { + "epoch": 0.61, + "grad_norm": 2.0785191719404517, + "learning_rate": 7.031414484131467e-06, + "loss": 1.0426, + "step": 25810 + }, + { + "epoch": 0.61, + "grad_norm": 1.9416970517549748, + "learning_rate": 7.030685848488595e-06, + "loss": 0.9899, + "step": 25811 + }, + { + "epoch": 0.61, + "grad_norm": 1.969834408079606, + "learning_rate": 7.029957230134015e-06, + "loss": 0.8401, + "step": 25812 + }, + { + "epoch": 0.61, + "grad_norm": 3.460558605934489, + "learning_rate": 7.029228629071958e-06, + "loss": 1.0082, + "step": 25813 + }, + { + "epoch": 0.61, + "grad_norm": 2.1866738577313387, + "learning_rate": 7.028500045306676e-06, + "loss": 1.0349, + "step": 25814 + }, + { + "epoch": 0.61, + "grad_norm": 2.001851214978402, + "learning_rate": 7.027771478842405e-06, + "loss": 0.9763, + "step": 25815 + }, + { + "epoch": 0.61, + "grad_norm": 1.962378431033546, + "learning_rate": 7.027042929683394e-06, + "loss": 0.9134, + "step": 25816 + }, + { + "epoch": 0.61, + "grad_norm": 2.4378200575058497, + "learning_rate": 7.0263143978338756e-06, + "loss": 0.9868, + "step": 25817 + }, + { + "epoch": 0.61, + "grad_norm": 2.667319252942325, + "learning_rate": 7.025585883298099e-06, + "loss": 0.9117, + "step": 25818 + }, + { + "epoch": 0.61, + "grad_norm": 1.8605773427703318, + "learning_rate": 7.0248573860803035e-06, + "loss": 1.0215, + "step": 25819 + }, + { + "epoch": 0.61, + "grad_norm": 2.841061470729714, + "learning_rate": 7.024128906184726e-06, + "loss": 1.002, + "step": 25820 + }, + { + "epoch": 0.61, + "grad_norm": 2.173523708601643, + "learning_rate": 7.0234004436156155e-06, + "loss": 0.9436, + "step": 25821 + }, + { + "epoch": 0.61, + "grad_norm": 1.1239535471981417, + "learning_rate": 7.02267199837721e-06, + "loss": 0.9031, + "step": 25822 + }, + { + "epoch": 0.61, + "grad_norm": 2.444083087091318, + "learning_rate": 7.021943570473751e-06, + "loss": 0.9114, + "step": 25823 + }, + { + "epoch": 0.61, + "grad_norm": 1.9901235721453783, + "learning_rate": 7.021215159909478e-06, + "loss": 1.1004, + "step": 25824 + }, + { + "epoch": 0.61, + "grad_norm": 2.6579612482446184, + "learning_rate": 7.020486766688633e-06, + "loss": 0.9159, + "step": 25825 + }, + { + "epoch": 0.61, + "grad_norm": 1.7776341284925976, + "learning_rate": 7.019758390815461e-06, + "loss": 0.9874, + "step": 25826 + }, + { + "epoch": 0.61, + "grad_norm": 1.9282949887597591, + "learning_rate": 7.019030032294196e-06, + "loss": 1.1274, + "step": 25827 + }, + { + "epoch": 0.61, + "grad_norm": 1.9051192495953562, + "learning_rate": 7.0183016911290816e-06, + "loss": 0.9992, + "step": 25828 + }, + { + "epoch": 0.61, + "grad_norm": 1.8977768478455566, + "learning_rate": 7.0175733673243586e-06, + "loss": 1.0126, + "step": 25829 + }, + { + "epoch": 0.61, + "grad_norm": 1.9654455721382051, + "learning_rate": 7.016845060884271e-06, + "loss": 0.8435, + "step": 25830 + }, + { + "epoch": 0.61, + "grad_norm": 2.548095174158708, + "learning_rate": 7.016116771813053e-06, + "loss": 1.0273, + "step": 25831 + }, + { + "epoch": 0.61, + "grad_norm": 2.0453010148858106, + "learning_rate": 7.015388500114947e-06, + "loss": 1.0963, + "step": 25832 + }, + { + "epoch": 0.61, + "grad_norm": 1.8978374687272943, + "learning_rate": 7.014660245794197e-06, + "loss": 1.1183, + "step": 25833 + }, + { + "epoch": 0.61, + "grad_norm": 2.2901305467108353, + "learning_rate": 7.013932008855039e-06, + "loss": 1.1244, + "step": 25834 + }, + { + "epoch": 0.61, + "grad_norm": 1.9327732026099573, + "learning_rate": 7.013203789301713e-06, + "loss": 0.9734, + "step": 25835 + }, + { + "epoch": 0.61, + "grad_norm": 1.9740480167196313, + "learning_rate": 7.01247558713846e-06, + "loss": 0.9456, + "step": 25836 + }, + { + "epoch": 0.61, + "grad_norm": 2.252902575723681, + "learning_rate": 7.011747402369522e-06, + "loss": 1.0165, + "step": 25837 + }, + { + "epoch": 0.61, + "grad_norm": 1.9635140672033304, + "learning_rate": 7.011019234999137e-06, + "loss": 0.9062, + "step": 25838 + }, + { + "epoch": 0.61, + "grad_norm": 1.9571866562190319, + "learning_rate": 7.010291085031542e-06, + "loss": 1.0715, + "step": 25839 + }, + { + "epoch": 0.61, + "grad_norm": 2.190308891144644, + "learning_rate": 7.0095629524709826e-06, + "loss": 1.0844, + "step": 25840 + }, + { + "epoch": 0.61, + "grad_norm": 2.1627154007082687, + "learning_rate": 7.008834837321689e-06, + "loss": 0.8571, + "step": 25841 + }, + { + "epoch": 0.61, + "grad_norm": 2.0617991474957647, + "learning_rate": 7.0081067395879085e-06, + "loss": 1.1346, + "step": 25842 + }, + { + "epoch": 0.61, + "grad_norm": 1.9790942591228708, + "learning_rate": 7.007378659273878e-06, + "loss": 0.9714, + "step": 25843 + }, + { + "epoch": 0.61, + "grad_norm": 1.9408448463437122, + "learning_rate": 7.006650596383839e-06, + "loss": 0.8594, + "step": 25844 + }, + { + "epoch": 0.61, + "grad_norm": 2.0090988723039023, + "learning_rate": 7.005922550922023e-06, + "loss": 0.9599, + "step": 25845 + }, + { + "epoch": 0.61, + "grad_norm": 1.935715740162904, + "learning_rate": 7.005194522892676e-06, + "loss": 1.0646, + "step": 25846 + }, + { + "epoch": 0.61, + "grad_norm": 1.8786383278464982, + "learning_rate": 7.004466512300037e-06, + "loss": 0.9824, + "step": 25847 + }, + { + "epoch": 0.61, + "grad_norm": 1.0807045429413886, + "learning_rate": 7.00373851914834e-06, + "loss": 0.9196, + "step": 25848 + }, + { + "epoch": 0.61, + "grad_norm": 2.153074675182182, + "learning_rate": 7.003010543441824e-06, + "loss": 1.0231, + "step": 25849 + }, + { + "epoch": 0.61, + "grad_norm": 2.228462860107957, + "learning_rate": 7.002282585184731e-06, + "loss": 1.0234, + "step": 25850 + }, + { + "epoch": 0.61, + "grad_norm": 1.848523119003516, + "learning_rate": 7.0015546443813e-06, + "loss": 1.0363, + "step": 25851 + }, + { + "epoch": 0.61, + "grad_norm": 3.535690967976346, + "learning_rate": 7.000826721035765e-06, + "loss": 0.8873, + "step": 25852 + }, + { + "epoch": 0.61, + "grad_norm": 2.6560208939699126, + "learning_rate": 7.000098815152365e-06, + "loss": 0.9022, + "step": 25853 + }, + { + "epoch": 0.61, + "grad_norm": 1.1083418220395347, + "learning_rate": 6.999370926735342e-06, + "loss": 0.9162, + "step": 25854 + }, + { + "epoch": 0.61, + "grad_norm": 1.9815602393189886, + "learning_rate": 6.998643055788932e-06, + "loss": 0.8658, + "step": 25855 + }, + { + "epoch": 0.61, + "grad_norm": 2.325935353346417, + "learning_rate": 6.997915202317372e-06, + "loss": 0.962, + "step": 25856 + }, + { + "epoch": 0.61, + "grad_norm": 2.1109612421391946, + "learning_rate": 6.997187366324895e-06, + "loss": 0.949, + "step": 25857 + }, + { + "epoch": 0.61, + "grad_norm": 2.120834607532025, + "learning_rate": 6.99645954781575e-06, + "loss": 1.0241, + "step": 25858 + }, + { + "epoch": 0.61, + "grad_norm": 1.9960634548637324, + "learning_rate": 6.9957317467941675e-06, + "loss": 0.9499, + "step": 25859 + }, + { + "epoch": 0.61, + "grad_norm": 1.8787501191214453, + "learning_rate": 6.995003963264382e-06, + "loss": 0.9846, + "step": 25860 + }, + { + "epoch": 0.61, + "grad_norm": 1.9567527198625283, + "learning_rate": 6.994276197230643e-06, + "loss": 0.9148, + "step": 25861 + }, + { + "epoch": 0.61, + "grad_norm": 2.115545864540985, + "learning_rate": 6.993548448697171e-06, + "loss": 1.0792, + "step": 25862 + }, + { + "epoch": 0.61, + "grad_norm": 2.2593433024438863, + "learning_rate": 6.992820717668215e-06, + "loss": 0.9674, + "step": 25863 + }, + { + "epoch": 0.61, + "grad_norm": 2.515142444862538, + "learning_rate": 6.992093004148006e-06, + "loss": 0.9005, + "step": 25864 + }, + { + "epoch": 0.61, + "grad_norm": 1.125066647194575, + "learning_rate": 6.991365308140792e-06, + "loss": 1.0033, + "step": 25865 + }, + { + "epoch": 0.61, + "grad_norm": 1.9550751029052205, + "learning_rate": 6.990637629650794e-06, + "loss": 0.8937, + "step": 25866 + }, + { + "epoch": 0.61, + "grad_norm": 1.973030004290441, + "learning_rate": 6.989909968682258e-06, + "loss": 1.0497, + "step": 25867 + }, + { + "epoch": 0.61, + "grad_norm": 2.2723504376891737, + "learning_rate": 6.989182325239423e-06, + "loss": 0.9786, + "step": 25868 + }, + { + "epoch": 0.61, + "grad_norm": 2.1679357808093505, + "learning_rate": 6.988454699326515e-06, + "loss": 1.0325, + "step": 25869 + }, + { + "epoch": 0.61, + "grad_norm": 2.262144982903013, + "learning_rate": 6.987727090947781e-06, + "loss": 0.9378, + "step": 25870 + }, + { + "epoch": 0.61, + "grad_norm": 2.871296129506017, + "learning_rate": 6.986999500107454e-06, + "loss": 1.0259, + "step": 25871 + }, + { + "epoch": 0.61, + "grad_norm": 2.567729772065329, + "learning_rate": 6.9862719268097715e-06, + "loss": 1.1021, + "step": 25872 + }, + { + "epoch": 0.61, + "grad_norm": 1.934233109415257, + "learning_rate": 6.985544371058963e-06, + "loss": 0.8986, + "step": 25873 + }, + { + "epoch": 0.61, + "grad_norm": 1.8863658194844948, + "learning_rate": 6.984816832859271e-06, + "loss": 0.9664, + "step": 25874 + }, + { + "epoch": 0.61, + "grad_norm": 2.378601457587425, + "learning_rate": 6.9840893122149346e-06, + "loss": 1.0031, + "step": 25875 + }, + { + "epoch": 0.61, + "grad_norm": 2.5050178506304643, + "learning_rate": 6.98336180913018e-06, + "loss": 1.0454, + "step": 25876 + }, + { + "epoch": 0.61, + "grad_norm": 2.014912315550699, + "learning_rate": 6.982634323609249e-06, + "loss": 0.946, + "step": 25877 + }, + { + "epoch": 0.61, + "grad_norm": 1.9625820076244502, + "learning_rate": 6.981906855656376e-06, + "loss": 1.0951, + "step": 25878 + }, + { + "epoch": 0.61, + "grad_norm": 1.8198610277882603, + "learning_rate": 6.9811794052758e-06, + "loss": 0.9115, + "step": 25879 + }, + { + "epoch": 0.61, + "grad_norm": 1.9495722071812265, + "learning_rate": 6.980451972471751e-06, + "loss": 0.9713, + "step": 25880 + }, + { + "epoch": 0.61, + "grad_norm": 2.2875189803282545, + "learning_rate": 6.979724557248464e-06, + "loss": 0.9687, + "step": 25881 + }, + { + "epoch": 0.61, + "grad_norm": 1.976786017148379, + "learning_rate": 6.978997159610181e-06, + "loss": 0.9415, + "step": 25882 + }, + { + "epoch": 0.61, + "grad_norm": 1.9847845995652387, + "learning_rate": 6.978269779561131e-06, + "loss": 0.9873, + "step": 25883 + }, + { + "epoch": 0.61, + "grad_norm": 1.8090465636129245, + "learning_rate": 6.977542417105552e-06, + "loss": 0.9173, + "step": 25884 + }, + { + "epoch": 0.61, + "grad_norm": 2.32349719279191, + "learning_rate": 6.9768150722476744e-06, + "loss": 0.8465, + "step": 25885 + }, + { + "epoch": 0.61, + "grad_norm": 2.25120697652324, + "learning_rate": 6.976087744991742e-06, + "loss": 0.9735, + "step": 25886 + }, + { + "epoch": 0.61, + "grad_norm": 1.8701519513447307, + "learning_rate": 6.975360435341981e-06, + "loss": 0.9655, + "step": 25887 + }, + { + "epoch": 0.61, + "grad_norm": 1.8761491738299578, + "learning_rate": 6.974633143302629e-06, + "loss": 0.9782, + "step": 25888 + }, + { + "epoch": 0.61, + "grad_norm": 2.4560707373312702, + "learning_rate": 6.973905868877923e-06, + "loss": 1.0527, + "step": 25889 + }, + { + "epoch": 0.61, + "grad_norm": 1.8265249002255175, + "learning_rate": 6.973178612072092e-06, + "loss": 0.9816, + "step": 25890 + }, + { + "epoch": 0.61, + "grad_norm": 2.028231938725353, + "learning_rate": 6.972451372889375e-06, + "loss": 1.0012, + "step": 25891 + }, + { + "epoch": 0.61, + "grad_norm": 1.9597448354359337, + "learning_rate": 6.971724151334005e-06, + "loss": 0.9693, + "step": 25892 + }, + { + "epoch": 0.61, + "grad_norm": 1.9473307157712478, + "learning_rate": 6.970996947410217e-06, + "loss": 1.1529, + "step": 25893 + }, + { + "epoch": 0.61, + "grad_norm": 2.0493718535834518, + "learning_rate": 6.9702697611222394e-06, + "loss": 0.9061, + "step": 25894 + }, + { + "epoch": 0.61, + "grad_norm": 2.187008500123429, + "learning_rate": 6.9695425924743125e-06, + "loss": 1.0271, + "step": 25895 + }, + { + "epoch": 0.61, + "grad_norm": 1.9464088068505903, + "learning_rate": 6.968815441470666e-06, + "loss": 0.9617, + "step": 25896 + }, + { + "epoch": 0.61, + "grad_norm": 2.082729752049409, + "learning_rate": 6.968088308115543e-06, + "loss": 1.007, + "step": 25897 + }, + { + "epoch": 0.61, + "grad_norm": 1.931767648747799, + "learning_rate": 6.967361192413163e-06, + "loss": 1.0072, + "step": 25898 + }, + { + "epoch": 0.61, + "grad_norm": 2.0635689419334176, + "learning_rate": 6.966634094367768e-06, + "loss": 1.0568, + "step": 25899 + }, + { + "epoch": 0.61, + "grad_norm": 1.9228539202270274, + "learning_rate": 6.96590701398359e-06, + "loss": 0.9937, + "step": 25900 + }, + { + "epoch": 0.61, + "grad_norm": 1.147554188145204, + "learning_rate": 6.965179951264862e-06, + "loss": 0.9209, + "step": 25901 + }, + { + "epoch": 0.61, + "grad_norm": 2.7803828375690487, + "learning_rate": 6.964452906215815e-06, + "loss": 1.0147, + "step": 25902 + }, + { + "epoch": 0.61, + "grad_norm": 2.5499877769901, + "learning_rate": 6.963725878840686e-06, + "loss": 0.9315, + "step": 25903 + }, + { + "epoch": 0.61, + "grad_norm": 2.0357764638931894, + "learning_rate": 6.9629988691437086e-06, + "loss": 1.0968, + "step": 25904 + }, + { + "epoch": 0.61, + "grad_norm": 2.3303896931704227, + "learning_rate": 6.962271877129111e-06, + "loss": 0.9559, + "step": 25905 + }, + { + "epoch": 0.61, + "grad_norm": 1.1323204694342874, + "learning_rate": 6.961544902801126e-06, + "loss": 0.9341, + "step": 25906 + }, + { + "epoch": 0.61, + "grad_norm": 2.056282491942675, + "learning_rate": 6.960817946163993e-06, + "loss": 0.918, + "step": 25907 + }, + { + "epoch": 0.61, + "grad_norm": 1.9791845279936473, + "learning_rate": 6.960091007221938e-06, + "loss": 1.0437, + "step": 25908 + }, + { + "epoch": 0.61, + "grad_norm": 2.0414796867475893, + "learning_rate": 6.959364085979192e-06, + "loss": 0.968, + "step": 25909 + }, + { + "epoch": 0.61, + "grad_norm": 1.1512877056280277, + "learning_rate": 6.958637182439996e-06, + "loss": 0.9674, + "step": 25910 + }, + { + "epoch": 0.61, + "grad_norm": 5.251801188102293, + "learning_rate": 6.957910296608576e-06, + "loss": 1.0351, + "step": 25911 + }, + { + "epoch": 0.61, + "grad_norm": 2.366581183705712, + "learning_rate": 6.957183428489165e-06, + "loss": 0.9791, + "step": 25912 + }, + { + "epoch": 0.61, + "grad_norm": 2.1791018147573986, + "learning_rate": 6.956456578085994e-06, + "loss": 0.8546, + "step": 25913 + }, + { + "epoch": 0.61, + "grad_norm": 1.9139051984601056, + "learning_rate": 6.955729745403301e-06, + "loss": 1.0512, + "step": 25914 + }, + { + "epoch": 0.61, + "grad_norm": 2.0306948425845532, + "learning_rate": 6.955002930445308e-06, + "loss": 1.0717, + "step": 25915 + }, + { + "epoch": 0.61, + "grad_norm": 2.1952866452194693, + "learning_rate": 6.954276133216254e-06, + "loss": 1.0103, + "step": 25916 + }, + { + "epoch": 0.61, + "grad_norm": 2.050362940995819, + "learning_rate": 6.953549353720366e-06, + "loss": 0.9201, + "step": 25917 + }, + { + "epoch": 0.61, + "grad_norm": 2.119953282612829, + "learning_rate": 6.952822591961883e-06, + "loss": 1.0106, + "step": 25918 + }, + { + "epoch": 0.61, + "grad_norm": 1.874336725194016, + "learning_rate": 6.9520958479450285e-06, + "loss": 0.9855, + "step": 25919 + }, + { + "epoch": 0.61, + "grad_norm": 2.01604755939589, + "learning_rate": 6.9513691216740385e-06, + "loss": 1.0813, + "step": 25920 + }, + { + "epoch": 0.61, + "grad_norm": 2.1748743352285107, + "learning_rate": 6.950642413153144e-06, + "loss": 0.9776, + "step": 25921 + }, + { + "epoch": 0.61, + "grad_norm": 1.9560542733324844, + "learning_rate": 6.949915722386571e-06, + "loss": 0.928, + "step": 25922 + }, + { + "epoch": 0.61, + "grad_norm": 1.8957736729484853, + "learning_rate": 6.949189049378555e-06, + "loss": 0.9432, + "step": 25923 + }, + { + "epoch": 0.61, + "grad_norm": 2.140958099983542, + "learning_rate": 6.9484623941333274e-06, + "loss": 1.0572, + "step": 25924 + }, + { + "epoch": 0.61, + "grad_norm": 2.417629593695635, + "learning_rate": 6.94773575665512e-06, + "loss": 1.0984, + "step": 25925 + }, + { + "epoch": 0.61, + "grad_norm": 1.9426857220412592, + "learning_rate": 6.947009136948157e-06, + "loss": 0.9986, + "step": 25926 + }, + { + "epoch": 0.61, + "grad_norm": 1.8867763984108543, + "learning_rate": 6.9462825350166744e-06, + "loss": 1.0219, + "step": 25927 + }, + { + "epoch": 0.61, + "grad_norm": 1.9676361350818825, + "learning_rate": 6.945555950864904e-06, + "loss": 1.0715, + "step": 25928 + }, + { + "epoch": 0.61, + "grad_norm": 1.8941723199650218, + "learning_rate": 6.944829384497073e-06, + "loss": 1.0459, + "step": 25929 + }, + { + "epoch": 0.61, + "grad_norm": 1.8023656555065315, + "learning_rate": 6.94410283591741e-06, + "loss": 1.0528, + "step": 25930 + }, + { + "epoch": 0.61, + "grad_norm": 1.9947870559515632, + "learning_rate": 6.943376305130149e-06, + "loss": 0.9256, + "step": 25931 + }, + { + "epoch": 0.61, + "grad_norm": 1.7686554376097934, + "learning_rate": 6.942649792139522e-06, + "loss": 0.8249, + "step": 25932 + }, + { + "epoch": 0.61, + "grad_norm": 1.9019703153988787, + "learning_rate": 6.941923296949753e-06, + "loss": 0.8343, + "step": 25933 + }, + { + "epoch": 0.61, + "grad_norm": 2.1131630989982835, + "learning_rate": 6.941196819565074e-06, + "loss": 1.0464, + "step": 25934 + }, + { + "epoch": 0.61, + "grad_norm": 2.3945727194509674, + "learning_rate": 6.940470359989719e-06, + "loss": 0.9459, + "step": 25935 + }, + { + "epoch": 0.61, + "grad_norm": 1.9196305574201493, + "learning_rate": 6.939743918227911e-06, + "loss": 1.0731, + "step": 25936 + }, + { + "epoch": 0.61, + "grad_norm": 2.007738031428494, + "learning_rate": 6.939017494283884e-06, + "loss": 0.9446, + "step": 25937 + }, + { + "epoch": 0.61, + "grad_norm": 1.9560635931218895, + "learning_rate": 6.938291088161864e-06, + "loss": 0.9725, + "step": 25938 + }, + { + "epoch": 0.61, + "grad_norm": 2.18307407022113, + "learning_rate": 6.937564699866086e-06, + "loss": 0.9945, + "step": 25939 + }, + { + "epoch": 0.61, + "grad_norm": 2.2269589407914654, + "learning_rate": 6.936838329400774e-06, + "loss": 1.0506, + "step": 25940 + }, + { + "epoch": 0.61, + "grad_norm": 1.9054923623449205, + "learning_rate": 6.93611197677016e-06, + "loss": 1.0314, + "step": 25941 + }, + { + "epoch": 0.61, + "grad_norm": 2.2116515910185575, + "learning_rate": 6.9353856419784734e-06, + "loss": 1.0483, + "step": 25942 + }, + { + "epoch": 0.61, + "grad_norm": 1.8747324419229467, + "learning_rate": 6.934659325029938e-06, + "loss": 0.8985, + "step": 25943 + }, + { + "epoch": 0.61, + "grad_norm": 1.9631353944498247, + "learning_rate": 6.933933025928788e-06, + "loss": 1.009, + "step": 25944 + }, + { + "epoch": 0.61, + "grad_norm": 2.1349679772686097, + "learning_rate": 6.933206744679249e-06, + "loss": 1.0607, + "step": 25945 + }, + { + "epoch": 0.61, + "grad_norm": 1.934210984256015, + "learning_rate": 6.932480481285558e-06, + "loss": 0.9575, + "step": 25946 + }, + { + "epoch": 0.61, + "grad_norm": 2.050405494802629, + "learning_rate": 6.931754235751929e-06, + "loss": 1.0322, + "step": 25947 + }, + { + "epoch": 0.61, + "grad_norm": 6.13151522084894, + "learning_rate": 6.931028008082602e-06, + "loss": 1.0888, + "step": 25948 + }, + { + "epoch": 0.61, + "grad_norm": 1.8853934698753612, + "learning_rate": 6.930301798281803e-06, + "loss": 1.0595, + "step": 25949 + }, + { + "epoch": 0.61, + "grad_norm": 2.0508897539226734, + "learning_rate": 6.929575606353757e-06, + "loss": 0.9908, + "step": 25950 + }, + { + "epoch": 0.61, + "grad_norm": 1.8906369048819616, + "learning_rate": 6.928849432302691e-06, + "loss": 1.0346, + "step": 25951 + }, + { + "epoch": 0.61, + "grad_norm": 2.112391331201184, + "learning_rate": 6.9281232761328384e-06, + "loss": 1.028, + "step": 25952 + }, + { + "epoch": 0.61, + "grad_norm": 1.826069091770501, + "learning_rate": 6.927397137848427e-06, + "loss": 1.0106, + "step": 25953 + }, + { + "epoch": 0.61, + "grad_norm": 2.057234427957575, + "learning_rate": 6.926671017453678e-06, + "loss": 0.9872, + "step": 25954 + }, + { + "epoch": 0.61, + "grad_norm": 1.8681778024885063, + "learning_rate": 6.925944914952825e-06, + "loss": 1.045, + "step": 25955 + }, + { + "epoch": 0.61, + "grad_norm": 2.033866777792556, + "learning_rate": 6.925218830350097e-06, + "loss": 1.0715, + "step": 25956 + }, + { + "epoch": 0.61, + "grad_norm": 1.876528219338504, + "learning_rate": 6.924492763649715e-06, + "loss": 0.9803, + "step": 25957 + }, + { + "epoch": 0.61, + "grad_norm": 1.9404047271820284, + "learning_rate": 6.923766714855909e-06, + "loss": 0.9345, + "step": 25958 + }, + { + "epoch": 0.61, + "grad_norm": 2.0762041645419007, + "learning_rate": 6.923040683972908e-06, + "loss": 1.1216, + "step": 25959 + }, + { + "epoch": 0.61, + "grad_norm": 2.14731284789299, + "learning_rate": 6.922314671004942e-06, + "loss": 1.0159, + "step": 25960 + }, + { + "epoch": 0.61, + "grad_norm": 1.922058761541814, + "learning_rate": 6.92158867595623e-06, + "loss": 1.1077, + "step": 25961 + }, + { + "epoch": 0.61, + "grad_norm": 2.120886877546415, + "learning_rate": 6.920862698831002e-06, + "loss": 1.0382, + "step": 25962 + }, + { + "epoch": 0.61, + "grad_norm": 1.0885427444930271, + "learning_rate": 6.920136739633494e-06, + "loss": 1.0008, + "step": 25963 + }, + { + "epoch": 0.61, + "grad_norm": 2.107315486420255, + "learning_rate": 6.919410798367917e-06, + "loss": 1.0099, + "step": 25964 + }, + { + "epoch": 0.61, + "grad_norm": 2.238674693038437, + "learning_rate": 6.9186848750385085e-06, + "loss": 1.029, + "step": 25965 + }, + { + "epoch": 0.61, + "grad_norm": 1.9897439678157378, + "learning_rate": 6.91795896964949e-06, + "loss": 0.9464, + "step": 25966 + }, + { + "epoch": 0.61, + "grad_norm": 2.0484759951594507, + "learning_rate": 6.917233082205094e-06, + "loss": 1.1189, + "step": 25967 + }, + { + "epoch": 0.61, + "grad_norm": 1.9996938591046498, + "learning_rate": 6.916507212709541e-06, + "loss": 1.0719, + "step": 25968 + }, + { + "epoch": 0.61, + "grad_norm": 1.789300220911036, + "learning_rate": 6.91578136116706e-06, + "loss": 1.0082, + "step": 25969 + }, + { + "epoch": 0.61, + "grad_norm": 2.6039122928470726, + "learning_rate": 6.915055527581878e-06, + "loss": 0.8846, + "step": 25970 + }, + { + "epoch": 0.61, + "grad_norm": 1.9199905424460204, + "learning_rate": 6.914329711958214e-06, + "loss": 1.0096, + "step": 25971 + }, + { + "epoch": 0.61, + "grad_norm": 1.8775784929822206, + "learning_rate": 6.913603914300302e-06, + "loss": 1.0379, + "step": 25972 + }, + { + "epoch": 0.61, + "grad_norm": 1.9136311973261353, + "learning_rate": 6.912878134612367e-06, + "loss": 1.0543, + "step": 25973 + }, + { + "epoch": 0.61, + "grad_norm": 2.385403808404834, + "learning_rate": 6.912152372898633e-06, + "loss": 1.072, + "step": 25974 + }, + { + "epoch": 0.61, + "grad_norm": 1.120901068592789, + "learning_rate": 6.911426629163323e-06, + "loss": 0.8962, + "step": 25975 + }, + { + "epoch": 0.61, + "grad_norm": 2.3231244865161114, + "learning_rate": 6.9107009034106655e-06, + "loss": 0.978, + "step": 25976 + }, + { + "epoch": 0.61, + "grad_norm": 2.4159613858990796, + "learning_rate": 6.909975195644887e-06, + "loss": 0.9107, + "step": 25977 + }, + { + "epoch": 0.61, + "grad_norm": 1.8752435093324757, + "learning_rate": 6.9092495058702105e-06, + "loss": 1.018, + "step": 25978 + }, + { + "epoch": 0.61, + "grad_norm": 2.0817869485466054, + "learning_rate": 6.9085238340908585e-06, + "loss": 0.9221, + "step": 25979 + }, + { + "epoch": 0.61, + "grad_norm": 2.258286753069635, + "learning_rate": 6.907798180311063e-06, + "loss": 1.0457, + "step": 25980 + }, + { + "epoch": 0.61, + "grad_norm": 1.1015988030691104, + "learning_rate": 6.907072544535048e-06, + "loss": 0.9225, + "step": 25981 + }, + { + "epoch": 0.61, + "grad_norm": 2.2826839849864764, + "learning_rate": 6.906346926767031e-06, + "loss": 1.0269, + "step": 25982 + }, + { + "epoch": 0.61, + "grad_norm": 1.7403251204715473, + "learning_rate": 6.905621327011241e-06, + "loss": 1.0246, + "step": 25983 + }, + { + "epoch": 0.61, + "grad_norm": 2.0876756339725375, + "learning_rate": 6.904895745271909e-06, + "loss": 0.976, + "step": 25984 + }, + { + "epoch": 0.61, + "grad_norm": 1.7863538537116217, + "learning_rate": 6.904170181553251e-06, + "loss": 0.9876, + "step": 25985 + }, + { + "epoch": 0.61, + "grad_norm": 2.13950441172588, + "learning_rate": 6.903444635859493e-06, + "loss": 0.9174, + "step": 25986 + }, + { + "epoch": 0.61, + "grad_norm": 2.0259182443015176, + "learning_rate": 6.902719108194859e-06, + "loss": 1.0092, + "step": 25987 + }, + { + "epoch": 0.61, + "grad_norm": 2.43876239549264, + "learning_rate": 6.90199359856358e-06, + "loss": 1.0501, + "step": 25988 + }, + { + "epoch": 0.61, + "grad_norm": 2.175814084099786, + "learning_rate": 6.901268106969872e-06, + "loss": 1.054, + "step": 25989 + }, + { + "epoch": 0.61, + "grad_norm": 1.8121966434934853, + "learning_rate": 6.9005426334179635e-06, + "loss": 0.9991, + "step": 25990 + }, + { + "epoch": 0.61, + "grad_norm": 1.0891204938137304, + "learning_rate": 6.899817177912078e-06, + "loss": 0.9741, + "step": 25991 + }, + { + "epoch": 0.61, + "grad_norm": 2.023679839929697, + "learning_rate": 6.899091740456434e-06, + "loss": 0.9847, + "step": 25992 + }, + { + "epoch": 0.61, + "grad_norm": 2.2508043233631554, + "learning_rate": 6.898366321055263e-06, + "loss": 1.0381, + "step": 25993 + }, + { + "epoch": 0.61, + "grad_norm": 2.1350619261200516, + "learning_rate": 6.897640919712782e-06, + "loss": 1.0783, + "step": 25994 + }, + { + "epoch": 0.61, + "grad_norm": 1.9661750798452793, + "learning_rate": 6.896915536433224e-06, + "loss": 0.9993, + "step": 25995 + }, + { + "epoch": 0.61, + "grad_norm": 2.1520363078920517, + "learning_rate": 6.8961901712208005e-06, + "loss": 1.0693, + "step": 25996 + }, + { + "epoch": 0.61, + "grad_norm": 2.0222060122554093, + "learning_rate": 6.895464824079743e-06, + "loss": 1.0413, + "step": 25997 + }, + { + "epoch": 0.61, + "grad_norm": 2.0479067382055622, + "learning_rate": 6.894739495014273e-06, + "loss": 0.9993, + "step": 25998 + }, + { + "epoch": 0.61, + "grad_norm": 1.1030103131018443, + "learning_rate": 6.894014184028611e-06, + "loss": 0.9084, + "step": 25999 + }, + { + "epoch": 0.61, + "grad_norm": 2.1177651516432916, + "learning_rate": 6.893288891126981e-06, + "loss": 0.9858, + "step": 26000 + }, + { + "epoch": 0.61, + "grad_norm": 2.044250952518681, + "learning_rate": 6.892563616313608e-06, + "loss": 0.9667, + "step": 26001 + }, + { + "epoch": 0.61, + "grad_norm": 1.9199777426575697, + "learning_rate": 6.891838359592714e-06, + "loss": 0.9459, + "step": 26002 + }, + { + "epoch": 0.61, + "grad_norm": 1.8752264195317432, + "learning_rate": 6.891113120968518e-06, + "loss": 1.049, + "step": 26003 + }, + { + "epoch": 0.61, + "grad_norm": 1.150612192784409, + "learning_rate": 6.890387900445248e-06, + "loss": 0.967, + "step": 26004 + }, + { + "epoch": 0.61, + "grad_norm": 1.983443281755184, + "learning_rate": 6.8896626980271265e-06, + "loss": 0.9052, + "step": 26005 + }, + { + "epoch": 0.61, + "grad_norm": 2.165551423516302, + "learning_rate": 6.8889375137183714e-06, + "loss": 0.9667, + "step": 26006 + }, + { + "epoch": 0.61, + "grad_norm": 2.1408240179447975, + "learning_rate": 6.888212347523204e-06, + "loss": 0.9673, + "step": 26007 + }, + { + "epoch": 0.61, + "grad_norm": 2.197362817204176, + "learning_rate": 6.887487199445852e-06, + "loss": 0.966, + "step": 26008 + }, + { + "epoch": 0.61, + "grad_norm": 1.952950497461162, + "learning_rate": 6.886762069490538e-06, + "loss": 0.9462, + "step": 26009 + }, + { + "epoch": 0.61, + "grad_norm": 1.9006615492906453, + "learning_rate": 6.886036957661477e-06, + "loss": 1.0165, + "step": 26010 + }, + { + "epoch": 0.61, + "grad_norm": 2.436274763931276, + "learning_rate": 6.885311863962894e-06, + "loss": 0.8366, + "step": 26011 + }, + { + "epoch": 0.61, + "grad_norm": 2.1180469394599895, + "learning_rate": 6.884586788399017e-06, + "loss": 0.9273, + "step": 26012 + }, + { + "epoch": 0.61, + "grad_norm": 1.8254057538129882, + "learning_rate": 6.883861730974057e-06, + "loss": 1.1195, + "step": 26013 + }, + { + "epoch": 0.61, + "grad_norm": 2.330885415449579, + "learning_rate": 6.883136691692241e-06, + "loss": 0.8967, + "step": 26014 + }, + { + "epoch": 0.61, + "grad_norm": 1.988761081743168, + "learning_rate": 6.882411670557788e-06, + "loss": 1.0015, + "step": 26015 + }, + { + "epoch": 0.61, + "grad_norm": 1.7875367471779111, + "learning_rate": 6.881686667574926e-06, + "loss": 0.9469, + "step": 26016 + }, + { + "epoch": 0.61, + "grad_norm": 2.135238807698117, + "learning_rate": 6.8809616827478695e-06, + "loss": 1.0251, + "step": 26017 + }, + { + "epoch": 0.61, + "grad_norm": 1.1053705799503653, + "learning_rate": 6.880236716080841e-06, + "loss": 0.9571, + "step": 26018 + }, + { + "epoch": 0.61, + "grad_norm": 2.610504872472078, + "learning_rate": 6.879511767578061e-06, + "loss": 1.1375, + "step": 26019 + }, + { + "epoch": 0.61, + "grad_norm": 3.8049979703749517, + "learning_rate": 6.878786837243755e-06, + "loss": 1.0682, + "step": 26020 + }, + { + "epoch": 0.61, + "grad_norm": 1.9387669170479032, + "learning_rate": 6.878061925082138e-06, + "loss": 1.0062, + "step": 26021 + }, + { + "epoch": 0.61, + "grad_norm": 2.335116422817839, + "learning_rate": 6.877337031097433e-06, + "loss": 0.9817, + "step": 26022 + }, + { + "epoch": 0.61, + "grad_norm": 1.8701071427675608, + "learning_rate": 6.8766121552938625e-06, + "loss": 1.0718, + "step": 26023 + }, + { + "epoch": 0.61, + "grad_norm": 2.089801697497319, + "learning_rate": 6.8758872976756406e-06, + "loss": 0.9928, + "step": 26024 + }, + { + "epoch": 0.61, + "grad_norm": 1.928567216805851, + "learning_rate": 6.875162458246994e-06, + "loss": 1.084, + "step": 26025 + }, + { + "epoch": 0.61, + "grad_norm": 1.9358616001896574, + "learning_rate": 6.874437637012141e-06, + "loss": 1.0678, + "step": 26026 + }, + { + "epoch": 0.61, + "grad_norm": 1.8611023494270216, + "learning_rate": 6.873712833975304e-06, + "loss": 1.0006, + "step": 26027 + }, + { + "epoch": 0.61, + "grad_norm": 1.9940085570496642, + "learning_rate": 6.872988049140695e-06, + "loss": 1.1044, + "step": 26028 + }, + { + "epoch": 0.61, + "grad_norm": 2.297711663682881, + "learning_rate": 6.872263282512544e-06, + "loss": 0.8721, + "step": 26029 + }, + { + "epoch": 0.61, + "grad_norm": 2.0757516188233085, + "learning_rate": 6.8715385340950655e-06, + "loss": 1.09, + "step": 26030 + }, + { + "epoch": 0.61, + "grad_norm": 2.5445805164787916, + "learning_rate": 6.87081380389248e-06, + "loss": 0.9975, + "step": 26031 + }, + { + "epoch": 0.61, + "grad_norm": 2.1287907849068315, + "learning_rate": 6.8700890919090044e-06, + "loss": 0.9747, + "step": 26032 + }, + { + "epoch": 0.61, + "grad_norm": 2.1229658432694807, + "learning_rate": 6.869364398148865e-06, + "loss": 0.8789, + "step": 26033 + }, + { + "epoch": 0.61, + "grad_norm": 1.8793915271363366, + "learning_rate": 6.868639722616277e-06, + "loss": 1.0446, + "step": 26034 + }, + { + "epoch": 0.61, + "grad_norm": 2.512827737882792, + "learning_rate": 6.8679150653154595e-06, + "loss": 1.0737, + "step": 26035 + }, + { + "epoch": 0.61, + "grad_norm": 2.053530876530434, + "learning_rate": 6.867190426250629e-06, + "loss": 1.0573, + "step": 26036 + }, + { + "epoch": 0.61, + "grad_norm": 2.1705412972032616, + "learning_rate": 6.8664658054260124e-06, + "loss": 1.0179, + "step": 26037 + }, + { + "epoch": 0.61, + "grad_norm": 1.9184385339690384, + "learning_rate": 6.865741202845821e-06, + "loss": 1.0078, + "step": 26038 + }, + { + "epoch": 0.61, + "grad_norm": 2.1001808068666636, + "learning_rate": 6.865016618514277e-06, + "loss": 0.9577, + "step": 26039 + }, + { + "epoch": 0.61, + "grad_norm": 2.086570309347559, + "learning_rate": 6.864292052435597e-06, + "loss": 1.0329, + "step": 26040 + }, + { + "epoch": 0.61, + "grad_norm": 1.9078115945705714, + "learning_rate": 6.863567504614006e-06, + "loss": 0.945, + "step": 26041 + }, + { + "epoch": 0.61, + "grad_norm": 1.884307371648079, + "learning_rate": 6.862842975053714e-06, + "loss": 1.0794, + "step": 26042 + }, + { + "epoch": 0.61, + "grad_norm": 1.8710992448806711, + "learning_rate": 6.862118463758944e-06, + "loss": 0.997, + "step": 26043 + }, + { + "epoch": 0.61, + "grad_norm": 1.9611295950551382, + "learning_rate": 6.8613939707339185e-06, + "loss": 0.9235, + "step": 26044 + }, + { + "epoch": 0.61, + "grad_norm": 2.1101450888606195, + "learning_rate": 6.860669495982844e-06, + "loss": 1.0169, + "step": 26045 + }, + { + "epoch": 0.61, + "grad_norm": 2.478206506650876, + "learning_rate": 6.859945039509948e-06, + "loss": 0.9406, + "step": 26046 + }, + { + "epoch": 0.61, + "grad_norm": 2.0551466045905187, + "learning_rate": 6.859220601319445e-06, + "loss": 1.0762, + "step": 26047 + }, + { + "epoch": 0.61, + "grad_norm": 2.2166686557667896, + "learning_rate": 6.8584961814155595e-06, + "loss": 0.9733, + "step": 26048 + }, + { + "epoch": 0.61, + "grad_norm": 1.978496675692434, + "learning_rate": 6.857771779802498e-06, + "loss": 0.9274, + "step": 26049 + }, + { + "epoch": 0.61, + "grad_norm": 2.16593050800235, + "learning_rate": 6.857047396484486e-06, + "loss": 0.8668, + "step": 26050 + }, + { + "epoch": 0.61, + "grad_norm": 1.0098318538491862, + "learning_rate": 6.856323031465741e-06, + "loss": 0.8604, + "step": 26051 + }, + { + "epoch": 0.61, + "grad_norm": 3.4227216734708943, + "learning_rate": 6.855598684750475e-06, + "loss": 0.9902, + "step": 26052 + }, + { + "epoch": 0.61, + "grad_norm": 1.984273713166427, + "learning_rate": 6.854874356342911e-06, + "loss": 1.1304, + "step": 26053 + }, + { + "epoch": 0.61, + "grad_norm": 2.072133852243532, + "learning_rate": 6.854150046247263e-06, + "loss": 0.9451, + "step": 26054 + }, + { + "epoch": 0.61, + "grad_norm": 2.316241760719901, + "learning_rate": 6.853425754467753e-06, + "loss": 1.0576, + "step": 26055 + }, + { + "epoch": 0.61, + "grad_norm": 1.1649581354673098, + "learning_rate": 6.852701481008589e-06, + "loss": 0.9831, + "step": 26056 + }, + { + "epoch": 0.61, + "grad_norm": 2.572617142025581, + "learning_rate": 6.851977225873996e-06, + "loss": 0.9131, + "step": 26057 + }, + { + "epoch": 0.61, + "grad_norm": 1.9840459313164027, + "learning_rate": 6.851252989068191e-06, + "loss": 0.9743, + "step": 26058 + }, + { + "epoch": 0.61, + "grad_norm": 2.0707815309262556, + "learning_rate": 6.8505287705953835e-06, + "loss": 1.0301, + "step": 26059 + }, + { + "epoch": 0.61, + "grad_norm": 2.0788475402624056, + "learning_rate": 6.849804570459795e-06, + "loss": 0.9758, + "step": 26060 + }, + { + "epoch": 0.61, + "grad_norm": 2.1378615961843246, + "learning_rate": 6.849080388665643e-06, + "loss": 0.9921, + "step": 26061 + }, + { + "epoch": 0.61, + "grad_norm": 2.727191097027227, + "learning_rate": 6.848356225217146e-06, + "loss": 0.9786, + "step": 26062 + }, + { + "epoch": 0.61, + "grad_norm": 1.0834499909536304, + "learning_rate": 6.8476320801185116e-06, + "loss": 0.9856, + "step": 26063 + }, + { + "epoch": 0.61, + "grad_norm": 2.1263177345624134, + "learning_rate": 6.846907953373962e-06, + "loss": 0.9176, + "step": 26064 + }, + { + "epoch": 0.61, + "grad_norm": 1.9897271872716298, + "learning_rate": 6.846183844987715e-06, + "loss": 0.9842, + "step": 26065 + }, + { + "epoch": 0.61, + "grad_norm": 1.061455236848023, + "learning_rate": 6.845459754963983e-06, + "loss": 0.9286, + "step": 26066 + }, + { + "epoch": 0.61, + "grad_norm": 1.9034945116872286, + "learning_rate": 6.844735683306982e-06, + "loss": 1.0412, + "step": 26067 + }, + { + "epoch": 0.61, + "grad_norm": 2.078964210831731, + "learning_rate": 6.844011630020928e-06, + "loss": 0.8851, + "step": 26068 + }, + { + "epoch": 0.61, + "grad_norm": 1.1077608297478614, + "learning_rate": 6.843287595110041e-06, + "loss": 0.9402, + "step": 26069 + }, + { + "epoch": 0.61, + "grad_norm": 1.8096223984899464, + "learning_rate": 6.842563578578531e-06, + "loss": 1.0154, + "step": 26070 + }, + { + "epoch": 0.61, + "grad_norm": 2.1800220534622783, + "learning_rate": 6.841839580430616e-06, + "loss": 1.0283, + "step": 26071 + }, + { + "epoch": 0.61, + "grad_norm": 1.9326031617141892, + "learning_rate": 6.8411156006705116e-06, + "loss": 0.8979, + "step": 26072 + }, + { + "epoch": 0.61, + "grad_norm": 1.032252194321849, + "learning_rate": 6.8403916393024285e-06, + "loss": 0.9131, + "step": 26073 + }, + { + "epoch": 0.61, + "grad_norm": 2.0772938563656598, + "learning_rate": 6.839667696330589e-06, + "loss": 0.9671, + "step": 26074 + }, + { + "epoch": 0.61, + "grad_norm": 2.5384318091129607, + "learning_rate": 6.838943771759204e-06, + "loss": 0.953, + "step": 26075 + }, + { + "epoch": 0.61, + "grad_norm": 1.1382036699310691, + "learning_rate": 6.838219865592491e-06, + "loss": 0.9177, + "step": 26076 + }, + { + "epoch": 0.61, + "grad_norm": 1.067508395643436, + "learning_rate": 6.837495977834658e-06, + "loss": 0.8912, + "step": 26077 + }, + { + "epoch": 0.61, + "grad_norm": 1.8360550391967652, + "learning_rate": 6.836772108489928e-06, + "loss": 0.9698, + "step": 26078 + }, + { + "epoch": 0.61, + "grad_norm": 1.7161268585028118, + "learning_rate": 6.836048257562513e-06, + "loss": 1.0382, + "step": 26079 + }, + { + "epoch": 0.61, + "grad_norm": 1.7201344273373451, + "learning_rate": 6.835324425056626e-06, + "loss": 0.9589, + "step": 26080 + }, + { + "epoch": 0.61, + "grad_norm": 2.105279202457331, + "learning_rate": 6.834600610976478e-06, + "loss": 1.0894, + "step": 26081 + }, + { + "epoch": 0.61, + "grad_norm": 1.8736549144802432, + "learning_rate": 6.8338768153262925e-06, + "loss": 0.9191, + "step": 26082 + }, + { + "epoch": 0.61, + "grad_norm": 1.9237593136259328, + "learning_rate": 6.833153038110279e-06, + "loss": 1.0323, + "step": 26083 + }, + { + "epoch": 0.61, + "grad_norm": 2.279717239339265, + "learning_rate": 6.8324292793326485e-06, + "loss": 0.8893, + "step": 26084 + }, + { + "epoch": 0.61, + "grad_norm": 1.0775118359211064, + "learning_rate": 6.831705538997617e-06, + "loss": 0.9896, + "step": 26085 + }, + { + "epoch": 0.61, + "grad_norm": 2.264601459505297, + "learning_rate": 6.830981817109401e-06, + "loss": 1.2, + "step": 26086 + }, + { + "epoch": 0.61, + "grad_norm": 2.158761123980154, + "learning_rate": 6.830258113672213e-06, + "loss": 1.0398, + "step": 26087 + }, + { + "epoch": 0.61, + "grad_norm": 1.8744674851538783, + "learning_rate": 6.829534428690263e-06, + "loss": 0.9619, + "step": 26088 + }, + { + "epoch": 0.61, + "grad_norm": 2.0726350626669587, + "learning_rate": 6.8288107621677655e-06, + "loss": 1.0963, + "step": 26089 + }, + { + "epoch": 0.61, + "grad_norm": 2.2759369276716765, + "learning_rate": 6.8280871141089415e-06, + "loss": 1.0552, + "step": 26090 + }, + { + "epoch": 0.61, + "grad_norm": 1.8325258538736526, + "learning_rate": 6.8273634845179955e-06, + "loss": 0.914, + "step": 26091 + }, + { + "epoch": 0.61, + "grad_norm": 1.9527022622500896, + "learning_rate": 6.8266398733991415e-06, + "loss": 0.9729, + "step": 26092 + }, + { + "epoch": 0.61, + "grad_norm": 2.091943033060448, + "learning_rate": 6.825916280756601e-06, + "loss": 0.9046, + "step": 26093 + }, + { + "epoch": 0.61, + "grad_norm": 1.9150040633418812, + "learning_rate": 6.8251927065945755e-06, + "loss": 0.9111, + "step": 26094 + }, + { + "epoch": 0.61, + "grad_norm": 1.1440294261282233, + "learning_rate": 6.824469150917286e-06, + "loss": 0.9256, + "step": 26095 + }, + { + "epoch": 0.61, + "grad_norm": 2.056530982022028, + "learning_rate": 6.8237456137289385e-06, + "loss": 0.9689, + "step": 26096 + }, + { + "epoch": 0.61, + "grad_norm": 1.0816481775903204, + "learning_rate": 6.823022095033756e-06, + "loss": 0.9866, + "step": 26097 + }, + { + "epoch": 0.61, + "grad_norm": 2.080120800704348, + "learning_rate": 6.822298594835939e-06, + "loss": 0.9955, + "step": 26098 + }, + { + "epoch": 0.61, + "grad_norm": 1.9589108153069252, + "learning_rate": 6.821575113139707e-06, + "loss": 0.9839, + "step": 26099 + }, + { + "epoch": 0.61, + "grad_norm": 2.125096205306847, + "learning_rate": 6.820851649949274e-06, + "loss": 1.0099, + "step": 26100 + }, + { + "epoch": 0.61, + "grad_norm": 1.96983924178413, + "learning_rate": 6.820128205268845e-06, + "loss": 1.0149, + "step": 26101 + }, + { + "epoch": 0.61, + "grad_norm": 2.030206359119399, + "learning_rate": 6.8194047791026375e-06, + "loss": 1.0608, + "step": 26102 + }, + { + "epoch": 0.61, + "grad_norm": 1.8871633638114338, + "learning_rate": 6.8186813714548625e-06, + "loss": 1.0888, + "step": 26103 + }, + { + "epoch": 0.61, + "grad_norm": 2.0121881064765677, + "learning_rate": 6.817957982329735e-06, + "loss": 1.0822, + "step": 26104 + }, + { + "epoch": 0.62, + "grad_norm": 2.0002540735728886, + "learning_rate": 6.817234611731458e-06, + "loss": 1.0232, + "step": 26105 + }, + { + "epoch": 0.62, + "grad_norm": 2.2802645400280097, + "learning_rate": 6.816511259664251e-06, + "loss": 0.9656, + "step": 26106 + }, + { + "epoch": 0.62, + "grad_norm": 2.1589136870456387, + "learning_rate": 6.8157879261323265e-06, + "loss": 0.9416, + "step": 26107 + }, + { + "epoch": 0.62, + "grad_norm": 1.9382467592504091, + "learning_rate": 6.81506461113989e-06, + "loss": 1.1155, + "step": 26108 + }, + { + "epoch": 0.62, + "grad_norm": 5.438535040025392, + "learning_rate": 6.814341314691153e-06, + "loss": 1.0971, + "step": 26109 + }, + { + "epoch": 0.62, + "grad_norm": 1.9914784625916921, + "learning_rate": 6.813618036790333e-06, + "loss": 1.114, + "step": 26110 + }, + { + "epoch": 0.62, + "grad_norm": 1.9006058538218757, + "learning_rate": 6.812894777441638e-06, + "loss": 0.919, + "step": 26111 + }, + { + "epoch": 0.62, + "grad_norm": 2.829353132294952, + "learning_rate": 6.812171536649278e-06, + "loss": 0.9672, + "step": 26112 + }, + { + "epoch": 0.62, + "grad_norm": 2.201419139575127, + "learning_rate": 6.811448314417461e-06, + "loss": 1.0658, + "step": 26113 + }, + { + "epoch": 0.62, + "grad_norm": 2.022031748533906, + "learning_rate": 6.8107251107504065e-06, + "loss": 0.9539, + "step": 26114 + }, + { + "epoch": 0.62, + "grad_norm": 2.1138146508649522, + "learning_rate": 6.810001925652319e-06, + "loss": 0.9886, + "step": 26115 + }, + { + "epoch": 0.62, + "grad_norm": 1.9026741066793178, + "learning_rate": 6.809278759127409e-06, + "loss": 0.8983, + "step": 26116 + }, + { + "epoch": 0.62, + "grad_norm": 1.8481860326241946, + "learning_rate": 6.808555611179886e-06, + "loss": 1.0575, + "step": 26117 + }, + { + "epoch": 0.62, + "grad_norm": 2.0329971818392223, + "learning_rate": 6.807832481813967e-06, + "loss": 1.0695, + "step": 26118 + }, + { + "epoch": 0.62, + "grad_norm": 1.9461193613658816, + "learning_rate": 6.807109371033856e-06, + "loss": 0.8121, + "step": 26119 + }, + { + "epoch": 0.62, + "grad_norm": 1.9212947189670093, + "learning_rate": 6.806386278843764e-06, + "loss": 0.9455, + "step": 26120 + }, + { + "epoch": 0.62, + "grad_norm": 2.2496171717087186, + "learning_rate": 6.8056632052479055e-06, + "loss": 1.0787, + "step": 26121 + }, + { + "epoch": 0.62, + "grad_norm": 1.952537713685237, + "learning_rate": 6.804940150250483e-06, + "loss": 0.9868, + "step": 26122 + }, + { + "epoch": 0.62, + "grad_norm": 2.1533981751401607, + "learning_rate": 6.804217113855712e-06, + "loss": 0.9914, + "step": 26123 + }, + { + "epoch": 0.62, + "grad_norm": 2.1805510716413967, + "learning_rate": 6.803494096067802e-06, + "loss": 1.0294, + "step": 26124 + }, + { + "epoch": 0.62, + "grad_norm": 2.0206211314360116, + "learning_rate": 6.802771096890962e-06, + "loss": 0.9559, + "step": 26125 + }, + { + "epoch": 0.62, + "grad_norm": 2.039874615873117, + "learning_rate": 6.8020481163293975e-06, + "loss": 1.0969, + "step": 26126 + }, + { + "epoch": 0.62, + "grad_norm": 2.057290768044714, + "learning_rate": 6.801325154387323e-06, + "loss": 1.0604, + "step": 26127 + }, + { + "epoch": 0.62, + "grad_norm": 2.1626764556807165, + "learning_rate": 6.800602211068947e-06, + "loss": 1.0078, + "step": 26128 + }, + { + "epoch": 0.62, + "grad_norm": 1.9304857695090574, + "learning_rate": 6.7998792863784775e-06, + "loss": 0.9713, + "step": 26129 + }, + { + "epoch": 0.62, + "grad_norm": 1.8612066245026566, + "learning_rate": 6.799156380320122e-06, + "loss": 1.1069, + "step": 26130 + }, + { + "epoch": 0.62, + "grad_norm": 1.1063621435867632, + "learning_rate": 6.798433492898093e-06, + "loss": 1.0144, + "step": 26131 + }, + { + "epoch": 0.62, + "grad_norm": 2.197040282297369, + "learning_rate": 6.797710624116601e-06, + "loss": 1.0212, + "step": 26132 + }, + { + "epoch": 0.62, + "grad_norm": 2.1034209485561792, + "learning_rate": 6.796987773979846e-06, + "loss": 1.1201, + "step": 26133 + }, + { + "epoch": 0.62, + "grad_norm": 1.1199345172629886, + "learning_rate": 6.796264942492044e-06, + "loss": 0.9795, + "step": 26134 + }, + { + "epoch": 0.62, + "grad_norm": 1.0499150751539674, + "learning_rate": 6.7955421296574045e-06, + "loss": 0.948, + "step": 26135 + }, + { + "epoch": 0.62, + "grad_norm": 2.204391707581324, + "learning_rate": 6.7948193354801306e-06, + "loss": 1.132, + "step": 26136 + }, + { + "epoch": 0.62, + "grad_norm": 1.8337817219127042, + "learning_rate": 6.794096559964434e-06, + "loss": 0.9665, + "step": 26137 + }, + { + "epoch": 0.62, + "grad_norm": 1.918439243406037, + "learning_rate": 6.7933738031145205e-06, + "loss": 1.0492, + "step": 26138 + }, + { + "epoch": 0.62, + "grad_norm": 2.155577904607442, + "learning_rate": 6.792651064934602e-06, + "loss": 0.9265, + "step": 26139 + }, + { + "epoch": 0.62, + "grad_norm": 2.28424538810014, + "learning_rate": 6.791928345428884e-06, + "loss": 0.9314, + "step": 26140 + }, + { + "epoch": 0.62, + "grad_norm": 5.031695957042852, + "learning_rate": 6.791205644601573e-06, + "loss": 0.8953, + "step": 26141 + }, + { + "epoch": 0.62, + "grad_norm": 2.1411273738349457, + "learning_rate": 6.790482962456879e-06, + "loss": 0.9295, + "step": 26142 + }, + { + "epoch": 0.62, + "grad_norm": 2.073613289809166, + "learning_rate": 6.789760298999013e-06, + "loss": 0.9046, + "step": 26143 + }, + { + "epoch": 0.62, + "grad_norm": 2.2879101940747417, + "learning_rate": 6.789037654232177e-06, + "loss": 0.9908, + "step": 26144 + }, + { + "epoch": 0.62, + "grad_norm": 2.0917920484095367, + "learning_rate": 6.788315028160577e-06, + "loss": 0.9499, + "step": 26145 + }, + { + "epoch": 0.62, + "grad_norm": 1.1622641983201036, + "learning_rate": 6.787592420788431e-06, + "loss": 0.9596, + "step": 26146 + }, + { + "epoch": 0.62, + "grad_norm": 1.9796685815129624, + "learning_rate": 6.786869832119933e-06, + "loss": 0.9045, + "step": 26147 + }, + { + "epoch": 0.62, + "grad_norm": 1.933510648803264, + "learning_rate": 6.786147262159297e-06, + "loss": 1.0374, + "step": 26148 + }, + { + "epoch": 0.62, + "grad_norm": 1.0568722334022027, + "learning_rate": 6.7854247109107285e-06, + "loss": 0.8778, + "step": 26149 + }, + { + "epoch": 0.62, + "grad_norm": 2.0336537142989566, + "learning_rate": 6.784702178378437e-06, + "loss": 0.9887, + "step": 26150 + }, + { + "epoch": 0.62, + "grad_norm": 2.0547850962085388, + "learning_rate": 6.783979664566627e-06, + "loss": 0.9985, + "step": 26151 + }, + { + "epoch": 0.62, + "grad_norm": 2.439044701974769, + "learning_rate": 6.783257169479504e-06, + "loss": 0.9305, + "step": 26152 + }, + { + "epoch": 0.62, + "grad_norm": 2.078047635092166, + "learning_rate": 6.782534693121281e-06, + "loss": 1.0284, + "step": 26153 + }, + { + "epoch": 0.62, + "grad_norm": 1.9978405101495396, + "learning_rate": 6.781812235496153e-06, + "loss": 1.0662, + "step": 26154 + }, + { + "epoch": 0.62, + "grad_norm": 1.97204835750482, + "learning_rate": 6.7810897966083365e-06, + "loss": 1.1189, + "step": 26155 + }, + { + "epoch": 0.62, + "grad_norm": 2.0762032985776395, + "learning_rate": 6.780367376462033e-06, + "loss": 0.9923, + "step": 26156 + }, + { + "epoch": 0.62, + "grad_norm": 2.036642328383474, + "learning_rate": 6.779644975061454e-06, + "loss": 1.0477, + "step": 26157 + }, + { + "epoch": 0.62, + "grad_norm": 2.5582952179629657, + "learning_rate": 6.778922592410796e-06, + "loss": 1.0022, + "step": 26158 + }, + { + "epoch": 0.62, + "grad_norm": 2.0417462459201845, + "learning_rate": 6.778200228514274e-06, + "loss": 0.9925, + "step": 26159 + }, + { + "epoch": 0.62, + "grad_norm": 2.035754525426218, + "learning_rate": 6.777477883376089e-06, + "loss": 0.8632, + "step": 26160 + }, + { + "epoch": 0.62, + "grad_norm": 1.9608813814239368, + "learning_rate": 6.776755557000449e-06, + "loss": 1.0502, + "step": 26161 + }, + { + "epoch": 0.62, + "grad_norm": 1.9848133090251783, + "learning_rate": 6.776033249391555e-06, + "loss": 0.9385, + "step": 26162 + }, + { + "epoch": 0.62, + "grad_norm": 1.998966223909432, + "learning_rate": 6.775310960553618e-06, + "loss": 1.0042, + "step": 26163 + }, + { + "epoch": 0.62, + "grad_norm": 1.9164187768386003, + "learning_rate": 6.774588690490844e-06, + "loss": 0.9727, + "step": 26164 + }, + { + "epoch": 0.62, + "grad_norm": 1.8990744292014168, + "learning_rate": 6.7738664392074345e-06, + "loss": 0.9707, + "step": 26165 + }, + { + "epoch": 0.62, + "grad_norm": 1.9259584509297087, + "learning_rate": 6.773144206707594e-06, + "loss": 1.0568, + "step": 26166 + }, + { + "epoch": 0.62, + "grad_norm": 1.0962879032235582, + "learning_rate": 6.772421992995533e-06, + "loss": 0.8571, + "step": 26167 + }, + { + "epoch": 0.62, + "grad_norm": 1.8964326407774852, + "learning_rate": 6.7716997980754505e-06, + "loss": 1.0239, + "step": 26168 + }, + { + "epoch": 0.62, + "grad_norm": 2.1148001266038126, + "learning_rate": 6.770977621951554e-06, + "loss": 1.0014, + "step": 26169 + }, + { + "epoch": 0.62, + "grad_norm": 2.222544039800083, + "learning_rate": 6.770255464628047e-06, + "loss": 1.0067, + "step": 26170 + }, + { + "epoch": 0.62, + "grad_norm": 1.0575917480637436, + "learning_rate": 6.769533326109139e-06, + "loss": 1.0413, + "step": 26171 + }, + { + "epoch": 0.62, + "grad_norm": 1.1091586027007045, + "learning_rate": 6.768811206399028e-06, + "loss": 0.9385, + "step": 26172 + }, + { + "epoch": 0.62, + "grad_norm": 19.86071155239374, + "learning_rate": 6.7680891055019215e-06, + "loss": 0.9522, + "step": 26173 + }, + { + "epoch": 0.62, + "grad_norm": 1.955460339192634, + "learning_rate": 6.767367023422025e-06, + "loss": 0.9527, + "step": 26174 + }, + { + "epoch": 0.62, + "grad_norm": 2.0903566464171317, + "learning_rate": 6.766644960163538e-06, + "loss": 0.9888, + "step": 26175 + }, + { + "epoch": 0.62, + "grad_norm": 1.8546387309187957, + "learning_rate": 6.765922915730669e-06, + "loss": 0.9423, + "step": 26176 + }, + { + "epoch": 0.62, + "grad_norm": 1.8952802838923384, + "learning_rate": 6.765200890127619e-06, + "loss": 0.9183, + "step": 26177 + }, + { + "epoch": 0.62, + "grad_norm": 1.9876215885940471, + "learning_rate": 6.7644788833586e-06, + "loss": 0.9719, + "step": 26178 + }, + { + "epoch": 0.62, + "grad_norm": 1.8690156258318509, + "learning_rate": 6.763756895427802e-06, + "loss": 0.9313, + "step": 26179 + }, + { + "epoch": 0.62, + "grad_norm": 2.337772290502343, + "learning_rate": 6.763034926339438e-06, + "loss": 1.0901, + "step": 26180 + }, + { + "epoch": 0.62, + "grad_norm": 2.219132993002937, + "learning_rate": 6.762312976097714e-06, + "loss": 1.0569, + "step": 26181 + }, + { + "epoch": 0.62, + "grad_norm": 1.8080289893029466, + "learning_rate": 6.7615910447068235e-06, + "loss": 0.9729, + "step": 26182 + }, + { + "epoch": 0.62, + "grad_norm": 2.2282328074101527, + "learning_rate": 6.7608691321709755e-06, + "loss": 1.0133, + "step": 26183 + }, + { + "epoch": 0.62, + "grad_norm": 2.238421751525724, + "learning_rate": 6.760147238494374e-06, + "loss": 0.9699, + "step": 26184 + }, + { + "epoch": 0.62, + "grad_norm": 1.8389109466013693, + "learning_rate": 6.759425363681223e-06, + "loss": 0.9576, + "step": 26185 + }, + { + "epoch": 0.62, + "grad_norm": 2.446166172338695, + "learning_rate": 6.758703507735721e-06, + "loss": 1.0795, + "step": 26186 + }, + { + "epoch": 0.62, + "grad_norm": 2.1572278355404304, + "learning_rate": 6.7579816706620724e-06, + "loss": 0.8774, + "step": 26187 + }, + { + "epoch": 0.62, + "grad_norm": 1.1158114608715366, + "learning_rate": 6.757259852464484e-06, + "loss": 0.9312, + "step": 26188 + }, + { + "epoch": 0.62, + "grad_norm": 5.693544170078042, + "learning_rate": 6.7565380531471545e-06, + "loss": 0.9649, + "step": 26189 + }, + { + "epoch": 0.62, + "grad_norm": 1.9324802297867276, + "learning_rate": 6.755816272714284e-06, + "loss": 1.0261, + "step": 26190 + }, + { + "epoch": 0.62, + "grad_norm": 1.7612784950623819, + "learning_rate": 6.755094511170082e-06, + "loss": 1.1039, + "step": 26191 + }, + { + "epoch": 0.62, + "grad_norm": 2.0370202595922016, + "learning_rate": 6.754372768518746e-06, + "loss": 1.0736, + "step": 26192 + }, + { + "epoch": 0.62, + "grad_norm": 2.0197020269188664, + "learning_rate": 6.7536510447644795e-06, + "loss": 0.9468, + "step": 26193 + }, + { + "epoch": 0.62, + "grad_norm": 2.1691014486138105, + "learning_rate": 6.752929339911483e-06, + "loss": 1.0883, + "step": 26194 + }, + { + "epoch": 0.62, + "grad_norm": 2.0812430069449004, + "learning_rate": 6.752207653963965e-06, + "loss": 0.8504, + "step": 26195 + }, + { + "epoch": 0.62, + "grad_norm": 2.2116279692818157, + "learning_rate": 6.7514859869261164e-06, + "loss": 1.0385, + "step": 26196 + }, + { + "epoch": 0.62, + "grad_norm": 2.2653873114417427, + "learning_rate": 6.750764338802147e-06, + "loss": 0.9223, + "step": 26197 + }, + { + "epoch": 0.62, + "grad_norm": 1.8280986579322727, + "learning_rate": 6.750042709596254e-06, + "loss": 1.0173, + "step": 26198 + }, + { + "epoch": 0.62, + "grad_norm": 2.6253298799825835, + "learning_rate": 6.749321099312644e-06, + "loss": 0.9864, + "step": 26199 + }, + { + "epoch": 0.62, + "grad_norm": 2.4886143219961947, + "learning_rate": 6.748599507955516e-06, + "loss": 0.9827, + "step": 26200 + }, + { + "epoch": 0.62, + "grad_norm": 2.030394583666157, + "learning_rate": 6.74787793552907e-06, + "loss": 1.0092, + "step": 26201 + }, + { + "epoch": 0.62, + "grad_norm": 2.103514693462764, + "learning_rate": 6.747156382037511e-06, + "loss": 1.0391, + "step": 26202 + }, + { + "epoch": 0.62, + "grad_norm": 2.603392796738327, + "learning_rate": 6.7464348474850324e-06, + "loss": 0.9713, + "step": 26203 + }, + { + "epoch": 0.62, + "grad_norm": 2.0151365207129444, + "learning_rate": 6.745713331875843e-06, + "loss": 1.0388, + "step": 26204 + }, + { + "epoch": 0.62, + "grad_norm": 2.1158969190235983, + "learning_rate": 6.74499183521414e-06, + "loss": 0.9943, + "step": 26205 + }, + { + "epoch": 0.62, + "grad_norm": 2.216593254921498, + "learning_rate": 6.744270357504128e-06, + "loss": 1.0741, + "step": 26206 + }, + { + "epoch": 0.62, + "grad_norm": 2.08125759978865, + "learning_rate": 6.743548898749999e-06, + "loss": 0.9569, + "step": 26207 + }, + { + "epoch": 0.62, + "grad_norm": 1.9119552406740374, + "learning_rate": 6.7428274589559626e-06, + "loss": 1.0508, + "step": 26208 + }, + { + "epoch": 0.62, + "grad_norm": 2.8736930655502726, + "learning_rate": 6.742106038126218e-06, + "loss": 0.9604, + "step": 26209 + }, + { + "epoch": 0.62, + "grad_norm": 2.0008068184929972, + "learning_rate": 6.741384636264961e-06, + "loss": 1.0952, + "step": 26210 + }, + { + "epoch": 0.62, + "grad_norm": 1.9149059987295198, + "learning_rate": 6.740663253376393e-06, + "loss": 0.9853, + "step": 26211 + }, + { + "epoch": 0.62, + "grad_norm": 2.2975355751584403, + "learning_rate": 6.7399418894647164e-06, + "loss": 0.8849, + "step": 26212 + }, + { + "epoch": 0.62, + "grad_norm": 2.29607464255342, + "learning_rate": 6.739220544534133e-06, + "loss": 1.0586, + "step": 26213 + }, + { + "epoch": 0.62, + "grad_norm": 2.123048134428772, + "learning_rate": 6.738499218588839e-06, + "loss": 1.0527, + "step": 26214 + }, + { + "epoch": 0.62, + "grad_norm": 2.342094070651211, + "learning_rate": 6.737777911633031e-06, + "loss": 1.0462, + "step": 26215 + }, + { + "epoch": 0.62, + "grad_norm": 2.1866759016643793, + "learning_rate": 6.737056623670918e-06, + "loss": 0.9388, + "step": 26216 + }, + { + "epoch": 0.62, + "grad_norm": 2.7011189014235644, + "learning_rate": 6.736335354706691e-06, + "loss": 0.962, + "step": 26217 + }, + { + "epoch": 0.62, + "grad_norm": 2.174841642494018, + "learning_rate": 6.7356141047445545e-06, + "loss": 1.0887, + "step": 26218 + }, + { + "epoch": 0.62, + "grad_norm": 2.0295252869931635, + "learning_rate": 6.734892873788703e-06, + "loss": 0.8848, + "step": 26219 + }, + { + "epoch": 0.62, + "grad_norm": 2.075782986174673, + "learning_rate": 6.7341716618433435e-06, + "loss": 0.9895, + "step": 26220 + }, + { + "epoch": 0.62, + "grad_norm": 1.9112511310320075, + "learning_rate": 6.733450468912669e-06, + "loss": 1.0703, + "step": 26221 + }, + { + "epoch": 0.62, + "grad_norm": 1.0160504362477816, + "learning_rate": 6.732729295000879e-06, + "loss": 0.9268, + "step": 26222 + }, + { + "epoch": 0.62, + "grad_norm": 2.103537967380819, + "learning_rate": 6.732008140112176e-06, + "loss": 0.9405, + "step": 26223 + }, + { + "epoch": 0.62, + "grad_norm": 1.9347904499863495, + "learning_rate": 6.731287004250752e-06, + "loss": 1.0406, + "step": 26224 + }, + { + "epoch": 0.62, + "grad_norm": 2.1662176263099044, + "learning_rate": 6.730565887420813e-06, + "loss": 0.9995, + "step": 26225 + }, + { + "epoch": 0.62, + "grad_norm": 2.2628301548199947, + "learning_rate": 6.729844789626552e-06, + "loss": 0.9276, + "step": 26226 + }, + { + "epoch": 0.62, + "grad_norm": 2.9669525222275217, + "learning_rate": 6.729123710872175e-06, + "loss": 1.0796, + "step": 26227 + }, + { + "epoch": 0.62, + "grad_norm": 2.594929057666486, + "learning_rate": 6.728402651161869e-06, + "loss": 1.1834, + "step": 26228 + }, + { + "epoch": 0.62, + "grad_norm": 1.8982544217606891, + "learning_rate": 6.727681610499842e-06, + "loss": 0.9207, + "step": 26229 + }, + { + "epoch": 0.62, + "grad_norm": 2.1777731669075755, + "learning_rate": 6.72696058889029e-06, + "loss": 0.9643, + "step": 26230 + }, + { + "epoch": 0.62, + "grad_norm": 2.1277921183231037, + "learning_rate": 6.726239586337408e-06, + "loss": 0.9576, + "step": 26231 + }, + { + "epoch": 0.62, + "grad_norm": 3.0326043755123155, + "learning_rate": 6.725518602845394e-06, + "loss": 0.9609, + "step": 26232 + }, + { + "epoch": 0.62, + "grad_norm": 2.0729187782999086, + "learning_rate": 6.7247976384184475e-06, + "loss": 0.9848, + "step": 26233 + }, + { + "epoch": 0.62, + "grad_norm": 2.3567240294189675, + "learning_rate": 6.72407669306077e-06, + "loss": 1.0384, + "step": 26234 + }, + { + "epoch": 0.62, + "grad_norm": 2.3137070903073718, + "learning_rate": 6.723355766776553e-06, + "loss": 1.0048, + "step": 26235 + }, + { + "epoch": 0.62, + "grad_norm": 2.032498148827352, + "learning_rate": 6.722634859569992e-06, + "loss": 0.9877, + "step": 26236 + }, + { + "epoch": 0.62, + "grad_norm": 1.8739163277810176, + "learning_rate": 6.721913971445295e-06, + "loss": 0.9036, + "step": 26237 + }, + { + "epoch": 0.62, + "grad_norm": 2.4010493500885093, + "learning_rate": 6.721193102406649e-06, + "loss": 1.0507, + "step": 26238 + }, + { + "epoch": 0.62, + "grad_norm": 2.204530546271686, + "learning_rate": 6.720472252458254e-06, + "loss": 1.0801, + "step": 26239 + }, + { + "epoch": 0.62, + "grad_norm": 1.9465757384725042, + "learning_rate": 6.719751421604309e-06, + "loss": 0.9191, + "step": 26240 + }, + { + "epoch": 0.62, + "grad_norm": 2.0210549430546187, + "learning_rate": 6.719030609849012e-06, + "loss": 0.9163, + "step": 26241 + }, + { + "epoch": 0.62, + "grad_norm": 2.4284109929116857, + "learning_rate": 6.718309817196556e-06, + "loss": 1.0504, + "step": 26242 + }, + { + "epoch": 0.62, + "grad_norm": 2.158628673047473, + "learning_rate": 6.717589043651136e-06, + "loss": 1.0029, + "step": 26243 + }, + { + "epoch": 0.62, + "grad_norm": 1.1451077083642298, + "learning_rate": 6.716868289216959e-06, + "loss": 0.8883, + "step": 26244 + }, + { + "epoch": 0.62, + "grad_norm": 1.938070032169238, + "learning_rate": 6.716147553898206e-06, + "loss": 1.0515, + "step": 26245 + }, + { + "epoch": 0.62, + "grad_norm": 2.0506695378621713, + "learning_rate": 6.715426837699086e-06, + "loss": 1.1188, + "step": 26246 + }, + { + "epoch": 0.62, + "grad_norm": 1.0874862754129753, + "learning_rate": 6.714706140623789e-06, + "loss": 0.9595, + "step": 26247 + }, + { + "epoch": 0.62, + "grad_norm": 2.1449375269337336, + "learning_rate": 6.7139854626765155e-06, + "loss": 1.089, + "step": 26248 + }, + { + "epoch": 0.62, + "grad_norm": 1.9089214746688592, + "learning_rate": 6.713264803861459e-06, + "loss": 1.045, + "step": 26249 + }, + { + "epoch": 0.62, + "grad_norm": 2.6386456707091615, + "learning_rate": 6.712544164182813e-06, + "loss": 0.8753, + "step": 26250 + }, + { + "epoch": 0.62, + "grad_norm": 2.166426814055111, + "learning_rate": 6.711823543644779e-06, + "loss": 0.9941, + "step": 26251 + }, + { + "epoch": 0.62, + "grad_norm": 2.1152773125835207, + "learning_rate": 6.711102942251544e-06, + "loss": 0.92, + "step": 26252 + }, + { + "epoch": 0.62, + "grad_norm": 1.0961699397655094, + "learning_rate": 6.710382360007313e-06, + "loss": 1.0121, + "step": 26253 + }, + { + "epoch": 0.62, + "grad_norm": 2.064140344343657, + "learning_rate": 6.7096617969162756e-06, + "loss": 0.9982, + "step": 26254 + }, + { + "epoch": 0.62, + "grad_norm": 1.8796384325583488, + "learning_rate": 6.708941252982632e-06, + "loss": 0.9928, + "step": 26255 + }, + { + "epoch": 0.62, + "grad_norm": 2.0587008964569913, + "learning_rate": 6.70822072821057e-06, + "loss": 0.9977, + "step": 26256 + }, + { + "epoch": 0.62, + "grad_norm": 2.04141421384477, + "learning_rate": 6.707500222604291e-06, + "loss": 1.0046, + "step": 26257 + }, + { + "epoch": 0.62, + "grad_norm": 2.050292095362202, + "learning_rate": 6.70677973616799e-06, + "loss": 0.9968, + "step": 26258 + }, + { + "epoch": 0.62, + "grad_norm": 1.884704502626666, + "learning_rate": 6.706059268905858e-06, + "loss": 0.9015, + "step": 26259 + }, + { + "epoch": 0.62, + "grad_norm": 1.831705322254846, + "learning_rate": 6.705338820822089e-06, + "loss": 0.9437, + "step": 26260 + }, + { + "epoch": 0.62, + "grad_norm": 1.9808709948884788, + "learning_rate": 6.704618391920884e-06, + "loss": 0.965, + "step": 26261 + }, + { + "epoch": 0.62, + "grad_norm": 1.966156722058635, + "learning_rate": 6.7038979822064355e-06, + "loss": 0.9707, + "step": 26262 + }, + { + "epoch": 0.62, + "grad_norm": 1.0341878372633124, + "learning_rate": 6.703177591682935e-06, + "loss": 0.9054, + "step": 26263 + }, + { + "epoch": 0.62, + "grad_norm": 1.057048221442727, + "learning_rate": 6.702457220354575e-06, + "loss": 0.939, + "step": 26264 + }, + { + "epoch": 0.62, + "grad_norm": 1.9013853769627473, + "learning_rate": 6.701736868225556e-06, + "loss": 0.9931, + "step": 26265 + }, + { + "epoch": 0.62, + "grad_norm": 2.0758392219668687, + "learning_rate": 6.701016535300071e-06, + "loss": 0.9781, + "step": 26266 + }, + { + "epoch": 0.62, + "grad_norm": 2.1682862514330754, + "learning_rate": 6.7002962215823095e-06, + "loss": 1.0406, + "step": 26267 + }, + { + "epoch": 0.62, + "grad_norm": 1.0876269153585614, + "learning_rate": 6.699575927076467e-06, + "loss": 1.0163, + "step": 26268 + }, + { + "epoch": 0.62, + "grad_norm": 5.2095471836127345, + "learning_rate": 6.698855651786742e-06, + "loss": 1.0254, + "step": 26269 + }, + { + "epoch": 0.62, + "grad_norm": 1.0821423780143022, + "learning_rate": 6.698135395717323e-06, + "loss": 0.9141, + "step": 26270 + }, + { + "epoch": 0.62, + "grad_norm": 1.951998262248992, + "learning_rate": 6.697415158872405e-06, + "loss": 1.0312, + "step": 26271 + }, + { + "epoch": 0.62, + "grad_norm": 2.143896592772452, + "learning_rate": 6.696694941256179e-06, + "loss": 0.9835, + "step": 26272 + }, + { + "epoch": 0.62, + "grad_norm": 2.3265216435914677, + "learning_rate": 6.695974742872844e-06, + "loss": 0.887, + "step": 26273 + }, + { + "epoch": 0.62, + "grad_norm": 2.527107962372583, + "learning_rate": 6.69525456372659e-06, + "loss": 1.116, + "step": 26274 + }, + { + "epoch": 0.62, + "grad_norm": 2.0118298234876852, + "learning_rate": 6.694534403821606e-06, + "loss": 0.9476, + "step": 26275 + }, + { + "epoch": 0.62, + "grad_norm": 2.0181778863831936, + "learning_rate": 6.6938142631620975e-06, + "loss": 0.996, + "step": 26276 + }, + { + "epoch": 0.62, + "grad_norm": 2.1536971806932397, + "learning_rate": 6.693094141752241e-06, + "loss": 0.9993, + "step": 26277 + }, + { + "epoch": 0.62, + "grad_norm": 1.8976316078934143, + "learning_rate": 6.692374039596241e-06, + "loss": 0.9169, + "step": 26278 + }, + { + "epoch": 0.62, + "grad_norm": 1.8220899307217302, + "learning_rate": 6.691653956698283e-06, + "loss": 1.1114, + "step": 26279 + }, + { + "epoch": 0.62, + "grad_norm": 2.943265120472267, + "learning_rate": 6.69093389306257e-06, + "loss": 1.0362, + "step": 26280 + }, + { + "epoch": 0.62, + "grad_norm": 1.9907607276547115, + "learning_rate": 6.690213848693282e-06, + "loss": 1.0525, + "step": 26281 + }, + { + "epoch": 0.62, + "grad_norm": 2.2240365550392287, + "learning_rate": 6.689493823594618e-06, + "loss": 1.0294, + "step": 26282 + }, + { + "epoch": 0.62, + "grad_norm": 2.108764884220005, + "learning_rate": 6.688773817770772e-06, + "loss": 0.9249, + "step": 26283 + }, + { + "epoch": 0.62, + "grad_norm": 2.092587478321984, + "learning_rate": 6.68805383122593e-06, + "loss": 0.9735, + "step": 26284 + }, + { + "epoch": 0.62, + "grad_norm": 2.0058237670896992, + "learning_rate": 6.687333863964285e-06, + "loss": 1.038, + "step": 26285 + }, + { + "epoch": 0.62, + "grad_norm": 1.9309057902997384, + "learning_rate": 6.686613915990034e-06, + "loss": 1.024, + "step": 26286 + }, + { + "epoch": 0.62, + "grad_norm": 2.4509844128860454, + "learning_rate": 6.685893987307368e-06, + "loss": 0.9966, + "step": 26287 + }, + { + "epoch": 0.62, + "grad_norm": 1.9196563386215995, + "learning_rate": 6.685174077920471e-06, + "loss": 1.1364, + "step": 26288 + }, + { + "epoch": 0.62, + "grad_norm": 2.0640906123460327, + "learning_rate": 6.684454187833543e-06, + "loss": 1.1362, + "step": 26289 + }, + { + "epoch": 0.62, + "grad_norm": 2.028081362373136, + "learning_rate": 6.683734317050774e-06, + "loss": 0.9388, + "step": 26290 + }, + { + "epoch": 0.62, + "grad_norm": 2.182217237985571, + "learning_rate": 6.683014465576353e-06, + "loss": 0.9249, + "step": 26291 + }, + { + "epoch": 0.62, + "grad_norm": 1.9345214372774155, + "learning_rate": 6.6822946334144675e-06, + "loss": 1.0565, + "step": 26292 + }, + { + "epoch": 0.62, + "grad_norm": 2.1524126163861053, + "learning_rate": 6.681574820569317e-06, + "loss": 1.072, + "step": 26293 + }, + { + "epoch": 0.62, + "grad_norm": 1.986929529569154, + "learning_rate": 6.68085502704509e-06, + "loss": 0.9204, + "step": 26294 + }, + { + "epoch": 0.62, + "grad_norm": 2.0179753089716783, + "learning_rate": 6.680135252845974e-06, + "loss": 1.0374, + "step": 26295 + }, + { + "epoch": 0.62, + "grad_norm": 2.1652410236207014, + "learning_rate": 6.6794154979761595e-06, + "loss": 0.9728, + "step": 26296 + }, + { + "epoch": 0.62, + "grad_norm": 2.037241622613207, + "learning_rate": 6.678695762439843e-06, + "loss": 1.0145, + "step": 26297 + }, + { + "epoch": 0.62, + "grad_norm": 2.6390130780160588, + "learning_rate": 6.677976046241209e-06, + "loss": 0.9764, + "step": 26298 + }, + { + "epoch": 0.62, + "grad_norm": 3.5797571861350734, + "learning_rate": 6.677256349384451e-06, + "loss": 0.962, + "step": 26299 + }, + { + "epoch": 0.62, + "grad_norm": 1.948835694108533, + "learning_rate": 6.676536671873755e-06, + "loss": 0.9258, + "step": 26300 + }, + { + "epoch": 0.62, + "grad_norm": 2.9268691368238273, + "learning_rate": 6.675817013713321e-06, + "loss": 0.9123, + "step": 26301 + }, + { + "epoch": 0.62, + "grad_norm": 2.7730547047269556, + "learning_rate": 6.675097374907329e-06, + "loss": 0.9633, + "step": 26302 + }, + { + "epoch": 0.62, + "grad_norm": 2.2400098272812783, + "learning_rate": 6.6743777554599735e-06, + "loss": 0.9908, + "step": 26303 + }, + { + "epoch": 0.62, + "grad_norm": 2.0060769011222397, + "learning_rate": 6.673658155375446e-06, + "loss": 1.0719, + "step": 26304 + }, + { + "epoch": 0.62, + "grad_norm": 1.1522504795038522, + "learning_rate": 6.672938574657928e-06, + "loss": 0.9705, + "step": 26305 + }, + { + "epoch": 0.62, + "grad_norm": 2.4436173390517637, + "learning_rate": 6.672219013311618e-06, + "loss": 0.8332, + "step": 26306 + }, + { + "epoch": 0.62, + "grad_norm": 2.2184590383206486, + "learning_rate": 6.671499471340704e-06, + "loss": 1.0487, + "step": 26307 + }, + { + "epoch": 0.62, + "grad_norm": 2.1638291274858057, + "learning_rate": 6.670779948749373e-06, + "loss": 1.0337, + "step": 26308 + }, + { + "epoch": 0.62, + "grad_norm": 2.057143305199096, + "learning_rate": 6.6700604455418125e-06, + "loss": 1.0061, + "step": 26309 + }, + { + "epoch": 0.62, + "grad_norm": 1.0507254903695797, + "learning_rate": 6.669340961722217e-06, + "loss": 1.0387, + "step": 26310 + }, + { + "epoch": 0.62, + "grad_norm": 2.0281071282836085, + "learning_rate": 6.668621497294774e-06, + "loss": 0.9809, + "step": 26311 + }, + { + "epoch": 0.62, + "grad_norm": 4.225185774319946, + "learning_rate": 6.667902052263671e-06, + "loss": 0.8702, + "step": 26312 + }, + { + "epoch": 0.62, + "grad_norm": 1.8041003245942802, + "learning_rate": 6.667182626633093e-06, + "loss": 0.8461, + "step": 26313 + }, + { + "epoch": 0.62, + "grad_norm": 1.8060326926552626, + "learning_rate": 6.666463220407235e-06, + "loss": 1.0211, + "step": 26314 + }, + { + "epoch": 0.62, + "grad_norm": 1.0782739690342524, + "learning_rate": 6.665743833590287e-06, + "loss": 0.9677, + "step": 26315 + }, + { + "epoch": 0.62, + "grad_norm": 2.109348400928426, + "learning_rate": 6.665024466186431e-06, + "loss": 1.0788, + "step": 26316 + }, + { + "epoch": 0.62, + "grad_norm": 1.962760212821987, + "learning_rate": 6.664305118199857e-06, + "loss": 1.0135, + "step": 26317 + }, + { + "epoch": 0.62, + "grad_norm": 1.9776346977091321, + "learning_rate": 6.663585789634758e-06, + "loss": 0.9897, + "step": 26318 + }, + { + "epoch": 0.62, + "grad_norm": 2.2004992427615546, + "learning_rate": 6.662866480495317e-06, + "loss": 0.992, + "step": 26319 + }, + { + "epoch": 0.62, + "grad_norm": 2.0324420865080555, + "learning_rate": 6.6621471907857234e-06, + "loss": 1.0017, + "step": 26320 + }, + { + "epoch": 0.62, + "grad_norm": 1.9989519516463006, + "learning_rate": 6.661427920510164e-06, + "loss": 0.9341, + "step": 26321 + }, + { + "epoch": 0.62, + "grad_norm": 1.8396965875864064, + "learning_rate": 6.660708669672833e-06, + "loss": 0.9103, + "step": 26322 + }, + { + "epoch": 0.62, + "grad_norm": 1.9981897262624295, + "learning_rate": 6.6599894382779104e-06, + "loss": 1.0059, + "step": 26323 + }, + { + "epoch": 0.62, + "grad_norm": 2.3498977298292574, + "learning_rate": 6.659270226329586e-06, + "loss": 1.127, + "step": 26324 + }, + { + "epoch": 0.62, + "grad_norm": 1.925690788215172, + "learning_rate": 6.658551033832052e-06, + "loss": 0.9141, + "step": 26325 + }, + { + "epoch": 0.62, + "grad_norm": 3.0454476034477733, + "learning_rate": 6.657831860789488e-06, + "loss": 0.9383, + "step": 26326 + }, + { + "epoch": 0.62, + "grad_norm": 2.1035983431091663, + "learning_rate": 6.657112707206086e-06, + "loss": 1.061, + "step": 26327 + }, + { + "epoch": 0.62, + "grad_norm": 2.194519876674366, + "learning_rate": 6.656393573086029e-06, + "loss": 1.0192, + "step": 26328 + }, + { + "epoch": 0.62, + "grad_norm": 1.9281721261842497, + "learning_rate": 6.655674458433514e-06, + "loss": 0.984, + "step": 26329 + }, + { + "epoch": 0.62, + "grad_norm": 2.0075885398710533, + "learning_rate": 6.6549553632527154e-06, + "loss": 1.0757, + "step": 26330 + }, + { + "epoch": 0.62, + "grad_norm": 2.0586987358716637, + "learning_rate": 6.6542362875478285e-06, + "loss": 0.9871, + "step": 26331 + }, + { + "epoch": 0.62, + "grad_norm": 2.681144261213384, + "learning_rate": 6.6535172313230375e-06, + "loss": 0.9205, + "step": 26332 + }, + { + "epoch": 0.62, + "grad_norm": 2.8974172770801356, + "learning_rate": 6.652798194582528e-06, + "loss": 0.9421, + "step": 26333 + }, + { + "epoch": 0.62, + "grad_norm": 2.075323210870123, + "learning_rate": 6.652079177330484e-06, + "loss": 0.9359, + "step": 26334 + }, + { + "epoch": 0.62, + "grad_norm": 2.5598102000159795, + "learning_rate": 6.6513601795710984e-06, + "loss": 1.0964, + "step": 26335 + }, + { + "epoch": 0.62, + "grad_norm": 2.2723657139727367, + "learning_rate": 6.650641201308554e-06, + "loss": 0.8912, + "step": 26336 + }, + { + "epoch": 0.62, + "grad_norm": 1.9011841449788363, + "learning_rate": 6.649922242547033e-06, + "loss": 0.9583, + "step": 26337 + }, + { + "epoch": 0.62, + "grad_norm": 2.0777206868440494, + "learning_rate": 6.64920330329073e-06, + "loss": 0.9342, + "step": 26338 + }, + { + "epoch": 0.62, + "grad_norm": 2.117127377634723, + "learning_rate": 6.648484383543826e-06, + "loss": 0.9084, + "step": 26339 + }, + { + "epoch": 0.62, + "grad_norm": 2.0191459360816655, + "learning_rate": 6.647765483310506e-06, + "loss": 0.8023, + "step": 26340 + }, + { + "epoch": 0.62, + "grad_norm": 1.9656480519706714, + "learning_rate": 6.647046602594953e-06, + "loss": 0.9246, + "step": 26341 + }, + { + "epoch": 0.62, + "grad_norm": 2.1985886055954147, + "learning_rate": 6.64632774140136e-06, + "loss": 0.9316, + "step": 26342 + }, + { + "epoch": 0.62, + "grad_norm": 1.8993217900008792, + "learning_rate": 6.645608899733911e-06, + "loss": 1.0826, + "step": 26343 + }, + { + "epoch": 0.62, + "grad_norm": 1.9515856994573015, + "learning_rate": 6.644890077596785e-06, + "loss": 1.0146, + "step": 26344 + }, + { + "epoch": 0.62, + "grad_norm": 2.1769216878181052, + "learning_rate": 6.64417127499417e-06, + "loss": 0.8816, + "step": 26345 + }, + { + "epoch": 0.62, + "grad_norm": 2.13388478268464, + "learning_rate": 6.643452491930257e-06, + "loss": 0.9307, + "step": 26346 + }, + { + "epoch": 0.62, + "grad_norm": 1.856521058079952, + "learning_rate": 6.642733728409223e-06, + "loss": 0.9494, + "step": 26347 + }, + { + "epoch": 0.62, + "grad_norm": 2.0218915247122133, + "learning_rate": 6.642014984435257e-06, + "loss": 1.0383, + "step": 26348 + }, + { + "epoch": 0.62, + "grad_norm": 2.09395515725389, + "learning_rate": 6.641296260012541e-06, + "loss": 1.038, + "step": 26349 + }, + { + "epoch": 0.62, + "grad_norm": 2.15025569815559, + "learning_rate": 6.640577555145266e-06, + "loss": 0.9242, + "step": 26350 + }, + { + "epoch": 0.62, + "grad_norm": 2.1667491130557592, + "learning_rate": 6.639858869837608e-06, + "loss": 0.9578, + "step": 26351 + }, + { + "epoch": 0.62, + "grad_norm": 2.1008009939047945, + "learning_rate": 6.639140204093757e-06, + "loss": 0.9116, + "step": 26352 + }, + { + "epoch": 0.62, + "grad_norm": 2.198529665814758, + "learning_rate": 6.638421557917898e-06, + "loss": 0.9707, + "step": 26353 + }, + { + "epoch": 0.62, + "grad_norm": 2.052948797834001, + "learning_rate": 6.637702931314208e-06, + "loss": 1.0554, + "step": 26354 + }, + { + "epoch": 0.62, + "grad_norm": 1.8466546209763335, + "learning_rate": 6.6369843242868795e-06, + "loss": 0.9159, + "step": 26355 + }, + { + "epoch": 0.62, + "grad_norm": 2.0672228347641055, + "learning_rate": 6.63626573684009e-06, + "loss": 0.9828, + "step": 26356 + }, + { + "epoch": 0.62, + "grad_norm": 1.8649904937983657, + "learning_rate": 6.635547168978032e-06, + "loss": 1.0682, + "step": 26357 + }, + { + "epoch": 0.62, + "grad_norm": 2.030333072576437, + "learning_rate": 6.6348286207048776e-06, + "loss": 0.9322, + "step": 26358 + }, + { + "epoch": 0.62, + "grad_norm": 2.0543048797805263, + "learning_rate": 6.634110092024819e-06, + "loss": 1.0559, + "step": 26359 + }, + { + "epoch": 0.62, + "grad_norm": 1.1816091864320784, + "learning_rate": 6.633391582942038e-06, + "loss": 0.9669, + "step": 26360 + }, + { + "epoch": 0.62, + "grad_norm": 2.3030491181875146, + "learning_rate": 6.632673093460715e-06, + "loss": 1.0083, + "step": 26361 + }, + { + "epoch": 0.62, + "grad_norm": 1.9243518916335347, + "learning_rate": 6.631954623585035e-06, + "loss": 1.119, + "step": 26362 + }, + { + "epoch": 0.62, + "grad_norm": 2.1848551364563233, + "learning_rate": 6.631236173319182e-06, + "loss": 0.8941, + "step": 26363 + }, + { + "epoch": 0.62, + "grad_norm": 2.3429312164895872, + "learning_rate": 6.630517742667341e-06, + "loss": 0.9793, + "step": 26364 + }, + { + "epoch": 0.62, + "grad_norm": 2.5862848797285127, + "learning_rate": 6.629799331633691e-06, + "loss": 1.0552, + "step": 26365 + }, + { + "epoch": 0.62, + "grad_norm": 2.4553448024662585, + "learning_rate": 6.629080940222414e-06, + "loss": 1.1229, + "step": 26366 + }, + { + "epoch": 0.62, + "grad_norm": 2.5153649631484094, + "learning_rate": 6.628362568437698e-06, + "loss": 0.88, + "step": 26367 + }, + { + "epoch": 0.62, + "grad_norm": 2.7503985428506526, + "learning_rate": 6.6276442162837215e-06, + "loss": 0.9461, + "step": 26368 + }, + { + "epoch": 0.62, + "grad_norm": 1.9085230711016157, + "learning_rate": 6.6269258837646675e-06, + "loss": 1.0045, + "step": 26369 + }, + { + "epoch": 0.62, + "grad_norm": 2.269606043876624, + "learning_rate": 6.626207570884717e-06, + "loss": 0.9157, + "step": 26370 + }, + { + "epoch": 0.62, + "grad_norm": 1.7495794567949232, + "learning_rate": 6.625489277648059e-06, + "loss": 0.9302, + "step": 26371 + }, + { + "epoch": 0.62, + "grad_norm": 2.913492255386941, + "learning_rate": 6.624771004058869e-06, + "loss": 1.0233, + "step": 26372 + }, + { + "epoch": 0.62, + "grad_norm": 2.5935037211255194, + "learning_rate": 6.624052750121328e-06, + "loss": 1.021, + "step": 26373 + }, + { + "epoch": 0.62, + "grad_norm": 2.0748546679787143, + "learning_rate": 6.623334515839627e-06, + "loss": 0.8768, + "step": 26374 + }, + { + "epoch": 0.62, + "grad_norm": 1.9948128162563674, + "learning_rate": 6.6226163012179345e-06, + "loss": 1.0311, + "step": 26375 + }, + { + "epoch": 0.62, + "grad_norm": 1.961799292342818, + "learning_rate": 6.621898106260443e-06, + "loss": 1.0934, + "step": 26376 + }, + { + "epoch": 0.62, + "grad_norm": 2.6268925295545134, + "learning_rate": 6.621179930971327e-06, + "loss": 1.0987, + "step": 26377 + }, + { + "epoch": 0.62, + "grad_norm": 1.9254632426357419, + "learning_rate": 6.620461775354776e-06, + "loss": 1.0157, + "step": 26378 + }, + { + "epoch": 0.62, + "grad_norm": 1.8999852014374317, + "learning_rate": 6.619743639414961e-06, + "loss": 1.0471, + "step": 26379 + }, + { + "epoch": 0.62, + "grad_norm": 2.2293821388404673, + "learning_rate": 6.619025523156073e-06, + "loss": 1.03, + "step": 26380 + }, + { + "epoch": 0.62, + "grad_norm": 1.8468992430821218, + "learning_rate": 6.618307426582289e-06, + "loss": 1.062, + "step": 26381 + }, + { + "epoch": 0.62, + "grad_norm": 2.009882707123369, + "learning_rate": 6.617589349697786e-06, + "loss": 0.8746, + "step": 26382 + }, + { + "epoch": 0.62, + "grad_norm": 1.9964579683373775, + "learning_rate": 6.61687129250675e-06, + "loss": 1.1064, + "step": 26383 + }, + { + "epoch": 0.62, + "grad_norm": 2.126777246369613, + "learning_rate": 6.616153255013363e-06, + "loss": 0.9912, + "step": 26384 + }, + { + "epoch": 0.62, + "grad_norm": 2.2844310097695155, + "learning_rate": 6.6154352372218015e-06, + "loss": 1.0682, + "step": 26385 + }, + { + "epoch": 0.62, + "grad_norm": 1.9001979415329178, + "learning_rate": 6.614717239136246e-06, + "loss": 0.8654, + "step": 26386 + }, + { + "epoch": 0.62, + "grad_norm": 2.058017107061124, + "learning_rate": 6.6139992607608795e-06, + "loss": 0.9759, + "step": 26387 + }, + { + "epoch": 0.62, + "grad_norm": 2.798820077487905, + "learning_rate": 6.613281302099881e-06, + "loss": 0.9939, + "step": 26388 + }, + { + "epoch": 0.62, + "grad_norm": 2.196870707308291, + "learning_rate": 6.612563363157434e-06, + "loss": 1.0123, + "step": 26389 + }, + { + "epoch": 0.62, + "grad_norm": 1.999395418511052, + "learning_rate": 6.61184544393771e-06, + "loss": 1.1185, + "step": 26390 + }, + { + "epoch": 0.62, + "grad_norm": 1.8875272984930613, + "learning_rate": 6.611127544444897e-06, + "loss": 1.0497, + "step": 26391 + }, + { + "epoch": 0.62, + "grad_norm": 1.9683267760189125, + "learning_rate": 6.610409664683176e-06, + "loss": 1.0644, + "step": 26392 + }, + { + "epoch": 0.62, + "grad_norm": 2.0866806205849255, + "learning_rate": 6.609691804656719e-06, + "loss": 0.9587, + "step": 26393 + }, + { + "epoch": 0.62, + "grad_norm": 1.1335639381004943, + "learning_rate": 6.608973964369709e-06, + "loss": 0.9835, + "step": 26394 + }, + { + "epoch": 0.62, + "grad_norm": 1.9690976413075632, + "learning_rate": 6.608256143826328e-06, + "loss": 0.9804, + "step": 26395 + }, + { + "epoch": 0.62, + "grad_norm": 2.3220756603354733, + "learning_rate": 6.607538343030755e-06, + "loss": 1.0292, + "step": 26396 + }, + { + "epoch": 0.62, + "grad_norm": 1.0813552212345026, + "learning_rate": 6.6068205619871664e-06, + "loss": 0.964, + "step": 26397 + }, + { + "epoch": 0.62, + "grad_norm": 1.9118361199901888, + "learning_rate": 6.606102800699741e-06, + "loss": 0.9658, + "step": 26398 + }, + { + "epoch": 0.62, + "grad_norm": 2.9791366862218074, + "learning_rate": 6.605385059172662e-06, + "loss": 0.9831, + "step": 26399 + }, + { + "epoch": 0.62, + "grad_norm": 2.060034316503745, + "learning_rate": 6.6046673374101054e-06, + "loss": 0.9622, + "step": 26400 + }, + { + "epoch": 0.62, + "grad_norm": 1.0784561313435617, + "learning_rate": 6.60394963541625e-06, + "loss": 0.9479, + "step": 26401 + }, + { + "epoch": 0.62, + "grad_norm": 1.905106678586031, + "learning_rate": 6.603231953195272e-06, + "loss": 1.0189, + "step": 26402 + }, + { + "epoch": 0.62, + "grad_norm": 2.143794792518167, + "learning_rate": 6.6025142907513596e-06, + "loss": 0.9525, + "step": 26403 + }, + { + "epoch": 0.62, + "grad_norm": 1.7653431879803028, + "learning_rate": 6.601796648088681e-06, + "loss": 0.9376, + "step": 26404 + }, + { + "epoch": 0.62, + "grad_norm": 2.0019519486074944, + "learning_rate": 6.601079025211418e-06, + "loss": 1.0477, + "step": 26405 + }, + { + "epoch": 0.62, + "grad_norm": 2.158614250062722, + "learning_rate": 6.600361422123751e-06, + "loss": 0.911, + "step": 26406 + }, + { + "epoch": 0.62, + "grad_norm": 2.0473149108199307, + "learning_rate": 6.599643838829852e-06, + "loss": 0.9702, + "step": 26407 + }, + { + "epoch": 0.62, + "grad_norm": 1.903736926063139, + "learning_rate": 6.598926275333906e-06, + "loss": 0.917, + "step": 26408 + }, + { + "epoch": 0.62, + "grad_norm": 1.0512473039363581, + "learning_rate": 6.598208731640085e-06, + "loss": 0.9649, + "step": 26409 + }, + { + "epoch": 0.62, + "grad_norm": 1.899396893342781, + "learning_rate": 6.597491207752577e-06, + "loss": 1.0787, + "step": 26410 + }, + { + "epoch": 0.62, + "grad_norm": 2.1780819294267375, + "learning_rate": 6.596773703675545e-06, + "loss": 0.9406, + "step": 26411 + }, + { + "epoch": 0.62, + "grad_norm": 1.7363608483791395, + "learning_rate": 6.596056219413176e-06, + "loss": 0.9335, + "step": 26412 + }, + { + "epoch": 0.62, + "grad_norm": 1.9015489515490538, + "learning_rate": 6.595338754969649e-06, + "loss": 1.0155, + "step": 26413 + }, + { + "epoch": 0.62, + "grad_norm": 1.9843816610425378, + "learning_rate": 6.594621310349134e-06, + "loss": 0.8996, + "step": 26414 + }, + { + "epoch": 0.62, + "grad_norm": 2.1100910774912336, + "learning_rate": 6.59390388555581e-06, + "loss": 0.9117, + "step": 26415 + }, + { + "epoch": 0.62, + "grad_norm": 2.426985245059931, + "learning_rate": 6.593186480593858e-06, + "loss": 1.1174, + "step": 26416 + }, + { + "epoch": 0.62, + "grad_norm": 2.558827812284894, + "learning_rate": 6.592469095467456e-06, + "loss": 0.9192, + "step": 26417 + }, + { + "epoch": 0.62, + "grad_norm": 2.887935006350032, + "learning_rate": 6.591751730180774e-06, + "loss": 1.039, + "step": 26418 + }, + { + "epoch": 0.62, + "grad_norm": 1.847218622959625, + "learning_rate": 6.591034384737992e-06, + "loss": 1.0333, + "step": 26419 + }, + { + "epoch": 0.62, + "grad_norm": 2.0767114868610004, + "learning_rate": 6.5903170591432896e-06, + "loss": 1.0798, + "step": 26420 + }, + { + "epoch": 0.62, + "grad_norm": 1.8216461511680184, + "learning_rate": 6.58959975340084e-06, + "loss": 1.0689, + "step": 26421 + }, + { + "epoch": 0.62, + "grad_norm": 1.845093132761, + "learning_rate": 6.588882467514819e-06, + "loss": 0.9787, + "step": 26422 + }, + { + "epoch": 0.62, + "grad_norm": 1.2017248706682837, + "learning_rate": 6.588165201489405e-06, + "loss": 1.0498, + "step": 26423 + }, + { + "epoch": 0.62, + "grad_norm": 2.462096037479424, + "learning_rate": 6.587447955328775e-06, + "loss": 1.0771, + "step": 26424 + }, + { + "epoch": 0.62, + "grad_norm": 1.9768224712961462, + "learning_rate": 6.586730729037102e-06, + "loss": 1.0544, + "step": 26425 + }, + { + "epoch": 0.62, + "grad_norm": 1.1065639125477094, + "learning_rate": 6.586013522618562e-06, + "loss": 0.9625, + "step": 26426 + }, + { + "epoch": 0.62, + "grad_norm": 2.0958426505515115, + "learning_rate": 6.585296336077337e-06, + "loss": 1.1783, + "step": 26427 + }, + { + "epoch": 0.62, + "grad_norm": 2.290948275588778, + "learning_rate": 6.584579169417592e-06, + "loss": 1.0325, + "step": 26428 + }, + { + "epoch": 0.62, + "grad_norm": 2.472279972887875, + "learning_rate": 6.58386202264351e-06, + "loss": 1.0225, + "step": 26429 + }, + { + "epoch": 0.62, + "grad_norm": 2.26355336984155, + "learning_rate": 6.583144895759264e-06, + "loss": 1.0224, + "step": 26430 + }, + { + "epoch": 0.62, + "grad_norm": 2.3978258334997475, + "learning_rate": 6.582427788769033e-06, + "loss": 0.8955, + "step": 26431 + }, + { + "epoch": 0.62, + "grad_norm": 2.198370424039887, + "learning_rate": 6.581710701676989e-06, + "loss": 0.963, + "step": 26432 + }, + { + "epoch": 0.62, + "grad_norm": 1.9765862028935075, + "learning_rate": 6.580993634487306e-06, + "loss": 1.0896, + "step": 26433 + }, + { + "epoch": 0.62, + "grad_norm": 2.106603239987414, + "learning_rate": 6.580276587204163e-06, + "loss": 0.9887, + "step": 26434 + }, + { + "epoch": 0.62, + "grad_norm": 1.8811972458600645, + "learning_rate": 6.579559559831729e-06, + "loss": 1.0089, + "step": 26435 + }, + { + "epoch": 0.62, + "grad_norm": 1.9659244556669593, + "learning_rate": 6.5788425523741825e-06, + "loss": 1.076, + "step": 26436 + }, + { + "epoch": 0.62, + "grad_norm": 2.291071225448566, + "learning_rate": 6.578125564835699e-06, + "loss": 1.0367, + "step": 26437 + }, + { + "epoch": 0.62, + "grad_norm": 1.9483938647505652, + "learning_rate": 6.5774085972204535e-06, + "loss": 1.1632, + "step": 26438 + }, + { + "epoch": 0.62, + "grad_norm": 2.2280437594069644, + "learning_rate": 6.576691649532615e-06, + "loss": 0.9781, + "step": 26439 + }, + { + "epoch": 0.62, + "grad_norm": 1.9960399200978491, + "learning_rate": 6.575974721776364e-06, + "loss": 0.9834, + "step": 26440 + }, + { + "epoch": 0.62, + "grad_norm": 1.9281482582562657, + "learning_rate": 6.575257813955873e-06, + "loss": 0.9813, + "step": 26441 + }, + { + "epoch": 0.62, + "grad_norm": 2.053523823448891, + "learning_rate": 6.574540926075314e-06, + "loss": 1.0048, + "step": 26442 + }, + { + "epoch": 0.62, + "grad_norm": 1.9932310612584028, + "learning_rate": 6.57382405813886e-06, + "loss": 0.9813, + "step": 26443 + }, + { + "epoch": 0.62, + "grad_norm": 2.304284042467197, + "learning_rate": 6.573107210150688e-06, + "loss": 1.0298, + "step": 26444 + }, + { + "epoch": 0.62, + "grad_norm": 1.975390443036688, + "learning_rate": 6.5723903821149746e-06, + "loss": 1.0506, + "step": 26445 + }, + { + "epoch": 0.62, + "grad_norm": 2.141196405293513, + "learning_rate": 6.5716735740358864e-06, + "loss": 1.0052, + "step": 26446 + }, + { + "epoch": 0.62, + "grad_norm": 2.1250423675233447, + "learning_rate": 6.570956785917599e-06, + "loss": 0.9579, + "step": 26447 + }, + { + "epoch": 0.62, + "grad_norm": 2.0320502621290375, + "learning_rate": 6.57024001776429e-06, + "loss": 1.004, + "step": 26448 + }, + { + "epoch": 0.62, + "grad_norm": 1.8996774710234623, + "learning_rate": 6.569523269580127e-06, + "loss": 1.077, + "step": 26449 + }, + { + "epoch": 0.62, + "grad_norm": 1.8211362846895398, + "learning_rate": 6.568806541369287e-06, + "loss": 0.9554, + "step": 26450 + }, + { + "epoch": 0.62, + "grad_norm": 1.9672701973061146, + "learning_rate": 6.568089833135939e-06, + "loss": 1.078, + "step": 26451 + }, + { + "epoch": 0.62, + "grad_norm": 1.9375067679994566, + "learning_rate": 6.567373144884262e-06, + "loss": 0.8582, + "step": 26452 + }, + { + "epoch": 0.62, + "grad_norm": 1.8553782758488928, + "learning_rate": 6.566656476618424e-06, + "loss": 1.0491, + "step": 26453 + }, + { + "epoch": 0.62, + "grad_norm": 2.8677821755868793, + "learning_rate": 6.565939828342598e-06, + "loss": 0.9802, + "step": 26454 + }, + { + "epoch": 0.62, + "grad_norm": 1.1704075757166519, + "learning_rate": 6.565223200060961e-06, + "loss": 0.9072, + "step": 26455 + }, + { + "epoch": 0.62, + "grad_norm": 1.1316825750042934, + "learning_rate": 6.564506591777677e-06, + "loss": 0.9327, + "step": 26456 + }, + { + "epoch": 0.62, + "grad_norm": 2.0952642697665493, + "learning_rate": 6.563790003496926e-06, + "loss": 0.8297, + "step": 26457 + }, + { + "epoch": 0.62, + "grad_norm": 1.0392383132260226, + "learning_rate": 6.563073435222874e-06, + "loss": 0.9569, + "step": 26458 + }, + { + "epoch": 0.62, + "grad_norm": 1.7758550256204335, + "learning_rate": 6.562356886959704e-06, + "loss": 0.9193, + "step": 26459 + }, + { + "epoch": 0.62, + "grad_norm": 2.1476713932210494, + "learning_rate": 6.561640358711574e-06, + "loss": 0.9773, + "step": 26460 + }, + { + "epoch": 0.62, + "grad_norm": 3.3048438199110537, + "learning_rate": 6.560923850482665e-06, + "loss": 0.9275, + "step": 26461 + }, + { + "epoch": 0.62, + "grad_norm": 1.9327665272966204, + "learning_rate": 6.560207362277146e-06, + "loss": 0.9117, + "step": 26462 + }, + { + "epoch": 0.62, + "grad_norm": 1.9365053596845476, + "learning_rate": 6.55949089409919e-06, + "loss": 0.923, + "step": 26463 + }, + { + "epoch": 0.62, + "grad_norm": 2.028085660837627, + "learning_rate": 6.558774445952962e-06, + "loss": 0.9253, + "step": 26464 + }, + { + "epoch": 0.62, + "grad_norm": 2.1491199206136993, + "learning_rate": 6.558058017842643e-06, + "loss": 0.9996, + "step": 26465 + }, + { + "epoch": 0.62, + "grad_norm": 1.9261334772495073, + "learning_rate": 6.5573416097724e-06, + "loss": 1.0401, + "step": 26466 + }, + { + "epoch": 0.62, + "grad_norm": 1.9427805492636816, + "learning_rate": 6.5566252217464045e-06, + "loss": 0.9755, + "step": 26467 + }, + { + "epoch": 0.62, + "grad_norm": 2.135127577521987, + "learning_rate": 6.555908853768824e-06, + "loss": 0.9533, + "step": 26468 + }, + { + "epoch": 0.62, + "grad_norm": 1.9660751220980188, + "learning_rate": 6.555192505843836e-06, + "loss": 1.082, + "step": 26469 + }, + { + "epoch": 0.62, + "grad_norm": 2.1443405647438114, + "learning_rate": 6.554476177975606e-06, + "loss": 1.0288, + "step": 26470 + }, + { + "epoch": 0.62, + "grad_norm": 2.079816941756685, + "learning_rate": 6.553759870168306e-06, + "loss": 1.0351, + "step": 26471 + }, + { + "epoch": 0.62, + "grad_norm": 2.262931851600509, + "learning_rate": 6.553043582426106e-06, + "loss": 0.9142, + "step": 26472 + }, + { + "epoch": 0.62, + "grad_norm": 2.0860236115308277, + "learning_rate": 6.552327314753182e-06, + "loss": 0.9427, + "step": 26473 + }, + { + "epoch": 0.62, + "grad_norm": 2.3989771289675943, + "learning_rate": 6.551611067153698e-06, + "loss": 1.0419, + "step": 26474 + }, + { + "epoch": 0.62, + "grad_norm": 1.8702638600951393, + "learning_rate": 6.550894839631823e-06, + "loss": 0.9072, + "step": 26475 + }, + { + "epoch": 0.62, + "grad_norm": 1.0893941305593289, + "learning_rate": 6.550178632191737e-06, + "loss": 0.9537, + "step": 26476 + }, + { + "epoch": 0.62, + "grad_norm": 3.046698281622499, + "learning_rate": 6.5494624448375964e-06, + "loss": 1.145, + "step": 26477 + }, + { + "epoch": 0.62, + "grad_norm": 2.213193042932566, + "learning_rate": 6.548746277573581e-06, + "loss": 1.1177, + "step": 26478 + }, + { + "epoch": 0.62, + "grad_norm": 1.8589554981376364, + "learning_rate": 6.548030130403854e-06, + "loss": 0.983, + "step": 26479 + }, + { + "epoch": 0.62, + "grad_norm": 2.3211723873247254, + "learning_rate": 6.5473140033325935e-06, + "loss": 0.9095, + "step": 26480 + }, + { + "epoch": 0.62, + "grad_norm": 1.9611164402959795, + "learning_rate": 6.546597896363962e-06, + "loss": 0.9882, + "step": 26481 + }, + { + "epoch": 0.62, + "grad_norm": 2.0694677076695753, + "learning_rate": 6.54588180950213e-06, + "loss": 1.0439, + "step": 26482 + }, + { + "epoch": 0.62, + "grad_norm": 2.256706682145168, + "learning_rate": 6.545165742751269e-06, + "loss": 0.978, + "step": 26483 + }, + { + "epoch": 0.62, + "grad_norm": 2.131560417252781, + "learning_rate": 6.544449696115545e-06, + "loss": 1.1097, + "step": 26484 + }, + { + "epoch": 0.62, + "grad_norm": 1.9447681916719552, + "learning_rate": 6.543733669599131e-06, + "loss": 1.0094, + "step": 26485 + }, + { + "epoch": 0.62, + "grad_norm": 1.1228871731611174, + "learning_rate": 6.543017663206192e-06, + "loss": 0.8386, + "step": 26486 + }, + { + "epoch": 0.62, + "grad_norm": 1.7603461785013517, + "learning_rate": 6.542301676940901e-06, + "loss": 0.9459, + "step": 26487 + }, + { + "epoch": 0.62, + "grad_norm": 1.8523571029861154, + "learning_rate": 6.541585710807421e-06, + "loss": 0.9946, + "step": 26488 + }, + { + "epoch": 0.62, + "grad_norm": 2.094492855295042, + "learning_rate": 6.540869764809925e-06, + "loss": 1.1653, + "step": 26489 + }, + { + "epoch": 0.62, + "grad_norm": 1.9076640248273002, + "learning_rate": 6.5401538389525835e-06, + "loss": 0.958, + "step": 26490 + }, + { + "epoch": 0.62, + "grad_norm": 2.2437986687835796, + "learning_rate": 6.539437933239559e-06, + "loss": 1.0578, + "step": 26491 + }, + { + "epoch": 0.62, + "grad_norm": 2.124217426605486, + "learning_rate": 6.53872204767502e-06, + "loss": 0.9493, + "step": 26492 + }, + { + "epoch": 0.62, + "grad_norm": 2.370187330838746, + "learning_rate": 6.538006182263139e-06, + "loss": 1.0148, + "step": 26493 + }, + { + "epoch": 0.62, + "grad_norm": 1.9179441536396455, + "learning_rate": 6.537290337008085e-06, + "loss": 0.976, + "step": 26494 + }, + { + "epoch": 0.62, + "grad_norm": 2.1958734912977063, + "learning_rate": 6.536574511914021e-06, + "loss": 1.0697, + "step": 26495 + }, + { + "epoch": 0.62, + "grad_norm": 2.430760397118953, + "learning_rate": 6.535858706985114e-06, + "loss": 1.0642, + "step": 26496 + }, + { + "epoch": 0.62, + "grad_norm": 1.0476969514211523, + "learning_rate": 6.5351429222255394e-06, + "loss": 0.9947, + "step": 26497 + }, + { + "epoch": 0.62, + "grad_norm": 1.0962456405154286, + "learning_rate": 6.5344271576394555e-06, + "loss": 0.9948, + "step": 26498 + }, + { + "epoch": 0.62, + "grad_norm": 2.2585962036318072, + "learning_rate": 6.533711413231036e-06, + "loss": 0.8965, + "step": 26499 + }, + { + "epoch": 0.62, + "grad_norm": 1.9647537468209835, + "learning_rate": 6.532995689004444e-06, + "loss": 1.0186, + "step": 26500 + }, + { + "epoch": 0.62, + "grad_norm": 1.8702132632374515, + "learning_rate": 6.532279984963852e-06, + "loss": 1.0065, + "step": 26501 + }, + { + "epoch": 0.62, + "grad_norm": 2.1105640278828233, + "learning_rate": 6.531564301113422e-06, + "loss": 1.0862, + "step": 26502 + }, + { + "epoch": 0.62, + "grad_norm": 1.8642522952686036, + "learning_rate": 6.530848637457322e-06, + "loss": 1.0006, + "step": 26503 + }, + { + "epoch": 0.62, + "grad_norm": 2.1048472762083623, + "learning_rate": 6.530132993999723e-06, + "loss": 0.9461, + "step": 26504 + }, + { + "epoch": 0.62, + "grad_norm": 1.9860991557690095, + "learning_rate": 6.529417370744783e-06, + "loss": 0.9947, + "step": 26505 + }, + { + "epoch": 0.62, + "grad_norm": 1.994082825287277, + "learning_rate": 6.528701767696676e-06, + "loss": 0.916, + "step": 26506 + }, + { + "epoch": 0.62, + "grad_norm": 2.0566943368572783, + "learning_rate": 6.527986184859565e-06, + "loss": 0.9098, + "step": 26507 + }, + { + "epoch": 0.62, + "grad_norm": 2.2279386152640535, + "learning_rate": 6.527270622237623e-06, + "loss": 1.099, + "step": 26508 + }, + { + "epoch": 0.62, + "grad_norm": 2.104060764744036, + "learning_rate": 6.526555079835004e-06, + "loss": 1.0658, + "step": 26509 + }, + { + "epoch": 0.62, + "grad_norm": 2.232714090512762, + "learning_rate": 6.525839557655884e-06, + "loss": 1.0669, + "step": 26510 + }, + { + "epoch": 0.62, + "grad_norm": 1.845292858666068, + "learning_rate": 6.525124055704425e-06, + "loss": 1.0522, + "step": 26511 + }, + { + "epoch": 0.62, + "grad_norm": 1.8207032128174225, + "learning_rate": 6.5244085739848e-06, + "loss": 0.9118, + "step": 26512 + }, + { + "epoch": 0.62, + "grad_norm": 2.11723174553911, + "learning_rate": 6.52369311250116e-06, + "loss": 1.0603, + "step": 26513 + }, + { + "epoch": 0.62, + "grad_norm": 1.9025438824457368, + "learning_rate": 6.522977671257683e-06, + "loss": 1.123, + "step": 26514 + }, + { + "epoch": 0.62, + "grad_norm": 1.0446729726787611, + "learning_rate": 6.522262250258533e-06, + "loss": 0.959, + "step": 26515 + }, + { + "epoch": 0.62, + "grad_norm": 2.4193583373790246, + "learning_rate": 6.521546849507872e-06, + "loss": 1.0487, + "step": 26516 + }, + { + "epoch": 0.62, + "grad_norm": 1.955982568656888, + "learning_rate": 6.520831469009863e-06, + "loss": 0.9898, + "step": 26517 + }, + { + "epoch": 0.62, + "grad_norm": 2.0493967502592234, + "learning_rate": 6.520116108768677e-06, + "loss": 0.9553, + "step": 26518 + }, + { + "epoch": 0.62, + "grad_norm": 2.0017999720723467, + "learning_rate": 6.519400768788478e-06, + "loss": 0.9628, + "step": 26519 + }, + { + "epoch": 0.62, + "grad_norm": 2.2374850415442045, + "learning_rate": 6.518685449073427e-06, + "loss": 0.9743, + "step": 26520 + }, + { + "epoch": 0.62, + "grad_norm": 2.2189545594826807, + "learning_rate": 6.517970149627692e-06, + "loss": 0.9239, + "step": 26521 + }, + { + "epoch": 0.62, + "grad_norm": 2.4298905093470755, + "learning_rate": 6.5172548704554405e-06, + "loss": 0.969, + "step": 26522 + }, + { + "epoch": 0.62, + "grad_norm": 2.336897976015313, + "learning_rate": 6.5165396115608315e-06, + "loss": 0.8209, + "step": 26523 + }, + { + "epoch": 0.62, + "grad_norm": 1.9314120343512624, + "learning_rate": 6.51582437294803e-06, + "loss": 0.8234, + "step": 26524 + }, + { + "epoch": 0.62, + "grad_norm": 2.200982988820218, + "learning_rate": 6.5151091546212034e-06, + "loss": 1.0461, + "step": 26525 + }, + { + "epoch": 0.62, + "grad_norm": 2.030173797648657, + "learning_rate": 6.5143939565845175e-06, + "loss": 0.9645, + "step": 26526 + }, + { + "epoch": 0.62, + "grad_norm": 1.114656284129864, + "learning_rate": 6.513678778842129e-06, + "loss": 0.9664, + "step": 26527 + }, + { + "epoch": 0.62, + "grad_norm": 2.0288532641206345, + "learning_rate": 6.512963621398207e-06, + "loss": 0.9119, + "step": 26528 + }, + { + "epoch": 0.63, + "grad_norm": 1.0665595917921586, + "learning_rate": 6.512248484256918e-06, + "loss": 0.9428, + "step": 26529 + }, + { + "epoch": 0.63, + "grad_norm": 2.1182348869129406, + "learning_rate": 6.51153336742242e-06, + "loss": 0.9616, + "step": 26530 + }, + { + "epoch": 0.63, + "grad_norm": 1.930482846890858, + "learning_rate": 6.51081827089888e-06, + "loss": 0.9546, + "step": 26531 + }, + { + "epoch": 0.63, + "grad_norm": 2.5898665225601425, + "learning_rate": 6.510103194690459e-06, + "loss": 1.09, + "step": 26532 + }, + { + "epoch": 0.63, + "grad_norm": 2.233090892532647, + "learning_rate": 6.509388138801326e-06, + "loss": 1.0315, + "step": 26533 + }, + { + "epoch": 0.63, + "grad_norm": 1.869381689004607, + "learning_rate": 6.508673103235637e-06, + "loss": 1.0689, + "step": 26534 + }, + { + "epoch": 0.63, + "grad_norm": 3.031308794227201, + "learning_rate": 6.507958087997559e-06, + "loss": 1.0891, + "step": 26535 + }, + { + "epoch": 0.63, + "grad_norm": 2.065784245127757, + "learning_rate": 6.507243093091258e-06, + "loss": 0.8898, + "step": 26536 + }, + { + "epoch": 0.63, + "grad_norm": 1.9083091947706718, + "learning_rate": 6.506528118520888e-06, + "loss": 1.1278, + "step": 26537 + }, + { + "epoch": 0.63, + "grad_norm": 1.9727771729052355, + "learning_rate": 6.50581316429062e-06, + "loss": 1.0442, + "step": 26538 + }, + { + "epoch": 0.63, + "grad_norm": 1.1303988553215925, + "learning_rate": 6.505098230404614e-06, + "loss": 1.0427, + "step": 26539 + }, + { + "epoch": 0.63, + "grad_norm": 3.783814080327487, + "learning_rate": 6.504383316867035e-06, + "loss": 1.0557, + "step": 26540 + }, + { + "epoch": 0.63, + "grad_norm": 2.2765407185176874, + "learning_rate": 6.503668423682037e-06, + "loss": 0.8488, + "step": 26541 + }, + { + "epoch": 0.63, + "grad_norm": 2.130327414286273, + "learning_rate": 6.502953550853792e-06, + "loss": 0.9418, + "step": 26542 + }, + { + "epoch": 0.63, + "grad_norm": 1.961107697239153, + "learning_rate": 6.50223869838646e-06, + "loss": 0.9997, + "step": 26543 + }, + { + "epoch": 0.63, + "grad_norm": 1.9298814184324529, + "learning_rate": 6.501523866284201e-06, + "loss": 1.0149, + "step": 26544 + }, + { + "epoch": 0.63, + "grad_norm": 2.3300342898430353, + "learning_rate": 6.500809054551173e-06, + "loss": 0.9425, + "step": 26545 + }, + { + "epoch": 0.63, + "grad_norm": 2.061604127565189, + "learning_rate": 6.500094263191548e-06, + "loss": 0.9462, + "step": 26546 + }, + { + "epoch": 0.63, + "grad_norm": 1.7832495313452328, + "learning_rate": 6.499379492209481e-06, + "loss": 1.1406, + "step": 26547 + }, + { + "epoch": 0.63, + "grad_norm": 1.102351238653976, + "learning_rate": 6.498664741609135e-06, + "loss": 1.0386, + "step": 26548 + }, + { + "epoch": 0.63, + "grad_norm": 2.203559182760545, + "learning_rate": 6.497950011394669e-06, + "loss": 1.0588, + "step": 26549 + }, + { + "epoch": 0.63, + "grad_norm": 1.9351895368889152, + "learning_rate": 6.497235301570251e-06, + "loss": 0.957, + "step": 26550 + }, + { + "epoch": 0.63, + "grad_norm": 2.3095265125754714, + "learning_rate": 6.496520612140033e-06, + "loss": 0.8954, + "step": 26551 + }, + { + "epoch": 0.63, + "grad_norm": 0.9943935833855404, + "learning_rate": 6.4958059431081845e-06, + "loss": 0.958, + "step": 26552 + }, + { + "epoch": 0.63, + "grad_norm": 2.2062292428359784, + "learning_rate": 6.4950912944788614e-06, + "loss": 0.9951, + "step": 26553 + }, + { + "epoch": 0.63, + "grad_norm": 2.2188419565706243, + "learning_rate": 6.494376666256229e-06, + "loss": 1.0223, + "step": 26554 + }, + { + "epoch": 0.63, + "grad_norm": 4.153715312827837, + "learning_rate": 6.493662058444443e-06, + "loss": 0.9266, + "step": 26555 + }, + { + "epoch": 0.63, + "grad_norm": 1.9111332760577842, + "learning_rate": 6.492947471047666e-06, + "loss": 0.9595, + "step": 26556 + }, + { + "epoch": 0.63, + "grad_norm": 1.9330240537453147, + "learning_rate": 6.492232904070065e-06, + "loss": 0.9306, + "step": 26557 + }, + { + "epoch": 0.63, + "grad_norm": 2.2275317279547755, + "learning_rate": 6.491518357515788e-06, + "loss": 1.0794, + "step": 26558 + }, + { + "epoch": 0.63, + "grad_norm": 2.3032323997502537, + "learning_rate": 6.490803831389003e-06, + "loss": 0.9745, + "step": 26559 + }, + { + "epoch": 0.63, + "grad_norm": 2.2401956604873314, + "learning_rate": 6.490089325693869e-06, + "loss": 0.9753, + "step": 26560 + }, + { + "epoch": 0.63, + "grad_norm": 1.9071417171700997, + "learning_rate": 6.489374840434551e-06, + "loss": 0.9889, + "step": 26561 + }, + { + "epoch": 0.63, + "grad_norm": 1.9112182142710659, + "learning_rate": 6.4886603756151985e-06, + "loss": 0.9512, + "step": 26562 + }, + { + "epoch": 0.63, + "grad_norm": 1.9530883650920186, + "learning_rate": 6.487945931239978e-06, + "loss": 0.9629, + "step": 26563 + }, + { + "epoch": 0.63, + "grad_norm": 1.95417549707378, + "learning_rate": 6.487231507313052e-06, + "loss": 0.8869, + "step": 26564 + }, + { + "epoch": 0.63, + "grad_norm": 1.8566226044934433, + "learning_rate": 6.486517103838574e-06, + "loss": 1.0107, + "step": 26565 + }, + { + "epoch": 0.63, + "grad_norm": 2.298700900150792, + "learning_rate": 6.485802720820702e-06, + "loss": 1.2001, + "step": 26566 + }, + { + "epoch": 0.63, + "grad_norm": 2.0634111917840428, + "learning_rate": 6.485088358263603e-06, + "loss": 1.0091, + "step": 26567 + }, + { + "epoch": 0.63, + "grad_norm": 1.9882130157972735, + "learning_rate": 6.4843740161714334e-06, + "loss": 1.0217, + "step": 26568 + }, + { + "epoch": 0.63, + "grad_norm": 2.900303351248414, + "learning_rate": 6.483659694548348e-06, + "loss": 0.9144, + "step": 26569 + }, + { + "epoch": 0.63, + "grad_norm": 1.9639024296229908, + "learning_rate": 6.4829453933985096e-06, + "loss": 1.0308, + "step": 26570 + }, + { + "epoch": 0.63, + "grad_norm": 2.165721938648338, + "learning_rate": 6.482231112726081e-06, + "loss": 0.9333, + "step": 26571 + }, + { + "epoch": 0.63, + "grad_norm": 1.9169040438981388, + "learning_rate": 6.4815168525352124e-06, + "loss": 0.9439, + "step": 26572 + }, + { + "epoch": 0.63, + "grad_norm": 2.191458906086769, + "learning_rate": 6.4808026128300665e-06, + "loss": 1.0451, + "step": 26573 + }, + { + "epoch": 0.63, + "grad_norm": 2.340290765420407, + "learning_rate": 6.480088393614802e-06, + "loss": 0.9139, + "step": 26574 + }, + { + "epoch": 0.63, + "grad_norm": 1.7733324716021817, + "learning_rate": 6.4793741948935795e-06, + "loss": 0.8976, + "step": 26575 + }, + { + "epoch": 0.63, + "grad_norm": 1.9259744432040389, + "learning_rate": 6.478660016670555e-06, + "loss": 0.9585, + "step": 26576 + }, + { + "epoch": 0.63, + "grad_norm": 1.8953447713997769, + "learning_rate": 6.4779458589498835e-06, + "loss": 0.9831, + "step": 26577 + }, + { + "epoch": 0.63, + "grad_norm": 2.356495642984164, + "learning_rate": 6.477231721735729e-06, + "loss": 1.0127, + "step": 26578 + }, + { + "epoch": 0.63, + "grad_norm": 1.8281873978023349, + "learning_rate": 6.4765176050322465e-06, + "loss": 0.9997, + "step": 26579 + }, + { + "epoch": 0.63, + "grad_norm": 2.0315647297503174, + "learning_rate": 6.4758035088435935e-06, + "loss": 0.9781, + "step": 26580 + }, + { + "epoch": 0.63, + "grad_norm": 2.096829762904437, + "learning_rate": 6.475089433173926e-06, + "loss": 0.8187, + "step": 26581 + }, + { + "epoch": 0.63, + "grad_norm": 1.10911204361789, + "learning_rate": 6.474375378027408e-06, + "loss": 0.9199, + "step": 26582 + }, + { + "epoch": 0.63, + "grad_norm": 3.1737176741674875, + "learning_rate": 6.473661343408191e-06, + "loss": 0.8975, + "step": 26583 + }, + { + "epoch": 0.63, + "grad_norm": 2.0704183853935545, + "learning_rate": 6.472947329320434e-06, + "loss": 1.0519, + "step": 26584 + }, + { + "epoch": 0.63, + "grad_norm": 2.162141648439775, + "learning_rate": 6.472233335768295e-06, + "loss": 1.0454, + "step": 26585 + }, + { + "epoch": 0.63, + "grad_norm": 2.7891805130303267, + "learning_rate": 6.471519362755927e-06, + "loss": 0.9823, + "step": 26586 + }, + { + "epoch": 0.63, + "grad_norm": 1.0140471720354436, + "learning_rate": 6.470805410287493e-06, + "loss": 0.8017, + "step": 26587 + }, + { + "epoch": 0.63, + "grad_norm": 2.7887330388239873, + "learning_rate": 6.470091478367147e-06, + "loss": 0.934, + "step": 26588 + }, + { + "epoch": 0.63, + "grad_norm": 1.9423311283222242, + "learning_rate": 6.469377566999049e-06, + "loss": 1.0121, + "step": 26589 + }, + { + "epoch": 0.63, + "grad_norm": 1.8900406370080522, + "learning_rate": 6.468663676187346e-06, + "loss": 1.0469, + "step": 26590 + }, + { + "epoch": 0.63, + "grad_norm": 1.920500735594087, + "learning_rate": 6.467949805936206e-06, + "loss": 1.0551, + "step": 26591 + }, + { + "epoch": 0.63, + "grad_norm": 2.0731284545630158, + "learning_rate": 6.46723595624978e-06, + "loss": 0.941, + "step": 26592 + }, + { + "epoch": 0.63, + "grad_norm": 1.1829699032440557, + "learning_rate": 6.4665221271322245e-06, + "loss": 0.9604, + "step": 26593 + }, + { + "epoch": 0.63, + "grad_norm": 2.2386426542348965, + "learning_rate": 6.465808318587693e-06, + "loss": 1.0248, + "step": 26594 + }, + { + "epoch": 0.63, + "grad_norm": 2.5002456879124426, + "learning_rate": 6.465094530620346e-06, + "loss": 0.9779, + "step": 26595 + }, + { + "epoch": 0.63, + "grad_norm": 1.9347429232206064, + "learning_rate": 6.464380763234341e-06, + "loss": 1.0689, + "step": 26596 + }, + { + "epoch": 0.63, + "grad_norm": 1.8171681722429456, + "learning_rate": 6.463667016433828e-06, + "loss": 1.0089, + "step": 26597 + }, + { + "epoch": 0.63, + "grad_norm": 3.4084760375591747, + "learning_rate": 6.462953290222963e-06, + "loss": 0.9873, + "step": 26598 + }, + { + "epoch": 0.63, + "grad_norm": 2.5150128799063434, + "learning_rate": 6.462239584605907e-06, + "loss": 1.0296, + "step": 26599 + }, + { + "epoch": 0.63, + "grad_norm": 2.2034477310747222, + "learning_rate": 6.461525899586812e-06, + "loss": 1.027, + "step": 26600 + }, + { + "epoch": 0.63, + "grad_norm": 2.037634191498605, + "learning_rate": 6.460812235169832e-06, + "loss": 1.0845, + "step": 26601 + }, + { + "epoch": 0.63, + "grad_norm": 1.0447235713610332, + "learning_rate": 6.460098591359123e-06, + "loss": 0.9146, + "step": 26602 + }, + { + "epoch": 0.63, + "grad_norm": 2.135752805331105, + "learning_rate": 6.459384968158844e-06, + "loss": 0.9728, + "step": 26603 + }, + { + "epoch": 0.63, + "grad_norm": 1.959798147217484, + "learning_rate": 6.458671365573144e-06, + "loss": 1.0285, + "step": 26604 + }, + { + "epoch": 0.63, + "grad_norm": 2.0532129378207853, + "learning_rate": 6.457957783606179e-06, + "loss": 1.0401, + "step": 26605 + }, + { + "epoch": 0.63, + "grad_norm": 2.0286136744724743, + "learning_rate": 6.457244222262111e-06, + "loss": 0.9834, + "step": 26606 + }, + { + "epoch": 0.63, + "grad_norm": 2.0199212907245494, + "learning_rate": 6.456530681545083e-06, + "loss": 1.0521, + "step": 26607 + }, + { + "epoch": 0.63, + "grad_norm": 2.6927833099271417, + "learning_rate": 6.455817161459257e-06, + "loss": 0.9616, + "step": 26608 + }, + { + "epoch": 0.63, + "grad_norm": 2.0254395487292327, + "learning_rate": 6.455103662008784e-06, + "loss": 0.9494, + "step": 26609 + }, + { + "epoch": 0.63, + "grad_norm": 2.3035334223182153, + "learning_rate": 6.4543901831978254e-06, + "loss": 1.0576, + "step": 26610 + }, + { + "epoch": 0.63, + "grad_norm": 1.1297671447193283, + "learning_rate": 6.453676725030524e-06, + "loss": 1.0601, + "step": 26611 + }, + { + "epoch": 0.63, + "grad_norm": 1.8390944472651891, + "learning_rate": 6.452963287511041e-06, + "loss": 0.9864, + "step": 26612 + }, + { + "epoch": 0.63, + "grad_norm": 1.9595483962040334, + "learning_rate": 6.452249870643531e-06, + "loss": 0.9985, + "step": 26613 + }, + { + "epoch": 0.63, + "grad_norm": 1.9811398014240023, + "learning_rate": 6.451536474432143e-06, + "loss": 0.878, + "step": 26614 + }, + { + "epoch": 0.63, + "grad_norm": 1.9149528039129766, + "learning_rate": 6.450823098881032e-06, + "loss": 0.9575, + "step": 26615 + }, + { + "epoch": 0.63, + "grad_norm": 2.1474777076173934, + "learning_rate": 6.450109743994353e-06, + "loss": 0.8754, + "step": 26616 + }, + { + "epoch": 0.63, + "grad_norm": 1.0751612388918226, + "learning_rate": 6.449396409776263e-06, + "loss": 1.0044, + "step": 26617 + }, + { + "epoch": 0.63, + "grad_norm": 1.9791628147650344, + "learning_rate": 6.448683096230906e-06, + "loss": 1.0147, + "step": 26618 + }, + { + "epoch": 0.63, + "grad_norm": 1.9843848318087323, + "learning_rate": 6.447969803362442e-06, + "loss": 0.9535, + "step": 26619 + }, + { + "epoch": 0.63, + "grad_norm": 2.2550419541252706, + "learning_rate": 6.447256531175026e-06, + "loss": 0.9083, + "step": 26620 + }, + { + "epoch": 0.63, + "grad_norm": 1.958286488885628, + "learning_rate": 6.446543279672803e-06, + "loss": 1.0585, + "step": 26621 + }, + { + "epoch": 0.63, + "grad_norm": 1.9508768651939858, + "learning_rate": 6.445830048859929e-06, + "loss": 1.1225, + "step": 26622 + }, + { + "epoch": 0.63, + "grad_norm": 1.746011300761917, + "learning_rate": 6.445116838740559e-06, + "loss": 1.0993, + "step": 26623 + }, + { + "epoch": 0.63, + "grad_norm": 1.8787011506235873, + "learning_rate": 6.444403649318849e-06, + "loss": 0.8613, + "step": 26624 + }, + { + "epoch": 0.63, + "grad_norm": 2.3668805241421156, + "learning_rate": 6.443690480598942e-06, + "loss": 0.9213, + "step": 26625 + }, + { + "epoch": 0.63, + "grad_norm": 1.8634495430494737, + "learning_rate": 6.4429773325849935e-06, + "loss": 0.9102, + "step": 26626 + }, + { + "epoch": 0.63, + "grad_norm": 2.040518786735785, + "learning_rate": 6.442264205281161e-06, + "loss": 0.9416, + "step": 26627 + }, + { + "epoch": 0.63, + "grad_norm": 2.088400500607105, + "learning_rate": 6.441551098691591e-06, + "loss": 0.8785, + "step": 26628 + }, + { + "epoch": 0.63, + "grad_norm": 2.336058197483541, + "learning_rate": 6.440838012820437e-06, + "loss": 0.9125, + "step": 26629 + }, + { + "epoch": 0.63, + "grad_norm": 2.0630124455227437, + "learning_rate": 6.440124947671848e-06, + "loss": 0.9097, + "step": 26630 + }, + { + "epoch": 0.63, + "grad_norm": 2.2481029668798223, + "learning_rate": 6.439411903249983e-06, + "loss": 0.9765, + "step": 26631 + }, + { + "epoch": 0.63, + "grad_norm": 1.9739769567023258, + "learning_rate": 6.438698879558989e-06, + "loss": 1.0277, + "step": 26632 + }, + { + "epoch": 0.63, + "grad_norm": 1.0499253309036387, + "learning_rate": 6.437985876603017e-06, + "loss": 0.9165, + "step": 26633 + }, + { + "epoch": 0.63, + "grad_norm": 1.938235949950221, + "learning_rate": 6.437272894386219e-06, + "loss": 1.0318, + "step": 26634 + }, + { + "epoch": 0.63, + "grad_norm": 1.9666142861923306, + "learning_rate": 6.4365599329127435e-06, + "loss": 0.9509, + "step": 26635 + }, + { + "epoch": 0.63, + "grad_norm": 2.5014793012377328, + "learning_rate": 6.435846992186746e-06, + "loss": 1.0523, + "step": 26636 + }, + { + "epoch": 0.63, + "grad_norm": 2.0016088017484104, + "learning_rate": 6.435134072212377e-06, + "loss": 0.853, + "step": 26637 + }, + { + "epoch": 0.63, + "grad_norm": 2.1888516201304076, + "learning_rate": 6.4344211729937855e-06, + "loss": 0.9771, + "step": 26638 + }, + { + "epoch": 0.63, + "grad_norm": 1.7763848109888882, + "learning_rate": 6.43370829453512e-06, + "loss": 1.002, + "step": 26639 + }, + { + "epoch": 0.63, + "grad_norm": 2.6163618836993923, + "learning_rate": 6.432995436840537e-06, + "loss": 1.0547, + "step": 26640 + }, + { + "epoch": 0.63, + "grad_norm": 2.1154995869348197, + "learning_rate": 6.43228259991418e-06, + "loss": 1.0292, + "step": 26641 + }, + { + "epoch": 0.63, + "grad_norm": 2.0262739499024756, + "learning_rate": 6.43156978376021e-06, + "loss": 1.0144, + "step": 26642 + }, + { + "epoch": 0.63, + "grad_norm": 1.9310170893680554, + "learning_rate": 6.430856988382766e-06, + "loss": 1.043, + "step": 26643 + }, + { + "epoch": 0.63, + "grad_norm": 2.1500950746290752, + "learning_rate": 6.430144213786002e-06, + "loss": 0.9863, + "step": 26644 + }, + { + "epoch": 0.63, + "grad_norm": 2.2168853298890445, + "learning_rate": 6.4294314599740715e-06, + "loss": 1.0283, + "step": 26645 + }, + { + "epoch": 0.63, + "grad_norm": 1.0616449028426596, + "learning_rate": 6.42871872695112e-06, + "loss": 0.9631, + "step": 26646 + }, + { + "epoch": 0.63, + "grad_norm": 2.052858170341152, + "learning_rate": 6.428006014721295e-06, + "loss": 1.0507, + "step": 26647 + }, + { + "epoch": 0.63, + "grad_norm": 2.033165794972546, + "learning_rate": 6.427293323288754e-06, + "loss": 0.869, + "step": 26648 + }, + { + "epoch": 0.63, + "grad_norm": 2.1228403237268463, + "learning_rate": 6.426580652657643e-06, + "loss": 1.0831, + "step": 26649 + }, + { + "epoch": 0.63, + "grad_norm": 2.0042386806361283, + "learning_rate": 6.4258680028321094e-06, + "loss": 0.9694, + "step": 26650 + }, + { + "epoch": 0.63, + "grad_norm": 2.2513321596683595, + "learning_rate": 6.425155373816303e-06, + "loss": 1.0283, + "step": 26651 + }, + { + "epoch": 0.63, + "grad_norm": 5.75486028446926, + "learning_rate": 6.424442765614376e-06, + "loss": 0.9782, + "step": 26652 + }, + { + "epoch": 0.63, + "grad_norm": 1.0814314820283808, + "learning_rate": 6.423730178230474e-06, + "loss": 0.9405, + "step": 26653 + }, + { + "epoch": 0.63, + "grad_norm": 2.355001074392143, + "learning_rate": 6.423017611668745e-06, + "loss": 1.0271, + "step": 26654 + }, + { + "epoch": 0.63, + "grad_norm": 2.0465623508979154, + "learning_rate": 6.422305065933342e-06, + "loss": 1.1025, + "step": 26655 + }, + { + "epoch": 0.63, + "grad_norm": 1.827830552967593, + "learning_rate": 6.421592541028414e-06, + "loss": 1.0061, + "step": 26656 + }, + { + "epoch": 0.63, + "grad_norm": 1.9167692343376928, + "learning_rate": 6.420880036958103e-06, + "loss": 1.0549, + "step": 26657 + }, + { + "epoch": 0.63, + "grad_norm": 2.119699755971627, + "learning_rate": 6.4201675537265615e-06, + "loss": 1.1275, + "step": 26658 + }, + { + "epoch": 0.63, + "grad_norm": 2.2806137881124435, + "learning_rate": 6.419455091337944e-06, + "loss": 1.0163, + "step": 26659 + }, + { + "epoch": 0.63, + "grad_norm": 2.155081464778438, + "learning_rate": 6.4187426497963855e-06, + "loss": 1.1983, + "step": 26660 + }, + { + "epoch": 0.63, + "grad_norm": 2.045927493348839, + "learning_rate": 6.418030229106045e-06, + "loss": 1.0043, + "step": 26661 + }, + { + "epoch": 0.63, + "grad_norm": 1.926064121030033, + "learning_rate": 6.417317829271062e-06, + "loss": 0.9089, + "step": 26662 + }, + { + "epoch": 0.63, + "grad_norm": 1.989016047806759, + "learning_rate": 6.4166054502955965e-06, + "loss": 1.0691, + "step": 26663 + }, + { + "epoch": 0.63, + "grad_norm": 2.036814887281774, + "learning_rate": 6.415893092183782e-06, + "loss": 0.9124, + "step": 26664 + }, + { + "epoch": 0.63, + "grad_norm": 1.104296942738874, + "learning_rate": 6.415180754939775e-06, + "loss": 0.8769, + "step": 26665 + }, + { + "epoch": 0.63, + "grad_norm": 1.1485249711480208, + "learning_rate": 6.414468438567722e-06, + "loss": 0.9499, + "step": 26666 + }, + { + "epoch": 0.63, + "grad_norm": 1.99914137184203, + "learning_rate": 6.413756143071765e-06, + "loss": 0.9384, + "step": 26667 + }, + { + "epoch": 0.63, + "grad_norm": 2.019763437786998, + "learning_rate": 6.413043868456057e-06, + "loss": 0.977, + "step": 26668 + }, + { + "epoch": 0.63, + "grad_norm": 2.3243678066197035, + "learning_rate": 6.4123316147247435e-06, + "loss": 0.9711, + "step": 26669 + }, + { + "epoch": 0.63, + "grad_norm": 1.975396465430924, + "learning_rate": 6.4116193818819735e-06, + "loss": 1.0307, + "step": 26670 + }, + { + "epoch": 0.63, + "grad_norm": 2.397642739971988, + "learning_rate": 6.410907169931888e-06, + "loss": 0.9821, + "step": 26671 + }, + { + "epoch": 0.63, + "grad_norm": 1.098218949036626, + "learning_rate": 6.410194978878638e-06, + "loss": 0.9439, + "step": 26672 + }, + { + "epoch": 0.63, + "grad_norm": 1.9218652772100224, + "learning_rate": 6.409482808726372e-06, + "loss": 0.9301, + "step": 26673 + }, + { + "epoch": 0.63, + "grad_norm": 2.304602680019411, + "learning_rate": 6.408770659479231e-06, + "loss": 0.9087, + "step": 26674 + }, + { + "epoch": 0.63, + "grad_norm": 1.9400588708615394, + "learning_rate": 6.408058531141363e-06, + "loss": 0.9654, + "step": 26675 + }, + { + "epoch": 0.63, + "grad_norm": 1.9997019508425076, + "learning_rate": 6.407346423716918e-06, + "loss": 0.9292, + "step": 26676 + }, + { + "epoch": 0.63, + "grad_norm": 1.0481221952984436, + "learning_rate": 6.406634337210041e-06, + "loss": 1.0074, + "step": 26677 + }, + { + "epoch": 0.63, + "grad_norm": 2.013783696915033, + "learning_rate": 6.405922271624874e-06, + "loss": 1.1231, + "step": 26678 + }, + { + "epoch": 0.63, + "grad_norm": 1.0828787706993765, + "learning_rate": 6.405210226965564e-06, + "loss": 1.0078, + "step": 26679 + }, + { + "epoch": 0.63, + "grad_norm": 2.2661418448589674, + "learning_rate": 6.404498203236261e-06, + "loss": 1.0728, + "step": 26680 + }, + { + "epoch": 0.63, + "grad_norm": 2.0211533703911124, + "learning_rate": 6.403786200441106e-06, + "loss": 0.9492, + "step": 26681 + }, + { + "epoch": 0.63, + "grad_norm": 1.1348246204721244, + "learning_rate": 6.403074218584248e-06, + "loss": 0.9802, + "step": 26682 + }, + { + "epoch": 0.63, + "grad_norm": 2.064680010441685, + "learning_rate": 6.402362257669827e-06, + "loss": 0.9293, + "step": 26683 + }, + { + "epoch": 0.63, + "grad_norm": 1.9860030010619825, + "learning_rate": 6.401650317701997e-06, + "loss": 1.0146, + "step": 26684 + }, + { + "epoch": 0.63, + "grad_norm": 2.28137433169674, + "learning_rate": 6.400938398684895e-06, + "loss": 1.0141, + "step": 26685 + }, + { + "epoch": 0.63, + "grad_norm": 2.1604685787174445, + "learning_rate": 6.40022650062267e-06, + "loss": 1.0052, + "step": 26686 + }, + { + "epoch": 0.63, + "grad_norm": 1.9833926072708794, + "learning_rate": 6.399514623519467e-06, + "loss": 1.0545, + "step": 26687 + }, + { + "epoch": 0.63, + "grad_norm": 1.991328220084571, + "learning_rate": 6.398802767379426e-06, + "loss": 1.0781, + "step": 26688 + }, + { + "epoch": 0.63, + "grad_norm": 1.9867924081530077, + "learning_rate": 6.398090932206698e-06, + "loss": 1.0531, + "step": 26689 + }, + { + "epoch": 0.63, + "grad_norm": 6.550921231665365, + "learning_rate": 6.397379118005423e-06, + "loss": 1.0622, + "step": 26690 + }, + { + "epoch": 0.63, + "grad_norm": 1.740427066616713, + "learning_rate": 6.396667324779754e-06, + "loss": 0.8958, + "step": 26691 + }, + { + "epoch": 0.63, + "grad_norm": 2.087783465811451, + "learning_rate": 6.395955552533821e-06, + "loss": 1.0274, + "step": 26692 + }, + { + "epoch": 0.63, + "grad_norm": 1.9777562347125508, + "learning_rate": 6.395243801271779e-06, + "loss": 0.9421, + "step": 26693 + }, + { + "epoch": 0.63, + "grad_norm": 1.88685687315517, + "learning_rate": 6.39453207099777e-06, + "loss": 0.9979, + "step": 26694 + }, + { + "epoch": 0.63, + "grad_norm": 2.12619291925362, + "learning_rate": 6.393820361715935e-06, + "loss": 1.0715, + "step": 26695 + }, + { + "epoch": 0.63, + "grad_norm": 1.9007469500617236, + "learning_rate": 6.393108673430418e-06, + "loss": 1.163, + "step": 26696 + }, + { + "epoch": 0.63, + "grad_norm": 1.7397259567989316, + "learning_rate": 6.392397006145366e-06, + "loss": 0.8594, + "step": 26697 + }, + { + "epoch": 0.63, + "grad_norm": 2.1475797476565135, + "learning_rate": 6.391685359864923e-06, + "loss": 0.9787, + "step": 26698 + }, + { + "epoch": 0.63, + "grad_norm": 2.3723267867704108, + "learning_rate": 6.390973734593227e-06, + "loss": 1.0028, + "step": 26699 + }, + { + "epoch": 0.63, + "grad_norm": 2.1526705067423033, + "learning_rate": 6.390262130334424e-06, + "loss": 1.0227, + "step": 26700 + }, + { + "epoch": 0.63, + "grad_norm": 2.260386695780777, + "learning_rate": 6.389550547092661e-06, + "loss": 1.0901, + "step": 26701 + }, + { + "epoch": 0.63, + "grad_norm": 1.825584295290862, + "learning_rate": 6.388838984872077e-06, + "loss": 0.8538, + "step": 26702 + }, + { + "epoch": 0.63, + "grad_norm": 1.930945278139287, + "learning_rate": 6.388127443676811e-06, + "loss": 0.963, + "step": 26703 + }, + { + "epoch": 0.63, + "grad_norm": 1.7635406533642368, + "learning_rate": 6.3874159235110144e-06, + "loss": 0.8302, + "step": 26704 + }, + { + "epoch": 0.63, + "grad_norm": 1.9863292995100572, + "learning_rate": 6.386704424378828e-06, + "loss": 1.0284, + "step": 26705 + }, + { + "epoch": 0.63, + "grad_norm": 2.023932021249923, + "learning_rate": 6.38599294628439e-06, + "loss": 0.9287, + "step": 26706 + }, + { + "epoch": 0.63, + "grad_norm": 2.3396803736368006, + "learning_rate": 6.385281489231843e-06, + "loss": 1.0192, + "step": 26707 + }, + { + "epoch": 0.63, + "grad_norm": 2.232452854755457, + "learning_rate": 6.384570053225338e-06, + "loss": 0.8408, + "step": 26708 + }, + { + "epoch": 0.63, + "grad_norm": 2.1134630770539924, + "learning_rate": 6.383858638269005e-06, + "loss": 1.0088, + "step": 26709 + }, + { + "epoch": 0.63, + "grad_norm": 1.820977665050037, + "learning_rate": 6.383147244366994e-06, + "loss": 1.0058, + "step": 26710 + }, + { + "epoch": 0.63, + "grad_norm": 2.3009284588459873, + "learning_rate": 6.382435871523442e-06, + "loss": 0.9547, + "step": 26711 + }, + { + "epoch": 0.63, + "grad_norm": 2.481126945193844, + "learning_rate": 6.3817245197425005e-06, + "loss": 0.9874, + "step": 26712 + }, + { + "epoch": 0.63, + "grad_norm": 1.9800096660206334, + "learning_rate": 6.3810131890282975e-06, + "loss": 1.2197, + "step": 26713 + }, + { + "epoch": 0.63, + "grad_norm": 1.9455824037823815, + "learning_rate": 6.3803018793849835e-06, + "loss": 0.9553, + "step": 26714 + }, + { + "epoch": 0.63, + "grad_norm": 2.065280194204469, + "learning_rate": 6.3795905908167e-06, + "loss": 1.0551, + "step": 26715 + }, + { + "epoch": 0.63, + "grad_norm": 1.9208064030570777, + "learning_rate": 6.378879323327583e-06, + "loss": 1.0439, + "step": 26716 + }, + { + "epoch": 0.63, + "grad_norm": 2.1522850307271155, + "learning_rate": 6.378168076921778e-06, + "loss": 1.0868, + "step": 26717 + }, + { + "epoch": 0.63, + "grad_norm": 2.1340055580648443, + "learning_rate": 6.377456851603426e-06, + "loss": 1.0052, + "step": 26718 + }, + { + "epoch": 0.63, + "grad_norm": 2.2711896386364994, + "learning_rate": 6.37674564737667e-06, + "loss": 1.0274, + "step": 26719 + }, + { + "epoch": 0.63, + "grad_norm": 2.1858216436629077, + "learning_rate": 6.376034464245641e-06, + "loss": 1.0397, + "step": 26720 + }, + { + "epoch": 0.63, + "grad_norm": 1.103793919863241, + "learning_rate": 6.375323302214491e-06, + "loss": 0.9492, + "step": 26721 + }, + { + "epoch": 0.63, + "grad_norm": 1.8911562777328825, + "learning_rate": 6.374612161287358e-06, + "loss": 0.912, + "step": 26722 + }, + { + "epoch": 0.63, + "grad_norm": 1.1382247063532094, + "learning_rate": 6.373901041468378e-06, + "loss": 0.9072, + "step": 26723 + }, + { + "epoch": 0.63, + "grad_norm": 1.8706148122953916, + "learning_rate": 6.373189942761693e-06, + "loss": 0.9648, + "step": 26724 + }, + { + "epoch": 0.63, + "grad_norm": 2.0686128087793416, + "learning_rate": 6.372478865171446e-06, + "loss": 0.9662, + "step": 26725 + }, + { + "epoch": 0.63, + "grad_norm": 1.9333978587441716, + "learning_rate": 6.371767808701778e-06, + "loss": 1.0939, + "step": 26726 + }, + { + "epoch": 0.63, + "grad_norm": 2.0886284413984275, + "learning_rate": 6.371056773356823e-06, + "loss": 0.9236, + "step": 26727 + }, + { + "epoch": 0.63, + "grad_norm": 1.9548450452149335, + "learning_rate": 6.370345759140724e-06, + "loss": 0.9295, + "step": 26728 + }, + { + "epoch": 0.63, + "grad_norm": 1.9453090490316323, + "learning_rate": 6.369634766057624e-06, + "loss": 1.1006, + "step": 26729 + }, + { + "epoch": 0.63, + "grad_norm": 1.982639894100845, + "learning_rate": 6.368923794111659e-06, + "loss": 0.9631, + "step": 26730 + }, + { + "epoch": 0.63, + "grad_norm": 2.18290604161694, + "learning_rate": 6.368212843306968e-06, + "loss": 0.9733, + "step": 26731 + }, + { + "epoch": 0.63, + "grad_norm": 1.8648860820292261, + "learning_rate": 6.36750191364769e-06, + "loss": 1.0444, + "step": 26732 + }, + { + "epoch": 0.63, + "grad_norm": 2.1029637830091055, + "learning_rate": 6.36679100513797e-06, + "loss": 1.0525, + "step": 26733 + }, + { + "epoch": 0.63, + "grad_norm": 1.9516646565520717, + "learning_rate": 6.366080117781941e-06, + "loss": 1.1185, + "step": 26734 + }, + { + "epoch": 0.63, + "grad_norm": 1.1176586586235167, + "learning_rate": 6.365369251583744e-06, + "loss": 0.9461, + "step": 26735 + }, + { + "epoch": 0.63, + "grad_norm": 2.7481755856532155, + "learning_rate": 6.364658406547519e-06, + "loss": 0.89, + "step": 26736 + }, + { + "epoch": 0.63, + "grad_norm": 1.9907239602126527, + "learning_rate": 6.363947582677401e-06, + "loss": 0.9632, + "step": 26737 + }, + { + "epoch": 0.63, + "grad_norm": 2.0723680809415357, + "learning_rate": 6.363236779977533e-06, + "loss": 1.0253, + "step": 26738 + }, + { + "epoch": 0.63, + "grad_norm": 2.4937944295189083, + "learning_rate": 6.362525998452051e-06, + "loss": 0.8126, + "step": 26739 + }, + { + "epoch": 0.63, + "grad_norm": 1.9591702473374695, + "learning_rate": 6.3618152381050975e-06, + "loss": 0.9058, + "step": 26740 + }, + { + "epoch": 0.63, + "grad_norm": 1.888466921660508, + "learning_rate": 6.3611044989408046e-06, + "loss": 1.0033, + "step": 26741 + }, + { + "epoch": 0.63, + "grad_norm": 2.258679035676054, + "learning_rate": 6.360393780963313e-06, + "loss": 1.0726, + "step": 26742 + }, + { + "epoch": 0.63, + "grad_norm": 1.9985397791069242, + "learning_rate": 6.359683084176764e-06, + "loss": 1.0701, + "step": 26743 + }, + { + "epoch": 0.63, + "grad_norm": 1.9983719938085889, + "learning_rate": 6.358972408585291e-06, + "loss": 1.0506, + "step": 26744 + }, + { + "epoch": 0.63, + "grad_norm": 2.0132901219386956, + "learning_rate": 6.358261754193032e-06, + "loss": 1.0008, + "step": 26745 + }, + { + "epoch": 0.63, + "grad_norm": 2.1396532080426085, + "learning_rate": 6.357551121004127e-06, + "loss": 0.9886, + "step": 26746 + }, + { + "epoch": 0.63, + "grad_norm": 2.30765785531763, + "learning_rate": 6.3568405090227125e-06, + "loss": 0.9629, + "step": 26747 + }, + { + "epoch": 0.63, + "grad_norm": 2.5005146278772923, + "learning_rate": 6.356129918252927e-06, + "loss": 0.9062, + "step": 26748 + }, + { + "epoch": 0.63, + "grad_norm": 2.0654815953633197, + "learning_rate": 6.3554193486989035e-06, + "loss": 0.9166, + "step": 26749 + }, + { + "epoch": 0.63, + "grad_norm": 1.087386461445137, + "learning_rate": 6.354708800364786e-06, + "loss": 0.9548, + "step": 26750 + }, + { + "epoch": 0.63, + "grad_norm": 2.2389543825168783, + "learning_rate": 6.353998273254707e-06, + "loss": 0.8422, + "step": 26751 + }, + { + "epoch": 0.63, + "grad_norm": 2.124695456184389, + "learning_rate": 6.353287767372802e-06, + "loss": 1.2176, + "step": 26752 + }, + { + "epoch": 0.63, + "grad_norm": 1.9382581603396265, + "learning_rate": 6.3525772827232115e-06, + "loss": 0.9273, + "step": 26753 + }, + { + "epoch": 0.63, + "grad_norm": 2.030379525262029, + "learning_rate": 6.351866819310073e-06, + "loss": 0.9234, + "step": 26754 + }, + { + "epoch": 0.63, + "grad_norm": 1.8704838156193984, + "learning_rate": 6.35115637713752e-06, + "loss": 1.0858, + "step": 26755 + }, + { + "epoch": 0.63, + "grad_norm": 1.899142437368334, + "learning_rate": 6.350445956209687e-06, + "loss": 1.0117, + "step": 26756 + }, + { + "epoch": 0.63, + "grad_norm": 1.8472882723725792, + "learning_rate": 6.349735556530718e-06, + "loss": 1.0298, + "step": 26757 + }, + { + "epoch": 0.63, + "grad_norm": 1.9732678768249412, + "learning_rate": 6.34902517810474e-06, + "loss": 0.96, + "step": 26758 + }, + { + "epoch": 0.63, + "grad_norm": 2.103734334119289, + "learning_rate": 6.348314820935895e-06, + "loss": 0.9513, + "step": 26759 + }, + { + "epoch": 0.63, + "grad_norm": 1.876841801806438, + "learning_rate": 6.347604485028314e-06, + "loss": 0.9796, + "step": 26760 + }, + { + "epoch": 0.63, + "grad_norm": 2.443782530906732, + "learning_rate": 6.346894170386143e-06, + "loss": 1.0597, + "step": 26761 + }, + { + "epoch": 0.63, + "grad_norm": 2.9271848837630645, + "learning_rate": 6.346183877013505e-06, + "loss": 0.9941, + "step": 26762 + }, + { + "epoch": 0.63, + "grad_norm": 1.7631312854168044, + "learning_rate": 6.345473604914542e-06, + "loss": 0.9799, + "step": 26763 + }, + { + "epoch": 0.63, + "grad_norm": 1.0946162846307481, + "learning_rate": 6.344763354093388e-06, + "loss": 0.8908, + "step": 26764 + }, + { + "epoch": 0.63, + "grad_norm": 1.9166399139690562, + "learning_rate": 6.344053124554183e-06, + "loss": 1.0685, + "step": 26765 + }, + { + "epoch": 0.63, + "grad_norm": 1.7966961412716227, + "learning_rate": 6.343342916301055e-06, + "loss": 0.8176, + "step": 26766 + }, + { + "epoch": 0.63, + "grad_norm": 1.9875382274461666, + "learning_rate": 6.342632729338143e-06, + "loss": 1.0243, + "step": 26767 + }, + { + "epoch": 0.63, + "grad_norm": 1.8951387779472806, + "learning_rate": 6.341922563669582e-06, + "loss": 0.9283, + "step": 26768 + }, + { + "epoch": 0.63, + "grad_norm": 2.58927963722353, + "learning_rate": 6.341212419299503e-06, + "loss": 1.0945, + "step": 26769 + }, + { + "epoch": 0.63, + "grad_norm": 1.9010037510527749, + "learning_rate": 6.340502296232046e-06, + "loss": 1.0743, + "step": 26770 + }, + { + "epoch": 0.63, + "grad_norm": 1.0999241079100774, + "learning_rate": 6.3397921944713435e-06, + "loss": 0.9631, + "step": 26771 + }, + { + "epoch": 0.63, + "grad_norm": 1.1070649589041537, + "learning_rate": 6.339082114021531e-06, + "loss": 0.9661, + "step": 26772 + }, + { + "epoch": 0.63, + "grad_norm": 1.8368131232375555, + "learning_rate": 6.3383720548867366e-06, + "loss": 1.0637, + "step": 26773 + }, + { + "epoch": 0.63, + "grad_norm": 1.8145561890901079, + "learning_rate": 6.3376620170711026e-06, + "loss": 0.9328, + "step": 26774 + }, + { + "epoch": 0.63, + "grad_norm": 1.8076010883135782, + "learning_rate": 6.336952000578761e-06, + "loss": 1.0072, + "step": 26775 + }, + { + "epoch": 0.63, + "grad_norm": 1.9925830829317877, + "learning_rate": 6.336242005413843e-06, + "loss": 1.0463, + "step": 26776 + }, + { + "epoch": 0.63, + "grad_norm": 1.809831666909575, + "learning_rate": 6.335532031580482e-06, + "loss": 0.9548, + "step": 26777 + }, + { + "epoch": 0.63, + "grad_norm": 2.0710094433238795, + "learning_rate": 6.334822079082815e-06, + "loss": 1.0844, + "step": 26778 + }, + { + "epoch": 0.63, + "grad_norm": 2.1438196612624014, + "learning_rate": 6.334112147924978e-06, + "loss": 1.1344, + "step": 26779 + }, + { + "epoch": 0.63, + "grad_norm": 1.8951735144614168, + "learning_rate": 6.333402238111097e-06, + "loss": 0.9255, + "step": 26780 + }, + { + "epoch": 0.63, + "grad_norm": 2.10948358020097, + "learning_rate": 6.332692349645307e-06, + "loss": 1.0833, + "step": 26781 + }, + { + "epoch": 0.63, + "grad_norm": 1.89747947344166, + "learning_rate": 6.331982482531748e-06, + "loss": 0.9505, + "step": 26782 + }, + { + "epoch": 0.63, + "grad_norm": 1.9344965332401012, + "learning_rate": 6.331272636774546e-06, + "loss": 0.9818, + "step": 26783 + }, + { + "epoch": 0.63, + "grad_norm": 1.882073126582259, + "learning_rate": 6.330562812377836e-06, + "loss": 0.976, + "step": 26784 + }, + { + "epoch": 0.63, + "grad_norm": 2.240574776675664, + "learning_rate": 6.329853009345748e-06, + "loss": 1.0687, + "step": 26785 + }, + { + "epoch": 0.63, + "grad_norm": 2.415766462134817, + "learning_rate": 6.329143227682423e-06, + "loss": 1.1436, + "step": 26786 + }, + { + "epoch": 0.63, + "grad_norm": 2.13149922160817, + "learning_rate": 6.328433467391986e-06, + "loss": 0.9389, + "step": 26787 + }, + { + "epoch": 0.63, + "grad_norm": 2.09502700588417, + "learning_rate": 6.3277237284785695e-06, + "loss": 1.1773, + "step": 26788 + }, + { + "epoch": 0.63, + "grad_norm": 1.9101951154897268, + "learning_rate": 6.327014010946315e-06, + "loss": 0.9721, + "step": 26789 + }, + { + "epoch": 0.63, + "grad_norm": 1.9426552514094224, + "learning_rate": 6.326304314799342e-06, + "loss": 0.9378, + "step": 26790 + }, + { + "epoch": 0.63, + "grad_norm": 2.053104367290389, + "learning_rate": 6.32559464004179e-06, + "loss": 0.992, + "step": 26791 + }, + { + "epoch": 0.63, + "grad_norm": 1.9084545401075492, + "learning_rate": 6.324884986677786e-06, + "loss": 1.0717, + "step": 26792 + }, + { + "epoch": 0.63, + "grad_norm": 2.164944993411514, + "learning_rate": 6.324175354711473e-06, + "loss": 1.0398, + "step": 26793 + }, + { + "epoch": 0.63, + "grad_norm": 1.9722404681520582, + "learning_rate": 6.323465744146967e-06, + "loss": 1.0012, + "step": 26794 + }, + { + "epoch": 0.63, + "grad_norm": 2.1505239704774315, + "learning_rate": 6.32275615498841e-06, + "loss": 1.0705, + "step": 26795 + }, + { + "epoch": 0.63, + "grad_norm": 2.438079429408983, + "learning_rate": 6.322046587239934e-06, + "loss": 0.9482, + "step": 26796 + }, + { + "epoch": 0.63, + "grad_norm": 1.896767353549462, + "learning_rate": 6.321337040905664e-06, + "loss": 0.8381, + "step": 26797 + }, + { + "epoch": 0.63, + "grad_norm": 2.078063942099204, + "learning_rate": 6.320627515989733e-06, + "loss": 1.0182, + "step": 26798 + }, + { + "epoch": 0.63, + "grad_norm": 2.127916172220132, + "learning_rate": 6.319918012496276e-06, + "loss": 1.0184, + "step": 26799 + }, + { + "epoch": 0.63, + "grad_norm": 1.857335933580865, + "learning_rate": 6.319208530429422e-06, + "loss": 0.9736, + "step": 26800 + }, + { + "epoch": 0.63, + "grad_norm": 2.193218269654082, + "learning_rate": 6.318499069793298e-06, + "loss": 1.0222, + "step": 26801 + }, + { + "epoch": 0.63, + "grad_norm": 1.8207809416983374, + "learning_rate": 6.317789630592039e-06, + "loss": 0.9847, + "step": 26802 + }, + { + "epoch": 0.63, + "grad_norm": 1.9080035541866462, + "learning_rate": 6.317080212829778e-06, + "loss": 1.0308, + "step": 26803 + }, + { + "epoch": 0.63, + "grad_norm": 2.0907410908861457, + "learning_rate": 6.31637081651064e-06, + "loss": 0.9946, + "step": 26804 + }, + { + "epoch": 0.63, + "grad_norm": 1.852124802799553, + "learning_rate": 6.315661441638754e-06, + "loss": 0.9289, + "step": 26805 + }, + { + "epoch": 0.63, + "grad_norm": 2.087154381173231, + "learning_rate": 6.314952088218256e-06, + "loss": 0.9737, + "step": 26806 + }, + { + "epoch": 0.63, + "grad_norm": 2.3916865231893247, + "learning_rate": 6.314242756253275e-06, + "loss": 0.9364, + "step": 26807 + }, + { + "epoch": 0.63, + "grad_norm": 2.054924770124837, + "learning_rate": 6.313533445747939e-06, + "loss": 1.1261, + "step": 26808 + }, + { + "epoch": 0.63, + "grad_norm": 1.135893464263834, + "learning_rate": 6.312824156706376e-06, + "loss": 0.978, + "step": 26809 + }, + { + "epoch": 0.63, + "grad_norm": 1.9267249626592065, + "learning_rate": 6.312114889132721e-06, + "loss": 1.0012, + "step": 26810 + }, + { + "epoch": 0.63, + "grad_norm": 1.9913209030745964, + "learning_rate": 6.311405643031101e-06, + "loss": 1.1365, + "step": 26811 + }, + { + "epoch": 0.63, + "grad_norm": 1.844302430828664, + "learning_rate": 6.310696418405643e-06, + "loss": 0.9535, + "step": 26812 + }, + { + "epoch": 0.63, + "grad_norm": 2.3719415237090526, + "learning_rate": 6.309987215260477e-06, + "loss": 1.1424, + "step": 26813 + }, + { + "epoch": 0.63, + "grad_norm": 2.4910419136482895, + "learning_rate": 6.309278033599738e-06, + "loss": 1.0213, + "step": 26814 + }, + { + "epoch": 0.63, + "grad_norm": 1.1303863425493044, + "learning_rate": 6.3085688734275485e-06, + "loss": 0.9302, + "step": 26815 + }, + { + "epoch": 0.63, + "grad_norm": 1.9300509663849987, + "learning_rate": 6.307859734748039e-06, + "loss": 1.0032, + "step": 26816 + }, + { + "epoch": 0.63, + "grad_norm": 1.8149629499379605, + "learning_rate": 6.307150617565343e-06, + "loss": 1.0503, + "step": 26817 + }, + { + "epoch": 0.63, + "grad_norm": 2.0526196514546036, + "learning_rate": 6.3064415218835795e-06, + "loss": 0.9142, + "step": 26818 + }, + { + "epoch": 0.63, + "grad_norm": 2.03149030735721, + "learning_rate": 6.305732447706885e-06, + "loss": 1.0392, + "step": 26819 + }, + { + "epoch": 0.63, + "grad_norm": 2.0834430155777457, + "learning_rate": 6.305023395039387e-06, + "loss": 1.0257, + "step": 26820 + }, + { + "epoch": 0.63, + "grad_norm": 2.029085035345832, + "learning_rate": 6.3043143638852134e-06, + "loss": 0.9777, + "step": 26821 + }, + { + "epoch": 0.63, + "grad_norm": 1.899498603417705, + "learning_rate": 6.303605354248488e-06, + "loss": 0.9773, + "step": 26822 + }, + { + "epoch": 0.63, + "grad_norm": 2.5740720573423905, + "learning_rate": 6.302896366133344e-06, + "loss": 1.0225, + "step": 26823 + }, + { + "epoch": 0.63, + "grad_norm": 1.964244734570357, + "learning_rate": 6.302187399543911e-06, + "loss": 1.1411, + "step": 26824 + }, + { + "epoch": 0.63, + "grad_norm": 2.1543010789876886, + "learning_rate": 6.301478454484311e-06, + "loss": 0.927, + "step": 26825 + }, + { + "epoch": 0.63, + "grad_norm": 2.0775986458397497, + "learning_rate": 6.300769530958673e-06, + "loss": 0.9505, + "step": 26826 + }, + { + "epoch": 0.63, + "grad_norm": 2.117411184745405, + "learning_rate": 6.300060628971127e-06, + "loss": 0.8295, + "step": 26827 + }, + { + "epoch": 0.63, + "grad_norm": 1.8210416459579448, + "learning_rate": 6.299351748525802e-06, + "loss": 1.086, + "step": 26828 + }, + { + "epoch": 0.63, + "grad_norm": 1.7838380940205159, + "learning_rate": 6.298642889626821e-06, + "loss": 0.9137, + "step": 26829 + }, + { + "epoch": 0.63, + "grad_norm": 1.9310097785246485, + "learning_rate": 6.297934052278309e-06, + "loss": 0.9365, + "step": 26830 + }, + { + "epoch": 0.63, + "grad_norm": 1.966115060924933, + "learning_rate": 6.297225236484402e-06, + "loss": 1.049, + "step": 26831 + }, + { + "epoch": 0.63, + "grad_norm": 2.0270512152528597, + "learning_rate": 6.296516442249221e-06, + "loss": 0.9714, + "step": 26832 + }, + { + "epoch": 0.63, + "grad_norm": 1.9620985982275418, + "learning_rate": 6.295807669576892e-06, + "loss": 0.9481, + "step": 26833 + }, + { + "epoch": 0.63, + "grad_norm": 2.3181425087977985, + "learning_rate": 6.295098918471543e-06, + "loss": 1.0357, + "step": 26834 + }, + { + "epoch": 0.63, + "grad_norm": 1.8714367956711266, + "learning_rate": 6.294390188937305e-06, + "loss": 0.9117, + "step": 26835 + }, + { + "epoch": 0.63, + "grad_norm": 2.0280906606239077, + "learning_rate": 6.293681480978298e-06, + "loss": 0.9561, + "step": 26836 + }, + { + "epoch": 0.63, + "grad_norm": 2.0068579426600848, + "learning_rate": 6.292972794598649e-06, + "loss": 1.0656, + "step": 26837 + }, + { + "epoch": 0.63, + "grad_norm": 2.252346510762234, + "learning_rate": 6.292264129802492e-06, + "loss": 0.9208, + "step": 26838 + }, + { + "epoch": 0.63, + "grad_norm": 2.2172619666798243, + "learning_rate": 6.291555486593941e-06, + "loss": 0.9227, + "step": 26839 + }, + { + "epoch": 0.63, + "grad_norm": 1.1025376801359636, + "learning_rate": 6.290846864977129e-06, + "loss": 0.9017, + "step": 26840 + }, + { + "epoch": 0.63, + "grad_norm": 1.984322716475203, + "learning_rate": 6.290138264956181e-06, + "loss": 1.0187, + "step": 26841 + }, + { + "epoch": 0.63, + "grad_norm": 1.9983378808567853, + "learning_rate": 6.289429686535226e-06, + "loss": 0.9719, + "step": 26842 + }, + { + "epoch": 0.63, + "grad_norm": 2.3696487138235915, + "learning_rate": 6.2887211297183825e-06, + "loss": 0.9214, + "step": 26843 + }, + { + "epoch": 0.63, + "grad_norm": 2.061080118419565, + "learning_rate": 6.28801259450978e-06, + "loss": 1.0495, + "step": 26844 + }, + { + "epoch": 0.63, + "grad_norm": 1.9344493851279472, + "learning_rate": 6.2873040809135465e-06, + "loss": 1.0644, + "step": 26845 + }, + { + "epoch": 0.63, + "grad_norm": 1.9841703485150268, + "learning_rate": 6.286595588933802e-06, + "loss": 1.0491, + "step": 26846 + }, + { + "epoch": 0.63, + "grad_norm": 1.85561926218724, + "learning_rate": 6.285887118574671e-06, + "loss": 0.9807, + "step": 26847 + }, + { + "epoch": 0.63, + "grad_norm": 1.9098756792294442, + "learning_rate": 6.285178669840284e-06, + "loss": 1.0027, + "step": 26848 + }, + { + "epoch": 0.63, + "grad_norm": 1.9157605182435449, + "learning_rate": 6.284470242734764e-06, + "loss": 1.0223, + "step": 26849 + }, + { + "epoch": 0.63, + "grad_norm": 1.8070846254313753, + "learning_rate": 6.28376183726223e-06, + "loss": 1.0061, + "step": 26850 + }, + { + "epoch": 0.63, + "grad_norm": 1.8846273813789742, + "learning_rate": 6.283053453426813e-06, + "loss": 0.9945, + "step": 26851 + }, + { + "epoch": 0.63, + "grad_norm": 2.5089293832690047, + "learning_rate": 6.282345091232639e-06, + "loss": 0.9942, + "step": 26852 + }, + { + "epoch": 0.63, + "grad_norm": 2.1999926391188525, + "learning_rate": 6.2816367506838265e-06, + "loss": 0.96, + "step": 26853 + }, + { + "epoch": 0.63, + "grad_norm": 2.2472920595386734, + "learning_rate": 6.280928431784499e-06, + "loss": 0.8936, + "step": 26854 + }, + { + "epoch": 0.63, + "grad_norm": 1.954072749861025, + "learning_rate": 6.280220134538787e-06, + "loss": 1.0233, + "step": 26855 + }, + { + "epoch": 0.63, + "grad_norm": 1.9278586194447564, + "learning_rate": 6.279511858950813e-06, + "loss": 0.8959, + "step": 26856 + }, + { + "epoch": 0.63, + "grad_norm": 2.038851678320436, + "learning_rate": 6.278803605024695e-06, + "loss": 1.0162, + "step": 26857 + }, + { + "epoch": 0.63, + "grad_norm": 2.24699091113075, + "learning_rate": 6.278095372764561e-06, + "loss": 0.8977, + "step": 26858 + }, + { + "epoch": 0.63, + "grad_norm": 2.1389954341795643, + "learning_rate": 6.277387162174536e-06, + "loss": 1.0124, + "step": 26859 + }, + { + "epoch": 0.63, + "grad_norm": 1.0663700268998377, + "learning_rate": 6.276678973258741e-06, + "loss": 0.9291, + "step": 26860 + }, + { + "epoch": 0.63, + "grad_norm": 2.292279302385372, + "learning_rate": 6.2759708060212985e-06, + "loss": 1.0386, + "step": 26861 + }, + { + "epoch": 0.63, + "grad_norm": 1.9245812021666815, + "learning_rate": 6.275262660466333e-06, + "loss": 0.9749, + "step": 26862 + }, + { + "epoch": 0.63, + "grad_norm": 1.9895723904807845, + "learning_rate": 6.274554536597969e-06, + "loss": 0.9824, + "step": 26863 + }, + { + "epoch": 0.63, + "grad_norm": 1.9822311317824564, + "learning_rate": 6.273846434420328e-06, + "loss": 0.9604, + "step": 26864 + }, + { + "epoch": 0.63, + "grad_norm": 1.8027237730913046, + "learning_rate": 6.2731383539375315e-06, + "loss": 1.0361, + "step": 26865 + }, + { + "epoch": 0.63, + "grad_norm": 1.0838998609107497, + "learning_rate": 6.272430295153705e-06, + "loss": 0.9779, + "step": 26866 + }, + { + "epoch": 0.63, + "grad_norm": 2.0930520235833923, + "learning_rate": 6.2717222580729665e-06, + "loss": 0.988, + "step": 26867 + }, + { + "epoch": 0.63, + "grad_norm": 2.000248823203481, + "learning_rate": 6.271014242699442e-06, + "loss": 0.9784, + "step": 26868 + }, + { + "epoch": 0.63, + "grad_norm": 1.9272574919606875, + "learning_rate": 6.270306249037254e-06, + "loss": 1.0985, + "step": 26869 + }, + { + "epoch": 0.63, + "grad_norm": 2.316953567460207, + "learning_rate": 6.269598277090527e-06, + "loss": 1.07, + "step": 26870 + }, + { + "epoch": 0.63, + "grad_norm": 1.8027660338937537, + "learning_rate": 6.268890326863375e-06, + "loss": 1.022, + "step": 26871 + }, + { + "epoch": 0.63, + "grad_norm": 1.875372831336271, + "learning_rate": 6.268182398359927e-06, + "loss": 1.0275, + "step": 26872 + }, + { + "epoch": 0.63, + "grad_norm": 2.4345612196785016, + "learning_rate": 6.267474491584304e-06, + "loss": 1.0645, + "step": 26873 + }, + { + "epoch": 0.63, + "grad_norm": 1.9617036316876633, + "learning_rate": 6.266766606540624e-06, + "loss": 0.9452, + "step": 26874 + }, + { + "epoch": 0.63, + "grad_norm": 2.334827692911877, + "learning_rate": 6.26605874323301e-06, + "loss": 1.0372, + "step": 26875 + }, + { + "epoch": 0.63, + "grad_norm": 1.9979763482344537, + "learning_rate": 6.265350901665586e-06, + "loss": 0.9023, + "step": 26876 + }, + { + "epoch": 0.63, + "grad_norm": 1.9745057731601918, + "learning_rate": 6.264643081842472e-06, + "loss": 0.9752, + "step": 26877 + }, + { + "epoch": 0.63, + "grad_norm": 2.0177902516838784, + "learning_rate": 6.263935283767787e-06, + "loss": 1.0232, + "step": 26878 + }, + { + "epoch": 0.63, + "grad_norm": 1.0730663451416602, + "learning_rate": 6.2632275074456515e-06, + "loss": 0.9004, + "step": 26879 + }, + { + "epoch": 0.63, + "grad_norm": 2.1192376111289986, + "learning_rate": 6.262519752880193e-06, + "loss": 0.9918, + "step": 26880 + }, + { + "epoch": 0.63, + "grad_norm": 2.12200890055911, + "learning_rate": 6.2618120200755264e-06, + "loss": 0.9675, + "step": 26881 + }, + { + "epoch": 0.63, + "grad_norm": 2.1901409254970488, + "learning_rate": 6.261104309035772e-06, + "loss": 1.0947, + "step": 26882 + }, + { + "epoch": 0.63, + "grad_norm": 1.9328024900895346, + "learning_rate": 6.260396619765051e-06, + "loss": 0.9423, + "step": 26883 + }, + { + "epoch": 0.63, + "grad_norm": 2.127516115865141, + "learning_rate": 6.259688952267489e-06, + "loss": 1.1253, + "step": 26884 + }, + { + "epoch": 0.63, + "grad_norm": 2.2927944545203394, + "learning_rate": 6.2589813065472005e-06, + "loss": 0.9115, + "step": 26885 + }, + { + "epoch": 0.63, + "grad_norm": 1.8875740853407477, + "learning_rate": 6.258273682608304e-06, + "loss": 0.688, + "step": 26886 + }, + { + "epoch": 0.63, + "grad_norm": 1.985086591294844, + "learning_rate": 6.257566080454925e-06, + "loss": 0.8781, + "step": 26887 + }, + { + "epoch": 0.63, + "grad_norm": 1.7774643908156247, + "learning_rate": 6.256858500091184e-06, + "loss": 0.8966, + "step": 26888 + }, + { + "epoch": 0.63, + "grad_norm": 2.097416184938361, + "learning_rate": 6.256150941521195e-06, + "loss": 1.0814, + "step": 26889 + }, + { + "epoch": 0.63, + "grad_norm": 1.97095336568175, + "learning_rate": 6.25544340474908e-06, + "loss": 0.9952, + "step": 26890 + }, + { + "epoch": 0.63, + "grad_norm": 1.8583068957913047, + "learning_rate": 6.2547358897789625e-06, + "loss": 1.0941, + "step": 26891 + }, + { + "epoch": 0.63, + "grad_norm": 2.553846671105853, + "learning_rate": 6.254028396614954e-06, + "loss": 1.0337, + "step": 26892 + }, + { + "epoch": 0.63, + "grad_norm": 2.0039215194202193, + "learning_rate": 6.2533209252611795e-06, + "loss": 1.0798, + "step": 26893 + }, + { + "epoch": 0.63, + "grad_norm": 1.7910269226120372, + "learning_rate": 6.252613475721756e-06, + "loss": 0.9359, + "step": 26894 + }, + { + "epoch": 0.63, + "grad_norm": 2.0136105823970523, + "learning_rate": 6.251906048000807e-06, + "loss": 0.8282, + "step": 26895 + }, + { + "epoch": 0.63, + "grad_norm": 2.157009699594481, + "learning_rate": 6.251198642102444e-06, + "loss": 0.883, + "step": 26896 + }, + { + "epoch": 0.63, + "grad_norm": 2.1626126474458207, + "learning_rate": 6.2504912580307905e-06, + "loss": 0.8161, + "step": 26897 + }, + { + "epoch": 0.63, + "grad_norm": 1.8947875739257516, + "learning_rate": 6.2497838957899645e-06, + "loss": 0.9669, + "step": 26898 + }, + { + "epoch": 0.63, + "grad_norm": 6.01211466060327, + "learning_rate": 6.249076555384082e-06, + "loss": 1.1173, + "step": 26899 + }, + { + "epoch": 0.63, + "grad_norm": 2.1991424533549164, + "learning_rate": 6.248369236817265e-06, + "loss": 0.973, + "step": 26900 + }, + { + "epoch": 0.63, + "grad_norm": 1.7649504915509209, + "learning_rate": 6.247661940093629e-06, + "loss": 0.9229, + "step": 26901 + }, + { + "epoch": 0.63, + "grad_norm": 2.2672300671262677, + "learning_rate": 6.246954665217295e-06, + "loss": 1.0357, + "step": 26902 + }, + { + "epoch": 0.63, + "grad_norm": 2.052741957309305, + "learning_rate": 6.2462474121923765e-06, + "loss": 0.9717, + "step": 26903 + }, + { + "epoch": 0.63, + "grad_norm": 1.877746539191077, + "learning_rate": 6.245540181022996e-06, + "loss": 0.9525, + "step": 26904 + }, + { + "epoch": 0.63, + "grad_norm": 1.9618936629490382, + "learning_rate": 6.2448329717132705e-06, + "loss": 0.9528, + "step": 26905 + }, + { + "epoch": 0.63, + "grad_norm": 2.126083028733555, + "learning_rate": 6.244125784267314e-06, + "loss": 0.9151, + "step": 26906 + }, + { + "epoch": 0.63, + "grad_norm": 2.1687022560981806, + "learning_rate": 6.243418618689246e-06, + "loss": 0.8927, + "step": 26907 + }, + { + "epoch": 0.63, + "grad_norm": 2.01485442247112, + "learning_rate": 6.242711474983183e-06, + "loss": 0.907, + "step": 26908 + }, + { + "epoch": 0.63, + "grad_norm": 2.9086832649465384, + "learning_rate": 6.242004353153248e-06, + "loss": 1.0596, + "step": 26909 + }, + { + "epoch": 0.63, + "grad_norm": 2.0925035192477193, + "learning_rate": 6.241297253203551e-06, + "loss": 1.0945, + "step": 26910 + }, + { + "epoch": 0.63, + "grad_norm": 3.8823929258726326, + "learning_rate": 6.240590175138208e-06, + "loss": 0.9324, + "step": 26911 + }, + { + "epoch": 0.63, + "grad_norm": 2.1349025258808734, + "learning_rate": 6.239883118961344e-06, + "loss": 1.0429, + "step": 26912 + }, + { + "epoch": 0.63, + "grad_norm": 1.8043779976941987, + "learning_rate": 6.23917608467707e-06, + "loss": 0.9811, + "step": 26913 + }, + { + "epoch": 0.63, + "grad_norm": 1.8623341777818996, + "learning_rate": 6.2384690722895016e-06, + "loss": 1.0849, + "step": 26914 + }, + { + "epoch": 0.63, + "grad_norm": 1.987932181994479, + "learning_rate": 6.237762081802757e-06, + "loss": 0.9488, + "step": 26915 + }, + { + "epoch": 0.63, + "grad_norm": 2.1091110976139307, + "learning_rate": 6.237055113220955e-06, + "loss": 1.0143, + "step": 26916 + }, + { + "epoch": 0.63, + "grad_norm": 2.3532771324076482, + "learning_rate": 6.236348166548208e-06, + "loss": 0.9887, + "step": 26917 + }, + { + "epoch": 0.63, + "grad_norm": 1.1977880090492043, + "learning_rate": 6.2356412417886335e-06, + "loss": 1.0062, + "step": 26918 + }, + { + "epoch": 0.63, + "grad_norm": 2.113672552915428, + "learning_rate": 6.23493433894635e-06, + "loss": 0.8738, + "step": 26919 + }, + { + "epoch": 0.63, + "grad_norm": 2.1533524046991186, + "learning_rate": 6.234227458025466e-06, + "loss": 1.074, + "step": 26920 + }, + { + "epoch": 0.63, + "grad_norm": 1.9397174585682528, + "learning_rate": 6.233520599030105e-06, + "loss": 1.0121, + "step": 26921 + }, + { + "epoch": 0.63, + "grad_norm": 2.177236724665678, + "learning_rate": 6.232813761964377e-06, + "loss": 0.9863, + "step": 26922 + }, + { + "epoch": 0.63, + "grad_norm": 1.80881914720415, + "learning_rate": 6.232106946832406e-06, + "loss": 1.08, + "step": 26923 + }, + { + "epoch": 0.63, + "grad_norm": 2.182809502547769, + "learning_rate": 6.231400153638294e-06, + "loss": 0.9493, + "step": 26924 + }, + { + "epoch": 0.63, + "grad_norm": 2.374918656942362, + "learning_rate": 6.230693382386167e-06, + "loss": 0.8557, + "step": 26925 + }, + { + "epoch": 0.63, + "grad_norm": 1.9941514084493592, + "learning_rate": 6.229986633080138e-06, + "loss": 0.9448, + "step": 26926 + }, + { + "epoch": 0.63, + "grad_norm": 2.0511274423628656, + "learning_rate": 6.229279905724319e-06, + "loss": 1.068, + "step": 26927 + }, + { + "epoch": 0.63, + "grad_norm": 2.369338243942641, + "learning_rate": 6.228573200322823e-06, + "loss": 0.9598, + "step": 26928 + }, + { + "epoch": 0.63, + "grad_norm": 2.1266597208494797, + "learning_rate": 6.22786651687977e-06, + "loss": 0.9964, + "step": 26929 + }, + { + "epoch": 0.63, + "grad_norm": 1.9331001981204246, + "learning_rate": 6.227159855399276e-06, + "loss": 1.046, + "step": 26930 + }, + { + "epoch": 0.63, + "grad_norm": 2.195429939750855, + "learning_rate": 6.226453215885448e-06, + "loss": 1.0377, + "step": 26931 + }, + { + "epoch": 0.63, + "grad_norm": 1.8953597836667873, + "learning_rate": 6.225746598342402e-06, + "loss": 1.0317, + "step": 26932 + }, + { + "epoch": 0.63, + "grad_norm": 1.9349149719986456, + "learning_rate": 6.22504000277426e-06, + "loss": 0.9017, + "step": 26933 + }, + { + "epoch": 0.63, + "grad_norm": 1.754914617011432, + "learning_rate": 6.224333429185127e-06, + "loss": 0.8579, + "step": 26934 + }, + { + "epoch": 0.63, + "grad_norm": 2.3426212688319596, + "learning_rate": 6.223626877579119e-06, + "loss": 0.95, + "step": 26935 + }, + { + "epoch": 0.63, + "grad_norm": 2.3179332259440972, + "learning_rate": 6.22292034796035e-06, + "loss": 1.0171, + "step": 26936 + }, + { + "epoch": 0.63, + "grad_norm": 3.2307207627345758, + "learning_rate": 6.22221384033294e-06, + "loss": 0.9911, + "step": 26937 + }, + { + "epoch": 0.63, + "grad_norm": 2.187422376877566, + "learning_rate": 6.221507354700993e-06, + "loss": 1.0265, + "step": 26938 + }, + { + "epoch": 0.63, + "grad_norm": 1.958074577125535, + "learning_rate": 6.2208008910686255e-06, + "loss": 1.1263, + "step": 26939 + }, + { + "epoch": 0.63, + "grad_norm": 0.9883516442300944, + "learning_rate": 6.220094449439957e-06, + "loss": 0.9797, + "step": 26940 + }, + { + "epoch": 0.63, + "grad_norm": 2.2097646389659467, + "learning_rate": 6.21938802981909e-06, + "loss": 0.9985, + "step": 26941 + }, + { + "epoch": 0.63, + "grad_norm": 1.0736566074080787, + "learning_rate": 6.218681632210144e-06, + "loss": 0.8803, + "step": 26942 + }, + { + "epoch": 0.63, + "grad_norm": 2.105424194890726, + "learning_rate": 6.21797525661723e-06, + "loss": 0.9506, + "step": 26943 + }, + { + "epoch": 0.63, + "grad_norm": 2.103297461074203, + "learning_rate": 6.217268903044467e-06, + "loss": 1.0531, + "step": 26944 + }, + { + "epoch": 0.63, + "grad_norm": 2.8655710219218125, + "learning_rate": 6.2165625714959545e-06, + "loss": 0.9663, + "step": 26945 + }, + { + "epoch": 0.63, + "grad_norm": 1.791440394937485, + "learning_rate": 6.215856261975816e-06, + "loss": 0.9079, + "step": 26946 + }, + { + "epoch": 0.63, + "grad_norm": 2.000306758147267, + "learning_rate": 6.215149974488161e-06, + "loss": 0.9148, + "step": 26947 + }, + { + "epoch": 0.63, + "grad_norm": 2.2015062664578453, + "learning_rate": 6.214443709037099e-06, + "loss": 1.0059, + "step": 26948 + }, + { + "epoch": 0.63, + "grad_norm": 2.204751518252436, + "learning_rate": 6.213737465626746e-06, + "loss": 1.0237, + "step": 26949 + }, + { + "epoch": 0.63, + "grad_norm": 2.2479716489511277, + "learning_rate": 6.213031244261212e-06, + "loss": 1.0065, + "step": 26950 + }, + { + "epoch": 0.63, + "grad_norm": 2.2146184352948874, + "learning_rate": 6.212325044944609e-06, + "loss": 0.9119, + "step": 26951 + }, + { + "epoch": 0.63, + "grad_norm": 2.0446163099941015, + "learning_rate": 6.2116188676810466e-06, + "loss": 1.0619, + "step": 26952 + }, + { + "epoch": 0.63, + "grad_norm": 3.170809430270333, + "learning_rate": 6.2109127124746405e-06, + "loss": 0.9193, + "step": 26953 + }, + { + "epoch": 0.64, + "grad_norm": 1.070596618732332, + "learning_rate": 6.210206579329502e-06, + "loss": 0.9432, + "step": 26954 + }, + { + "epoch": 0.64, + "grad_norm": 1.1028540765630785, + "learning_rate": 6.209500468249738e-06, + "loss": 0.9778, + "step": 26955 + }, + { + "epoch": 0.64, + "grad_norm": 2.185348991990783, + "learning_rate": 6.208794379239462e-06, + "loss": 0.9607, + "step": 26956 + }, + { + "epoch": 0.64, + "grad_norm": 2.052855984382149, + "learning_rate": 6.208088312302785e-06, + "loss": 0.8971, + "step": 26957 + }, + { + "epoch": 0.64, + "grad_norm": 1.9353249885325763, + "learning_rate": 6.207382267443823e-06, + "loss": 1.0002, + "step": 26958 + }, + { + "epoch": 0.64, + "grad_norm": 2.3048137000352864, + "learning_rate": 6.2066762446666785e-06, + "loss": 1.0492, + "step": 26959 + }, + { + "epoch": 0.64, + "grad_norm": 1.734648582603258, + "learning_rate": 6.205970243975464e-06, + "loss": 0.8828, + "step": 26960 + }, + { + "epoch": 0.64, + "grad_norm": 1.978716571531746, + "learning_rate": 6.205264265374295e-06, + "loss": 0.9637, + "step": 26961 + }, + { + "epoch": 0.64, + "grad_norm": 2.017037567533826, + "learning_rate": 6.2045583088672785e-06, + "loss": 1.0796, + "step": 26962 + }, + { + "epoch": 0.64, + "grad_norm": 2.340549216966251, + "learning_rate": 6.2038523744585254e-06, + "loss": 0.9235, + "step": 26963 + }, + { + "epoch": 0.64, + "grad_norm": 2.033507470906669, + "learning_rate": 6.203146462152142e-06, + "loss": 0.9765, + "step": 26964 + }, + { + "epoch": 0.64, + "grad_norm": 2.1780528946444115, + "learning_rate": 6.202440571952246e-06, + "loss": 1.0197, + "step": 26965 + }, + { + "epoch": 0.64, + "grad_norm": 1.9213382169522606, + "learning_rate": 6.201734703862942e-06, + "loss": 0.9835, + "step": 26966 + }, + { + "epoch": 0.64, + "grad_norm": 2.172515544626889, + "learning_rate": 6.201028857888342e-06, + "loss": 1.0194, + "step": 26967 + }, + { + "epoch": 0.64, + "grad_norm": 2.027328363559563, + "learning_rate": 6.200323034032556e-06, + "loss": 1.0415, + "step": 26968 + }, + { + "epoch": 0.64, + "grad_norm": 1.9103204422818325, + "learning_rate": 6.1996172322996885e-06, + "loss": 0.9201, + "step": 26969 + }, + { + "epoch": 0.64, + "grad_norm": 1.9617750924788717, + "learning_rate": 6.1989114526938535e-06, + "loss": 1.0448, + "step": 26970 + }, + { + "epoch": 0.64, + "grad_norm": 2.056339885238922, + "learning_rate": 6.198205695219157e-06, + "loss": 0.9481, + "step": 26971 + }, + { + "epoch": 0.64, + "grad_norm": 2.294453187117257, + "learning_rate": 6.197499959879718e-06, + "loss": 1.0086, + "step": 26972 + }, + { + "epoch": 0.64, + "grad_norm": 2.260917243194523, + "learning_rate": 6.1967942466796315e-06, + "loss": 0.9816, + "step": 26973 + }, + { + "epoch": 0.64, + "grad_norm": 1.7902759449503576, + "learning_rate": 6.196088555623016e-06, + "loss": 1.1419, + "step": 26974 + }, + { + "epoch": 0.64, + "grad_norm": 2.199388840362798, + "learning_rate": 6.1953828867139784e-06, + "loss": 1.0498, + "step": 26975 + }, + { + "epoch": 0.64, + "grad_norm": 1.9483693564769922, + "learning_rate": 6.194677239956625e-06, + "loss": 0.9, + "step": 26976 + }, + { + "epoch": 0.64, + "grad_norm": 2.0206067178366602, + "learning_rate": 6.193971615355062e-06, + "loss": 1.0227, + "step": 26977 + }, + { + "epoch": 0.64, + "grad_norm": 2.9821085459646253, + "learning_rate": 6.1932660129134035e-06, + "loss": 0.8701, + "step": 26978 + }, + { + "epoch": 0.64, + "grad_norm": 1.9626161855557611, + "learning_rate": 6.1925604326357595e-06, + "loss": 1.0442, + "step": 26979 + }, + { + "epoch": 0.64, + "grad_norm": 2.26494413413602, + "learning_rate": 6.191854874526229e-06, + "loss": 0.8577, + "step": 26980 + }, + { + "epoch": 0.64, + "grad_norm": 2.1445393418852516, + "learning_rate": 6.191149338588924e-06, + "loss": 0.9421, + "step": 26981 + }, + { + "epoch": 0.64, + "grad_norm": 2.4317559037784657, + "learning_rate": 6.190443824827958e-06, + "loss": 1.0397, + "step": 26982 + }, + { + "epoch": 0.64, + "grad_norm": 3.5029197291731697, + "learning_rate": 6.189738333247433e-06, + "loss": 0.9839, + "step": 26983 + }, + { + "epoch": 0.64, + "grad_norm": 2.0877542683152113, + "learning_rate": 6.189032863851454e-06, + "loss": 0.962, + "step": 26984 + }, + { + "epoch": 0.64, + "grad_norm": 1.8353511675527436, + "learning_rate": 6.1883274166441356e-06, + "loss": 0.9706, + "step": 26985 + }, + { + "epoch": 0.64, + "grad_norm": 1.9009310379346487, + "learning_rate": 6.187621991629582e-06, + "loss": 1.0047, + "step": 26986 + }, + { + "epoch": 0.64, + "grad_norm": 1.9016135381405872, + "learning_rate": 6.1869165888119e-06, + "loss": 0.9403, + "step": 26987 + }, + { + "epoch": 0.64, + "grad_norm": 2.4721172372953486, + "learning_rate": 6.186211208195194e-06, + "loss": 0.8439, + "step": 26988 + }, + { + "epoch": 0.64, + "grad_norm": 1.872848609762264, + "learning_rate": 6.18550584978358e-06, + "loss": 0.965, + "step": 26989 + }, + { + "epoch": 0.64, + "grad_norm": 2.151465241101997, + "learning_rate": 6.184800513581153e-06, + "loss": 1.0107, + "step": 26990 + }, + { + "epoch": 0.64, + "grad_norm": 1.9644862102321203, + "learning_rate": 6.184095199592026e-06, + "loss": 1.047, + "step": 26991 + }, + { + "epoch": 0.64, + "grad_norm": 2.081614623289448, + "learning_rate": 6.183389907820303e-06, + "loss": 0.9467, + "step": 26992 + }, + { + "epoch": 0.64, + "grad_norm": 1.8557074080874176, + "learning_rate": 6.1826846382701e-06, + "loss": 0.9402, + "step": 26993 + }, + { + "epoch": 0.64, + "grad_norm": 3.027648960963671, + "learning_rate": 6.1819793909455076e-06, + "loss": 0.9719, + "step": 26994 + }, + { + "epoch": 0.64, + "grad_norm": 1.930925212191452, + "learning_rate": 6.181274165850644e-06, + "loss": 0.9179, + "step": 26995 + }, + { + "epoch": 0.64, + "grad_norm": 1.7109770959415789, + "learning_rate": 6.18056896298961e-06, + "loss": 0.9762, + "step": 26996 + }, + { + "epoch": 0.64, + "grad_norm": 2.0245817155740817, + "learning_rate": 6.179863782366512e-06, + "loss": 0.9804, + "step": 26997 + }, + { + "epoch": 0.64, + "grad_norm": 2.1678020312238777, + "learning_rate": 6.179158623985457e-06, + "loss": 0.9629, + "step": 26998 + }, + { + "epoch": 0.64, + "grad_norm": 2.318851641838101, + "learning_rate": 6.178453487850551e-06, + "loss": 1.0064, + "step": 26999 + }, + { + "epoch": 0.64, + "grad_norm": 1.0979484013618264, + "learning_rate": 6.1777483739659e-06, + "loss": 0.9421, + "step": 27000 + }, + { + "epoch": 0.64, + "grad_norm": 1.976637638177779, + "learning_rate": 6.177043282335603e-06, + "loss": 1.0022, + "step": 27001 + }, + { + "epoch": 0.64, + "grad_norm": 1.9280232850241543, + "learning_rate": 6.1763382129637725e-06, + "loss": 0.9681, + "step": 27002 + }, + { + "epoch": 0.64, + "grad_norm": 2.0787041770599677, + "learning_rate": 6.175633165854515e-06, + "loss": 1.0672, + "step": 27003 + }, + { + "epoch": 0.64, + "grad_norm": 2.493905967698106, + "learning_rate": 6.17492814101193e-06, + "loss": 0.8471, + "step": 27004 + }, + { + "epoch": 0.64, + "grad_norm": 1.997476360784263, + "learning_rate": 6.174223138440121e-06, + "loss": 1.0629, + "step": 27005 + }, + { + "epoch": 0.64, + "grad_norm": 1.9310421152039348, + "learning_rate": 6.173518158143199e-06, + "loss": 0.9459, + "step": 27006 + }, + { + "epoch": 0.64, + "grad_norm": 2.1243980424865625, + "learning_rate": 6.172813200125267e-06, + "loss": 0.924, + "step": 27007 + }, + { + "epoch": 0.64, + "grad_norm": 1.9662238596071622, + "learning_rate": 6.172108264390428e-06, + "loss": 0.9891, + "step": 27008 + }, + { + "epoch": 0.64, + "grad_norm": 1.9658058634282412, + "learning_rate": 6.171403350942783e-06, + "loss": 0.9813, + "step": 27009 + }, + { + "epoch": 0.64, + "grad_norm": 2.1683452582118834, + "learning_rate": 6.170698459786442e-06, + "loss": 1.0511, + "step": 27010 + }, + { + "epoch": 0.64, + "grad_norm": 2.13839989885875, + "learning_rate": 6.169993590925509e-06, + "loss": 1.0341, + "step": 27011 + }, + { + "epoch": 0.64, + "grad_norm": 2.563820315155818, + "learning_rate": 6.169288744364084e-06, + "loss": 0.9732, + "step": 27012 + }, + { + "epoch": 0.64, + "grad_norm": 1.782156034341308, + "learning_rate": 6.16858392010627e-06, + "loss": 0.9906, + "step": 27013 + }, + { + "epoch": 0.64, + "grad_norm": 1.9765295996114067, + "learning_rate": 6.167879118156179e-06, + "loss": 0.871, + "step": 27014 + }, + { + "epoch": 0.64, + "grad_norm": 2.2124806251665072, + "learning_rate": 6.167174338517905e-06, + "loss": 0.9982, + "step": 27015 + }, + { + "epoch": 0.64, + "grad_norm": 2.25359502605115, + "learning_rate": 6.166469581195558e-06, + "loss": 0.9624, + "step": 27016 + }, + { + "epoch": 0.64, + "grad_norm": 2.0344220074107926, + "learning_rate": 6.165764846193234e-06, + "loss": 1.0372, + "step": 27017 + }, + { + "epoch": 0.64, + "grad_norm": 2.142496814625871, + "learning_rate": 6.165060133515046e-06, + "loss": 0.9838, + "step": 27018 + }, + { + "epoch": 0.64, + "grad_norm": 2.1918485300030177, + "learning_rate": 6.164355443165091e-06, + "loss": 1.0433, + "step": 27019 + }, + { + "epoch": 0.64, + "grad_norm": 2.1205297796289138, + "learning_rate": 6.163650775147471e-06, + "loss": 1.0485, + "step": 27020 + }, + { + "epoch": 0.64, + "grad_norm": 1.941961790082599, + "learning_rate": 6.162946129466295e-06, + "loss": 0.9426, + "step": 27021 + }, + { + "epoch": 0.64, + "grad_norm": 1.8979908937619387, + "learning_rate": 6.162241506125657e-06, + "loss": 0.9115, + "step": 27022 + }, + { + "epoch": 0.64, + "grad_norm": 2.018262052755171, + "learning_rate": 6.161536905129666e-06, + "loss": 0.8953, + "step": 27023 + }, + { + "epoch": 0.64, + "grad_norm": 2.1122397188158892, + "learning_rate": 6.160832326482418e-06, + "loss": 1.0726, + "step": 27024 + }, + { + "epoch": 0.64, + "grad_norm": 2.265939379586688, + "learning_rate": 6.160127770188029e-06, + "loss": 0.9383, + "step": 27025 + }, + { + "epoch": 0.64, + "grad_norm": 1.0287578877598424, + "learning_rate": 6.1594232362505835e-06, + "loss": 0.9152, + "step": 27026 + }, + { + "epoch": 0.64, + "grad_norm": 2.336761876051539, + "learning_rate": 6.158718724674194e-06, + "loss": 1.0884, + "step": 27027 + }, + { + "epoch": 0.64, + "grad_norm": 1.8579756680090638, + "learning_rate": 6.158014235462964e-06, + "loss": 0.7824, + "step": 27028 + }, + { + "epoch": 0.64, + "grad_norm": 2.135423376261339, + "learning_rate": 6.157309768620987e-06, + "loss": 1.1068, + "step": 27029 + }, + { + "epoch": 0.64, + "grad_norm": 2.187731449089107, + "learning_rate": 6.156605324152369e-06, + "loss": 1.0115, + "step": 27030 + }, + { + "epoch": 0.64, + "grad_norm": 1.78556925113194, + "learning_rate": 6.155900902061212e-06, + "loss": 1.0155, + "step": 27031 + }, + { + "epoch": 0.64, + "grad_norm": 1.9827015286071656, + "learning_rate": 6.155196502351621e-06, + "loss": 1.104, + "step": 27032 + }, + { + "epoch": 0.64, + "grad_norm": 2.1999077189917435, + "learning_rate": 6.154492125027688e-06, + "loss": 1.0996, + "step": 27033 + }, + { + "epoch": 0.64, + "grad_norm": 1.8426525237897016, + "learning_rate": 6.1537877700935215e-06, + "loss": 1.012, + "step": 27034 + }, + { + "epoch": 0.64, + "grad_norm": 2.0431649800364085, + "learning_rate": 6.153083437553222e-06, + "loss": 0.9556, + "step": 27035 + }, + { + "epoch": 0.64, + "grad_norm": 2.1193270678181233, + "learning_rate": 6.152379127410887e-06, + "loss": 1.1272, + "step": 27036 + }, + { + "epoch": 0.64, + "grad_norm": 2.221433902705911, + "learning_rate": 6.151674839670616e-06, + "loss": 0.9173, + "step": 27037 + }, + { + "epoch": 0.64, + "grad_norm": 2.147046937887273, + "learning_rate": 6.1509705743365154e-06, + "loss": 1.0705, + "step": 27038 + }, + { + "epoch": 0.64, + "grad_norm": 2.1919686928854407, + "learning_rate": 6.150266331412684e-06, + "loss": 0.8489, + "step": 27039 + }, + { + "epoch": 0.64, + "grad_norm": 2.149183916076475, + "learning_rate": 6.14956211090322e-06, + "loss": 0.9584, + "step": 27040 + }, + { + "epoch": 0.64, + "grad_norm": 2.0858295132334637, + "learning_rate": 6.148857912812222e-06, + "loss": 1.0073, + "step": 27041 + }, + { + "epoch": 0.64, + "grad_norm": 1.867165562992635, + "learning_rate": 6.148153737143798e-06, + "loss": 0.9871, + "step": 27042 + }, + { + "epoch": 0.64, + "grad_norm": 2.463039775317049, + "learning_rate": 6.1474495839020365e-06, + "loss": 0.9179, + "step": 27043 + }, + { + "epoch": 0.64, + "grad_norm": 2.3236586617158013, + "learning_rate": 6.1467454530910454e-06, + "loss": 1.0594, + "step": 27044 + }, + { + "epoch": 0.64, + "grad_norm": 2.3781288205078037, + "learning_rate": 6.146041344714919e-06, + "loss": 0.9449, + "step": 27045 + }, + { + "epoch": 0.64, + "grad_norm": 1.839255229582142, + "learning_rate": 6.145337258777766e-06, + "loss": 1.0061, + "step": 27046 + }, + { + "epoch": 0.64, + "grad_norm": 2.097925936910616, + "learning_rate": 6.144633195283677e-06, + "loss": 1.1344, + "step": 27047 + }, + { + "epoch": 0.64, + "grad_norm": 2.84725022242142, + "learning_rate": 6.1439291542367544e-06, + "loss": 0.9434, + "step": 27048 + }, + { + "epoch": 0.64, + "grad_norm": 1.93967865034117, + "learning_rate": 6.143225135641097e-06, + "loss": 0.8885, + "step": 27049 + }, + { + "epoch": 0.64, + "grad_norm": 1.8509767087059048, + "learning_rate": 6.142521139500803e-06, + "loss": 0.9132, + "step": 27050 + }, + { + "epoch": 0.64, + "grad_norm": 2.056683240464384, + "learning_rate": 6.1418171658199725e-06, + "loss": 0.9676, + "step": 27051 + }, + { + "epoch": 0.64, + "grad_norm": 2.3459773206140313, + "learning_rate": 6.141113214602705e-06, + "loss": 0.9692, + "step": 27052 + }, + { + "epoch": 0.64, + "grad_norm": 1.8500638332230128, + "learning_rate": 6.1404092858531e-06, + "loss": 0.9031, + "step": 27053 + }, + { + "epoch": 0.64, + "grad_norm": 1.8116980377276242, + "learning_rate": 6.139705379575248e-06, + "loss": 0.9543, + "step": 27054 + }, + { + "epoch": 0.64, + "grad_norm": 2.013778605185621, + "learning_rate": 6.139001495773258e-06, + "loss": 1.149, + "step": 27055 + }, + { + "epoch": 0.64, + "grad_norm": 1.8857017231979987, + "learning_rate": 6.138297634451224e-06, + "loss": 0.8565, + "step": 27056 + }, + { + "epoch": 0.64, + "grad_norm": 1.9067692140544064, + "learning_rate": 6.137593795613242e-06, + "loss": 0.9931, + "step": 27057 + }, + { + "epoch": 0.64, + "grad_norm": 2.046724919443561, + "learning_rate": 6.136889979263409e-06, + "loss": 0.9549, + "step": 27058 + }, + { + "epoch": 0.64, + "grad_norm": 1.8173867391639198, + "learning_rate": 6.13618618540583e-06, + "loss": 0.8967, + "step": 27059 + }, + { + "epoch": 0.64, + "grad_norm": 2.2301153730419587, + "learning_rate": 6.135482414044598e-06, + "loss": 1.0448, + "step": 27060 + }, + { + "epoch": 0.64, + "grad_norm": 2.089107094372909, + "learning_rate": 6.1347786651838105e-06, + "loss": 1.089, + "step": 27061 + }, + { + "epoch": 0.64, + "grad_norm": 1.111837431386775, + "learning_rate": 6.134074938827562e-06, + "loss": 1.013, + "step": 27062 + }, + { + "epoch": 0.64, + "grad_norm": 1.908743923120554, + "learning_rate": 6.133371234979958e-06, + "loss": 1.002, + "step": 27063 + }, + { + "epoch": 0.64, + "grad_norm": 1.9846888465526527, + "learning_rate": 6.1326675536450885e-06, + "loss": 1.0562, + "step": 27064 + }, + { + "epoch": 0.64, + "grad_norm": 1.9624769411860286, + "learning_rate": 6.1319638948270536e-06, + "loss": 1.0232, + "step": 27065 + }, + { + "epoch": 0.64, + "grad_norm": 1.9117930431178694, + "learning_rate": 6.131260258529947e-06, + "loss": 1.0594, + "step": 27066 + }, + { + "epoch": 0.64, + "grad_norm": 2.416824925732943, + "learning_rate": 6.130556644757871e-06, + "loss": 0.9504, + "step": 27067 + }, + { + "epoch": 0.64, + "grad_norm": 2.131516477282848, + "learning_rate": 6.129853053514918e-06, + "loss": 1.1643, + "step": 27068 + }, + { + "epoch": 0.64, + "grad_norm": 1.954278554781427, + "learning_rate": 6.129149484805183e-06, + "loss": 1.0487, + "step": 27069 + }, + { + "epoch": 0.64, + "grad_norm": 1.081362068297611, + "learning_rate": 6.1284459386327735e-06, + "loss": 0.9303, + "step": 27070 + }, + { + "epoch": 0.64, + "grad_norm": 1.928160112892443, + "learning_rate": 6.1277424150017695e-06, + "loss": 0.9803, + "step": 27071 + }, + { + "epoch": 0.64, + "grad_norm": 2.156408632682719, + "learning_rate": 6.127038913916278e-06, + "loss": 1.039, + "step": 27072 + }, + { + "epoch": 0.64, + "grad_norm": 2.3357454041729127, + "learning_rate": 6.12633543538039e-06, + "loss": 1.1854, + "step": 27073 + }, + { + "epoch": 0.64, + "grad_norm": 1.0436050164394826, + "learning_rate": 6.125631979398209e-06, + "loss": 0.9268, + "step": 27074 + }, + { + "epoch": 0.64, + "grad_norm": 2.015224444363396, + "learning_rate": 6.124928545973819e-06, + "loss": 0.958, + "step": 27075 + }, + { + "epoch": 0.64, + "grad_norm": 2.115912423426106, + "learning_rate": 6.124225135111325e-06, + "loss": 0.9855, + "step": 27076 + }, + { + "epoch": 0.64, + "grad_norm": 1.8343311365139432, + "learning_rate": 6.12352174681482e-06, + "loss": 0.8511, + "step": 27077 + }, + { + "epoch": 0.64, + "grad_norm": 2.1099987740237323, + "learning_rate": 6.122818381088398e-06, + "loss": 1.0433, + "step": 27078 + }, + { + "epoch": 0.64, + "grad_norm": 2.105914372742452, + "learning_rate": 6.1221150379361514e-06, + "loss": 0.9604, + "step": 27079 + }, + { + "epoch": 0.64, + "grad_norm": 1.845925330607006, + "learning_rate": 6.1214117173621815e-06, + "loss": 0.8513, + "step": 27080 + }, + { + "epoch": 0.64, + "grad_norm": 1.8717252067978716, + "learning_rate": 6.120708419370583e-06, + "loss": 1.0672, + "step": 27081 + }, + { + "epoch": 0.64, + "grad_norm": 1.9253215891669657, + "learning_rate": 6.120005143965443e-06, + "loss": 1.0252, + "step": 27082 + }, + { + "epoch": 0.64, + "grad_norm": 1.870968713071717, + "learning_rate": 6.1193018911508635e-06, + "loss": 0.8624, + "step": 27083 + }, + { + "epoch": 0.64, + "grad_norm": 2.2676286630149067, + "learning_rate": 6.118598660930941e-06, + "loss": 0.977, + "step": 27084 + }, + { + "epoch": 0.64, + "grad_norm": 1.9629029042301134, + "learning_rate": 6.117895453309762e-06, + "loss": 1.05, + "step": 27085 + }, + { + "epoch": 0.64, + "grad_norm": 2.109025877631471, + "learning_rate": 6.117192268291423e-06, + "loss": 0.8931, + "step": 27086 + }, + { + "epoch": 0.64, + "grad_norm": 2.0141991713241985, + "learning_rate": 6.116489105880023e-06, + "loss": 0.8466, + "step": 27087 + }, + { + "epoch": 0.64, + "grad_norm": 2.0333622893954035, + "learning_rate": 6.115785966079653e-06, + "loss": 1.169, + "step": 27088 + }, + { + "epoch": 0.64, + "grad_norm": 1.1572072790044172, + "learning_rate": 6.115082848894407e-06, + "loss": 0.918, + "step": 27089 + }, + { + "epoch": 0.64, + "grad_norm": 1.9148100064797409, + "learning_rate": 6.114379754328375e-06, + "loss": 0.9376, + "step": 27090 + }, + { + "epoch": 0.64, + "grad_norm": 2.090724861931955, + "learning_rate": 6.113676682385661e-06, + "loss": 0.9068, + "step": 27091 + }, + { + "epoch": 0.64, + "grad_norm": 1.0846693877292184, + "learning_rate": 6.112973633070346e-06, + "loss": 0.9523, + "step": 27092 + }, + { + "epoch": 0.64, + "grad_norm": 2.1716323541611433, + "learning_rate": 6.11227060638653e-06, + "loss": 0.9898, + "step": 27093 + }, + { + "epoch": 0.64, + "grad_norm": 1.159798087198631, + "learning_rate": 6.111567602338304e-06, + "loss": 0.944, + "step": 27094 + }, + { + "epoch": 0.64, + "grad_norm": 2.0605223705091182, + "learning_rate": 6.110864620929766e-06, + "loss": 0.9646, + "step": 27095 + }, + { + "epoch": 0.64, + "grad_norm": 2.0305227021830023, + "learning_rate": 6.110161662165005e-06, + "loss": 1.0338, + "step": 27096 + }, + { + "epoch": 0.64, + "grad_norm": 2.050808429424378, + "learning_rate": 6.109458726048113e-06, + "loss": 0.9767, + "step": 27097 + }, + { + "epoch": 0.64, + "grad_norm": 1.8826227477078923, + "learning_rate": 6.108755812583187e-06, + "loss": 0.9229, + "step": 27098 + }, + { + "epoch": 0.64, + "grad_norm": 1.8015353248950243, + "learning_rate": 6.108052921774312e-06, + "loss": 0.8153, + "step": 27099 + }, + { + "epoch": 0.64, + "grad_norm": 1.8170650101563468, + "learning_rate": 6.107350053625587e-06, + "loss": 1.0021, + "step": 27100 + }, + { + "epoch": 0.64, + "grad_norm": 2.19779925224376, + "learning_rate": 6.106647208141101e-06, + "loss": 1.0605, + "step": 27101 + }, + { + "epoch": 0.64, + "grad_norm": 2.128669276248939, + "learning_rate": 6.105944385324951e-06, + "loss": 1.0322, + "step": 27102 + }, + { + "epoch": 0.64, + "grad_norm": 2.1792190466664043, + "learning_rate": 6.105241585181222e-06, + "loss": 1.0147, + "step": 27103 + }, + { + "epoch": 0.64, + "grad_norm": 1.9821474388035394, + "learning_rate": 6.104538807714011e-06, + "loss": 1.064, + "step": 27104 + }, + { + "epoch": 0.64, + "grad_norm": 2.1117770333561228, + "learning_rate": 6.10383605292741e-06, + "loss": 1.0014, + "step": 27105 + }, + { + "epoch": 0.64, + "grad_norm": 2.0658593547028183, + "learning_rate": 6.103133320825507e-06, + "loss": 1.1267, + "step": 27106 + }, + { + "epoch": 0.64, + "grad_norm": 5.225381472882134, + "learning_rate": 6.102430611412395e-06, + "loss": 1.0268, + "step": 27107 + }, + { + "epoch": 0.64, + "grad_norm": 1.9363131959665556, + "learning_rate": 6.101727924692167e-06, + "loss": 0.8469, + "step": 27108 + }, + { + "epoch": 0.64, + "grad_norm": 1.9671600655423944, + "learning_rate": 6.101025260668916e-06, + "loss": 0.9455, + "step": 27109 + }, + { + "epoch": 0.64, + "grad_norm": 1.9560033138706816, + "learning_rate": 6.100322619346726e-06, + "loss": 1.0206, + "step": 27110 + }, + { + "epoch": 0.64, + "grad_norm": 2.0022360115688635, + "learning_rate": 6.099620000729692e-06, + "loss": 0.9656, + "step": 27111 + }, + { + "epoch": 0.64, + "grad_norm": 1.0553364981066335, + "learning_rate": 6.098917404821909e-06, + "loss": 0.9407, + "step": 27112 + }, + { + "epoch": 0.64, + "grad_norm": 2.2501977907688135, + "learning_rate": 6.098214831627462e-06, + "loss": 1.0319, + "step": 27113 + }, + { + "epoch": 0.64, + "grad_norm": 1.900767330980765, + "learning_rate": 6.097512281150444e-06, + "loss": 1.0452, + "step": 27114 + }, + { + "epoch": 0.64, + "grad_norm": 1.851557892580376, + "learning_rate": 6.096809753394941e-06, + "loss": 0.9395, + "step": 27115 + }, + { + "epoch": 0.64, + "grad_norm": 2.3750948454676846, + "learning_rate": 6.0961072483650526e-06, + "loss": 0.9756, + "step": 27116 + }, + { + "epoch": 0.64, + "grad_norm": 1.0706390300221924, + "learning_rate": 6.095404766064863e-06, + "loss": 0.9957, + "step": 27117 + }, + { + "epoch": 0.64, + "grad_norm": 1.9088725408213052, + "learning_rate": 6.094702306498459e-06, + "loss": 1.0506, + "step": 27118 + }, + { + "epoch": 0.64, + "grad_norm": 1.871709440153178, + "learning_rate": 6.093999869669942e-06, + "loss": 0.9888, + "step": 27119 + }, + { + "epoch": 0.64, + "grad_norm": 2.058831090609425, + "learning_rate": 6.0932974555833875e-06, + "loss": 1.0771, + "step": 27120 + }, + { + "epoch": 0.64, + "grad_norm": 2.056196263595433, + "learning_rate": 6.092595064242894e-06, + "loss": 0.9853, + "step": 27121 + }, + { + "epoch": 0.64, + "grad_norm": 1.9027317155027668, + "learning_rate": 6.091892695652548e-06, + "loss": 1.037, + "step": 27122 + }, + { + "epoch": 0.64, + "grad_norm": 2.0156130157343606, + "learning_rate": 6.091190349816446e-06, + "loss": 0.949, + "step": 27123 + }, + { + "epoch": 0.64, + "grad_norm": 2.27881796771374, + "learning_rate": 6.090488026738666e-06, + "loss": 1.0494, + "step": 27124 + }, + { + "epoch": 0.64, + "grad_norm": 2.6382473886364015, + "learning_rate": 6.089785726423303e-06, + "loss": 1.0053, + "step": 27125 + }, + { + "epoch": 0.64, + "grad_norm": 1.878143897537718, + "learning_rate": 6.089083448874446e-06, + "loss": 0.9642, + "step": 27126 + }, + { + "epoch": 0.64, + "grad_norm": 2.9278473367046374, + "learning_rate": 6.088381194096184e-06, + "loss": 0.9949, + "step": 27127 + }, + { + "epoch": 0.64, + "grad_norm": 1.12954221948101, + "learning_rate": 6.087678962092603e-06, + "loss": 0.8624, + "step": 27128 + }, + { + "epoch": 0.64, + "grad_norm": 1.8057610733035612, + "learning_rate": 6.0869767528677944e-06, + "loss": 0.9355, + "step": 27129 + }, + { + "epoch": 0.64, + "grad_norm": 1.9215557995844714, + "learning_rate": 6.086274566425848e-06, + "loss": 0.9533, + "step": 27130 + }, + { + "epoch": 0.64, + "grad_norm": 1.8435610864702818, + "learning_rate": 6.085572402770848e-06, + "loss": 1.0472, + "step": 27131 + }, + { + "epoch": 0.64, + "grad_norm": 2.167797514772904, + "learning_rate": 6.084870261906883e-06, + "loss": 0.9731, + "step": 27132 + }, + { + "epoch": 0.64, + "grad_norm": 1.8054077205413317, + "learning_rate": 6.084168143838046e-06, + "loss": 1.1595, + "step": 27133 + }, + { + "epoch": 0.64, + "grad_norm": 2.0694720742499086, + "learning_rate": 6.083466048568423e-06, + "loss": 1.0357, + "step": 27134 + }, + { + "epoch": 0.64, + "grad_norm": 2.233461026503989, + "learning_rate": 6.082763976102097e-06, + "loss": 0.9846, + "step": 27135 + }, + { + "epoch": 0.64, + "grad_norm": 2.8000586661118496, + "learning_rate": 6.082061926443162e-06, + "loss": 1.0036, + "step": 27136 + }, + { + "epoch": 0.64, + "grad_norm": 1.819685411942459, + "learning_rate": 6.081359899595702e-06, + "loss": 0.9857, + "step": 27137 + }, + { + "epoch": 0.64, + "grad_norm": 2.035608427071146, + "learning_rate": 6.080657895563807e-06, + "loss": 1.0017, + "step": 27138 + }, + { + "epoch": 0.64, + "grad_norm": 1.9631688432637926, + "learning_rate": 6.079955914351558e-06, + "loss": 1.1007, + "step": 27139 + }, + { + "epoch": 0.64, + "grad_norm": 2.013559142868992, + "learning_rate": 6.079253955963049e-06, + "loss": 0.9938, + "step": 27140 + }, + { + "epoch": 0.64, + "grad_norm": 1.9808466190383662, + "learning_rate": 6.078552020402367e-06, + "loss": 1.0405, + "step": 27141 + }, + { + "epoch": 0.64, + "grad_norm": 1.0466241547138653, + "learning_rate": 6.0778501076735956e-06, + "loss": 0.9099, + "step": 27142 + }, + { + "epoch": 0.64, + "grad_norm": 2.57348655729272, + "learning_rate": 6.077148217780818e-06, + "loss": 0.8883, + "step": 27143 + }, + { + "epoch": 0.64, + "grad_norm": 1.8982828996152536, + "learning_rate": 6.076446350728131e-06, + "loss": 1.0292, + "step": 27144 + }, + { + "epoch": 0.64, + "grad_norm": 1.9367285922526705, + "learning_rate": 6.0757445065196145e-06, + "loss": 1.0585, + "step": 27145 + }, + { + "epoch": 0.64, + "grad_norm": 1.0784091728281466, + "learning_rate": 6.075042685159356e-06, + "loss": 0.8967, + "step": 27146 + }, + { + "epoch": 0.64, + "grad_norm": 2.1582770096529673, + "learning_rate": 6.074340886651438e-06, + "loss": 0.891, + "step": 27147 + }, + { + "epoch": 0.64, + "grad_norm": 1.8781447315974644, + "learning_rate": 6.073639110999955e-06, + "loss": 1.0014, + "step": 27148 + }, + { + "epoch": 0.64, + "grad_norm": 1.9707406095261135, + "learning_rate": 6.072937358208986e-06, + "loss": 1.0058, + "step": 27149 + }, + { + "epoch": 0.64, + "grad_norm": 1.173791721568158, + "learning_rate": 6.072235628282619e-06, + "loss": 1.0747, + "step": 27150 + }, + { + "epoch": 0.64, + "grad_norm": 1.9946657226414486, + "learning_rate": 6.071533921224941e-06, + "loss": 0.9685, + "step": 27151 + }, + { + "epoch": 0.64, + "grad_norm": 1.9562152595718314, + "learning_rate": 6.0708322370400345e-06, + "loss": 0.8334, + "step": 27152 + }, + { + "epoch": 0.64, + "grad_norm": 2.1490612849753106, + "learning_rate": 6.0701305757319866e-06, + "loss": 0.9932, + "step": 27153 + }, + { + "epoch": 0.64, + "grad_norm": 2.158464856049661, + "learning_rate": 6.069428937304883e-06, + "loss": 0.9588, + "step": 27154 + }, + { + "epoch": 0.64, + "grad_norm": 2.111054497688574, + "learning_rate": 6.06872732176281e-06, + "loss": 0.9828, + "step": 27155 + }, + { + "epoch": 0.64, + "grad_norm": 2.0082161377984384, + "learning_rate": 6.068025729109849e-06, + "loss": 0.9364, + "step": 27156 + }, + { + "epoch": 0.64, + "grad_norm": 2.234973728571203, + "learning_rate": 6.067324159350088e-06, + "loss": 0.939, + "step": 27157 + }, + { + "epoch": 0.64, + "grad_norm": 2.12908618761528, + "learning_rate": 6.066622612487611e-06, + "loss": 0.9476, + "step": 27158 + }, + { + "epoch": 0.64, + "grad_norm": 1.8726059680103613, + "learning_rate": 6.065921088526503e-06, + "loss": 0.8871, + "step": 27159 + }, + { + "epoch": 0.64, + "grad_norm": 1.19667492968767, + "learning_rate": 6.065219587470845e-06, + "loss": 0.8983, + "step": 27160 + }, + { + "epoch": 0.64, + "grad_norm": 2.1219800413341066, + "learning_rate": 6.064518109324727e-06, + "loss": 0.8606, + "step": 27161 + }, + { + "epoch": 0.64, + "grad_norm": 1.9398732555054743, + "learning_rate": 6.063816654092232e-06, + "loss": 1.0051, + "step": 27162 + }, + { + "epoch": 0.64, + "grad_norm": 1.9375643121040045, + "learning_rate": 6.06311522177744e-06, + "loss": 1.0443, + "step": 27163 + }, + { + "epoch": 0.64, + "grad_norm": 1.8832247251147403, + "learning_rate": 6.062413812384436e-06, + "loss": 1.0984, + "step": 27164 + }, + { + "epoch": 0.64, + "grad_norm": 2.1567945877288057, + "learning_rate": 6.06171242591731e-06, + "loss": 1.0332, + "step": 27165 + }, + { + "epoch": 0.64, + "grad_norm": 1.8638766911056601, + "learning_rate": 6.061011062380139e-06, + "loss": 0.9354, + "step": 27166 + }, + { + "epoch": 0.64, + "grad_norm": 2.0746019779078657, + "learning_rate": 6.060309721777006e-06, + "loss": 1.0102, + "step": 27167 + }, + { + "epoch": 0.64, + "grad_norm": 1.1521152425353247, + "learning_rate": 6.059608404111998e-06, + "loss": 0.9378, + "step": 27168 + }, + { + "epoch": 0.64, + "grad_norm": 2.1661697712616936, + "learning_rate": 6.058907109389202e-06, + "loss": 1.0827, + "step": 27169 + }, + { + "epoch": 0.64, + "grad_norm": 2.80505537118277, + "learning_rate": 6.058205837612694e-06, + "loss": 0.9379, + "step": 27170 + }, + { + "epoch": 0.64, + "grad_norm": 2.422603507415029, + "learning_rate": 6.057504588786556e-06, + "loss": 1.0012, + "step": 27171 + }, + { + "epoch": 0.64, + "grad_norm": 1.9041686999148963, + "learning_rate": 6.0568033629148814e-06, + "loss": 1.0134, + "step": 27172 + }, + { + "epoch": 0.64, + "grad_norm": 1.8634725472229376, + "learning_rate": 6.05610216000174e-06, + "loss": 1.0166, + "step": 27173 + }, + { + "epoch": 0.64, + "grad_norm": 2.814731843365226, + "learning_rate": 6.055400980051223e-06, + "loss": 1.0329, + "step": 27174 + }, + { + "epoch": 0.64, + "grad_norm": 1.8763484066283078, + "learning_rate": 6.054699823067407e-06, + "loss": 1.0037, + "step": 27175 + }, + { + "epoch": 0.64, + "grad_norm": 2.0284586901932373, + "learning_rate": 6.053998689054384e-06, + "loss": 0.9498, + "step": 27176 + }, + { + "epoch": 0.64, + "grad_norm": 2.100338168852219, + "learning_rate": 6.053297578016224e-06, + "loss": 0.8357, + "step": 27177 + }, + { + "epoch": 0.64, + "grad_norm": 1.9739481508534744, + "learning_rate": 6.052596489957017e-06, + "loss": 0.9613, + "step": 27178 + }, + { + "epoch": 0.64, + "grad_norm": 2.037816731893028, + "learning_rate": 6.051895424880847e-06, + "loss": 1.0298, + "step": 27179 + }, + { + "epoch": 0.64, + "grad_norm": 2.4702260297456773, + "learning_rate": 6.051194382791785e-06, + "loss": 0.7792, + "step": 27180 + }, + { + "epoch": 0.64, + "grad_norm": 2.104689363925359, + "learning_rate": 6.050493363693924e-06, + "loss": 0.813, + "step": 27181 + }, + { + "epoch": 0.64, + "grad_norm": 1.8737882366029295, + "learning_rate": 6.049792367591338e-06, + "loss": 0.9367, + "step": 27182 + }, + { + "epoch": 0.64, + "grad_norm": 1.9128196178238643, + "learning_rate": 6.049091394488116e-06, + "loss": 0.9813, + "step": 27183 + }, + { + "epoch": 0.64, + "grad_norm": 1.2298372597214782, + "learning_rate": 6.048390444388329e-06, + "loss": 0.9754, + "step": 27184 + }, + { + "epoch": 0.64, + "grad_norm": 2.024591961868022, + "learning_rate": 6.047689517296067e-06, + "loss": 0.9384, + "step": 27185 + }, + { + "epoch": 0.64, + "grad_norm": 2.3666558532251987, + "learning_rate": 6.04698861321541e-06, + "loss": 1.0018, + "step": 27186 + }, + { + "epoch": 0.64, + "grad_norm": 2.0551109858871013, + "learning_rate": 6.046287732150433e-06, + "loss": 0.9502, + "step": 27187 + }, + { + "epoch": 0.64, + "grad_norm": 2.7874252027448274, + "learning_rate": 6.045586874105221e-06, + "loss": 0.9338, + "step": 27188 + }, + { + "epoch": 0.64, + "grad_norm": 2.083461847464141, + "learning_rate": 6.044886039083855e-06, + "loss": 1.0941, + "step": 27189 + }, + { + "epoch": 0.64, + "grad_norm": 1.1185178354278036, + "learning_rate": 6.044185227090415e-06, + "loss": 0.9841, + "step": 27190 + }, + { + "epoch": 0.64, + "grad_norm": 2.074946055196712, + "learning_rate": 6.0434844381289814e-06, + "loss": 1.217, + "step": 27191 + }, + { + "epoch": 0.64, + "grad_norm": 1.9710880084224423, + "learning_rate": 6.042783672203632e-06, + "loss": 0.9076, + "step": 27192 + }, + { + "epoch": 0.64, + "grad_norm": 1.9244907402300104, + "learning_rate": 6.042082929318452e-06, + "loss": 0.9667, + "step": 27193 + }, + { + "epoch": 0.64, + "grad_norm": 2.0176709715518757, + "learning_rate": 6.041382209477516e-06, + "loss": 1.068, + "step": 27194 + }, + { + "epoch": 0.64, + "grad_norm": 2.0319908464580934, + "learning_rate": 6.040681512684907e-06, + "loss": 0.9654, + "step": 27195 + }, + { + "epoch": 0.64, + "grad_norm": 1.929319177099639, + "learning_rate": 6.039980838944701e-06, + "loss": 0.9135, + "step": 27196 + }, + { + "epoch": 0.64, + "grad_norm": 2.240931593143325, + "learning_rate": 6.039280188260985e-06, + "loss": 0.9809, + "step": 27197 + }, + { + "epoch": 0.64, + "grad_norm": 1.8562417488123475, + "learning_rate": 6.0385795606378315e-06, + "loss": 0.9654, + "step": 27198 + }, + { + "epoch": 0.64, + "grad_norm": 2.024825480959104, + "learning_rate": 6.037878956079322e-06, + "loss": 1.1183, + "step": 27199 + }, + { + "epoch": 0.64, + "grad_norm": 1.838567774469529, + "learning_rate": 6.037178374589536e-06, + "loss": 0.9725, + "step": 27200 + }, + { + "epoch": 0.64, + "grad_norm": 1.979268713337734, + "learning_rate": 6.0364778161725505e-06, + "loss": 0.8938, + "step": 27201 + }, + { + "epoch": 0.64, + "grad_norm": 1.0535656154428323, + "learning_rate": 6.035777280832447e-06, + "loss": 0.9335, + "step": 27202 + }, + { + "epoch": 0.64, + "grad_norm": 1.8818697594759235, + "learning_rate": 6.035076768573303e-06, + "loss": 1.061, + "step": 27203 + }, + { + "epoch": 0.64, + "grad_norm": 2.375319462841604, + "learning_rate": 6.034376279399199e-06, + "loss": 0.9338, + "step": 27204 + }, + { + "epoch": 0.64, + "grad_norm": 2.837177283276479, + "learning_rate": 6.033675813314208e-06, + "loss": 0.9663, + "step": 27205 + }, + { + "epoch": 0.64, + "grad_norm": 2.0038205210596933, + "learning_rate": 6.032975370322414e-06, + "loss": 0.9591, + "step": 27206 + }, + { + "epoch": 0.64, + "grad_norm": 2.3492395175004637, + "learning_rate": 6.032274950427896e-06, + "loss": 1.0378, + "step": 27207 + }, + { + "epoch": 0.64, + "grad_norm": 1.8638193305230573, + "learning_rate": 6.031574553634727e-06, + "loss": 0.9656, + "step": 27208 + }, + { + "epoch": 0.64, + "grad_norm": 1.8555559122794556, + "learning_rate": 6.030874179946986e-06, + "loss": 1.1009, + "step": 27209 + }, + { + "epoch": 0.64, + "grad_norm": 2.260138094690173, + "learning_rate": 6.0301738293687525e-06, + "loss": 1.0129, + "step": 27210 + }, + { + "epoch": 0.64, + "grad_norm": 2.0746803582735502, + "learning_rate": 6.029473501904108e-06, + "loss": 0.9417, + "step": 27211 + }, + { + "epoch": 0.64, + "grad_norm": 1.0716981650021922, + "learning_rate": 6.028773197557121e-06, + "loss": 1.0194, + "step": 27212 + }, + { + "epoch": 0.64, + "grad_norm": 2.1107434562473855, + "learning_rate": 6.028072916331874e-06, + "loss": 0.9734, + "step": 27213 + }, + { + "epoch": 0.64, + "grad_norm": 1.9583128146967634, + "learning_rate": 6.027372658232448e-06, + "loss": 0.9698, + "step": 27214 + }, + { + "epoch": 0.64, + "grad_norm": 4.269698547307843, + "learning_rate": 6.026672423262914e-06, + "loss": 0.9438, + "step": 27215 + }, + { + "epoch": 0.64, + "grad_norm": 2.390469367352642, + "learning_rate": 6.025972211427346e-06, + "loss": 0.9406, + "step": 27216 + }, + { + "epoch": 0.64, + "grad_norm": 2.248113185975906, + "learning_rate": 6.025272022729831e-06, + "loss": 0.9895, + "step": 27217 + }, + { + "epoch": 0.64, + "grad_norm": 2.987353574287949, + "learning_rate": 6.024571857174443e-06, + "loss": 0.9544, + "step": 27218 + }, + { + "epoch": 0.64, + "grad_norm": 2.1413783769470554, + "learning_rate": 6.0238717147652525e-06, + "loss": 1.0133, + "step": 27219 + }, + { + "epoch": 0.64, + "grad_norm": 2.0570032600660157, + "learning_rate": 6.023171595506338e-06, + "loss": 1.0416, + "step": 27220 + }, + { + "epoch": 0.64, + "grad_norm": 2.604947672783175, + "learning_rate": 6.022471499401785e-06, + "loss": 0.9122, + "step": 27221 + }, + { + "epoch": 0.64, + "grad_norm": 1.9592277184600633, + "learning_rate": 6.021771426455656e-06, + "loss": 1.0494, + "step": 27222 + }, + { + "epoch": 0.64, + "grad_norm": 1.9166691468918065, + "learning_rate": 6.0210713766720344e-06, + "loss": 0.9282, + "step": 27223 + }, + { + "epoch": 0.64, + "grad_norm": 2.0290629681317784, + "learning_rate": 6.020371350054992e-06, + "loss": 1.0857, + "step": 27224 + }, + { + "epoch": 0.64, + "grad_norm": 2.015115022253649, + "learning_rate": 6.019671346608615e-06, + "loss": 1.0216, + "step": 27225 + }, + { + "epoch": 0.64, + "grad_norm": 1.9026195650634872, + "learning_rate": 6.018971366336965e-06, + "loss": 0.8677, + "step": 27226 + }, + { + "epoch": 0.64, + "grad_norm": 1.808702724090389, + "learning_rate": 6.018271409244127e-06, + "loss": 0.91, + "step": 27227 + }, + { + "epoch": 0.64, + "grad_norm": 1.8646639952925088, + "learning_rate": 6.017571475334174e-06, + "loss": 1.0203, + "step": 27228 + }, + { + "epoch": 0.64, + "grad_norm": 1.9243125741257898, + "learning_rate": 6.016871564611177e-06, + "loss": 0.9136, + "step": 27229 + }, + { + "epoch": 0.64, + "grad_norm": 2.0354077764685727, + "learning_rate": 6.016171677079218e-06, + "loss": 1.071, + "step": 27230 + }, + { + "epoch": 0.64, + "grad_norm": 2.005454449120598, + "learning_rate": 6.0154718127423685e-06, + "loss": 0.9934, + "step": 27231 + }, + { + "epoch": 0.64, + "grad_norm": 2.063360038645605, + "learning_rate": 6.014771971604705e-06, + "loss": 1.0578, + "step": 27232 + }, + { + "epoch": 0.64, + "grad_norm": 1.9332909066322426, + "learning_rate": 6.014072153670296e-06, + "loss": 1.0779, + "step": 27233 + }, + { + "epoch": 0.64, + "grad_norm": 1.8965028769035508, + "learning_rate": 6.0133723589432235e-06, + "loss": 0.9812, + "step": 27234 + }, + { + "epoch": 0.64, + "grad_norm": 2.3185068281346135, + "learning_rate": 6.012672587427562e-06, + "loss": 1.0106, + "step": 27235 + }, + { + "epoch": 0.64, + "grad_norm": 1.9050450298085475, + "learning_rate": 6.01197283912738e-06, + "loss": 0.9148, + "step": 27236 + }, + { + "epoch": 0.64, + "grad_norm": 2.1974991222929505, + "learning_rate": 6.011273114046754e-06, + "loss": 1.0829, + "step": 27237 + }, + { + "epoch": 0.64, + "grad_norm": 2.4715511637332264, + "learning_rate": 6.010573412189759e-06, + "loss": 1.1496, + "step": 27238 + }, + { + "epoch": 0.64, + "grad_norm": 1.8543072606916506, + "learning_rate": 6.009873733560472e-06, + "loss": 1.0166, + "step": 27239 + }, + { + "epoch": 0.64, + "grad_norm": 2.040622813580884, + "learning_rate": 6.00917407816296e-06, + "loss": 0.9529, + "step": 27240 + }, + { + "epoch": 0.64, + "grad_norm": 1.9388058340123564, + "learning_rate": 6.0084744460013e-06, + "loss": 1.1027, + "step": 27241 + }, + { + "epoch": 0.64, + "grad_norm": 3.0146751391813433, + "learning_rate": 6.007774837079567e-06, + "loss": 1.0219, + "step": 27242 + }, + { + "epoch": 0.64, + "grad_norm": 2.423350558294729, + "learning_rate": 6.007075251401833e-06, + "loss": 0.942, + "step": 27243 + }, + { + "epoch": 0.64, + "grad_norm": 3.7107426825005576, + "learning_rate": 6.00637568897217e-06, + "loss": 0.9697, + "step": 27244 + }, + { + "epoch": 0.64, + "grad_norm": 1.9047472896127036, + "learning_rate": 6.0056761497946505e-06, + "loss": 0.9969, + "step": 27245 + }, + { + "epoch": 0.64, + "grad_norm": 2.066271141345997, + "learning_rate": 6.004976633873354e-06, + "loss": 1.1494, + "step": 27246 + }, + { + "epoch": 0.64, + "grad_norm": 2.0545488094754343, + "learning_rate": 6.004277141212345e-06, + "loss": 0.9714, + "step": 27247 + }, + { + "epoch": 0.64, + "grad_norm": 1.7797618027189683, + "learning_rate": 6.003577671815699e-06, + "loss": 0.9226, + "step": 27248 + }, + { + "epoch": 0.64, + "grad_norm": 1.1010842892721249, + "learning_rate": 6.002878225687492e-06, + "loss": 0.8783, + "step": 27249 + }, + { + "epoch": 0.64, + "grad_norm": 2.01098687610848, + "learning_rate": 6.002178802831789e-06, + "loss": 1.0269, + "step": 27250 + }, + { + "epoch": 0.64, + "grad_norm": 1.0491770229719772, + "learning_rate": 6.001479403252669e-06, + "loss": 0.926, + "step": 27251 + }, + { + "epoch": 0.64, + "grad_norm": 2.2178128806738453, + "learning_rate": 6.000780026954199e-06, + "loss": 0.8149, + "step": 27252 + }, + { + "epoch": 0.64, + "grad_norm": 2.3122562298287566, + "learning_rate": 6.000080673940461e-06, + "loss": 1.0066, + "step": 27253 + }, + { + "epoch": 0.64, + "grad_norm": 2.2341988520571885, + "learning_rate": 5.999381344215514e-06, + "loss": 1.1182, + "step": 27254 + }, + { + "epoch": 0.64, + "grad_norm": 2.0710685576412606, + "learning_rate": 5.998682037783437e-06, + "loss": 0.9643, + "step": 27255 + }, + { + "epoch": 0.64, + "grad_norm": 2.066627415170507, + "learning_rate": 5.997982754648297e-06, + "loss": 1.0655, + "step": 27256 + }, + { + "epoch": 0.64, + "grad_norm": 1.9103918793826888, + "learning_rate": 5.997283494814175e-06, + "loss": 1.0567, + "step": 27257 + }, + { + "epoch": 0.64, + "grad_norm": 2.05639240099635, + "learning_rate": 5.99658425828513e-06, + "loss": 1.0242, + "step": 27258 + }, + { + "epoch": 0.64, + "grad_norm": 1.0369005512728597, + "learning_rate": 5.99588504506524e-06, + "loss": 0.8684, + "step": 27259 + }, + { + "epoch": 0.64, + "grad_norm": 2.09593404955688, + "learning_rate": 5.995185855158577e-06, + "loss": 1.0857, + "step": 27260 + }, + { + "epoch": 0.64, + "grad_norm": 2.174013139476552, + "learning_rate": 5.994486688569209e-06, + "loss": 1.0164, + "step": 27261 + }, + { + "epoch": 0.64, + "grad_norm": 1.110036229807467, + "learning_rate": 5.993787545301204e-06, + "loss": 0.9399, + "step": 27262 + }, + { + "epoch": 0.64, + "grad_norm": 2.1457061221230114, + "learning_rate": 5.99308842535864e-06, + "loss": 1.0985, + "step": 27263 + }, + { + "epoch": 0.64, + "grad_norm": 1.8876990076760065, + "learning_rate": 5.9923893287455845e-06, + "loss": 0.9454, + "step": 27264 + }, + { + "epoch": 0.64, + "grad_norm": 1.1299733102604848, + "learning_rate": 5.991690255466105e-06, + "loss": 0.8883, + "step": 27265 + }, + { + "epoch": 0.64, + "grad_norm": 2.9241199995930063, + "learning_rate": 5.990991205524274e-06, + "loss": 1.014, + "step": 27266 + }, + { + "epoch": 0.64, + "grad_norm": 2.7017649402822905, + "learning_rate": 5.990292178924164e-06, + "loss": 1.0614, + "step": 27267 + }, + { + "epoch": 0.64, + "grad_norm": 2.203048849417273, + "learning_rate": 5.98959317566984e-06, + "loss": 1.0268, + "step": 27268 + }, + { + "epoch": 0.64, + "grad_norm": 1.896184229892444, + "learning_rate": 5.988894195765372e-06, + "loss": 0.95, + "step": 27269 + }, + { + "epoch": 0.64, + "grad_norm": 2.0505596890428657, + "learning_rate": 5.988195239214835e-06, + "loss": 1.0431, + "step": 27270 + }, + { + "epoch": 0.64, + "grad_norm": 2.021107045953086, + "learning_rate": 5.987496306022297e-06, + "loss": 0.8902, + "step": 27271 + }, + { + "epoch": 0.64, + "grad_norm": 1.9280336592392624, + "learning_rate": 5.986797396191824e-06, + "loss": 0.9182, + "step": 27272 + }, + { + "epoch": 0.64, + "grad_norm": 2.068928738360845, + "learning_rate": 5.986098509727485e-06, + "loss": 0.8826, + "step": 27273 + }, + { + "epoch": 0.64, + "grad_norm": 1.089834945033851, + "learning_rate": 5.985399646633356e-06, + "loss": 1.016, + "step": 27274 + }, + { + "epoch": 0.64, + "grad_norm": 2.0228424209811005, + "learning_rate": 5.984700806913497e-06, + "loss": 0.9786, + "step": 27275 + }, + { + "epoch": 0.64, + "grad_norm": 1.9120514872373426, + "learning_rate": 5.9840019905719815e-06, + "loss": 1.012, + "step": 27276 + }, + { + "epoch": 0.64, + "grad_norm": 2.11450288494201, + "learning_rate": 5.983303197612878e-06, + "loss": 1.023, + "step": 27277 + }, + { + "epoch": 0.64, + "grad_norm": 1.8890784984327071, + "learning_rate": 5.9826044280402574e-06, + "loss": 0.9638, + "step": 27278 + }, + { + "epoch": 0.64, + "grad_norm": 2.1333115856388623, + "learning_rate": 5.981905681858184e-06, + "loss": 1.0249, + "step": 27279 + }, + { + "epoch": 0.64, + "grad_norm": 1.897402531166181, + "learning_rate": 5.981206959070728e-06, + "loss": 1.0479, + "step": 27280 + }, + { + "epoch": 0.64, + "grad_norm": 2.052982374605691, + "learning_rate": 5.980508259681958e-06, + "loss": 0.91, + "step": 27281 + }, + { + "epoch": 0.64, + "grad_norm": 1.805223216444014, + "learning_rate": 5.9798095836959405e-06, + "loss": 0.9957, + "step": 27282 + }, + { + "epoch": 0.64, + "grad_norm": 2.023491556893529, + "learning_rate": 5.979110931116744e-06, + "loss": 0.9656, + "step": 27283 + }, + { + "epoch": 0.64, + "grad_norm": 2.001471785505159, + "learning_rate": 5.978412301948438e-06, + "loss": 1.0137, + "step": 27284 + }, + { + "epoch": 0.64, + "grad_norm": 2.03794086816908, + "learning_rate": 5.977713696195089e-06, + "loss": 0.9878, + "step": 27285 + }, + { + "epoch": 0.64, + "grad_norm": 1.9637959939001826, + "learning_rate": 5.977015113860762e-06, + "loss": 1.0471, + "step": 27286 + }, + { + "epoch": 0.64, + "grad_norm": 2.2445138512185006, + "learning_rate": 5.9763165549495286e-06, + "loss": 0.9702, + "step": 27287 + }, + { + "epoch": 0.64, + "grad_norm": 2.2218858611678725, + "learning_rate": 5.975618019465454e-06, + "loss": 0.9896, + "step": 27288 + }, + { + "epoch": 0.64, + "grad_norm": 1.963148641649935, + "learning_rate": 5.974919507412607e-06, + "loss": 1.0441, + "step": 27289 + }, + { + "epoch": 0.64, + "grad_norm": 2.0536739313833863, + "learning_rate": 5.974221018795048e-06, + "loss": 0.9947, + "step": 27290 + }, + { + "epoch": 0.64, + "grad_norm": 1.9982522379305787, + "learning_rate": 5.9735225536168505e-06, + "loss": 0.9212, + "step": 27291 + }, + { + "epoch": 0.64, + "grad_norm": 2.1418632207450017, + "learning_rate": 5.972824111882083e-06, + "loss": 0.9297, + "step": 27292 + }, + { + "epoch": 0.64, + "grad_norm": 1.942560370487041, + "learning_rate": 5.972125693594806e-06, + "loss": 0.9006, + "step": 27293 + }, + { + "epoch": 0.64, + "grad_norm": 1.100475873157007, + "learning_rate": 5.971427298759087e-06, + "loss": 0.9174, + "step": 27294 + }, + { + "epoch": 0.64, + "grad_norm": 2.189226618451469, + "learning_rate": 5.970728927378997e-06, + "loss": 1.0147, + "step": 27295 + }, + { + "epoch": 0.64, + "grad_norm": 2.2754800630559147, + "learning_rate": 5.970030579458597e-06, + "loss": 1.027, + "step": 27296 + }, + { + "epoch": 0.64, + "grad_norm": 2.044337546673301, + "learning_rate": 5.969332255001955e-06, + "loss": 1.0113, + "step": 27297 + }, + { + "epoch": 0.64, + "grad_norm": 2.3194493713246223, + "learning_rate": 5.968633954013134e-06, + "loss": 0.8794, + "step": 27298 + }, + { + "epoch": 0.64, + "grad_norm": 1.9548763095718207, + "learning_rate": 5.967935676496208e-06, + "loss": 0.9425, + "step": 27299 + }, + { + "epoch": 0.64, + "grad_norm": 1.9741185228612683, + "learning_rate": 5.967237422455234e-06, + "loss": 1.0805, + "step": 27300 + }, + { + "epoch": 0.64, + "grad_norm": 1.8395346030765125, + "learning_rate": 5.966539191894279e-06, + "loss": 0.921, + "step": 27301 + }, + { + "epoch": 0.64, + "grad_norm": 1.9118379229374198, + "learning_rate": 5.965840984817415e-06, + "loss": 1.0253, + "step": 27302 + }, + { + "epoch": 0.64, + "grad_norm": 1.8843803434450388, + "learning_rate": 5.965142801228696e-06, + "loss": 1.0149, + "step": 27303 + }, + { + "epoch": 0.64, + "grad_norm": 2.133767965679497, + "learning_rate": 5.964444641132194e-06, + "loss": 0.9956, + "step": 27304 + }, + { + "epoch": 0.64, + "grad_norm": 2.0189265661576967, + "learning_rate": 5.9637465045319734e-06, + "loss": 0.9774, + "step": 27305 + }, + { + "epoch": 0.64, + "grad_norm": 1.9896908604965498, + "learning_rate": 5.963048391432102e-06, + "loss": 0.874, + "step": 27306 + }, + { + "epoch": 0.64, + "grad_norm": 2.066994240642726, + "learning_rate": 5.962350301836637e-06, + "loss": 1.0858, + "step": 27307 + }, + { + "epoch": 0.64, + "grad_norm": 2.100084800122097, + "learning_rate": 5.961652235749648e-06, + "loss": 1.0442, + "step": 27308 + }, + { + "epoch": 0.64, + "grad_norm": 1.15409384473285, + "learning_rate": 5.9609541931752015e-06, + "loss": 0.9525, + "step": 27309 + }, + { + "epoch": 0.64, + "grad_norm": 1.9558072526423314, + "learning_rate": 5.960256174117355e-06, + "loss": 0.8692, + "step": 27310 + }, + { + "epoch": 0.64, + "grad_norm": 1.9566739066844805, + "learning_rate": 5.959558178580174e-06, + "loss": 0.8709, + "step": 27311 + }, + { + "epoch": 0.64, + "grad_norm": 1.0590817890940802, + "learning_rate": 5.958860206567727e-06, + "loss": 0.9726, + "step": 27312 + }, + { + "epoch": 0.64, + "grad_norm": 2.539795210736636, + "learning_rate": 5.958162258084077e-06, + "loss": 0.8409, + "step": 27313 + }, + { + "epoch": 0.64, + "grad_norm": 2.0134759728111407, + "learning_rate": 5.957464333133282e-06, + "loss": 1.1147, + "step": 27314 + }, + { + "epoch": 0.64, + "grad_norm": 2.0704533387359, + "learning_rate": 5.9567664317194115e-06, + "loss": 0.9016, + "step": 27315 + }, + { + "epoch": 0.64, + "grad_norm": 1.1094948716865316, + "learning_rate": 5.956068553846528e-06, + "loss": 0.923, + "step": 27316 + }, + { + "epoch": 0.64, + "grad_norm": 2.1708697773169403, + "learning_rate": 5.955370699518694e-06, + "loss": 1.0221, + "step": 27317 + }, + { + "epoch": 0.64, + "grad_norm": 1.8606432578418306, + "learning_rate": 5.954672868739968e-06, + "loss": 0.9514, + "step": 27318 + }, + { + "epoch": 0.64, + "grad_norm": 2.0405972209338707, + "learning_rate": 5.95397506151442e-06, + "loss": 0.9987, + "step": 27319 + }, + { + "epoch": 0.64, + "grad_norm": 2.608582949994452, + "learning_rate": 5.953277277846112e-06, + "loss": 0.9317, + "step": 27320 + }, + { + "epoch": 0.64, + "grad_norm": 1.8170265160323114, + "learning_rate": 5.952579517739103e-06, + "loss": 0.8915, + "step": 27321 + }, + { + "epoch": 0.64, + "grad_norm": 1.9350218587113064, + "learning_rate": 5.951881781197456e-06, + "loss": 0.9372, + "step": 27322 + }, + { + "epoch": 0.64, + "grad_norm": 2.1287413942307425, + "learning_rate": 5.95118406822524e-06, + "loss": 0.9873, + "step": 27323 + }, + { + "epoch": 0.64, + "grad_norm": 2.1907587424991815, + "learning_rate": 5.950486378826508e-06, + "loss": 1.1156, + "step": 27324 + }, + { + "epoch": 0.64, + "grad_norm": 2.0052620202340745, + "learning_rate": 5.949788713005326e-06, + "loss": 0.8592, + "step": 27325 + }, + { + "epoch": 0.64, + "grad_norm": 1.8797711396179062, + "learning_rate": 5.949091070765755e-06, + "loss": 1.1005, + "step": 27326 + }, + { + "epoch": 0.64, + "grad_norm": 1.1437981219783138, + "learning_rate": 5.948393452111863e-06, + "loss": 0.787, + "step": 27327 + }, + { + "epoch": 0.64, + "grad_norm": 1.8903344975855612, + "learning_rate": 5.9476958570477036e-06, + "loss": 0.8385, + "step": 27328 + }, + { + "epoch": 0.64, + "grad_norm": 2.055517171019728, + "learning_rate": 5.946998285577343e-06, + "loss": 0.8744, + "step": 27329 + }, + { + "epoch": 0.64, + "grad_norm": 2.1503098398620284, + "learning_rate": 5.946300737704845e-06, + "loss": 0.9882, + "step": 27330 + }, + { + "epoch": 0.64, + "grad_norm": 1.9621370097733164, + "learning_rate": 5.945603213434261e-06, + "loss": 1.0677, + "step": 27331 + }, + { + "epoch": 0.64, + "grad_norm": 2.113331722522326, + "learning_rate": 5.9449057127696616e-06, + "loss": 1.0406, + "step": 27332 + }, + { + "epoch": 0.64, + "grad_norm": 2.120462172091649, + "learning_rate": 5.944208235715105e-06, + "loss": 0.8836, + "step": 27333 + }, + { + "epoch": 0.64, + "grad_norm": 2.0201547561052173, + "learning_rate": 5.943510782274655e-06, + "loss": 0.9163, + "step": 27334 + }, + { + "epoch": 0.64, + "grad_norm": 2.0395176023264505, + "learning_rate": 5.942813352452365e-06, + "loss": 1.0286, + "step": 27335 + }, + { + "epoch": 0.64, + "grad_norm": 2.321650008992858, + "learning_rate": 5.9421159462523e-06, + "loss": 0.9918, + "step": 27336 + }, + { + "epoch": 0.64, + "grad_norm": 2.5733175094319005, + "learning_rate": 5.941418563678525e-06, + "loss": 1.0548, + "step": 27337 + }, + { + "epoch": 0.64, + "grad_norm": 2.4171677329152743, + "learning_rate": 5.940721204735094e-06, + "loss": 1.0459, + "step": 27338 + }, + { + "epoch": 0.64, + "grad_norm": 2.283657806774085, + "learning_rate": 5.940023869426067e-06, + "loss": 0.9619, + "step": 27339 + }, + { + "epoch": 0.64, + "grad_norm": 2.643505208620218, + "learning_rate": 5.939326557755508e-06, + "loss": 1.0304, + "step": 27340 + }, + { + "epoch": 0.64, + "grad_norm": 1.842226982588423, + "learning_rate": 5.938629269727477e-06, + "loss": 0.9475, + "step": 27341 + }, + { + "epoch": 0.64, + "grad_norm": 1.9386731773520436, + "learning_rate": 5.937932005346031e-06, + "loss": 1.0619, + "step": 27342 + }, + { + "epoch": 0.64, + "grad_norm": 2.082160537406641, + "learning_rate": 5.93723476461523e-06, + "loss": 1.0339, + "step": 27343 + }, + { + "epoch": 0.64, + "grad_norm": 1.9731774333295948, + "learning_rate": 5.936537547539136e-06, + "loss": 0.9336, + "step": 27344 + }, + { + "epoch": 0.64, + "grad_norm": 1.0684134425633696, + "learning_rate": 5.9358403541218055e-06, + "loss": 0.9036, + "step": 27345 + }, + { + "epoch": 0.64, + "grad_norm": 2.0751401321704184, + "learning_rate": 5.9351431843673e-06, + "loss": 0.8956, + "step": 27346 + }, + { + "epoch": 0.64, + "grad_norm": 1.7839161929371101, + "learning_rate": 5.934446038279676e-06, + "loss": 1.0024, + "step": 27347 + }, + { + "epoch": 0.64, + "grad_norm": 2.204153510623638, + "learning_rate": 5.933748915862997e-06, + "loss": 1.0501, + "step": 27348 + }, + { + "epoch": 0.64, + "grad_norm": 2.287069519149325, + "learning_rate": 5.933051817121318e-06, + "loss": 0.9209, + "step": 27349 + }, + { + "epoch": 0.64, + "grad_norm": 1.9763140658920426, + "learning_rate": 5.9323547420586965e-06, + "loss": 0.9472, + "step": 27350 + }, + { + "epoch": 0.64, + "grad_norm": 2.366153889632363, + "learning_rate": 5.9316576906792e-06, + "loss": 0.9701, + "step": 27351 + }, + { + "epoch": 0.64, + "grad_norm": 1.9007289093236581, + "learning_rate": 5.930960662986873e-06, + "loss": 1.0171, + "step": 27352 + }, + { + "epoch": 0.64, + "grad_norm": 1.873996633233112, + "learning_rate": 5.930263658985785e-06, + "loss": 1.0032, + "step": 27353 + }, + { + "epoch": 0.64, + "grad_norm": 1.792330476924904, + "learning_rate": 5.9295666786799874e-06, + "loss": 0.9956, + "step": 27354 + }, + { + "epoch": 0.64, + "grad_norm": 1.9973593933995704, + "learning_rate": 5.928869722073549e-06, + "loss": 0.9201, + "step": 27355 + }, + { + "epoch": 0.64, + "grad_norm": 1.8299978000942443, + "learning_rate": 5.928172789170512e-06, + "loss": 0.9886, + "step": 27356 + }, + { + "epoch": 0.64, + "grad_norm": 1.0955321113471261, + "learning_rate": 5.927475879974946e-06, + "loss": 0.9954, + "step": 27357 + }, + { + "epoch": 0.64, + "grad_norm": 2.055774963903069, + "learning_rate": 5.926778994490905e-06, + "loss": 1.0732, + "step": 27358 + }, + { + "epoch": 0.64, + "grad_norm": 1.8553219499363613, + "learning_rate": 5.926082132722445e-06, + "loss": 1.0063, + "step": 27359 + }, + { + "epoch": 0.64, + "grad_norm": 2.3080026889381635, + "learning_rate": 5.925385294673623e-06, + "loss": 1.0716, + "step": 27360 + }, + { + "epoch": 0.64, + "grad_norm": 5.306932830689402, + "learning_rate": 5.924688480348498e-06, + "loss": 1.1735, + "step": 27361 + }, + { + "epoch": 0.64, + "grad_norm": 2.0609374573322925, + "learning_rate": 5.92399168975113e-06, + "loss": 0.9858, + "step": 27362 + }, + { + "epoch": 0.64, + "grad_norm": 2.488306069643299, + "learning_rate": 5.923294922885569e-06, + "loss": 0.9447, + "step": 27363 + }, + { + "epoch": 0.64, + "grad_norm": 2.0667016443790716, + "learning_rate": 5.922598179755879e-06, + "loss": 0.9552, + "step": 27364 + }, + { + "epoch": 0.64, + "grad_norm": 3.0961728100713777, + "learning_rate": 5.921901460366113e-06, + "loss": 1.0044, + "step": 27365 + }, + { + "epoch": 0.64, + "grad_norm": 2.1514038900118204, + "learning_rate": 5.921204764720327e-06, + "loss": 0.9524, + "step": 27366 + }, + { + "epoch": 0.64, + "grad_norm": 1.9129681547310513, + "learning_rate": 5.920508092822575e-06, + "loss": 0.9283, + "step": 27367 + }, + { + "epoch": 0.64, + "grad_norm": 1.1556848372306556, + "learning_rate": 5.919811444676919e-06, + "loss": 0.9793, + "step": 27368 + }, + { + "epoch": 0.64, + "grad_norm": 1.1218435126324857, + "learning_rate": 5.919114820287416e-06, + "loss": 0.9747, + "step": 27369 + }, + { + "epoch": 0.64, + "grad_norm": 2.6476825108144575, + "learning_rate": 5.918418219658115e-06, + "loss": 1.0449, + "step": 27370 + }, + { + "epoch": 0.64, + "grad_norm": 2.0334511937311754, + "learning_rate": 5.917721642793074e-06, + "loss": 0.9735, + "step": 27371 + }, + { + "epoch": 0.64, + "grad_norm": 2.71787597397373, + "learning_rate": 5.917025089696355e-06, + "loss": 0.946, + "step": 27372 + }, + { + "epoch": 0.64, + "grad_norm": 1.9752667984723684, + "learning_rate": 5.916328560372004e-06, + "loss": 1.0661, + "step": 27373 + }, + { + "epoch": 0.64, + "grad_norm": 5.524044990180547, + "learning_rate": 5.915632054824083e-06, + "loss": 0.9103, + "step": 27374 + }, + { + "epoch": 0.64, + "grad_norm": 2.1560889294923116, + "learning_rate": 5.914935573056641e-06, + "loss": 0.859, + "step": 27375 + }, + { + "epoch": 0.64, + "grad_norm": 1.1145730124442106, + "learning_rate": 5.914239115073743e-06, + "loss": 0.9151, + "step": 27376 + }, + { + "epoch": 0.64, + "grad_norm": 2.114782808240999, + "learning_rate": 5.913542680879437e-06, + "loss": 1.1149, + "step": 27377 + }, + { + "epoch": 0.65, + "grad_norm": 1.1359306707547496, + "learning_rate": 5.912846270477779e-06, + "loss": 0.9718, + "step": 27378 + }, + { + "epoch": 0.65, + "grad_norm": 2.160901780735742, + "learning_rate": 5.912149883872822e-06, + "loss": 0.96, + "step": 27379 + }, + { + "epoch": 0.65, + "grad_norm": 1.9742618666753458, + "learning_rate": 5.911453521068626e-06, + "loss": 0.9704, + "step": 27380 + }, + { + "epoch": 0.65, + "grad_norm": 1.0835970351950468, + "learning_rate": 5.91075718206924e-06, + "loss": 1.0098, + "step": 27381 + }, + { + "epoch": 0.65, + "grad_norm": 1.8105375661708405, + "learning_rate": 5.910060866878719e-06, + "loss": 1.003, + "step": 27382 + }, + { + "epoch": 0.65, + "grad_norm": 2.0941098813767645, + "learning_rate": 5.909364575501123e-06, + "loss": 1.0026, + "step": 27383 + }, + { + "epoch": 0.65, + "grad_norm": 2.1947612292487553, + "learning_rate": 5.908668307940497e-06, + "loss": 1.0428, + "step": 27384 + }, + { + "epoch": 0.65, + "grad_norm": 1.9545057619589234, + "learning_rate": 5.9079720642009004e-06, + "loss": 1.0491, + "step": 27385 + }, + { + "epoch": 0.65, + "grad_norm": 2.315676229352013, + "learning_rate": 5.907275844286386e-06, + "loss": 0.962, + "step": 27386 + }, + { + "epoch": 0.65, + "grad_norm": 1.9745962168918274, + "learning_rate": 5.9065796482010104e-06, + "loss": 1.0248, + "step": 27387 + }, + { + "epoch": 0.65, + "grad_norm": 2.0347487690158403, + "learning_rate": 5.905883475948819e-06, + "loss": 0.9217, + "step": 27388 + }, + { + "epoch": 0.65, + "grad_norm": 2.002551547430307, + "learning_rate": 5.905187327533873e-06, + "loss": 1.0329, + "step": 27389 + }, + { + "epoch": 0.65, + "grad_norm": 2.1882137815751292, + "learning_rate": 5.904491202960225e-06, + "loss": 0.9981, + "step": 27390 + }, + { + "epoch": 0.65, + "grad_norm": 2.1055614834960528, + "learning_rate": 5.903795102231923e-06, + "loss": 1.0547, + "step": 27391 + }, + { + "epoch": 0.65, + "grad_norm": 2.0028405065039583, + "learning_rate": 5.903099025353021e-06, + "loss": 0.9849, + "step": 27392 + }, + { + "epoch": 0.65, + "grad_norm": 1.8315502979932283, + "learning_rate": 5.902402972327575e-06, + "loss": 0.895, + "step": 27393 + }, + { + "epoch": 0.65, + "grad_norm": 2.5117534179727365, + "learning_rate": 5.9017069431596395e-06, + "loss": 1.0843, + "step": 27394 + }, + { + "epoch": 0.65, + "grad_norm": 1.9874444190724265, + "learning_rate": 5.901010937853261e-06, + "loss": 0.943, + "step": 27395 + }, + { + "epoch": 0.65, + "grad_norm": 1.0950014488556146, + "learning_rate": 5.900314956412492e-06, + "loss": 0.9545, + "step": 27396 + }, + { + "epoch": 0.65, + "grad_norm": 1.9219872979272747, + "learning_rate": 5.8996189988413925e-06, + "loss": 0.8913, + "step": 27397 + }, + { + "epoch": 0.65, + "grad_norm": 2.090763277009082, + "learning_rate": 5.898923065144006e-06, + "loss": 1.0875, + "step": 27398 + }, + { + "epoch": 0.65, + "grad_norm": 1.7579904953589276, + "learning_rate": 5.8982271553243866e-06, + "loss": 0.9754, + "step": 27399 + }, + { + "epoch": 0.65, + "grad_norm": 1.9570384435892667, + "learning_rate": 5.89753126938659e-06, + "loss": 1.0435, + "step": 27400 + }, + { + "epoch": 0.65, + "grad_norm": 2.307501227596725, + "learning_rate": 5.896835407334665e-06, + "loss": 1.0651, + "step": 27401 + }, + { + "epoch": 0.65, + "grad_norm": 1.9650323607344984, + "learning_rate": 5.8961395691726635e-06, + "loss": 0.8521, + "step": 27402 + }, + { + "epoch": 0.65, + "grad_norm": 2.221893248825977, + "learning_rate": 5.8954437549046325e-06, + "loss": 0.9026, + "step": 27403 + }, + { + "epoch": 0.65, + "grad_norm": 1.0998310602290025, + "learning_rate": 5.894747964534635e-06, + "loss": 0.9409, + "step": 27404 + }, + { + "epoch": 0.65, + "grad_norm": 2.500601921691168, + "learning_rate": 5.8940521980667086e-06, + "loss": 0.9786, + "step": 27405 + }, + { + "epoch": 0.65, + "grad_norm": 1.927778015311869, + "learning_rate": 5.893356455504911e-06, + "loss": 0.8996, + "step": 27406 + }, + { + "epoch": 0.65, + "grad_norm": 2.046824529031041, + "learning_rate": 5.892660736853291e-06, + "loss": 1.0873, + "step": 27407 + }, + { + "epoch": 0.65, + "grad_norm": 2.028953401444918, + "learning_rate": 5.891965042115907e-06, + "loss": 1.142, + "step": 27408 + }, + { + "epoch": 0.65, + "grad_norm": 1.985956451926228, + "learning_rate": 5.891269371296798e-06, + "loss": 1.0207, + "step": 27409 + }, + { + "epoch": 0.65, + "grad_norm": 2.1110661640343737, + "learning_rate": 5.89057372440002e-06, + "loss": 1.0501, + "step": 27410 + }, + { + "epoch": 0.65, + "grad_norm": 2.074346418322113, + "learning_rate": 5.889878101429624e-06, + "loss": 0.961, + "step": 27411 + }, + { + "epoch": 0.65, + "grad_norm": 2.117994045642447, + "learning_rate": 5.889182502389657e-06, + "loss": 1.0772, + "step": 27412 + }, + { + "epoch": 0.65, + "grad_norm": 2.1257514485617985, + "learning_rate": 5.88848692728417e-06, + "loss": 0.9436, + "step": 27413 + }, + { + "epoch": 0.65, + "grad_norm": 1.864133598651644, + "learning_rate": 5.887791376117216e-06, + "loss": 0.9353, + "step": 27414 + }, + { + "epoch": 0.65, + "grad_norm": 1.8440418768802096, + "learning_rate": 5.887095848892846e-06, + "loss": 0.9948, + "step": 27415 + }, + { + "epoch": 0.65, + "grad_norm": 1.943923807245593, + "learning_rate": 5.886400345615098e-06, + "loss": 1.0407, + "step": 27416 + }, + { + "epoch": 0.65, + "grad_norm": 1.11644875732975, + "learning_rate": 5.885704866288034e-06, + "loss": 0.9856, + "step": 27417 + }, + { + "epoch": 0.65, + "grad_norm": 2.029761120165467, + "learning_rate": 5.885009410915701e-06, + "loss": 1.007, + "step": 27418 + }, + { + "epoch": 0.65, + "grad_norm": 2.458739234252535, + "learning_rate": 5.8843139795021446e-06, + "loss": 1.0282, + "step": 27419 + }, + { + "epoch": 0.65, + "grad_norm": 1.121916495205954, + "learning_rate": 5.883618572051411e-06, + "loss": 1.0167, + "step": 27420 + }, + { + "epoch": 0.65, + "grad_norm": 2.2828155387251785, + "learning_rate": 5.882923188567558e-06, + "loss": 0.8485, + "step": 27421 + }, + { + "epoch": 0.65, + "grad_norm": 2.1524040025292903, + "learning_rate": 5.88222782905463e-06, + "loss": 0.952, + "step": 27422 + }, + { + "epoch": 0.65, + "grad_norm": 1.1168276259909333, + "learning_rate": 5.881532493516673e-06, + "loss": 0.9336, + "step": 27423 + }, + { + "epoch": 0.65, + "grad_norm": 2.5574383695904284, + "learning_rate": 5.880837181957737e-06, + "loss": 0.9056, + "step": 27424 + }, + { + "epoch": 0.65, + "grad_norm": 1.8642276943325697, + "learning_rate": 5.880141894381873e-06, + "loss": 1.0015, + "step": 27425 + }, + { + "epoch": 0.65, + "grad_norm": 2.0073657260037607, + "learning_rate": 5.879446630793126e-06, + "loss": 0.9927, + "step": 27426 + }, + { + "epoch": 0.65, + "grad_norm": 1.1135256371459152, + "learning_rate": 5.878751391195546e-06, + "loss": 0.8998, + "step": 27427 + }, + { + "epoch": 0.65, + "grad_norm": 2.4226029192844116, + "learning_rate": 5.8780561755931785e-06, + "loss": 1.0102, + "step": 27428 + }, + { + "epoch": 0.65, + "grad_norm": 1.8360616087815014, + "learning_rate": 5.877360983990077e-06, + "loss": 1.0173, + "step": 27429 + }, + { + "epoch": 0.65, + "grad_norm": 2.2694526746372805, + "learning_rate": 5.876665816390281e-06, + "loss": 0.967, + "step": 27430 + }, + { + "epoch": 0.65, + "grad_norm": 1.9717215711416023, + "learning_rate": 5.875970672797844e-06, + "loss": 0.9541, + "step": 27431 + }, + { + "epoch": 0.65, + "grad_norm": 2.009375792625747, + "learning_rate": 5.8752755532168125e-06, + "loss": 0.8537, + "step": 27432 + }, + { + "epoch": 0.65, + "grad_norm": 2.940369482945957, + "learning_rate": 5.874580457651229e-06, + "loss": 1.0617, + "step": 27433 + }, + { + "epoch": 0.65, + "grad_norm": 2.089425927244749, + "learning_rate": 5.8738853861051455e-06, + "loss": 0.9946, + "step": 27434 + }, + { + "epoch": 0.65, + "grad_norm": 1.947133170365576, + "learning_rate": 5.873190338582608e-06, + "loss": 0.8753, + "step": 27435 + }, + { + "epoch": 0.65, + "grad_norm": 2.3985985911894314, + "learning_rate": 5.8724953150876655e-06, + "loss": 0.9945, + "step": 27436 + }, + { + "epoch": 0.65, + "grad_norm": 2.119207225963754, + "learning_rate": 5.871800315624357e-06, + "loss": 1.0691, + "step": 27437 + }, + { + "epoch": 0.65, + "grad_norm": 2.1441225616130644, + "learning_rate": 5.871105340196737e-06, + "loss": 0.8829, + "step": 27438 + }, + { + "epoch": 0.65, + "grad_norm": 2.0944520583165707, + "learning_rate": 5.87041038880885e-06, + "loss": 0.9171, + "step": 27439 + }, + { + "epoch": 0.65, + "grad_norm": 1.864060463533504, + "learning_rate": 5.86971546146474e-06, + "loss": 0.8685, + "step": 27440 + }, + { + "epoch": 0.65, + "grad_norm": 2.186528007699331, + "learning_rate": 5.869020558168452e-06, + "loss": 0.9757, + "step": 27441 + }, + { + "epoch": 0.65, + "grad_norm": 1.9434131764028435, + "learning_rate": 5.868325678924036e-06, + "loss": 0.9225, + "step": 27442 + }, + { + "epoch": 0.65, + "grad_norm": 1.9369905821995923, + "learning_rate": 5.867630823735537e-06, + "loss": 1.0596, + "step": 27443 + }, + { + "epoch": 0.65, + "grad_norm": 2.1960994851510076, + "learning_rate": 5.866935992606999e-06, + "loss": 0.9908, + "step": 27444 + }, + { + "epoch": 0.65, + "grad_norm": 1.8820737255225262, + "learning_rate": 5.866241185542466e-06, + "loss": 0.9905, + "step": 27445 + }, + { + "epoch": 0.65, + "grad_norm": 2.2404623768867324, + "learning_rate": 5.865546402545989e-06, + "loss": 1.0136, + "step": 27446 + }, + { + "epoch": 0.65, + "grad_norm": 2.0071179511433375, + "learning_rate": 5.864851643621607e-06, + "loss": 1.0106, + "step": 27447 + }, + { + "epoch": 0.65, + "grad_norm": 3.8078513703567936, + "learning_rate": 5.864156908773369e-06, + "loss": 0.9507, + "step": 27448 + }, + { + "epoch": 0.65, + "grad_norm": 2.235577631043428, + "learning_rate": 5.863462198005318e-06, + "loss": 0.9411, + "step": 27449 + }, + { + "epoch": 0.65, + "grad_norm": 1.1771127952855889, + "learning_rate": 5.862767511321502e-06, + "loss": 0.9518, + "step": 27450 + }, + { + "epoch": 0.65, + "grad_norm": 4.643709073974975, + "learning_rate": 5.862072848725961e-06, + "loss": 0.9558, + "step": 27451 + }, + { + "epoch": 0.65, + "grad_norm": 2.095969995609799, + "learning_rate": 5.8613782102227416e-06, + "loss": 1.0431, + "step": 27452 + }, + { + "epoch": 0.65, + "grad_norm": 2.0166448593109974, + "learning_rate": 5.860683595815894e-06, + "loss": 0.9536, + "step": 27453 + }, + { + "epoch": 0.65, + "grad_norm": 2.0251816326308063, + "learning_rate": 5.85998900550945e-06, + "loss": 0.9938, + "step": 27454 + }, + { + "epoch": 0.65, + "grad_norm": 2.038918319944946, + "learning_rate": 5.8592944393074634e-06, + "loss": 0.9375, + "step": 27455 + }, + { + "epoch": 0.65, + "grad_norm": 1.925124712798742, + "learning_rate": 5.858599897213975e-06, + "loss": 1.0108, + "step": 27456 + }, + { + "epoch": 0.65, + "grad_norm": 1.957019405176862, + "learning_rate": 5.857905379233033e-06, + "loss": 0.9715, + "step": 27457 + }, + { + "epoch": 0.65, + "grad_norm": 1.9071242946072067, + "learning_rate": 5.857210885368672e-06, + "loss": 0.8681, + "step": 27458 + }, + { + "epoch": 0.65, + "grad_norm": 1.9075257179756748, + "learning_rate": 5.856516415624942e-06, + "loss": 1.0936, + "step": 27459 + }, + { + "epoch": 0.65, + "grad_norm": 1.1652304556293476, + "learning_rate": 5.855821970005888e-06, + "loss": 0.9551, + "step": 27460 + }, + { + "epoch": 0.65, + "grad_norm": 2.202434231024144, + "learning_rate": 5.855127548515547e-06, + "loss": 1.1211, + "step": 27461 + }, + { + "epoch": 0.65, + "grad_norm": 2.1386253604446583, + "learning_rate": 5.8544331511579664e-06, + "loss": 1.0712, + "step": 27462 + }, + { + "epoch": 0.65, + "grad_norm": 3.2288921365843657, + "learning_rate": 5.85373877793719e-06, + "loss": 1.0421, + "step": 27463 + }, + { + "epoch": 0.65, + "grad_norm": 2.08724996702164, + "learning_rate": 5.8530444288572595e-06, + "loss": 1.0041, + "step": 27464 + }, + { + "epoch": 0.65, + "grad_norm": 1.9502312478301465, + "learning_rate": 5.852350103922214e-06, + "loss": 0.9251, + "step": 27465 + }, + { + "epoch": 0.65, + "grad_norm": 1.0512150808169205, + "learning_rate": 5.851655803136101e-06, + "loss": 0.9494, + "step": 27466 + }, + { + "epoch": 0.65, + "grad_norm": 1.989477996199769, + "learning_rate": 5.850961526502964e-06, + "loss": 0.8968, + "step": 27467 + }, + { + "epoch": 0.65, + "grad_norm": 2.0733565630473816, + "learning_rate": 5.850267274026839e-06, + "loss": 0.8632, + "step": 27468 + }, + { + "epoch": 0.65, + "grad_norm": 2.155115440131855, + "learning_rate": 5.849573045711771e-06, + "loss": 1.1692, + "step": 27469 + }, + { + "epoch": 0.65, + "grad_norm": 1.9412119804676593, + "learning_rate": 5.848878841561805e-06, + "loss": 0.8444, + "step": 27470 + }, + { + "epoch": 0.65, + "grad_norm": 8.84670795715027, + "learning_rate": 5.848184661580982e-06, + "loss": 0.9745, + "step": 27471 + }, + { + "epoch": 0.65, + "grad_norm": 2.2012500452886985, + "learning_rate": 5.847490505773341e-06, + "loss": 0.8066, + "step": 27472 + }, + { + "epoch": 0.65, + "grad_norm": 1.9419847815856859, + "learning_rate": 5.846796374142923e-06, + "loss": 0.9427, + "step": 27473 + }, + { + "epoch": 0.65, + "grad_norm": 2.567394407923774, + "learning_rate": 5.8461022666937735e-06, + "loss": 0.8717, + "step": 27474 + }, + { + "epoch": 0.65, + "grad_norm": 1.9446997687698946, + "learning_rate": 5.8454081834299324e-06, + "loss": 0.9729, + "step": 27475 + }, + { + "epoch": 0.65, + "grad_norm": 8.590403524137189, + "learning_rate": 5.844714124355438e-06, + "loss": 1.0772, + "step": 27476 + }, + { + "epoch": 0.65, + "grad_norm": 2.1809457747723586, + "learning_rate": 5.844020089474334e-06, + "loss": 1.0925, + "step": 27477 + }, + { + "epoch": 0.65, + "grad_norm": 1.9849202224146363, + "learning_rate": 5.8433260787906624e-06, + "loss": 0.8975, + "step": 27478 + }, + { + "epoch": 0.65, + "grad_norm": 2.07971410196499, + "learning_rate": 5.84263209230846e-06, + "loss": 1.1295, + "step": 27479 + }, + { + "epoch": 0.65, + "grad_norm": 1.8055421879386369, + "learning_rate": 5.841938130031772e-06, + "loss": 0.9973, + "step": 27480 + }, + { + "epoch": 0.65, + "grad_norm": 2.206751572560944, + "learning_rate": 5.841244191964637e-06, + "loss": 0.9062, + "step": 27481 + }, + { + "epoch": 0.65, + "grad_norm": 2.2199616529019512, + "learning_rate": 5.8405502781110925e-06, + "loss": 0.9697, + "step": 27482 + }, + { + "epoch": 0.65, + "grad_norm": 1.8204657246859033, + "learning_rate": 5.839856388475182e-06, + "loss": 0.9294, + "step": 27483 + }, + { + "epoch": 0.65, + "grad_norm": 2.4944472684531322, + "learning_rate": 5.839162523060944e-06, + "loss": 0.8764, + "step": 27484 + }, + { + "epoch": 0.65, + "grad_norm": 1.944747171089729, + "learning_rate": 5.8384686818724205e-06, + "loss": 1.0458, + "step": 27485 + }, + { + "epoch": 0.65, + "grad_norm": 2.075330611586683, + "learning_rate": 5.83777486491365e-06, + "loss": 1.1113, + "step": 27486 + }, + { + "epoch": 0.65, + "grad_norm": 1.1081141873051403, + "learning_rate": 5.837081072188669e-06, + "loss": 0.9189, + "step": 27487 + }, + { + "epoch": 0.65, + "grad_norm": 1.9929155527029823, + "learning_rate": 5.836387303701525e-06, + "loss": 1.1529, + "step": 27488 + }, + { + "epoch": 0.65, + "grad_norm": 2.1929926325741387, + "learning_rate": 5.835693559456248e-06, + "loss": 0.9844, + "step": 27489 + }, + { + "epoch": 0.65, + "grad_norm": 2.177710993388947, + "learning_rate": 5.8349998394568845e-06, + "loss": 1.0196, + "step": 27490 + }, + { + "epoch": 0.65, + "grad_norm": 1.8037623976154922, + "learning_rate": 5.834306143707467e-06, + "loss": 1.0511, + "step": 27491 + }, + { + "epoch": 0.65, + "grad_norm": 2.0882156515405814, + "learning_rate": 5.833612472212043e-06, + "loss": 1.0864, + "step": 27492 + }, + { + "epoch": 0.65, + "grad_norm": 1.8150284161055845, + "learning_rate": 5.83291882497464e-06, + "loss": 0.8769, + "step": 27493 + }, + { + "epoch": 0.65, + "grad_norm": 2.0889098967590547, + "learning_rate": 5.832225201999308e-06, + "loss": 1.0309, + "step": 27494 + }, + { + "epoch": 0.65, + "grad_norm": 2.053696073697348, + "learning_rate": 5.83153160329008e-06, + "loss": 0.9614, + "step": 27495 + }, + { + "epoch": 0.65, + "grad_norm": 2.023128441387889, + "learning_rate": 5.830838028850992e-06, + "loss": 1.0515, + "step": 27496 + }, + { + "epoch": 0.65, + "grad_norm": 1.8323091271310659, + "learning_rate": 5.830144478686084e-06, + "loss": 0.9166, + "step": 27497 + }, + { + "epoch": 0.65, + "grad_norm": 1.94349238141811, + "learning_rate": 5.829450952799398e-06, + "loss": 0.9519, + "step": 27498 + }, + { + "epoch": 0.65, + "grad_norm": 2.003767477332509, + "learning_rate": 5.828757451194971e-06, + "loss": 0.9754, + "step": 27499 + }, + { + "epoch": 0.65, + "grad_norm": 1.8248979208586376, + "learning_rate": 5.828063973876834e-06, + "loss": 1.0403, + "step": 27500 + }, + { + "epoch": 0.65, + "grad_norm": 1.8961895309335235, + "learning_rate": 5.827370520849029e-06, + "loss": 1.0861, + "step": 27501 + }, + { + "epoch": 0.65, + "grad_norm": 2.060265579559881, + "learning_rate": 5.826677092115598e-06, + "loss": 0.8316, + "step": 27502 + }, + { + "epoch": 0.65, + "grad_norm": 2.1830765611457714, + "learning_rate": 5.8259836876805745e-06, + "loss": 1.1385, + "step": 27503 + }, + { + "epoch": 0.65, + "grad_norm": 2.376427611558431, + "learning_rate": 5.82529030754799e-06, + "loss": 1.003, + "step": 27504 + }, + { + "epoch": 0.65, + "grad_norm": 1.9603631537513677, + "learning_rate": 5.824596951721888e-06, + "loss": 1.1355, + "step": 27505 + }, + { + "epoch": 0.65, + "grad_norm": 2.1868473837811697, + "learning_rate": 5.823903620206308e-06, + "loss": 0.921, + "step": 27506 + }, + { + "epoch": 0.65, + "grad_norm": 1.9975413607504626, + "learning_rate": 5.8232103130052855e-06, + "loss": 0.9674, + "step": 27507 + }, + { + "epoch": 0.65, + "grad_norm": 2.151334184509482, + "learning_rate": 5.8225170301228485e-06, + "loss": 0.9675, + "step": 27508 + }, + { + "epoch": 0.65, + "grad_norm": 2.3199220142735397, + "learning_rate": 5.8218237715630396e-06, + "loss": 0.938, + "step": 27509 + }, + { + "epoch": 0.65, + "grad_norm": 1.854170895340715, + "learning_rate": 5.821130537329902e-06, + "loss": 0.9701, + "step": 27510 + }, + { + "epoch": 0.65, + "grad_norm": 2.139729821782779, + "learning_rate": 5.820437327427464e-06, + "loss": 0.9796, + "step": 27511 + }, + { + "epoch": 0.65, + "grad_norm": 1.9694706399365873, + "learning_rate": 5.819744141859759e-06, + "loss": 0.9896, + "step": 27512 + }, + { + "epoch": 0.65, + "grad_norm": 1.9581206985999262, + "learning_rate": 5.8190509806308316e-06, + "loss": 0.8464, + "step": 27513 + }, + { + "epoch": 0.65, + "grad_norm": 2.113388323726536, + "learning_rate": 5.8183578437447085e-06, + "loss": 0.8746, + "step": 27514 + }, + { + "epoch": 0.65, + "grad_norm": 1.9672881084011913, + "learning_rate": 5.817664731205429e-06, + "loss": 1.0686, + "step": 27515 + }, + { + "epoch": 0.65, + "grad_norm": 2.078127990649696, + "learning_rate": 5.816971643017036e-06, + "loss": 0.9613, + "step": 27516 + }, + { + "epoch": 0.65, + "grad_norm": 1.9777677400260214, + "learning_rate": 5.816278579183556e-06, + "loss": 0.9824, + "step": 27517 + }, + { + "epoch": 0.65, + "grad_norm": 1.7507151847685545, + "learning_rate": 5.815585539709023e-06, + "loss": 1.0207, + "step": 27518 + }, + { + "epoch": 0.65, + "grad_norm": 3.7969563266007316, + "learning_rate": 5.814892524597477e-06, + "loss": 0.88, + "step": 27519 + }, + { + "epoch": 0.65, + "grad_norm": 1.0772482303985162, + "learning_rate": 5.814199533852959e-06, + "loss": 0.9152, + "step": 27520 + }, + { + "epoch": 0.65, + "grad_norm": 1.7022887844774586, + "learning_rate": 5.8135065674794885e-06, + "loss": 0.9771, + "step": 27521 + }, + { + "epoch": 0.65, + "grad_norm": 2.1773981023972486, + "learning_rate": 5.812813625481108e-06, + "loss": 0.9513, + "step": 27522 + }, + { + "epoch": 0.65, + "grad_norm": 2.4696017201147735, + "learning_rate": 5.812120707861852e-06, + "loss": 0.9713, + "step": 27523 + }, + { + "epoch": 0.65, + "grad_norm": 1.07308755247443, + "learning_rate": 5.811427814625765e-06, + "loss": 1.0054, + "step": 27524 + }, + { + "epoch": 0.65, + "grad_norm": 1.7978002157162318, + "learning_rate": 5.8107349457768615e-06, + "loss": 0.8319, + "step": 27525 + }, + { + "epoch": 0.65, + "grad_norm": 2.022304803222096, + "learning_rate": 5.810042101319185e-06, + "loss": 0.9489, + "step": 27526 + }, + { + "epoch": 0.65, + "grad_norm": 2.433200542891736, + "learning_rate": 5.8093492812567754e-06, + "loss": 1.0112, + "step": 27527 + }, + { + "epoch": 0.65, + "grad_norm": 3.2530755139465843, + "learning_rate": 5.808656485593657e-06, + "loss": 0.9949, + "step": 27528 + }, + { + "epoch": 0.65, + "grad_norm": 2.180525835668541, + "learning_rate": 5.80796371433387e-06, + "loss": 1.0553, + "step": 27529 + }, + { + "epoch": 0.65, + "grad_norm": 2.0371318969734236, + "learning_rate": 5.807270967481442e-06, + "loss": 0.9233, + "step": 27530 + }, + { + "epoch": 0.65, + "grad_norm": 1.9473402621558342, + "learning_rate": 5.8065782450404155e-06, + "loss": 1.0144, + "step": 27531 + }, + { + "epoch": 0.65, + "grad_norm": 2.189847373751011, + "learning_rate": 5.805885547014812e-06, + "loss": 0.9148, + "step": 27532 + }, + { + "epoch": 0.65, + "grad_norm": 2.0472902310642644, + "learning_rate": 5.805192873408676e-06, + "loss": 0.9525, + "step": 27533 + }, + { + "epoch": 0.65, + "grad_norm": 2.4114423377795275, + "learning_rate": 5.804500224226034e-06, + "loss": 0.9084, + "step": 27534 + }, + { + "epoch": 0.65, + "grad_norm": 1.8636662391130827, + "learning_rate": 5.803807599470917e-06, + "loss": 0.9243, + "step": 27535 + }, + { + "epoch": 0.65, + "grad_norm": 2.036042203648112, + "learning_rate": 5.80311499914736e-06, + "loss": 1.0276, + "step": 27536 + }, + { + "epoch": 0.65, + "grad_norm": 1.041793825700579, + "learning_rate": 5.8024224232594e-06, + "loss": 0.9953, + "step": 27537 + }, + { + "epoch": 0.65, + "grad_norm": 1.8939292329733288, + "learning_rate": 5.801729871811066e-06, + "loss": 1.127, + "step": 27538 + }, + { + "epoch": 0.65, + "grad_norm": 1.957525428250214, + "learning_rate": 5.801037344806386e-06, + "loss": 1.063, + "step": 27539 + }, + { + "epoch": 0.65, + "grad_norm": 1.990540773102845, + "learning_rate": 5.800344842249396e-06, + "loss": 1.1043, + "step": 27540 + }, + { + "epoch": 0.65, + "grad_norm": 2.125354074902837, + "learning_rate": 5.79965236414413e-06, + "loss": 1.0173, + "step": 27541 + }, + { + "epoch": 0.65, + "grad_norm": 1.1241585235430416, + "learning_rate": 5.798959910494622e-06, + "loss": 0.9074, + "step": 27542 + }, + { + "epoch": 0.65, + "grad_norm": 2.1670480324914427, + "learning_rate": 5.798267481304891e-06, + "loss": 0.9593, + "step": 27543 + }, + { + "epoch": 0.65, + "grad_norm": 1.9171606333314435, + "learning_rate": 5.797575076578978e-06, + "loss": 0.893, + "step": 27544 + }, + { + "epoch": 0.65, + "grad_norm": 2.362504538372925, + "learning_rate": 5.79688269632092e-06, + "loss": 0.9492, + "step": 27545 + }, + { + "epoch": 0.65, + "grad_norm": 2.711382663381566, + "learning_rate": 5.79619034053474e-06, + "loss": 0.9608, + "step": 27546 + }, + { + "epoch": 0.65, + "grad_norm": 1.9155488262192248, + "learning_rate": 5.795498009224465e-06, + "loss": 0.9274, + "step": 27547 + }, + { + "epoch": 0.65, + "grad_norm": 2.569961373135364, + "learning_rate": 5.794805702394138e-06, + "loss": 0.9546, + "step": 27548 + }, + { + "epoch": 0.65, + "grad_norm": 1.0416483593365122, + "learning_rate": 5.794113420047779e-06, + "loss": 0.9043, + "step": 27549 + }, + { + "epoch": 0.65, + "grad_norm": 1.9497549366146931, + "learning_rate": 5.793421162189427e-06, + "loss": 1.025, + "step": 27550 + }, + { + "epoch": 0.65, + "grad_norm": 2.3725563975895345, + "learning_rate": 5.792728928823105e-06, + "loss": 1.018, + "step": 27551 + }, + { + "epoch": 0.65, + "grad_norm": 1.061206921122872, + "learning_rate": 5.792036719952851e-06, + "loss": 0.9899, + "step": 27552 + }, + { + "epoch": 0.65, + "grad_norm": 1.7629791287526022, + "learning_rate": 5.791344535582687e-06, + "loss": 0.8298, + "step": 27553 + }, + { + "epoch": 0.65, + "grad_norm": 2.171151563535092, + "learning_rate": 5.790652375716653e-06, + "loss": 1.0596, + "step": 27554 + }, + { + "epoch": 0.65, + "grad_norm": 2.147971562232276, + "learning_rate": 5.789960240358772e-06, + "loss": 1.0266, + "step": 27555 + }, + { + "epoch": 0.65, + "grad_norm": 2.4279591777754472, + "learning_rate": 5.78926812951307e-06, + "loss": 0.9334, + "step": 27556 + }, + { + "epoch": 0.65, + "grad_norm": 2.5023467964337827, + "learning_rate": 5.788576043183585e-06, + "loss": 0.9283, + "step": 27557 + }, + { + "epoch": 0.65, + "grad_norm": 1.077876517718029, + "learning_rate": 5.787883981374347e-06, + "loss": 0.9847, + "step": 27558 + }, + { + "epoch": 0.65, + "grad_norm": 2.0588846443931, + "learning_rate": 5.787191944089381e-06, + "loss": 1.0377, + "step": 27559 + }, + { + "epoch": 0.65, + "grad_norm": 2.0179057181906095, + "learning_rate": 5.786499931332714e-06, + "loss": 1.047, + "step": 27560 + }, + { + "epoch": 0.65, + "grad_norm": 2.0344356236447196, + "learning_rate": 5.785807943108377e-06, + "loss": 0.951, + "step": 27561 + }, + { + "epoch": 0.65, + "grad_norm": 2.3726032444830256, + "learning_rate": 5.785115979420406e-06, + "loss": 1.055, + "step": 27562 + }, + { + "epoch": 0.65, + "grad_norm": 2.0507125015883507, + "learning_rate": 5.784424040272823e-06, + "loss": 1.0232, + "step": 27563 + }, + { + "epoch": 0.65, + "grad_norm": 2.273625244216179, + "learning_rate": 5.7837321256696555e-06, + "loss": 0.9715, + "step": 27564 + }, + { + "epoch": 0.65, + "grad_norm": 1.8147303070022585, + "learning_rate": 5.783040235614933e-06, + "loss": 0.9882, + "step": 27565 + }, + { + "epoch": 0.65, + "grad_norm": 1.7905775041838687, + "learning_rate": 5.782348370112689e-06, + "loss": 0.9401, + "step": 27566 + }, + { + "epoch": 0.65, + "grad_norm": 1.0511224165823425, + "learning_rate": 5.7816565291669504e-06, + "loss": 0.9433, + "step": 27567 + }, + { + "epoch": 0.65, + "grad_norm": 1.0893543509285672, + "learning_rate": 5.780964712781736e-06, + "loss": 0.9088, + "step": 27568 + }, + { + "epoch": 0.65, + "grad_norm": 2.375876955391552, + "learning_rate": 5.780272920961087e-06, + "loss": 0.9471, + "step": 27569 + }, + { + "epoch": 0.65, + "grad_norm": 1.1199401080059692, + "learning_rate": 5.779581153709019e-06, + "loss": 0.8983, + "step": 27570 + }, + { + "epoch": 0.65, + "grad_norm": 1.9968629693258793, + "learning_rate": 5.778889411029571e-06, + "loss": 0.9414, + "step": 27571 + }, + { + "epoch": 0.65, + "grad_norm": 2.2783856333489676, + "learning_rate": 5.778197692926761e-06, + "loss": 0.9203, + "step": 27572 + }, + { + "epoch": 0.65, + "grad_norm": 1.8783312113261155, + "learning_rate": 5.777505999404625e-06, + "loss": 0.9851, + "step": 27573 + }, + { + "epoch": 0.65, + "grad_norm": 2.1739989961789523, + "learning_rate": 5.776814330467182e-06, + "loss": 1.0233, + "step": 27574 + }, + { + "epoch": 0.65, + "grad_norm": 1.9534657115591432, + "learning_rate": 5.776122686118466e-06, + "loss": 1.0025, + "step": 27575 + }, + { + "epoch": 0.65, + "grad_norm": 2.06071947672057, + "learning_rate": 5.775431066362503e-06, + "loss": 0.8849, + "step": 27576 + }, + { + "epoch": 0.65, + "grad_norm": 1.0660105317545556, + "learning_rate": 5.774739471203311e-06, + "loss": 0.8995, + "step": 27577 + }, + { + "epoch": 0.65, + "grad_norm": 1.9368602729181044, + "learning_rate": 5.774047900644926e-06, + "loss": 0.8549, + "step": 27578 + }, + { + "epoch": 0.65, + "grad_norm": 2.600402287023277, + "learning_rate": 5.773356354691374e-06, + "loss": 0.9988, + "step": 27579 + }, + { + "epoch": 0.65, + "grad_norm": 2.1301816923436947, + "learning_rate": 5.772664833346681e-06, + "loss": 0.9631, + "step": 27580 + }, + { + "epoch": 0.65, + "grad_norm": 1.8942034280442541, + "learning_rate": 5.771973336614867e-06, + "loss": 1.0132, + "step": 27581 + }, + { + "epoch": 0.65, + "grad_norm": 1.985225727038087, + "learning_rate": 5.7712818644999624e-06, + "loss": 1.0216, + "step": 27582 + }, + { + "epoch": 0.65, + "grad_norm": 1.9115228294582263, + "learning_rate": 5.7705904170059976e-06, + "loss": 1.0737, + "step": 27583 + }, + { + "epoch": 0.65, + "grad_norm": 2.4793623148036756, + "learning_rate": 5.769898994136996e-06, + "loss": 1.126, + "step": 27584 + }, + { + "epoch": 0.65, + "grad_norm": 2.2355384310045134, + "learning_rate": 5.769207595896975e-06, + "loss": 1.1355, + "step": 27585 + }, + { + "epoch": 0.65, + "grad_norm": 2.295680403244025, + "learning_rate": 5.768516222289968e-06, + "loss": 1.0216, + "step": 27586 + }, + { + "epoch": 0.65, + "grad_norm": 1.971627875828659, + "learning_rate": 5.767824873320005e-06, + "loss": 1.0426, + "step": 27587 + }, + { + "epoch": 0.65, + "grad_norm": 1.9176527592015802, + "learning_rate": 5.767133548991104e-06, + "loss": 0.898, + "step": 27588 + }, + { + "epoch": 0.65, + "grad_norm": 2.0604198922899286, + "learning_rate": 5.766442249307287e-06, + "loss": 1.0776, + "step": 27589 + }, + { + "epoch": 0.65, + "grad_norm": 1.889988367791961, + "learning_rate": 5.765750974272588e-06, + "loss": 0.9394, + "step": 27590 + }, + { + "epoch": 0.65, + "grad_norm": 1.124272601937533, + "learning_rate": 5.765059723891024e-06, + "loss": 0.9689, + "step": 27591 + }, + { + "epoch": 0.65, + "grad_norm": 1.0640649584305577, + "learning_rate": 5.764368498166626e-06, + "loss": 0.9009, + "step": 27592 + }, + { + "epoch": 0.65, + "grad_norm": 1.9629349684569743, + "learning_rate": 5.763677297103411e-06, + "loss": 1.0069, + "step": 27593 + }, + { + "epoch": 0.65, + "grad_norm": 2.191724323306038, + "learning_rate": 5.7629861207054135e-06, + "loss": 1.0407, + "step": 27594 + }, + { + "epoch": 0.65, + "grad_norm": 2.2085780516072613, + "learning_rate": 5.7622949689766496e-06, + "loss": 0.8623, + "step": 27595 + }, + { + "epoch": 0.65, + "grad_norm": 1.127018824861295, + "learning_rate": 5.761603841921147e-06, + "loss": 0.9426, + "step": 27596 + }, + { + "epoch": 0.65, + "grad_norm": 1.9964574477373223, + "learning_rate": 5.76091273954293e-06, + "loss": 0.986, + "step": 27597 + }, + { + "epoch": 0.65, + "grad_norm": 1.0018644315705587, + "learning_rate": 5.760221661846017e-06, + "loss": 0.9723, + "step": 27598 + }, + { + "epoch": 0.65, + "grad_norm": 2.0003471210649213, + "learning_rate": 5.7595306088344364e-06, + "loss": 1.1225, + "step": 27599 + }, + { + "epoch": 0.65, + "grad_norm": 2.1920155058219297, + "learning_rate": 5.758839580512214e-06, + "loss": 0.9403, + "step": 27600 + }, + { + "epoch": 0.65, + "grad_norm": 1.9778274513053729, + "learning_rate": 5.7581485768833715e-06, + "loss": 0.8315, + "step": 27601 + }, + { + "epoch": 0.65, + "grad_norm": 1.868839731295309, + "learning_rate": 5.757457597951925e-06, + "loss": 0.9902, + "step": 27602 + }, + { + "epoch": 0.65, + "grad_norm": 1.9065352307343204, + "learning_rate": 5.756766643721905e-06, + "loss": 1.0396, + "step": 27603 + }, + { + "epoch": 0.65, + "grad_norm": 2.06024793630432, + "learning_rate": 5.756075714197338e-06, + "loss": 0.9609, + "step": 27604 + }, + { + "epoch": 0.65, + "grad_norm": 2.111131890994014, + "learning_rate": 5.755384809382239e-06, + "loss": 1.016, + "step": 27605 + }, + { + "epoch": 0.65, + "grad_norm": 1.9475027079201475, + "learning_rate": 5.754693929280633e-06, + "loss": 1.0188, + "step": 27606 + }, + { + "epoch": 0.65, + "grad_norm": 2.1997673470952748, + "learning_rate": 5.754003073896539e-06, + "loss": 0.9606, + "step": 27607 + }, + { + "epoch": 0.65, + "grad_norm": 2.0825880423989904, + "learning_rate": 5.75331224323399e-06, + "loss": 1.0844, + "step": 27608 + }, + { + "epoch": 0.65, + "grad_norm": 2.2756430962684995, + "learning_rate": 5.7526214372970014e-06, + "loss": 0.9806, + "step": 27609 + }, + { + "epoch": 0.65, + "grad_norm": 2.376073964831407, + "learning_rate": 5.751930656089591e-06, + "loss": 0.8221, + "step": 27610 + }, + { + "epoch": 0.65, + "grad_norm": 2.1848383906028523, + "learning_rate": 5.751239899615789e-06, + "loss": 0.9118, + "step": 27611 + }, + { + "epoch": 0.65, + "grad_norm": 2.325629861400365, + "learning_rate": 5.75054916787961e-06, + "loss": 0.9311, + "step": 27612 + }, + { + "epoch": 0.65, + "grad_norm": 1.9536727336321893, + "learning_rate": 5.7498584608850786e-06, + "loss": 1.0295, + "step": 27613 + }, + { + "epoch": 0.65, + "grad_norm": 1.981685993012503, + "learning_rate": 5.749167778636221e-06, + "loss": 1.0527, + "step": 27614 + }, + { + "epoch": 0.65, + "grad_norm": 1.9741660677801538, + "learning_rate": 5.748477121137055e-06, + "loss": 1.0303, + "step": 27615 + }, + { + "epoch": 0.65, + "grad_norm": 2.094963028927839, + "learning_rate": 5.7477864883915956e-06, + "loss": 1.0723, + "step": 27616 + }, + { + "epoch": 0.65, + "grad_norm": 1.920316214304223, + "learning_rate": 5.74709588040387e-06, + "loss": 0.8161, + "step": 27617 + }, + { + "epoch": 0.65, + "grad_norm": 2.2690454255836343, + "learning_rate": 5.746405297177908e-06, + "loss": 1.0397, + "step": 27618 + }, + { + "epoch": 0.65, + "grad_norm": 2.0555456369769707, + "learning_rate": 5.745714738717711e-06, + "loss": 0.9591, + "step": 27619 + }, + { + "epoch": 0.65, + "grad_norm": 2.200304931732376, + "learning_rate": 5.745024205027312e-06, + "loss": 1.0331, + "step": 27620 + }, + { + "epoch": 0.65, + "grad_norm": 4.979532815702395, + "learning_rate": 5.744333696110729e-06, + "loss": 1.1262, + "step": 27621 + }, + { + "epoch": 0.65, + "grad_norm": 2.2945023131851348, + "learning_rate": 5.743643211971989e-06, + "loss": 1.1049, + "step": 27622 + }, + { + "epoch": 0.65, + "grad_norm": 1.8186617511751149, + "learning_rate": 5.742952752615099e-06, + "loss": 0.8761, + "step": 27623 + }, + { + "epoch": 0.65, + "grad_norm": 2.176273660803513, + "learning_rate": 5.742262318044084e-06, + "loss": 1.0859, + "step": 27624 + }, + { + "epoch": 0.65, + "grad_norm": 1.7324048466689566, + "learning_rate": 5.74157190826297e-06, + "loss": 0.9514, + "step": 27625 + }, + { + "epoch": 0.65, + "grad_norm": 1.8993589748739457, + "learning_rate": 5.740881523275771e-06, + "loss": 1.0189, + "step": 27626 + }, + { + "epoch": 0.65, + "grad_norm": 1.9124924112423387, + "learning_rate": 5.74019116308651e-06, + "loss": 0.9716, + "step": 27627 + }, + { + "epoch": 0.65, + "grad_norm": 1.8846943216103742, + "learning_rate": 5.7395008276992015e-06, + "loss": 1.0353, + "step": 27628 + }, + { + "epoch": 0.65, + "grad_norm": 1.958375129290414, + "learning_rate": 5.738810517117872e-06, + "loss": 1.0121, + "step": 27629 + }, + { + "epoch": 0.65, + "grad_norm": 2.323268989341845, + "learning_rate": 5.7381202313465334e-06, + "loss": 0.9671, + "step": 27630 + }, + { + "epoch": 0.65, + "grad_norm": 2.012442365553281, + "learning_rate": 5.73742997038921e-06, + "loss": 0.9486, + "step": 27631 + }, + { + "epoch": 0.65, + "grad_norm": 2.079048624390534, + "learning_rate": 5.736739734249916e-06, + "loss": 1.0002, + "step": 27632 + }, + { + "epoch": 0.65, + "grad_norm": 1.8276732077027522, + "learning_rate": 5.736049522932679e-06, + "loss": 1.1209, + "step": 27633 + }, + { + "epoch": 0.65, + "grad_norm": 2.0475708272667386, + "learning_rate": 5.735359336441504e-06, + "loss": 0.8174, + "step": 27634 + }, + { + "epoch": 0.65, + "grad_norm": 2.0292573375173006, + "learning_rate": 5.734669174780423e-06, + "loss": 0.9555, + "step": 27635 + }, + { + "epoch": 0.65, + "grad_norm": 1.9606766418596482, + "learning_rate": 5.733979037953449e-06, + "loss": 1.045, + "step": 27636 + }, + { + "epoch": 0.65, + "grad_norm": 2.205632874363672, + "learning_rate": 5.733288925964595e-06, + "loss": 0.9162, + "step": 27637 + }, + { + "epoch": 0.65, + "grad_norm": 1.876713268274644, + "learning_rate": 5.732598838817883e-06, + "loss": 0.8967, + "step": 27638 + }, + { + "epoch": 0.65, + "grad_norm": 1.9995663870626292, + "learning_rate": 5.731908776517336e-06, + "loss": 1.0003, + "step": 27639 + }, + { + "epoch": 0.65, + "grad_norm": 2.1354652283203985, + "learning_rate": 5.731218739066968e-06, + "loss": 0.9678, + "step": 27640 + }, + { + "epoch": 0.65, + "grad_norm": 1.9982916762642768, + "learning_rate": 5.730528726470792e-06, + "loss": 0.9957, + "step": 27641 + }, + { + "epoch": 0.65, + "grad_norm": 2.1009047465600887, + "learning_rate": 5.7298387387328295e-06, + "loss": 0.8685, + "step": 27642 + }, + { + "epoch": 0.65, + "grad_norm": 2.0482277900909662, + "learning_rate": 5.729148775857101e-06, + "loss": 1.058, + "step": 27643 + }, + { + "epoch": 0.65, + "grad_norm": 2.346497742266498, + "learning_rate": 5.72845883784762e-06, + "loss": 0.8757, + "step": 27644 + }, + { + "epoch": 0.65, + "grad_norm": 1.788499467072871, + "learning_rate": 5.7277689247084016e-06, + "loss": 0.8797, + "step": 27645 + }, + { + "epoch": 0.65, + "grad_norm": 1.8938958777675208, + "learning_rate": 5.727079036443464e-06, + "loss": 0.9343, + "step": 27646 + }, + { + "epoch": 0.65, + "grad_norm": 1.8582482950084505, + "learning_rate": 5.726389173056829e-06, + "loss": 0.9568, + "step": 27647 + }, + { + "epoch": 0.65, + "grad_norm": 1.7876321581006778, + "learning_rate": 5.725699334552509e-06, + "loss": 1.0444, + "step": 27648 + }, + { + "epoch": 0.65, + "grad_norm": 1.8492536239182151, + "learning_rate": 5.725009520934517e-06, + "loss": 1.1086, + "step": 27649 + }, + { + "epoch": 0.65, + "grad_norm": 2.270539643669551, + "learning_rate": 5.724319732206878e-06, + "loss": 1.0073, + "step": 27650 + }, + { + "epoch": 0.65, + "grad_norm": 2.331706502681662, + "learning_rate": 5.723629968373598e-06, + "loss": 0.9775, + "step": 27651 + }, + { + "epoch": 0.65, + "grad_norm": 2.1066422268733906, + "learning_rate": 5.722940229438703e-06, + "loss": 1.1261, + "step": 27652 + }, + { + "epoch": 0.65, + "grad_norm": 1.1231480397394424, + "learning_rate": 5.722250515406199e-06, + "loss": 0.9403, + "step": 27653 + }, + { + "epoch": 0.65, + "grad_norm": 2.300917204279127, + "learning_rate": 5.721560826280112e-06, + "loss": 1.0952, + "step": 27654 + }, + { + "epoch": 0.65, + "grad_norm": 1.0585200976512998, + "learning_rate": 5.720871162064447e-06, + "loss": 0.9264, + "step": 27655 + }, + { + "epoch": 0.65, + "grad_norm": 1.994304527534667, + "learning_rate": 5.7201815227632305e-06, + "loss": 0.9516, + "step": 27656 + }, + { + "epoch": 0.65, + "grad_norm": 1.9147219201981964, + "learning_rate": 5.719491908380471e-06, + "loss": 0.9689, + "step": 27657 + }, + { + "epoch": 0.65, + "grad_norm": 2.1518819068113824, + "learning_rate": 5.718802318920181e-06, + "loss": 1.0252, + "step": 27658 + }, + { + "epoch": 0.65, + "grad_norm": 2.0014279925882774, + "learning_rate": 5.71811275438638e-06, + "loss": 0.9961, + "step": 27659 + }, + { + "epoch": 0.65, + "grad_norm": 2.101744379802339, + "learning_rate": 5.717423214783085e-06, + "loss": 0.9084, + "step": 27660 + }, + { + "epoch": 0.65, + "grad_norm": 1.7986478805842858, + "learning_rate": 5.7167337001143085e-06, + "loss": 0.9411, + "step": 27661 + }, + { + "epoch": 0.65, + "grad_norm": 1.7672331159789776, + "learning_rate": 5.716044210384061e-06, + "loss": 1.1788, + "step": 27662 + }, + { + "epoch": 0.65, + "grad_norm": 1.9092465532891316, + "learning_rate": 5.7153547455963595e-06, + "loss": 1.0225, + "step": 27663 + }, + { + "epoch": 0.65, + "grad_norm": 1.8497150461701866, + "learning_rate": 5.714665305755223e-06, + "loss": 1.0936, + "step": 27664 + }, + { + "epoch": 0.65, + "grad_norm": 1.7992859273629813, + "learning_rate": 5.7139758908646625e-06, + "loss": 1.092, + "step": 27665 + }, + { + "epoch": 0.65, + "grad_norm": 2.0405767576985347, + "learning_rate": 5.7132865009286875e-06, + "loss": 1.0296, + "step": 27666 + }, + { + "epoch": 0.65, + "grad_norm": 1.8761320975532878, + "learning_rate": 5.712597135951314e-06, + "loss": 0.9207, + "step": 27667 + }, + { + "epoch": 0.65, + "grad_norm": 2.0218036028679562, + "learning_rate": 5.711907795936564e-06, + "loss": 0.9354, + "step": 27668 + }, + { + "epoch": 0.65, + "grad_norm": 2.0826074246264747, + "learning_rate": 5.711218480888443e-06, + "loss": 1.1137, + "step": 27669 + }, + { + "epoch": 0.65, + "grad_norm": 2.0940047925740655, + "learning_rate": 5.710529190810962e-06, + "loss": 1.045, + "step": 27670 + }, + { + "epoch": 0.65, + "grad_norm": 1.839681834577283, + "learning_rate": 5.709839925708141e-06, + "loss": 0.9341, + "step": 27671 + }, + { + "epoch": 0.65, + "grad_norm": 1.9154092559398657, + "learning_rate": 5.709150685583987e-06, + "loss": 0.9285, + "step": 27672 + }, + { + "epoch": 0.65, + "grad_norm": 1.9306116730750669, + "learning_rate": 5.708461470442521e-06, + "loss": 1.1412, + "step": 27673 + }, + { + "epoch": 0.65, + "grad_norm": 2.6929347026433352, + "learning_rate": 5.7077722802877465e-06, + "loss": 0.867, + "step": 27674 + }, + { + "epoch": 0.65, + "grad_norm": 1.991728163338416, + "learning_rate": 5.7070831151236836e-06, + "loss": 1.0593, + "step": 27675 + }, + { + "epoch": 0.65, + "grad_norm": 2.202074462214831, + "learning_rate": 5.706393974954338e-06, + "loss": 0.9774, + "step": 27676 + }, + { + "epoch": 0.65, + "grad_norm": 1.997736622544975, + "learning_rate": 5.705704859783732e-06, + "loss": 1.1103, + "step": 27677 + }, + { + "epoch": 0.65, + "grad_norm": 1.8824683852841073, + "learning_rate": 5.70501576961587e-06, + "loss": 1.07, + "step": 27678 + }, + { + "epoch": 0.65, + "grad_norm": 1.9501002258797917, + "learning_rate": 5.7043267044547636e-06, + "loss": 0.9742, + "step": 27679 + }, + { + "epoch": 0.65, + "grad_norm": 1.8499563265244436, + "learning_rate": 5.703637664304425e-06, + "loss": 0.8394, + "step": 27680 + }, + { + "epoch": 0.65, + "grad_norm": 2.2868689084158427, + "learning_rate": 5.702948649168874e-06, + "loss": 0.8946, + "step": 27681 + }, + { + "epoch": 0.65, + "grad_norm": 1.9289950504972602, + "learning_rate": 5.702259659052115e-06, + "loss": 0.9182, + "step": 27682 + }, + { + "epoch": 0.65, + "grad_norm": 2.3498026753660257, + "learning_rate": 5.701570693958159e-06, + "loss": 1.0809, + "step": 27683 + }, + { + "epoch": 0.65, + "grad_norm": 2.429577117835841, + "learning_rate": 5.700881753891017e-06, + "loss": 0.9879, + "step": 27684 + }, + { + "epoch": 0.65, + "grad_norm": 2.0070391379805916, + "learning_rate": 5.700192838854707e-06, + "loss": 1.0276, + "step": 27685 + }, + { + "epoch": 0.65, + "grad_norm": 1.7345293620457696, + "learning_rate": 5.6995039488532355e-06, + "loss": 0.9884, + "step": 27686 + }, + { + "epoch": 0.65, + "grad_norm": 2.833070151245756, + "learning_rate": 5.6988150838906096e-06, + "loss": 1.0593, + "step": 27687 + }, + { + "epoch": 0.65, + "grad_norm": 1.9439133700311761, + "learning_rate": 5.698126243970845e-06, + "loss": 1.0239, + "step": 27688 + }, + { + "epoch": 0.65, + "grad_norm": 1.9028655088517687, + "learning_rate": 5.697437429097954e-06, + "loss": 1.0414, + "step": 27689 + }, + { + "epoch": 0.65, + "grad_norm": 2.0818675446072628, + "learning_rate": 5.6967486392759455e-06, + "loss": 1.0464, + "step": 27690 + }, + { + "epoch": 0.65, + "grad_norm": 2.018636519298671, + "learning_rate": 5.696059874508823e-06, + "loss": 1.0165, + "step": 27691 + }, + { + "epoch": 0.65, + "grad_norm": 1.0965458046417993, + "learning_rate": 5.695371134800609e-06, + "loss": 0.9186, + "step": 27692 + }, + { + "epoch": 0.65, + "grad_norm": 1.8614689975924406, + "learning_rate": 5.694682420155302e-06, + "loss": 0.9602, + "step": 27693 + }, + { + "epoch": 0.65, + "grad_norm": 1.0685996347931672, + "learning_rate": 5.693993730576921e-06, + "loss": 0.9693, + "step": 27694 + }, + { + "epoch": 0.65, + "grad_norm": 1.845634026470211, + "learning_rate": 5.6933050660694674e-06, + "loss": 1.0244, + "step": 27695 + }, + { + "epoch": 0.65, + "grad_norm": 2.158805675023342, + "learning_rate": 5.69261642663696e-06, + "loss": 1.063, + "step": 27696 + }, + { + "epoch": 0.65, + "grad_norm": 2.1719715501423047, + "learning_rate": 5.6919278122834e-06, + "loss": 1.0427, + "step": 27697 + }, + { + "epoch": 0.65, + "grad_norm": 2.0245518105595353, + "learning_rate": 5.691239223012802e-06, + "loss": 0.9194, + "step": 27698 + }, + { + "epoch": 0.65, + "grad_norm": 2.1342102341103595, + "learning_rate": 5.690550658829175e-06, + "loss": 0.9141, + "step": 27699 + }, + { + "epoch": 0.65, + "grad_norm": 1.8741227941628091, + "learning_rate": 5.689862119736522e-06, + "loss": 1.03, + "step": 27700 + }, + { + "epoch": 0.65, + "grad_norm": 2.6444786116194474, + "learning_rate": 5.689173605738856e-06, + "loss": 0.9482, + "step": 27701 + }, + { + "epoch": 0.65, + "grad_norm": 1.7037792745417901, + "learning_rate": 5.688485116840187e-06, + "loss": 1.0708, + "step": 27702 + }, + { + "epoch": 0.65, + "grad_norm": 2.2384035943956837, + "learning_rate": 5.68779665304453e-06, + "loss": 1.0205, + "step": 27703 + }, + { + "epoch": 0.65, + "grad_norm": 3.027750430478573, + "learning_rate": 5.687108214355877e-06, + "loss": 0.9703, + "step": 27704 + }, + { + "epoch": 0.65, + "grad_norm": 2.880600858096878, + "learning_rate": 5.686419800778246e-06, + "loss": 0.9306, + "step": 27705 + }, + { + "epoch": 0.65, + "grad_norm": 2.3917293051552764, + "learning_rate": 5.685731412315649e-06, + "loss": 1.061, + "step": 27706 + }, + { + "epoch": 0.65, + "grad_norm": 1.9277214969114995, + "learning_rate": 5.685043048972091e-06, + "loss": 0.9561, + "step": 27707 + }, + { + "epoch": 0.65, + "grad_norm": 1.8988884975886087, + "learning_rate": 5.684354710751573e-06, + "loss": 0.9203, + "step": 27708 + }, + { + "epoch": 0.65, + "grad_norm": 1.7838329073179784, + "learning_rate": 5.683666397658108e-06, + "loss": 0.9642, + "step": 27709 + }, + { + "epoch": 0.65, + "grad_norm": 1.9949415719421406, + "learning_rate": 5.6829781096957086e-06, + "loss": 0.9731, + "step": 27710 + }, + { + "epoch": 0.65, + "grad_norm": 1.8840983821211992, + "learning_rate": 5.682289846868372e-06, + "loss": 1.0792, + "step": 27711 + }, + { + "epoch": 0.65, + "grad_norm": 2.0671258209412957, + "learning_rate": 5.681601609180117e-06, + "loss": 0.89, + "step": 27712 + }, + { + "epoch": 0.65, + "grad_norm": 2.019878749768816, + "learning_rate": 5.680913396634943e-06, + "loss": 0.9389, + "step": 27713 + }, + { + "epoch": 0.65, + "grad_norm": 2.011768844591872, + "learning_rate": 5.680225209236855e-06, + "loss": 1.0475, + "step": 27714 + }, + { + "epoch": 0.65, + "grad_norm": 1.8009763070784683, + "learning_rate": 5.6795370469898625e-06, + "loss": 1.1575, + "step": 27715 + }, + { + "epoch": 0.65, + "grad_norm": 1.7659937591650372, + "learning_rate": 5.678848909897979e-06, + "loss": 0.9362, + "step": 27716 + }, + { + "epoch": 0.65, + "grad_norm": 1.9205806670059464, + "learning_rate": 5.678160797965204e-06, + "loss": 0.9248, + "step": 27717 + }, + { + "epoch": 0.65, + "grad_norm": 2.136680768918089, + "learning_rate": 5.677472711195542e-06, + "loss": 0.9837, + "step": 27718 + }, + { + "epoch": 0.65, + "grad_norm": 2.6792039238085628, + "learning_rate": 5.676784649593001e-06, + "loss": 0.9852, + "step": 27719 + }, + { + "epoch": 0.65, + "grad_norm": 2.1260044684613426, + "learning_rate": 5.676096613161597e-06, + "loss": 1.094, + "step": 27720 + }, + { + "epoch": 0.65, + "grad_norm": 1.9620098621814746, + "learning_rate": 5.67540860190532e-06, + "loss": 0.8991, + "step": 27721 + }, + { + "epoch": 0.65, + "grad_norm": 2.269549267766279, + "learning_rate": 5.674720615828182e-06, + "loss": 1.0049, + "step": 27722 + }, + { + "epoch": 0.65, + "grad_norm": 1.117709097171965, + "learning_rate": 5.674032654934192e-06, + "loss": 0.9546, + "step": 27723 + }, + { + "epoch": 0.65, + "grad_norm": 2.5255640996281357, + "learning_rate": 5.673344719227356e-06, + "loss": 1.0027, + "step": 27724 + }, + { + "epoch": 0.65, + "grad_norm": 1.8004422068397015, + "learning_rate": 5.672656808711677e-06, + "loss": 0.989, + "step": 27725 + }, + { + "epoch": 0.65, + "grad_norm": 2.1017908922558783, + "learning_rate": 5.671968923391156e-06, + "loss": 1.0622, + "step": 27726 + }, + { + "epoch": 0.65, + "grad_norm": 2.063718472430717, + "learning_rate": 5.671281063269807e-06, + "loss": 0.9846, + "step": 27727 + }, + { + "epoch": 0.65, + "grad_norm": 1.9626502512603725, + "learning_rate": 5.670593228351625e-06, + "loss": 0.9337, + "step": 27728 + }, + { + "epoch": 0.65, + "grad_norm": 1.8838867028834418, + "learning_rate": 5.669905418640624e-06, + "loss": 0.9569, + "step": 27729 + }, + { + "epoch": 0.65, + "grad_norm": 1.923657109071002, + "learning_rate": 5.669217634140802e-06, + "loss": 0.8666, + "step": 27730 + }, + { + "epoch": 0.65, + "grad_norm": 1.8086399003349904, + "learning_rate": 5.668529874856168e-06, + "loss": 1.0085, + "step": 27731 + }, + { + "epoch": 0.65, + "grad_norm": 1.1206952865192836, + "learning_rate": 5.667842140790722e-06, + "loss": 0.9662, + "step": 27732 + }, + { + "epoch": 0.65, + "grad_norm": 1.8157214102973904, + "learning_rate": 5.667154431948474e-06, + "loss": 0.9982, + "step": 27733 + }, + { + "epoch": 0.65, + "grad_norm": 2.1180152629752844, + "learning_rate": 5.666466748333424e-06, + "loss": 1.0955, + "step": 27734 + }, + { + "epoch": 0.65, + "grad_norm": 1.0877271134260103, + "learning_rate": 5.665779089949574e-06, + "loss": 0.9948, + "step": 27735 + }, + { + "epoch": 0.65, + "grad_norm": 1.9215901819662966, + "learning_rate": 5.665091456800929e-06, + "loss": 0.8477, + "step": 27736 + }, + { + "epoch": 0.65, + "grad_norm": 2.0479988183681135, + "learning_rate": 5.664403848891499e-06, + "loss": 1.0742, + "step": 27737 + }, + { + "epoch": 0.65, + "grad_norm": 1.9343870261241014, + "learning_rate": 5.663716266225281e-06, + "loss": 1.0127, + "step": 27738 + }, + { + "epoch": 0.65, + "grad_norm": 1.8557502298421529, + "learning_rate": 5.663028708806277e-06, + "loss": 0.9763, + "step": 27739 + }, + { + "epoch": 0.65, + "grad_norm": 1.9823052575782052, + "learning_rate": 5.662341176638493e-06, + "loss": 1.0034, + "step": 27740 + }, + { + "epoch": 0.65, + "grad_norm": 2.067833062062028, + "learning_rate": 5.661653669725935e-06, + "loss": 1.0602, + "step": 27741 + }, + { + "epoch": 0.65, + "grad_norm": 1.933056382324237, + "learning_rate": 5.660966188072603e-06, + "loss": 0.9896, + "step": 27742 + }, + { + "epoch": 0.65, + "grad_norm": 2.2199840637546697, + "learning_rate": 5.660278731682496e-06, + "loss": 0.8648, + "step": 27743 + }, + { + "epoch": 0.65, + "grad_norm": 2.3995537313551547, + "learning_rate": 5.65959130055962e-06, + "loss": 1.0424, + "step": 27744 + }, + { + "epoch": 0.65, + "grad_norm": 2.0918490913562953, + "learning_rate": 5.658903894707981e-06, + "loss": 0.9623, + "step": 27745 + }, + { + "epoch": 0.65, + "grad_norm": 2.1109941974532305, + "learning_rate": 5.65821651413158e-06, + "loss": 0.8625, + "step": 27746 + }, + { + "epoch": 0.65, + "grad_norm": 2.071824451646394, + "learning_rate": 5.65752915883441e-06, + "loss": 0.9799, + "step": 27747 + }, + { + "epoch": 0.65, + "grad_norm": 1.9621520140216724, + "learning_rate": 5.656841828820486e-06, + "loss": 0.9182, + "step": 27748 + }, + { + "epoch": 0.65, + "grad_norm": 1.9983101391212226, + "learning_rate": 5.6561545240937995e-06, + "loss": 0.9566, + "step": 27749 + }, + { + "epoch": 0.65, + "grad_norm": 1.0866957989013268, + "learning_rate": 5.655467244658361e-06, + "loss": 0.9402, + "step": 27750 + }, + { + "epoch": 0.65, + "grad_norm": 2.1676581434106392, + "learning_rate": 5.654779990518163e-06, + "loss": 1.1126, + "step": 27751 + }, + { + "epoch": 0.65, + "grad_norm": 2.0619850396343833, + "learning_rate": 5.6540927616772165e-06, + "loss": 0.9539, + "step": 27752 + }, + { + "epoch": 0.65, + "grad_norm": 2.167095328210555, + "learning_rate": 5.653405558139513e-06, + "loss": 1.088, + "step": 27753 + }, + { + "epoch": 0.65, + "grad_norm": 1.1802669656823244, + "learning_rate": 5.652718379909062e-06, + "loss": 1.0165, + "step": 27754 + }, + { + "epoch": 0.65, + "grad_norm": 1.9030738524691362, + "learning_rate": 5.652031226989858e-06, + "loss": 0.9157, + "step": 27755 + }, + { + "epoch": 0.65, + "grad_norm": 2.030144523956342, + "learning_rate": 5.651344099385908e-06, + "loss": 1.0199, + "step": 27756 + }, + { + "epoch": 0.65, + "grad_norm": 2.035310266921603, + "learning_rate": 5.650656997101205e-06, + "loss": 0.9814, + "step": 27757 + }, + { + "epoch": 0.65, + "grad_norm": 1.8716090418958737, + "learning_rate": 5.64996992013976e-06, + "loss": 1.01, + "step": 27758 + }, + { + "epoch": 0.65, + "grad_norm": 1.9491267410715227, + "learning_rate": 5.6492828685055656e-06, + "loss": 1.0132, + "step": 27759 + }, + { + "epoch": 0.65, + "grad_norm": 1.9158405691587395, + "learning_rate": 5.64859584220262e-06, + "loss": 0.9961, + "step": 27760 + }, + { + "epoch": 0.65, + "grad_norm": 1.0678008409196575, + "learning_rate": 5.647908841234927e-06, + "loss": 0.9131, + "step": 27761 + }, + { + "epoch": 0.65, + "grad_norm": 2.2804573388245433, + "learning_rate": 5.64722186560649e-06, + "loss": 0.8686, + "step": 27762 + }, + { + "epoch": 0.65, + "grad_norm": 2.070343399829413, + "learning_rate": 5.6465349153213065e-06, + "loss": 0.9556, + "step": 27763 + }, + { + "epoch": 0.65, + "grad_norm": 2.0371110167272724, + "learning_rate": 5.6458479903833705e-06, + "loss": 1.2343, + "step": 27764 + }, + { + "epoch": 0.65, + "grad_norm": 1.985439506282038, + "learning_rate": 5.645161090796687e-06, + "loss": 0.929, + "step": 27765 + }, + { + "epoch": 0.65, + "grad_norm": 1.9497037378107078, + "learning_rate": 5.644474216565257e-06, + "loss": 1.1008, + "step": 27766 + }, + { + "epoch": 0.65, + "grad_norm": 1.926428065224715, + "learning_rate": 5.643787367693077e-06, + "loss": 0.9357, + "step": 27767 + }, + { + "epoch": 0.65, + "grad_norm": 2.0647144046273724, + "learning_rate": 5.643100544184142e-06, + "loss": 1.01, + "step": 27768 + }, + { + "epoch": 0.65, + "grad_norm": 1.1038331121549652, + "learning_rate": 5.642413746042457e-06, + "loss": 0.9183, + "step": 27769 + }, + { + "epoch": 0.65, + "grad_norm": 1.9927337598790362, + "learning_rate": 5.6417269732720204e-06, + "loss": 0.878, + "step": 27770 + }, + { + "epoch": 0.65, + "grad_norm": 1.9807126648135287, + "learning_rate": 5.64104022587683e-06, + "loss": 1.0522, + "step": 27771 + }, + { + "epoch": 0.65, + "grad_norm": 1.7361657670287223, + "learning_rate": 5.64035350386088e-06, + "loss": 1.0475, + "step": 27772 + }, + { + "epoch": 0.65, + "grad_norm": 1.914388209846568, + "learning_rate": 5.6396668072281745e-06, + "loss": 0.8122, + "step": 27773 + }, + { + "epoch": 0.65, + "grad_norm": 2.1390380162745917, + "learning_rate": 5.6389801359827075e-06, + "loss": 0.936, + "step": 27774 + }, + { + "epoch": 0.65, + "grad_norm": 2.0151592470134063, + "learning_rate": 5.638293490128481e-06, + "loss": 1.0754, + "step": 27775 + }, + { + "epoch": 0.65, + "grad_norm": 1.128181999439464, + "learning_rate": 5.6376068696694875e-06, + "loss": 0.9207, + "step": 27776 + }, + { + "epoch": 0.65, + "grad_norm": 1.807869699060815, + "learning_rate": 5.636920274609731e-06, + "loss": 0.9053, + "step": 27777 + }, + { + "epoch": 0.65, + "grad_norm": 1.9475788599386883, + "learning_rate": 5.636233704953205e-06, + "loss": 0.9602, + "step": 27778 + }, + { + "epoch": 0.65, + "grad_norm": 1.9423805483013246, + "learning_rate": 5.635547160703909e-06, + "loss": 0.9271, + "step": 27779 + }, + { + "epoch": 0.65, + "grad_norm": 2.045125763956841, + "learning_rate": 5.63486064186584e-06, + "loss": 1.0221, + "step": 27780 + }, + { + "epoch": 0.65, + "grad_norm": 2.265014585116197, + "learning_rate": 5.63417414844299e-06, + "loss": 1.0347, + "step": 27781 + }, + { + "epoch": 0.65, + "grad_norm": 1.9553467772470787, + "learning_rate": 5.633487680439362e-06, + "loss": 1.103, + "step": 27782 + }, + { + "epoch": 0.65, + "grad_norm": 1.9856137546683565, + "learning_rate": 5.632801237858954e-06, + "loss": 0.973, + "step": 27783 + }, + { + "epoch": 0.65, + "grad_norm": 1.9052241396355316, + "learning_rate": 5.632114820705759e-06, + "loss": 1.1449, + "step": 27784 + }, + { + "epoch": 0.65, + "grad_norm": 4.869590459147996, + "learning_rate": 5.6314284289837705e-06, + "loss": 1.0579, + "step": 27785 + }, + { + "epoch": 0.65, + "grad_norm": 2.032797177187004, + "learning_rate": 5.630742062696989e-06, + "loss": 0.9599, + "step": 27786 + }, + { + "epoch": 0.65, + "grad_norm": 1.8580711474200813, + "learning_rate": 5.630055721849414e-06, + "loss": 0.9252, + "step": 27787 + }, + { + "epoch": 0.65, + "grad_norm": 2.083147649747927, + "learning_rate": 5.62936940644504e-06, + "loss": 0.8598, + "step": 27788 + }, + { + "epoch": 0.65, + "grad_norm": 2.00081165056768, + "learning_rate": 5.6286831164878556e-06, + "loss": 1.0407, + "step": 27789 + }, + { + "epoch": 0.65, + "grad_norm": 2.1080746709102773, + "learning_rate": 5.627996851981863e-06, + "loss": 0.9489, + "step": 27790 + }, + { + "epoch": 0.65, + "grad_norm": 2.9076751295495655, + "learning_rate": 5.627310612931058e-06, + "loss": 0.9541, + "step": 27791 + }, + { + "epoch": 0.65, + "grad_norm": 2.16308565774834, + "learning_rate": 5.626624399339438e-06, + "loss": 0.9358, + "step": 27792 + }, + { + "epoch": 0.65, + "grad_norm": 1.9818857616715382, + "learning_rate": 5.62593821121099e-06, + "loss": 0.8995, + "step": 27793 + }, + { + "epoch": 0.65, + "grad_norm": 2.3096150677197604, + "learning_rate": 5.62525204854972e-06, + "loss": 1.0151, + "step": 27794 + }, + { + "epoch": 0.65, + "grad_norm": 1.9013636202795876, + "learning_rate": 5.624565911359612e-06, + "loss": 0.9033, + "step": 27795 + }, + { + "epoch": 0.65, + "grad_norm": 2.0773709280697292, + "learning_rate": 5.6238797996446714e-06, + "loss": 1.0331, + "step": 27796 + }, + { + "epoch": 0.65, + "grad_norm": 1.9330696549245248, + "learning_rate": 5.623193713408883e-06, + "loss": 0.9796, + "step": 27797 + }, + { + "epoch": 0.65, + "grad_norm": 1.936054897448225, + "learning_rate": 5.6225076526562525e-06, + "loss": 1.0702, + "step": 27798 + }, + { + "epoch": 0.65, + "grad_norm": 3.023798042939769, + "learning_rate": 5.621821617390763e-06, + "loss": 0.9159, + "step": 27799 + }, + { + "epoch": 0.65, + "grad_norm": 2.3858101041107527, + "learning_rate": 5.621135607616413e-06, + "loss": 1.0463, + "step": 27800 + }, + { + "epoch": 0.65, + "grad_norm": 1.979928878530693, + "learning_rate": 5.620449623337208e-06, + "loss": 0.9844, + "step": 27801 + }, + { + "epoch": 0.65, + "grad_norm": 2.315063583068282, + "learning_rate": 5.619763664557123e-06, + "loss": 1.0415, + "step": 27802 + }, + { + "epoch": 0.66, + "grad_norm": 1.899230856996123, + "learning_rate": 5.61907773128016e-06, + "loss": 1.0899, + "step": 27803 + }, + { + "epoch": 0.66, + "grad_norm": 1.8171437294297805, + "learning_rate": 5.6183918235103144e-06, + "loss": 1.0343, + "step": 27804 + }, + { + "epoch": 0.66, + "grad_norm": 2.072173209898131, + "learning_rate": 5.617705941251586e-06, + "loss": 1.0573, + "step": 27805 + }, + { + "epoch": 0.66, + "grad_norm": 1.7633632258351668, + "learning_rate": 5.617020084507954e-06, + "loss": 0.9526, + "step": 27806 + }, + { + "epoch": 0.66, + "grad_norm": 2.0721695409277396, + "learning_rate": 5.616334253283418e-06, + "loss": 0.9751, + "step": 27807 + }, + { + "epoch": 0.66, + "grad_norm": 1.7226177989659375, + "learning_rate": 5.615648447581976e-06, + "loss": 0.9025, + "step": 27808 + }, + { + "epoch": 0.66, + "grad_norm": 2.0997067826985334, + "learning_rate": 5.614962667407613e-06, + "loss": 1.0742, + "step": 27809 + }, + { + "epoch": 0.66, + "grad_norm": 1.0263375139393198, + "learning_rate": 5.614276912764329e-06, + "loss": 0.9825, + "step": 27810 + }, + { + "epoch": 0.66, + "grad_norm": 1.7789883989993287, + "learning_rate": 5.61359118365611e-06, + "loss": 0.8634, + "step": 27811 + }, + { + "epoch": 0.66, + "grad_norm": 2.1544745564139194, + "learning_rate": 5.612905480086954e-06, + "loss": 0.9644, + "step": 27812 + }, + { + "epoch": 0.66, + "grad_norm": 2.26678589365523, + "learning_rate": 5.61221980206085e-06, + "loss": 1.0838, + "step": 27813 + }, + { + "epoch": 0.66, + "grad_norm": 2.305290547976187, + "learning_rate": 5.611534149581794e-06, + "loss": 1.0121, + "step": 27814 + }, + { + "epoch": 0.66, + "grad_norm": 1.094139301146204, + "learning_rate": 5.610848522653776e-06, + "loss": 1.0503, + "step": 27815 + }, + { + "epoch": 0.66, + "grad_norm": 1.1367990718971985, + "learning_rate": 5.610162921280783e-06, + "loss": 0.9403, + "step": 27816 + }, + { + "epoch": 0.66, + "grad_norm": 2.1332192478503185, + "learning_rate": 5.609477345466811e-06, + "loss": 0.9906, + "step": 27817 + }, + { + "epoch": 0.66, + "grad_norm": 1.8749359055332742, + "learning_rate": 5.608791795215857e-06, + "loss": 0.9843, + "step": 27818 + }, + { + "epoch": 0.66, + "grad_norm": 2.186338256155951, + "learning_rate": 5.608106270531907e-06, + "loss": 1.0079, + "step": 27819 + }, + { + "epoch": 0.66, + "grad_norm": 2.0429991109900527, + "learning_rate": 5.6074207714189475e-06, + "loss": 1.0175, + "step": 27820 + }, + { + "epoch": 0.66, + "grad_norm": 3.7046929855790744, + "learning_rate": 5.606735297880977e-06, + "loss": 0.9458, + "step": 27821 + }, + { + "epoch": 0.66, + "grad_norm": 2.20515198780267, + "learning_rate": 5.606049849921987e-06, + "loss": 0.9041, + "step": 27822 + }, + { + "epoch": 0.66, + "grad_norm": 1.8973680729702198, + "learning_rate": 5.6053644275459665e-06, + "loss": 0.9692, + "step": 27823 + }, + { + "epoch": 0.66, + "grad_norm": 2.0239010507938637, + "learning_rate": 5.604679030756901e-06, + "loss": 1.0613, + "step": 27824 + }, + { + "epoch": 0.66, + "grad_norm": 2.103123214618717, + "learning_rate": 5.603993659558786e-06, + "loss": 1.1346, + "step": 27825 + }, + { + "epoch": 0.66, + "grad_norm": 2.0439001599877535, + "learning_rate": 5.603308313955617e-06, + "loss": 0.977, + "step": 27826 + }, + { + "epoch": 0.66, + "grad_norm": 2.031583591151863, + "learning_rate": 5.602622993951377e-06, + "loss": 1.1006, + "step": 27827 + }, + { + "epoch": 0.66, + "grad_norm": 2.1967258961913405, + "learning_rate": 5.601937699550055e-06, + "loss": 0.9398, + "step": 27828 + }, + { + "epoch": 0.66, + "grad_norm": 2.24691484318395, + "learning_rate": 5.601252430755648e-06, + "loss": 1.0022, + "step": 27829 + }, + { + "epoch": 0.66, + "grad_norm": 1.8117586623212643, + "learning_rate": 5.600567187572139e-06, + "loss": 1.0265, + "step": 27830 + }, + { + "epoch": 0.66, + "grad_norm": 2.1377281341760326, + "learning_rate": 5.599881970003522e-06, + "loss": 0.8769, + "step": 27831 + }, + { + "epoch": 0.66, + "grad_norm": 1.8147522828691902, + "learning_rate": 5.599196778053784e-06, + "loss": 0.9329, + "step": 27832 + }, + { + "epoch": 0.66, + "grad_norm": 2.0870763472818044, + "learning_rate": 5.5985116117269175e-06, + "loss": 0.9415, + "step": 27833 + }, + { + "epoch": 0.66, + "grad_norm": 2.118304959236986, + "learning_rate": 5.5978264710269085e-06, + "loss": 1.0149, + "step": 27834 + }, + { + "epoch": 0.66, + "grad_norm": 2.1247980523875887, + "learning_rate": 5.597141355957749e-06, + "loss": 1.0692, + "step": 27835 + }, + { + "epoch": 0.66, + "grad_norm": 1.9400870007863895, + "learning_rate": 5.5964562665234275e-06, + "loss": 1.0066, + "step": 27836 + }, + { + "epoch": 0.66, + "grad_norm": 2.073409257164088, + "learning_rate": 5.595771202727928e-06, + "loss": 0.9449, + "step": 27837 + }, + { + "epoch": 0.66, + "grad_norm": 2.182128555210757, + "learning_rate": 5.595086164575243e-06, + "loss": 1.0918, + "step": 27838 + }, + { + "epoch": 0.66, + "grad_norm": 2.1654807521190325, + "learning_rate": 5.5944011520693645e-06, + "loss": 1.1357, + "step": 27839 + }, + { + "epoch": 0.66, + "grad_norm": 1.9819873338261478, + "learning_rate": 5.593716165214276e-06, + "loss": 1.0509, + "step": 27840 + }, + { + "epoch": 0.66, + "grad_norm": 2.224523171407484, + "learning_rate": 5.593031204013965e-06, + "loss": 0.9292, + "step": 27841 + }, + { + "epoch": 0.66, + "grad_norm": 2.1055497936166483, + "learning_rate": 5.59234626847242e-06, + "loss": 1.0204, + "step": 27842 + }, + { + "epoch": 0.66, + "grad_norm": 2.2576737104437763, + "learning_rate": 5.591661358593636e-06, + "loss": 1.0187, + "step": 27843 + }, + { + "epoch": 0.66, + "grad_norm": 1.9074118320218076, + "learning_rate": 5.590976474381595e-06, + "loss": 0.9368, + "step": 27844 + }, + { + "epoch": 0.66, + "grad_norm": 1.8711976438599243, + "learning_rate": 5.59029161584028e-06, + "loss": 0.8829, + "step": 27845 + }, + { + "epoch": 0.66, + "grad_norm": 1.8004515043695448, + "learning_rate": 5.589606782973683e-06, + "loss": 0.9039, + "step": 27846 + }, + { + "epoch": 0.66, + "grad_norm": 2.0502400861571606, + "learning_rate": 5.588921975785797e-06, + "loss": 0.9791, + "step": 27847 + }, + { + "epoch": 0.66, + "grad_norm": 2.0304501037146956, + "learning_rate": 5.588237194280603e-06, + "loss": 1.0651, + "step": 27848 + }, + { + "epoch": 0.66, + "grad_norm": 1.9753159990560303, + "learning_rate": 5.5875524384620846e-06, + "loss": 0.8484, + "step": 27849 + }, + { + "epoch": 0.66, + "grad_norm": 1.0415809174102353, + "learning_rate": 5.586867708334236e-06, + "loss": 0.942, + "step": 27850 + }, + { + "epoch": 0.66, + "grad_norm": 2.117554191045695, + "learning_rate": 5.5861830039010375e-06, + "loss": 1.0254, + "step": 27851 + }, + { + "epoch": 0.66, + "grad_norm": 2.0184917971318828, + "learning_rate": 5.585498325166484e-06, + "loss": 1.0142, + "step": 27852 + }, + { + "epoch": 0.66, + "grad_norm": 1.9458899887963674, + "learning_rate": 5.584813672134553e-06, + "loss": 0.8517, + "step": 27853 + }, + { + "epoch": 0.66, + "grad_norm": 1.8606224664425834, + "learning_rate": 5.584129044809237e-06, + "loss": 0.9613, + "step": 27854 + }, + { + "epoch": 0.66, + "grad_norm": 2.081017329177391, + "learning_rate": 5.583444443194517e-06, + "loss": 0.9895, + "step": 27855 + }, + { + "epoch": 0.66, + "grad_norm": 2.0970147636460963, + "learning_rate": 5.582759867294385e-06, + "loss": 1.0383, + "step": 27856 + }, + { + "epoch": 0.66, + "grad_norm": 1.0641362042185054, + "learning_rate": 5.5820753171128235e-06, + "loss": 0.8846, + "step": 27857 + }, + { + "epoch": 0.66, + "grad_norm": 2.342066215933094, + "learning_rate": 5.581390792653815e-06, + "loss": 0.965, + "step": 27858 + }, + { + "epoch": 0.66, + "grad_norm": 2.0657217684981792, + "learning_rate": 5.580706293921348e-06, + "loss": 1.0263, + "step": 27859 + }, + { + "epoch": 0.66, + "grad_norm": 1.975964297128067, + "learning_rate": 5.580021820919411e-06, + "loss": 0.8972, + "step": 27860 + }, + { + "epoch": 0.66, + "grad_norm": 1.131531155143417, + "learning_rate": 5.5793373736519876e-06, + "loss": 0.9329, + "step": 27861 + }, + { + "epoch": 0.66, + "grad_norm": 1.9484224968604467, + "learning_rate": 5.578652952123056e-06, + "loss": 0.96, + "step": 27862 + }, + { + "epoch": 0.66, + "grad_norm": 2.1335051946785955, + "learning_rate": 5.577968556336608e-06, + "loss": 0.9634, + "step": 27863 + }, + { + "epoch": 0.66, + "grad_norm": 1.048910968077665, + "learning_rate": 5.577284186296631e-06, + "loss": 0.9209, + "step": 27864 + }, + { + "epoch": 0.66, + "grad_norm": 1.0759858494846166, + "learning_rate": 5.576599842007105e-06, + "loss": 0.9622, + "step": 27865 + }, + { + "epoch": 0.66, + "grad_norm": 2.2482012496651373, + "learning_rate": 5.575915523472012e-06, + "loss": 0.9746, + "step": 27866 + }, + { + "epoch": 0.66, + "grad_norm": 1.8161284247063063, + "learning_rate": 5.575231230695339e-06, + "loss": 0.9845, + "step": 27867 + }, + { + "epoch": 0.66, + "grad_norm": 1.9591004145565856, + "learning_rate": 5.5745469636810746e-06, + "loss": 1.0608, + "step": 27868 + }, + { + "epoch": 0.66, + "grad_norm": 1.8296701152403785, + "learning_rate": 5.573862722433199e-06, + "loss": 0.9525, + "step": 27869 + }, + { + "epoch": 0.66, + "grad_norm": 1.8093650568164286, + "learning_rate": 5.573178506955692e-06, + "loss": 0.9595, + "step": 27870 + }, + { + "epoch": 0.66, + "grad_norm": 2.026332338734364, + "learning_rate": 5.572494317252545e-06, + "loss": 1.1073, + "step": 27871 + }, + { + "epoch": 0.66, + "grad_norm": 2.108934083085655, + "learning_rate": 5.571810153327734e-06, + "loss": 1.0546, + "step": 27872 + }, + { + "epoch": 0.66, + "grad_norm": 2.0572959904258097, + "learning_rate": 5.5711260151852494e-06, + "loss": 0.9191, + "step": 27873 + }, + { + "epoch": 0.66, + "grad_norm": 1.7242915845165607, + "learning_rate": 5.570441902829069e-06, + "loss": 0.8863, + "step": 27874 + }, + { + "epoch": 0.66, + "grad_norm": 1.8542828390236332, + "learning_rate": 5.569757816263181e-06, + "loss": 0.8569, + "step": 27875 + }, + { + "epoch": 0.66, + "grad_norm": 2.4723945970481043, + "learning_rate": 5.569073755491561e-06, + "loss": 1.0737, + "step": 27876 + }, + { + "epoch": 0.66, + "grad_norm": 2.726578740765643, + "learning_rate": 5.568389720518201e-06, + "loss": 0.9133, + "step": 27877 + }, + { + "epoch": 0.66, + "grad_norm": 1.9859111907669686, + "learning_rate": 5.567705711347075e-06, + "loss": 0.9283, + "step": 27878 + }, + { + "epoch": 0.66, + "grad_norm": 1.8860728393391513, + "learning_rate": 5.5670217279821745e-06, + "loss": 0.8981, + "step": 27879 + }, + { + "epoch": 0.66, + "grad_norm": 2.1564511326479052, + "learning_rate": 5.566337770427472e-06, + "loss": 0.9522, + "step": 27880 + }, + { + "epoch": 0.66, + "grad_norm": 2.1082972347217854, + "learning_rate": 5.565653838686959e-06, + "loss": 1.1073, + "step": 27881 + }, + { + "epoch": 0.66, + "grad_norm": 1.9741650843477885, + "learning_rate": 5.564969932764615e-06, + "loss": 1.0432, + "step": 27882 + }, + { + "epoch": 0.66, + "grad_norm": 2.7388672761074484, + "learning_rate": 5.5642860526644135e-06, + "loss": 0.9961, + "step": 27883 + }, + { + "epoch": 0.66, + "grad_norm": 2.3378499169139837, + "learning_rate": 5.563602198390343e-06, + "loss": 1.0912, + "step": 27884 + }, + { + "epoch": 0.66, + "grad_norm": 1.868929875260553, + "learning_rate": 5.562918369946386e-06, + "loss": 1.0758, + "step": 27885 + }, + { + "epoch": 0.66, + "grad_norm": 2.001895067605921, + "learning_rate": 5.562234567336531e-06, + "loss": 0.9434, + "step": 27886 + }, + { + "epoch": 0.66, + "grad_norm": 2.0028371465211157, + "learning_rate": 5.561550790564744e-06, + "loss": 1.0357, + "step": 27887 + }, + { + "epoch": 0.66, + "grad_norm": 2.0336126890438186, + "learning_rate": 5.560867039635012e-06, + "loss": 1.0044, + "step": 27888 + }, + { + "epoch": 0.66, + "grad_norm": 1.9597559052107125, + "learning_rate": 5.560183314551323e-06, + "loss": 1.0156, + "step": 27889 + }, + { + "epoch": 0.66, + "grad_norm": 2.098164980994891, + "learning_rate": 5.559499615317652e-06, + "loss": 0.9888, + "step": 27890 + }, + { + "epoch": 0.66, + "grad_norm": 2.2206550010408774, + "learning_rate": 5.558815941937975e-06, + "loss": 0.9055, + "step": 27891 + }, + { + "epoch": 0.66, + "grad_norm": 2.218933540085021, + "learning_rate": 5.558132294416278e-06, + "loss": 0.9567, + "step": 27892 + }, + { + "epoch": 0.66, + "grad_norm": 2.180159727851228, + "learning_rate": 5.557448672756546e-06, + "loss": 0.9747, + "step": 27893 + }, + { + "epoch": 0.66, + "grad_norm": 1.907047750804241, + "learning_rate": 5.556765076962749e-06, + "loss": 1.0048, + "step": 27894 + }, + { + "epoch": 0.66, + "grad_norm": 1.8872765929219268, + "learning_rate": 5.556081507038877e-06, + "loss": 0.9597, + "step": 27895 + }, + { + "epoch": 0.66, + "grad_norm": 1.9225197173201989, + "learning_rate": 5.555397962988905e-06, + "loss": 1.0111, + "step": 27896 + }, + { + "epoch": 0.66, + "grad_norm": 1.9641713831429009, + "learning_rate": 5.5547144448168115e-06, + "loss": 0.9236, + "step": 27897 + }, + { + "epoch": 0.66, + "grad_norm": 2.106147152004652, + "learning_rate": 5.554030952526574e-06, + "loss": 1.0855, + "step": 27898 + }, + { + "epoch": 0.66, + "grad_norm": 1.956675999176574, + "learning_rate": 5.553347486122182e-06, + "loss": 0.9834, + "step": 27899 + }, + { + "epoch": 0.66, + "grad_norm": 1.7939620735592992, + "learning_rate": 5.55266404560761e-06, + "loss": 1.0799, + "step": 27900 + }, + { + "epoch": 0.66, + "grad_norm": 1.1597116577552062, + "learning_rate": 5.5519806309868304e-06, + "loss": 0.9539, + "step": 27901 + }, + { + "epoch": 0.66, + "grad_norm": 1.9095336048208698, + "learning_rate": 5.551297242263827e-06, + "loss": 0.8584, + "step": 27902 + }, + { + "epoch": 0.66, + "grad_norm": 2.028244757542197, + "learning_rate": 5.550613879442589e-06, + "loss": 0.8843, + "step": 27903 + }, + { + "epoch": 0.66, + "grad_norm": 1.8901005162265756, + "learning_rate": 5.5499305425270765e-06, + "loss": 0.9674, + "step": 27904 + }, + { + "epoch": 0.66, + "grad_norm": 2.1039817892555663, + "learning_rate": 5.549247231521278e-06, + "loss": 0.9349, + "step": 27905 + }, + { + "epoch": 0.66, + "grad_norm": 1.7194007136191607, + "learning_rate": 5.54856394642917e-06, + "loss": 0.9828, + "step": 27906 + }, + { + "epoch": 0.66, + "grad_norm": 2.959494689175879, + "learning_rate": 5.547880687254739e-06, + "loss": 0.8517, + "step": 27907 + }, + { + "epoch": 0.66, + "grad_norm": 1.9762210774880231, + "learning_rate": 5.5471974540019536e-06, + "loss": 1.0606, + "step": 27908 + }, + { + "epoch": 0.66, + "grad_norm": 1.1067035758512018, + "learning_rate": 5.54651424667479e-06, + "loss": 0.9126, + "step": 27909 + }, + { + "epoch": 0.66, + "grad_norm": 1.849880093142821, + "learning_rate": 5.545831065277236e-06, + "loss": 0.8992, + "step": 27910 + }, + { + "epoch": 0.66, + "grad_norm": 1.8197455638036961, + "learning_rate": 5.545147909813259e-06, + "loss": 1.0347, + "step": 27911 + }, + { + "epoch": 0.66, + "grad_norm": 1.9247349976126835, + "learning_rate": 5.544464780286845e-06, + "loss": 1.0715, + "step": 27912 + }, + { + "epoch": 0.66, + "grad_norm": 1.9589784459036708, + "learning_rate": 5.543781676701964e-06, + "loss": 1.0116, + "step": 27913 + }, + { + "epoch": 0.66, + "grad_norm": 1.988043981603723, + "learning_rate": 5.543098599062602e-06, + "loss": 0.8126, + "step": 27914 + }, + { + "epoch": 0.66, + "grad_norm": 2.048615249787993, + "learning_rate": 5.542415547372724e-06, + "loss": 0.9182, + "step": 27915 + }, + { + "epoch": 0.66, + "grad_norm": 1.950940685374829, + "learning_rate": 5.54173252163632e-06, + "loss": 0.9616, + "step": 27916 + }, + { + "epoch": 0.66, + "grad_norm": 2.090886745372976, + "learning_rate": 5.541049521857361e-06, + "loss": 0.8771, + "step": 27917 + }, + { + "epoch": 0.66, + "grad_norm": 2.3702955820315035, + "learning_rate": 5.5403665480398196e-06, + "loss": 0.9823, + "step": 27918 + }, + { + "epoch": 0.66, + "grad_norm": 2.1012071362003075, + "learning_rate": 5.539683600187675e-06, + "loss": 0.9317, + "step": 27919 + }, + { + "epoch": 0.66, + "grad_norm": 1.9363269487403958, + "learning_rate": 5.539000678304909e-06, + "loss": 0.9266, + "step": 27920 + }, + { + "epoch": 0.66, + "grad_norm": 1.92326433015849, + "learning_rate": 5.538317782395493e-06, + "loss": 0.9728, + "step": 27921 + }, + { + "epoch": 0.66, + "grad_norm": 1.1289473486137833, + "learning_rate": 5.537634912463399e-06, + "loss": 0.9396, + "step": 27922 + }, + { + "epoch": 0.66, + "grad_norm": 1.9403010875523636, + "learning_rate": 5.536952068512609e-06, + "loss": 1.0012, + "step": 27923 + }, + { + "epoch": 0.66, + "grad_norm": 2.0075027106461345, + "learning_rate": 5.536269250547099e-06, + "loss": 0.7759, + "step": 27924 + }, + { + "epoch": 0.66, + "grad_norm": 1.098330782700061, + "learning_rate": 5.535586458570844e-06, + "loss": 0.9422, + "step": 27925 + }, + { + "epoch": 0.66, + "grad_norm": 3.0665233688782223, + "learning_rate": 5.534903692587814e-06, + "loss": 0.9285, + "step": 27926 + }, + { + "epoch": 0.66, + "grad_norm": 2.079014832545295, + "learning_rate": 5.534220952601987e-06, + "loss": 0.9642, + "step": 27927 + }, + { + "epoch": 0.66, + "grad_norm": 1.1097917534762725, + "learning_rate": 5.533538238617345e-06, + "loss": 0.9085, + "step": 27928 + }, + { + "epoch": 0.66, + "grad_norm": 2.0072466633590964, + "learning_rate": 5.532855550637856e-06, + "loss": 0.8743, + "step": 27929 + }, + { + "epoch": 0.66, + "grad_norm": 1.827948062754745, + "learning_rate": 5.532172888667494e-06, + "loss": 1.0021, + "step": 27930 + }, + { + "epoch": 0.66, + "grad_norm": 1.916183554244025, + "learning_rate": 5.531490252710239e-06, + "loss": 1.0146, + "step": 27931 + }, + { + "epoch": 0.66, + "grad_norm": 1.707861485382962, + "learning_rate": 5.53080764277006e-06, + "loss": 0.8942, + "step": 27932 + }, + { + "epoch": 0.66, + "grad_norm": 2.1369148496537975, + "learning_rate": 5.530125058850935e-06, + "loss": 1.0136, + "step": 27933 + }, + { + "epoch": 0.66, + "grad_norm": 1.0899300375529133, + "learning_rate": 5.529442500956834e-06, + "loss": 1.0029, + "step": 27934 + }, + { + "epoch": 0.66, + "grad_norm": 2.2891906030714044, + "learning_rate": 5.528759969091739e-06, + "loss": 1.0335, + "step": 27935 + }, + { + "epoch": 0.66, + "grad_norm": 1.1011936900035555, + "learning_rate": 5.5280774632596145e-06, + "loss": 0.9029, + "step": 27936 + }, + { + "epoch": 0.66, + "grad_norm": 2.1312780303331853, + "learning_rate": 5.5273949834644425e-06, + "loss": 1.0694, + "step": 27937 + }, + { + "epoch": 0.66, + "grad_norm": 2.712223093151268, + "learning_rate": 5.526712529710194e-06, + "loss": 0.8789, + "step": 27938 + }, + { + "epoch": 0.66, + "grad_norm": 2.022701133005903, + "learning_rate": 5.526030102000836e-06, + "loss": 0.9463, + "step": 27939 + }, + { + "epoch": 0.66, + "grad_norm": 2.12807518066442, + "learning_rate": 5.525347700340349e-06, + "loss": 0.974, + "step": 27940 + }, + { + "epoch": 0.66, + "grad_norm": 2.0001606578672546, + "learning_rate": 5.524665324732706e-06, + "loss": 0.918, + "step": 27941 + }, + { + "epoch": 0.66, + "grad_norm": 2.3424684299952094, + "learning_rate": 5.52398297518188e-06, + "loss": 1.0046, + "step": 27942 + }, + { + "epoch": 0.66, + "grad_norm": 1.8477208569030041, + "learning_rate": 5.523300651691837e-06, + "loss": 0.8147, + "step": 27943 + }, + { + "epoch": 0.66, + "grad_norm": 1.8668745114749197, + "learning_rate": 5.522618354266556e-06, + "loss": 0.9925, + "step": 27944 + }, + { + "epoch": 0.66, + "grad_norm": 2.0112132013014485, + "learning_rate": 5.521936082910012e-06, + "loss": 0.866, + "step": 27945 + }, + { + "epoch": 0.66, + "grad_norm": 2.2165689950979526, + "learning_rate": 5.521253837626175e-06, + "loss": 1.0025, + "step": 27946 + }, + { + "epoch": 0.66, + "grad_norm": 2.7041883729361293, + "learning_rate": 5.5205716184190106e-06, + "loss": 1.0716, + "step": 27947 + }, + { + "epoch": 0.66, + "grad_norm": 1.8612540655373118, + "learning_rate": 5.519889425292496e-06, + "loss": 1.0615, + "step": 27948 + }, + { + "epoch": 0.66, + "grad_norm": 1.1146654982553743, + "learning_rate": 5.5192072582506086e-06, + "loss": 0.9584, + "step": 27949 + }, + { + "epoch": 0.66, + "grad_norm": 2.6654833326883365, + "learning_rate": 5.5185251172973155e-06, + "loss": 1.1205, + "step": 27950 + }, + { + "epoch": 0.66, + "grad_norm": 2.148628472380238, + "learning_rate": 5.517843002436583e-06, + "loss": 0.9702, + "step": 27951 + }, + { + "epoch": 0.66, + "grad_norm": 3.502055139739164, + "learning_rate": 5.517160913672393e-06, + "loss": 0.911, + "step": 27952 + }, + { + "epoch": 0.66, + "grad_norm": 3.0505313600758464, + "learning_rate": 5.516478851008705e-06, + "loss": 0.9765, + "step": 27953 + }, + { + "epoch": 0.66, + "grad_norm": 2.0477881847560457, + "learning_rate": 5.515796814449502e-06, + "loss": 0.889, + "step": 27954 + }, + { + "epoch": 0.66, + "grad_norm": 1.862881942751488, + "learning_rate": 5.515114803998745e-06, + "loss": 1.0075, + "step": 27955 + }, + { + "epoch": 0.66, + "grad_norm": 1.838743543621039, + "learning_rate": 5.5144328196604145e-06, + "loss": 0.9941, + "step": 27956 + }, + { + "epoch": 0.66, + "grad_norm": 1.945992783510945, + "learning_rate": 5.513750861438472e-06, + "loss": 0.8966, + "step": 27957 + }, + { + "epoch": 0.66, + "grad_norm": 2.123783498892095, + "learning_rate": 5.5130689293368955e-06, + "loss": 1.0761, + "step": 27958 + }, + { + "epoch": 0.66, + "grad_norm": 2.5162293366477027, + "learning_rate": 5.5123870233596536e-06, + "loss": 1.0259, + "step": 27959 + }, + { + "epoch": 0.66, + "grad_norm": 2.00269576231339, + "learning_rate": 5.511705143510711e-06, + "loss": 1.041, + "step": 27960 + }, + { + "epoch": 0.66, + "grad_norm": 1.7843823407875203, + "learning_rate": 5.51102328979404e-06, + "loss": 0.9201, + "step": 27961 + }, + { + "epoch": 0.66, + "grad_norm": 2.051774026601492, + "learning_rate": 5.510341462213619e-06, + "loss": 0.8799, + "step": 27962 + }, + { + "epoch": 0.66, + "grad_norm": 2.1180021883463045, + "learning_rate": 5.50965966077341e-06, + "loss": 1.1062, + "step": 27963 + }, + { + "epoch": 0.66, + "grad_norm": 1.9807644634804256, + "learning_rate": 5.50897788547738e-06, + "loss": 0.8806, + "step": 27964 + }, + { + "epoch": 0.66, + "grad_norm": 2.053399293402519, + "learning_rate": 5.508296136329504e-06, + "loss": 0.9175, + "step": 27965 + }, + { + "epoch": 0.66, + "grad_norm": 2.0021383094038354, + "learning_rate": 5.507614413333753e-06, + "loss": 0.9711, + "step": 27966 + }, + { + "epoch": 0.66, + "grad_norm": 2.093901699000869, + "learning_rate": 5.506932716494093e-06, + "loss": 1.0228, + "step": 27967 + }, + { + "epoch": 0.66, + "grad_norm": 2.8429602964862077, + "learning_rate": 5.50625104581449e-06, + "loss": 0.8578, + "step": 27968 + }, + { + "epoch": 0.66, + "grad_norm": 2.494566049099365, + "learning_rate": 5.505569401298916e-06, + "loss": 0.9642, + "step": 27969 + }, + { + "epoch": 0.66, + "grad_norm": 2.0646266978197128, + "learning_rate": 5.504887782951343e-06, + "loss": 0.9231, + "step": 27970 + }, + { + "epoch": 0.66, + "grad_norm": 2.0416186815254105, + "learning_rate": 5.504206190775736e-06, + "loss": 0.9184, + "step": 27971 + }, + { + "epoch": 0.66, + "grad_norm": 2.15748898042036, + "learning_rate": 5.503524624776061e-06, + "loss": 0.9652, + "step": 27972 + }, + { + "epoch": 0.66, + "grad_norm": 2.238253312098817, + "learning_rate": 5.502843084956293e-06, + "loss": 1.0726, + "step": 27973 + }, + { + "epoch": 0.66, + "grad_norm": 1.8168503733627748, + "learning_rate": 5.502161571320393e-06, + "loss": 1.0212, + "step": 27974 + }, + { + "epoch": 0.66, + "grad_norm": 2.1243195658369043, + "learning_rate": 5.501480083872335e-06, + "loss": 0.926, + "step": 27975 + }, + { + "epoch": 0.66, + "grad_norm": 1.976646414608605, + "learning_rate": 5.500798622616081e-06, + "loss": 1.0636, + "step": 27976 + }, + { + "epoch": 0.66, + "grad_norm": 1.8327749074629887, + "learning_rate": 5.500117187555604e-06, + "loss": 0.9764, + "step": 27977 + }, + { + "epoch": 0.66, + "grad_norm": 1.9050803142845554, + "learning_rate": 5.499435778694868e-06, + "loss": 1.0015, + "step": 27978 + }, + { + "epoch": 0.66, + "grad_norm": 2.2549724810908454, + "learning_rate": 5.4987543960378445e-06, + "loss": 0.8633, + "step": 27979 + }, + { + "epoch": 0.66, + "grad_norm": 2.147740532683346, + "learning_rate": 5.498073039588498e-06, + "loss": 1.067, + "step": 27980 + }, + { + "epoch": 0.66, + "grad_norm": 1.8491084329057332, + "learning_rate": 5.497391709350792e-06, + "loss": 0.9495, + "step": 27981 + }, + { + "epoch": 0.66, + "grad_norm": 1.7968572250943482, + "learning_rate": 5.496710405328698e-06, + "loss": 0.9032, + "step": 27982 + }, + { + "epoch": 0.66, + "grad_norm": 1.9329091314075926, + "learning_rate": 5.496029127526179e-06, + "loss": 1.0258, + "step": 27983 + }, + { + "epoch": 0.66, + "grad_norm": 2.2474180208053385, + "learning_rate": 5.495347875947215e-06, + "loss": 1.0044, + "step": 27984 + }, + { + "epoch": 0.66, + "grad_norm": 3.4830126497292526, + "learning_rate": 5.494666650595752e-06, + "loss": 0.9226, + "step": 27985 + }, + { + "epoch": 0.66, + "grad_norm": 1.81729861650674, + "learning_rate": 5.4939854514757675e-06, + "loss": 0.932, + "step": 27986 + }, + { + "epoch": 0.66, + "grad_norm": 2.0611665378142225, + "learning_rate": 5.493304278591229e-06, + "loss": 1.0787, + "step": 27987 + }, + { + "epoch": 0.66, + "grad_norm": 2.259086859937878, + "learning_rate": 5.4926231319461e-06, + "loss": 0.9605, + "step": 27988 + }, + { + "epoch": 0.66, + "grad_norm": 1.9527670004094284, + "learning_rate": 5.491942011544342e-06, + "loss": 1.0195, + "step": 27989 + }, + { + "epoch": 0.66, + "grad_norm": 2.0346272522624265, + "learning_rate": 5.4912609173899255e-06, + "loss": 0.9589, + "step": 27990 + }, + { + "epoch": 0.66, + "grad_norm": 2.05020341770721, + "learning_rate": 5.490579849486819e-06, + "loss": 0.9846, + "step": 27991 + }, + { + "epoch": 0.66, + "grad_norm": 2.020447120750598, + "learning_rate": 5.48989880783898e-06, + "loss": 0.9983, + "step": 27992 + }, + { + "epoch": 0.66, + "grad_norm": 1.102013341819569, + "learning_rate": 5.489217792450383e-06, + "loss": 0.9561, + "step": 27993 + }, + { + "epoch": 0.66, + "grad_norm": 4.233107946366416, + "learning_rate": 5.488536803324989e-06, + "loss": 1.0459, + "step": 27994 + }, + { + "epoch": 0.66, + "grad_norm": 2.1518943536195496, + "learning_rate": 5.487855840466756e-06, + "loss": 1.0027, + "step": 27995 + }, + { + "epoch": 0.66, + "grad_norm": 1.9729039192522195, + "learning_rate": 5.4871749038796554e-06, + "loss": 1.0831, + "step": 27996 + }, + { + "epoch": 0.66, + "grad_norm": 1.8481145873976186, + "learning_rate": 5.486493993567654e-06, + "loss": 0.9518, + "step": 27997 + }, + { + "epoch": 0.66, + "grad_norm": 1.058167539597041, + "learning_rate": 5.485813109534716e-06, + "loss": 0.9726, + "step": 27998 + }, + { + "epoch": 0.66, + "grad_norm": 1.9039745503072294, + "learning_rate": 5.485132251784797e-06, + "loss": 1.004, + "step": 27999 + }, + { + "epoch": 0.66, + "grad_norm": 1.8742366202246068, + "learning_rate": 5.484451420321869e-06, + "loss": 1.0195, + "step": 28000 + }, + { + "epoch": 0.66, + "grad_norm": 2.097892874252881, + "learning_rate": 5.483770615149898e-06, + "loss": 0.9839, + "step": 28001 + }, + { + "epoch": 0.66, + "grad_norm": 2.1132908003832767, + "learning_rate": 5.483089836272844e-06, + "loss": 0.9984, + "step": 28002 + }, + { + "epoch": 0.66, + "grad_norm": 2.442433882302934, + "learning_rate": 5.482409083694667e-06, + "loss": 1.0202, + "step": 28003 + }, + { + "epoch": 0.66, + "grad_norm": 1.109365126362704, + "learning_rate": 5.481728357419336e-06, + "loss": 0.955, + "step": 28004 + }, + { + "epoch": 0.66, + "grad_norm": 1.9697806703037133, + "learning_rate": 5.481047657450816e-06, + "loss": 0.9487, + "step": 28005 + }, + { + "epoch": 0.66, + "grad_norm": 2.3205052568264106, + "learning_rate": 5.480366983793066e-06, + "loss": 1.1368, + "step": 28006 + }, + { + "epoch": 0.66, + "grad_norm": 2.164262126974765, + "learning_rate": 5.479686336450047e-06, + "loss": 1.0179, + "step": 28007 + }, + { + "epoch": 0.66, + "grad_norm": 2.0894777682184413, + "learning_rate": 5.479005715425725e-06, + "loss": 1.0543, + "step": 28008 + }, + { + "epoch": 0.66, + "grad_norm": 1.8008957811434507, + "learning_rate": 5.4783251207240685e-06, + "loss": 0.9441, + "step": 28009 + }, + { + "epoch": 0.66, + "grad_norm": 2.0760789011111087, + "learning_rate": 5.477644552349033e-06, + "loss": 1.1264, + "step": 28010 + }, + { + "epoch": 0.66, + "grad_norm": 2.036252055690164, + "learning_rate": 5.476964010304579e-06, + "loss": 0.9739, + "step": 28011 + }, + { + "epoch": 0.66, + "grad_norm": 2.0888378249142674, + "learning_rate": 5.476283494594677e-06, + "loss": 1.0256, + "step": 28012 + }, + { + "epoch": 0.66, + "grad_norm": 1.9007251307639945, + "learning_rate": 5.475603005223281e-06, + "loss": 1.015, + "step": 28013 + }, + { + "epoch": 0.66, + "grad_norm": 1.8309653578266818, + "learning_rate": 5.474922542194361e-06, + "loss": 0.9519, + "step": 28014 + }, + { + "epoch": 0.66, + "grad_norm": 2.048905230651775, + "learning_rate": 5.4742421055118686e-06, + "loss": 0.9635, + "step": 28015 + }, + { + "epoch": 0.66, + "grad_norm": 2.690142544206149, + "learning_rate": 5.473561695179777e-06, + "loss": 1.0917, + "step": 28016 + }, + { + "epoch": 0.66, + "grad_norm": 2.032939354073487, + "learning_rate": 5.472881311202039e-06, + "loss": 0.9308, + "step": 28017 + }, + { + "epoch": 0.66, + "grad_norm": 2.3314930146543036, + "learning_rate": 5.472200953582622e-06, + "loss": 1.0258, + "step": 28018 + }, + { + "epoch": 0.66, + "grad_norm": 1.9208694385153897, + "learning_rate": 5.471520622325484e-06, + "loss": 1.096, + "step": 28019 + }, + { + "epoch": 0.66, + "grad_norm": 2.403305627721048, + "learning_rate": 5.470840317434584e-06, + "loss": 1.0262, + "step": 28020 + }, + { + "epoch": 0.66, + "grad_norm": 1.9952382240281064, + "learning_rate": 5.470160038913884e-06, + "loss": 0.9579, + "step": 28021 + }, + { + "epoch": 0.66, + "grad_norm": 3.254642434490359, + "learning_rate": 5.469479786767351e-06, + "loss": 1.1068, + "step": 28022 + }, + { + "epoch": 0.66, + "grad_norm": 2.058648484884939, + "learning_rate": 5.468799560998942e-06, + "loss": 0.9981, + "step": 28023 + }, + { + "epoch": 0.66, + "grad_norm": 2.3630503467788504, + "learning_rate": 5.468119361612612e-06, + "loss": 0.9978, + "step": 28024 + }, + { + "epoch": 0.66, + "grad_norm": 2.9007289471606192, + "learning_rate": 5.467439188612326e-06, + "loss": 1.0961, + "step": 28025 + }, + { + "epoch": 0.66, + "grad_norm": 2.018679962252544, + "learning_rate": 5.4667590420020475e-06, + "loss": 0.9762, + "step": 28026 + }, + { + "epoch": 0.66, + "grad_norm": 2.217854233680573, + "learning_rate": 5.466078921785733e-06, + "loss": 0.9444, + "step": 28027 + }, + { + "epoch": 0.66, + "grad_norm": 1.8343003740150974, + "learning_rate": 5.465398827967339e-06, + "loss": 0.9121, + "step": 28028 + }, + { + "epoch": 0.66, + "grad_norm": 2.2849712897094827, + "learning_rate": 5.464718760550829e-06, + "loss": 1.0328, + "step": 28029 + }, + { + "epoch": 0.66, + "grad_norm": 2.3308736048018415, + "learning_rate": 5.464038719540166e-06, + "loss": 1.127, + "step": 28030 + }, + { + "epoch": 0.66, + "grad_norm": 1.075507804395864, + "learning_rate": 5.463358704939305e-06, + "loss": 0.9687, + "step": 28031 + }, + { + "epoch": 0.66, + "grad_norm": 2.3238207658656624, + "learning_rate": 5.462678716752201e-06, + "loss": 0.9999, + "step": 28032 + }, + { + "epoch": 0.66, + "grad_norm": 2.136812533943479, + "learning_rate": 5.461998754982825e-06, + "loss": 0.9855, + "step": 28033 + }, + { + "epoch": 0.66, + "grad_norm": 2.159910680911184, + "learning_rate": 5.461318819635122e-06, + "loss": 0.9666, + "step": 28034 + }, + { + "epoch": 0.66, + "grad_norm": 2.0499036443830776, + "learning_rate": 5.4606389107130634e-06, + "loss": 1.0846, + "step": 28035 + }, + { + "epoch": 0.66, + "grad_norm": 2.764956974770126, + "learning_rate": 5.459959028220597e-06, + "loss": 0.7849, + "step": 28036 + }, + { + "epoch": 0.66, + "grad_norm": 1.12375274606534, + "learning_rate": 5.459279172161691e-06, + "loss": 0.9741, + "step": 28037 + }, + { + "epoch": 0.66, + "grad_norm": 1.970689098659841, + "learning_rate": 5.458599342540296e-06, + "loss": 1.0404, + "step": 28038 + }, + { + "epoch": 0.66, + "grad_norm": 1.9922451154708085, + "learning_rate": 5.457919539360377e-06, + "loss": 1.001, + "step": 28039 + }, + { + "epoch": 0.66, + "grad_norm": 1.7777540757674588, + "learning_rate": 5.457239762625888e-06, + "loss": 0.8878, + "step": 28040 + }, + { + "epoch": 0.66, + "grad_norm": 1.9671643315428773, + "learning_rate": 5.456560012340784e-06, + "loss": 0.9198, + "step": 28041 + }, + { + "epoch": 0.66, + "grad_norm": 1.964809817185614, + "learning_rate": 5.4558802885090255e-06, + "loss": 1.0246, + "step": 28042 + }, + { + "epoch": 0.66, + "grad_norm": 1.9952307347640457, + "learning_rate": 5.455200591134574e-06, + "loss": 1.0255, + "step": 28043 + }, + { + "epoch": 0.66, + "grad_norm": 2.0519281114207937, + "learning_rate": 5.4545209202213845e-06, + "loss": 0.9209, + "step": 28044 + }, + { + "epoch": 0.66, + "grad_norm": 2.636443254242162, + "learning_rate": 5.453841275773408e-06, + "loss": 0.9932, + "step": 28045 + }, + { + "epoch": 0.66, + "grad_norm": 1.9386568019295585, + "learning_rate": 5.453161657794609e-06, + "loss": 0.9914, + "step": 28046 + }, + { + "epoch": 0.66, + "grad_norm": 2.7402065741990556, + "learning_rate": 5.452482066288945e-06, + "loss": 1.185, + "step": 28047 + }, + { + "epoch": 0.66, + "grad_norm": 2.0372882794860554, + "learning_rate": 5.45180250126037e-06, + "loss": 1.0645, + "step": 28048 + }, + { + "epoch": 0.66, + "grad_norm": 8.066194407144023, + "learning_rate": 5.451122962712838e-06, + "loss": 1.0099, + "step": 28049 + }, + { + "epoch": 0.66, + "grad_norm": 1.1464824589172287, + "learning_rate": 5.450443450650306e-06, + "loss": 0.9411, + "step": 28050 + }, + { + "epoch": 0.66, + "grad_norm": 2.1732306410120295, + "learning_rate": 5.449763965076739e-06, + "loss": 0.8481, + "step": 28051 + }, + { + "epoch": 0.66, + "grad_norm": 1.887380034584337, + "learning_rate": 5.449084505996085e-06, + "loss": 0.9198, + "step": 28052 + }, + { + "epoch": 0.66, + "grad_norm": 1.9228093287700876, + "learning_rate": 5.4484050734122995e-06, + "loss": 0.8598, + "step": 28053 + }, + { + "epoch": 0.66, + "grad_norm": 1.7817953371236068, + "learning_rate": 5.447725667329343e-06, + "loss": 1.0126, + "step": 28054 + }, + { + "epoch": 0.66, + "grad_norm": 2.1182120035478786, + "learning_rate": 5.447046287751165e-06, + "loss": 1.0188, + "step": 28055 + }, + { + "epoch": 0.66, + "grad_norm": 2.0306292604983893, + "learning_rate": 5.446366934681731e-06, + "loss": 0.9177, + "step": 28056 + }, + { + "epoch": 0.66, + "grad_norm": 1.8122941622976756, + "learning_rate": 5.445687608124984e-06, + "loss": 1.0804, + "step": 28057 + }, + { + "epoch": 0.66, + "grad_norm": 1.9729174174912436, + "learning_rate": 5.445008308084892e-06, + "loss": 1.0354, + "step": 28058 + }, + { + "epoch": 0.66, + "grad_norm": 2.015902928039984, + "learning_rate": 5.444329034565397e-06, + "loss": 0.9275, + "step": 28059 + }, + { + "epoch": 0.66, + "grad_norm": 2.0716258836962895, + "learning_rate": 5.443649787570467e-06, + "loss": 1.0335, + "step": 28060 + }, + { + "epoch": 0.66, + "grad_norm": 1.993377735596562, + "learning_rate": 5.4429705671040486e-06, + "loss": 1.0506, + "step": 28061 + }, + { + "epoch": 0.66, + "grad_norm": 2.1199499858973203, + "learning_rate": 5.442291373170095e-06, + "loss": 1.0143, + "step": 28062 + }, + { + "epoch": 0.66, + "grad_norm": 1.94844249768771, + "learning_rate": 5.441612205772564e-06, + "loss": 1.0179, + "step": 28063 + }, + { + "epoch": 0.66, + "grad_norm": 2.775533403016608, + "learning_rate": 5.440933064915414e-06, + "loss": 0.9755, + "step": 28064 + }, + { + "epoch": 0.66, + "grad_norm": 2.8174343496134315, + "learning_rate": 5.440253950602594e-06, + "loss": 1.0114, + "step": 28065 + }, + { + "epoch": 0.66, + "grad_norm": 1.8160315719379414, + "learning_rate": 5.439574862838057e-06, + "loss": 1.0164, + "step": 28066 + }, + { + "epoch": 0.66, + "grad_norm": 1.8937328543389464, + "learning_rate": 5.438895801625758e-06, + "loss": 0.9415, + "step": 28067 + }, + { + "epoch": 0.66, + "grad_norm": 1.7692455959595241, + "learning_rate": 5.438216766969655e-06, + "loss": 0.9879, + "step": 28068 + }, + { + "epoch": 0.66, + "grad_norm": 1.0582179272490508, + "learning_rate": 5.437537758873699e-06, + "loss": 0.9539, + "step": 28069 + }, + { + "epoch": 0.66, + "grad_norm": 1.1611548486644623, + "learning_rate": 5.436858777341839e-06, + "loss": 0.9629, + "step": 28070 + }, + { + "epoch": 0.66, + "grad_norm": 2.263217706021994, + "learning_rate": 5.436179822378031e-06, + "loss": 0.9433, + "step": 28071 + }, + { + "epoch": 0.66, + "grad_norm": 2.310866519099514, + "learning_rate": 5.435500893986234e-06, + "loss": 1.0686, + "step": 28072 + }, + { + "epoch": 0.66, + "grad_norm": 1.9092173520809592, + "learning_rate": 5.434821992170396e-06, + "loss": 0.947, + "step": 28073 + }, + { + "epoch": 0.66, + "grad_norm": 1.8378985681727016, + "learning_rate": 5.434143116934465e-06, + "loss": 0.9149, + "step": 28074 + }, + { + "epoch": 0.66, + "grad_norm": 2.688881482629218, + "learning_rate": 5.433464268282401e-06, + "loss": 0.8689, + "step": 28075 + }, + { + "epoch": 0.66, + "grad_norm": 2.0640494456977527, + "learning_rate": 5.4327854462181515e-06, + "loss": 0.8438, + "step": 28076 + }, + { + "epoch": 0.66, + "grad_norm": 1.9325063607045005, + "learning_rate": 5.432106650745675e-06, + "loss": 1.0023, + "step": 28077 + }, + { + "epoch": 0.66, + "grad_norm": 2.1676054420823405, + "learning_rate": 5.431427881868914e-06, + "loss": 0.9915, + "step": 28078 + }, + { + "epoch": 0.66, + "grad_norm": 2.216945498322019, + "learning_rate": 5.430749139591833e-06, + "loss": 1.0441, + "step": 28079 + }, + { + "epoch": 0.66, + "grad_norm": 2.0380679405306137, + "learning_rate": 5.430070423918371e-06, + "loss": 1.0732, + "step": 28080 + }, + { + "epoch": 0.66, + "grad_norm": 2.0772211136543723, + "learning_rate": 5.429391734852487e-06, + "loss": 1.0251, + "step": 28081 + }, + { + "epoch": 0.66, + "grad_norm": 2.114616493729626, + "learning_rate": 5.428713072398139e-06, + "loss": 1.0117, + "step": 28082 + }, + { + "epoch": 0.66, + "grad_norm": 1.9246073247999471, + "learning_rate": 5.4280344365592615e-06, + "loss": 0.9698, + "step": 28083 + }, + { + "epoch": 0.66, + "grad_norm": 1.9736149910731404, + "learning_rate": 5.427355827339818e-06, + "loss": 0.9312, + "step": 28084 + }, + { + "epoch": 0.66, + "grad_norm": 2.25278766657352, + "learning_rate": 5.4266772447437535e-06, + "loss": 0.8853, + "step": 28085 + }, + { + "epoch": 0.66, + "grad_norm": 1.9579763387947267, + "learning_rate": 5.4259986887750315e-06, + "loss": 0.848, + "step": 28086 + }, + { + "epoch": 0.66, + "grad_norm": 2.0266415859482403, + "learning_rate": 5.4253201594375835e-06, + "loss": 0.933, + "step": 28087 + }, + { + "epoch": 0.66, + "grad_norm": 3.9297053843744427, + "learning_rate": 5.424641656735372e-06, + "loss": 0.911, + "step": 28088 + }, + { + "epoch": 0.66, + "grad_norm": 1.8722074505200317, + "learning_rate": 5.423963180672349e-06, + "loss": 1.0648, + "step": 28089 + }, + { + "epoch": 0.66, + "grad_norm": 2.065231326567967, + "learning_rate": 5.4232847312524565e-06, + "loss": 1.1245, + "step": 28090 + }, + { + "epoch": 0.66, + "grad_norm": 1.9570675585285338, + "learning_rate": 5.4226063084796536e-06, + "loss": 1.0385, + "step": 28091 + }, + { + "epoch": 0.66, + "grad_norm": 1.977950861404948, + "learning_rate": 5.421927912357881e-06, + "loss": 0.8668, + "step": 28092 + }, + { + "epoch": 0.66, + "grad_norm": 2.0494460066506424, + "learning_rate": 5.421249542891098e-06, + "loss": 1.0884, + "step": 28093 + }, + { + "epoch": 0.66, + "grad_norm": 1.976750353605085, + "learning_rate": 5.420571200083246e-06, + "loss": 0.9591, + "step": 28094 + }, + { + "epoch": 0.66, + "grad_norm": 2.4607244252687823, + "learning_rate": 5.419892883938282e-06, + "loss": 0.8708, + "step": 28095 + }, + { + "epoch": 0.66, + "grad_norm": 2.6497294367092947, + "learning_rate": 5.419214594460152e-06, + "loss": 1.0398, + "step": 28096 + }, + { + "epoch": 0.66, + "grad_norm": 2.7858033683252517, + "learning_rate": 5.418536331652802e-06, + "loss": 0.9469, + "step": 28097 + }, + { + "epoch": 0.66, + "grad_norm": 1.0649685898844046, + "learning_rate": 5.417858095520182e-06, + "loss": 0.9336, + "step": 28098 + }, + { + "epoch": 0.66, + "grad_norm": 2.2904503941951946, + "learning_rate": 5.417179886066249e-06, + "loss": 1.0075, + "step": 28099 + }, + { + "epoch": 0.66, + "grad_norm": 2.15612940077404, + "learning_rate": 5.416501703294943e-06, + "loss": 0.8883, + "step": 28100 + }, + { + "epoch": 0.66, + "grad_norm": 1.1628516257057095, + "learning_rate": 5.415823547210213e-06, + "loss": 0.9737, + "step": 28101 + }, + { + "epoch": 0.66, + "grad_norm": 2.1787021386504484, + "learning_rate": 5.415145417816008e-06, + "loss": 0.9928, + "step": 28102 + }, + { + "epoch": 0.66, + "grad_norm": 2.411077246135754, + "learning_rate": 5.414467315116283e-06, + "loss": 1.032, + "step": 28103 + }, + { + "epoch": 0.66, + "grad_norm": 1.7596071527971808, + "learning_rate": 5.4137892391149815e-06, + "loss": 0.8138, + "step": 28104 + }, + { + "epoch": 0.66, + "grad_norm": 2.3251584866013557, + "learning_rate": 5.413111189816047e-06, + "loss": 1.1156, + "step": 28105 + }, + { + "epoch": 0.66, + "grad_norm": 3.2902807236911404, + "learning_rate": 5.41243316722343e-06, + "loss": 1.0031, + "step": 28106 + }, + { + "epoch": 0.66, + "grad_norm": 2.127551591139475, + "learning_rate": 5.411755171341085e-06, + "loss": 0.9574, + "step": 28107 + }, + { + "epoch": 0.66, + "grad_norm": 1.9570212108745748, + "learning_rate": 5.411077202172953e-06, + "loss": 1.0269, + "step": 28108 + }, + { + "epoch": 0.66, + "grad_norm": 2.2592606137837077, + "learning_rate": 5.410399259722978e-06, + "loss": 1.1407, + "step": 28109 + }, + { + "epoch": 0.66, + "grad_norm": 1.1195582000675748, + "learning_rate": 5.409721343995116e-06, + "loss": 0.9967, + "step": 28110 + }, + { + "epoch": 0.66, + "grad_norm": 2.157956968052571, + "learning_rate": 5.409043454993307e-06, + "loss": 0.9034, + "step": 28111 + }, + { + "epoch": 0.66, + "grad_norm": 2.2420698833997688, + "learning_rate": 5.408365592721503e-06, + "loss": 1.0301, + "step": 28112 + }, + { + "epoch": 0.66, + "grad_norm": 1.793002761594067, + "learning_rate": 5.407687757183645e-06, + "loss": 1.0481, + "step": 28113 + }, + { + "epoch": 0.66, + "grad_norm": 1.7738690378362731, + "learning_rate": 5.407009948383686e-06, + "loss": 0.9232, + "step": 28114 + }, + { + "epoch": 0.66, + "grad_norm": 1.98131817820625, + "learning_rate": 5.406332166325565e-06, + "loss": 0.9408, + "step": 28115 + }, + { + "epoch": 0.66, + "grad_norm": 1.9068737487439558, + "learning_rate": 5.405654411013237e-06, + "loss": 0.9949, + "step": 28116 + }, + { + "epoch": 0.66, + "grad_norm": 1.8508487737456374, + "learning_rate": 5.404976682450644e-06, + "loss": 1.0653, + "step": 28117 + }, + { + "epoch": 0.66, + "grad_norm": 2.3516813921220074, + "learning_rate": 5.404298980641727e-06, + "loss": 0.9308, + "step": 28118 + }, + { + "epoch": 0.66, + "grad_norm": 1.9332466796574925, + "learning_rate": 5.403621305590436e-06, + "loss": 1.0245, + "step": 28119 + }, + { + "epoch": 0.66, + "grad_norm": 2.2553592226517534, + "learning_rate": 5.402943657300722e-06, + "loss": 0.9074, + "step": 28120 + }, + { + "epoch": 0.66, + "grad_norm": 1.8729850767052454, + "learning_rate": 5.402266035776525e-06, + "loss": 1.0174, + "step": 28121 + }, + { + "epoch": 0.66, + "grad_norm": 2.004318541481431, + "learning_rate": 5.401588441021786e-06, + "loss": 0.9645, + "step": 28122 + }, + { + "epoch": 0.66, + "grad_norm": 2.282895624580645, + "learning_rate": 5.400910873040456e-06, + "loss": 0.9638, + "step": 28123 + }, + { + "epoch": 0.66, + "grad_norm": 1.914052898397497, + "learning_rate": 5.400233331836482e-06, + "loss": 0.9972, + "step": 28124 + }, + { + "epoch": 0.66, + "grad_norm": 2.1036461831240745, + "learning_rate": 5.399555817413806e-06, + "loss": 0.9738, + "step": 28125 + }, + { + "epoch": 0.66, + "grad_norm": 3.3365723932425952, + "learning_rate": 5.398878329776369e-06, + "loss": 0.9237, + "step": 28126 + }, + { + "epoch": 0.66, + "grad_norm": 2.0918965452822142, + "learning_rate": 5.398200868928118e-06, + "loss": 0.9572, + "step": 28127 + }, + { + "epoch": 0.66, + "grad_norm": 1.7641235699358613, + "learning_rate": 5.397523434873004e-06, + "loss": 0.9527, + "step": 28128 + }, + { + "epoch": 0.66, + "grad_norm": 1.742534451085962, + "learning_rate": 5.3968460276149645e-06, + "loss": 0.9144, + "step": 28129 + }, + { + "epoch": 0.66, + "grad_norm": 2.075713812229635, + "learning_rate": 5.396168647157942e-06, + "loss": 1.0359, + "step": 28130 + }, + { + "epoch": 0.66, + "grad_norm": 2.052991492484111, + "learning_rate": 5.395491293505881e-06, + "loss": 0.9224, + "step": 28131 + }, + { + "epoch": 0.66, + "grad_norm": 2.484849942610612, + "learning_rate": 5.3948139666627335e-06, + "loss": 0.9955, + "step": 28132 + }, + { + "epoch": 0.66, + "grad_norm": 1.9034265523523908, + "learning_rate": 5.394136666632437e-06, + "loss": 0.9025, + "step": 28133 + }, + { + "epoch": 0.66, + "grad_norm": 2.1289563885086724, + "learning_rate": 5.3934593934189295e-06, + "loss": 1.0611, + "step": 28134 + }, + { + "epoch": 0.66, + "grad_norm": 1.9330735944103565, + "learning_rate": 5.392782147026166e-06, + "loss": 0.8476, + "step": 28135 + }, + { + "epoch": 0.66, + "grad_norm": 1.9489676852785993, + "learning_rate": 5.392104927458077e-06, + "loss": 0.9848, + "step": 28136 + }, + { + "epoch": 0.66, + "grad_norm": 2.1888647591074575, + "learning_rate": 5.3914277347186174e-06, + "loss": 1.0764, + "step": 28137 + }, + { + "epoch": 0.66, + "grad_norm": 2.0653114720145163, + "learning_rate": 5.39075056881172e-06, + "loss": 0.9205, + "step": 28138 + }, + { + "epoch": 0.66, + "grad_norm": 1.9927459672655061, + "learning_rate": 5.390073429741336e-06, + "loss": 0.888, + "step": 28139 + }, + { + "epoch": 0.66, + "grad_norm": 1.0699491660758909, + "learning_rate": 5.389396317511401e-06, + "loss": 0.9183, + "step": 28140 + }, + { + "epoch": 0.66, + "grad_norm": 1.9377892866110713, + "learning_rate": 5.388719232125863e-06, + "loss": 1.0474, + "step": 28141 + }, + { + "epoch": 0.66, + "grad_norm": 2.1112323925871155, + "learning_rate": 5.388042173588662e-06, + "loss": 0.9274, + "step": 28142 + }, + { + "epoch": 0.66, + "grad_norm": 2.1599684577810634, + "learning_rate": 5.387365141903735e-06, + "loss": 0.9486, + "step": 28143 + }, + { + "epoch": 0.66, + "grad_norm": 2.015330521551614, + "learning_rate": 5.386688137075031e-06, + "loss": 0.9423, + "step": 28144 + }, + { + "epoch": 0.66, + "grad_norm": 1.9171845479987564, + "learning_rate": 5.38601115910649e-06, + "loss": 0.9905, + "step": 28145 + }, + { + "epoch": 0.66, + "grad_norm": 2.047700769873946, + "learning_rate": 5.385334208002055e-06, + "loss": 1.0688, + "step": 28146 + }, + { + "epoch": 0.66, + "grad_norm": 2.0599038620231496, + "learning_rate": 5.384657283765661e-06, + "loss": 0.9721, + "step": 28147 + }, + { + "epoch": 0.66, + "grad_norm": 1.9198715889404996, + "learning_rate": 5.383980386401253e-06, + "loss": 1.0616, + "step": 28148 + }, + { + "epoch": 0.66, + "grad_norm": 2.124105016194082, + "learning_rate": 5.383303515912777e-06, + "loss": 0.9384, + "step": 28149 + }, + { + "epoch": 0.66, + "grad_norm": 2.484316345431884, + "learning_rate": 5.382626672304168e-06, + "loss": 0.9364, + "step": 28150 + }, + { + "epoch": 0.66, + "grad_norm": 1.9503678288274537, + "learning_rate": 5.381949855579365e-06, + "loss": 0.9884, + "step": 28151 + }, + { + "epoch": 0.66, + "grad_norm": 2.2396552854607754, + "learning_rate": 5.381273065742313e-06, + "loss": 0.9655, + "step": 28152 + }, + { + "epoch": 0.66, + "grad_norm": 1.0793657714084284, + "learning_rate": 5.380596302796954e-06, + "loss": 0.9341, + "step": 28153 + }, + { + "epoch": 0.66, + "grad_norm": 2.194290522800197, + "learning_rate": 5.379919566747227e-06, + "loss": 1.0051, + "step": 28154 + }, + { + "epoch": 0.66, + "grad_norm": 2.6564523025568256, + "learning_rate": 5.379242857597067e-06, + "loss": 0.8634, + "step": 28155 + }, + { + "epoch": 0.66, + "grad_norm": 1.8669863611484177, + "learning_rate": 5.378566175350421e-06, + "loss": 1.1625, + "step": 28156 + }, + { + "epoch": 0.66, + "grad_norm": 2.238823151648309, + "learning_rate": 5.377889520011224e-06, + "loss": 0.9125, + "step": 28157 + }, + { + "epoch": 0.66, + "grad_norm": 2.130746964777234, + "learning_rate": 5.377212891583419e-06, + "loss": 1.1574, + "step": 28158 + }, + { + "epoch": 0.66, + "grad_norm": 2.02689799491842, + "learning_rate": 5.376536290070942e-06, + "loss": 0.9635, + "step": 28159 + }, + { + "epoch": 0.66, + "grad_norm": 1.9873985875227147, + "learning_rate": 5.375859715477738e-06, + "loss": 1.0883, + "step": 28160 + }, + { + "epoch": 0.66, + "grad_norm": 2.191647295032315, + "learning_rate": 5.3751831678077385e-06, + "loss": 0.9366, + "step": 28161 + }, + { + "epoch": 0.66, + "grad_norm": 1.922175007466966, + "learning_rate": 5.3745066470648895e-06, + "loss": 0.8701, + "step": 28162 + }, + { + "epoch": 0.66, + "grad_norm": 1.9169620717054372, + "learning_rate": 5.373830153253129e-06, + "loss": 1.0142, + "step": 28163 + }, + { + "epoch": 0.66, + "grad_norm": 2.107216111488813, + "learning_rate": 5.373153686376389e-06, + "loss": 1.0152, + "step": 28164 + }, + { + "epoch": 0.66, + "grad_norm": 2.326367065766605, + "learning_rate": 5.3724772464386145e-06, + "loss": 1.0563, + "step": 28165 + }, + { + "epoch": 0.66, + "grad_norm": 1.8917414268133834, + "learning_rate": 5.371800833443741e-06, + "loss": 0.9017, + "step": 28166 + }, + { + "epoch": 0.66, + "grad_norm": 1.7199816365703327, + "learning_rate": 5.371124447395717e-06, + "loss": 1.0265, + "step": 28167 + }, + { + "epoch": 0.66, + "grad_norm": 1.9245380217373507, + "learning_rate": 5.370448088298463e-06, + "loss": 1.0176, + "step": 28168 + }, + { + "epoch": 0.66, + "grad_norm": 2.0571541864999947, + "learning_rate": 5.369771756155926e-06, + "loss": 1.1402, + "step": 28169 + }, + { + "epoch": 0.66, + "grad_norm": 2.073791142050572, + "learning_rate": 5.3690954509720484e-06, + "loss": 1.049, + "step": 28170 + }, + { + "epoch": 0.66, + "grad_norm": 2.648090515116162, + "learning_rate": 5.368419172750763e-06, + "loss": 1.0096, + "step": 28171 + }, + { + "epoch": 0.66, + "grad_norm": 1.972112367384093, + "learning_rate": 5.367742921496004e-06, + "loss": 1.1825, + "step": 28172 + }, + { + "epoch": 0.66, + "grad_norm": 1.135504670904011, + "learning_rate": 5.367066697211711e-06, + "loss": 0.9944, + "step": 28173 + }, + { + "epoch": 0.66, + "grad_norm": 1.9892664931645834, + "learning_rate": 5.366390499901828e-06, + "loss": 1.0471, + "step": 28174 + }, + { + "epoch": 0.66, + "grad_norm": 1.114693502676042, + "learning_rate": 5.365714329570286e-06, + "loss": 0.9484, + "step": 28175 + }, + { + "epoch": 0.66, + "grad_norm": 5.6553176602887545, + "learning_rate": 5.3650381862210185e-06, + "loss": 0.8628, + "step": 28176 + }, + { + "epoch": 0.66, + "grad_norm": 2.3511122670764015, + "learning_rate": 5.364362069857969e-06, + "loss": 1.0176, + "step": 28177 + }, + { + "epoch": 0.66, + "grad_norm": 3.5401481239356323, + "learning_rate": 5.363685980485069e-06, + "loss": 0.9691, + "step": 28178 + }, + { + "epoch": 0.66, + "grad_norm": 2.1557072905304184, + "learning_rate": 5.363009918106256e-06, + "loss": 1.0507, + "step": 28179 + }, + { + "epoch": 0.66, + "grad_norm": 2.0566941402180206, + "learning_rate": 5.362333882725471e-06, + "loss": 1.0005, + "step": 28180 + }, + { + "epoch": 0.66, + "grad_norm": 1.0823471807948923, + "learning_rate": 5.361657874346647e-06, + "loss": 0.9328, + "step": 28181 + }, + { + "epoch": 0.66, + "grad_norm": 1.949328098163635, + "learning_rate": 5.360981892973716e-06, + "loss": 0.8626, + "step": 28182 + }, + { + "epoch": 0.66, + "grad_norm": 2.085717348784416, + "learning_rate": 5.3603059386106145e-06, + "loss": 0.9702, + "step": 28183 + }, + { + "epoch": 0.66, + "grad_norm": 2.608406248307682, + "learning_rate": 5.359630011261291e-06, + "loss": 1.0238, + "step": 28184 + }, + { + "epoch": 0.66, + "grad_norm": 2.0142303121659717, + "learning_rate": 5.3589541109296615e-06, + "loss": 1.0249, + "step": 28185 + }, + { + "epoch": 0.66, + "grad_norm": 3.0412179884774644, + "learning_rate": 5.358278237619672e-06, + "loss": 0.9721, + "step": 28186 + }, + { + "epoch": 0.66, + "grad_norm": 2.0414087469631865, + "learning_rate": 5.357602391335255e-06, + "loss": 1.0592, + "step": 28187 + }, + { + "epoch": 0.66, + "grad_norm": 2.23579378777284, + "learning_rate": 5.35692657208035e-06, + "loss": 0.9458, + "step": 28188 + }, + { + "epoch": 0.66, + "grad_norm": 1.1202897439458446, + "learning_rate": 5.356250779858889e-06, + "loss": 1.0312, + "step": 28189 + }, + { + "epoch": 0.66, + "grad_norm": 2.1877596006090894, + "learning_rate": 5.3555750146748034e-06, + "loss": 1.0463, + "step": 28190 + }, + { + "epoch": 0.66, + "grad_norm": 2.328200153943553, + "learning_rate": 5.354899276532033e-06, + "loss": 0.9401, + "step": 28191 + }, + { + "epoch": 0.66, + "grad_norm": 2.0443594505480913, + "learning_rate": 5.354223565434507e-06, + "loss": 1.0671, + "step": 28192 + }, + { + "epoch": 0.66, + "grad_norm": 1.8482642485848817, + "learning_rate": 5.353547881386165e-06, + "loss": 1.0253, + "step": 28193 + }, + { + "epoch": 0.66, + "grad_norm": 2.3090364238311034, + "learning_rate": 5.3528722243909346e-06, + "loss": 0.9578, + "step": 28194 + }, + { + "epoch": 0.66, + "grad_norm": 2.287174536251575, + "learning_rate": 5.352196594452756e-06, + "loss": 1.0378, + "step": 28195 + }, + { + "epoch": 0.66, + "grad_norm": 1.8824860181673542, + "learning_rate": 5.351520991575558e-06, + "loss": 1.0384, + "step": 28196 + }, + { + "epoch": 0.66, + "grad_norm": 2.015498685482876, + "learning_rate": 5.3508454157632815e-06, + "loss": 1.2073, + "step": 28197 + }, + { + "epoch": 0.66, + "grad_norm": 1.1081091957133828, + "learning_rate": 5.350169867019853e-06, + "loss": 0.9546, + "step": 28198 + }, + { + "epoch": 0.66, + "grad_norm": 1.9028332362515004, + "learning_rate": 5.349494345349204e-06, + "loss": 1.0505, + "step": 28199 + }, + { + "epoch": 0.66, + "grad_norm": 2.0412224531327947, + "learning_rate": 5.348818850755272e-06, + "loss": 1.021, + "step": 28200 + }, + { + "epoch": 0.66, + "grad_norm": 1.928381284596736, + "learning_rate": 5.348143383241992e-06, + "loss": 0.9725, + "step": 28201 + }, + { + "epoch": 0.66, + "grad_norm": 2.1477901991072232, + "learning_rate": 5.347467942813296e-06, + "loss": 0.9822, + "step": 28202 + }, + { + "epoch": 0.66, + "grad_norm": 1.9601357391694674, + "learning_rate": 5.346792529473108e-06, + "loss": 1.0536, + "step": 28203 + }, + { + "epoch": 0.66, + "grad_norm": 1.848829530107499, + "learning_rate": 5.346117143225369e-06, + "loss": 0.9985, + "step": 28204 + }, + { + "epoch": 0.66, + "grad_norm": 1.9820236614731028, + "learning_rate": 5.3454417840740125e-06, + "loss": 0.962, + "step": 28205 + }, + { + "epoch": 0.66, + "grad_norm": 1.8721943343088259, + "learning_rate": 5.344766452022967e-06, + "loss": 1.1142, + "step": 28206 + }, + { + "epoch": 0.66, + "grad_norm": 2.0322193219405555, + "learning_rate": 5.344091147076162e-06, + "loss": 0.9913, + "step": 28207 + }, + { + "epoch": 0.66, + "grad_norm": 2.159946310622636, + "learning_rate": 5.343415869237533e-06, + "loss": 0.8517, + "step": 28208 + }, + { + "epoch": 0.66, + "grad_norm": 1.9665887051494582, + "learning_rate": 5.342740618511013e-06, + "loss": 1.0154, + "step": 28209 + }, + { + "epoch": 0.66, + "grad_norm": 1.9400434048936344, + "learning_rate": 5.342065394900533e-06, + "loss": 1.0968, + "step": 28210 + }, + { + "epoch": 0.66, + "grad_norm": 2.415048613010716, + "learning_rate": 5.3413901984100195e-06, + "loss": 0.7097, + "step": 28211 + }, + { + "epoch": 0.66, + "grad_norm": 2.0075029630191183, + "learning_rate": 5.340715029043411e-06, + "loss": 1.0202, + "step": 28212 + }, + { + "epoch": 0.66, + "grad_norm": 1.9490962050900777, + "learning_rate": 5.3400398868046285e-06, + "loss": 0.8882, + "step": 28213 + }, + { + "epoch": 0.66, + "grad_norm": 1.9735930557130272, + "learning_rate": 5.339364771697615e-06, + "loss": 1.0862, + "step": 28214 + }, + { + "epoch": 0.66, + "grad_norm": 2.022670807602795, + "learning_rate": 5.338689683726291e-06, + "loss": 1.0185, + "step": 28215 + }, + { + "epoch": 0.66, + "grad_norm": 2.1079225797156758, + "learning_rate": 5.338014622894596e-06, + "loss": 1.1494, + "step": 28216 + }, + { + "epoch": 0.66, + "grad_norm": 1.8939890810820654, + "learning_rate": 5.337339589206452e-06, + "loss": 0.9512, + "step": 28217 + }, + { + "epoch": 0.66, + "grad_norm": 2.083915500499314, + "learning_rate": 5.3366645826657956e-06, + "loss": 0.957, + "step": 28218 + }, + { + "epoch": 0.66, + "grad_norm": 1.8660554978329238, + "learning_rate": 5.3359896032765545e-06, + "loss": 1.0445, + "step": 28219 + }, + { + "epoch": 0.66, + "grad_norm": 2.046593315058183, + "learning_rate": 5.335314651042656e-06, + "loss": 1.0379, + "step": 28220 + }, + { + "epoch": 0.66, + "grad_norm": 2.1387999969800897, + "learning_rate": 5.334639725968031e-06, + "loss": 1.0593, + "step": 28221 + }, + { + "epoch": 0.66, + "grad_norm": 1.8516600550179791, + "learning_rate": 5.333964828056615e-06, + "loss": 0.8939, + "step": 28222 + }, + { + "epoch": 0.66, + "grad_norm": 2.052718713493744, + "learning_rate": 5.333289957312333e-06, + "loss": 1.004, + "step": 28223 + }, + { + "epoch": 0.66, + "grad_norm": 1.6475019612374358, + "learning_rate": 5.332615113739111e-06, + "loss": 0.9641, + "step": 28224 + }, + { + "epoch": 0.66, + "grad_norm": 2.280812360111094, + "learning_rate": 5.3319402973408805e-06, + "loss": 0.9415, + "step": 28225 + }, + { + "epoch": 0.66, + "grad_norm": 1.8235333976462358, + "learning_rate": 5.331265508121577e-06, + "loss": 0.9277, + "step": 28226 + }, + { + "epoch": 0.67, + "grad_norm": 1.978720057899204, + "learning_rate": 5.330590746085122e-06, + "loss": 0.9262, + "step": 28227 + }, + { + "epoch": 0.67, + "grad_norm": 1.8010707043871663, + "learning_rate": 5.329916011235443e-06, + "loss": 0.8735, + "step": 28228 + }, + { + "epoch": 0.67, + "grad_norm": 2.193101991925373, + "learning_rate": 5.329241303576471e-06, + "loss": 0.9723, + "step": 28229 + }, + { + "epoch": 0.67, + "grad_norm": 2.0571707907559036, + "learning_rate": 5.328566623112139e-06, + "loss": 1.0042, + "step": 28230 + }, + { + "epoch": 0.67, + "grad_norm": 1.900563065542957, + "learning_rate": 5.327891969846371e-06, + "loss": 0.8855, + "step": 28231 + }, + { + "epoch": 0.67, + "grad_norm": 2.0948999638801276, + "learning_rate": 5.327217343783091e-06, + "loss": 0.9848, + "step": 28232 + }, + { + "epoch": 0.67, + "grad_norm": 2.3222508664620682, + "learning_rate": 5.326542744926235e-06, + "loss": 0.9794, + "step": 28233 + }, + { + "epoch": 0.67, + "grad_norm": 2.0317523064710263, + "learning_rate": 5.325868173279725e-06, + "loss": 0.9183, + "step": 28234 + }, + { + "epoch": 0.67, + "grad_norm": 1.9256947813052943, + "learning_rate": 5.325193628847492e-06, + "loss": 1.0387, + "step": 28235 + }, + { + "epoch": 0.67, + "grad_norm": 2.0180110577687014, + "learning_rate": 5.324519111633457e-06, + "loss": 1.1342, + "step": 28236 + }, + { + "epoch": 0.67, + "grad_norm": 1.9375987730738726, + "learning_rate": 5.323844621641556e-06, + "loss": 1.115, + "step": 28237 + }, + { + "epoch": 0.67, + "grad_norm": 0.9953519343769607, + "learning_rate": 5.32317015887571e-06, + "loss": 0.9046, + "step": 28238 + }, + { + "epoch": 0.67, + "grad_norm": 2.165901303069182, + "learning_rate": 5.322495723339851e-06, + "loss": 0.9628, + "step": 28239 + }, + { + "epoch": 0.67, + "grad_norm": 2.0347738658397985, + "learning_rate": 5.321821315037903e-06, + "loss": 1.0343, + "step": 28240 + }, + { + "epoch": 0.67, + "grad_norm": 1.9469989958274558, + "learning_rate": 5.3211469339737885e-06, + "loss": 0.9506, + "step": 28241 + }, + { + "epoch": 0.67, + "grad_norm": 2.0998574791187727, + "learning_rate": 5.320472580151437e-06, + "loss": 1.048, + "step": 28242 + }, + { + "epoch": 0.67, + "grad_norm": 2.197440950180549, + "learning_rate": 5.31979825357478e-06, + "loss": 1.0449, + "step": 28243 + }, + { + "epoch": 0.67, + "grad_norm": 2.0983750239134245, + "learning_rate": 5.31912395424774e-06, + "loss": 0.9162, + "step": 28244 + }, + { + "epoch": 0.67, + "grad_norm": 2.1164093596791855, + "learning_rate": 5.318449682174238e-06, + "loss": 1.0022, + "step": 28245 + }, + { + "epoch": 0.67, + "grad_norm": 2.302743783212488, + "learning_rate": 5.317775437358203e-06, + "loss": 0.9869, + "step": 28246 + }, + { + "epoch": 0.67, + "grad_norm": 2.254946157846882, + "learning_rate": 5.317101219803566e-06, + "loss": 0.9127, + "step": 28247 + }, + { + "epoch": 0.67, + "grad_norm": 1.8766295431242315, + "learning_rate": 5.316427029514251e-06, + "loss": 0.9793, + "step": 28248 + }, + { + "epoch": 0.67, + "grad_norm": 2.127745313565518, + "learning_rate": 5.315752866494174e-06, + "loss": 0.8543, + "step": 28249 + }, + { + "epoch": 0.67, + "grad_norm": 2.1319237909156734, + "learning_rate": 5.315078730747268e-06, + "loss": 1.1419, + "step": 28250 + }, + { + "epoch": 0.67, + "grad_norm": 2.1930278330482396, + "learning_rate": 5.3144046222774595e-06, + "loss": 0.9374, + "step": 28251 + }, + { + "epoch": 0.67, + "grad_norm": 2.2761552220825862, + "learning_rate": 5.313730541088671e-06, + "loss": 0.9787, + "step": 28252 + }, + { + "epoch": 0.67, + "grad_norm": 2.526654813696179, + "learning_rate": 5.313056487184824e-06, + "loss": 0.9741, + "step": 28253 + }, + { + "epoch": 0.67, + "grad_norm": 2.2201351790023427, + "learning_rate": 5.3123824605698444e-06, + "loss": 1.0604, + "step": 28254 + }, + { + "epoch": 0.67, + "grad_norm": 1.829605177111851, + "learning_rate": 5.311708461247664e-06, + "loss": 0.8948, + "step": 28255 + }, + { + "epoch": 0.67, + "grad_norm": 2.3599302359131475, + "learning_rate": 5.3110344892222e-06, + "loss": 1.1163, + "step": 28256 + }, + { + "epoch": 0.67, + "grad_norm": 2.0414266077839365, + "learning_rate": 5.3103605444973735e-06, + "loss": 1.0774, + "step": 28257 + }, + { + "epoch": 0.67, + "grad_norm": 1.915415677460578, + "learning_rate": 5.309686627077117e-06, + "loss": 1.0014, + "step": 28258 + }, + { + "epoch": 0.67, + "grad_norm": 2.040604973051428, + "learning_rate": 5.3090127369653465e-06, + "loss": 1.0152, + "step": 28259 + }, + { + "epoch": 0.67, + "grad_norm": 1.9122056490548944, + "learning_rate": 5.308338874165993e-06, + "loss": 1.1004, + "step": 28260 + }, + { + "epoch": 0.67, + "grad_norm": 1.1042379445860637, + "learning_rate": 5.307665038682972e-06, + "loss": 0.9017, + "step": 28261 + }, + { + "epoch": 0.67, + "grad_norm": 2.11218982349836, + "learning_rate": 5.306991230520213e-06, + "loss": 1.0151, + "step": 28262 + }, + { + "epoch": 0.67, + "grad_norm": 2.004331209295333, + "learning_rate": 5.306317449681635e-06, + "loss": 1.163, + "step": 28263 + }, + { + "epoch": 0.67, + "grad_norm": 2.2804010116619366, + "learning_rate": 5.305643696171161e-06, + "loss": 0.9398, + "step": 28264 + }, + { + "epoch": 0.67, + "grad_norm": 1.1313514421290183, + "learning_rate": 5.304969969992725e-06, + "loss": 0.919, + "step": 28265 + }, + { + "epoch": 0.67, + "grad_norm": 2.164906924138024, + "learning_rate": 5.304296271150231e-06, + "loss": 0.9223, + "step": 28266 + }, + { + "epoch": 0.67, + "grad_norm": 1.9756627409672045, + "learning_rate": 5.303622599647613e-06, + "loss": 1.0686, + "step": 28267 + }, + { + "epoch": 0.67, + "grad_norm": 2.3554582405310573, + "learning_rate": 5.302948955488789e-06, + "loss": 0.9737, + "step": 28268 + }, + { + "epoch": 0.67, + "grad_norm": 1.9773481515873468, + "learning_rate": 5.302275338677692e-06, + "loss": 1.0183, + "step": 28269 + }, + { + "epoch": 0.67, + "grad_norm": 2.159355655238827, + "learning_rate": 5.301601749218227e-06, + "loss": 1.0707, + "step": 28270 + }, + { + "epoch": 0.67, + "grad_norm": 2.2259996652303804, + "learning_rate": 5.300928187114323e-06, + "loss": 0.9992, + "step": 28271 + }, + { + "epoch": 0.67, + "grad_norm": 2.1159372214308827, + "learning_rate": 5.30025465236991e-06, + "loss": 0.9952, + "step": 28272 + }, + { + "epoch": 0.67, + "grad_norm": 1.9795314476051251, + "learning_rate": 5.299581144988899e-06, + "loss": 0.9258, + "step": 28273 + }, + { + "epoch": 0.67, + "grad_norm": 1.8581307627353265, + "learning_rate": 5.298907664975213e-06, + "loss": 1.0001, + "step": 28274 + }, + { + "epoch": 0.67, + "grad_norm": 1.8460803430455242, + "learning_rate": 5.298234212332773e-06, + "loss": 1.0622, + "step": 28275 + }, + { + "epoch": 0.67, + "grad_norm": 2.0095634237090345, + "learning_rate": 5.297560787065506e-06, + "loss": 0.849, + "step": 28276 + }, + { + "epoch": 0.67, + "grad_norm": 1.990357693306521, + "learning_rate": 5.296887389177326e-06, + "loss": 1.0046, + "step": 28277 + }, + { + "epoch": 0.67, + "grad_norm": 1.8977327355828495, + "learning_rate": 5.296214018672161e-06, + "loss": 1.048, + "step": 28278 + }, + { + "epoch": 0.67, + "grad_norm": 1.7891733501431526, + "learning_rate": 5.295540675553926e-06, + "loss": 0.8933, + "step": 28279 + }, + { + "epoch": 0.67, + "grad_norm": 2.044973870709856, + "learning_rate": 5.294867359826539e-06, + "loss": 1.1414, + "step": 28280 + }, + { + "epoch": 0.67, + "grad_norm": 2.589545099485456, + "learning_rate": 5.294194071493925e-06, + "loss": 0.945, + "step": 28281 + }, + { + "epoch": 0.67, + "grad_norm": 1.9811587577693808, + "learning_rate": 5.293520810560006e-06, + "loss": 1.0264, + "step": 28282 + }, + { + "epoch": 0.67, + "grad_norm": 2.0047507657177572, + "learning_rate": 5.292847577028698e-06, + "loss": 1.0446, + "step": 28283 + }, + { + "epoch": 0.67, + "grad_norm": 2.079385083387633, + "learning_rate": 5.292174370903919e-06, + "loss": 1.04, + "step": 28284 + }, + { + "epoch": 0.67, + "grad_norm": 1.9799228183570763, + "learning_rate": 5.29150119218959e-06, + "loss": 1.0187, + "step": 28285 + }, + { + "epoch": 0.67, + "grad_norm": 1.7706989602280419, + "learning_rate": 5.290828040889637e-06, + "loss": 0.9199, + "step": 28286 + }, + { + "epoch": 0.67, + "grad_norm": 4.159067098218179, + "learning_rate": 5.2901549170079736e-06, + "loss": 0.8883, + "step": 28287 + }, + { + "epoch": 0.67, + "grad_norm": 1.8300517499136586, + "learning_rate": 5.2894818205485144e-06, + "loss": 1.0095, + "step": 28288 + }, + { + "epoch": 0.67, + "grad_norm": 2.255590027824064, + "learning_rate": 5.288808751515184e-06, + "loss": 1.0333, + "step": 28289 + }, + { + "epoch": 0.67, + "grad_norm": 2.005809411814428, + "learning_rate": 5.288135709911906e-06, + "loss": 0.9937, + "step": 28290 + }, + { + "epoch": 0.67, + "grad_norm": 1.8930073923963906, + "learning_rate": 5.287462695742593e-06, + "loss": 0.8861, + "step": 28291 + }, + { + "epoch": 0.67, + "grad_norm": 1.8744565796524153, + "learning_rate": 5.2867897090111594e-06, + "loss": 0.9367, + "step": 28292 + }, + { + "epoch": 0.67, + "grad_norm": 2.4797975534574546, + "learning_rate": 5.286116749721533e-06, + "loss": 0.9625, + "step": 28293 + }, + { + "epoch": 0.67, + "grad_norm": 1.924046651173332, + "learning_rate": 5.285443817877622e-06, + "loss": 0.9056, + "step": 28294 + }, + { + "epoch": 0.67, + "grad_norm": 2.206366124614244, + "learning_rate": 5.284770913483355e-06, + "loss": 1.1013, + "step": 28295 + }, + { + "epoch": 0.67, + "grad_norm": 2.103082419658047, + "learning_rate": 5.28409803654264e-06, + "loss": 0.9623, + "step": 28296 + }, + { + "epoch": 0.67, + "grad_norm": 1.9092442938783207, + "learning_rate": 5.283425187059403e-06, + "loss": 0.9616, + "step": 28297 + }, + { + "epoch": 0.67, + "grad_norm": 1.7488690953283592, + "learning_rate": 5.282752365037554e-06, + "loss": 0.9673, + "step": 28298 + }, + { + "epoch": 0.67, + "grad_norm": 2.5291692683164237, + "learning_rate": 5.282079570481018e-06, + "loss": 1.1324, + "step": 28299 + }, + { + "epoch": 0.67, + "grad_norm": 2.4011398253754006, + "learning_rate": 5.2814068033937096e-06, + "loss": 1.0943, + "step": 28300 + }, + { + "epoch": 0.67, + "grad_norm": 1.9806359992142428, + "learning_rate": 5.28073406377954e-06, + "loss": 0.8365, + "step": 28301 + }, + { + "epoch": 0.67, + "grad_norm": 1.9261818558232817, + "learning_rate": 5.280061351642429e-06, + "loss": 0.9753, + "step": 28302 + }, + { + "epoch": 0.67, + "grad_norm": 2.022059459937188, + "learning_rate": 5.279388666986301e-06, + "loss": 1.0787, + "step": 28303 + }, + { + "epoch": 0.67, + "grad_norm": 2.12945561145355, + "learning_rate": 5.278716009815067e-06, + "loss": 0.9405, + "step": 28304 + }, + { + "epoch": 0.67, + "grad_norm": 2.7158221287157174, + "learning_rate": 5.278043380132638e-06, + "loss": 0.9048, + "step": 28305 + }, + { + "epoch": 0.67, + "grad_norm": 2.226332101906028, + "learning_rate": 5.277370777942936e-06, + "loss": 0.857, + "step": 28306 + }, + { + "epoch": 0.67, + "grad_norm": 1.6747834774264674, + "learning_rate": 5.276698203249882e-06, + "loss": 0.9068, + "step": 28307 + }, + { + "epoch": 0.67, + "grad_norm": 1.8441287001337576, + "learning_rate": 5.276025656057384e-06, + "loss": 0.884, + "step": 28308 + }, + { + "epoch": 0.67, + "grad_norm": 1.8908854206189665, + "learning_rate": 5.275353136369357e-06, + "loss": 0.9056, + "step": 28309 + }, + { + "epoch": 0.67, + "grad_norm": 1.9194363147888072, + "learning_rate": 5.27468064418972e-06, + "loss": 0.9145, + "step": 28310 + }, + { + "epoch": 0.67, + "grad_norm": 1.903098020373117, + "learning_rate": 5.274008179522392e-06, + "loss": 0.9691, + "step": 28311 + }, + { + "epoch": 0.67, + "grad_norm": 2.1603047440482035, + "learning_rate": 5.273335742371284e-06, + "loss": 1.0605, + "step": 28312 + }, + { + "epoch": 0.67, + "grad_norm": 2.1523654466303945, + "learning_rate": 5.272663332740308e-06, + "loss": 1.0113, + "step": 28313 + }, + { + "epoch": 0.67, + "grad_norm": 1.0947958718409505, + "learning_rate": 5.271990950633388e-06, + "loss": 0.8207, + "step": 28314 + }, + { + "epoch": 0.67, + "grad_norm": 3.4421333794147615, + "learning_rate": 5.271318596054429e-06, + "loss": 0.9782, + "step": 28315 + }, + { + "epoch": 0.67, + "grad_norm": 2.343928243942311, + "learning_rate": 5.270646269007355e-06, + "loss": 1.0161, + "step": 28316 + }, + { + "epoch": 0.67, + "grad_norm": 1.034613912704123, + "learning_rate": 5.269973969496069e-06, + "loss": 0.9017, + "step": 28317 + }, + { + "epoch": 0.67, + "grad_norm": 1.8258530117174128, + "learning_rate": 5.269301697524498e-06, + "loss": 0.9164, + "step": 28318 + }, + { + "epoch": 0.67, + "grad_norm": 2.0294277574161432, + "learning_rate": 5.268629453096547e-06, + "loss": 0.9644, + "step": 28319 + }, + { + "epoch": 0.67, + "grad_norm": 2.1225711550317863, + "learning_rate": 5.267957236216137e-06, + "loss": 1.0903, + "step": 28320 + }, + { + "epoch": 0.67, + "grad_norm": 2.0807792436759853, + "learning_rate": 5.267285046887176e-06, + "loss": 0.9264, + "step": 28321 + }, + { + "epoch": 0.67, + "grad_norm": 1.9894980912610523, + "learning_rate": 5.2666128851135775e-06, + "loss": 0.9637, + "step": 28322 + }, + { + "epoch": 0.67, + "grad_norm": 2.033903360255239, + "learning_rate": 5.265940750899256e-06, + "loss": 1.0325, + "step": 28323 + }, + { + "epoch": 0.67, + "grad_norm": 2.0940375795558768, + "learning_rate": 5.265268644248133e-06, + "loss": 0.9193, + "step": 28324 + }, + { + "epoch": 0.67, + "grad_norm": 2.3342429580457136, + "learning_rate": 5.264596565164113e-06, + "loss": 0.8713, + "step": 28325 + }, + { + "epoch": 0.67, + "grad_norm": 1.833610993428371, + "learning_rate": 5.263924513651107e-06, + "loss": 0.8661, + "step": 28326 + }, + { + "epoch": 0.67, + "grad_norm": 1.9582123297618255, + "learning_rate": 5.263252489713032e-06, + "loss": 0.9608, + "step": 28327 + }, + { + "epoch": 0.67, + "grad_norm": 2.091842288571473, + "learning_rate": 5.262580493353805e-06, + "loss": 1.1954, + "step": 28328 + }, + { + "epoch": 0.67, + "grad_norm": 2.5898568903185484, + "learning_rate": 5.261908524577334e-06, + "loss": 0.9934, + "step": 28329 + }, + { + "epoch": 0.67, + "grad_norm": 1.1160172011067546, + "learning_rate": 5.261236583387528e-06, + "loss": 0.9688, + "step": 28330 + }, + { + "epoch": 0.67, + "grad_norm": 2.0823139138358773, + "learning_rate": 5.260564669788302e-06, + "loss": 1.0529, + "step": 28331 + }, + { + "epoch": 0.67, + "grad_norm": 1.9507252399883086, + "learning_rate": 5.259892783783573e-06, + "loss": 1.0417, + "step": 28332 + }, + { + "epoch": 0.67, + "grad_norm": 1.8670564701779433, + "learning_rate": 5.259220925377249e-06, + "loss": 1.0075, + "step": 28333 + }, + { + "epoch": 0.67, + "grad_norm": 2.060709804041484, + "learning_rate": 5.258549094573239e-06, + "loss": 0.9784, + "step": 28334 + }, + { + "epoch": 0.67, + "grad_norm": 2.000250583998216, + "learning_rate": 5.257877291375459e-06, + "loss": 0.9304, + "step": 28335 + }, + { + "epoch": 0.67, + "grad_norm": 1.8495635516780435, + "learning_rate": 5.257205515787817e-06, + "loss": 0.8729, + "step": 28336 + }, + { + "epoch": 0.67, + "grad_norm": 2.0002921768085966, + "learning_rate": 5.256533767814229e-06, + "loss": 0.9411, + "step": 28337 + }, + { + "epoch": 0.67, + "grad_norm": 2.21506950196721, + "learning_rate": 5.2558620474586e-06, + "loss": 0.9822, + "step": 28338 + }, + { + "epoch": 0.67, + "grad_norm": 1.9133835479996602, + "learning_rate": 5.255190354724846e-06, + "loss": 0.9428, + "step": 28339 + }, + { + "epoch": 0.67, + "grad_norm": 2.208356187425895, + "learning_rate": 5.254518689616874e-06, + "loss": 1.0481, + "step": 28340 + }, + { + "epoch": 0.67, + "grad_norm": 1.9675482190362183, + "learning_rate": 5.2538470521385996e-06, + "loss": 1.0306, + "step": 28341 + }, + { + "epoch": 0.67, + "grad_norm": 2.098419286428019, + "learning_rate": 5.253175442293929e-06, + "loss": 1.0807, + "step": 28342 + }, + { + "epoch": 0.67, + "grad_norm": 2.7470787062111586, + "learning_rate": 5.252503860086773e-06, + "loss": 0.9451, + "step": 28343 + }, + { + "epoch": 0.67, + "grad_norm": 1.933604705061849, + "learning_rate": 5.2518323055210405e-06, + "loss": 1.0187, + "step": 28344 + }, + { + "epoch": 0.67, + "grad_norm": 1.8708267882562015, + "learning_rate": 5.251160778600648e-06, + "loss": 0.9564, + "step": 28345 + }, + { + "epoch": 0.67, + "grad_norm": 1.9709909421371206, + "learning_rate": 5.250489279329501e-06, + "loss": 0.936, + "step": 28346 + }, + { + "epoch": 0.67, + "grad_norm": 1.9932640303867992, + "learning_rate": 5.249817807711505e-06, + "loss": 1.0357, + "step": 28347 + }, + { + "epoch": 0.67, + "grad_norm": 2.0862961912318165, + "learning_rate": 5.249146363750574e-06, + "loss": 1.0719, + "step": 28348 + }, + { + "epoch": 0.67, + "grad_norm": 2.409679682876112, + "learning_rate": 5.24847494745062e-06, + "loss": 0.9716, + "step": 28349 + }, + { + "epoch": 0.67, + "grad_norm": 2.2469761062778617, + "learning_rate": 5.24780355881555e-06, + "loss": 0.991, + "step": 28350 + }, + { + "epoch": 0.67, + "grad_norm": 1.96955652813926, + "learning_rate": 5.247132197849267e-06, + "loss": 0.9493, + "step": 28351 + }, + { + "epoch": 0.67, + "grad_norm": 1.8597261826972467, + "learning_rate": 5.246460864555687e-06, + "loss": 1.0271, + "step": 28352 + }, + { + "epoch": 0.67, + "grad_norm": 1.9846417996753947, + "learning_rate": 5.245789558938719e-06, + "loss": 1.0012, + "step": 28353 + }, + { + "epoch": 0.67, + "grad_norm": 2.153871041810675, + "learning_rate": 5.245118281002271e-06, + "loss": 0.9188, + "step": 28354 + }, + { + "epoch": 0.67, + "grad_norm": 1.890199468665414, + "learning_rate": 5.244447030750244e-06, + "loss": 0.9556, + "step": 28355 + }, + { + "epoch": 0.67, + "grad_norm": 2.010180793518598, + "learning_rate": 5.243775808186557e-06, + "loss": 0.9946, + "step": 28356 + }, + { + "epoch": 0.67, + "grad_norm": 1.8435695321705614, + "learning_rate": 5.2431046133151085e-06, + "loss": 0.779, + "step": 28357 + }, + { + "epoch": 0.67, + "grad_norm": 2.0070257069320725, + "learning_rate": 5.242433446139816e-06, + "loss": 1.0339, + "step": 28358 + }, + { + "epoch": 0.67, + "grad_norm": 1.9255304395681785, + "learning_rate": 5.241762306664578e-06, + "loss": 1.0513, + "step": 28359 + }, + { + "epoch": 0.67, + "grad_norm": 2.686428565076894, + "learning_rate": 5.241091194893309e-06, + "loss": 1.0529, + "step": 28360 + }, + { + "epoch": 0.67, + "grad_norm": 1.8983364053778755, + "learning_rate": 5.24042011082991e-06, + "loss": 1.0345, + "step": 28361 + }, + { + "epoch": 0.67, + "grad_norm": 2.1833520237916675, + "learning_rate": 5.2397490544782935e-06, + "loss": 0.9562, + "step": 28362 + }, + { + "epoch": 0.67, + "grad_norm": 1.9230480366158411, + "learning_rate": 5.2390780258423715e-06, + "loss": 1.0255, + "step": 28363 + }, + { + "epoch": 0.67, + "grad_norm": 2.5177240818624522, + "learning_rate": 5.238407024926038e-06, + "loss": 0.9469, + "step": 28364 + }, + { + "epoch": 0.67, + "grad_norm": 2.1138415525362007, + "learning_rate": 5.237736051733205e-06, + "loss": 1.0189, + "step": 28365 + }, + { + "epoch": 0.67, + "grad_norm": 1.8845565288930948, + "learning_rate": 5.23706510626778e-06, + "loss": 0.9274, + "step": 28366 + }, + { + "epoch": 0.67, + "grad_norm": 1.9978565174794283, + "learning_rate": 5.236394188533679e-06, + "loss": 0.974, + "step": 28367 + }, + { + "epoch": 0.67, + "grad_norm": 1.9256315815152503, + "learning_rate": 5.2357232985347905e-06, + "loss": 0.994, + "step": 28368 + }, + { + "epoch": 0.67, + "grad_norm": 1.9258748037462752, + "learning_rate": 5.23505243627503e-06, + "loss": 1.0404, + "step": 28369 + }, + { + "epoch": 0.67, + "grad_norm": 1.0702897116400012, + "learning_rate": 5.234381601758306e-06, + "loss": 0.8869, + "step": 28370 + }, + { + "epoch": 0.67, + "grad_norm": 1.0866654922405354, + "learning_rate": 5.233710794988518e-06, + "loss": 1.0053, + "step": 28371 + }, + { + "epoch": 0.67, + "grad_norm": 2.1216022176627236, + "learning_rate": 5.233040015969577e-06, + "loss": 1.063, + "step": 28372 + }, + { + "epoch": 0.67, + "grad_norm": 2.022896947894389, + "learning_rate": 5.232369264705384e-06, + "loss": 0.8895, + "step": 28373 + }, + { + "epoch": 0.67, + "grad_norm": 1.0977217028573774, + "learning_rate": 5.23169854119985e-06, + "loss": 0.9129, + "step": 28374 + }, + { + "epoch": 0.67, + "grad_norm": 1.9441992561019803, + "learning_rate": 5.231027845456873e-06, + "loss": 0.9815, + "step": 28375 + }, + { + "epoch": 0.67, + "grad_norm": 2.2702985034366456, + "learning_rate": 5.230357177480365e-06, + "loss": 0.9097, + "step": 28376 + }, + { + "epoch": 0.67, + "grad_norm": 1.0791145938845934, + "learning_rate": 5.229686537274223e-06, + "loss": 0.8392, + "step": 28377 + }, + { + "epoch": 0.67, + "grad_norm": 2.210451252390552, + "learning_rate": 5.229015924842362e-06, + "loss": 1.0073, + "step": 28378 + }, + { + "epoch": 0.67, + "grad_norm": 1.9672793996371163, + "learning_rate": 5.228345340188677e-06, + "loss": 1.0156, + "step": 28379 + }, + { + "epoch": 0.67, + "grad_norm": 2.0054529493571, + "learning_rate": 5.227674783317078e-06, + "loss": 0.9808, + "step": 28380 + }, + { + "epoch": 0.67, + "grad_norm": 1.8562997334995315, + "learning_rate": 5.227004254231468e-06, + "loss": 1.038, + "step": 28381 + }, + { + "epoch": 0.67, + "grad_norm": 1.925093068429096, + "learning_rate": 5.226333752935747e-06, + "loss": 0.9449, + "step": 28382 + }, + { + "epoch": 0.67, + "grad_norm": 2.220256765593577, + "learning_rate": 5.2256632794338216e-06, + "loss": 0.9914, + "step": 28383 + }, + { + "epoch": 0.67, + "grad_norm": 2.0070987707969854, + "learning_rate": 5.224992833729601e-06, + "loss": 0.8353, + "step": 28384 + }, + { + "epoch": 0.67, + "grad_norm": 1.8092460757434365, + "learning_rate": 5.2243224158269835e-06, + "loss": 1.0274, + "step": 28385 + }, + { + "epoch": 0.67, + "grad_norm": 2.078535949302616, + "learning_rate": 5.223652025729868e-06, + "loss": 1.0338, + "step": 28386 + }, + { + "epoch": 0.67, + "grad_norm": 1.9993369927411073, + "learning_rate": 5.222981663442164e-06, + "loss": 0.9976, + "step": 28387 + }, + { + "epoch": 0.67, + "grad_norm": 2.1217166952276614, + "learning_rate": 5.222311328967775e-06, + "loss": 1.0196, + "step": 28388 + }, + { + "epoch": 0.67, + "grad_norm": 2.0181610960658474, + "learning_rate": 5.221641022310604e-06, + "loss": 1.0846, + "step": 28389 + }, + { + "epoch": 0.67, + "grad_norm": 1.8416601155033383, + "learning_rate": 5.220970743474547e-06, + "loss": 1.09, + "step": 28390 + }, + { + "epoch": 0.67, + "grad_norm": 1.8872187806537257, + "learning_rate": 5.220300492463511e-06, + "loss": 0.8979, + "step": 28391 + }, + { + "epoch": 0.67, + "grad_norm": 2.100621687679317, + "learning_rate": 5.219630269281404e-06, + "loss": 1.0583, + "step": 28392 + }, + { + "epoch": 0.67, + "grad_norm": 2.192146006944805, + "learning_rate": 5.218960073932122e-06, + "loss": 1.0271, + "step": 28393 + }, + { + "epoch": 0.67, + "grad_norm": 2.0201958146468897, + "learning_rate": 5.218289906419564e-06, + "loss": 0.9338, + "step": 28394 + }, + { + "epoch": 0.67, + "grad_norm": 0.9960354956253743, + "learning_rate": 5.217619766747641e-06, + "loss": 0.9333, + "step": 28395 + }, + { + "epoch": 0.67, + "grad_norm": 2.0999447963284394, + "learning_rate": 5.2169496549202444e-06, + "loss": 0.9414, + "step": 28396 + }, + { + "epoch": 0.67, + "grad_norm": 1.8133864305265026, + "learning_rate": 5.216279570941287e-06, + "loss": 0.8993, + "step": 28397 + }, + { + "epoch": 0.67, + "grad_norm": 2.2647684505726557, + "learning_rate": 5.2156095148146594e-06, + "loss": 0.8552, + "step": 28398 + }, + { + "epoch": 0.67, + "grad_norm": 1.8357887500123238, + "learning_rate": 5.214939486544271e-06, + "loss": 0.9429, + "step": 28399 + }, + { + "epoch": 0.67, + "grad_norm": 2.0518598839790982, + "learning_rate": 5.214269486134017e-06, + "loss": 1.0423, + "step": 28400 + }, + { + "epoch": 0.67, + "grad_norm": 1.8266177048857766, + "learning_rate": 5.213599513587805e-06, + "loss": 1.0124, + "step": 28401 + }, + { + "epoch": 0.67, + "grad_norm": 1.0932294112241085, + "learning_rate": 5.212929568909532e-06, + "loss": 1.0225, + "step": 28402 + }, + { + "epoch": 0.67, + "grad_norm": 2.1404159635917854, + "learning_rate": 5.212259652103095e-06, + "loss": 1.045, + "step": 28403 + }, + { + "epoch": 0.67, + "grad_norm": 3.615936787914556, + "learning_rate": 5.211589763172398e-06, + "loss": 0.8595, + "step": 28404 + }, + { + "epoch": 0.67, + "grad_norm": 2.0729112867539325, + "learning_rate": 5.210919902121344e-06, + "loss": 0.9803, + "step": 28405 + }, + { + "epoch": 0.67, + "grad_norm": 2.1964553719791415, + "learning_rate": 5.2102500689538316e-06, + "loss": 1.0296, + "step": 28406 + }, + { + "epoch": 0.67, + "grad_norm": 2.091554603651522, + "learning_rate": 5.2095802636737556e-06, + "loss": 1.0464, + "step": 28407 + }, + { + "epoch": 0.67, + "grad_norm": 3.4870342137998933, + "learning_rate": 5.20891048628502e-06, + "loss": 0.9397, + "step": 28408 + }, + { + "epoch": 0.67, + "grad_norm": 1.9659022691027892, + "learning_rate": 5.208240736791528e-06, + "loss": 1.018, + "step": 28409 + }, + { + "epoch": 0.67, + "grad_norm": 2.195384213769116, + "learning_rate": 5.207571015197177e-06, + "loss": 0.9832, + "step": 28410 + }, + { + "epoch": 0.67, + "grad_norm": 2.692580263535635, + "learning_rate": 5.206901321505859e-06, + "loss": 0.9187, + "step": 28411 + }, + { + "epoch": 0.67, + "grad_norm": 1.9243615485633696, + "learning_rate": 5.2062316557214785e-06, + "loss": 0.9096, + "step": 28412 + }, + { + "epoch": 0.67, + "grad_norm": 1.8631775166618219, + "learning_rate": 5.2055620178479405e-06, + "loss": 0.9891, + "step": 28413 + }, + { + "epoch": 0.67, + "grad_norm": 2.1051002130802496, + "learning_rate": 5.204892407889138e-06, + "loss": 1.0073, + "step": 28414 + }, + { + "epoch": 0.67, + "grad_norm": 2.1626343015587866, + "learning_rate": 5.204222825848966e-06, + "loss": 1.1905, + "step": 28415 + }, + { + "epoch": 0.67, + "grad_norm": 1.9642613516649838, + "learning_rate": 5.203553271731332e-06, + "loss": 1.0332, + "step": 28416 + }, + { + "epoch": 0.67, + "grad_norm": 1.897224599046364, + "learning_rate": 5.202883745540124e-06, + "loss": 0.9161, + "step": 28417 + }, + { + "epoch": 0.67, + "grad_norm": 1.8069313043580786, + "learning_rate": 5.2022142472792494e-06, + "loss": 1.0342, + "step": 28418 + }, + { + "epoch": 0.67, + "grad_norm": 1.9049330291296132, + "learning_rate": 5.2015447769525985e-06, + "loss": 0.9557, + "step": 28419 + }, + { + "epoch": 0.67, + "grad_norm": 1.0700881130667677, + "learning_rate": 5.2008753345640775e-06, + "loss": 0.9462, + "step": 28420 + }, + { + "epoch": 0.67, + "grad_norm": 1.8574563212765645, + "learning_rate": 5.200205920117576e-06, + "loss": 0.9807, + "step": 28421 + }, + { + "epoch": 0.67, + "grad_norm": 1.9037039933009998, + "learning_rate": 5.199536533616999e-06, + "loss": 0.9761, + "step": 28422 + }, + { + "epoch": 0.67, + "grad_norm": 2.066203456258964, + "learning_rate": 5.19886717506624e-06, + "loss": 0.8134, + "step": 28423 + }, + { + "epoch": 0.67, + "grad_norm": 1.9722414200202796, + "learning_rate": 5.198197844469193e-06, + "loss": 0.94, + "step": 28424 + }, + { + "epoch": 0.67, + "grad_norm": 1.8712885672830701, + "learning_rate": 5.197528541829758e-06, + "loss": 0.9414, + "step": 28425 + }, + { + "epoch": 0.67, + "grad_norm": 2.037950705582583, + "learning_rate": 5.196859267151835e-06, + "loss": 0.9882, + "step": 28426 + }, + { + "epoch": 0.67, + "grad_norm": 1.9173689243452499, + "learning_rate": 5.196190020439319e-06, + "loss": 0.8926, + "step": 28427 + }, + { + "epoch": 0.67, + "grad_norm": 1.9856198665968965, + "learning_rate": 5.195520801696101e-06, + "loss": 0.9896, + "step": 28428 + }, + { + "epoch": 0.67, + "grad_norm": 2.003339090916633, + "learning_rate": 5.194851610926083e-06, + "loss": 0.9643, + "step": 28429 + }, + { + "epoch": 0.67, + "grad_norm": 1.1126238707848715, + "learning_rate": 5.194182448133163e-06, + "loss": 0.9364, + "step": 28430 + }, + { + "epoch": 0.67, + "grad_norm": 1.8766930455895312, + "learning_rate": 5.1935133133212345e-06, + "loss": 0.9305, + "step": 28431 + }, + { + "epoch": 0.67, + "grad_norm": 1.8227763574394245, + "learning_rate": 5.19284420649419e-06, + "loss": 0.8735, + "step": 28432 + }, + { + "epoch": 0.67, + "grad_norm": 2.0102840581272248, + "learning_rate": 5.192175127655929e-06, + "loss": 1.0426, + "step": 28433 + }, + { + "epoch": 0.67, + "grad_norm": 1.856051569680838, + "learning_rate": 5.191506076810349e-06, + "loss": 0.8719, + "step": 28434 + }, + { + "epoch": 0.67, + "grad_norm": 3.134607724494111, + "learning_rate": 5.190837053961344e-06, + "loss": 0.9463, + "step": 28435 + }, + { + "epoch": 0.67, + "grad_norm": 1.955153077862906, + "learning_rate": 5.190168059112804e-06, + "loss": 0.9402, + "step": 28436 + }, + { + "epoch": 0.67, + "grad_norm": 2.070404991409577, + "learning_rate": 5.189499092268632e-06, + "loss": 1.0635, + "step": 28437 + }, + { + "epoch": 0.67, + "grad_norm": 1.9283239176110933, + "learning_rate": 5.188830153432716e-06, + "loss": 0.9899, + "step": 28438 + }, + { + "epoch": 0.67, + "grad_norm": 1.7967001709421995, + "learning_rate": 5.188161242608957e-06, + "loss": 1.0361, + "step": 28439 + }, + { + "epoch": 0.67, + "grad_norm": 1.1329438281392594, + "learning_rate": 5.187492359801243e-06, + "loss": 0.9834, + "step": 28440 + }, + { + "epoch": 0.67, + "grad_norm": 2.1793494130053155, + "learning_rate": 5.186823505013476e-06, + "loss": 0.9751, + "step": 28441 + }, + { + "epoch": 0.67, + "grad_norm": 1.7502253902286038, + "learning_rate": 5.186154678249542e-06, + "loss": 0.926, + "step": 28442 + }, + { + "epoch": 0.67, + "grad_norm": 1.9981649162106905, + "learning_rate": 5.185485879513344e-06, + "loss": 0.9254, + "step": 28443 + }, + { + "epoch": 0.67, + "grad_norm": 2.2389375021476305, + "learning_rate": 5.184817108808773e-06, + "loss": 1.0173, + "step": 28444 + }, + { + "epoch": 0.67, + "grad_norm": 1.8708879515930181, + "learning_rate": 5.184148366139715e-06, + "loss": 0.9059, + "step": 28445 + }, + { + "epoch": 0.67, + "grad_norm": 1.152703235812224, + "learning_rate": 5.18347965151007e-06, + "loss": 0.9325, + "step": 28446 + }, + { + "epoch": 0.67, + "grad_norm": 1.7957844493824897, + "learning_rate": 5.182810964923733e-06, + "loss": 0.9139, + "step": 28447 + }, + { + "epoch": 0.67, + "grad_norm": 2.002771719720535, + "learning_rate": 5.182142306384602e-06, + "loss": 1.0317, + "step": 28448 + }, + { + "epoch": 0.67, + "grad_norm": 1.9204290072895347, + "learning_rate": 5.181473675896557e-06, + "loss": 1.0879, + "step": 28449 + }, + { + "epoch": 0.67, + "grad_norm": 2.233250102545255, + "learning_rate": 5.180805073463497e-06, + "loss": 1.0934, + "step": 28450 + }, + { + "epoch": 0.67, + "grad_norm": 1.8508461401901237, + "learning_rate": 5.18013649908932e-06, + "loss": 0.9832, + "step": 28451 + }, + { + "epoch": 0.67, + "grad_norm": 2.941111114786906, + "learning_rate": 5.179467952777915e-06, + "loss": 1.1, + "step": 28452 + }, + { + "epoch": 0.67, + "grad_norm": 1.869453558814954, + "learning_rate": 5.178799434533168e-06, + "loss": 0.848, + "step": 28453 + }, + { + "epoch": 0.67, + "grad_norm": 1.8961976548005635, + "learning_rate": 5.178130944358978e-06, + "loss": 1.0633, + "step": 28454 + }, + { + "epoch": 0.67, + "grad_norm": 1.7759743954041072, + "learning_rate": 5.17746248225924e-06, + "loss": 0.9978, + "step": 28455 + }, + { + "epoch": 0.67, + "grad_norm": 1.833887124745855, + "learning_rate": 5.176794048237842e-06, + "loss": 0.9905, + "step": 28456 + }, + { + "epoch": 0.67, + "grad_norm": 1.0548094009003726, + "learning_rate": 5.176125642298673e-06, + "loss": 0.8887, + "step": 28457 + }, + { + "epoch": 0.67, + "grad_norm": 1.8579615632212452, + "learning_rate": 5.1754572644456315e-06, + "loss": 0.9809, + "step": 28458 + }, + { + "epoch": 0.67, + "grad_norm": 1.7955136369814888, + "learning_rate": 5.1747889146826e-06, + "loss": 0.8889, + "step": 28459 + }, + { + "epoch": 0.67, + "grad_norm": 2.3112260396907462, + "learning_rate": 5.174120593013476e-06, + "loss": 0.9858, + "step": 28460 + }, + { + "epoch": 0.67, + "grad_norm": 4.040423733486312, + "learning_rate": 5.173452299442154e-06, + "loss": 0.9741, + "step": 28461 + }, + { + "epoch": 0.67, + "grad_norm": 1.8271271008877312, + "learning_rate": 5.17278403397252e-06, + "loss": 0.9892, + "step": 28462 + }, + { + "epoch": 0.67, + "grad_norm": 2.1421320565201043, + "learning_rate": 5.172115796608462e-06, + "loss": 1.0716, + "step": 28463 + }, + { + "epoch": 0.67, + "grad_norm": 1.9233039177475848, + "learning_rate": 5.171447587353875e-06, + "loss": 0.9729, + "step": 28464 + }, + { + "epoch": 0.67, + "grad_norm": 2.227182101321699, + "learning_rate": 5.170779406212656e-06, + "loss": 0.9574, + "step": 28465 + }, + { + "epoch": 0.67, + "grad_norm": 1.901256551080554, + "learning_rate": 5.170111253188681e-06, + "loss": 0.9649, + "step": 28466 + }, + { + "epoch": 0.67, + "grad_norm": 2.5115270084437595, + "learning_rate": 5.169443128285846e-06, + "loss": 0.903, + "step": 28467 + }, + { + "epoch": 0.67, + "grad_norm": 2.405609939027054, + "learning_rate": 5.1687750315080446e-06, + "loss": 0.9908, + "step": 28468 + }, + { + "epoch": 0.67, + "grad_norm": 1.8728042036122117, + "learning_rate": 5.168106962859166e-06, + "loss": 0.921, + "step": 28469 + }, + { + "epoch": 0.67, + "grad_norm": 1.1239471457298091, + "learning_rate": 5.1674389223431e-06, + "loss": 0.9435, + "step": 28470 + }, + { + "epoch": 0.67, + "grad_norm": 1.9750185866378205, + "learning_rate": 5.16677090996373e-06, + "loss": 1.0653, + "step": 28471 + }, + { + "epoch": 0.67, + "grad_norm": 2.2733956152664225, + "learning_rate": 5.166102925724954e-06, + "loss": 0.9758, + "step": 28472 + }, + { + "epoch": 0.67, + "grad_norm": 1.0853326704240691, + "learning_rate": 5.165434969630654e-06, + "loss": 0.9438, + "step": 28473 + }, + { + "epoch": 0.67, + "grad_norm": 2.0990833502910293, + "learning_rate": 5.164767041684724e-06, + "loss": 1.0415, + "step": 28474 + }, + { + "epoch": 0.67, + "grad_norm": 2.044761078288876, + "learning_rate": 5.1640991418910494e-06, + "loss": 1.0148, + "step": 28475 + }, + { + "epoch": 0.67, + "grad_norm": 2.0261794785803127, + "learning_rate": 5.163431270253523e-06, + "loss": 0.9344, + "step": 28476 + }, + { + "epoch": 0.67, + "grad_norm": 1.9706168522112069, + "learning_rate": 5.162763426776027e-06, + "loss": 1.0456, + "step": 28477 + }, + { + "epoch": 0.67, + "grad_norm": 1.9861238542434076, + "learning_rate": 5.162095611462458e-06, + "loss": 0.9396, + "step": 28478 + }, + { + "epoch": 0.67, + "grad_norm": 1.094903735597644, + "learning_rate": 5.161427824316703e-06, + "loss": 0.948, + "step": 28479 + }, + { + "epoch": 0.67, + "grad_norm": 1.9183756577304274, + "learning_rate": 5.160760065342639e-06, + "loss": 1.0039, + "step": 28480 + }, + { + "epoch": 0.67, + "grad_norm": 1.8995086630428486, + "learning_rate": 5.160092334544165e-06, + "loss": 1.0731, + "step": 28481 + }, + { + "epoch": 0.67, + "grad_norm": 1.9479853445825348, + "learning_rate": 5.159424631925168e-06, + "loss": 1.0698, + "step": 28482 + }, + { + "epoch": 0.67, + "grad_norm": 1.890001740929258, + "learning_rate": 5.158756957489533e-06, + "loss": 0.9789, + "step": 28483 + }, + { + "epoch": 0.67, + "grad_norm": 2.1129887119703032, + "learning_rate": 5.158089311241143e-06, + "loss": 1.0088, + "step": 28484 + }, + { + "epoch": 0.67, + "grad_norm": 1.1364671846223038, + "learning_rate": 5.157421693183893e-06, + "loss": 0.9315, + "step": 28485 + }, + { + "epoch": 0.67, + "grad_norm": 1.8939726929093288, + "learning_rate": 5.156754103321669e-06, + "loss": 0.9968, + "step": 28486 + }, + { + "epoch": 0.67, + "grad_norm": 1.8371315701984345, + "learning_rate": 5.156086541658356e-06, + "loss": 0.9402, + "step": 28487 + }, + { + "epoch": 0.67, + "grad_norm": 1.8203266390365964, + "learning_rate": 5.155419008197837e-06, + "loss": 1.0509, + "step": 28488 + }, + { + "epoch": 0.67, + "grad_norm": 2.1691517399958755, + "learning_rate": 5.154751502944002e-06, + "loss": 0.9825, + "step": 28489 + }, + { + "epoch": 0.67, + "grad_norm": 2.010130331671637, + "learning_rate": 5.154084025900742e-06, + "loss": 1.063, + "step": 28490 + }, + { + "epoch": 0.67, + "grad_norm": 1.8595458109268892, + "learning_rate": 5.153416577071939e-06, + "loss": 1.0954, + "step": 28491 + }, + { + "epoch": 0.67, + "grad_norm": 2.077694250065588, + "learning_rate": 5.152749156461475e-06, + "loss": 1.0514, + "step": 28492 + }, + { + "epoch": 0.67, + "grad_norm": 2.083318081392646, + "learning_rate": 5.152081764073244e-06, + "loss": 0.9689, + "step": 28493 + }, + { + "epoch": 0.67, + "grad_norm": 1.0913229386669088, + "learning_rate": 5.1514143999111234e-06, + "loss": 0.9891, + "step": 28494 + }, + { + "epoch": 0.67, + "grad_norm": 2.1478100573007968, + "learning_rate": 5.150747063979008e-06, + "loss": 0.868, + "step": 28495 + }, + { + "epoch": 0.67, + "grad_norm": 1.9930354288389969, + "learning_rate": 5.150079756280774e-06, + "loss": 0.9468, + "step": 28496 + }, + { + "epoch": 0.67, + "grad_norm": 1.8931976106361716, + "learning_rate": 5.149412476820314e-06, + "loss": 0.958, + "step": 28497 + }, + { + "epoch": 0.67, + "grad_norm": 1.0712034074728507, + "learning_rate": 5.148745225601509e-06, + "loss": 0.9596, + "step": 28498 + }, + { + "epoch": 0.67, + "grad_norm": 1.9909061760037883, + "learning_rate": 5.148078002628247e-06, + "loss": 0.9981, + "step": 28499 + }, + { + "epoch": 0.67, + "grad_norm": 2.064865770756573, + "learning_rate": 5.147410807904408e-06, + "loss": 0.8723, + "step": 28500 + }, + { + "epoch": 0.67, + "grad_norm": 1.86272492940692, + "learning_rate": 5.146743641433884e-06, + "loss": 0.9232, + "step": 28501 + }, + { + "epoch": 0.67, + "grad_norm": 2.0166947992121327, + "learning_rate": 5.146076503220549e-06, + "loss": 0.8773, + "step": 28502 + }, + { + "epoch": 0.67, + "grad_norm": 1.8774870194796025, + "learning_rate": 5.145409393268299e-06, + "loss": 0.9657, + "step": 28503 + }, + { + "epoch": 0.67, + "grad_norm": 1.8700018822927176, + "learning_rate": 5.144742311581013e-06, + "loss": 0.9065, + "step": 28504 + }, + { + "epoch": 0.67, + "grad_norm": 2.0726369126083197, + "learning_rate": 5.144075258162569e-06, + "loss": 1.0863, + "step": 28505 + }, + { + "epoch": 0.67, + "grad_norm": 1.9704395103461196, + "learning_rate": 5.143408233016855e-06, + "loss": 0.8617, + "step": 28506 + }, + { + "epoch": 0.67, + "grad_norm": 1.0502681572961545, + "learning_rate": 5.142741236147762e-06, + "loss": 0.9401, + "step": 28507 + }, + { + "epoch": 0.67, + "grad_norm": 1.9375470950282638, + "learning_rate": 5.142074267559167e-06, + "loss": 1.0057, + "step": 28508 + }, + { + "epoch": 0.67, + "grad_norm": 2.4405837988268253, + "learning_rate": 5.141407327254951e-06, + "loss": 1.0254, + "step": 28509 + }, + { + "epoch": 0.67, + "grad_norm": 2.1154068684317955, + "learning_rate": 5.140740415238996e-06, + "loss": 1.0805, + "step": 28510 + }, + { + "epoch": 0.67, + "grad_norm": 2.4547502138250707, + "learning_rate": 5.140073531515195e-06, + "loss": 0.9873, + "step": 28511 + }, + { + "epoch": 0.67, + "grad_norm": 2.2424295945591695, + "learning_rate": 5.139406676087424e-06, + "loss": 0.9758, + "step": 28512 + }, + { + "epoch": 0.67, + "grad_norm": 2.0036295935464175, + "learning_rate": 5.138739848959563e-06, + "loss": 1.0837, + "step": 28513 + }, + { + "epoch": 0.67, + "grad_norm": 1.0551010587070042, + "learning_rate": 5.138073050135497e-06, + "loss": 0.9364, + "step": 28514 + }, + { + "epoch": 0.67, + "grad_norm": 1.8985658141618285, + "learning_rate": 5.137406279619113e-06, + "loss": 0.8879, + "step": 28515 + }, + { + "epoch": 0.67, + "grad_norm": 1.9424345525982158, + "learning_rate": 5.136739537414291e-06, + "loss": 0.9804, + "step": 28516 + }, + { + "epoch": 0.67, + "grad_norm": 1.9837184348054004, + "learning_rate": 5.136072823524905e-06, + "loss": 1.0716, + "step": 28517 + }, + { + "epoch": 0.67, + "grad_norm": 2.2806161088488723, + "learning_rate": 5.1354061379548455e-06, + "loss": 1.1036, + "step": 28518 + }, + { + "epoch": 0.67, + "grad_norm": 2.042335455974631, + "learning_rate": 5.134739480707991e-06, + "loss": 0.8404, + "step": 28519 + }, + { + "epoch": 0.67, + "grad_norm": 2.303755841822159, + "learning_rate": 5.1340728517882256e-06, + "loss": 1.0583, + "step": 28520 + }, + { + "epoch": 0.67, + "grad_norm": 1.8168938757109183, + "learning_rate": 5.133406251199424e-06, + "loss": 1.0502, + "step": 28521 + }, + { + "epoch": 0.67, + "grad_norm": 2.118644110023375, + "learning_rate": 5.132739678945478e-06, + "loss": 0.9854, + "step": 28522 + }, + { + "epoch": 0.67, + "grad_norm": 1.863085858103787, + "learning_rate": 5.132073135030255e-06, + "loss": 0.9035, + "step": 28523 + }, + { + "epoch": 0.67, + "grad_norm": 1.891228501998069, + "learning_rate": 5.13140661945765e-06, + "loss": 0.9944, + "step": 28524 + }, + { + "epoch": 0.67, + "grad_norm": 2.06868262934867, + "learning_rate": 5.1307401322315375e-06, + "loss": 0.9025, + "step": 28525 + }, + { + "epoch": 0.67, + "grad_norm": 2.109864388756703, + "learning_rate": 5.1300736733557936e-06, + "loss": 1.0088, + "step": 28526 + }, + { + "epoch": 0.67, + "grad_norm": 2.014648333083523, + "learning_rate": 5.129407242834301e-06, + "loss": 1.0121, + "step": 28527 + }, + { + "epoch": 0.67, + "grad_norm": 1.9029706504836796, + "learning_rate": 5.128740840670946e-06, + "loss": 1.0193, + "step": 28528 + }, + { + "epoch": 0.67, + "grad_norm": 2.074413865439324, + "learning_rate": 5.128074466869603e-06, + "loss": 0.8329, + "step": 28529 + }, + { + "epoch": 0.67, + "grad_norm": 2.660545851256118, + "learning_rate": 5.127408121434152e-06, + "loss": 0.9839, + "step": 28530 + }, + { + "epoch": 0.67, + "grad_norm": 1.8463005649335869, + "learning_rate": 5.126741804368471e-06, + "loss": 1.0564, + "step": 28531 + }, + { + "epoch": 0.67, + "grad_norm": 2.536898682048629, + "learning_rate": 5.1260755156764474e-06, + "loss": 0.9839, + "step": 28532 + }, + { + "epoch": 0.67, + "grad_norm": 2.302164166359065, + "learning_rate": 5.125409255361954e-06, + "loss": 1.0653, + "step": 28533 + }, + { + "epoch": 0.67, + "grad_norm": 1.7120159225494942, + "learning_rate": 5.124743023428867e-06, + "loss": 0.9147, + "step": 28534 + }, + { + "epoch": 0.67, + "grad_norm": 1.890350601065467, + "learning_rate": 5.124076819881071e-06, + "loss": 1.0194, + "step": 28535 + }, + { + "epoch": 0.67, + "grad_norm": 1.8478222918175393, + "learning_rate": 5.123410644722446e-06, + "loss": 0.9743, + "step": 28536 + }, + { + "epoch": 0.67, + "grad_norm": 5.838819216440073, + "learning_rate": 5.1227444979568694e-06, + "loss": 0.8622, + "step": 28537 + }, + { + "epoch": 0.67, + "grad_norm": 2.078977451118514, + "learning_rate": 5.122078379588214e-06, + "loss": 1.0877, + "step": 28538 + }, + { + "epoch": 0.67, + "grad_norm": 2.7427439249920558, + "learning_rate": 5.121412289620367e-06, + "loss": 1.0434, + "step": 28539 + }, + { + "epoch": 0.67, + "grad_norm": 1.2588051338257453, + "learning_rate": 5.120746228057197e-06, + "loss": 0.9845, + "step": 28540 + }, + { + "epoch": 0.67, + "grad_norm": 2.5033665168687596, + "learning_rate": 5.120080194902593e-06, + "loss": 0.9985, + "step": 28541 + }, + { + "epoch": 0.67, + "grad_norm": 2.2331435532341164, + "learning_rate": 5.119414190160421e-06, + "loss": 0.9161, + "step": 28542 + }, + { + "epoch": 0.67, + "grad_norm": 2.038877940366535, + "learning_rate": 5.11874821383457e-06, + "loss": 0.8322, + "step": 28543 + }, + { + "epoch": 0.67, + "grad_norm": 3.279048489310901, + "learning_rate": 5.118082265928908e-06, + "loss": 1.0455, + "step": 28544 + }, + { + "epoch": 0.67, + "grad_norm": 2.0307749505080888, + "learning_rate": 5.117416346447317e-06, + "loss": 1.0006, + "step": 28545 + }, + { + "epoch": 0.67, + "grad_norm": 2.1848120097394874, + "learning_rate": 5.116750455393681e-06, + "loss": 0.9126, + "step": 28546 + }, + { + "epoch": 0.67, + "grad_norm": 1.8701702827505924, + "learning_rate": 5.116084592771861e-06, + "loss": 1.0369, + "step": 28547 + }, + { + "epoch": 0.67, + "grad_norm": 2.1123243851036464, + "learning_rate": 5.115418758585745e-06, + "loss": 1.0086, + "step": 28548 + }, + { + "epoch": 0.67, + "grad_norm": 2.340280922364076, + "learning_rate": 5.114752952839207e-06, + "loss": 1.0144, + "step": 28549 + }, + { + "epoch": 0.67, + "grad_norm": 1.898764715159067, + "learning_rate": 5.114087175536131e-06, + "loss": 0.9695, + "step": 28550 + }, + { + "epoch": 0.67, + "grad_norm": 1.9327560896421536, + "learning_rate": 5.113421426680378e-06, + "loss": 1.0284, + "step": 28551 + }, + { + "epoch": 0.67, + "grad_norm": 2.1342295168071566, + "learning_rate": 5.112755706275833e-06, + "loss": 1.0631, + "step": 28552 + }, + { + "epoch": 0.67, + "grad_norm": 2.134100337096841, + "learning_rate": 5.112090014326376e-06, + "loss": 1.0413, + "step": 28553 + }, + { + "epoch": 0.67, + "grad_norm": 2.3189680140947266, + "learning_rate": 5.111424350835879e-06, + "loss": 1.0459, + "step": 28554 + }, + { + "epoch": 0.67, + "grad_norm": 1.9665130599392762, + "learning_rate": 5.110758715808213e-06, + "loss": 0.938, + "step": 28555 + }, + { + "epoch": 0.67, + "grad_norm": 2.304494661107091, + "learning_rate": 5.1100931092472564e-06, + "loss": 1.0507, + "step": 28556 + }, + { + "epoch": 0.67, + "grad_norm": 1.846114548488518, + "learning_rate": 5.10942753115689e-06, + "loss": 0.9653, + "step": 28557 + }, + { + "epoch": 0.67, + "grad_norm": 2.1711690055631165, + "learning_rate": 5.1087619815409815e-06, + "loss": 0.9491, + "step": 28558 + }, + { + "epoch": 0.67, + "grad_norm": 1.9824792375627742, + "learning_rate": 5.1080964604034136e-06, + "loss": 0.9485, + "step": 28559 + }, + { + "epoch": 0.67, + "grad_norm": 1.7990767068247722, + "learning_rate": 5.107430967748056e-06, + "loss": 1.0416, + "step": 28560 + }, + { + "epoch": 0.67, + "grad_norm": 2.17586119457198, + "learning_rate": 5.106765503578781e-06, + "loss": 0.975, + "step": 28561 + }, + { + "epoch": 0.67, + "grad_norm": 2.071875161836044, + "learning_rate": 5.1061000678994665e-06, + "loss": 1.0311, + "step": 28562 + }, + { + "epoch": 0.67, + "grad_norm": 2.2911103926974925, + "learning_rate": 5.10543466071399e-06, + "loss": 1.0635, + "step": 28563 + }, + { + "epoch": 0.67, + "grad_norm": 2.6781146274606984, + "learning_rate": 5.104769282026223e-06, + "loss": 0.9491, + "step": 28564 + }, + { + "epoch": 0.67, + "grad_norm": 1.9217099431683469, + "learning_rate": 5.104103931840035e-06, + "loss": 0.9744, + "step": 28565 + }, + { + "epoch": 0.67, + "grad_norm": 1.1203200669752997, + "learning_rate": 5.103438610159303e-06, + "loss": 0.9401, + "step": 28566 + }, + { + "epoch": 0.67, + "grad_norm": 1.961433607204058, + "learning_rate": 5.102773316987905e-06, + "loss": 0.9078, + "step": 28567 + }, + { + "epoch": 0.67, + "grad_norm": 1.0872085749369267, + "learning_rate": 5.102108052329714e-06, + "loss": 0.9267, + "step": 28568 + }, + { + "epoch": 0.67, + "grad_norm": 2.215808394796367, + "learning_rate": 5.1014428161885956e-06, + "loss": 0.9581, + "step": 28569 + }, + { + "epoch": 0.67, + "grad_norm": 2.3934114381015856, + "learning_rate": 5.1007776085684265e-06, + "loss": 0.9287, + "step": 28570 + }, + { + "epoch": 0.67, + "grad_norm": 2.0408476841429306, + "learning_rate": 5.100112429473087e-06, + "loss": 0.9335, + "step": 28571 + }, + { + "epoch": 0.67, + "grad_norm": 2.3473095640854043, + "learning_rate": 5.0994472789064434e-06, + "loss": 1.0701, + "step": 28572 + }, + { + "epoch": 0.67, + "grad_norm": 2.5406131949830844, + "learning_rate": 5.098782156872365e-06, + "loss": 0.8881, + "step": 28573 + }, + { + "epoch": 0.67, + "grad_norm": 1.9836779411098169, + "learning_rate": 5.0981170633747326e-06, + "loss": 0.9617, + "step": 28574 + }, + { + "epoch": 0.67, + "grad_norm": 2.1721713545505796, + "learning_rate": 5.097451998417411e-06, + "loss": 0.8489, + "step": 28575 + }, + { + "epoch": 0.67, + "grad_norm": 2.239979685412937, + "learning_rate": 5.09678696200428e-06, + "loss": 0.9881, + "step": 28576 + }, + { + "epoch": 0.67, + "grad_norm": 1.8163416080611452, + "learning_rate": 5.096121954139204e-06, + "loss": 1.0494, + "step": 28577 + }, + { + "epoch": 0.67, + "grad_norm": 1.8494070195680232, + "learning_rate": 5.095456974826061e-06, + "loss": 0.9424, + "step": 28578 + }, + { + "epoch": 0.67, + "grad_norm": 2.0052383447182667, + "learning_rate": 5.094792024068718e-06, + "loss": 1.0443, + "step": 28579 + }, + { + "epoch": 0.67, + "grad_norm": 1.148866409740734, + "learning_rate": 5.0941271018710514e-06, + "loss": 0.9456, + "step": 28580 + }, + { + "epoch": 0.67, + "grad_norm": 1.0388208977593185, + "learning_rate": 5.093462208236931e-06, + "loss": 0.9982, + "step": 28581 + }, + { + "epoch": 0.67, + "grad_norm": 1.8465882466047356, + "learning_rate": 5.092797343170223e-06, + "loss": 1.0394, + "step": 28582 + }, + { + "epoch": 0.67, + "grad_norm": 1.8849263438942876, + "learning_rate": 5.092132506674801e-06, + "loss": 1.0967, + "step": 28583 + }, + { + "epoch": 0.67, + "grad_norm": 2.020010824947761, + "learning_rate": 5.091467698754543e-06, + "loss": 1.0065, + "step": 28584 + }, + { + "epoch": 0.67, + "grad_norm": 1.7770763828841876, + "learning_rate": 5.090802919413314e-06, + "loss": 0.9943, + "step": 28585 + }, + { + "epoch": 0.67, + "grad_norm": 1.091972904887087, + "learning_rate": 5.09013816865498e-06, + "loss": 0.9754, + "step": 28586 + }, + { + "epoch": 0.67, + "grad_norm": 1.9323034119725677, + "learning_rate": 5.089473446483416e-06, + "loss": 1.0848, + "step": 28587 + }, + { + "epoch": 0.67, + "grad_norm": 1.0946719677504078, + "learning_rate": 5.0888087529024965e-06, + "loss": 1.0312, + "step": 28588 + }, + { + "epoch": 0.67, + "grad_norm": 2.212279977489549, + "learning_rate": 5.088144087916087e-06, + "loss": 0.912, + "step": 28589 + }, + { + "epoch": 0.67, + "grad_norm": 1.8959693803656505, + "learning_rate": 5.087479451528055e-06, + "loss": 0.954, + "step": 28590 + }, + { + "epoch": 0.67, + "grad_norm": 1.8534898661932917, + "learning_rate": 5.086814843742272e-06, + "loss": 1.0672, + "step": 28591 + }, + { + "epoch": 0.67, + "grad_norm": 1.9243251290470904, + "learning_rate": 5.086150264562612e-06, + "loss": 1.118, + "step": 28592 + }, + { + "epoch": 0.67, + "grad_norm": 2.0696858374206877, + "learning_rate": 5.08548571399294e-06, + "loss": 0.9503, + "step": 28593 + }, + { + "epoch": 0.67, + "grad_norm": 2.0046481425668254, + "learning_rate": 5.0848211920371236e-06, + "loss": 1.1069, + "step": 28594 + }, + { + "epoch": 0.67, + "grad_norm": 1.9445710284351183, + "learning_rate": 5.084156698699038e-06, + "loss": 0.9554, + "step": 28595 + }, + { + "epoch": 0.67, + "grad_norm": 2.014551650291254, + "learning_rate": 5.0834922339825445e-06, + "loss": 0.8817, + "step": 28596 + }, + { + "epoch": 0.67, + "grad_norm": 1.808635272526262, + "learning_rate": 5.08282779789152e-06, + "loss": 0.9858, + "step": 28597 + }, + { + "epoch": 0.67, + "grad_norm": 1.9854548570957122, + "learning_rate": 5.082163390429824e-06, + "loss": 0.9796, + "step": 28598 + }, + { + "epoch": 0.67, + "grad_norm": 2.226833716613622, + "learning_rate": 5.081499011601334e-06, + "loss": 0.9747, + "step": 28599 + }, + { + "epoch": 0.67, + "grad_norm": 2.445193240289928, + "learning_rate": 5.0808346614099106e-06, + "loss": 0.9577, + "step": 28600 + }, + { + "epoch": 0.67, + "grad_norm": 1.9874886670113763, + "learning_rate": 5.080170339859429e-06, + "loss": 0.9247, + "step": 28601 + }, + { + "epoch": 0.67, + "grad_norm": 2.9245134652055764, + "learning_rate": 5.079506046953753e-06, + "loss": 1.0114, + "step": 28602 + }, + { + "epoch": 0.67, + "grad_norm": 2.0712753745754897, + "learning_rate": 5.078841782696746e-06, + "loss": 0.9612, + "step": 28603 + }, + { + "epoch": 0.67, + "grad_norm": 2.732868153193244, + "learning_rate": 5.07817754709228e-06, + "loss": 1.0822, + "step": 28604 + }, + { + "epoch": 0.67, + "grad_norm": 1.938239461321672, + "learning_rate": 5.0775133401442275e-06, + "loss": 0.9361, + "step": 28605 + }, + { + "epoch": 0.67, + "grad_norm": 1.930155548435495, + "learning_rate": 5.0768491618564504e-06, + "loss": 0.9918, + "step": 28606 + }, + { + "epoch": 0.67, + "grad_norm": 1.811217704265094, + "learning_rate": 5.076185012232813e-06, + "loss": 0.8479, + "step": 28607 + }, + { + "epoch": 0.67, + "grad_norm": 1.8323576753045634, + "learning_rate": 5.0755208912771835e-06, + "loss": 1.0089, + "step": 28608 + }, + { + "epoch": 0.67, + "grad_norm": 2.0386778343730847, + "learning_rate": 5.074856798993436e-06, + "loss": 1.0088, + "step": 28609 + }, + { + "epoch": 0.67, + "grad_norm": 2.137476249495427, + "learning_rate": 5.0741927353854305e-06, + "loss": 1.0256, + "step": 28610 + }, + { + "epoch": 0.67, + "grad_norm": 2.0946653257102184, + "learning_rate": 5.073528700457031e-06, + "loss": 0.9252, + "step": 28611 + }, + { + "epoch": 0.67, + "grad_norm": 1.9976875563798953, + "learning_rate": 5.0728646942121074e-06, + "loss": 1.1245, + "step": 28612 + }, + { + "epoch": 0.67, + "grad_norm": 1.9604609619135707, + "learning_rate": 5.0722007166545294e-06, + "loss": 1.0414, + "step": 28613 + }, + { + "epoch": 0.67, + "grad_norm": 2.4076088904456543, + "learning_rate": 5.071536767788158e-06, + "loss": 0.9172, + "step": 28614 + }, + { + "epoch": 0.67, + "grad_norm": 1.9133522368999207, + "learning_rate": 5.070872847616857e-06, + "loss": 0.8526, + "step": 28615 + }, + { + "epoch": 0.67, + "grad_norm": 1.898531508055928, + "learning_rate": 5.070208956144498e-06, + "loss": 0.9561, + "step": 28616 + }, + { + "epoch": 0.67, + "grad_norm": 2.114563706296451, + "learning_rate": 5.0695450933749395e-06, + "loss": 1.008, + "step": 28617 + }, + { + "epoch": 0.67, + "grad_norm": 2.0658338979774054, + "learning_rate": 5.068881259312054e-06, + "loss": 0.961, + "step": 28618 + }, + { + "epoch": 0.67, + "grad_norm": 1.9465790483876446, + "learning_rate": 5.068217453959698e-06, + "loss": 1.078, + "step": 28619 + }, + { + "epoch": 0.67, + "grad_norm": 1.7496130617436743, + "learning_rate": 5.067553677321748e-06, + "loss": 0.909, + "step": 28620 + }, + { + "epoch": 0.67, + "grad_norm": 2.0045012398610815, + "learning_rate": 5.066889929402057e-06, + "loss": 1.0169, + "step": 28621 + }, + { + "epoch": 0.67, + "grad_norm": 1.0415904255746755, + "learning_rate": 5.066226210204498e-06, + "loss": 0.9737, + "step": 28622 + }, + { + "epoch": 0.67, + "grad_norm": 2.0420401033711544, + "learning_rate": 5.065562519732929e-06, + "loss": 0.9265, + "step": 28623 + }, + { + "epoch": 0.67, + "grad_norm": 2.132137146027347, + "learning_rate": 5.064898857991221e-06, + "loss": 0.9725, + "step": 28624 + }, + { + "epoch": 0.67, + "grad_norm": 1.9166041225909756, + "learning_rate": 5.0642352249832295e-06, + "loss": 1.017, + "step": 28625 + }, + { + "epoch": 0.67, + "grad_norm": 1.8621071854289013, + "learning_rate": 5.0635716207128274e-06, + "loss": 1.0137, + "step": 28626 + }, + { + "epoch": 0.67, + "grad_norm": 2.9780677316381583, + "learning_rate": 5.0629080451838745e-06, + "loss": 0.944, + "step": 28627 + }, + { + "epoch": 0.67, + "grad_norm": 2.2273849326739934, + "learning_rate": 5.062244498400228e-06, + "loss": 1.0658, + "step": 28628 + }, + { + "epoch": 0.67, + "grad_norm": 2.084161605198847, + "learning_rate": 5.0615809803657615e-06, + "loss": 1.0242, + "step": 28629 + }, + { + "epoch": 0.67, + "grad_norm": 2.111988150939204, + "learning_rate": 5.060917491084331e-06, + "loss": 0.8796, + "step": 28630 + }, + { + "epoch": 0.67, + "grad_norm": 1.864063651420988, + "learning_rate": 5.06025403055981e-06, + "loss": 0.9868, + "step": 28631 + }, + { + "epoch": 0.67, + "grad_norm": 1.9054383032300999, + "learning_rate": 5.059590598796047e-06, + "loss": 1.0264, + "step": 28632 + }, + { + "epoch": 0.67, + "grad_norm": 2.757811298704426, + "learning_rate": 5.058927195796912e-06, + "loss": 1.0301, + "step": 28633 + }, + { + "epoch": 0.67, + "grad_norm": 2.2594882088822645, + "learning_rate": 5.0582638215662715e-06, + "loss": 0.9101, + "step": 28634 + }, + { + "epoch": 0.67, + "grad_norm": 2.0108162881111484, + "learning_rate": 5.057600476107982e-06, + "loss": 0.9876, + "step": 28635 + }, + { + "epoch": 0.67, + "grad_norm": 2.084211763643289, + "learning_rate": 5.0569371594259056e-06, + "loss": 0.9866, + "step": 28636 + }, + { + "epoch": 0.67, + "grad_norm": 1.914718992974199, + "learning_rate": 5.056273871523904e-06, + "loss": 1.0488, + "step": 28637 + }, + { + "epoch": 0.67, + "grad_norm": 1.9672765917436548, + "learning_rate": 5.055610612405846e-06, + "loss": 1.035, + "step": 28638 + }, + { + "epoch": 0.67, + "grad_norm": 1.0655458196465155, + "learning_rate": 5.054947382075588e-06, + "loss": 0.979, + "step": 28639 + }, + { + "epoch": 0.67, + "grad_norm": 2.874584795565802, + "learning_rate": 5.054284180536988e-06, + "loss": 0.9915, + "step": 28640 + }, + { + "epoch": 0.67, + "grad_norm": 1.8684237455341923, + "learning_rate": 5.053621007793916e-06, + "loss": 0.9713, + "step": 28641 + }, + { + "epoch": 0.67, + "grad_norm": 2.0069524021185248, + "learning_rate": 5.052957863850223e-06, + "loss": 0.9868, + "step": 28642 + }, + { + "epoch": 0.67, + "grad_norm": 2.1235234557807687, + "learning_rate": 5.052294748709776e-06, + "loss": 0.9693, + "step": 28643 + }, + { + "epoch": 0.67, + "grad_norm": 1.0711321962579188, + "learning_rate": 5.051631662376441e-06, + "loss": 0.9507, + "step": 28644 + }, + { + "epoch": 0.67, + "grad_norm": 2.06743403525813, + "learning_rate": 5.050968604854071e-06, + "loss": 1.1597, + "step": 28645 + }, + { + "epoch": 0.67, + "grad_norm": 2.097352914348952, + "learning_rate": 5.050305576146526e-06, + "loss": 0.9654, + "step": 28646 + }, + { + "epoch": 0.67, + "grad_norm": 2.0715542904033977, + "learning_rate": 5.0496425762576685e-06, + "loss": 1.1078, + "step": 28647 + }, + { + "epoch": 0.67, + "grad_norm": 1.0452250397377438, + "learning_rate": 5.048979605191367e-06, + "loss": 0.8861, + "step": 28648 + }, + { + "epoch": 0.67, + "grad_norm": 1.0996499342069923, + "learning_rate": 5.0483166629514654e-06, + "loss": 0.9374, + "step": 28649 + }, + { + "epoch": 0.67, + "grad_norm": 1.7620171145731962, + "learning_rate": 5.047653749541832e-06, + "loss": 0.9005, + "step": 28650 + }, + { + "epoch": 0.67, + "grad_norm": 2.030028878187273, + "learning_rate": 5.046990864966328e-06, + "loss": 1.0054, + "step": 28651 + }, + { + "epoch": 0.68, + "grad_norm": 2.7179440200831384, + "learning_rate": 5.046328009228817e-06, + "loss": 0.9588, + "step": 28652 + }, + { + "epoch": 0.68, + "grad_norm": 2.0749526830089384, + "learning_rate": 5.045665182333146e-06, + "loss": 1.0011, + "step": 28653 + }, + { + "epoch": 0.68, + "grad_norm": 2.0478807691361416, + "learning_rate": 5.045002384283179e-06, + "loss": 1.0241, + "step": 28654 + }, + { + "epoch": 0.68, + "grad_norm": 2.1042845874760094, + "learning_rate": 5.044339615082783e-06, + "loss": 1.0005, + "step": 28655 + }, + { + "epoch": 0.68, + "grad_norm": 2.276192561234483, + "learning_rate": 5.043676874735807e-06, + "loss": 1.0543, + "step": 28656 + }, + { + "epoch": 0.68, + "grad_norm": 2.3844960747313375, + "learning_rate": 5.043014163246116e-06, + "loss": 0.9358, + "step": 28657 + }, + { + "epoch": 0.68, + "grad_norm": 1.7318355085804327, + "learning_rate": 5.042351480617562e-06, + "loss": 1.0104, + "step": 28658 + }, + { + "epoch": 0.68, + "grad_norm": 1.951077504782714, + "learning_rate": 5.041688826854012e-06, + "loss": 1.0448, + "step": 28659 + }, + { + "epoch": 0.68, + "grad_norm": 1.820199062047319, + "learning_rate": 5.041026201959314e-06, + "loss": 0.9402, + "step": 28660 + }, + { + "epoch": 0.68, + "grad_norm": 2.02080667199729, + "learning_rate": 5.040363605937337e-06, + "loss": 0.9894, + "step": 28661 + }, + { + "epoch": 0.68, + "grad_norm": 3.9692429212148177, + "learning_rate": 5.039701038791932e-06, + "loss": 1.0397, + "step": 28662 + }, + { + "epoch": 0.68, + "grad_norm": 2.025323323168071, + "learning_rate": 5.039038500526955e-06, + "loss": 1.0811, + "step": 28663 + }, + { + "epoch": 0.68, + "grad_norm": 1.1019427315616686, + "learning_rate": 5.038375991146266e-06, + "loss": 0.9087, + "step": 28664 + }, + { + "epoch": 0.68, + "grad_norm": 2.213966245258748, + "learning_rate": 5.037713510653727e-06, + "loss": 1.0298, + "step": 28665 + }, + { + "epoch": 0.68, + "grad_norm": 1.982342748988998, + "learning_rate": 5.03705105905319e-06, + "loss": 0.9914, + "step": 28666 + }, + { + "epoch": 0.68, + "grad_norm": 2.214224037172914, + "learning_rate": 5.0363886363485105e-06, + "loss": 1.116, + "step": 28667 + }, + { + "epoch": 0.68, + "grad_norm": 2.1862411738511156, + "learning_rate": 5.035726242543546e-06, + "loss": 0.9848, + "step": 28668 + }, + { + "epoch": 0.68, + "grad_norm": 2.1799539134899653, + "learning_rate": 5.0350638776421615e-06, + "loss": 0.9559, + "step": 28669 + }, + { + "epoch": 0.68, + "grad_norm": 1.8868134368214464, + "learning_rate": 5.034401541648204e-06, + "loss": 0.8572, + "step": 28670 + }, + { + "epoch": 0.68, + "grad_norm": 2.1742310517941768, + "learning_rate": 5.033739234565531e-06, + "loss": 0.9867, + "step": 28671 + }, + { + "epoch": 0.68, + "grad_norm": 3.325967575293288, + "learning_rate": 5.033076956398e-06, + "loss": 0.9539, + "step": 28672 + }, + { + "epoch": 0.68, + "grad_norm": 2.0826624949801156, + "learning_rate": 5.032414707149471e-06, + "loss": 1.0872, + "step": 28673 + }, + { + "epoch": 0.68, + "grad_norm": 2.628203181285806, + "learning_rate": 5.031752486823797e-06, + "loss": 1.0166, + "step": 28674 + }, + { + "epoch": 0.68, + "grad_norm": 2.2306876615851317, + "learning_rate": 5.031090295424828e-06, + "loss": 1.0531, + "step": 28675 + }, + { + "epoch": 0.68, + "grad_norm": 2.23001202972769, + "learning_rate": 5.030428132956428e-06, + "loss": 0.8741, + "step": 28676 + }, + { + "epoch": 0.68, + "grad_norm": 2.151002530172683, + "learning_rate": 5.029765999422447e-06, + "loss": 1.0357, + "step": 28677 + }, + { + "epoch": 0.68, + "grad_norm": 1.9746433698716264, + "learning_rate": 5.029103894826745e-06, + "loss": 1.0, + "step": 28678 + }, + { + "epoch": 0.68, + "grad_norm": 2.146095153705199, + "learning_rate": 5.0284418191731684e-06, + "loss": 0.9559, + "step": 28679 + }, + { + "epoch": 0.68, + "grad_norm": 2.0438462342499695, + "learning_rate": 5.027779772465583e-06, + "loss": 0.8791, + "step": 28680 + }, + { + "epoch": 0.68, + "grad_norm": 1.8190861586556446, + "learning_rate": 5.027117754707833e-06, + "loss": 1.006, + "step": 28681 + }, + { + "epoch": 0.68, + "grad_norm": 2.354975587269127, + "learning_rate": 5.026455765903783e-06, + "loss": 0.9233, + "step": 28682 + }, + { + "epoch": 0.68, + "grad_norm": 1.9175361171580818, + "learning_rate": 5.025793806057282e-06, + "loss": 0.9692, + "step": 28683 + }, + { + "epoch": 0.68, + "grad_norm": 2.973025008192481, + "learning_rate": 5.025131875172179e-06, + "loss": 0.9945, + "step": 28684 + }, + { + "epoch": 0.68, + "grad_norm": 2.0174835233445036, + "learning_rate": 5.024469973252334e-06, + "loss": 0.9724, + "step": 28685 + }, + { + "epoch": 0.68, + "grad_norm": 2.9055538366347355, + "learning_rate": 5.023808100301604e-06, + "loss": 0.9319, + "step": 28686 + }, + { + "epoch": 0.68, + "grad_norm": 1.066909783928651, + "learning_rate": 5.023146256323839e-06, + "loss": 0.9259, + "step": 28687 + }, + { + "epoch": 0.68, + "grad_norm": 1.86523944737958, + "learning_rate": 5.022484441322888e-06, + "loss": 0.7999, + "step": 28688 + }, + { + "epoch": 0.68, + "grad_norm": 1.9702119075654, + "learning_rate": 5.021822655302608e-06, + "loss": 1.0675, + "step": 28689 + }, + { + "epoch": 0.68, + "grad_norm": 1.7925900068874348, + "learning_rate": 5.021160898266857e-06, + "loss": 1.0488, + "step": 28690 + }, + { + "epoch": 0.68, + "grad_norm": 1.9800428935698384, + "learning_rate": 5.020499170219483e-06, + "loss": 1.0071, + "step": 28691 + }, + { + "epoch": 0.68, + "grad_norm": 4.443138623046354, + "learning_rate": 5.019837471164335e-06, + "loss": 0.9388, + "step": 28692 + }, + { + "epoch": 0.68, + "grad_norm": 2.218257372946558, + "learning_rate": 5.019175801105272e-06, + "loss": 0.9276, + "step": 28693 + }, + { + "epoch": 0.68, + "grad_norm": 2.4341676498495133, + "learning_rate": 5.018514160046146e-06, + "loss": 0.8619, + "step": 28694 + }, + { + "epoch": 0.68, + "grad_norm": 1.9950937754867377, + "learning_rate": 5.017852547990809e-06, + "loss": 1.002, + "step": 28695 + }, + { + "epoch": 0.68, + "grad_norm": 1.056414137230597, + "learning_rate": 5.0171909649431075e-06, + "loss": 0.9227, + "step": 28696 + }, + { + "epoch": 0.68, + "grad_norm": 1.9187296976768962, + "learning_rate": 5.0165294109069e-06, + "loss": 1.0539, + "step": 28697 + }, + { + "epoch": 0.68, + "grad_norm": 2.139531084896564, + "learning_rate": 5.015867885886035e-06, + "loss": 1.005, + "step": 28698 + }, + { + "epoch": 0.68, + "grad_norm": 2.138434813699285, + "learning_rate": 5.0152063898843685e-06, + "loss": 1.0749, + "step": 28699 + }, + { + "epoch": 0.68, + "grad_norm": 1.1373939105708974, + "learning_rate": 5.014544922905744e-06, + "loss": 0.9237, + "step": 28700 + }, + { + "epoch": 0.68, + "grad_norm": 1.8344484417230786, + "learning_rate": 5.013883484954021e-06, + "loss": 0.9299, + "step": 28701 + }, + { + "epoch": 0.68, + "grad_norm": 1.9039716742456543, + "learning_rate": 5.013222076033043e-06, + "loss": 0.9372, + "step": 28702 + }, + { + "epoch": 0.68, + "grad_norm": 1.0962158175980932, + "learning_rate": 5.012560696146669e-06, + "loss": 0.9985, + "step": 28703 + }, + { + "epoch": 0.68, + "grad_norm": 1.878714936984138, + "learning_rate": 5.011899345298745e-06, + "loss": 1.0131, + "step": 28704 + }, + { + "epoch": 0.68, + "grad_norm": 2.182021229889884, + "learning_rate": 5.011238023493119e-06, + "loss": 1.0271, + "step": 28705 + }, + { + "epoch": 0.68, + "grad_norm": 1.929387976083125, + "learning_rate": 5.010576730733645e-06, + "loss": 0.9338, + "step": 28706 + }, + { + "epoch": 0.68, + "grad_norm": 2.7326195142128187, + "learning_rate": 5.0099154670241756e-06, + "loss": 1.0028, + "step": 28707 + }, + { + "epoch": 0.68, + "grad_norm": 2.1845457367032655, + "learning_rate": 5.009254232368559e-06, + "loss": 1.06, + "step": 28708 + }, + { + "epoch": 0.68, + "grad_norm": 2.360812700957806, + "learning_rate": 5.0085930267706404e-06, + "loss": 0.9718, + "step": 28709 + }, + { + "epoch": 0.68, + "grad_norm": 1.8186358276648182, + "learning_rate": 5.007931850234273e-06, + "loss": 1.029, + "step": 28710 + }, + { + "epoch": 0.68, + "grad_norm": 2.077243668238849, + "learning_rate": 5.0072707027633115e-06, + "loss": 0.9725, + "step": 28711 + }, + { + "epoch": 0.68, + "grad_norm": 1.1402017604442403, + "learning_rate": 5.006609584361601e-06, + "loss": 0.9618, + "step": 28712 + }, + { + "epoch": 0.68, + "grad_norm": 1.7837915654782521, + "learning_rate": 5.005948495032985e-06, + "loss": 0.9216, + "step": 28713 + }, + { + "epoch": 0.68, + "grad_norm": 2.8666448125434254, + "learning_rate": 5.0052874347813186e-06, + "loss": 0.957, + "step": 28714 + }, + { + "epoch": 0.68, + "grad_norm": 1.9983244612335278, + "learning_rate": 5.004626403610454e-06, + "loss": 0.8591, + "step": 28715 + }, + { + "epoch": 0.68, + "grad_norm": 1.899590345920221, + "learning_rate": 5.0039654015242355e-06, + "loss": 1.0432, + "step": 28716 + }, + { + "epoch": 0.68, + "grad_norm": 1.981557184410606, + "learning_rate": 5.00330442852651e-06, + "loss": 0.9749, + "step": 28717 + }, + { + "epoch": 0.68, + "grad_norm": 2.111676688230515, + "learning_rate": 5.00264348462113e-06, + "loss": 0.8921, + "step": 28718 + }, + { + "epoch": 0.68, + "grad_norm": 1.9323765052817479, + "learning_rate": 5.001982569811938e-06, + "loss": 1.0123, + "step": 28719 + }, + { + "epoch": 0.68, + "grad_norm": 1.9916527965510267, + "learning_rate": 5.001321684102791e-06, + "loss": 0.9691, + "step": 28720 + }, + { + "epoch": 0.68, + "grad_norm": 1.8896632425580893, + "learning_rate": 5.000660827497527e-06, + "loss": 1.0781, + "step": 28721 + }, + { + "epoch": 0.68, + "grad_norm": 1.1467924378389347, + "learning_rate": 5.000000000000003e-06, + "loss": 0.9091, + "step": 28722 + }, + { + "epoch": 0.68, + "grad_norm": 2.25084307040998, + "learning_rate": 4.999339201614057e-06, + "loss": 0.959, + "step": 28723 + }, + { + "epoch": 0.68, + "grad_norm": 1.1018934569630898, + "learning_rate": 4.998678432343545e-06, + "loss": 0.9603, + "step": 28724 + }, + { + "epoch": 0.68, + "grad_norm": 1.9790029613992397, + "learning_rate": 4.998017692192312e-06, + "loss": 1.0081, + "step": 28725 + }, + { + "epoch": 0.68, + "grad_norm": 1.1225192329422007, + "learning_rate": 4.9973569811641985e-06, + "loss": 0.939, + "step": 28726 + }, + { + "epoch": 0.68, + "grad_norm": 2.1243396855742414, + "learning_rate": 4.996696299263057e-06, + "loss": 0.8917, + "step": 28727 + }, + { + "epoch": 0.68, + "grad_norm": 1.1475214033225347, + "learning_rate": 4.996035646492734e-06, + "loss": 0.9696, + "step": 28728 + }, + { + "epoch": 0.68, + "grad_norm": 2.044711634682276, + "learning_rate": 4.995375022857083e-06, + "loss": 1.0717, + "step": 28729 + }, + { + "epoch": 0.68, + "grad_norm": 2.033063727230281, + "learning_rate": 4.994714428359936e-06, + "loss": 0.8557, + "step": 28730 + }, + { + "epoch": 0.68, + "grad_norm": 2.0551348656315307, + "learning_rate": 4.994053863005145e-06, + "loss": 0.8703, + "step": 28731 + }, + { + "epoch": 0.68, + "grad_norm": 1.98196618851472, + "learning_rate": 4.993393326796562e-06, + "loss": 0.9778, + "step": 28732 + }, + { + "epoch": 0.68, + "grad_norm": 1.8677818159782886, + "learning_rate": 4.992732819738028e-06, + "loss": 1.0778, + "step": 28733 + }, + { + "epoch": 0.68, + "grad_norm": 2.003634264610622, + "learning_rate": 4.992072341833385e-06, + "loss": 1.0129, + "step": 28734 + }, + { + "epoch": 0.68, + "grad_norm": 2.186120863412963, + "learning_rate": 4.991411893086482e-06, + "loss": 0.9732, + "step": 28735 + }, + { + "epoch": 0.68, + "grad_norm": 1.9263354215243185, + "learning_rate": 4.990751473501168e-06, + "loss": 0.9107, + "step": 28736 + }, + { + "epoch": 0.68, + "grad_norm": 1.8942303095297452, + "learning_rate": 4.990091083081285e-06, + "loss": 1.0471, + "step": 28737 + }, + { + "epoch": 0.68, + "grad_norm": 2.276095857532518, + "learning_rate": 4.989430721830676e-06, + "loss": 0.9055, + "step": 28738 + }, + { + "epoch": 0.68, + "grad_norm": 1.9805824681651105, + "learning_rate": 4.988770389753189e-06, + "loss": 0.939, + "step": 28739 + }, + { + "epoch": 0.68, + "grad_norm": 1.0239905081998069, + "learning_rate": 4.988110086852664e-06, + "loss": 0.9476, + "step": 28740 + }, + { + "epoch": 0.68, + "grad_norm": 2.0136726849523816, + "learning_rate": 4.9874498131329495e-06, + "loss": 1.004, + "step": 28741 + }, + { + "epoch": 0.68, + "grad_norm": 2.379181211378594, + "learning_rate": 4.986789568597893e-06, + "loss": 0.9163, + "step": 28742 + }, + { + "epoch": 0.68, + "grad_norm": 1.9206917825650611, + "learning_rate": 4.986129353251335e-06, + "loss": 1.1135, + "step": 28743 + }, + { + "epoch": 0.68, + "grad_norm": 1.8061906233518672, + "learning_rate": 4.9854691670971145e-06, + "loss": 0.9629, + "step": 28744 + }, + { + "epoch": 0.68, + "grad_norm": 2.1365384432999504, + "learning_rate": 4.9848090101390815e-06, + "loss": 1.0153, + "step": 28745 + }, + { + "epoch": 0.68, + "grad_norm": 1.9050106514914502, + "learning_rate": 4.9841488823810805e-06, + "loss": 1.0505, + "step": 28746 + }, + { + "epoch": 0.68, + "grad_norm": 2.090995349844528, + "learning_rate": 4.983488783826953e-06, + "loss": 0.9593, + "step": 28747 + }, + { + "epoch": 0.68, + "grad_norm": 2.00244852205234, + "learning_rate": 4.9828287144805385e-06, + "loss": 0.8981, + "step": 28748 + }, + { + "epoch": 0.68, + "grad_norm": 1.7449276116847734, + "learning_rate": 4.9821686743456835e-06, + "loss": 0.95, + "step": 28749 + }, + { + "epoch": 0.68, + "grad_norm": 2.26649795683317, + "learning_rate": 4.981508663426236e-06, + "loss": 1.1526, + "step": 28750 + }, + { + "epoch": 0.68, + "grad_norm": 1.8282269025078528, + "learning_rate": 4.9808486817260314e-06, + "loss": 0.9613, + "step": 28751 + }, + { + "epoch": 0.68, + "grad_norm": 1.8910324817882087, + "learning_rate": 4.980188729248913e-06, + "loss": 0.9797, + "step": 28752 + }, + { + "epoch": 0.68, + "grad_norm": 2.018623566571172, + "learning_rate": 4.979528805998724e-06, + "loss": 0.9343, + "step": 28753 + }, + { + "epoch": 0.68, + "grad_norm": 2.2068009773322546, + "learning_rate": 4.978868911979312e-06, + "loss": 0.8232, + "step": 28754 + }, + { + "epoch": 0.68, + "grad_norm": 1.9487965052763108, + "learning_rate": 4.9782090471945155e-06, + "loss": 0.9803, + "step": 28755 + }, + { + "epoch": 0.68, + "grad_norm": 1.7523647688447153, + "learning_rate": 4.977549211648172e-06, + "loss": 0.854, + "step": 28756 + }, + { + "epoch": 0.68, + "grad_norm": 2.081650324866609, + "learning_rate": 4.976889405344129e-06, + "loss": 0.9321, + "step": 28757 + }, + { + "epoch": 0.68, + "grad_norm": 1.9100538987693423, + "learning_rate": 4.9762296282862245e-06, + "loss": 0.8382, + "step": 28758 + }, + { + "epoch": 0.68, + "grad_norm": 2.3682257650143437, + "learning_rate": 4.9755698804783045e-06, + "loss": 0.9204, + "step": 28759 + }, + { + "epoch": 0.68, + "grad_norm": 2.72756149577781, + "learning_rate": 4.974910161924204e-06, + "loss": 1.0105, + "step": 28760 + }, + { + "epoch": 0.68, + "grad_norm": 1.9523317007467307, + "learning_rate": 4.974250472627771e-06, + "loss": 0.9921, + "step": 28761 + }, + { + "epoch": 0.68, + "grad_norm": 2.0257063769440533, + "learning_rate": 4.973590812592839e-06, + "loss": 1.0083, + "step": 28762 + }, + { + "epoch": 0.68, + "grad_norm": 1.7293346046220652, + "learning_rate": 4.9729311818232565e-06, + "loss": 0.9101, + "step": 28763 + }, + { + "epoch": 0.68, + "grad_norm": 1.9402300931665564, + "learning_rate": 4.97227158032286e-06, + "loss": 0.9843, + "step": 28764 + }, + { + "epoch": 0.68, + "grad_norm": 1.9290577831333078, + "learning_rate": 4.971612008095486e-06, + "loss": 0.9547, + "step": 28765 + }, + { + "epoch": 0.68, + "grad_norm": 1.963197345160801, + "learning_rate": 4.970952465144979e-06, + "loss": 1.0302, + "step": 28766 + }, + { + "epoch": 0.68, + "grad_norm": 2.059617906018719, + "learning_rate": 4.970292951475183e-06, + "loss": 0.988, + "step": 28767 + }, + { + "epoch": 0.68, + "grad_norm": 1.952678615223618, + "learning_rate": 4.969633467089934e-06, + "loss": 1.005, + "step": 28768 + }, + { + "epoch": 0.68, + "grad_norm": 2.041180475435839, + "learning_rate": 4.968974011993067e-06, + "loss": 1.0545, + "step": 28769 + }, + { + "epoch": 0.68, + "grad_norm": 2.6378393391645987, + "learning_rate": 4.968314586188427e-06, + "loss": 0.8921, + "step": 28770 + }, + { + "epoch": 0.68, + "grad_norm": 1.0362085852659382, + "learning_rate": 4.9676551896798565e-06, + "loss": 0.8788, + "step": 28771 + }, + { + "epoch": 0.68, + "grad_norm": 1.9659809757768574, + "learning_rate": 4.96699582247119e-06, + "loss": 0.9128, + "step": 28772 + }, + { + "epoch": 0.68, + "grad_norm": 2.035945463880862, + "learning_rate": 4.9663364845662635e-06, + "loss": 1.1239, + "step": 28773 + }, + { + "epoch": 0.68, + "grad_norm": 1.9951150311119012, + "learning_rate": 4.965677175968919e-06, + "loss": 0.9685, + "step": 28774 + }, + { + "epoch": 0.68, + "grad_norm": 1.9162662240875894, + "learning_rate": 4.965017896683001e-06, + "loss": 1.0124, + "step": 28775 + }, + { + "epoch": 0.68, + "grad_norm": 1.8878482373932388, + "learning_rate": 4.9643586467123416e-06, + "loss": 0.9472, + "step": 28776 + }, + { + "epoch": 0.68, + "grad_norm": 1.1487289688543547, + "learning_rate": 4.963699426060777e-06, + "loss": 0.8845, + "step": 28777 + }, + { + "epoch": 0.68, + "grad_norm": 2.204183413236907, + "learning_rate": 4.963040234732154e-06, + "loss": 0.9902, + "step": 28778 + }, + { + "epoch": 0.68, + "grad_norm": 2.344433954797157, + "learning_rate": 4.9623810727303e-06, + "loss": 1.0392, + "step": 28779 + }, + { + "epoch": 0.68, + "grad_norm": 2.0751128540486063, + "learning_rate": 4.961721940059062e-06, + "loss": 1.1058, + "step": 28780 + }, + { + "epoch": 0.68, + "grad_norm": 2.197902103005011, + "learning_rate": 4.961062836722271e-06, + "loss": 0.8143, + "step": 28781 + }, + { + "epoch": 0.68, + "grad_norm": 2.216365082317783, + "learning_rate": 4.960403762723771e-06, + "loss": 0.9705, + "step": 28782 + }, + { + "epoch": 0.68, + "grad_norm": 2.2024090404950294, + "learning_rate": 4.9597447180673916e-06, + "loss": 1.0862, + "step": 28783 + }, + { + "epoch": 0.68, + "grad_norm": 2.0725691682694705, + "learning_rate": 4.959085702756979e-06, + "loss": 0.9445, + "step": 28784 + }, + { + "epoch": 0.68, + "grad_norm": 2.0315184403786084, + "learning_rate": 4.958426716796364e-06, + "loss": 0.9547, + "step": 28785 + }, + { + "epoch": 0.68, + "grad_norm": 2.195777387475315, + "learning_rate": 4.957767760189382e-06, + "loss": 1.048, + "step": 28786 + }, + { + "epoch": 0.68, + "grad_norm": 1.0428192957136408, + "learning_rate": 4.957108832939872e-06, + "loss": 0.9946, + "step": 28787 + }, + { + "epoch": 0.68, + "grad_norm": 1.8816388127902481, + "learning_rate": 4.956449935051675e-06, + "loss": 1.0435, + "step": 28788 + }, + { + "epoch": 0.68, + "grad_norm": 2.257985901608724, + "learning_rate": 4.9557910665286236e-06, + "loss": 0.847, + "step": 28789 + }, + { + "epoch": 0.68, + "grad_norm": 2.040519561100066, + "learning_rate": 4.955132227374548e-06, + "loss": 1.0485, + "step": 28790 + }, + { + "epoch": 0.68, + "grad_norm": 2.0150245233072988, + "learning_rate": 4.954473417593292e-06, + "loss": 0.9538, + "step": 28791 + }, + { + "epoch": 0.68, + "grad_norm": 1.8429135718833949, + "learning_rate": 4.953814637188692e-06, + "loss": 0.9614, + "step": 28792 + }, + { + "epoch": 0.68, + "grad_norm": 2.104357545400092, + "learning_rate": 4.95315588616458e-06, + "loss": 0.9365, + "step": 28793 + }, + { + "epoch": 0.68, + "grad_norm": 1.9722008715971167, + "learning_rate": 4.95249716452479e-06, + "loss": 0.9544, + "step": 28794 + }, + { + "epoch": 0.68, + "grad_norm": 2.0372693992424282, + "learning_rate": 4.951838472273158e-06, + "loss": 0.9949, + "step": 28795 + }, + { + "epoch": 0.68, + "grad_norm": 2.4731815449278702, + "learning_rate": 4.951179809413526e-06, + "loss": 0.9229, + "step": 28796 + }, + { + "epoch": 0.68, + "grad_norm": 2.268771547723871, + "learning_rate": 4.950521175949722e-06, + "loss": 1.1415, + "step": 28797 + }, + { + "epoch": 0.68, + "grad_norm": 2.1995308841498185, + "learning_rate": 4.9498625718855786e-06, + "loss": 1.0125, + "step": 28798 + }, + { + "epoch": 0.68, + "grad_norm": 1.9529538952861432, + "learning_rate": 4.94920399722494e-06, + "loss": 0.993, + "step": 28799 + }, + { + "epoch": 0.68, + "grad_norm": 3.6062082988902824, + "learning_rate": 4.948545451971629e-06, + "loss": 1.1039, + "step": 28800 + }, + { + "epoch": 0.68, + "grad_norm": 2.103005720341363, + "learning_rate": 4.94788693612949e-06, + "loss": 1.082, + "step": 28801 + }, + { + "epoch": 0.68, + "grad_norm": 1.8370579790424348, + "learning_rate": 4.947228449702348e-06, + "loss": 1.0442, + "step": 28802 + }, + { + "epoch": 0.68, + "grad_norm": 3.0389713483449246, + "learning_rate": 4.946569992694047e-06, + "loss": 0.9963, + "step": 28803 + }, + { + "epoch": 0.68, + "grad_norm": 2.210748968367596, + "learning_rate": 4.945911565108411e-06, + "loss": 1.1228, + "step": 28804 + }, + { + "epoch": 0.68, + "grad_norm": 2.075807439056867, + "learning_rate": 4.945253166949282e-06, + "loss": 1.0158, + "step": 28805 + }, + { + "epoch": 0.68, + "grad_norm": 2.083432406743263, + "learning_rate": 4.944594798220489e-06, + "loss": 0.9407, + "step": 28806 + }, + { + "epoch": 0.68, + "grad_norm": 1.0783392703630754, + "learning_rate": 4.943936458925862e-06, + "loss": 0.8845, + "step": 28807 + }, + { + "epoch": 0.68, + "grad_norm": 2.1346549420312884, + "learning_rate": 4.943278149069237e-06, + "loss": 1.0145, + "step": 28808 + }, + { + "epoch": 0.68, + "grad_norm": 2.0362495337450826, + "learning_rate": 4.942619868654453e-06, + "loss": 1.168, + "step": 28809 + }, + { + "epoch": 0.68, + "grad_norm": 2.0547399344208324, + "learning_rate": 4.9419616176853355e-06, + "loss": 1.0654, + "step": 28810 + }, + { + "epoch": 0.68, + "grad_norm": 1.869923758650262, + "learning_rate": 4.941303396165716e-06, + "loss": 1.0032, + "step": 28811 + }, + { + "epoch": 0.68, + "grad_norm": 3.754904754262046, + "learning_rate": 4.940645204099429e-06, + "loss": 0.8826, + "step": 28812 + }, + { + "epoch": 0.68, + "grad_norm": 1.98505962580766, + "learning_rate": 4.939987041490312e-06, + "loss": 0.9454, + "step": 28813 + }, + { + "epoch": 0.68, + "grad_norm": 2.043963234937219, + "learning_rate": 4.939328908342192e-06, + "loss": 0.9297, + "step": 28814 + }, + { + "epoch": 0.68, + "grad_norm": 2.1886873499067745, + "learning_rate": 4.938670804658897e-06, + "loss": 1.0244, + "step": 28815 + }, + { + "epoch": 0.68, + "grad_norm": 1.9416541065097528, + "learning_rate": 4.938012730444264e-06, + "loss": 0.9437, + "step": 28816 + }, + { + "epoch": 0.68, + "grad_norm": 2.875520075967029, + "learning_rate": 4.937354685702126e-06, + "loss": 1.0162, + "step": 28817 + }, + { + "epoch": 0.68, + "grad_norm": 1.9849051076392954, + "learning_rate": 4.936696670436313e-06, + "loss": 1.0067, + "step": 28818 + }, + { + "epoch": 0.68, + "grad_norm": 1.9400341840844173, + "learning_rate": 4.93603868465065e-06, + "loss": 0.9428, + "step": 28819 + }, + { + "epoch": 0.68, + "grad_norm": 2.003995335933632, + "learning_rate": 4.935380728348978e-06, + "loss": 0.9679, + "step": 28820 + }, + { + "epoch": 0.68, + "grad_norm": 1.8861754469508258, + "learning_rate": 4.9347228015351185e-06, + "loss": 0.9395, + "step": 28821 + }, + { + "epoch": 0.68, + "grad_norm": 2.6329798838888117, + "learning_rate": 4.934064904212911e-06, + "loss": 0.954, + "step": 28822 + }, + { + "epoch": 0.68, + "grad_norm": 2.429934579641841, + "learning_rate": 4.9334070363861765e-06, + "loss": 1.0214, + "step": 28823 + }, + { + "epoch": 0.68, + "grad_norm": 2.2439706579680014, + "learning_rate": 4.932749198058755e-06, + "loss": 0.9964, + "step": 28824 + }, + { + "epoch": 0.68, + "grad_norm": 2.0380605783790378, + "learning_rate": 4.932091389234467e-06, + "loss": 0.9517, + "step": 28825 + }, + { + "epoch": 0.68, + "grad_norm": 1.8926174908442284, + "learning_rate": 4.931433609917149e-06, + "loss": 0.9298, + "step": 28826 + }, + { + "epoch": 0.68, + "grad_norm": 1.9719743550958768, + "learning_rate": 4.930775860110636e-06, + "loss": 1.0641, + "step": 28827 + }, + { + "epoch": 0.68, + "grad_norm": 1.9386566044868772, + "learning_rate": 4.930118139818742e-06, + "loss": 0.9596, + "step": 28828 + }, + { + "epoch": 0.68, + "grad_norm": 1.9429436943561524, + "learning_rate": 4.929460449045307e-06, + "loss": 0.997, + "step": 28829 + }, + { + "epoch": 0.68, + "grad_norm": 1.0946164327799532, + "learning_rate": 4.928802787794158e-06, + "loss": 0.9515, + "step": 28830 + }, + { + "epoch": 0.68, + "grad_norm": 1.9959025405724402, + "learning_rate": 4.9281451560691325e-06, + "loss": 0.9534, + "step": 28831 + }, + { + "epoch": 0.68, + "grad_norm": 1.868685924770543, + "learning_rate": 4.927487553874044e-06, + "loss": 0.9489, + "step": 28832 + }, + { + "epoch": 0.68, + "grad_norm": 1.966605687153668, + "learning_rate": 4.926829981212728e-06, + "loss": 1.0096, + "step": 28833 + }, + { + "epoch": 0.68, + "grad_norm": 2.072674212647965, + "learning_rate": 4.9261724380890195e-06, + "loss": 0.9733, + "step": 28834 + }, + { + "epoch": 0.68, + "grad_norm": 2.145462883638162, + "learning_rate": 4.92551492450674e-06, + "loss": 0.9942, + "step": 28835 + }, + { + "epoch": 0.68, + "grad_norm": 1.932864757109339, + "learning_rate": 4.924857440469716e-06, + "loss": 0.8527, + "step": 28836 + }, + { + "epoch": 0.68, + "grad_norm": 1.8732827342509195, + "learning_rate": 4.9241999859817794e-06, + "loss": 0.9312, + "step": 28837 + }, + { + "epoch": 0.68, + "grad_norm": 2.0312802698786827, + "learning_rate": 4.923542561046759e-06, + "loss": 1.1794, + "step": 28838 + }, + { + "epoch": 0.68, + "grad_norm": 1.7855230652723426, + "learning_rate": 4.9228851656684794e-06, + "loss": 0.9925, + "step": 28839 + }, + { + "epoch": 0.68, + "grad_norm": 1.9706811014037406, + "learning_rate": 4.922227799850772e-06, + "loss": 0.9597, + "step": 28840 + }, + { + "epoch": 0.68, + "grad_norm": 2.0770217726749403, + "learning_rate": 4.921570463597464e-06, + "loss": 1.1488, + "step": 28841 + }, + { + "epoch": 0.68, + "grad_norm": 1.8014381645276594, + "learning_rate": 4.920913156912374e-06, + "loss": 0.8453, + "step": 28842 + }, + { + "epoch": 0.68, + "grad_norm": 2.0440085773090777, + "learning_rate": 4.920255879799338e-06, + "loss": 1.0176, + "step": 28843 + }, + { + "epoch": 0.68, + "grad_norm": 1.9661301604597834, + "learning_rate": 4.919598632262183e-06, + "loss": 0.9787, + "step": 28844 + }, + { + "epoch": 0.68, + "grad_norm": 2.248653226624898, + "learning_rate": 4.918941414304733e-06, + "loss": 1.0771, + "step": 28845 + }, + { + "epoch": 0.68, + "grad_norm": 1.9607766136958862, + "learning_rate": 4.918284225930812e-06, + "loss": 0.9887, + "step": 28846 + }, + { + "epoch": 0.68, + "grad_norm": 1.9108923798331754, + "learning_rate": 4.917627067144248e-06, + "loss": 1.0151, + "step": 28847 + }, + { + "epoch": 0.68, + "grad_norm": 1.8514658841344933, + "learning_rate": 4.916969937948874e-06, + "loss": 0.9398, + "step": 28848 + }, + { + "epoch": 0.68, + "grad_norm": 1.8389166457118735, + "learning_rate": 4.916312838348508e-06, + "loss": 0.9586, + "step": 28849 + }, + { + "epoch": 0.68, + "grad_norm": 2.5190356145959565, + "learning_rate": 4.915655768346975e-06, + "loss": 0.9879, + "step": 28850 + }, + { + "epoch": 0.68, + "grad_norm": 1.9419774981839675, + "learning_rate": 4.914998727948105e-06, + "loss": 1.1151, + "step": 28851 + }, + { + "epoch": 0.68, + "grad_norm": 1.889339591755828, + "learning_rate": 4.9143417171557255e-06, + "loss": 1.0949, + "step": 28852 + }, + { + "epoch": 0.68, + "grad_norm": 2.066002653958542, + "learning_rate": 4.913684735973657e-06, + "loss": 1.0203, + "step": 28853 + }, + { + "epoch": 0.68, + "grad_norm": 1.8889761090934936, + "learning_rate": 4.913027784405724e-06, + "loss": 1.0367, + "step": 28854 + }, + { + "epoch": 0.68, + "grad_norm": 2.144284675565926, + "learning_rate": 4.912370862455757e-06, + "loss": 0.9934, + "step": 28855 + }, + { + "epoch": 0.68, + "grad_norm": 1.9420370614824394, + "learning_rate": 4.911713970127574e-06, + "loss": 1.0415, + "step": 28856 + }, + { + "epoch": 0.68, + "grad_norm": 2.221904054462991, + "learning_rate": 4.911057107425007e-06, + "loss": 1.0142, + "step": 28857 + }, + { + "epoch": 0.68, + "grad_norm": 2.1443651330211027, + "learning_rate": 4.910400274351873e-06, + "loss": 0.8943, + "step": 28858 + }, + { + "epoch": 0.68, + "grad_norm": 1.8298201788851651, + "learning_rate": 4.909743470912004e-06, + "loss": 0.9664, + "step": 28859 + }, + { + "epoch": 0.68, + "grad_norm": 2.024200681002486, + "learning_rate": 4.909086697109215e-06, + "loss": 0.8893, + "step": 28860 + }, + { + "epoch": 0.68, + "grad_norm": 1.7830967542378013, + "learning_rate": 4.90842995294734e-06, + "loss": 0.85, + "step": 28861 + }, + { + "epoch": 0.68, + "grad_norm": 1.985465688657303, + "learning_rate": 4.907773238430198e-06, + "loss": 0.9561, + "step": 28862 + }, + { + "epoch": 0.68, + "grad_norm": 1.9464600592487153, + "learning_rate": 4.907116553561608e-06, + "loss": 1.0025, + "step": 28863 + }, + { + "epoch": 0.68, + "grad_norm": 1.7702256054455627, + "learning_rate": 4.906459898345397e-06, + "loss": 0.8853, + "step": 28864 + }, + { + "epoch": 0.68, + "grad_norm": 1.9184788494903489, + "learning_rate": 4.905803272785394e-06, + "loss": 0.9222, + "step": 28865 + }, + { + "epoch": 0.68, + "grad_norm": 1.949566835607361, + "learning_rate": 4.905146676885417e-06, + "loss": 1.0264, + "step": 28866 + }, + { + "epoch": 0.68, + "grad_norm": 2.0305419541785827, + "learning_rate": 4.904490110649285e-06, + "loss": 0.9728, + "step": 28867 + }, + { + "epoch": 0.68, + "grad_norm": 2.040964149380384, + "learning_rate": 4.903833574080825e-06, + "loss": 1.0568, + "step": 28868 + }, + { + "epoch": 0.68, + "grad_norm": 2.1842535344632026, + "learning_rate": 4.903177067183863e-06, + "loss": 1.0094, + "step": 28869 + }, + { + "epoch": 0.68, + "grad_norm": 2.0744983374997257, + "learning_rate": 4.902520589962218e-06, + "loss": 0.9782, + "step": 28870 + }, + { + "epoch": 0.68, + "grad_norm": 1.0831093573928228, + "learning_rate": 4.9018641424197075e-06, + "loss": 0.9416, + "step": 28871 + }, + { + "epoch": 0.68, + "grad_norm": 1.9371014310535823, + "learning_rate": 4.901207724560158e-06, + "loss": 0.9704, + "step": 28872 + }, + { + "epoch": 0.68, + "grad_norm": 1.9764699342928715, + "learning_rate": 4.900551336387395e-06, + "loss": 0.9458, + "step": 28873 + }, + { + "epoch": 0.68, + "grad_norm": 1.870662584530651, + "learning_rate": 4.899894977905239e-06, + "loss": 1.0033, + "step": 28874 + }, + { + "epoch": 0.68, + "grad_norm": 1.887933476649609, + "learning_rate": 4.8992386491175025e-06, + "loss": 1.0798, + "step": 28875 + }, + { + "epoch": 0.68, + "grad_norm": 2.0640747741797028, + "learning_rate": 4.898582350028015e-06, + "loss": 1.0336, + "step": 28876 + }, + { + "epoch": 0.68, + "grad_norm": 2.142666578423039, + "learning_rate": 4.897926080640599e-06, + "loss": 1.0003, + "step": 28877 + }, + { + "epoch": 0.68, + "grad_norm": 2.005824746898103, + "learning_rate": 4.897269840959072e-06, + "loss": 1.1063, + "step": 28878 + }, + { + "epoch": 0.68, + "grad_norm": 1.766306884058466, + "learning_rate": 4.896613630987254e-06, + "loss": 0.9381, + "step": 28879 + }, + { + "epoch": 0.68, + "grad_norm": 1.9286083469388593, + "learning_rate": 4.89595745072897e-06, + "loss": 1.0043, + "step": 28880 + }, + { + "epoch": 0.68, + "grad_norm": 1.0990065081082085, + "learning_rate": 4.8953013001880324e-06, + "loss": 0.9134, + "step": 28881 + }, + { + "epoch": 0.68, + "grad_norm": 2.0167659067355466, + "learning_rate": 4.894645179368272e-06, + "loss": 0.9095, + "step": 28882 + }, + { + "epoch": 0.68, + "grad_norm": 1.8174778244963088, + "learning_rate": 4.8939890882735e-06, + "loss": 0.9749, + "step": 28883 + }, + { + "epoch": 0.68, + "grad_norm": 1.9900471003277702, + "learning_rate": 4.893333026907543e-06, + "loss": 1.0239, + "step": 28884 + }, + { + "epoch": 0.68, + "grad_norm": 2.1509797436332057, + "learning_rate": 4.8926769952742145e-06, + "loss": 0.8784, + "step": 28885 + }, + { + "epoch": 0.68, + "grad_norm": 2.4328882891572188, + "learning_rate": 4.892020993377341e-06, + "loss": 1.0785, + "step": 28886 + }, + { + "epoch": 0.68, + "grad_norm": 1.1345122470155402, + "learning_rate": 4.891365021220739e-06, + "loss": 0.9457, + "step": 28887 + }, + { + "epoch": 0.68, + "grad_norm": 1.0860924531954355, + "learning_rate": 4.8907090788082225e-06, + "loss": 0.9924, + "step": 28888 + }, + { + "epoch": 0.68, + "grad_norm": 2.0079830683764355, + "learning_rate": 4.890053166143617e-06, + "loss": 1.1425, + "step": 28889 + }, + { + "epoch": 0.68, + "grad_norm": 1.9840472592061584, + "learning_rate": 4.889397283230742e-06, + "loss": 0.8823, + "step": 28890 + }, + { + "epoch": 0.68, + "grad_norm": 1.131502616684971, + "learning_rate": 4.888741430073414e-06, + "loss": 0.8847, + "step": 28891 + }, + { + "epoch": 0.68, + "grad_norm": 2.1584029987612525, + "learning_rate": 4.888085606675449e-06, + "loss": 0.8453, + "step": 28892 + }, + { + "epoch": 0.68, + "grad_norm": 2.039102040732783, + "learning_rate": 4.8874298130406665e-06, + "loss": 1.0359, + "step": 28893 + }, + { + "epoch": 0.68, + "grad_norm": 1.7813341162452296, + "learning_rate": 4.886774049172891e-06, + "loss": 0.978, + "step": 28894 + }, + { + "epoch": 0.68, + "grad_norm": 1.904027145552272, + "learning_rate": 4.886118315075935e-06, + "loss": 0.869, + "step": 28895 + }, + { + "epoch": 0.68, + "grad_norm": 2.0945097781092405, + "learning_rate": 4.885462610753614e-06, + "loss": 1.1321, + "step": 28896 + }, + { + "epoch": 0.68, + "grad_norm": 2.100643726996654, + "learning_rate": 4.884806936209749e-06, + "loss": 0.9715, + "step": 28897 + }, + { + "epoch": 0.68, + "grad_norm": 1.907319751612394, + "learning_rate": 4.884151291448161e-06, + "loss": 0.9167, + "step": 28898 + }, + { + "epoch": 0.68, + "grad_norm": 1.9185631226424373, + "learning_rate": 4.883495676472665e-06, + "loss": 1.054, + "step": 28899 + }, + { + "epoch": 0.68, + "grad_norm": 2.0004442409488905, + "learning_rate": 4.882840091287072e-06, + "loss": 0.9646, + "step": 28900 + }, + { + "epoch": 0.68, + "grad_norm": 1.9914685955970775, + "learning_rate": 4.882184535895208e-06, + "loss": 1.0007, + "step": 28901 + }, + { + "epoch": 0.68, + "grad_norm": 2.1529579955804046, + "learning_rate": 4.881529010300882e-06, + "loss": 0.8101, + "step": 28902 + }, + { + "epoch": 0.68, + "grad_norm": 1.1182170976113488, + "learning_rate": 4.880873514507919e-06, + "loss": 0.9871, + "step": 28903 + }, + { + "epoch": 0.68, + "grad_norm": 2.000652793314451, + "learning_rate": 4.880218048520127e-06, + "loss": 0.994, + "step": 28904 + }, + { + "epoch": 0.68, + "grad_norm": 1.8541112561149804, + "learning_rate": 4.87956261234133e-06, + "loss": 0.8287, + "step": 28905 + }, + { + "epoch": 0.68, + "grad_norm": 2.6127846696672288, + "learning_rate": 4.878907205975336e-06, + "loss": 0.9065, + "step": 28906 + }, + { + "epoch": 0.68, + "grad_norm": 2.029347777875561, + "learning_rate": 4.87825182942597e-06, + "loss": 1.03, + "step": 28907 + }, + { + "epoch": 0.68, + "grad_norm": 2.0082800435378165, + "learning_rate": 4.877596482697045e-06, + "loss": 0.9681, + "step": 28908 + }, + { + "epoch": 0.68, + "grad_norm": 2.0747251877918345, + "learning_rate": 4.87694116579237e-06, + "loss": 1.0157, + "step": 28909 + }, + { + "epoch": 0.68, + "grad_norm": 2.084042755116324, + "learning_rate": 4.876285878715764e-06, + "loss": 1.0188, + "step": 28910 + }, + { + "epoch": 0.68, + "grad_norm": 1.1033182157482038, + "learning_rate": 4.875630621471045e-06, + "loss": 0.9483, + "step": 28911 + }, + { + "epoch": 0.68, + "grad_norm": 2.056925378958801, + "learning_rate": 4.874975394062034e-06, + "loss": 1.0316, + "step": 28912 + }, + { + "epoch": 0.68, + "grad_norm": 2.0623784447110793, + "learning_rate": 4.874320196492533e-06, + "loss": 1.1468, + "step": 28913 + }, + { + "epoch": 0.68, + "grad_norm": 1.091125864693669, + "learning_rate": 4.873665028766361e-06, + "loss": 0.9231, + "step": 28914 + }, + { + "epoch": 0.68, + "grad_norm": 1.0792936599642486, + "learning_rate": 4.873009890887338e-06, + "loss": 1.0799, + "step": 28915 + }, + { + "epoch": 0.68, + "grad_norm": 2.0914108178794577, + "learning_rate": 4.872354782859274e-06, + "loss": 1.0502, + "step": 28916 + }, + { + "epoch": 0.68, + "grad_norm": 2.0250260199990526, + "learning_rate": 4.8716997046859816e-06, + "loss": 1.0566, + "step": 28917 + }, + { + "epoch": 0.68, + "grad_norm": 2.2556819257673806, + "learning_rate": 4.871044656371276e-06, + "loss": 0.9682, + "step": 28918 + }, + { + "epoch": 0.68, + "grad_norm": 2.0536167686184634, + "learning_rate": 4.870389637918976e-06, + "loss": 1.0671, + "step": 28919 + }, + { + "epoch": 0.68, + "grad_norm": 1.9897304663015198, + "learning_rate": 4.869734649332892e-06, + "loss": 0.8646, + "step": 28920 + }, + { + "epoch": 0.68, + "grad_norm": 1.7648761508138615, + "learning_rate": 4.8690796906168334e-06, + "loss": 0.7623, + "step": 28921 + }, + { + "epoch": 0.68, + "grad_norm": 2.140889639467548, + "learning_rate": 4.868424761774622e-06, + "loss": 0.9628, + "step": 28922 + }, + { + "epoch": 0.68, + "grad_norm": 2.1978312784706606, + "learning_rate": 4.867769862810061e-06, + "loss": 1.1551, + "step": 28923 + }, + { + "epoch": 0.68, + "grad_norm": 2.0821352808783047, + "learning_rate": 4.8671149937269694e-06, + "loss": 0.8867, + "step": 28924 + }, + { + "epoch": 0.68, + "grad_norm": 2.007515327980676, + "learning_rate": 4.866460154529163e-06, + "loss": 0.9442, + "step": 28925 + }, + { + "epoch": 0.68, + "grad_norm": 1.9154606436693766, + "learning_rate": 4.86580534522045e-06, + "loss": 1.0247, + "step": 28926 + }, + { + "epoch": 0.68, + "grad_norm": 2.0701835106307174, + "learning_rate": 4.86515056580464e-06, + "loss": 0.9128, + "step": 28927 + }, + { + "epoch": 0.68, + "grad_norm": 2.2199834611174336, + "learning_rate": 4.864495816285549e-06, + "loss": 1.0177, + "step": 28928 + }, + { + "epoch": 0.68, + "grad_norm": 2.0627954988861856, + "learning_rate": 4.863841096666998e-06, + "loss": 1.1888, + "step": 28929 + }, + { + "epoch": 0.68, + "grad_norm": 2.756959797265547, + "learning_rate": 4.863186406952782e-06, + "loss": 1.0915, + "step": 28930 + }, + { + "epoch": 0.68, + "grad_norm": 2.3101136002072256, + "learning_rate": 4.862531747146722e-06, + "loss": 0.9052, + "step": 28931 + }, + { + "epoch": 0.68, + "grad_norm": 1.9003533985024232, + "learning_rate": 4.861877117252627e-06, + "loss": 0.9216, + "step": 28932 + }, + { + "epoch": 0.68, + "grad_norm": 2.2913435947567526, + "learning_rate": 4.861222517274319e-06, + "loss": 0.963, + "step": 28933 + }, + { + "epoch": 0.68, + "grad_norm": 1.8066629687422526, + "learning_rate": 4.860567947215592e-06, + "loss": 0.947, + "step": 28934 + }, + { + "epoch": 0.68, + "grad_norm": 1.9170960739716183, + "learning_rate": 4.8599134070802656e-06, + "loss": 0.9482, + "step": 28935 + }, + { + "epoch": 0.68, + "grad_norm": 2.074281398642921, + "learning_rate": 4.859258896872155e-06, + "loss": 1.0156, + "step": 28936 + }, + { + "epoch": 0.68, + "grad_norm": 2.2137334252725265, + "learning_rate": 4.858604416595063e-06, + "loss": 0.9911, + "step": 28937 + }, + { + "epoch": 0.68, + "grad_norm": 2.0989956934781295, + "learning_rate": 4.857949966252806e-06, + "loss": 0.8494, + "step": 28938 + }, + { + "epoch": 0.68, + "grad_norm": 2.0872953876064377, + "learning_rate": 4.85729554584919e-06, + "loss": 0.8439, + "step": 28939 + }, + { + "epoch": 0.68, + "grad_norm": 2.440039031194322, + "learning_rate": 4.856641155388031e-06, + "loss": 0.9268, + "step": 28940 + }, + { + "epoch": 0.68, + "grad_norm": 1.9918152340367181, + "learning_rate": 4.8559867948731325e-06, + "loss": 0.8206, + "step": 28941 + }, + { + "epoch": 0.68, + "grad_norm": 1.9088400394733127, + "learning_rate": 4.85533246430831e-06, + "loss": 1.0607, + "step": 28942 + }, + { + "epoch": 0.68, + "grad_norm": 1.8981728613994755, + "learning_rate": 4.854678163697371e-06, + "loss": 0.9342, + "step": 28943 + }, + { + "epoch": 0.68, + "grad_norm": 1.928109257078703, + "learning_rate": 4.854023893044121e-06, + "loss": 0.9459, + "step": 28944 + }, + { + "epoch": 0.68, + "grad_norm": 2.295025476947772, + "learning_rate": 4.853369652352373e-06, + "loss": 1.0258, + "step": 28945 + }, + { + "epoch": 0.68, + "grad_norm": 2.164042928177677, + "learning_rate": 4.85271544162594e-06, + "loss": 0.9052, + "step": 28946 + }, + { + "epoch": 0.68, + "grad_norm": 1.0929194029887894, + "learning_rate": 4.8520612608686265e-06, + "loss": 0.9138, + "step": 28947 + }, + { + "epoch": 0.68, + "grad_norm": 1.1116559969824582, + "learning_rate": 4.8514071100842395e-06, + "loss": 0.9007, + "step": 28948 + }, + { + "epoch": 0.68, + "grad_norm": 1.9042688476505136, + "learning_rate": 4.85075298927659e-06, + "loss": 0.9351, + "step": 28949 + }, + { + "epoch": 0.68, + "grad_norm": 2.4447837559205037, + "learning_rate": 4.850098898449491e-06, + "loss": 0.9882, + "step": 28950 + }, + { + "epoch": 0.68, + "grad_norm": 1.8128587785429375, + "learning_rate": 4.849444837606747e-06, + "loss": 0.9982, + "step": 28951 + }, + { + "epoch": 0.68, + "grad_norm": 2.1637294732307915, + "learning_rate": 4.848790806752162e-06, + "loss": 0.9626, + "step": 28952 + }, + { + "epoch": 0.68, + "grad_norm": 3.1605832982979005, + "learning_rate": 4.848136805889546e-06, + "loss": 0.9938, + "step": 28953 + }, + { + "epoch": 0.68, + "grad_norm": 2.040124724302968, + "learning_rate": 4.8474828350227135e-06, + "loss": 1.1, + "step": 28954 + }, + { + "epoch": 0.68, + "grad_norm": 1.9561634137060022, + "learning_rate": 4.846828894155468e-06, + "loss": 0.991, + "step": 28955 + }, + { + "epoch": 0.68, + "grad_norm": 2.410966431244978, + "learning_rate": 4.8461749832916104e-06, + "loss": 1.0196, + "step": 28956 + }, + { + "epoch": 0.68, + "grad_norm": 1.052155081736127, + "learning_rate": 4.845521102434959e-06, + "loss": 0.9656, + "step": 28957 + }, + { + "epoch": 0.68, + "grad_norm": 2.055564654612906, + "learning_rate": 4.844867251589311e-06, + "loss": 0.9137, + "step": 28958 + }, + { + "epoch": 0.68, + "grad_norm": 2.0143981366449966, + "learning_rate": 4.844213430758483e-06, + "loss": 1.0151, + "step": 28959 + }, + { + "epoch": 0.68, + "grad_norm": 1.9528038250779998, + "learning_rate": 4.8435596399462725e-06, + "loss": 1.0735, + "step": 28960 + }, + { + "epoch": 0.68, + "grad_norm": 2.6194294730047636, + "learning_rate": 4.842905879156493e-06, + "loss": 0.9024, + "step": 28961 + }, + { + "epoch": 0.68, + "grad_norm": 2.1600139536130265, + "learning_rate": 4.842252148392945e-06, + "loss": 0.9958, + "step": 28962 + }, + { + "epoch": 0.68, + "grad_norm": 1.6665978059460345, + "learning_rate": 4.841598447659443e-06, + "loss": 0.8697, + "step": 28963 + }, + { + "epoch": 0.68, + "grad_norm": 2.8320341390562116, + "learning_rate": 4.840944776959786e-06, + "loss": 1.101, + "step": 28964 + }, + { + "epoch": 0.68, + "grad_norm": 2.007559814444218, + "learning_rate": 4.840291136297779e-06, + "loss": 1.0797, + "step": 28965 + }, + { + "epoch": 0.68, + "grad_norm": 2.119105657165959, + "learning_rate": 4.83963752567723e-06, + "loss": 0.9597, + "step": 28966 + }, + { + "epoch": 0.68, + "grad_norm": 1.9506400776665462, + "learning_rate": 4.838983945101949e-06, + "loss": 0.9988, + "step": 28967 + }, + { + "epoch": 0.68, + "grad_norm": 1.9656855067475612, + "learning_rate": 4.8383303945757374e-06, + "loss": 0.9989, + "step": 28968 + }, + { + "epoch": 0.68, + "grad_norm": 1.9472982001306087, + "learning_rate": 4.837676874102397e-06, + "loss": 1.1094, + "step": 28969 + }, + { + "epoch": 0.68, + "grad_norm": 2.0516512920076115, + "learning_rate": 4.837023383685736e-06, + "loss": 0.9702, + "step": 28970 + }, + { + "epoch": 0.68, + "grad_norm": 2.01247353832316, + "learning_rate": 4.836369923329564e-06, + "loss": 0.889, + "step": 28971 + }, + { + "epoch": 0.68, + "grad_norm": 2.039199467558624, + "learning_rate": 4.83571649303768e-06, + "loss": 0.9788, + "step": 28972 + }, + { + "epoch": 0.68, + "grad_norm": 1.8375233787240246, + "learning_rate": 4.835063092813886e-06, + "loss": 1.0079, + "step": 28973 + }, + { + "epoch": 0.68, + "grad_norm": 1.9378956198075281, + "learning_rate": 4.83440972266199e-06, + "loss": 0.8687, + "step": 28974 + }, + { + "epoch": 0.68, + "grad_norm": 1.7263706586703393, + "learning_rate": 4.833756382585799e-06, + "loss": 1.0528, + "step": 28975 + }, + { + "epoch": 0.68, + "grad_norm": 1.9342762342287947, + "learning_rate": 4.8331030725891145e-06, + "loss": 1.0859, + "step": 28976 + }, + { + "epoch": 0.68, + "grad_norm": 1.997906163580194, + "learning_rate": 4.832449792675737e-06, + "loss": 1.0318, + "step": 28977 + }, + { + "epoch": 0.68, + "grad_norm": 2.202022829089422, + "learning_rate": 4.831796542849476e-06, + "loss": 0.9589, + "step": 28978 + }, + { + "epoch": 0.68, + "grad_norm": 1.6900350042709036, + "learning_rate": 4.831143323114127e-06, + "loss": 0.9185, + "step": 28979 + }, + { + "epoch": 0.68, + "grad_norm": 1.979240063497917, + "learning_rate": 4.8304901334735035e-06, + "loss": 0.9856, + "step": 28980 + }, + { + "epoch": 0.68, + "grad_norm": 1.8601600289932236, + "learning_rate": 4.8298369739313975e-06, + "loss": 0.914, + "step": 28981 + }, + { + "epoch": 0.68, + "grad_norm": 1.8611668159580397, + "learning_rate": 4.829183844491623e-06, + "loss": 1.0161, + "step": 28982 + }, + { + "epoch": 0.68, + "grad_norm": 2.0338141906303835, + "learning_rate": 4.828530745157972e-06, + "loss": 0.9612, + "step": 28983 + }, + { + "epoch": 0.68, + "grad_norm": 2.30698278665787, + "learning_rate": 4.827877675934256e-06, + "loss": 0.9051, + "step": 28984 + }, + { + "epoch": 0.68, + "grad_norm": 3.096330915763256, + "learning_rate": 4.827224636824275e-06, + "loss": 1.0137, + "step": 28985 + }, + { + "epoch": 0.68, + "grad_norm": 1.912135987832069, + "learning_rate": 4.826571627831824e-06, + "loss": 1.0172, + "step": 28986 + }, + { + "epoch": 0.68, + "grad_norm": 2.3470948191334955, + "learning_rate": 4.825918648960712e-06, + "loss": 0.9963, + "step": 28987 + }, + { + "epoch": 0.68, + "grad_norm": 2.1618757788316447, + "learning_rate": 4.825265700214744e-06, + "loss": 1.0044, + "step": 28988 + }, + { + "epoch": 0.68, + "grad_norm": 2.0814918147805224, + "learning_rate": 4.824612781597717e-06, + "loss": 0.998, + "step": 28989 + }, + { + "epoch": 0.68, + "grad_norm": 2.435698615286186, + "learning_rate": 4.823959893113429e-06, + "loss": 1.0688, + "step": 28990 + }, + { + "epoch": 0.68, + "grad_norm": 2.0740062386771108, + "learning_rate": 4.823307034765686e-06, + "loss": 1.0152, + "step": 28991 + }, + { + "epoch": 0.68, + "grad_norm": 2.2701371400012587, + "learning_rate": 4.822654206558292e-06, + "loss": 1.1029, + "step": 28992 + }, + { + "epoch": 0.68, + "grad_norm": 1.0435281623529997, + "learning_rate": 4.822001408495043e-06, + "loss": 0.8973, + "step": 28993 + }, + { + "epoch": 0.68, + "grad_norm": 2.2310692914845296, + "learning_rate": 4.821348640579738e-06, + "loss": 0.9919, + "step": 28994 + }, + { + "epoch": 0.68, + "grad_norm": 2.108834350251038, + "learning_rate": 4.820695902816181e-06, + "loss": 1.1328, + "step": 28995 + }, + { + "epoch": 0.68, + "grad_norm": 1.7462850942940997, + "learning_rate": 4.820043195208176e-06, + "loss": 1.0121, + "step": 28996 + }, + { + "epoch": 0.68, + "grad_norm": 1.867778111814677, + "learning_rate": 4.819390517759519e-06, + "loss": 0.9739, + "step": 28997 + }, + { + "epoch": 0.68, + "grad_norm": 1.0886754818727566, + "learning_rate": 4.818737870474007e-06, + "loss": 1.0054, + "step": 28998 + }, + { + "epoch": 0.68, + "grad_norm": 1.9746576019971578, + "learning_rate": 4.818085253355443e-06, + "loss": 1.0568, + "step": 28999 + }, + { + "epoch": 0.68, + "grad_norm": 2.2383193330911917, + "learning_rate": 4.8174326664076324e-06, + "loss": 0.8918, + "step": 29000 + }, + { + "epoch": 0.68, + "grad_norm": 2.038182488968428, + "learning_rate": 4.816780109634368e-06, + "loss": 0.8927, + "step": 29001 + }, + { + "epoch": 0.68, + "grad_norm": 1.7679945193153996, + "learning_rate": 4.816127583039448e-06, + "loss": 0.8989, + "step": 29002 + }, + { + "epoch": 0.68, + "grad_norm": 1.0997362783526725, + "learning_rate": 4.8154750866266785e-06, + "loss": 0.9232, + "step": 29003 + }, + { + "epoch": 0.68, + "grad_norm": 2.1194744541546533, + "learning_rate": 4.814822620399849e-06, + "loss": 0.999, + "step": 29004 + }, + { + "epoch": 0.68, + "grad_norm": 1.8820881961534115, + "learning_rate": 4.81417018436277e-06, + "loss": 0.8978, + "step": 29005 + }, + { + "epoch": 0.68, + "grad_norm": 1.1092562122869423, + "learning_rate": 4.8135177785192275e-06, + "loss": 0.9775, + "step": 29006 + }, + { + "epoch": 0.68, + "grad_norm": 2.5599775135967273, + "learning_rate": 4.812865402873032e-06, + "loss": 0.9322, + "step": 29007 + }, + { + "epoch": 0.68, + "grad_norm": 1.0630088889797096, + "learning_rate": 4.812213057427972e-06, + "loss": 0.9293, + "step": 29008 + }, + { + "epoch": 0.68, + "grad_norm": 1.1661347856687962, + "learning_rate": 4.811560742187851e-06, + "loss": 1.0158, + "step": 29009 + }, + { + "epoch": 0.68, + "grad_norm": 2.088063600884728, + "learning_rate": 4.810908457156472e-06, + "loss": 1.0839, + "step": 29010 + }, + { + "epoch": 0.68, + "grad_norm": 1.9208718374022162, + "learning_rate": 4.810256202337621e-06, + "loss": 0.9488, + "step": 29011 + }, + { + "epoch": 0.68, + "grad_norm": 2.1221706516534518, + "learning_rate": 4.8096039777351e-06, + "loss": 0.8944, + "step": 29012 + }, + { + "epoch": 0.68, + "grad_norm": 2.0353610621323894, + "learning_rate": 4.808951783352709e-06, + "loss": 1.0409, + "step": 29013 + }, + { + "epoch": 0.68, + "grad_norm": 2.074467656072957, + "learning_rate": 4.808299619194251e-06, + "loss": 0.9344, + "step": 29014 + }, + { + "epoch": 0.68, + "grad_norm": 2.110249538726926, + "learning_rate": 4.807647485263508e-06, + "loss": 1.1456, + "step": 29015 + }, + { + "epoch": 0.68, + "grad_norm": 2.2575550781808986, + "learning_rate": 4.8069953815642865e-06, + "loss": 0.9737, + "step": 29016 + }, + { + "epoch": 0.68, + "grad_norm": 2.9734286501284695, + "learning_rate": 4.806343308100385e-06, + "loss": 0.9234, + "step": 29017 + }, + { + "epoch": 0.68, + "grad_norm": 2.0023933772075946, + "learning_rate": 4.8056912648755975e-06, + "loss": 0.9075, + "step": 29018 + }, + { + "epoch": 0.68, + "grad_norm": 1.8156264386001066, + "learning_rate": 4.805039251893716e-06, + "loss": 1.1153, + "step": 29019 + }, + { + "epoch": 0.68, + "grad_norm": 2.0112577300135435, + "learning_rate": 4.80438726915854e-06, + "loss": 1.0412, + "step": 29020 + }, + { + "epoch": 0.68, + "grad_norm": 2.136436654739898, + "learning_rate": 4.803735316673871e-06, + "loss": 1.1393, + "step": 29021 + }, + { + "epoch": 0.68, + "grad_norm": 2.3193647020463617, + "learning_rate": 4.803083394443497e-06, + "loss": 1.0165, + "step": 29022 + }, + { + "epoch": 0.68, + "grad_norm": 2.1003892393362125, + "learning_rate": 4.802431502471219e-06, + "loss": 0.9646, + "step": 29023 + }, + { + "epoch": 0.68, + "grad_norm": 1.8069104681311814, + "learning_rate": 4.801779640760831e-06, + "loss": 0.9452, + "step": 29024 + }, + { + "epoch": 0.68, + "grad_norm": 2.305807922064838, + "learning_rate": 4.801127809316123e-06, + "loss": 0.9004, + "step": 29025 + }, + { + "epoch": 0.68, + "grad_norm": 1.9516916589816617, + "learning_rate": 4.8004760081408966e-06, + "loss": 1.0251, + "step": 29026 + }, + { + "epoch": 0.68, + "grad_norm": 1.9266744554288633, + "learning_rate": 4.799824237238947e-06, + "loss": 1.0913, + "step": 29027 + }, + { + "epoch": 0.68, + "grad_norm": 1.927796342483254, + "learning_rate": 4.7991724966140674e-06, + "loss": 1.105, + "step": 29028 + }, + { + "epoch": 0.68, + "grad_norm": 2.2246897017718825, + "learning_rate": 4.798520786270049e-06, + "loss": 0.9412, + "step": 29029 + }, + { + "epoch": 0.68, + "grad_norm": 2.0324173146472924, + "learning_rate": 4.79786910621069e-06, + "loss": 1.1864, + "step": 29030 + }, + { + "epoch": 0.68, + "grad_norm": 2.122439656348836, + "learning_rate": 4.79721745643979e-06, + "loss": 0.9882, + "step": 29031 + }, + { + "epoch": 0.68, + "grad_norm": 1.9674070879817886, + "learning_rate": 4.79656583696113e-06, + "loss": 0.9748, + "step": 29032 + }, + { + "epoch": 0.68, + "grad_norm": 2.174656186838533, + "learning_rate": 4.795914247778511e-06, + "loss": 1.1488, + "step": 29033 + }, + { + "epoch": 0.68, + "grad_norm": 2.2230948625491895, + "learning_rate": 4.795262688895728e-06, + "loss": 1.0424, + "step": 29034 + }, + { + "epoch": 0.68, + "grad_norm": 1.969446590135935, + "learning_rate": 4.794611160316576e-06, + "loss": 1.0621, + "step": 29035 + }, + { + "epoch": 0.68, + "grad_norm": 1.9047572850914354, + "learning_rate": 4.793959662044846e-06, + "loss": 0.8946, + "step": 29036 + }, + { + "epoch": 0.68, + "grad_norm": 2.493824333177907, + "learning_rate": 4.793308194084327e-06, + "loss": 0.9903, + "step": 29037 + }, + { + "epoch": 0.68, + "grad_norm": 2.346854419836673, + "learning_rate": 4.792656756438821e-06, + "loss": 0.9412, + "step": 29038 + }, + { + "epoch": 0.68, + "grad_norm": 1.9502132189424777, + "learning_rate": 4.7920053491121114e-06, + "loss": 1.1127, + "step": 29039 + }, + { + "epoch": 0.68, + "grad_norm": 1.9107561761855216, + "learning_rate": 4.791353972107999e-06, + "loss": 0.954, + "step": 29040 + }, + { + "epoch": 0.68, + "grad_norm": 2.108413293498161, + "learning_rate": 4.790702625430269e-06, + "loss": 0.9644, + "step": 29041 + }, + { + "epoch": 0.68, + "grad_norm": 2.3004429248466787, + "learning_rate": 4.7900513090827225e-06, + "loss": 1.1282, + "step": 29042 + }, + { + "epoch": 0.68, + "grad_norm": 2.7296261392865944, + "learning_rate": 4.789400023069142e-06, + "loss": 0.9852, + "step": 29043 + }, + { + "epoch": 0.68, + "grad_norm": 1.7322307168601285, + "learning_rate": 4.788748767393328e-06, + "loss": 1.0371, + "step": 29044 + }, + { + "epoch": 0.68, + "grad_norm": 2.1322724907359234, + "learning_rate": 4.788097542059069e-06, + "loss": 0.9099, + "step": 29045 + }, + { + "epoch": 0.68, + "grad_norm": 1.9578541352901737, + "learning_rate": 4.787446347070152e-06, + "loss": 1.1703, + "step": 29046 + }, + { + "epoch": 0.68, + "grad_norm": 2.0244164494446566, + "learning_rate": 4.7867951824303725e-06, + "loss": 1.0224, + "step": 29047 + }, + { + "epoch": 0.68, + "grad_norm": 1.8903369119431088, + "learning_rate": 4.786144048143526e-06, + "loss": 0.934, + "step": 29048 + }, + { + "epoch": 0.68, + "grad_norm": 1.8971457188263157, + "learning_rate": 4.785492944213399e-06, + "loss": 1.0918, + "step": 29049 + }, + { + "epoch": 0.68, + "grad_norm": 1.8845316217858747, + "learning_rate": 4.784841870643781e-06, + "loss": 0.9347, + "step": 29050 + }, + { + "epoch": 0.68, + "grad_norm": 3.5073942023853264, + "learning_rate": 4.784190827438462e-06, + "loss": 1.0603, + "step": 29051 + }, + { + "epoch": 0.68, + "grad_norm": 1.8311265650946638, + "learning_rate": 4.783539814601242e-06, + "loss": 1.0884, + "step": 29052 + }, + { + "epoch": 0.68, + "grad_norm": 2.100017940834961, + "learning_rate": 4.782888832135903e-06, + "loss": 1.0431, + "step": 29053 + }, + { + "epoch": 0.68, + "grad_norm": 2.1806229789509404, + "learning_rate": 4.782237880046232e-06, + "loss": 1.004, + "step": 29054 + }, + { + "epoch": 0.68, + "grad_norm": 2.134976177921955, + "learning_rate": 4.781586958336025e-06, + "loss": 0.9789, + "step": 29055 + }, + { + "epoch": 0.68, + "grad_norm": 1.9628261094142683, + "learning_rate": 4.780936067009075e-06, + "loss": 1.116, + "step": 29056 + }, + { + "epoch": 0.68, + "grad_norm": 1.8952955528053521, + "learning_rate": 4.7802852060691676e-06, + "loss": 0.8569, + "step": 29057 + }, + { + "epoch": 0.68, + "grad_norm": 1.8986963829722916, + "learning_rate": 4.779634375520088e-06, + "loss": 0.9626, + "step": 29058 + }, + { + "epoch": 0.68, + "grad_norm": 2.035518868770604, + "learning_rate": 4.778983575365634e-06, + "loss": 0.9007, + "step": 29059 + }, + { + "epoch": 0.68, + "grad_norm": 1.9372726278654993, + "learning_rate": 4.778332805609587e-06, + "loss": 0.9017, + "step": 29060 + }, + { + "epoch": 0.68, + "grad_norm": 2.074703672850295, + "learning_rate": 4.777682066255743e-06, + "loss": 1.1293, + "step": 29061 + }, + { + "epoch": 0.68, + "grad_norm": 2.109993206712459, + "learning_rate": 4.777031357307884e-06, + "loss": 1.0623, + "step": 29062 + }, + { + "epoch": 0.68, + "grad_norm": 1.9614630322412527, + "learning_rate": 4.776380678769805e-06, + "loss": 1.038, + "step": 29063 + }, + { + "epoch": 0.68, + "grad_norm": 1.7740388475101792, + "learning_rate": 4.775730030645288e-06, + "loss": 1.0385, + "step": 29064 + }, + { + "epoch": 0.68, + "grad_norm": 1.9880086086787592, + "learning_rate": 4.7750794129381285e-06, + "loss": 1.0305, + "step": 29065 + }, + { + "epoch": 0.68, + "grad_norm": 1.9921282044169388, + "learning_rate": 4.774428825652111e-06, + "loss": 0.9459, + "step": 29066 + }, + { + "epoch": 0.68, + "grad_norm": 2.284305882158431, + "learning_rate": 4.77377826879102e-06, + "loss": 1.047, + "step": 29067 + }, + { + "epoch": 0.68, + "grad_norm": 1.9959469447768037, + "learning_rate": 4.773127742358645e-06, + "loss": 1.0365, + "step": 29068 + }, + { + "epoch": 0.68, + "grad_norm": 2.2496747916648308, + "learning_rate": 4.772477246358781e-06, + "loss": 0.9533, + "step": 29069 + }, + { + "epoch": 0.68, + "grad_norm": 2.2297675105405674, + "learning_rate": 4.771826780795208e-06, + "loss": 1.0724, + "step": 29070 + }, + { + "epoch": 0.68, + "grad_norm": 2.2606514975357936, + "learning_rate": 4.771176345671711e-06, + "loss": 1.042, + "step": 29071 + }, + { + "epoch": 0.68, + "grad_norm": 2.0417060151541033, + "learning_rate": 4.770525940992081e-06, + "loss": 0.9129, + "step": 29072 + }, + { + "epoch": 0.68, + "grad_norm": 1.968778238444051, + "learning_rate": 4.769875566760108e-06, + "loss": 0.885, + "step": 29073 + }, + { + "epoch": 0.68, + "grad_norm": 2.1634221802669744, + "learning_rate": 4.769225222979575e-06, + "loss": 0.9272, + "step": 29074 + }, + { + "epoch": 0.68, + "grad_norm": 1.8136214867028932, + "learning_rate": 4.768574909654267e-06, + "loss": 0.9633, + "step": 29075 + }, + { + "epoch": 0.69, + "grad_norm": 2.963787962107601, + "learning_rate": 4.76792462678797e-06, + "loss": 1.0677, + "step": 29076 + }, + { + "epoch": 0.69, + "grad_norm": 1.9451097414812848, + "learning_rate": 4.767274374384476e-06, + "loss": 0.9895, + "step": 29077 + }, + { + "epoch": 0.69, + "grad_norm": 2.5004567036481276, + "learning_rate": 4.766624152447569e-06, + "loss": 0.86, + "step": 29078 + }, + { + "epoch": 0.69, + "grad_norm": 2.1115557531786693, + "learning_rate": 4.765973960981029e-06, + "loss": 0.8379, + "step": 29079 + }, + { + "epoch": 0.69, + "grad_norm": 2.0400710327086657, + "learning_rate": 4.765323799988648e-06, + "loss": 0.866, + "step": 29080 + }, + { + "epoch": 0.69, + "grad_norm": 1.9915087194491676, + "learning_rate": 4.764673669474206e-06, + "loss": 0.9678, + "step": 29081 + }, + { + "epoch": 0.69, + "grad_norm": 1.078966363899713, + "learning_rate": 4.764023569441495e-06, + "loss": 0.9643, + "step": 29082 + }, + { + "epoch": 0.69, + "grad_norm": 2.097208436124762, + "learning_rate": 4.763373499894293e-06, + "loss": 1.1511, + "step": 29083 + }, + { + "epoch": 0.69, + "grad_norm": 2.1158957539330348, + "learning_rate": 4.762723460836392e-06, + "loss": 0.9375, + "step": 29084 + }, + { + "epoch": 0.69, + "grad_norm": 1.050148462851586, + "learning_rate": 4.762073452271569e-06, + "loss": 0.8816, + "step": 29085 + }, + { + "epoch": 0.69, + "grad_norm": 2.0854332277296255, + "learning_rate": 4.761423474203617e-06, + "loss": 1.0214, + "step": 29086 + }, + { + "epoch": 0.69, + "grad_norm": 1.9801970696183278, + "learning_rate": 4.760773526636315e-06, + "loss": 0.9211, + "step": 29087 + }, + { + "epoch": 0.69, + "grad_norm": 1.9637522870125736, + "learning_rate": 4.760123609573446e-06, + "loss": 0.9445, + "step": 29088 + }, + { + "epoch": 0.69, + "grad_norm": 2.1044693154554728, + "learning_rate": 4.759473723018796e-06, + "loss": 0.9892, + "step": 29089 + }, + { + "epoch": 0.69, + "grad_norm": 1.981685196892046, + "learning_rate": 4.758823866976152e-06, + "loss": 0.9669, + "step": 29090 + }, + { + "epoch": 0.69, + "grad_norm": 1.823631504353449, + "learning_rate": 4.758174041449295e-06, + "loss": 0.9634, + "step": 29091 + }, + { + "epoch": 0.69, + "grad_norm": 2.6521773403743873, + "learning_rate": 4.7575242464420045e-06, + "loss": 1.129, + "step": 29092 + }, + { + "epoch": 0.69, + "grad_norm": 1.8304233710095494, + "learning_rate": 4.756874481958068e-06, + "loss": 1.049, + "step": 29093 + }, + { + "epoch": 0.69, + "grad_norm": 2.2844515131522045, + "learning_rate": 4.756224748001272e-06, + "loss": 1.0935, + "step": 29094 + }, + { + "epoch": 0.69, + "grad_norm": 1.8264726325470049, + "learning_rate": 4.755575044575396e-06, + "loss": 0.8921, + "step": 29095 + }, + { + "epoch": 0.69, + "grad_norm": 1.8888890899419435, + "learning_rate": 4.754925371684217e-06, + "loss": 0.8839, + "step": 29096 + }, + { + "epoch": 0.69, + "grad_norm": 2.510454488259718, + "learning_rate": 4.754275729331524e-06, + "loss": 0.8357, + "step": 29097 + }, + { + "epoch": 0.69, + "grad_norm": 1.927744837609821, + "learning_rate": 4.753626117521103e-06, + "loss": 1.0959, + "step": 29098 + }, + { + "epoch": 0.69, + "grad_norm": 1.9074302021132334, + "learning_rate": 4.752976536256731e-06, + "loss": 0.8886, + "step": 29099 + }, + { + "epoch": 0.69, + "grad_norm": 1.8667186552198816, + "learning_rate": 4.752326985542187e-06, + "loss": 0.9522, + "step": 29100 + }, + { + "epoch": 0.69, + "grad_norm": 1.9499556695540685, + "learning_rate": 4.75167746538126e-06, + "loss": 1.0679, + "step": 29101 + }, + { + "epoch": 0.69, + "grad_norm": 2.4916872437824953, + "learning_rate": 4.751027975777726e-06, + "loss": 1.1295, + "step": 29102 + }, + { + "epoch": 0.69, + "grad_norm": 2.1346559493090664, + "learning_rate": 4.750378516735372e-06, + "loss": 1.0089, + "step": 29103 + }, + { + "epoch": 0.69, + "grad_norm": 1.952370067131318, + "learning_rate": 4.7497290882579724e-06, + "loss": 0.8705, + "step": 29104 + }, + { + "epoch": 0.69, + "grad_norm": 1.979044402463713, + "learning_rate": 4.749079690349316e-06, + "loss": 0.9112, + "step": 29105 + }, + { + "epoch": 0.69, + "grad_norm": 2.0862730788007147, + "learning_rate": 4.748430323013176e-06, + "loss": 1.0622, + "step": 29106 + }, + { + "epoch": 0.69, + "grad_norm": 1.9973064996448862, + "learning_rate": 4.747780986253337e-06, + "loss": 1.0667, + "step": 29107 + }, + { + "epoch": 0.69, + "grad_norm": 2.167606187157736, + "learning_rate": 4.747131680073589e-06, + "loss": 1.0, + "step": 29108 + }, + { + "epoch": 0.69, + "grad_norm": 1.8121269069192754, + "learning_rate": 4.746482404477695e-06, + "loss": 0.87, + "step": 29109 + }, + { + "epoch": 0.69, + "grad_norm": 2.250773304742688, + "learning_rate": 4.745833159469443e-06, + "loss": 0.9826, + "step": 29110 + }, + { + "epoch": 0.69, + "grad_norm": 1.9440585410461693, + "learning_rate": 4.745183945052615e-06, + "loss": 0.8951, + "step": 29111 + }, + { + "epoch": 0.69, + "grad_norm": 2.050572221184901, + "learning_rate": 4.744534761230996e-06, + "loss": 0.8605, + "step": 29112 + }, + { + "epoch": 0.69, + "grad_norm": 1.9421151582360254, + "learning_rate": 4.7438856080083525e-06, + "loss": 0.9114, + "step": 29113 + }, + { + "epoch": 0.69, + "grad_norm": 1.8282725007715286, + "learning_rate": 4.743236485388471e-06, + "loss": 0.9004, + "step": 29114 + }, + { + "epoch": 0.69, + "grad_norm": 2.2867305895574126, + "learning_rate": 4.742587393375131e-06, + "loss": 1.0351, + "step": 29115 + }, + { + "epoch": 0.69, + "grad_norm": 1.9598068922553595, + "learning_rate": 4.741938331972119e-06, + "loss": 0.9177, + "step": 29116 + }, + { + "epoch": 0.69, + "grad_norm": 2.243103153562676, + "learning_rate": 4.7412893011831995e-06, + "loss": 0.9567, + "step": 29117 + }, + { + "epoch": 0.69, + "grad_norm": 1.7583814443302932, + "learning_rate": 4.7406403010121595e-06, + "loss": 0.9063, + "step": 29118 + }, + { + "epoch": 0.69, + "grad_norm": 1.9185568562534885, + "learning_rate": 4.7399913314627785e-06, + "loss": 1.0356, + "step": 29119 + }, + { + "epoch": 0.69, + "grad_norm": 2.155988096186254, + "learning_rate": 4.739342392538831e-06, + "loss": 1.168, + "step": 29120 + }, + { + "epoch": 0.69, + "grad_norm": 2.0398281188879674, + "learning_rate": 4.738693484244101e-06, + "loss": 0.9543, + "step": 29121 + }, + { + "epoch": 0.69, + "grad_norm": 2.0556716443100966, + "learning_rate": 4.738044606582359e-06, + "loss": 1.1289, + "step": 29122 + }, + { + "epoch": 0.69, + "grad_norm": 4.140221740771042, + "learning_rate": 4.737395759557392e-06, + "loss": 1.0047, + "step": 29123 + }, + { + "epoch": 0.69, + "grad_norm": 2.0079723614105935, + "learning_rate": 4.736746943172969e-06, + "loss": 0.966, + "step": 29124 + }, + { + "epoch": 0.69, + "grad_norm": 1.895380512357351, + "learning_rate": 4.736098157432874e-06, + "loss": 1.0073, + "step": 29125 + }, + { + "epoch": 0.69, + "grad_norm": 2.1865460246420056, + "learning_rate": 4.735449402340882e-06, + "loss": 1.0432, + "step": 29126 + }, + { + "epoch": 0.69, + "grad_norm": 1.7819456521832144, + "learning_rate": 4.7348006779007685e-06, + "loss": 0.8048, + "step": 29127 + }, + { + "epoch": 0.69, + "grad_norm": 1.9792270757515378, + "learning_rate": 4.734151984116311e-06, + "loss": 1.0165, + "step": 29128 + }, + { + "epoch": 0.69, + "grad_norm": 2.140853127395631, + "learning_rate": 4.733503320991292e-06, + "loss": 0.9452, + "step": 29129 + }, + { + "epoch": 0.69, + "grad_norm": 2.2180913198575327, + "learning_rate": 4.732854688529484e-06, + "loss": 0.8591, + "step": 29130 + }, + { + "epoch": 0.69, + "grad_norm": 1.9616017689999052, + "learning_rate": 4.73220608673466e-06, + "loss": 0.8663, + "step": 29131 + }, + { + "epoch": 0.69, + "grad_norm": 2.1327280627817906, + "learning_rate": 4.7315575156106e-06, + "loss": 0.9344, + "step": 29132 + }, + { + "epoch": 0.69, + "grad_norm": 1.9264387585214056, + "learning_rate": 4.730908975161083e-06, + "loss": 0.9743, + "step": 29133 + }, + { + "epoch": 0.69, + "grad_norm": 2.3294549753858718, + "learning_rate": 4.730260465389882e-06, + "loss": 0.9359, + "step": 29134 + }, + { + "epoch": 0.69, + "grad_norm": 1.081065835826356, + "learning_rate": 4.729611986300771e-06, + "loss": 1.0157, + "step": 29135 + }, + { + "epoch": 0.69, + "grad_norm": 1.9118609992209203, + "learning_rate": 4.728963537897526e-06, + "loss": 1.0279, + "step": 29136 + }, + { + "epoch": 0.69, + "grad_norm": 2.6924444976931072, + "learning_rate": 4.728315120183928e-06, + "loss": 0.9157, + "step": 29137 + }, + { + "epoch": 0.69, + "grad_norm": 2.3367918877680562, + "learning_rate": 4.727666733163748e-06, + "loss": 1.0174, + "step": 29138 + }, + { + "epoch": 0.69, + "grad_norm": 1.9489259479028715, + "learning_rate": 4.7270183768407585e-06, + "loss": 1.0216, + "step": 29139 + }, + { + "epoch": 0.69, + "grad_norm": 1.056120617178794, + "learning_rate": 4.726370051218741e-06, + "loss": 0.9445, + "step": 29140 + }, + { + "epoch": 0.69, + "grad_norm": 1.913499574818801, + "learning_rate": 4.725721756301463e-06, + "loss": 1.0751, + "step": 29141 + }, + { + "epoch": 0.69, + "grad_norm": 1.9947332715006916, + "learning_rate": 4.725073492092707e-06, + "loss": 1.0918, + "step": 29142 + }, + { + "epoch": 0.69, + "grad_norm": 1.8885571218475463, + "learning_rate": 4.724425258596238e-06, + "loss": 0.9656, + "step": 29143 + }, + { + "epoch": 0.69, + "grad_norm": 2.0345708241626395, + "learning_rate": 4.72377705581584e-06, + "loss": 1.0329, + "step": 29144 + }, + { + "epoch": 0.69, + "grad_norm": 2.163161383555252, + "learning_rate": 4.723128883755279e-06, + "loss": 0.9205, + "step": 29145 + }, + { + "epoch": 0.69, + "grad_norm": 2.8046711497673678, + "learning_rate": 4.722480742418335e-06, + "loss": 0.9938, + "step": 29146 + }, + { + "epoch": 0.69, + "grad_norm": 2.094172432716528, + "learning_rate": 4.721832631808781e-06, + "loss": 1.01, + "step": 29147 + }, + { + "epoch": 0.69, + "grad_norm": 1.8575253350016943, + "learning_rate": 4.721184551930382e-06, + "loss": 1.0304, + "step": 29148 + }, + { + "epoch": 0.69, + "grad_norm": 1.8232738690568504, + "learning_rate": 4.720536502786919e-06, + "loss": 0.9514, + "step": 29149 + }, + { + "epoch": 0.69, + "grad_norm": 2.04190276204163, + "learning_rate": 4.719888484382168e-06, + "loss": 0.995, + "step": 29150 + }, + { + "epoch": 0.69, + "grad_norm": 2.0327402086253765, + "learning_rate": 4.719240496719897e-06, + "loss": 0.9861, + "step": 29151 + }, + { + "epoch": 0.69, + "grad_norm": 2.2442560985842754, + "learning_rate": 4.718592539803877e-06, + "loss": 0.9098, + "step": 29152 + }, + { + "epoch": 0.69, + "grad_norm": 2.130261941092099, + "learning_rate": 4.717944613637883e-06, + "loss": 0.9559, + "step": 29153 + }, + { + "epoch": 0.69, + "grad_norm": 2.0451462105357554, + "learning_rate": 4.717296718225691e-06, + "loss": 1.044, + "step": 29154 + }, + { + "epoch": 0.69, + "grad_norm": 1.9048111789009285, + "learning_rate": 4.7166488535710696e-06, + "loss": 0.9971, + "step": 29155 + }, + { + "epoch": 0.69, + "grad_norm": 2.384260798868933, + "learning_rate": 4.716001019677788e-06, + "loss": 0.9935, + "step": 29156 + }, + { + "epoch": 0.69, + "grad_norm": 2.536674056174897, + "learning_rate": 4.715353216549622e-06, + "loss": 1.0061, + "step": 29157 + }, + { + "epoch": 0.69, + "grad_norm": 1.162378616057986, + "learning_rate": 4.714705444190346e-06, + "loss": 0.9441, + "step": 29158 + }, + { + "epoch": 0.69, + "grad_norm": 2.047146751044733, + "learning_rate": 4.714057702603728e-06, + "loss": 1.0563, + "step": 29159 + }, + { + "epoch": 0.69, + "grad_norm": 1.9209191615019938, + "learning_rate": 4.713409991793536e-06, + "loss": 1.0533, + "step": 29160 + }, + { + "epoch": 0.69, + "grad_norm": 2.1248071467755913, + "learning_rate": 4.712762311763549e-06, + "loss": 1.0185, + "step": 29161 + }, + { + "epoch": 0.69, + "grad_norm": 1.8909665229795785, + "learning_rate": 4.71211466251753e-06, + "loss": 0.9591, + "step": 29162 + }, + { + "epoch": 0.69, + "grad_norm": 1.8259230194255345, + "learning_rate": 4.711467044059256e-06, + "loss": 1.0444, + "step": 29163 + }, + { + "epoch": 0.69, + "grad_norm": 2.0654079515021366, + "learning_rate": 4.710819456392492e-06, + "loss": 0.9667, + "step": 29164 + }, + { + "epoch": 0.69, + "grad_norm": 2.0463518427810294, + "learning_rate": 4.710171899521015e-06, + "loss": 0.9075, + "step": 29165 + }, + { + "epoch": 0.69, + "grad_norm": 2.2185173272606487, + "learning_rate": 4.709524373448589e-06, + "loss": 1.0518, + "step": 29166 + }, + { + "epoch": 0.69, + "grad_norm": 1.954930727458464, + "learning_rate": 4.708876878178991e-06, + "loss": 1.0479, + "step": 29167 + }, + { + "epoch": 0.69, + "grad_norm": 2.1108111161287515, + "learning_rate": 4.708229413715987e-06, + "loss": 0.9548, + "step": 29168 + }, + { + "epoch": 0.69, + "grad_norm": 1.9811697398599222, + "learning_rate": 4.707581980063342e-06, + "loss": 0.9405, + "step": 29169 + }, + { + "epoch": 0.69, + "grad_norm": 1.935751829064631, + "learning_rate": 4.706934577224831e-06, + "loss": 1.1162, + "step": 29170 + }, + { + "epoch": 0.69, + "grad_norm": 1.2286247164613513, + "learning_rate": 4.706287205204225e-06, + "loss": 0.9225, + "step": 29171 + }, + { + "epoch": 0.69, + "grad_norm": 2.3243576914866826, + "learning_rate": 4.705639864005292e-06, + "loss": 1.1108, + "step": 29172 + }, + { + "epoch": 0.69, + "grad_norm": 2.065500514527271, + "learning_rate": 4.704992553631796e-06, + "loss": 0.9816, + "step": 29173 + }, + { + "epoch": 0.69, + "grad_norm": 2.076406215238682, + "learning_rate": 4.704345274087511e-06, + "loss": 1.0754, + "step": 29174 + }, + { + "epoch": 0.69, + "grad_norm": 2.135497644558942, + "learning_rate": 4.703698025376207e-06, + "loss": 0.9637, + "step": 29175 + }, + { + "epoch": 0.69, + "grad_norm": 2.234869719532083, + "learning_rate": 4.703050807501649e-06, + "loss": 0.9203, + "step": 29176 + }, + { + "epoch": 0.69, + "grad_norm": 1.9890997063157527, + "learning_rate": 4.702403620467604e-06, + "loss": 0.9558, + "step": 29177 + }, + { + "epoch": 0.69, + "grad_norm": 1.9369056821445454, + "learning_rate": 4.701756464277841e-06, + "loss": 1.0144, + "step": 29178 + }, + { + "epoch": 0.69, + "grad_norm": 2.039836760296748, + "learning_rate": 4.701109338936134e-06, + "loss": 0.8582, + "step": 29179 + }, + { + "epoch": 0.69, + "grad_norm": 1.8148824460518231, + "learning_rate": 4.700462244446246e-06, + "loss": 0.9356, + "step": 29180 + }, + { + "epoch": 0.69, + "grad_norm": 1.940753225616437, + "learning_rate": 4.699815180811939e-06, + "loss": 1.0531, + "step": 29181 + }, + { + "epoch": 0.69, + "grad_norm": 1.898126390385872, + "learning_rate": 4.699168148036992e-06, + "loss": 1.008, + "step": 29182 + }, + { + "epoch": 0.69, + "grad_norm": 2.8297876568654288, + "learning_rate": 4.698521146125161e-06, + "loss": 0.9165, + "step": 29183 + }, + { + "epoch": 0.69, + "grad_norm": 2.242079123779553, + "learning_rate": 4.697874175080224e-06, + "loss": 0.9784, + "step": 29184 + }, + { + "epoch": 0.69, + "grad_norm": 1.8209355232333357, + "learning_rate": 4.6972272349059365e-06, + "loss": 0.9223, + "step": 29185 + }, + { + "epoch": 0.69, + "grad_norm": 2.0740371946689007, + "learning_rate": 4.696580325606075e-06, + "loss": 1.0191, + "step": 29186 + }, + { + "epoch": 0.69, + "grad_norm": 1.902872817336782, + "learning_rate": 4.695933447184398e-06, + "loss": 0.912, + "step": 29187 + }, + { + "epoch": 0.69, + "grad_norm": 1.9016442761913708, + "learning_rate": 4.695286599644681e-06, + "loss": 0.9758, + "step": 29188 + }, + { + "epoch": 0.69, + "grad_norm": 1.938777504461041, + "learning_rate": 4.694639782990683e-06, + "loss": 0.9193, + "step": 29189 + }, + { + "epoch": 0.69, + "grad_norm": 2.073767567965832, + "learning_rate": 4.69399299722617e-06, + "loss": 1.0443, + "step": 29190 + }, + { + "epoch": 0.69, + "grad_norm": 2.148198745205745, + "learning_rate": 4.693346242354908e-06, + "loss": 0.9733, + "step": 29191 + }, + { + "epoch": 0.69, + "grad_norm": 2.052885610969045, + "learning_rate": 4.692699518380664e-06, + "loss": 1.0684, + "step": 29192 + }, + { + "epoch": 0.69, + "grad_norm": 2.2105468542421214, + "learning_rate": 4.692052825307213e-06, + "loss": 1.0136, + "step": 29193 + }, + { + "epoch": 0.69, + "grad_norm": 2.1929527787551764, + "learning_rate": 4.691406163138302e-06, + "loss": 0.9903, + "step": 29194 + }, + { + "epoch": 0.69, + "grad_norm": 2.0308254803587724, + "learning_rate": 4.690759531877706e-06, + "loss": 0.8758, + "step": 29195 + }, + { + "epoch": 0.69, + "grad_norm": 2.993528485594151, + "learning_rate": 4.690112931529191e-06, + "loss": 0.9826, + "step": 29196 + }, + { + "epoch": 0.69, + "grad_norm": 3.005509076332573, + "learning_rate": 4.689466362096521e-06, + "loss": 0.9082, + "step": 29197 + }, + { + "epoch": 0.69, + "grad_norm": 2.166731096335718, + "learning_rate": 4.688819823583454e-06, + "loss": 0.9057, + "step": 29198 + }, + { + "epoch": 0.69, + "grad_norm": 2.0981846578173062, + "learning_rate": 4.688173315993761e-06, + "loss": 1.0101, + "step": 29199 + }, + { + "epoch": 0.69, + "grad_norm": 1.892947781757021, + "learning_rate": 4.687526839331208e-06, + "loss": 0.875, + "step": 29200 + }, + { + "epoch": 0.69, + "grad_norm": 2.1407204460093303, + "learning_rate": 4.686880393599556e-06, + "loss": 0.9537, + "step": 29201 + }, + { + "epoch": 0.69, + "grad_norm": 2.8374365387830194, + "learning_rate": 4.686233978802563e-06, + "loss": 1.0525, + "step": 29202 + }, + { + "epoch": 0.69, + "grad_norm": 2.376910264710967, + "learning_rate": 4.685587594944003e-06, + "loss": 1.0353, + "step": 29203 + }, + { + "epoch": 0.69, + "grad_norm": 2.0498060086270127, + "learning_rate": 4.684941242027631e-06, + "loss": 0.9616, + "step": 29204 + }, + { + "epoch": 0.69, + "grad_norm": 1.9753387825027868, + "learning_rate": 4.684294920057214e-06, + "loss": 1.0285, + "step": 29205 + }, + { + "epoch": 0.69, + "grad_norm": 2.1044934341356165, + "learning_rate": 4.683648629036518e-06, + "loss": 0.9859, + "step": 29206 + }, + { + "epoch": 0.69, + "grad_norm": 2.0861015097963653, + "learning_rate": 4.683002368969303e-06, + "loss": 1.0886, + "step": 29207 + }, + { + "epoch": 0.69, + "grad_norm": 1.8455969070973934, + "learning_rate": 4.682356139859327e-06, + "loss": 0.8873, + "step": 29208 + }, + { + "epoch": 0.69, + "grad_norm": 3.1721396389415424, + "learning_rate": 4.681709941710358e-06, + "loss": 1.0295, + "step": 29209 + }, + { + "epoch": 0.69, + "grad_norm": 1.9387785311227599, + "learning_rate": 4.681063774526166e-06, + "loss": 0.9084, + "step": 29210 + }, + { + "epoch": 0.69, + "grad_norm": 1.9192996695257432, + "learning_rate": 4.680417638310496e-06, + "loss": 1.1092, + "step": 29211 + }, + { + "epoch": 0.69, + "grad_norm": 2.1782831047023326, + "learning_rate": 4.679771533067119e-06, + "loss": 1.0874, + "step": 29212 + }, + { + "epoch": 0.69, + "grad_norm": 2.3627031766065745, + "learning_rate": 4.679125458799796e-06, + "loss": 1.0344, + "step": 29213 + }, + { + "epoch": 0.69, + "grad_norm": 2.047651933659491, + "learning_rate": 4.678479415512297e-06, + "loss": 1.0367, + "step": 29214 + }, + { + "epoch": 0.69, + "grad_norm": 2.00664857769423, + "learning_rate": 4.677833403208367e-06, + "loss": 1.0756, + "step": 29215 + }, + { + "epoch": 0.69, + "grad_norm": 2.077838571371611, + "learning_rate": 4.6771874218917765e-06, + "loss": 0.9107, + "step": 29216 + }, + { + "epoch": 0.69, + "grad_norm": 1.969113447284231, + "learning_rate": 4.676541471566291e-06, + "loss": 1.003, + "step": 29217 + }, + { + "epoch": 0.69, + "grad_norm": 1.1443224790155746, + "learning_rate": 4.675895552235662e-06, + "loss": 0.8708, + "step": 29218 + }, + { + "epoch": 0.69, + "grad_norm": 1.0723821714776798, + "learning_rate": 4.675249663903658e-06, + "loss": 0.9449, + "step": 29219 + }, + { + "epoch": 0.69, + "grad_norm": 2.100433454028999, + "learning_rate": 4.674603806574033e-06, + "loss": 1.0505, + "step": 29220 + }, + { + "epoch": 0.69, + "grad_norm": 1.042322430033079, + "learning_rate": 4.673957980250553e-06, + "loss": 0.8681, + "step": 29221 + }, + { + "epoch": 0.69, + "grad_norm": 3.039555442071122, + "learning_rate": 4.673312184936973e-06, + "loss": 0.9578, + "step": 29222 + }, + { + "epoch": 0.69, + "grad_norm": 1.7798149450449143, + "learning_rate": 4.672666420637061e-06, + "loss": 0.9983, + "step": 29223 + }, + { + "epoch": 0.69, + "grad_norm": 2.019946577074728, + "learning_rate": 4.67202068735457e-06, + "loss": 0.9833, + "step": 29224 + }, + { + "epoch": 0.69, + "grad_norm": 1.7759969181489257, + "learning_rate": 4.6713749850932576e-06, + "loss": 0.8831, + "step": 29225 + }, + { + "epoch": 0.69, + "grad_norm": 1.9461388942736186, + "learning_rate": 4.670729313856887e-06, + "loss": 0.8904, + "step": 29226 + }, + { + "epoch": 0.69, + "grad_norm": 1.731897209832281, + "learning_rate": 4.670083673649223e-06, + "loss": 0.9867, + "step": 29227 + }, + { + "epoch": 0.69, + "grad_norm": 1.9819829143572334, + "learning_rate": 4.669438064474018e-06, + "loss": 0.9056, + "step": 29228 + }, + { + "epoch": 0.69, + "grad_norm": 2.045650634291143, + "learning_rate": 4.668792486335029e-06, + "loss": 0.8779, + "step": 29229 + }, + { + "epoch": 0.69, + "grad_norm": 1.8103753879807445, + "learning_rate": 4.668146939236017e-06, + "loss": 0.9058, + "step": 29230 + }, + { + "epoch": 0.69, + "grad_norm": 1.04094831844466, + "learning_rate": 4.667501423180746e-06, + "loss": 0.8947, + "step": 29231 + }, + { + "epoch": 0.69, + "grad_norm": 1.1589237005200075, + "learning_rate": 4.666855938172969e-06, + "loss": 0.958, + "step": 29232 + }, + { + "epoch": 0.69, + "grad_norm": 2.154709172049082, + "learning_rate": 4.666210484216444e-06, + "loss": 0.8702, + "step": 29233 + }, + { + "epoch": 0.69, + "grad_norm": 1.7585342080667583, + "learning_rate": 4.665565061314927e-06, + "loss": 0.9769, + "step": 29234 + }, + { + "epoch": 0.69, + "grad_norm": 2.7189003987836737, + "learning_rate": 4.6649196694721845e-06, + "loss": 1.0932, + "step": 29235 + }, + { + "epoch": 0.69, + "grad_norm": 1.9726152583386105, + "learning_rate": 4.664274308691968e-06, + "loss": 0.9965, + "step": 29236 + }, + { + "epoch": 0.69, + "grad_norm": 2.097325314668689, + "learning_rate": 4.663628978978033e-06, + "loss": 1.0168, + "step": 29237 + }, + { + "epoch": 0.69, + "grad_norm": 1.0854371147836548, + "learning_rate": 4.662983680334144e-06, + "loss": 0.9176, + "step": 29238 + }, + { + "epoch": 0.69, + "grad_norm": 2.6284267520881803, + "learning_rate": 4.662338412764048e-06, + "loss": 0.9432, + "step": 29239 + }, + { + "epoch": 0.69, + "grad_norm": 2.1935436205575862, + "learning_rate": 4.661693176271513e-06, + "loss": 1.1093, + "step": 29240 + }, + { + "epoch": 0.69, + "grad_norm": 2.023582164104162, + "learning_rate": 4.661047970860285e-06, + "loss": 0.9326, + "step": 29241 + }, + { + "epoch": 0.69, + "grad_norm": 2.4531205782425265, + "learning_rate": 4.66040279653413e-06, + "loss": 0.9816, + "step": 29242 + }, + { + "epoch": 0.69, + "grad_norm": 1.9594896780625006, + "learning_rate": 4.659757653296797e-06, + "loss": 0.8989, + "step": 29243 + }, + { + "epoch": 0.69, + "grad_norm": 2.2031892773347868, + "learning_rate": 4.65911254115205e-06, + "loss": 0.9019, + "step": 29244 + }, + { + "epoch": 0.69, + "grad_norm": 2.1417656557097486, + "learning_rate": 4.658467460103636e-06, + "loss": 0.8937, + "step": 29245 + }, + { + "epoch": 0.69, + "grad_norm": 1.9960185457179356, + "learning_rate": 4.657822410155321e-06, + "loss": 1.0239, + "step": 29246 + }, + { + "epoch": 0.69, + "grad_norm": 2.3146752383191216, + "learning_rate": 4.65717739131085e-06, + "loss": 0.9724, + "step": 29247 + }, + { + "epoch": 0.69, + "grad_norm": 1.9716031130861003, + "learning_rate": 4.656532403573987e-06, + "loss": 1.0332, + "step": 29248 + }, + { + "epoch": 0.69, + "grad_norm": 2.3528178738838776, + "learning_rate": 4.655887446948485e-06, + "loss": 0.9342, + "step": 29249 + }, + { + "epoch": 0.69, + "grad_norm": 1.8645865549921241, + "learning_rate": 4.655242521438095e-06, + "loss": 1.0213, + "step": 29250 + }, + { + "epoch": 0.69, + "grad_norm": 1.9420090468454736, + "learning_rate": 4.654597627046574e-06, + "loss": 0.9679, + "step": 29251 + }, + { + "epoch": 0.69, + "grad_norm": 2.066666290393285, + "learning_rate": 4.653952763777682e-06, + "loss": 1.1591, + "step": 29252 + }, + { + "epoch": 0.69, + "grad_norm": 1.9377821491662046, + "learning_rate": 4.653307931635169e-06, + "loss": 0.9042, + "step": 29253 + }, + { + "epoch": 0.69, + "grad_norm": 1.1720916077976122, + "learning_rate": 4.652663130622788e-06, + "loss": 0.875, + "step": 29254 + }, + { + "epoch": 0.69, + "grad_norm": 1.1236993115618312, + "learning_rate": 4.652018360744294e-06, + "loss": 1.0139, + "step": 29255 + }, + { + "epoch": 0.69, + "grad_norm": 1.9935522488200454, + "learning_rate": 4.651373622003445e-06, + "loss": 0.8904, + "step": 29256 + }, + { + "epoch": 0.69, + "grad_norm": 1.9640025761323623, + "learning_rate": 4.650728914403994e-06, + "loss": 0.8506, + "step": 29257 + }, + { + "epoch": 0.69, + "grad_norm": 2.882619976787638, + "learning_rate": 4.650084237949688e-06, + "loss": 0.9079, + "step": 29258 + }, + { + "epoch": 0.69, + "grad_norm": 2.368142017969781, + "learning_rate": 4.649439592644286e-06, + "loss": 0.8088, + "step": 29259 + }, + { + "epoch": 0.69, + "grad_norm": 1.8250380715441197, + "learning_rate": 4.648794978491543e-06, + "loss": 0.9332, + "step": 29260 + }, + { + "epoch": 0.69, + "grad_norm": 2.116084440285558, + "learning_rate": 4.6481503954952114e-06, + "loss": 1.0989, + "step": 29261 + }, + { + "epoch": 0.69, + "grad_norm": 1.7863283198953168, + "learning_rate": 4.647505843659038e-06, + "loss": 1.0011, + "step": 29262 + }, + { + "epoch": 0.69, + "grad_norm": 1.9208470813060934, + "learning_rate": 4.6468613229867845e-06, + "loss": 0.9026, + "step": 29263 + }, + { + "epoch": 0.69, + "grad_norm": 2.1492613196967056, + "learning_rate": 4.646216833482195e-06, + "loss": 0.8865, + "step": 29264 + }, + { + "epoch": 0.69, + "grad_norm": 1.9550660608874035, + "learning_rate": 4.645572375149029e-06, + "loss": 0.9412, + "step": 29265 + }, + { + "epoch": 0.69, + "grad_norm": 1.7541694644721875, + "learning_rate": 4.644927947991034e-06, + "loss": 0.9602, + "step": 29266 + }, + { + "epoch": 0.69, + "grad_norm": 2.209200156041738, + "learning_rate": 4.644283552011965e-06, + "loss": 0.9331, + "step": 29267 + }, + { + "epoch": 0.69, + "grad_norm": 2.204288638566999, + "learning_rate": 4.6436391872155714e-06, + "loss": 0.9777, + "step": 29268 + }, + { + "epoch": 0.69, + "grad_norm": 2.003987723129451, + "learning_rate": 4.642994853605609e-06, + "loss": 1.0248, + "step": 29269 + }, + { + "epoch": 0.69, + "grad_norm": 2.094388465783808, + "learning_rate": 4.642350551185827e-06, + "loss": 0.9543, + "step": 29270 + }, + { + "epoch": 0.69, + "grad_norm": 1.9436011322763007, + "learning_rate": 4.6417062799599725e-06, + "loss": 0.9746, + "step": 29271 + }, + { + "epoch": 0.69, + "grad_norm": 2.1224384177310234, + "learning_rate": 4.641062039931799e-06, + "loss": 0.9257, + "step": 29272 + }, + { + "epoch": 0.69, + "grad_norm": 1.860706733481166, + "learning_rate": 4.640417831105064e-06, + "loss": 0.9694, + "step": 29273 + }, + { + "epoch": 0.69, + "grad_norm": 2.016125598023575, + "learning_rate": 4.6397736534835136e-06, + "loss": 0.8944, + "step": 29274 + }, + { + "epoch": 0.69, + "grad_norm": 2.120748364225476, + "learning_rate": 4.639129507070893e-06, + "loss": 0.9347, + "step": 29275 + }, + { + "epoch": 0.69, + "grad_norm": 1.9898657634648784, + "learning_rate": 4.638485391870958e-06, + "loss": 0.9047, + "step": 29276 + }, + { + "epoch": 0.69, + "grad_norm": 2.176495151687021, + "learning_rate": 4.637841307887463e-06, + "loss": 0.9202, + "step": 29277 + }, + { + "epoch": 0.69, + "grad_norm": 5.987930899862621, + "learning_rate": 4.637197255124153e-06, + "loss": 1.0535, + "step": 29278 + }, + { + "epoch": 0.69, + "grad_norm": 1.9657499588099305, + "learning_rate": 4.636553233584775e-06, + "loss": 1.0176, + "step": 29279 + }, + { + "epoch": 0.69, + "grad_norm": 1.9395885040923304, + "learning_rate": 4.635909243273082e-06, + "loss": 1.0209, + "step": 29280 + }, + { + "epoch": 0.69, + "grad_norm": 2.1064059584765404, + "learning_rate": 4.635265284192827e-06, + "loss": 0.903, + "step": 29281 + }, + { + "epoch": 0.69, + "grad_norm": 1.0596630798549087, + "learning_rate": 4.634621356347756e-06, + "loss": 0.9817, + "step": 29282 + }, + { + "epoch": 0.69, + "grad_norm": 2.287139211536601, + "learning_rate": 4.633977459741614e-06, + "loss": 1.0578, + "step": 29283 + }, + { + "epoch": 0.69, + "grad_norm": 2.605003457204817, + "learning_rate": 4.6333335943781586e-06, + "loss": 0.8919, + "step": 29284 + }, + { + "epoch": 0.69, + "grad_norm": 1.9681315687771197, + "learning_rate": 4.632689760261131e-06, + "loss": 0.9893, + "step": 29285 + }, + { + "epoch": 0.69, + "grad_norm": 2.1065926226189053, + "learning_rate": 4.632045957394286e-06, + "loss": 0.9414, + "step": 29286 + }, + { + "epoch": 0.69, + "grad_norm": 2.289768497653377, + "learning_rate": 4.631402185781364e-06, + "loss": 1.0168, + "step": 29287 + }, + { + "epoch": 0.69, + "grad_norm": 1.922446314767624, + "learning_rate": 4.630758445426123e-06, + "loss": 1.0543, + "step": 29288 + }, + { + "epoch": 0.69, + "grad_norm": 2.056504143608766, + "learning_rate": 4.630114736332302e-06, + "loss": 0.976, + "step": 29289 + }, + { + "epoch": 0.69, + "grad_norm": 1.9430373665470113, + "learning_rate": 4.6294710585036536e-06, + "loss": 1.0495, + "step": 29290 + }, + { + "epoch": 0.69, + "grad_norm": 2.042191956534554, + "learning_rate": 4.628827411943933e-06, + "loss": 0.9181, + "step": 29291 + }, + { + "epoch": 0.69, + "grad_norm": 1.7404502158553765, + "learning_rate": 4.628183796656871e-06, + "loss": 0.9857, + "step": 29292 + }, + { + "epoch": 0.69, + "grad_norm": 2.415391394199891, + "learning_rate": 4.627540212646223e-06, + "loss": 0.9863, + "step": 29293 + }, + { + "epoch": 0.69, + "grad_norm": 3.22015018846505, + "learning_rate": 4.626896659915739e-06, + "loss": 0.9179, + "step": 29294 + }, + { + "epoch": 0.69, + "grad_norm": 2.066339010021723, + "learning_rate": 4.626253138469169e-06, + "loss": 1.0184, + "step": 29295 + }, + { + "epoch": 0.69, + "grad_norm": 2.4854413895965, + "learning_rate": 4.625609648310248e-06, + "loss": 0.9504, + "step": 29296 + }, + { + "epoch": 0.69, + "grad_norm": 2.102717140570213, + "learning_rate": 4.62496618944273e-06, + "loss": 1.0107, + "step": 29297 + }, + { + "epoch": 0.69, + "grad_norm": 2.035777068564405, + "learning_rate": 4.624322761870363e-06, + "loss": 1.016, + "step": 29298 + }, + { + "epoch": 0.69, + "grad_norm": 1.97045463021064, + "learning_rate": 4.623679365596891e-06, + "loss": 1.0476, + "step": 29299 + }, + { + "epoch": 0.69, + "grad_norm": 1.9965806802692354, + "learning_rate": 4.623036000626056e-06, + "loss": 0.9676, + "step": 29300 + }, + { + "epoch": 0.69, + "grad_norm": 3.8823883356963313, + "learning_rate": 4.62239266696161e-06, + "loss": 0.9329, + "step": 29301 + }, + { + "epoch": 0.69, + "grad_norm": 1.7655742401131047, + "learning_rate": 4.621749364607297e-06, + "loss": 0.9786, + "step": 29302 + }, + { + "epoch": 0.69, + "grad_norm": 2.113672391594087, + "learning_rate": 4.6211060935668595e-06, + "loss": 1.0008, + "step": 29303 + }, + { + "epoch": 0.69, + "grad_norm": 1.8906192756055835, + "learning_rate": 4.620462853844049e-06, + "loss": 1.0087, + "step": 29304 + }, + { + "epoch": 0.69, + "grad_norm": 1.8980216340523866, + "learning_rate": 4.619819645442607e-06, + "loss": 0.8791, + "step": 29305 + }, + { + "epoch": 0.69, + "grad_norm": 1.0869692944128586, + "learning_rate": 4.619176468366274e-06, + "loss": 1.0156, + "step": 29306 + }, + { + "epoch": 0.69, + "grad_norm": 1.154895489484568, + "learning_rate": 4.6185333226188e-06, + "loss": 0.9708, + "step": 29307 + }, + { + "epoch": 0.69, + "grad_norm": 2.0137423380197843, + "learning_rate": 4.617890208203932e-06, + "loss": 1.0379, + "step": 29308 + }, + { + "epoch": 0.69, + "grad_norm": 1.971128704650902, + "learning_rate": 4.617247125125411e-06, + "loss": 1.0453, + "step": 29309 + }, + { + "epoch": 0.69, + "grad_norm": 1.8776778214895509, + "learning_rate": 4.616604073386978e-06, + "loss": 0.9424, + "step": 29310 + }, + { + "epoch": 0.69, + "grad_norm": 2.175711206352283, + "learning_rate": 4.6159610529923805e-06, + "loss": 0.967, + "step": 29311 + }, + { + "epoch": 0.69, + "grad_norm": 4.033556414745318, + "learning_rate": 4.61531806394537e-06, + "loss": 0.8982, + "step": 29312 + }, + { + "epoch": 0.69, + "grad_norm": 2.144405403856997, + "learning_rate": 4.614675106249674e-06, + "loss": 0.9502, + "step": 29313 + }, + { + "epoch": 0.69, + "grad_norm": 1.908117428007601, + "learning_rate": 4.614032179909045e-06, + "loss": 0.9275, + "step": 29314 + }, + { + "epoch": 0.69, + "grad_norm": 2.196621313253905, + "learning_rate": 4.613389284927225e-06, + "loss": 0.9064, + "step": 29315 + }, + { + "epoch": 0.69, + "grad_norm": 1.7918827441833693, + "learning_rate": 4.612746421307963e-06, + "loss": 0.9378, + "step": 29316 + }, + { + "epoch": 0.69, + "grad_norm": 2.068879030028936, + "learning_rate": 4.612103589054997e-06, + "loss": 0.9096, + "step": 29317 + }, + { + "epoch": 0.69, + "grad_norm": 1.1269088557504974, + "learning_rate": 4.611460788172065e-06, + "loss": 0.9114, + "step": 29318 + }, + { + "epoch": 0.69, + "grad_norm": 1.9861201937750377, + "learning_rate": 4.610818018662919e-06, + "loss": 0.9022, + "step": 29319 + }, + { + "epoch": 0.69, + "grad_norm": 2.081351786412349, + "learning_rate": 4.6101752805312926e-06, + "loss": 0.9842, + "step": 29320 + }, + { + "epoch": 0.69, + "grad_norm": 1.9811266226383817, + "learning_rate": 4.609532573780936e-06, + "loss": 0.9674, + "step": 29321 + }, + { + "epoch": 0.69, + "grad_norm": 1.845046230136715, + "learning_rate": 4.608889898415583e-06, + "loss": 1.0694, + "step": 29322 + }, + { + "epoch": 0.69, + "grad_norm": 2.229670889840346, + "learning_rate": 4.608247254438983e-06, + "loss": 0.917, + "step": 29323 + }, + { + "epoch": 0.69, + "grad_norm": 2.0370385460778877, + "learning_rate": 4.607604641854872e-06, + "loss": 0.9268, + "step": 29324 + }, + { + "epoch": 0.69, + "grad_norm": 1.111264075292275, + "learning_rate": 4.606962060666997e-06, + "loss": 0.9327, + "step": 29325 + }, + { + "epoch": 0.69, + "grad_norm": 1.8731239267447757, + "learning_rate": 4.606319510879097e-06, + "loss": 0.963, + "step": 29326 + }, + { + "epoch": 0.69, + "grad_norm": 1.1780376462968936, + "learning_rate": 4.605676992494909e-06, + "loss": 0.9059, + "step": 29327 + }, + { + "epoch": 0.69, + "grad_norm": 1.903444718221429, + "learning_rate": 4.605034505518176e-06, + "loss": 1.1583, + "step": 29328 + }, + { + "epoch": 0.69, + "grad_norm": 1.9552372714986543, + "learning_rate": 4.604392049952645e-06, + "loss": 0.9686, + "step": 29329 + }, + { + "epoch": 0.69, + "grad_norm": 2.0171196946779686, + "learning_rate": 4.603749625802051e-06, + "loss": 1.0157, + "step": 29330 + }, + { + "epoch": 0.69, + "grad_norm": 1.1471076864939584, + "learning_rate": 4.603107233070131e-06, + "loss": 0.9632, + "step": 29331 + }, + { + "epoch": 0.69, + "grad_norm": 2.249720246213721, + "learning_rate": 4.60246487176063e-06, + "loss": 1.1722, + "step": 29332 + }, + { + "epoch": 0.69, + "grad_norm": 1.056980110471148, + "learning_rate": 4.601822541877291e-06, + "loss": 0.9665, + "step": 29333 + }, + { + "epoch": 0.69, + "grad_norm": 2.4283130104641204, + "learning_rate": 4.601180243423851e-06, + "loss": 1.0573, + "step": 29334 + }, + { + "epoch": 0.69, + "grad_norm": 1.8907726186863143, + "learning_rate": 4.600537976404044e-06, + "loss": 0.9549, + "step": 29335 + }, + { + "epoch": 0.69, + "grad_norm": 1.0265869822660079, + "learning_rate": 4.5998957408216135e-06, + "loss": 0.9287, + "step": 29336 + }, + { + "epoch": 0.69, + "grad_norm": 2.4064055157329745, + "learning_rate": 4.599253536680305e-06, + "loss": 0.9397, + "step": 29337 + }, + { + "epoch": 0.69, + "grad_norm": 1.9196246877368646, + "learning_rate": 4.598611363983853e-06, + "loss": 1.0205, + "step": 29338 + }, + { + "epoch": 0.69, + "grad_norm": 1.9575391427302078, + "learning_rate": 4.59796922273599e-06, + "loss": 1.0579, + "step": 29339 + }, + { + "epoch": 0.69, + "grad_norm": 1.895296943735643, + "learning_rate": 4.597327112940465e-06, + "loss": 0.9219, + "step": 29340 + }, + { + "epoch": 0.69, + "grad_norm": 2.031463938345187, + "learning_rate": 4.596685034601008e-06, + "loss": 1.0231, + "step": 29341 + }, + { + "epoch": 0.69, + "grad_norm": 1.837443286156764, + "learning_rate": 4.596042987721365e-06, + "loss": 1.1013, + "step": 29342 + }, + { + "epoch": 0.69, + "grad_norm": 1.9265032894143357, + "learning_rate": 4.595400972305267e-06, + "loss": 1.0177, + "step": 29343 + }, + { + "epoch": 0.69, + "grad_norm": 1.9763028160129588, + "learning_rate": 4.594758988356459e-06, + "loss": 0.9911, + "step": 29344 + }, + { + "epoch": 0.69, + "grad_norm": 3.1709516624040384, + "learning_rate": 4.594117035878672e-06, + "loss": 0.9932, + "step": 29345 + }, + { + "epoch": 0.69, + "grad_norm": 2.046821481324911, + "learning_rate": 4.593475114875652e-06, + "loss": 1.0385, + "step": 29346 + }, + { + "epoch": 0.69, + "grad_norm": 2.1979316196956544, + "learning_rate": 4.592833225351131e-06, + "loss": 1.0466, + "step": 29347 + }, + { + "epoch": 0.69, + "grad_norm": 1.8607849904025655, + "learning_rate": 4.592191367308842e-06, + "loss": 1.0809, + "step": 29348 + }, + { + "epoch": 0.69, + "grad_norm": 2.105739239071736, + "learning_rate": 4.5915495407525275e-06, + "loss": 1.1037, + "step": 29349 + }, + { + "epoch": 0.69, + "grad_norm": 2.101337076659778, + "learning_rate": 4.590907745685927e-06, + "loss": 1.0519, + "step": 29350 + }, + { + "epoch": 0.69, + "grad_norm": 1.9726252764687433, + "learning_rate": 4.590265982112776e-06, + "loss": 0.9905, + "step": 29351 + }, + { + "epoch": 0.69, + "grad_norm": 2.178186098855325, + "learning_rate": 4.589624250036804e-06, + "loss": 1.0216, + "step": 29352 + }, + { + "epoch": 0.69, + "grad_norm": 1.779217108326628, + "learning_rate": 4.588982549461752e-06, + "loss": 0.8722, + "step": 29353 + }, + { + "epoch": 0.69, + "grad_norm": 2.219314636201093, + "learning_rate": 4.588340880391362e-06, + "loss": 1.0118, + "step": 29354 + }, + { + "epoch": 0.69, + "grad_norm": 1.1193579843202346, + "learning_rate": 4.587699242829364e-06, + "loss": 1.0392, + "step": 29355 + }, + { + "epoch": 0.69, + "grad_norm": 2.3903273261949254, + "learning_rate": 4.58705763677949e-06, + "loss": 1.0387, + "step": 29356 + }, + { + "epoch": 0.69, + "grad_norm": 2.5523300547534498, + "learning_rate": 4.586416062245481e-06, + "loss": 1.091, + "step": 29357 + }, + { + "epoch": 0.69, + "grad_norm": 2.251095302784736, + "learning_rate": 4.5857745192310755e-06, + "loss": 0.8922, + "step": 29358 + }, + { + "epoch": 0.69, + "grad_norm": 1.8876886946152172, + "learning_rate": 4.585133007740005e-06, + "loss": 1.0464, + "step": 29359 + }, + { + "epoch": 0.69, + "grad_norm": 3.071096231535504, + "learning_rate": 4.584491527775999e-06, + "loss": 0.9467, + "step": 29360 + }, + { + "epoch": 0.69, + "grad_norm": 1.934790246252503, + "learning_rate": 4.583850079342803e-06, + "loss": 0.9406, + "step": 29361 + }, + { + "epoch": 0.69, + "grad_norm": 2.1335249395817395, + "learning_rate": 4.583208662444143e-06, + "loss": 0.9529, + "step": 29362 + }, + { + "epoch": 0.69, + "grad_norm": 2.0318878151345894, + "learning_rate": 4.58256727708376e-06, + "loss": 1.0767, + "step": 29363 + }, + { + "epoch": 0.69, + "grad_norm": 1.8774939857084907, + "learning_rate": 4.5819259232653824e-06, + "loss": 0.9862, + "step": 29364 + }, + { + "epoch": 0.69, + "grad_norm": 2.262045254368371, + "learning_rate": 4.581284600992751e-06, + "loss": 0.8485, + "step": 29365 + }, + { + "epoch": 0.69, + "grad_norm": 2.0921268845055905, + "learning_rate": 4.580643310269592e-06, + "loss": 1.0508, + "step": 29366 + }, + { + "epoch": 0.69, + "grad_norm": 2.0023139041203, + "learning_rate": 4.580002051099648e-06, + "loss": 0.9081, + "step": 29367 + }, + { + "epoch": 0.69, + "grad_norm": 1.8655103130473787, + "learning_rate": 4.579360823486644e-06, + "loss": 0.9734, + "step": 29368 + }, + { + "epoch": 0.69, + "grad_norm": 2.334303097430637, + "learning_rate": 4.57871962743432e-06, + "loss": 1.0371, + "step": 29369 + }, + { + "epoch": 0.69, + "grad_norm": 1.8189231314136658, + "learning_rate": 4.578078462946404e-06, + "loss": 0.9011, + "step": 29370 + }, + { + "epoch": 0.69, + "grad_norm": 2.141677812905303, + "learning_rate": 4.577437330026635e-06, + "loss": 0.947, + "step": 29371 + }, + { + "epoch": 0.69, + "grad_norm": 2.389164282598154, + "learning_rate": 4.576796228678744e-06, + "loss": 0.8093, + "step": 29372 + }, + { + "epoch": 0.69, + "grad_norm": 1.9838057831749392, + "learning_rate": 4.576155158906457e-06, + "loss": 1.1504, + "step": 29373 + }, + { + "epoch": 0.69, + "grad_norm": 1.8939748026743193, + "learning_rate": 4.575514120713513e-06, + "loss": 0.8982, + "step": 29374 + }, + { + "epoch": 0.69, + "grad_norm": 1.1507533971610282, + "learning_rate": 4.574873114103645e-06, + "loss": 0.9262, + "step": 29375 + }, + { + "epoch": 0.69, + "grad_norm": 2.1061185549973636, + "learning_rate": 4.574232139080586e-06, + "loss": 1.0756, + "step": 29376 + }, + { + "epoch": 0.69, + "grad_norm": 1.8513873663548208, + "learning_rate": 4.573591195648059e-06, + "loss": 0.9033, + "step": 29377 + }, + { + "epoch": 0.69, + "grad_norm": 1.0559290452785477, + "learning_rate": 4.572950283809803e-06, + "loss": 0.942, + "step": 29378 + }, + { + "epoch": 0.69, + "grad_norm": 2.0443996332553596, + "learning_rate": 4.572309403569552e-06, + "loss": 0.9453, + "step": 29379 + }, + { + "epoch": 0.69, + "grad_norm": 2.4098763886885304, + "learning_rate": 4.571668554931035e-06, + "loss": 0.9396, + "step": 29380 + }, + { + "epoch": 0.69, + "grad_norm": 2.104833828141606, + "learning_rate": 4.571027737897979e-06, + "loss": 0.9607, + "step": 29381 + }, + { + "epoch": 0.69, + "grad_norm": 2.3059695296768092, + "learning_rate": 4.570386952474116e-06, + "loss": 1.0276, + "step": 29382 + }, + { + "epoch": 0.69, + "grad_norm": 1.1012228854821688, + "learning_rate": 4.569746198663184e-06, + "loss": 1.0306, + "step": 29383 + }, + { + "epoch": 0.69, + "grad_norm": 3.0722499514772306, + "learning_rate": 4.569105476468909e-06, + "loss": 0.9358, + "step": 29384 + }, + { + "epoch": 0.69, + "grad_norm": 2.087957557434159, + "learning_rate": 4.568464785895018e-06, + "loss": 0.9511, + "step": 29385 + }, + { + "epoch": 0.69, + "grad_norm": 2.394709087320821, + "learning_rate": 4.567824126945248e-06, + "loss": 0.9425, + "step": 29386 + }, + { + "epoch": 0.69, + "grad_norm": 2.0025310443390056, + "learning_rate": 4.567183499623322e-06, + "loss": 0.9397, + "step": 29387 + }, + { + "epoch": 0.69, + "grad_norm": 2.105107008856554, + "learning_rate": 4.5665429039329735e-06, + "loss": 0.9315, + "step": 29388 + }, + { + "epoch": 0.69, + "grad_norm": 2.0236483959026543, + "learning_rate": 4.565902339877936e-06, + "loss": 0.872, + "step": 29389 + }, + { + "epoch": 0.69, + "grad_norm": 2.045044593979432, + "learning_rate": 4.5652618074619355e-06, + "loss": 0.9004, + "step": 29390 + }, + { + "epoch": 0.69, + "grad_norm": 2.101811536605331, + "learning_rate": 4.564621306688698e-06, + "loss": 1.0867, + "step": 29391 + }, + { + "epoch": 0.69, + "grad_norm": 1.8859432836515413, + "learning_rate": 4.563980837561955e-06, + "loss": 0.9647, + "step": 29392 + }, + { + "epoch": 0.69, + "grad_norm": 2.2732923753658425, + "learning_rate": 4.563340400085445e-06, + "loss": 0.9536, + "step": 29393 + }, + { + "epoch": 0.69, + "grad_norm": 2.474358098415479, + "learning_rate": 4.562699994262881e-06, + "loss": 0.9574, + "step": 29394 + }, + { + "epoch": 0.69, + "grad_norm": 1.9951533552006036, + "learning_rate": 4.5620596200979996e-06, + "loss": 1.0656, + "step": 29395 + }, + { + "epoch": 0.69, + "grad_norm": 2.3959311372897116, + "learning_rate": 4.561419277594528e-06, + "loss": 1.0985, + "step": 29396 + }, + { + "epoch": 0.69, + "grad_norm": 3.218156647041317, + "learning_rate": 4.560778966756203e-06, + "loss": 0.8858, + "step": 29397 + }, + { + "epoch": 0.69, + "grad_norm": 2.2677212190698195, + "learning_rate": 4.560138687586737e-06, + "loss": 1.0213, + "step": 29398 + }, + { + "epoch": 0.69, + "grad_norm": 2.102009796115186, + "learning_rate": 4.5594984400898655e-06, + "loss": 0.9894, + "step": 29399 + }, + { + "epoch": 0.69, + "grad_norm": 1.8806325732653424, + "learning_rate": 4.5588582242693215e-06, + "loss": 0.9633, + "step": 29400 + }, + { + "epoch": 0.69, + "grad_norm": 2.0510706363587303, + "learning_rate": 4.558218040128822e-06, + "loss": 0.966, + "step": 29401 + }, + { + "epoch": 0.69, + "grad_norm": 1.0885177960518242, + "learning_rate": 4.5575778876721055e-06, + "loss": 0.8975, + "step": 29402 + }, + { + "epoch": 0.69, + "grad_norm": 1.9406774453414855, + "learning_rate": 4.556937766902888e-06, + "loss": 0.9506, + "step": 29403 + }, + { + "epoch": 0.69, + "grad_norm": 1.9756635274938592, + "learning_rate": 4.556297677824907e-06, + "loss": 0.9155, + "step": 29404 + }, + { + "epoch": 0.69, + "grad_norm": 2.226867404983209, + "learning_rate": 4.555657620441881e-06, + "loss": 0.9293, + "step": 29405 + }, + { + "epoch": 0.69, + "grad_norm": 1.9772664434994134, + "learning_rate": 4.555017594757543e-06, + "loss": 1.088, + "step": 29406 + }, + { + "epoch": 0.69, + "grad_norm": 1.9593266663966737, + "learning_rate": 4.554377600775617e-06, + "loss": 1.0415, + "step": 29407 + }, + { + "epoch": 0.69, + "grad_norm": 2.0160383838582234, + "learning_rate": 4.553737638499824e-06, + "loss": 1.0721, + "step": 29408 + }, + { + "epoch": 0.69, + "grad_norm": 1.1168610898972093, + "learning_rate": 4.553097707933894e-06, + "loss": 0.9623, + "step": 29409 + }, + { + "epoch": 0.69, + "grad_norm": 2.573478781049696, + "learning_rate": 4.55245780908156e-06, + "loss": 0.9731, + "step": 29410 + }, + { + "epoch": 0.69, + "grad_norm": 1.9709664668359088, + "learning_rate": 4.551817941946539e-06, + "loss": 1.0506, + "step": 29411 + }, + { + "epoch": 0.69, + "grad_norm": 1.8077012698264767, + "learning_rate": 4.5511781065325564e-06, + "loss": 1.0465, + "step": 29412 + }, + { + "epoch": 0.69, + "grad_norm": 2.186646497848986, + "learning_rate": 4.550538302843339e-06, + "loss": 0.9893, + "step": 29413 + }, + { + "epoch": 0.69, + "grad_norm": 2.146368665978109, + "learning_rate": 4.5498985308826174e-06, + "loss": 1.0078, + "step": 29414 + }, + { + "epoch": 0.69, + "grad_norm": 2.244234058289754, + "learning_rate": 4.549258790654112e-06, + "loss": 1.1113, + "step": 29415 + }, + { + "epoch": 0.69, + "grad_norm": 1.0975763676422126, + "learning_rate": 4.548619082161543e-06, + "loss": 0.9673, + "step": 29416 + }, + { + "epoch": 0.69, + "grad_norm": 1.9339129683550738, + "learning_rate": 4.54797940540864e-06, + "loss": 0.9524, + "step": 29417 + }, + { + "epoch": 0.69, + "grad_norm": 2.3605876425955534, + "learning_rate": 4.547339760399131e-06, + "loss": 0.9533, + "step": 29418 + }, + { + "epoch": 0.69, + "grad_norm": 2.102303424410892, + "learning_rate": 4.5467001471367355e-06, + "loss": 0.9255, + "step": 29419 + }, + { + "epoch": 0.69, + "grad_norm": 1.8413170294870636, + "learning_rate": 4.546060565625175e-06, + "loss": 0.9429, + "step": 29420 + }, + { + "epoch": 0.69, + "grad_norm": 1.8600312959780174, + "learning_rate": 4.545421015868181e-06, + "loss": 0.8851, + "step": 29421 + }, + { + "epoch": 0.69, + "grad_norm": 2.404019448470323, + "learning_rate": 4.5447814978694685e-06, + "loss": 0.8481, + "step": 29422 + }, + { + "epoch": 0.69, + "grad_norm": 2.0005274347026343, + "learning_rate": 4.544142011632769e-06, + "loss": 1.0224, + "step": 29423 + }, + { + "epoch": 0.69, + "grad_norm": 1.8995523908032246, + "learning_rate": 4.543502557161797e-06, + "loss": 0.898, + "step": 29424 + }, + { + "epoch": 0.69, + "grad_norm": 1.1014612608745726, + "learning_rate": 4.542863134460286e-06, + "loss": 0.955, + "step": 29425 + }, + { + "epoch": 0.69, + "grad_norm": 2.8213211882922056, + "learning_rate": 4.542223743531949e-06, + "loss": 0.9037, + "step": 29426 + }, + { + "epoch": 0.69, + "grad_norm": 1.7968574898841034, + "learning_rate": 4.541584384380518e-06, + "loss": 0.8854, + "step": 29427 + }, + { + "epoch": 0.69, + "grad_norm": 2.148233858774401, + "learning_rate": 4.540945057009709e-06, + "loss": 0.9072, + "step": 29428 + }, + { + "epoch": 0.69, + "grad_norm": 1.0854289507292822, + "learning_rate": 4.540305761423244e-06, + "loss": 0.9672, + "step": 29429 + }, + { + "epoch": 0.69, + "grad_norm": 1.9820285205447303, + "learning_rate": 4.5396664976248476e-06, + "loss": 0.9938, + "step": 29430 + }, + { + "epoch": 0.69, + "grad_norm": 1.8713025971261523, + "learning_rate": 4.539027265618244e-06, + "loss": 0.9688, + "step": 29431 + }, + { + "epoch": 0.69, + "grad_norm": 1.9056210759199679, + "learning_rate": 4.5383880654071525e-06, + "loss": 1.0521, + "step": 29432 + }, + { + "epoch": 0.69, + "grad_norm": 1.9576420181891476, + "learning_rate": 4.537748896995292e-06, + "loss": 0.9577, + "step": 29433 + }, + { + "epoch": 0.69, + "grad_norm": 4.962350883731299, + "learning_rate": 4.537109760386387e-06, + "loss": 0.9294, + "step": 29434 + }, + { + "epoch": 0.69, + "grad_norm": 2.0581734963370644, + "learning_rate": 4.536470655584161e-06, + "loss": 1.0202, + "step": 29435 + }, + { + "epoch": 0.69, + "grad_norm": 1.1541034740997689, + "learning_rate": 4.535831582592333e-06, + "loss": 0.9107, + "step": 29436 + }, + { + "epoch": 0.69, + "grad_norm": 2.0333229755293996, + "learning_rate": 4.53519254141462e-06, + "loss": 0.8834, + "step": 29437 + }, + { + "epoch": 0.69, + "grad_norm": 1.1437723321604019, + "learning_rate": 4.5345535320547465e-06, + "loss": 0.9261, + "step": 29438 + }, + { + "epoch": 0.69, + "grad_norm": 1.9085233417479357, + "learning_rate": 4.533914554516435e-06, + "loss": 0.8912, + "step": 29439 + }, + { + "epoch": 0.69, + "grad_norm": 2.123324941794163, + "learning_rate": 4.533275608803405e-06, + "loss": 1.1748, + "step": 29440 + }, + { + "epoch": 0.69, + "grad_norm": 1.9151938919654092, + "learning_rate": 4.532636694919371e-06, + "loss": 1.0454, + "step": 29441 + }, + { + "epoch": 0.69, + "grad_norm": 2.0240276404795177, + "learning_rate": 4.531997812868061e-06, + "loss": 1.0, + "step": 29442 + }, + { + "epoch": 0.69, + "grad_norm": 1.9347804905290293, + "learning_rate": 4.5313589626531864e-06, + "loss": 1.1541, + "step": 29443 + }, + { + "epoch": 0.69, + "grad_norm": 1.9408912717434654, + "learning_rate": 4.5307201442784755e-06, + "loss": 0.8889, + "step": 29444 + }, + { + "epoch": 0.69, + "grad_norm": 2.091311540496909, + "learning_rate": 4.53008135774764e-06, + "loss": 0.947, + "step": 29445 + }, + { + "epoch": 0.69, + "grad_norm": 1.9447050614790433, + "learning_rate": 4.529442603064406e-06, + "loss": 1.0427, + "step": 29446 + }, + { + "epoch": 0.69, + "grad_norm": 1.0391372902697793, + "learning_rate": 4.528803880232485e-06, + "loss": 1.0374, + "step": 29447 + }, + { + "epoch": 0.69, + "grad_norm": 2.3948215872366645, + "learning_rate": 4.528165189255605e-06, + "loss": 0.9613, + "step": 29448 + }, + { + "epoch": 0.69, + "grad_norm": 1.946800628321458, + "learning_rate": 4.527526530137478e-06, + "loss": 1.0526, + "step": 29449 + }, + { + "epoch": 0.69, + "grad_norm": 2.0807396606881716, + "learning_rate": 4.526887902881822e-06, + "loss": 0.9351, + "step": 29450 + }, + { + "epoch": 0.69, + "grad_norm": 2.0380844959067494, + "learning_rate": 4.526249307492355e-06, + "loss": 1.0695, + "step": 29451 + }, + { + "epoch": 0.69, + "grad_norm": 2.564314466502597, + "learning_rate": 4.525610743972804e-06, + "loss": 0.9775, + "step": 29452 + }, + { + "epoch": 0.69, + "grad_norm": 2.0463433391931143, + "learning_rate": 4.524972212326878e-06, + "loss": 1.0111, + "step": 29453 + }, + { + "epoch": 0.69, + "grad_norm": 2.3077569319707307, + "learning_rate": 4.524333712558294e-06, + "loss": 0.8264, + "step": 29454 + }, + { + "epoch": 0.69, + "grad_norm": 1.7469625633702435, + "learning_rate": 4.5236952446707735e-06, + "loss": 1.0964, + "step": 29455 + }, + { + "epoch": 0.69, + "grad_norm": 2.270604586128371, + "learning_rate": 4.523056808668037e-06, + "loss": 1.1064, + "step": 29456 + }, + { + "epoch": 0.69, + "grad_norm": 2.043252948883829, + "learning_rate": 4.5224184045537964e-06, + "loss": 1.0879, + "step": 29457 + }, + { + "epoch": 0.69, + "grad_norm": 1.9329105191212848, + "learning_rate": 4.521780032331768e-06, + "loss": 1.038, + "step": 29458 + }, + { + "epoch": 0.69, + "grad_norm": 2.060604337370133, + "learning_rate": 4.521141692005669e-06, + "loss": 0.9759, + "step": 29459 + }, + { + "epoch": 0.69, + "grad_norm": 2.246513741244407, + "learning_rate": 4.5205033835792226e-06, + "loss": 1.0186, + "step": 29460 + }, + { + "epoch": 0.69, + "grad_norm": 2.270865406107587, + "learning_rate": 4.519865107056139e-06, + "loss": 0.9543, + "step": 29461 + }, + { + "epoch": 0.69, + "grad_norm": 2.0344713829916197, + "learning_rate": 4.519226862440134e-06, + "loss": 0.9175, + "step": 29462 + }, + { + "epoch": 0.69, + "grad_norm": 2.1765623899132565, + "learning_rate": 4.518588649734927e-06, + "loss": 0.9645, + "step": 29463 + }, + { + "epoch": 0.69, + "grad_norm": 1.89156699444425, + "learning_rate": 4.51795046894423e-06, + "loss": 0.9104, + "step": 29464 + }, + { + "epoch": 0.69, + "grad_norm": 1.177940631514222, + "learning_rate": 4.517312320071764e-06, + "loss": 0.9564, + "step": 29465 + }, + { + "epoch": 0.69, + "grad_norm": 1.9510387389790453, + "learning_rate": 4.5166742031212375e-06, + "loss": 1.0121, + "step": 29466 + }, + { + "epoch": 0.69, + "grad_norm": 1.9727402760059494, + "learning_rate": 4.516036118096374e-06, + "loss": 0.9906, + "step": 29467 + }, + { + "epoch": 0.69, + "grad_norm": 2.1727391462498957, + "learning_rate": 4.515398065000881e-06, + "loss": 0.965, + "step": 29468 + }, + { + "epoch": 0.69, + "grad_norm": 2.0678493568984884, + "learning_rate": 4.51476004383848e-06, + "loss": 1.0592, + "step": 29469 + }, + { + "epoch": 0.69, + "grad_norm": 2.0140548236354197, + "learning_rate": 4.514122054612883e-06, + "loss": 0.9751, + "step": 29470 + }, + { + "epoch": 0.69, + "grad_norm": 2.523125319354786, + "learning_rate": 4.5134840973278e-06, + "loss": 1.1552, + "step": 29471 + }, + { + "epoch": 0.69, + "grad_norm": 1.998582393559585, + "learning_rate": 4.51284617198695e-06, + "loss": 0.8499, + "step": 29472 + }, + { + "epoch": 0.69, + "grad_norm": 1.1826590812926787, + "learning_rate": 4.51220827859405e-06, + "loss": 0.9276, + "step": 29473 + }, + { + "epoch": 0.69, + "grad_norm": 1.8472153387567871, + "learning_rate": 4.5115704171528105e-06, + "loss": 0.919, + "step": 29474 + }, + { + "epoch": 0.69, + "grad_norm": 1.9046668951300472, + "learning_rate": 4.510932587666943e-06, + "loss": 1.1453, + "step": 29475 + }, + { + "epoch": 0.69, + "grad_norm": 2.0838130323658435, + "learning_rate": 4.510294790140164e-06, + "loss": 1.0044, + "step": 29476 + }, + { + "epoch": 0.69, + "grad_norm": 1.887530412023358, + "learning_rate": 4.509657024576189e-06, + "loss": 1.1012, + "step": 29477 + }, + { + "epoch": 0.69, + "grad_norm": 1.8018715794112705, + "learning_rate": 4.50901929097873e-06, + "loss": 0.8972, + "step": 29478 + }, + { + "epoch": 0.69, + "grad_norm": 2.0830870889134774, + "learning_rate": 4.508381589351496e-06, + "loss": 0.9154, + "step": 29479 + }, + { + "epoch": 0.69, + "grad_norm": 2.220669597160426, + "learning_rate": 4.507743919698202e-06, + "loss": 0.9783, + "step": 29480 + }, + { + "epoch": 0.69, + "grad_norm": 1.8825266392789282, + "learning_rate": 4.507106282022565e-06, + "loss": 0.9834, + "step": 29481 + }, + { + "epoch": 0.69, + "grad_norm": 1.897405395036017, + "learning_rate": 4.506468676328295e-06, + "loss": 0.9629, + "step": 29482 + }, + { + "epoch": 0.69, + "grad_norm": 1.943260399416211, + "learning_rate": 4.5058311026191e-06, + "loss": 0.9202, + "step": 29483 + }, + { + "epoch": 0.69, + "grad_norm": 1.8975308764178038, + "learning_rate": 4.5051935608986994e-06, + "loss": 1.1505, + "step": 29484 + }, + { + "epoch": 0.69, + "grad_norm": 2.057926732353467, + "learning_rate": 4.504556051170797e-06, + "loss": 1.0279, + "step": 29485 + }, + { + "epoch": 0.69, + "grad_norm": 2.0804328838682964, + "learning_rate": 4.50391857343911e-06, + "loss": 1.0233, + "step": 29486 + }, + { + "epoch": 0.69, + "grad_norm": 1.1664920527245952, + "learning_rate": 4.5032811277073525e-06, + "loss": 0.9154, + "step": 29487 + }, + { + "epoch": 0.69, + "grad_norm": 2.455205570079076, + "learning_rate": 4.502643713979232e-06, + "loss": 1.0767, + "step": 29488 + }, + { + "epoch": 0.69, + "grad_norm": 1.937034697368711, + "learning_rate": 4.5020063322584585e-06, + "loss": 1.086, + "step": 29489 + }, + { + "epoch": 0.69, + "grad_norm": 2.5680874363272097, + "learning_rate": 4.501368982548743e-06, + "loss": 1.0028, + "step": 29490 + }, + { + "epoch": 0.69, + "grad_norm": 1.9450751957247139, + "learning_rate": 4.500731664853801e-06, + "loss": 1.0024, + "step": 29491 + }, + { + "epoch": 0.69, + "grad_norm": 2.0436365634347062, + "learning_rate": 4.500094379177342e-06, + "loss": 0.9897, + "step": 29492 + }, + { + "epoch": 0.69, + "grad_norm": 2.1543400830166393, + "learning_rate": 4.499457125523071e-06, + "loss": 1.14, + "step": 29493 + }, + { + "epoch": 0.69, + "grad_norm": 1.9565425667644532, + "learning_rate": 4.498819903894701e-06, + "loss": 0.869, + "step": 29494 + }, + { + "epoch": 0.69, + "grad_norm": 2.0372770323662377, + "learning_rate": 4.498182714295951e-06, + "loss": 0.8819, + "step": 29495 + }, + { + "epoch": 0.69, + "grad_norm": 2.048818485389157, + "learning_rate": 4.497545556730516e-06, + "loss": 1.1121, + "step": 29496 + }, + { + "epoch": 0.69, + "grad_norm": 1.9254192134463906, + "learning_rate": 4.496908431202113e-06, + "loss": 0.9764, + "step": 29497 + }, + { + "epoch": 0.69, + "grad_norm": 4.163235053710152, + "learning_rate": 4.49627133771445e-06, + "loss": 0.9159, + "step": 29498 + }, + { + "epoch": 0.69, + "grad_norm": 2.657733598811484, + "learning_rate": 4.495634276271244e-06, + "loss": 1.0071, + "step": 29499 + }, + { + "epoch": 0.69, + "grad_norm": 1.8199064876767055, + "learning_rate": 4.494997246876196e-06, + "loss": 0.8709, + "step": 29500 + }, + { + "epoch": 0.7, + "grad_norm": 1.926870010430436, + "learning_rate": 4.494360249533013e-06, + "loss": 1.0442, + "step": 29501 + }, + { + "epoch": 0.7, + "grad_norm": 2.1239022642408116, + "learning_rate": 4.493723284245413e-06, + "loss": 1.1013, + "step": 29502 + }, + { + "epoch": 0.7, + "grad_norm": 1.93656493487312, + "learning_rate": 4.493086351017093e-06, + "loss": 1.0738, + "step": 29503 + }, + { + "epoch": 0.7, + "grad_norm": 1.1262844978505024, + "learning_rate": 4.4924494498517735e-06, + "loss": 0.9599, + "step": 29504 + }, + { + "epoch": 0.7, + "grad_norm": 1.0400159251187038, + "learning_rate": 4.491812580753152e-06, + "loss": 0.9277, + "step": 29505 + }, + { + "epoch": 0.7, + "grad_norm": 1.0247597622480125, + "learning_rate": 4.491175743724947e-06, + "loss": 0.9409, + "step": 29506 + }, + { + "epoch": 0.7, + "grad_norm": 1.9582199595336687, + "learning_rate": 4.490538938770855e-06, + "loss": 1.1074, + "step": 29507 + }, + { + "epoch": 0.7, + "grad_norm": 1.8327270550904653, + "learning_rate": 4.489902165894595e-06, + "loss": 0.9584, + "step": 29508 + }, + { + "epoch": 0.7, + "grad_norm": 2.5376095990796723, + "learning_rate": 4.489265425099868e-06, + "loss": 0.9385, + "step": 29509 + }, + { + "epoch": 0.7, + "grad_norm": 1.9861470596388808, + "learning_rate": 4.48862871639038e-06, + "loss": 1.0524, + "step": 29510 + }, + { + "epoch": 0.7, + "grad_norm": 2.040937725889453, + "learning_rate": 4.487992039769838e-06, + "loss": 1.0955, + "step": 29511 + }, + { + "epoch": 0.7, + "grad_norm": 2.1844740239103744, + "learning_rate": 4.487355395241958e-06, + "loss": 1.046, + "step": 29512 + }, + { + "epoch": 0.7, + "grad_norm": 1.7748009623887975, + "learning_rate": 4.486718782810439e-06, + "loss": 1.0167, + "step": 29513 + }, + { + "epoch": 0.7, + "grad_norm": 2.70452550634026, + "learning_rate": 4.486082202478984e-06, + "loss": 0.8679, + "step": 29514 + }, + { + "epoch": 0.7, + "grad_norm": 1.812201898694704, + "learning_rate": 4.485445654251305e-06, + "loss": 0.8936, + "step": 29515 + }, + { + "epoch": 0.7, + "grad_norm": 1.9676435110585075, + "learning_rate": 4.484809138131111e-06, + "loss": 0.9669, + "step": 29516 + }, + { + "epoch": 0.7, + "grad_norm": 1.7391691419551485, + "learning_rate": 4.484172654122104e-06, + "loss": 1.019, + "step": 29517 + }, + { + "epoch": 0.7, + "grad_norm": 2.1345795467311177, + "learning_rate": 4.483536202227987e-06, + "loss": 0.9287, + "step": 29518 + }, + { + "epoch": 0.7, + "grad_norm": 2.011204745773445, + "learning_rate": 4.482899782452468e-06, + "loss": 0.9655, + "step": 29519 + }, + { + "epoch": 0.7, + "grad_norm": 2.079328534900915, + "learning_rate": 4.482263394799258e-06, + "loss": 1.0269, + "step": 29520 + }, + { + "epoch": 0.7, + "grad_norm": 1.9920448734263763, + "learning_rate": 4.481627039272056e-06, + "loss": 1.0291, + "step": 29521 + }, + { + "epoch": 0.7, + "grad_norm": 1.8847559015734783, + "learning_rate": 4.480990715874564e-06, + "loss": 0.8167, + "step": 29522 + }, + { + "epoch": 0.7, + "grad_norm": 1.0892253239721517, + "learning_rate": 4.480354424610498e-06, + "loss": 0.9184, + "step": 29523 + }, + { + "epoch": 0.7, + "grad_norm": 1.9209396723809555, + "learning_rate": 4.47971816548355e-06, + "loss": 1.0511, + "step": 29524 + }, + { + "epoch": 0.7, + "grad_norm": 1.9169670336283233, + "learning_rate": 4.479081938497435e-06, + "loss": 0.9282, + "step": 29525 + }, + { + "epoch": 0.7, + "grad_norm": 2.021310096866815, + "learning_rate": 4.478445743655849e-06, + "loss": 0.9304, + "step": 29526 + }, + { + "epoch": 0.7, + "grad_norm": 1.9008200394667336, + "learning_rate": 4.477809580962503e-06, + "loss": 1.0253, + "step": 29527 + }, + { + "epoch": 0.7, + "grad_norm": 2.2557133189722673, + "learning_rate": 4.477173450421096e-06, + "loss": 0.8696, + "step": 29528 + }, + { + "epoch": 0.7, + "grad_norm": 1.961148857722176, + "learning_rate": 4.476537352035336e-06, + "loss": 0.931, + "step": 29529 + }, + { + "epoch": 0.7, + "grad_norm": 1.9039809346372416, + "learning_rate": 4.475901285808924e-06, + "loss": 0.9458, + "step": 29530 + }, + { + "epoch": 0.7, + "grad_norm": 1.93053580544536, + "learning_rate": 4.4752652517455605e-06, + "loss": 0.8003, + "step": 29531 + }, + { + "epoch": 0.7, + "grad_norm": 1.8117169506893542, + "learning_rate": 4.4746292498489515e-06, + "loss": 0.9561, + "step": 29532 + }, + { + "epoch": 0.7, + "grad_norm": 1.9455051926320535, + "learning_rate": 4.4739932801228046e-06, + "loss": 0.8532, + "step": 29533 + }, + { + "epoch": 0.7, + "grad_norm": 1.8714390246490638, + "learning_rate": 4.473357342570819e-06, + "loss": 0.962, + "step": 29534 + }, + { + "epoch": 0.7, + "grad_norm": 1.7633206600239486, + "learning_rate": 4.472721437196692e-06, + "loss": 1.0294, + "step": 29535 + }, + { + "epoch": 0.7, + "grad_norm": 2.2557738698715233, + "learning_rate": 4.472085564004131e-06, + "loss": 0.9101, + "step": 29536 + }, + { + "epoch": 0.7, + "grad_norm": 2.006614062824722, + "learning_rate": 4.47144972299684e-06, + "loss": 0.9067, + "step": 29537 + }, + { + "epoch": 0.7, + "grad_norm": 2.2298790062068887, + "learning_rate": 4.470813914178521e-06, + "loss": 1.0562, + "step": 29538 + }, + { + "epoch": 0.7, + "grad_norm": 2.0678429776186507, + "learning_rate": 4.47017813755287e-06, + "loss": 1.0354, + "step": 29539 + }, + { + "epoch": 0.7, + "grad_norm": 1.139836300044621, + "learning_rate": 4.4695423931235915e-06, + "loss": 0.8772, + "step": 29540 + }, + { + "epoch": 0.7, + "grad_norm": 1.903738835500599, + "learning_rate": 4.4689066808943925e-06, + "loss": 0.9662, + "step": 29541 + }, + { + "epoch": 0.7, + "grad_norm": 1.0942727546929976, + "learning_rate": 4.46827100086897e-06, + "loss": 1.0291, + "step": 29542 + }, + { + "epoch": 0.7, + "grad_norm": 2.186770990992331, + "learning_rate": 4.467635353051021e-06, + "loss": 1.104, + "step": 29543 + }, + { + "epoch": 0.7, + "grad_norm": 1.0555597096245843, + "learning_rate": 4.466999737444254e-06, + "loss": 0.9136, + "step": 29544 + }, + { + "epoch": 0.7, + "grad_norm": 1.8646430564945375, + "learning_rate": 4.466364154052364e-06, + "loss": 0.8965, + "step": 29545 + }, + { + "epoch": 0.7, + "grad_norm": 1.995277439661138, + "learning_rate": 4.4657286028790555e-06, + "loss": 0.9108, + "step": 29546 + }, + { + "epoch": 0.7, + "grad_norm": 2.0063041041473175, + "learning_rate": 4.465093083928025e-06, + "loss": 0.9099, + "step": 29547 + }, + { + "epoch": 0.7, + "grad_norm": 1.9431619147282044, + "learning_rate": 4.464457597202978e-06, + "loss": 0.9327, + "step": 29548 + }, + { + "epoch": 0.7, + "grad_norm": 2.6991760541910788, + "learning_rate": 4.4638221427076075e-06, + "loss": 0.8565, + "step": 29549 + }, + { + "epoch": 0.7, + "grad_norm": 1.9257010988246248, + "learning_rate": 4.4631867204456215e-06, + "loss": 0.9367, + "step": 29550 + }, + { + "epoch": 0.7, + "grad_norm": 1.9844383615417895, + "learning_rate": 4.462551330420715e-06, + "loss": 0.8755, + "step": 29551 + }, + { + "epoch": 0.7, + "grad_norm": 1.8363450041336815, + "learning_rate": 4.461915972636584e-06, + "loss": 0.9041, + "step": 29552 + }, + { + "epoch": 0.7, + "grad_norm": 2.1502944247410944, + "learning_rate": 4.461280647096932e-06, + "loss": 0.9608, + "step": 29553 + }, + { + "epoch": 0.7, + "grad_norm": 1.0901472297656545, + "learning_rate": 4.46064535380546e-06, + "loss": 0.9613, + "step": 29554 + }, + { + "epoch": 0.7, + "grad_norm": 2.09241538126597, + "learning_rate": 4.460010092765866e-06, + "loss": 0.9731, + "step": 29555 + }, + { + "epoch": 0.7, + "grad_norm": 1.9979784551541986, + "learning_rate": 4.4593748639818425e-06, + "loss": 0.9799, + "step": 29556 + }, + { + "epoch": 0.7, + "grad_norm": 2.4160940548626093, + "learning_rate": 4.458739667457094e-06, + "loss": 0.858, + "step": 29557 + }, + { + "epoch": 0.7, + "grad_norm": 1.9285581914559253, + "learning_rate": 4.4581045031953195e-06, + "loss": 0.9049, + "step": 29558 + }, + { + "epoch": 0.7, + "grad_norm": 2.038705976876727, + "learning_rate": 4.457469371200216e-06, + "loss": 1.0308, + "step": 29559 + }, + { + "epoch": 0.7, + "grad_norm": 1.8868533144720632, + "learning_rate": 4.456834271475477e-06, + "loss": 0.8981, + "step": 29560 + }, + { + "epoch": 0.7, + "grad_norm": 1.9405775079280327, + "learning_rate": 4.456199204024803e-06, + "loss": 1.0035, + "step": 29561 + }, + { + "epoch": 0.7, + "grad_norm": 1.8914892834646617, + "learning_rate": 4.455564168851897e-06, + "loss": 0.9232, + "step": 29562 + }, + { + "epoch": 0.7, + "grad_norm": 2.2980715248767476, + "learning_rate": 4.454929165960451e-06, + "loss": 1.0037, + "step": 29563 + }, + { + "epoch": 0.7, + "grad_norm": 2.1015738832930353, + "learning_rate": 4.45429419535416e-06, + "loss": 0.9526, + "step": 29564 + }, + { + "epoch": 0.7, + "grad_norm": 2.733305348025484, + "learning_rate": 4.453659257036727e-06, + "loss": 0.9914, + "step": 29565 + }, + { + "epoch": 0.7, + "grad_norm": 2.078971077762715, + "learning_rate": 4.453024351011842e-06, + "loss": 1.1204, + "step": 29566 + }, + { + "epoch": 0.7, + "grad_norm": 2.024588420170861, + "learning_rate": 4.452389477283209e-06, + "loss": 1.0953, + "step": 29567 + }, + { + "epoch": 0.7, + "grad_norm": 3.0208351889855, + "learning_rate": 4.4517546358545175e-06, + "loss": 1.0381, + "step": 29568 + }, + { + "epoch": 0.7, + "grad_norm": 1.054479017005868, + "learning_rate": 4.451119826729471e-06, + "loss": 0.9658, + "step": 29569 + }, + { + "epoch": 0.7, + "grad_norm": 1.1640300196158122, + "learning_rate": 4.450485049911757e-06, + "loss": 0.9883, + "step": 29570 + }, + { + "epoch": 0.7, + "grad_norm": 1.9525044467488988, + "learning_rate": 4.449850305405082e-06, + "loss": 1.0573, + "step": 29571 + }, + { + "epoch": 0.7, + "grad_norm": 1.775101851567332, + "learning_rate": 4.449215593213133e-06, + "loss": 0.935, + "step": 29572 + }, + { + "epoch": 0.7, + "grad_norm": 1.9668751334537913, + "learning_rate": 4.448580913339605e-06, + "loss": 1.1447, + "step": 29573 + }, + { + "epoch": 0.7, + "grad_norm": 1.736322922961635, + "learning_rate": 4.447946265788197e-06, + "loss": 0.959, + "step": 29574 + }, + { + "epoch": 0.7, + "grad_norm": 1.8267866042824243, + "learning_rate": 4.447311650562604e-06, + "loss": 0.8711, + "step": 29575 + }, + { + "epoch": 0.7, + "grad_norm": 1.1414546294814618, + "learning_rate": 4.446677067666527e-06, + "loss": 0.9212, + "step": 29576 + }, + { + "epoch": 0.7, + "grad_norm": 2.3590209834781724, + "learning_rate": 4.446042517103647e-06, + "loss": 0.9598, + "step": 29577 + }, + { + "epoch": 0.7, + "grad_norm": 1.7791876461030198, + "learning_rate": 4.445407998877667e-06, + "loss": 0.9979, + "step": 29578 + }, + { + "epoch": 0.7, + "grad_norm": 1.9502905492040308, + "learning_rate": 4.444773512992282e-06, + "loss": 0.9999, + "step": 29579 + }, + { + "epoch": 0.7, + "grad_norm": 2.63705114922931, + "learning_rate": 4.4441390594511855e-06, + "loss": 0.9689, + "step": 29580 + }, + { + "epoch": 0.7, + "grad_norm": 1.89076605578002, + "learning_rate": 4.443504638258066e-06, + "loss": 1.0727, + "step": 29581 + }, + { + "epoch": 0.7, + "grad_norm": 1.9757143394691659, + "learning_rate": 4.442870249416622e-06, + "loss": 1.0222, + "step": 29582 + }, + { + "epoch": 0.7, + "grad_norm": 1.906340086505541, + "learning_rate": 4.442235892930551e-06, + "loss": 0.9591, + "step": 29583 + }, + { + "epoch": 0.7, + "grad_norm": 1.965382856576646, + "learning_rate": 4.441601568803538e-06, + "loss": 0.9284, + "step": 29584 + }, + { + "epoch": 0.7, + "grad_norm": 1.9150267684022508, + "learning_rate": 4.440967277039284e-06, + "loss": 1.0529, + "step": 29585 + }, + { + "epoch": 0.7, + "grad_norm": 1.9327590709830331, + "learning_rate": 4.4403330176414795e-06, + "loss": 0.9143, + "step": 29586 + }, + { + "epoch": 0.7, + "grad_norm": 1.8977092416271775, + "learning_rate": 4.439698790613811e-06, + "loss": 1.0366, + "step": 29587 + }, + { + "epoch": 0.7, + "grad_norm": 2.026112432322201, + "learning_rate": 4.439064595959978e-06, + "loss": 0.8766, + "step": 29588 + }, + { + "epoch": 0.7, + "grad_norm": 2.1772655027216956, + "learning_rate": 4.4384304336836745e-06, + "loss": 1.0149, + "step": 29589 + }, + { + "epoch": 0.7, + "grad_norm": 2.062203423386756, + "learning_rate": 4.437796303788589e-06, + "loss": 0.8456, + "step": 29590 + }, + { + "epoch": 0.7, + "grad_norm": 1.942231009989828, + "learning_rate": 4.437162206278413e-06, + "loss": 0.9668, + "step": 29591 + }, + { + "epoch": 0.7, + "grad_norm": 2.637414608985941, + "learning_rate": 4.436528141156838e-06, + "loss": 0.9784, + "step": 29592 + }, + { + "epoch": 0.7, + "grad_norm": 1.9324493501578792, + "learning_rate": 4.435894108427564e-06, + "loss": 0.9335, + "step": 29593 + }, + { + "epoch": 0.7, + "grad_norm": 2.2291364237657145, + "learning_rate": 4.43526010809427e-06, + "loss": 1.1299, + "step": 29594 + }, + { + "epoch": 0.7, + "grad_norm": 1.0829775328096682, + "learning_rate": 4.434626140160654e-06, + "loss": 0.9723, + "step": 29595 + }, + { + "epoch": 0.7, + "grad_norm": 1.0567760975628882, + "learning_rate": 4.4339922046304055e-06, + "loss": 0.8749, + "step": 29596 + }, + { + "epoch": 0.7, + "grad_norm": 1.9827196097296718, + "learning_rate": 4.43335830150722e-06, + "loss": 0.8723, + "step": 29597 + }, + { + "epoch": 0.7, + "grad_norm": 2.339214220960232, + "learning_rate": 4.432724430794786e-06, + "loss": 0.9804, + "step": 29598 + }, + { + "epoch": 0.7, + "grad_norm": 1.9028041057872167, + "learning_rate": 4.432090592496788e-06, + "loss": 1.0603, + "step": 29599 + }, + { + "epoch": 0.7, + "grad_norm": 1.706147357820832, + "learning_rate": 4.431456786616927e-06, + "loss": 0.9905, + "step": 29600 + }, + { + "epoch": 0.7, + "grad_norm": 2.0072429535445138, + "learning_rate": 4.4308230131588826e-06, + "loss": 1.0673, + "step": 29601 + }, + { + "epoch": 0.7, + "grad_norm": 1.9542642863539883, + "learning_rate": 4.430189272126353e-06, + "loss": 0.9608, + "step": 29602 + }, + { + "epoch": 0.7, + "grad_norm": 2.384634897300797, + "learning_rate": 4.429555563523021e-06, + "loss": 0.9981, + "step": 29603 + }, + { + "epoch": 0.7, + "grad_norm": 1.9469767183281643, + "learning_rate": 4.4289218873525855e-06, + "loss": 0.9904, + "step": 29604 + }, + { + "epoch": 0.7, + "grad_norm": 1.8783426713440756, + "learning_rate": 4.428288243618725e-06, + "loss": 0.9652, + "step": 29605 + }, + { + "epoch": 0.7, + "grad_norm": 1.7119035069030863, + "learning_rate": 4.427654632325139e-06, + "loss": 0.9307, + "step": 29606 + }, + { + "epoch": 0.7, + "grad_norm": 1.0715150674061793, + "learning_rate": 4.427021053475512e-06, + "loss": 0.9588, + "step": 29607 + }, + { + "epoch": 0.7, + "grad_norm": 1.838993917865084, + "learning_rate": 4.426387507073529e-06, + "loss": 0.956, + "step": 29608 + }, + { + "epoch": 0.7, + "grad_norm": 2.206115163406468, + "learning_rate": 4.425753993122882e-06, + "loss": 1.0287, + "step": 29609 + }, + { + "epoch": 0.7, + "grad_norm": 1.9302115797722608, + "learning_rate": 4.425120511627263e-06, + "loss": 0.9109, + "step": 29610 + }, + { + "epoch": 0.7, + "grad_norm": 1.944419751499525, + "learning_rate": 4.424487062590359e-06, + "loss": 0.8929, + "step": 29611 + }, + { + "epoch": 0.7, + "grad_norm": 2.3592941528088476, + "learning_rate": 4.423853646015852e-06, + "loss": 1.0877, + "step": 29612 + }, + { + "epoch": 0.7, + "grad_norm": 2.062973837717271, + "learning_rate": 4.423220261907435e-06, + "loss": 1.0147, + "step": 29613 + }, + { + "epoch": 0.7, + "grad_norm": 2.009146622151796, + "learning_rate": 4.422586910268798e-06, + "loss": 1.0517, + "step": 29614 + }, + { + "epoch": 0.7, + "grad_norm": 2.052034282259457, + "learning_rate": 4.421953591103627e-06, + "loss": 0.9711, + "step": 29615 + }, + { + "epoch": 0.7, + "grad_norm": 1.954230929131189, + "learning_rate": 4.421320304415604e-06, + "loss": 0.9346, + "step": 29616 + }, + { + "epoch": 0.7, + "grad_norm": 1.8604487019556633, + "learning_rate": 4.42068705020842e-06, + "loss": 1.07, + "step": 29617 + }, + { + "epoch": 0.7, + "grad_norm": 2.3924172042720055, + "learning_rate": 4.420053828485765e-06, + "loss": 1.0385, + "step": 29618 + }, + { + "epoch": 0.7, + "grad_norm": 1.951271883661139, + "learning_rate": 4.419420639251325e-06, + "loss": 1.0135, + "step": 29619 + }, + { + "epoch": 0.7, + "grad_norm": 2.4638390867382425, + "learning_rate": 4.4187874825087806e-06, + "loss": 0.9269, + "step": 29620 + }, + { + "epoch": 0.7, + "grad_norm": 2.519783459744628, + "learning_rate": 4.418154358261822e-06, + "loss": 0.994, + "step": 29621 + }, + { + "epoch": 0.7, + "grad_norm": 1.9138179366146204, + "learning_rate": 4.417521266514141e-06, + "loss": 1.0195, + "step": 29622 + }, + { + "epoch": 0.7, + "grad_norm": 1.9148732149537255, + "learning_rate": 4.416888207269418e-06, + "loss": 1.0185, + "step": 29623 + }, + { + "epoch": 0.7, + "grad_norm": 1.9425799251577205, + "learning_rate": 4.4162551805313345e-06, + "loss": 0.9763, + "step": 29624 + }, + { + "epoch": 0.7, + "grad_norm": 1.992811991362479, + "learning_rate": 4.4156221863035854e-06, + "loss": 1.0093, + "step": 29625 + }, + { + "epoch": 0.7, + "grad_norm": 2.0050539134401464, + "learning_rate": 4.414989224589848e-06, + "loss": 0.9012, + "step": 29626 + }, + { + "epoch": 0.7, + "grad_norm": 1.1310835559079362, + "learning_rate": 4.414356295393816e-06, + "loss": 0.9541, + "step": 29627 + }, + { + "epoch": 0.7, + "grad_norm": 2.1706264100618005, + "learning_rate": 4.413723398719167e-06, + "loss": 1.0294, + "step": 29628 + }, + { + "epoch": 0.7, + "grad_norm": 2.107842076014897, + "learning_rate": 4.413090534569591e-06, + "loss": 1.0432, + "step": 29629 + }, + { + "epoch": 0.7, + "grad_norm": 1.9147954796947486, + "learning_rate": 4.412457702948768e-06, + "loss": 1.0379, + "step": 29630 + }, + { + "epoch": 0.7, + "grad_norm": 1.9734469753436779, + "learning_rate": 4.411824903860387e-06, + "loss": 1.0777, + "step": 29631 + }, + { + "epoch": 0.7, + "grad_norm": 2.1529393965864623, + "learning_rate": 4.411192137308134e-06, + "loss": 1.0449, + "step": 29632 + }, + { + "epoch": 0.7, + "grad_norm": 1.915631247431694, + "learning_rate": 4.410559403295683e-06, + "loss": 1.0315, + "step": 29633 + }, + { + "epoch": 0.7, + "grad_norm": 2.0360896831514026, + "learning_rate": 4.409926701826726e-06, + "loss": 1.0145, + "step": 29634 + }, + { + "epoch": 0.7, + "grad_norm": 2.227050509324211, + "learning_rate": 4.4092940329049495e-06, + "loss": 1.0451, + "step": 29635 + }, + { + "epoch": 0.7, + "grad_norm": 2.4145956271513, + "learning_rate": 4.408661396534033e-06, + "loss": 0.9948, + "step": 29636 + }, + { + "epoch": 0.7, + "grad_norm": 2.030758626701164, + "learning_rate": 4.4080287927176565e-06, + "loss": 1.065, + "step": 29637 + }, + { + "epoch": 0.7, + "grad_norm": 2.349941583741833, + "learning_rate": 4.407396221459507e-06, + "loss": 1.0072, + "step": 29638 + }, + { + "epoch": 0.7, + "grad_norm": 1.156138137512125, + "learning_rate": 4.40676368276327e-06, + "loss": 0.9292, + "step": 29639 + }, + { + "epoch": 0.7, + "grad_norm": 1.876150046130645, + "learning_rate": 4.406131176632626e-06, + "loss": 0.8841, + "step": 29640 + }, + { + "epoch": 0.7, + "grad_norm": 1.1279048006606713, + "learning_rate": 4.405498703071254e-06, + "loss": 0.9008, + "step": 29641 + }, + { + "epoch": 0.7, + "grad_norm": 1.946477560044419, + "learning_rate": 4.404866262082839e-06, + "loss": 1.1104, + "step": 29642 + }, + { + "epoch": 0.7, + "grad_norm": 1.871032775196976, + "learning_rate": 4.404233853671069e-06, + "loss": 1.1111, + "step": 29643 + }, + { + "epoch": 0.7, + "grad_norm": 2.1272568478181335, + "learning_rate": 4.403601477839619e-06, + "loss": 0.9293, + "step": 29644 + }, + { + "epoch": 0.7, + "grad_norm": 1.968169178286017, + "learning_rate": 4.402969134592171e-06, + "loss": 1.006, + "step": 29645 + }, + { + "epoch": 0.7, + "grad_norm": 2.247008393067426, + "learning_rate": 4.402336823932412e-06, + "loss": 0.9385, + "step": 29646 + }, + { + "epoch": 0.7, + "grad_norm": 2.039770104161978, + "learning_rate": 4.401704545864016e-06, + "loss": 0.9872, + "step": 29647 + }, + { + "epoch": 0.7, + "grad_norm": 2.1228964656034854, + "learning_rate": 4.401072300390672e-06, + "loss": 1.0018, + "step": 29648 + }, + { + "epoch": 0.7, + "grad_norm": 2.1009691193500113, + "learning_rate": 4.4004400875160545e-06, + "loss": 0.9797, + "step": 29649 + }, + { + "epoch": 0.7, + "grad_norm": 2.5054696857685865, + "learning_rate": 4.399807907243851e-06, + "loss": 0.9446, + "step": 29650 + }, + { + "epoch": 0.7, + "grad_norm": 2.023412302497332, + "learning_rate": 4.399175759577737e-06, + "loss": 1.0207, + "step": 29651 + }, + { + "epoch": 0.7, + "grad_norm": 2.2416738359339123, + "learning_rate": 4.398543644521396e-06, + "loss": 1.0067, + "step": 29652 + }, + { + "epoch": 0.7, + "grad_norm": 1.8901410951516904, + "learning_rate": 4.397911562078508e-06, + "loss": 1.0573, + "step": 29653 + }, + { + "epoch": 0.7, + "grad_norm": 1.209307484327011, + "learning_rate": 4.39727951225275e-06, + "loss": 0.9678, + "step": 29654 + }, + { + "epoch": 0.7, + "grad_norm": 2.487306225846964, + "learning_rate": 4.3966474950478036e-06, + "loss": 1.1176, + "step": 29655 + }, + { + "epoch": 0.7, + "grad_norm": 2.050035099178008, + "learning_rate": 4.396015510467353e-06, + "loss": 0.9196, + "step": 29656 + }, + { + "epoch": 0.7, + "grad_norm": 1.8230857206049287, + "learning_rate": 4.395383558515075e-06, + "loss": 0.936, + "step": 29657 + }, + { + "epoch": 0.7, + "grad_norm": 2.0391840027320525, + "learning_rate": 4.394751639194644e-06, + "loss": 1.1101, + "step": 29658 + }, + { + "epoch": 0.7, + "grad_norm": 1.8054714531472817, + "learning_rate": 4.394119752509743e-06, + "loss": 1.0983, + "step": 29659 + }, + { + "epoch": 0.7, + "grad_norm": 1.859562787096935, + "learning_rate": 4.393487898464056e-06, + "loss": 0.8645, + "step": 29660 + }, + { + "epoch": 0.7, + "grad_norm": 2.145673522965517, + "learning_rate": 4.392856077061258e-06, + "loss": 0.983, + "step": 29661 + }, + { + "epoch": 0.7, + "grad_norm": 1.9064186927344644, + "learning_rate": 4.3922242883050226e-06, + "loss": 1.0499, + "step": 29662 + }, + { + "epoch": 0.7, + "grad_norm": 1.0832397314306002, + "learning_rate": 4.391592532199034e-06, + "loss": 0.9363, + "step": 29663 + }, + { + "epoch": 0.7, + "grad_norm": 1.1284565510461506, + "learning_rate": 4.390960808746971e-06, + "loss": 1.0122, + "step": 29664 + }, + { + "epoch": 0.7, + "grad_norm": 1.9857417451252637, + "learning_rate": 4.390329117952511e-06, + "loss": 1.1616, + "step": 29665 + }, + { + "epoch": 0.7, + "grad_norm": 1.1228854554724899, + "learning_rate": 4.389697459819327e-06, + "loss": 0.9219, + "step": 29666 + }, + { + "epoch": 0.7, + "grad_norm": 2.0003993544390877, + "learning_rate": 4.389065834351105e-06, + "loss": 1.0305, + "step": 29667 + }, + { + "epoch": 0.7, + "grad_norm": 1.8750389606679807, + "learning_rate": 4.3884342415515135e-06, + "loss": 0.9785, + "step": 29668 + }, + { + "epoch": 0.7, + "grad_norm": 2.1898634657497933, + "learning_rate": 4.387802681424234e-06, + "loss": 0.8939, + "step": 29669 + }, + { + "epoch": 0.7, + "grad_norm": 1.9529919672005682, + "learning_rate": 4.387171153972949e-06, + "loss": 0.9696, + "step": 29670 + }, + { + "epoch": 0.7, + "grad_norm": 2.071518187280204, + "learning_rate": 4.38653965920133e-06, + "loss": 0.9709, + "step": 29671 + }, + { + "epoch": 0.7, + "grad_norm": 2.0057224584000646, + "learning_rate": 4.3859081971130516e-06, + "loss": 0.9757, + "step": 29672 + }, + { + "epoch": 0.7, + "grad_norm": 2.0718855149698476, + "learning_rate": 4.385276767711792e-06, + "loss": 0.9042, + "step": 29673 + }, + { + "epoch": 0.7, + "grad_norm": 2.1468957830195983, + "learning_rate": 4.384645371001236e-06, + "loss": 1.0816, + "step": 29674 + }, + { + "epoch": 0.7, + "grad_norm": 2.225968924046808, + "learning_rate": 4.384014006985044e-06, + "loss": 1.0357, + "step": 29675 + }, + { + "epoch": 0.7, + "grad_norm": 1.0900629472174845, + "learning_rate": 4.383382675666903e-06, + "loss": 0.9028, + "step": 29676 + }, + { + "epoch": 0.7, + "grad_norm": 1.8421707740872502, + "learning_rate": 4.382751377050485e-06, + "loss": 0.9681, + "step": 29677 + }, + { + "epoch": 0.7, + "grad_norm": 1.9757030357366627, + "learning_rate": 4.382120111139474e-06, + "loss": 1.0621, + "step": 29678 + }, + { + "epoch": 0.7, + "grad_norm": 1.786344901130104, + "learning_rate": 4.3814888779375305e-06, + "loss": 1.0535, + "step": 29679 + }, + { + "epoch": 0.7, + "grad_norm": 2.2518864299818326, + "learning_rate": 4.380857677448338e-06, + "loss": 1.0587, + "step": 29680 + }, + { + "epoch": 0.7, + "grad_norm": 2.076039024612594, + "learning_rate": 4.380226509675576e-06, + "loss": 0.9516, + "step": 29681 + }, + { + "epoch": 0.7, + "grad_norm": 1.1378171189431354, + "learning_rate": 4.379595374622909e-06, + "loss": 0.9629, + "step": 29682 + }, + { + "epoch": 0.7, + "grad_norm": 2.5654208984885156, + "learning_rate": 4.378964272294022e-06, + "loss": 0.8901, + "step": 29683 + }, + { + "epoch": 0.7, + "grad_norm": 1.9668299276582442, + "learning_rate": 4.37833320269258e-06, + "loss": 0.9098, + "step": 29684 + }, + { + "epoch": 0.7, + "grad_norm": 1.8203179382761774, + "learning_rate": 4.377702165822266e-06, + "loss": 1.0348, + "step": 29685 + }, + { + "epoch": 0.7, + "grad_norm": 2.0752056371768592, + "learning_rate": 4.377071161686745e-06, + "loss": 1.038, + "step": 29686 + }, + { + "epoch": 0.7, + "grad_norm": 1.1236216330705648, + "learning_rate": 4.376440190289703e-06, + "loss": 0.8757, + "step": 29687 + }, + { + "epoch": 0.7, + "grad_norm": 2.0036267386338396, + "learning_rate": 4.375809251634804e-06, + "loss": 1.0654, + "step": 29688 + }, + { + "epoch": 0.7, + "grad_norm": 2.020814253471306, + "learning_rate": 4.37517834572572e-06, + "loss": 1.0594, + "step": 29689 + }, + { + "epoch": 0.7, + "grad_norm": 2.200564777471107, + "learning_rate": 4.374547472566129e-06, + "loss": 0.961, + "step": 29690 + }, + { + "epoch": 0.7, + "grad_norm": 1.8718833787101488, + "learning_rate": 4.373916632159708e-06, + "loss": 1.0236, + "step": 29691 + }, + { + "epoch": 0.7, + "grad_norm": 1.914499042563938, + "learning_rate": 4.373285824510125e-06, + "loss": 0.9979, + "step": 29692 + }, + { + "epoch": 0.7, + "grad_norm": 2.184724035810148, + "learning_rate": 4.372655049621051e-06, + "loss": 0.9359, + "step": 29693 + }, + { + "epoch": 0.7, + "grad_norm": 2.030896175269021, + "learning_rate": 4.372024307496159e-06, + "loss": 1.0285, + "step": 29694 + }, + { + "epoch": 0.7, + "grad_norm": 1.9237425903122578, + "learning_rate": 4.371393598139128e-06, + "loss": 0.8894, + "step": 29695 + }, + { + "epoch": 0.7, + "grad_norm": 1.1216660482843408, + "learning_rate": 4.370762921553625e-06, + "loss": 0.9803, + "step": 29696 + }, + { + "epoch": 0.7, + "grad_norm": 2.0169730063743794, + "learning_rate": 4.370132277743319e-06, + "loss": 0.9485, + "step": 29697 + }, + { + "epoch": 0.7, + "grad_norm": 1.8727617428385837, + "learning_rate": 4.369501666711885e-06, + "loss": 0.9471, + "step": 29698 + }, + { + "epoch": 0.7, + "grad_norm": 2.0195731185432, + "learning_rate": 4.368871088462998e-06, + "loss": 0.9435, + "step": 29699 + }, + { + "epoch": 0.7, + "grad_norm": 2.0497964471058006, + "learning_rate": 4.368240543000327e-06, + "loss": 1.0095, + "step": 29700 + }, + { + "epoch": 0.7, + "grad_norm": 2.2769654133115473, + "learning_rate": 4.36761003032754e-06, + "loss": 1.1198, + "step": 29701 + }, + { + "epoch": 0.7, + "grad_norm": 2.024546455124677, + "learning_rate": 4.366979550448313e-06, + "loss": 0.9241, + "step": 29702 + }, + { + "epoch": 0.7, + "grad_norm": 1.937497742472764, + "learning_rate": 4.366349103366311e-06, + "loss": 0.8831, + "step": 29703 + }, + { + "epoch": 0.7, + "grad_norm": 1.888590838558755, + "learning_rate": 4.365718689085211e-06, + "loss": 1.0252, + "step": 29704 + }, + { + "epoch": 0.7, + "grad_norm": 2.448397768130514, + "learning_rate": 4.365088307608678e-06, + "loss": 0.942, + "step": 29705 + }, + { + "epoch": 0.7, + "grad_norm": 1.989881221301789, + "learning_rate": 4.364457958940388e-06, + "loss": 0.961, + "step": 29706 + }, + { + "epoch": 0.7, + "grad_norm": 1.0925517594569647, + "learning_rate": 4.3638276430840045e-06, + "loss": 0.9342, + "step": 29707 + }, + { + "epoch": 0.7, + "grad_norm": 1.9506785835516878, + "learning_rate": 4.363197360043204e-06, + "loss": 0.8679, + "step": 29708 + }, + { + "epoch": 0.7, + "grad_norm": 1.1387387595686116, + "learning_rate": 4.362567109821654e-06, + "loss": 0.9418, + "step": 29709 + }, + { + "epoch": 0.7, + "grad_norm": 1.9211294883580712, + "learning_rate": 4.3619368924230185e-06, + "loss": 1.0494, + "step": 29710 + }, + { + "epoch": 0.7, + "grad_norm": 1.9851851984090745, + "learning_rate": 4.361306707850972e-06, + "loss": 0.9542, + "step": 29711 + }, + { + "epoch": 0.7, + "grad_norm": 1.9987120505800338, + "learning_rate": 4.360676556109186e-06, + "loss": 1.0374, + "step": 29712 + }, + { + "epoch": 0.7, + "grad_norm": 2.1254185365643523, + "learning_rate": 4.360046437201327e-06, + "loss": 0.9476, + "step": 29713 + }, + { + "epoch": 0.7, + "grad_norm": 2.2099493186989134, + "learning_rate": 4.3594163511310596e-06, + "loss": 1.0645, + "step": 29714 + }, + { + "epoch": 0.7, + "grad_norm": 2.0358643237001877, + "learning_rate": 4.358786297902056e-06, + "loss": 0.9554, + "step": 29715 + }, + { + "epoch": 0.7, + "grad_norm": 2.3753598228939947, + "learning_rate": 4.3581562775179875e-06, + "loss": 0.8824, + "step": 29716 + }, + { + "epoch": 0.7, + "grad_norm": 1.1769977825481, + "learning_rate": 4.35752628998252e-06, + "loss": 0.9216, + "step": 29717 + }, + { + "epoch": 0.7, + "grad_norm": 1.9721406416240521, + "learning_rate": 4.356896335299317e-06, + "loss": 1.0748, + "step": 29718 + }, + { + "epoch": 0.7, + "grad_norm": 2.1200464674274175, + "learning_rate": 4.35626641347205e-06, + "loss": 0.9109, + "step": 29719 + }, + { + "epoch": 0.7, + "grad_norm": 1.9204665286560532, + "learning_rate": 4.3556365245043905e-06, + "loss": 1.0575, + "step": 29720 + }, + { + "epoch": 0.7, + "grad_norm": 3.267654344699538, + "learning_rate": 4.355006668400003e-06, + "loss": 0.936, + "step": 29721 + }, + { + "epoch": 0.7, + "grad_norm": 2.1078435970381473, + "learning_rate": 4.354376845162549e-06, + "loss": 1.0358, + "step": 29722 + }, + { + "epoch": 0.7, + "grad_norm": 2.157971537903413, + "learning_rate": 4.3537470547957054e-06, + "loss": 1.0625, + "step": 29723 + }, + { + "epoch": 0.7, + "grad_norm": 1.9181802758877615, + "learning_rate": 4.353117297303129e-06, + "loss": 0.887, + "step": 29724 + }, + { + "epoch": 0.7, + "grad_norm": 2.021268712278686, + "learning_rate": 4.352487572688495e-06, + "loss": 0.9734, + "step": 29725 + }, + { + "epoch": 0.7, + "grad_norm": 2.211674572966806, + "learning_rate": 4.351857880955464e-06, + "loss": 0.9876, + "step": 29726 + }, + { + "epoch": 0.7, + "grad_norm": 1.0610096583218989, + "learning_rate": 4.351228222107706e-06, + "loss": 0.9892, + "step": 29727 + }, + { + "epoch": 0.7, + "grad_norm": 1.9892996490781496, + "learning_rate": 4.350598596148884e-06, + "loss": 1.0446, + "step": 29728 + }, + { + "epoch": 0.7, + "grad_norm": 1.0580294564823496, + "learning_rate": 4.349969003082669e-06, + "loss": 0.954, + "step": 29729 + }, + { + "epoch": 0.7, + "grad_norm": 1.9448721118616463, + "learning_rate": 4.349339442912723e-06, + "loss": 0.927, + "step": 29730 + }, + { + "epoch": 0.7, + "grad_norm": 2.614362275334639, + "learning_rate": 4.3487099156427075e-06, + "loss": 0.9994, + "step": 29731 + }, + { + "epoch": 0.7, + "grad_norm": 1.9443065178497683, + "learning_rate": 4.348080421276293e-06, + "loss": 0.8983, + "step": 29732 + }, + { + "epoch": 0.7, + "grad_norm": 2.1780390985976688, + "learning_rate": 4.347450959817146e-06, + "loss": 0.9875, + "step": 29733 + }, + { + "epoch": 0.7, + "grad_norm": 1.7294070616256367, + "learning_rate": 4.34682153126893e-06, + "loss": 0.987, + "step": 29734 + }, + { + "epoch": 0.7, + "grad_norm": 2.220864029274688, + "learning_rate": 4.346192135635305e-06, + "loss": 0.9196, + "step": 29735 + }, + { + "epoch": 0.7, + "grad_norm": 2.3500122700358848, + "learning_rate": 4.345562772919939e-06, + "loss": 1.0387, + "step": 29736 + }, + { + "epoch": 0.7, + "grad_norm": 1.965704801937364, + "learning_rate": 4.344933443126501e-06, + "loss": 1.05, + "step": 29737 + }, + { + "epoch": 0.7, + "grad_norm": 1.797673522089336, + "learning_rate": 4.344304146258651e-06, + "loss": 0.9119, + "step": 29738 + }, + { + "epoch": 0.7, + "grad_norm": 2.2920146009209095, + "learning_rate": 4.343674882320048e-06, + "loss": 0.893, + "step": 29739 + }, + { + "epoch": 0.7, + "grad_norm": 1.126442637946026, + "learning_rate": 4.343045651314361e-06, + "loss": 0.9035, + "step": 29740 + }, + { + "epoch": 0.7, + "grad_norm": 2.2617378670538417, + "learning_rate": 4.342416453245259e-06, + "loss": 1.0054, + "step": 29741 + }, + { + "epoch": 0.7, + "grad_norm": 2.3471058361550368, + "learning_rate": 4.341787288116397e-06, + "loss": 0.9665, + "step": 29742 + }, + { + "epoch": 0.7, + "grad_norm": 1.7951522187463211, + "learning_rate": 4.341158155931437e-06, + "loss": 1.0078, + "step": 29743 + }, + { + "epoch": 0.7, + "grad_norm": 2.0802663823127108, + "learning_rate": 4.3405290566940475e-06, + "loss": 1.0018, + "step": 29744 + }, + { + "epoch": 0.7, + "grad_norm": 2.3024321265518424, + "learning_rate": 4.339899990407892e-06, + "loss": 1.005, + "step": 29745 + }, + { + "epoch": 0.7, + "grad_norm": 2.2887729513114543, + "learning_rate": 4.339270957076631e-06, + "loss": 1.0828, + "step": 29746 + }, + { + "epoch": 0.7, + "grad_norm": 1.9859698778044728, + "learning_rate": 4.338641956703923e-06, + "loss": 0.9185, + "step": 29747 + }, + { + "epoch": 0.7, + "grad_norm": 1.8302153363005778, + "learning_rate": 4.338012989293438e-06, + "loss": 1.0246, + "step": 29748 + }, + { + "epoch": 0.7, + "grad_norm": 2.0407635306655023, + "learning_rate": 4.33738405484883e-06, + "loss": 0.9604, + "step": 29749 + }, + { + "epoch": 0.7, + "grad_norm": 2.8403316976197184, + "learning_rate": 4.3367551533737685e-06, + "loss": 0.9475, + "step": 29750 + }, + { + "epoch": 0.7, + "grad_norm": 1.8806829809252874, + "learning_rate": 4.336126284871908e-06, + "loss": 0.9996, + "step": 29751 + }, + { + "epoch": 0.7, + "grad_norm": 2.5162653024854715, + "learning_rate": 4.335497449346916e-06, + "loss": 0.9294, + "step": 29752 + }, + { + "epoch": 0.7, + "grad_norm": 2.000289243568848, + "learning_rate": 4.334868646802448e-06, + "loss": 1.0082, + "step": 29753 + }, + { + "epoch": 0.7, + "grad_norm": 2.2462834055275183, + "learning_rate": 4.334239877242172e-06, + "loss": 1.0252, + "step": 29754 + }, + { + "epoch": 0.7, + "grad_norm": 2.531610828068181, + "learning_rate": 4.333611140669746e-06, + "loss": 1.0582, + "step": 29755 + }, + { + "epoch": 0.7, + "grad_norm": 1.8096555046711231, + "learning_rate": 4.332982437088825e-06, + "loss": 0.9679, + "step": 29756 + }, + { + "epoch": 0.7, + "grad_norm": 1.8606535205935248, + "learning_rate": 4.332353766503075e-06, + "loss": 1.0306, + "step": 29757 + }, + { + "epoch": 0.7, + "grad_norm": 1.939918544215946, + "learning_rate": 4.331725128916157e-06, + "loss": 0.8885, + "step": 29758 + }, + { + "epoch": 0.7, + "grad_norm": 2.0340606061241204, + "learning_rate": 4.331096524331735e-06, + "loss": 0.9094, + "step": 29759 + }, + { + "epoch": 0.7, + "grad_norm": 2.1079538738775248, + "learning_rate": 4.330467952753458e-06, + "loss": 1.0269, + "step": 29760 + }, + { + "epoch": 0.7, + "grad_norm": 1.9234479480121869, + "learning_rate": 4.329839414184991e-06, + "loss": 0.9056, + "step": 29761 + }, + { + "epoch": 0.7, + "grad_norm": 1.929924882253242, + "learning_rate": 4.329210908629997e-06, + "loss": 0.9394, + "step": 29762 + }, + { + "epoch": 0.7, + "grad_norm": 2.0669350288369945, + "learning_rate": 4.328582436092133e-06, + "loss": 1.0409, + "step": 29763 + }, + { + "epoch": 0.7, + "grad_norm": 1.940760695125413, + "learning_rate": 4.3279539965750535e-06, + "loss": 0.906, + "step": 29764 + }, + { + "epoch": 0.7, + "grad_norm": 1.1389037905381403, + "learning_rate": 4.327325590082423e-06, + "loss": 0.9151, + "step": 29765 + }, + { + "epoch": 0.7, + "grad_norm": 1.9254497056250406, + "learning_rate": 4.326697216617901e-06, + "loss": 0.9171, + "step": 29766 + }, + { + "epoch": 0.7, + "grad_norm": 1.894237256650888, + "learning_rate": 4.326068876185141e-06, + "loss": 0.9164, + "step": 29767 + }, + { + "epoch": 0.7, + "grad_norm": 2.101718002432749, + "learning_rate": 4.3254405687878075e-06, + "loss": 0.9758, + "step": 29768 + }, + { + "epoch": 0.7, + "grad_norm": 2.7967583219735874, + "learning_rate": 4.324812294429557e-06, + "loss": 0.9743, + "step": 29769 + }, + { + "epoch": 0.7, + "grad_norm": 2.164229956105861, + "learning_rate": 4.324184053114041e-06, + "loss": 1.0788, + "step": 29770 + }, + { + "epoch": 0.7, + "grad_norm": 2.09994780136192, + "learning_rate": 4.323555844844922e-06, + "loss": 1.0803, + "step": 29771 + }, + { + "epoch": 0.7, + "grad_norm": 2.2050323703316765, + "learning_rate": 4.3229276696258625e-06, + "loss": 0.8454, + "step": 29772 + }, + { + "epoch": 0.7, + "grad_norm": 2.192560889821807, + "learning_rate": 4.3222995274605155e-06, + "loss": 0.8919, + "step": 29773 + }, + { + "epoch": 0.7, + "grad_norm": 2.1273988192421496, + "learning_rate": 4.321671418352534e-06, + "loss": 1.0405, + "step": 29774 + }, + { + "epoch": 0.7, + "grad_norm": 2.2787133177119268, + "learning_rate": 4.3210433423055795e-06, + "loss": 0.8524, + "step": 29775 + }, + { + "epoch": 0.7, + "grad_norm": 1.7550743977065406, + "learning_rate": 4.320415299323316e-06, + "loss": 1.0278, + "step": 29776 + }, + { + "epoch": 0.7, + "grad_norm": 2.555974692052747, + "learning_rate": 4.319787289409385e-06, + "loss": 0.9453, + "step": 29777 + }, + { + "epoch": 0.7, + "grad_norm": 1.9147812554531873, + "learning_rate": 4.319159312567451e-06, + "loss": 1.0147, + "step": 29778 + }, + { + "epoch": 0.7, + "grad_norm": 2.2054574680681673, + "learning_rate": 4.318531368801171e-06, + "loss": 0.9024, + "step": 29779 + }, + { + "epoch": 0.7, + "grad_norm": 4.029435350418845, + "learning_rate": 4.317903458114202e-06, + "loss": 0.9637, + "step": 29780 + }, + { + "epoch": 0.7, + "grad_norm": 2.0789190249631133, + "learning_rate": 4.317275580510199e-06, + "loss": 1.1004, + "step": 29781 + }, + { + "epoch": 0.7, + "grad_norm": 1.972590206593954, + "learning_rate": 4.316647735992814e-06, + "loss": 1.1246, + "step": 29782 + }, + { + "epoch": 0.7, + "grad_norm": 1.884250252618651, + "learning_rate": 4.316019924565708e-06, + "loss": 0.9557, + "step": 29783 + }, + { + "epoch": 0.7, + "grad_norm": 2.013391256319961, + "learning_rate": 4.315392146232529e-06, + "loss": 0.957, + "step": 29784 + }, + { + "epoch": 0.7, + "grad_norm": 2.073522102337504, + "learning_rate": 4.31476440099694e-06, + "loss": 1.0219, + "step": 29785 + }, + { + "epoch": 0.7, + "grad_norm": 2.1014076757745292, + "learning_rate": 4.31413668886259e-06, + "loss": 0.9454, + "step": 29786 + }, + { + "epoch": 0.7, + "grad_norm": 2.10475022875894, + "learning_rate": 4.31350900983314e-06, + "loss": 0.9916, + "step": 29787 + }, + { + "epoch": 0.7, + "grad_norm": 2.0300684887835527, + "learning_rate": 4.312881363912237e-06, + "loss": 1.0844, + "step": 29788 + }, + { + "epoch": 0.7, + "grad_norm": 2.148040884103761, + "learning_rate": 4.312253751103544e-06, + "loss": 1.0683, + "step": 29789 + }, + { + "epoch": 0.7, + "grad_norm": 2.4154860301366345, + "learning_rate": 4.311626171410707e-06, + "loss": 0.937, + "step": 29790 + }, + { + "epoch": 0.7, + "grad_norm": 1.9530620561232348, + "learning_rate": 4.310998624837383e-06, + "loss": 1.1018, + "step": 29791 + }, + { + "epoch": 0.7, + "grad_norm": 1.939655620879317, + "learning_rate": 4.310371111387225e-06, + "loss": 0.9747, + "step": 29792 + }, + { + "epoch": 0.7, + "grad_norm": 2.597492021912068, + "learning_rate": 4.309743631063892e-06, + "loss": 0.9926, + "step": 29793 + }, + { + "epoch": 0.7, + "grad_norm": 1.8530151633790857, + "learning_rate": 4.309116183871032e-06, + "loss": 0.9295, + "step": 29794 + }, + { + "epoch": 0.7, + "grad_norm": 1.9056885244288693, + "learning_rate": 4.308488769812296e-06, + "loss": 0.9238, + "step": 29795 + }, + { + "epoch": 0.7, + "grad_norm": 2.055395554043428, + "learning_rate": 4.30786138889134e-06, + "loss": 0.8601, + "step": 29796 + }, + { + "epoch": 0.7, + "grad_norm": 2.586820675164034, + "learning_rate": 4.30723404111182e-06, + "loss": 1.0362, + "step": 29797 + }, + { + "epoch": 0.7, + "grad_norm": 1.9941311578217646, + "learning_rate": 4.306606726477388e-06, + "loss": 1.0124, + "step": 29798 + }, + { + "epoch": 0.7, + "grad_norm": 3.4287753575134112, + "learning_rate": 4.3059794449916896e-06, + "loss": 1.0228, + "step": 29799 + }, + { + "epoch": 0.7, + "grad_norm": 1.9602826579776178, + "learning_rate": 4.305352196658381e-06, + "loss": 1.0237, + "step": 29800 + }, + { + "epoch": 0.7, + "grad_norm": 1.8403761856007812, + "learning_rate": 4.3047249814811185e-06, + "loss": 0.9824, + "step": 29801 + }, + { + "epoch": 0.7, + "grad_norm": 2.034206071841398, + "learning_rate": 4.30409779946355e-06, + "loss": 1.036, + "step": 29802 + }, + { + "epoch": 0.7, + "grad_norm": 1.9793579998698236, + "learning_rate": 4.303470650609325e-06, + "loss": 1.0093, + "step": 29803 + }, + { + "epoch": 0.7, + "grad_norm": 2.9248563944205594, + "learning_rate": 4.3028435349221e-06, + "loss": 0.8195, + "step": 29804 + }, + { + "epoch": 0.7, + "grad_norm": 1.1118503452502082, + "learning_rate": 4.302216452405521e-06, + "loss": 0.9713, + "step": 29805 + }, + { + "epoch": 0.7, + "grad_norm": 2.0470977966911033, + "learning_rate": 4.301589403063244e-06, + "loss": 0.8937, + "step": 29806 + }, + { + "epoch": 0.7, + "grad_norm": 1.1511688549963575, + "learning_rate": 4.300962386898915e-06, + "loss": 0.8989, + "step": 29807 + }, + { + "epoch": 0.7, + "grad_norm": 1.7482545827157265, + "learning_rate": 4.3003354039161906e-06, + "loss": 0.9965, + "step": 29808 + }, + { + "epoch": 0.7, + "grad_norm": 1.7699751392585574, + "learning_rate": 4.299708454118714e-06, + "loss": 0.8817, + "step": 29809 + }, + { + "epoch": 0.7, + "grad_norm": 1.8352743635335023, + "learning_rate": 4.299081537510143e-06, + "loss": 0.9584, + "step": 29810 + }, + { + "epoch": 0.7, + "grad_norm": 2.0389148576036447, + "learning_rate": 4.2984546540941245e-06, + "loss": 1.0911, + "step": 29811 + }, + { + "epoch": 0.7, + "grad_norm": 1.890728795006784, + "learning_rate": 4.297827803874304e-06, + "loss": 0.9225, + "step": 29812 + }, + { + "epoch": 0.7, + "grad_norm": 1.991759318847549, + "learning_rate": 4.297200986854335e-06, + "loss": 1.0171, + "step": 29813 + }, + { + "epoch": 0.7, + "grad_norm": 1.1011984779337, + "learning_rate": 4.296574203037872e-06, + "loss": 0.9877, + "step": 29814 + }, + { + "epoch": 0.7, + "grad_norm": 2.1148329218010167, + "learning_rate": 4.295947452428559e-06, + "loss": 1.0441, + "step": 29815 + }, + { + "epoch": 0.7, + "grad_norm": 1.917878114565192, + "learning_rate": 4.295320735030043e-06, + "loss": 1.0453, + "step": 29816 + }, + { + "epoch": 0.7, + "grad_norm": 1.9569185777978317, + "learning_rate": 4.294694050845976e-06, + "loss": 1.0823, + "step": 29817 + }, + { + "epoch": 0.7, + "grad_norm": 1.9304166091595265, + "learning_rate": 4.294067399880009e-06, + "loss": 0.9794, + "step": 29818 + }, + { + "epoch": 0.7, + "grad_norm": 2.144974852673453, + "learning_rate": 4.293440782135788e-06, + "loss": 0.9601, + "step": 29819 + }, + { + "epoch": 0.7, + "grad_norm": 2.2666631101755663, + "learning_rate": 4.292814197616958e-06, + "loss": 1.0096, + "step": 29820 + }, + { + "epoch": 0.7, + "grad_norm": 1.13769110292057, + "learning_rate": 4.29218764632717e-06, + "loss": 0.9644, + "step": 29821 + }, + { + "epoch": 0.7, + "grad_norm": 2.481320907333446, + "learning_rate": 4.2915611282700775e-06, + "loss": 1.0179, + "step": 29822 + }, + { + "epoch": 0.7, + "grad_norm": 1.748567909303075, + "learning_rate": 4.290934643449324e-06, + "loss": 0.9314, + "step": 29823 + }, + { + "epoch": 0.7, + "grad_norm": 2.6220910865344353, + "learning_rate": 4.290308191868551e-06, + "loss": 0.9269, + "step": 29824 + }, + { + "epoch": 0.7, + "grad_norm": 2.211232813078434, + "learning_rate": 4.289681773531415e-06, + "loss": 1.0042, + "step": 29825 + }, + { + "epoch": 0.7, + "grad_norm": 2.183860230691422, + "learning_rate": 4.289055388441558e-06, + "loss": 0.932, + "step": 29826 + }, + { + "epoch": 0.7, + "grad_norm": 2.060656886870244, + "learning_rate": 4.28842903660263e-06, + "loss": 1.0033, + "step": 29827 + }, + { + "epoch": 0.7, + "grad_norm": 1.8230883637387394, + "learning_rate": 4.2878027180182745e-06, + "loss": 0.9938, + "step": 29828 + }, + { + "epoch": 0.7, + "grad_norm": 1.841650189144001, + "learning_rate": 4.287176432692143e-06, + "loss": 1.1108, + "step": 29829 + }, + { + "epoch": 0.7, + "grad_norm": 1.1522815370333932, + "learning_rate": 4.286550180627874e-06, + "loss": 0.9444, + "step": 29830 + }, + { + "epoch": 0.7, + "grad_norm": 2.1018157415539704, + "learning_rate": 4.285923961829124e-06, + "loss": 1.0414, + "step": 29831 + }, + { + "epoch": 0.7, + "grad_norm": 1.8862947988242202, + "learning_rate": 4.285297776299533e-06, + "loss": 0.9652, + "step": 29832 + }, + { + "epoch": 0.7, + "grad_norm": 1.0800174805833682, + "learning_rate": 4.284671624042744e-06, + "loss": 0.9388, + "step": 29833 + }, + { + "epoch": 0.7, + "grad_norm": 1.9788071468453219, + "learning_rate": 4.284045505062406e-06, + "loss": 1.043, + "step": 29834 + }, + { + "epoch": 0.7, + "grad_norm": 1.8271274661018195, + "learning_rate": 4.283419419362169e-06, + "loss": 0.9953, + "step": 29835 + }, + { + "epoch": 0.7, + "grad_norm": 2.03513692347398, + "learning_rate": 4.282793366945675e-06, + "loss": 1.0229, + "step": 29836 + }, + { + "epoch": 0.7, + "grad_norm": 2.588544016981481, + "learning_rate": 4.282167347816562e-06, + "loss": 0.9213, + "step": 29837 + }, + { + "epoch": 0.7, + "grad_norm": 1.9911295456502422, + "learning_rate": 4.281541361978483e-06, + "loss": 0.9082, + "step": 29838 + }, + { + "epoch": 0.7, + "grad_norm": 1.206055818588097, + "learning_rate": 4.280915409435084e-06, + "loss": 0.9824, + "step": 29839 + }, + { + "epoch": 0.7, + "grad_norm": 1.8829671874845848, + "learning_rate": 4.280289490190005e-06, + "loss": 0.9921, + "step": 29840 + }, + { + "epoch": 0.7, + "grad_norm": 1.8302662233303706, + "learning_rate": 4.279663604246888e-06, + "loss": 1.006, + "step": 29841 + }, + { + "epoch": 0.7, + "grad_norm": 1.1906817862045476, + "learning_rate": 4.27903775160938e-06, + "loss": 1.0355, + "step": 29842 + }, + { + "epoch": 0.7, + "grad_norm": 2.0598120025298265, + "learning_rate": 4.27841193228113e-06, + "loss": 0.9873, + "step": 29843 + }, + { + "epoch": 0.7, + "grad_norm": 2.223690852478706, + "learning_rate": 4.277786146265778e-06, + "loss": 0.8648, + "step": 29844 + }, + { + "epoch": 0.7, + "grad_norm": 2.2491257646747447, + "learning_rate": 4.277160393566962e-06, + "loss": 1.1795, + "step": 29845 + }, + { + "epoch": 0.7, + "grad_norm": 1.0915823951416441, + "learning_rate": 4.2765346741883325e-06, + "loss": 0.9149, + "step": 29846 + }, + { + "epoch": 0.7, + "grad_norm": 1.875030176471231, + "learning_rate": 4.275908988133528e-06, + "loss": 0.9788, + "step": 29847 + }, + { + "epoch": 0.7, + "grad_norm": 1.824657988648434, + "learning_rate": 4.275283335406197e-06, + "loss": 0.8925, + "step": 29848 + }, + { + "epoch": 0.7, + "grad_norm": 2.0714812315500324, + "learning_rate": 4.274657716009976e-06, + "loss": 1.0785, + "step": 29849 + }, + { + "epoch": 0.7, + "grad_norm": 2.340257923384295, + "learning_rate": 4.274032129948512e-06, + "loss": 1.0044, + "step": 29850 + }, + { + "epoch": 0.7, + "grad_norm": 1.1375200421063016, + "learning_rate": 4.273406577225444e-06, + "loss": 0.9866, + "step": 29851 + }, + { + "epoch": 0.7, + "grad_norm": 1.9770959749801005, + "learning_rate": 4.272781057844418e-06, + "loss": 1.0129, + "step": 29852 + }, + { + "epoch": 0.7, + "grad_norm": 2.1092030616286026, + "learning_rate": 4.272155571809076e-06, + "loss": 1.1115, + "step": 29853 + }, + { + "epoch": 0.7, + "grad_norm": 1.8770182179735935, + "learning_rate": 4.271530119123052e-06, + "loss": 1.0189, + "step": 29854 + }, + { + "epoch": 0.7, + "grad_norm": 1.9954995114838112, + "learning_rate": 4.270904699789994e-06, + "loss": 0.8496, + "step": 29855 + }, + { + "epoch": 0.7, + "grad_norm": 1.90855860473817, + "learning_rate": 4.270279313813542e-06, + "loss": 1.1075, + "step": 29856 + }, + { + "epoch": 0.7, + "grad_norm": 2.172654461011181, + "learning_rate": 4.269653961197346e-06, + "loss": 0.9359, + "step": 29857 + }, + { + "epoch": 0.7, + "grad_norm": 2.491067816631866, + "learning_rate": 4.269028641945032e-06, + "loss": 0.9274, + "step": 29858 + }, + { + "epoch": 0.7, + "grad_norm": 2.664139210248204, + "learning_rate": 4.2684033560602465e-06, + "loss": 0.9135, + "step": 29859 + }, + { + "epoch": 0.7, + "grad_norm": 2.3329015091237544, + "learning_rate": 4.267778103546632e-06, + "loss": 1.0337, + "step": 29860 + }, + { + "epoch": 0.7, + "grad_norm": 1.8999985588047283, + "learning_rate": 4.267152884407835e-06, + "loss": 1.0088, + "step": 29861 + }, + { + "epoch": 0.7, + "grad_norm": 1.8414689808248559, + "learning_rate": 4.266527698647482e-06, + "loss": 1.0485, + "step": 29862 + }, + { + "epoch": 0.7, + "grad_norm": 1.870235389081515, + "learning_rate": 4.265902546269219e-06, + "loss": 0.969, + "step": 29863 + }, + { + "epoch": 0.7, + "grad_norm": 1.932475164622618, + "learning_rate": 4.265277427276692e-06, + "loss": 0.9868, + "step": 29864 + }, + { + "epoch": 0.7, + "grad_norm": 1.866307587344124, + "learning_rate": 4.264652341673531e-06, + "loss": 1.0542, + "step": 29865 + }, + { + "epoch": 0.7, + "grad_norm": 2.12357499321424, + "learning_rate": 4.2640272894633825e-06, + "loss": 1.052, + "step": 29866 + }, + { + "epoch": 0.7, + "grad_norm": 2.2089702006432512, + "learning_rate": 4.2634022706498815e-06, + "loss": 0.892, + "step": 29867 + }, + { + "epoch": 0.7, + "grad_norm": 1.9451892553540777, + "learning_rate": 4.26277728523667e-06, + "loss": 0.969, + "step": 29868 + }, + { + "epoch": 0.7, + "grad_norm": 1.8046434536942306, + "learning_rate": 4.262152333227383e-06, + "loss": 0.9755, + "step": 29869 + }, + { + "epoch": 0.7, + "grad_norm": 1.9916738427035616, + "learning_rate": 4.2615274146256654e-06, + "loss": 1.125, + "step": 29870 + }, + { + "epoch": 0.7, + "grad_norm": 2.096248394679963, + "learning_rate": 4.2609025294351525e-06, + "loss": 0.989, + "step": 29871 + }, + { + "epoch": 0.7, + "grad_norm": 2.568643508177753, + "learning_rate": 4.260277677659477e-06, + "loss": 0.9257, + "step": 29872 + }, + { + "epoch": 0.7, + "grad_norm": 1.9007459875013673, + "learning_rate": 4.2596528593022835e-06, + "loss": 0.8152, + "step": 29873 + }, + { + "epoch": 0.7, + "grad_norm": 1.919321779518357, + "learning_rate": 4.259028074367211e-06, + "loss": 1.081, + "step": 29874 + }, + { + "epoch": 0.7, + "grad_norm": 1.1515136155272145, + "learning_rate": 4.258403322857895e-06, + "loss": 0.9895, + "step": 29875 + }, + { + "epoch": 0.7, + "grad_norm": 1.9177698815997823, + "learning_rate": 4.257778604777969e-06, + "loss": 0.9579, + "step": 29876 + }, + { + "epoch": 0.7, + "grad_norm": 2.153165681740835, + "learning_rate": 4.257153920131074e-06, + "loss": 1.0978, + "step": 29877 + }, + { + "epoch": 0.7, + "grad_norm": 1.8166918210920322, + "learning_rate": 4.256529268920851e-06, + "loss": 0.922, + "step": 29878 + }, + { + "epoch": 0.7, + "grad_norm": 2.0058989300398244, + "learning_rate": 4.255904651150931e-06, + "loss": 0.916, + "step": 29879 + }, + { + "epoch": 0.7, + "grad_norm": 2.1144810853815263, + "learning_rate": 4.255280066824952e-06, + "loss": 0.8492, + "step": 29880 + }, + { + "epoch": 0.7, + "grad_norm": 1.944304365881774, + "learning_rate": 4.254655515946549e-06, + "loss": 0.9489, + "step": 29881 + }, + { + "epoch": 0.7, + "grad_norm": 1.97659693439884, + "learning_rate": 4.254030998519365e-06, + "loss": 0.9481, + "step": 29882 + }, + { + "epoch": 0.7, + "grad_norm": 2.3426189322465203, + "learning_rate": 4.253406514547031e-06, + "loss": 0.9942, + "step": 29883 + }, + { + "epoch": 0.7, + "grad_norm": 1.912269507892901, + "learning_rate": 4.25278206403318e-06, + "loss": 0.8717, + "step": 29884 + }, + { + "epoch": 0.7, + "grad_norm": 2.017279969652992, + "learning_rate": 4.252157646981457e-06, + "loss": 0.9811, + "step": 29885 + }, + { + "epoch": 0.7, + "grad_norm": 1.8453224114309115, + "learning_rate": 4.251533263395486e-06, + "loss": 0.8922, + "step": 29886 + }, + { + "epoch": 0.7, + "grad_norm": 2.2669925195358993, + "learning_rate": 4.2509089132789126e-06, + "loss": 0.9539, + "step": 29887 + }, + { + "epoch": 0.7, + "grad_norm": 2.121984484793762, + "learning_rate": 4.250284596635365e-06, + "loss": 0.9354, + "step": 29888 + }, + { + "epoch": 0.7, + "grad_norm": 2.1243442600037348, + "learning_rate": 4.249660313468483e-06, + "loss": 0.9802, + "step": 29889 + }, + { + "epoch": 0.7, + "grad_norm": 1.957441154494276, + "learning_rate": 4.2490360637818965e-06, + "loss": 0.9201, + "step": 29890 + }, + { + "epoch": 0.7, + "grad_norm": 2.073952062404546, + "learning_rate": 4.248411847579245e-06, + "loss": 1.0307, + "step": 29891 + }, + { + "epoch": 0.7, + "grad_norm": 1.9706265954034017, + "learning_rate": 4.247787664864163e-06, + "loss": 1.0553, + "step": 29892 + }, + { + "epoch": 0.7, + "grad_norm": 1.9175740780261206, + "learning_rate": 4.247163515640277e-06, + "loss": 0.9145, + "step": 29893 + }, + { + "epoch": 0.7, + "grad_norm": 2.0842090253666004, + "learning_rate": 4.2465393999112256e-06, + "loss": 1.0427, + "step": 29894 + }, + { + "epoch": 0.7, + "grad_norm": 1.7367881523126016, + "learning_rate": 4.245915317680649e-06, + "loss": 0.9042, + "step": 29895 + }, + { + "epoch": 0.7, + "grad_norm": 2.1211891984890765, + "learning_rate": 4.245291268952174e-06, + "loss": 0.874, + "step": 29896 + }, + { + "epoch": 0.7, + "grad_norm": 2.053477978168524, + "learning_rate": 4.244667253729431e-06, + "loss": 1.0598, + "step": 29897 + }, + { + "epoch": 0.7, + "grad_norm": 1.9633999507718274, + "learning_rate": 4.244043272016059e-06, + "loss": 0.9649, + "step": 29898 + }, + { + "epoch": 0.7, + "grad_norm": 1.9103490773826295, + "learning_rate": 4.243419323815692e-06, + "loss": 0.8809, + "step": 29899 + }, + { + "epoch": 0.7, + "grad_norm": 1.9107966381985841, + "learning_rate": 4.24279540913196e-06, + "loss": 0.9661, + "step": 29900 + }, + { + "epoch": 0.7, + "grad_norm": 2.010053085116105, + "learning_rate": 4.242171527968492e-06, + "loss": 1.0164, + "step": 29901 + }, + { + "epoch": 0.7, + "grad_norm": 2.439725682339237, + "learning_rate": 4.241547680328924e-06, + "loss": 0.9519, + "step": 29902 + }, + { + "epoch": 0.7, + "grad_norm": 3.1326327328076533, + "learning_rate": 4.240923866216894e-06, + "loss": 0.9867, + "step": 29903 + }, + { + "epoch": 0.7, + "grad_norm": 1.939244626709532, + "learning_rate": 4.240300085636027e-06, + "loss": 0.9521, + "step": 29904 + }, + { + "epoch": 0.7, + "grad_norm": 3.4742158296472874, + "learning_rate": 4.239676338589953e-06, + "loss": 0.9362, + "step": 29905 + }, + { + "epoch": 0.7, + "grad_norm": 2.3763872477433363, + "learning_rate": 4.239052625082311e-06, + "loss": 0.943, + "step": 29906 + }, + { + "epoch": 0.7, + "grad_norm": 1.880193932289668, + "learning_rate": 4.238428945116725e-06, + "loss": 0.9515, + "step": 29907 + }, + { + "epoch": 0.7, + "grad_norm": 2.1982293047033896, + "learning_rate": 4.237805298696832e-06, + "loss": 0.971, + "step": 29908 + }, + { + "epoch": 0.7, + "grad_norm": 1.0926222187356815, + "learning_rate": 4.23718168582626e-06, + "loss": 0.8941, + "step": 29909 + }, + { + "epoch": 0.7, + "grad_norm": 1.8646221281376023, + "learning_rate": 4.236558106508641e-06, + "loss": 0.8431, + "step": 29910 + }, + { + "epoch": 0.7, + "grad_norm": 2.135795621755133, + "learning_rate": 4.235934560747604e-06, + "loss": 1.0056, + "step": 29911 + }, + { + "epoch": 0.7, + "grad_norm": 2.0286684107177027, + "learning_rate": 4.2353110485467845e-06, + "loss": 1.0215, + "step": 29912 + }, + { + "epoch": 0.7, + "grad_norm": 1.9817264146214078, + "learning_rate": 4.234687569909808e-06, + "loss": 0.7415, + "step": 29913 + }, + { + "epoch": 0.7, + "grad_norm": 3.195206925824128, + "learning_rate": 4.234064124840302e-06, + "loss": 1.071, + "step": 29914 + }, + { + "epoch": 0.7, + "grad_norm": 2.3238275028352327, + "learning_rate": 4.233440713341901e-06, + "loss": 1.072, + "step": 29915 + }, + { + "epoch": 0.7, + "grad_norm": 2.028675730746734, + "learning_rate": 4.232817335418236e-06, + "loss": 0.9068, + "step": 29916 + }, + { + "epoch": 0.7, + "grad_norm": 1.855402967516638, + "learning_rate": 4.232193991072937e-06, + "loss": 0.978, + "step": 29917 + }, + { + "epoch": 0.7, + "grad_norm": 2.165785609480793, + "learning_rate": 4.231570680309626e-06, + "loss": 0.9647, + "step": 29918 + }, + { + "epoch": 0.7, + "grad_norm": 1.9978830871916056, + "learning_rate": 4.230947403131935e-06, + "loss": 0.9354, + "step": 29919 + }, + { + "epoch": 0.7, + "grad_norm": 1.1195511570506245, + "learning_rate": 4.2303241595435e-06, + "loss": 0.962, + "step": 29920 + }, + { + "epoch": 0.7, + "grad_norm": 1.8578878273543245, + "learning_rate": 4.229700949547944e-06, + "loss": 0.9648, + "step": 29921 + }, + { + "epoch": 0.7, + "grad_norm": 2.1132047182334714, + "learning_rate": 4.229077773148892e-06, + "loss": 0.8902, + "step": 29922 + }, + { + "epoch": 0.7, + "grad_norm": 1.935789955144643, + "learning_rate": 4.228454630349977e-06, + "loss": 0.9007, + "step": 29923 + }, + { + "epoch": 0.7, + "grad_norm": 1.9060856812333145, + "learning_rate": 4.227831521154829e-06, + "loss": 1.063, + "step": 29924 + }, + { + "epoch": 0.71, + "grad_norm": 2.0473547271055805, + "learning_rate": 4.227208445567075e-06, + "loss": 0.9546, + "step": 29925 + }, + { + "epoch": 0.71, + "grad_norm": 2.0985510034889003, + "learning_rate": 4.226585403590335e-06, + "loss": 0.9217, + "step": 29926 + }, + { + "epoch": 0.71, + "grad_norm": 1.084890048147267, + "learning_rate": 4.225962395228248e-06, + "loss": 0.9168, + "step": 29927 + }, + { + "epoch": 0.71, + "grad_norm": 1.9773129877917675, + "learning_rate": 4.225339420484432e-06, + "loss": 1.0471, + "step": 29928 + }, + { + "epoch": 0.71, + "grad_norm": 2.2424711978768723, + "learning_rate": 4.224716479362521e-06, + "loss": 0.9501, + "step": 29929 + }, + { + "epoch": 0.71, + "grad_norm": 2.2493041603933572, + "learning_rate": 4.2240935718661365e-06, + "loss": 1.0469, + "step": 29930 + }, + { + "epoch": 0.71, + "grad_norm": 2.096571121949702, + "learning_rate": 4.223470697998911e-06, + "loss": 0.8445, + "step": 29931 + }, + { + "epoch": 0.71, + "grad_norm": 2.2132443636033168, + "learning_rate": 4.222847857764464e-06, + "loss": 1.0026, + "step": 29932 + }, + { + "epoch": 0.71, + "grad_norm": 2.2012319044679125, + "learning_rate": 4.22222505116643e-06, + "loss": 1.052, + "step": 29933 + }, + { + "epoch": 0.71, + "grad_norm": 2.417824978552171, + "learning_rate": 4.221602278208431e-06, + "loss": 0.936, + "step": 29934 + }, + { + "epoch": 0.71, + "grad_norm": 2.1224829775695397, + "learning_rate": 4.220979538894089e-06, + "loss": 0.9147, + "step": 29935 + }, + { + "epoch": 0.71, + "grad_norm": 1.0744481809485074, + "learning_rate": 4.220356833227034e-06, + "loss": 0.9703, + "step": 29936 + }, + { + "epoch": 0.71, + "grad_norm": 2.1900882864126694, + "learning_rate": 4.2197341612108944e-06, + "loss": 0.9416, + "step": 29937 + }, + { + "epoch": 0.71, + "grad_norm": 1.9994308949065966, + "learning_rate": 4.219111522849293e-06, + "loss": 0.9681, + "step": 29938 + }, + { + "epoch": 0.71, + "grad_norm": 2.1112209960836683, + "learning_rate": 4.21848891814585e-06, + "loss": 0.8675, + "step": 29939 + }, + { + "epoch": 0.71, + "grad_norm": 1.8984295492633685, + "learning_rate": 4.217866347104196e-06, + "loss": 1.0136, + "step": 29940 + }, + { + "epoch": 0.71, + "grad_norm": 1.8976204784076547, + "learning_rate": 4.217243809727958e-06, + "loss": 1.0002, + "step": 29941 + }, + { + "epoch": 0.71, + "grad_norm": 1.8977824168001374, + "learning_rate": 4.216621306020757e-06, + "loss": 0.9965, + "step": 29942 + }, + { + "epoch": 0.71, + "grad_norm": 1.983481516769719, + "learning_rate": 4.215998835986215e-06, + "loss": 0.9901, + "step": 29943 + }, + { + "epoch": 0.71, + "grad_norm": 2.418264439903293, + "learning_rate": 4.215376399627959e-06, + "loss": 1.0758, + "step": 29944 + }, + { + "epoch": 0.71, + "grad_norm": 2.134806467696899, + "learning_rate": 4.214753996949616e-06, + "loss": 0.9423, + "step": 29945 + }, + { + "epoch": 0.71, + "grad_norm": 1.1406485560431034, + "learning_rate": 4.2141316279548065e-06, + "loss": 0.9478, + "step": 29946 + }, + { + "epoch": 0.71, + "grad_norm": 1.9796012806953673, + "learning_rate": 4.21350929264715e-06, + "loss": 1.01, + "step": 29947 + }, + { + "epoch": 0.71, + "grad_norm": 3.078288976935681, + "learning_rate": 4.21288699103028e-06, + "loss": 1.0667, + "step": 29948 + }, + { + "epoch": 0.71, + "grad_norm": 1.8046505030693667, + "learning_rate": 4.21226472310781e-06, + "loss": 0.9606, + "step": 29949 + }, + { + "epoch": 0.71, + "grad_norm": 2.363186532479086, + "learning_rate": 4.211642488883369e-06, + "loss": 0.8208, + "step": 29950 + }, + { + "epoch": 0.71, + "grad_norm": 2.3394755774732845, + "learning_rate": 4.211020288360576e-06, + "loss": 0.9603, + "step": 29951 + }, + { + "epoch": 0.71, + "grad_norm": 2.3577471045443823, + "learning_rate": 4.210398121543059e-06, + "loss": 0.8331, + "step": 29952 + }, + { + "epoch": 0.71, + "grad_norm": 2.2373946380765593, + "learning_rate": 4.2097759884344335e-06, + "loss": 1.0992, + "step": 29953 + }, + { + "epoch": 0.71, + "grad_norm": 2.210990790227853, + "learning_rate": 4.209153889038325e-06, + "loss": 0.9863, + "step": 29954 + }, + { + "epoch": 0.71, + "grad_norm": 2.1540499348531994, + "learning_rate": 4.208531823358364e-06, + "loss": 1.0782, + "step": 29955 + }, + { + "epoch": 0.71, + "grad_norm": 2.5513990849815507, + "learning_rate": 4.207909791398156e-06, + "loss": 0.8667, + "step": 29956 + }, + { + "epoch": 0.71, + "grad_norm": 1.1082779111955419, + "learning_rate": 4.207287793161332e-06, + "loss": 0.9857, + "step": 29957 + }, + { + "epoch": 0.71, + "grad_norm": 2.4266216773335505, + "learning_rate": 4.2066658286515115e-06, + "loss": 0.87, + "step": 29958 + }, + { + "epoch": 0.71, + "grad_norm": 2.3922557567898872, + "learning_rate": 4.206043897872324e-06, + "loss": 0.9377, + "step": 29959 + }, + { + "epoch": 0.71, + "grad_norm": 1.029836828817424, + "learning_rate": 4.2054220008273765e-06, + "loss": 0.9557, + "step": 29960 + }, + { + "epoch": 0.71, + "grad_norm": 2.029775601245927, + "learning_rate": 4.204800137520296e-06, + "loss": 1.0391, + "step": 29961 + }, + { + "epoch": 0.71, + "grad_norm": 1.8327609961985494, + "learning_rate": 4.204178307954708e-06, + "loss": 1.0427, + "step": 29962 + }, + { + "epoch": 0.71, + "grad_norm": 2.319097892491383, + "learning_rate": 4.203556512134225e-06, + "loss": 0.9804, + "step": 29963 + }, + { + "epoch": 0.71, + "grad_norm": 1.8805180628142528, + "learning_rate": 4.202934750062475e-06, + "loss": 0.9914, + "step": 29964 + }, + { + "epoch": 0.71, + "grad_norm": 1.1055940836702116, + "learning_rate": 4.202313021743071e-06, + "loss": 1.0218, + "step": 29965 + }, + { + "epoch": 0.71, + "grad_norm": 1.9559921145965038, + "learning_rate": 4.201691327179639e-06, + "loss": 0.9513, + "step": 29966 + }, + { + "epoch": 0.71, + "grad_norm": 1.9388707115765376, + "learning_rate": 4.201069666375792e-06, + "loss": 0.9084, + "step": 29967 + }, + { + "epoch": 0.71, + "grad_norm": 1.100789687193381, + "learning_rate": 4.2004480393351565e-06, + "loss": 0.9731, + "step": 29968 + }, + { + "epoch": 0.71, + "grad_norm": 2.2412849614545167, + "learning_rate": 4.19982644606135e-06, + "loss": 0.9993, + "step": 29969 + }, + { + "epoch": 0.71, + "grad_norm": 1.8211308916866928, + "learning_rate": 4.199204886557986e-06, + "loss": 0.9767, + "step": 29970 + }, + { + "epoch": 0.71, + "grad_norm": 2.0428638668380157, + "learning_rate": 4.198583360828688e-06, + "loss": 0.9965, + "step": 29971 + }, + { + "epoch": 0.71, + "grad_norm": 2.3514965365352736, + "learning_rate": 4.197961868877078e-06, + "loss": 0.9625, + "step": 29972 + }, + { + "epoch": 0.71, + "grad_norm": 1.0297279606559275, + "learning_rate": 4.19734041070677e-06, + "loss": 0.9384, + "step": 29973 + }, + { + "epoch": 0.71, + "grad_norm": 2.0025464041789345, + "learning_rate": 4.19671898632138e-06, + "loss": 1.0998, + "step": 29974 + }, + { + "epoch": 0.71, + "grad_norm": 2.2173860574641595, + "learning_rate": 4.1960975957245285e-06, + "loss": 1.0039, + "step": 29975 + }, + { + "epoch": 0.71, + "grad_norm": 2.049889224282451, + "learning_rate": 4.19547623891984e-06, + "loss": 1.0147, + "step": 29976 + }, + { + "epoch": 0.71, + "grad_norm": 2.6255663103189986, + "learning_rate": 4.194854915910925e-06, + "loss": 1.0534, + "step": 29977 + }, + { + "epoch": 0.71, + "grad_norm": 1.9335839356789868, + "learning_rate": 4.194233626701399e-06, + "loss": 0.9362, + "step": 29978 + }, + { + "epoch": 0.71, + "grad_norm": 2.2024235137028634, + "learning_rate": 4.193612371294883e-06, + "loss": 0.9718, + "step": 29979 + }, + { + "epoch": 0.71, + "grad_norm": 1.8625936879045897, + "learning_rate": 4.1929911496949985e-06, + "loss": 0.9621, + "step": 29980 + }, + { + "epoch": 0.71, + "grad_norm": 2.0327166788219824, + "learning_rate": 4.192369961905357e-06, + "loss": 1.1028, + "step": 29981 + }, + { + "epoch": 0.71, + "grad_norm": 2.0268258784391873, + "learning_rate": 4.1917488079295735e-06, + "loss": 1.1162, + "step": 29982 + }, + { + "epoch": 0.71, + "grad_norm": 2.7738425684344556, + "learning_rate": 4.191127687771266e-06, + "loss": 1.0308, + "step": 29983 + }, + { + "epoch": 0.71, + "grad_norm": 1.9266056637789837, + "learning_rate": 4.190506601434057e-06, + "loss": 0.9394, + "step": 29984 + }, + { + "epoch": 0.71, + "grad_norm": 1.9210176236984924, + "learning_rate": 4.1898855489215574e-06, + "loss": 0.8704, + "step": 29985 + }, + { + "epoch": 0.71, + "grad_norm": 1.9044927005461008, + "learning_rate": 4.18926453023738e-06, + "loss": 0.9892, + "step": 29986 + }, + { + "epoch": 0.71, + "grad_norm": 1.9749136080413803, + "learning_rate": 4.188643545385148e-06, + "loss": 1.0413, + "step": 29987 + }, + { + "epoch": 0.71, + "grad_norm": 1.926789505113521, + "learning_rate": 4.1880225943684695e-06, + "loss": 1.1521, + "step": 29988 + }, + { + "epoch": 0.71, + "grad_norm": 1.9787942413866098, + "learning_rate": 4.187401677190966e-06, + "loss": 0.9999, + "step": 29989 + }, + { + "epoch": 0.71, + "grad_norm": 1.858395135593939, + "learning_rate": 4.186780793856247e-06, + "loss": 0.9828, + "step": 29990 + }, + { + "epoch": 0.71, + "grad_norm": 2.23641677195583, + "learning_rate": 4.186159944367936e-06, + "loss": 0.8326, + "step": 29991 + }, + { + "epoch": 0.71, + "grad_norm": 2.271527400330913, + "learning_rate": 4.185539128729636e-06, + "loss": 0.9919, + "step": 29992 + }, + { + "epoch": 0.71, + "grad_norm": 1.8378218836602371, + "learning_rate": 4.184918346944974e-06, + "loss": 0.8472, + "step": 29993 + }, + { + "epoch": 0.71, + "grad_norm": 2.2064472042852956, + "learning_rate": 4.1842975990175575e-06, + "loss": 1.1721, + "step": 29994 + }, + { + "epoch": 0.71, + "grad_norm": 2.170815141909993, + "learning_rate": 4.183676884950998e-06, + "loss": 0.9125, + "step": 29995 + }, + { + "epoch": 0.71, + "grad_norm": 1.777436668657849, + "learning_rate": 4.183056204748913e-06, + "loss": 1.0325, + "step": 29996 + }, + { + "epoch": 0.71, + "grad_norm": 2.0311900477319216, + "learning_rate": 4.18243555841492e-06, + "loss": 1.0217, + "step": 29997 + }, + { + "epoch": 0.71, + "grad_norm": 2.0474011190035344, + "learning_rate": 4.181814945952629e-06, + "loss": 0.9466, + "step": 29998 + }, + { + "epoch": 0.71, + "grad_norm": 2.1523092760091913, + "learning_rate": 4.18119436736565e-06, + "loss": 0.9845, + "step": 29999 + }, + { + "epoch": 0.71, + "grad_norm": 1.984183356298319, + "learning_rate": 4.1805738226575995e-06, + "loss": 0.8632, + "step": 30000 + }, + { + "epoch": 0.71, + "grad_norm": 1.9053048901012366, + "learning_rate": 4.179953311832094e-06, + "loss": 0.9194, + "step": 30001 + }, + { + "epoch": 0.71, + "grad_norm": 2.450970095520978, + "learning_rate": 4.179332834892743e-06, + "loss": 1.0627, + "step": 30002 + }, + { + "epoch": 0.71, + "grad_norm": 1.121153691355809, + "learning_rate": 4.178712391843156e-06, + "loss": 0.9618, + "step": 30003 + }, + { + "epoch": 0.71, + "grad_norm": 1.8772357640788153, + "learning_rate": 4.178091982686948e-06, + "loss": 0.9121, + "step": 30004 + }, + { + "epoch": 0.71, + "grad_norm": 2.1804885025908263, + "learning_rate": 4.177471607427736e-06, + "loss": 1.0376, + "step": 30005 + }, + { + "epoch": 0.71, + "grad_norm": 2.182470466068739, + "learning_rate": 4.176851266069126e-06, + "loss": 1.004, + "step": 30006 + }, + { + "epoch": 0.71, + "grad_norm": 2.9477862501968586, + "learning_rate": 4.176230958614729e-06, + "loss": 1.0529, + "step": 30007 + }, + { + "epoch": 0.71, + "grad_norm": 2.0470546463171364, + "learning_rate": 4.1756106850681624e-06, + "loss": 1.0956, + "step": 30008 + }, + { + "epoch": 0.71, + "grad_norm": 2.0349027622010514, + "learning_rate": 4.174990445433032e-06, + "loss": 0.9488, + "step": 30009 + }, + { + "epoch": 0.71, + "grad_norm": 2.3392664732792228, + "learning_rate": 4.174370239712952e-06, + "loss": 0.8897, + "step": 30010 + }, + { + "epoch": 0.71, + "grad_norm": 2.1946941710081584, + "learning_rate": 4.173750067911532e-06, + "loss": 0.92, + "step": 30011 + }, + { + "epoch": 0.71, + "grad_norm": 2.1425585779560703, + "learning_rate": 4.173129930032387e-06, + "loss": 1.047, + "step": 30012 + }, + { + "epoch": 0.71, + "grad_norm": 1.98293819104732, + "learning_rate": 4.17250982607912e-06, + "loss": 0.9778, + "step": 30013 + }, + { + "epoch": 0.71, + "grad_norm": 1.8684272959536283, + "learning_rate": 4.17188975605535e-06, + "loss": 0.9989, + "step": 30014 + }, + { + "epoch": 0.71, + "grad_norm": 1.9104805389020523, + "learning_rate": 4.171269719964681e-06, + "loss": 0.9879, + "step": 30015 + }, + { + "epoch": 0.71, + "grad_norm": 1.9483955811701956, + "learning_rate": 4.170649717810723e-06, + "loss": 1.0165, + "step": 30016 + }, + { + "epoch": 0.71, + "grad_norm": 3.0634166415184905, + "learning_rate": 4.170029749597087e-06, + "loss": 1.0426, + "step": 30017 + }, + { + "epoch": 0.71, + "grad_norm": 2.194499137171718, + "learning_rate": 4.169409815327388e-06, + "loss": 0.9723, + "step": 30018 + }, + { + "epoch": 0.71, + "grad_norm": 1.9933194617837917, + "learning_rate": 4.168789915005233e-06, + "loss": 0.8371, + "step": 30019 + }, + { + "epoch": 0.71, + "grad_norm": 1.9164444348650747, + "learning_rate": 4.168170048634223e-06, + "loss": 0.9586, + "step": 30020 + }, + { + "epoch": 0.71, + "grad_norm": 1.0477844387430182, + "learning_rate": 4.167550216217974e-06, + "loss": 0.9223, + "step": 30021 + }, + { + "epoch": 0.71, + "grad_norm": 2.4090448304275043, + "learning_rate": 4.1669304177600964e-06, + "loss": 0.9157, + "step": 30022 + }, + { + "epoch": 0.71, + "grad_norm": 1.8670380962275277, + "learning_rate": 4.166310653264199e-06, + "loss": 1.0673, + "step": 30023 + }, + { + "epoch": 0.71, + "grad_norm": 1.9930080506254775, + "learning_rate": 4.1656909227338826e-06, + "loss": 1.0409, + "step": 30024 + }, + { + "epoch": 0.71, + "grad_norm": 2.0109516561436176, + "learning_rate": 4.1650712261727614e-06, + "loss": 1.0402, + "step": 30025 + }, + { + "epoch": 0.71, + "grad_norm": 2.1075059961900986, + "learning_rate": 4.164451563584446e-06, + "loss": 1.0776, + "step": 30026 + }, + { + "epoch": 0.71, + "grad_norm": 1.8142123863363884, + "learning_rate": 4.163831934972541e-06, + "loss": 0.8927, + "step": 30027 + }, + { + "epoch": 0.71, + "grad_norm": 1.8989782307738816, + "learning_rate": 4.163212340340651e-06, + "loss": 1.0164, + "step": 30028 + }, + { + "epoch": 0.71, + "grad_norm": 1.8772587457279228, + "learning_rate": 4.16259277969239e-06, + "loss": 1.1033, + "step": 30029 + }, + { + "epoch": 0.71, + "grad_norm": 1.8769060857398054, + "learning_rate": 4.161973253031359e-06, + "loss": 0.9691, + "step": 30030 + }, + { + "epoch": 0.71, + "grad_norm": 2.16966560410171, + "learning_rate": 4.161353760361171e-06, + "loss": 0.871, + "step": 30031 + }, + { + "epoch": 0.71, + "grad_norm": 1.8780830167463114, + "learning_rate": 4.160734301685427e-06, + "loss": 0.9002, + "step": 30032 + }, + { + "epoch": 0.71, + "grad_norm": 1.9612804190535387, + "learning_rate": 4.160114877007739e-06, + "loss": 0.8748, + "step": 30033 + }, + { + "epoch": 0.71, + "grad_norm": 2.25890548847589, + "learning_rate": 4.159495486331708e-06, + "loss": 0.9377, + "step": 30034 + }, + { + "epoch": 0.71, + "grad_norm": 1.8954298962965468, + "learning_rate": 4.1588761296609485e-06, + "loss": 0.9813, + "step": 30035 + }, + { + "epoch": 0.71, + "grad_norm": 2.019697605130575, + "learning_rate": 4.158256806999059e-06, + "loss": 0.9682, + "step": 30036 + }, + { + "epoch": 0.71, + "grad_norm": 1.988835226102963, + "learning_rate": 4.157637518349646e-06, + "loss": 0.9304, + "step": 30037 + }, + { + "epoch": 0.71, + "grad_norm": 2.2400425754637316, + "learning_rate": 4.1570182637163155e-06, + "loss": 0.9301, + "step": 30038 + }, + { + "epoch": 0.71, + "grad_norm": 2.3866121162741445, + "learning_rate": 4.156399043102675e-06, + "loss": 0.9029, + "step": 30039 + }, + { + "epoch": 0.71, + "grad_norm": 2.720500108198707, + "learning_rate": 4.155779856512338e-06, + "loss": 0.9589, + "step": 30040 + }, + { + "epoch": 0.71, + "grad_norm": 2.306994980754817, + "learning_rate": 4.155160703948892e-06, + "loss": 1.0075, + "step": 30041 + }, + { + "epoch": 0.71, + "grad_norm": 1.8466332110523112, + "learning_rate": 4.154541585415951e-06, + "loss": 1.007, + "step": 30042 + }, + { + "epoch": 0.71, + "grad_norm": 1.8578132604127424, + "learning_rate": 4.153922500917123e-06, + "loss": 1.0224, + "step": 30043 + }, + { + "epoch": 0.71, + "grad_norm": 2.2610163540952826, + "learning_rate": 4.15330345045601e-06, + "loss": 0.963, + "step": 30044 + }, + { + "epoch": 0.71, + "grad_norm": 2.007615850935995, + "learning_rate": 4.152684434036212e-06, + "loss": 0.8033, + "step": 30045 + }, + { + "epoch": 0.71, + "grad_norm": 2.0060907380661726, + "learning_rate": 4.152065451661334e-06, + "loss": 1.0537, + "step": 30046 + }, + { + "epoch": 0.71, + "grad_norm": 2.1228837356446646, + "learning_rate": 4.151446503334988e-06, + "loss": 0.9125, + "step": 30047 + }, + { + "epoch": 0.71, + "grad_norm": 1.9166568852981425, + "learning_rate": 4.150827589060769e-06, + "loss": 0.9544, + "step": 30048 + }, + { + "epoch": 0.71, + "grad_norm": 1.1000733703035486, + "learning_rate": 4.150208708842286e-06, + "loss": 0.9231, + "step": 30049 + }, + { + "epoch": 0.71, + "grad_norm": 2.0534102235139025, + "learning_rate": 4.149589862683141e-06, + "loss": 0.8417, + "step": 30050 + }, + { + "epoch": 0.71, + "grad_norm": 1.1070177978439975, + "learning_rate": 4.1489710505869305e-06, + "loss": 0.9323, + "step": 30051 + }, + { + "epoch": 0.71, + "grad_norm": 1.836524912398372, + "learning_rate": 4.148352272557264e-06, + "loss": 0.9903, + "step": 30052 + }, + { + "epoch": 0.71, + "grad_norm": 2.203139840872471, + "learning_rate": 4.147733528597748e-06, + "loss": 1.0305, + "step": 30053 + }, + { + "epoch": 0.71, + "grad_norm": 2.196996671004728, + "learning_rate": 4.147114818711979e-06, + "loss": 0.7646, + "step": 30054 + }, + { + "epoch": 0.71, + "grad_norm": 1.078978147073221, + "learning_rate": 4.146496142903557e-06, + "loss": 0.9745, + "step": 30055 + }, + { + "epoch": 0.71, + "grad_norm": 2.0209824502165987, + "learning_rate": 4.145877501176088e-06, + "loss": 0.881, + "step": 30056 + }, + { + "epoch": 0.71, + "grad_norm": 2.9579870423040813, + "learning_rate": 4.145258893533179e-06, + "loss": 1.0723, + "step": 30057 + }, + { + "epoch": 0.71, + "grad_norm": 1.9608688617349561, + "learning_rate": 4.144640319978422e-06, + "loss": 0.9292, + "step": 30058 + }, + { + "epoch": 0.71, + "grad_norm": 1.9766546897441322, + "learning_rate": 4.144021780515421e-06, + "loss": 0.9187, + "step": 30059 + }, + { + "epoch": 0.71, + "grad_norm": 2.323596903164052, + "learning_rate": 4.1434032751477794e-06, + "loss": 0.9434, + "step": 30060 + }, + { + "epoch": 0.71, + "grad_norm": 2.121402647514161, + "learning_rate": 4.142784803879102e-06, + "loss": 0.9908, + "step": 30061 + }, + { + "epoch": 0.71, + "grad_norm": 2.03029677337274, + "learning_rate": 4.142166366712985e-06, + "loss": 1.0342, + "step": 30062 + }, + { + "epoch": 0.71, + "grad_norm": 2.0782268784224103, + "learning_rate": 4.141547963653027e-06, + "loss": 1.0069, + "step": 30063 + }, + { + "epoch": 0.71, + "grad_norm": 2.061249630082119, + "learning_rate": 4.140929594702836e-06, + "loss": 0.9964, + "step": 30064 + }, + { + "epoch": 0.71, + "grad_norm": 2.0985110926798405, + "learning_rate": 4.140311259866003e-06, + "loss": 1.0767, + "step": 30065 + }, + { + "epoch": 0.71, + "grad_norm": 2.102011020519178, + "learning_rate": 4.139692959146138e-06, + "loss": 0.921, + "step": 30066 + }, + { + "epoch": 0.71, + "grad_norm": 1.9783797519880462, + "learning_rate": 4.139074692546831e-06, + "loss": 1.0083, + "step": 30067 + }, + { + "epoch": 0.71, + "grad_norm": 1.9384927063132795, + "learning_rate": 4.138456460071691e-06, + "loss": 0.866, + "step": 30068 + }, + { + "epoch": 0.71, + "grad_norm": 2.1553047895851885, + "learning_rate": 4.137838261724308e-06, + "loss": 0.9357, + "step": 30069 + }, + { + "epoch": 0.71, + "grad_norm": 1.9113547783935847, + "learning_rate": 4.137220097508292e-06, + "loss": 0.9138, + "step": 30070 + }, + { + "epoch": 0.71, + "grad_norm": 1.9732707294075218, + "learning_rate": 4.136601967427236e-06, + "loss": 1.0766, + "step": 30071 + }, + { + "epoch": 0.71, + "grad_norm": 1.0819463782644487, + "learning_rate": 4.135983871484735e-06, + "loss": 0.9887, + "step": 30072 + }, + { + "epoch": 0.71, + "grad_norm": 2.0102870925450818, + "learning_rate": 4.135365809684394e-06, + "loss": 0.9679, + "step": 30073 + }, + { + "epoch": 0.71, + "grad_norm": 2.1426474610865918, + "learning_rate": 4.134747782029814e-06, + "loss": 0.9942, + "step": 30074 + }, + { + "epoch": 0.71, + "grad_norm": 4.672735990043772, + "learning_rate": 4.134129788524588e-06, + "loss": 1.0597, + "step": 30075 + }, + { + "epoch": 0.71, + "grad_norm": 2.0248245690465203, + "learning_rate": 4.133511829172313e-06, + "loss": 1.0192, + "step": 30076 + }, + { + "epoch": 0.71, + "grad_norm": 2.0329601267338084, + "learning_rate": 4.132893903976588e-06, + "loss": 1.0774, + "step": 30077 + }, + { + "epoch": 0.71, + "grad_norm": 1.8621539055865488, + "learning_rate": 4.132276012941016e-06, + "loss": 0.9645, + "step": 30078 + }, + { + "epoch": 0.71, + "grad_norm": 1.7470549081130708, + "learning_rate": 4.1316581560691924e-06, + "loss": 1.0392, + "step": 30079 + }, + { + "epoch": 0.71, + "grad_norm": 2.8331874582032888, + "learning_rate": 4.131040333364708e-06, + "loss": 0.9477, + "step": 30080 + }, + { + "epoch": 0.71, + "grad_norm": 1.9042451936482345, + "learning_rate": 4.130422544831165e-06, + "loss": 1.0434, + "step": 30081 + }, + { + "epoch": 0.71, + "grad_norm": 2.068999060406477, + "learning_rate": 4.129804790472164e-06, + "loss": 1.0945, + "step": 30082 + }, + { + "epoch": 0.71, + "grad_norm": 2.2220499980335235, + "learning_rate": 4.129187070291298e-06, + "loss": 1.0748, + "step": 30083 + }, + { + "epoch": 0.71, + "grad_norm": 2.0820442333720193, + "learning_rate": 4.128569384292161e-06, + "loss": 0.895, + "step": 30084 + }, + { + "epoch": 0.71, + "grad_norm": 2.1152148881540316, + "learning_rate": 4.127951732478353e-06, + "loss": 0.9606, + "step": 30085 + }, + { + "epoch": 0.71, + "grad_norm": 1.9062756250004635, + "learning_rate": 4.127334114853467e-06, + "loss": 0.9813, + "step": 30086 + }, + { + "epoch": 0.71, + "grad_norm": 2.0307482301411284, + "learning_rate": 4.126716531421105e-06, + "loss": 0.985, + "step": 30087 + }, + { + "epoch": 0.71, + "grad_norm": 1.1600742120427614, + "learning_rate": 4.126098982184854e-06, + "loss": 0.9311, + "step": 30088 + }, + { + "epoch": 0.71, + "grad_norm": 2.0952876084368945, + "learning_rate": 4.125481467148319e-06, + "loss": 0.8982, + "step": 30089 + }, + { + "epoch": 0.71, + "grad_norm": 2.032114656438641, + "learning_rate": 4.124863986315087e-06, + "loss": 0.9275, + "step": 30090 + }, + { + "epoch": 0.71, + "grad_norm": 2.0376509160819998, + "learning_rate": 4.1242465396887595e-06, + "loss": 0.9726, + "step": 30091 + }, + { + "epoch": 0.71, + "grad_norm": 1.9455032359852467, + "learning_rate": 4.12362912727293e-06, + "loss": 0.8913, + "step": 30092 + }, + { + "epoch": 0.71, + "grad_norm": 1.0482009446767968, + "learning_rate": 4.123011749071188e-06, + "loss": 0.9371, + "step": 30093 + }, + { + "epoch": 0.71, + "grad_norm": 1.0989536308651995, + "learning_rate": 4.122394405087132e-06, + "loss": 0.9667, + "step": 30094 + }, + { + "epoch": 0.71, + "grad_norm": 1.8246017283524614, + "learning_rate": 4.121777095324361e-06, + "loss": 0.9305, + "step": 30095 + }, + { + "epoch": 0.71, + "grad_norm": 1.9080031784611289, + "learning_rate": 4.121159819786465e-06, + "loss": 1.0845, + "step": 30096 + }, + { + "epoch": 0.71, + "grad_norm": 1.8208087193118347, + "learning_rate": 4.120542578477033e-06, + "loss": 0.9771, + "step": 30097 + }, + { + "epoch": 0.71, + "grad_norm": 1.9129151018650556, + "learning_rate": 4.119925371399665e-06, + "loss": 0.7755, + "step": 30098 + }, + { + "epoch": 0.71, + "grad_norm": 1.079024653045839, + "learning_rate": 4.119308198557955e-06, + "loss": 0.9846, + "step": 30099 + }, + { + "epoch": 0.71, + "grad_norm": 2.021073172275222, + "learning_rate": 4.1186910599554954e-06, + "loss": 0.9512, + "step": 30100 + }, + { + "epoch": 0.71, + "grad_norm": 2.3353473336839587, + "learning_rate": 4.118073955595873e-06, + "loss": 0.9561, + "step": 30101 + }, + { + "epoch": 0.71, + "grad_norm": 1.9064202426838583, + "learning_rate": 4.117456885482689e-06, + "loss": 0.9817, + "step": 30102 + }, + { + "epoch": 0.71, + "grad_norm": 1.8242664177770598, + "learning_rate": 4.116839849619536e-06, + "loss": 1.0431, + "step": 30103 + }, + { + "epoch": 0.71, + "grad_norm": 1.8252983546370172, + "learning_rate": 4.116222848010004e-06, + "loss": 1.0398, + "step": 30104 + }, + { + "epoch": 0.71, + "grad_norm": 2.810048652485514, + "learning_rate": 4.1156058806576815e-06, + "loss": 0.9111, + "step": 30105 + }, + { + "epoch": 0.71, + "grad_norm": 2.5177543557554922, + "learning_rate": 4.114988947566167e-06, + "loss": 1.037, + "step": 30106 + }, + { + "epoch": 0.71, + "grad_norm": 1.9958803292339684, + "learning_rate": 4.114372048739048e-06, + "loss": 1.0399, + "step": 30107 + }, + { + "epoch": 0.71, + "grad_norm": 1.821468265578991, + "learning_rate": 4.113755184179922e-06, + "loss": 0.9988, + "step": 30108 + }, + { + "epoch": 0.71, + "grad_norm": 1.987974645888112, + "learning_rate": 4.113138353892373e-06, + "loss": 0.88, + "step": 30109 + }, + { + "epoch": 0.71, + "grad_norm": 2.052904991904684, + "learning_rate": 4.11252155788e-06, + "loss": 0.934, + "step": 30110 + }, + { + "epoch": 0.71, + "grad_norm": 2.0628327953273966, + "learning_rate": 4.1119047961463855e-06, + "loss": 1.0383, + "step": 30111 + }, + { + "epoch": 0.71, + "grad_norm": 2.57636031829208, + "learning_rate": 4.1112880686951306e-06, + "loss": 0.9811, + "step": 30112 + }, + { + "epoch": 0.71, + "grad_norm": 4.4419807854950015, + "learning_rate": 4.110671375529817e-06, + "loss": 0.9975, + "step": 30113 + }, + { + "epoch": 0.71, + "grad_norm": 1.1296176088116021, + "learning_rate": 4.110054716654042e-06, + "loss": 0.9815, + "step": 30114 + }, + { + "epoch": 0.71, + "grad_norm": 2.0932783017615044, + "learning_rate": 4.109438092071389e-06, + "loss": 0.9918, + "step": 30115 + }, + { + "epoch": 0.71, + "grad_norm": 1.9578573095955552, + "learning_rate": 4.108821501785458e-06, + "loss": 1.0977, + "step": 30116 + }, + { + "epoch": 0.71, + "grad_norm": 1.9782930402041528, + "learning_rate": 4.1082049457998316e-06, + "loss": 1.0584, + "step": 30117 + }, + { + "epoch": 0.71, + "grad_norm": 3.2568849515944174, + "learning_rate": 4.107588424118098e-06, + "loss": 0.9051, + "step": 30118 + }, + { + "epoch": 0.71, + "grad_norm": 1.8514258643881687, + "learning_rate": 4.106971936743851e-06, + "loss": 1.0161, + "step": 30119 + }, + { + "epoch": 0.71, + "grad_norm": 1.9752903665686805, + "learning_rate": 4.106355483680683e-06, + "loss": 0.9256, + "step": 30120 + }, + { + "epoch": 0.71, + "grad_norm": 3.1538182822260423, + "learning_rate": 4.105739064932178e-06, + "loss": 0.8449, + "step": 30121 + }, + { + "epoch": 0.71, + "grad_norm": 2.3307275422674874, + "learning_rate": 4.105122680501923e-06, + "loss": 0.969, + "step": 30122 + }, + { + "epoch": 0.71, + "grad_norm": 2.0677738826087704, + "learning_rate": 4.1045063303935105e-06, + "loss": 1.0481, + "step": 30123 + }, + { + "epoch": 0.71, + "grad_norm": 1.8240368145330543, + "learning_rate": 4.103890014610532e-06, + "loss": 0.9312, + "step": 30124 + }, + { + "epoch": 0.71, + "grad_norm": 2.2552736168992884, + "learning_rate": 4.103273733156572e-06, + "loss": 0.895, + "step": 30125 + }, + { + "epoch": 0.71, + "grad_norm": 1.069227327858077, + "learning_rate": 4.102657486035216e-06, + "loss": 0.8918, + "step": 30126 + }, + { + "epoch": 0.71, + "grad_norm": 1.9273837640551272, + "learning_rate": 4.102041273250055e-06, + "loss": 0.9559, + "step": 30127 + }, + { + "epoch": 0.71, + "grad_norm": 1.7474351724372248, + "learning_rate": 4.1014250948046805e-06, + "loss": 0.9542, + "step": 30128 + }, + { + "epoch": 0.71, + "grad_norm": 1.9153933938922048, + "learning_rate": 4.100808950702677e-06, + "loss": 0.8692, + "step": 30129 + }, + { + "epoch": 0.71, + "grad_norm": 3.5083971081346648, + "learning_rate": 4.100192840947628e-06, + "loss": 0.9138, + "step": 30130 + }, + { + "epoch": 0.71, + "grad_norm": 2.0127196045575197, + "learning_rate": 4.099576765543128e-06, + "loss": 1.0816, + "step": 30131 + }, + { + "epoch": 0.71, + "grad_norm": 2.264295633181646, + "learning_rate": 4.098960724492756e-06, + "loss": 0.9794, + "step": 30132 + }, + { + "epoch": 0.71, + "grad_norm": 1.9602519273355987, + "learning_rate": 4.098344717800105e-06, + "loss": 0.979, + "step": 30133 + }, + { + "epoch": 0.71, + "grad_norm": 2.1114779606874974, + "learning_rate": 4.0977287454687575e-06, + "loss": 1.0067, + "step": 30134 + }, + { + "epoch": 0.71, + "grad_norm": 1.9088302456535795, + "learning_rate": 4.097112807502306e-06, + "loss": 1.0818, + "step": 30135 + }, + { + "epoch": 0.71, + "grad_norm": 2.1162475732176764, + "learning_rate": 4.096496903904328e-06, + "loss": 0.906, + "step": 30136 + }, + { + "epoch": 0.71, + "grad_norm": 1.969998285854766, + "learning_rate": 4.095881034678414e-06, + "loss": 1.0523, + "step": 30137 + }, + { + "epoch": 0.71, + "grad_norm": 1.1095848944806548, + "learning_rate": 4.095265199828157e-06, + "loss": 0.9476, + "step": 30138 + }, + { + "epoch": 0.71, + "grad_norm": 2.0719407620307835, + "learning_rate": 4.094649399357129e-06, + "loss": 1.0249, + "step": 30139 + }, + { + "epoch": 0.71, + "grad_norm": 1.8800023704756248, + "learning_rate": 4.0940336332689216e-06, + "loss": 1.0983, + "step": 30140 + }, + { + "epoch": 0.71, + "grad_norm": 2.0349989201473435, + "learning_rate": 4.093417901567119e-06, + "loss": 0.9961, + "step": 30141 + }, + { + "epoch": 0.71, + "grad_norm": 1.9834172263445842, + "learning_rate": 4.092802204255316e-06, + "loss": 1.0477, + "step": 30142 + }, + { + "epoch": 0.71, + "grad_norm": 1.8707439388284395, + "learning_rate": 4.092186541337081e-06, + "loss": 1.0781, + "step": 30143 + }, + { + "epoch": 0.71, + "grad_norm": 2.00889462665138, + "learning_rate": 4.091570912816006e-06, + "loss": 1.0482, + "step": 30144 + }, + { + "epoch": 0.71, + "grad_norm": 2.013740819396878, + "learning_rate": 4.090955318695679e-06, + "loss": 0.8743, + "step": 30145 + }, + { + "epoch": 0.71, + "grad_norm": 1.952525790088247, + "learning_rate": 4.090339758979679e-06, + "loss": 1.0521, + "step": 30146 + }, + { + "epoch": 0.71, + "grad_norm": 2.4437330148581027, + "learning_rate": 4.089724233671594e-06, + "loss": 0.9888, + "step": 30147 + }, + { + "epoch": 0.71, + "grad_norm": 1.1075205548706115, + "learning_rate": 4.089108742775003e-06, + "loss": 0.9201, + "step": 30148 + }, + { + "epoch": 0.71, + "grad_norm": 2.2121198232159154, + "learning_rate": 4.088493286293495e-06, + "loss": 0.9689, + "step": 30149 + }, + { + "epoch": 0.71, + "grad_norm": 1.8757484566428804, + "learning_rate": 4.0878778642306475e-06, + "loss": 0.8737, + "step": 30150 + }, + { + "epoch": 0.71, + "grad_norm": 1.1640170745739067, + "learning_rate": 4.08726247659005e-06, + "loss": 0.9807, + "step": 30151 + }, + { + "epoch": 0.71, + "grad_norm": 1.9526677106537913, + "learning_rate": 4.086647123375282e-06, + "loss": 1.0476, + "step": 30152 + }, + { + "epoch": 0.71, + "grad_norm": 2.9278405364628, + "learning_rate": 4.086031804589925e-06, + "loss": 0.9924, + "step": 30153 + }, + { + "epoch": 0.71, + "grad_norm": 1.9917193113904783, + "learning_rate": 4.085416520237562e-06, + "loss": 0.9723, + "step": 30154 + }, + { + "epoch": 0.71, + "grad_norm": 1.8611064765909142, + "learning_rate": 4.0848012703217785e-06, + "loss": 0.965, + "step": 30155 + }, + { + "epoch": 0.71, + "grad_norm": 2.052619508928581, + "learning_rate": 4.0841860548461565e-06, + "loss": 0.9125, + "step": 30156 + }, + { + "epoch": 0.71, + "grad_norm": 1.0833537984760067, + "learning_rate": 4.083570873814273e-06, + "loss": 0.9313, + "step": 30157 + }, + { + "epoch": 0.71, + "grad_norm": 1.7674530631155383, + "learning_rate": 4.082955727229712e-06, + "loss": 1.0305, + "step": 30158 + }, + { + "epoch": 0.71, + "grad_norm": 1.750225541083425, + "learning_rate": 4.08234061509606e-06, + "loss": 0.9246, + "step": 30159 + }, + { + "epoch": 0.71, + "grad_norm": 1.9516642910214683, + "learning_rate": 4.081725537416894e-06, + "loss": 0.9993, + "step": 30160 + }, + { + "epoch": 0.71, + "grad_norm": 1.959329521926783, + "learning_rate": 4.0811104941957924e-06, + "loss": 0.9608, + "step": 30161 + }, + { + "epoch": 0.71, + "grad_norm": 2.1689834932730747, + "learning_rate": 4.0804954854363395e-06, + "loss": 0.9349, + "step": 30162 + }, + { + "epoch": 0.71, + "grad_norm": 1.0320095037720174, + "learning_rate": 4.079880511142119e-06, + "loss": 0.936, + "step": 30163 + }, + { + "epoch": 0.71, + "grad_norm": 1.076048430811394, + "learning_rate": 4.07926557131671e-06, + "loss": 0.9921, + "step": 30164 + }, + { + "epoch": 0.71, + "grad_norm": 1.9335426091405477, + "learning_rate": 4.078650665963686e-06, + "loss": 1.0418, + "step": 30165 + }, + { + "epoch": 0.71, + "grad_norm": 2.1031282452534183, + "learning_rate": 4.078035795086636e-06, + "loss": 0.9461, + "step": 30166 + }, + { + "epoch": 0.71, + "grad_norm": 1.9937607004437552, + "learning_rate": 4.077420958689134e-06, + "loss": 1.0215, + "step": 30167 + }, + { + "epoch": 0.71, + "grad_norm": 1.0810204102475807, + "learning_rate": 4.076806156774765e-06, + "loss": 0.9272, + "step": 30168 + }, + { + "epoch": 0.71, + "grad_norm": 1.9108833781040253, + "learning_rate": 4.076191389347103e-06, + "loss": 0.9248, + "step": 30169 + }, + { + "epoch": 0.71, + "grad_norm": 2.2287687238671365, + "learning_rate": 4.075576656409733e-06, + "loss": 1.0045, + "step": 30170 + }, + { + "epoch": 0.71, + "grad_norm": 2.0970454320091574, + "learning_rate": 4.074961957966228e-06, + "loss": 0.9673, + "step": 30171 + }, + { + "epoch": 0.71, + "grad_norm": 1.8392145672134972, + "learning_rate": 4.0743472940201746e-06, + "loss": 1.0241, + "step": 30172 + }, + { + "epoch": 0.71, + "grad_norm": 1.89721349104251, + "learning_rate": 4.073732664575146e-06, + "loss": 0.9963, + "step": 30173 + }, + { + "epoch": 0.71, + "grad_norm": 1.0756844416886433, + "learning_rate": 4.073118069634719e-06, + "loss": 0.8978, + "step": 30174 + }, + { + "epoch": 0.71, + "grad_norm": 1.9303037310072713, + "learning_rate": 4.072503509202476e-06, + "loss": 0.9447, + "step": 30175 + }, + { + "epoch": 0.71, + "grad_norm": 1.9379386022067477, + "learning_rate": 4.071888983281996e-06, + "loss": 0.8539, + "step": 30176 + }, + { + "epoch": 0.71, + "grad_norm": 1.7570482311831772, + "learning_rate": 4.071274491876856e-06, + "loss": 0.9235, + "step": 30177 + }, + { + "epoch": 0.71, + "grad_norm": 2.12563859016818, + "learning_rate": 4.07066003499063e-06, + "loss": 1.0374, + "step": 30178 + }, + { + "epoch": 0.71, + "grad_norm": 2.5169205123006804, + "learning_rate": 4.070045612626898e-06, + "loss": 0.8921, + "step": 30179 + }, + { + "epoch": 0.71, + "grad_norm": 1.9883178273595006, + "learning_rate": 4.069431224789242e-06, + "loss": 1.0326, + "step": 30180 + }, + { + "epoch": 0.71, + "grad_norm": 3.686758460296366, + "learning_rate": 4.068816871481234e-06, + "loss": 0.8722, + "step": 30181 + }, + { + "epoch": 0.71, + "grad_norm": 2.0580367829161856, + "learning_rate": 4.0682025527064486e-06, + "loss": 0.8768, + "step": 30182 + }, + { + "epoch": 0.71, + "grad_norm": 2.077683468213137, + "learning_rate": 4.067588268468467e-06, + "loss": 0.953, + "step": 30183 + }, + { + "epoch": 0.71, + "grad_norm": 2.0876014134923344, + "learning_rate": 4.066974018770866e-06, + "loss": 0.9146, + "step": 30184 + }, + { + "epoch": 0.71, + "grad_norm": 2.0729762259145676, + "learning_rate": 4.0663598036172225e-06, + "loss": 1.0663, + "step": 30185 + }, + { + "epoch": 0.71, + "grad_norm": 2.1285500272146898, + "learning_rate": 4.065745623011106e-06, + "loss": 0.835, + "step": 30186 + }, + { + "epoch": 0.71, + "grad_norm": 1.9681545353538026, + "learning_rate": 4.065131476956101e-06, + "loss": 0.929, + "step": 30187 + }, + { + "epoch": 0.71, + "grad_norm": 1.9584771747193652, + "learning_rate": 4.0645173654557766e-06, + "loss": 1.0694, + "step": 30188 + }, + { + "epoch": 0.71, + "grad_norm": 1.9245851139638055, + "learning_rate": 4.063903288513714e-06, + "loss": 0.9289, + "step": 30189 + }, + { + "epoch": 0.71, + "grad_norm": 2.197823921112596, + "learning_rate": 4.063289246133484e-06, + "loss": 1.017, + "step": 30190 + }, + { + "epoch": 0.71, + "grad_norm": 1.8117639101826886, + "learning_rate": 4.062675238318665e-06, + "loss": 0.9214, + "step": 30191 + }, + { + "epoch": 0.71, + "grad_norm": 1.9299023618172209, + "learning_rate": 4.062061265072829e-06, + "loss": 0.9533, + "step": 30192 + }, + { + "epoch": 0.71, + "grad_norm": 2.0392257008062513, + "learning_rate": 4.061447326399554e-06, + "loss": 1.1333, + "step": 30193 + }, + { + "epoch": 0.71, + "grad_norm": 1.9980241112563615, + "learning_rate": 4.060833422302412e-06, + "loss": 0.9835, + "step": 30194 + }, + { + "epoch": 0.71, + "grad_norm": 1.9437482666396189, + "learning_rate": 4.060219552784976e-06, + "loss": 1.072, + "step": 30195 + }, + { + "epoch": 0.71, + "grad_norm": 1.1144012934257304, + "learning_rate": 4.059605717850823e-06, + "loss": 0.9335, + "step": 30196 + }, + { + "epoch": 0.71, + "grad_norm": 2.0161480575536177, + "learning_rate": 4.058991917503529e-06, + "loss": 0.9023, + "step": 30197 + }, + { + "epoch": 0.71, + "grad_norm": 2.038471703835611, + "learning_rate": 4.058378151746665e-06, + "loss": 0.9706, + "step": 30198 + }, + { + "epoch": 0.71, + "grad_norm": 2.1493962584295017, + "learning_rate": 4.057764420583801e-06, + "loss": 1.0013, + "step": 30199 + }, + { + "epoch": 0.71, + "grad_norm": 2.085440745522933, + "learning_rate": 4.057150724018514e-06, + "loss": 1.0251, + "step": 30200 + }, + { + "epoch": 0.71, + "grad_norm": 2.656253800319309, + "learning_rate": 4.0565370620543795e-06, + "loss": 1.1017, + "step": 30201 + }, + { + "epoch": 0.71, + "grad_norm": 2.1074775756606683, + "learning_rate": 4.055923434694968e-06, + "loss": 1.0742, + "step": 30202 + }, + { + "epoch": 0.71, + "grad_norm": 2.6615537373528544, + "learning_rate": 4.05530984194385e-06, + "loss": 0.8906, + "step": 30203 + }, + { + "epoch": 0.71, + "grad_norm": 2.099793019358566, + "learning_rate": 4.054696283804599e-06, + "loss": 1.135, + "step": 30204 + }, + { + "epoch": 0.71, + "grad_norm": 1.849291731125952, + "learning_rate": 4.054082760280792e-06, + "loss": 1.0368, + "step": 30205 + }, + { + "epoch": 0.71, + "grad_norm": 5.727669943820638, + "learning_rate": 4.053469271375998e-06, + "loss": 0.9723, + "step": 30206 + }, + { + "epoch": 0.71, + "grad_norm": 2.348563515836407, + "learning_rate": 4.052855817093785e-06, + "loss": 0.9546, + "step": 30207 + }, + { + "epoch": 0.71, + "grad_norm": 2.137949378959926, + "learning_rate": 4.0522423974377325e-06, + "loss": 1.077, + "step": 30208 + }, + { + "epoch": 0.71, + "grad_norm": 1.9799989657120127, + "learning_rate": 4.051629012411403e-06, + "loss": 1.0469, + "step": 30209 + }, + { + "epoch": 0.71, + "grad_norm": 1.7970154339396422, + "learning_rate": 4.051015662018376e-06, + "loss": 0.8677, + "step": 30210 + }, + { + "epoch": 0.71, + "grad_norm": 1.8497357269924317, + "learning_rate": 4.050402346262217e-06, + "loss": 1.0483, + "step": 30211 + }, + { + "epoch": 0.71, + "grad_norm": 1.8000348152149153, + "learning_rate": 4.049789065146502e-06, + "loss": 0.9517, + "step": 30212 + }, + { + "epoch": 0.71, + "grad_norm": 1.7548971371923137, + "learning_rate": 4.049175818674795e-06, + "loss": 0.991, + "step": 30213 + }, + { + "epoch": 0.71, + "grad_norm": 1.9016058319703901, + "learning_rate": 4.048562606850674e-06, + "loss": 0.8981, + "step": 30214 + }, + { + "epoch": 0.71, + "grad_norm": 2.5846021265925616, + "learning_rate": 4.047949429677706e-06, + "loss": 1.0121, + "step": 30215 + }, + { + "epoch": 0.71, + "grad_norm": 1.8816342290259196, + "learning_rate": 4.0473362871594576e-06, + "loss": 0.9389, + "step": 30216 + }, + { + "epoch": 0.71, + "grad_norm": 1.9210496851648207, + "learning_rate": 4.0467231792995024e-06, + "loss": 1.0122, + "step": 30217 + }, + { + "epoch": 0.71, + "grad_norm": 1.6682920110578663, + "learning_rate": 4.046110106101412e-06, + "loss": 0.8487, + "step": 30218 + }, + { + "epoch": 0.71, + "grad_norm": 1.989369509580324, + "learning_rate": 4.045497067568753e-06, + "loss": 0.9816, + "step": 30219 + }, + { + "epoch": 0.71, + "grad_norm": 1.9042629263370465, + "learning_rate": 4.0448840637050934e-06, + "loss": 0.8947, + "step": 30220 + }, + { + "epoch": 0.71, + "grad_norm": 1.7479608184260773, + "learning_rate": 4.044271094514003e-06, + "loss": 1.1591, + "step": 30221 + }, + { + "epoch": 0.71, + "grad_norm": 2.109992749877353, + "learning_rate": 4.043658159999055e-06, + "loss": 1.0382, + "step": 30222 + }, + { + "epoch": 0.71, + "grad_norm": 2.0852964628766943, + "learning_rate": 4.043045260163816e-06, + "loss": 1.0079, + "step": 30223 + }, + { + "epoch": 0.71, + "grad_norm": 2.352243798833214, + "learning_rate": 4.0424323950118495e-06, + "loss": 1.04, + "step": 30224 + }, + { + "epoch": 0.71, + "grad_norm": 1.1092434686773853, + "learning_rate": 4.041819564546727e-06, + "loss": 0.9312, + "step": 30225 + }, + { + "epoch": 0.71, + "grad_norm": 1.9188807454785446, + "learning_rate": 4.041206768772023e-06, + "loss": 0.9835, + "step": 30226 + }, + { + "epoch": 0.71, + "grad_norm": 1.875910870329235, + "learning_rate": 4.040594007691297e-06, + "loss": 0.9745, + "step": 30227 + }, + { + "epoch": 0.71, + "grad_norm": 1.9903808974770527, + "learning_rate": 4.039981281308117e-06, + "loss": 0.9741, + "step": 30228 + }, + { + "epoch": 0.71, + "grad_norm": 1.7867322287981, + "learning_rate": 4.039368589626057e-06, + "loss": 0.9223, + "step": 30229 + }, + { + "epoch": 0.71, + "grad_norm": 1.876560906173107, + "learning_rate": 4.038755932648677e-06, + "loss": 1.0342, + "step": 30230 + }, + { + "epoch": 0.71, + "grad_norm": 1.999751300920673, + "learning_rate": 4.03814331037955e-06, + "loss": 0.9432, + "step": 30231 + }, + { + "epoch": 0.71, + "grad_norm": 2.36381932243398, + "learning_rate": 4.037530722822236e-06, + "loss": 0.9879, + "step": 30232 + }, + { + "epoch": 0.71, + "grad_norm": 2.0405432278767277, + "learning_rate": 4.036918169980311e-06, + "loss": 0.9929, + "step": 30233 + }, + { + "epoch": 0.71, + "grad_norm": 2.0359270722001592, + "learning_rate": 4.03630565185733e-06, + "loss": 0.9356, + "step": 30234 + }, + { + "epoch": 0.71, + "grad_norm": 2.109472595215461, + "learning_rate": 4.035693168456869e-06, + "loss": 1.0167, + "step": 30235 + }, + { + "epoch": 0.71, + "grad_norm": 1.9099425353418742, + "learning_rate": 4.035080719782491e-06, + "loss": 0.8629, + "step": 30236 + }, + { + "epoch": 0.71, + "grad_norm": 2.0627291354097412, + "learning_rate": 4.0344683058377645e-06, + "loss": 0.9099, + "step": 30237 + }, + { + "epoch": 0.71, + "grad_norm": 2.1194241458234204, + "learning_rate": 4.033855926626247e-06, + "loss": 0.9807, + "step": 30238 + }, + { + "epoch": 0.71, + "grad_norm": 1.9261488977039232, + "learning_rate": 4.033243582151509e-06, + "loss": 0.9054, + "step": 30239 + }, + { + "epoch": 0.71, + "grad_norm": 1.943845928687732, + "learning_rate": 4.032631272417123e-06, + "loss": 0.8855, + "step": 30240 + }, + { + "epoch": 0.71, + "grad_norm": 1.9691488387323044, + "learning_rate": 4.03201899742664e-06, + "loss": 0.9694, + "step": 30241 + }, + { + "epoch": 0.71, + "grad_norm": 2.158370706904606, + "learning_rate": 4.031406757183632e-06, + "loss": 1.1222, + "step": 30242 + }, + { + "epoch": 0.71, + "grad_norm": 2.2883171272766027, + "learning_rate": 4.030794551691663e-06, + "loss": 0.9429, + "step": 30243 + }, + { + "epoch": 0.71, + "grad_norm": 1.9709920435099646, + "learning_rate": 4.030182380954303e-06, + "loss": 1.0156, + "step": 30244 + }, + { + "epoch": 0.71, + "grad_norm": 2.0632207236174445, + "learning_rate": 4.02957024497511e-06, + "loss": 0.8939, + "step": 30245 + }, + { + "epoch": 0.71, + "grad_norm": 1.8727932940819407, + "learning_rate": 4.028958143757647e-06, + "loss": 1.1118, + "step": 30246 + }, + { + "epoch": 0.71, + "grad_norm": 2.0956674194582803, + "learning_rate": 4.028346077305482e-06, + "loss": 0.9696, + "step": 30247 + }, + { + "epoch": 0.71, + "grad_norm": 2.5634079831350514, + "learning_rate": 4.027734045622172e-06, + "loss": 1.0874, + "step": 30248 + }, + { + "epoch": 0.71, + "grad_norm": 2.2129291728240434, + "learning_rate": 4.027122048711292e-06, + "loss": 0.9211, + "step": 30249 + }, + { + "epoch": 0.71, + "grad_norm": 2.028886960796381, + "learning_rate": 4.026510086576393e-06, + "loss": 0.9076, + "step": 30250 + }, + { + "epoch": 0.71, + "grad_norm": 2.0038126169906385, + "learning_rate": 4.0258981592210464e-06, + "loss": 0.9901, + "step": 30251 + }, + { + "epoch": 0.71, + "grad_norm": 2.308630601613819, + "learning_rate": 4.0252862666488105e-06, + "loss": 0.9904, + "step": 30252 + }, + { + "epoch": 0.71, + "grad_norm": 1.992565509050997, + "learning_rate": 4.024674408863252e-06, + "loss": 0.9213, + "step": 30253 + }, + { + "epoch": 0.71, + "grad_norm": 1.7857496984543524, + "learning_rate": 4.02406258586793e-06, + "loss": 0.8401, + "step": 30254 + }, + { + "epoch": 0.71, + "grad_norm": 2.1508250360746537, + "learning_rate": 4.023450797666405e-06, + "loss": 0.9735, + "step": 30255 + }, + { + "epoch": 0.71, + "grad_norm": 1.9824911643354652, + "learning_rate": 4.022839044262241e-06, + "loss": 1.0583, + "step": 30256 + }, + { + "epoch": 0.71, + "grad_norm": 2.070579711127487, + "learning_rate": 4.022227325659005e-06, + "loss": 0.9866, + "step": 30257 + }, + { + "epoch": 0.71, + "grad_norm": 2.0603121098344497, + "learning_rate": 4.0216156418602535e-06, + "loss": 0.9454, + "step": 30258 + }, + { + "epoch": 0.71, + "grad_norm": 2.1173548610282555, + "learning_rate": 4.021003992869545e-06, + "loss": 1.1109, + "step": 30259 + }, + { + "epoch": 0.71, + "grad_norm": 2.3723952059905473, + "learning_rate": 4.020392378690444e-06, + "loss": 1.0631, + "step": 30260 + }, + { + "epoch": 0.71, + "grad_norm": 1.1062288117728467, + "learning_rate": 4.019780799326515e-06, + "loss": 0.9549, + "step": 30261 + }, + { + "epoch": 0.71, + "grad_norm": 1.1390447515199222, + "learning_rate": 4.019169254781315e-06, + "loss": 0.9418, + "step": 30262 + }, + { + "epoch": 0.71, + "grad_norm": 1.0477183541500674, + "learning_rate": 4.018557745058403e-06, + "loss": 0.938, + "step": 30263 + }, + { + "epoch": 0.71, + "grad_norm": 2.094987101013358, + "learning_rate": 4.01794627016134e-06, + "loss": 0.8666, + "step": 30264 + }, + { + "epoch": 0.71, + "grad_norm": 1.8757522508227775, + "learning_rate": 4.017334830093693e-06, + "loss": 0.9765, + "step": 30265 + }, + { + "epoch": 0.71, + "grad_norm": 1.8497328399812047, + "learning_rate": 4.016723424859016e-06, + "loss": 1.0098, + "step": 30266 + }, + { + "epoch": 0.71, + "grad_norm": 1.884723588865077, + "learning_rate": 4.016112054460866e-06, + "loss": 1.1267, + "step": 30267 + }, + { + "epoch": 0.71, + "grad_norm": 2.5654830564945637, + "learning_rate": 4.0155007189028086e-06, + "loss": 0.9415, + "step": 30268 + }, + { + "epoch": 0.71, + "grad_norm": 1.99323884392891, + "learning_rate": 4.014889418188398e-06, + "loss": 1.0275, + "step": 30269 + }, + { + "epoch": 0.71, + "grad_norm": 2.1477684721936066, + "learning_rate": 4.014278152321199e-06, + "loss": 1.0389, + "step": 30270 + }, + { + "epoch": 0.71, + "grad_norm": 1.9777146945479351, + "learning_rate": 4.013666921304765e-06, + "loss": 0.9651, + "step": 30271 + }, + { + "epoch": 0.71, + "grad_norm": 2.378745395604389, + "learning_rate": 4.013055725142661e-06, + "loss": 0.7783, + "step": 30272 + }, + { + "epoch": 0.71, + "grad_norm": 2.1523819828488775, + "learning_rate": 4.012444563838437e-06, + "loss": 0.9438, + "step": 30273 + }, + { + "epoch": 0.71, + "grad_norm": 2.4342662663119166, + "learning_rate": 4.011833437395661e-06, + "loss": 0.9713, + "step": 30274 + }, + { + "epoch": 0.71, + "grad_norm": 2.0255275609214576, + "learning_rate": 4.011222345817886e-06, + "loss": 1.1024, + "step": 30275 + }, + { + "epoch": 0.71, + "grad_norm": 2.0742693908799144, + "learning_rate": 4.0106112891086675e-06, + "loss": 1.0171, + "step": 30276 + }, + { + "epoch": 0.71, + "grad_norm": 2.017646849542862, + "learning_rate": 4.010000267271566e-06, + "loss": 1.1071, + "step": 30277 + }, + { + "epoch": 0.71, + "grad_norm": 2.0313127482790914, + "learning_rate": 4.009389280310144e-06, + "loss": 0.9741, + "step": 30278 + }, + { + "epoch": 0.71, + "grad_norm": 1.05605020031347, + "learning_rate": 4.008778328227953e-06, + "loss": 0.9206, + "step": 30279 + }, + { + "epoch": 0.71, + "grad_norm": 2.068075582713855, + "learning_rate": 4.008167411028548e-06, + "loss": 1.1514, + "step": 30280 + }, + { + "epoch": 0.71, + "grad_norm": 2.234149744360228, + "learning_rate": 4.007556528715489e-06, + "loss": 0.9822, + "step": 30281 + }, + { + "epoch": 0.71, + "grad_norm": 2.115749775377965, + "learning_rate": 4.006945681292338e-06, + "loss": 0.9642, + "step": 30282 + }, + { + "epoch": 0.71, + "grad_norm": 1.9324346228914133, + "learning_rate": 4.0063348687626455e-06, + "loss": 0.899, + "step": 30283 + }, + { + "epoch": 0.71, + "grad_norm": 2.2022386927833644, + "learning_rate": 4.005724091129965e-06, + "loss": 0.9547, + "step": 30284 + }, + { + "epoch": 0.71, + "grad_norm": 1.8000232649964036, + "learning_rate": 4.005113348397858e-06, + "loss": 1.0592, + "step": 30285 + }, + { + "epoch": 0.71, + "grad_norm": 1.9793414121764281, + "learning_rate": 4.004502640569882e-06, + "loss": 0.9142, + "step": 30286 + }, + { + "epoch": 0.71, + "grad_norm": 2.1761850576458537, + "learning_rate": 4.003891967649589e-06, + "loss": 0.9457, + "step": 30287 + }, + { + "epoch": 0.71, + "grad_norm": 1.9256073646405685, + "learning_rate": 4.003281329640533e-06, + "loss": 1.0328, + "step": 30288 + }, + { + "epoch": 0.71, + "grad_norm": 2.058448954594031, + "learning_rate": 4.002670726546274e-06, + "loss": 0.9845, + "step": 30289 + }, + { + "epoch": 0.71, + "grad_norm": 2.068463936577133, + "learning_rate": 4.002060158370361e-06, + "loss": 0.9946, + "step": 30290 + }, + { + "epoch": 0.71, + "grad_norm": 1.8486594969258843, + "learning_rate": 4.001449625116357e-06, + "loss": 0.934, + "step": 30291 + }, + { + "epoch": 0.71, + "grad_norm": 2.019887292036032, + "learning_rate": 4.000839126787809e-06, + "loss": 0.9561, + "step": 30292 + }, + { + "epoch": 0.71, + "grad_norm": 1.9196051380471206, + "learning_rate": 4.000228663388278e-06, + "loss": 0.9685, + "step": 30293 + }, + { + "epoch": 0.71, + "grad_norm": 2.0864723524211892, + "learning_rate": 3.999618234921312e-06, + "loss": 1.1392, + "step": 30294 + }, + { + "epoch": 0.71, + "grad_norm": 1.0616747604398107, + "learning_rate": 3.999007841390472e-06, + "loss": 0.9381, + "step": 30295 + }, + { + "epoch": 0.71, + "grad_norm": 2.3030622052685468, + "learning_rate": 3.998397482799307e-06, + "loss": 0.9563, + "step": 30296 + }, + { + "epoch": 0.71, + "grad_norm": 1.995443473638675, + "learning_rate": 3.9977871591513705e-06, + "loss": 0.9606, + "step": 30297 + }, + { + "epoch": 0.71, + "grad_norm": 2.0845534651223905, + "learning_rate": 3.9971768704502156e-06, + "loss": 0.9713, + "step": 30298 + }, + { + "epoch": 0.71, + "grad_norm": 2.3083027536850436, + "learning_rate": 3.996566616699402e-06, + "loss": 0.9998, + "step": 30299 + }, + { + "epoch": 0.71, + "grad_norm": 1.8842282601453482, + "learning_rate": 3.995956397902478e-06, + "loss": 0.9539, + "step": 30300 + }, + { + "epoch": 0.71, + "grad_norm": 2.072246721983254, + "learning_rate": 3.995346214062993e-06, + "loss": 1.0206, + "step": 30301 + }, + { + "epoch": 0.71, + "grad_norm": 1.790752695917277, + "learning_rate": 3.994736065184503e-06, + "loss": 0.9104, + "step": 30302 + }, + { + "epoch": 0.71, + "grad_norm": 1.1164269023554019, + "learning_rate": 3.9941259512705654e-06, + "loss": 0.9104, + "step": 30303 + }, + { + "epoch": 0.71, + "grad_norm": 2.1189330071823256, + "learning_rate": 3.993515872324727e-06, + "loss": 1.0435, + "step": 30304 + }, + { + "epoch": 0.71, + "grad_norm": 1.991818763054262, + "learning_rate": 3.992905828350538e-06, + "loss": 0.8355, + "step": 30305 + }, + { + "epoch": 0.71, + "grad_norm": 2.006687321520101, + "learning_rate": 3.992295819351553e-06, + "loss": 0.9999, + "step": 30306 + }, + { + "epoch": 0.71, + "grad_norm": 1.946203905188726, + "learning_rate": 3.991685845331326e-06, + "loss": 0.9315, + "step": 30307 + }, + { + "epoch": 0.71, + "grad_norm": 1.1368204204285939, + "learning_rate": 3.9910759062934065e-06, + "loss": 0.9461, + "step": 30308 + }, + { + "epoch": 0.71, + "grad_norm": 2.1477201604795977, + "learning_rate": 3.990466002241344e-06, + "loss": 0.9423, + "step": 30309 + }, + { + "epoch": 0.71, + "grad_norm": 2.153033366406625, + "learning_rate": 3.989856133178692e-06, + "loss": 0.9736, + "step": 30310 + }, + { + "epoch": 0.71, + "grad_norm": 1.855724377716409, + "learning_rate": 3.989246299108999e-06, + "loss": 0.8198, + "step": 30311 + }, + { + "epoch": 0.71, + "grad_norm": 2.065374296654055, + "learning_rate": 3.988636500035819e-06, + "loss": 1.1382, + "step": 30312 + }, + { + "epoch": 0.71, + "grad_norm": 2.095728363221103, + "learning_rate": 3.988026735962698e-06, + "loss": 0.9554, + "step": 30313 + }, + { + "epoch": 0.71, + "grad_norm": 1.9063204391135664, + "learning_rate": 3.987417006893191e-06, + "loss": 1.0149, + "step": 30314 + }, + { + "epoch": 0.71, + "grad_norm": 1.9117750898720318, + "learning_rate": 3.986807312830843e-06, + "loss": 1.0421, + "step": 30315 + }, + { + "epoch": 0.71, + "grad_norm": 1.950522665850352, + "learning_rate": 3.98619765377921e-06, + "loss": 0.9663, + "step": 30316 + }, + { + "epoch": 0.71, + "grad_norm": 1.875807829645321, + "learning_rate": 3.985588029741838e-06, + "loss": 1.0561, + "step": 30317 + }, + { + "epoch": 0.71, + "grad_norm": 2.1727087147508652, + "learning_rate": 3.984978440722274e-06, + "loss": 1.0463, + "step": 30318 + }, + { + "epoch": 0.71, + "grad_norm": 2.4599130637855082, + "learning_rate": 3.9843688867240695e-06, + "loss": 0.9736, + "step": 30319 + }, + { + "epoch": 0.71, + "grad_norm": 1.87248823252538, + "learning_rate": 3.983759367750772e-06, + "loss": 1.0986, + "step": 30320 + }, + { + "epoch": 0.71, + "grad_norm": 1.8197874775451124, + "learning_rate": 3.9831498838059415e-06, + "loss": 1.0542, + "step": 30321 + }, + { + "epoch": 0.71, + "grad_norm": 2.544220596472816, + "learning_rate": 3.982540434893109e-06, + "loss": 0.931, + "step": 30322 + }, + { + "epoch": 0.71, + "grad_norm": 1.9161666711906975, + "learning_rate": 3.9819310210158324e-06, + "loss": 0.9255, + "step": 30323 + }, + { + "epoch": 0.71, + "grad_norm": 1.994008426856554, + "learning_rate": 3.981321642177662e-06, + "loss": 0.9583, + "step": 30324 + }, + { + "epoch": 0.71, + "grad_norm": 2.0404343197027734, + "learning_rate": 3.980712298382141e-06, + "loss": 1.0722, + "step": 30325 + }, + { + "epoch": 0.71, + "grad_norm": 2.093874378188618, + "learning_rate": 3.980102989632816e-06, + "loss": 0.9574, + "step": 30326 + }, + { + "epoch": 0.71, + "grad_norm": 1.8454996432051811, + "learning_rate": 3.979493715933238e-06, + "loss": 0.9012, + "step": 30327 + }, + { + "epoch": 0.71, + "grad_norm": 1.9618186583328745, + "learning_rate": 3.978884477286956e-06, + "loss": 0.898, + "step": 30328 + }, + { + "epoch": 0.71, + "grad_norm": 2.208644700485238, + "learning_rate": 3.9782752736975165e-06, + "loss": 1.0563, + "step": 30329 + }, + { + "epoch": 0.71, + "grad_norm": 1.9601413833098857, + "learning_rate": 3.97766610516846e-06, + "loss": 1.0503, + "step": 30330 + }, + { + "epoch": 0.71, + "grad_norm": 2.3507333692616856, + "learning_rate": 3.977056971703342e-06, + "loss": 0.9571, + "step": 30331 + }, + { + "epoch": 0.71, + "grad_norm": 2.348674596050007, + "learning_rate": 3.9764478733057e-06, + "loss": 0.9386, + "step": 30332 + }, + { + "epoch": 0.71, + "grad_norm": 4.718982493241982, + "learning_rate": 3.975838809979087e-06, + "loss": 0.886, + "step": 30333 + }, + { + "epoch": 0.71, + "grad_norm": 2.017580091086812, + "learning_rate": 3.975229781727052e-06, + "loss": 1.0423, + "step": 30334 + }, + { + "epoch": 0.71, + "grad_norm": 2.0837110620165054, + "learning_rate": 3.974620788553135e-06, + "loss": 1.0962, + "step": 30335 + }, + { + "epoch": 0.71, + "grad_norm": 2.287932459575928, + "learning_rate": 3.97401183046088e-06, + "loss": 0.9284, + "step": 30336 + }, + { + "epoch": 0.71, + "grad_norm": 1.7305904908283183, + "learning_rate": 3.973402907453837e-06, + "loss": 1.0174, + "step": 30337 + }, + { + "epoch": 0.71, + "grad_norm": 2.3423135917548654, + "learning_rate": 3.972794019535556e-06, + "loss": 0.8954, + "step": 30338 + }, + { + "epoch": 0.71, + "grad_norm": 2.985888370406542, + "learning_rate": 3.97218516670957e-06, + "loss": 1.0064, + "step": 30339 + }, + { + "epoch": 0.71, + "grad_norm": 1.8691037452431647, + "learning_rate": 3.971576348979431e-06, + "loss": 0.9055, + "step": 30340 + }, + { + "epoch": 0.71, + "grad_norm": 1.8211580547689028, + "learning_rate": 3.970967566348682e-06, + "loss": 0.9435, + "step": 30341 + }, + { + "epoch": 0.71, + "grad_norm": 1.8334141135777426, + "learning_rate": 3.970358818820873e-06, + "loss": 0.869, + "step": 30342 + }, + { + "epoch": 0.71, + "grad_norm": 1.957893335772559, + "learning_rate": 3.9697501063995435e-06, + "loss": 1.0902, + "step": 30343 + }, + { + "epoch": 0.71, + "grad_norm": 2.0961454050102386, + "learning_rate": 3.969141429088235e-06, + "loss": 0.9719, + "step": 30344 + }, + { + "epoch": 0.71, + "grad_norm": 1.8414987270179017, + "learning_rate": 3.968532786890496e-06, + "loss": 0.9652, + "step": 30345 + }, + { + "epoch": 0.71, + "grad_norm": 1.9391724197511693, + "learning_rate": 3.967924179809868e-06, + "loss": 0.9734, + "step": 30346 + }, + { + "epoch": 0.71, + "grad_norm": 2.003191290691539, + "learning_rate": 3.9673156078498964e-06, + "loss": 0.8672, + "step": 30347 + }, + { + "epoch": 0.71, + "grad_norm": 1.8620381915995097, + "learning_rate": 3.966707071014121e-06, + "loss": 0.8853, + "step": 30348 + }, + { + "epoch": 0.71, + "grad_norm": 1.8268105705119444, + "learning_rate": 3.966098569306091e-06, + "loss": 1.0597, + "step": 30349 + }, + { + "epoch": 0.72, + "grad_norm": 2.1121703789471145, + "learning_rate": 3.9654901027293405e-06, + "loss": 0.9762, + "step": 30350 + }, + { + "epoch": 0.72, + "grad_norm": 1.0256312769099747, + "learning_rate": 3.964881671287422e-06, + "loss": 0.9625, + "step": 30351 + }, + { + "epoch": 0.72, + "grad_norm": 2.7509422984814536, + "learning_rate": 3.964273274983872e-06, + "loss": 0.9064, + "step": 30352 + }, + { + "epoch": 0.72, + "grad_norm": 2.2589476321924753, + "learning_rate": 3.963664913822231e-06, + "loss": 0.9353, + "step": 30353 + }, + { + "epoch": 0.72, + "grad_norm": 2.1850527175099725, + "learning_rate": 3.963056587806044e-06, + "loss": 0.984, + "step": 30354 + }, + { + "epoch": 0.72, + "grad_norm": 2.020399674277722, + "learning_rate": 3.962448296938854e-06, + "loss": 0.9331, + "step": 30355 + }, + { + "epoch": 0.72, + "grad_norm": 2.01295700088727, + "learning_rate": 3.961840041224204e-06, + "loss": 0.9328, + "step": 30356 + }, + { + "epoch": 0.72, + "grad_norm": 2.1031564741964335, + "learning_rate": 3.96123182066563e-06, + "loss": 1.0343, + "step": 30357 + }, + { + "epoch": 0.72, + "grad_norm": 2.153086706387258, + "learning_rate": 3.960623635266675e-06, + "loss": 1.0343, + "step": 30358 + }, + { + "epoch": 0.72, + "grad_norm": 1.198913622574887, + "learning_rate": 3.960015485030885e-06, + "loss": 0.8793, + "step": 30359 + }, + { + "epoch": 0.72, + "grad_norm": 1.8008198930479176, + "learning_rate": 3.959407369961796e-06, + "loss": 0.8916, + "step": 30360 + }, + { + "epoch": 0.72, + "grad_norm": 1.9093324361731114, + "learning_rate": 3.958799290062947e-06, + "loss": 0.8559, + "step": 30361 + }, + { + "epoch": 0.72, + "grad_norm": 2.536603400872716, + "learning_rate": 3.958191245337881e-06, + "loss": 0.8808, + "step": 30362 + }, + { + "epoch": 0.72, + "grad_norm": 2.0989084853745252, + "learning_rate": 3.957583235790142e-06, + "loss": 1.0563, + "step": 30363 + }, + { + "epoch": 0.72, + "grad_norm": 2.783973175315735, + "learning_rate": 3.956975261423267e-06, + "loss": 1.1318, + "step": 30364 + }, + { + "epoch": 0.72, + "grad_norm": 1.9729607311108894, + "learning_rate": 3.956367322240792e-06, + "loss": 0.9142, + "step": 30365 + }, + { + "epoch": 0.72, + "grad_norm": 1.8031704115458855, + "learning_rate": 3.955759418246259e-06, + "loss": 1.0323, + "step": 30366 + }, + { + "epoch": 0.72, + "grad_norm": 1.873997183690886, + "learning_rate": 3.955151549443211e-06, + "loss": 1.0277, + "step": 30367 + }, + { + "epoch": 0.72, + "grad_norm": 2.1343333544174694, + "learning_rate": 3.954543715835185e-06, + "loss": 0.9924, + "step": 30368 + }, + { + "epoch": 0.72, + "grad_norm": 1.9643283693254017, + "learning_rate": 3.953935917425716e-06, + "loss": 0.8508, + "step": 30369 + }, + { + "epoch": 0.72, + "grad_norm": 2.1504162688116506, + "learning_rate": 3.953328154218351e-06, + "loss": 0.9222, + "step": 30370 + }, + { + "epoch": 0.72, + "grad_norm": 1.0638359316913906, + "learning_rate": 3.952720426216618e-06, + "loss": 0.9325, + "step": 30371 + }, + { + "epoch": 0.72, + "grad_norm": 1.1550662591761414, + "learning_rate": 3.952112733424067e-06, + "loss": 0.8996, + "step": 30372 + }, + { + "epoch": 0.72, + "grad_norm": 1.1267489272256765, + "learning_rate": 3.951505075844225e-06, + "loss": 0.9998, + "step": 30373 + }, + { + "epoch": 0.72, + "grad_norm": 1.8009810111923055, + "learning_rate": 3.95089745348064e-06, + "loss": 1.0618, + "step": 30374 + }, + { + "epoch": 0.72, + "grad_norm": 1.998203084685382, + "learning_rate": 3.950289866336842e-06, + "loss": 0.9594, + "step": 30375 + }, + { + "epoch": 0.72, + "grad_norm": 2.0214516048493425, + "learning_rate": 3.949682314416374e-06, + "loss": 0.9944, + "step": 30376 + }, + { + "epoch": 0.72, + "grad_norm": 2.1225479955323063, + "learning_rate": 3.9490747977227715e-06, + "loss": 0.9587, + "step": 30377 + }, + { + "epoch": 0.72, + "grad_norm": 2.187735951738982, + "learning_rate": 3.948467316259569e-06, + "loss": 1.0272, + "step": 30378 + }, + { + "epoch": 0.72, + "grad_norm": 2.1056107516741176, + "learning_rate": 3.947859870030304e-06, + "loss": 0.9477, + "step": 30379 + }, + { + "epoch": 0.72, + "grad_norm": 2.1693664369635477, + "learning_rate": 3.947252459038519e-06, + "loss": 0.9564, + "step": 30380 + }, + { + "epoch": 0.72, + "grad_norm": 2.040720490513213, + "learning_rate": 3.946645083287747e-06, + "loss": 1.1318, + "step": 30381 + }, + { + "epoch": 0.72, + "grad_norm": 1.9310535397192932, + "learning_rate": 3.946037742781519e-06, + "loss": 1.0833, + "step": 30382 + }, + { + "epoch": 0.72, + "grad_norm": 2.277248305627987, + "learning_rate": 3.945430437523376e-06, + "loss": 0.9552, + "step": 30383 + }, + { + "epoch": 0.72, + "grad_norm": 2.1817977678135803, + "learning_rate": 3.9448231675168584e-06, + "loss": 0.9136, + "step": 30384 + }, + { + "epoch": 0.72, + "grad_norm": 1.508527213278365, + "learning_rate": 3.944215932765498e-06, + "loss": 0.9114, + "step": 30385 + }, + { + "epoch": 0.72, + "grad_norm": 1.7187994992513937, + "learning_rate": 3.943608733272824e-06, + "loss": 0.9582, + "step": 30386 + }, + { + "epoch": 0.72, + "grad_norm": 1.9887243457623054, + "learning_rate": 3.943001569042378e-06, + "loss": 0.8074, + "step": 30387 + }, + { + "epoch": 0.72, + "grad_norm": 1.9981866577433534, + "learning_rate": 3.942394440077699e-06, + "loss": 0.936, + "step": 30388 + }, + { + "epoch": 0.72, + "grad_norm": 2.0032270942245143, + "learning_rate": 3.9417873463823165e-06, + "loss": 1.0371, + "step": 30389 + }, + { + "epoch": 0.72, + "grad_norm": 1.961409702857139, + "learning_rate": 3.941180287959763e-06, + "loss": 1.0461, + "step": 30390 + }, + { + "epoch": 0.72, + "grad_norm": 1.9347174974571242, + "learning_rate": 3.94057326481358e-06, + "loss": 1.0012, + "step": 30391 + }, + { + "epoch": 0.72, + "grad_norm": 2.1035035331364362, + "learning_rate": 3.939966276947293e-06, + "loss": 0.9224, + "step": 30392 + }, + { + "epoch": 0.72, + "grad_norm": 2.024109118156511, + "learning_rate": 3.939359324364445e-06, + "loss": 0.871, + "step": 30393 + }, + { + "epoch": 0.72, + "grad_norm": 2.519473181654208, + "learning_rate": 3.938752407068562e-06, + "loss": 0.9399, + "step": 30394 + }, + { + "epoch": 0.72, + "grad_norm": 2.1447497537304145, + "learning_rate": 3.9381455250631864e-06, + "loss": 0.9983, + "step": 30395 + }, + { + "epoch": 0.72, + "grad_norm": 1.1129890290519129, + "learning_rate": 3.937538678351842e-06, + "loss": 0.8377, + "step": 30396 + }, + { + "epoch": 0.72, + "grad_norm": 2.2466644535436977, + "learning_rate": 3.936931866938071e-06, + "loss": 1.0456, + "step": 30397 + }, + { + "epoch": 0.72, + "grad_norm": 1.9656712398061436, + "learning_rate": 3.9363250908254015e-06, + "loss": 1.0301, + "step": 30398 + }, + { + "epoch": 0.72, + "grad_norm": 2.2943484988103755, + "learning_rate": 3.9357183500173635e-06, + "loss": 1.0722, + "step": 30399 + }, + { + "epoch": 0.72, + "grad_norm": 2.1981933242818865, + "learning_rate": 3.935111644517493e-06, + "loss": 1.0617, + "step": 30400 + }, + { + "epoch": 0.72, + "grad_norm": 2.0375435375419277, + "learning_rate": 3.934504974329326e-06, + "loss": 0.9976, + "step": 30401 + }, + { + "epoch": 0.72, + "grad_norm": 2.0858716911925446, + "learning_rate": 3.933898339456392e-06, + "loss": 1.0333, + "step": 30402 + }, + { + "epoch": 0.72, + "grad_norm": 1.8998054905196844, + "learning_rate": 3.9332917399022196e-06, + "loss": 1.0506, + "step": 30403 + }, + { + "epoch": 0.72, + "grad_norm": 2.0815199439761654, + "learning_rate": 3.9326851756703425e-06, + "loss": 1.1353, + "step": 30404 + }, + { + "epoch": 0.72, + "grad_norm": 1.9130799392978093, + "learning_rate": 3.932078646764297e-06, + "loss": 0.9653, + "step": 30405 + }, + { + "epoch": 0.72, + "grad_norm": 2.0736765555848238, + "learning_rate": 3.93147215318761e-06, + "loss": 0.9458, + "step": 30406 + }, + { + "epoch": 0.72, + "grad_norm": 2.1507949268214173, + "learning_rate": 3.930865694943809e-06, + "loss": 0.9921, + "step": 30407 + }, + { + "epoch": 0.72, + "grad_norm": 2.205262495787348, + "learning_rate": 3.930259272036432e-06, + "loss": 0.9284, + "step": 30408 + }, + { + "epoch": 0.72, + "grad_norm": 2.036341547386663, + "learning_rate": 3.9296528844690085e-06, + "loss": 1.1123, + "step": 30409 + }, + { + "epoch": 0.72, + "grad_norm": 3.9711777399433723, + "learning_rate": 3.9290465322450685e-06, + "loss": 0.8621, + "step": 30410 + }, + { + "epoch": 0.72, + "grad_norm": 2.169144286596216, + "learning_rate": 3.9284402153681385e-06, + "loss": 1.0626, + "step": 30411 + }, + { + "epoch": 0.72, + "grad_norm": 1.1462578753859736, + "learning_rate": 3.927833933841755e-06, + "loss": 0.9017, + "step": 30412 + }, + { + "epoch": 0.72, + "grad_norm": 2.0145991555573564, + "learning_rate": 3.927227687669442e-06, + "loss": 1.1554, + "step": 30413 + }, + { + "epoch": 0.72, + "grad_norm": 2.090209847541427, + "learning_rate": 3.9266214768547335e-06, + "loss": 0.9982, + "step": 30414 + }, + { + "epoch": 0.72, + "grad_norm": 1.9933021656783692, + "learning_rate": 3.926015301401156e-06, + "loss": 1.0705, + "step": 30415 + }, + { + "epoch": 0.72, + "grad_norm": 2.0869507854436633, + "learning_rate": 3.925409161312243e-06, + "loss": 0.9022, + "step": 30416 + }, + { + "epoch": 0.72, + "grad_norm": 1.8934642273137894, + "learning_rate": 3.924803056591517e-06, + "loss": 0.9513, + "step": 30417 + }, + { + "epoch": 0.72, + "grad_norm": 2.5885612085013805, + "learning_rate": 3.924196987242512e-06, + "loss": 0.8629, + "step": 30418 + }, + { + "epoch": 0.72, + "grad_norm": 1.8546948123858435, + "learning_rate": 3.923590953268762e-06, + "loss": 0.8411, + "step": 30419 + }, + { + "epoch": 0.72, + "grad_norm": 2.046899722697106, + "learning_rate": 3.922984954673783e-06, + "loss": 0.9935, + "step": 30420 + }, + { + "epoch": 0.72, + "grad_norm": 2.1296196036604127, + "learning_rate": 3.92237899146111e-06, + "loss": 0.8473, + "step": 30421 + }, + { + "epoch": 0.72, + "grad_norm": 1.9525933301406742, + "learning_rate": 3.921773063634269e-06, + "loss": 1.0377, + "step": 30422 + }, + { + "epoch": 0.72, + "grad_norm": 1.8365124024899728, + "learning_rate": 3.921167171196798e-06, + "loss": 0.8634, + "step": 30423 + }, + { + "epoch": 0.72, + "grad_norm": 2.0138987777925794, + "learning_rate": 3.920561314152208e-06, + "loss": 0.9551, + "step": 30424 + }, + { + "epoch": 0.72, + "grad_norm": 1.086858292007399, + "learning_rate": 3.919955492504036e-06, + "loss": 0.9757, + "step": 30425 + }, + { + "epoch": 0.72, + "grad_norm": 2.016599819792209, + "learning_rate": 3.91934970625581e-06, + "loss": 1.002, + "step": 30426 + }, + { + "epoch": 0.72, + "grad_norm": 1.8652906602211292, + "learning_rate": 3.918743955411057e-06, + "loss": 0.968, + "step": 30427 + }, + { + "epoch": 0.72, + "grad_norm": 1.9249158201306662, + "learning_rate": 3.918138239973297e-06, + "loss": 0.8706, + "step": 30428 + }, + { + "epoch": 0.72, + "grad_norm": 1.8664068172145794, + "learning_rate": 3.917532559946062e-06, + "loss": 0.98, + "step": 30429 + }, + { + "epoch": 0.72, + "grad_norm": 2.0017273007934495, + "learning_rate": 3.916926915332882e-06, + "loss": 0.9929, + "step": 30430 + }, + { + "epoch": 0.72, + "grad_norm": 2.401058499952754, + "learning_rate": 3.916321306137276e-06, + "loss": 0.9688, + "step": 30431 + }, + { + "epoch": 0.72, + "grad_norm": 2.117062139362831, + "learning_rate": 3.915715732362777e-06, + "loss": 0.9353, + "step": 30432 + }, + { + "epoch": 0.72, + "grad_norm": 2.040575273023466, + "learning_rate": 3.915110194012907e-06, + "loss": 0.9139, + "step": 30433 + }, + { + "epoch": 0.72, + "grad_norm": 1.9981306858793975, + "learning_rate": 3.914504691091188e-06, + "loss": 0.9353, + "step": 30434 + }, + { + "epoch": 0.72, + "grad_norm": 1.8910954223644487, + "learning_rate": 3.913899223601149e-06, + "loss": 1.1237, + "step": 30435 + }, + { + "epoch": 0.72, + "grad_norm": 1.8934262847819092, + "learning_rate": 3.91329379154632e-06, + "loss": 0.8967, + "step": 30436 + }, + { + "epoch": 0.72, + "grad_norm": 1.9681626363319256, + "learning_rate": 3.912688394930221e-06, + "loss": 1.052, + "step": 30437 + }, + { + "epoch": 0.72, + "grad_norm": 2.5089049846611147, + "learning_rate": 3.912083033756374e-06, + "loss": 0.994, + "step": 30438 + }, + { + "epoch": 0.72, + "grad_norm": 1.1803340515730583, + "learning_rate": 3.911477708028308e-06, + "loss": 0.8678, + "step": 30439 + }, + { + "epoch": 0.72, + "grad_norm": 2.0066912905222876, + "learning_rate": 3.910872417749548e-06, + "loss": 0.9589, + "step": 30440 + }, + { + "epoch": 0.72, + "grad_norm": 2.2793387655970316, + "learning_rate": 3.910267162923618e-06, + "loss": 0.8824, + "step": 30441 + }, + { + "epoch": 0.72, + "grad_norm": 2.1393251241173195, + "learning_rate": 3.909661943554036e-06, + "loss": 0.9529, + "step": 30442 + }, + { + "epoch": 0.72, + "grad_norm": 2.0995828926279305, + "learning_rate": 3.909056759644331e-06, + "loss": 0.9081, + "step": 30443 + }, + { + "epoch": 0.72, + "grad_norm": 2.154565312083766, + "learning_rate": 3.9084516111980285e-06, + "loss": 0.9761, + "step": 30444 + }, + { + "epoch": 0.72, + "grad_norm": 2.3197315535186935, + "learning_rate": 3.9078464982186505e-06, + "loss": 1.0448, + "step": 30445 + }, + { + "epoch": 0.72, + "grad_norm": 1.85605798738783, + "learning_rate": 3.907241420709716e-06, + "loss": 0.9466, + "step": 30446 + }, + { + "epoch": 0.72, + "grad_norm": 2.434617407603766, + "learning_rate": 3.906636378674752e-06, + "loss": 0.9469, + "step": 30447 + }, + { + "epoch": 0.72, + "grad_norm": 2.404782141136564, + "learning_rate": 3.906031372117278e-06, + "loss": 0.8844, + "step": 30448 + }, + { + "epoch": 0.72, + "grad_norm": 2.2518971071033596, + "learning_rate": 3.905426401040823e-06, + "loss": 0.8934, + "step": 30449 + }, + { + "epoch": 0.72, + "grad_norm": 2.0625832818115923, + "learning_rate": 3.9048214654489e-06, + "loss": 0.9285, + "step": 30450 + }, + { + "epoch": 0.72, + "grad_norm": 1.9196866243989332, + "learning_rate": 3.90421656534504e-06, + "loss": 1.0791, + "step": 30451 + }, + { + "epoch": 0.72, + "grad_norm": 1.104049474576322, + "learning_rate": 3.903611700732758e-06, + "loss": 0.9001, + "step": 30452 + }, + { + "epoch": 0.72, + "grad_norm": 2.06889118797469, + "learning_rate": 3.903006871615582e-06, + "loss": 0.934, + "step": 30453 + }, + { + "epoch": 0.72, + "grad_norm": 1.9621640949058763, + "learning_rate": 3.9024020779970305e-06, + "loss": 0.7973, + "step": 30454 + }, + { + "epoch": 0.72, + "grad_norm": 1.1521265538454544, + "learning_rate": 3.901797319880622e-06, + "loss": 0.9007, + "step": 30455 + }, + { + "epoch": 0.72, + "grad_norm": 1.9352840955285928, + "learning_rate": 3.901192597269878e-06, + "loss": 0.9561, + "step": 30456 + }, + { + "epoch": 0.72, + "grad_norm": 2.0250101613844644, + "learning_rate": 3.9005879101683266e-06, + "loss": 0.9672, + "step": 30457 + }, + { + "epoch": 0.72, + "grad_norm": 2.108191182598543, + "learning_rate": 3.899983258579483e-06, + "loss": 0.9842, + "step": 30458 + }, + { + "epoch": 0.72, + "grad_norm": 2.0755439209681006, + "learning_rate": 3.899378642506863e-06, + "loss": 1.0287, + "step": 30459 + }, + { + "epoch": 0.72, + "grad_norm": 1.9730201685303932, + "learning_rate": 3.898774061953995e-06, + "loss": 1.0507, + "step": 30460 + }, + { + "epoch": 0.72, + "grad_norm": 2.0825613979142665, + "learning_rate": 3.8981695169243986e-06, + "loss": 1.0887, + "step": 30461 + }, + { + "epoch": 0.72, + "grad_norm": 2.012102558735044, + "learning_rate": 3.89756500742159e-06, + "loss": 1.0119, + "step": 30462 + }, + { + "epoch": 0.72, + "grad_norm": 2.1733484850270104, + "learning_rate": 3.896960533449087e-06, + "loss": 0.9131, + "step": 30463 + }, + { + "epoch": 0.72, + "grad_norm": 1.9354342264014057, + "learning_rate": 3.896356095010412e-06, + "loss": 0.9192, + "step": 30464 + }, + { + "epoch": 0.72, + "grad_norm": 2.0368176710992065, + "learning_rate": 3.895751692109087e-06, + "loss": 0.9541, + "step": 30465 + }, + { + "epoch": 0.72, + "grad_norm": 2.0466249921023607, + "learning_rate": 3.895147324748629e-06, + "loss": 1.0173, + "step": 30466 + }, + { + "epoch": 0.72, + "grad_norm": 2.154487012234993, + "learning_rate": 3.894542992932553e-06, + "loss": 0.9458, + "step": 30467 + }, + { + "epoch": 0.72, + "grad_norm": 1.876981696061239, + "learning_rate": 3.893938696664383e-06, + "loss": 1.0692, + "step": 30468 + }, + { + "epoch": 0.72, + "grad_norm": 2.6335974265731794, + "learning_rate": 3.893334435947632e-06, + "loss": 0.8821, + "step": 30469 + }, + { + "epoch": 0.72, + "grad_norm": 2.105717069317638, + "learning_rate": 3.892730210785825e-06, + "loss": 0.9369, + "step": 30470 + }, + { + "epoch": 0.72, + "grad_norm": 2.137055538081934, + "learning_rate": 3.892126021182473e-06, + "loss": 0.7939, + "step": 30471 + }, + { + "epoch": 0.72, + "grad_norm": 1.9678002453863985, + "learning_rate": 3.891521867141099e-06, + "loss": 0.9283, + "step": 30472 + }, + { + "epoch": 0.72, + "grad_norm": 2.0570513403706827, + "learning_rate": 3.890917748665217e-06, + "loss": 0.9636, + "step": 30473 + }, + { + "epoch": 0.72, + "grad_norm": 2.090444404969385, + "learning_rate": 3.890313665758348e-06, + "loss": 1.0442, + "step": 30474 + }, + { + "epoch": 0.72, + "grad_norm": 2.5152779823502156, + "learning_rate": 3.889709618424008e-06, + "loss": 0.9086, + "step": 30475 + }, + { + "epoch": 0.72, + "grad_norm": 1.9601796969681584, + "learning_rate": 3.889105606665709e-06, + "loss": 0.8696, + "step": 30476 + }, + { + "epoch": 0.72, + "grad_norm": 1.099914164040607, + "learning_rate": 3.888501630486973e-06, + "loss": 0.9042, + "step": 30477 + }, + { + "epoch": 0.72, + "grad_norm": 3.2336984147881305, + "learning_rate": 3.887897689891318e-06, + "loss": 0.8358, + "step": 30478 + }, + { + "epoch": 0.72, + "grad_norm": 2.1017936294495017, + "learning_rate": 3.887293784882258e-06, + "loss": 0.9848, + "step": 30479 + }, + { + "epoch": 0.72, + "grad_norm": 2.075822286362624, + "learning_rate": 3.886689915463305e-06, + "loss": 1.1779, + "step": 30480 + }, + { + "epoch": 0.72, + "grad_norm": 2.2484952864596544, + "learning_rate": 3.886086081637979e-06, + "loss": 0.9877, + "step": 30481 + }, + { + "epoch": 0.72, + "grad_norm": 2.3879590613333614, + "learning_rate": 3.8854822834097985e-06, + "loss": 1.0379, + "step": 30482 + }, + { + "epoch": 0.72, + "grad_norm": 1.073168079717708, + "learning_rate": 3.884878520782277e-06, + "loss": 0.9567, + "step": 30483 + }, + { + "epoch": 0.72, + "grad_norm": 1.846309769934094, + "learning_rate": 3.884274793758924e-06, + "loss": 0.9986, + "step": 30484 + }, + { + "epoch": 0.72, + "grad_norm": 2.0976641524288566, + "learning_rate": 3.88367110234326e-06, + "loss": 1.0861, + "step": 30485 + }, + { + "epoch": 0.72, + "grad_norm": 2.228980692045866, + "learning_rate": 3.883067446538803e-06, + "loss": 1.1456, + "step": 30486 + }, + { + "epoch": 0.72, + "grad_norm": 2.019920916451961, + "learning_rate": 3.8824638263490635e-06, + "loss": 1.0693, + "step": 30487 + }, + { + "epoch": 0.72, + "grad_norm": 1.99007783534794, + "learning_rate": 3.881860241777553e-06, + "loss": 1.0308, + "step": 30488 + }, + { + "epoch": 0.72, + "grad_norm": 2.296127099446879, + "learning_rate": 3.88125669282779e-06, + "loss": 0.9932, + "step": 30489 + }, + { + "epoch": 0.72, + "grad_norm": 1.9735663992444399, + "learning_rate": 3.88065317950329e-06, + "loss": 0.9899, + "step": 30490 + }, + { + "epoch": 0.72, + "grad_norm": 1.9572580223031384, + "learning_rate": 3.8800497018075645e-06, + "loss": 0.8561, + "step": 30491 + }, + { + "epoch": 0.72, + "grad_norm": 2.0371178188404997, + "learning_rate": 3.879446259744124e-06, + "loss": 1.0442, + "step": 30492 + }, + { + "epoch": 0.72, + "grad_norm": 1.8923503383854559, + "learning_rate": 3.878842853316489e-06, + "loss": 0.8926, + "step": 30493 + }, + { + "epoch": 0.72, + "grad_norm": 2.1541582757464095, + "learning_rate": 3.878239482528164e-06, + "loss": 1.008, + "step": 30494 + }, + { + "epoch": 0.72, + "grad_norm": 2.0004125114730456, + "learning_rate": 3.877636147382672e-06, + "loss": 0.9653, + "step": 30495 + }, + { + "epoch": 0.72, + "grad_norm": 2.0379998177930236, + "learning_rate": 3.877032847883516e-06, + "loss": 1.0353, + "step": 30496 + }, + { + "epoch": 0.72, + "grad_norm": 1.7641987100860381, + "learning_rate": 3.876429584034218e-06, + "loss": 0.9761, + "step": 30497 + }, + { + "epoch": 0.72, + "grad_norm": 1.8751069466195982, + "learning_rate": 3.875826355838282e-06, + "loss": 1.0838, + "step": 30498 + }, + { + "epoch": 0.72, + "grad_norm": 1.969340170012824, + "learning_rate": 3.8752231632992254e-06, + "loss": 0.9782, + "step": 30499 + }, + { + "epoch": 0.72, + "grad_norm": 1.9443734104084243, + "learning_rate": 3.87462000642056e-06, + "loss": 0.9896, + "step": 30500 + }, + { + "epoch": 0.72, + "grad_norm": 2.065031582260817, + "learning_rate": 3.874016885205793e-06, + "loss": 1.1395, + "step": 30501 + }, + { + "epoch": 0.72, + "grad_norm": 2.138832973592018, + "learning_rate": 3.873413799658438e-06, + "loss": 1.0586, + "step": 30502 + }, + { + "epoch": 0.72, + "grad_norm": 2.886509984240151, + "learning_rate": 3.872810749782009e-06, + "loss": 1.041, + "step": 30503 + }, + { + "epoch": 0.72, + "grad_norm": 1.9866388830013746, + "learning_rate": 3.8722077355800225e-06, + "loss": 1.0422, + "step": 30504 + }, + { + "epoch": 0.72, + "grad_norm": 1.9536582409586651, + "learning_rate": 3.871604757055974e-06, + "loss": 1.0163, + "step": 30505 + }, + { + "epoch": 0.72, + "grad_norm": 1.9229537697921708, + "learning_rate": 3.8710018142133844e-06, + "loss": 0.9035, + "step": 30506 + }, + { + "epoch": 0.72, + "grad_norm": 1.9164683760246985, + "learning_rate": 3.870398907055766e-06, + "loss": 1.0781, + "step": 30507 + }, + { + "epoch": 0.72, + "grad_norm": 2.0811305260500266, + "learning_rate": 3.869796035586625e-06, + "loss": 0.9838, + "step": 30508 + }, + { + "epoch": 0.72, + "grad_norm": 2.0704579493424675, + "learning_rate": 3.869193199809469e-06, + "loss": 1.027, + "step": 30509 + }, + { + "epoch": 0.72, + "grad_norm": 2.320299047774677, + "learning_rate": 3.8685903997278115e-06, + "loss": 0.9116, + "step": 30510 + }, + { + "epoch": 0.72, + "grad_norm": 2.001411353922984, + "learning_rate": 3.867987635345165e-06, + "loss": 1.001, + "step": 30511 + }, + { + "epoch": 0.72, + "grad_norm": 2.7479928525808046, + "learning_rate": 3.867384906665036e-06, + "loss": 0.8504, + "step": 30512 + }, + { + "epoch": 0.72, + "grad_norm": 2.0625078084194595, + "learning_rate": 3.8667822136909304e-06, + "loss": 0.8919, + "step": 30513 + }, + { + "epoch": 0.72, + "grad_norm": 1.9729516834983238, + "learning_rate": 3.866179556426364e-06, + "loss": 1.0584, + "step": 30514 + }, + { + "epoch": 0.72, + "grad_norm": 1.7227555073906407, + "learning_rate": 3.865576934874839e-06, + "loss": 0.918, + "step": 30515 + }, + { + "epoch": 0.72, + "grad_norm": 1.8085757560663918, + "learning_rate": 3.864974349039866e-06, + "loss": 0.8645, + "step": 30516 + }, + { + "epoch": 0.72, + "grad_norm": 3.1580882783631465, + "learning_rate": 3.8643717989249585e-06, + "loss": 1.0951, + "step": 30517 + }, + { + "epoch": 0.72, + "grad_norm": 2.117785261034184, + "learning_rate": 3.863769284533622e-06, + "loss": 1.084, + "step": 30518 + }, + { + "epoch": 0.72, + "grad_norm": 2.096841356381361, + "learning_rate": 3.8631668058693585e-06, + "loss": 0.9768, + "step": 30519 + }, + { + "epoch": 0.72, + "grad_norm": 2.1376068211063224, + "learning_rate": 3.862564362935681e-06, + "loss": 1.1054, + "step": 30520 + }, + { + "epoch": 0.72, + "grad_norm": 2.143861158920385, + "learning_rate": 3.8619619557361044e-06, + "loss": 0.9195, + "step": 30521 + }, + { + "epoch": 0.72, + "grad_norm": 2.2214353963665525, + "learning_rate": 3.861359584274122e-06, + "loss": 0.9988, + "step": 30522 + }, + { + "epoch": 0.72, + "grad_norm": 2.0558447821818775, + "learning_rate": 3.860757248553247e-06, + "loss": 1.1497, + "step": 30523 + }, + { + "epoch": 0.72, + "grad_norm": 1.934092230434053, + "learning_rate": 3.8601549485769865e-06, + "loss": 0.9798, + "step": 30524 + }, + { + "epoch": 0.72, + "grad_norm": 1.8649315792986803, + "learning_rate": 3.8595526843488515e-06, + "loss": 1.0379, + "step": 30525 + }, + { + "epoch": 0.72, + "grad_norm": 2.214041354518228, + "learning_rate": 3.858950455872344e-06, + "loss": 1.1163, + "step": 30526 + }, + { + "epoch": 0.72, + "grad_norm": 1.109380769357824, + "learning_rate": 3.8583482631509695e-06, + "loss": 1.0154, + "step": 30527 + }, + { + "epoch": 0.72, + "grad_norm": 1.8481494756730732, + "learning_rate": 3.857746106188238e-06, + "loss": 0.9103, + "step": 30528 + }, + { + "epoch": 0.72, + "grad_norm": 1.8501692796053768, + "learning_rate": 3.85714398498765e-06, + "loss": 1.0397, + "step": 30529 + }, + { + "epoch": 0.72, + "grad_norm": 1.8188243116890426, + "learning_rate": 3.8565418995527185e-06, + "loss": 0.9958, + "step": 30530 + }, + { + "epoch": 0.72, + "grad_norm": 2.0486021026293826, + "learning_rate": 3.8559398498869416e-06, + "loss": 1.0177, + "step": 30531 + }, + { + "epoch": 0.72, + "grad_norm": 2.5122037009874223, + "learning_rate": 3.855337835993831e-06, + "loss": 1.0476, + "step": 30532 + }, + { + "epoch": 0.72, + "grad_norm": 2.0152669526691036, + "learning_rate": 3.8547358578768854e-06, + "loss": 1.0399, + "step": 30533 + }, + { + "epoch": 0.72, + "grad_norm": 1.9965903000253395, + "learning_rate": 3.854133915539617e-06, + "loss": 0.8323, + "step": 30534 + }, + { + "epoch": 0.72, + "grad_norm": 1.895151237917863, + "learning_rate": 3.853532008985527e-06, + "loss": 1.0032, + "step": 30535 + }, + { + "epoch": 0.72, + "grad_norm": 2.2626570093101144, + "learning_rate": 3.852930138218115e-06, + "loss": 1.0164, + "step": 30536 + }, + { + "epoch": 0.72, + "grad_norm": 2.3751715205827937, + "learning_rate": 3.852328303240891e-06, + "loss": 0.9397, + "step": 30537 + }, + { + "epoch": 0.72, + "grad_norm": 1.1293837015520867, + "learning_rate": 3.8517265040573606e-06, + "loss": 0.9523, + "step": 30538 + }, + { + "epoch": 0.72, + "grad_norm": 1.954563948569104, + "learning_rate": 3.851124740671026e-06, + "loss": 1.0809, + "step": 30539 + }, + { + "epoch": 0.72, + "grad_norm": 2.0103303557956003, + "learning_rate": 3.850523013085385e-06, + "loss": 0.9688, + "step": 30540 + }, + { + "epoch": 0.72, + "grad_norm": 2.8404333093042764, + "learning_rate": 3.849921321303947e-06, + "loss": 0.9986, + "step": 30541 + }, + { + "epoch": 0.72, + "grad_norm": 2.352141129613736, + "learning_rate": 3.849319665330218e-06, + "loss": 0.9284, + "step": 30542 + }, + { + "epoch": 0.72, + "grad_norm": 1.791984720630317, + "learning_rate": 3.848718045167695e-06, + "loss": 0.993, + "step": 30543 + }, + { + "epoch": 0.72, + "grad_norm": 2.120437204543949, + "learning_rate": 3.848116460819882e-06, + "loss": 0.9498, + "step": 30544 + }, + { + "epoch": 0.72, + "grad_norm": 1.830420124342842, + "learning_rate": 3.8475149122902815e-06, + "loss": 0.9181, + "step": 30545 + }, + { + "epoch": 0.72, + "grad_norm": 1.945209121733988, + "learning_rate": 3.846913399582401e-06, + "loss": 0.7739, + "step": 30546 + }, + { + "epoch": 0.72, + "grad_norm": 1.9243096504199309, + "learning_rate": 3.846311922699738e-06, + "loss": 0.9442, + "step": 30547 + }, + { + "epoch": 0.72, + "grad_norm": 1.8834947342654351, + "learning_rate": 3.845710481645793e-06, + "loss": 0.8709, + "step": 30548 + }, + { + "epoch": 0.72, + "grad_norm": 2.15062214133907, + "learning_rate": 3.845109076424071e-06, + "loss": 1.0129, + "step": 30549 + }, + { + "epoch": 0.72, + "grad_norm": 1.9011414562395585, + "learning_rate": 3.844507707038071e-06, + "loss": 1.0126, + "step": 30550 + }, + { + "epoch": 0.72, + "grad_norm": 1.9757235447764998, + "learning_rate": 3.843906373491299e-06, + "loss": 1.0016, + "step": 30551 + }, + { + "epoch": 0.72, + "grad_norm": 1.8962646084738173, + "learning_rate": 3.843305075787249e-06, + "loss": 1.0454, + "step": 30552 + }, + { + "epoch": 0.72, + "grad_norm": 2.504523190863726, + "learning_rate": 3.8427038139294304e-06, + "loss": 1.0196, + "step": 30553 + }, + { + "epoch": 0.72, + "grad_norm": 1.9586121090384965, + "learning_rate": 3.842102587921336e-06, + "loss": 1.018, + "step": 30554 + }, + { + "epoch": 0.72, + "grad_norm": 1.9522639419806855, + "learning_rate": 3.841501397766474e-06, + "loss": 0.7912, + "step": 30555 + }, + { + "epoch": 0.72, + "grad_norm": 2.130075610288062, + "learning_rate": 3.84090024346834e-06, + "loss": 1.0524, + "step": 30556 + }, + { + "epoch": 0.72, + "grad_norm": 1.9673650775760685, + "learning_rate": 3.84029912503043e-06, + "loss": 1.1957, + "step": 30557 + }, + { + "epoch": 0.72, + "grad_norm": 2.128358378476443, + "learning_rate": 3.83969804245625e-06, + "loss": 0.9591, + "step": 30558 + }, + { + "epoch": 0.72, + "grad_norm": 1.936746205447953, + "learning_rate": 3.839096995749302e-06, + "loss": 1.0808, + "step": 30559 + }, + { + "epoch": 0.72, + "grad_norm": 2.0315805944451624, + "learning_rate": 3.838495984913081e-06, + "loss": 0.9555, + "step": 30560 + }, + { + "epoch": 0.72, + "grad_norm": 1.906447536413277, + "learning_rate": 3.8378950099510835e-06, + "loss": 0.8366, + "step": 30561 + }, + { + "epoch": 0.72, + "grad_norm": 1.134406336672214, + "learning_rate": 3.837294070866812e-06, + "loss": 0.9319, + "step": 30562 + }, + { + "epoch": 0.72, + "grad_norm": 2.14380456308741, + "learning_rate": 3.83669316766377e-06, + "loss": 0.987, + "step": 30563 + }, + { + "epoch": 0.72, + "grad_norm": 1.9517100726003362, + "learning_rate": 3.8360923003454505e-06, + "loss": 0.8544, + "step": 30564 + }, + { + "epoch": 0.72, + "grad_norm": 2.3875680314212, + "learning_rate": 3.835491468915351e-06, + "loss": 1.0197, + "step": 30565 + }, + { + "epoch": 0.72, + "grad_norm": 2.098450855734021, + "learning_rate": 3.834890673376971e-06, + "loss": 1.0301, + "step": 30566 + }, + { + "epoch": 0.72, + "grad_norm": 2.036501399075822, + "learning_rate": 3.834289913733813e-06, + "loss": 0.8873, + "step": 30567 + }, + { + "epoch": 0.72, + "grad_norm": 2.3273823885032847, + "learning_rate": 3.83368918998937e-06, + "loss": 1.0423, + "step": 30568 + }, + { + "epoch": 0.72, + "grad_norm": 1.8065581989816577, + "learning_rate": 3.833088502147139e-06, + "loss": 1.0166, + "step": 30569 + }, + { + "epoch": 0.72, + "grad_norm": 1.0561129350580358, + "learning_rate": 3.832487850210622e-06, + "loss": 0.953, + "step": 30570 + }, + { + "epoch": 0.72, + "grad_norm": 1.881374464381141, + "learning_rate": 3.83188723418331e-06, + "loss": 1.0003, + "step": 30571 + }, + { + "epoch": 0.72, + "grad_norm": 1.8323624238876741, + "learning_rate": 3.831286654068706e-06, + "loss": 0.8772, + "step": 30572 + }, + { + "epoch": 0.72, + "grad_norm": 1.1532520113269684, + "learning_rate": 3.830686109870302e-06, + "loss": 0.906, + "step": 30573 + }, + { + "epoch": 0.72, + "grad_norm": 2.061884572067638, + "learning_rate": 3.830085601591599e-06, + "loss": 0.9688, + "step": 30574 + }, + { + "epoch": 0.72, + "grad_norm": 1.9211153251943924, + "learning_rate": 3.829485129236088e-06, + "loss": 0.9267, + "step": 30575 + }, + { + "epoch": 0.72, + "grad_norm": 2.149208278458897, + "learning_rate": 3.8288846928072705e-06, + "loss": 0.9069, + "step": 30576 + }, + { + "epoch": 0.72, + "grad_norm": 2.0601213751469536, + "learning_rate": 3.8282842923086415e-06, + "loss": 1.0169, + "step": 30577 + }, + { + "epoch": 0.72, + "grad_norm": 2.904823688027729, + "learning_rate": 3.827683927743692e-06, + "loss": 1.0401, + "step": 30578 + }, + { + "epoch": 0.72, + "grad_norm": 1.944949930535151, + "learning_rate": 3.8270835991159195e-06, + "loss": 0.8145, + "step": 30579 + }, + { + "epoch": 0.72, + "grad_norm": 1.932112506573199, + "learning_rate": 3.826483306428824e-06, + "loss": 1.095, + "step": 30580 + }, + { + "epoch": 0.72, + "grad_norm": 1.9406347442407652, + "learning_rate": 3.8258830496858975e-06, + "loss": 1.0237, + "step": 30581 + }, + { + "epoch": 0.72, + "grad_norm": 1.8501537764517981, + "learning_rate": 3.825282828890632e-06, + "loss": 1.0514, + "step": 30582 + }, + { + "epoch": 0.72, + "grad_norm": 2.38496376237873, + "learning_rate": 3.824682644046524e-06, + "loss": 0.9876, + "step": 30583 + }, + { + "epoch": 0.72, + "grad_norm": 1.8663549149346732, + "learning_rate": 3.824082495157073e-06, + "loss": 0.987, + "step": 30584 + }, + { + "epoch": 0.72, + "grad_norm": 2.1306972598244216, + "learning_rate": 3.823482382225768e-06, + "loss": 1.0143, + "step": 30585 + }, + { + "epoch": 0.72, + "grad_norm": 1.984602486391189, + "learning_rate": 3.822882305256101e-06, + "loss": 0.8769, + "step": 30586 + }, + { + "epoch": 0.72, + "grad_norm": 1.932605881920545, + "learning_rate": 3.822282264251569e-06, + "loss": 1.007, + "step": 30587 + }, + { + "epoch": 0.72, + "grad_norm": 1.8514550659241016, + "learning_rate": 3.8216822592156685e-06, + "loss": 0.8665, + "step": 30588 + }, + { + "epoch": 0.72, + "grad_norm": 2.0509533602155274, + "learning_rate": 3.8210822901518904e-06, + "loss": 0.9372, + "step": 30589 + }, + { + "epoch": 0.72, + "grad_norm": 2.0128458869984507, + "learning_rate": 3.820482357063724e-06, + "loss": 0.9631, + "step": 30590 + }, + { + "epoch": 0.72, + "grad_norm": 2.0226858313087375, + "learning_rate": 3.819882459954668e-06, + "loss": 1.0387, + "step": 30591 + }, + { + "epoch": 0.72, + "grad_norm": 1.8851595836233281, + "learning_rate": 3.819282598828211e-06, + "loss": 1.0098, + "step": 30592 + }, + { + "epoch": 0.72, + "grad_norm": 1.9168018242538105, + "learning_rate": 3.8186827736878505e-06, + "loss": 0.9166, + "step": 30593 + }, + { + "epoch": 0.72, + "grad_norm": 1.9419382821550688, + "learning_rate": 3.8180829845370714e-06, + "loss": 1.0765, + "step": 30594 + }, + { + "epoch": 0.72, + "grad_norm": 1.9298161911861398, + "learning_rate": 3.817483231379375e-06, + "loss": 1.0026, + "step": 30595 + }, + { + "epoch": 0.72, + "grad_norm": 1.9124142982780536, + "learning_rate": 3.816883514218245e-06, + "loss": 1.0496, + "step": 30596 + }, + { + "epoch": 0.72, + "grad_norm": 1.877759916871037, + "learning_rate": 3.816283833057179e-06, + "loss": 1.0104, + "step": 30597 + }, + { + "epoch": 0.72, + "grad_norm": 1.0559966436183206, + "learning_rate": 3.815684187899668e-06, + "loss": 0.9812, + "step": 30598 + }, + { + "epoch": 0.72, + "grad_norm": 2.007460308866875, + "learning_rate": 3.815084578749197e-06, + "loss": 0.9054, + "step": 30599 + }, + { + "epoch": 0.72, + "grad_norm": 2.297211930684162, + "learning_rate": 3.814485005609262e-06, + "loss": 1.0048, + "step": 30600 + }, + { + "epoch": 0.72, + "grad_norm": 1.9949298789155288, + "learning_rate": 3.8138854684833537e-06, + "loss": 1.252, + "step": 30601 + }, + { + "epoch": 0.72, + "grad_norm": 2.0479573533788527, + "learning_rate": 3.8132859673749688e-06, + "loss": 0.9348, + "step": 30602 + }, + { + "epoch": 0.72, + "grad_norm": 1.9304927582060316, + "learning_rate": 3.8126865022875858e-06, + "loss": 0.9569, + "step": 30603 + }, + { + "epoch": 0.72, + "grad_norm": 2.005063633979731, + "learning_rate": 3.812087073224701e-06, + "loss": 0.883, + "step": 30604 + }, + { + "epoch": 0.72, + "grad_norm": 2.5511197483319523, + "learning_rate": 3.8114876801898036e-06, + "loss": 0.9843, + "step": 30605 + }, + { + "epoch": 0.72, + "grad_norm": 2.1783384937615127, + "learning_rate": 3.810888323186391e-06, + "loss": 0.9843, + "step": 30606 + }, + { + "epoch": 0.72, + "grad_norm": 1.8838342506833494, + "learning_rate": 3.8102890022179393e-06, + "loss": 0.8563, + "step": 30607 + }, + { + "epoch": 0.72, + "grad_norm": 1.987026175986644, + "learning_rate": 3.809689717287944e-06, + "loss": 0.9921, + "step": 30608 + }, + { + "epoch": 0.72, + "grad_norm": 1.1185855656144057, + "learning_rate": 3.8090904683998997e-06, + "loss": 0.9555, + "step": 30609 + }, + { + "epoch": 0.72, + "grad_norm": 1.787747570314614, + "learning_rate": 3.8084912555572907e-06, + "loss": 0.9541, + "step": 30610 + }, + { + "epoch": 0.72, + "grad_norm": 2.0231765786882834, + "learning_rate": 3.807892078763602e-06, + "loss": 1.001, + "step": 30611 + }, + { + "epoch": 0.72, + "grad_norm": 1.8626301843053628, + "learning_rate": 3.8072929380223256e-06, + "loss": 1.0836, + "step": 30612 + }, + { + "epoch": 0.72, + "grad_norm": 2.337887690378162, + "learning_rate": 3.8066938333369542e-06, + "loss": 1.0231, + "step": 30613 + }, + { + "epoch": 0.72, + "grad_norm": 2.31076613355884, + "learning_rate": 3.806094764710968e-06, + "loss": 1.0152, + "step": 30614 + }, + { + "epoch": 0.72, + "grad_norm": 1.9208917511017347, + "learning_rate": 3.8054957321478623e-06, + "loss": 0.9738, + "step": 30615 + }, + { + "epoch": 0.72, + "grad_norm": 2.870054092156625, + "learning_rate": 3.8048967356511234e-06, + "loss": 1.0387, + "step": 30616 + }, + { + "epoch": 0.72, + "grad_norm": 2.1355038506845605, + "learning_rate": 3.8042977752242315e-06, + "loss": 1.0372, + "step": 30617 + }, + { + "epoch": 0.72, + "grad_norm": 1.9497940781214023, + "learning_rate": 3.803698850870681e-06, + "loss": 1.0869, + "step": 30618 + }, + { + "epoch": 0.72, + "grad_norm": 1.1717179904501112, + "learning_rate": 3.8030999625939604e-06, + "loss": 0.9029, + "step": 30619 + }, + { + "epoch": 0.72, + "grad_norm": 2.1526357794262374, + "learning_rate": 3.802501110397553e-06, + "loss": 1.0305, + "step": 30620 + }, + { + "epoch": 0.72, + "grad_norm": 2.03109649166407, + "learning_rate": 3.8019022942849436e-06, + "loss": 0.9283, + "step": 30621 + }, + { + "epoch": 0.72, + "grad_norm": 1.8656273501537526, + "learning_rate": 3.8013035142596213e-06, + "loss": 0.9598, + "step": 30622 + }, + { + "epoch": 0.72, + "grad_norm": 2.1687748930928854, + "learning_rate": 3.8007047703250756e-06, + "loss": 0.9501, + "step": 30623 + }, + { + "epoch": 0.72, + "grad_norm": 2.0753441673073025, + "learning_rate": 3.8001060624847896e-06, + "loss": 1.0384, + "step": 30624 + }, + { + "epoch": 0.72, + "grad_norm": 1.790824986590163, + "learning_rate": 3.7995073907422454e-06, + "loss": 1.0909, + "step": 30625 + }, + { + "epoch": 0.72, + "grad_norm": 1.9290586428547323, + "learning_rate": 3.798908755100932e-06, + "loss": 0.902, + "step": 30626 + }, + { + "epoch": 0.72, + "grad_norm": 1.9397304338820682, + "learning_rate": 3.798310155564339e-06, + "loss": 0.8655, + "step": 30627 + }, + { + "epoch": 0.72, + "grad_norm": 1.8627480954581195, + "learning_rate": 3.7977115921359477e-06, + "loss": 0.9184, + "step": 30628 + }, + { + "epoch": 0.72, + "grad_norm": 1.8808915289242778, + "learning_rate": 3.797113064819239e-06, + "loss": 0.952, + "step": 30629 + }, + { + "epoch": 0.72, + "grad_norm": 2.383048542440009, + "learning_rate": 3.7965145736177058e-06, + "loss": 1.0266, + "step": 30630 + }, + { + "epoch": 0.72, + "grad_norm": 2.082837264306079, + "learning_rate": 3.795916118534826e-06, + "loss": 1.054, + "step": 30631 + }, + { + "epoch": 0.72, + "grad_norm": 2.0813954641331742, + "learning_rate": 3.795317699574089e-06, + "loss": 0.9352, + "step": 30632 + }, + { + "epoch": 0.72, + "grad_norm": 1.903382636224709, + "learning_rate": 3.7947193167389728e-06, + "loss": 0.9892, + "step": 30633 + }, + { + "epoch": 0.72, + "grad_norm": 2.261597306753885, + "learning_rate": 3.7941209700329706e-06, + "loss": 0.8879, + "step": 30634 + }, + { + "epoch": 0.72, + "grad_norm": 2.0275472809171453, + "learning_rate": 3.7935226594595564e-06, + "loss": 0.948, + "step": 30635 + }, + { + "epoch": 0.72, + "grad_norm": 1.9634838988107222, + "learning_rate": 3.7929243850222206e-06, + "loss": 0.9705, + "step": 30636 + }, + { + "epoch": 0.72, + "grad_norm": 1.2252410158240565, + "learning_rate": 3.7923261467244455e-06, + "loss": 1.0022, + "step": 30637 + }, + { + "epoch": 0.72, + "grad_norm": 4.653623936310346, + "learning_rate": 3.79172794456971e-06, + "loss": 0.8623, + "step": 30638 + }, + { + "epoch": 0.72, + "grad_norm": 1.9377257009914288, + "learning_rate": 3.791129778561499e-06, + "loss": 1.0702, + "step": 30639 + }, + { + "epoch": 0.72, + "grad_norm": 1.9303411264091328, + "learning_rate": 3.7905316487032995e-06, + "loss": 1.0316, + "step": 30640 + }, + { + "epoch": 0.72, + "grad_norm": 1.9102677402207007, + "learning_rate": 3.7899335549985903e-06, + "loss": 0.8807, + "step": 30641 + }, + { + "epoch": 0.72, + "grad_norm": 2.3558396180696537, + "learning_rate": 3.7893354974508513e-06, + "loss": 1.1184, + "step": 30642 + }, + { + "epoch": 0.72, + "grad_norm": 2.380986270053789, + "learning_rate": 3.7887374760635675e-06, + "loss": 1.0279, + "step": 30643 + }, + { + "epoch": 0.72, + "grad_norm": 1.1100348543677736, + "learning_rate": 3.7881394908402237e-06, + "loss": 0.825, + "step": 30644 + }, + { + "epoch": 0.72, + "grad_norm": 2.037074986087098, + "learning_rate": 3.7875415417842974e-06, + "loss": 0.9581, + "step": 30645 + }, + { + "epoch": 0.72, + "grad_norm": 2.0407138077697, + "learning_rate": 3.786943628899269e-06, + "loss": 0.9684, + "step": 30646 + }, + { + "epoch": 0.72, + "grad_norm": 1.0559248133549133, + "learning_rate": 3.7863457521886215e-06, + "loss": 0.8811, + "step": 30647 + }, + { + "epoch": 0.72, + "grad_norm": 1.875749433582564, + "learning_rate": 3.7857479116558393e-06, + "loss": 0.9367, + "step": 30648 + }, + { + "epoch": 0.72, + "grad_norm": 1.9514920271658311, + "learning_rate": 3.7851501073044005e-06, + "loss": 0.9503, + "step": 30649 + }, + { + "epoch": 0.72, + "grad_norm": 1.9805202878316204, + "learning_rate": 3.7845523391377815e-06, + "loss": 0.9756, + "step": 30650 + }, + { + "epoch": 0.72, + "grad_norm": 1.9541016059896454, + "learning_rate": 3.7839546071594703e-06, + "loss": 0.979, + "step": 30651 + }, + { + "epoch": 0.72, + "grad_norm": 1.8476741445382534, + "learning_rate": 3.7833569113729394e-06, + "loss": 1.0233, + "step": 30652 + }, + { + "epoch": 0.72, + "grad_norm": 1.9614876642880272, + "learning_rate": 3.782759251781677e-06, + "loss": 1.0767, + "step": 30653 + }, + { + "epoch": 0.72, + "grad_norm": 1.825056208008692, + "learning_rate": 3.782161628389155e-06, + "loss": 0.8856, + "step": 30654 + }, + { + "epoch": 0.72, + "grad_norm": 1.9001081162105138, + "learning_rate": 3.78156404119886e-06, + "loss": 1.0435, + "step": 30655 + }, + { + "epoch": 0.72, + "grad_norm": 2.118054787906762, + "learning_rate": 3.780966490214264e-06, + "loss": 1.0393, + "step": 30656 + }, + { + "epoch": 0.72, + "grad_norm": 1.7967739856874567, + "learning_rate": 3.7803689754388538e-06, + "loss": 1.0745, + "step": 30657 + }, + { + "epoch": 0.72, + "grad_norm": 2.0302495184366007, + "learning_rate": 3.7797714968761045e-06, + "loss": 1.0366, + "step": 30658 + }, + { + "epoch": 0.72, + "grad_norm": 1.8725935670354494, + "learning_rate": 3.7791740545294907e-06, + "loss": 0.8643, + "step": 30659 + }, + { + "epoch": 0.72, + "grad_norm": 3.632678810666862, + "learning_rate": 3.7785766484024954e-06, + "loss": 0.9251, + "step": 30660 + }, + { + "epoch": 0.72, + "grad_norm": 1.9057847405728363, + "learning_rate": 3.7779792784986e-06, + "loss": 0.9687, + "step": 30661 + }, + { + "epoch": 0.72, + "grad_norm": 1.102743553249433, + "learning_rate": 3.77738194482128e-06, + "loss": 0.973, + "step": 30662 + }, + { + "epoch": 0.72, + "grad_norm": 1.794123543792556, + "learning_rate": 3.7767846473740077e-06, + "loss": 0.9169, + "step": 30663 + }, + { + "epoch": 0.72, + "grad_norm": 1.9978660142613422, + "learning_rate": 3.776187386160266e-06, + "loss": 0.8781, + "step": 30664 + }, + { + "epoch": 0.72, + "grad_norm": 2.197069517179222, + "learning_rate": 3.7755901611835356e-06, + "loss": 1.0142, + "step": 30665 + }, + { + "epoch": 0.72, + "grad_norm": 1.0618873822663504, + "learning_rate": 3.774992972447289e-06, + "loss": 0.9996, + "step": 30666 + }, + { + "epoch": 0.72, + "grad_norm": 1.929841356500096, + "learning_rate": 3.7743958199550013e-06, + "loss": 0.8878, + "step": 30667 + }, + { + "epoch": 0.72, + "grad_norm": 1.933474357126741, + "learning_rate": 3.773798703710152e-06, + "loss": 0.9675, + "step": 30668 + }, + { + "epoch": 0.72, + "grad_norm": 2.195743493878186, + "learning_rate": 3.7732016237162216e-06, + "loss": 0.956, + "step": 30669 + }, + { + "epoch": 0.72, + "grad_norm": 2.528303843330769, + "learning_rate": 3.772604579976682e-06, + "loss": 0.9943, + "step": 30670 + }, + { + "epoch": 0.72, + "grad_norm": 2.3966603064431498, + "learning_rate": 3.772007572495007e-06, + "loss": 1.0078, + "step": 30671 + }, + { + "epoch": 0.72, + "grad_norm": 1.1357734984184407, + "learning_rate": 3.7714106012746786e-06, + "loss": 0.9587, + "step": 30672 + }, + { + "epoch": 0.72, + "grad_norm": 2.114822514744542, + "learning_rate": 3.7708136663191673e-06, + "loss": 1.1616, + "step": 30673 + }, + { + "epoch": 0.72, + "grad_norm": 1.9710757452185559, + "learning_rate": 3.7702167676319534e-06, + "loss": 1.0846, + "step": 30674 + }, + { + "epoch": 0.72, + "grad_norm": 1.8498074396302455, + "learning_rate": 3.769619905216507e-06, + "loss": 0.9585, + "step": 30675 + }, + { + "epoch": 0.72, + "grad_norm": 1.8865557067282215, + "learning_rate": 3.7690230790763094e-06, + "loss": 0.9186, + "step": 30676 + }, + { + "epoch": 0.72, + "grad_norm": 1.9711253782576164, + "learning_rate": 3.768426289214828e-06, + "loss": 0.9805, + "step": 30677 + }, + { + "epoch": 0.72, + "grad_norm": 1.96404825972303, + "learning_rate": 3.7678295356355456e-06, + "loss": 0.8993, + "step": 30678 + }, + { + "epoch": 0.72, + "grad_norm": 1.9959796620497992, + "learning_rate": 3.7672328183419327e-06, + "loss": 0.9247, + "step": 30679 + }, + { + "epoch": 0.72, + "grad_norm": 1.0551685046589838, + "learning_rate": 3.7666361373374607e-06, + "loss": 0.9094, + "step": 30680 + }, + { + "epoch": 0.72, + "grad_norm": 2.2003489809769943, + "learning_rate": 3.766039492625605e-06, + "loss": 0.93, + "step": 30681 + }, + { + "epoch": 0.72, + "grad_norm": 2.033535688856843, + "learning_rate": 3.7654428842098457e-06, + "loss": 0.9322, + "step": 30682 + }, + { + "epoch": 0.72, + "grad_norm": 2.2724992880394552, + "learning_rate": 3.764846312093652e-06, + "loss": 0.9209, + "step": 30683 + }, + { + "epoch": 0.72, + "grad_norm": 2.0974701601515338, + "learning_rate": 3.764249776280493e-06, + "loss": 0.9294, + "step": 30684 + }, + { + "epoch": 0.72, + "grad_norm": 1.0984319071920692, + "learning_rate": 3.7636532767738455e-06, + "loss": 1.0202, + "step": 30685 + }, + { + "epoch": 0.72, + "grad_norm": 2.465273765575065, + "learning_rate": 3.7630568135771873e-06, + "loss": 0.9682, + "step": 30686 + }, + { + "epoch": 0.72, + "grad_norm": 1.9735008905367442, + "learning_rate": 3.762460386693987e-06, + "loss": 1.0531, + "step": 30687 + }, + { + "epoch": 0.72, + "grad_norm": 1.901323126942182, + "learning_rate": 3.7618639961277124e-06, + "loss": 0.8961, + "step": 30688 + }, + { + "epoch": 0.72, + "grad_norm": 1.8012535553893543, + "learning_rate": 3.7612676418818415e-06, + "loss": 0.9106, + "step": 30689 + }, + { + "epoch": 0.72, + "grad_norm": 1.9482983130542568, + "learning_rate": 3.760671323959849e-06, + "loss": 0.8631, + "step": 30690 + }, + { + "epoch": 0.72, + "grad_norm": 1.090927513629906, + "learning_rate": 3.7600750423652043e-06, + "loss": 1.0048, + "step": 30691 + }, + { + "epoch": 0.72, + "grad_norm": 2.0274108363868955, + "learning_rate": 3.7594787971013736e-06, + "loss": 0.8525, + "step": 30692 + }, + { + "epoch": 0.72, + "grad_norm": 2.3031521278510203, + "learning_rate": 3.7588825881718372e-06, + "loss": 0.8566, + "step": 30693 + }, + { + "epoch": 0.72, + "grad_norm": 2.047718063754668, + "learning_rate": 3.758286415580058e-06, + "loss": 1.0217, + "step": 30694 + }, + { + "epoch": 0.72, + "grad_norm": 2.2840422426478684, + "learning_rate": 3.7576902793295158e-06, + "loss": 0.9732, + "step": 30695 + }, + { + "epoch": 0.72, + "grad_norm": 1.9309760930764146, + "learning_rate": 3.7570941794236725e-06, + "loss": 0.9934, + "step": 30696 + }, + { + "epoch": 0.72, + "grad_norm": 2.1577427674079583, + "learning_rate": 3.7564981158660073e-06, + "loss": 1.001, + "step": 30697 + }, + { + "epoch": 0.72, + "grad_norm": 1.9469402965372182, + "learning_rate": 3.755902088659983e-06, + "loss": 1.0062, + "step": 30698 + }, + { + "epoch": 0.72, + "grad_norm": 2.0039364888463997, + "learning_rate": 3.755306097809074e-06, + "loss": 1.0996, + "step": 30699 + }, + { + "epoch": 0.72, + "grad_norm": 1.2703844290175166, + "learning_rate": 3.7547101433167567e-06, + "loss": 1.0945, + "step": 30700 + }, + { + "epoch": 0.72, + "grad_norm": 2.0224230444778364, + "learning_rate": 3.7541142251864883e-06, + "loss": 1.0031, + "step": 30701 + }, + { + "epoch": 0.72, + "grad_norm": 2.139819325761724, + "learning_rate": 3.7535183434217427e-06, + "loss": 1.0144, + "step": 30702 + }, + { + "epoch": 0.72, + "grad_norm": 2.5874614831198834, + "learning_rate": 3.7529224980259914e-06, + "loss": 0.9308, + "step": 30703 + }, + { + "epoch": 0.72, + "grad_norm": 1.9648449693471262, + "learning_rate": 3.75232668900271e-06, + "loss": 0.9448, + "step": 30704 + }, + { + "epoch": 0.72, + "grad_norm": 1.084863319155742, + "learning_rate": 3.751730916355354e-06, + "loss": 0.9507, + "step": 30705 + }, + { + "epoch": 0.72, + "grad_norm": 3.400241436744898, + "learning_rate": 3.7511351800873987e-06, + "loss": 0.9604, + "step": 30706 + }, + { + "epoch": 0.72, + "grad_norm": 2.0709781797299565, + "learning_rate": 3.7505394802023163e-06, + "loss": 0.9774, + "step": 30707 + }, + { + "epoch": 0.72, + "grad_norm": 1.8520223319686888, + "learning_rate": 3.749943816703572e-06, + "loss": 0.9281, + "step": 30708 + }, + { + "epoch": 0.72, + "grad_norm": 2.3313830382733753, + "learning_rate": 3.7493481895946295e-06, + "loss": 1.0202, + "step": 30709 + }, + { + "epoch": 0.72, + "grad_norm": 2.1077915994017613, + "learning_rate": 3.7487525988789595e-06, + "loss": 0.9179, + "step": 30710 + }, + { + "epoch": 0.72, + "grad_norm": 1.9310249260656487, + "learning_rate": 3.7481570445600357e-06, + "loss": 1.0891, + "step": 30711 + }, + { + "epoch": 0.72, + "grad_norm": 3.2870174614756893, + "learning_rate": 3.7475615266413178e-06, + "loss": 1.0421, + "step": 30712 + }, + { + "epoch": 0.72, + "grad_norm": 1.9736433374548021, + "learning_rate": 3.746966045126279e-06, + "loss": 0.9726, + "step": 30713 + }, + { + "epoch": 0.72, + "grad_norm": 2.024570403327959, + "learning_rate": 3.7463706000183843e-06, + "loss": 1.0408, + "step": 30714 + }, + { + "epoch": 0.72, + "grad_norm": 1.1116652970287568, + "learning_rate": 3.745775191321096e-06, + "loss": 0.9834, + "step": 30715 + }, + { + "epoch": 0.72, + "grad_norm": 1.9272968891042237, + "learning_rate": 3.745179819037884e-06, + "loss": 1.063, + "step": 30716 + }, + { + "epoch": 0.72, + "grad_norm": 1.9483451678610726, + "learning_rate": 3.74458448317222e-06, + "loss": 1.0142, + "step": 30717 + }, + { + "epoch": 0.72, + "grad_norm": 1.984969530628435, + "learning_rate": 3.7439891837275644e-06, + "loss": 1.0377, + "step": 30718 + }, + { + "epoch": 0.72, + "grad_norm": 4.060975259253397, + "learning_rate": 3.743393920707381e-06, + "loss": 0.9761, + "step": 30719 + }, + { + "epoch": 0.72, + "grad_norm": 1.8545654067982116, + "learning_rate": 3.7427986941151394e-06, + "loss": 0.8918, + "step": 30720 + }, + { + "epoch": 0.72, + "grad_norm": 2.4772803660572564, + "learning_rate": 3.742203503954308e-06, + "loss": 1.0322, + "step": 30721 + }, + { + "epoch": 0.72, + "grad_norm": 2.105599233683602, + "learning_rate": 3.7416083502283494e-06, + "loss": 0.9534, + "step": 30722 + }, + { + "epoch": 0.72, + "grad_norm": 1.978675816971413, + "learning_rate": 3.7410132329407258e-06, + "loss": 0.9938, + "step": 30723 + }, + { + "epoch": 0.72, + "grad_norm": 2.163326258258312, + "learning_rate": 3.7404181520949033e-06, + "loss": 0.9588, + "step": 30724 + }, + { + "epoch": 0.72, + "grad_norm": 1.9855786351776974, + "learning_rate": 3.7398231076943513e-06, + "loss": 0.9993, + "step": 30725 + }, + { + "epoch": 0.72, + "grad_norm": 1.0424015474261918, + "learning_rate": 3.7392280997425325e-06, + "loss": 0.9192, + "step": 30726 + }, + { + "epoch": 0.72, + "grad_norm": 1.0648297921930847, + "learning_rate": 3.7386331282429045e-06, + "loss": 0.9524, + "step": 30727 + }, + { + "epoch": 0.72, + "grad_norm": 2.1440003861499215, + "learning_rate": 3.7380381931989384e-06, + "loss": 0.8673, + "step": 30728 + }, + { + "epoch": 0.72, + "grad_norm": 1.8009562186275403, + "learning_rate": 3.7374432946140994e-06, + "loss": 0.9318, + "step": 30729 + }, + { + "epoch": 0.72, + "grad_norm": 1.8136405158150959, + "learning_rate": 3.7368484324918473e-06, + "loss": 0.8856, + "step": 30730 + }, + { + "epoch": 0.72, + "grad_norm": 2.013745311323915, + "learning_rate": 3.7362536068356448e-06, + "loss": 0.8337, + "step": 30731 + }, + { + "epoch": 0.72, + "grad_norm": 2.162131307621202, + "learning_rate": 3.7356588176489595e-06, + "loss": 1.0574, + "step": 30732 + }, + { + "epoch": 0.72, + "grad_norm": 1.931088974608659, + "learning_rate": 3.7350640649352477e-06, + "loss": 0.9191, + "step": 30733 + }, + { + "epoch": 0.72, + "grad_norm": 1.913280384440088, + "learning_rate": 3.73446934869798e-06, + "loss": 1.0148, + "step": 30734 + }, + { + "epoch": 0.72, + "grad_norm": 1.0242050056904732, + "learning_rate": 3.7338746689406123e-06, + "loss": 0.9038, + "step": 30735 + }, + { + "epoch": 0.72, + "grad_norm": 1.9061397076968736, + "learning_rate": 3.7332800256666134e-06, + "loss": 0.9096, + "step": 30736 + }, + { + "epoch": 0.72, + "grad_norm": 1.0580332488906625, + "learning_rate": 3.7326854188794393e-06, + "loss": 0.9537, + "step": 30737 + }, + { + "epoch": 0.72, + "grad_norm": 1.8836246080532952, + "learning_rate": 3.732090848582558e-06, + "loss": 0.9611, + "step": 30738 + }, + { + "epoch": 0.72, + "grad_norm": 1.8894387653744453, + "learning_rate": 3.731496314779428e-06, + "loss": 1.0525, + "step": 30739 + }, + { + "epoch": 0.72, + "grad_norm": 1.856208829978108, + "learning_rate": 3.7309018174735078e-06, + "loss": 0.9913, + "step": 30740 + }, + { + "epoch": 0.72, + "grad_norm": 2.289855678633677, + "learning_rate": 3.7303073566682625e-06, + "loss": 0.9838, + "step": 30741 + }, + { + "epoch": 0.72, + "grad_norm": 1.953115700461183, + "learning_rate": 3.7297129323671553e-06, + "loss": 1.0265, + "step": 30742 + }, + { + "epoch": 0.72, + "grad_norm": 2.5155234581526242, + "learning_rate": 3.7291185445736445e-06, + "loss": 1.0342, + "step": 30743 + }, + { + "epoch": 0.72, + "grad_norm": 1.9615941021740526, + "learning_rate": 3.7285241932911874e-06, + "loss": 0.8484, + "step": 30744 + }, + { + "epoch": 0.72, + "grad_norm": 1.7861103193447716, + "learning_rate": 3.7279298785232488e-06, + "loss": 0.9823, + "step": 30745 + }, + { + "epoch": 0.72, + "grad_norm": 3.3691702884252788, + "learning_rate": 3.7273356002732908e-06, + "loss": 1.0242, + "step": 30746 + }, + { + "epoch": 0.72, + "grad_norm": 2.0852542083547787, + "learning_rate": 3.726741358544771e-06, + "loss": 1.0159, + "step": 30747 + }, + { + "epoch": 0.72, + "grad_norm": 1.7946710879623522, + "learning_rate": 3.7261471533411464e-06, + "loss": 0.9136, + "step": 30748 + }, + { + "epoch": 0.72, + "grad_norm": 2.370711594513205, + "learning_rate": 3.7255529846658777e-06, + "loss": 1.0389, + "step": 30749 + }, + { + "epoch": 0.72, + "grad_norm": 1.023638936027017, + "learning_rate": 3.724958852522431e-06, + "loss": 0.8629, + "step": 30750 + }, + { + "epoch": 0.72, + "grad_norm": 2.1606579544167666, + "learning_rate": 3.7243647569142606e-06, + "loss": 1.169, + "step": 30751 + }, + { + "epoch": 0.72, + "grad_norm": 2.2037493919196183, + "learning_rate": 3.7237706978448206e-06, + "loss": 1.1708, + "step": 30752 + }, + { + "epoch": 0.72, + "grad_norm": 1.9521976889808792, + "learning_rate": 3.723176675317579e-06, + "loss": 0.9167, + "step": 30753 + }, + { + "epoch": 0.72, + "grad_norm": 2.532529658965619, + "learning_rate": 3.722582689335986e-06, + "loss": 1.0249, + "step": 30754 + }, + { + "epoch": 0.72, + "grad_norm": 1.8770762737077817, + "learning_rate": 3.721988739903508e-06, + "loss": 0.9822, + "step": 30755 + }, + { + "epoch": 0.72, + "grad_norm": 2.6010414544958844, + "learning_rate": 3.7213948270235954e-06, + "loss": 0.9188, + "step": 30756 + }, + { + "epoch": 0.72, + "grad_norm": 1.819065192354045, + "learning_rate": 3.7208009506997135e-06, + "loss": 0.7878, + "step": 30757 + }, + { + "epoch": 0.72, + "grad_norm": 2.1660682237136157, + "learning_rate": 3.720207110935312e-06, + "loss": 0.9366, + "step": 30758 + }, + { + "epoch": 0.72, + "grad_norm": 1.9048764236414046, + "learning_rate": 3.719613307733858e-06, + "loss": 1.0132, + "step": 30759 + }, + { + "epoch": 0.72, + "grad_norm": 1.9387716627827405, + "learning_rate": 3.719019541098803e-06, + "loss": 1.0632, + "step": 30760 + }, + { + "epoch": 0.72, + "grad_norm": 2.1687655898308194, + "learning_rate": 3.7184258110336e-06, + "loss": 0.9618, + "step": 30761 + }, + { + "epoch": 0.72, + "grad_norm": 2.1444774548848216, + "learning_rate": 3.717832117541712e-06, + "loss": 0.9096, + "step": 30762 + }, + { + "epoch": 0.72, + "grad_norm": 1.84158214822234, + "learning_rate": 3.7172384606265976e-06, + "loss": 1.0527, + "step": 30763 + }, + { + "epoch": 0.72, + "grad_norm": 2.2319664089496127, + "learning_rate": 3.71664484029171e-06, + "loss": 0.9663, + "step": 30764 + }, + { + "epoch": 0.72, + "grad_norm": 2.0897936757051303, + "learning_rate": 3.7160512565405016e-06, + "loss": 0.9837, + "step": 30765 + }, + { + "epoch": 0.72, + "grad_norm": 2.0639761508061576, + "learning_rate": 3.7154577093764334e-06, + "loss": 0.9716, + "step": 30766 + }, + { + "epoch": 0.72, + "grad_norm": 1.9760340099047287, + "learning_rate": 3.714864198802962e-06, + "loss": 0.9996, + "step": 30767 + }, + { + "epoch": 0.72, + "grad_norm": 1.99555337906504, + "learning_rate": 3.714270724823542e-06, + "loss": 0.9583, + "step": 30768 + }, + { + "epoch": 0.72, + "grad_norm": 2.1141659106572246, + "learning_rate": 3.713677287441624e-06, + "loss": 0.9479, + "step": 30769 + }, + { + "epoch": 0.72, + "grad_norm": 1.8232516339337779, + "learning_rate": 3.7130838866606665e-06, + "loss": 1.0783, + "step": 30770 + }, + { + "epoch": 0.72, + "grad_norm": 2.101423459478944, + "learning_rate": 3.7124905224841303e-06, + "loss": 0.9766, + "step": 30771 + }, + { + "epoch": 0.72, + "grad_norm": 2.1854324744464497, + "learning_rate": 3.7118971949154624e-06, + "loss": 1.171, + "step": 30772 + }, + { + "epoch": 0.72, + "grad_norm": 1.8912123121821733, + "learning_rate": 3.7113039039581177e-06, + "loss": 0.9286, + "step": 30773 + }, + { + "epoch": 0.73, + "grad_norm": 1.0988907189175454, + "learning_rate": 3.7107106496155565e-06, + "loss": 0.9946, + "step": 30774 + }, + { + "epoch": 0.73, + "grad_norm": 1.9228320718942777, + "learning_rate": 3.7101174318912257e-06, + "loss": 1.0283, + "step": 30775 + }, + { + "epoch": 0.73, + "grad_norm": 1.8949469299582686, + "learning_rate": 3.7095242507885854e-06, + "loss": 0.8574, + "step": 30776 + }, + { + "epoch": 0.73, + "grad_norm": 1.8980972567452743, + "learning_rate": 3.7089311063110823e-06, + "loss": 0.9693, + "step": 30777 + }, + { + "epoch": 0.73, + "grad_norm": 1.9637816024941186, + "learning_rate": 3.708337998462178e-06, + "loss": 1.0118, + "step": 30778 + }, + { + "epoch": 0.73, + "grad_norm": 2.2797839177394437, + "learning_rate": 3.707744927245319e-06, + "loss": 0.9824, + "step": 30779 + }, + { + "epoch": 0.73, + "grad_norm": 1.9247674326941984, + "learning_rate": 3.707151892663964e-06, + "loss": 1.017, + "step": 30780 + }, + { + "epoch": 0.73, + "grad_norm": 2.578193042680853, + "learning_rate": 3.7065588947215636e-06, + "loss": 0.9069, + "step": 30781 + }, + { + "epoch": 0.73, + "grad_norm": 2.024995582067636, + "learning_rate": 3.705965933421566e-06, + "loss": 1.0217, + "step": 30782 + }, + { + "epoch": 0.73, + "grad_norm": 1.0747021993903696, + "learning_rate": 3.7053730087674267e-06, + "loss": 0.9532, + "step": 30783 + }, + { + "epoch": 0.73, + "grad_norm": 2.0707731868000017, + "learning_rate": 3.704780120762599e-06, + "loss": 1.0451, + "step": 30784 + }, + { + "epoch": 0.73, + "grad_norm": 2.429728164976513, + "learning_rate": 3.704187269410542e-06, + "loss": 0.8903, + "step": 30785 + }, + { + "epoch": 0.73, + "grad_norm": 2.162163568888349, + "learning_rate": 3.7035944547146917e-06, + "loss": 1.0567, + "step": 30786 + }, + { + "epoch": 0.73, + "grad_norm": 1.8163248169498165, + "learning_rate": 3.7030016766785095e-06, + "loss": 0.9232, + "step": 30787 + }, + { + "epoch": 0.73, + "grad_norm": 1.9073670923179, + "learning_rate": 3.702408935305447e-06, + "loss": 0.965, + "step": 30788 + }, + { + "epoch": 0.73, + "grad_norm": 2.036783501371662, + "learning_rate": 3.701816230598955e-06, + "loss": 1.0942, + "step": 30789 + }, + { + "epoch": 0.73, + "grad_norm": 1.8974571168476262, + "learning_rate": 3.701223562562478e-06, + "loss": 0.9729, + "step": 30790 + }, + { + "epoch": 0.73, + "grad_norm": 2.117467462541688, + "learning_rate": 3.7006309311994727e-06, + "loss": 0.9677, + "step": 30791 + }, + { + "epoch": 0.73, + "grad_norm": 2.600955577849334, + "learning_rate": 3.7000383365133916e-06, + "loss": 0.982, + "step": 30792 + }, + { + "epoch": 0.73, + "grad_norm": 2.130158447098388, + "learning_rate": 3.6994457785076823e-06, + "loss": 1.0309, + "step": 30793 + }, + { + "epoch": 0.73, + "grad_norm": 1.9049130529811436, + "learning_rate": 3.6988532571857895e-06, + "loss": 0.9091, + "step": 30794 + }, + { + "epoch": 0.73, + "grad_norm": 2.371986129714937, + "learning_rate": 3.698260772551172e-06, + "loss": 1.056, + "step": 30795 + }, + { + "epoch": 0.73, + "grad_norm": 1.8803852913339587, + "learning_rate": 3.697668324607272e-06, + "loss": 0.7944, + "step": 30796 + }, + { + "epoch": 0.73, + "grad_norm": 1.737179852863359, + "learning_rate": 3.697075913357543e-06, + "loss": 0.8178, + "step": 30797 + }, + { + "epoch": 0.73, + "grad_norm": 2.003778195030903, + "learning_rate": 3.6964835388054367e-06, + "loss": 1.0601, + "step": 30798 + }, + { + "epoch": 0.73, + "grad_norm": 1.9716754159177297, + "learning_rate": 3.695891200954398e-06, + "loss": 1.0175, + "step": 30799 + }, + { + "epoch": 0.73, + "grad_norm": 1.966730965482584, + "learning_rate": 3.695298899807874e-06, + "loss": 1.1174, + "step": 30800 + }, + { + "epoch": 0.73, + "grad_norm": 2.1766483979774214, + "learning_rate": 3.694706635369316e-06, + "loss": 1.0022, + "step": 30801 + }, + { + "epoch": 0.73, + "grad_norm": 2.1145138287097907, + "learning_rate": 3.6941144076421786e-06, + "loss": 0.9777, + "step": 30802 + }, + { + "epoch": 0.73, + "grad_norm": 1.824884370327566, + "learning_rate": 3.693522216629897e-06, + "loss": 0.9301, + "step": 30803 + }, + { + "epoch": 0.73, + "grad_norm": 2.2390031119686067, + "learning_rate": 3.6929300623359255e-06, + "loss": 1.0243, + "step": 30804 + }, + { + "epoch": 0.73, + "grad_norm": 2.0909199441479367, + "learning_rate": 3.692337944763711e-06, + "loss": 1.0915, + "step": 30805 + }, + { + "epoch": 0.73, + "grad_norm": 2.1055373295008213, + "learning_rate": 3.6917458639167105e-06, + "loss": 1.02, + "step": 30806 + }, + { + "epoch": 0.73, + "grad_norm": 2.0772516681619124, + "learning_rate": 3.691153819798354e-06, + "loss": 0.9266, + "step": 30807 + }, + { + "epoch": 0.73, + "grad_norm": 1.8813106734707274, + "learning_rate": 3.690561812412099e-06, + "loss": 1.0952, + "step": 30808 + }, + { + "epoch": 0.73, + "grad_norm": 1.0979145391416552, + "learning_rate": 3.689969841761393e-06, + "loss": 0.9451, + "step": 30809 + }, + { + "epoch": 0.73, + "grad_norm": 2.040971127166458, + "learning_rate": 3.689377907849677e-06, + "loss": 0.948, + "step": 30810 + }, + { + "epoch": 0.73, + "grad_norm": 2.251052240404912, + "learning_rate": 3.688786010680405e-06, + "loss": 1.0034, + "step": 30811 + }, + { + "epoch": 0.73, + "grad_norm": 1.8368027365360058, + "learning_rate": 3.688194150257015e-06, + "loss": 0.9345, + "step": 30812 + }, + { + "epoch": 0.73, + "grad_norm": 2.0694991765663633, + "learning_rate": 3.687602326582961e-06, + "loss": 1.0196, + "step": 30813 + }, + { + "epoch": 0.73, + "grad_norm": 2.5414929716770964, + "learning_rate": 3.6870105396616795e-06, + "loss": 0.9131, + "step": 30814 + }, + { + "epoch": 0.73, + "grad_norm": 1.960974257080034, + "learning_rate": 3.6864187894966263e-06, + "loss": 0.8253, + "step": 30815 + }, + { + "epoch": 0.73, + "grad_norm": 2.0197210046926704, + "learning_rate": 3.685827076091242e-06, + "loss": 0.965, + "step": 30816 + }, + { + "epoch": 0.73, + "grad_norm": 1.815112214188042, + "learning_rate": 3.6852353994489665e-06, + "loss": 0.8656, + "step": 30817 + }, + { + "epoch": 0.73, + "grad_norm": 1.742194465443667, + "learning_rate": 3.684643759573251e-06, + "loss": 0.8796, + "step": 30818 + }, + { + "epoch": 0.73, + "grad_norm": 1.8653372274996163, + "learning_rate": 3.6840521564675424e-06, + "loss": 0.9895, + "step": 30819 + }, + { + "epoch": 0.73, + "grad_norm": 2.4596435411965114, + "learning_rate": 3.6834605901352804e-06, + "loss": 0.9233, + "step": 30820 + }, + { + "epoch": 0.73, + "grad_norm": 2.0051237214518856, + "learning_rate": 3.6828690605799087e-06, + "loss": 1.0921, + "step": 30821 + }, + { + "epoch": 0.73, + "grad_norm": 2.0432577492904596, + "learning_rate": 3.6822775678048727e-06, + "loss": 0.8739, + "step": 30822 + }, + { + "epoch": 0.73, + "grad_norm": 2.031587468137463, + "learning_rate": 3.6816861118136204e-06, + "loss": 0.9774, + "step": 30823 + }, + { + "epoch": 0.73, + "grad_norm": 2.172089378107581, + "learning_rate": 3.6810946926095914e-06, + "loss": 0.9555, + "step": 30824 + }, + { + "epoch": 0.73, + "grad_norm": 2.0256821743043836, + "learning_rate": 3.680503310196227e-06, + "loss": 1.0204, + "step": 30825 + }, + { + "epoch": 0.73, + "grad_norm": 2.0684713514795923, + "learning_rate": 3.679911964576972e-06, + "loss": 0.9666, + "step": 30826 + }, + { + "epoch": 0.73, + "grad_norm": 1.9967680028410637, + "learning_rate": 3.6793206557552753e-06, + "loss": 0.9414, + "step": 30827 + }, + { + "epoch": 0.73, + "grad_norm": 2.8241001828256236, + "learning_rate": 3.678729383734574e-06, + "loss": 0.9807, + "step": 30828 + }, + { + "epoch": 0.73, + "grad_norm": 1.0529289622004478, + "learning_rate": 3.678138148518308e-06, + "loss": 0.9245, + "step": 30829 + }, + { + "epoch": 0.73, + "grad_norm": 1.9783657535942296, + "learning_rate": 3.677546950109927e-06, + "loss": 1.155, + "step": 30830 + }, + { + "epoch": 0.73, + "grad_norm": 3.6331052247226197, + "learning_rate": 3.676955788512866e-06, + "loss": 0.8502, + "step": 30831 + }, + { + "epoch": 0.73, + "grad_norm": 2.001802040307148, + "learning_rate": 3.6763646637305726e-06, + "loss": 0.9707, + "step": 30832 + }, + { + "epoch": 0.73, + "grad_norm": 2.0885155131194053, + "learning_rate": 3.6757735757664835e-06, + "loss": 0.9841, + "step": 30833 + }, + { + "epoch": 0.73, + "grad_norm": 1.067643303509335, + "learning_rate": 3.675182524624047e-06, + "loss": 0.9014, + "step": 30834 + }, + { + "epoch": 0.73, + "grad_norm": 2.9737851998004516, + "learning_rate": 3.6745915103066952e-06, + "loss": 0.9809, + "step": 30835 + }, + { + "epoch": 0.73, + "grad_norm": 1.0877582972088617, + "learning_rate": 3.6740005328178786e-06, + "loss": 0.9012, + "step": 30836 + }, + { + "epoch": 0.73, + "grad_norm": 1.9891144718883502, + "learning_rate": 3.6734095921610336e-06, + "loss": 1.0248, + "step": 30837 + }, + { + "epoch": 0.73, + "grad_norm": 2.0129905350671162, + "learning_rate": 3.6728186883395978e-06, + "loss": 0.9918, + "step": 30838 + }, + { + "epoch": 0.73, + "grad_norm": 2.0549494517214915, + "learning_rate": 3.672227821357014e-06, + "loss": 0.9201, + "step": 30839 + }, + { + "epoch": 0.73, + "grad_norm": 1.8326067082353374, + "learning_rate": 3.671636991216727e-06, + "loss": 1.0776, + "step": 30840 + }, + { + "epoch": 0.73, + "grad_norm": 2.2822069275692463, + "learning_rate": 3.6710461979221735e-06, + "loss": 1.0161, + "step": 30841 + }, + { + "epoch": 0.73, + "grad_norm": 2.0245371698072674, + "learning_rate": 3.670455441476788e-06, + "loss": 0.8318, + "step": 30842 + }, + { + "epoch": 0.73, + "grad_norm": 2.0635264271489624, + "learning_rate": 3.6698647218840157e-06, + "loss": 0.9828, + "step": 30843 + }, + { + "epoch": 0.73, + "grad_norm": 1.8932740261458814, + "learning_rate": 3.669274039147299e-06, + "loss": 0.8362, + "step": 30844 + }, + { + "epoch": 0.73, + "grad_norm": 2.0021237990268395, + "learning_rate": 3.6686833932700717e-06, + "loss": 0.9923, + "step": 30845 + }, + { + "epoch": 0.73, + "grad_norm": 2.013151747703562, + "learning_rate": 3.6680927842557724e-06, + "loss": 0.882, + "step": 30846 + }, + { + "epoch": 0.73, + "grad_norm": 2.129040417468148, + "learning_rate": 3.667502212107841e-06, + "loss": 0.9741, + "step": 30847 + }, + { + "epoch": 0.73, + "grad_norm": 2.2352930785636613, + "learning_rate": 3.6669116768297187e-06, + "loss": 0.9288, + "step": 30848 + }, + { + "epoch": 0.73, + "grad_norm": 2.2707393378479575, + "learning_rate": 3.6663211784248433e-06, + "loss": 1.0724, + "step": 30849 + }, + { + "epoch": 0.73, + "grad_norm": 2.1061416677044407, + "learning_rate": 3.665730716896647e-06, + "loss": 0.8504, + "step": 30850 + }, + { + "epoch": 0.73, + "grad_norm": 2.3068702125837146, + "learning_rate": 3.6651402922485715e-06, + "loss": 0.9179, + "step": 30851 + }, + { + "epoch": 0.73, + "grad_norm": 1.936583456132759, + "learning_rate": 3.6645499044840594e-06, + "loss": 0.9104, + "step": 30852 + }, + { + "epoch": 0.73, + "grad_norm": 1.8028247409966784, + "learning_rate": 3.6639595536065418e-06, + "loss": 0.9653, + "step": 30853 + }, + { + "epoch": 0.73, + "grad_norm": 1.7781206768614788, + "learning_rate": 3.6633692396194554e-06, + "loss": 0.9309, + "step": 30854 + }, + { + "epoch": 0.73, + "grad_norm": 2.09238735959782, + "learning_rate": 3.6627789625262434e-06, + "loss": 0.9655, + "step": 30855 + }, + { + "epoch": 0.73, + "grad_norm": 2.4973183969460386, + "learning_rate": 3.662188722330334e-06, + "loss": 0.8536, + "step": 30856 + }, + { + "epoch": 0.73, + "grad_norm": 1.941685145165518, + "learning_rate": 3.6615985190351722e-06, + "loss": 1.1077, + "step": 30857 + }, + { + "epoch": 0.73, + "grad_norm": 2.014818278557888, + "learning_rate": 3.6610083526441876e-06, + "loss": 0.9482, + "step": 30858 + }, + { + "epoch": 0.73, + "grad_norm": 2.5820310209594486, + "learning_rate": 3.660418223160822e-06, + "loss": 0.9653, + "step": 30859 + }, + { + "epoch": 0.73, + "grad_norm": 1.8587909673324499, + "learning_rate": 3.6598281305885052e-06, + "loss": 1.0111, + "step": 30860 + }, + { + "epoch": 0.73, + "grad_norm": 1.9727967675484697, + "learning_rate": 3.65923807493068e-06, + "loss": 1.0764, + "step": 30861 + }, + { + "epoch": 0.73, + "grad_norm": 1.9627298524714536, + "learning_rate": 3.658648056190778e-06, + "loss": 0.9395, + "step": 30862 + }, + { + "epoch": 0.73, + "grad_norm": 2.2104352948925956, + "learning_rate": 3.6580580743722315e-06, + "loss": 0.8676, + "step": 30863 + }, + { + "epoch": 0.73, + "grad_norm": 1.9022003900423123, + "learning_rate": 3.6574681294784788e-06, + "loss": 1.1187, + "step": 30864 + }, + { + "epoch": 0.73, + "grad_norm": 2.005988102160588, + "learning_rate": 3.6568782215129574e-06, + "loss": 1.0098, + "step": 30865 + }, + { + "epoch": 0.73, + "grad_norm": 2.2224954521692477, + "learning_rate": 3.6562883504791004e-06, + "loss": 0.9104, + "step": 30866 + }, + { + "epoch": 0.73, + "grad_norm": 1.983798183324142, + "learning_rate": 3.655698516380336e-06, + "loss": 1.0591, + "step": 30867 + }, + { + "epoch": 0.73, + "grad_norm": 1.860643858487282, + "learning_rate": 3.655108719220104e-06, + "loss": 0.9348, + "step": 30868 + }, + { + "epoch": 0.73, + "grad_norm": 2.1185757537901155, + "learning_rate": 3.6545189590018416e-06, + "loss": 1.0454, + "step": 30869 + }, + { + "epoch": 0.73, + "grad_norm": 2.676493136960372, + "learning_rate": 3.653929235728979e-06, + "loss": 1.0211, + "step": 30870 + }, + { + "epoch": 0.73, + "grad_norm": 2.2898138172384006, + "learning_rate": 3.653339549404946e-06, + "loss": 1.0479, + "step": 30871 + }, + { + "epoch": 0.73, + "grad_norm": 2.3213561659788198, + "learning_rate": 3.652749900033179e-06, + "loss": 0.9488, + "step": 30872 + }, + { + "epoch": 0.73, + "grad_norm": 2.1081052542516843, + "learning_rate": 3.6521602876171147e-06, + "loss": 1.0068, + "step": 30873 + }, + { + "epoch": 0.73, + "grad_norm": 1.7524382102822373, + "learning_rate": 3.651570712160184e-06, + "loss": 0.7527, + "step": 30874 + }, + { + "epoch": 0.73, + "grad_norm": 1.9013490878068044, + "learning_rate": 3.650981173665814e-06, + "loss": 0.9397, + "step": 30875 + }, + { + "epoch": 0.73, + "grad_norm": 1.1264041004578433, + "learning_rate": 3.6503916721374456e-06, + "loss": 0.9057, + "step": 30876 + }, + { + "epoch": 0.73, + "grad_norm": 1.952368147953467, + "learning_rate": 3.6498022075785034e-06, + "loss": 1.0246, + "step": 30877 + }, + { + "epoch": 0.73, + "grad_norm": 1.9589879900807678, + "learning_rate": 3.649212779992427e-06, + "loss": 0.8966, + "step": 30878 + }, + { + "epoch": 0.73, + "grad_norm": 2.2994325547320025, + "learning_rate": 3.6486233893826405e-06, + "loss": 1.016, + "step": 30879 + }, + { + "epoch": 0.73, + "grad_norm": 1.9641986017888968, + "learning_rate": 3.6480340357525832e-06, + "loss": 1.1149, + "step": 30880 + }, + { + "epoch": 0.73, + "grad_norm": 1.0995566381289028, + "learning_rate": 3.6474447191056793e-06, + "loss": 0.9755, + "step": 30881 + }, + { + "epoch": 0.73, + "grad_norm": 2.45935145852544, + "learning_rate": 3.6468554394453636e-06, + "loss": 0.9914, + "step": 30882 + }, + { + "epoch": 0.73, + "grad_norm": 2.0543291835371416, + "learning_rate": 3.6462661967750744e-06, + "loss": 0.9613, + "step": 30883 + }, + { + "epoch": 0.73, + "grad_norm": 2.0761996653815076, + "learning_rate": 3.645676991098227e-06, + "loss": 0.9439, + "step": 30884 + }, + { + "epoch": 0.73, + "grad_norm": 2.197470514917257, + "learning_rate": 3.6450878224182608e-06, + "loss": 1.0062, + "step": 30885 + }, + { + "epoch": 0.73, + "grad_norm": 1.122284685312276, + "learning_rate": 3.6444986907386035e-06, + "loss": 0.9962, + "step": 30886 + }, + { + "epoch": 0.73, + "grad_norm": 2.0989461351383136, + "learning_rate": 3.6439095960626957e-06, + "loss": 0.8323, + "step": 30887 + }, + { + "epoch": 0.73, + "grad_norm": 1.8194862929778184, + "learning_rate": 3.643320538393951e-06, + "loss": 0.9328, + "step": 30888 + }, + { + "epoch": 0.73, + "grad_norm": 2.537964192861456, + "learning_rate": 3.642731517735807e-06, + "loss": 0.9923, + "step": 30889 + }, + { + "epoch": 0.73, + "grad_norm": 2.129765395492129, + "learning_rate": 3.642142534091695e-06, + "loss": 0.9359, + "step": 30890 + }, + { + "epoch": 0.73, + "grad_norm": 1.0630062909543465, + "learning_rate": 3.641553587465043e-06, + "loss": 0.903, + "step": 30891 + }, + { + "epoch": 0.73, + "grad_norm": 1.1311034022050614, + "learning_rate": 3.6409646778592756e-06, + "loss": 0.867, + "step": 30892 + }, + { + "epoch": 0.73, + "grad_norm": 1.8703132479310831, + "learning_rate": 3.640375805277825e-06, + "loss": 1.0111, + "step": 30893 + }, + { + "epoch": 0.73, + "grad_norm": 1.0964247783292613, + "learning_rate": 3.639786969724123e-06, + "loss": 0.9325, + "step": 30894 + }, + { + "epoch": 0.73, + "grad_norm": 2.7360874556639154, + "learning_rate": 3.6391981712015922e-06, + "loss": 0.8876, + "step": 30895 + }, + { + "epoch": 0.73, + "grad_norm": 2.021664786350687, + "learning_rate": 3.638609409713666e-06, + "loss": 1.0772, + "step": 30896 + }, + { + "epoch": 0.73, + "grad_norm": 1.9600033279124678, + "learning_rate": 3.63802068526377e-06, + "loss": 0.9716, + "step": 30897 + }, + { + "epoch": 0.73, + "grad_norm": 1.9173544578101307, + "learning_rate": 3.6374319978553284e-06, + "loss": 0.9913, + "step": 30898 + }, + { + "epoch": 0.73, + "grad_norm": 1.9043785544142364, + "learning_rate": 3.636843347491772e-06, + "loss": 1.0945, + "step": 30899 + }, + { + "epoch": 0.73, + "grad_norm": 2.041683969914221, + "learning_rate": 3.636254734176532e-06, + "loss": 0.9425, + "step": 30900 + }, + { + "epoch": 0.73, + "grad_norm": 2.2265897814016262, + "learning_rate": 3.6356661579130314e-06, + "loss": 0.9379, + "step": 30901 + }, + { + "epoch": 0.73, + "grad_norm": 2.695208477694302, + "learning_rate": 3.635077618704694e-06, + "loss": 0.8525, + "step": 30902 + }, + { + "epoch": 0.73, + "grad_norm": 2.069547662394496, + "learning_rate": 3.63448911655495e-06, + "loss": 0.9476, + "step": 30903 + }, + { + "epoch": 0.73, + "grad_norm": 1.9875866668814208, + "learning_rate": 3.633900651467228e-06, + "loss": 1.0054, + "step": 30904 + }, + { + "epoch": 0.73, + "grad_norm": 1.8459891872138572, + "learning_rate": 3.6333122234449523e-06, + "loss": 0.9021, + "step": 30905 + }, + { + "epoch": 0.73, + "grad_norm": 1.870432417648511, + "learning_rate": 3.632723832491546e-06, + "loss": 0.9266, + "step": 30906 + }, + { + "epoch": 0.73, + "grad_norm": 2.3500314636229356, + "learning_rate": 3.632135478610437e-06, + "loss": 1.1151, + "step": 30907 + }, + { + "epoch": 0.73, + "grad_norm": 2.129779648873184, + "learning_rate": 3.6315471618050546e-06, + "loss": 0.9063, + "step": 30908 + }, + { + "epoch": 0.73, + "grad_norm": 1.9358854762708186, + "learning_rate": 3.6309588820788213e-06, + "loss": 0.9078, + "step": 30909 + }, + { + "epoch": 0.73, + "grad_norm": 2.021393074747346, + "learning_rate": 3.6303706394351567e-06, + "loss": 1.0631, + "step": 30910 + }, + { + "epoch": 0.73, + "grad_norm": 1.9862370103242288, + "learning_rate": 3.6297824338774957e-06, + "loss": 1.0461, + "step": 30911 + }, + { + "epoch": 0.73, + "grad_norm": 2.1003302223873686, + "learning_rate": 3.6291942654092547e-06, + "loss": 1.0853, + "step": 30912 + }, + { + "epoch": 0.73, + "grad_norm": 2.116442387034906, + "learning_rate": 3.628606134033865e-06, + "loss": 1.0717, + "step": 30913 + }, + { + "epoch": 0.73, + "grad_norm": 2.0824513977639993, + "learning_rate": 3.6280180397547438e-06, + "loss": 0.9147, + "step": 30914 + }, + { + "epoch": 0.73, + "grad_norm": 2.5709123174720427, + "learning_rate": 3.6274299825753225e-06, + "loss": 1.1, + "step": 30915 + }, + { + "epoch": 0.73, + "grad_norm": 1.9033176586999687, + "learning_rate": 3.626841962499018e-06, + "loss": 1.0234, + "step": 30916 + }, + { + "epoch": 0.73, + "grad_norm": 1.894385595228311, + "learning_rate": 3.6262539795292604e-06, + "loss": 1.0152, + "step": 30917 + }, + { + "epoch": 0.73, + "grad_norm": 2.1877499665339295, + "learning_rate": 3.6256660336694705e-06, + "loss": 0.9868, + "step": 30918 + }, + { + "epoch": 0.73, + "grad_norm": 2.200047355833559, + "learning_rate": 3.625078124923067e-06, + "loss": 1.0737, + "step": 30919 + }, + { + "epoch": 0.73, + "grad_norm": 1.8860061043872176, + "learning_rate": 3.624490253293477e-06, + "loss": 0.9069, + "step": 30920 + }, + { + "epoch": 0.73, + "grad_norm": 2.230899246061913, + "learning_rate": 3.6239024187841266e-06, + "loss": 1.1396, + "step": 30921 + }, + { + "epoch": 0.73, + "grad_norm": 1.955349597389803, + "learning_rate": 3.623314621398436e-06, + "loss": 0.958, + "step": 30922 + }, + { + "epoch": 0.73, + "grad_norm": 2.0842212922863763, + "learning_rate": 3.6227268611398214e-06, + "loss": 1.0278, + "step": 30923 + }, + { + "epoch": 0.73, + "grad_norm": 2.283289125387547, + "learning_rate": 3.622139138011711e-06, + "loss": 1.0263, + "step": 30924 + }, + { + "epoch": 0.73, + "grad_norm": 1.099142759546807, + "learning_rate": 3.62155145201753e-06, + "loss": 0.9926, + "step": 30925 + }, + { + "epoch": 0.73, + "grad_norm": 1.1132518197653547, + "learning_rate": 3.6209638031606942e-06, + "loss": 0.9669, + "step": 30926 + }, + { + "epoch": 0.73, + "grad_norm": 1.8859511250540621, + "learning_rate": 3.6203761914446243e-06, + "loss": 1.052, + "step": 30927 + }, + { + "epoch": 0.73, + "grad_norm": 1.8856459675914343, + "learning_rate": 3.6197886168727436e-06, + "loss": 0.9871, + "step": 30928 + }, + { + "epoch": 0.73, + "grad_norm": 2.081635539877599, + "learning_rate": 3.6192010794484766e-06, + "loss": 0.9151, + "step": 30929 + }, + { + "epoch": 0.73, + "grad_norm": 2.1641694698337313, + "learning_rate": 3.618613579175242e-06, + "loss": 1.0459, + "step": 30930 + }, + { + "epoch": 0.73, + "grad_norm": 1.9365114038769702, + "learning_rate": 3.618026116056456e-06, + "loss": 1.0186, + "step": 30931 + }, + { + "epoch": 0.73, + "grad_norm": 2.124013145246453, + "learning_rate": 3.617438690095545e-06, + "loss": 0.8633, + "step": 30932 + }, + { + "epoch": 0.73, + "grad_norm": 2.0240988235216575, + "learning_rate": 3.6168513012959227e-06, + "loss": 0.8376, + "step": 30933 + }, + { + "epoch": 0.73, + "grad_norm": 1.8458101472404569, + "learning_rate": 3.6162639496610173e-06, + "loss": 0.9787, + "step": 30934 + }, + { + "epoch": 0.73, + "grad_norm": 2.086559693873032, + "learning_rate": 3.615676635194242e-06, + "loss": 0.9672, + "step": 30935 + }, + { + "epoch": 0.73, + "grad_norm": 1.854591500617149, + "learning_rate": 3.6150893578990198e-06, + "loss": 0.9888, + "step": 30936 + }, + { + "epoch": 0.73, + "grad_norm": 1.8349392358248782, + "learning_rate": 3.6145021177787666e-06, + "loss": 0.8462, + "step": 30937 + }, + { + "epoch": 0.73, + "grad_norm": 1.1053013014470452, + "learning_rate": 3.613914914836907e-06, + "loss": 0.9781, + "step": 30938 + }, + { + "epoch": 0.73, + "grad_norm": 1.9718796406412078, + "learning_rate": 3.613327749076856e-06, + "loss": 0.9614, + "step": 30939 + }, + { + "epoch": 0.73, + "grad_norm": 1.0504235424116495, + "learning_rate": 3.6127406205020298e-06, + "loss": 0.9336, + "step": 30940 + }, + { + "epoch": 0.73, + "grad_norm": 1.937796297869759, + "learning_rate": 3.6121535291158506e-06, + "loss": 1.1775, + "step": 30941 + }, + { + "epoch": 0.73, + "grad_norm": 2.635127750734087, + "learning_rate": 3.6115664749217383e-06, + "loss": 0.9969, + "step": 30942 + }, + { + "epoch": 0.73, + "grad_norm": 2.113320906438007, + "learning_rate": 3.6109794579231094e-06, + "loss": 0.9597, + "step": 30943 + }, + { + "epoch": 0.73, + "grad_norm": 1.9918649832553823, + "learning_rate": 3.610392478123377e-06, + "loss": 0.9642, + "step": 30944 + }, + { + "epoch": 0.73, + "grad_norm": 4.462094743597452, + "learning_rate": 3.609805535525962e-06, + "loss": 0.971, + "step": 30945 + }, + { + "epoch": 0.73, + "grad_norm": 2.0034859941161582, + "learning_rate": 3.6092186301342858e-06, + "loss": 0.9638, + "step": 30946 + }, + { + "epoch": 0.73, + "grad_norm": 1.933674629041933, + "learning_rate": 3.608631761951763e-06, + "loss": 1.064, + "step": 30947 + }, + { + "epoch": 0.73, + "grad_norm": 1.9660776208445845, + "learning_rate": 3.6080449309818045e-06, + "loss": 1.0976, + "step": 30948 + }, + { + "epoch": 0.73, + "grad_norm": 1.8235722673629942, + "learning_rate": 3.607458137227834e-06, + "loss": 0.9107, + "step": 30949 + }, + { + "epoch": 0.73, + "grad_norm": 1.8341603903049322, + "learning_rate": 3.6068713806932678e-06, + "loss": 0.9777, + "step": 30950 + }, + { + "epoch": 0.73, + "grad_norm": 1.9795084781145769, + "learning_rate": 3.6062846613815217e-06, + "loss": 0.9335, + "step": 30951 + }, + { + "epoch": 0.73, + "grad_norm": 2.072153161859503, + "learning_rate": 3.6056979792960067e-06, + "loss": 1.0252, + "step": 30952 + }, + { + "epoch": 0.73, + "grad_norm": 2.462465077436379, + "learning_rate": 3.6051113344401457e-06, + "loss": 0.775, + "step": 30953 + }, + { + "epoch": 0.73, + "grad_norm": 1.7813906139631202, + "learning_rate": 3.604524726817348e-06, + "loss": 1.0016, + "step": 30954 + }, + { + "epoch": 0.73, + "grad_norm": 2.0656260602229826, + "learning_rate": 3.603938156431035e-06, + "loss": 1.0565, + "step": 30955 + }, + { + "epoch": 0.73, + "grad_norm": 2.1278144454868375, + "learning_rate": 3.6033516232846154e-06, + "loss": 0.9596, + "step": 30956 + }, + { + "epoch": 0.73, + "grad_norm": 2.4531600416335966, + "learning_rate": 3.602765127381511e-06, + "loss": 0.8984, + "step": 30957 + }, + { + "epoch": 0.73, + "grad_norm": 1.9556438046163072, + "learning_rate": 3.6021786687251315e-06, + "loss": 1.131, + "step": 30958 + }, + { + "epoch": 0.73, + "grad_norm": 3.229917368989746, + "learning_rate": 3.601592247318896e-06, + "loss": 0.8694, + "step": 30959 + }, + { + "epoch": 0.73, + "grad_norm": 2.685515601582935, + "learning_rate": 3.601005863166216e-06, + "loss": 1.0498, + "step": 30960 + }, + { + "epoch": 0.73, + "grad_norm": 2.1890880756165565, + "learning_rate": 3.600419516270501e-06, + "loss": 0.9822, + "step": 30961 + }, + { + "epoch": 0.73, + "grad_norm": 2.414607968053347, + "learning_rate": 3.5998332066351716e-06, + "loss": 0.9405, + "step": 30962 + }, + { + "epoch": 0.73, + "grad_norm": 1.832806874578165, + "learning_rate": 3.599246934263643e-06, + "loss": 0.9377, + "step": 30963 + }, + { + "epoch": 0.73, + "grad_norm": 2.0361805125236336, + "learning_rate": 3.598660699159324e-06, + "loss": 0.9786, + "step": 30964 + }, + { + "epoch": 0.73, + "grad_norm": 1.8406598629578435, + "learning_rate": 3.5980745013256258e-06, + "loss": 1.0053, + "step": 30965 + }, + { + "epoch": 0.73, + "grad_norm": 2.140376069459617, + "learning_rate": 3.5974883407659655e-06, + "loss": 1.0946, + "step": 30966 + }, + { + "epoch": 0.73, + "grad_norm": 2.0229213940803645, + "learning_rate": 3.5969022174837576e-06, + "loss": 1.0636, + "step": 30967 + }, + { + "epoch": 0.73, + "grad_norm": 1.9915934709025627, + "learning_rate": 3.5963161314824123e-06, + "loss": 0.9933, + "step": 30968 + }, + { + "epoch": 0.73, + "grad_norm": 1.6721217958449008, + "learning_rate": 3.5957300827653385e-06, + "loss": 1.008, + "step": 30969 + }, + { + "epoch": 0.73, + "grad_norm": 1.8702400632783962, + "learning_rate": 3.595144071335952e-06, + "loss": 0.9142, + "step": 30970 + }, + { + "epoch": 0.73, + "grad_norm": 1.992985718358419, + "learning_rate": 3.594558097197668e-06, + "loss": 1.0292, + "step": 30971 + }, + { + "epoch": 0.73, + "grad_norm": 2.2442183576555763, + "learning_rate": 3.5939721603538946e-06, + "loss": 1.0363, + "step": 30972 + }, + { + "epoch": 0.73, + "grad_norm": 1.074346464107822, + "learning_rate": 3.5933862608080405e-06, + "loss": 0.9441, + "step": 30973 + }, + { + "epoch": 0.73, + "grad_norm": 2.3958044976420307, + "learning_rate": 3.592800398563522e-06, + "loss": 0.9432, + "step": 30974 + }, + { + "epoch": 0.73, + "grad_norm": 1.786915685496801, + "learning_rate": 3.5922145736237457e-06, + "loss": 0.9851, + "step": 30975 + }, + { + "epoch": 0.73, + "grad_norm": 1.8969324572856219, + "learning_rate": 3.5916287859921273e-06, + "loss": 0.9259, + "step": 30976 + }, + { + "epoch": 0.73, + "grad_norm": 2.001746907603675, + "learning_rate": 3.5910430356720727e-06, + "loss": 1.1014, + "step": 30977 + }, + { + "epoch": 0.73, + "grad_norm": 2.751268188655102, + "learning_rate": 3.590457322666997e-06, + "loss": 0.8764, + "step": 30978 + }, + { + "epoch": 0.73, + "grad_norm": 2.113764729828117, + "learning_rate": 3.589871646980305e-06, + "loss": 0.845, + "step": 30979 + }, + { + "epoch": 0.73, + "grad_norm": 1.1013972220982329, + "learning_rate": 3.5892860086154103e-06, + "loss": 0.9673, + "step": 30980 + }, + { + "epoch": 0.73, + "grad_norm": 1.8635103624163454, + "learning_rate": 3.588700407575725e-06, + "loss": 0.9337, + "step": 30981 + }, + { + "epoch": 0.73, + "grad_norm": 1.768077829498413, + "learning_rate": 3.588114843864655e-06, + "loss": 0.938, + "step": 30982 + }, + { + "epoch": 0.73, + "grad_norm": 2.249824213809962, + "learning_rate": 3.5875293174856073e-06, + "loss": 0.8656, + "step": 30983 + }, + { + "epoch": 0.73, + "grad_norm": 1.9635252510256491, + "learning_rate": 3.5869438284419946e-06, + "loss": 1.0012, + "step": 30984 + }, + { + "epoch": 0.73, + "grad_norm": 2.0824846547639795, + "learning_rate": 3.586358376737231e-06, + "loss": 1.0161, + "step": 30985 + }, + { + "epoch": 0.73, + "grad_norm": 2.0649135227580366, + "learning_rate": 3.585772962374714e-06, + "loss": 0.9844, + "step": 30986 + }, + { + "epoch": 0.73, + "grad_norm": 2.454277331607442, + "learning_rate": 3.5851875853578568e-06, + "loss": 1.0372, + "step": 30987 + }, + { + "epoch": 0.73, + "grad_norm": 1.9182202398628636, + "learning_rate": 3.584602245690069e-06, + "loss": 1.0851, + "step": 30988 + }, + { + "epoch": 0.73, + "grad_norm": 2.066465887924458, + "learning_rate": 3.584016943374764e-06, + "loss": 0.9571, + "step": 30989 + }, + { + "epoch": 0.73, + "grad_norm": 2.136221137871119, + "learning_rate": 3.583431678415337e-06, + "loss": 1.0314, + "step": 30990 + }, + { + "epoch": 0.73, + "grad_norm": 1.9038572132753944, + "learning_rate": 3.582846450815202e-06, + "loss": 1.0129, + "step": 30991 + }, + { + "epoch": 0.73, + "grad_norm": 2.202871776673276, + "learning_rate": 3.5822612605777706e-06, + "loss": 1.1394, + "step": 30992 + }, + { + "epoch": 0.73, + "grad_norm": 2.130010221173922, + "learning_rate": 3.581676107706442e-06, + "loss": 1.147, + "step": 30993 + }, + { + "epoch": 0.73, + "grad_norm": 1.8588114312514399, + "learning_rate": 3.581090992204631e-06, + "loss": 1.0115, + "step": 30994 + }, + { + "epoch": 0.73, + "grad_norm": 1.9218615068905989, + "learning_rate": 3.580505914075737e-06, + "loss": 0.9987, + "step": 30995 + }, + { + "epoch": 0.73, + "grad_norm": 1.9911027160269554, + "learning_rate": 3.5799208733231727e-06, + "loss": 1.0455, + "step": 30996 + }, + { + "epoch": 0.73, + "grad_norm": 2.8102754575499067, + "learning_rate": 3.579335869950339e-06, + "loss": 1.0358, + "step": 30997 + }, + { + "epoch": 0.73, + "grad_norm": 2.210942801647772, + "learning_rate": 3.578750903960647e-06, + "loss": 0.9025, + "step": 30998 + }, + { + "epoch": 0.73, + "grad_norm": 2.1945546796226325, + "learning_rate": 3.5781659753575015e-06, + "loss": 0.999, + "step": 30999 + }, + { + "epoch": 0.73, + "grad_norm": 1.9762001765672, + "learning_rate": 3.5775810841443025e-06, + "loss": 0.902, + "step": 31000 + }, + { + "epoch": 0.73, + "grad_norm": 1.0279508413297773, + "learning_rate": 3.5769962303244597e-06, + "loss": 0.8538, + "step": 31001 + }, + { + "epoch": 0.73, + "grad_norm": 1.953918761368115, + "learning_rate": 3.5764114139013828e-06, + "loss": 0.8674, + "step": 31002 + }, + { + "epoch": 0.73, + "grad_norm": 1.938425413653201, + "learning_rate": 3.575826634878471e-06, + "loss": 0.8649, + "step": 31003 + }, + { + "epoch": 0.73, + "grad_norm": 2.3198800051663215, + "learning_rate": 3.575241893259128e-06, + "loss": 1.016, + "step": 31004 + }, + { + "epoch": 0.73, + "grad_norm": 2.325881292164371, + "learning_rate": 3.5746571890467607e-06, + "loss": 1.0369, + "step": 31005 + }, + { + "epoch": 0.73, + "grad_norm": 1.9544184639638156, + "learning_rate": 3.574072522244776e-06, + "loss": 0.9172, + "step": 31006 + }, + { + "epoch": 0.73, + "grad_norm": 2.6159020411869176, + "learning_rate": 3.5734878928565763e-06, + "loss": 1.0081, + "step": 31007 + }, + { + "epoch": 0.73, + "grad_norm": 1.1375262654659728, + "learning_rate": 3.572903300885561e-06, + "loss": 0.9704, + "step": 31008 + }, + { + "epoch": 0.73, + "grad_norm": 2.121908862748087, + "learning_rate": 3.572318746335137e-06, + "loss": 0.9119, + "step": 31009 + }, + { + "epoch": 0.73, + "grad_norm": 1.8474441730715805, + "learning_rate": 3.571734229208712e-06, + "loss": 0.898, + "step": 31010 + }, + { + "epoch": 0.73, + "grad_norm": 2.0010370194883853, + "learning_rate": 3.571149749509686e-06, + "loss": 0.919, + "step": 31011 + }, + { + "epoch": 0.73, + "grad_norm": 1.1990870900912873, + "learning_rate": 3.570565307241457e-06, + "loss": 0.9119, + "step": 31012 + }, + { + "epoch": 0.73, + "grad_norm": 1.9045082503161617, + "learning_rate": 3.5699809024074373e-06, + "loss": 1.0727, + "step": 31013 + }, + { + "epoch": 0.73, + "grad_norm": 2.0275474773400846, + "learning_rate": 3.5693965350110194e-06, + "loss": 0.9653, + "step": 31014 + }, + { + "epoch": 0.73, + "grad_norm": 2.0338450752717345, + "learning_rate": 3.568812205055615e-06, + "loss": 1.0198, + "step": 31015 + }, + { + "epoch": 0.73, + "grad_norm": 2.267444107517071, + "learning_rate": 3.5682279125446185e-06, + "loss": 0.9979, + "step": 31016 + }, + { + "epoch": 0.73, + "grad_norm": 2.057268327442805, + "learning_rate": 3.567643657481439e-06, + "loss": 0.9629, + "step": 31017 + }, + { + "epoch": 0.73, + "grad_norm": 1.8095969526529616, + "learning_rate": 3.5670594398694713e-06, + "loss": 0.9447, + "step": 31018 + }, + { + "epoch": 0.73, + "grad_norm": 1.8344600942068372, + "learning_rate": 3.5664752597121233e-06, + "loss": 1.1097, + "step": 31019 + }, + { + "epoch": 0.73, + "grad_norm": 1.9947525622256097, + "learning_rate": 3.565891117012793e-06, + "loss": 0.9397, + "step": 31020 + }, + { + "epoch": 0.73, + "grad_norm": 1.958626614745856, + "learning_rate": 3.5653070117748777e-06, + "loss": 0.9744, + "step": 31021 + }, + { + "epoch": 0.73, + "grad_norm": 2.1828393499865553, + "learning_rate": 3.5647229440017827e-06, + "loss": 0.9959, + "step": 31022 + }, + { + "epoch": 0.73, + "grad_norm": 1.9775960912252109, + "learning_rate": 3.564138913696912e-06, + "loss": 1.035, + "step": 31023 + }, + { + "epoch": 0.73, + "grad_norm": 2.0262515525111264, + "learning_rate": 3.5635549208636623e-06, + "loss": 1.032, + "step": 31024 + }, + { + "epoch": 0.73, + "grad_norm": 1.932980949778522, + "learning_rate": 3.562970965505429e-06, + "loss": 0.8458, + "step": 31025 + }, + { + "epoch": 0.73, + "grad_norm": 1.8329296884212978, + "learning_rate": 3.5623870476256174e-06, + "loss": 1.0368, + "step": 31026 + }, + { + "epoch": 0.73, + "grad_norm": 2.5189656066716646, + "learning_rate": 3.56180316722763e-06, + "loss": 1.0142, + "step": 31027 + }, + { + "epoch": 0.73, + "grad_norm": 1.0871671166130608, + "learning_rate": 3.5612193243148617e-06, + "loss": 0.9012, + "step": 31028 + }, + { + "epoch": 0.73, + "grad_norm": 1.9270747062397637, + "learning_rate": 3.5606355188907116e-06, + "loss": 0.9962, + "step": 31029 + }, + { + "epoch": 0.73, + "grad_norm": 2.1522230853978987, + "learning_rate": 3.5600517509585784e-06, + "loss": 1.0694, + "step": 31030 + }, + { + "epoch": 0.73, + "grad_norm": 2.0477918302140687, + "learning_rate": 3.5594680205218667e-06, + "loss": 1.066, + "step": 31031 + }, + { + "epoch": 0.73, + "grad_norm": 1.895622239646166, + "learning_rate": 3.558884327583971e-06, + "loss": 0.9061, + "step": 31032 + }, + { + "epoch": 0.73, + "grad_norm": 1.950222901405599, + "learning_rate": 3.5583006721482873e-06, + "loss": 0.854, + "step": 31033 + }, + { + "epoch": 0.73, + "grad_norm": 2.2102187468703276, + "learning_rate": 3.5577170542182195e-06, + "loss": 1.0855, + "step": 31034 + }, + { + "epoch": 0.73, + "grad_norm": 2.1626081714042886, + "learning_rate": 3.5571334737971596e-06, + "loss": 1.1274, + "step": 31035 + }, + { + "epoch": 0.73, + "grad_norm": 1.9690198062990611, + "learning_rate": 3.556549930888511e-06, + "loss": 0.9536, + "step": 31036 + }, + { + "epoch": 0.73, + "grad_norm": 2.0329776364274013, + "learning_rate": 3.555966425495666e-06, + "loss": 0.9285, + "step": 31037 + }, + { + "epoch": 0.73, + "grad_norm": 2.0061117811335647, + "learning_rate": 3.5553829576220277e-06, + "loss": 1.0679, + "step": 31038 + }, + { + "epoch": 0.73, + "grad_norm": 2.0122551244641023, + "learning_rate": 3.554799527270987e-06, + "loss": 0.9525, + "step": 31039 + }, + { + "epoch": 0.73, + "grad_norm": 1.1342638804613383, + "learning_rate": 3.554216134445947e-06, + "loss": 0.9077, + "step": 31040 + }, + { + "epoch": 0.73, + "grad_norm": 2.185641716982168, + "learning_rate": 3.553632779150301e-06, + "loss": 0.9425, + "step": 31041 + }, + { + "epoch": 0.73, + "grad_norm": 2.250123273101768, + "learning_rate": 3.5530494613874433e-06, + "loss": 0.7867, + "step": 31042 + }, + { + "epoch": 0.73, + "grad_norm": 1.9726298287954525, + "learning_rate": 3.552466181160772e-06, + "loss": 1.1025, + "step": 31043 + }, + { + "epoch": 0.73, + "grad_norm": 2.1768537999070117, + "learning_rate": 3.5518829384736888e-06, + "loss": 0.969, + "step": 31044 + }, + { + "epoch": 0.73, + "grad_norm": 2.0740740612449824, + "learning_rate": 3.551299733329583e-06, + "loss": 0.8423, + "step": 31045 + }, + { + "epoch": 0.73, + "grad_norm": 2.1542072823303218, + "learning_rate": 3.5507165657318487e-06, + "loss": 1.0087, + "step": 31046 + }, + { + "epoch": 0.73, + "grad_norm": 1.9298724959863778, + "learning_rate": 3.5501334356838846e-06, + "loss": 0.9528, + "step": 31047 + }, + { + "epoch": 0.73, + "grad_norm": 2.135956817592706, + "learning_rate": 3.549550343189089e-06, + "loss": 0.9003, + "step": 31048 + }, + { + "epoch": 0.73, + "grad_norm": 2.025690041314649, + "learning_rate": 3.548967288250853e-06, + "loss": 1.0943, + "step": 31049 + }, + { + "epoch": 0.73, + "grad_norm": 2.17810622160399, + "learning_rate": 3.548384270872568e-06, + "loss": 0.9762, + "step": 31050 + }, + { + "epoch": 0.73, + "grad_norm": 2.163763827842237, + "learning_rate": 3.5478012910576332e-06, + "loss": 1.0479, + "step": 31051 + }, + { + "epoch": 0.73, + "grad_norm": 2.7887879381421596, + "learning_rate": 3.547218348809446e-06, + "loss": 0.9655, + "step": 31052 + }, + { + "epoch": 0.73, + "grad_norm": 1.896139615074536, + "learning_rate": 3.5466354441313954e-06, + "loss": 0.9682, + "step": 31053 + }, + { + "epoch": 0.73, + "grad_norm": 1.990002731552635, + "learning_rate": 3.5460525770268726e-06, + "loss": 0.9884, + "step": 31054 + }, + { + "epoch": 0.73, + "grad_norm": 2.140392877900268, + "learning_rate": 3.5454697474992796e-06, + "loss": 1.019, + "step": 31055 + }, + { + "epoch": 0.73, + "grad_norm": 1.9584128076319638, + "learning_rate": 3.5448869555520016e-06, + "loss": 0.9993, + "step": 31056 + }, + { + "epoch": 0.73, + "grad_norm": 2.245441374264969, + "learning_rate": 3.5443042011884377e-06, + "loss": 0.9284, + "step": 31057 + }, + { + "epoch": 0.73, + "grad_norm": 2.137498987513785, + "learning_rate": 3.543721484411976e-06, + "loss": 0.9155, + "step": 31058 + }, + { + "epoch": 0.73, + "grad_norm": 2.0050640319855293, + "learning_rate": 3.543138805226015e-06, + "loss": 0.9757, + "step": 31059 + }, + { + "epoch": 0.73, + "grad_norm": 1.0406200768371132, + "learning_rate": 3.542556163633941e-06, + "loss": 0.8917, + "step": 31060 + }, + { + "epoch": 0.73, + "grad_norm": 2.8409793175062674, + "learning_rate": 3.5419735596391537e-06, + "loss": 0.9866, + "step": 31061 + }, + { + "epoch": 0.73, + "grad_norm": 1.8934738810591893, + "learning_rate": 3.5413909932450398e-06, + "loss": 0.9214, + "step": 31062 + }, + { + "epoch": 0.73, + "grad_norm": 1.7329257814977155, + "learning_rate": 3.540808464454989e-06, + "loss": 0.9953, + "step": 31063 + }, + { + "epoch": 0.73, + "grad_norm": 1.964460307781539, + "learning_rate": 3.5402259732723953e-06, + "loss": 1.2077, + "step": 31064 + }, + { + "epoch": 0.73, + "grad_norm": 2.7288658101091845, + "learning_rate": 3.539643519700653e-06, + "loss": 0.9956, + "step": 31065 + }, + { + "epoch": 0.73, + "grad_norm": 2.007345589431436, + "learning_rate": 3.5390611037431564e-06, + "loss": 0.9815, + "step": 31066 + }, + { + "epoch": 0.73, + "grad_norm": 2.0745523973756024, + "learning_rate": 3.538478725403286e-06, + "loss": 0.9739, + "step": 31067 + }, + { + "epoch": 0.73, + "grad_norm": 1.1485299155134616, + "learning_rate": 3.5378963846844383e-06, + "loss": 0.9584, + "step": 31068 + }, + { + "epoch": 0.73, + "grad_norm": 2.048662131670057, + "learning_rate": 3.5373140815900065e-06, + "loss": 0.9806, + "step": 31069 + }, + { + "epoch": 0.73, + "grad_norm": 2.601038024346889, + "learning_rate": 3.5367318161233777e-06, + "loss": 0.8812, + "step": 31070 + }, + { + "epoch": 0.73, + "grad_norm": 2.120703866859766, + "learning_rate": 3.5361495882879406e-06, + "loss": 1.0933, + "step": 31071 + }, + { + "epoch": 0.73, + "grad_norm": 1.9238418623995341, + "learning_rate": 3.5355673980870853e-06, + "loss": 0.9872, + "step": 31072 + }, + { + "epoch": 0.73, + "grad_norm": 1.9513578446216795, + "learning_rate": 3.5349852455242086e-06, + "loss": 0.928, + "step": 31073 + }, + { + "epoch": 0.73, + "grad_norm": 2.4698606947151514, + "learning_rate": 3.534403130602693e-06, + "loss": 0.8722, + "step": 31074 + }, + { + "epoch": 0.73, + "grad_norm": 1.9210453937210843, + "learning_rate": 3.5338210533259264e-06, + "loss": 1.155, + "step": 31075 + }, + { + "epoch": 0.73, + "grad_norm": 1.883117438266781, + "learning_rate": 3.5332390136973036e-06, + "loss": 1.0529, + "step": 31076 + }, + { + "epoch": 0.73, + "grad_norm": 2.7841745169647436, + "learning_rate": 3.5326570117202076e-06, + "loss": 1.0753, + "step": 31077 + }, + { + "epoch": 0.73, + "grad_norm": 1.77794274707263, + "learning_rate": 3.5320750473980304e-06, + "loss": 0.9529, + "step": 31078 + }, + { + "epoch": 0.73, + "grad_norm": 2.267588827679222, + "learning_rate": 3.531493120734163e-06, + "loss": 0.9339, + "step": 31079 + }, + { + "epoch": 0.73, + "grad_norm": 2.633859824882057, + "learning_rate": 3.5309112317319916e-06, + "loss": 0.887, + "step": 31080 + }, + { + "epoch": 0.73, + "grad_norm": 1.9198553607033573, + "learning_rate": 3.530329380394898e-06, + "loss": 1.1408, + "step": 31081 + }, + { + "epoch": 0.73, + "grad_norm": 1.8923262874164732, + "learning_rate": 3.529747566726276e-06, + "loss": 0.9553, + "step": 31082 + }, + { + "epoch": 0.73, + "grad_norm": 1.6599352462307388, + "learning_rate": 3.5291657907295184e-06, + "loss": 0.9554, + "step": 31083 + }, + { + "epoch": 0.73, + "grad_norm": 1.8813738860710885, + "learning_rate": 3.5285840524079994e-06, + "loss": 0.9194, + "step": 31084 + }, + { + "epoch": 0.73, + "grad_norm": 1.0992539629281346, + "learning_rate": 3.528002351765114e-06, + "loss": 0.9103, + "step": 31085 + }, + { + "epoch": 0.73, + "grad_norm": 2.0839439406489926, + "learning_rate": 3.527420688804247e-06, + "loss": 0.9883, + "step": 31086 + }, + { + "epoch": 0.73, + "grad_norm": 1.9250753331944972, + "learning_rate": 3.526839063528793e-06, + "loss": 1.0205, + "step": 31087 + }, + { + "epoch": 0.73, + "grad_norm": 2.061854992690961, + "learning_rate": 3.5262574759421243e-06, + "loss": 1.2178, + "step": 31088 + }, + { + "epoch": 0.73, + "grad_norm": 2.000055173815081, + "learning_rate": 3.525675926047634e-06, + "loss": 0.8292, + "step": 31089 + }, + { + "epoch": 0.73, + "grad_norm": 1.0573943707824651, + "learning_rate": 3.525094413848712e-06, + "loss": 0.8617, + "step": 31090 + }, + { + "epoch": 0.73, + "grad_norm": 2.1497452541187374, + "learning_rate": 3.5245129393487365e-06, + "loss": 0.99, + "step": 31091 + }, + { + "epoch": 0.73, + "grad_norm": 1.1357181662153715, + "learning_rate": 3.5239315025511e-06, + "loss": 1.0002, + "step": 31092 + }, + { + "epoch": 0.73, + "grad_norm": 1.1806471009373134, + "learning_rate": 3.523350103459182e-06, + "loss": 0.9058, + "step": 31093 + }, + { + "epoch": 0.73, + "grad_norm": 1.9519363895709958, + "learning_rate": 3.522768742076372e-06, + "loss": 1.0844, + "step": 31094 + }, + { + "epoch": 0.73, + "grad_norm": 2.268048575755188, + "learning_rate": 3.5221874184060513e-06, + "loss": 0.9966, + "step": 31095 + }, + { + "epoch": 0.73, + "grad_norm": 2.34770760464157, + "learning_rate": 3.5216061324516093e-06, + "loss": 0.9499, + "step": 31096 + }, + { + "epoch": 0.73, + "grad_norm": 1.8792737709925447, + "learning_rate": 3.5210248842164274e-06, + "loss": 1.0747, + "step": 31097 + }, + { + "epoch": 0.73, + "grad_norm": 2.0634750870800023, + "learning_rate": 3.520443673703886e-06, + "loss": 0.9092, + "step": 31098 + }, + { + "epoch": 0.73, + "grad_norm": 2.1083158984422625, + "learning_rate": 3.5198625009173726e-06, + "loss": 1.1112, + "step": 31099 + }, + { + "epoch": 0.73, + "grad_norm": 2.1299916251187145, + "learning_rate": 3.5192813658602754e-06, + "loss": 1.106, + "step": 31100 + }, + { + "epoch": 0.73, + "grad_norm": 1.9617066598904769, + "learning_rate": 3.518700268535974e-06, + "loss": 0.9275, + "step": 31101 + }, + { + "epoch": 0.73, + "grad_norm": 2.0726651032277066, + "learning_rate": 3.5181192089478477e-06, + "loss": 0.9484, + "step": 31102 + }, + { + "epoch": 0.73, + "grad_norm": 1.7867471675745854, + "learning_rate": 3.517538187099284e-06, + "loss": 0.846, + "step": 31103 + }, + { + "epoch": 0.73, + "grad_norm": 2.010837749625355, + "learning_rate": 3.5169572029936682e-06, + "loss": 1.0895, + "step": 31104 + }, + { + "epoch": 0.73, + "grad_norm": 1.9202806009344724, + "learning_rate": 3.516376256634381e-06, + "loss": 1.0601, + "step": 31105 + }, + { + "epoch": 0.73, + "grad_norm": 1.0168258401732746, + "learning_rate": 3.5157953480247985e-06, + "loss": 0.9275, + "step": 31106 + }, + { + "epoch": 0.73, + "grad_norm": 2.0725848008762635, + "learning_rate": 3.5152144771683107e-06, + "loss": 0.9756, + "step": 31107 + }, + { + "epoch": 0.73, + "grad_norm": 2.3167588009786595, + "learning_rate": 3.5146336440682995e-06, + "loss": 0.8762, + "step": 31108 + }, + { + "epoch": 0.73, + "grad_norm": 2.2886783316948116, + "learning_rate": 3.5140528487281443e-06, + "loss": 1.0281, + "step": 31109 + }, + { + "epoch": 0.73, + "grad_norm": 2.0892424946262937, + "learning_rate": 3.5134720911512233e-06, + "loss": 0.9191, + "step": 31110 + }, + { + "epoch": 0.73, + "grad_norm": 1.9800951543488845, + "learning_rate": 3.512891371340922e-06, + "loss": 1.0075, + "step": 31111 + }, + { + "epoch": 0.73, + "grad_norm": 1.8582021889148952, + "learning_rate": 3.5123106893006242e-06, + "loss": 1.0209, + "step": 31112 + }, + { + "epoch": 0.73, + "grad_norm": 2.4709890658127107, + "learning_rate": 3.5117300450337066e-06, + "loss": 0.7906, + "step": 31113 + }, + { + "epoch": 0.73, + "grad_norm": 2.050297409951291, + "learning_rate": 3.5111494385435484e-06, + "loss": 1.0023, + "step": 31114 + }, + { + "epoch": 0.73, + "grad_norm": 2.10715534876527, + "learning_rate": 3.5105688698335363e-06, + "loss": 1.0239, + "step": 31115 + }, + { + "epoch": 0.73, + "grad_norm": 1.6945458500632384, + "learning_rate": 3.509988338907042e-06, + "loss": 1.0418, + "step": 31116 + }, + { + "epoch": 0.73, + "grad_norm": 2.0960143787278915, + "learning_rate": 3.5094078457674553e-06, + "loss": 0.9992, + "step": 31117 + }, + { + "epoch": 0.73, + "grad_norm": 2.1123910446256304, + "learning_rate": 3.5088273904181457e-06, + "loss": 0.8687, + "step": 31118 + }, + { + "epoch": 0.73, + "grad_norm": 1.9934391243586849, + "learning_rate": 3.5082469728625034e-06, + "loss": 1.0113, + "step": 31119 + }, + { + "epoch": 0.73, + "grad_norm": 2.048429098901318, + "learning_rate": 3.5076665931038976e-06, + "loss": 1.0588, + "step": 31120 + }, + { + "epoch": 0.73, + "grad_norm": 1.8704218560241619, + "learning_rate": 3.5070862511457172e-06, + "loss": 1.0026, + "step": 31121 + }, + { + "epoch": 0.73, + "grad_norm": 2.3869029900158476, + "learning_rate": 3.5065059469913356e-06, + "loss": 0.9484, + "step": 31122 + }, + { + "epoch": 0.73, + "grad_norm": 2.064981692564537, + "learning_rate": 3.5059256806441276e-06, + "loss": 0.8582, + "step": 31123 + }, + { + "epoch": 0.73, + "grad_norm": 1.9377137403205003, + "learning_rate": 3.505345452107478e-06, + "loss": 0.966, + "step": 31124 + }, + { + "epoch": 0.73, + "grad_norm": 2.0328868912967515, + "learning_rate": 3.504765261384766e-06, + "loss": 0.9216, + "step": 31125 + }, + { + "epoch": 0.73, + "grad_norm": 1.9034224582359067, + "learning_rate": 3.5041851084793665e-06, + "loss": 0.9205, + "step": 31126 + }, + { + "epoch": 0.73, + "grad_norm": 1.828506162724789, + "learning_rate": 3.503604993394656e-06, + "loss": 0.9125, + "step": 31127 + }, + { + "epoch": 0.73, + "grad_norm": 1.819282502095058, + "learning_rate": 3.5030249161340124e-06, + "loss": 0.9361, + "step": 31128 + }, + { + "epoch": 0.73, + "grad_norm": 2.0686085031341896, + "learning_rate": 3.502444876700819e-06, + "loss": 1.0242, + "step": 31129 + }, + { + "epoch": 0.73, + "grad_norm": 1.954504137211363, + "learning_rate": 3.5018648750984473e-06, + "loss": 1.0013, + "step": 31130 + }, + { + "epoch": 0.73, + "grad_norm": 1.8628328382842543, + "learning_rate": 3.5012849113302737e-06, + "loss": 1.0741, + "step": 31131 + }, + { + "epoch": 0.73, + "grad_norm": 1.9459456659076455, + "learning_rate": 3.5007049853996767e-06, + "loss": 1.0204, + "step": 31132 + }, + { + "epoch": 0.73, + "grad_norm": 2.371332379474832, + "learning_rate": 3.5001250973100353e-06, + "loss": 0.8942, + "step": 31133 + }, + { + "epoch": 0.73, + "grad_norm": 1.94272892372599, + "learning_rate": 3.499545247064724e-06, + "loss": 0.8515, + "step": 31134 + }, + { + "epoch": 0.73, + "grad_norm": 1.9433201297674312, + "learning_rate": 3.4989654346671143e-06, + "loss": 0.9791, + "step": 31135 + }, + { + "epoch": 0.73, + "grad_norm": 1.9235791373212983, + "learning_rate": 3.4983856601205903e-06, + "loss": 1.0054, + "step": 31136 + }, + { + "epoch": 0.73, + "grad_norm": 1.978446243976982, + "learning_rate": 3.49780592342852e-06, + "loss": 0.974, + "step": 31137 + }, + { + "epoch": 0.73, + "grad_norm": 1.9538406893056468, + "learning_rate": 3.4972262245942847e-06, + "loss": 0.9743, + "step": 31138 + }, + { + "epoch": 0.73, + "grad_norm": 1.0787560283012712, + "learning_rate": 3.4966465636212543e-06, + "loss": 0.9874, + "step": 31139 + }, + { + "epoch": 0.73, + "grad_norm": 1.062165355092745, + "learning_rate": 3.49606694051281e-06, + "loss": 0.9689, + "step": 31140 + }, + { + "epoch": 0.73, + "grad_norm": 1.920178157867435, + "learning_rate": 3.4954873552723188e-06, + "loss": 0.8535, + "step": 31141 + }, + { + "epoch": 0.73, + "grad_norm": 1.9267618858523459, + "learning_rate": 3.494907807903164e-06, + "loss": 0.8977, + "step": 31142 + }, + { + "epoch": 0.73, + "grad_norm": 2.8216872063904486, + "learning_rate": 3.4943282984087147e-06, + "loss": 0.9622, + "step": 31143 + }, + { + "epoch": 0.73, + "grad_norm": 2.100375595949314, + "learning_rate": 3.4937488267923426e-06, + "loss": 0.9341, + "step": 31144 + }, + { + "epoch": 0.73, + "grad_norm": 1.7848824173241224, + "learning_rate": 3.493169393057425e-06, + "loss": 0.9279, + "step": 31145 + }, + { + "epoch": 0.73, + "grad_norm": 1.9307839067471237, + "learning_rate": 3.4925899972073387e-06, + "loss": 0.9701, + "step": 31146 + }, + { + "epoch": 0.73, + "grad_norm": 1.0236862558023836, + "learning_rate": 3.4920106392454523e-06, + "loss": 0.989, + "step": 31147 + }, + { + "epoch": 0.73, + "grad_norm": 1.8397002534983562, + "learning_rate": 3.491431319175138e-06, + "loss": 0.9104, + "step": 31148 + }, + { + "epoch": 0.73, + "grad_norm": 3.0088755615445746, + "learning_rate": 3.490852036999771e-06, + "loss": 0.9937, + "step": 31149 + }, + { + "epoch": 0.73, + "grad_norm": 2.302360764719073, + "learning_rate": 3.4902727927227277e-06, + "loss": 0.9615, + "step": 31150 + }, + { + "epoch": 0.73, + "grad_norm": 1.8290044557137872, + "learning_rate": 3.489693586347378e-06, + "loss": 1.0075, + "step": 31151 + }, + { + "epoch": 0.73, + "grad_norm": 2.0538441896630175, + "learning_rate": 3.4891144178770885e-06, + "loss": 0.9986, + "step": 31152 + }, + { + "epoch": 0.73, + "grad_norm": 1.7879420469443232, + "learning_rate": 3.488535287315237e-06, + "loss": 1.001, + "step": 31153 + }, + { + "epoch": 0.73, + "grad_norm": 2.183334337575865, + "learning_rate": 3.4879561946651974e-06, + "loss": 0.9962, + "step": 31154 + }, + { + "epoch": 0.73, + "grad_norm": 2.0427789397073948, + "learning_rate": 3.487377139930339e-06, + "loss": 0.9016, + "step": 31155 + }, + { + "epoch": 0.73, + "grad_norm": 1.959950561809589, + "learning_rate": 3.486798123114029e-06, + "loss": 1.034, + "step": 31156 + }, + { + "epoch": 0.73, + "grad_norm": 2.5423466920319853, + "learning_rate": 3.486219144219646e-06, + "loss": 1.0443, + "step": 31157 + }, + { + "epoch": 0.73, + "grad_norm": 1.9437300865827014, + "learning_rate": 3.485640203250553e-06, + "loss": 0.8649, + "step": 31158 + }, + { + "epoch": 0.73, + "grad_norm": 2.012469263723182, + "learning_rate": 3.4850613002101295e-06, + "loss": 0.9226, + "step": 31159 + }, + { + "epoch": 0.73, + "grad_norm": 1.9287283999219913, + "learning_rate": 3.4844824351017394e-06, + "loss": 0.8555, + "step": 31160 + }, + { + "epoch": 0.73, + "grad_norm": 1.9770113767256903, + "learning_rate": 3.483903607928757e-06, + "loss": 0.9884, + "step": 31161 + }, + { + "epoch": 0.73, + "grad_norm": 1.10479834177037, + "learning_rate": 3.4833248186945477e-06, + "loss": 0.9154, + "step": 31162 + }, + { + "epoch": 0.73, + "grad_norm": 1.0691835890418955, + "learning_rate": 3.4827460674024847e-06, + "loss": 0.9548, + "step": 31163 + }, + { + "epoch": 0.73, + "grad_norm": 1.9700975300441732, + "learning_rate": 3.4821673540559442e-06, + "loss": 0.9897, + "step": 31164 + }, + { + "epoch": 0.73, + "grad_norm": 1.7220016319532818, + "learning_rate": 3.481588678658282e-06, + "loss": 0.9275, + "step": 31165 + }, + { + "epoch": 0.73, + "grad_norm": 2.245204233898019, + "learning_rate": 3.4810100412128743e-06, + "loss": 0.9413, + "step": 31166 + }, + { + "epoch": 0.73, + "grad_norm": 2.117088816492196, + "learning_rate": 3.48043144172309e-06, + "loss": 0.9428, + "step": 31167 + }, + { + "epoch": 0.73, + "grad_norm": 2.218945523635056, + "learning_rate": 3.479852880192305e-06, + "loss": 0.9452, + "step": 31168 + }, + { + "epoch": 0.73, + "grad_norm": 2.0552897711718088, + "learning_rate": 3.479274356623874e-06, + "loss": 1.0279, + "step": 31169 + }, + { + "epoch": 0.73, + "grad_norm": 2.0239844589792546, + "learning_rate": 3.4786958710211714e-06, + "loss": 0.9845, + "step": 31170 + }, + { + "epoch": 0.73, + "grad_norm": 1.8330586253110972, + "learning_rate": 3.4781174233875705e-06, + "loss": 0.9004, + "step": 31171 + }, + { + "epoch": 0.73, + "grad_norm": 1.842105650036429, + "learning_rate": 3.477539013726433e-06, + "loss": 1.0009, + "step": 31172 + }, + { + "epoch": 0.73, + "grad_norm": 1.883976187998154, + "learning_rate": 3.4769606420411273e-06, + "loss": 0.9902, + "step": 31173 + }, + { + "epoch": 0.73, + "grad_norm": 1.8763059012254835, + "learning_rate": 3.4763823083350202e-06, + "loss": 1.1114, + "step": 31174 + }, + { + "epoch": 0.73, + "grad_norm": 3.0302945454999777, + "learning_rate": 3.4758040126114845e-06, + "loss": 1.0763, + "step": 31175 + }, + { + "epoch": 0.73, + "grad_norm": 2.162089286149179, + "learning_rate": 3.47522575487388e-06, + "loss": 1.0368, + "step": 31176 + }, + { + "epoch": 0.73, + "grad_norm": 1.943208999662484, + "learning_rate": 3.4746475351255806e-06, + "loss": 0.962, + "step": 31177 + }, + { + "epoch": 0.73, + "grad_norm": 2.263239694618094, + "learning_rate": 3.474069353369949e-06, + "loss": 1.0432, + "step": 31178 + }, + { + "epoch": 0.73, + "grad_norm": 2.4142636892166807, + "learning_rate": 3.473491209610347e-06, + "loss": 0.9107, + "step": 31179 + }, + { + "epoch": 0.73, + "grad_norm": 1.9630776126485041, + "learning_rate": 3.4729131038501475e-06, + "loss": 0.9965, + "step": 31180 + }, + { + "epoch": 0.73, + "grad_norm": 1.0585861190383612, + "learning_rate": 3.4723350360927165e-06, + "loss": 0.8408, + "step": 31181 + }, + { + "epoch": 0.73, + "grad_norm": 1.9399162561373966, + "learning_rate": 3.4717570063414174e-06, + "loss": 0.9132, + "step": 31182 + }, + { + "epoch": 0.73, + "grad_norm": 2.237643910295683, + "learning_rate": 3.4711790145996127e-06, + "loss": 0.9018, + "step": 31183 + }, + { + "epoch": 0.73, + "grad_norm": 2.115098612561537, + "learning_rate": 3.470601060870671e-06, + "loss": 0.9354, + "step": 31184 + }, + { + "epoch": 0.73, + "grad_norm": 2.0491874230226697, + "learning_rate": 3.4700231451579634e-06, + "loss": 1.0562, + "step": 31185 + }, + { + "epoch": 0.73, + "grad_norm": 2.2973718655189725, + "learning_rate": 3.4694452674648425e-06, + "loss": 0.9742, + "step": 31186 + }, + { + "epoch": 0.73, + "grad_norm": 1.9554301157259926, + "learning_rate": 3.4688674277946775e-06, + "loss": 1.0858, + "step": 31187 + }, + { + "epoch": 0.73, + "grad_norm": 2.0331774638074602, + "learning_rate": 3.468289626150836e-06, + "loss": 1.0455, + "step": 31188 + }, + { + "epoch": 0.73, + "grad_norm": 2.1455434305510095, + "learning_rate": 3.4677118625366825e-06, + "loss": 1.0706, + "step": 31189 + }, + { + "epoch": 0.73, + "grad_norm": 2.410255040693652, + "learning_rate": 3.4671341369555776e-06, + "loss": 0.918, + "step": 31190 + }, + { + "epoch": 0.73, + "grad_norm": 2.02200216075132, + "learning_rate": 3.4665564494108837e-06, + "loss": 1.0724, + "step": 31191 + }, + { + "epoch": 0.73, + "grad_norm": 2.019191561885005, + "learning_rate": 3.46597879990597e-06, + "loss": 0.9536, + "step": 31192 + }, + { + "epoch": 0.73, + "grad_norm": 1.8925774865105087, + "learning_rate": 3.465401188444193e-06, + "loss": 1.1023, + "step": 31193 + }, + { + "epoch": 0.73, + "grad_norm": 1.0599381417061458, + "learning_rate": 3.464823615028923e-06, + "loss": 0.9329, + "step": 31194 + }, + { + "epoch": 0.73, + "grad_norm": 1.9663448557743815, + "learning_rate": 3.4642460796635146e-06, + "loss": 0.9268, + "step": 31195 + }, + { + "epoch": 0.73, + "grad_norm": 1.9318245277332031, + "learning_rate": 3.4636685823513384e-06, + "loss": 1.0588, + "step": 31196 + }, + { + "epoch": 0.73, + "grad_norm": 3.1911277842840735, + "learning_rate": 3.4630911230957486e-06, + "loss": 0.9796, + "step": 31197 + }, + { + "epoch": 0.74, + "grad_norm": 1.8737804002789658, + "learning_rate": 3.4625137019001164e-06, + "loss": 1.0444, + "step": 31198 + }, + { + "epoch": 0.74, + "grad_norm": 1.8567050425936524, + "learning_rate": 3.461936318767799e-06, + "loss": 0.7809, + "step": 31199 + }, + { + "epoch": 0.74, + "grad_norm": 2.2612171239630054, + "learning_rate": 3.4613589737021557e-06, + "loss": 1.0077, + "step": 31200 + }, + { + "epoch": 0.74, + "grad_norm": 1.9862740691227396, + "learning_rate": 3.4607816667065497e-06, + "loss": 0.871, + "step": 31201 + }, + { + "epoch": 0.74, + "grad_norm": 2.2032452620873437, + "learning_rate": 3.460204397784347e-06, + "loss": 0.8897, + "step": 31202 + }, + { + "epoch": 0.74, + "grad_norm": 2.2130282974513813, + "learning_rate": 3.459627166938904e-06, + "loss": 0.8975, + "step": 31203 + }, + { + "epoch": 0.74, + "grad_norm": 1.904874225906291, + "learning_rate": 3.4590499741735796e-06, + "loss": 0.941, + "step": 31204 + }, + { + "epoch": 0.74, + "grad_norm": 1.9503633959661297, + "learning_rate": 3.4584728194917363e-06, + "loss": 0.9774, + "step": 31205 + }, + { + "epoch": 0.74, + "grad_norm": 1.9155146843683704, + "learning_rate": 3.4578957028967388e-06, + "loss": 1.0138, + "step": 31206 + }, + { + "epoch": 0.74, + "grad_norm": 1.8859102774536218, + "learning_rate": 3.4573186243919443e-06, + "loss": 1.0569, + "step": 31207 + }, + { + "epoch": 0.74, + "grad_norm": 2.1224754629623623, + "learning_rate": 3.4567415839807083e-06, + "loss": 1.0905, + "step": 31208 + }, + { + "epoch": 0.74, + "grad_norm": 2.0162858200893368, + "learning_rate": 3.4561645816663945e-06, + "loss": 1.0162, + "step": 31209 + }, + { + "epoch": 0.74, + "grad_norm": 2.1104748186581066, + "learning_rate": 3.455587617452365e-06, + "loss": 0.8527, + "step": 31210 + }, + { + "epoch": 0.74, + "grad_norm": 1.890328532562195, + "learning_rate": 3.455010691341978e-06, + "loss": 0.9151, + "step": 31211 + }, + { + "epoch": 0.74, + "grad_norm": 1.9639519945126167, + "learning_rate": 3.454433803338586e-06, + "loss": 0.7507, + "step": 31212 + }, + { + "epoch": 0.74, + "grad_norm": 2.0214617441126865, + "learning_rate": 3.453856953445557e-06, + "loss": 1.1083, + "step": 31213 + }, + { + "epoch": 0.74, + "grad_norm": 1.978624349357654, + "learning_rate": 3.453280141666241e-06, + "loss": 1.0729, + "step": 31214 + }, + { + "epoch": 0.74, + "grad_norm": 2.6406293552795868, + "learning_rate": 3.4527033680040047e-06, + "loss": 0.857, + "step": 31215 + }, + { + "epoch": 0.74, + "grad_norm": 1.9132539209023853, + "learning_rate": 3.4521266324621983e-06, + "loss": 0.8764, + "step": 31216 + }, + { + "epoch": 0.74, + "grad_norm": 1.8629853047434466, + "learning_rate": 3.451549935044187e-06, + "loss": 1.1164, + "step": 31217 + }, + { + "epoch": 0.74, + "grad_norm": 2.414913686213901, + "learning_rate": 3.450973275753322e-06, + "loss": 0.9796, + "step": 31218 + }, + { + "epoch": 0.74, + "grad_norm": 1.8688160047388866, + "learning_rate": 3.450396654592968e-06, + "loss": 0.9356, + "step": 31219 + }, + { + "epoch": 0.74, + "grad_norm": 1.8683433069094084, + "learning_rate": 3.4498200715664775e-06, + "loss": 1.0186, + "step": 31220 + }, + { + "epoch": 0.74, + "grad_norm": 1.966306065575997, + "learning_rate": 3.449243526677205e-06, + "loss": 1.0569, + "step": 31221 + }, + { + "epoch": 0.74, + "grad_norm": 2.0743182709807964, + "learning_rate": 3.448667019928511e-06, + "loss": 0.9948, + "step": 31222 + }, + { + "epoch": 0.74, + "grad_norm": 1.7133360512484288, + "learning_rate": 3.448090551323755e-06, + "loss": 0.8954, + "step": 31223 + }, + { + "epoch": 0.74, + "grad_norm": 1.8939263936589115, + "learning_rate": 3.44751412086629e-06, + "loss": 0.8902, + "step": 31224 + }, + { + "epoch": 0.74, + "grad_norm": 2.2267180655117147, + "learning_rate": 3.4469377285594687e-06, + "loss": 0.9185, + "step": 31225 + }, + { + "epoch": 0.74, + "grad_norm": 2.0574707049039307, + "learning_rate": 3.4463613744066516e-06, + "loss": 1.0407, + "step": 31226 + }, + { + "epoch": 0.74, + "grad_norm": 2.1455762044435245, + "learning_rate": 3.445785058411195e-06, + "loss": 0.8924, + "step": 31227 + }, + { + "epoch": 0.74, + "grad_norm": 2.114615114218923, + "learning_rate": 3.4452087805764544e-06, + "loss": 1.047, + "step": 31228 + }, + { + "epoch": 0.74, + "grad_norm": 2.221612951303389, + "learning_rate": 3.4446325409057792e-06, + "loss": 1.0091, + "step": 31229 + }, + { + "epoch": 0.74, + "grad_norm": 2.009575692746697, + "learning_rate": 3.4440563394025294e-06, + "loss": 0.8974, + "step": 31230 + }, + { + "epoch": 0.74, + "grad_norm": 1.9771776424791982, + "learning_rate": 3.443480176070062e-06, + "loss": 0.9678, + "step": 31231 + }, + { + "epoch": 0.74, + "grad_norm": 2.0748481737828945, + "learning_rate": 3.4429040509117295e-06, + "loss": 0.9617, + "step": 31232 + }, + { + "epoch": 0.74, + "grad_norm": 1.8472583129387805, + "learning_rate": 3.4423279639308815e-06, + "loss": 1.017, + "step": 31233 + }, + { + "epoch": 0.74, + "grad_norm": 2.1162121913151, + "learning_rate": 3.4417519151308763e-06, + "loss": 0.9728, + "step": 31234 + }, + { + "epoch": 0.74, + "grad_norm": 2.3292349569240227, + "learning_rate": 3.441175904515072e-06, + "loss": 0.9608, + "step": 31235 + }, + { + "epoch": 0.74, + "grad_norm": 2.0930133253539562, + "learning_rate": 3.4405999320868165e-06, + "loss": 1.1009, + "step": 31236 + }, + { + "epoch": 0.74, + "grad_norm": 2.6973891553656015, + "learning_rate": 3.440023997849462e-06, + "loss": 0.9923, + "step": 31237 + }, + { + "epoch": 0.74, + "grad_norm": 1.835917641463827, + "learning_rate": 3.4394481018063685e-06, + "loss": 0.988, + "step": 31238 + }, + { + "epoch": 0.74, + "grad_norm": 2.411745402157496, + "learning_rate": 3.4388722439608822e-06, + "loss": 1.0286, + "step": 31239 + }, + { + "epoch": 0.74, + "grad_norm": 1.978881023275745, + "learning_rate": 3.438296424316362e-06, + "loss": 0.8646, + "step": 31240 + }, + { + "epoch": 0.74, + "grad_norm": 1.973125250107833, + "learning_rate": 3.4377206428761534e-06, + "loss": 0.9322, + "step": 31241 + }, + { + "epoch": 0.74, + "grad_norm": 1.1109398304301847, + "learning_rate": 3.4371448996436163e-06, + "loss": 0.9374, + "step": 31242 + }, + { + "epoch": 0.74, + "grad_norm": 1.1193721358690196, + "learning_rate": 3.436569194622096e-06, + "loss": 0.9521, + "step": 31243 + }, + { + "epoch": 0.74, + "grad_norm": 1.9006538096768053, + "learning_rate": 3.4359935278149513e-06, + "loss": 0.9956, + "step": 31244 + }, + { + "epoch": 0.74, + "grad_norm": 2.1436194737013454, + "learning_rate": 3.4354178992255306e-06, + "loss": 0.9782, + "step": 31245 + }, + { + "epoch": 0.74, + "grad_norm": 1.9140992847138774, + "learning_rate": 3.4348423088571815e-06, + "loss": 1.0035, + "step": 31246 + }, + { + "epoch": 0.74, + "grad_norm": 2.03978729395008, + "learning_rate": 3.4342667567132593e-06, + "loss": 1.0518, + "step": 31247 + }, + { + "epoch": 0.74, + "grad_norm": 2.0870573418481104, + "learning_rate": 3.433691242797118e-06, + "loss": 0.9692, + "step": 31248 + }, + { + "epoch": 0.74, + "grad_norm": 1.0514140883489949, + "learning_rate": 3.433115767112105e-06, + "loss": 0.9344, + "step": 31249 + }, + { + "epoch": 0.74, + "grad_norm": 2.4259084021872535, + "learning_rate": 3.4325403296615677e-06, + "loss": 0.9247, + "step": 31250 + }, + { + "epoch": 0.74, + "grad_norm": 2.4409837858892316, + "learning_rate": 3.43196493044886e-06, + "loss": 0.9807, + "step": 31251 + }, + { + "epoch": 0.74, + "grad_norm": 2.0467587791887567, + "learning_rate": 3.4313895694773357e-06, + "loss": 0.9934, + "step": 31252 + }, + { + "epoch": 0.74, + "grad_norm": 1.889476881815377, + "learning_rate": 3.4308142467503404e-06, + "loss": 0.9368, + "step": 31253 + }, + { + "epoch": 0.74, + "grad_norm": 1.9992293704620252, + "learning_rate": 3.4302389622712207e-06, + "loss": 0.9512, + "step": 31254 + }, + { + "epoch": 0.74, + "grad_norm": 1.9780243515975118, + "learning_rate": 3.42966371604333e-06, + "loss": 0.8993, + "step": 31255 + }, + { + "epoch": 0.74, + "grad_norm": 1.8428928513945855, + "learning_rate": 3.429088508070021e-06, + "loss": 0.9393, + "step": 31256 + }, + { + "epoch": 0.74, + "grad_norm": 2.001882538714098, + "learning_rate": 3.428513338354639e-06, + "loss": 1.0204, + "step": 31257 + }, + { + "epoch": 0.74, + "grad_norm": 1.7393498991379706, + "learning_rate": 3.4279382069005283e-06, + "loss": 1.0445, + "step": 31258 + }, + { + "epoch": 0.74, + "grad_norm": 1.9464204155792388, + "learning_rate": 3.4273631137110464e-06, + "loss": 1.0631, + "step": 31259 + }, + { + "epoch": 0.74, + "grad_norm": 2.0808017253409004, + "learning_rate": 3.4267880587895343e-06, + "loss": 1.0377, + "step": 31260 + }, + { + "epoch": 0.74, + "grad_norm": 1.9483433684029945, + "learning_rate": 3.4262130421393424e-06, + "loss": 0.9814, + "step": 31261 + }, + { + "epoch": 0.74, + "grad_norm": 2.4028018069871204, + "learning_rate": 3.425638063763823e-06, + "loss": 1.1156, + "step": 31262 + }, + { + "epoch": 0.74, + "grad_norm": 1.8444358389353956, + "learning_rate": 3.42506312366632e-06, + "loss": 0.9423, + "step": 31263 + }, + { + "epoch": 0.74, + "grad_norm": 1.8453162448838596, + "learning_rate": 3.4244882218501775e-06, + "loss": 1.0283, + "step": 31264 + }, + { + "epoch": 0.74, + "grad_norm": 1.9598536808329627, + "learning_rate": 3.423913358318747e-06, + "loss": 1.04, + "step": 31265 + }, + { + "epoch": 0.74, + "grad_norm": 2.1349782802905835, + "learning_rate": 3.423338533075381e-06, + "loss": 1.0627, + "step": 31266 + }, + { + "epoch": 0.74, + "grad_norm": 1.7171818764796691, + "learning_rate": 3.422763746123414e-06, + "loss": 0.9166, + "step": 31267 + }, + { + "epoch": 0.74, + "grad_norm": 2.1304179754088968, + "learning_rate": 3.422188997466199e-06, + "loss": 0.9134, + "step": 31268 + }, + { + "epoch": 0.74, + "grad_norm": 1.9617213266802782, + "learning_rate": 3.421614287107082e-06, + "loss": 0.9752, + "step": 31269 + }, + { + "epoch": 0.74, + "grad_norm": 1.9407009419077894, + "learning_rate": 3.4210396150494152e-06, + "loss": 0.9553, + "step": 31270 + }, + { + "epoch": 0.74, + "grad_norm": 1.8008652479229812, + "learning_rate": 3.4204649812965327e-06, + "loss": 1.0194, + "step": 31271 + }, + { + "epoch": 0.74, + "grad_norm": 2.0760293714964844, + "learning_rate": 3.4198903858517863e-06, + "loss": 1.069, + "step": 31272 + }, + { + "epoch": 0.74, + "grad_norm": 1.98366923223341, + "learning_rate": 3.4193158287185248e-06, + "loss": 0.8836, + "step": 31273 + }, + { + "epoch": 0.74, + "grad_norm": 1.8351542698737637, + "learning_rate": 3.4187413099000854e-06, + "loss": 0.9871, + "step": 31274 + }, + { + "epoch": 0.74, + "grad_norm": 2.2260015463885057, + "learning_rate": 3.4181668293998216e-06, + "loss": 1.0291, + "step": 31275 + }, + { + "epoch": 0.74, + "grad_norm": 2.2796633060358187, + "learning_rate": 3.4175923872210715e-06, + "loss": 1.0131, + "step": 31276 + }, + { + "epoch": 0.74, + "grad_norm": 1.0404963280625836, + "learning_rate": 3.4170179833671847e-06, + "loss": 0.9004, + "step": 31277 + }, + { + "epoch": 0.74, + "grad_norm": 1.9004870154565396, + "learning_rate": 3.4164436178415007e-06, + "loss": 0.9701, + "step": 31278 + }, + { + "epoch": 0.74, + "grad_norm": 1.736199641797551, + "learning_rate": 3.41586929064737e-06, + "loss": 0.9435, + "step": 31279 + }, + { + "epoch": 0.74, + "grad_norm": 2.288195270615825, + "learning_rate": 3.4152950017881325e-06, + "loss": 0.9739, + "step": 31280 + }, + { + "epoch": 0.74, + "grad_norm": 1.990519848491139, + "learning_rate": 3.414720751267129e-06, + "loss": 0.8555, + "step": 31281 + }, + { + "epoch": 0.74, + "grad_norm": 2.161829558799179, + "learning_rate": 3.414146539087706e-06, + "loss": 1.0628, + "step": 31282 + }, + { + "epoch": 0.74, + "grad_norm": 1.9676681903943736, + "learning_rate": 3.4135723652532093e-06, + "loss": 1.0708, + "step": 31283 + }, + { + "epoch": 0.74, + "grad_norm": 1.9280424857374618, + "learning_rate": 3.4129982297669807e-06, + "loss": 1.0284, + "step": 31284 + }, + { + "epoch": 0.74, + "grad_norm": 1.8779282971583797, + "learning_rate": 3.412424132632358e-06, + "loss": 1.0164, + "step": 31285 + }, + { + "epoch": 0.74, + "grad_norm": 2.5008592125613522, + "learning_rate": 3.411850073852688e-06, + "loss": 1.087, + "step": 31286 + }, + { + "epoch": 0.74, + "grad_norm": 2.0193552953032263, + "learning_rate": 3.411276053431315e-06, + "loss": 1.0737, + "step": 31287 + }, + { + "epoch": 0.74, + "grad_norm": 2.063182521117771, + "learning_rate": 3.41070207137158e-06, + "loss": 0.9244, + "step": 31288 + }, + { + "epoch": 0.74, + "grad_norm": 1.850137611349359, + "learning_rate": 3.4101281276768194e-06, + "loss": 0.9635, + "step": 31289 + }, + { + "epoch": 0.74, + "grad_norm": 1.9718073921167985, + "learning_rate": 3.4095542223503795e-06, + "loss": 0.9243, + "step": 31290 + }, + { + "epoch": 0.74, + "grad_norm": 2.0926282349917753, + "learning_rate": 3.4089803553956058e-06, + "loss": 0.9151, + "step": 31291 + }, + { + "epoch": 0.74, + "grad_norm": 1.7766766649503611, + "learning_rate": 3.4084065268158338e-06, + "loss": 0.8952, + "step": 31292 + }, + { + "epoch": 0.74, + "grad_norm": 1.9999832117923937, + "learning_rate": 3.4078327366144025e-06, + "loss": 1.001, + "step": 31293 + }, + { + "epoch": 0.74, + "grad_norm": 2.09399071308757, + "learning_rate": 3.4072589847946603e-06, + "loss": 0.9597, + "step": 31294 + }, + { + "epoch": 0.74, + "grad_norm": 2.0740835860587103, + "learning_rate": 3.40668527135994e-06, + "loss": 0.9014, + "step": 31295 + }, + { + "epoch": 0.74, + "grad_norm": 1.9381259323649993, + "learning_rate": 3.406111596313588e-06, + "loss": 0.9797, + "step": 31296 + }, + { + "epoch": 0.74, + "grad_norm": 2.0428642748601233, + "learning_rate": 3.4055379596589387e-06, + "loss": 0.9931, + "step": 31297 + }, + { + "epoch": 0.74, + "grad_norm": 1.950236391516901, + "learning_rate": 3.404964361399339e-06, + "loss": 1.1341, + "step": 31298 + }, + { + "epoch": 0.74, + "grad_norm": 2.158555721752809, + "learning_rate": 3.4043908015381212e-06, + "loss": 1.0581, + "step": 31299 + }, + { + "epoch": 0.74, + "grad_norm": 2.282477781970498, + "learning_rate": 3.4038172800786317e-06, + "loss": 1.107, + "step": 31300 + }, + { + "epoch": 0.74, + "grad_norm": 4.775272380768134, + "learning_rate": 3.403243797024206e-06, + "loss": 0.9739, + "step": 31301 + }, + { + "epoch": 0.74, + "grad_norm": 1.0908485866133502, + "learning_rate": 3.40267035237818e-06, + "loss": 0.964, + "step": 31302 + }, + { + "epoch": 0.74, + "grad_norm": 2.270420572286607, + "learning_rate": 3.4020969461438966e-06, + "loss": 1.2012, + "step": 31303 + }, + { + "epoch": 0.74, + "grad_norm": 1.9309134968602948, + "learning_rate": 3.401523578324697e-06, + "loss": 1.0052, + "step": 31304 + }, + { + "epoch": 0.74, + "grad_norm": 1.8810516639286512, + "learning_rate": 3.4009502489239157e-06, + "loss": 0.7757, + "step": 31305 + }, + { + "epoch": 0.74, + "grad_norm": 2.7868194991547517, + "learning_rate": 3.4003769579448875e-06, + "loss": 0.9737, + "step": 31306 + }, + { + "epoch": 0.74, + "grad_norm": 2.335957943705658, + "learning_rate": 3.399803705390955e-06, + "loss": 0.9889, + "step": 31307 + }, + { + "epoch": 0.74, + "grad_norm": 1.8114297961506696, + "learning_rate": 3.3992304912654573e-06, + "loss": 1.001, + "step": 31308 + }, + { + "epoch": 0.74, + "grad_norm": 1.9612404358488096, + "learning_rate": 3.3986573155717307e-06, + "loss": 0.8695, + "step": 31309 + }, + { + "epoch": 0.74, + "grad_norm": 1.8360143841334937, + "learning_rate": 3.3980841783131067e-06, + "loss": 0.9858, + "step": 31310 + }, + { + "epoch": 0.74, + "grad_norm": 2.0513483756271613, + "learning_rate": 3.3975110794929267e-06, + "loss": 0.9775, + "step": 31311 + }, + { + "epoch": 0.74, + "grad_norm": 2.011710226286265, + "learning_rate": 3.396938019114532e-06, + "loss": 0.9678, + "step": 31312 + }, + { + "epoch": 0.74, + "grad_norm": 1.049822982916985, + "learning_rate": 3.396364997181254e-06, + "loss": 0.9536, + "step": 31313 + }, + { + "epoch": 0.74, + "grad_norm": 1.9020340521635384, + "learning_rate": 3.395792013696426e-06, + "loss": 1.0688, + "step": 31314 + }, + { + "epoch": 0.74, + "grad_norm": 2.7407908635034475, + "learning_rate": 3.3952190686633914e-06, + "loss": 0.9598, + "step": 31315 + }, + { + "epoch": 0.74, + "grad_norm": 2.0864683767800862, + "learning_rate": 3.3946461620854788e-06, + "loss": 1.1016, + "step": 31316 + }, + { + "epoch": 0.74, + "grad_norm": 1.9371858162287252, + "learning_rate": 3.3940732939660303e-06, + "loss": 0.8414, + "step": 31317 + }, + { + "epoch": 0.74, + "grad_norm": 1.8908533409029897, + "learning_rate": 3.393500464308377e-06, + "loss": 1.0496, + "step": 31318 + }, + { + "epoch": 0.74, + "grad_norm": 1.9431748790240537, + "learning_rate": 3.392927673115857e-06, + "loss": 0.9589, + "step": 31319 + }, + { + "epoch": 0.74, + "grad_norm": 2.2508360728901438, + "learning_rate": 3.3923549203918017e-06, + "loss": 1.0126, + "step": 31320 + }, + { + "epoch": 0.74, + "grad_norm": 3.2020942479667522, + "learning_rate": 3.391782206139551e-06, + "loss": 0.8872, + "step": 31321 + }, + { + "epoch": 0.74, + "grad_norm": 1.998603814262407, + "learning_rate": 3.391209530362436e-06, + "loss": 1.0295, + "step": 31322 + }, + { + "epoch": 0.74, + "grad_norm": 2.0964440876124972, + "learning_rate": 3.390636893063789e-06, + "loss": 0.9027, + "step": 31323 + }, + { + "epoch": 0.74, + "grad_norm": 2.4218787368237984, + "learning_rate": 3.390064294246945e-06, + "loss": 0.9273, + "step": 31324 + }, + { + "epoch": 0.74, + "grad_norm": 2.145215754279336, + "learning_rate": 3.3894917339152444e-06, + "loss": 0.9808, + "step": 31325 + }, + { + "epoch": 0.74, + "grad_norm": 2.013470022165198, + "learning_rate": 3.388919212072015e-06, + "loss": 1.0153, + "step": 31326 + }, + { + "epoch": 0.74, + "grad_norm": 1.9286687079768026, + "learning_rate": 3.3883467287205874e-06, + "loss": 0.8919, + "step": 31327 + }, + { + "epoch": 0.74, + "grad_norm": 1.916167694656984, + "learning_rate": 3.387774283864298e-06, + "loss": 1.0595, + "step": 31328 + }, + { + "epoch": 0.74, + "grad_norm": 1.8768266920333647, + "learning_rate": 3.387201877506484e-06, + "loss": 0.9878, + "step": 31329 + }, + { + "epoch": 0.74, + "grad_norm": 1.8051012606532228, + "learning_rate": 3.3866295096504744e-06, + "loss": 0.9197, + "step": 31330 + }, + { + "epoch": 0.74, + "grad_norm": 1.8578968998790721, + "learning_rate": 3.386057180299599e-06, + "loss": 0.9343, + "step": 31331 + }, + { + "epoch": 0.74, + "grad_norm": 1.82203126120369, + "learning_rate": 3.3854848894571913e-06, + "loss": 1.0244, + "step": 31332 + }, + { + "epoch": 0.74, + "grad_norm": 1.89647200334346, + "learning_rate": 3.384912637126588e-06, + "loss": 1.0271, + "step": 31333 + }, + { + "epoch": 0.74, + "grad_norm": 2.0879947683885596, + "learning_rate": 3.3843404233111187e-06, + "loss": 0.8717, + "step": 31334 + }, + { + "epoch": 0.74, + "grad_norm": 1.1507245821331076, + "learning_rate": 3.38376824801411e-06, + "loss": 0.956, + "step": 31335 + }, + { + "epoch": 0.74, + "grad_norm": 2.203826941590737, + "learning_rate": 3.3831961112389e-06, + "loss": 0.844, + "step": 31336 + }, + { + "epoch": 0.74, + "grad_norm": 1.9401864734281193, + "learning_rate": 3.3826240129888143e-06, + "loss": 0.9317, + "step": 31337 + }, + { + "epoch": 0.74, + "grad_norm": 2.0306495358977688, + "learning_rate": 3.3820519532671905e-06, + "loss": 1.0748, + "step": 31338 + }, + { + "epoch": 0.74, + "grad_norm": 2.694799104806116, + "learning_rate": 3.3814799320773517e-06, + "loss": 0.8948, + "step": 31339 + }, + { + "epoch": 0.74, + "grad_norm": 1.9506152082542458, + "learning_rate": 3.3809079494226347e-06, + "loss": 0.9748, + "step": 31340 + }, + { + "epoch": 0.74, + "grad_norm": 2.3183749435604266, + "learning_rate": 3.380336005306365e-06, + "loss": 1.0033, + "step": 31341 + }, + { + "epoch": 0.74, + "grad_norm": 2.2549779054333836, + "learning_rate": 3.379764099731877e-06, + "loss": 1.0819, + "step": 31342 + }, + { + "epoch": 0.74, + "grad_norm": 2.096338655456868, + "learning_rate": 3.3791922327024994e-06, + "loss": 0.9758, + "step": 31343 + }, + { + "epoch": 0.74, + "grad_norm": 1.9975702035529719, + "learning_rate": 3.3786204042215576e-06, + "loss": 0.9863, + "step": 31344 + }, + { + "epoch": 0.74, + "grad_norm": 1.8597319544207196, + "learning_rate": 3.378048614292383e-06, + "loss": 0.8927, + "step": 31345 + }, + { + "epoch": 0.74, + "grad_norm": 3.436033720859403, + "learning_rate": 3.3774768629183065e-06, + "loss": 0.8687, + "step": 31346 + }, + { + "epoch": 0.74, + "grad_norm": 2.2222487016503853, + "learning_rate": 3.376905150102663e-06, + "loss": 0.9847, + "step": 31347 + }, + { + "epoch": 0.74, + "grad_norm": 1.8996978575947239, + "learning_rate": 3.3763334758487674e-06, + "loss": 0.9914, + "step": 31348 + }, + { + "epoch": 0.74, + "grad_norm": 2.209765253556719, + "learning_rate": 3.375761840159956e-06, + "loss": 0.9791, + "step": 31349 + }, + { + "epoch": 0.74, + "grad_norm": 1.8773326077082961, + "learning_rate": 3.3751902430395558e-06, + "loss": 0.9912, + "step": 31350 + }, + { + "epoch": 0.74, + "grad_norm": 1.852954533871686, + "learning_rate": 3.3746186844909025e-06, + "loss": 1.0102, + "step": 31351 + }, + { + "epoch": 0.74, + "grad_norm": 1.9792746044149727, + "learning_rate": 3.374047164517309e-06, + "loss": 1.1039, + "step": 31352 + }, + { + "epoch": 0.74, + "grad_norm": 2.0720365444974127, + "learning_rate": 3.3734756831221116e-06, + "loss": 0.8879, + "step": 31353 + }, + { + "epoch": 0.74, + "grad_norm": 2.6750021828869244, + "learning_rate": 3.37290424030864e-06, + "loss": 1.0503, + "step": 31354 + }, + { + "epoch": 0.74, + "grad_norm": 1.9954802709760198, + "learning_rate": 3.3723328360802187e-06, + "loss": 0.8834, + "step": 31355 + }, + { + "epoch": 0.74, + "grad_norm": 1.9767125363946159, + "learning_rate": 3.371761470440169e-06, + "loss": 0.9554, + "step": 31356 + }, + { + "epoch": 0.74, + "grad_norm": 3.2425377523004184, + "learning_rate": 3.371190143391824e-06, + "loss": 0.9279, + "step": 31357 + }, + { + "epoch": 0.74, + "grad_norm": 1.79415033349455, + "learning_rate": 3.370618854938511e-06, + "loss": 0.9587, + "step": 31358 + }, + { + "epoch": 0.74, + "grad_norm": 2.2088464188248254, + "learning_rate": 3.3700476050835506e-06, + "loss": 0.9915, + "step": 31359 + }, + { + "epoch": 0.74, + "grad_norm": 1.9841563867499499, + "learning_rate": 3.3694763938302764e-06, + "loss": 0.9391, + "step": 31360 + }, + { + "epoch": 0.74, + "grad_norm": 2.0212734672154125, + "learning_rate": 3.368905221182008e-06, + "loss": 0.9737, + "step": 31361 + }, + { + "epoch": 0.74, + "grad_norm": 1.9036394030263535, + "learning_rate": 3.3683340871420712e-06, + "loss": 1.0541, + "step": 31362 + }, + { + "epoch": 0.74, + "grad_norm": 1.9229400599506967, + "learning_rate": 3.367762991713792e-06, + "loss": 0.9954, + "step": 31363 + }, + { + "epoch": 0.74, + "grad_norm": 2.038980252266951, + "learning_rate": 3.3671919349005e-06, + "loss": 1.0121, + "step": 31364 + }, + { + "epoch": 0.74, + "grad_norm": 1.8838130124148222, + "learning_rate": 3.3666209167055163e-06, + "loss": 0.9539, + "step": 31365 + }, + { + "epoch": 0.74, + "grad_norm": 2.041020470709273, + "learning_rate": 3.366049937132162e-06, + "loss": 0.9855, + "step": 31366 + }, + { + "epoch": 0.74, + "grad_norm": 1.8493348584340064, + "learning_rate": 3.3654789961837666e-06, + "loss": 0.9013, + "step": 31367 + }, + { + "epoch": 0.74, + "grad_norm": 2.048205242235734, + "learning_rate": 3.364908093863658e-06, + "loss": 1.0266, + "step": 31368 + }, + { + "epoch": 0.74, + "grad_norm": 2.022423950968732, + "learning_rate": 3.3643372301751498e-06, + "loss": 1.0175, + "step": 31369 + }, + { + "epoch": 0.74, + "grad_norm": 1.7568916776602825, + "learning_rate": 3.3637664051215703e-06, + "loss": 0.8361, + "step": 31370 + }, + { + "epoch": 0.74, + "grad_norm": 1.901631594361472, + "learning_rate": 3.363195618706243e-06, + "loss": 1.019, + "step": 31371 + }, + { + "epoch": 0.74, + "grad_norm": 1.8233883335086754, + "learning_rate": 3.3626248709324973e-06, + "loss": 0.9897, + "step": 31372 + }, + { + "epoch": 0.74, + "grad_norm": 2.059204420245392, + "learning_rate": 3.3620541618036496e-06, + "loss": 0.9352, + "step": 31373 + }, + { + "epoch": 0.74, + "grad_norm": 2.007349956308477, + "learning_rate": 3.361483491323021e-06, + "loss": 0.8768, + "step": 31374 + }, + { + "epoch": 0.74, + "grad_norm": 1.1459892681724733, + "learning_rate": 3.36091285949394e-06, + "loss": 0.9021, + "step": 31375 + }, + { + "epoch": 0.74, + "grad_norm": 2.0859576799813144, + "learning_rate": 3.3603422663197236e-06, + "loss": 0.9636, + "step": 31376 + }, + { + "epoch": 0.74, + "grad_norm": 1.8305844365582995, + "learning_rate": 3.3597717118037e-06, + "loss": 1.0324, + "step": 31377 + }, + { + "epoch": 0.74, + "grad_norm": 5.1667060291106885, + "learning_rate": 3.3592011959491844e-06, + "loss": 1.075, + "step": 31378 + }, + { + "epoch": 0.74, + "grad_norm": 1.9667649188376701, + "learning_rate": 3.3586307187595045e-06, + "loss": 0.9016, + "step": 31379 + }, + { + "epoch": 0.74, + "grad_norm": 2.7324256165842176, + "learning_rate": 3.358060280237977e-06, + "loss": 1.094, + "step": 31380 + }, + { + "epoch": 0.74, + "grad_norm": 1.9834820862621474, + "learning_rate": 3.3574898803879286e-06, + "loss": 0.9755, + "step": 31381 + }, + { + "epoch": 0.74, + "grad_norm": 2.5094923867948107, + "learning_rate": 3.3569195192126758e-06, + "loss": 0.9889, + "step": 31382 + }, + { + "epoch": 0.74, + "grad_norm": 1.778667253103596, + "learning_rate": 3.356349196715538e-06, + "loss": 0.7998, + "step": 31383 + }, + { + "epoch": 0.74, + "grad_norm": 2.1865525850684837, + "learning_rate": 3.35577891289984e-06, + "loss": 1.0038, + "step": 31384 + }, + { + "epoch": 0.74, + "grad_norm": 1.8647926364213552, + "learning_rate": 3.355208667768902e-06, + "loss": 1.0013, + "step": 31385 + }, + { + "epoch": 0.74, + "grad_norm": 2.0122059491383033, + "learning_rate": 3.354638461326043e-06, + "loss": 0.9708, + "step": 31386 + }, + { + "epoch": 0.74, + "grad_norm": 2.060245066275407, + "learning_rate": 3.3540682935745804e-06, + "loss": 1.0355, + "step": 31387 + }, + { + "epoch": 0.74, + "grad_norm": 1.8224030318436684, + "learning_rate": 3.3534981645178356e-06, + "loss": 0.9886, + "step": 31388 + }, + { + "epoch": 0.74, + "grad_norm": 2.0058995732000406, + "learning_rate": 3.352928074159133e-06, + "loss": 1.1772, + "step": 31389 + }, + { + "epoch": 0.74, + "grad_norm": 1.8896692727662807, + "learning_rate": 3.3523580225017872e-06, + "loss": 0.9622, + "step": 31390 + }, + { + "epoch": 0.74, + "grad_norm": 1.9666294350516187, + "learning_rate": 3.351788009549114e-06, + "loss": 1.0088, + "step": 31391 + }, + { + "epoch": 0.74, + "grad_norm": 1.8188746926519228, + "learning_rate": 3.3512180353044356e-06, + "loss": 0.965, + "step": 31392 + }, + { + "epoch": 0.74, + "grad_norm": 1.95762589397148, + "learning_rate": 3.3506480997710743e-06, + "loss": 1.0761, + "step": 31393 + }, + { + "epoch": 0.74, + "grad_norm": 2.313091871073124, + "learning_rate": 3.350078202952345e-06, + "loss": 0.9891, + "step": 31394 + }, + { + "epoch": 0.74, + "grad_norm": 1.058401527913276, + "learning_rate": 3.349508344851562e-06, + "loss": 0.9309, + "step": 31395 + }, + { + "epoch": 0.74, + "grad_norm": 1.9120582735930767, + "learning_rate": 3.3489385254720507e-06, + "loss": 0.818, + "step": 31396 + }, + { + "epoch": 0.74, + "grad_norm": 1.9386414941171803, + "learning_rate": 3.348368744817121e-06, + "loss": 1.0437, + "step": 31397 + }, + { + "epoch": 0.74, + "grad_norm": 1.9229085170579654, + "learning_rate": 3.3477990028900986e-06, + "loss": 1.1411, + "step": 31398 + }, + { + "epoch": 0.74, + "grad_norm": 2.049449947887883, + "learning_rate": 3.347229299694292e-06, + "loss": 0.994, + "step": 31399 + }, + { + "epoch": 0.74, + "grad_norm": 1.8299891089914273, + "learning_rate": 3.346659635233027e-06, + "loss": 1.0294, + "step": 31400 + }, + { + "epoch": 0.74, + "grad_norm": 2.3263408162540147, + "learning_rate": 3.346090009509613e-06, + "loss": 1.0893, + "step": 31401 + }, + { + "epoch": 0.74, + "grad_norm": 2.0042529302382306, + "learning_rate": 3.3455204225273718e-06, + "loss": 0.9737, + "step": 31402 + }, + { + "epoch": 0.74, + "grad_norm": 1.9753262023937284, + "learning_rate": 3.3449508742896176e-06, + "loss": 0.8591, + "step": 31403 + }, + { + "epoch": 0.74, + "grad_norm": 1.9806765695592112, + "learning_rate": 3.3443813647996638e-06, + "loss": 0.9922, + "step": 31404 + }, + { + "epoch": 0.74, + "grad_norm": 2.1070772967391918, + "learning_rate": 3.343811894060829e-06, + "loss": 1.0435, + "step": 31405 + }, + { + "epoch": 0.74, + "grad_norm": 1.842825222500699, + "learning_rate": 3.3432424620764304e-06, + "loss": 0.9903, + "step": 31406 + }, + { + "epoch": 0.74, + "grad_norm": 1.8025730495217542, + "learning_rate": 3.342673068849783e-06, + "loss": 1.0725, + "step": 31407 + }, + { + "epoch": 0.74, + "grad_norm": 2.0978288930213798, + "learning_rate": 3.342103714384197e-06, + "loss": 1.0092, + "step": 31408 + }, + { + "epoch": 0.74, + "grad_norm": 1.10233931974809, + "learning_rate": 3.3415343986829908e-06, + "loss": 0.9255, + "step": 31409 + }, + { + "epoch": 0.74, + "grad_norm": 1.9496773745714795, + "learning_rate": 3.340965121749483e-06, + "loss": 1.057, + "step": 31410 + }, + { + "epoch": 0.74, + "grad_norm": 1.9633425295585143, + "learning_rate": 3.3403958835869843e-06, + "loss": 1.0443, + "step": 31411 + }, + { + "epoch": 0.74, + "grad_norm": 2.183855388319646, + "learning_rate": 3.339826684198806e-06, + "loss": 0.9708, + "step": 31412 + }, + { + "epoch": 0.74, + "grad_norm": 1.0831967005073153, + "learning_rate": 3.3392575235882654e-06, + "loss": 0.9687, + "step": 31413 + }, + { + "epoch": 0.74, + "grad_norm": 1.9199565554947167, + "learning_rate": 3.3386884017586794e-06, + "loss": 1.0098, + "step": 31414 + }, + { + "epoch": 0.74, + "grad_norm": 1.9557690338045204, + "learning_rate": 3.3381193187133587e-06, + "loss": 0.9641, + "step": 31415 + }, + { + "epoch": 0.74, + "grad_norm": 1.1088882825900046, + "learning_rate": 3.3375502744556134e-06, + "loss": 1.0147, + "step": 31416 + }, + { + "epoch": 0.74, + "grad_norm": 1.8395202615809156, + "learning_rate": 3.3369812689887617e-06, + "loss": 0.8537, + "step": 31417 + }, + { + "epoch": 0.74, + "grad_norm": 2.048995684248593, + "learning_rate": 3.336412302316112e-06, + "loss": 1.122, + "step": 31418 + }, + { + "epoch": 0.74, + "grad_norm": 2.2245737046571232, + "learning_rate": 3.3358433744409835e-06, + "loss": 1.0155, + "step": 31419 + }, + { + "epoch": 0.74, + "grad_norm": 2.1241360626834993, + "learning_rate": 3.3352744853666807e-06, + "loss": 1.0007, + "step": 31420 + }, + { + "epoch": 0.74, + "grad_norm": 1.1752984150220314, + "learning_rate": 3.3347056350965235e-06, + "loss": 0.9632, + "step": 31421 + }, + { + "epoch": 0.74, + "grad_norm": 2.132238129397204, + "learning_rate": 3.334136823633817e-06, + "loss": 0.8995, + "step": 31422 + }, + { + "epoch": 0.74, + "grad_norm": 2.130322359298378, + "learning_rate": 3.333568050981879e-06, + "loss": 0.9003, + "step": 31423 + }, + { + "epoch": 0.74, + "grad_norm": 2.0047137464501215, + "learning_rate": 3.332999317144019e-06, + "loss": 0.9308, + "step": 31424 + }, + { + "epoch": 0.74, + "grad_norm": 3.5297759607807144, + "learning_rate": 3.3324306221235435e-06, + "loss": 0.9189, + "step": 31425 + }, + { + "epoch": 0.74, + "grad_norm": 1.8594899998492058, + "learning_rate": 3.3318619659237684e-06, + "loss": 0.8599, + "step": 31426 + }, + { + "epoch": 0.74, + "grad_norm": 2.5052083643232046, + "learning_rate": 3.331293348548007e-06, + "loss": 0.9032, + "step": 31427 + }, + { + "epoch": 0.74, + "grad_norm": 2.064156777123804, + "learning_rate": 3.330724769999568e-06, + "loss": 1.0473, + "step": 31428 + }, + { + "epoch": 0.74, + "grad_norm": 1.8654044127476974, + "learning_rate": 3.3301562302817568e-06, + "loss": 0.9078, + "step": 31429 + }, + { + "epoch": 0.74, + "grad_norm": 1.9305398940425618, + "learning_rate": 3.3295877293978875e-06, + "loss": 0.9074, + "step": 31430 + }, + { + "epoch": 0.74, + "grad_norm": 1.8416884673314653, + "learning_rate": 3.3290192673512747e-06, + "loss": 0.9238, + "step": 31431 + }, + { + "epoch": 0.74, + "grad_norm": 2.3554912442703735, + "learning_rate": 3.3284508441452225e-06, + "loss": 0.9687, + "step": 31432 + }, + { + "epoch": 0.74, + "grad_norm": 2.0063996452745196, + "learning_rate": 3.3278824597830383e-06, + "loss": 1.0337, + "step": 31433 + }, + { + "epoch": 0.74, + "grad_norm": 2.0255936784773825, + "learning_rate": 3.3273141142680356e-06, + "loss": 1.0232, + "step": 31434 + }, + { + "epoch": 0.74, + "grad_norm": 2.1358711306663105, + "learning_rate": 3.3267458076035255e-06, + "loss": 0.9331, + "step": 31435 + }, + { + "epoch": 0.74, + "grad_norm": 1.9472856575577295, + "learning_rate": 3.3261775397928143e-06, + "loss": 1.0434, + "step": 31436 + }, + { + "epoch": 0.74, + "grad_norm": 2.2678153834308854, + "learning_rate": 3.3256093108392064e-06, + "loss": 1.031, + "step": 31437 + }, + { + "epoch": 0.74, + "grad_norm": 2.543087630653752, + "learning_rate": 3.3250411207460188e-06, + "loss": 0.9788, + "step": 31438 + }, + { + "epoch": 0.74, + "grad_norm": 1.927796124060678, + "learning_rate": 3.32447296951655e-06, + "loss": 1.0216, + "step": 31439 + }, + { + "epoch": 0.74, + "grad_norm": 1.8492340708005472, + "learning_rate": 3.323904857154118e-06, + "loss": 0.945, + "step": 31440 + }, + { + "epoch": 0.74, + "grad_norm": 1.2412590740467988, + "learning_rate": 3.323336783662021e-06, + "loss": 0.9521, + "step": 31441 + }, + { + "epoch": 0.74, + "grad_norm": 1.1167481137126771, + "learning_rate": 3.3227687490435746e-06, + "loss": 0.9274, + "step": 31442 + }, + { + "epoch": 0.74, + "grad_norm": 2.2973744577081248, + "learning_rate": 3.3222007533020793e-06, + "loss": 1.0222, + "step": 31443 + }, + { + "epoch": 0.74, + "grad_norm": 2.2457491311260083, + "learning_rate": 3.3216327964408456e-06, + "loss": 0.9133, + "step": 31444 + }, + { + "epoch": 0.74, + "grad_norm": 1.9411562214376943, + "learning_rate": 3.321064878463187e-06, + "loss": 1.0422, + "step": 31445 + }, + { + "epoch": 0.74, + "grad_norm": 1.0663492173097988, + "learning_rate": 3.3204969993723955e-06, + "loss": 0.9054, + "step": 31446 + }, + { + "epoch": 0.74, + "grad_norm": 1.894634029461527, + "learning_rate": 3.3199291591717865e-06, + "loss": 1.0256, + "step": 31447 + }, + { + "epoch": 0.74, + "grad_norm": 2.1895170572527722, + "learning_rate": 3.3193613578646633e-06, + "loss": 1.0695, + "step": 31448 + }, + { + "epoch": 0.74, + "grad_norm": 2.2081987973463826, + "learning_rate": 3.3187935954543417e-06, + "loss": 0.9852, + "step": 31449 + }, + { + "epoch": 0.74, + "grad_norm": 2.501157462129021, + "learning_rate": 3.3182258719441107e-06, + "loss": 1.1184, + "step": 31450 + }, + { + "epoch": 0.74, + "grad_norm": 2.141969893727237, + "learning_rate": 3.3176581873372848e-06, + "loss": 0.9969, + "step": 31451 + }, + { + "epoch": 0.74, + "grad_norm": 2.016948519028812, + "learning_rate": 3.3170905416371723e-06, + "loss": 0.9246, + "step": 31452 + }, + { + "epoch": 0.74, + "grad_norm": 1.8092086750238077, + "learning_rate": 3.316522934847074e-06, + "loss": 0.9615, + "step": 31453 + }, + { + "epoch": 0.74, + "grad_norm": 1.9181213014752052, + "learning_rate": 3.3159553669702917e-06, + "loss": 0.9518, + "step": 31454 + }, + { + "epoch": 0.74, + "grad_norm": 2.0082851668881667, + "learning_rate": 3.3153878380101332e-06, + "loss": 1.032, + "step": 31455 + }, + { + "epoch": 0.74, + "grad_norm": 1.9888005613645297, + "learning_rate": 3.314820347969907e-06, + "loss": 1.006, + "step": 31456 + }, + { + "epoch": 0.74, + "grad_norm": 2.0794024137812355, + "learning_rate": 3.3142528968529097e-06, + "loss": 0.9149, + "step": 31457 + }, + { + "epoch": 0.74, + "grad_norm": 1.8767142342908012, + "learning_rate": 3.313685484662452e-06, + "loss": 0.8915, + "step": 31458 + }, + { + "epoch": 0.74, + "grad_norm": 1.8649388139040235, + "learning_rate": 3.313118111401834e-06, + "loss": 0.9155, + "step": 31459 + }, + { + "epoch": 0.74, + "grad_norm": 2.0825791807676803, + "learning_rate": 3.312550777074356e-06, + "loss": 1.0036, + "step": 31460 + }, + { + "epoch": 0.74, + "grad_norm": 2.1153558708734748, + "learning_rate": 3.311983481683325e-06, + "loss": 1.0287, + "step": 31461 + }, + { + "epoch": 0.74, + "grad_norm": 2.7389053112818713, + "learning_rate": 3.311416225232047e-06, + "loss": 1.0272, + "step": 31462 + }, + { + "epoch": 0.74, + "grad_norm": 2.0439356513986575, + "learning_rate": 3.3108490077238207e-06, + "loss": 0.8948, + "step": 31463 + }, + { + "epoch": 0.74, + "grad_norm": 2.123658217348588, + "learning_rate": 3.3102818291619467e-06, + "loss": 1.0347, + "step": 31464 + }, + { + "epoch": 0.74, + "grad_norm": 1.8142088077828962, + "learning_rate": 3.309714689549729e-06, + "loss": 0.8636, + "step": 31465 + }, + { + "epoch": 0.74, + "grad_norm": 2.0330528283657787, + "learning_rate": 3.3091475888904777e-06, + "loss": 1.0732, + "step": 31466 + }, + { + "epoch": 0.74, + "grad_norm": 4.525426237721891, + "learning_rate": 3.3085805271874803e-06, + "loss": 0.9089, + "step": 31467 + }, + { + "epoch": 0.74, + "grad_norm": 1.9767044860627152, + "learning_rate": 3.3080135044440454e-06, + "loss": 0.9065, + "step": 31468 + }, + { + "epoch": 0.74, + "grad_norm": 2.1032714123651157, + "learning_rate": 3.3074465206634755e-06, + "loss": 1.0343, + "step": 31469 + }, + { + "epoch": 0.74, + "grad_norm": 1.121777004268625, + "learning_rate": 3.3068795758490736e-06, + "loss": 0.9052, + "step": 31470 + }, + { + "epoch": 0.74, + "grad_norm": 2.0278720937868586, + "learning_rate": 3.306312670004137e-06, + "loss": 0.9851, + "step": 31471 + }, + { + "epoch": 0.74, + "grad_norm": 1.812159861990259, + "learning_rate": 3.3057458031319654e-06, + "loss": 0.9979, + "step": 31472 + }, + { + "epoch": 0.74, + "grad_norm": 2.099504908468996, + "learning_rate": 3.30517897523586e-06, + "loss": 0.8931, + "step": 31473 + }, + { + "epoch": 0.74, + "grad_norm": 1.9792602726196549, + "learning_rate": 3.304612186319126e-06, + "loss": 0.9253, + "step": 31474 + }, + { + "epoch": 0.74, + "grad_norm": 2.065446675796625, + "learning_rate": 3.304045436385059e-06, + "loss": 0.8621, + "step": 31475 + }, + { + "epoch": 0.74, + "grad_norm": 2.006658763791449, + "learning_rate": 3.303478725436956e-06, + "loss": 0.9871, + "step": 31476 + }, + { + "epoch": 0.74, + "grad_norm": 1.8629211262955376, + "learning_rate": 3.3029120534781244e-06, + "loss": 0.9538, + "step": 31477 + }, + { + "epoch": 0.74, + "grad_norm": 1.8792254670955502, + "learning_rate": 3.3023454205118545e-06, + "loss": 0.9749, + "step": 31478 + }, + { + "epoch": 0.74, + "grad_norm": 1.1502449924258402, + "learning_rate": 3.301778826541454e-06, + "loss": 1.0275, + "step": 31479 + }, + { + "epoch": 0.74, + "grad_norm": 1.9024591464775253, + "learning_rate": 3.3012122715702145e-06, + "loss": 0.9059, + "step": 31480 + }, + { + "epoch": 0.74, + "grad_norm": 1.9948689336761174, + "learning_rate": 3.3006457556014415e-06, + "loss": 1.1396, + "step": 31481 + }, + { + "epoch": 0.74, + "grad_norm": 1.9084209967614956, + "learning_rate": 3.300079278638426e-06, + "loss": 1.0287, + "step": 31482 + }, + { + "epoch": 0.74, + "grad_norm": 2.103999189662361, + "learning_rate": 3.2995128406844733e-06, + "loss": 0.8767, + "step": 31483 + }, + { + "epoch": 0.74, + "grad_norm": 2.7272362634867737, + "learning_rate": 3.2989464417428785e-06, + "loss": 0.8716, + "step": 31484 + }, + { + "epoch": 0.74, + "grad_norm": 1.9502733175114662, + "learning_rate": 3.2983800818169363e-06, + "loss": 1.0262, + "step": 31485 + }, + { + "epoch": 0.74, + "grad_norm": 2.019179906202853, + "learning_rate": 3.297813760909946e-06, + "loss": 0.8607, + "step": 31486 + }, + { + "epoch": 0.74, + "grad_norm": 2.1527710291347755, + "learning_rate": 3.2972474790252097e-06, + "loss": 1.0995, + "step": 31487 + }, + { + "epoch": 0.74, + "grad_norm": 2.219637771043954, + "learning_rate": 3.2966812361660204e-06, + "loss": 1.0157, + "step": 31488 + }, + { + "epoch": 0.74, + "grad_norm": 2.032515768061318, + "learning_rate": 3.2961150323356715e-06, + "loss": 0.9167, + "step": 31489 + }, + { + "epoch": 0.74, + "grad_norm": 2.228638021892518, + "learning_rate": 3.2955488675374635e-06, + "loss": 0.8957, + "step": 31490 + }, + { + "epoch": 0.74, + "grad_norm": 1.899147053659471, + "learning_rate": 3.294982741774697e-06, + "loss": 1.0443, + "step": 31491 + }, + { + "epoch": 0.74, + "grad_norm": 1.143486285790405, + "learning_rate": 3.294416655050663e-06, + "loss": 0.9286, + "step": 31492 + }, + { + "epoch": 0.74, + "grad_norm": 1.7664810278456617, + "learning_rate": 3.293850607368655e-06, + "loss": 0.91, + "step": 31493 + }, + { + "epoch": 0.74, + "grad_norm": 1.905025391609504, + "learning_rate": 3.2932845987319704e-06, + "loss": 0.9835, + "step": 31494 + }, + { + "epoch": 0.74, + "grad_norm": 2.3284414426552047, + "learning_rate": 3.2927186291439117e-06, + "loss": 1.0192, + "step": 31495 + }, + { + "epoch": 0.74, + "grad_norm": 1.9442657882436094, + "learning_rate": 3.292152698607768e-06, + "loss": 0.8678, + "step": 31496 + }, + { + "epoch": 0.74, + "grad_norm": 1.8025336857745597, + "learning_rate": 3.291586807126831e-06, + "loss": 1.0472, + "step": 31497 + }, + { + "epoch": 0.74, + "grad_norm": 1.9162011609014198, + "learning_rate": 3.2910209547044027e-06, + "loss": 0.9547, + "step": 31498 + }, + { + "epoch": 0.74, + "grad_norm": 2.0080209697581375, + "learning_rate": 3.290455141343771e-06, + "loss": 0.9823, + "step": 31499 + }, + { + "epoch": 0.74, + "grad_norm": 2.048764156455711, + "learning_rate": 3.289889367048238e-06, + "loss": 1.0141, + "step": 31500 + }, + { + "epoch": 0.74, + "grad_norm": 1.750556527819812, + "learning_rate": 3.28932363182109e-06, + "loss": 0.9608, + "step": 31501 + }, + { + "epoch": 0.74, + "grad_norm": 1.7532071677627485, + "learning_rate": 3.2887579356656274e-06, + "loss": 0.9669, + "step": 31502 + }, + { + "epoch": 0.74, + "grad_norm": 1.8255846540541016, + "learning_rate": 3.2881922785851374e-06, + "loss": 0.9513, + "step": 31503 + }, + { + "epoch": 0.74, + "grad_norm": 2.1232369119839873, + "learning_rate": 3.28762666058292e-06, + "loss": 0.9003, + "step": 31504 + }, + { + "epoch": 0.74, + "grad_norm": 2.2942949332499762, + "learning_rate": 3.2870610816622663e-06, + "loss": 0.9209, + "step": 31505 + }, + { + "epoch": 0.74, + "grad_norm": 2.189418938890539, + "learning_rate": 3.286495541826463e-06, + "loss": 0.9544, + "step": 31506 + }, + { + "epoch": 0.74, + "grad_norm": 2.621084815861446, + "learning_rate": 3.28593004107881e-06, + "loss": 1.0686, + "step": 31507 + }, + { + "epoch": 0.74, + "grad_norm": 1.935784776980426, + "learning_rate": 3.2853645794226007e-06, + "loss": 0.8863, + "step": 31508 + }, + { + "epoch": 0.74, + "grad_norm": 2.0150434743619954, + "learning_rate": 3.284799156861125e-06, + "loss": 0.8261, + "step": 31509 + }, + { + "epoch": 0.74, + "grad_norm": 7.789292290989246, + "learning_rate": 3.2842337733976704e-06, + "loss": 0.9953, + "step": 31510 + }, + { + "epoch": 0.74, + "grad_norm": 2.0362761756342143, + "learning_rate": 3.2836684290355336e-06, + "loss": 0.9198, + "step": 31511 + }, + { + "epoch": 0.74, + "grad_norm": 2.3356874486203076, + "learning_rate": 3.2831031237780086e-06, + "loss": 1.0337, + "step": 31512 + }, + { + "epoch": 0.74, + "grad_norm": 2.060812653106089, + "learning_rate": 3.2825378576283827e-06, + "loss": 0.9168, + "step": 31513 + }, + { + "epoch": 0.74, + "grad_norm": 1.9409964064577288, + "learning_rate": 3.2819726305899467e-06, + "loss": 0.9683, + "step": 31514 + }, + { + "epoch": 0.74, + "grad_norm": 2.061693336320068, + "learning_rate": 3.2814074426659925e-06, + "loss": 0.9249, + "step": 31515 + }, + { + "epoch": 0.74, + "grad_norm": 2.2914069603498786, + "learning_rate": 3.280842293859814e-06, + "loss": 1.0466, + "step": 31516 + }, + { + "epoch": 0.74, + "grad_norm": 1.9716458741652243, + "learning_rate": 3.2802771841746984e-06, + "loss": 0.9881, + "step": 31517 + }, + { + "epoch": 0.74, + "grad_norm": 2.2310009473217387, + "learning_rate": 3.2797121136139345e-06, + "loss": 0.998, + "step": 31518 + }, + { + "epoch": 0.74, + "grad_norm": 2.074002668203282, + "learning_rate": 3.279147082180817e-06, + "loss": 1.0049, + "step": 31519 + }, + { + "epoch": 0.74, + "grad_norm": 4.131444469909517, + "learning_rate": 3.278582089878629e-06, + "loss": 0.9398, + "step": 31520 + }, + { + "epoch": 0.74, + "grad_norm": 1.8906644164855253, + "learning_rate": 3.2780171367106684e-06, + "loss": 0.8954, + "step": 31521 + }, + { + "epoch": 0.74, + "grad_norm": 1.1608569992269475, + "learning_rate": 3.2774522226802162e-06, + "loss": 0.9879, + "step": 31522 + }, + { + "epoch": 0.74, + "grad_norm": 1.13325634525741, + "learning_rate": 3.2768873477905696e-06, + "loss": 0.8954, + "step": 31523 + }, + { + "epoch": 0.74, + "grad_norm": 2.0195000425192733, + "learning_rate": 3.2763225120450094e-06, + "loss": 1.0711, + "step": 31524 + }, + { + "epoch": 0.74, + "grad_norm": 2.088247129141574, + "learning_rate": 3.275757715446831e-06, + "loss": 0.8027, + "step": 31525 + }, + { + "epoch": 0.74, + "grad_norm": 2.0574244105417763, + "learning_rate": 3.2751929579993203e-06, + "loss": 0.9685, + "step": 31526 + }, + { + "epoch": 0.74, + "grad_norm": 1.9603752072379086, + "learning_rate": 3.2746282397057627e-06, + "loss": 0.9481, + "step": 31527 + }, + { + "epoch": 0.74, + "grad_norm": 2.0181878391643955, + "learning_rate": 3.274063560569448e-06, + "loss": 1.0304, + "step": 31528 + }, + { + "epoch": 0.74, + "grad_norm": 2.4331271361799245, + "learning_rate": 3.2734989205936684e-06, + "loss": 1.0206, + "step": 31529 + }, + { + "epoch": 0.74, + "grad_norm": 2.5677480687202436, + "learning_rate": 3.2729343197817078e-06, + "loss": 0.9496, + "step": 31530 + }, + { + "epoch": 0.74, + "grad_norm": 1.9991057146757405, + "learning_rate": 3.2723697581368496e-06, + "loss": 1.0623, + "step": 31531 + }, + { + "epoch": 0.74, + "grad_norm": 1.798685847357565, + "learning_rate": 3.271805235662384e-06, + "loss": 0.9389, + "step": 31532 + }, + { + "epoch": 0.74, + "grad_norm": 2.372352595767081, + "learning_rate": 3.271240752361603e-06, + "loss": 0.8747, + "step": 31533 + }, + { + "epoch": 0.74, + "grad_norm": 1.1391650063664491, + "learning_rate": 3.270676308237788e-06, + "loss": 0.9538, + "step": 31534 + }, + { + "epoch": 0.74, + "grad_norm": 2.1426103860291934, + "learning_rate": 3.270111903294222e-06, + "loss": 0.845, + "step": 31535 + }, + { + "epoch": 0.74, + "grad_norm": 1.8543409633146486, + "learning_rate": 3.2695475375341967e-06, + "loss": 0.9497, + "step": 31536 + }, + { + "epoch": 0.74, + "grad_norm": 2.0775798654229813, + "learning_rate": 3.2689832109609985e-06, + "loss": 0.933, + "step": 31537 + }, + { + "epoch": 0.74, + "grad_norm": 2.205612713545331, + "learning_rate": 3.268418923577912e-06, + "loss": 0.9733, + "step": 31538 + }, + { + "epoch": 0.74, + "grad_norm": 2.120530451924026, + "learning_rate": 3.2678546753882177e-06, + "loss": 1.1103, + "step": 31539 + }, + { + "epoch": 0.74, + "grad_norm": 2.296944597201088, + "learning_rate": 3.2672904663952078e-06, + "loss": 1.0139, + "step": 31540 + }, + { + "epoch": 0.74, + "grad_norm": 1.062944038076666, + "learning_rate": 3.2667262966021617e-06, + "loss": 0.9261, + "step": 31541 + }, + { + "epoch": 0.74, + "grad_norm": 2.6893050525110036, + "learning_rate": 3.2661621660123666e-06, + "loss": 0.9808, + "step": 31542 + }, + { + "epoch": 0.74, + "grad_norm": 1.0735316607498966, + "learning_rate": 3.2655980746291114e-06, + "loss": 0.9894, + "step": 31543 + }, + { + "epoch": 0.74, + "grad_norm": 2.0641908085934793, + "learning_rate": 3.2650340224556754e-06, + "loss": 0.9857, + "step": 31544 + }, + { + "epoch": 0.74, + "grad_norm": 2.3054254946469133, + "learning_rate": 3.264470009495341e-06, + "loss": 0.8722, + "step": 31545 + }, + { + "epoch": 0.74, + "grad_norm": 2.0476727067603733, + "learning_rate": 3.263906035751393e-06, + "loss": 0.9092, + "step": 31546 + }, + { + "epoch": 0.74, + "grad_norm": 1.9679846561509537, + "learning_rate": 3.263342101227125e-06, + "loss": 0.8559, + "step": 31547 + }, + { + "epoch": 0.74, + "grad_norm": 2.420693214794262, + "learning_rate": 3.2627782059258053e-06, + "loss": 0.9246, + "step": 31548 + }, + { + "epoch": 0.74, + "grad_norm": 3.2516650785427297, + "learning_rate": 3.262214349850723e-06, + "loss": 0.9043, + "step": 31549 + }, + { + "epoch": 0.74, + "grad_norm": 3.3753289948087573, + "learning_rate": 3.261650533005162e-06, + "loss": 0.8919, + "step": 31550 + }, + { + "epoch": 0.74, + "grad_norm": 5.265370318072841, + "learning_rate": 3.261086755392412e-06, + "loss": 0.9779, + "step": 31551 + }, + { + "epoch": 0.74, + "grad_norm": 2.2575478994771365, + "learning_rate": 3.2605230170157407e-06, + "loss": 1.0511, + "step": 31552 + }, + { + "epoch": 0.74, + "grad_norm": 1.9433213079260914, + "learning_rate": 3.2599593178784395e-06, + "loss": 0.9838, + "step": 31553 + }, + { + "epoch": 0.74, + "grad_norm": 2.1431978100788918, + "learning_rate": 3.2593956579837914e-06, + "loss": 1.042, + "step": 31554 + }, + { + "epoch": 0.74, + "grad_norm": 2.0044626065262388, + "learning_rate": 3.2588320373350736e-06, + "loss": 1.0273, + "step": 31555 + }, + { + "epoch": 0.74, + "grad_norm": 4.227289503773173, + "learning_rate": 3.2582684559355727e-06, + "loss": 1.0666, + "step": 31556 + }, + { + "epoch": 0.74, + "grad_norm": 1.9257889100883834, + "learning_rate": 3.257704913788564e-06, + "loss": 1.0115, + "step": 31557 + }, + { + "epoch": 0.74, + "grad_norm": 2.006540844213878, + "learning_rate": 3.257141410897334e-06, + "loss": 1.041, + "step": 31558 + }, + { + "epoch": 0.74, + "grad_norm": 1.9810808278009937, + "learning_rate": 3.2565779472651595e-06, + "loss": 1.0275, + "step": 31559 + }, + { + "epoch": 0.74, + "grad_norm": 1.7460600465816978, + "learning_rate": 3.2560145228953267e-06, + "loss": 0.8675, + "step": 31560 + }, + { + "epoch": 0.74, + "grad_norm": 2.087681703222475, + "learning_rate": 3.2554511377911115e-06, + "loss": 1.0655, + "step": 31561 + }, + { + "epoch": 0.74, + "grad_norm": 1.1226725847103491, + "learning_rate": 3.254887791955792e-06, + "loss": 1.0114, + "step": 31562 + }, + { + "epoch": 0.74, + "grad_norm": 1.9860780837774987, + "learning_rate": 3.2543244853926515e-06, + "loss": 1.0486, + "step": 31563 + }, + { + "epoch": 0.74, + "grad_norm": 2.3002752827908, + "learning_rate": 3.2537612181049726e-06, + "loss": 0.9844, + "step": 31564 + }, + { + "epoch": 0.74, + "grad_norm": 1.92066801933506, + "learning_rate": 3.253197990096031e-06, + "loss": 1.0699, + "step": 31565 + }, + { + "epoch": 0.74, + "grad_norm": 2.004119732009198, + "learning_rate": 3.2526348013691043e-06, + "loss": 0.9803, + "step": 31566 + }, + { + "epoch": 0.74, + "grad_norm": 1.8190780350376137, + "learning_rate": 3.2520716519274743e-06, + "loss": 0.8832, + "step": 31567 + }, + { + "epoch": 0.74, + "grad_norm": 1.8389572135698529, + "learning_rate": 3.251508541774422e-06, + "loss": 1.0334, + "step": 31568 + }, + { + "epoch": 0.74, + "grad_norm": 2.218955930997203, + "learning_rate": 3.250945470913224e-06, + "loss": 0.8795, + "step": 31569 + }, + { + "epoch": 0.74, + "grad_norm": 2.1775421356019016, + "learning_rate": 3.250382439347154e-06, + "loss": 0.9026, + "step": 31570 + }, + { + "epoch": 0.74, + "grad_norm": 1.738310097872925, + "learning_rate": 3.2498194470794954e-06, + "loss": 0.9637, + "step": 31571 + }, + { + "epoch": 0.74, + "grad_norm": 1.894769447016527, + "learning_rate": 3.2492564941135273e-06, + "loss": 0.9728, + "step": 31572 + }, + { + "epoch": 0.74, + "grad_norm": 1.948372399309723, + "learning_rate": 3.2486935804525264e-06, + "loss": 0.8248, + "step": 31573 + }, + { + "epoch": 0.74, + "grad_norm": 1.9431839279669476, + "learning_rate": 3.2481307060997645e-06, + "loss": 1.0558, + "step": 31574 + }, + { + "epoch": 0.74, + "grad_norm": 1.8723519104884043, + "learning_rate": 3.2475678710585267e-06, + "loss": 0.8601, + "step": 31575 + }, + { + "epoch": 0.74, + "grad_norm": 2.2129927777434504, + "learning_rate": 3.247005075332084e-06, + "loss": 1.0818, + "step": 31576 + }, + { + "epoch": 0.74, + "grad_norm": 2.7932975941416602, + "learning_rate": 3.246442318923718e-06, + "loss": 1.0834, + "step": 31577 + }, + { + "epoch": 0.74, + "grad_norm": 2.1991766149353427, + "learning_rate": 3.245879601836701e-06, + "loss": 0.9663, + "step": 31578 + }, + { + "epoch": 0.74, + "grad_norm": 1.9698866337996295, + "learning_rate": 3.2453169240743132e-06, + "loss": 0.9261, + "step": 31579 + }, + { + "epoch": 0.74, + "grad_norm": 2.0170351271271074, + "learning_rate": 3.244754285639826e-06, + "loss": 1.0426, + "step": 31580 + }, + { + "epoch": 0.74, + "grad_norm": 2.165587750810518, + "learning_rate": 3.244191686536522e-06, + "loss": 1.011, + "step": 31581 + }, + { + "epoch": 0.74, + "grad_norm": 2.1858359488499537, + "learning_rate": 3.2436291267676725e-06, + "loss": 0.8629, + "step": 31582 + }, + { + "epoch": 0.74, + "grad_norm": 1.9330206494345, + "learning_rate": 3.243066606336549e-06, + "loss": 1.013, + "step": 31583 + }, + { + "epoch": 0.74, + "grad_norm": 1.969427603576648, + "learning_rate": 3.242504125246432e-06, + "loss": 0.8959, + "step": 31584 + }, + { + "epoch": 0.74, + "grad_norm": 2.1291827179814766, + "learning_rate": 3.241941683500598e-06, + "loss": 0.9157, + "step": 31585 + }, + { + "epoch": 0.74, + "grad_norm": 1.9819066554980944, + "learning_rate": 3.241379281102319e-06, + "loss": 1.0496, + "step": 31586 + }, + { + "epoch": 0.74, + "grad_norm": 1.9742563423316484, + "learning_rate": 3.2408169180548675e-06, + "loss": 0.9385, + "step": 31587 + }, + { + "epoch": 0.74, + "grad_norm": 1.0373193164439252, + "learning_rate": 3.2402545943615195e-06, + "loss": 0.9544, + "step": 31588 + }, + { + "epoch": 0.74, + "grad_norm": 1.966726980830745, + "learning_rate": 3.2396923100255515e-06, + "loss": 0.9713, + "step": 31589 + }, + { + "epoch": 0.74, + "grad_norm": 1.8574404872875325, + "learning_rate": 3.2391300650502356e-06, + "loss": 0.9516, + "step": 31590 + }, + { + "epoch": 0.74, + "grad_norm": 2.0564258181144046, + "learning_rate": 3.2385678594388416e-06, + "loss": 1.0707, + "step": 31591 + }, + { + "epoch": 0.74, + "grad_norm": 2.0461483743537943, + "learning_rate": 3.238005693194647e-06, + "loss": 0.925, + "step": 31592 + }, + { + "epoch": 0.74, + "grad_norm": 1.81401050094299, + "learning_rate": 3.2374435663209268e-06, + "loss": 0.991, + "step": 31593 + }, + { + "epoch": 0.74, + "grad_norm": 2.085783216991554, + "learning_rate": 3.23688147882095e-06, + "loss": 1.0544, + "step": 31594 + }, + { + "epoch": 0.74, + "grad_norm": 1.8983828164273322, + "learning_rate": 3.2363194306979884e-06, + "loss": 1.0328, + "step": 31595 + }, + { + "epoch": 0.74, + "grad_norm": 1.1177243835236017, + "learning_rate": 3.235757421955317e-06, + "loss": 0.907, + "step": 31596 + }, + { + "epoch": 0.74, + "grad_norm": 2.0803875841832786, + "learning_rate": 3.235195452596209e-06, + "loss": 0.9597, + "step": 31597 + }, + { + "epoch": 0.74, + "grad_norm": 1.80653687308073, + "learning_rate": 3.2346335226239357e-06, + "loss": 0.9171, + "step": 31598 + }, + { + "epoch": 0.74, + "grad_norm": 1.9858128753596156, + "learning_rate": 3.234071632041765e-06, + "loss": 1.0042, + "step": 31599 + }, + { + "epoch": 0.74, + "grad_norm": 1.774631309279258, + "learning_rate": 3.2335097808529736e-06, + "loss": 0.8975, + "step": 31600 + }, + { + "epoch": 0.74, + "grad_norm": 1.8557915990182703, + "learning_rate": 3.232947969060828e-06, + "loss": 1.0578, + "step": 31601 + }, + { + "epoch": 0.74, + "grad_norm": 2.0600837175691256, + "learning_rate": 3.2323861966686055e-06, + "loss": 1.2041, + "step": 31602 + }, + { + "epoch": 0.74, + "grad_norm": 2.6858903000332366, + "learning_rate": 3.231824463679569e-06, + "loss": 0.9585, + "step": 31603 + }, + { + "epoch": 0.74, + "grad_norm": 1.8149656104813037, + "learning_rate": 3.231262770096997e-06, + "loss": 1.0364, + "step": 31604 + }, + { + "epoch": 0.74, + "grad_norm": 2.014638461050368, + "learning_rate": 3.230701115924152e-06, + "loss": 0.9721, + "step": 31605 + }, + { + "epoch": 0.74, + "grad_norm": 1.9247720249662028, + "learning_rate": 3.2301395011643124e-06, + "loss": 0.8894, + "step": 31606 + }, + { + "epoch": 0.74, + "grad_norm": 2.0007657944372954, + "learning_rate": 3.2295779258207426e-06, + "loss": 0.9812, + "step": 31607 + }, + { + "epoch": 0.74, + "grad_norm": 1.6939636338117436, + "learning_rate": 3.229016389896712e-06, + "loss": 0.861, + "step": 31608 + }, + { + "epoch": 0.74, + "grad_norm": 1.113981831697712, + "learning_rate": 3.2284548933954905e-06, + "loss": 0.8325, + "step": 31609 + }, + { + "epoch": 0.74, + "grad_norm": 2.1576739494236397, + "learning_rate": 3.227893436320353e-06, + "loss": 1.0565, + "step": 31610 + }, + { + "epoch": 0.74, + "grad_norm": 2.062121376524584, + "learning_rate": 3.2273320186745625e-06, + "loss": 0.9132, + "step": 31611 + }, + { + "epoch": 0.74, + "grad_norm": 1.9058340407343484, + "learning_rate": 3.2267706404613864e-06, + "loss": 0.8662, + "step": 31612 + }, + { + "epoch": 0.74, + "grad_norm": 1.83176637335542, + "learning_rate": 3.2262093016840955e-06, + "loss": 0.8683, + "step": 31613 + }, + { + "epoch": 0.74, + "grad_norm": 1.1236655782527005, + "learning_rate": 3.2256480023459633e-06, + "loss": 0.8857, + "step": 31614 + }, + { + "epoch": 0.74, + "grad_norm": 1.8816473934596496, + "learning_rate": 3.225086742450252e-06, + "loss": 1.0214, + "step": 31615 + }, + { + "epoch": 0.74, + "grad_norm": 2.0521331019022324, + "learning_rate": 3.224525522000228e-06, + "loss": 0.9111, + "step": 31616 + }, + { + "epoch": 0.74, + "grad_norm": 2.160781197616874, + "learning_rate": 3.2239643409991607e-06, + "loss": 0.8627, + "step": 31617 + }, + { + "epoch": 0.74, + "grad_norm": 2.0142120470818674, + "learning_rate": 3.2234031994503223e-06, + "loss": 0.8966, + "step": 31618 + }, + { + "epoch": 0.74, + "grad_norm": 1.962886430726386, + "learning_rate": 3.2228420973569753e-06, + "loss": 0.9485, + "step": 31619 + }, + { + "epoch": 0.74, + "grad_norm": 2.0592084337121426, + "learning_rate": 3.2222810347223845e-06, + "loss": 0.9994, + "step": 31620 + }, + { + "epoch": 0.74, + "grad_norm": 2.1846116786377547, + "learning_rate": 3.2217200115498216e-06, + "loss": 0.9353, + "step": 31621 + }, + { + "epoch": 0.74, + "grad_norm": 2.550025947188757, + "learning_rate": 3.2211590278425486e-06, + "loss": 0.9729, + "step": 31622 + }, + { + "epoch": 0.75, + "grad_norm": 2.1978358371271716, + "learning_rate": 3.2205980836038363e-06, + "loss": 0.9975, + "step": 31623 + }, + { + "epoch": 0.75, + "grad_norm": 1.9267493623977452, + "learning_rate": 3.2200371788369457e-06, + "loss": 1.0589, + "step": 31624 + }, + { + "epoch": 0.75, + "grad_norm": 1.1476857417639692, + "learning_rate": 3.219476313545148e-06, + "loss": 0.9522, + "step": 31625 + }, + { + "epoch": 0.75, + "grad_norm": 2.051670970751611, + "learning_rate": 3.2189154877317032e-06, + "loss": 0.8618, + "step": 31626 + }, + { + "epoch": 0.75, + "grad_norm": 2.1566837726870496, + "learning_rate": 3.2183547013998816e-06, + "loss": 0.9612, + "step": 31627 + }, + { + "epoch": 0.75, + "grad_norm": 2.018003963383089, + "learning_rate": 3.217793954552946e-06, + "loss": 0.8065, + "step": 31628 + }, + { + "epoch": 0.75, + "grad_norm": 2.0678413084963374, + "learning_rate": 3.217233247194158e-06, + "loss": 0.9151, + "step": 31629 + }, + { + "epoch": 0.75, + "grad_norm": 1.1501297776448764, + "learning_rate": 3.216672579326786e-06, + "loss": 0.9295, + "step": 31630 + }, + { + "epoch": 0.75, + "grad_norm": 1.08696510215669, + "learning_rate": 3.216111950954094e-06, + "loss": 0.9479, + "step": 31631 + }, + { + "epoch": 0.75, + "grad_norm": 2.604690214716624, + "learning_rate": 3.2155513620793523e-06, + "loss": 0.9959, + "step": 31632 + }, + { + "epoch": 0.75, + "grad_norm": 2.0888684387234493, + "learning_rate": 3.2149908127058116e-06, + "loss": 0.9999, + "step": 31633 + }, + { + "epoch": 0.75, + "grad_norm": 1.9572169332754659, + "learning_rate": 3.214430302836743e-06, + "loss": 0.9393, + "step": 31634 + }, + { + "epoch": 0.75, + "grad_norm": 2.0469985333253824, + "learning_rate": 3.2138698324754126e-06, + "loss": 1.1024, + "step": 31635 + }, + { + "epoch": 0.75, + "grad_norm": 2.124479609097525, + "learning_rate": 3.2133094016250797e-06, + "loss": 1.0276, + "step": 31636 + }, + { + "epoch": 0.75, + "grad_norm": 1.7171618729463223, + "learning_rate": 3.212749010289006e-06, + "loss": 1.0045, + "step": 31637 + }, + { + "epoch": 0.75, + "grad_norm": 1.6913044752419608, + "learning_rate": 3.2121886584704553e-06, + "loss": 1.0368, + "step": 31638 + }, + { + "epoch": 0.75, + "grad_norm": 2.1397449965868294, + "learning_rate": 3.2116283461726948e-06, + "loss": 1.0528, + "step": 31639 + }, + { + "epoch": 0.75, + "grad_norm": 2.036404596115274, + "learning_rate": 3.21106807339898e-06, + "loss": 0.9834, + "step": 31640 + }, + { + "epoch": 0.75, + "grad_norm": 1.982481370347892, + "learning_rate": 3.210507840152579e-06, + "loss": 0.9682, + "step": 31641 + }, + { + "epoch": 0.75, + "grad_norm": 2.016084421943994, + "learning_rate": 3.209947646436752e-06, + "loss": 0.8865, + "step": 31642 + }, + { + "epoch": 0.75, + "grad_norm": 2.206699590849828, + "learning_rate": 3.2093874922547563e-06, + "loss": 0.999, + "step": 31643 + }, + { + "epoch": 0.75, + "grad_norm": 1.8069623041518714, + "learning_rate": 3.2088273776098554e-06, + "loss": 0.9042, + "step": 31644 + }, + { + "epoch": 0.75, + "grad_norm": 1.9089615316704447, + "learning_rate": 3.2082673025053157e-06, + "loss": 0.9513, + "step": 31645 + }, + { + "epoch": 0.75, + "grad_norm": 1.8990313539121273, + "learning_rate": 3.207707266944393e-06, + "loss": 1.0282, + "step": 31646 + }, + { + "epoch": 0.75, + "grad_norm": 2.295071744692126, + "learning_rate": 3.2071472709303466e-06, + "loss": 0.9655, + "step": 31647 + }, + { + "epoch": 0.75, + "grad_norm": 1.9650434754876323, + "learning_rate": 3.2065873144664393e-06, + "loss": 0.9125, + "step": 31648 + }, + { + "epoch": 0.75, + "grad_norm": 1.8791292744942274, + "learning_rate": 3.2060273975559374e-06, + "loss": 0.8753, + "step": 31649 + }, + { + "epoch": 0.75, + "grad_norm": 1.8920281378005475, + "learning_rate": 3.205467520202089e-06, + "loss": 0.8762, + "step": 31650 + }, + { + "epoch": 0.75, + "grad_norm": 2.1464452890745904, + "learning_rate": 3.2049076824081595e-06, + "loss": 0.8883, + "step": 31651 + }, + { + "epoch": 0.75, + "grad_norm": 1.9177351040131094, + "learning_rate": 3.204347884177409e-06, + "loss": 0.946, + "step": 31652 + }, + { + "epoch": 0.75, + "grad_norm": 2.2230222452859145, + "learning_rate": 3.203788125513099e-06, + "loss": 1.0075, + "step": 31653 + }, + { + "epoch": 0.75, + "grad_norm": 2.094848221800558, + "learning_rate": 3.2032284064184872e-06, + "loss": 0.9424, + "step": 31654 + }, + { + "epoch": 0.75, + "grad_norm": 2.9535972427582893, + "learning_rate": 3.202668726896828e-06, + "loss": 0.9192, + "step": 31655 + }, + { + "epoch": 0.75, + "grad_norm": 2.1022007694852696, + "learning_rate": 3.202109086951386e-06, + "loss": 0.9545, + "step": 31656 + }, + { + "epoch": 0.75, + "grad_norm": 1.9868779933437557, + "learning_rate": 3.2015494865854144e-06, + "loss": 1.1074, + "step": 31657 + }, + { + "epoch": 0.75, + "grad_norm": 2.468951544030673, + "learning_rate": 3.200989925802177e-06, + "loss": 1.0609, + "step": 31658 + }, + { + "epoch": 0.75, + "grad_norm": 2.14974662370379, + "learning_rate": 3.200430404604926e-06, + "loss": 0.9168, + "step": 31659 + }, + { + "epoch": 0.75, + "grad_norm": 1.9206640231282763, + "learning_rate": 3.199870922996925e-06, + "loss": 1.003, + "step": 31660 + }, + { + "epoch": 0.75, + "grad_norm": 2.4097289875772856, + "learning_rate": 3.199311480981425e-06, + "loss": 0.979, + "step": 31661 + }, + { + "epoch": 0.75, + "grad_norm": 2.0484809210256762, + "learning_rate": 3.1987520785616898e-06, + "loss": 0.9569, + "step": 31662 + }, + { + "epoch": 0.75, + "grad_norm": 1.9916717586321728, + "learning_rate": 3.198192715740973e-06, + "loss": 1.0482, + "step": 31663 + }, + { + "epoch": 0.75, + "grad_norm": 2.062957142814974, + "learning_rate": 3.197633392522529e-06, + "loss": 0.8986, + "step": 31664 + }, + { + "epoch": 0.75, + "grad_norm": 2.1184167837584873, + "learning_rate": 3.1970741089096168e-06, + "loss": 0.9491, + "step": 31665 + }, + { + "epoch": 0.75, + "grad_norm": 1.976521841321766, + "learning_rate": 3.196514864905497e-06, + "loss": 1.091, + "step": 31666 + }, + { + "epoch": 0.75, + "grad_norm": 1.9752022709625379, + "learning_rate": 3.19595566051342e-06, + "loss": 0.9679, + "step": 31667 + }, + { + "epoch": 0.75, + "grad_norm": 2.489790585560434, + "learning_rate": 3.1953964957366413e-06, + "loss": 0.9995, + "step": 31668 + }, + { + "epoch": 0.75, + "grad_norm": 2.187618308295308, + "learning_rate": 3.1948373705784173e-06, + "loss": 0.9322, + "step": 31669 + }, + { + "epoch": 0.75, + "grad_norm": 2.230678190575367, + "learning_rate": 3.1942782850420086e-06, + "loss": 0.9077, + "step": 31670 + }, + { + "epoch": 0.75, + "grad_norm": 2.184651434700545, + "learning_rate": 3.193719239130666e-06, + "loss": 1.0424, + "step": 31671 + }, + { + "epoch": 0.75, + "grad_norm": 1.9756789461343012, + "learning_rate": 3.193160232847642e-06, + "loss": 0.9734, + "step": 31672 + }, + { + "epoch": 0.75, + "grad_norm": 1.113490173119153, + "learning_rate": 3.1926012661961935e-06, + "loss": 0.9372, + "step": 31673 + }, + { + "epoch": 0.75, + "grad_norm": 1.841233259398368, + "learning_rate": 3.1920423391795795e-06, + "loss": 0.9669, + "step": 31674 + }, + { + "epoch": 0.75, + "grad_norm": 2.5598117255331414, + "learning_rate": 3.1914834518010506e-06, + "loss": 0.9221, + "step": 31675 + }, + { + "epoch": 0.75, + "grad_norm": 1.1344259267812307, + "learning_rate": 3.1909246040638562e-06, + "loss": 0.8968, + "step": 31676 + }, + { + "epoch": 0.75, + "grad_norm": 3.656299436496849, + "learning_rate": 3.1903657959712585e-06, + "loss": 1.0737, + "step": 31677 + }, + { + "epoch": 0.75, + "grad_norm": 1.99217941291888, + "learning_rate": 3.1898070275265035e-06, + "loss": 1.073, + "step": 31678 + }, + { + "epoch": 0.75, + "grad_norm": 1.8808223077254866, + "learning_rate": 3.189248298732852e-06, + "loss": 0.9415, + "step": 31679 + }, + { + "epoch": 0.75, + "grad_norm": 2.304963135400586, + "learning_rate": 3.1886896095935494e-06, + "loss": 0.9246, + "step": 31680 + }, + { + "epoch": 0.75, + "grad_norm": 1.945771133359716, + "learning_rate": 3.188130960111856e-06, + "loss": 0.9052, + "step": 31681 + }, + { + "epoch": 0.75, + "grad_norm": 2.339463652448396, + "learning_rate": 3.1875723502910173e-06, + "loss": 1.0717, + "step": 31682 + }, + { + "epoch": 0.75, + "grad_norm": 2.0243195602208948, + "learning_rate": 3.1870137801342917e-06, + "loss": 0.9289, + "step": 31683 + }, + { + "epoch": 0.75, + "grad_norm": 1.9688683664174835, + "learning_rate": 3.1864552496449297e-06, + "loss": 0.9626, + "step": 31684 + }, + { + "epoch": 0.75, + "grad_norm": 2.046531160381946, + "learning_rate": 3.185896758826178e-06, + "loss": 1.0734, + "step": 31685 + }, + { + "epoch": 0.75, + "grad_norm": 1.9466992992394179, + "learning_rate": 3.185338307681294e-06, + "loss": 1.0474, + "step": 31686 + }, + { + "epoch": 0.75, + "grad_norm": 2.460590778157638, + "learning_rate": 3.18477989621353e-06, + "loss": 0.9827, + "step": 31687 + }, + { + "epoch": 0.75, + "grad_norm": 2.127646244187257, + "learning_rate": 3.1842215244261356e-06, + "loss": 1.0633, + "step": 31688 + }, + { + "epoch": 0.75, + "grad_norm": 1.928265715529068, + "learning_rate": 3.1836631923223583e-06, + "loss": 0.9371, + "step": 31689 + }, + { + "epoch": 0.75, + "grad_norm": 1.9849034641979686, + "learning_rate": 3.1831048999054513e-06, + "loss": 0.9486, + "step": 31690 + }, + { + "epoch": 0.75, + "grad_norm": 2.008253283626486, + "learning_rate": 3.1825466471786693e-06, + "loss": 0.9826, + "step": 31691 + }, + { + "epoch": 0.75, + "grad_norm": 2.003821881857793, + "learning_rate": 3.18198843414526e-06, + "loss": 0.8774, + "step": 31692 + }, + { + "epoch": 0.75, + "grad_norm": 2.360175238321897, + "learning_rate": 3.1814302608084678e-06, + "loss": 0.9451, + "step": 31693 + }, + { + "epoch": 0.75, + "grad_norm": 1.1538167076212988, + "learning_rate": 3.180872127171548e-06, + "loss": 0.8716, + "step": 31694 + }, + { + "epoch": 0.75, + "grad_norm": 1.8687524926934873, + "learning_rate": 3.1803140332377535e-06, + "loss": 1.0377, + "step": 31695 + }, + { + "epoch": 0.75, + "grad_norm": 1.987756097042124, + "learning_rate": 3.1797559790103293e-06, + "loss": 0.9728, + "step": 31696 + }, + { + "epoch": 0.75, + "grad_norm": 2.2481288091276244, + "learning_rate": 3.179197964492522e-06, + "loss": 1.0256, + "step": 31697 + }, + { + "epoch": 0.75, + "grad_norm": 1.7166129589841836, + "learning_rate": 3.1786399896875862e-06, + "loss": 0.9986, + "step": 31698 + }, + { + "epoch": 0.75, + "grad_norm": 2.157984379000611, + "learning_rate": 3.1780820545987647e-06, + "loss": 1.0591, + "step": 31699 + }, + { + "epoch": 0.75, + "grad_norm": 1.8195317697145514, + "learning_rate": 3.1775241592293127e-06, + "loss": 1.1316, + "step": 31700 + }, + { + "epoch": 0.75, + "grad_norm": 4.299825915364999, + "learning_rate": 3.176966303582473e-06, + "loss": 0.9016, + "step": 31701 + }, + { + "epoch": 0.75, + "grad_norm": 1.8729548984862938, + "learning_rate": 3.1764084876614965e-06, + "loss": 1.001, + "step": 31702 + }, + { + "epoch": 0.75, + "grad_norm": 1.132521849374388, + "learning_rate": 3.175850711469629e-06, + "loss": 0.9684, + "step": 31703 + }, + { + "epoch": 0.75, + "grad_norm": 2.0058059738669054, + "learning_rate": 3.1752929750101216e-06, + "loss": 1.0046, + "step": 31704 + }, + { + "epoch": 0.75, + "grad_norm": 2.2694748804384903, + "learning_rate": 3.1747352782862185e-06, + "loss": 0.8928, + "step": 31705 + }, + { + "epoch": 0.75, + "grad_norm": 3.4324894232868735, + "learning_rate": 3.1741776213011643e-06, + "loss": 0.9842, + "step": 31706 + }, + { + "epoch": 0.75, + "grad_norm": 1.858699107240585, + "learning_rate": 3.17362000405821e-06, + "loss": 0.9557, + "step": 31707 + }, + { + "epoch": 0.75, + "grad_norm": 1.8783525986547829, + "learning_rate": 3.1730624265606037e-06, + "loss": 0.8454, + "step": 31708 + }, + { + "epoch": 0.75, + "grad_norm": 1.954546718755387, + "learning_rate": 3.1725048888115893e-06, + "loss": 1.0097, + "step": 31709 + }, + { + "epoch": 0.75, + "grad_norm": 2.347682329168281, + "learning_rate": 3.1719473908144106e-06, + "loss": 0.9624, + "step": 31710 + }, + { + "epoch": 0.75, + "grad_norm": 2.1847949058578173, + "learning_rate": 3.1713899325723153e-06, + "loss": 1.027, + "step": 31711 + }, + { + "epoch": 0.75, + "grad_norm": 2.0036968064583767, + "learning_rate": 3.1708325140885533e-06, + "loss": 0.8321, + "step": 31712 + }, + { + "epoch": 0.75, + "grad_norm": 1.9969928269405508, + "learning_rate": 3.1702751353663664e-06, + "loss": 1.0347, + "step": 31713 + }, + { + "epoch": 0.75, + "grad_norm": 2.3684080459324135, + "learning_rate": 3.1697177964089963e-06, + "loss": 0.9197, + "step": 31714 + }, + { + "epoch": 0.75, + "grad_norm": 1.9668099776790917, + "learning_rate": 3.1691604972196922e-06, + "loss": 1.0585, + "step": 31715 + }, + { + "epoch": 0.75, + "grad_norm": 1.843476844953827, + "learning_rate": 3.168603237801701e-06, + "loss": 0.9322, + "step": 31716 + }, + { + "epoch": 0.75, + "grad_norm": 2.2625071210026313, + "learning_rate": 3.168046018158264e-06, + "loss": 1.0129, + "step": 31717 + }, + { + "epoch": 0.75, + "grad_norm": 2.208152427033152, + "learning_rate": 3.1674888382926237e-06, + "loss": 0.9806, + "step": 31718 + }, + { + "epoch": 0.75, + "grad_norm": 1.8205124192445075, + "learning_rate": 3.1669316982080268e-06, + "loss": 0.955, + "step": 31719 + }, + { + "epoch": 0.75, + "grad_norm": 1.8258187697207315, + "learning_rate": 3.1663745979077186e-06, + "loss": 1.0131, + "step": 31720 + }, + { + "epoch": 0.75, + "grad_norm": 1.9896081778933727, + "learning_rate": 3.165817537394943e-06, + "loss": 0.9719, + "step": 31721 + }, + { + "epoch": 0.75, + "grad_norm": 3.0278258123673845, + "learning_rate": 3.1652605166729367e-06, + "loss": 1.0016, + "step": 31722 + }, + { + "epoch": 0.75, + "grad_norm": 1.9714480447520641, + "learning_rate": 3.164703535744951e-06, + "loss": 1.0035, + "step": 31723 + }, + { + "epoch": 0.75, + "grad_norm": 1.8786544874010642, + "learning_rate": 3.1641465946142223e-06, + "loss": 0.8607, + "step": 31724 + }, + { + "epoch": 0.75, + "grad_norm": 1.044807779130388, + "learning_rate": 3.1635896932839994e-06, + "loss": 0.9238, + "step": 31725 + }, + { + "epoch": 0.75, + "grad_norm": 2.0492488367881188, + "learning_rate": 3.1630328317575187e-06, + "loss": 0.9234, + "step": 31726 + }, + { + "epoch": 0.75, + "grad_norm": 1.0622176486993882, + "learning_rate": 3.162476010038028e-06, + "loss": 0.9388, + "step": 31727 + }, + { + "epoch": 0.75, + "grad_norm": 1.9905918377532992, + "learning_rate": 3.1619192281287647e-06, + "loss": 0.9338, + "step": 31728 + }, + { + "epoch": 0.75, + "grad_norm": 1.9195171998238134, + "learning_rate": 3.1613624860329716e-06, + "loss": 0.8088, + "step": 31729 + }, + { + "epoch": 0.75, + "grad_norm": 2.084646565323913, + "learning_rate": 3.1608057837538976e-06, + "loss": 0.8418, + "step": 31730 + }, + { + "epoch": 0.75, + "grad_norm": 1.9286719761305393, + "learning_rate": 3.1602491212947717e-06, + "loss": 0.9748, + "step": 31731 + }, + { + "epoch": 0.75, + "grad_norm": 2.00319764206218, + "learning_rate": 3.1596924986588397e-06, + "loss": 0.9711, + "step": 31732 + }, + { + "epoch": 0.75, + "grad_norm": 2.0509176210066404, + "learning_rate": 3.1591359158493453e-06, + "loss": 1.0874, + "step": 31733 + }, + { + "epoch": 0.75, + "grad_norm": 1.8928888198023728, + "learning_rate": 3.1585793728695325e-06, + "loss": 1.1449, + "step": 31734 + }, + { + "epoch": 0.75, + "grad_norm": 2.13067506793788, + "learning_rate": 3.15802286972263e-06, + "loss": 1.0179, + "step": 31735 + }, + { + "epoch": 0.75, + "grad_norm": 2.0103795287631137, + "learning_rate": 3.1574664064118856e-06, + "loss": 0.8945, + "step": 31736 + }, + { + "epoch": 0.75, + "grad_norm": 2.0274453400114862, + "learning_rate": 3.15690998294054e-06, + "loss": 0.9441, + "step": 31737 + }, + { + "epoch": 0.75, + "grad_norm": 1.970395945919708, + "learning_rate": 3.1563535993118287e-06, + "loss": 0.8269, + "step": 31738 + }, + { + "epoch": 0.75, + "grad_norm": 4.320112140160965, + "learning_rate": 3.155797255528996e-06, + "loss": 1.0883, + "step": 31739 + }, + { + "epoch": 0.75, + "grad_norm": 1.8102641925322993, + "learning_rate": 3.1552409515952764e-06, + "loss": 0.8839, + "step": 31740 + }, + { + "epoch": 0.75, + "grad_norm": 1.9843248546123533, + "learning_rate": 3.1546846875139136e-06, + "loss": 0.979, + "step": 31741 + }, + { + "epoch": 0.75, + "grad_norm": 1.9625224867816793, + "learning_rate": 3.15412846328814e-06, + "loss": 1.0107, + "step": 31742 + }, + { + "epoch": 0.75, + "grad_norm": 1.9455043069083904, + "learning_rate": 3.153572278921202e-06, + "loss": 0.9238, + "step": 31743 + }, + { + "epoch": 0.75, + "grad_norm": 2.1434131953610183, + "learning_rate": 3.1530161344163324e-06, + "loss": 0.8963, + "step": 31744 + }, + { + "epoch": 0.75, + "grad_norm": 1.1242158189263816, + "learning_rate": 3.1524600297767682e-06, + "loss": 0.9461, + "step": 31745 + }, + { + "epoch": 0.75, + "grad_norm": 2.182608700742159, + "learning_rate": 3.1519039650057502e-06, + "loss": 0.894, + "step": 31746 + }, + { + "epoch": 0.75, + "grad_norm": 2.1834918406330908, + "learning_rate": 3.1513479401065174e-06, + "loss": 0.9265, + "step": 31747 + }, + { + "epoch": 0.75, + "grad_norm": 2.19345901694755, + "learning_rate": 3.1507919550823063e-06, + "loss": 0.8645, + "step": 31748 + }, + { + "epoch": 0.75, + "grad_norm": 2.480284276057394, + "learning_rate": 3.150236009936349e-06, + "loss": 1.0302, + "step": 31749 + }, + { + "epoch": 0.75, + "grad_norm": 1.9785514148824723, + "learning_rate": 3.149680104671886e-06, + "loss": 1.0061, + "step": 31750 + }, + { + "epoch": 0.75, + "grad_norm": 2.0406307996149735, + "learning_rate": 3.149124239292158e-06, + "loss": 1.0145, + "step": 31751 + }, + { + "epoch": 0.75, + "grad_norm": 2.586431978055168, + "learning_rate": 3.148568413800397e-06, + "loss": 0.9954, + "step": 31752 + }, + { + "epoch": 0.75, + "grad_norm": 1.0728826273053262, + "learning_rate": 3.148012628199837e-06, + "loss": 0.9841, + "step": 31753 + }, + { + "epoch": 0.75, + "grad_norm": 3.948785659007644, + "learning_rate": 3.147456882493717e-06, + "loss": 0.9358, + "step": 31754 + }, + { + "epoch": 0.75, + "grad_norm": 1.7929566903425562, + "learning_rate": 3.1469011766852754e-06, + "loss": 1.0257, + "step": 31755 + }, + { + "epoch": 0.75, + "grad_norm": 2.0076042684929303, + "learning_rate": 3.146345510777745e-06, + "loss": 1.0774, + "step": 31756 + }, + { + "epoch": 0.75, + "grad_norm": 2.808115405700406, + "learning_rate": 3.1457898847743573e-06, + "loss": 1.0835, + "step": 31757 + }, + { + "epoch": 0.75, + "grad_norm": 2.1818746340523782, + "learning_rate": 3.1452342986783545e-06, + "loss": 1.0263, + "step": 31758 + }, + { + "epoch": 0.75, + "grad_norm": 2.007272179958338, + "learning_rate": 3.1446787524929644e-06, + "loss": 0.9309, + "step": 31759 + }, + { + "epoch": 0.75, + "grad_norm": 1.1451955813470267, + "learning_rate": 3.144123246221428e-06, + "loss": 0.9315, + "step": 31760 + }, + { + "epoch": 0.75, + "grad_norm": 2.5236895331279356, + "learning_rate": 3.1435677798669738e-06, + "loss": 0.9119, + "step": 31761 + }, + { + "epoch": 0.75, + "grad_norm": 2.0152642645516847, + "learning_rate": 3.1430123534328415e-06, + "loss": 0.9738, + "step": 31762 + }, + { + "epoch": 0.75, + "grad_norm": 2.1064939985144573, + "learning_rate": 3.142456966922259e-06, + "loss": 1.0257, + "step": 31763 + }, + { + "epoch": 0.75, + "grad_norm": 1.097008059082421, + "learning_rate": 3.1419016203384666e-06, + "loss": 0.8774, + "step": 31764 + }, + { + "epoch": 0.75, + "grad_norm": 1.8826340712861624, + "learning_rate": 3.1413463136846945e-06, + "loss": 0.9984, + "step": 31765 + }, + { + "epoch": 0.75, + "grad_norm": 1.990179651158941, + "learning_rate": 3.140791046964172e-06, + "loss": 1.0343, + "step": 31766 + }, + { + "epoch": 0.75, + "grad_norm": 2.253804017090036, + "learning_rate": 3.1402358201801355e-06, + "loss": 0.9116, + "step": 31767 + }, + { + "epoch": 0.75, + "grad_norm": 1.8381847529835023, + "learning_rate": 3.1396806333358208e-06, + "loss": 0.9708, + "step": 31768 + }, + { + "epoch": 0.75, + "grad_norm": 1.8625640966644867, + "learning_rate": 3.1391254864344577e-06, + "loss": 1.0111, + "step": 31769 + }, + { + "epoch": 0.75, + "grad_norm": 2.4871422947148423, + "learning_rate": 3.1385703794792742e-06, + "loss": 0.8607, + "step": 31770 + }, + { + "epoch": 0.75, + "grad_norm": 2.654319703261486, + "learning_rate": 3.138015312473508e-06, + "loss": 0.9202, + "step": 31771 + }, + { + "epoch": 0.75, + "grad_norm": 1.0111547808547694, + "learning_rate": 3.1374602854203908e-06, + "loss": 0.8616, + "step": 31772 + }, + { + "epoch": 0.75, + "grad_norm": 2.006761854549278, + "learning_rate": 3.1369052983231527e-06, + "loss": 1.1617, + "step": 31773 + }, + { + "epoch": 0.75, + "grad_norm": 2.234175321353115, + "learning_rate": 3.1363503511850222e-06, + "loss": 1.0669, + "step": 31774 + }, + { + "epoch": 0.75, + "grad_norm": 1.1636387719662176, + "learning_rate": 3.135795444009232e-06, + "loss": 0.9579, + "step": 31775 + }, + { + "epoch": 0.75, + "grad_norm": 2.019547655621859, + "learning_rate": 3.135240576799018e-06, + "loss": 0.9312, + "step": 31776 + }, + { + "epoch": 0.75, + "grad_norm": 2.058913642727343, + "learning_rate": 3.1346857495576066e-06, + "loss": 1.0766, + "step": 31777 + }, + { + "epoch": 0.75, + "grad_norm": 1.0999967928421073, + "learning_rate": 3.134130962288224e-06, + "loss": 0.9611, + "step": 31778 + }, + { + "epoch": 0.75, + "grad_norm": 1.8633720687709576, + "learning_rate": 3.1335762149941095e-06, + "loss": 0.85, + "step": 31779 + }, + { + "epoch": 0.75, + "grad_norm": 2.5881298560553043, + "learning_rate": 3.133021507678484e-06, + "loss": 0.972, + "step": 31780 + }, + { + "epoch": 0.75, + "grad_norm": 1.8701442300718751, + "learning_rate": 3.1324668403445845e-06, + "loss": 0.9937, + "step": 31781 + }, + { + "epoch": 0.75, + "grad_norm": 2.0381378909536902, + "learning_rate": 3.1319122129956337e-06, + "loss": 1.0662, + "step": 31782 + }, + { + "epoch": 0.75, + "grad_norm": 2.7400559568132445, + "learning_rate": 3.1313576256348687e-06, + "loss": 1.0646, + "step": 31783 + }, + { + "epoch": 0.75, + "grad_norm": 1.9985131980869628, + "learning_rate": 3.1308030782655106e-06, + "loss": 1.0622, + "step": 31784 + }, + { + "epoch": 0.75, + "grad_norm": 1.828722877172157, + "learning_rate": 3.130248570890795e-06, + "loss": 0.9004, + "step": 31785 + }, + { + "epoch": 0.75, + "grad_norm": 2.0573586904446524, + "learning_rate": 3.129694103513946e-06, + "loss": 1.0594, + "step": 31786 + }, + { + "epoch": 0.75, + "grad_norm": 1.929791497589834, + "learning_rate": 3.129139676138191e-06, + "loss": 0.9287, + "step": 31787 + }, + { + "epoch": 0.75, + "grad_norm": 1.854054453062543, + "learning_rate": 3.12858528876676e-06, + "loss": 0.9528, + "step": 31788 + }, + { + "epoch": 0.75, + "grad_norm": 2.098425070850271, + "learning_rate": 3.128030941402883e-06, + "loss": 0.9859, + "step": 31789 + }, + { + "epoch": 0.75, + "grad_norm": 2.7033068949387986, + "learning_rate": 3.1274766340497876e-06, + "loss": 1.045, + "step": 31790 + }, + { + "epoch": 0.75, + "grad_norm": 2.120026446807433, + "learning_rate": 3.126922366710694e-06, + "loss": 0.8289, + "step": 31791 + }, + { + "epoch": 0.75, + "grad_norm": 1.8783137435967794, + "learning_rate": 3.1263681393888347e-06, + "loss": 0.9069, + "step": 31792 + }, + { + "epoch": 0.75, + "grad_norm": 1.9034202676903722, + "learning_rate": 3.1258139520874396e-06, + "loss": 1.0096, + "step": 31793 + }, + { + "epoch": 0.75, + "grad_norm": 1.8325301956381785, + "learning_rate": 3.125259804809733e-06, + "loss": 0.9522, + "step": 31794 + }, + { + "epoch": 0.75, + "grad_norm": 2.21876083353749, + "learning_rate": 3.1247056975589364e-06, + "loss": 1.0183, + "step": 31795 + }, + { + "epoch": 0.75, + "grad_norm": 2.17408495111636, + "learning_rate": 3.124151630338279e-06, + "loss": 0.9816, + "step": 31796 + }, + { + "epoch": 0.75, + "grad_norm": 2.044138776318906, + "learning_rate": 3.123597603150992e-06, + "loss": 0.9861, + "step": 31797 + }, + { + "epoch": 0.75, + "grad_norm": 1.9730770318546147, + "learning_rate": 3.1230436160002963e-06, + "loss": 1.027, + "step": 31798 + }, + { + "epoch": 0.75, + "grad_norm": 2.0101821199543397, + "learning_rate": 3.1224896688894146e-06, + "loss": 1.0494, + "step": 31799 + }, + { + "epoch": 0.75, + "grad_norm": 1.9480749914183777, + "learning_rate": 3.1219357618215785e-06, + "loss": 0.9519, + "step": 31800 + }, + { + "epoch": 0.75, + "grad_norm": 1.939132691450278, + "learning_rate": 3.1213818948000074e-06, + "loss": 1.032, + "step": 31801 + }, + { + "epoch": 0.75, + "grad_norm": 2.1156561306430235, + "learning_rate": 3.120828067827931e-06, + "loss": 1.0007, + "step": 31802 + }, + { + "epoch": 0.75, + "grad_norm": 2.40578822869819, + "learning_rate": 3.1202742809085694e-06, + "loss": 0.9106, + "step": 31803 + }, + { + "epoch": 0.75, + "grad_norm": 2.5507384819853027, + "learning_rate": 3.119720534045152e-06, + "loss": 1.1381, + "step": 31804 + }, + { + "epoch": 0.75, + "grad_norm": 2.2623649442629947, + "learning_rate": 3.119166827240896e-06, + "loss": 0.9265, + "step": 31805 + }, + { + "epoch": 0.75, + "grad_norm": 1.8481521529439895, + "learning_rate": 3.118613160499032e-06, + "loss": 0.9381, + "step": 31806 + }, + { + "epoch": 0.75, + "grad_norm": 1.8248623473581163, + "learning_rate": 3.1180595338227827e-06, + "loss": 0.9912, + "step": 31807 + }, + { + "epoch": 0.75, + "grad_norm": 2.0022252666607177, + "learning_rate": 3.1175059472153648e-06, + "loss": 1.0201, + "step": 31808 + }, + { + "epoch": 0.75, + "grad_norm": 1.951240989269549, + "learning_rate": 3.1169524006800066e-06, + "loss": 1.1573, + "step": 31809 + }, + { + "epoch": 0.75, + "grad_norm": 2.050094879503577, + "learning_rate": 3.116398894219934e-06, + "loss": 1.0189, + "step": 31810 + }, + { + "epoch": 0.75, + "grad_norm": 1.9020367983388151, + "learning_rate": 3.1158454278383666e-06, + "loss": 0.8976, + "step": 31811 + }, + { + "epoch": 0.75, + "grad_norm": 1.9411953405204068, + "learning_rate": 3.1152920015385225e-06, + "loss": 0.951, + "step": 31812 + }, + { + "epoch": 0.75, + "grad_norm": 2.1494658784327147, + "learning_rate": 3.114738615323628e-06, + "loss": 1.1656, + "step": 31813 + }, + { + "epoch": 0.75, + "grad_norm": 2.184328485442685, + "learning_rate": 3.1141852691969098e-06, + "loss": 0.9208, + "step": 31814 + }, + { + "epoch": 0.75, + "grad_norm": 1.8560356117479757, + "learning_rate": 3.113631963161584e-06, + "loss": 0.841, + "step": 31815 + }, + { + "epoch": 0.75, + "grad_norm": 2.1423221914396087, + "learning_rate": 3.1130786972208694e-06, + "loss": 0.8952, + "step": 31816 + }, + { + "epoch": 0.75, + "grad_norm": 1.7407882739589544, + "learning_rate": 3.112525471377992e-06, + "loss": 0.9839, + "step": 31817 + }, + { + "epoch": 0.75, + "grad_norm": 1.9807089104148683, + "learning_rate": 3.1119722856361744e-06, + "loss": 1.0768, + "step": 31818 + }, + { + "epoch": 0.75, + "grad_norm": 2.320317185231417, + "learning_rate": 3.111419139998635e-06, + "loss": 1.0139, + "step": 31819 + }, + { + "epoch": 0.75, + "grad_norm": 1.9012852512979215, + "learning_rate": 3.1108660344685913e-06, + "loss": 0.9156, + "step": 31820 + }, + { + "epoch": 0.75, + "grad_norm": 1.9244020871763812, + "learning_rate": 3.1103129690492694e-06, + "loss": 1.0052, + "step": 31821 + }, + { + "epoch": 0.75, + "grad_norm": 1.9055517579702659, + "learning_rate": 3.109759943743884e-06, + "loss": 1.0272, + "step": 31822 + }, + { + "epoch": 0.75, + "grad_norm": 1.0825269054244266, + "learning_rate": 3.1092069585556573e-06, + "loss": 0.9594, + "step": 31823 + }, + { + "epoch": 0.75, + "grad_norm": 2.1523036428399602, + "learning_rate": 3.1086540134878117e-06, + "loss": 0.9153, + "step": 31824 + }, + { + "epoch": 0.75, + "grad_norm": 2.3272997179256674, + "learning_rate": 3.1081011085435653e-06, + "loss": 1.0368, + "step": 31825 + }, + { + "epoch": 0.75, + "grad_norm": 1.8437671619536886, + "learning_rate": 3.1075482437261326e-06, + "loss": 1.0597, + "step": 31826 + }, + { + "epoch": 0.75, + "grad_norm": 1.1016244517906983, + "learning_rate": 3.106995419038736e-06, + "loss": 0.9248, + "step": 31827 + }, + { + "epoch": 0.75, + "grad_norm": 2.025683801960223, + "learning_rate": 3.106442634484601e-06, + "loss": 0.8951, + "step": 31828 + }, + { + "epoch": 0.75, + "grad_norm": 2.068775998324751, + "learning_rate": 3.105889890066933e-06, + "loss": 1.0403, + "step": 31829 + }, + { + "epoch": 0.75, + "grad_norm": 2.2072976302381035, + "learning_rate": 3.105337185788957e-06, + "loss": 0.9425, + "step": 31830 + }, + { + "epoch": 0.75, + "grad_norm": 1.9200253572739878, + "learning_rate": 3.1047845216538895e-06, + "loss": 0.9949, + "step": 31831 + }, + { + "epoch": 0.75, + "grad_norm": 2.157579190806324, + "learning_rate": 3.1042318976649575e-06, + "loss": 0.8682, + "step": 31832 + }, + { + "epoch": 0.75, + "grad_norm": 2.0498964366170775, + "learning_rate": 3.1036793138253627e-06, + "loss": 0.9273, + "step": 31833 + }, + { + "epoch": 0.75, + "grad_norm": 1.8682171300079884, + "learning_rate": 3.103126770138332e-06, + "loss": 0.7299, + "step": 31834 + }, + { + "epoch": 0.75, + "grad_norm": 2.056686176934928, + "learning_rate": 3.1025742666070823e-06, + "loss": 1.0933, + "step": 31835 + }, + { + "epoch": 0.75, + "grad_norm": 1.868790759952599, + "learning_rate": 3.102021803234827e-06, + "loss": 0.9867, + "step": 31836 + }, + { + "epoch": 0.75, + "grad_norm": 2.0958630309954827, + "learning_rate": 3.1014693800247865e-06, + "loss": 0.9616, + "step": 31837 + }, + { + "epoch": 0.75, + "grad_norm": 1.9198210664072541, + "learning_rate": 3.1009169969801735e-06, + "loss": 1.0501, + "step": 31838 + }, + { + "epoch": 0.75, + "grad_norm": 1.794153863292003, + "learning_rate": 3.100364654104208e-06, + "loss": 0.9316, + "step": 31839 + }, + { + "epoch": 0.75, + "grad_norm": 2.088677836445851, + "learning_rate": 3.099812351400101e-06, + "loss": 0.9889, + "step": 31840 + }, + { + "epoch": 0.75, + "grad_norm": 1.8493235048930454, + "learning_rate": 3.0992600888710745e-06, + "loss": 1.0614, + "step": 31841 + }, + { + "epoch": 0.75, + "grad_norm": 2.015200508546285, + "learning_rate": 3.098707866520341e-06, + "loss": 0.7757, + "step": 31842 + }, + { + "epoch": 0.75, + "grad_norm": 1.9856162107105848, + "learning_rate": 3.098155684351111e-06, + "loss": 0.9616, + "step": 31843 + }, + { + "epoch": 0.75, + "grad_norm": 2.157317283880491, + "learning_rate": 3.097603542366604e-06, + "loss": 1.029, + "step": 31844 + }, + { + "epoch": 0.75, + "grad_norm": 2.1756063314800995, + "learning_rate": 3.097051440570038e-06, + "loss": 0.9564, + "step": 31845 + }, + { + "epoch": 0.75, + "grad_norm": 2.2095175927160713, + "learning_rate": 3.0964993789646236e-06, + "loss": 0.9218, + "step": 31846 + }, + { + "epoch": 0.75, + "grad_norm": 2.0873644318958084, + "learning_rate": 3.0959473575535725e-06, + "loss": 0.9895, + "step": 31847 + }, + { + "epoch": 0.75, + "grad_norm": 2.2986992755931936, + "learning_rate": 3.0953953763401013e-06, + "loss": 1.0217, + "step": 31848 + }, + { + "epoch": 0.75, + "grad_norm": 3.078552300849808, + "learning_rate": 3.094843435327427e-06, + "loss": 0.9506, + "step": 31849 + }, + { + "epoch": 0.75, + "grad_norm": 2.132512359342426, + "learning_rate": 3.0942915345187617e-06, + "loss": 0.9307, + "step": 31850 + }, + { + "epoch": 0.75, + "grad_norm": 1.1491310320923909, + "learning_rate": 3.093739673917313e-06, + "loss": 0.9553, + "step": 31851 + }, + { + "epoch": 0.75, + "grad_norm": 2.3365902255872326, + "learning_rate": 3.0931878535262994e-06, + "loss": 0.8866, + "step": 31852 + }, + { + "epoch": 0.75, + "grad_norm": 2.1773003982911714, + "learning_rate": 3.0926360733489345e-06, + "loss": 1.14, + "step": 31853 + }, + { + "epoch": 0.75, + "grad_norm": 1.1279669261127434, + "learning_rate": 3.092084333388431e-06, + "loss": 0.955, + "step": 31854 + }, + { + "epoch": 0.75, + "grad_norm": 2.199907841127119, + "learning_rate": 3.0915326336479947e-06, + "loss": 0.9149, + "step": 31855 + }, + { + "epoch": 0.75, + "grad_norm": 1.9032043921565975, + "learning_rate": 3.0909809741308437e-06, + "loss": 0.9985, + "step": 31856 + }, + { + "epoch": 0.75, + "grad_norm": 1.9202654167754405, + "learning_rate": 3.090429354840192e-06, + "loss": 1.0003, + "step": 31857 + }, + { + "epoch": 0.75, + "grad_norm": 2.023314644250962, + "learning_rate": 3.0898777757792477e-06, + "loss": 1.0163, + "step": 31858 + }, + { + "epoch": 0.75, + "grad_norm": 2.2068373369168084, + "learning_rate": 3.089326236951219e-06, + "loss": 0.9426, + "step": 31859 + }, + { + "epoch": 0.75, + "grad_norm": 2.173430388537201, + "learning_rate": 3.0887747383593245e-06, + "loss": 0.9694, + "step": 31860 + }, + { + "epoch": 0.75, + "grad_norm": 2.3853588798600933, + "learning_rate": 3.088223280006769e-06, + "loss": 1.0717, + "step": 31861 + }, + { + "epoch": 0.75, + "grad_norm": 1.8679814675063526, + "learning_rate": 3.087671861896768e-06, + "loss": 0.9104, + "step": 31862 + }, + { + "epoch": 0.75, + "grad_norm": 1.928320180747952, + "learning_rate": 3.0871204840325276e-06, + "loss": 0.9879, + "step": 31863 + }, + { + "epoch": 0.75, + "grad_norm": 2.119225752346846, + "learning_rate": 3.086569146417263e-06, + "loss": 1.0541, + "step": 31864 + }, + { + "epoch": 0.75, + "grad_norm": 1.8402362024514378, + "learning_rate": 3.086017849054178e-06, + "loss": 0.9879, + "step": 31865 + }, + { + "epoch": 0.75, + "grad_norm": 1.93935364391718, + "learning_rate": 3.08546659194649e-06, + "loss": 1.0122, + "step": 31866 + }, + { + "epoch": 0.75, + "grad_norm": 2.0894433051222836, + "learning_rate": 3.0849153750974036e-06, + "loss": 1.0984, + "step": 31867 + }, + { + "epoch": 0.75, + "grad_norm": 2.003887332116819, + "learning_rate": 3.084364198510127e-06, + "loss": 1.0663, + "step": 31868 + }, + { + "epoch": 0.75, + "grad_norm": 2.0689726529097046, + "learning_rate": 3.0838130621878704e-06, + "loss": 0.8782, + "step": 31869 + }, + { + "epoch": 0.75, + "grad_norm": 2.077772935061975, + "learning_rate": 3.0832619661338482e-06, + "loss": 1.1158, + "step": 31870 + }, + { + "epoch": 0.75, + "grad_norm": 1.0841925295522936, + "learning_rate": 3.0827109103512643e-06, + "loss": 0.9623, + "step": 31871 + }, + { + "epoch": 0.75, + "grad_norm": 2.09074238146847, + "learning_rate": 3.082159894843324e-06, + "loss": 0.9076, + "step": 31872 + }, + { + "epoch": 0.75, + "grad_norm": 2.4427472417133984, + "learning_rate": 3.081608919613238e-06, + "loss": 0.9229, + "step": 31873 + }, + { + "epoch": 0.75, + "grad_norm": 2.242990986562006, + "learning_rate": 3.0810579846642196e-06, + "loss": 0.8472, + "step": 31874 + }, + { + "epoch": 0.75, + "grad_norm": 1.9671801271330733, + "learning_rate": 3.0805070899994715e-06, + "loss": 1.0193, + "step": 31875 + }, + { + "epoch": 0.75, + "grad_norm": 2.0176406437067844, + "learning_rate": 3.079956235622199e-06, + "loss": 0.9804, + "step": 31876 + }, + { + "epoch": 0.75, + "grad_norm": 1.975209490553453, + "learning_rate": 3.079405421535613e-06, + "loss": 0.9648, + "step": 31877 + }, + { + "epoch": 0.75, + "grad_norm": 1.7709297274756302, + "learning_rate": 3.0788546477429214e-06, + "loss": 1.1093, + "step": 31878 + }, + { + "epoch": 0.75, + "grad_norm": 1.8929252362174838, + "learning_rate": 3.0783039142473305e-06, + "loss": 1.0025, + "step": 31879 + }, + { + "epoch": 0.75, + "grad_norm": 2.342116697848796, + "learning_rate": 3.0777532210520413e-06, + "loss": 0.9355, + "step": 31880 + }, + { + "epoch": 0.75, + "grad_norm": 1.0583038621582879, + "learning_rate": 3.077202568160268e-06, + "loss": 0.8865, + "step": 31881 + }, + { + "epoch": 0.75, + "grad_norm": 1.9990331941770907, + "learning_rate": 3.0766519555752095e-06, + "loss": 0.9732, + "step": 31882 + }, + { + "epoch": 0.75, + "grad_norm": 2.099637260917183, + "learning_rate": 3.0761013833000785e-06, + "loss": 1.0011, + "step": 31883 + }, + { + "epoch": 0.75, + "grad_norm": 1.8675040091267845, + "learning_rate": 3.075550851338075e-06, + "loss": 0.9397, + "step": 31884 + }, + { + "epoch": 0.75, + "grad_norm": 2.0963331700417736, + "learning_rate": 3.0750003596924095e-06, + "loss": 0.8959, + "step": 31885 + }, + { + "epoch": 0.75, + "grad_norm": 1.751611476802801, + "learning_rate": 3.0744499083662803e-06, + "loss": 0.9615, + "step": 31886 + }, + { + "epoch": 0.75, + "grad_norm": 2.0041311591075246, + "learning_rate": 3.0738994973629e-06, + "loss": 1.0146, + "step": 31887 + }, + { + "epoch": 0.75, + "grad_norm": 2.104264515645135, + "learning_rate": 3.07334912668547e-06, + "loss": 0.8947, + "step": 31888 + }, + { + "epoch": 0.75, + "grad_norm": 1.833759291319644, + "learning_rate": 3.0727987963371906e-06, + "loss": 0.9142, + "step": 31889 + }, + { + "epoch": 0.75, + "grad_norm": 1.9154634485714774, + "learning_rate": 3.0722485063212703e-06, + "loss": 0.8658, + "step": 31890 + }, + { + "epoch": 0.75, + "grad_norm": 2.1942034777102197, + "learning_rate": 3.071698256640915e-06, + "loss": 0.9435, + "step": 31891 + }, + { + "epoch": 0.75, + "grad_norm": 2.2246568264894924, + "learning_rate": 3.0711480472993262e-06, + "loss": 1.0312, + "step": 31892 + }, + { + "epoch": 0.75, + "grad_norm": 2.006158506550348, + "learning_rate": 3.070597878299705e-06, + "loss": 0.8851, + "step": 31893 + }, + { + "epoch": 0.75, + "grad_norm": 1.9581032453857967, + "learning_rate": 3.070047749645255e-06, + "loss": 0.9711, + "step": 31894 + }, + { + "epoch": 0.75, + "grad_norm": 2.033854538521517, + "learning_rate": 3.0694976613391845e-06, + "loss": 0.9818, + "step": 31895 + }, + { + "epoch": 0.75, + "grad_norm": 2.227772800578038, + "learning_rate": 3.068947613384694e-06, + "loss": 1.0836, + "step": 31896 + }, + { + "epoch": 0.75, + "grad_norm": 1.9419074640210283, + "learning_rate": 3.068397605784981e-06, + "loss": 0.9657, + "step": 31897 + }, + { + "epoch": 0.75, + "grad_norm": 2.0022971869426645, + "learning_rate": 3.067847638543251e-06, + "loss": 0.8671, + "step": 31898 + }, + { + "epoch": 0.75, + "grad_norm": 1.981260202613123, + "learning_rate": 3.0672977116627102e-06, + "loss": 0.998, + "step": 31899 + }, + { + "epoch": 0.75, + "grad_norm": 1.9439309374114775, + "learning_rate": 3.066747825146558e-06, + "loss": 0.8044, + "step": 31900 + }, + { + "epoch": 0.75, + "grad_norm": 1.0910080472407797, + "learning_rate": 3.06619797899799e-06, + "loss": 0.9831, + "step": 31901 + }, + { + "epoch": 0.75, + "grad_norm": 1.9067064692301556, + "learning_rate": 3.065648173220217e-06, + "loss": 0.901, + "step": 31902 + }, + { + "epoch": 0.75, + "grad_norm": 2.060433165296362, + "learning_rate": 3.0650984078164325e-06, + "loss": 0.9464, + "step": 31903 + }, + { + "epoch": 0.75, + "grad_norm": 2.001658603705577, + "learning_rate": 3.064548682789843e-06, + "loss": 1.0547, + "step": 31904 + }, + { + "epoch": 0.75, + "grad_norm": 1.7909957747882983, + "learning_rate": 3.0639989981436447e-06, + "loss": 0.9286, + "step": 31905 + }, + { + "epoch": 0.75, + "grad_norm": 2.520696598950765, + "learning_rate": 3.063449353881043e-06, + "loss": 0.9422, + "step": 31906 + }, + { + "epoch": 0.75, + "grad_norm": 2.1486818996048296, + "learning_rate": 3.0628997500052314e-06, + "loss": 1.1521, + "step": 31907 + }, + { + "epoch": 0.75, + "grad_norm": 2.1751750941936927, + "learning_rate": 3.062350186519417e-06, + "loss": 1.0815, + "step": 31908 + }, + { + "epoch": 0.75, + "grad_norm": 2.0496879098998924, + "learning_rate": 3.061800663426797e-06, + "loss": 1.0138, + "step": 31909 + }, + { + "epoch": 0.75, + "grad_norm": 2.1242603826450286, + "learning_rate": 3.0612511807305665e-06, + "loss": 0.8481, + "step": 31910 + }, + { + "epoch": 0.75, + "grad_norm": 2.097077277246488, + "learning_rate": 3.060701738433929e-06, + "loss": 1.0055, + "step": 31911 + }, + { + "epoch": 0.75, + "grad_norm": 1.7698994752354271, + "learning_rate": 3.060152336540082e-06, + "loss": 0.924, + "step": 31912 + }, + { + "epoch": 0.75, + "grad_norm": 2.11465055009411, + "learning_rate": 3.0596029750522316e-06, + "loss": 0.9642, + "step": 31913 + }, + { + "epoch": 0.75, + "grad_norm": 1.8235168123142294, + "learning_rate": 3.059053653973564e-06, + "loss": 0.8476, + "step": 31914 + }, + { + "epoch": 0.75, + "grad_norm": 2.1040326257269615, + "learning_rate": 3.0585043733072838e-06, + "loss": 1.0956, + "step": 31915 + }, + { + "epoch": 0.75, + "grad_norm": 1.0198432160293622, + "learning_rate": 3.0579551330565915e-06, + "loss": 0.9121, + "step": 31916 + }, + { + "epoch": 0.75, + "grad_norm": 2.3070222578679225, + "learning_rate": 3.057405933224682e-06, + "loss": 0.8466, + "step": 31917 + }, + { + "epoch": 0.75, + "grad_norm": 2.2710369705576396, + "learning_rate": 3.0568567738147505e-06, + "loss": 0.993, + "step": 31918 + }, + { + "epoch": 0.75, + "grad_norm": 1.830943321338145, + "learning_rate": 3.0563076548299965e-06, + "loss": 0.9743, + "step": 31919 + }, + { + "epoch": 0.75, + "grad_norm": 1.8582005929824397, + "learning_rate": 3.055758576273621e-06, + "loss": 1.0162, + "step": 31920 + }, + { + "epoch": 0.75, + "grad_norm": 2.042080075066077, + "learning_rate": 3.0552095381488144e-06, + "loss": 0.9821, + "step": 31921 + }, + { + "epoch": 0.75, + "grad_norm": 2.0021851781939097, + "learning_rate": 3.0546605404587803e-06, + "loss": 0.9399, + "step": 31922 + }, + { + "epoch": 0.75, + "grad_norm": 1.7511446240416344, + "learning_rate": 3.0541115832067115e-06, + "loss": 0.8761, + "step": 31923 + }, + { + "epoch": 0.75, + "grad_norm": 2.085806102891458, + "learning_rate": 3.0535626663957996e-06, + "loss": 0.9361, + "step": 31924 + }, + { + "epoch": 0.75, + "grad_norm": 2.0142804646251675, + "learning_rate": 3.0530137900292456e-06, + "loss": 1.0202, + "step": 31925 + }, + { + "epoch": 0.75, + "grad_norm": 1.9217655580313246, + "learning_rate": 3.052464954110249e-06, + "loss": 0.996, + "step": 31926 + }, + { + "epoch": 0.75, + "grad_norm": 1.1243731369431318, + "learning_rate": 3.051916158642e-06, + "loss": 0.9257, + "step": 31927 + }, + { + "epoch": 0.75, + "grad_norm": 1.951755154883771, + "learning_rate": 3.051367403627692e-06, + "loss": 0.9576, + "step": 31928 + }, + { + "epoch": 0.75, + "grad_norm": 2.258091330326241, + "learning_rate": 3.0508186890705214e-06, + "loss": 0.9763, + "step": 31929 + }, + { + "epoch": 0.75, + "grad_norm": 2.05857003886431, + "learning_rate": 3.050270014973693e-06, + "loss": 0.8605, + "step": 31930 + }, + { + "epoch": 0.75, + "grad_norm": 1.8381579072498484, + "learning_rate": 3.049721381340386e-06, + "loss": 0.9522, + "step": 31931 + }, + { + "epoch": 0.75, + "grad_norm": 1.1016022797809752, + "learning_rate": 3.049172788173801e-06, + "loss": 0.9208, + "step": 31932 + }, + { + "epoch": 0.75, + "grad_norm": 1.081311752498763, + "learning_rate": 3.0486242354771322e-06, + "loss": 0.9332, + "step": 31933 + }, + { + "epoch": 0.75, + "grad_norm": 1.0718278742633287, + "learning_rate": 3.0480757232535773e-06, + "loss": 0.9497, + "step": 31934 + }, + { + "epoch": 0.75, + "grad_norm": 1.9340542906491058, + "learning_rate": 3.047527251506327e-06, + "loss": 1.0487, + "step": 31935 + }, + { + "epoch": 0.75, + "grad_norm": 2.0013712983048704, + "learning_rate": 3.0469788202385707e-06, + "loss": 0.8934, + "step": 31936 + }, + { + "epoch": 0.75, + "grad_norm": 3.626675351904018, + "learning_rate": 3.0464304294535086e-06, + "loss": 0.9699, + "step": 31937 + }, + { + "epoch": 0.75, + "grad_norm": 2.2014474628030403, + "learning_rate": 3.0458820791543263e-06, + "loss": 0.8776, + "step": 31938 + }, + { + "epoch": 0.75, + "grad_norm": 1.8720670062271239, + "learning_rate": 3.045333769344223e-06, + "loss": 1.0436, + "step": 31939 + }, + { + "epoch": 0.75, + "grad_norm": 1.7502068526129455, + "learning_rate": 3.044785500026386e-06, + "loss": 0.7808, + "step": 31940 + }, + { + "epoch": 0.75, + "grad_norm": 2.082260775730057, + "learning_rate": 3.0442372712040123e-06, + "loss": 1.033, + "step": 31941 + }, + { + "epoch": 0.75, + "grad_norm": 1.9533295561244959, + "learning_rate": 3.043689082880289e-06, + "loss": 1.0433, + "step": 31942 + }, + { + "epoch": 0.75, + "grad_norm": 2.845505582459808, + "learning_rate": 3.043140935058412e-06, + "loss": 0.8843, + "step": 31943 + }, + { + "epoch": 0.75, + "grad_norm": 1.8248753063036016, + "learning_rate": 3.0425928277415716e-06, + "loss": 0.9722, + "step": 31944 + }, + { + "epoch": 0.75, + "grad_norm": 2.30732754977412, + "learning_rate": 3.0420447609329564e-06, + "loss": 1.0376, + "step": 31945 + }, + { + "epoch": 0.75, + "grad_norm": 1.9761284549628104, + "learning_rate": 3.041496734635758e-06, + "loss": 1.0016, + "step": 31946 + }, + { + "epoch": 0.75, + "grad_norm": 1.917844433018639, + "learning_rate": 3.040948748853172e-06, + "loss": 0.9598, + "step": 31947 + }, + { + "epoch": 0.75, + "grad_norm": 2.009127088539459, + "learning_rate": 3.0404008035883857e-06, + "loss": 1.0212, + "step": 31948 + }, + { + "epoch": 0.75, + "grad_norm": 2.03172215799353, + "learning_rate": 3.0398528988445875e-06, + "loss": 0.9483, + "step": 31949 + }, + { + "epoch": 0.75, + "grad_norm": 2.045101359594109, + "learning_rate": 3.0393050346249677e-06, + "loss": 1.018, + "step": 31950 + }, + { + "epoch": 0.75, + "grad_norm": 1.8737339822589398, + "learning_rate": 3.0387572109327214e-06, + "loss": 0.9466, + "step": 31951 + }, + { + "epoch": 0.75, + "grad_norm": 2.0634660714916797, + "learning_rate": 3.0382094277710337e-06, + "loss": 1.0431, + "step": 31952 + }, + { + "epoch": 0.75, + "grad_norm": 2.203573664844378, + "learning_rate": 3.037661685143093e-06, + "loss": 0.9715, + "step": 31953 + }, + { + "epoch": 0.75, + "grad_norm": 2.03543193324154, + "learning_rate": 3.037113983052089e-06, + "loss": 1.0676, + "step": 31954 + }, + { + "epoch": 0.75, + "grad_norm": 2.7150771507268243, + "learning_rate": 3.036566321501215e-06, + "loss": 0.9894, + "step": 31955 + }, + { + "epoch": 0.75, + "grad_norm": 1.8994492451564577, + "learning_rate": 3.036018700493658e-06, + "loss": 0.8358, + "step": 31956 + }, + { + "epoch": 0.75, + "grad_norm": 2.1097483868513596, + "learning_rate": 3.0354711200326004e-06, + "loss": 0.9947, + "step": 31957 + }, + { + "epoch": 0.75, + "grad_norm": 1.9741007086498394, + "learning_rate": 3.0349235801212383e-06, + "loss": 1.0637, + "step": 31958 + }, + { + "epoch": 0.75, + "grad_norm": 1.9645112798684297, + "learning_rate": 3.0343760807627533e-06, + "loss": 0.9463, + "step": 31959 + }, + { + "epoch": 0.75, + "grad_norm": 2.0508262756491145, + "learning_rate": 3.0338286219603385e-06, + "loss": 0.9982, + "step": 31960 + }, + { + "epoch": 0.75, + "grad_norm": 2.040067499509482, + "learning_rate": 3.033281203717177e-06, + "loss": 1.0494, + "step": 31961 + }, + { + "epoch": 0.75, + "grad_norm": 1.958649329023429, + "learning_rate": 3.0327338260364604e-06, + "loss": 0.9956, + "step": 31962 + }, + { + "epoch": 0.75, + "grad_norm": 1.919466293025377, + "learning_rate": 3.032186488921369e-06, + "loss": 0.9765, + "step": 31963 + }, + { + "epoch": 0.75, + "grad_norm": 1.7870294948480498, + "learning_rate": 3.0316391923750986e-06, + "loss": 1.0527, + "step": 31964 + }, + { + "epoch": 0.75, + "grad_norm": 2.400502800260315, + "learning_rate": 3.031091936400831e-06, + "loss": 0.9659, + "step": 31965 + }, + { + "epoch": 0.75, + "grad_norm": 2.36538111157118, + "learning_rate": 3.030544721001748e-06, + "loss": 1.0437, + "step": 31966 + }, + { + "epoch": 0.75, + "grad_norm": 1.9511668321660793, + "learning_rate": 3.0299975461810406e-06, + "loss": 0.8743, + "step": 31967 + }, + { + "epoch": 0.75, + "grad_norm": 2.4233649197014504, + "learning_rate": 3.029450411941898e-06, + "loss": 0.9655, + "step": 31968 + }, + { + "epoch": 0.75, + "grad_norm": 1.8732687717510634, + "learning_rate": 3.0289033182875015e-06, + "loss": 1.0178, + "step": 31969 + }, + { + "epoch": 0.75, + "grad_norm": 2.0196237463355122, + "learning_rate": 3.028356265221033e-06, + "loss": 0.9879, + "step": 31970 + }, + { + "epoch": 0.75, + "grad_norm": 2.7718486605285713, + "learning_rate": 3.0278092527456826e-06, + "loss": 0.8756, + "step": 31971 + }, + { + "epoch": 0.75, + "grad_norm": 1.9820717332069582, + "learning_rate": 3.0272622808646366e-06, + "loss": 0.9533, + "step": 31972 + }, + { + "epoch": 0.75, + "grad_norm": 1.834777154329427, + "learning_rate": 3.0267153495810776e-06, + "loss": 1.0172, + "step": 31973 + }, + { + "epoch": 0.75, + "grad_norm": 3.432091436644234, + "learning_rate": 3.0261684588981866e-06, + "loss": 0.9097, + "step": 31974 + }, + { + "epoch": 0.75, + "grad_norm": 1.8560887385089628, + "learning_rate": 3.0256216088191502e-06, + "loss": 0.9148, + "step": 31975 + }, + { + "epoch": 0.75, + "grad_norm": 1.971161683137006, + "learning_rate": 3.0250747993471563e-06, + "loss": 1.0199, + "step": 31976 + }, + { + "epoch": 0.75, + "grad_norm": 1.9777564506195313, + "learning_rate": 3.024528030485384e-06, + "loss": 1.1535, + "step": 31977 + }, + { + "epoch": 0.75, + "grad_norm": 1.0742058523444191, + "learning_rate": 3.023981302237016e-06, + "loss": 0.9464, + "step": 31978 + }, + { + "epoch": 0.75, + "grad_norm": 2.203326159028949, + "learning_rate": 3.0234346146052374e-06, + "loss": 0.9669, + "step": 31979 + }, + { + "epoch": 0.75, + "grad_norm": 2.330756918834734, + "learning_rate": 3.0228879675932343e-06, + "loss": 0.82, + "step": 31980 + }, + { + "epoch": 0.75, + "grad_norm": 1.89708262251229, + "learning_rate": 3.0223413612041865e-06, + "loss": 0.9564, + "step": 31981 + }, + { + "epoch": 0.75, + "grad_norm": 1.9563331940180697, + "learning_rate": 3.021794795441272e-06, + "loss": 1.0142, + "step": 31982 + }, + { + "epoch": 0.75, + "grad_norm": 1.8139151282704815, + "learning_rate": 3.0212482703076817e-06, + "loss": 1.0574, + "step": 31983 + }, + { + "epoch": 0.75, + "grad_norm": 2.1413117516853126, + "learning_rate": 3.0207017858065903e-06, + "loss": 0.8854, + "step": 31984 + }, + { + "epoch": 0.75, + "grad_norm": 2.05236676959766, + "learning_rate": 3.0201553419411857e-06, + "loss": 1.0063, + "step": 31985 + }, + { + "epoch": 0.75, + "grad_norm": 1.9730221728821666, + "learning_rate": 3.0196089387146433e-06, + "loss": 1.026, + "step": 31986 + }, + { + "epoch": 0.75, + "grad_norm": 2.205586597788508, + "learning_rate": 3.0190625761301506e-06, + "loss": 1.1015, + "step": 31987 + }, + { + "epoch": 0.75, + "grad_norm": 1.7561785176644216, + "learning_rate": 3.018516254190883e-06, + "loss": 1.004, + "step": 31988 + }, + { + "epoch": 0.75, + "grad_norm": 1.9846191007370346, + "learning_rate": 3.017969972900027e-06, + "loss": 0.8554, + "step": 31989 + }, + { + "epoch": 0.75, + "grad_norm": 1.8940878693386325, + "learning_rate": 3.0174237322607593e-06, + "loss": 0.921, + "step": 31990 + }, + { + "epoch": 0.75, + "grad_norm": 2.084402386983707, + "learning_rate": 3.016877532276259e-06, + "loss": 0.8726, + "step": 31991 + }, + { + "epoch": 0.75, + "grad_norm": 2.0549319547585205, + "learning_rate": 3.016331372949709e-06, + "loss": 0.8402, + "step": 31992 + }, + { + "epoch": 0.75, + "grad_norm": 1.8088991893231767, + "learning_rate": 3.0157852542842914e-06, + "loss": 0.991, + "step": 31993 + }, + { + "epoch": 0.75, + "grad_norm": 1.7449968798996414, + "learning_rate": 3.015239176283183e-06, + "loss": 0.8769, + "step": 31994 + }, + { + "epoch": 0.75, + "grad_norm": 2.92964853380039, + "learning_rate": 3.0146931389495604e-06, + "loss": 0.9703, + "step": 31995 + }, + { + "epoch": 0.75, + "grad_norm": 2.201012446264333, + "learning_rate": 3.0141471422866065e-06, + "loss": 0.9858, + "step": 31996 + }, + { + "epoch": 0.75, + "grad_norm": 2.085922776149726, + "learning_rate": 3.0136011862975022e-06, + "loss": 1.0578, + "step": 31997 + }, + { + "epoch": 0.75, + "grad_norm": 2.1643999857387026, + "learning_rate": 3.0130552709854235e-06, + "loss": 1.0578, + "step": 31998 + }, + { + "epoch": 0.75, + "grad_norm": 1.0206891269550646, + "learning_rate": 3.0125093963535458e-06, + "loss": 0.8383, + "step": 31999 + }, + { + "epoch": 0.75, + "grad_norm": 2.2245656689284905, + "learning_rate": 3.0119635624050515e-06, + "loss": 0.9961, + "step": 32000 + }, + { + "epoch": 0.75, + "grad_norm": 2.2685796897017836, + "learning_rate": 3.0114177691431203e-06, + "loss": 0.8795, + "step": 32001 + }, + { + "epoch": 0.75, + "grad_norm": 1.8306953817696325, + "learning_rate": 3.0108720165709273e-06, + "loss": 1.028, + "step": 32002 + }, + { + "epoch": 0.75, + "grad_norm": 2.0056896878502894, + "learning_rate": 3.0103263046916464e-06, + "loss": 0.9139, + "step": 32003 + }, + { + "epoch": 0.75, + "grad_norm": 2.235045397286962, + "learning_rate": 3.009780633508463e-06, + "loss": 0.8859, + "step": 32004 + }, + { + "epoch": 0.75, + "grad_norm": 1.9267998229544903, + "learning_rate": 3.0092350030245465e-06, + "loss": 0.9177, + "step": 32005 + }, + { + "epoch": 0.75, + "grad_norm": 1.0838829322765642, + "learning_rate": 3.00868941324308e-06, + "loss": 0.9798, + "step": 32006 + }, + { + "epoch": 0.75, + "grad_norm": 1.952225091444457, + "learning_rate": 3.0081438641672334e-06, + "loss": 0.9463, + "step": 32007 + }, + { + "epoch": 0.75, + "grad_norm": 1.9821825238839632, + "learning_rate": 3.0075983558001898e-06, + "loss": 0.9308, + "step": 32008 + }, + { + "epoch": 0.75, + "grad_norm": 2.0704889827870088, + "learning_rate": 3.00705288814512e-06, + "loss": 0.8043, + "step": 32009 + }, + { + "epoch": 0.75, + "grad_norm": 1.0789922056842023, + "learning_rate": 3.0065074612052024e-06, + "loss": 0.9591, + "step": 32010 + }, + { + "epoch": 0.75, + "grad_norm": 1.0808959520692294, + "learning_rate": 3.0059620749836182e-06, + "loss": 0.9026, + "step": 32011 + }, + { + "epoch": 0.75, + "grad_norm": 1.0581626740198355, + "learning_rate": 3.0054167294835314e-06, + "loss": 0.9612, + "step": 32012 + }, + { + "epoch": 0.75, + "grad_norm": 1.9582749956204673, + "learning_rate": 3.004871424708121e-06, + "loss": 0.9491, + "step": 32013 + }, + { + "epoch": 0.75, + "grad_norm": 1.852108602290021, + "learning_rate": 3.0043261606605656e-06, + "loss": 0.9754, + "step": 32014 + }, + { + "epoch": 0.75, + "grad_norm": 1.0856664711951192, + "learning_rate": 3.003780937344044e-06, + "loss": 0.8825, + "step": 32015 + }, + { + "epoch": 0.75, + "grad_norm": 2.1710009491698745, + "learning_rate": 3.0032357547617187e-06, + "loss": 1.0236, + "step": 32016 + }, + { + "epoch": 0.75, + "grad_norm": 1.9088543073162758, + "learning_rate": 3.00269061291677e-06, + "loss": 1.0177, + "step": 32017 + }, + { + "epoch": 0.75, + "grad_norm": 3.622395122808904, + "learning_rate": 3.0021455118123745e-06, + "loss": 1.0098, + "step": 32018 + }, + { + "epoch": 0.75, + "grad_norm": 1.9563301597589358, + "learning_rate": 3.0016004514517005e-06, + "loss": 0.9709, + "step": 32019 + }, + { + "epoch": 0.75, + "grad_norm": 1.9644244211205364, + "learning_rate": 3.001055431837927e-06, + "loss": 0.9639, + "step": 32020 + }, + { + "epoch": 0.75, + "grad_norm": 2.155050636839275, + "learning_rate": 3.0005104529742214e-06, + "loss": 1.0259, + "step": 32021 + }, + { + "epoch": 0.75, + "grad_norm": 2.113926742877537, + "learning_rate": 2.999965514863764e-06, + "loss": 0.9803, + "step": 32022 + }, + { + "epoch": 0.75, + "grad_norm": 1.9521259793305976, + "learning_rate": 2.9994206175097196e-06, + "loss": 1.0141, + "step": 32023 + }, + { + "epoch": 0.75, + "grad_norm": 2.3557351642855933, + "learning_rate": 2.9988757609152686e-06, + "loss": 1.0065, + "step": 32024 + }, + { + "epoch": 0.75, + "grad_norm": 2.021431483392251, + "learning_rate": 2.9983309450835785e-06, + "loss": 0.9038, + "step": 32025 + }, + { + "epoch": 0.75, + "grad_norm": 2.531750829899147, + "learning_rate": 2.9977861700178202e-06, + "loss": 0.936, + "step": 32026 + }, + { + "epoch": 0.75, + "grad_norm": 1.9956024509180306, + "learning_rate": 2.9972414357211667e-06, + "loss": 1.0944, + "step": 32027 + }, + { + "epoch": 0.75, + "grad_norm": 1.9123046625380056, + "learning_rate": 2.996696742196794e-06, + "loss": 0.9567, + "step": 32028 + }, + { + "epoch": 0.75, + "grad_norm": 1.7491064366125917, + "learning_rate": 2.9961520894478703e-06, + "loss": 0.9792, + "step": 32029 + }, + { + "epoch": 0.75, + "grad_norm": 1.7627825978243326, + "learning_rate": 2.9956074774775636e-06, + "loss": 0.8497, + "step": 32030 + }, + { + "epoch": 0.75, + "grad_norm": 2.062883380164037, + "learning_rate": 2.995062906289047e-06, + "loss": 0.7515, + "step": 32031 + }, + { + "epoch": 0.75, + "grad_norm": 1.8207190970463196, + "learning_rate": 2.9945183758854956e-06, + "loss": 0.9956, + "step": 32032 + }, + { + "epoch": 0.75, + "grad_norm": 1.796412016863358, + "learning_rate": 2.9939738862700764e-06, + "loss": 1.0124, + "step": 32033 + }, + { + "epoch": 0.75, + "grad_norm": 2.1834562321029716, + "learning_rate": 2.9934294374459548e-06, + "loss": 0.7872, + "step": 32034 + }, + { + "epoch": 0.75, + "grad_norm": 1.1123545194252682, + "learning_rate": 2.992885029416307e-06, + "loss": 0.9177, + "step": 32035 + }, + { + "epoch": 0.75, + "grad_norm": 2.1410023829101616, + "learning_rate": 2.9923406621843034e-06, + "loss": 1.0906, + "step": 32036 + }, + { + "epoch": 0.75, + "grad_norm": 2.112190821261937, + "learning_rate": 2.9917963357531112e-06, + "loss": 0.8651, + "step": 32037 + }, + { + "epoch": 0.75, + "grad_norm": 1.8301870438082803, + "learning_rate": 2.9912520501258958e-06, + "loss": 0.8602, + "step": 32038 + }, + { + "epoch": 0.75, + "grad_norm": 2.207658251382416, + "learning_rate": 2.9907078053058325e-06, + "loss": 0.9653, + "step": 32039 + }, + { + "epoch": 0.75, + "grad_norm": 2.3245311867158684, + "learning_rate": 2.9901636012960856e-06, + "loss": 0.9399, + "step": 32040 + }, + { + "epoch": 0.75, + "grad_norm": 1.0775200598531045, + "learning_rate": 2.9896194380998276e-06, + "loss": 0.9194, + "step": 32041 + }, + { + "epoch": 0.75, + "grad_norm": 2.5374267808053617, + "learning_rate": 2.989075315720222e-06, + "loss": 0.895, + "step": 32042 + }, + { + "epoch": 0.75, + "grad_norm": 1.0871552300899758, + "learning_rate": 2.988531234160443e-06, + "loss": 0.9818, + "step": 32043 + }, + { + "epoch": 0.75, + "grad_norm": 1.7954079595816388, + "learning_rate": 2.9879871934236525e-06, + "loss": 1.0202, + "step": 32044 + }, + { + "epoch": 0.75, + "grad_norm": 1.963057957581625, + "learning_rate": 2.9874431935130223e-06, + "loss": 1.0446, + "step": 32045 + }, + { + "epoch": 0.75, + "grad_norm": 2.1624518020359877, + "learning_rate": 2.9868992344317184e-06, + "loss": 1.0992, + "step": 32046 + }, + { + "epoch": 0.76, + "grad_norm": 1.1985142545852003, + "learning_rate": 2.986355316182905e-06, + "loss": 0.9361, + "step": 32047 + }, + { + "epoch": 0.76, + "grad_norm": 2.054098881804714, + "learning_rate": 2.98581143876975e-06, + "loss": 0.9651, + "step": 32048 + }, + { + "epoch": 0.76, + "grad_norm": 1.9970799345610108, + "learning_rate": 2.985267602195426e-06, + "loss": 0.9666, + "step": 32049 + }, + { + "epoch": 0.76, + "grad_norm": 1.8224376135288094, + "learning_rate": 2.9847238064630947e-06, + "loss": 1.0186, + "step": 32050 + }, + { + "epoch": 0.76, + "grad_norm": 1.9189928466306416, + "learning_rate": 2.984180051575918e-06, + "loss": 1.0343, + "step": 32051 + }, + { + "epoch": 0.76, + "grad_norm": 1.9285242351874272, + "learning_rate": 2.983636337537067e-06, + "loss": 0.9387, + "step": 32052 + }, + { + "epoch": 0.76, + "grad_norm": 2.1134004609952295, + "learning_rate": 2.9830926643497106e-06, + "loss": 0.9909, + "step": 32053 + }, + { + "epoch": 0.76, + "grad_norm": 2.078045359834184, + "learning_rate": 2.982549032017009e-06, + "loss": 0.9039, + "step": 32054 + }, + { + "epoch": 0.76, + "grad_norm": 3.3953196037580677, + "learning_rate": 2.9820054405421263e-06, + "loss": 0.8837, + "step": 32055 + }, + { + "epoch": 0.76, + "grad_norm": 2.0614032688313944, + "learning_rate": 2.981461889928229e-06, + "loss": 1.0436, + "step": 32056 + }, + { + "epoch": 0.76, + "grad_norm": 1.9519317617863825, + "learning_rate": 2.980918380178486e-06, + "loss": 0.9942, + "step": 32057 + }, + { + "epoch": 0.76, + "grad_norm": 1.9663612362464988, + "learning_rate": 2.980374911296058e-06, + "loss": 0.9485, + "step": 32058 + }, + { + "epoch": 0.76, + "grad_norm": 1.126304741789229, + "learning_rate": 2.979831483284107e-06, + "loss": 0.9581, + "step": 32059 + }, + { + "epoch": 0.76, + "grad_norm": 1.9205304362022548, + "learning_rate": 2.9792880961458028e-06, + "loss": 0.9515, + "step": 32060 + }, + { + "epoch": 0.76, + "grad_norm": 1.8547583582851332, + "learning_rate": 2.9787447498843024e-06, + "loss": 0.8603, + "step": 32061 + }, + { + "epoch": 0.76, + "grad_norm": 1.9348005643279924, + "learning_rate": 2.9782014445027774e-06, + "loss": 0.9638, + "step": 32062 + }, + { + "epoch": 0.76, + "grad_norm": 2.2047966002571835, + "learning_rate": 2.9776581800043823e-06, + "loss": 0.951, + "step": 32063 + }, + { + "epoch": 0.76, + "grad_norm": 2.6828824222175256, + "learning_rate": 2.9771149563922884e-06, + "loss": 1.0927, + "step": 32064 + }, + { + "epoch": 0.76, + "grad_norm": 2.120075403985181, + "learning_rate": 2.9765717736696507e-06, + "loss": 0.8533, + "step": 32065 + }, + { + "epoch": 0.76, + "grad_norm": 1.852859735636134, + "learning_rate": 2.9760286318396394e-06, + "loss": 0.9916, + "step": 32066 + }, + { + "epoch": 0.76, + "grad_norm": 1.8603451223712497, + "learning_rate": 2.975485530905413e-06, + "loss": 1.0278, + "step": 32067 + }, + { + "epoch": 0.76, + "grad_norm": 2.1921448614026944, + "learning_rate": 2.9749424708701313e-06, + "loss": 0.9744, + "step": 32068 + }, + { + "epoch": 0.76, + "grad_norm": 2.0863555185078058, + "learning_rate": 2.9743994517369577e-06, + "loss": 1.0522, + "step": 32069 + }, + { + "epoch": 0.76, + "grad_norm": 1.0613278004277602, + "learning_rate": 2.973856473509058e-06, + "loss": 1.0138, + "step": 32070 + }, + { + "epoch": 0.76, + "grad_norm": 1.1504969325111496, + "learning_rate": 2.9733135361895905e-06, + "loss": 0.967, + "step": 32071 + }, + { + "epoch": 0.76, + "grad_norm": 3.203518995471182, + "learning_rate": 2.9727706397817136e-06, + "loss": 1.0064, + "step": 32072 + }, + { + "epoch": 0.76, + "grad_norm": 2.043846241426796, + "learning_rate": 2.9722277842885893e-06, + "loss": 1.0082, + "step": 32073 + }, + { + "epoch": 0.76, + "grad_norm": 2.090642157456246, + "learning_rate": 2.971684969713384e-06, + "loss": 1.0122, + "step": 32074 + }, + { + "epoch": 0.76, + "grad_norm": 2.1271590266954954, + "learning_rate": 2.9711421960592546e-06, + "loss": 1.0345, + "step": 32075 + }, + { + "epoch": 0.76, + "grad_norm": 1.9165316141154443, + "learning_rate": 2.9705994633293567e-06, + "loss": 0.9808, + "step": 32076 + }, + { + "epoch": 0.76, + "grad_norm": 1.0178794656045278, + "learning_rate": 2.9700567715268546e-06, + "loss": 0.9911, + "step": 32077 + }, + { + "epoch": 0.76, + "grad_norm": 2.477299324637753, + "learning_rate": 2.9695141206549107e-06, + "loss": 0.9346, + "step": 32078 + }, + { + "epoch": 0.76, + "grad_norm": 1.8558008003515896, + "learning_rate": 2.9689715107166806e-06, + "loss": 0.888, + "step": 32079 + }, + { + "epoch": 0.76, + "grad_norm": 1.9786181622198293, + "learning_rate": 2.968428941715321e-06, + "loss": 0.8629, + "step": 32080 + }, + { + "epoch": 0.76, + "grad_norm": 1.7851715705452058, + "learning_rate": 2.9678864136539985e-06, + "loss": 1.011, + "step": 32081 + }, + { + "epoch": 0.76, + "grad_norm": 2.0798147809977676, + "learning_rate": 2.967343926535864e-06, + "loss": 1.0485, + "step": 32082 + }, + { + "epoch": 0.76, + "grad_norm": 2.7466072818866962, + "learning_rate": 2.966801480364083e-06, + "loss": 0.9993, + "step": 32083 + }, + { + "epoch": 0.76, + "grad_norm": 1.9669776156201717, + "learning_rate": 2.9662590751418075e-06, + "loss": 1.0394, + "step": 32084 + }, + { + "epoch": 0.76, + "grad_norm": 1.8702453508359367, + "learning_rate": 2.9657167108722018e-06, + "loss": 0.8998, + "step": 32085 + }, + { + "epoch": 0.76, + "grad_norm": 1.921091649859856, + "learning_rate": 2.965174387558416e-06, + "loss": 0.9078, + "step": 32086 + }, + { + "epoch": 0.76, + "grad_norm": 1.896209710534718, + "learning_rate": 2.9646321052036164e-06, + "loss": 1.0873, + "step": 32087 + }, + { + "epoch": 0.76, + "grad_norm": 1.976818796784511, + "learning_rate": 2.964089863810956e-06, + "loss": 1.0175, + "step": 32088 + }, + { + "epoch": 0.76, + "grad_norm": 1.0796184188791225, + "learning_rate": 2.9635476633835893e-06, + "loss": 0.9382, + "step": 32089 + }, + { + "epoch": 0.76, + "grad_norm": 2.26910496740518, + "learning_rate": 2.963005503924674e-06, + "loss": 1.0212, + "step": 32090 + }, + { + "epoch": 0.76, + "grad_norm": 1.9626706239668072, + "learning_rate": 2.962463385437373e-06, + "loss": 0.9513, + "step": 32091 + }, + { + "epoch": 0.76, + "grad_norm": 1.7950638063376831, + "learning_rate": 2.961921307924838e-06, + "loss": 0.9365, + "step": 32092 + }, + { + "epoch": 0.76, + "grad_norm": 1.9219433754876463, + "learning_rate": 2.961379271390222e-06, + "loss": 0.8869, + "step": 32093 + }, + { + "epoch": 0.76, + "grad_norm": 2.163541822367414, + "learning_rate": 2.9608372758366842e-06, + "loss": 0.9235, + "step": 32094 + }, + { + "epoch": 0.76, + "grad_norm": 2.2783322339087353, + "learning_rate": 2.9602953212673813e-06, + "loss": 0.9886, + "step": 32095 + }, + { + "epoch": 0.76, + "grad_norm": 1.9018657204764253, + "learning_rate": 2.959753407685473e-06, + "loss": 0.9999, + "step": 32096 + }, + { + "epoch": 0.76, + "grad_norm": 1.0490380939211925, + "learning_rate": 2.9592115350941032e-06, + "loss": 0.9039, + "step": 32097 + }, + { + "epoch": 0.76, + "grad_norm": 2.55934897926355, + "learning_rate": 2.958669703496434e-06, + "loss": 0.9966, + "step": 32098 + }, + { + "epoch": 0.76, + "grad_norm": 1.9397889137223032, + "learning_rate": 2.958127912895621e-06, + "loss": 0.9934, + "step": 32099 + }, + { + "epoch": 0.76, + "grad_norm": 2.1394171655627665, + "learning_rate": 2.9575861632948168e-06, + "loss": 0.9778, + "step": 32100 + }, + { + "epoch": 0.76, + "grad_norm": 2.2651874521547657, + "learning_rate": 2.957044454697172e-06, + "loss": 0.9687, + "step": 32101 + }, + { + "epoch": 0.76, + "grad_norm": 1.808454629983138, + "learning_rate": 2.956502787105845e-06, + "loss": 0.9878, + "step": 32102 + }, + { + "epoch": 0.76, + "grad_norm": 1.938130064535659, + "learning_rate": 2.9559611605239914e-06, + "loss": 1.007, + "step": 32103 + }, + { + "epoch": 0.76, + "grad_norm": 2.312752198872017, + "learning_rate": 2.955419574954762e-06, + "loss": 0.9451, + "step": 32104 + }, + { + "epoch": 0.76, + "grad_norm": 2.1031974612298594, + "learning_rate": 2.954878030401308e-06, + "loss": 0.969, + "step": 32105 + }, + { + "epoch": 0.76, + "grad_norm": 2.1139825137320596, + "learning_rate": 2.9543365268667866e-06, + "loss": 0.9815, + "step": 32106 + }, + { + "epoch": 0.76, + "grad_norm": 2.0814354683331024, + "learning_rate": 2.9537950643543455e-06, + "loss": 0.9577, + "step": 32107 + }, + { + "epoch": 0.76, + "grad_norm": 2.537085888917622, + "learning_rate": 2.953253642867141e-06, + "loss": 1.017, + "step": 32108 + }, + { + "epoch": 0.76, + "grad_norm": 1.8327185813394933, + "learning_rate": 2.952712262408327e-06, + "loss": 1.0073, + "step": 32109 + }, + { + "epoch": 0.76, + "grad_norm": 2.0709510865944227, + "learning_rate": 2.9521709229810535e-06, + "loss": 0.8612, + "step": 32110 + }, + { + "epoch": 0.76, + "grad_norm": 1.9190132810965395, + "learning_rate": 2.951629624588469e-06, + "loss": 0.9594, + "step": 32111 + }, + { + "epoch": 0.76, + "grad_norm": 2.7244429958809517, + "learning_rate": 2.951088367233729e-06, + "loss": 0.9113, + "step": 32112 + }, + { + "epoch": 0.76, + "grad_norm": 1.8988364867027456, + "learning_rate": 2.95054715091999e-06, + "loss": 0.939, + "step": 32113 + }, + { + "epoch": 0.76, + "grad_norm": 1.8898027126041377, + "learning_rate": 2.9500059756503907e-06, + "loss": 0.9796, + "step": 32114 + }, + { + "epoch": 0.76, + "grad_norm": 2.483176221536024, + "learning_rate": 2.949464841428089e-06, + "loss": 0.968, + "step": 32115 + }, + { + "epoch": 0.76, + "grad_norm": 1.1563365892177555, + "learning_rate": 2.948923748256236e-06, + "loss": 0.921, + "step": 32116 + }, + { + "epoch": 0.76, + "grad_norm": 1.8261328540677273, + "learning_rate": 2.9483826961379837e-06, + "loss": 0.915, + "step": 32117 + }, + { + "epoch": 0.76, + "grad_norm": 5.944608767539441, + "learning_rate": 2.947841685076479e-06, + "loss": 0.9056, + "step": 32118 + }, + { + "epoch": 0.76, + "grad_norm": 2.0856430492386036, + "learning_rate": 2.94730071507487e-06, + "loss": 1.0191, + "step": 32119 + }, + { + "epoch": 0.76, + "grad_norm": 2.297186802157373, + "learning_rate": 2.946759786136313e-06, + "loss": 0.9359, + "step": 32120 + }, + { + "epoch": 0.76, + "grad_norm": 2.09588252052546, + "learning_rate": 2.9462188982639494e-06, + "loss": 1.0523, + "step": 32121 + }, + { + "epoch": 0.76, + "grad_norm": 1.9988369132563102, + "learning_rate": 2.9456780514609364e-06, + "loss": 0.9183, + "step": 32122 + }, + { + "epoch": 0.76, + "grad_norm": 9.867975554701923, + "learning_rate": 2.9451372457304163e-06, + "loss": 0.884, + "step": 32123 + }, + { + "epoch": 0.76, + "grad_norm": 1.8276868891524447, + "learning_rate": 2.944596481075542e-06, + "loss": 0.9281, + "step": 32124 + }, + { + "epoch": 0.76, + "grad_norm": 1.9946632440662655, + "learning_rate": 2.94405575749946e-06, + "loss": 0.9328, + "step": 32125 + }, + { + "epoch": 0.76, + "grad_norm": 1.1236524887440933, + "learning_rate": 2.943515075005321e-06, + "loss": 0.9055, + "step": 32126 + }, + { + "epoch": 0.76, + "grad_norm": 1.9709175121320717, + "learning_rate": 2.9429744335962717e-06, + "loss": 1.035, + "step": 32127 + }, + { + "epoch": 0.76, + "grad_norm": 2.174671560961176, + "learning_rate": 2.9424338332754553e-06, + "loss": 0.991, + "step": 32128 + }, + { + "epoch": 0.76, + "grad_norm": 2.3147594099350184, + "learning_rate": 2.941893274046025e-06, + "loss": 0.8019, + "step": 32129 + }, + { + "epoch": 0.76, + "grad_norm": 1.0249947932848167, + "learning_rate": 2.941352755911129e-06, + "loss": 0.8742, + "step": 32130 + }, + { + "epoch": 0.76, + "grad_norm": 2.3335659182183974, + "learning_rate": 2.9408122788739124e-06, + "loss": 0.9144, + "step": 32131 + }, + { + "epoch": 0.76, + "grad_norm": 2.0185310362268583, + "learning_rate": 2.940271842937518e-06, + "loss": 0.8843, + "step": 32132 + }, + { + "epoch": 0.76, + "grad_norm": 2.0377897704034407, + "learning_rate": 2.9397314481050953e-06, + "loss": 1.0914, + "step": 32133 + }, + { + "epoch": 0.76, + "grad_norm": 2.1590127839461957, + "learning_rate": 2.9391910943797963e-06, + "loss": 1.0082, + "step": 32134 + }, + { + "epoch": 0.76, + "grad_norm": 1.9617593878977095, + "learning_rate": 2.93865078176476e-06, + "loss": 1.0316, + "step": 32135 + }, + { + "epoch": 0.76, + "grad_norm": 2.123863265147803, + "learning_rate": 2.9381105102631334e-06, + "loss": 0.9736, + "step": 32136 + }, + { + "epoch": 0.76, + "grad_norm": 1.9372913208319704, + "learning_rate": 2.9375702798780626e-06, + "loss": 0.9934, + "step": 32137 + }, + { + "epoch": 0.76, + "grad_norm": 2.095056286117202, + "learning_rate": 2.937030090612696e-06, + "loss": 1.0494, + "step": 32138 + }, + { + "epoch": 0.76, + "grad_norm": 1.8900095112284896, + "learning_rate": 2.9364899424701765e-06, + "loss": 0.8039, + "step": 32139 + }, + { + "epoch": 0.76, + "grad_norm": 2.9628448941719356, + "learning_rate": 2.9359498354536465e-06, + "loss": 0.9272, + "step": 32140 + }, + { + "epoch": 0.76, + "grad_norm": 1.8777668045916598, + "learning_rate": 2.935409769566255e-06, + "loss": 0.8923, + "step": 32141 + }, + { + "epoch": 0.76, + "grad_norm": 2.0645136560409845, + "learning_rate": 2.934869744811142e-06, + "loss": 0.9615, + "step": 32142 + }, + { + "epoch": 0.76, + "grad_norm": 1.0917172515936522, + "learning_rate": 2.934329761191457e-06, + "loss": 0.8277, + "step": 32143 + }, + { + "epoch": 0.76, + "grad_norm": 2.153356912035887, + "learning_rate": 2.9337898187103374e-06, + "loss": 0.8926, + "step": 32144 + }, + { + "epoch": 0.76, + "grad_norm": 2.327326913025526, + "learning_rate": 2.933249917370935e-06, + "loss": 1.0064, + "step": 32145 + }, + { + "epoch": 0.76, + "grad_norm": 2.002051643878261, + "learning_rate": 2.9327100571763843e-06, + "loss": 1.0032, + "step": 32146 + }, + { + "epoch": 0.76, + "grad_norm": 2.366687910687675, + "learning_rate": 2.9321702381298378e-06, + "loss": 0.9237, + "step": 32147 + }, + { + "epoch": 0.76, + "grad_norm": 1.6601143201974549, + "learning_rate": 2.9316304602344324e-06, + "loss": 0.9574, + "step": 32148 + }, + { + "epoch": 0.76, + "grad_norm": 3.027129768530851, + "learning_rate": 2.9310907234933104e-06, + "loss": 0.9776, + "step": 32149 + }, + { + "epoch": 0.76, + "grad_norm": 1.8712068434714024, + "learning_rate": 2.930551027909615e-06, + "loss": 0.9389, + "step": 32150 + }, + { + "epoch": 0.76, + "grad_norm": 1.9980263961840683, + "learning_rate": 2.930011373486494e-06, + "loss": 1.0379, + "step": 32151 + }, + { + "epoch": 0.76, + "grad_norm": 1.1022424192165403, + "learning_rate": 2.9294717602270837e-06, + "loss": 0.9883, + "step": 32152 + }, + { + "epoch": 0.76, + "grad_norm": 1.8842518627542484, + "learning_rate": 2.9289321881345257e-06, + "loss": 0.9525, + "step": 32153 + }, + { + "epoch": 0.76, + "grad_norm": 1.9705617303851302, + "learning_rate": 2.9283926572119625e-06, + "loss": 0.8508, + "step": 32154 + }, + { + "epoch": 0.76, + "grad_norm": 1.7549244543249864, + "learning_rate": 2.927853167462539e-06, + "loss": 0.9491, + "step": 32155 + }, + { + "epoch": 0.76, + "grad_norm": 1.9518164912568277, + "learning_rate": 2.927313718889393e-06, + "loss": 1.023, + "step": 32156 + }, + { + "epoch": 0.76, + "grad_norm": 2.096055653323982, + "learning_rate": 2.9267743114956635e-06, + "loss": 0.9537, + "step": 32157 + }, + { + "epoch": 0.76, + "grad_norm": 1.0256990901436878, + "learning_rate": 2.926234945284493e-06, + "loss": 0.9245, + "step": 32158 + }, + { + "epoch": 0.76, + "grad_norm": 3.33222033998588, + "learning_rate": 2.9256956202590248e-06, + "loss": 0.9378, + "step": 32159 + }, + { + "epoch": 0.76, + "grad_norm": 1.9764124628769197, + "learning_rate": 2.9251563364223967e-06, + "loss": 0.9443, + "step": 32160 + }, + { + "epoch": 0.76, + "grad_norm": 2.1590668985479304, + "learning_rate": 2.924617093777745e-06, + "loss": 0.939, + "step": 32161 + }, + { + "epoch": 0.76, + "grad_norm": 2.120211516519536, + "learning_rate": 2.9240778923282163e-06, + "loss": 1.0078, + "step": 32162 + }, + { + "epoch": 0.76, + "grad_norm": 2.0578020245452704, + "learning_rate": 2.9235387320769427e-06, + "loss": 1.0872, + "step": 32163 + }, + { + "epoch": 0.76, + "grad_norm": 2.35627513111702, + "learning_rate": 2.92299961302707e-06, + "loss": 0.9838, + "step": 32164 + }, + { + "epoch": 0.76, + "grad_norm": 1.086096871125554, + "learning_rate": 2.9224605351817314e-06, + "loss": 0.9185, + "step": 32165 + }, + { + "epoch": 0.76, + "grad_norm": 1.8083019948808907, + "learning_rate": 2.9219214985440715e-06, + "loss": 0.9626, + "step": 32166 + }, + { + "epoch": 0.76, + "grad_norm": 2.0782445164243266, + "learning_rate": 2.9213825031172216e-06, + "loss": 0.9449, + "step": 32167 + }, + { + "epoch": 0.76, + "grad_norm": 1.958763183759805, + "learning_rate": 2.9208435489043276e-06, + "loss": 0.9721, + "step": 32168 + }, + { + "epoch": 0.76, + "grad_norm": 2.0976043415928287, + "learning_rate": 2.920304635908524e-06, + "loss": 1.0852, + "step": 32169 + }, + { + "epoch": 0.76, + "grad_norm": 1.8755039496488597, + "learning_rate": 2.9197657641329447e-06, + "loss": 1.0155, + "step": 32170 + }, + { + "epoch": 0.76, + "grad_norm": 2.0109073164444915, + "learning_rate": 2.91922693358073e-06, + "loss": 1.1642, + "step": 32171 + }, + { + "epoch": 0.76, + "grad_norm": 2.181467005823065, + "learning_rate": 2.918688144255022e-06, + "loss": 0.8883, + "step": 32172 + }, + { + "epoch": 0.76, + "grad_norm": 1.907000869236376, + "learning_rate": 2.9181493961589526e-06, + "loss": 1.0156, + "step": 32173 + }, + { + "epoch": 0.76, + "grad_norm": 1.069470507066824, + "learning_rate": 2.917610689295657e-06, + "loss": 0.8596, + "step": 32174 + }, + { + "epoch": 0.76, + "grad_norm": 3.6252370089015655, + "learning_rate": 2.9170720236682737e-06, + "loss": 0.9443, + "step": 32175 + }, + { + "epoch": 0.76, + "grad_norm": 2.0365337520236455, + "learning_rate": 2.9165333992799428e-06, + "loss": 0.8849, + "step": 32176 + }, + { + "epoch": 0.76, + "grad_norm": 3.0009018732459696, + "learning_rate": 2.915994816133797e-06, + "loss": 1.0282, + "step": 32177 + }, + { + "epoch": 0.76, + "grad_norm": 2.1638568900352504, + "learning_rate": 2.9154562742329695e-06, + "loss": 1.0759, + "step": 32178 + }, + { + "epoch": 0.76, + "grad_norm": 1.8432363723123772, + "learning_rate": 2.9149177735805977e-06, + "loss": 1.0605, + "step": 32179 + }, + { + "epoch": 0.76, + "grad_norm": 1.1115922741730284, + "learning_rate": 2.9143793141798214e-06, + "loss": 0.9573, + "step": 32180 + }, + { + "epoch": 0.76, + "grad_norm": 1.8840068389186864, + "learning_rate": 2.9138408960337725e-06, + "loss": 0.8592, + "step": 32181 + }, + { + "epoch": 0.76, + "grad_norm": 1.9088080104767426, + "learning_rate": 2.913302519145582e-06, + "loss": 0.9632, + "step": 32182 + }, + { + "epoch": 0.76, + "grad_norm": 2.19251632623727, + "learning_rate": 2.9127641835183906e-06, + "loss": 0.9284, + "step": 32183 + }, + { + "epoch": 0.76, + "grad_norm": 1.9864413075379235, + "learning_rate": 2.912225889155327e-06, + "loss": 0.9211, + "step": 32184 + }, + { + "epoch": 0.76, + "grad_norm": 1.9226658633599019, + "learning_rate": 2.9116876360595304e-06, + "loss": 0.9957, + "step": 32185 + }, + { + "epoch": 0.76, + "grad_norm": 1.9335374526536566, + "learning_rate": 2.9111494242341297e-06, + "loss": 0.9562, + "step": 32186 + }, + { + "epoch": 0.76, + "grad_norm": 1.9470770596551754, + "learning_rate": 2.910611253682265e-06, + "loss": 0.965, + "step": 32187 + }, + { + "epoch": 0.76, + "grad_norm": 2.0597863137191674, + "learning_rate": 2.9100731244070635e-06, + "loss": 1.0521, + "step": 32188 + }, + { + "epoch": 0.76, + "grad_norm": 1.0472653136418302, + "learning_rate": 2.909535036411664e-06, + "loss": 0.8938, + "step": 32189 + }, + { + "epoch": 0.76, + "grad_norm": 1.9375239594770082, + "learning_rate": 2.9089969896991956e-06, + "loss": 0.8901, + "step": 32190 + }, + { + "epoch": 0.76, + "grad_norm": 1.8886058452749614, + "learning_rate": 2.9084589842727883e-06, + "loss": 1.0478, + "step": 32191 + }, + { + "epoch": 0.76, + "grad_norm": 1.9798725125789145, + "learning_rate": 2.907921020135578e-06, + "loss": 1.0174, + "step": 32192 + }, + { + "epoch": 0.76, + "grad_norm": 1.9400000440922667, + "learning_rate": 2.907383097290697e-06, + "loss": 0.8479, + "step": 32193 + }, + { + "epoch": 0.76, + "grad_norm": 1.8794083364340912, + "learning_rate": 2.906845215741283e-06, + "loss": 0.984, + "step": 32194 + }, + { + "epoch": 0.76, + "grad_norm": 1.941738371292923, + "learning_rate": 2.906307375490457e-06, + "loss": 0.9939, + "step": 32195 + }, + { + "epoch": 0.76, + "grad_norm": 2.6019664529008932, + "learning_rate": 2.905769576541354e-06, + "loss": 1.0562, + "step": 32196 + }, + { + "epoch": 0.76, + "grad_norm": 1.9434344674942592, + "learning_rate": 2.9052318188971097e-06, + "loss": 1.0496, + "step": 32197 + }, + { + "epoch": 0.76, + "grad_norm": 2.069830946147575, + "learning_rate": 2.904694102560852e-06, + "loss": 0.9386, + "step": 32198 + }, + { + "epoch": 0.76, + "grad_norm": 2.2662216786106444, + "learning_rate": 2.904156427535709e-06, + "loss": 1.0463, + "step": 32199 + }, + { + "epoch": 0.76, + "grad_norm": 1.9607915107984244, + "learning_rate": 2.9036187938248126e-06, + "loss": 1.0174, + "step": 32200 + }, + { + "epoch": 0.76, + "grad_norm": 1.9442021343892468, + "learning_rate": 2.903081201431299e-06, + "loss": 0.94, + "step": 32201 + }, + { + "epoch": 0.76, + "grad_norm": 2.1549631671303455, + "learning_rate": 2.9025436503582893e-06, + "loss": 0.9804, + "step": 32202 + }, + { + "epoch": 0.76, + "grad_norm": 1.8944568025817268, + "learning_rate": 2.9020061406089207e-06, + "loss": 0.9658, + "step": 32203 + }, + { + "epoch": 0.76, + "grad_norm": 2.1355390089582524, + "learning_rate": 2.9014686721863192e-06, + "loss": 1.0377, + "step": 32204 + }, + { + "epoch": 0.76, + "grad_norm": 1.878944045966014, + "learning_rate": 2.900931245093612e-06, + "loss": 0.9968, + "step": 32205 + }, + { + "epoch": 0.76, + "grad_norm": 1.922317584095137, + "learning_rate": 2.900393859333931e-06, + "loss": 1.0678, + "step": 32206 + }, + { + "epoch": 0.76, + "grad_norm": 1.8558805092263808, + "learning_rate": 2.8998565149104075e-06, + "loss": 0.9269, + "step": 32207 + }, + { + "epoch": 0.76, + "grad_norm": 2.1958352447377183, + "learning_rate": 2.899319211826167e-06, + "loss": 0.9415, + "step": 32208 + }, + { + "epoch": 0.76, + "grad_norm": 2.0892745582032144, + "learning_rate": 2.898781950084335e-06, + "loss": 0.9348, + "step": 32209 + }, + { + "epoch": 0.76, + "grad_norm": 2.065097942268753, + "learning_rate": 2.8982447296880423e-06, + "loss": 0.9611, + "step": 32210 + }, + { + "epoch": 0.76, + "grad_norm": 2.8642449107173795, + "learning_rate": 2.897707550640424e-06, + "loss": 1.0128, + "step": 32211 + }, + { + "epoch": 0.76, + "grad_norm": 1.940465500831802, + "learning_rate": 2.8971704129445944e-06, + "loss": 0.9821, + "step": 32212 + }, + { + "epoch": 0.76, + "grad_norm": 2.4317802714651973, + "learning_rate": 2.8966333166036864e-06, + "loss": 1.0526, + "step": 32213 + }, + { + "epoch": 0.76, + "grad_norm": 2.0838133225922437, + "learning_rate": 2.89609626162083e-06, + "loss": 0.9092, + "step": 32214 + }, + { + "epoch": 0.76, + "grad_norm": 2.331992854907706, + "learning_rate": 2.8955592479991524e-06, + "loss": 1.0159, + "step": 32215 + }, + { + "epoch": 0.76, + "grad_norm": 1.9998569377986593, + "learning_rate": 2.895022275741778e-06, + "loss": 0.9913, + "step": 32216 + }, + { + "epoch": 0.76, + "grad_norm": 1.9223226476211377, + "learning_rate": 2.8944853448518295e-06, + "loss": 0.9404, + "step": 32217 + }, + { + "epoch": 0.76, + "grad_norm": 1.7327695490163695, + "learning_rate": 2.893948455332438e-06, + "loss": 0.9279, + "step": 32218 + }, + { + "epoch": 0.76, + "grad_norm": 1.9025725656913064, + "learning_rate": 2.8934116071867314e-06, + "loss": 1.0321, + "step": 32219 + }, + { + "epoch": 0.76, + "grad_norm": 2.2547837899992964, + "learning_rate": 2.892874800417832e-06, + "loss": 0.9322, + "step": 32220 + }, + { + "epoch": 0.76, + "grad_norm": 2.1389164930608047, + "learning_rate": 2.892338035028862e-06, + "loss": 0.9081, + "step": 32221 + }, + { + "epoch": 0.76, + "grad_norm": 2.6896876930706353, + "learning_rate": 2.8918013110229535e-06, + "loss": 1.1321, + "step": 32222 + }, + { + "epoch": 0.76, + "grad_norm": 2.2196882011121315, + "learning_rate": 2.8912646284032264e-06, + "loss": 0.9206, + "step": 32223 + }, + { + "epoch": 0.76, + "grad_norm": 2.113338228738245, + "learning_rate": 2.890727987172809e-06, + "loss": 1.0308, + "step": 32224 + }, + { + "epoch": 0.76, + "grad_norm": 2.0195589255035924, + "learning_rate": 2.8901913873348208e-06, + "loss": 0.9103, + "step": 32225 + }, + { + "epoch": 0.76, + "grad_norm": 1.941861009919936, + "learning_rate": 2.889654828892393e-06, + "loss": 0.9285, + "step": 32226 + }, + { + "epoch": 0.76, + "grad_norm": 2.2571693785452083, + "learning_rate": 2.889118311848643e-06, + "loss": 0.7786, + "step": 32227 + }, + { + "epoch": 0.76, + "grad_norm": 2.1155744586986684, + "learning_rate": 2.8885818362066997e-06, + "loss": 0.8838, + "step": 32228 + }, + { + "epoch": 0.76, + "grad_norm": 2.1727956385327207, + "learning_rate": 2.8880454019696846e-06, + "loss": 0.9213, + "step": 32229 + }, + { + "epoch": 0.76, + "grad_norm": 1.8509994329219184, + "learning_rate": 2.8875090091407187e-06, + "loss": 0.943, + "step": 32230 + }, + { + "epoch": 0.76, + "grad_norm": 2.3024726480114337, + "learning_rate": 2.8869726577229264e-06, + "loss": 0.9265, + "step": 32231 + }, + { + "epoch": 0.76, + "grad_norm": 2.4931708588677486, + "learning_rate": 2.8864363477194336e-06, + "loss": 1.0603, + "step": 32232 + }, + { + "epoch": 0.76, + "grad_norm": 1.9310762002765613, + "learning_rate": 2.8859000791333613e-06, + "loss": 0.9784, + "step": 32233 + }, + { + "epoch": 0.76, + "grad_norm": 2.0594631862906523, + "learning_rate": 2.8853638519678285e-06, + "loss": 1.0579, + "step": 32234 + }, + { + "epoch": 0.76, + "grad_norm": 2.005323455445341, + "learning_rate": 2.884827666225959e-06, + "loss": 1.0999, + "step": 32235 + }, + { + "epoch": 0.76, + "grad_norm": 1.9910169578749484, + "learning_rate": 2.884291521910878e-06, + "loss": 0.9494, + "step": 32236 + }, + { + "epoch": 0.76, + "grad_norm": 2.018114858692215, + "learning_rate": 2.8837554190257057e-06, + "loss": 0.8882, + "step": 32237 + }, + { + "epoch": 0.76, + "grad_norm": 1.8879773520184553, + "learning_rate": 2.883219357573559e-06, + "loss": 0.9087, + "step": 32238 + }, + { + "epoch": 0.76, + "grad_norm": 1.923318861468979, + "learning_rate": 2.8826833375575615e-06, + "loss": 0.8392, + "step": 32239 + }, + { + "epoch": 0.76, + "grad_norm": 2.0280391036035392, + "learning_rate": 2.882147358980839e-06, + "loss": 0.8507, + "step": 32240 + }, + { + "epoch": 0.76, + "grad_norm": 2.0026796381080905, + "learning_rate": 2.8816114218465073e-06, + "loss": 1.0506, + "step": 32241 + }, + { + "epoch": 0.76, + "grad_norm": 1.8877423540703218, + "learning_rate": 2.881075526157685e-06, + "loss": 1.0685, + "step": 32242 + }, + { + "epoch": 0.76, + "grad_norm": 2.08249836700749, + "learning_rate": 2.880539671917497e-06, + "loss": 1.0552, + "step": 32243 + }, + { + "epoch": 0.76, + "grad_norm": 2.1224003553037605, + "learning_rate": 2.880003859129059e-06, + "loss": 0.9514, + "step": 32244 + }, + { + "epoch": 0.76, + "grad_norm": 2.0104893795375527, + "learning_rate": 2.879468087795495e-06, + "loss": 1.0191, + "step": 32245 + }, + { + "epoch": 0.76, + "grad_norm": 1.0159274217242253, + "learning_rate": 2.87893235791992e-06, + "loss": 0.8376, + "step": 32246 + }, + { + "epoch": 0.76, + "grad_norm": 1.9119093122285922, + "learning_rate": 2.8783966695054575e-06, + "loss": 1.1956, + "step": 32247 + }, + { + "epoch": 0.76, + "grad_norm": 2.03615391626559, + "learning_rate": 2.877861022555222e-06, + "loss": 0.9192, + "step": 32248 + }, + { + "epoch": 0.76, + "grad_norm": 1.8852179867687509, + "learning_rate": 2.8773254170723363e-06, + "loss": 0.993, + "step": 32249 + }, + { + "epoch": 0.76, + "grad_norm": 2.1974831129097296, + "learning_rate": 2.876789853059918e-06, + "loss": 1.0195, + "step": 32250 + }, + { + "epoch": 0.76, + "grad_norm": 1.8302866906958448, + "learning_rate": 2.876254330521081e-06, + "loss": 0.9606, + "step": 32251 + }, + { + "epoch": 0.76, + "grad_norm": 2.0125473973145382, + "learning_rate": 2.8757188494589474e-06, + "loss": 1.0733, + "step": 32252 + }, + { + "epoch": 0.76, + "grad_norm": 1.9533959360553113, + "learning_rate": 2.875183409876636e-06, + "loss": 0.9987, + "step": 32253 + }, + { + "epoch": 0.76, + "grad_norm": 1.8274144617298036, + "learning_rate": 2.8746480117772626e-06, + "loss": 1.0255, + "step": 32254 + }, + { + "epoch": 0.76, + "grad_norm": 1.1239381433831768, + "learning_rate": 2.874112655163942e-06, + "loss": 0.8911, + "step": 32255 + }, + { + "epoch": 0.76, + "grad_norm": 1.9481613001368718, + "learning_rate": 2.8735773400397926e-06, + "loss": 1.0327, + "step": 32256 + }, + { + "epoch": 0.76, + "grad_norm": 1.8892434670710048, + "learning_rate": 2.8730420664079363e-06, + "loss": 1.0205, + "step": 32257 + }, + { + "epoch": 0.76, + "grad_norm": 1.8799802957996117, + "learning_rate": 2.8725068342714847e-06, + "loss": 0.8925, + "step": 32258 + }, + { + "epoch": 0.76, + "grad_norm": 1.9087664334981544, + "learning_rate": 2.871971643633552e-06, + "loss": 0.964, + "step": 32259 + }, + { + "epoch": 0.76, + "grad_norm": 2.167549408166281, + "learning_rate": 2.871436494497256e-06, + "loss": 1.0034, + "step": 32260 + }, + { + "epoch": 0.76, + "grad_norm": 1.9867411148845626, + "learning_rate": 2.8709013868657186e-06, + "loss": 0.9032, + "step": 32261 + }, + { + "epoch": 0.76, + "grad_norm": 2.1522775874487747, + "learning_rate": 2.8703663207420494e-06, + "loss": 0.9645, + "step": 32262 + }, + { + "epoch": 0.76, + "grad_norm": 2.486131460194846, + "learning_rate": 2.869831296129362e-06, + "loss": 1.2083, + "step": 32263 + }, + { + "epoch": 0.76, + "grad_norm": 2.124224092936978, + "learning_rate": 2.869296313030776e-06, + "loss": 0.9369, + "step": 32264 + }, + { + "epoch": 0.76, + "grad_norm": 2.4598313318663134, + "learning_rate": 2.8687613714494023e-06, + "loss": 0.9604, + "step": 32265 + }, + { + "epoch": 0.76, + "grad_norm": 1.939404738870425, + "learning_rate": 2.868226471388361e-06, + "loss": 0.8891, + "step": 32266 + }, + { + "epoch": 0.76, + "grad_norm": 2.3103733088245244, + "learning_rate": 2.867691612850759e-06, + "loss": 0.9056, + "step": 32267 + }, + { + "epoch": 0.76, + "grad_norm": 1.0928278901797535, + "learning_rate": 2.8671567958397183e-06, + "loss": 0.9077, + "step": 32268 + }, + { + "epoch": 0.76, + "grad_norm": 1.9620124457325072, + "learning_rate": 2.8666220203583462e-06, + "loss": 0.9185, + "step": 32269 + }, + { + "epoch": 0.76, + "grad_norm": 2.1639429151462983, + "learning_rate": 2.866087286409761e-06, + "loss": 0.9764, + "step": 32270 + }, + { + "epoch": 0.76, + "grad_norm": 3.2434125634631794, + "learning_rate": 2.865552593997075e-06, + "loss": 0.9823, + "step": 32271 + }, + { + "epoch": 0.76, + "grad_norm": 2.0619909124320586, + "learning_rate": 2.8650179431233984e-06, + "loss": 0.8705, + "step": 32272 + }, + { + "epoch": 0.76, + "grad_norm": 1.9837614982757898, + "learning_rate": 2.8644833337918444e-06, + "loss": 0.9962, + "step": 32273 + }, + { + "epoch": 0.76, + "grad_norm": 2.3542973268836653, + "learning_rate": 2.863948766005531e-06, + "loss": 0.9204, + "step": 32274 + }, + { + "epoch": 0.76, + "grad_norm": 2.857091819316014, + "learning_rate": 2.8634142397675678e-06, + "loss": 0.9353, + "step": 32275 + }, + { + "epoch": 0.76, + "grad_norm": 1.9889760353888681, + "learning_rate": 2.862879755081063e-06, + "loss": 0.8165, + "step": 32276 + }, + { + "epoch": 0.76, + "grad_norm": 2.2100457941318186, + "learning_rate": 2.8623453119491316e-06, + "loss": 0.9794, + "step": 32277 + }, + { + "epoch": 0.76, + "grad_norm": 1.0610016449087738, + "learning_rate": 2.861810910374888e-06, + "loss": 0.892, + "step": 32278 + }, + { + "epoch": 0.76, + "grad_norm": 2.1048118270016447, + "learning_rate": 2.8612765503614424e-06, + "loss": 0.8921, + "step": 32279 + }, + { + "epoch": 0.76, + "grad_norm": 1.9053543129624313, + "learning_rate": 2.8607422319119017e-06, + "loss": 0.9903, + "step": 32280 + }, + { + "epoch": 0.76, + "grad_norm": 2.2436005779911206, + "learning_rate": 2.8602079550293792e-06, + "loss": 1.0017, + "step": 32281 + }, + { + "epoch": 0.76, + "grad_norm": 1.0975070967423084, + "learning_rate": 2.85967371971699e-06, + "loss": 0.9895, + "step": 32282 + }, + { + "epoch": 0.76, + "grad_norm": 1.7208418082564898, + "learning_rate": 2.859139525977841e-06, + "loss": 0.9022, + "step": 32283 + }, + { + "epoch": 0.76, + "grad_norm": 2.024539411931434, + "learning_rate": 2.8586053738150388e-06, + "loss": 0.9312, + "step": 32284 + }, + { + "epoch": 0.76, + "grad_norm": 2.469704042353135, + "learning_rate": 2.858071263231701e-06, + "loss": 0.8956, + "step": 32285 + }, + { + "epoch": 0.76, + "grad_norm": 1.997008274672375, + "learning_rate": 2.8575371942309293e-06, + "loss": 0.8338, + "step": 32286 + }, + { + "epoch": 0.76, + "grad_norm": 2.1235074222154418, + "learning_rate": 2.8570031668158403e-06, + "loss": 1.054, + "step": 32287 + }, + { + "epoch": 0.76, + "grad_norm": 2.0203261048273022, + "learning_rate": 2.8564691809895383e-06, + "loss": 0.9699, + "step": 32288 + }, + { + "epoch": 0.76, + "grad_norm": 1.8197702207146162, + "learning_rate": 2.8559352367551375e-06, + "loss": 0.9936, + "step": 32289 + }, + { + "epoch": 0.76, + "grad_norm": 2.0367660142795936, + "learning_rate": 2.855401334115739e-06, + "loss": 0.9265, + "step": 32290 + }, + { + "epoch": 0.76, + "grad_norm": 2.0936927756453394, + "learning_rate": 2.8548674730744576e-06, + "loss": 0.9839, + "step": 32291 + }, + { + "epoch": 0.76, + "grad_norm": 1.9906615998067683, + "learning_rate": 2.8543336536344047e-06, + "loss": 1.043, + "step": 32292 + }, + { + "epoch": 0.76, + "grad_norm": 2.159098678766071, + "learning_rate": 2.8537998757986773e-06, + "loss": 0.9607, + "step": 32293 + }, + { + "epoch": 0.76, + "grad_norm": 1.963166311698703, + "learning_rate": 2.853266139570391e-06, + "loss": 1.019, + "step": 32294 + }, + { + "epoch": 0.76, + "grad_norm": 2.1402891522751877, + "learning_rate": 2.8527324449526505e-06, + "loss": 1.0529, + "step": 32295 + }, + { + "epoch": 0.76, + "grad_norm": 2.002957535491532, + "learning_rate": 2.8521987919485705e-06, + "loss": 0.9936, + "step": 32296 + }, + { + "epoch": 0.76, + "grad_norm": 1.9510439940855746, + "learning_rate": 2.851665180561247e-06, + "loss": 1.0125, + "step": 32297 + }, + { + "epoch": 0.76, + "grad_norm": 1.1025588376549964, + "learning_rate": 2.8511316107937912e-06, + "loss": 0.9599, + "step": 32298 + }, + { + "epoch": 0.76, + "grad_norm": 1.063407177802837, + "learning_rate": 2.850598082649314e-06, + "loss": 0.9007, + "step": 32299 + }, + { + "epoch": 0.76, + "grad_norm": 1.0501546827450245, + "learning_rate": 2.850064596130915e-06, + "loss": 0.9489, + "step": 32300 + }, + { + "epoch": 0.76, + "grad_norm": 1.1193805478181904, + "learning_rate": 2.849531151241707e-06, + "loss": 0.9921, + "step": 32301 + }, + { + "epoch": 0.76, + "grad_norm": 2.0168339503044765, + "learning_rate": 2.8489977479847886e-06, + "loss": 0.965, + "step": 32302 + }, + { + "epoch": 0.76, + "grad_norm": 1.9333500305138713, + "learning_rate": 2.8484643863632722e-06, + "loss": 0.9956, + "step": 32303 + }, + { + "epoch": 0.76, + "grad_norm": 1.1368223813672216, + "learning_rate": 2.8479310663802574e-06, + "loss": 0.9195, + "step": 32304 + }, + { + "epoch": 0.76, + "grad_norm": 2.0602448566397316, + "learning_rate": 2.847397788038856e-06, + "loss": 1.0337, + "step": 32305 + }, + { + "epoch": 0.76, + "grad_norm": 2.2254308242274012, + "learning_rate": 2.8468645513421688e-06, + "loss": 1.0099, + "step": 32306 + }, + { + "epoch": 0.76, + "grad_norm": 2.1053172538710587, + "learning_rate": 2.8463313562932983e-06, + "loss": 0.7998, + "step": 32307 + }, + { + "epoch": 0.76, + "grad_norm": 1.9787171169205602, + "learning_rate": 2.8457982028953502e-06, + "loss": 1.0484, + "step": 32308 + }, + { + "epoch": 0.76, + "grad_norm": 1.8582235070805393, + "learning_rate": 2.8452650911514336e-06, + "loss": 0.9986, + "step": 32309 + }, + { + "epoch": 0.76, + "grad_norm": 2.0905487995834884, + "learning_rate": 2.844732021064649e-06, + "loss": 0.9347, + "step": 32310 + }, + { + "epoch": 0.76, + "grad_norm": 2.2431807168633777, + "learning_rate": 2.8441989926380966e-06, + "loss": 0.9021, + "step": 32311 + }, + { + "epoch": 0.76, + "grad_norm": 3.221594014810835, + "learning_rate": 2.843666005874883e-06, + "loss": 0.9784, + "step": 32312 + }, + { + "epoch": 0.76, + "grad_norm": 1.9952424536253783, + "learning_rate": 2.843133060778115e-06, + "loss": 0.9748, + "step": 32313 + }, + { + "epoch": 0.76, + "grad_norm": 1.9795222989376484, + "learning_rate": 2.842600157350892e-06, + "loss": 1.0276, + "step": 32314 + }, + { + "epoch": 0.76, + "grad_norm": 1.8972783111232958, + "learning_rate": 2.8420672955963136e-06, + "loss": 1.1481, + "step": 32315 + }, + { + "epoch": 0.76, + "grad_norm": 2.4860382180757568, + "learning_rate": 2.8415344755174855e-06, + "loss": 0.9962, + "step": 32316 + }, + { + "epoch": 0.76, + "grad_norm": 2.073686458694598, + "learning_rate": 2.8410016971175136e-06, + "loss": 0.8898, + "step": 32317 + }, + { + "epoch": 0.76, + "grad_norm": 1.8904555181722362, + "learning_rate": 2.840468960399496e-06, + "loss": 0.9444, + "step": 32318 + }, + { + "epoch": 0.76, + "grad_norm": 2.0211222001680733, + "learning_rate": 2.8399362653665318e-06, + "loss": 0.9446, + "step": 32319 + }, + { + "epoch": 0.76, + "grad_norm": 2.1440728415048116, + "learning_rate": 2.8394036120217296e-06, + "loss": 0.9263, + "step": 32320 + }, + { + "epoch": 0.76, + "grad_norm": 1.8130296732094136, + "learning_rate": 2.838871000368182e-06, + "loss": 1.0338, + "step": 32321 + }, + { + "epoch": 0.76, + "grad_norm": 2.108602050496901, + "learning_rate": 2.8383384304089978e-06, + "loss": 0.9569, + "step": 32322 + }, + { + "epoch": 0.76, + "grad_norm": 2.2066344513176213, + "learning_rate": 2.8378059021472725e-06, + "loss": 1.0429, + "step": 32323 + }, + { + "epoch": 0.76, + "grad_norm": 1.9637329302507023, + "learning_rate": 2.8372734155861105e-06, + "loss": 0.8844, + "step": 32324 + }, + { + "epoch": 0.76, + "grad_norm": 2.138588591264771, + "learning_rate": 2.8367409707286085e-06, + "loss": 0.8951, + "step": 32325 + }, + { + "epoch": 0.76, + "grad_norm": 1.8553421971178923, + "learning_rate": 2.8362085675778708e-06, + "loss": 0.988, + "step": 32326 + }, + { + "epoch": 0.76, + "grad_norm": 1.9251860772856912, + "learning_rate": 2.8356762061369945e-06, + "loss": 1.0063, + "step": 32327 + }, + { + "epoch": 0.76, + "grad_norm": 2.222898574610497, + "learning_rate": 2.8351438864090774e-06, + "loss": 1.0119, + "step": 32328 + }, + { + "epoch": 0.76, + "grad_norm": 2.1057058668157667, + "learning_rate": 2.834611608397221e-06, + "loss": 0.9842, + "step": 32329 + }, + { + "epoch": 0.76, + "grad_norm": 2.1235069097375723, + "learning_rate": 2.8340793721045266e-06, + "loss": 0.9899, + "step": 32330 + }, + { + "epoch": 0.76, + "grad_norm": 2.1622506273635906, + "learning_rate": 2.8335471775340907e-06, + "loss": 0.989, + "step": 32331 + }, + { + "epoch": 0.76, + "grad_norm": 1.9826748302523536, + "learning_rate": 2.8330150246890097e-06, + "loss": 1.0297, + "step": 32332 + }, + { + "epoch": 0.76, + "grad_norm": 1.816641509423106, + "learning_rate": 2.832482913572384e-06, + "loss": 1.05, + "step": 32333 + }, + { + "epoch": 0.76, + "grad_norm": 2.105271364555715, + "learning_rate": 2.8319508441873143e-06, + "loss": 1.0552, + "step": 32334 + }, + { + "epoch": 0.76, + "grad_norm": 1.9165483916641324, + "learning_rate": 2.831418816536897e-06, + "loss": 0.964, + "step": 32335 + }, + { + "epoch": 0.76, + "grad_norm": 2.1531930744224397, + "learning_rate": 2.8308868306242253e-06, + "loss": 0.975, + "step": 32336 + }, + { + "epoch": 0.76, + "grad_norm": 1.9343686323646543, + "learning_rate": 2.8303548864524012e-06, + "loss": 0.9912, + "step": 32337 + }, + { + "epoch": 0.76, + "grad_norm": 1.9620520087289104, + "learning_rate": 2.829822984024523e-06, + "loss": 1.0428, + "step": 32338 + }, + { + "epoch": 0.76, + "grad_norm": 1.108053466389056, + "learning_rate": 2.829291123343686e-06, + "loss": 0.8991, + "step": 32339 + }, + { + "epoch": 0.76, + "grad_norm": 2.026867403888047, + "learning_rate": 2.8287593044129825e-06, + "loss": 1.1668, + "step": 32340 + }, + { + "epoch": 0.76, + "grad_norm": 2.0735549434805938, + "learning_rate": 2.828227527235513e-06, + "loss": 1.0169, + "step": 32341 + }, + { + "epoch": 0.76, + "grad_norm": 1.8357578373566705, + "learning_rate": 2.8276957918143766e-06, + "loss": 0.9648, + "step": 32342 + }, + { + "epoch": 0.76, + "grad_norm": 2.5780517751958754, + "learning_rate": 2.827164098152666e-06, + "loss": 0.9531, + "step": 32343 + }, + { + "epoch": 0.76, + "grad_norm": 2.0082932368210296, + "learning_rate": 2.8266324462534733e-06, + "loss": 0.9852, + "step": 32344 + }, + { + "epoch": 0.76, + "grad_norm": 2.072212255347492, + "learning_rate": 2.8261008361199017e-06, + "loss": 1.0453, + "step": 32345 + }, + { + "epoch": 0.76, + "grad_norm": 1.823827211081459, + "learning_rate": 2.8255692677550385e-06, + "loss": 0.8313, + "step": 32346 + }, + { + "epoch": 0.76, + "grad_norm": 1.037665044727961, + "learning_rate": 2.8250377411619845e-06, + "loss": 0.9703, + "step": 32347 + }, + { + "epoch": 0.76, + "grad_norm": 2.186295688297203, + "learning_rate": 2.8245062563438306e-06, + "loss": 0.9816, + "step": 32348 + }, + { + "epoch": 0.76, + "grad_norm": 2.568472380654081, + "learning_rate": 2.823974813303676e-06, + "loss": 0.9609, + "step": 32349 + }, + { + "epoch": 0.76, + "grad_norm": 2.499978735580012, + "learning_rate": 2.8234434120446084e-06, + "loss": 0.9458, + "step": 32350 + }, + { + "epoch": 0.76, + "grad_norm": 1.7341835074936358, + "learning_rate": 2.8229120525697285e-06, + "loss": 0.9716, + "step": 32351 + }, + { + "epoch": 0.76, + "grad_norm": 2.080522280495725, + "learning_rate": 2.8223807348821273e-06, + "loss": 0.9304, + "step": 32352 + }, + { + "epoch": 0.76, + "grad_norm": 1.0395581471149518, + "learning_rate": 2.821849458984894e-06, + "loss": 0.9184, + "step": 32353 + }, + { + "epoch": 0.76, + "grad_norm": 1.9306529974101794, + "learning_rate": 2.8213182248811265e-06, + "loss": 0.8735, + "step": 32354 + }, + { + "epoch": 0.76, + "grad_norm": 2.0824983604669844, + "learning_rate": 2.8207870325739196e-06, + "loss": 1.0598, + "step": 32355 + }, + { + "epoch": 0.76, + "grad_norm": 1.9066758093081635, + "learning_rate": 2.820255882066364e-06, + "loss": 0.9717, + "step": 32356 + }, + { + "epoch": 0.76, + "grad_norm": 1.149383631419269, + "learning_rate": 2.8197247733615474e-06, + "loss": 0.9571, + "step": 32357 + }, + { + "epoch": 0.76, + "grad_norm": 2.107108328481405, + "learning_rate": 2.8191937064625675e-06, + "loss": 1.0956, + "step": 32358 + }, + { + "epoch": 0.76, + "grad_norm": 1.8193768028557398, + "learning_rate": 2.818662681372518e-06, + "loss": 0.9659, + "step": 32359 + }, + { + "epoch": 0.76, + "grad_norm": 2.207566139085243, + "learning_rate": 2.8181316980944884e-06, + "loss": 1.0206, + "step": 32360 + }, + { + "epoch": 0.76, + "grad_norm": 1.1122877294044797, + "learning_rate": 2.8176007566315665e-06, + "loss": 0.9544, + "step": 32361 + }, + { + "epoch": 0.76, + "grad_norm": 1.9863893287085448, + "learning_rate": 2.817069856986846e-06, + "loss": 1.0368, + "step": 32362 + }, + { + "epoch": 0.76, + "grad_norm": 2.077260591077876, + "learning_rate": 2.816538999163423e-06, + "loss": 0.9171, + "step": 32363 + }, + { + "epoch": 0.76, + "grad_norm": 1.979274753281213, + "learning_rate": 2.8160081831643836e-06, + "loss": 0.9009, + "step": 32364 + }, + { + "epoch": 0.76, + "grad_norm": 1.964408081330571, + "learning_rate": 2.8154774089928172e-06, + "loss": 1.0541, + "step": 32365 + }, + { + "epoch": 0.76, + "grad_norm": 1.9905498702516002, + "learning_rate": 2.8149466766518175e-06, + "loss": 0.956, + "step": 32366 + }, + { + "epoch": 0.76, + "grad_norm": 1.8752560702693015, + "learning_rate": 2.814415986144471e-06, + "loss": 0.8083, + "step": 32367 + }, + { + "epoch": 0.76, + "grad_norm": 1.8503279866026454, + "learning_rate": 2.8138853374738717e-06, + "loss": 0.8826, + "step": 32368 + }, + { + "epoch": 0.76, + "grad_norm": 1.106122451478351, + "learning_rate": 2.813354730643104e-06, + "loss": 0.9601, + "step": 32369 + }, + { + "epoch": 0.76, + "grad_norm": 2.059926398923506, + "learning_rate": 2.8128241656552637e-06, + "loss": 1.0228, + "step": 32370 + }, + { + "epoch": 0.76, + "grad_norm": 1.980833040129415, + "learning_rate": 2.812293642513434e-06, + "loss": 0.8531, + "step": 32371 + }, + { + "epoch": 0.76, + "grad_norm": 1.9425362593610398, + "learning_rate": 2.8117631612207084e-06, + "loss": 1.0136, + "step": 32372 + }, + { + "epoch": 0.76, + "grad_norm": 2.200899146809948, + "learning_rate": 2.811232721780174e-06, + "loss": 0.9153, + "step": 32373 + }, + { + "epoch": 0.76, + "grad_norm": 1.952802077420064, + "learning_rate": 2.810702324194915e-06, + "loss": 0.8587, + "step": 32374 + }, + { + "epoch": 0.76, + "grad_norm": 1.196746047765208, + "learning_rate": 2.810171968468024e-06, + "loss": 0.9284, + "step": 32375 + }, + { + "epoch": 0.76, + "grad_norm": 2.1379331653123956, + "learning_rate": 2.809641654602587e-06, + "loss": 0.9922, + "step": 32376 + }, + { + "epoch": 0.76, + "grad_norm": 2.520889697802686, + "learning_rate": 2.8091113826017004e-06, + "loss": 0.902, + "step": 32377 + }, + { + "epoch": 0.76, + "grad_norm": 1.9024797488630683, + "learning_rate": 2.808581152468436e-06, + "loss": 0.9098, + "step": 32378 + }, + { + "epoch": 0.76, + "grad_norm": 2.1915195876917357, + "learning_rate": 2.8080509642058894e-06, + "loss": 0.9458, + "step": 32379 + }, + { + "epoch": 0.76, + "grad_norm": 1.8805963651970081, + "learning_rate": 2.8075208178171507e-06, + "loss": 1.017, + "step": 32380 + }, + { + "epoch": 0.76, + "grad_norm": 1.7304697360327648, + "learning_rate": 2.8069907133053033e-06, + "loss": 0.8582, + "step": 32381 + }, + { + "epoch": 0.76, + "grad_norm": 2.041553966010646, + "learning_rate": 2.8064606506734292e-06, + "loss": 0.8728, + "step": 32382 + }, + { + "epoch": 0.76, + "grad_norm": 3.321263042175292, + "learning_rate": 2.805930629924619e-06, + "loss": 0.8986, + "step": 32383 + }, + { + "epoch": 0.76, + "grad_norm": 1.0748319815359355, + "learning_rate": 2.8054006510619615e-06, + "loss": 0.9162, + "step": 32384 + }, + { + "epoch": 0.76, + "grad_norm": 2.116202226347129, + "learning_rate": 2.804870714088539e-06, + "loss": 1.0343, + "step": 32385 + }, + { + "epoch": 0.76, + "grad_norm": 1.9522893888387804, + "learning_rate": 2.804340819007434e-06, + "loss": 1.0087, + "step": 32386 + }, + { + "epoch": 0.76, + "grad_norm": 2.0618118480132765, + "learning_rate": 2.803810965821737e-06, + "loss": 0.9746, + "step": 32387 + }, + { + "epoch": 0.76, + "grad_norm": 2.068260471137924, + "learning_rate": 2.8032811545345294e-06, + "loss": 1.017, + "step": 32388 + }, + { + "epoch": 0.76, + "grad_norm": 1.0298968394064152, + "learning_rate": 2.8027513851488963e-06, + "loss": 0.8722, + "step": 32389 + }, + { + "epoch": 0.76, + "grad_norm": 1.9894430620158814, + "learning_rate": 2.802221657667927e-06, + "loss": 0.8827, + "step": 32390 + }, + { + "epoch": 0.76, + "grad_norm": 2.129364934119446, + "learning_rate": 2.8016919720947024e-06, + "loss": 1.1669, + "step": 32391 + }, + { + "epoch": 0.76, + "grad_norm": 2.006385271432774, + "learning_rate": 2.801162328432302e-06, + "loss": 0.9283, + "step": 32392 + }, + { + "epoch": 0.76, + "grad_norm": 1.8623107148419498, + "learning_rate": 2.800632726683814e-06, + "loss": 1.1017, + "step": 32393 + }, + { + "epoch": 0.76, + "grad_norm": 2.005642419887375, + "learning_rate": 2.8001031668523273e-06, + "loss": 1.0686, + "step": 32394 + }, + { + "epoch": 0.76, + "grad_norm": 1.0978462139281668, + "learning_rate": 2.799573648940913e-06, + "loss": 0.9275, + "step": 32395 + }, + { + "epoch": 0.76, + "grad_norm": 2.536860293224869, + "learning_rate": 2.7990441729526617e-06, + "loss": 1.0865, + "step": 32396 + }, + { + "epoch": 0.76, + "grad_norm": 2.00655850873553, + "learning_rate": 2.798514738890654e-06, + "loss": 1.0064, + "step": 32397 + }, + { + "epoch": 0.76, + "grad_norm": 2.119860673016226, + "learning_rate": 2.7979853467579777e-06, + "loss": 0.9861, + "step": 32398 + }, + { + "epoch": 0.76, + "grad_norm": 2.5029319643359296, + "learning_rate": 2.7974559965577097e-06, + "loss": 0.9143, + "step": 32399 + }, + { + "epoch": 0.76, + "grad_norm": 1.8243239721373405, + "learning_rate": 2.7969266882929304e-06, + "loss": 0.9013, + "step": 32400 + }, + { + "epoch": 0.76, + "grad_norm": 2.0990349308679637, + "learning_rate": 2.796397421966728e-06, + "loss": 0.9516, + "step": 32401 + }, + { + "epoch": 0.76, + "grad_norm": 1.8292608127500352, + "learning_rate": 2.7958681975821767e-06, + "loss": 0.9731, + "step": 32402 + }, + { + "epoch": 0.76, + "grad_norm": 2.197830094947305, + "learning_rate": 2.795339015142364e-06, + "loss": 0.8842, + "step": 32403 + }, + { + "epoch": 0.76, + "grad_norm": 2.0349992517738413, + "learning_rate": 2.794809874650367e-06, + "loss": 0.9912, + "step": 32404 + }, + { + "epoch": 0.76, + "grad_norm": 3.9954883280778426, + "learning_rate": 2.7942807761092704e-06, + "loss": 1.0368, + "step": 32405 + }, + { + "epoch": 0.76, + "grad_norm": 2.035312079634287, + "learning_rate": 2.7937517195221498e-06, + "loss": 0.9701, + "step": 32406 + }, + { + "epoch": 0.76, + "grad_norm": 2.15683904645185, + "learning_rate": 2.79322270489209e-06, + "loss": 0.9478, + "step": 32407 + }, + { + "epoch": 0.76, + "grad_norm": 2.153452649666971, + "learning_rate": 2.7926937322221705e-06, + "loss": 0.962, + "step": 32408 + }, + { + "epoch": 0.76, + "grad_norm": 2.6463942205374202, + "learning_rate": 2.7921648015154655e-06, + "loss": 0.9475, + "step": 32409 + }, + { + "epoch": 0.76, + "grad_norm": 1.9265898847896041, + "learning_rate": 2.7916359127750605e-06, + "loss": 0.8897, + "step": 32410 + }, + { + "epoch": 0.76, + "grad_norm": 1.8825474675953238, + "learning_rate": 2.791107066004035e-06, + "loss": 1.0268, + "step": 32411 + }, + { + "epoch": 0.76, + "grad_norm": 2.0297568567940116, + "learning_rate": 2.7905782612054676e-06, + "loss": 0.9027, + "step": 32412 + }, + { + "epoch": 0.76, + "grad_norm": 2.0102315376330724, + "learning_rate": 2.790049498382432e-06, + "loss": 0.9305, + "step": 32413 + }, + { + "epoch": 0.76, + "grad_norm": 2.0100033539417597, + "learning_rate": 2.7895207775380116e-06, + "loss": 0.9963, + "step": 32414 + }, + { + "epoch": 0.76, + "grad_norm": 2.1389982960740266, + "learning_rate": 2.788992098675287e-06, + "loss": 0.8092, + "step": 32415 + }, + { + "epoch": 0.76, + "grad_norm": 1.9835615825256219, + "learning_rate": 2.788463461797333e-06, + "loss": 1.0636, + "step": 32416 + }, + { + "epoch": 0.76, + "grad_norm": 1.9141445336960348, + "learning_rate": 2.787934866907225e-06, + "loss": 0.9482, + "step": 32417 + }, + { + "epoch": 0.76, + "grad_norm": 1.9597292019085624, + "learning_rate": 2.787406314008043e-06, + "loss": 0.9518, + "step": 32418 + }, + { + "epoch": 0.76, + "grad_norm": 2.103697566259169, + "learning_rate": 2.7868778031028688e-06, + "loss": 0.8956, + "step": 32419 + }, + { + "epoch": 0.76, + "grad_norm": 1.0293641796763688, + "learning_rate": 2.7863493341947756e-06, + "loss": 0.9054, + "step": 32420 + }, + { + "epoch": 0.76, + "grad_norm": 2.377044514795162, + "learning_rate": 2.785820907286837e-06, + "loss": 0.8703, + "step": 32421 + }, + { + "epoch": 0.76, + "grad_norm": 2.0121688320770903, + "learning_rate": 2.785292522382137e-06, + "loss": 0.9573, + "step": 32422 + }, + { + "epoch": 0.76, + "grad_norm": 1.888029516531719, + "learning_rate": 2.7847641794837432e-06, + "loss": 0.9353, + "step": 32423 + }, + { + "epoch": 0.76, + "grad_norm": 2.2542885577167886, + "learning_rate": 2.784235878594741e-06, + "loss": 1.0455, + "step": 32424 + }, + { + "epoch": 0.76, + "grad_norm": 2.278818892316105, + "learning_rate": 2.7837076197181987e-06, + "loss": 1.0114, + "step": 32425 + }, + { + "epoch": 0.76, + "grad_norm": 1.9969654481259553, + "learning_rate": 2.783179402857198e-06, + "loss": 1.14, + "step": 32426 + }, + { + "epoch": 0.76, + "grad_norm": 2.027249853053582, + "learning_rate": 2.782651228014809e-06, + "loss": 0.9027, + "step": 32427 + }, + { + "epoch": 0.76, + "grad_norm": 1.7689787672460895, + "learning_rate": 2.7821230951941125e-06, + "loss": 1.046, + "step": 32428 + }, + { + "epoch": 0.76, + "grad_norm": 1.9735025331504963, + "learning_rate": 2.781595004398181e-06, + "loss": 0.9875, + "step": 32429 + }, + { + "epoch": 0.76, + "grad_norm": 1.907845642887049, + "learning_rate": 2.781066955630085e-06, + "loss": 1.0391, + "step": 32430 + }, + { + "epoch": 0.76, + "grad_norm": 2.2472258233823075, + "learning_rate": 2.7805389488929025e-06, + "loss": 0.8783, + "step": 32431 + }, + { + "epoch": 0.76, + "grad_norm": 2.062653927428057, + "learning_rate": 2.7800109841897105e-06, + "loss": 0.8649, + "step": 32432 + }, + { + "epoch": 0.76, + "grad_norm": 2.187140982406803, + "learning_rate": 2.7794830615235812e-06, + "loss": 1.0701, + "step": 32433 + }, + { + "epoch": 0.76, + "grad_norm": 2.054652362761464, + "learning_rate": 2.7789551808975843e-06, + "loss": 1.0292, + "step": 32434 + }, + { + "epoch": 0.76, + "grad_norm": 1.9230822852304528, + "learning_rate": 2.7784273423147966e-06, + "loss": 0.9147, + "step": 32435 + }, + { + "epoch": 0.76, + "grad_norm": 2.018264794214661, + "learning_rate": 2.777899545778294e-06, + "loss": 0.8661, + "step": 32436 + }, + { + "epoch": 0.76, + "grad_norm": 2.136993774283288, + "learning_rate": 2.777371791291147e-06, + "loss": 0.9397, + "step": 32437 + }, + { + "epoch": 0.76, + "grad_norm": 1.8440991656742294, + "learning_rate": 2.7768440788564243e-06, + "loss": 0.9086, + "step": 32438 + }, + { + "epoch": 0.76, + "grad_norm": 1.939787717754416, + "learning_rate": 2.7763164084772033e-06, + "loss": 0.9299, + "step": 32439 + }, + { + "epoch": 0.76, + "grad_norm": 1.905174244967999, + "learning_rate": 2.7757887801565577e-06, + "loss": 0.8987, + "step": 32440 + }, + { + "epoch": 0.76, + "grad_norm": 1.8842640098086154, + "learning_rate": 2.7752611938975573e-06, + "loss": 0.9449, + "step": 32441 + }, + { + "epoch": 0.76, + "grad_norm": 2.0305171216409454, + "learning_rate": 2.7747336497032697e-06, + "loss": 0.7925, + "step": 32442 + }, + { + "epoch": 0.76, + "grad_norm": 2.306767420025315, + "learning_rate": 2.7742061475767735e-06, + "loss": 1.0048, + "step": 32443 + }, + { + "epoch": 0.76, + "grad_norm": 1.8917404004090068, + "learning_rate": 2.773678687521134e-06, + "loss": 0.9465, + "step": 32444 + }, + { + "epoch": 0.76, + "grad_norm": 1.9549663207576289, + "learning_rate": 2.773151269539428e-06, + "loss": 1.0066, + "step": 32445 + }, + { + "epoch": 0.76, + "grad_norm": 2.324209344267847, + "learning_rate": 2.7726238936347204e-06, + "loss": 0.9264, + "step": 32446 + }, + { + "epoch": 0.76, + "grad_norm": 1.950226373047869, + "learning_rate": 2.7720965598100867e-06, + "loss": 1.0535, + "step": 32447 + }, + { + "epoch": 0.76, + "grad_norm": 2.2303018665660708, + "learning_rate": 2.7715692680685933e-06, + "loss": 1.0098, + "step": 32448 + }, + { + "epoch": 0.76, + "grad_norm": 1.973715556050959, + "learning_rate": 2.7710420184133148e-06, + "loss": 0.9479, + "step": 32449 + }, + { + "epoch": 0.76, + "grad_norm": 1.7309884058418452, + "learning_rate": 2.7705148108473177e-06, + "loss": 1.0028, + "step": 32450 + }, + { + "epoch": 0.76, + "grad_norm": 1.9375008666705684, + "learning_rate": 2.769987645373671e-06, + "loss": 0.9955, + "step": 32451 + }, + { + "epoch": 0.76, + "grad_norm": 1.8175219991650335, + "learning_rate": 2.769460521995443e-06, + "loss": 0.9855, + "step": 32452 + }, + { + "epoch": 0.76, + "grad_norm": 1.941580045407398, + "learning_rate": 2.7689334407157086e-06, + "loss": 1.0502, + "step": 32453 + }, + { + "epoch": 0.76, + "grad_norm": 2.2395772917129104, + "learning_rate": 2.7684064015375333e-06, + "loss": 0.9392, + "step": 32454 + }, + { + "epoch": 0.76, + "grad_norm": 2.583842863796496, + "learning_rate": 2.7678794044639835e-06, + "loss": 1.0611, + "step": 32455 + }, + { + "epoch": 0.76, + "grad_norm": 1.913248019730115, + "learning_rate": 2.7673524494981275e-06, + "loss": 1.1362, + "step": 32456 + }, + { + "epoch": 0.76, + "grad_norm": 1.9462590861273013, + "learning_rate": 2.7668255366430397e-06, + "loss": 1.0717, + "step": 32457 + }, + { + "epoch": 0.76, + "grad_norm": 2.3749694698733674, + "learning_rate": 2.7662986659017833e-06, + "loss": 1.0924, + "step": 32458 + }, + { + "epoch": 0.76, + "grad_norm": 2.3240009955170513, + "learning_rate": 2.7657718372774233e-06, + "loss": 0.9867, + "step": 32459 + }, + { + "epoch": 0.76, + "grad_norm": 1.9266251714373288, + "learning_rate": 2.7652450507730297e-06, + "loss": 0.9845, + "step": 32460 + }, + { + "epoch": 0.76, + "grad_norm": 2.0008960673825875, + "learning_rate": 2.764718306391673e-06, + "loss": 1.0345, + "step": 32461 + }, + { + "epoch": 0.76, + "grad_norm": 2.0998859218957087, + "learning_rate": 2.764191604136417e-06, + "loss": 1.1683, + "step": 32462 + }, + { + "epoch": 0.76, + "grad_norm": 2.0550935488159956, + "learning_rate": 2.7636649440103247e-06, + "loss": 0.9022, + "step": 32463 + }, + { + "epoch": 0.76, + "grad_norm": 2.0134641120902033, + "learning_rate": 2.763138326016467e-06, + "loss": 0.9875, + "step": 32464 + }, + { + "epoch": 0.76, + "grad_norm": 1.97647664035368, + "learning_rate": 2.762611750157912e-06, + "loss": 1.0578, + "step": 32465 + }, + { + "epoch": 0.76, + "grad_norm": 2.0842134860679766, + "learning_rate": 2.762085216437722e-06, + "loss": 0.9321, + "step": 32466 + }, + { + "epoch": 0.76, + "grad_norm": 2.1654472063153642, + "learning_rate": 2.7615587248589614e-06, + "loss": 1.0187, + "step": 32467 + }, + { + "epoch": 0.76, + "grad_norm": 1.907532007054257, + "learning_rate": 2.7610322754247e-06, + "loss": 0.9484, + "step": 32468 + }, + { + "epoch": 0.76, + "grad_norm": 1.950192994573378, + "learning_rate": 2.760505868137997e-06, + "loss": 1.0001, + "step": 32469 + }, + { + "epoch": 0.76, + "grad_norm": 1.951203333230683, + "learning_rate": 2.7599795030019237e-06, + "loss": 0.9146, + "step": 32470 + }, + { + "epoch": 0.76, + "grad_norm": 1.9623888007686427, + "learning_rate": 2.75945318001954e-06, + "loss": 0.9753, + "step": 32471 + }, + { + "epoch": 0.77, + "grad_norm": 1.7885340452564118, + "learning_rate": 2.7589268991939145e-06, + "loss": 0.9284, + "step": 32472 + }, + { + "epoch": 0.77, + "grad_norm": 1.1788390882235904, + "learning_rate": 2.758400660528107e-06, + "loss": 0.9489, + "step": 32473 + }, + { + "epoch": 0.77, + "grad_norm": 1.7175728696814727, + "learning_rate": 2.757874464025182e-06, + "loss": 0.8779, + "step": 32474 + }, + { + "epoch": 0.77, + "grad_norm": 2.1039278072406526, + "learning_rate": 2.7573483096882114e-06, + "loss": 0.9363, + "step": 32475 + }, + { + "epoch": 0.77, + "grad_norm": 1.7860439824242844, + "learning_rate": 2.7568221975202468e-06, + "loss": 1.0818, + "step": 32476 + }, + { + "epoch": 0.77, + "grad_norm": 2.3827832717549042, + "learning_rate": 2.756296127524355e-06, + "loss": 0.9773, + "step": 32477 + }, + { + "epoch": 0.77, + "grad_norm": 1.9656350123797863, + "learning_rate": 2.755770099703602e-06, + "loss": 0.9867, + "step": 32478 + }, + { + "epoch": 0.77, + "grad_norm": 1.8391412140451053, + "learning_rate": 2.755244114061054e-06, + "loss": 0.9769, + "step": 32479 + }, + { + "epoch": 0.77, + "grad_norm": 1.9780721835909105, + "learning_rate": 2.7547181705997626e-06, + "loss": 0.9549, + "step": 32480 + }, + { + "epoch": 0.77, + "grad_norm": 2.0710903933047127, + "learning_rate": 2.754192269322795e-06, + "loss": 0.8071, + "step": 32481 + }, + { + "epoch": 0.77, + "grad_norm": 2.0789881208754895, + "learning_rate": 2.7536664102332177e-06, + "loss": 0.9853, + "step": 32482 + }, + { + "epoch": 0.77, + "grad_norm": 1.9635250285735508, + "learning_rate": 2.7531405933340894e-06, + "loss": 0.9323, + "step": 32483 + }, + { + "epoch": 0.77, + "grad_norm": 1.9759302006091322, + "learning_rate": 2.7526148186284674e-06, + "loss": 0.8441, + "step": 32484 + }, + { + "epoch": 0.77, + "grad_norm": 1.9291483660700854, + "learning_rate": 2.752089086119416e-06, + "loss": 0.9212, + "step": 32485 + }, + { + "epoch": 0.77, + "grad_norm": 2.6878502294930744, + "learning_rate": 2.7515633958099998e-06, + "loss": 0.9359, + "step": 32486 + }, + { + "epoch": 0.77, + "grad_norm": 1.9100956269699703, + "learning_rate": 2.7510377477032735e-06, + "loss": 0.9519, + "step": 32487 + }, + { + "epoch": 0.77, + "grad_norm": 2.1291566344979818, + "learning_rate": 2.750512141802304e-06, + "loss": 0.9994, + "step": 32488 + }, + { + "epoch": 0.77, + "grad_norm": 1.9622614589993523, + "learning_rate": 2.7499865781101466e-06, + "loss": 0.9802, + "step": 32489 + }, + { + "epoch": 0.77, + "grad_norm": 2.069583013959294, + "learning_rate": 2.7494610566298608e-06, + "loss": 0.9338, + "step": 32490 + }, + { + "epoch": 0.77, + "grad_norm": 2.143733995319838, + "learning_rate": 2.7489355773645078e-06, + "loss": 0.9519, + "step": 32491 + }, + { + "epoch": 0.77, + "grad_norm": 2.108961476636918, + "learning_rate": 2.7484101403171505e-06, + "loss": 0.9173, + "step": 32492 + }, + { + "epoch": 0.77, + "grad_norm": 1.9603678300430953, + "learning_rate": 2.7478847454908454e-06, + "loss": 0.9535, + "step": 32493 + }, + { + "epoch": 0.77, + "grad_norm": 2.8282292623262624, + "learning_rate": 2.747359392888648e-06, + "loss": 0.9841, + "step": 32494 + }, + { + "epoch": 0.77, + "grad_norm": 1.1022419186983914, + "learning_rate": 2.74683408251362e-06, + "loss": 0.9758, + "step": 32495 + }, + { + "epoch": 0.77, + "grad_norm": 2.044132714809346, + "learning_rate": 2.7463088143688234e-06, + "loss": 0.9097, + "step": 32496 + }, + { + "epoch": 0.77, + "grad_norm": 1.8094698396325546, + "learning_rate": 2.7457835884573137e-06, + "loss": 0.947, + "step": 32497 + }, + { + "epoch": 0.77, + "grad_norm": 1.9851700134036672, + "learning_rate": 2.7452584047821452e-06, + "loss": 1.0, + "step": 32498 + }, + { + "epoch": 0.77, + "grad_norm": 1.9273851393784966, + "learning_rate": 2.7447332633463795e-06, + "loss": 0.9093, + "step": 32499 + }, + { + "epoch": 0.77, + "grad_norm": 2.195069323496988, + "learning_rate": 2.744208164153077e-06, + "loss": 0.9699, + "step": 32500 + }, + { + "epoch": 0.77, + "grad_norm": 2.0639649862258245, + "learning_rate": 2.7436831072052915e-06, + "loss": 0.8515, + "step": 32501 + }, + { + "epoch": 0.77, + "grad_norm": 2.197793973130776, + "learning_rate": 2.7431580925060762e-06, + "loss": 1.0625, + "step": 32502 + }, + { + "epoch": 0.77, + "grad_norm": 1.1269527238823627, + "learning_rate": 2.7426331200584966e-06, + "loss": 1.0017, + "step": 32503 + }, + { + "epoch": 0.77, + "grad_norm": 1.074531864352534, + "learning_rate": 2.742108189865601e-06, + "loss": 0.9633, + "step": 32504 + }, + { + "epoch": 0.77, + "grad_norm": 1.8499157086948683, + "learning_rate": 2.7415833019304517e-06, + "loss": 0.9986, + "step": 32505 + }, + { + "epoch": 0.77, + "grad_norm": 1.9526876788490064, + "learning_rate": 2.7410584562561004e-06, + "loss": 0.9194, + "step": 32506 + }, + { + "epoch": 0.77, + "grad_norm": 1.9683711604223342, + "learning_rate": 2.7405336528456073e-06, + "loss": 0.936, + "step": 32507 + }, + { + "epoch": 0.77, + "grad_norm": 1.9870691117635115, + "learning_rate": 2.740008891702023e-06, + "loss": 1.0528, + "step": 32508 + }, + { + "epoch": 0.77, + "grad_norm": 1.9817315419431714, + "learning_rate": 2.7394841728284083e-06, + "loss": 0.9489, + "step": 32509 + }, + { + "epoch": 0.77, + "grad_norm": 1.9475683727343116, + "learning_rate": 2.738959496227814e-06, + "loss": 0.927, + "step": 32510 + }, + { + "epoch": 0.77, + "grad_norm": 1.9982550906177776, + "learning_rate": 2.738434861903295e-06, + "loss": 0.9747, + "step": 32511 + }, + { + "epoch": 0.77, + "grad_norm": 2.087214665034616, + "learning_rate": 2.7379102698579075e-06, + "loss": 1.0931, + "step": 32512 + }, + { + "epoch": 0.77, + "grad_norm": 2.1506122584577496, + "learning_rate": 2.7373857200947073e-06, + "loss": 1.0499, + "step": 32513 + }, + { + "epoch": 0.77, + "grad_norm": 1.8784465493858182, + "learning_rate": 2.736861212616747e-06, + "loss": 1.0159, + "step": 32514 + }, + { + "epoch": 0.77, + "grad_norm": 2.0691937549192665, + "learning_rate": 2.7363367474270774e-06, + "loss": 1.0202, + "step": 32515 + }, + { + "epoch": 0.77, + "grad_norm": 1.1800691307254507, + "learning_rate": 2.7358123245287548e-06, + "loss": 0.9522, + "step": 32516 + }, + { + "epoch": 0.77, + "grad_norm": 2.2461541490104815, + "learning_rate": 2.7352879439248358e-06, + "loss": 1.1687, + "step": 32517 + }, + { + "epoch": 0.77, + "grad_norm": 2.5637925243934077, + "learning_rate": 2.73476360561837e-06, + "loss": 0.8871, + "step": 32518 + }, + { + "epoch": 0.77, + "grad_norm": 1.084391896329396, + "learning_rate": 2.734239309612409e-06, + "loss": 0.9093, + "step": 32519 + }, + { + "epoch": 0.77, + "grad_norm": 1.783904547557092, + "learning_rate": 2.7337150559100055e-06, + "loss": 0.9386, + "step": 32520 + }, + { + "epoch": 0.77, + "grad_norm": 1.8933169281198383, + "learning_rate": 2.7331908445142165e-06, + "loss": 0.9805, + "step": 32521 + }, + { + "epoch": 0.77, + "grad_norm": 1.964763280067264, + "learning_rate": 2.7326666754280918e-06, + "loss": 0.9776, + "step": 32522 + }, + { + "epoch": 0.77, + "grad_norm": 1.8295273095030127, + "learning_rate": 2.7321425486546792e-06, + "loss": 0.9008, + "step": 32523 + }, + { + "epoch": 0.77, + "grad_norm": 1.9607231365042685, + "learning_rate": 2.7316184641970367e-06, + "loss": 0.9467, + "step": 32524 + }, + { + "epoch": 0.77, + "grad_norm": 2.0689903833072796, + "learning_rate": 2.7310944220582092e-06, + "loss": 0.9347, + "step": 32525 + }, + { + "epoch": 0.77, + "grad_norm": 2.074442164351325, + "learning_rate": 2.7305704222412543e-06, + "loss": 1.0218, + "step": 32526 + }, + { + "epoch": 0.77, + "grad_norm": 2.293446120008178, + "learning_rate": 2.7300464647492177e-06, + "loss": 0.9826, + "step": 32527 + }, + { + "epoch": 0.77, + "grad_norm": 1.85877968714485, + "learning_rate": 2.729522549585154e-06, + "loss": 0.8969, + "step": 32528 + }, + { + "epoch": 0.77, + "grad_norm": 1.9999870201151582, + "learning_rate": 2.7289986767521105e-06, + "loss": 0.8668, + "step": 32529 + }, + { + "epoch": 0.77, + "grad_norm": 2.2365507841666172, + "learning_rate": 2.7284748462531394e-06, + "loss": 0.97, + "step": 32530 + }, + { + "epoch": 0.77, + "grad_norm": 2.0478708186037857, + "learning_rate": 2.7279510580912916e-06, + "loss": 0.97, + "step": 32531 + }, + { + "epoch": 0.77, + "grad_norm": 1.9700457024087497, + "learning_rate": 2.7274273122696118e-06, + "loss": 0.9257, + "step": 32532 + }, + { + "epoch": 0.77, + "grad_norm": 1.9656697597926351, + "learning_rate": 2.7269036087911525e-06, + "loss": 0.8457, + "step": 32533 + }, + { + "epoch": 0.77, + "grad_norm": 1.808222787809996, + "learning_rate": 2.726379947658966e-06, + "loss": 0.9766, + "step": 32534 + }, + { + "epoch": 0.77, + "grad_norm": 2.5806149345533904, + "learning_rate": 2.7258563288760976e-06, + "loss": 0.9907, + "step": 32535 + }, + { + "epoch": 0.77, + "grad_norm": 2.7967523632521014, + "learning_rate": 2.725332752445594e-06, + "loss": 1.0477, + "step": 32536 + }, + { + "epoch": 0.77, + "grad_norm": 2.2929841423655297, + "learning_rate": 2.7248092183705065e-06, + "loss": 1.0439, + "step": 32537 + }, + { + "epoch": 0.77, + "grad_norm": 2.0272413882100744, + "learning_rate": 2.7242857266538857e-06, + "loss": 1.0462, + "step": 32538 + }, + { + "epoch": 0.77, + "grad_norm": 2.3908641724664874, + "learning_rate": 2.7237622772987772e-06, + "loss": 0.9587, + "step": 32539 + }, + { + "epoch": 0.77, + "grad_norm": 1.0876516453356162, + "learning_rate": 2.7232388703082258e-06, + "loss": 0.9321, + "step": 32540 + }, + { + "epoch": 0.77, + "grad_norm": 2.4708089791679777, + "learning_rate": 2.7227155056852806e-06, + "loss": 1.0464, + "step": 32541 + }, + { + "epoch": 0.77, + "grad_norm": 1.9305265654563522, + "learning_rate": 2.7221921834329923e-06, + "loss": 1.0503, + "step": 32542 + }, + { + "epoch": 0.77, + "grad_norm": 1.9112048980395377, + "learning_rate": 2.7216689035544064e-06, + "loss": 1.0338, + "step": 32543 + }, + { + "epoch": 0.77, + "grad_norm": 2.297298252054911, + "learning_rate": 2.7211456660525647e-06, + "loss": 0.8759, + "step": 32544 + }, + { + "epoch": 0.77, + "grad_norm": 1.8995374675025563, + "learning_rate": 2.7206224709305206e-06, + "loss": 0.8667, + "step": 32545 + }, + { + "epoch": 0.77, + "grad_norm": 1.9304952575400927, + "learning_rate": 2.7200993181913137e-06, + "loss": 1.0659, + "step": 32546 + }, + { + "epoch": 0.77, + "grad_norm": 1.8923931942460193, + "learning_rate": 2.719576207837997e-06, + "loss": 0.9386, + "step": 32547 + }, + { + "epoch": 0.77, + "grad_norm": 2.098591001463785, + "learning_rate": 2.7190531398736098e-06, + "loss": 0.9168, + "step": 32548 + }, + { + "epoch": 0.77, + "grad_norm": 2.289452145526965, + "learning_rate": 2.718530114301202e-06, + "loss": 0.8705, + "step": 32549 + }, + { + "epoch": 0.77, + "grad_norm": 1.8983251960085665, + "learning_rate": 2.7180071311238152e-06, + "loss": 1.0601, + "step": 32550 + }, + { + "epoch": 0.77, + "grad_norm": 2.4958146501518432, + "learning_rate": 2.717484190344499e-06, + "loss": 1.0418, + "step": 32551 + }, + { + "epoch": 0.77, + "grad_norm": 1.955457621607318, + "learning_rate": 2.716961291966296e-06, + "loss": 1.069, + "step": 32552 + }, + { + "epoch": 0.77, + "grad_norm": 2.279966142878354, + "learning_rate": 2.7164384359922467e-06, + "loss": 0.9124, + "step": 32553 + }, + { + "epoch": 0.77, + "grad_norm": 2.0457732516252003, + "learning_rate": 2.715915622425398e-06, + "loss": 0.9772, + "step": 32554 + }, + { + "epoch": 0.77, + "grad_norm": 2.168533299051254, + "learning_rate": 2.715392851268799e-06, + "loss": 1.0848, + "step": 32555 + }, + { + "epoch": 0.77, + "grad_norm": 2.7743522183307263, + "learning_rate": 2.714870122525488e-06, + "loss": 1.0406, + "step": 32556 + }, + { + "epoch": 0.77, + "grad_norm": 2.105460836224466, + "learning_rate": 2.7143474361985065e-06, + "loss": 0.9124, + "step": 32557 + }, + { + "epoch": 0.77, + "grad_norm": 1.9572247305565416, + "learning_rate": 2.7138247922909024e-06, + "loss": 0.974, + "step": 32558 + }, + { + "epoch": 0.77, + "grad_norm": 2.0367226951423367, + "learning_rate": 2.713302190805719e-06, + "loss": 0.9361, + "step": 32559 + }, + { + "epoch": 0.77, + "grad_norm": 1.9794560008277107, + "learning_rate": 2.7127796317459976e-06, + "loss": 1.0002, + "step": 32560 + }, + { + "epoch": 0.77, + "grad_norm": 1.8892526443727828, + "learning_rate": 2.7122571151147757e-06, + "loss": 1.0496, + "step": 32561 + }, + { + "epoch": 0.77, + "grad_norm": 1.9460543713754783, + "learning_rate": 2.7117346409151023e-06, + "loss": 1.0441, + "step": 32562 + }, + { + "epoch": 0.77, + "grad_norm": 2.0515074815512593, + "learning_rate": 2.7112122091500194e-06, + "loss": 0.903, + "step": 32563 + }, + { + "epoch": 0.77, + "grad_norm": 1.0248420419531907, + "learning_rate": 2.710689819822566e-06, + "loss": 0.9144, + "step": 32564 + }, + { + "epoch": 0.77, + "grad_norm": 1.1005067709050709, + "learning_rate": 2.7101674729357817e-06, + "loss": 0.9605, + "step": 32565 + }, + { + "epoch": 0.77, + "grad_norm": 2.1714192580046845, + "learning_rate": 2.7096451684927138e-06, + "loss": 1.0925, + "step": 32566 + }, + { + "epoch": 0.77, + "grad_norm": 1.958278879763581, + "learning_rate": 2.7091229064963964e-06, + "loss": 0.9598, + "step": 32567 + }, + { + "epoch": 0.77, + "grad_norm": 2.333266105820483, + "learning_rate": 2.708600686949876e-06, + "loss": 0.9868, + "step": 32568 + }, + { + "epoch": 0.77, + "grad_norm": 1.9777462469876113, + "learning_rate": 2.708078509856189e-06, + "loss": 1.0161, + "step": 32569 + }, + { + "epoch": 0.77, + "grad_norm": 2.0609561322780245, + "learning_rate": 2.70755637521838e-06, + "loss": 0.9659, + "step": 32570 + }, + { + "epoch": 0.77, + "grad_norm": 2.553423925211851, + "learning_rate": 2.7070342830394836e-06, + "loss": 0.9989, + "step": 32571 + }, + { + "epoch": 0.77, + "grad_norm": 2.7595248469966975, + "learning_rate": 2.7065122333225426e-06, + "loss": 0.9634, + "step": 32572 + }, + { + "epoch": 0.77, + "grad_norm": 1.9629080426453962, + "learning_rate": 2.705990226070602e-06, + "loss": 0.9367, + "step": 32573 + }, + { + "epoch": 0.77, + "grad_norm": 2.2619364344742423, + "learning_rate": 2.70546826128669e-06, + "loss": 0.8472, + "step": 32574 + }, + { + "epoch": 0.77, + "grad_norm": 2.018759728829367, + "learning_rate": 2.7049463389738506e-06, + "loss": 1.0128, + "step": 32575 + }, + { + "epoch": 0.77, + "grad_norm": 1.7981980103692508, + "learning_rate": 2.704424459135123e-06, + "loss": 1.0078, + "step": 32576 + }, + { + "epoch": 0.77, + "grad_norm": 2.211423299305408, + "learning_rate": 2.703902621773553e-06, + "loss": 1.0414, + "step": 32577 + }, + { + "epoch": 0.77, + "grad_norm": 1.8911442299198307, + "learning_rate": 2.7033808268921658e-06, + "loss": 1.0219, + "step": 32578 + }, + { + "epoch": 0.77, + "grad_norm": 2.0306636683904045, + "learning_rate": 2.702859074494004e-06, + "loss": 0.9691, + "step": 32579 + }, + { + "epoch": 0.77, + "grad_norm": 2.1057644275495186, + "learning_rate": 2.7023373645821115e-06, + "loss": 1.0916, + "step": 32580 + }, + { + "epoch": 0.77, + "grad_norm": 1.9778989498539046, + "learning_rate": 2.7018156971595177e-06, + "loss": 0.9384, + "step": 32581 + }, + { + "epoch": 0.77, + "grad_norm": 1.9039665002832202, + "learning_rate": 2.701294072229267e-06, + "loss": 1.0052, + "step": 32582 + }, + { + "epoch": 0.77, + "grad_norm": 2.079857697918701, + "learning_rate": 2.7007724897943887e-06, + "loss": 0.9186, + "step": 32583 + }, + { + "epoch": 0.77, + "grad_norm": 1.9150001979927496, + "learning_rate": 2.7002509498579275e-06, + "loss": 0.9617, + "step": 32584 + }, + { + "epoch": 0.77, + "grad_norm": 2.2208392055214023, + "learning_rate": 2.6997294524229145e-06, + "loss": 1.0649, + "step": 32585 + }, + { + "epoch": 0.77, + "grad_norm": 2.2140958416155634, + "learning_rate": 2.699207997492389e-06, + "loss": 0.9687, + "step": 32586 + }, + { + "epoch": 0.77, + "grad_norm": 1.9213941540022854, + "learning_rate": 2.6986865850693877e-06, + "loss": 1.0996, + "step": 32587 + }, + { + "epoch": 0.77, + "grad_norm": 2.4868186400714505, + "learning_rate": 2.6981652151569404e-06, + "loss": 1.0951, + "step": 32588 + }, + { + "epoch": 0.77, + "grad_norm": 2.317331385654737, + "learning_rate": 2.6976438877580867e-06, + "loss": 1.0868, + "step": 32589 + }, + { + "epoch": 0.77, + "grad_norm": 2.0787919143372835, + "learning_rate": 2.6971226028758655e-06, + "loss": 0.9525, + "step": 32590 + }, + { + "epoch": 0.77, + "grad_norm": 1.9302986875727772, + "learning_rate": 2.696601360513309e-06, + "loss": 0.9974, + "step": 32591 + }, + { + "epoch": 0.77, + "grad_norm": 2.127528862428391, + "learning_rate": 2.6960801606734487e-06, + "loss": 0.9531, + "step": 32592 + }, + { + "epoch": 0.77, + "grad_norm": 2.151851808991944, + "learning_rate": 2.6955590033593217e-06, + "loss": 0.9996, + "step": 32593 + }, + { + "epoch": 0.77, + "grad_norm": 2.163636169405421, + "learning_rate": 2.695037888573966e-06, + "loss": 1.0666, + "step": 32594 + }, + { + "epoch": 0.77, + "grad_norm": 2.0502757725604313, + "learning_rate": 2.6945168163204116e-06, + "loss": 0.9573, + "step": 32595 + }, + { + "epoch": 0.77, + "grad_norm": 2.035873841665013, + "learning_rate": 2.693995786601691e-06, + "loss": 0.867, + "step": 32596 + }, + { + "epoch": 0.77, + "grad_norm": 2.1314920223584632, + "learning_rate": 2.6934747994208397e-06, + "loss": 0.9858, + "step": 32597 + }, + { + "epoch": 0.77, + "grad_norm": 2.5830292269066026, + "learning_rate": 2.692953854780894e-06, + "loss": 1.0052, + "step": 32598 + }, + { + "epoch": 0.77, + "grad_norm": 1.9203612826685201, + "learning_rate": 2.692432952684886e-06, + "loss": 0.9392, + "step": 32599 + }, + { + "epoch": 0.77, + "grad_norm": 1.8809489714593062, + "learning_rate": 2.691912093135842e-06, + "loss": 0.9349, + "step": 32600 + }, + { + "epoch": 0.77, + "grad_norm": 1.9981949901404037, + "learning_rate": 2.6913912761368e-06, + "loss": 0.9505, + "step": 32601 + }, + { + "epoch": 0.77, + "grad_norm": 1.9680231419271574, + "learning_rate": 2.6908705016907956e-06, + "loss": 1.0318, + "step": 32602 + }, + { + "epoch": 0.77, + "grad_norm": 1.8928506789528259, + "learning_rate": 2.6903497698008564e-06, + "loss": 0.9092, + "step": 32603 + }, + { + "epoch": 0.77, + "grad_norm": 1.7841048653003893, + "learning_rate": 2.689829080470011e-06, + "loss": 0.9768, + "step": 32604 + }, + { + "epoch": 0.77, + "grad_norm": 1.7188445084229131, + "learning_rate": 2.6893084337013e-06, + "loss": 0.9453, + "step": 32605 + }, + { + "epoch": 0.77, + "grad_norm": 2.134302766731856, + "learning_rate": 2.6887878294977456e-06, + "loss": 0.9806, + "step": 32606 + }, + { + "epoch": 0.77, + "grad_norm": 2.151663237010153, + "learning_rate": 2.688267267862387e-06, + "loss": 1.0595, + "step": 32607 + }, + { + "epoch": 0.77, + "grad_norm": 1.8349660362119014, + "learning_rate": 2.687746748798248e-06, + "loss": 1.0824, + "step": 32608 + }, + { + "epoch": 0.77, + "grad_norm": 2.193581920235145, + "learning_rate": 2.687226272308365e-06, + "loss": 0.96, + "step": 32609 + }, + { + "epoch": 0.77, + "grad_norm": 2.19238520879126, + "learning_rate": 2.6867058383957633e-06, + "loss": 0.994, + "step": 32610 + }, + { + "epoch": 0.77, + "grad_norm": 2.053526367150696, + "learning_rate": 2.6861854470634784e-06, + "loss": 1.1399, + "step": 32611 + }, + { + "epoch": 0.77, + "grad_norm": 2.1598789163597836, + "learning_rate": 2.685665098314537e-06, + "loss": 0.9045, + "step": 32612 + }, + { + "epoch": 0.77, + "grad_norm": 1.8948813241562605, + "learning_rate": 2.685144792151966e-06, + "loss": 0.9405, + "step": 32613 + }, + { + "epoch": 0.77, + "grad_norm": 2.08301864559506, + "learning_rate": 2.6846245285787985e-06, + "loss": 1.0045, + "step": 32614 + }, + { + "epoch": 0.77, + "grad_norm": 2.0670846999846852, + "learning_rate": 2.6841043075980653e-06, + "loss": 0.9797, + "step": 32615 + }, + { + "epoch": 0.77, + "grad_norm": 2.0512611993521936, + "learning_rate": 2.683584129212794e-06, + "loss": 0.9221, + "step": 32616 + }, + { + "epoch": 0.77, + "grad_norm": 2.143500683979686, + "learning_rate": 2.6830639934260084e-06, + "loss": 0.962, + "step": 32617 + }, + { + "epoch": 0.77, + "grad_norm": 2.334343209225334, + "learning_rate": 2.6825439002407405e-06, + "loss": 0.8659, + "step": 32618 + }, + { + "epoch": 0.77, + "grad_norm": 1.9592678913425956, + "learning_rate": 2.682023849660023e-06, + "loss": 1.0277, + "step": 32619 + }, + { + "epoch": 0.77, + "grad_norm": 2.035031349553599, + "learning_rate": 2.681503841686879e-06, + "loss": 0.9843, + "step": 32620 + }, + { + "epoch": 0.77, + "grad_norm": 1.8748669395817619, + "learning_rate": 2.6809838763243325e-06, + "loss": 0.9515, + "step": 32621 + }, + { + "epoch": 0.77, + "grad_norm": 2.0688351632362925, + "learning_rate": 2.680463953575416e-06, + "loss": 1.0011, + "step": 32622 + }, + { + "epoch": 0.77, + "grad_norm": 1.9171091134294715, + "learning_rate": 2.679944073443158e-06, + "loss": 0.9426, + "step": 32623 + }, + { + "epoch": 0.77, + "grad_norm": 1.9792189797694695, + "learning_rate": 2.6794242359305843e-06, + "loss": 0.9751, + "step": 32624 + }, + { + "epoch": 0.77, + "grad_norm": 1.9639118017367359, + "learning_rate": 2.6789044410407162e-06, + "loss": 1.0572, + "step": 32625 + }, + { + "epoch": 0.77, + "grad_norm": 1.7780356994431943, + "learning_rate": 2.6783846887765886e-06, + "loss": 0.8892, + "step": 32626 + }, + { + "epoch": 0.77, + "grad_norm": 2.9021895275882135, + "learning_rate": 2.6778649791412194e-06, + "loss": 0.8917, + "step": 32627 + }, + { + "epoch": 0.77, + "grad_norm": 4.864145737340799, + "learning_rate": 2.6773453121376414e-06, + "loss": 1.0211, + "step": 32628 + }, + { + "epoch": 0.77, + "grad_norm": 2.0626118017368755, + "learning_rate": 2.6768256877688735e-06, + "loss": 0.9042, + "step": 32629 + }, + { + "epoch": 0.77, + "grad_norm": 2.222290063316892, + "learning_rate": 2.6763061060379492e-06, + "loss": 1.0505, + "step": 32630 + }, + { + "epoch": 0.77, + "grad_norm": 2.101250619041304, + "learning_rate": 2.6757865669478854e-06, + "loss": 0.9849, + "step": 32631 + }, + { + "epoch": 0.77, + "grad_norm": 1.8725121746004465, + "learning_rate": 2.6752670705017148e-06, + "loss": 0.8034, + "step": 32632 + }, + { + "epoch": 0.77, + "grad_norm": 1.8802531605827129, + "learning_rate": 2.6747476167024587e-06, + "loss": 1.034, + "step": 32633 + }, + { + "epoch": 0.77, + "grad_norm": 2.052445157652933, + "learning_rate": 2.674228205553138e-06, + "loss": 0.9789, + "step": 32634 + }, + { + "epoch": 0.77, + "grad_norm": 2.209164093716899, + "learning_rate": 2.673708837056779e-06, + "loss": 0.9377, + "step": 32635 + }, + { + "epoch": 0.77, + "grad_norm": 1.9098251388313674, + "learning_rate": 2.6731895112164096e-06, + "loss": 0.8811, + "step": 32636 + }, + { + "epoch": 0.77, + "grad_norm": 1.840958533422672, + "learning_rate": 2.6726702280350513e-06, + "loss": 1.0776, + "step": 32637 + }, + { + "epoch": 0.77, + "grad_norm": 1.888458395834172, + "learning_rate": 2.672150987515724e-06, + "loss": 0.9684, + "step": 32638 + }, + { + "epoch": 0.77, + "grad_norm": 2.2585936454372453, + "learning_rate": 2.671631789661453e-06, + "loss": 1.0106, + "step": 32639 + }, + { + "epoch": 0.77, + "grad_norm": 2.100838785192255, + "learning_rate": 2.6711126344752647e-06, + "loss": 1.101, + "step": 32640 + }, + { + "epoch": 0.77, + "grad_norm": 1.0733552305687277, + "learning_rate": 2.6705935219601788e-06, + "loss": 0.9663, + "step": 32641 + }, + { + "epoch": 0.77, + "grad_norm": 1.929309565440023, + "learning_rate": 2.670074452119216e-06, + "loss": 0.92, + "step": 32642 + }, + { + "epoch": 0.77, + "grad_norm": 2.430184238795455, + "learning_rate": 2.669555424955399e-06, + "loss": 1.0322, + "step": 32643 + }, + { + "epoch": 0.77, + "grad_norm": 1.8497164490330464, + "learning_rate": 2.669036440471754e-06, + "loss": 1.0922, + "step": 32644 + }, + { + "epoch": 0.77, + "grad_norm": 1.1047171103138964, + "learning_rate": 2.6685174986713012e-06, + "loss": 0.9824, + "step": 32645 + }, + { + "epoch": 0.77, + "grad_norm": 2.005034670151264, + "learning_rate": 2.667998599557057e-06, + "loss": 0.8728, + "step": 32646 + }, + { + "epoch": 0.77, + "grad_norm": 1.8958783141381068, + "learning_rate": 2.667479743132049e-06, + "loss": 1.069, + "step": 32647 + }, + { + "epoch": 0.77, + "grad_norm": 2.527212303922536, + "learning_rate": 2.6669609293992914e-06, + "loss": 0.9829, + "step": 32648 + }, + { + "epoch": 0.77, + "grad_norm": 1.9859214737329538, + "learning_rate": 2.6664421583618138e-06, + "loss": 0.8939, + "step": 32649 + }, + { + "epoch": 0.77, + "grad_norm": 1.8315581202484146, + "learning_rate": 2.665923430022628e-06, + "loss": 0.9175, + "step": 32650 + }, + { + "epoch": 0.77, + "grad_norm": 2.1449296723197833, + "learning_rate": 2.6654047443847607e-06, + "loss": 0.9193, + "step": 32651 + }, + { + "epoch": 0.77, + "grad_norm": 1.9077882683027008, + "learning_rate": 2.6648861014512263e-06, + "loss": 1.0039, + "step": 32652 + }, + { + "epoch": 0.77, + "grad_norm": 2.00212537896704, + "learning_rate": 2.664367501225049e-06, + "loss": 1.0517, + "step": 32653 + }, + { + "epoch": 0.77, + "grad_norm": 2.041085022107016, + "learning_rate": 2.6638489437092486e-06, + "loss": 1.2068, + "step": 32654 + }, + { + "epoch": 0.77, + "grad_norm": 2.1921656477662186, + "learning_rate": 2.6633304289068384e-06, + "loss": 0.9624, + "step": 32655 + }, + { + "epoch": 0.77, + "grad_norm": 1.8745080332882527, + "learning_rate": 2.662811956820841e-06, + "loss": 0.9251, + "step": 32656 + }, + { + "epoch": 0.77, + "grad_norm": 1.9869430654666935, + "learning_rate": 2.662293527454275e-06, + "loss": 0.9753, + "step": 32657 + }, + { + "epoch": 0.77, + "grad_norm": 4.310245517965674, + "learning_rate": 2.6617751408101655e-06, + "loss": 0.8503, + "step": 32658 + }, + { + "epoch": 0.77, + "grad_norm": 2.131304854742126, + "learning_rate": 2.661256796891518e-06, + "loss": 1.0261, + "step": 32659 + }, + { + "epoch": 0.77, + "grad_norm": 1.1339877249855155, + "learning_rate": 2.6607384957013583e-06, + "loss": 0.9259, + "step": 32660 + }, + { + "epoch": 0.77, + "grad_norm": 2.2520693764907866, + "learning_rate": 2.6602202372427046e-06, + "loss": 1.0179, + "step": 32661 + }, + { + "epoch": 0.77, + "grad_norm": 2.0282927438013063, + "learning_rate": 2.659702021518572e-06, + "loss": 1.0168, + "step": 32662 + }, + { + "epoch": 0.77, + "grad_norm": 2.1026685868506223, + "learning_rate": 2.6591838485319754e-06, + "loss": 1.0062, + "step": 32663 + }, + { + "epoch": 0.77, + "grad_norm": 2.2009804385577048, + "learning_rate": 2.658665718285934e-06, + "loss": 1.0303, + "step": 32664 + }, + { + "epoch": 0.77, + "grad_norm": 1.8886879082115582, + "learning_rate": 2.6581476307834697e-06, + "loss": 0.9388, + "step": 32665 + }, + { + "epoch": 0.77, + "grad_norm": 1.7598082017828542, + "learning_rate": 2.6576295860275927e-06, + "loss": 1.009, + "step": 32666 + }, + { + "epoch": 0.77, + "grad_norm": 2.0709180867269326, + "learning_rate": 2.657111584021318e-06, + "loss": 1.0261, + "step": 32667 + }, + { + "epoch": 0.77, + "grad_norm": 1.0819599595280278, + "learning_rate": 2.6565936247676684e-06, + "loss": 0.9201, + "step": 32668 + }, + { + "epoch": 0.77, + "grad_norm": 3.1990122793250957, + "learning_rate": 2.656075708269652e-06, + "loss": 0.914, + "step": 32669 + }, + { + "epoch": 0.77, + "grad_norm": 1.888761855282853, + "learning_rate": 2.655557834530288e-06, + "loss": 0.9748, + "step": 32670 + }, + { + "epoch": 0.77, + "grad_norm": 1.920800733639013, + "learning_rate": 2.6550400035525936e-06, + "loss": 1.0216, + "step": 32671 + }, + { + "epoch": 0.77, + "grad_norm": 2.529166202317883, + "learning_rate": 2.6545222153395822e-06, + "loss": 1.0513, + "step": 32672 + }, + { + "epoch": 0.77, + "grad_norm": 1.9393022439886032, + "learning_rate": 2.6540044698942655e-06, + "loss": 0.7656, + "step": 32673 + }, + { + "epoch": 0.77, + "grad_norm": 2.227356918310084, + "learning_rate": 2.6534867672196594e-06, + "loss": 1.1277, + "step": 32674 + }, + { + "epoch": 0.77, + "grad_norm": 2.7801135015159733, + "learning_rate": 2.652969107318787e-06, + "loss": 1.0784, + "step": 32675 + }, + { + "epoch": 0.77, + "grad_norm": 1.9988837659221037, + "learning_rate": 2.6524514901946473e-06, + "loss": 0.8917, + "step": 32676 + }, + { + "epoch": 0.77, + "grad_norm": 1.8892035321579976, + "learning_rate": 2.651933915850261e-06, + "loss": 0.979, + "step": 32677 + }, + { + "epoch": 0.77, + "grad_norm": 2.074749180017059, + "learning_rate": 2.651416384288642e-06, + "loss": 0.8962, + "step": 32678 + }, + { + "epoch": 0.77, + "grad_norm": 1.821961776394524, + "learning_rate": 2.6508988955128067e-06, + "loss": 0.8697, + "step": 32679 + }, + { + "epoch": 0.77, + "grad_norm": 2.1368370016892544, + "learning_rate": 2.6503814495257652e-06, + "loss": 0.9893, + "step": 32680 + }, + { + "epoch": 0.77, + "grad_norm": 1.9696667455972239, + "learning_rate": 2.649864046330526e-06, + "loss": 0.9864, + "step": 32681 + }, + { + "epoch": 0.77, + "grad_norm": 2.10559334656102, + "learning_rate": 2.6493466859301074e-06, + "loss": 1.1622, + "step": 32682 + }, + { + "epoch": 0.77, + "grad_norm": 1.1152880237355418, + "learning_rate": 2.6488293683275177e-06, + "loss": 0.9276, + "step": 32683 + }, + { + "epoch": 0.77, + "grad_norm": 2.1763098386481454, + "learning_rate": 2.6483120935257744e-06, + "loss": 0.9207, + "step": 32684 + }, + { + "epoch": 0.77, + "grad_norm": 1.0656972449964517, + "learning_rate": 2.6477948615278813e-06, + "loss": 0.9356, + "step": 32685 + }, + { + "epoch": 0.77, + "grad_norm": 1.9589750457514907, + "learning_rate": 2.6472776723368577e-06, + "loss": 0.9453, + "step": 32686 + }, + { + "epoch": 0.77, + "grad_norm": 2.0975989565004394, + "learning_rate": 2.646760525955708e-06, + "loss": 0.9412, + "step": 32687 + }, + { + "epoch": 0.77, + "grad_norm": 1.9142822415616858, + "learning_rate": 2.64624342238745e-06, + "loss": 0.8068, + "step": 32688 + }, + { + "epoch": 0.77, + "grad_norm": 1.0929577237838235, + "learning_rate": 2.6457263616350915e-06, + "loss": 0.9557, + "step": 32689 + }, + { + "epoch": 0.77, + "grad_norm": 1.1241190864050292, + "learning_rate": 2.645209343701638e-06, + "loss": 1.0204, + "step": 32690 + }, + { + "epoch": 0.77, + "grad_norm": 1.9591276578998498, + "learning_rate": 2.644692368590105e-06, + "loss": 1.117, + "step": 32691 + }, + { + "epoch": 0.77, + "grad_norm": 2.228497139177455, + "learning_rate": 2.6441754363035045e-06, + "loss": 1.0528, + "step": 32692 + }, + { + "epoch": 0.77, + "grad_norm": 2.3247941517024304, + "learning_rate": 2.6436585468448438e-06, + "loss": 0.9496, + "step": 32693 + }, + { + "epoch": 0.77, + "grad_norm": 1.8339887088048021, + "learning_rate": 2.643141700217129e-06, + "loss": 0.9447, + "step": 32694 + }, + { + "epoch": 0.77, + "grad_norm": 2.149874574277011, + "learning_rate": 2.6426248964233725e-06, + "loss": 0.9249, + "step": 32695 + }, + { + "epoch": 0.77, + "grad_norm": 2.468357703307513, + "learning_rate": 2.6421081354665856e-06, + "loss": 1.034, + "step": 32696 + }, + { + "epoch": 0.77, + "grad_norm": 1.8403743237149652, + "learning_rate": 2.6415914173497746e-06, + "loss": 1.0037, + "step": 32697 + }, + { + "epoch": 0.77, + "grad_norm": 2.4110827307523164, + "learning_rate": 2.6410747420759455e-06, + "loss": 1.0245, + "step": 32698 + }, + { + "epoch": 0.77, + "grad_norm": 3.525142022861934, + "learning_rate": 2.6405581096481083e-06, + "loss": 0.9581, + "step": 32699 + }, + { + "epoch": 0.77, + "grad_norm": 1.8306501744722434, + "learning_rate": 2.6400415200692754e-06, + "loss": 0.9377, + "step": 32700 + }, + { + "epoch": 0.77, + "grad_norm": 2.80678451795717, + "learning_rate": 2.63952497334245e-06, + "loss": 0.9502, + "step": 32701 + }, + { + "epoch": 0.77, + "grad_norm": 1.8994474063633524, + "learning_rate": 2.639008469470639e-06, + "loss": 1.002, + "step": 32702 + }, + { + "epoch": 0.77, + "grad_norm": 1.8789041121623355, + "learning_rate": 2.638492008456853e-06, + "loss": 0.9993, + "step": 32703 + }, + { + "epoch": 0.77, + "grad_norm": 2.03819104254007, + "learning_rate": 2.6379755903040947e-06, + "loss": 1.0148, + "step": 32704 + }, + { + "epoch": 0.77, + "grad_norm": 2.073966285953805, + "learning_rate": 2.6374592150153755e-06, + "loss": 1.0454, + "step": 32705 + }, + { + "epoch": 0.77, + "grad_norm": 2.463682584219895, + "learning_rate": 2.6369428825936973e-06, + "loss": 0.9819, + "step": 32706 + }, + { + "epoch": 0.77, + "grad_norm": 2.1986837579084613, + "learning_rate": 2.636426593042072e-06, + "loss": 0.89, + "step": 32707 + }, + { + "epoch": 0.77, + "grad_norm": 2.0631488452591022, + "learning_rate": 2.635910346363499e-06, + "loss": 1.0213, + "step": 32708 + }, + { + "epoch": 0.77, + "grad_norm": 1.9440361713450212, + "learning_rate": 2.6353941425609907e-06, + "loss": 1.054, + "step": 32709 + }, + { + "epoch": 0.77, + "grad_norm": 1.8480104588641026, + "learning_rate": 2.634877981637548e-06, + "loss": 0.9403, + "step": 32710 + }, + { + "epoch": 0.77, + "grad_norm": 2.110310338465886, + "learning_rate": 2.6343618635961765e-06, + "loss": 1.0358, + "step": 32711 + }, + { + "epoch": 0.77, + "grad_norm": 2.060889799161252, + "learning_rate": 2.6338457884398805e-06, + "loss": 0.9562, + "step": 32712 + }, + { + "epoch": 0.77, + "grad_norm": 2.0193796026538946, + "learning_rate": 2.6333297561716696e-06, + "loss": 0.9352, + "step": 32713 + }, + { + "epoch": 0.77, + "grad_norm": 1.9149975014718452, + "learning_rate": 2.6328137667945463e-06, + "loss": 0.9625, + "step": 32714 + }, + { + "epoch": 0.77, + "grad_norm": 1.076856817938424, + "learning_rate": 2.6322978203115102e-06, + "loss": 0.8677, + "step": 32715 + }, + { + "epoch": 0.77, + "grad_norm": 1.9775278537338397, + "learning_rate": 2.6317819167255676e-06, + "loss": 0.9742, + "step": 32716 + }, + { + "epoch": 0.77, + "grad_norm": 1.9088362599541713, + "learning_rate": 2.6312660560397273e-06, + "loss": 1.0135, + "step": 32717 + }, + { + "epoch": 0.77, + "grad_norm": 2.1119728486873623, + "learning_rate": 2.6307502382569894e-06, + "loss": 0.9111, + "step": 32718 + }, + { + "epoch": 0.77, + "grad_norm": 1.7976030858443608, + "learning_rate": 2.6302344633803534e-06, + "loss": 1.0193, + "step": 32719 + }, + { + "epoch": 0.77, + "grad_norm": 1.9099069296230102, + "learning_rate": 2.6297187314128247e-06, + "loss": 0.883, + "step": 32720 + }, + { + "epoch": 0.77, + "grad_norm": 1.9502936137514795, + "learning_rate": 2.629203042357411e-06, + "loss": 0.9243, + "step": 32721 + }, + { + "epoch": 0.77, + "grad_norm": 1.9858959753194791, + "learning_rate": 2.62868739621711e-06, + "loss": 0.9188, + "step": 32722 + }, + { + "epoch": 0.77, + "grad_norm": 1.83930897913451, + "learning_rate": 2.6281717929949226e-06, + "loss": 0.9034, + "step": 32723 + }, + { + "epoch": 0.77, + "grad_norm": 1.1315649079808072, + "learning_rate": 2.6276562326938516e-06, + "loss": 0.8735, + "step": 32724 + }, + { + "epoch": 0.77, + "grad_norm": 1.8595554219568144, + "learning_rate": 2.6271407153169036e-06, + "loss": 0.9592, + "step": 32725 + }, + { + "epoch": 0.77, + "grad_norm": 2.102477703536868, + "learning_rate": 2.6266252408670776e-06, + "loss": 1.1869, + "step": 32726 + }, + { + "epoch": 0.77, + "grad_norm": 1.9047342808595173, + "learning_rate": 2.626109809347369e-06, + "loss": 0.9891, + "step": 32727 + }, + { + "epoch": 0.77, + "grad_norm": 1.755324699467869, + "learning_rate": 2.6255944207607885e-06, + "loss": 0.8532, + "step": 32728 + }, + { + "epoch": 0.77, + "grad_norm": 1.9675889603751167, + "learning_rate": 2.6250790751103282e-06, + "loss": 0.9713, + "step": 32729 + }, + { + "epoch": 0.77, + "grad_norm": 1.7857670422513772, + "learning_rate": 2.6245637723989946e-06, + "loss": 0.8889, + "step": 32730 + }, + { + "epoch": 0.77, + "grad_norm": 1.129486592528107, + "learning_rate": 2.6240485126297834e-06, + "loss": 0.9407, + "step": 32731 + }, + { + "epoch": 0.77, + "grad_norm": 2.0602659864523996, + "learning_rate": 2.6235332958057004e-06, + "loss": 0.9281, + "step": 32732 + }, + { + "epoch": 0.77, + "grad_norm": 2.225292718464362, + "learning_rate": 2.6230181219297378e-06, + "loss": 1.0185, + "step": 32733 + }, + { + "epoch": 0.77, + "grad_norm": 2.0969332722442213, + "learning_rate": 2.6225029910049037e-06, + "loss": 1.0847, + "step": 32734 + }, + { + "epoch": 0.77, + "grad_norm": 1.9830045658224198, + "learning_rate": 2.621987903034192e-06, + "loss": 0.9954, + "step": 32735 + }, + { + "epoch": 0.77, + "grad_norm": 1.9939216893829732, + "learning_rate": 2.6214728580205997e-06, + "loss": 1.0295, + "step": 32736 + }, + { + "epoch": 0.77, + "grad_norm": 2.862441416191295, + "learning_rate": 2.620957855967128e-06, + "loss": 0.8664, + "step": 32737 + }, + { + "epoch": 0.77, + "grad_norm": 1.8653522722571914, + "learning_rate": 2.620442896876779e-06, + "loss": 0.9918, + "step": 32738 + }, + { + "epoch": 0.77, + "grad_norm": 2.2297808339789573, + "learning_rate": 2.619927980752548e-06, + "loss": 0.9373, + "step": 32739 + }, + { + "epoch": 0.77, + "grad_norm": 2.4958031998097328, + "learning_rate": 2.6194131075974294e-06, + "loss": 0.9785, + "step": 32740 + }, + { + "epoch": 0.77, + "grad_norm": 1.1188104337563887, + "learning_rate": 2.6188982774144245e-06, + "loss": 0.9466, + "step": 32741 + }, + { + "epoch": 0.77, + "grad_norm": 1.973936953484927, + "learning_rate": 2.618383490206534e-06, + "loss": 0.9393, + "step": 32742 + }, + { + "epoch": 0.77, + "grad_norm": 1.8799727905092465, + "learning_rate": 2.6178687459767514e-06, + "loss": 1.0104, + "step": 32743 + }, + { + "epoch": 0.77, + "grad_norm": 2.0900611666305564, + "learning_rate": 2.617354044728072e-06, + "loss": 1.1186, + "step": 32744 + }, + { + "epoch": 0.77, + "grad_norm": 1.071995167766496, + "learning_rate": 2.6168393864634935e-06, + "loss": 0.8971, + "step": 32745 + }, + { + "epoch": 0.77, + "grad_norm": 2.177247335087554, + "learning_rate": 2.6163247711860173e-06, + "loss": 0.9956, + "step": 32746 + }, + { + "epoch": 0.77, + "grad_norm": 2.2630209225459605, + "learning_rate": 2.615810198898635e-06, + "loss": 0.9214, + "step": 32747 + }, + { + "epoch": 0.77, + "grad_norm": 2.3867142855981758, + "learning_rate": 2.615295669604342e-06, + "loss": 1.0426, + "step": 32748 + }, + { + "epoch": 0.77, + "grad_norm": 1.9381496533337739, + "learning_rate": 2.6147811833061376e-06, + "loss": 0.8768, + "step": 32749 + }, + { + "epoch": 0.77, + "grad_norm": 2.145697282334906, + "learning_rate": 2.614266740007012e-06, + "loss": 0.9465, + "step": 32750 + }, + { + "epoch": 0.77, + "grad_norm": 2.6485356367898505, + "learning_rate": 2.6137523397099672e-06, + "loss": 0.9465, + "step": 32751 + }, + { + "epoch": 0.77, + "grad_norm": 3.6055802766376037, + "learning_rate": 2.6132379824179923e-06, + "loss": 0.8841, + "step": 32752 + }, + { + "epoch": 0.77, + "grad_norm": 1.8517653251947594, + "learning_rate": 2.612723668134086e-06, + "loss": 0.9452, + "step": 32753 + }, + { + "epoch": 0.77, + "grad_norm": 2.1117203232511126, + "learning_rate": 2.6122093968612393e-06, + "loss": 0.9419, + "step": 32754 + }, + { + "epoch": 0.77, + "grad_norm": 1.9629639407837376, + "learning_rate": 2.611695168602448e-06, + "loss": 1.0487, + "step": 32755 + }, + { + "epoch": 0.77, + "grad_norm": 1.8392371095161775, + "learning_rate": 2.6111809833607126e-06, + "loss": 1.095, + "step": 32756 + }, + { + "epoch": 0.77, + "grad_norm": 1.9935194896649027, + "learning_rate": 2.6106668411390156e-06, + "loss": 0.9977, + "step": 32757 + }, + { + "epoch": 0.77, + "grad_norm": 1.9987478428480314, + "learning_rate": 2.610152741940355e-06, + "loss": 0.9662, + "step": 32758 + }, + { + "epoch": 0.77, + "grad_norm": 2.4118328468796353, + "learning_rate": 2.609638685767725e-06, + "loss": 0.9951, + "step": 32759 + }, + { + "epoch": 0.77, + "grad_norm": 1.900147110150913, + "learning_rate": 2.609124672624124e-06, + "loss": 0.8928, + "step": 32760 + }, + { + "epoch": 0.77, + "grad_norm": 2.2763700028524414, + "learning_rate": 2.608610702512533e-06, + "loss": 0.9853, + "step": 32761 + }, + { + "epoch": 0.77, + "grad_norm": 1.157482058049579, + "learning_rate": 2.608096775435951e-06, + "loss": 0.917, + "step": 32762 + }, + { + "epoch": 0.77, + "grad_norm": 1.954856344795237, + "learning_rate": 2.607582891397372e-06, + "loss": 0.8645, + "step": 32763 + }, + { + "epoch": 0.77, + "grad_norm": 2.094373353332016, + "learning_rate": 2.6070690503997866e-06, + "loss": 0.8295, + "step": 32764 + }, + { + "epoch": 0.77, + "grad_norm": 2.2685228746641033, + "learning_rate": 2.6065552524461823e-06, + "loss": 1.0613, + "step": 32765 + }, + { + "epoch": 0.77, + "grad_norm": 2.2306437925262483, + "learning_rate": 2.6060414975395543e-06, + "loss": 0.7852, + "step": 32766 + }, + { + "epoch": 0.77, + "grad_norm": 1.9870419306292533, + "learning_rate": 2.6055277856828964e-06, + "loss": 0.8939, + "step": 32767 + }, + { + "epoch": 0.77, + "grad_norm": 1.9908634530850389, + "learning_rate": 2.605014116879193e-06, + "loss": 1.0935, + "step": 32768 + }, + { + "epoch": 0.77, + "grad_norm": 1.1607023761539133, + "learning_rate": 2.604500491131443e-06, + "loss": 0.9895, + "step": 32769 + }, + { + "epoch": 0.77, + "grad_norm": 2.1564917102818595, + "learning_rate": 2.603986908442632e-06, + "loss": 1.0383, + "step": 32770 + }, + { + "epoch": 0.77, + "grad_norm": 1.9353685168138568, + "learning_rate": 2.6034733688157475e-06, + "loss": 0.9234, + "step": 32771 + }, + { + "epoch": 0.77, + "grad_norm": 1.970706588977892, + "learning_rate": 2.602959872253783e-06, + "loss": 0.9314, + "step": 32772 + }, + { + "epoch": 0.77, + "grad_norm": 3.8170005178626587, + "learning_rate": 2.602446418759731e-06, + "loss": 0.9452, + "step": 32773 + }, + { + "epoch": 0.77, + "grad_norm": 2.17971068255853, + "learning_rate": 2.6019330083365778e-06, + "loss": 0.9858, + "step": 32774 + }, + { + "epoch": 0.77, + "grad_norm": 1.872809238457354, + "learning_rate": 2.6014196409873104e-06, + "loss": 0.9857, + "step": 32775 + }, + { + "epoch": 0.77, + "grad_norm": 1.9667562597270887, + "learning_rate": 2.6009063167149195e-06, + "loss": 0.9889, + "step": 32776 + }, + { + "epoch": 0.77, + "grad_norm": 2.2913309928622856, + "learning_rate": 2.600393035522398e-06, + "loss": 1.0578, + "step": 32777 + }, + { + "epoch": 0.77, + "grad_norm": 1.8419854038810541, + "learning_rate": 2.599879797412731e-06, + "loss": 0.858, + "step": 32778 + }, + { + "epoch": 0.77, + "grad_norm": 2.117805038766255, + "learning_rate": 2.599366602388903e-06, + "loss": 0.9149, + "step": 32779 + }, + { + "epoch": 0.77, + "grad_norm": 1.949833790221739, + "learning_rate": 2.5988534504539065e-06, + "loss": 1.0169, + "step": 32780 + }, + { + "epoch": 0.77, + "grad_norm": 2.0886034563841362, + "learning_rate": 2.598340341610731e-06, + "loss": 0.9431, + "step": 32781 + }, + { + "epoch": 0.77, + "grad_norm": 3.143445654622762, + "learning_rate": 2.597827275862361e-06, + "loss": 0.9009, + "step": 32782 + }, + { + "epoch": 0.77, + "grad_norm": 1.974923735058843, + "learning_rate": 2.597314253211781e-06, + "loss": 0.9789, + "step": 32783 + }, + { + "epoch": 0.77, + "grad_norm": 2.0435728276165954, + "learning_rate": 2.596801273661985e-06, + "loss": 1.0003, + "step": 32784 + }, + { + "epoch": 0.77, + "grad_norm": 2.3825847694745748, + "learning_rate": 2.596288337215952e-06, + "loss": 0.8729, + "step": 32785 + }, + { + "epoch": 0.77, + "grad_norm": 1.9420871835689344, + "learning_rate": 2.595775443876676e-06, + "loss": 0.8301, + "step": 32786 + }, + { + "epoch": 0.77, + "grad_norm": 2.5245034171583987, + "learning_rate": 2.5952625936471354e-06, + "loss": 0.91, + "step": 32787 + }, + { + "epoch": 0.77, + "grad_norm": 2.161840407144417, + "learning_rate": 2.594749786530324e-06, + "loss": 0.9617, + "step": 32788 + }, + { + "epoch": 0.77, + "grad_norm": 2.2259055789378195, + "learning_rate": 2.5942370225292214e-06, + "loss": 0.9321, + "step": 32789 + }, + { + "epoch": 0.77, + "grad_norm": 2.1172760396387855, + "learning_rate": 2.593724301646817e-06, + "loss": 0.9771, + "step": 32790 + }, + { + "epoch": 0.77, + "grad_norm": 1.9935346475690676, + "learning_rate": 2.593211623886096e-06, + "loss": 0.9471, + "step": 32791 + }, + { + "epoch": 0.77, + "grad_norm": 1.9516018287812895, + "learning_rate": 2.5926989892500376e-06, + "loss": 1.0018, + "step": 32792 + }, + { + "epoch": 0.77, + "grad_norm": 2.0001381580961826, + "learning_rate": 2.5921863977416316e-06, + "loss": 1.0397, + "step": 32793 + }, + { + "epoch": 0.77, + "grad_norm": 2.0819523737801306, + "learning_rate": 2.5916738493638638e-06, + "loss": 0.8906, + "step": 32794 + }, + { + "epoch": 0.77, + "grad_norm": 1.899511134871378, + "learning_rate": 2.591161344119717e-06, + "loss": 0.9546, + "step": 32795 + }, + { + "epoch": 0.77, + "grad_norm": 2.0282267358128996, + "learning_rate": 2.590648882012172e-06, + "loss": 0.8943, + "step": 32796 + }, + { + "epoch": 0.77, + "grad_norm": 2.027529918053171, + "learning_rate": 2.5901364630442136e-06, + "loss": 0.986, + "step": 32797 + }, + { + "epoch": 0.77, + "grad_norm": 2.0457621510800044, + "learning_rate": 2.5896240872188305e-06, + "loss": 0.935, + "step": 32798 + }, + { + "epoch": 0.77, + "grad_norm": 2.073724602533122, + "learning_rate": 2.589111754539002e-06, + "loss": 0.9779, + "step": 32799 + }, + { + "epoch": 0.77, + "grad_norm": 1.0650071759358908, + "learning_rate": 2.5885994650077086e-06, + "loss": 0.9132, + "step": 32800 + }, + { + "epoch": 0.77, + "grad_norm": 2.0417650954855153, + "learning_rate": 2.588087218627935e-06, + "loss": 1.0627, + "step": 32801 + }, + { + "epoch": 0.77, + "grad_norm": 2.098954820625065, + "learning_rate": 2.587575015402667e-06, + "loss": 0.977, + "step": 32802 + }, + { + "epoch": 0.77, + "grad_norm": 1.9418206593522191, + "learning_rate": 2.587062855334884e-06, + "loss": 1.111, + "step": 32803 + }, + { + "epoch": 0.77, + "grad_norm": 2.867685658736506, + "learning_rate": 2.5865507384275646e-06, + "loss": 1.0575, + "step": 32804 + }, + { + "epoch": 0.77, + "grad_norm": 1.1215883935931399, + "learning_rate": 2.586038664683698e-06, + "loss": 0.8795, + "step": 32805 + }, + { + "epoch": 0.77, + "grad_norm": 1.9252656627056646, + "learning_rate": 2.585526634106258e-06, + "loss": 0.9944, + "step": 32806 + }, + { + "epoch": 0.77, + "grad_norm": 2.029658135336791, + "learning_rate": 2.585014646698232e-06, + "loss": 1.0945, + "step": 32807 + }, + { + "epoch": 0.77, + "grad_norm": 2.0595684464131017, + "learning_rate": 2.584502702462596e-06, + "loss": 0.9774, + "step": 32808 + }, + { + "epoch": 0.77, + "grad_norm": 2.1420569543420074, + "learning_rate": 2.5839908014023353e-06, + "loss": 0.9357, + "step": 32809 + }, + { + "epoch": 0.77, + "grad_norm": 2.096940726823809, + "learning_rate": 2.5834789435204245e-06, + "loss": 1.0285, + "step": 32810 + }, + { + "epoch": 0.77, + "grad_norm": 1.985217907500054, + "learning_rate": 2.582967128819851e-06, + "loss": 0.9371, + "step": 32811 + }, + { + "epoch": 0.77, + "grad_norm": 2.118249631387075, + "learning_rate": 2.5824553573035914e-06, + "loss": 0.9444, + "step": 32812 + }, + { + "epoch": 0.77, + "grad_norm": 1.9212168927495272, + "learning_rate": 2.581943628974621e-06, + "loss": 0.9296, + "step": 32813 + }, + { + "epoch": 0.77, + "grad_norm": 2.0209792235058113, + "learning_rate": 2.581431943835925e-06, + "loss": 1.0172, + "step": 32814 + }, + { + "epoch": 0.77, + "grad_norm": 1.8974739964860865, + "learning_rate": 2.5809203018904828e-06, + "loss": 0.9118, + "step": 32815 + }, + { + "epoch": 0.77, + "grad_norm": 2.1682428586212, + "learning_rate": 2.5804087031412705e-06, + "loss": 0.9584, + "step": 32816 + }, + { + "epoch": 0.77, + "grad_norm": 1.1369415361677435, + "learning_rate": 2.579897147591266e-06, + "loss": 0.9542, + "step": 32817 + }, + { + "epoch": 0.77, + "grad_norm": 2.113066369729551, + "learning_rate": 2.57938563524345e-06, + "loss": 1.0386, + "step": 32818 + }, + { + "epoch": 0.77, + "grad_norm": 1.759088123782499, + "learning_rate": 2.578874166100802e-06, + "loss": 1.1005, + "step": 32819 + }, + { + "epoch": 0.77, + "grad_norm": 1.9058847818290807, + "learning_rate": 2.578362740166298e-06, + "loss": 0.8886, + "step": 32820 + }, + { + "epoch": 0.77, + "grad_norm": 1.99747324133881, + "learning_rate": 2.5778513574429143e-06, + "loss": 1.0007, + "step": 32821 + }, + { + "epoch": 0.77, + "grad_norm": 1.9432721854064663, + "learning_rate": 2.577340017933628e-06, + "loss": 1.0067, + "step": 32822 + }, + { + "epoch": 0.77, + "grad_norm": 1.954226899571726, + "learning_rate": 2.5768287216414224e-06, + "loss": 1.0632, + "step": 32823 + }, + { + "epoch": 0.77, + "grad_norm": 2.1066058309494444, + "learning_rate": 2.5763174685692693e-06, + "loss": 1.0818, + "step": 32824 + }, + { + "epoch": 0.77, + "grad_norm": 1.9847998169046814, + "learning_rate": 2.5758062587201436e-06, + "loss": 0.8333, + "step": 32825 + }, + { + "epoch": 0.77, + "grad_norm": 1.942286406891642, + "learning_rate": 2.5752950920970278e-06, + "loss": 1.0425, + "step": 32826 + }, + { + "epoch": 0.77, + "grad_norm": 1.8205430140425918, + "learning_rate": 2.574783968702891e-06, + "loss": 0.8896, + "step": 32827 + }, + { + "epoch": 0.77, + "grad_norm": 2.1393598327491214, + "learning_rate": 2.5742728885407153e-06, + "loss": 1.1326, + "step": 32828 + }, + { + "epoch": 0.77, + "grad_norm": 1.9724952411331687, + "learning_rate": 2.573761851613471e-06, + "loss": 1.0621, + "step": 32829 + }, + { + "epoch": 0.77, + "grad_norm": 1.124118111374231, + "learning_rate": 2.57325085792414e-06, + "loss": 0.9223, + "step": 32830 + }, + { + "epoch": 0.77, + "grad_norm": 2.0787348467518045, + "learning_rate": 2.57273990747569e-06, + "loss": 0.9918, + "step": 32831 + }, + { + "epoch": 0.77, + "grad_norm": 1.8812658632241712, + "learning_rate": 2.5722290002711025e-06, + "loss": 0.9328, + "step": 32832 + }, + { + "epoch": 0.77, + "grad_norm": 2.5108323536361628, + "learning_rate": 2.57171813631335e-06, + "loss": 1.0285, + "step": 32833 + }, + { + "epoch": 0.77, + "grad_norm": 2.1775639520769743, + "learning_rate": 2.571207315605404e-06, + "loss": 1.0804, + "step": 32834 + }, + { + "epoch": 0.77, + "grad_norm": 1.0448996734265934, + "learning_rate": 2.5706965381502392e-06, + "loss": 0.9422, + "step": 32835 + }, + { + "epoch": 0.77, + "grad_norm": 2.0167907426367013, + "learning_rate": 2.5701858039508354e-06, + "loss": 1.0445, + "step": 32836 + }, + { + "epoch": 0.77, + "grad_norm": 2.005119788652767, + "learning_rate": 2.5696751130101614e-06, + "loss": 1.0075, + "step": 32837 + }, + { + "epoch": 0.77, + "grad_norm": 1.8275169781252207, + "learning_rate": 2.5691644653311887e-06, + "loss": 0.9627, + "step": 32838 + }, + { + "epoch": 0.77, + "grad_norm": 2.217858778161114, + "learning_rate": 2.5686538609168932e-06, + "loss": 0.8642, + "step": 32839 + }, + { + "epoch": 0.77, + "grad_norm": 2.3740230214688602, + "learning_rate": 2.568143299770247e-06, + "loss": 1.0016, + "step": 32840 + }, + { + "epoch": 0.77, + "grad_norm": 1.8706968870032075, + "learning_rate": 2.5676327818942305e-06, + "loss": 0.8554, + "step": 32841 + }, + { + "epoch": 0.77, + "grad_norm": 1.9684917865667901, + "learning_rate": 2.567122307291803e-06, + "loss": 0.8483, + "step": 32842 + }, + { + "epoch": 0.77, + "grad_norm": 2.0097063386465615, + "learning_rate": 2.566611875965941e-06, + "loss": 0.9668, + "step": 32843 + }, + { + "epoch": 0.77, + "grad_norm": 4.045770044957472, + "learning_rate": 2.5661014879196223e-06, + "loss": 1.0457, + "step": 32844 + }, + { + "epoch": 0.77, + "grad_norm": 1.9225299558019446, + "learning_rate": 2.5655911431558145e-06, + "loss": 0.9334, + "step": 32845 + }, + { + "epoch": 0.77, + "grad_norm": 1.9266744196584864, + "learning_rate": 2.565080841677484e-06, + "loss": 0.8578, + "step": 32846 + }, + { + "epoch": 0.77, + "grad_norm": 1.8907357700169558, + "learning_rate": 2.564570583487609e-06, + "loss": 0.9915, + "step": 32847 + }, + { + "epoch": 0.77, + "grad_norm": 2.0302309418965336, + "learning_rate": 2.5640603685891597e-06, + "loss": 1.0594, + "step": 32848 + }, + { + "epoch": 0.77, + "grad_norm": 2.214234050904368, + "learning_rate": 2.563550196985105e-06, + "loss": 1.0441, + "step": 32849 + }, + { + "epoch": 0.77, + "grad_norm": 2.061777636230761, + "learning_rate": 2.5630400686784128e-06, + "loss": 1.0084, + "step": 32850 + }, + { + "epoch": 0.77, + "grad_norm": 1.1974996414689862, + "learning_rate": 2.5625299836720597e-06, + "loss": 0.9973, + "step": 32851 + }, + { + "epoch": 0.77, + "grad_norm": 1.0905370091455506, + "learning_rate": 2.5620199419690074e-06, + "loss": 0.9315, + "step": 32852 + }, + { + "epoch": 0.77, + "grad_norm": 1.8304969458859766, + "learning_rate": 2.56150994357223e-06, + "loss": 0.9036, + "step": 32853 + }, + { + "epoch": 0.77, + "grad_norm": 1.879001070542558, + "learning_rate": 2.5609999884847003e-06, + "loss": 0.8439, + "step": 32854 + }, + { + "epoch": 0.77, + "grad_norm": 1.936307585839914, + "learning_rate": 2.560490076709383e-06, + "loss": 1.0386, + "step": 32855 + }, + { + "epoch": 0.77, + "grad_norm": 5.999369120779566, + "learning_rate": 2.559980208249245e-06, + "loss": 1.08, + "step": 32856 + }, + { + "epoch": 0.77, + "grad_norm": 1.9967641936260168, + "learning_rate": 2.5594703831072586e-06, + "loss": 0.9317, + "step": 32857 + }, + { + "epoch": 0.77, + "grad_norm": 2.6837314495076634, + "learning_rate": 2.5589606012863968e-06, + "loss": 0.9773, + "step": 32858 + }, + { + "epoch": 0.77, + "grad_norm": 2.141926551070432, + "learning_rate": 2.5584508627896155e-06, + "loss": 0.8623, + "step": 32859 + }, + { + "epoch": 0.77, + "grad_norm": 1.989181328292398, + "learning_rate": 2.5579411676198897e-06, + "loss": 0.9578, + "step": 32860 + }, + { + "epoch": 0.77, + "grad_norm": 2.7945106590038686, + "learning_rate": 2.5574315157801867e-06, + "loss": 1.0303, + "step": 32861 + }, + { + "epoch": 0.77, + "grad_norm": 1.0641354604151776, + "learning_rate": 2.55692190727348e-06, + "loss": 0.8997, + "step": 32862 + }, + { + "epoch": 0.77, + "grad_norm": 1.9489834953821357, + "learning_rate": 2.556412342102723e-06, + "loss": 0.9471, + "step": 32863 + }, + { + "epoch": 0.77, + "grad_norm": 1.9518746640387632, + "learning_rate": 2.5559028202708913e-06, + "loss": 1.1294, + "step": 32864 + }, + { + "epoch": 0.77, + "grad_norm": 2.7833295659991713, + "learning_rate": 2.555393341780953e-06, + "loss": 1.0336, + "step": 32865 + }, + { + "epoch": 0.77, + "grad_norm": 2.0404445072730235, + "learning_rate": 2.554883906635868e-06, + "loss": 0.9971, + "step": 32866 + }, + { + "epoch": 0.77, + "grad_norm": 2.0872787005073166, + "learning_rate": 2.554374514838609e-06, + "loss": 0.9475, + "step": 32867 + }, + { + "epoch": 0.77, + "grad_norm": 2.0710662178312598, + "learning_rate": 2.5538651663921367e-06, + "loss": 0.9376, + "step": 32868 + }, + { + "epoch": 0.77, + "grad_norm": 1.878287117634404, + "learning_rate": 2.5533558612994213e-06, + "loss": 1.0629, + "step": 32869 + }, + { + "epoch": 0.77, + "grad_norm": 1.8123661207111263, + "learning_rate": 2.5528465995634223e-06, + "loss": 0.9456, + "step": 32870 + }, + { + "epoch": 0.77, + "grad_norm": 2.2935913740781233, + "learning_rate": 2.552337381187111e-06, + "loss": 0.8428, + "step": 32871 + }, + { + "epoch": 0.77, + "grad_norm": 2.0000429999233824, + "learning_rate": 2.5518282061734502e-06, + "loss": 0.9578, + "step": 32872 + }, + { + "epoch": 0.77, + "grad_norm": 2.0152524311081605, + "learning_rate": 2.5513190745254e-06, + "loss": 0.9161, + "step": 32873 + }, + { + "epoch": 0.77, + "grad_norm": 2.288074345117689, + "learning_rate": 2.550809986245929e-06, + "loss": 1.0487, + "step": 32874 + }, + { + "epoch": 0.77, + "grad_norm": 2.023799564500672, + "learning_rate": 2.5503009413380033e-06, + "loss": 1.0191, + "step": 32875 + }, + { + "epoch": 0.77, + "grad_norm": 1.8221708460392094, + "learning_rate": 2.5497919398045833e-06, + "loss": 0.9416, + "step": 32876 + }, + { + "epoch": 0.77, + "grad_norm": 2.2369921493856673, + "learning_rate": 2.549282981648631e-06, + "loss": 0.9963, + "step": 32877 + }, + { + "epoch": 0.77, + "grad_norm": 2.2602338280098127, + "learning_rate": 2.548774066873112e-06, + "loss": 1.0026, + "step": 32878 + }, + { + "epoch": 0.77, + "grad_norm": 1.9094748142070064, + "learning_rate": 2.548265195480992e-06, + "loss": 0.9424, + "step": 32879 + }, + { + "epoch": 0.77, + "grad_norm": 1.989352568168355, + "learning_rate": 2.5477563674752327e-06, + "loss": 0.9773, + "step": 32880 + }, + { + "epoch": 0.77, + "grad_norm": 1.9036361965316253, + "learning_rate": 2.5472475828587906e-06, + "loss": 1.1436, + "step": 32881 + }, + { + "epoch": 0.77, + "grad_norm": 1.93170451277003, + "learning_rate": 2.546738841634633e-06, + "loss": 1.0212, + "step": 32882 + }, + { + "epoch": 0.77, + "grad_norm": 1.0664001164897483, + "learning_rate": 2.5462301438057245e-06, + "loss": 1.0363, + "step": 32883 + }, + { + "epoch": 0.77, + "grad_norm": 2.13453229294743, + "learning_rate": 2.545721489375024e-06, + "loss": 0.9187, + "step": 32884 + }, + { + "epoch": 0.77, + "grad_norm": 1.8953863832933693, + "learning_rate": 2.54521287834549e-06, + "loss": 0.8767, + "step": 32885 + }, + { + "epoch": 0.77, + "grad_norm": 2.050522847471749, + "learning_rate": 2.5447043107200895e-06, + "loss": 1.0267, + "step": 32886 + }, + { + "epoch": 0.77, + "grad_norm": 2.1619222915633065, + "learning_rate": 2.5441957865017784e-06, + "loss": 0.9173, + "step": 32887 + }, + { + "epoch": 0.77, + "grad_norm": 1.801670753629381, + "learning_rate": 2.543687305693523e-06, + "loss": 0.9691, + "step": 32888 + }, + { + "epoch": 0.77, + "grad_norm": 1.90769262559644, + "learning_rate": 2.543178868298277e-06, + "loss": 1.0239, + "step": 32889 + }, + { + "epoch": 0.77, + "grad_norm": 2.200365823961153, + "learning_rate": 2.542670474319008e-06, + "loss": 0.9676, + "step": 32890 + }, + { + "epoch": 0.77, + "grad_norm": 1.860226023498241, + "learning_rate": 2.542162123758669e-06, + "loss": 0.901, + "step": 32891 + }, + { + "epoch": 0.77, + "grad_norm": 3.5475822760502176, + "learning_rate": 2.5416538166202267e-06, + "loss": 0.9403, + "step": 32892 + }, + { + "epoch": 0.77, + "grad_norm": 1.9883673755582516, + "learning_rate": 2.5411455529066363e-06, + "loss": 0.9736, + "step": 32893 + }, + { + "epoch": 0.77, + "grad_norm": 2.0140622375268373, + "learning_rate": 2.5406373326208554e-06, + "loss": 0.9869, + "step": 32894 + }, + { + "epoch": 0.77, + "grad_norm": 2.0187669508854578, + "learning_rate": 2.5401291557658457e-06, + "loss": 1.078, + "step": 32895 + }, + { + "epoch": 0.78, + "grad_norm": 2.212402058973696, + "learning_rate": 2.5396210223445684e-06, + "loss": 0.9906, + "step": 32896 + }, + { + "epoch": 0.78, + "grad_norm": 2.0752417593975294, + "learning_rate": 2.5391129323599796e-06, + "loss": 1.007, + "step": 32897 + }, + { + "epoch": 0.78, + "grad_norm": 2.089605715934708, + "learning_rate": 2.5386048858150346e-06, + "loss": 1.0734, + "step": 32898 + }, + { + "epoch": 0.78, + "grad_norm": 1.9752007272465624, + "learning_rate": 2.5380968827126937e-06, + "loss": 0.9634, + "step": 32899 + }, + { + "epoch": 0.78, + "grad_norm": 2.2377893041863297, + "learning_rate": 2.5375889230559192e-06, + "loss": 0.9842, + "step": 32900 + }, + { + "epoch": 0.78, + "grad_norm": 2.062489219659028, + "learning_rate": 2.5370810068476637e-06, + "loss": 0.9661, + "step": 32901 + }, + { + "epoch": 0.78, + "grad_norm": 1.912359454804296, + "learning_rate": 2.536573134090883e-06, + "loss": 0.9553, + "step": 32902 + }, + { + "epoch": 0.78, + "grad_norm": 2.6314311408034543, + "learning_rate": 2.5360653047885352e-06, + "loss": 0.956, + "step": 32903 + }, + { + "epoch": 0.78, + "grad_norm": 1.8913864759117036, + "learning_rate": 2.535557518943582e-06, + "loss": 0.9872, + "step": 32904 + }, + { + "epoch": 0.78, + "grad_norm": 2.060823297999083, + "learning_rate": 2.5350497765589765e-06, + "loss": 0.9192, + "step": 32905 + }, + { + "epoch": 0.78, + "grad_norm": 1.8554959391810355, + "learning_rate": 2.5345420776376706e-06, + "loss": 1.0273, + "step": 32906 + }, + { + "epoch": 0.78, + "grad_norm": 2.054477727401955, + "learning_rate": 2.5340344221826287e-06, + "loss": 0.9459, + "step": 32907 + }, + { + "epoch": 0.78, + "grad_norm": 1.1286894536898817, + "learning_rate": 2.533526810196798e-06, + "loss": 0.9205, + "step": 32908 + }, + { + "epoch": 0.78, + "grad_norm": 1.1056604475066039, + "learning_rate": 2.5330192416831413e-06, + "loss": 0.9542, + "step": 32909 + }, + { + "epoch": 0.78, + "grad_norm": 1.9030515298477555, + "learning_rate": 2.5325117166446065e-06, + "loss": 0.8979, + "step": 32910 + }, + { + "epoch": 0.78, + "grad_norm": 2.0467520981301948, + "learning_rate": 2.5320042350841577e-06, + "loss": 1.0662, + "step": 32911 + }, + { + "epoch": 0.78, + "grad_norm": 1.9312691487435305, + "learning_rate": 2.53149679700474e-06, + "loss": 0.9847, + "step": 32912 + }, + { + "epoch": 0.78, + "grad_norm": 1.7034979595623332, + "learning_rate": 2.5309894024093163e-06, + "loss": 0.961, + "step": 32913 + }, + { + "epoch": 0.78, + "grad_norm": 2.016430986593366, + "learning_rate": 2.5304820513008364e-06, + "loss": 1.0205, + "step": 32914 + }, + { + "epoch": 0.78, + "grad_norm": 2.009117490625359, + "learning_rate": 2.529974743682252e-06, + "loss": 1.0437, + "step": 32915 + }, + { + "epoch": 0.78, + "grad_norm": 2.3195340394696, + "learning_rate": 2.52946747955652e-06, + "loss": 0.8027, + "step": 32916 + }, + { + "epoch": 0.78, + "grad_norm": 1.9556578145112602, + "learning_rate": 2.528960258926596e-06, + "loss": 0.9529, + "step": 32917 + }, + { + "epoch": 0.78, + "grad_norm": 1.7516247815486483, + "learning_rate": 2.52845308179543e-06, + "loss": 0.9655, + "step": 32918 + }, + { + "epoch": 0.78, + "grad_norm": 2.1684234978819488, + "learning_rate": 2.527945948165974e-06, + "loss": 0.9372, + "step": 32919 + }, + { + "epoch": 0.78, + "grad_norm": 2.1795808926698537, + "learning_rate": 2.5274388580411813e-06, + "loss": 1.0374, + "step": 32920 + }, + { + "epoch": 0.78, + "grad_norm": 2.1549158695259174, + "learning_rate": 2.5269318114240092e-06, + "loss": 0.9838, + "step": 32921 + }, + { + "epoch": 0.78, + "grad_norm": 1.1133478679091438, + "learning_rate": 2.5264248083174047e-06, + "loss": 0.8885, + "step": 32922 + }, + { + "epoch": 0.78, + "grad_norm": 2.2126903532743545, + "learning_rate": 2.5259178487243197e-06, + "loss": 0.9776, + "step": 32923 + }, + { + "epoch": 0.78, + "grad_norm": 1.9507452416734565, + "learning_rate": 2.525410932647706e-06, + "loss": 0.9872, + "step": 32924 + }, + { + "epoch": 0.78, + "grad_norm": 2.1137356550525044, + "learning_rate": 2.5249040600905195e-06, + "loss": 0.912, + "step": 32925 + }, + { + "epoch": 0.78, + "grad_norm": 1.9966037127044562, + "learning_rate": 2.5243972310557087e-06, + "loss": 1.0114, + "step": 32926 + }, + { + "epoch": 0.78, + "grad_norm": 1.9321580884155019, + "learning_rate": 2.523890445546221e-06, + "loss": 0.9354, + "step": 32927 + }, + { + "epoch": 0.78, + "grad_norm": 1.8180466180892352, + "learning_rate": 2.523383703565012e-06, + "loss": 1.038, + "step": 32928 + }, + { + "epoch": 0.78, + "grad_norm": 2.015821536041323, + "learning_rate": 2.5228770051150274e-06, + "loss": 1.0027, + "step": 32929 + }, + { + "epoch": 0.78, + "grad_norm": 1.061128169431732, + "learning_rate": 2.5223703501992234e-06, + "loss": 0.8838, + "step": 32930 + }, + { + "epoch": 0.78, + "grad_norm": 2.0477478813334895, + "learning_rate": 2.521863738820545e-06, + "loss": 0.9788, + "step": 32931 + }, + { + "epoch": 0.78, + "grad_norm": 1.943969050696239, + "learning_rate": 2.521357170981944e-06, + "loss": 0.9164, + "step": 32932 + }, + { + "epoch": 0.78, + "grad_norm": 2.068998766132696, + "learning_rate": 2.520850646686368e-06, + "loss": 0.9794, + "step": 32933 + }, + { + "epoch": 0.78, + "grad_norm": 1.8669201146815428, + "learning_rate": 2.52034416593677e-06, + "loss": 0.8644, + "step": 32934 + }, + { + "epoch": 0.78, + "grad_norm": 1.9955548711008775, + "learning_rate": 2.5198377287360954e-06, + "loss": 1.0814, + "step": 32935 + }, + { + "epoch": 0.78, + "grad_norm": 2.058286132326904, + "learning_rate": 2.519331335087292e-06, + "loss": 0.9778, + "step": 32936 + }, + { + "epoch": 0.78, + "grad_norm": 1.2081578972127667, + "learning_rate": 2.5188249849933087e-06, + "loss": 0.9782, + "step": 32937 + }, + { + "epoch": 0.78, + "grad_norm": 1.9728686200092682, + "learning_rate": 2.5183186784570957e-06, + "loss": 0.865, + "step": 32938 + }, + { + "epoch": 0.78, + "grad_norm": 2.1160431133216813, + "learning_rate": 2.5178124154816052e-06, + "loss": 0.9164, + "step": 32939 + }, + { + "epoch": 0.78, + "grad_norm": 2.0143915095540055, + "learning_rate": 2.5173061960697744e-06, + "loss": 0.9937, + "step": 32940 + }, + { + "epoch": 0.78, + "grad_norm": 2.1842798436185897, + "learning_rate": 2.516800020224556e-06, + "loss": 0.9393, + "step": 32941 + }, + { + "epoch": 0.78, + "grad_norm": 1.8744326796168085, + "learning_rate": 2.5162938879488996e-06, + "loss": 0.9477, + "step": 32942 + }, + { + "epoch": 0.78, + "grad_norm": 1.1215120618814665, + "learning_rate": 2.5157877992457493e-06, + "loss": 0.9287, + "step": 32943 + }, + { + "epoch": 0.78, + "grad_norm": 2.30803170695586, + "learning_rate": 2.5152817541180484e-06, + "loss": 0.8198, + "step": 32944 + }, + { + "epoch": 0.78, + "grad_norm": 1.9487882586808827, + "learning_rate": 2.5147757525687477e-06, + "loss": 1.0773, + "step": 32945 + }, + { + "epoch": 0.78, + "grad_norm": 2.0055333592703657, + "learning_rate": 2.5142697946007955e-06, + "loss": 0.9783, + "step": 32946 + }, + { + "epoch": 0.78, + "grad_norm": 2.617138381041593, + "learning_rate": 2.5137638802171337e-06, + "loss": 0.954, + "step": 32947 + }, + { + "epoch": 0.78, + "grad_norm": 1.7689364747642389, + "learning_rate": 2.513258009420707e-06, + "loss": 0.9145, + "step": 32948 + }, + { + "epoch": 0.78, + "grad_norm": 2.5089668243059537, + "learning_rate": 2.512752182214464e-06, + "loss": 1.117, + "step": 32949 + }, + { + "epoch": 0.78, + "grad_norm": 1.9560841582783035, + "learning_rate": 2.5122463986013458e-06, + "loss": 0.9828, + "step": 32950 + }, + { + "epoch": 0.78, + "grad_norm": 1.117065612272711, + "learning_rate": 2.5117406585842998e-06, + "loss": 0.9603, + "step": 32951 + }, + { + "epoch": 0.78, + "grad_norm": 2.0375169547906262, + "learning_rate": 2.5112349621662736e-06, + "loss": 1.0307, + "step": 32952 + }, + { + "epoch": 0.78, + "grad_norm": 3.1014124046660325, + "learning_rate": 2.510729309350208e-06, + "loss": 0.992, + "step": 32953 + }, + { + "epoch": 0.78, + "grad_norm": 2.178948119870998, + "learning_rate": 2.510223700139044e-06, + "loss": 1.0244, + "step": 32954 + }, + { + "epoch": 0.78, + "grad_norm": 1.9486086589160907, + "learning_rate": 2.509718134535729e-06, + "loss": 0.9151, + "step": 32955 + }, + { + "epoch": 0.78, + "grad_norm": 1.8487607656733507, + "learning_rate": 2.5092126125432137e-06, + "loss": 1.0158, + "step": 32956 + }, + { + "epoch": 0.78, + "grad_norm": 1.732278777691036, + "learning_rate": 2.5087071341644265e-06, + "loss": 1.0128, + "step": 32957 + }, + { + "epoch": 0.78, + "grad_norm": 1.9798987630455804, + "learning_rate": 2.5082016994023185e-06, + "loss": 1.0175, + "step": 32958 + }, + { + "epoch": 0.78, + "grad_norm": 2.0406211891693706, + "learning_rate": 2.5076963082598327e-06, + "loss": 1.0561, + "step": 32959 + }, + { + "epoch": 0.78, + "grad_norm": 1.881269574122918, + "learning_rate": 2.507190960739917e-06, + "loss": 1.0428, + "step": 32960 + }, + { + "epoch": 0.78, + "grad_norm": 2.1080293850871583, + "learning_rate": 2.5066856568455012e-06, + "loss": 1.0087, + "step": 32961 + }, + { + "epoch": 0.78, + "grad_norm": 2.194526512675436, + "learning_rate": 2.506180396579534e-06, + "loss": 0.9717, + "step": 32962 + }, + { + "epoch": 0.78, + "grad_norm": 1.887257394821854, + "learning_rate": 2.505675179944956e-06, + "loss": 0.9144, + "step": 32963 + }, + { + "epoch": 0.78, + "grad_norm": 2.0009232830924426, + "learning_rate": 2.5051700069447136e-06, + "loss": 0.9963, + "step": 32964 + }, + { + "epoch": 0.78, + "grad_norm": 1.8321849392771454, + "learning_rate": 2.504664877581745e-06, + "loss": 0.9332, + "step": 32965 + }, + { + "epoch": 0.78, + "grad_norm": 2.024154236281499, + "learning_rate": 2.5041597918589867e-06, + "loss": 0.9905, + "step": 32966 + }, + { + "epoch": 0.78, + "grad_norm": 1.9202514103200135, + "learning_rate": 2.5036547497793863e-06, + "loss": 1.1057, + "step": 32967 + }, + { + "epoch": 0.78, + "grad_norm": 1.7660083097533354, + "learning_rate": 2.503149751345878e-06, + "loss": 0.8729, + "step": 32968 + }, + { + "epoch": 0.78, + "grad_norm": 1.9968667380095988, + "learning_rate": 2.5026447965614086e-06, + "loss": 0.9112, + "step": 32969 + }, + { + "epoch": 0.78, + "grad_norm": 2.2930067141889587, + "learning_rate": 2.5021398854289124e-06, + "loss": 0.8886, + "step": 32970 + }, + { + "epoch": 0.78, + "grad_norm": 2.36264899590327, + "learning_rate": 2.501635017951334e-06, + "loss": 0.883, + "step": 32971 + }, + { + "epoch": 0.78, + "grad_norm": 2.527077015457078, + "learning_rate": 2.501130194131608e-06, + "loss": 0.9935, + "step": 32972 + }, + { + "epoch": 0.78, + "grad_norm": 1.9989606364541348, + "learning_rate": 2.500625413972678e-06, + "loss": 0.8651, + "step": 32973 + }, + { + "epoch": 0.78, + "grad_norm": 1.9439625334793749, + "learning_rate": 2.5001206774774813e-06, + "loss": 0.9468, + "step": 32974 + }, + { + "epoch": 0.78, + "grad_norm": 1.8857190248261029, + "learning_rate": 2.4996159846489544e-06, + "loss": 1.0321, + "step": 32975 + }, + { + "epoch": 0.78, + "grad_norm": 1.9745089946893875, + "learning_rate": 2.4991113354900374e-06, + "loss": 0.8777, + "step": 32976 + }, + { + "epoch": 0.78, + "grad_norm": 1.9438791521611525, + "learning_rate": 2.4986067300036708e-06, + "loss": 0.9817, + "step": 32977 + }, + { + "epoch": 0.78, + "grad_norm": 2.0818738494308784, + "learning_rate": 2.498102168192792e-06, + "loss": 1.0209, + "step": 32978 + }, + { + "epoch": 0.78, + "grad_norm": 1.8986496809233715, + "learning_rate": 2.4975976500603348e-06, + "loss": 1.0411, + "step": 32979 + }, + { + "epoch": 0.78, + "grad_norm": 2.2329126229138057, + "learning_rate": 2.4970931756092387e-06, + "loss": 1.0164, + "step": 32980 + }, + { + "epoch": 0.78, + "grad_norm": 2.6099862883318, + "learning_rate": 2.4965887448424443e-06, + "loss": 0.9248, + "step": 32981 + }, + { + "epoch": 0.78, + "grad_norm": 2.457161704047265, + "learning_rate": 2.4960843577628858e-06, + "loss": 1.0643, + "step": 32982 + }, + { + "epoch": 0.78, + "grad_norm": 1.098580065414673, + "learning_rate": 2.4955800143734966e-06, + "loss": 0.9227, + "step": 32983 + }, + { + "epoch": 0.78, + "grad_norm": 3.2545056310521514, + "learning_rate": 2.4950757146772175e-06, + "loss": 1.1169, + "step": 32984 + }, + { + "epoch": 0.78, + "grad_norm": 2.1393800565632057, + "learning_rate": 2.4945714586769854e-06, + "loss": 1.003, + "step": 32985 + }, + { + "epoch": 0.78, + "grad_norm": 2.9436461871198283, + "learning_rate": 2.4940672463757353e-06, + "loss": 0.9093, + "step": 32986 + }, + { + "epoch": 0.78, + "grad_norm": 2.035552612664729, + "learning_rate": 2.493563077776399e-06, + "loss": 0.9819, + "step": 32987 + }, + { + "epoch": 0.78, + "grad_norm": 1.8216978056140498, + "learning_rate": 2.4930589528819172e-06, + "loss": 1.0494, + "step": 32988 + }, + { + "epoch": 0.78, + "grad_norm": 1.92751027828229, + "learning_rate": 2.492554871695221e-06, + "loss": 1.0539, + "step": 32989 + }, + { + "epoch": 0.78, + "grad_norm": 2.0557052237813798, + "learning_rate": 2.4920508342192485e-06, + "loss": 0.8882, + "step": 32990 + }, + { + "epoch": 0.78, + "grad_norm": 1.7986061331207355, + "learning_rate": 2.491546840456931e-06, + "loss": 0.9953, + "step": 32991 + }, + { + "epoch": 0.78, + "grad_norm": 2.1892149956949183, + "learning_rate": 2.491042890411208e-06, + "loss": 1.0065, + "step": 32992 + }, + { + "epoch": 0.78, + "grad_norm": 2.098169357928822, + "learning_rate": 2.490538984085008e-06, + "loss": 0.9468, + "step": 32993 + }, + { + "epoch": 0.78, + "grad_norm": 2.1999236306128247, + "learning_rate": 2.4900351214812703e-06, + "loss": 1.0771, + "step": 32994 + }, + { + "epoch": 0.78, + "grad_norm": 1.8131873194014758, + "learning_rate": 2.4895313026029243e-06, + "loss": 1.0025, + "step": 32995 + }, + { + "epoch": 0.78, + "grad_norm": 1.6912092558455596, + "learning_rate": 2.489027527452903e-06, + "loss": 0.9862, + "step": 32996 + }, + { + "epoch": 0.78, + "grad_norm": 2.126643679179657, + "learning_rate": 2.4885237960341413e-06, + "loss": 0.8739, + "step": 32997 + }, + { + "epoch": 0.78, + "grad_norm": 1.8686256869049245, + "learning_rate": 2.488020108349576e-06, + "loss": 1.0039, + "step": 32998 + }, + { + "epoch": 0.78, + "grad_norm": 2.7971781230573463, + "learning_rate": 2.4875164644021343e-06, + "loss": 0.9321, + "step": 32999 + }, + { + "epoch": 0.78, + "grad_norm": 2.0061764262820274, + "learning_rate": 2.487012864194748e-06, + "loss": 1.0071, + "step": 33000 + }, + { + "epoch": 0.78, + "grad_norm": 1.9424746279738712, + "learning_rate": 2.4865093077303514e-06, + "loss": 0.9487, + "step": 33001 + }, + { + "epoch": 0.78, + "grad_norm": 2.159465689870635, + "learning_rate": 2.48600579501188e-06, + "loss": 1.0352, + "step": 33002 + }, + { + "epoch": 0.78, + "grad_norm": 1.9876169539315898, + "learning_rate": 2.4855023260422605e-06, + "loss": 1.0464, + "step": 33003 + }, + { + "epoch": 0.78, + "grad_norm": 1.8954884882747207, + "learning_rate": 2.4849989008244224e-06, + "loss": 0.9897, + "step": 33004 + }, + { + "epoch": 0.78, + "grad_norm": 1.923987706459314, + "learning_rate": 2.4844955193613007e-06, + "loss": 0.9184, + "step": 33005 + }, + { + "epoch": 0.78, + "grad_norm": 1.9915940808632373, + "learning_rate": 2.483992181655829e-06, + "loss": 0.9074, + "step": 33006 + }, + { + "epoch": 0.78, + "grad_norm": 1.8690304190534606, + "learning_rate": 2.483488887710933e-06, + "loss": 0.9726, + "step": 33007 + }, + { + "epoch": 0.78, + "grad_norm": 1.8122061392813837, + "learning_rate": 2.4829856375295426e-06, + "loss": 0.995, + "step": 33008 + }, + { + "epoch": 0.78, + "grad_norm": 2.102861360226613, + "learning_rate": 2.482482431114591e-06, + "loss": 0.932, + "step": 33009 + }, + { + "epoch": 0.78, + "grad_norm": 2.2441471677025753, + "learning_rate": 2.4819792684690046e-06, + "loss": 0.8871, + "step": 33010 + }, + { + "epoch": 0.78, + "grad_norm": 2.0133979409387464, + "learning_rate": 2.4814761495957183e-06, + "loss": 0.9424, + "step": 33011 + }, + { + "epoch": 0.78, + "grad_norm": 2.5199991419210694, + "learning_rate": 2.4809730744976556e-06, + "loss": 1.0337, + "step": 33012 + }, + { + "epoch": 0.78, + "grad_norm": 1.8965704013350426, + "learning_rate": 2.480470043177751e-06, + "loss": 1.0181, + "step": 33013 + }, + { + "epoch": 0.78, + "grad_norm": 1.0604576160874084, + "learning_rate": 2.4799670556389267e-06, + "loss": 0.8845, + "step": 33014 + }, + { + "epoch": 0.78, + "grad_norm": 2.0552311246566166, + "learning_rate": 2.479464111884118e-06, + "loss": 0.9502, + "step": 33015 + }, + { + "epoch": 0.78, + "grad_norm": 1.969348537386223, + "learning_rate": 2.478961211916251e-06, + "loss": 0.9372, + "step": 33016 + }, + { + "epoch": 0.78, + "grad_norm": 1.9911355105590547, + "learning_rate": 2.4784583557382492e-06, + "loss": 0.94, + "step": 33017 + }, + { + "epoch": 0.78, + "grad_norm": 2.153129625635549, + "learning_rate": 2.4779555433530446e-06, + "loss": 1.1555, + "step": 33018 + }, + { + "epoch": 0.78, + "grad_norm": 2.18767881961072, + "learning_rate": 2.4774527747635667e-06, + "loss": 1.0448, + "step": 33019 + }, + { + "epoch": 0.78, + "grad_norm": 2.2026868540496207, + "learning_rate": 2.4769500499727417e-06, + "loss": 1.1557, + "step": 33020 + }, + { + "epoch": 0.78, + "grad_norm": 1.8997908958800382, + "learning_rate": 2.4764473689834912e-06, + "loss": 0.885, + "step": 33021 + }, + { + "epoch": 0.78, + "grad_norm": 1.9565489256948785, + "learning_rate": 2.475944731798746e-06, + "loss": 0.9027, + "step": 33022 + }, + { + "epoch": 0.78, + "grad_norm": 2.066493690868017, + "learning_rate": 2.4754421384214355e-06, + "loss": 1.0945, + "step": 33023 + }, + { + "epoch": 0.78, + "grad_norm": 1.9196274809146279, + "learning_rate": 2.4749395888544837e-06, + "loss": 1.0108, + "step": 33024 + }, + { + "epoch": 0.78, + "grad_norm": 1.7684938778683748, + "learning_rate": 2.4744370831008125e-06, + "loss": 0.9525, + "step": 33025 + }, + { + "epoch": 0.78, + "grad_norm": 2.332067986689, + "learning_rate": 2.473934621163352e-06, + "loss": 1.0179, + "step": 33026 + }, + { + "epoch": 0.78, + "grad_norm": 2.117580538994434, + "learning_rate": 2.4734322030450286e-06, + "loss": 1.0649, + "step": 33027 + }, + { + "epoch": 0.78, + "grad_norm": 1.8991320203904327, + "learning_rate": 2.472929828748767e-06, + "loss": 0.986, + "step": 33028 + }, + { + "epoch": 0.78, + "grad_norm": 2.474449179909249, + "learning_rate": 2.4724274982774866e-06, + "loss": 1.0291, + "step": 33029 + }, + { + "epoch": 0.78, + "grad_norm": 1.897375332010031, + "learning_rate": 2.4719252116341207e-06, + "loss": 0.8997, + "step": 33030 + }, + { + "epoch": 0.78, + "grad_norm": 2.2588323231523724, + "learning_rate": 2.4714229688215864e-06, + "loss": 1.0666, + "step": 33031 + }, + { + "epoch": 0.78, + "grad_norm": 1.9187134876916536, + "learning_rate": 2.470920769842814e-06, + "loss": 1.0554, + "step": 33032 + }, + { + "epoch": 0.78, + "grad_norm": 2.2120365156294737, + "learning_rate": 2.4704186147007216e-06, + "loss": 1.0533, + "step": 33033 + }, + { + "epoch": 0.78, + "grad_norm": 2.253187383713671, + "learning_rate": 2.4699165033982377e-06, + "loss": 0.9275, + "step": 33034 + }, + { + "epoch": 0.78, + "grad_norm": 1.9889207258476123, + "learning_rate": 2.469414435938282e-06, + "loss": 0.9961, + "step": 33035 + }, + { + "epoch": 0.78, + "grad_norm": 2.0664803296964243, + "learning_rate": 2.4689124123237795e-06, + "loss": 0.9205, + "step": 33036 + }, + { + "epoch": 0.78, + "grad_norm": 2.0660181457396325, + "learning_rate": 2.4684104325576595e-06, + "loss": 0.985, + "step": 33037 + }, + { + "epoch": 0.78, + "grad_norm": 5.935177307557906, + "learning_rate": 2.467908496642831e-06, + "loss": 1.072, + "step": 33038 + }, + { + "epoch": 0.78, + "grad_norm": 2.012022110428896, + "learning_rate": 2.4674066045822256e-06, + "loss": 1.1291, + "step": 33039 + }, + { + "epoch": 0.78, + "grad_norm": 2.385670744493368, + "learning_rate": 2.466904756378763e-06, + "loss": 1.151, + "step": 33040 + }, + { + "epoch": 0.78, + "grad_norm": 1.8309983846273357, + "learning_rate": 2.4664029520353716e-06, + "loss": 0.8499, + "step": 33041 + }, + { + "epoch": 0.78, + "grad_norm": 2.1205422757408443, + "learning_rate": 2.4659011915549626e-06, + "loss": 0.9117, + "step": 33042 + }, + { + "epoch": 0.78, + "grad_norm": 2.026369933160967, + "learning_rate": 2.465399474940461e-06, + "loss": 0.916, + "step": 33043 + }, + { + "epoch": 0.78, + "grad_norm": 2.3128354139024694, + "learning_rate": 2.4648978021947923e-06, + "loss": 0.9765, + "step": 33044 + }, + { + "epoch": 0.78, + "grad_norm": 1.755662686557169, + "learning_rate": 2.464396173320873e-06, + "loss": 0.9872, + "step": 33045 + }, + { + "epoch": 0.78, + "grad_norm": 1.9470821460852545, + "learning_rate": 2.4638945883216236e-06, + "loss": 0.9937, + "step": 33046 + }, + { + "epoch": 0.78, + "grad_norm": 1.927347781091494, + "learning_rate": 2.4633930471999657e-06, + "loss": 1.067, + "step": 33047 + }, + { + "epoch": 0.78, + "grad_norm": 2.2189112538770246, + "learning_rate": 2.4628915499588213e-06, + "loss": 1.1463, + "step": 33048 + }, + { + "epoch": 0.78, + "grad_norm": 2.0733041368290617, + "learning_rate": 2.4623900966011062e-06, + "loss": 0.963, + "step": 33049 + }, + { + "epoch": 0.78, + "grad_norm": 1.9828272817817296, + "learning_rate": 2.4618886871297454e-06, + "loss": 1.0782, + "step": 33050 + }, + { + "epoch": 0.78, + "grad_norm": 1.148599946038793, + "learning_rate": 2.461387321547656e-06, + "loss": 1.0111, + "step": 33051 + }, + { + "epoch": 0.78, + "grad_norm": 2.1230255280950705, + "learning_rate": 2.4608859998577515e-06, + "loss": 1.1243, + "step": 33052 + }, + { + "epoch": 0.78, + "grad_norm": 1.8316956010027425, + "learning_rate": 2.4603847220629575e-06, + "loss": 0.9561, + "step": 33053 + }, + { + "epoch": 0.78, + "grad_norm": 2.2595265501265414, + "learning_rate": 2.4598834881661914e-06, + "loss": 1.0926, + "step": 33054 + }, + { + "epoch": 0.78, + "grad_norm": 2.4058991698847105, + "learning_rate": 2.459382298170373e-06, + "loss": 0.8957, + "step": 33055 + }, + { + "epoch": 0.78, + "grad_norm": 1.178873791028176, + "learning_rate": 2.4588811520784138e-06, + "loss": 0.9221, + "step": 33056 + }, + { + "epoch": 0.78, + "grad_norm": 1.1043772758883503, + "learning_rate": 2.4583800498932365e-06, + "loss": 0.9734, + "step": 33057 + }, + { + "epoch": 0.78, + "grad_norm": 1.0939009095076782, + "learning_rate": 2.4578789916177614e-06, + "loss": 0.9227, + "step": 33058 + }, + { + "epoch": 0.78, + "grad_norm": 1.9632828784916707, + "learning_rate": 2.4573779772549023e-06, + "loss": 0.9633, + "step": 33059 + }, + { + "epoch": 0.78, + "grad_norm": 1.9143862791661046, + "learning_rate": 2.456877006807573e-06, + "loss": 0.9691, + "step": 33060 + }, + { + "epoch": 0.78, + "grad_norm": 3.126739209903907, + "learning_rate": 2.456376080278695e-06, + "loss": 0.9892, + "step": 33061 + }, + { + "epoch": 0.78, + "grad_norm": 1.9836772260721047, + "learning_rate": 2.4558751976711857e-06, + "loss": 1.002, + "step": 33062 + }, + { + "epoch": 0.78, + "grad_norm": 1.841160467199044, + "learning_rate": 2.45537435898796e-06, + "loss": 0.9191, + "step": 33063 + }, + { + "epoch": 0.78, + "grad_norm": 1.972762007994336, + "learning_rate": 2.4548735642319297e-06, + "loss": 0.9918, + "step": 33064 + }, + { + "epoch": 0.78, + "grad_norm": 2.0160176506699377, + "learning_rate": 2.4543728134060186e-06, + "loss": 1.005, + "step": 33065 + }, + { + "epoch": 0.78, + "grad_norm": 2.0761606838753677, + "learning_rate": 2.4538721065131344e-06, + "loss": 0.8814, + "step": 33066 + }, + { + "epoch": 0.78, + "grad_norm": 1.8927811907598726, + "learning_rate": 2.453371443556198e-06, + "loss": 0.9743, + "step": 33067 + }, + { + "epoch": 0.78, + "grad_norm": 1.9865435521672636, + "learning_rate": 2.4528708245381203e-06, + "loss": 1.0306, + "step": 33068 + }, + { + "epoch": 0.78, + "grad_norm": 2.342034611593299, + "learning_rate": 2.4523702494618206e-06, + "loss": 0.9018, + "step": 33069 + }, + { + "epoch": 0.78, + "grad_norm": 1.7372642291784737, + "learning_rate": 2.4518697183302076e-06, + "loss": 0.8291, + "step": 33070 + }, + { + "epoch": 0.78, + "grad_norm": 1.1419321590845972, + "learning_rate": 2.4513692311462024e-06, + "loss": 0.9026, + "step": 33071 + }, + { + "epoch": 0.78, + "grad_norm": 2.008900688901291, + "learning_rate": 2.4508687879127145e-06, + "loss": 0.8765, + "step": 33072 + }, + { + "epoch": 0.78, + "grad_norm": 1.9525574654704696, + "learning_rate": 2.4503683886326557e-06, + "loss": 1.0902, + "step": 33073 + }, + { + "epoch": 0.78, + "grad_norm": 2.029001644626416, + "learning_rate": 2.449868033308943e-06, + "loss": 0.9209, + "step": 33074 + }, + { + "epoch": 0.78, + "grad_norm": 1.9009649889023505, + "learning_rate": 2.449367721944492e-06, + "loss": 1.0529, + "step": 33075 + }, + { + "epoch": 0.78, + "grad_norm": 2.03690296411577, + "learning_rate": 2.448867454542212e-06, + "loss": 1.0817, + "step": 33076 + }, + { + "epoch": 0.78, + "grad_norm": 1.891902578623563, + "learning_rate": 2.4483672311050134e-06, + "loss": 0.9661, + "step": 33077 + }, + { + "epoch": 0.78, + "grad_norm": 1.8586627294434377, + "learning_rate": 2.4478670516358116e-06, + "loss": 0.9261, + "step": 33078 + }, + { + "epoch": 0.78, + "grad_norm": 2.400362868425953, + "learning_rate": 2.4473669161375225e-06, + "loss": 1.0168, + "step": 33079 + }, + { + "epoch": 0.78, + "grad_norm": 1.9573056930141475, + "learning_rate": 2.4468668246130534e-06, + "loss": 1.0482, + "step": 33080 + }, + { + "epoch": 0.78, + "grad_norm": 1.9621059285693132, + "learning_rate": 2.4463667770653144e-06, + "loss": 0.9572, + "step": 33081 + }, + { + "epoch": 0.78, + "grad_norm": 1.7503186529429307, + "learning_rate": 2.4458667734972196e-06, + "loss": 0.9035, + "step": 33082 + }, + { + "epoch": 0.78, + "grad_norm": 2.0193333069935195, + "learning_rate": 2.4453668139116824e-06, + "loss": 0.9299, + "step": 33083 + }, + { + "epoch": 0.78, + "grad_norm": 2.3964092232945386, + "learning_rate": 2.444866898311612e-06, + "loss": 0.9324, + "step": 33084 + }, + { + "epoch": 0.78, + "grad_norm": 1.8085772749694218, + "learning_rate": 2.4443670266999155e-06, + "loss": 0.8958, + "step": 33085 + }, + { + "epoch": 0.78, + "grad_norm": 1.9278381806689964, + "learning_rate": 2.443867199079506e-06, + "loss": 1.012, + "step": 33086 + }, + { + "epoch": 0.78, + "grad_norm": 2.1179779765942297, + "learning_rate": 2.443367415453297e-06, + "loss": 1.0046, + "step": 33087 + }, + { + "epoch": 0.78, + "grad_norm": 1.0337936480768375, + "learning_rate": 2.4428676758241953e-06, + "loss": 0.9082, + "step": 33088 + }, + { + "epoch": 0.78, + "grad_norm": 2.0687191302382844, + "learning_rate": 2.4423679801951073e-06, + "loss": 1.0681, + "step": 33089 + }, + { + "epoch": 0.78, + "grad_norm": 1.124087490109428, + "learning_rate": 2.441868328568948e-06, + "loss": 0.976, + "step": 33090 + }, + { + "epoch": 0.78, + "grad_norm": 1.9773939497416586, + "learning_rate": 2.4413687209486223e-06, + "loss": 0.9346, + "step": 33091 + }, + { + "epoch": 0.78, + "grad_norm": 1.0865382151341914, + "learning_rate": 2.4408691573370438e-06, + "loss": 0.8492, + "step": 33092 + }, + { + "epoch": 0.78, + "grad_norm": 1.8994289317934188, + "learning_rate": 2.4403696377371146e-06, + "loss": 0.9897, + "step": 33093 + }, + { + "epoch": 0.78, + "grad_norm": 1.920122125976439, + "learning_rate": 2.4398701621517507e-06, + "loss": 1.0505, + "step": 33094 + }, + { + "epoch": 0.78, + "grad_norm": 1.9239466510244132, + "learning_rate": 2.439370730583852e-06, + "loss": 0.9877, + "step": 33095 + }, + { + "epoch": 0.78, + "grad_norm": 1.047818586114121, + "learning_rate": 2.4388713430363343e-06, + "loss": 0.8969, + "step": 33096 + }, + { + "epoch": 0.78, + "grad_norm": 2.0676372687496514, + "learning_rate": 2.4383719995121013e-06, + "loss": 1.0939, + "step": 33097 + }, + { + "epoch": 0.78, + "grad_norm": 2.415028766034764, + "learning_rate": 2.437872700014058e-06, + "loss": 0.951, + "step": 33098 + }, + { + "epoch": 0.78, + "grad_norm": 1.7313970235253946, + "learning_rate": 2.4373734445451127e-06, + "loss": 0.8962, + "step": 33099 + }, + { + "epoch": 0.78, + "grad_norm": 1.9294135964941224, + "learning_rate": 2.4368742331081775e-06, + "loss": 0.9158, + "step": 33100 + }, + { + "epoch": 0.78, + "grad_norm": 1.9827214699951845, + "learning_rate": 2.4363750657061546e-06, + "loss": 0.8957, + "step": 33101 + }, + { + "epoch": 0.78, + "grad_norm": 1.906439466726878, + "learning_rate": 2.4358759423419476e-06, + "loss": 0.8747, + "step": 33102 + }, + { + "epoch": 0.78, + "grad_norm": 2.1358790994775916, + "learning_rate": 2.4353768630184658e-06, + "loss": 0.9846, + "step": 33103 + }, + { + "epoch": 0.78, + "grad_norm": 2.425526759850779, + "learning_rate": 2.4348778277386177e-06, + "loss": 1.0066, + "step": 33104 + }, + { + "epoch": 0.78, + "grad_norm": 1.9206315610574471, + "learning_rate": 2.4343788365053066e-06, + "loss": 1.0708, + "step": 33105 + }, + { + "epoch": 0.78, + "grad_norm": 2.028355224336011, + "learning_rate": 2.4338798893214332e-06, + "loss": 1.0159, + "step": 33106 + }, + { + "epoch": 0.78, + "grad_norm": 2.7303270121891186, + "learning_rate": 2.433380986189907e-06, + "loss": 0.9285, + "step": 33107 + }, + { + "epoch": 0.78, + "grad_norm": 1.9028481699841824, + "learning_rate": 2.432882127113635e-06, + "loss": 0.9334, + "step": 33108 + }, + { + "epoch": 0.78, + "grad_norm": 1.79042921318819, + "learning_rate": 2.4323833120955197e-06, + "loss": 0.9711, + "step": 33109 + }, + { + "epoch": 0.78, + "grad_norm": 1.8735813307173574, + "learning_rate": 2.4318845411384608e-06, + "loss": 0.9761, + "step": 33110 + }, + { + "epoch": 0.78, + "grad_norm": 2.2556679561864454, + "learning_rate": 2.4313858142453695e-06, + "loss": 0.9158, + "step": 33111 + }, + { + "epoch": 0.78, + "grad_norm": 1.9699052807369768, + "learning_rate": 2.4308871314191427e-06, + "loss": 1.0107, + "step": 33112 + }, + { + "epoch": 0.78, + "grad_norm": 1.9191353183327837, + "learning_rate": 2.430388492662691e-06, + "loss": 0.9347, + "step": 33113 + }, + { + "epoch": 0.78, + "grad_norm": 2.0313794349274206, + "learning_rate": 2.429889897978911e-06, + "loss": 0.9382, + "step": 33114 + }, + { + "epoch": 0.78, + "grad_norm": 1.0681014799594837, + "learning_rate": 2.4293913473707108e-06, + "loss": 0.9725, + "step": 33115 + }, + { + "epoch": 0.78, + "grad_norm": 2.2767225053757865, + "learning_rate": 2.4288928408409885e-06, + "loss": 0.9706, + "step": 33116 + }, + { + "epoch": 0.78, + "grad_norm": 1.0866517427749782, + "learning_rate": 2.4283943783926523e-06, + "loss": 0.907, + "step": 33117 + }, + { + "epoch": 0.78, + "grad_norm": 2.171435498739869, + "learning_rate": 2.427895960028601e-06, + "loss": 1.031, + "step": 33118 + }, + { + "epoch": 0.78, + "grad_norm": 1.8858687619722125, + "learning_rate": 2.427397585751734e-06, + "loss": 0.9422, + "step": 33119 + }, + { + "epoch": 0.78, + "grad_norm": 1.9819322267928465, + "learning_rate": 2.4268992555649553e-06, + "loss": 0.9963, + "step": 33120 + }, + { + "epoch": 0.78, + "grad_norm": 2.094459414063942, + "learning_rate": 2.4264009694711665e-06, + "loss": 1.0108, + "step": 33121 + }, + { + "epoch": 0.78, + "grad_norm": 1.1173584077269008, + "learning_rate": 2.4259027274732748e-06, + "loss": 0.9456, + "step": 33122 + }, + { + "epoch": 0.78, + "grad_norm": 1.8496739950618926, + "learning_rate": 2.42540452957417e-06, + "loss": 0.933, + "step": 33123 + }, + { + "epoch": 0.78, + "grad_norm": 2.0825582723155707, + "learning_rate": 2.4249063757767587e-06, + "loss": 1.0861, + "step": 33124 + }, + { + "epoch": 0.78, + "grad_norm": 2.1620944908020885, + "learning_rate": 2.424408266083943e-06, + "loss": 0.9852, + "step": 33125 + }, + { + "epoch": 0.78, + "grad_norm": 2.32033030897669, + "learning_rate": 2.4239102004986203e-06, + "loss": 0.9112, + "step": 33126 + }, + { + "epoch": 0.78, + "grad_norm": 2.5509342325537916, + "learning_rate": 2.4234121790236887e-06, + "loss": 1.0258, + "step": 33127 + }, + { + "epoch": 0.78, + "grad_norm": 1.104051088605466, + "learning_rate": 2.42291420166205e-06, + "loss": 0.8876, + "step": 33128 + }, + { + "epoch": 0.78, + "grad_norm": 1.9622009697368983, + "learning_rate": 2.4224162684166064e-06, + "loss": 1.0522, + "step": 33129 + }, + { + "epoch": 0.78, + "grad_norm": 2.210247877703756, + "learning_rate": 2.4219183792902544e-06, + "loss": 1.0195, + "step": 33130 + }, + { + "epoch": 0.78, + "grad_norm": 2.430130199712346, + "learning_rate": 2.4214205342858897e-06, + "loss": 0.958, + "step": 33131 + }, + { + "epoch": 0.78, + "grad_norm": 1.741299985565153, + "learning_rate": 2.4209227334064167e-06, + "loss": 0.9592, + "step": 33132 + }, + { + "epoch": 0.78, + "grad_norm": 2.150054448989563, + "learning_rate": 2.4204249766547295e-06, + "loss": 0.9213, + "step": 33133 + }, + { + "epoch": 0.78, + "grad_norm": 1.795729754850857, + "learning_rate": 2.4199272640337256e-06, + "loss": 1.0522, + "step": 33134 + }, + { + "epoch": 0.78, + "grad_norm": 1.8811663466746638, + "learning_rate": 2.41942959554631e-06, + "loss": 0.8432, + "step": 33135 + }, + { + "epoch": 0.78, + "grad_norm": 2.3079967564469497, + "learning_rate": 2.418931971195374e-06, + "loss": 0.9647, + "step": 33136 + }, + { + "epoch": 0.78, + "grad_norm": 1.9816032346030614, + "learning_rate": 2.4184343909838135e-06, + "loss": 0.9758, + "step": 33137 + }, + { + "epoch": 0.78, + "grad_norm": 2.7146315916270005, + "learning_rate": 2.4179368549145286e-06, + "loss": 0.873, + "step": 33138 + }, + { + "epoch": 0.78, + "grad_norm": 1.8544181923971652, + "learning_rate": 2.4174393629904214e-06, + "loss": 1.042, + "step": 33139 + }, + { + "epoch": 0.78, + "grad_norm": 2.1100073195580924, + "learning_rate": 2.416941915214377e-06, + "loss": 0.9926, + "step": 33140 + }, + { + "epoch": 0.78, + "grad_norm": 2.284129604575008, + "learning_rate": 2.4164445115892966e-06, + "loss": 0.9246, + "step": 33141 + }, + { + "epoch": 0.78, + "grad_norm": 2.1797989553793475, + "learning_rate": 2.415947152118079e-06, + "loss": 0.8977, + "step": 33142 + }, + { + "epoch": 0.78, + "grad_norm": 1.854763941918923, + "learning_rate": 2.4154498368036226e-06, + "loss": 1.0087, + "step": 33143 + }, + { + "epoch": 0.78, + "grad_norm": 2.1626964295896918, + "learning_rate": 2.4149525656488128e-06, + "loss": 1.0024, + "step": 33144 + }, + { + "epoch": 0.78, + "grad_norm": 1.9188113469761592, + "learning_rate": 2.4144553386565504e-06, + "loss": 1.0077, + "step": 33145 + }, + { + "epoch": 0.78, + "grad_norm": 1.8254415507736286, + "learning_rate": 2.4139581558297344e-06, + "loss": 0.9717, + "step": 33146 + }, + { + "epoch": 0.78, + "grad_norm": 1.770452525945243, + "learning_rate": 2.4134610171712525e-06, + "loss": 1.005, + "step": 33147 + }, + { + "epoch": 0.78, + "grad_norm": 1.8716916040520655, + "learning_rate": 2.412963922684004e-06, + "loss": 1.0005, + "step": 33148 + }, + { + "epoch": 0.78, + "grad_norm": 1.7675216360180903, + "learning_rate": 2.41246687237088e-06, + "loss": 0.8957, + "step": 33149 + }, + { + "epoch": 0.78, + "grad_norm": 1.908747438185086, + "learning_rate": 2.411969866234779e-06, + "loss": 0.8507, + "step": 33150 + }, + { + "epoch": 0.78, + "grad_norm": 1.999963045477322, + "learning_rate": 2.411472904278589e-06, + "loss": 0.8976, + "step": 33151 + }, + { + "epoch": 0.78, + "grad_norm": 2.0141244310620845, + "learning_rate": 2.4109759865052084e-06, + "loss": 0.9456, + "step": 33152 + }, + { + "epoch": 0.78, + "grad_norm": 1.854590617234898, + "learning_rate": 2.4104791129175286e-06, + "loss": 1.0745, + "step": 33153 + }, + { + "epoch": 0.78, + "grad_norm": 2.0757776355505557, + "learning_rate": 2.4099822835184392e-06, + "loss": 0.873, + "step": 33154 + }, + { + "epoch": 0.78, + "grad_norm": 1.0537329146468222, + "learning_rate": 2.409485498310835e-06, + "loss": 0.945, + "step": 33155 + }, + { + "epoch": 0.78, + "grad_norm": 2.5020445032540883, + "learning_rate": 2.4089887572976134e-06, + "loss": 0.9443, + "step": 33156 + }, + { + "epoch": 0.78, + "grad_norm": 2.18357353668121, + "learning_rate": 2.408492060481663e-06, + "loss": 0.9814, + "step": 33157 + }, + { + "epoch": 0.78, + "grad_norm": 1.9568550070716695, + "learning_rate": 2.407995407865872e-06, + "loss": 0.8913, + "step": 33158 + }, + { + "epoch": 0.78, + "grad_norm": 1.904454465055227, + "learning_rate": 2.407498799453134e-06, + "loss": 0.9728, + "step": 33159 + }, + { + "epoch": 0.78, + "grad_norm": 2.0822972964501933, + "learning_rate": 2.407002235246346e-06, + "loss": 0.9109, + "step": 33160 + }, + { + "epoch": 0.78, + "grad_norm": 1.9011738749423617, + "learning_rate": 2.406505715248395e-06, + "loss": 0.9926, + "step": 33161 + }, + { + "epoch": 0.78, + "grad_norm": 2.516886085528942, + "learning_rate": 2.406009239462168e-06, + "loss": 0.988, + "step": 33162 + }, + { + "epoch": 0.78, + "grad_norm": 1.8974758147364714, + "learning_rate": 2.40551280789056e-06, + "loss": 1.1414, + "step": 33163 + }, + { + "epoch": 0.78, + "grad_norm": 1.134727772444138, + "learning_rate": 2.4050164205364633e-06, + "loss": 0.9005, + "step": 33164 + }, + { + "epoch": 0.78, + "grad_norm": 1.7994730555125353, + "learning_rate": 2.4045200774027653e-06, + "loss": 0.8047, + "step": 33165 + }, + { + "epoch": 0.78, + "grad_norm": 2.183204129471875, + "learning_rate": 2.404023778492354e-06, + "loss": 0.9809, + "step": 33166 + }, + { + "epoch": 0.78, + "grad_norm": 2.5790294837015555, + "learning_rate": 2.4035275238081236e-06, + "loss": 0.8835, + "step": 33167 + }, + { + "epoch": 0.78, + "grad_norm": 1.823634019277569, + "learning_rate": 2.4030313133529583e-06, + "loss": 0.9467, + "step": 33168 + }, + { + "epoch": 0.78, + "grad_norm": 2.001115471815538, + "learning_rate": 2.402535147129753e-06, + "loss": 1.0277, + "step": 33169 + }, + { + "epoch": 0.78, + "grad_norm": 1.8014626635996667, + "learning_rate": 2.4020390251413893e-06, + "loss": 1.0632, + "step": 33170 + }, + { + "epoch": 0.78, + "grad_norm": 2.7865863343435997, + "learning_rate": 2.4015429473907637e-06, + "loss": 0.9397, + "step": 33171 + }, + { + "epoch": 0.78, + "grad_norm": 1.9604753037661475, + "learning_rate": 2.4010469138807578e-06, + "loss": 0.9418, + "step": 33172 + }, + { + "epoch": 0.78, + "grad_norm": 1.8478988156568357, + "learning_rate": 2.4005509246142655e-06, + "loss": 0.9894, + "step": 33173 + }, + { + "epoch": 0.78, + "grad_norm": 1.8677890179782157, + "learning_rate": 2.400054979594172e-06, + "loss": 1.0695, + "step": 33174 + }, + { + "epoch": 0.78, + "grad_norm": 1.9146672066656094, + "learning_rate": 2.3995590788233604e-06, + "loss": 0.8828, + "step": 33175 + }, + { + "epoch": 0.78, + "grad_norm": 1.9692173985566521, + "learning_rate": 2.399063222304723e-06, + "loss": 0.9266, + "step": 33176 + }, + { + "epoch": 0.78, + "grad_norm": 1.8237644248569078, + "learning_rate": 2.3985674100411484e-06, + "loss": 1.0137, + "step": 33177 + }, + { + "epoch": 0.78, + "grad_norm": 1.0629653761730402, + "learning_rate": 2.398071642035521e-06, + "loss": 0.9401, + "step": 33178 + }, + { + "epoch": 0.78, + "grad_norm": 2.1591665870548584, + "learning_rate": 2.3975759182907244e-06, + "loss": 0.9697, + "step": 33179 + }, + { + "epoch": 0.78, + "grad_norm": 2.233612173959467, + "learning_rate": 2.3970802388096483e-06, + "loss": 0.8387, + "step": 33180 + }, + { + "epoch": 0.78, + "grad_norm": 2.0495324805986623, + "learning_rate": 2.39658460359518e-06, + "loss": 0.9633, + "step": 33181 + }, + { + "epoch": 0.78, + "grad_norm": 2.2428033817040163, + "learning_rate": 2.3960890126502034e-06, + "loss": 0.9205, + "step": 33182 + }, + { + "epoch": 0.78, + "grad_norm": 2.168943142578537, + "learning_rate": 2.3955934659776013e-06, + "loss": 0.9709, + "step": 33183 + }, + { + "epoch": 0.78, + "grad_norm": 2.0487592066663938, + "learning_rate": 2.3950979635802617e-06, + "loss": 0.9644, + "step": 33184 + }, + { + "epoch": 0.78, + "grad_norm": 1.8265517407078793, + "learning_rate": 2.3946025054610722e-06, + "loss": 1.0499, + "step": 33185 + }, + { + "epoch": 0.78, + "grad_norm": 2.0090158723460774, + "learning_rate": 2.394107091622915e-06, + "loss": 1.0742, + "step": 33186 + }, + { + "epoch": 0.78, + "grad_norm": 1.1081700383774564, + "learning_rate": 2.3936117220686717e-06, + "loss": 0.9057, + "step": 33187 + }, + { + "epoch": 0.78, + "grad_norm": 1.9622130442681778, + "learning_rate": 2.3931163968012315e-06, + "loss": 1.0671, + "step": 33188 + }, + { + "epoch": 0.78, + "grad_norm": 2.097968353052738, + "learning_rate": 2.3926211158234725e-06, + "loss": 0.9019, + "step": 33189 + }, + { + "epoch": 0.78, + "grad_norm": 2.264200001777955, + "learning_rate": 2.392125879138286e-06, + "loss": 1.0865, + "step": 33190 + }, + { + "epoch": 0.78, + "grad_norm": 2.0693793013234054, + "learning_rate": 2.3916306867485474e-06, + "loss": 1.0565, + "step": 33191 + }, + { + "epoch": 0.78, + "grad_norm": 2.0267854723035956, + "learning_rate": 2.3911355386571466e-06, + "loss": 0.882, + "step": 33192 + }, + { + "epoch": 0.78, + "grad_norm": 1.9019310232293634, + "learning_rate": 2.390640434866961e-06, + "loss": 0.9099, + "step": 33193 + }, + { + "epoch": 0.78, + "grad_norm": 2.072379481911695, + "learning_rate": 2.390145375380879e-06, + "loss": 1.1148, + "step": 33194 + }, + { + "epoch": 0.78, + "grad_norm": 1.0469873202620352, + "learning_rate": 2.3896503602017794e-06, + "loss": 0.977, + "step": 33195 + }, + { + "epoch": 0.78, + "grad_norm": 1.118453246831091, + "learning_rate": 2.3891553893325424e-06, + "loss": 0.8908, + "step": 33196 + }, + { + "epoch": 0.78, + "grad_norm": 1.9248277992059009, + "learning_rate": 2.388660462776051e-06, + "loss": 1.0265, + "step": 33197 + }, + { + "epoch": 0.78, + "grad_norm": 2.008892611267377, + "learning_rate": 2.388165580535192e-06, + "loss": 1.0371, + "step": 33198 + }, + { + "epoch": 0.78, + "grad_norm": 2.434093276737296, + "learning_rate": 2.387670742612843e-06, + "loss": 0.9877, + "step": 33199 + }, + { + "epoch": 0.78, + "grad_norm": 1.9407354430978718, + "learning_rate": 2.387175949011882e-06, + "loss": 0.9228, + "step": 33200 + }, + { + "epoch": 0.78, + "grad_norm": 1.8013128155615792, + "learning_rate": 2.3866811997351925e-06, + "loss": 0.8853, + "step": 33201 + }, + { + "epoch": 0.78, + "grad_norm": 1.7326514718626642, + "learning_rate": 2.3861864947856583e-06, + "loss": 0.983, + "step": 33202 + }, + { + "epoch": 0.78, + "grad_norm": 1.972554101046038, + "learning_rate": 2.385691834166157e-06, + "loss": 0.9916, + "step": 33203 + }, + { + "epoch": 0.78, + "grad_norm": 1.8498741356177586, + "learning_rate": 2.385197217879566e-06, + "loss": 0.8676, + "step": 33204 + }, + { + "epoch": 0.78, + "grad_norm": 1.9747711930345684, + "learning_rate": 2.384702645928767e-06, + "loss": 0.8579, + "step": 33205 + }, + { + "epoch": 0.78, + "grad_norm": 1.9983719284463872, + "learning_rate": 2.3842081183166434e-06, + "loss": 1.0735, + "step": 33206 + }, + { + "epoch": 0.78, + "grad_norm": 1.8751072694596487, + "learning_rate": 2.383713635046071e-06, + "loss": 1.0244, + "step": 33207 + }, + { + "epoch": 0.78, + "grad_norm": 1.9547039792473864, + "learning_rate": 2.383219196119927e-06, + "loss": 0.9997, + "step": 33208 + }, + { + "epoch": 0.78, + "grad_norm": 1.1064290560422865, + "learning_rate": 2.3827248015410907e-06, + "loss": 0.9382, + "step": 33209 + }, + { + "epoch": 0.78, + "grad_norm": 2.456153555941005, + "learning_rate": 2.382230451312446e-06, + "loss": 0.951, + "step": 33210 + }, + { + "epoch": 0.78, + "grad_norm": 2.0714089462562777, + "learning_rate": 2.381736145436867e-06, + "loss": 1.0838, + "step": 33211 + }, + { + "epoch": 0.78, + "grad_norm": 2.0812027817879595, + "learning_rate": 2.38124188391723e-06, + "loss": 1.0064, + "step": 33212 + }, + { + "epoch": 0.78, + "grad_norm": 1.8309811493411645, + "learning_rate": 2.3807476667564168e-06, + "loss": 1.0306, + "step": 33213 + }, + { + "epoch": 0.78, + "grad_norm": 2.18081357326374, + "learning_rate": 2.3802534939573006e-06, + "loss": 1.0319, + "step": 33214 + }, + { + "epoch": 0.78, + "grad_norm": 2.075688877370441, + "learning_rate": 2.3797593655227637e-06, + "loss": 1.0945, + "step": 33215 + }, + { + "epoch": 0.78, + "grad_norm": 1.765763652234455, + "learning_rate": 2.379265281455677e-06, + "loss": 0.8923, + "step": 33216 + }, + { + "epoch": 0.78, + "grad_norm": 2.3936138196354433, + "learning_rate": 2.378771241758924e-06, + "loss": 0.9475, + "step": 33217 + }, + { + "epoch": 0.78, + "grad_norm": 1.9504102388859192, + "learning_rate": 2.378277246435374e-06, + "loss": 0.9143, + "step": 33218 + }, + { + "epoch": 0.78, + "grad_norm": 2.038789728845491, + "learning_rate": 2.3777832954879075e-06, + "loss": 0.9707, + "step": 33219 + }, + { + "epoch": 0.78, + "grad_norm": 1.8490963090062094, + "learning_rate": 2.377289388919406e-06, + "loss": 0.99, + "step": 33220 + }, + { + "epoch": 0.78, + "grad_norm": 2.0236405609946404, + "learning_rate": 2.3767955267327325e-06, + "loss": 0.9187, + "step": 33221 + }, + { + "epoch": 0.78, + "grad_norm": 1.1160001036255942, + "learning_rate": 2.3763017089307692e-06, + "loss": 0.8945, + "step": 33222 + }, + { + "epoch": 0.78, + "grad_norm": 2.2102685219442386, + "learning_rate": 2.375807935516391e-06, + "loss": 0.9473, + "step": 33223 + }, + { + "epoch": 0.78, + "grad_norm": 1.8353487297035025, + "learning_rate": 2.3753142064924785e-06, + "loss": 0.9121, + "step": 33224 + }, + { + "epoch": 0.78, + "grad_norm": 1.9509059215338744, + "learning_rate": 2.374820521861895e-06, + "loss": 1.1067, + "step": 33225 + }, + { + "epoch": 0.78, + "grad_norm": 2.091129254875266, + "learning_rate": 2.37432688162752e-06, + "loss": 1.1004, + "step": 33226 + }, + { + "epoch": 0.78, + "grad_norm": 1.1388091044317115, + "learning_rate": 2.3738332857922307e-06, + "loss": 0.9992, + "step": 33227 + }, + { + "epoch": 0.78, + "grad_norm": 3.0040643320076295, + "learning_rate": 2.3733397343588994e-06, + "loss": 0.9456, + "step": 33228 + }, + { + "epoch": 0.78, + "grad_norm": 2.0504139497489144, + "learning_rate": 2.3728462273303954e-06, + "loss": 1.0163, + "step": 33229 + }, + { + "epoch": 0.78, + "grad_norm": 1.9798011648621012, + "learning_rate": 2.3723527647095956e-06, + "loss": 1.0198, + "step": 33230 + }, + { + "epoch": 0.78, + "grad_norm": 2.3624434751669043, + "learning_rate": 2.3718593464993757e-06, + "loss": 0.9999, + "step": 33231 + }, + { + "epoch": 0.78, + "grad_norm": 2.2763307573405576, + "learning_rate": 2.3713659727026016e-06, + "loss": 1.0514, + "step": 33232 + }, + { + "epoch": 0.78, + "grad_norm": 2.1535179929756585, + "learning_rate": 2.370872643322154e-06, + "loss": 1.02, + "step": 33233 + }, + { + "epoch": 0.78, + "grad_norm": 1.9318646772563013, + "learning_rate": 2.3703793583609013e-06, + "loss": 0.9483, + "step": 33234 + }, + { + "epoch": 0.78, + "grad_norm": 1.8441308978548814, + "learning_rate": 2.3698861178217125e-06, + "loss": 1.0727, + "step": 33235 + }, + { + "epoch": 0.78, + "grad_norm": 1.9837005941269952, + "learning_rate": 2.369392921707463e-06, + "loss": 1.0279, + "step": 33236 + }, + { + "epoch": 0.78, + "grad_norm": 1.7868541426483118, + "learning_rate": 2.368899770021026e-06, + "loss": 0.9149, + "step": 33237 + }, + { + "epoch": 0.78, + "grad_norm": 2.399564271788961, + "learning_rate": 2.36840666276527e-06, + "loss": 1.0665, + "step": 33238 + }, + { + "epoch": 0.78, + "grad_norm": 1.0147104885698552, + "learning_rate": 2.3679135999430646e-06, + "loss": 0.92, + "step": 33239 + }, + { + "epoch": 0.78, + "grad_norm": 1.8923448576582191, + "learning_rate": 2.3674205815572816e-06, + "loss": 0.8764, + "step": 33240 + }, + { + "epoch": 0.78, + "grad_norm": 1.0207359717700333, + "learning_rate": 2.3669276076107985e-06, + "loss": 0.868, + "step": 33241 + }, + { + "epoch": 0.78, + "grad_norm": 1.8435916913490884, + "learning_rate": 2.366434678106474e-06, + "loss": 0.9997, + "step": 33242 + }, + { + "epoch": 0.78, + "grad_norm": 1.6785416129344017, + "learning_rate": 2.365941793047184e-06, + "loss": 0.7951, + "step": 33243 + }, + { + "epoch": 0.78, + "grad_norm": 1.9450854206976864, + "learning_rate": 2.365448952435797e-06, + "loss": 0.9673, + "step": 33244 + }, + { + "epoch": 0.78, + "grad_norm": 1.9964971050841063, + "learning_rate": 2.3649561562751853e-06, + "loss": 0.9888, + "step": 33245 + }, + { + "epoch": 0.78, + "grad_norm": 1.8455975126396449, + "learning_rate": 2.364463404568218e-06, + "loss": 0.7867, + "step": 33246 + }, + { + "epoch": 0.78, + "grad_norm": 2.375352194495897, + "learning_rate": 2.3639706973177577e-06, + "loss": 0.9791, + "step": 33247 + }, + { + "epoch": 0.78, + "grad_norm": 2.4415479019079376, + "learning_rate": 2.3634780345266805e-06, + "loss": 1.003, + "step": 33248 + }, + { + "epoch": 0.78, + "grad_norm": 2.1437528412737694, + "learning_rate": 2.3629854161978505e-06, + "loss": 0.8832, + "step": 33249 + }, + { + "epoch": 0.78, + "grad_norm": 1.91188590772059, + "learning_rate": 2.3624928423341387e-06, + "loss": 1.0414, + "step": 33250 + }, + { + "epoch": 0.78, + "grad_norm": 1.7886471236869472, + "learning_rate": 2.3620003129384095e-06, + "loss": 0.8823, + "step": 33251 + }, + { + "epoch": 0.78, + "grad_norm": 2.0788130571220043, + "learning_rate": 2.361507828013535e-06, + "loss": 0.9722, + "step": 33252 + }, + { + "epoch": 0.78, + "grad_norm": 1.903661082636826, + "learning_rate": 2.3610153875623777e-06, + "loss": 1.0096, + "step": 33253 + }, + { + "epoch": 0.78, + "grad_norm": 1.8957450778146276, + "learning_rate": 2.3605229915878093e-06, + "loss": 0.9498, + "step": 33254 + }, + { + "epoch": 0.78, + "grad_norm": 2.103399823878852, + "learning_rate": 2.3600306400926955e-06, + "loss": 0.9286, + "step": 33255 + }, + { + "epoch": 0.78, + "grad_norm": 1.7831542171187966, + "learning_rate": 2.359538333079898e-06, + "loss": 0.9731, + "step": 33256 + }, + { + "epoch": 0.78, + "grad_norm": 2.1609655638523626, + "learning_rate": 2.359046070552289e-06, + "loss": 0.9176, + "step": 33257 + }, + { + "epoch": 0.78, + "grad_norm": 2.289223270865167, + "learning_rate": 2.358553852512735e-06, + "loss": 0.9312, + "step": 33258 + }, + { + "epoch": 0.78, + "grad_norm": 1.9536144384229608, + "learning_rate": 2.3580616789640997e-06, + "loss": 1.0177, + "step": 33259 + }, + { + "epoch": 0.78, + "grad_norm": 1.1163834230396383, + "learning_rate": 2.3575695499092455e-06, + "loss": 0.9082, + "step": 33260 + }, + { + "epoch": 0.78, + "grad_norm": 2.352882517050254, + "learning_rate": 2.3570774653510407e-06, + "loss": 1.0453, + "step": 33261 + }, + { + "epoch": 0.78, + "grad_norm": 2.1709499125656384, + "learning_rate": 2.3565854252923547e-06, + "loss": 0.9731, + "step": 33262 + }, + { + "epoch": 0.78, + "grad_norm": 1.8365819101924954, + "learning_rate": 2.356093429736047e-06, + "loss": 0.9447, + "step": 33263 + }, + { + "epoch": 0.78, + "grad_norm": 1.9109126979320963, + "learning_rate": 2.355601478684981e-06, + "loss": 1.0527, + "step": 33264 + }, + { + "epoch": 0.78, + "grad_norm": 1.9119430592401767, + "learning_rate": 2.355109572142024e-06, + "loss": 0.9434, + "step": 33265 + }, + { + "epoch": 0.78, + "grad_norm": 1.8992085368944167, + "learning_rate": 2.3546177101100422e-06, + "loss": 0.9606, + "step": 33266 + }, + { + "epoch": 0.78, + "grad_norm": 2.006377732624061, + "learning_rate": 2.354125892591895e-06, + "loss": 1.0134, + "step": 33267 + }, + { + "epoch": 0.78, + "grad_norm": 1.8941479665743164, + "learning_rate": 2.3536341195904455e-06, + "loss": 0.9282, + "step": 33268 + }, + { + "epoch": 0.78, + "grad_norm": 1.835293647554238, + "learning_rate": 2.3531423911085626e-06, + "loss": 1.0555, + "step": 33269 + }, + { + "epoch": 0.78, + "grad_norm": 2.2983977221577616, + "learning_rate": 2.3526507071491023e-06, + "loss": 1.0691, + "step": 33270 + }, + { + "epoch": 0.78, + "grad_norm": 1.0808680314209218, + "learning_rate": 2.352159067714933e-06, + "loss": 0.9003, + "step": 33271 + }, + { + "epoch": 0.78, + "grad_norm": 1.9489249162528193, + "learning_rate": 2.3516674728089127e-06, + "loss": 0.9709, + "step": 33272 + }, + { + "epoch": 0.78, + "grad_norm": 2.0680590423688496, + "learning_rate": 2.351175922433908e-06, + "loss": 0.9769, + "step": 33273 + }, + { + "epoch": 0.78, + "grad_norm": 2.046050297898226, + "learning_rate": 2.3506844165927765e-06, + "loss": 0.9511, + "step": 33274 + }, + { + "epoch": 0.78, + "grad_norm": 1.89804891159357, + "learning_rate": 2.350192955288385e-06, + "loss": 0.8723, + "step": 33275 + }, + { + "epoch": 0.78, + "grad_norm": 1.9674940618495231, + "learning_rate": 2.349701538523591e-06, + "loss": 0.876, + "step": 33276 + }, + { + "epoch": 0.78, + "grad_norm": 1.9755135632305416, + "learning_rate": 2.3492101663012546e-06, + "loss": 0.9275, + "step": 33277 + }, + { + "epoch": 0.78, + "grad_norm": 2.296897949346168, + "learning_rate": 2.3487188386242387e-06, + "loss": 0.9169, + "step": 33278 + }, + { + "epoch": 0.78, + "grad_norm": 1.8555394041261861, + "learning_rate": 2.348227555495407e-06, + "loss": 0.8973, + "step": 33279 + }, + { + "epoch": 0.78, + "grad_norm": 1.867131571488235, + "learning_rate": 2.347736316917617e-06, + "loss": 0.9894, + "step": 33280 + }, + { + "epoch": 0.78, + "grad_norm": 1.0444795374152587, + "learning_rate": 2.3472451228937254e-06, + "loss": 0.9081, + "step": 33281 + }, + { + "epoch": 0.78, + "grad_norm": 1.8784787404582335, + "learning_rate": 2.3467539734265966e-06, + "loss": 1.0191, + "step": 33282 + }, + { + "epoch": 0.78, + "grad_norm": 1.985392026165221, + "learning_rate": 2.3462628685190913e-06, + "loss": 0.9633, + "step": 33283 + }, + { + "epoch": 0.78, + "grad_norm": 1.7187852505027625, + "learning_rate": 2.3457718081740664e-06, + "loss": 0.9078, + "step": 33284 + }, + { + "epoch": 0.78, + "grad_norm": 1.8067195869923214, + "learning_rate": 2.345280792394379e-06, + "loss": 0.9536, + "step": 33285 + }, + { + "epoch": 0.78, + "grad_norm": 2.11064153095308, + "learning_rate": 2.3447898211828903e-06, + "loss": 1.1251, + "step": 33286 + }, + { + "epoch": 0.78, + "grad_norm": 1.7121793017983744, + "learning_rate": 2.3442988945424617e-06, + "loss": 1.0438, + "step": 33287 + }, + { + "epoch": 0.78, + "grad_norm": 1.9915338921883732, + "learning_rate": 2.3438080124759476e-06, + "loss": 0.969, + "step": 33288 + }, + { + "epoch": 0.78, + "grad_norm": 2.01795111445001, + "learning_rate": 2.343317174986206e-06, + "loss": 0.9606, + "step": 33289 + }, + { + "epoch": 0.78, + "grad_norm": 1.1253246983912357, + "learning_rate": 2.342826382076098e-06, + "loss": 0.9149, + "step": 33290 + }, + { + "epoch": 0.78, + "grad_norm": 1.0975375507953384, + "learning_rate": 2.342335633748476e-06, + "loss": 0.9553, + "step": 33291 + }, + { + "epoch": 0.78, + "grad_norm": 2.139017737958579, + "learning_rate": 2.341844930006203e-06, + "loss": 0.9747, + "step": 33292 + }, + { + "epoch": 0.78, + "grad_norm": 2.2249140603006845, + "learning_rate": 2.341354270852131e-06, + "loss": 0.9908, + "step": 33293 + }, + { + "epoch": 0.78, + "grad_norm": 1.947444489547048, + "learning_rate": 2.3408636562891218e-06, + "loss": 1.1929, + "step": 33294 + }, + { + "epoch": 0.78, + "grad_norm": 2.2884659127037015, + "learning_rate": 2.3403730863200256e-06, + "loss": 0.9063, + "step": 33295 + }, + { + "epoch": 0.78, + "grad_norm": 1.8883364757095769, + "learning_rate": 2.339882560947706e-06, + "loss": 0.9384, + "step": 33296 + }, + { + "epoch": 0.78, + "grad_norm": 2.1609468005111925, + "learning_rate": 2.3393920801750157e-06, + "loss": 1.0903, + "step": 33297 + }, + { + "epoch": 0.78, + "grad_norm": 1.7979442568913524, + "learning_rate": 2.3389016440048063e-06, + "loss": 1.0337, + "step": 33298 + }, + { + "epoch": 0.78, + "grad_norm": 1.8990454214293484, + "learning_rate": 2.338411252439937e-06, + "loss": 0.8741, + "step": 33299 + }, + { + "epoch": 0.78, + "grad_norm": 1.9176170843467475, + "learning_rate": 2.337920905483266e-06, + "loss": 0.9657, + "step": 33300 + }, + { + "epoch": 0.78, + "grad_norm": 2.8125465407428196, + "learning_rate": 2.337430603137646e-06, + "loss": 0.9444, + "step": 33301 + }, + { + "epoch": 0.78, + "grad_norm": 1.932742810732447, + "learning_rate": 2.336940345405928e-06, + "loss": 1.0604, + "step": 33302 + }, + { + "epoch": 0.78, + "grad_norm": 1.0209410777792274, + "learning_rate": 2.3364501322909685e-06, + "loss": 0.906, + "step": 33303 + }, + { + "epoch": 0.78, + "grad_norm": 2.276736276485789, + "learning_rate": 2.3359599637956264e-06, + "loss": 0.9626, + "step": 33304 + }, + { + "epoch": 0.78, + "grad_norm": 2.0445956945329677, + "learning_rate": 2.3354698399227504e-06, + "loss": 1.0792, + "step": 33305 + }, + { + "epoch": 0.78, + "grad_norm": 2.2624679433905523, + "learning_rate": 2.3349797606751945e-06, + "loss": 1.0468, + "step": 33306 + }, + { + "epoch": 0.78, + "grad_norm": 1.1222747698629374, + "learning_rate": 2.334489726055812e-06, + "loss": 0.8754, + "step": 33307 + }, + { + "epoch": 0.78, + "grad_norm": 1.8591791313877672, + "learning_rate": 2.3339997360674606e-06, + "loss": 1.0777, + "step": 33308 + }, + { + "epoch": 0.78, + "grad_norm": 1.850514731131822, + "learning_rate": 2.333509790712989e-06, + "loss": 0.9476, + "step": 33309 + }, + { + "epoch": 0.78, + "grad_norm": 1.8280993118733064, + "learning_rate": 2.3330198899952484e-06, + "loss": 0.9734, + "step": 33310 + }, + { + "epoch": 0.78, + "grad_norm": 1.8467956677513608, + "learning_rate": 2.3325300339170954e-06, + "loss": 1.0164, + "step": 33311 + }, + { + "epoch": 0.78, + "grad_norm": 1.8793340773855822, + "learning_rate": 2.3320402224813776e-06, + "loss": 0.9786, + "step": 33312 + }, + { + "epoch": 0.78, + "grad_norm": 2.8475199955915556, + "learning_rate": 2.3315504556909517e-06, + "loss": 1.0227, + "step": 33313 + }, + { + "epoch": 0.78, + "grad_norm": 2.2976553746445205, + "learning_rate": 2.3310607335486645e-06, + "loss": 0.8922, + "step": 33314 + }, + { + "epoch": 0.78, + "grad_norm": 1.9978099163466618, + "learning_rate": 2.3305710560573723e-06, + "loss": 1.0117, + "step": 33315 + }, + { + "epoch": 0.78, + "grad_norm": 1.8540655339481789, + "learning_rate": 2.330081423219921e-06, + "loss": 0.9803, + "step": 33316 + }, + { + "epoch": 0.78, + "grad_norm": 2.3875678935796305, + "learning_rate": 2.3295918350391632e-06, + "loss": 0.9625, + "step": 33317 + }, + { + "epoch": 0.78, + "grad_norm": 1.9400973246547628, + "learning_rate": 2.3291022915179575e-06, + "loss": 0.9108, + "step": 33318 + }, + { + "epoch": 0.78, + "grad_norm": 1.8433811214754348, + "learning_rate": 2.3286127926591397e-06, + "loss": 0.9631, + "step": 33319 + }, + { + "epoch": 0.78, + "grad_norm": 2.0959652250628005, + "learning_rate": 2.3281233384655665e-06, + "loss": 0.9912, + "step": 33320 + }, + { + "epoch": 0.79, + "grad_norm": 1.1249010252379796, + "learning_rate": 2.3276339289400895e-06, + "loss": 0.9444, + "step": 33321 + }, + { + "epoch": 0.79, + "grad_norm": 2.0265863365568926, + "learning_rate": 2.3271445640855618e-06, + "loss": 0.8724, + "step": 33322 + }, + { + "epoch": 0.79, + "grad_norm": 1.8864942983826907, + "learning_rate": 2.326655243904822e-06, + "loss": 0.9461, + "step": 33323 + }, + { + "epoch": 0.79, + "grad_norm": 2.2185737240136887, + "learning_rate": 2.326165968400724e-06, + "loss": 0.9432, + "step": 33324 + }, + { + "epoch": 0.79, + "grad_norm": 2.1233383518630085, + "learning_rate": 2.325676737576121e-06, + "loss": 1.0762, + "step": 33325 + }, + { + "epoch": 0.79, + "grad_norm": 2.087613902887768, + "learning_rate": 2.3251875514338575e-06, + "loss": 0.9117, + "step": 33326 + }, + { + "epoch": 0.79, + "grad_norm": 1.9155573984156589, + "learning_rate": 2.324698409976779e-06, + "loss": 0.9495, + "step": 33327 + }, + { + "epoch": 0.79, + "grad_norm": 2.085712423508724, + "learning_rate": 2.324209313207736e-06, + "loss": 1.01, + "step": 33328 + }, + { + "epoch": 0.79, + "grad_norm": 2.284658124217253, + "learning_rate": 2.3237202611295795e-06, + "loss": 1.0506, + "step": 33329 + }, + { + "epoch": 0.79, + "grad_norm": 2.0340479422607434, + "learning_rate": 2.323231253745151e-06, + "loss": 0.9495, + "step": 33330 + }, + { + "epoch": 0.79, + "grad_norm": 1.9110087597866248, + "learning_rate": 2.322742291057304e-06, + "loss": 0.8725, + "step": 33331 + }, + { + "epoch": 0.79, + "grad_norm": 1.1702441774986343, + "learning_rate": 2.322253373068879e-06, + "loss": 1.0184, + "step": 33332 + }, + { + "epoch": 0.79, + "grad_norm": 1.8370539540593642, + "learning_rate": 2.321764499782727e-06, + "loss": 1.087, + "step": 33333 + }, + { + "epoch": 0.79, + "grad_norm": 2.1182043304546507, + "learning_rate": 2.3212756712016913e-06, + "loss": 0.9371, + "step": 33334 + }, + { + "epoch": 0.79, + "grad_norm": 2.0954275552205335, + "learning_rate": 2.3207868873286232e-06, + "loss": 1.0911, + "step": 33335 + }, + { + "epoch": 0.79, + "grad_norm": 2.0461497179823267, + "learning_rate": 2.3202981481663646e-06, + "loss": 1.0349, + "step": 33336 + }, + { + "epoch": 0.79, + "grad_norm": 2.272040927842134, + "learning_rate": 2.319809453717757e-06, + "loss": 0.9873, + "step": 33337 + }, + { + "epoch": 0.79, + "grad_norm": 2.0865631625178853, + "learning_rate": 2.319320803985653e-06, + "loss": 1.0495, + "step": 33338 + }, + { + "epoch": 0.79, + "grad_norm": 1.9566021901481732, + "learning_rate": 2.3188321989728955e-06, + "loss": 0.9236, + "step": 33339 + }, + { + "epoch": 0.79, + "grad_norm": 2.4645410160904837, + "learning_rate": 2.31834363868233e-06, + "loss": 0.8304, + "step": 33340 + }, + { + "epoch": 0.79, + "grad_norm": 2.015473366895361, + "learning_rate": 2.3178551231167968e-06, + "loss": 0.9584, + "step": 33341 + }, + { + "epoch": 0.79, + "grad_norm": 2.0444109680603826, + "learning_rate": 2.317366652279143e-06, + "loss": 0.9752, + "step": 33342 + }, + { + "epoch": 0.79, + "grad_norm": 1.8975114623011844, + "learning_rate": 2.3168782261722156e-06, + "loss": 1.0692, + "step": 33343 + }, + { + "epoch": 0.79, + "grad_norm": 1.8547764678765242, + "learning_rate": 2.3163898447988562e-06, + "loss": 0.904, + "step": 33344 + }, + { + "epoch": 0.79, + "grad_norm": 2.405972607507184, + "learning_rate": 2.315901508161904e-06, + "loss": 0.9748, + "step": 33345 + }, + { + "epoch": 0.79, + "grad_norm": 2.1622050213786133, + "learning_rate": 2.315413216264206e-06, + "loss": 0.9186, + "step": 33346 + }, + { + "epoch": 0.79, + "grad_norm": 1.9482816223106814, + "learning_rate": 2.3149249691086094e-06, + "loss": 1.075, + "step": 33347 + }, + { + "epoch": 0.79, + "grad_norm": 2.057760463930519, + "learning_rate": 2.3144367666979517e-06, + "loss": 0.9435, + "step": 33348 + }, + { + "epoch": 0.79, + "grad_norm": 2.1794087815428855, + "learning_rate": 2.3139486090350727e-06, + "loss": 0.9951, + "step": 33349 + }, + { + "epoch": 0.79, + "grad_norm": 1.6988407316055905, + "learning_rate": 2.3134604961228225e-06, + "loss": 0.8143, + "step": 33350 + }, + { + "epoch": 0.79, + "grad_norm": 2.199876260439828, + "learning_rate": 2.3129724279640355e-06, + "loss": 0.9293, + "step": 33351 + }, + { + "epoch": 0.79, + "grad_norm": 2.2011011649852303, + "learning_rate": 2.312484404561559e-06, + "loss": 1.0817, + "step": 33352 + }, + { + "epoch": 0.79, + "grad_norm": 1.864233824614057, + "learning_rate": 2.311996425918229e-06, + "loss": 0.9603, + "step": 33353 + }, + { + "epoch": 0.79, + "grad_norm": 1.9805491249973903, + "learning_rate": 2.3115084920368935e-06, + "loss": 1.0215, + "step": 33354 + }, + { + "epoch": 0.79, + "grad_norm": 2.0110633540572036, + "learning_rate": 2.3110206029203872e-06, + "loss": 1.0391, + "step": 33355 + }, + { + "epoch": 0.79, + "grad_norm": 2.105907266795235, + "learning_rate": 2.310532758571555e-06, + "loss": 1.0481, + "step": 33356 + }, + { + "epoch": 0.79, + "grad_norm": 1.8998625736931933, + "learning_rate": 2.310044958993236e-06, + "loss": 0.9217, + "step": 33357 + }, + { + "epoch": 0.79, + "grad_norm": 1.9542988451666223, + "learning_rate": 2.3095572041882675e-06, + "loss": 1.0138, + "step": 33358 + }, + { + "epoch": 0.79, + "grad_norm": 2.038300718931507, + "learning_rate": 2.3090694941594914e-06, + "loss": 1.0318, + "step": 33359 + }, + { + "epoch": 0.79, + "grad_norm": 1.1324962256670785, + "learning_rate": 2.3085818289097506e-06, + "loss": 0.9939, + "step": 33360 + }, + { + "epoch": 0.79, + "grad_norm": 1.9585094157811767, + "learning_rate": 2.308094208441881e-06, + "loss": 0.993, + "step": 33361 + }, + { + "epoch": 0.79, + "grad_norm": 1.966502030332843, + "learning_rate": 2.307606632758719e-06, + "loss": 1.0289, + "step": 33362 + }, + { + "epoch": 0.79, + "grad_norm": 2.122372414405117, + "learning_rate": 2.307119101863108e-06, + "loss": 0.968, + "step": 33363 + }, + { + "epoch": 0.79, + "grad_norm": 1.9942663201705848, + "learning_rate": 2.306631615757886e-06, + "loss": 0.941, + "step": 33364 + }, + { + "epoch": 0.79, + "grad_norm": 2.0141173738502025, + "learning_rate": 2.306144174445891e-06, + "loss": 1.0716, + "step": 33365 + }, + { + "epoch": 0.79, + "grad_norm": 1.0747865955497942, + "learning_rate": 2.3056567779299588e-06, + "loss": 0.893, + "step": 33366 + }, + { + "epoch": 0.79, + "grad_norm": 2.1374969720426873, + "learning_rate": 2.305169426212928e-06, + "loss": 0.9837, + "step": 33367 + }, + { + "epoch": 0.79, + "grad_norm": 1.8829510405823264, + "learning_rate": 2.30468211929764e-06, + "loss": 1.0848, + "step": 33368 + }, + { + "epoch": 0.79, + "grad_norm": 1.8883275708868508, + "learning_rate": 2.304194857186929e-06, + "loss": 0.95, + "step": 33369 + }, + { + "epoch": 0.79, + "grad_norm": 1.9324409181359412, + "learning_rate": 2.303707639883629e-06, + "loss": 0.9949, + "step": 33370 + }, + { + "epoch": 0.79, + "grad_norm": 1.80337919084677, + "learning_rate": 2.303220467390582e-06, + "loss": 1.0028, + "step": 33371 + }, + { + "epoch": 0.79, + "grad_norm": 1.8774502340414305, + "learning_rate": 2.30273333971062e-06, + "loss": 0.9607, + "step": 33372 + }, + { + "epoch": 0.79, + "grad_norm": 1.0940373798875593, + "learning_rate": 2.3022462568465843e-06, + "loss": 0.9193, + "step": 33373 + }, + { + "epoch": 0.79, + "grad_norm": 2.0867184379715584, + "learning_rate": 2.301759218801305e-06, + "loss": 1.0677, + "step": 33374 + }, + { + "epoch": 0.79, + "grad_norm": 1.8813291124775962, + "learning_rate": 2.3012722255776234e-06, + "loss": 1.0166, + "step": 33375 + }, + { + "epoch": 0.79, + "grad_norm": 2.21543430257727, + "learning_rate": 2.300785277178369e-06, + "loss": 1.004, + "step": 33376 + }, + { + "epoch": 0.79, + "grad_norm": 2.3527389204341422, + "learning_rate": 2.3002983736063843e-06, + "loss": 0.8743, + "step": 33377 + }, + { + "epoch": 0.79, + "grad_norm": 2.4220252317317272, + "learning_rate": 2.2998115148644997e-06, + "loss": 0.9514, + "step": 33378 + }, + { + "epoch": 0.79, + "grad_norm": 2.176089966033135, + "learning_rate": 2.2993247009555463e-06, + "loss": 0.9882, + "step": 33379 + }, + { + "epoch": 0.79, + "grad_norm": 1.9874911251883063, + "learning_rate": 2.298837931882364e-06, + "loss": 0.9746, + "step": 33380 + }, + { + "epoch": 0.79, + "grad_norm": 1.8096998123865577, + "learning_rate": 2.2983512076477875e-06, + "loss": 1.0358, + "step": 33381 + }, + { + "epoch": 0.79, + "grad_norm": 1.9695890459720458, + "learning_rate": 2.2978645282546487e-06, + "loss": 0.9648, + "step": 33382 + }, + { + "epoch": 0.79, + "grad_norm": 1.9576913771747815, + "learning_rate": 2.2973778937057787e-06, + "loss": 1.1002, + "step": 33383 + }, + { + "epoch": 0.79, + "grad_norm": 1.8560306229762962, + "learning_rate": 2.296891304004013e-06, + "loss": 0.9615, + "step": 33384 + }, + { + "epoch": 0.79, + "grad_norm": 2.0623467254761714, + "learning_rate": 2.296404759152188e-06, + "loss": 1.0146, + "step": 33385 + }, + { + "epoch": 0.79, + "grad_norm": 2.0083643203902914, + "learning_rate": 2.2959182591531327e-06, + "loss": 1.091, + "step": 33386 + }, + { + "epoch": 0.79, + "grad_norm": 2.555161794246587, + "learning_rate": 2.295431804009678e-06, + "loss": 1.0172, + "step": 33387 + }, + { + "epoch": 0.79, + "grad_norm": 1.8466476108447172, + "learning_rate": 2.294945393724658e-06, + "loss": 1.0264, + "step": 33388 + }, + { + "epoch": 0.79, + "grad_norm": 3.2019716736383197, + "learning_rate": 2.2944590283009083e-06, + "loss": 1.045, + "step": 33389 + }, + { + "epoch": 0.79, + "grad_norm": 2.185772203244963, + "learning_rate": 2.2939727077412578e-06, + "loss": 0.8992, + "step": 33390 + }, + { + "epoch": 0.79, + "grad_norm": 2.1834657376491573, + "learning_rate": 2.293486432048535e-06, + "loss": 0.9794, + "step": 33391 + }, + { + "epoch": 0.79, + "grad_norm": 2.7321177706880286, + "learning_rate": 2.293000201225577e-06, + "loss": 0.9755, + "step": 33392 + }, + { + "epoch": 0.79, + "grad_norm": 2.1851556502795906, + "learning_rate": 2.292514015275209e-06, + "loss": 0.9395, + "step": 33393 + }, + { + "epoch": 0.79, + "grad_norm": 2.18397767433598, + "learning_rate": 2.2920278742002677e-06, + "loss": 0.9447, + "step": 33394 + }, + { + "epoch": 0.79, + "grad_norm": 1.9208712206809535, + "learning_rate": 2.2915417780035763e-06, + "loss": 0.9692, + "step": 33395 + }, + { + "epoch": 0.79, + "grad_norm": 1.9118947536614046, + "learning_rate": 2.2910557266879742e-06, + "loss": 0.851, + "step": 33396 + }, + { + "epoch": 0.79, + "grad_norm": 2.025161514490882, + "learning_rate": 2.2905697202562817e-06, + "loss": 0.9676, + "step": 33397 + }, + { + "epoch": 0.79, + "grad_norm": 2.055874802225571, + "learning_rate": 2.2900837587113366e-06, + "loss": 0.9413, + "step": 33398 + }, + { + "epoch": 0.79, + "grad_norm": 1.9974785034331974, + "learning_rate": 2.289597842055964e-06, + "loss": 0.9239, + "step": 33399 + }, + { + "epoch": 0.79, + "grad_norm": 1.910719392486105, + "learning_rate": 2.289111970292992e-06, + "loss": 0.9209, + "step": 33400 + }, + { + "epoch": 0.79, + "grad_norm": 1.8666938360826626, + "learning_rate": 2.28862614342525e-06, + "loss": 0.8793, + "step": 33401 + }, + { + "epoch": 0.79, + "grad_norm": 1.9022814358726436, + "learning_rate": 2.288140361455572e-06, + "loss": 0.9399, + "step": 33402 + }, + { + "epoch": 0.79, + "grad_norm": 2.1437563240869264, + "learning_rate": 2.2876546243867826e-06, + "loss": 0.9397, + "step": 33403 + }, + { + "epoch": 0.79, + "grad_norm": 2.5321507191545125, + "learning_rate": 2.287168932221706e-06, + "loss": 0.9867, + "step": 33404 + }, + { + "epoch": 0.79, + "grad_norm": 1.9064898612372132, + "learning_rate": 2.286683284963174e-06, + "loss": 0.9761, + "step": 33405 + }, + { + "epoch": 0.79, + "grad_norm": 2.129910427663954, + "learning_rate": 2.2861976826140164e-06, + "loss": 0.9406, + "step": 33406 + }, + { + "epoch": 0.79, + "grad_norm": 2.066153058407826, + "learning_rate": 2.2857121251770585e-06, + "loss": 1.0776, + "step": 33407 + }, + { + "epoch": 0.79, + "grad_norm": 2.163886714782289, + "learning_rate": 2.285226612655124e-06, + "loss": 1.1478, + "step": 33408 + }, + { + "epoch": 0.79, + "grad_norm": 2.1915733953428074, + "learning_rate": 2.2847411450510426e-06, + "loss": 1.0581, + "step": 33409 + }, + { + "epoch": 0.79, + "grad_norm": 1.8079527470555297, + "learning_rate": 2.284255722367643e-06, + "loss": 0.9389, + "step": 33410 + }, + { + "epoch": 0.79, + "grad_norm": 2.037156472297157, + "learning_rate": 2.2837703446077496e-06, + "loss": 0.9451, + "step": 33411 + }, + { + "epoch": 0.79, + "grad_norm": 2.3645563048465728, + "learning_rate": 2.2832850117741866e-06, + "loss": 0.9778, + "step": 33412 + }, + { + "epoch": 0.79, + "grad_norm": 2.4482810546504385, + "learning_rate": 2.282799723869783e-06, + "loss": 0.9713, + "step": 33413 + }, + { + "epoch": 0.79, + "grad_norm": 2.0441901963142732, + "learning_rate": 2.2823144808973596e-06, + "loss": 1.0771, + "step": 33414 + }, + { + "epoch": 0.79, + "grad_norm": 1.8491481656131865, + "learning_rate": 2.2818292828597457e-06, + "loss": 0.8323, + "step": 33415 + }, + { + "epoch": 0.79, + "grad_norm": 2.481886357199959, + "learning_rate": 2.2813441297597673e-06, + "loss": 0.9083, + "step": 33416 + }, + { + "epoch": 0.79, + "grad_norm": 2.1016510050512216, + "learning_rate": 2.280859021600247e-06, + "loss": 1.0554, + "step": 33417 + }, + { + "epoch": 0.79, + "grad_norm": 2.221188984955755, + "learning_rate": 2.2803739583840066e-06, + "loss": 0.8035, + "step": 33418 + }, + { + "epoch": 0.79, + "grad_norm": 2.185752946826134, + "learning_rate": 2.2798889401138735e-06, + "loss": 0.8733, + "step": 33419 + }, + { + "epoch": 0.79, + "grad_norm": 2.335573040397273, + "learning_rate": 2.279403966792676e-06, + "loss": 1.0572, + "step": 33420 + }, + { + "epoch": 0.79, + "grad_norm": 1.8241727865180233, + "learning_rate": 2.2789190384232286e-06, + "loss": 0.9002, + "step": 33421 + }, + { + "epoch": 0.79, + "grad_norm": 1.844778780478333, + "learning_rate": 2.2784341550083577e-06, + "loss": 1.0642, + "step": 33422 + }, + { + "epoch": 0.79, + "grad_norm": 2.021680513946936, + "learning_rate": 2.277949316550888e-06, + "loss": 1.0956, + "step": 33423 + }, + { + "epoch": 0.79, + "grad_norm": 2.093224981500376, + "learning_rate": 2.2774645230536485e-06, + "loss": 0.8366, + "step": 33424 + }, + { + "epoch": 0.79, + "grad_norm": 2.5520022672453933, + "learning_rate": 2.276979774519449e-06, + "loss": 1.0947, + "step": 33425 + }, + { + "epoch": 0.79, + "grad_norm": 1.9625687579639104, + "learning_rate": 2.276495070951118e-06, + "loss": 0.9973, + "step": 33426 + }, + { + "epoch": 0.79, + "grad_norm": 2.0939494696743894, + "learning_rate": 2.2760104123514814e-06, + "loss": 1.0023, + "step": 33427 + }, + { + "epoch": 0.79, + "grad_norm": 2.2058728770288156, + "learning_rate": 2.275525798723355e-06, + "loss": 0.9888, + "step": 33428 + }, + { + "epoch": 0.79, + "grad_norm": 2.2741090411126756, + "learning_rate": 2.2750412300695656e-06, + "loss": 0.8707, + "step": 33429 + }, + { + "epoch": 0.79, + "grad_norm": 1.9532671276474303, + "learning_rate": 2.2745567063929285e-06, + "loss": 1.136, + "step": 33430 + }, + { + "epoch": 0.79, + "grad_norm": 2.0244298625859436, + "learning_rate": 2.2740722276962714e-06, + "loss": 1.0721, + "step": 33431 + }, + { + "epoch": 0.79, + "grad_norm": 2.4968000216936317, + "learning_rate": 2.273587793982409e-06, + "loss": 1.0247, + "step": 33432 + }, + { + "epoch": 0.79, + "grad_norm": 2.1367596549576766, + "learning_rate": 2.273103405254167e-06, + "loss": 0.9643, + "step": 33433 + }, + { + "epoch": 0.79, + "grad_norm": 1.8172072152670777, + "learning_rate": 2.2726190615143638e-06, + "loss": 0.8417, + "step": 33434 + }, + { + "epoch": 0.79, + "grad_norm": 1.961406577806621, + "learning_rate": 2.272134762765815e-06, + "loss": 0.8664, + "step": 33435 + }, + { + "epoch": 0.79, + "grad_norm": 2.3234289431181905, + "learning_rate": 2.271650509011346e-06, + "loss": 0.9964, + "step": 33436 + }, + { + "epoch": 0.79, + "grad_norm": 1.9120434438810987, + "learning_rate": 2.2711663002537754e-06, + "loss": 1.101, + "step": 33437 + }, + { + "epoch": 0.79, + "grad_norm": 1.8805779789286603, + "learning_rate": 2.270682136495923e-06, + "loss": 0.9934, + "step": 33438 + }, + { + "epoch": 0.79, + "grad_norm": 2.2481622236513514, + "learning_rate": 2.2701980177406026e-06, + "loss": 0.9396, + "step": 33439 + }, + { + "epoch": 0.79, + "grad_norm": 2.0580792419169844, + "learning_rate": 2.2697139439906367e-06, + "loss": 0.9986, + "step": 33440 + }, + { + "epoch": 0.79, + "grad_norm": 2.1329830072811213, + "learning_rate": 2.269229915248846e-06, + "loss": 0.9303, + "step": 33441 + }, + { + "epoch": 0.79, + "grad_norm": 2.0547540187266775, + "learning_rate": 2.2687459315180473e-06, + "loss": 1.0252, + "step": 33442 + }, + { + "epoch": 0.79, + "grad_norm": 2.2593821281148196, + "learning_rate": 2.268261992801054e-06, + "loss": 1.0084, + "step": 33443 + }, + { + "epoch": 0.79, + "grad_norm": 1.061091263885511, + "learning_rate": 2.267778099100687e-06, + "loss": 0.8848, + "step": 33444 + }, + { + "epoch": 0.79, + "grad_norm": 1.9724147771587572, + "learning_rate": 2.267294250419767e-06, + "loss": 1.0544, + "step": 33445 + }, + { + "epoch": 0.79, + "grad_norm": 1.8752438723769194, + "learning_rate": 2.2668104467611087e-06, + "loss": 0.9687, + "step": 33446 + }, + { + "epoch": 0.79, + "grad_norm": 2.0893226932095375, + "learning_rate": 2.2663266881275247e-06, + "loss": 0.9423, + "step": 33447 + }, + { + "epoch": 0.79, + "grad_norm": 2.0664552791400035, + "learning_rate": 2.265842974521838e-06, + "loss": 1.04, + "step": 33448 + }, + { + "epoch": 0.79, + "grad_norm": 2.1027188938440142, + "learning_rate": 2.265359305946859e-06, + "loss": 0.933, + "step": 33449 + }, + { + "epoch": 0.79, + "grad_norm": 1.8315961042408797, + "learning_rate": 2.2648756824054097e-06, + "loss": 0.9068, + "step": 33450 + }, + { + "epoch": 0.79, + "grad_norm": 3.127859710752634, + "learning_rate": 2.2643921039003013e-06, + "loss": 0.9547, + "step": 33451 + }, + { + "epoch": 0.79, + "grad_norm": 1.8542801846047785, + "learning_rate": 2.263908570434352e-06, + "loss": 0.8998, + "step": 33452 + }, + { + "epoch": 0.79, + "grad_norm": 1.9354792709650526, + "learning_rate": 2.2634250820103752e-06, + "loss": 0.8851, + "step": 33453 + }, + { + "epoch": 0.79, + "grad_norm": 2.3043430355836527, + "learning_rate": 2.26294163863119e-06, + "loss": 1.0352, + "step": 33454 + }, + { + "epoch": 0.79, + "grad_norm": 2.14610647789948, + "learning_rate": 2.2624582402996066e-06, + "loss": 0.9885, + "step": 33455 + }, + { + "epoch": 0.79, + "grad_norm": 2.2992982506212427, + "learning_rate": 2.261974887018439e-06, + "loss": 0.9667, + "step": 33456 + }, + { + "epoch": 0.79, + "grad_norm": 3.3546100690832255, + "learning_rate": 2.2614915787905024e-06, + "loss": 0.9774, + "step": 33457 + }, + { + "epoch": 0.79, + "grad_norm": 3.168334658906937, + "learning_rate": 2.261008315618616e-06, + "loss": 0.9357, + "step": 33458 + }, + { + "epoch": 0.79, + "grad_norm": 2.01764282546804, + "learning_rate": 2.260525097505588e-06, + "loss": 0.846, + "step": 33459 + }, + { + "epoch": 0.79, + "grad_norm": 2.2252713070375654, + "learning_rate": 2.2600419244542316e-06, + "loss": 0.9581, + "step": 33460 + }, + { + "epoch": 0.79, + "grad_norm": 3.748665478915377, + "learning_rate": 2.2595587964673614e-06, + "loss": 1.0002, + "step": 33461 + }, + { + "epoch": 0.79, + "grad_norm": 1.9047577919445358, + "learning_rate": 2.2590757135477927e-06, + "loss": 0.9097, + "step": 33462 + }, + { + "epoch": 0.79, + "grad_norm": 3.2150016163042148, + "learning_rate": 2.2585926756983355e-06, + "loss": 0.9423, + "step": 33463 + }, + { + "epoch": 0.79, + "grad_norm": 4.381478976075012, + "learning_rate": 2.2581096829218008e-06, + "loss": 0.939, + "step": 33464 + }, + { + "epoch": 0.79, + "grad_norm": 2.704819572784611, + "learning_rate": 2.2576267352210022e-06, + "loss": 0.9734, + "step": 33465 + }, + { + "epoch": 0.79, + "grad_norm": 1.1038856895427602, + "learning_rate": 2.257143832598755e-06, + "loss": 0.9233, + "step": 33466 + }, + { + "epoch": 0.79, + "grad_norm": 2.0537623333497157, + "learning_rate": 2.256660975057867e-06, + "loss": 0.9346, + "step": 33467 + }, + { + "epoch": 0.79, + "grad_norm": 1.854245534672897, + "learning_rate": 2.256178162601148e-06, + "loss": 0.9721, + "step": 33468 + }, + { + "epoch": 0.79, + "grad_norm": 1.9765515059751528, + "learning_rate": 2.255695395231412e-06, + "loss": 0.9319, + "step": 33469 + }, + { + "epoch": 0.79, + "grad_norm": 1.9275823675607453, + "learning_rate": 2.255212672951471e-06, + "loss": 0.9335, + "step": 33470 + }, + { + "epoch": 0.79, + "grad_norm": 1.8238822707866493, + "learning_rate": 2.2547299957641345e-06, + "loss": 0.969, + "step": 33471 + }, + { + "epoch": 0.79, + "grad_norm": 1.898419542973367, + "learning_rate": 2.2542473636722095e-06, + "loss": 0.8868, + "step": 33472 + }, + { + "epoch": 0.79, + "grad_norm": 1.7288452968225558, + "learning_rate": 2.2537647766785107e-06, + "loss": 0.9116, + "step": 33473 + }, + { + "epoch": 0.79, + "grad_norm": 1.919017282803444, + "learning_rate": 2.253282234785843e-06, + "loss": 0.9901, + "step": 33474 + }, + { + "epoch": 0.79, + "grad_norm": 2.055061593194975, + "learning_rate": 2.2527997379970214e-06, + "loss": 0.9098, + "step": 33475 + }, + { + "epoch": 0.79, + "grad_norm": 2.29077424365612, + "learning_rate": 2.25231728631485e-06, + "loss": 1.0832, + "step": 33476 + }, + { + "epoch": 0.79, + "grad_norm": 1.9627186823260352, + "learning_rate": 2.251834879742143e-06, + "loss": 0.8839, + "step": 33477 + }, + { + "epoch": 0.79, + "grad_norm": 2.005863495551337, + "learning_rate": 2.251352518281704e-06, + "loss": 0.9495, + "step": 33478 + }, + { + "epoch": 0.79, + "grad_norm": 2.001388789977359, + "learning_rate": 2.250870201936346e-06, + "loss": 0.9379, + "step": 33479 + }, + { + "epoch": 0.79, + "grad_norm": 2.262021793786385, + "learning_rate": 2.250387930708875e-06, + "loss": 1.0335, + "step": 33480 + }, + { + "epoch": 0.79, + "grad_norm": 1.1421073723247457, + "learning_rate": 2.2499057046020966e-06, + "loss": 0.943, + "step": 33481 + }, + { + "epoch": 0.79, + "grad_norm": 1.8758980917318142, + "learning_rate": 2.24942352361882e-06, + "loss": 0.9253, + "step": 33482 + }, + { + "epoch": 0.79, + "grad_norm": 1.960772842261071, + "learning_rate": 2.248941387761857e-06, + "loss": 1.011, + "step": 33483 + }, + { + "epoch": 0.79, + "grad_norm": 1.972747654201767, + "learning_rate": 2.2484592970340115e-06, + "loss": 0.9281, + "step": 33484 + }, + { + "epoch": 0.79, + "grad_norm": 2.225741479735261, + "learning_rate": 2.247977251438086e-06, + "loss": 0.947, + "step": 33485 + }, + { + "epoch": 0.79, + "grad_norm": 1.9594308150952857, + "learning_rate": 2.247495250976892e-06, + "loss": 0.9017, + "step": 33486 + }, + { + "epoch": 0.79, + "grad_norm": 1.8232425486396222, + "learning_rate": 2.2470132956532376e-06, + "loss": 0.8754, + "step": 33487 + }, + { + "epoch": 0.79, + "grad_norm": 2.539762810725305, + "learning_rate": 2.2465313854699265e-06, + "loss": 0.9392, + "step": 33488 + }, + { + "epoch": 0.79, + "grad_norm": 2.0900852704418837, + "learning_rate": 2.2460495204297615e-06, + "loss": 0.9258, + "step": 33489 + }, + { + "epoch": 0.79, + "grad_norm": 2.0963223869260093, + "learning_rate": 2.2455677005355513e-06, + "loss": 0.9601, + "step": 33490 + }, + { + "epoch": 0.79, + "grad_norm": 2.1257010570182424, + "learning_rate": 2.245085925790104e-06, + "loss": 0.9972, + "step": 33491 + }, + { + "epoch": 0.79, + "grad_norm": 2.055473694232187, + "learning_rate": 2.2446041961962207e-06, + "loss": 0.8921, + "step": 33492 + }, + { + "epoch": 0.79, + "grad_norm": 2.419759074421929, + "learning_rate": 2.2441225117567056e-06, + "loss": 0.9951, + "step": 33493 + }, + { + "epoch": 0.79, + "grad_norm": 2.2462309409827177, + "learning_rate": 2.2436408724743664e-06, + "loss": 1.0786, + "step": 33494 + }, + { + "epoch": 0.79, + "grad_norm": 1.1337704651849156, + "learning_rate": 2.243159278352003e-06, + "loss": 0.9312, + "step": 33495 + }, + { + "epoch": 0.79, + "grad_norm": 2.0028216304022597, + "learning_rate": 2.242677729392425e-06, + "loss": 0.9415, + "step": 33496 + }, + { + "epoch": 0.79, + "grad_norm": 1.8657937932947555, + "learning_rate": 2.242196225598431e-06, + "loss": 0.9606, + "step": 33497 + }, + { + "epoch": 0.79, + "grad_norm": 1.9713444616103213, + "learning_rate": 2.241714766972828e-06, + "loss": 0.8618, + "step": 33498 + }, + { + "epoch": 0.79, + "grad_norm": 2.5539714009367147, + "learning_rate": 2.241233353518416e-06, + "loss": 0.8947, + "step": 33499 + }, + { + "epoch": 0.79, + "grad_norm": 1.8203530397884153, + "learning_rate": 2.240751985238e-06, + "loss": 0.8937, + "step": 33500 + }, + { + "epoch": 0.79, + "grad_norm": 2.934780890515078, + "learning_rate": 2.2402706621343873e-06, + "loss": 1.0336, + "step": 33501 + }, + { + "epoch": 0.79, + "grad_norm": 2.147739153200654, + "learning_rate": 2.2397893842103714e-06, + "loss": 0.8638, + "step": 33502 + }, + { + "epoch": 0.79, + "grad_norm": 1.871197643237781, + "learning_rate": 2.2393081514687564e-06, + "loss": 0.9693, + "step": 33503 + }, + { + "epoch": 0.79, + "grad_norm": 2.0142663305792885, + "learning_rate": 2.2388269639123482e-06, + "loss": 1.0162, + "step": 33504 + }, + { + "epoch": 0.79, + "grad_norm": 1.9120151321937506, + "learning_rate": 2.2383458215439513e-06, + "loss": 1.1196, + "step": 33505 + }, + { + "epoch": 0.79, + "grad_norm": 2.6038866182984113, + "learning_rate": 2.2378647243663563e-06, + "loss": 0.9558, + "step": 33506 + }, + { + "epoch": 0.79, + "grad_norm": 1.948178674623458, + "learning_rate": 2.2373836723823714e-06, + "loss": 0.9351, + "step": 33507 + }, + { + "epoch": 0.79, + "grad_norm": 3.170611687166311, + "learning_rate": 2.2369026655947988e-06, + "loss": 0.9329, + "step": 33508 + }, + { + "epoch": 0.79, + "grad_norm": 1.962739968923552, + "learning_rate": 2.236421704006436e-06, + "loss": 0.9596, + "step": 33509 + }, + { + "epoch": 0.79, + "grad_norm": 2.4201228797806555, + "learning_rate": 2.2359407876200822e-06, + "loss": 0.9513, + "step": 33510 + }, + { + "epoch": 0.79, + "grad_norm": 1.9858543868783363, + "learning_rate": 2.235459916438538e-06, + "loss": 0.9366, + "step": 33511 + }, + { + "epoch": 0.79, + "grad_norm": 2.7484531297166113, + "learning_rate": 2.2349790904646085e-06, + "loss": 0.9478, + "step": 33512 + }, + { + "epoch": 0.79, + "grad_norm": 1.9989765190382445, + "learning_rate": 2.2344983097010854e-06, + "loss": 1.0301, + "step": 33513 + }, + { + "epoch": 0.79, + "grad_norm": 2.0148011846324607, + "learning_rate": 2.234017574150774e-06, + "loss": 0.9658, + "step": 33514 + }, + { + "epoch": 0.79, + "grad_norm": 2.242513119427437, + "learning_rate": 2.2335368838164715e-06, + "loss": 0.9134, + "step": 33515 + }, + { + "epoch": 0.79, + "grad_norm": 2.030356911873966, + "learning_rate": 2.2330562387009745e-06, + "loss": 1.0761, + "step": 33516 + }, + { + "epoch": 0.79, + "grad_norm": 2.082942196281542, + "learning_rate": 2.2325756388070817e-06, + "loss": 0.8442, + "step": 33517 + }, + { + "epoch": 0.79, + "grad_norm": 1.1715841192599061, + "learning_rate": 2.2320950841375956e-06, + "loss": 0.9436, + "step": 33518 + }, + { + "epoch": 0.79, + "grad_norm": 2.2831150143780854, + "learning_rate": 2.2316145746953112e-06, + "loss": 0.8945, + "step": 33519 + }, + { + "epoch": 0.79, + "grad_norm": 2.080466970065326, + "learning_rate": 2.231134110483023e-06, + "loss": 1.1062, + "step": 33520 + }, + { + "epoch": 0.79, + "grad_norm": 1.9956871591302867, + "learning_rate": 2.2306536915035326e-06, + "loss": 1.0876, + "step": 33521 + }, + { + "epoch": 0.79, + "grad_norm": 2.1159026354186925, + "learning_rate": 2.2301733177596407e-06, + "loss": 0.9666, + "step": 33522 + }, + { + "epoch": 0.79, + "grad_norm": 2.1386792815855316, + "learning_rate": 2.229692989254135e-06, + "loss": 0.9183, + "step": 33523 + }, + { + "epoch": 0.79, + "grad_norm": 2.0950704839114516, + "learning_rate": 2.229212705989816e-06, + "loss": 0.9395, + "step": 33524 + }, + { + "epoch": 0.79, + "grad_norm": 2.0468577267513095, + "learning_rate": 2.228732467969481e-06, + "loss": 0.9918, + "step": 33525 + }, + { + "epoch": 0.79, + "grad_norm": 2.011818210911528, + "learning_rate": 2.228252275195929e-06, + "loss": 1.0164, + "step": 33526 + }, + { + "epoch": 0.79, + "grad_norm": 1.7826225970779563, + "learning_rate": 2.2277721276719532e-06, + "loss": 0.9237, + "step": 33527 + }, + { + "epoch": 0.79, + "grad_norm": 1.916216168640955, + "learning_rate": 2.227292025400345e-06, + "loss": 0.9053, + "step": 33528 + }, + { + "epoch": 0.79, + "grad_norm": 1.0627948567921268, + "learning_rate": 2.226811968383906e-06, + "loss": 0.946, + "step": 33529 + }, + { + "epoch": 0.79, + "grad_norm": 1.9799565482546726, + "learning_rate": 2.226331956625427e-06, + "loss": 0.9662, + "step": 33530 + }, + { + "epoch": 0.79, + "grad_norm": 2.080519020787158, + "learning_rate": 2.2258519901277066e-06, + "loss": 0.8912, + "step": 33531 + }, + { + "epoch": 0.79, + "grad_norm": 2.0455774960258206, + "learning_rate": 2.2253720688935344e-06, + "loss": 1.0239, + "step": 33532 + }, + { + "epoch": 0.79, + "grad_norm": 2.059313475926399, + "learning_rate": 2.2248921929257106e-06, + "loss": 1.0128, + "step": 33533 + }, + { + "epoch": 0.79, + "grad_norm": 1.8715623781907538, + "learning_rate": 2.2244123622270218e-06, + "loss": 1.0414, + "step": 33534 + }, + { + "epoch": 0.79, + "grad_norm": 2.1388151040845043, + "learning_rate": 2.223932576800271e-06, + "loss": 0.9766, + "step": 33535 + }, + { + "epoch": 0.79, + "grad_norm": 1.8238078634597465, + "learning_rate": 2.2234528366482445e-06, + "loss": 0.9636, + "step": 33536 + }, + { + "epoch": 0.79, + "grad_norm": 1.8231361537875532, + "learning_rate": 2.2229731417737354e-06, + "loss": 0.9519, + "step": 33537 + }, + { + "epoch": 0.79, + "grad_norm": 1.8784668555739, + "learning_rate": 2.2224934921795392e-06, + "loss": 1.0017, + "step": 33538 + }, + { + "epoch": 0.79, + "grad_norm": 2.2289858834038374, + "learning_rate": 2.2220138878684516e-06, + "loss": 0.9615, + "step": 33539 + }, + { + "epoch": 0.79, + "grad_norm": 1.9111524633528856, + "learning_rate": 2.221534328843261e-06, + "loss": 0.931, + "step": 33540 + }, + { + "epoch": 0.79, + "grad_norm": 1.1196881152575062, + "learning_rate": 2.221054815106757e-06, + "loss": 0.8694, + "step": 33541 + }, + { + "epoch": 0.79, + "grad_norm": 1.8104569561063175, + "learning_rate": 2.2205753466617353e-06, + "loss": 1.0146, + "step": 33542 + }, + { + "epoch": 0.79, + "grad_norm": 2.094855772603889, + "learning_rate": 2.2200959235109886e-06, + "loss": 0.9413, + "step": 33543 + }, + { + "epoch": 0.79, + "grad_norm": 2.87767002359302, + "learning_rate": 2.219616545657307e-06, + "loss": 0.8759, + "step": 33544 + }, + { + "epoch": 0.79, + "grad_norm": 2.3249264024486407, + "learning_rate": 2.2191372131034793e-06, + "loss": 0.9954, + "step": 33545 + }, + { + "epoch": 0.79, + "grad_norm": 2.2903913652932686, + "learning_rate": 2.2186579258522966e-06, + "loss": 0.9365, + "step": 33546 + }, + { + "epoch": 0.79, + "grad_norm": 2.1684911215640117, + "learning_rate": 2.2181786839065544e-06, + "loss": 0.961, + "step": 33547 + }, + { + "epoch": 0.79, + "grad_norm": 2.063880872789368, + "learning_rate": 2.2176994872690404e-06, + "loss": 1.0039, + "step": 33548 + }, + { + "epoch": 0.79, + "grad_norm": 1.917852772728584, + "learning_rate": 2.2172203359425405e-06, + "loss": 0.9167, + "step": 33549 + }, + { + "epoch": 0.79, + "grad_norm": 1.951375434827706, + "learning_rate": 2.2167412299298497e-06, + "loss": 0.8464, + "step": 33550 + }, + { + "epoch": 0.79, + "grad_norm": 2.197474223301782, + "learning_rate": 2.2162621692337537e-06, + "loss": 1.088, + "step": 33551 + }, + { + "epoch": 0.79, + "grad_norm": 2.16534787362222, + "learning_rate": 2.2157831538570464e-06, + "loss": 0.9478, + "step": 33552 + }, + { + "epoch": 0.79, + "grad_norm": 2.1018814303357978, + "learning_rate": 2.2153041838025114e-06, + "loss": 0.9223, + "step": 33553 + }, + { + "epoch": 0.79, + "grad_norm": 2.0093383374811933, + "learning_rate": 2.2148252590729424e-06, + "loss": 0.9679, + "step": 33554 + }, + { + "epoch": 0.79, + "grad_norm": 1.0993274332350642, + "learning_rate": 2.2143463796711228e-06, + "loss": 0.9793, + "step": 33555 + }, + { + "epoch": 0.79, + "grad_norm": 1.10814837458159, + "learning_rate": 2.213867545599846e-06, + "loss": 0.965, + "step": 33556 + }, + { + "epoch": 0.79, + "grad_norm": 2.0807197888480893, + "learning_rate": 2.2133887568618972e-06, + "loss": 1.0908, + "step": 33557 + }, + { + "epoch": 0.79, + "grad_norm": 2.037664625745664, + "learning_rate": 2.2129100134600624e-06, + "loss": 0.9666, + "step": 33558 + }, + { + "epoch": 0.79, + "grad_norm": 1.9836847629463108, + "learning_rate": 2.21243131539713e-06, + "loss": 0.9776, + "step": 33559 + }, + { + "epoch": 0.79, + "grad_norm": 2.1365209411411574, + "learning_rate": 2.2119526626758914e-06, + "loss": 0.9378, + "step": 33560 + }, + { + "epoch": 0.79, + "grad_norm": 2.274373837275403, + "learning_rate": 2.21147405529913e-06, + "loss": 0.9195, + "step": 33561 + }, + { + "epoch": 0.79, + "grad_norm": 2.141435985430929, + "learning_rate": 2.2109954932696285e-06, + "loss": 1.0029, + "step": 33562 + }, + { + "epoch": 0.79, + "grad_norm": 2.007099262420099, + "learning_rate": 2.210516976590179e-06, + "loss": 0.8779, + "step": 33563 + }, + { + "epoch": 0.79, + "grad_norm": 2.164526422618696, + "learning_rate": 2.2100385052635677e-06, + "loss": 1.021, + "step": 33564 + }, + { + "epoch": 0.79, + "grad_norm": 2.215229996783841, + "learning_rate": 2.209560079292579e-06, + "loss": 0.9932, + "step": 33565 + }, + { + "epoch": 0.79, + "grad_norm": 2.031260745600099, + "learning_rate": 2.2090816986799934e-06, + "loss": 0.8418, + "step": 33566 + }, + { + "epoch": 0.79, + "grad_norm": 2.0979408008670424, + "learning_rate": 2.208603363428602e-06, + "loss": 0.8391, + "step": 33567 + }, + { + "epoch": 0.79, + "grad_norm": 2.419074902857499, + "learning_rate": 2.2081250735411917e-06, + "loss": 1.0339, + "step": 33568 + }, + { + "epoch": 0.79, + "grad_norm": 2.316977877670368, + "learning_rate": 2.2076468290205433e-06, + "loss": 0.8483, + "step": 33569 + }, + { + "epoch": 0.79, + "grad_norm": 2.2360182856722886, + "learning_rate": 2.2071686298694393e-06, + "loss": 0.9429, + "step": 33570 + }, + { + "epoch": 0.79, + "grad_norm": 1.090102056921724, + "learning_rate": 2.2066904760906704e-06, + "loss": 0.9607, + "step": 33571 + }, + { + "epoch": 0.79, + "grad_norm": 2.121631286145915, + "learning_rate": 2.206212367687014e-06, + "loss": 0.9181, + "step": 33572 + }, + { + "epoch": 0.79, + "grad_norm": 1.135216640352985, + "learning_rate": 2.205734304661259e-06, + "loss": 0.8933, + "step": 33573 + }, + { + "epoch": 0.79, + "grad_norm": 2.105946667838297, + "learning_rate": 2.2052562870161832e-06, + "loss": 0.8788, + "step": 33574 + }, + { + "epoch": 0.79, + "grad_norm": 2.4193475040089396, + "learning_rate": 2.2047783147545763e-06, + "loss": 1.0067, + "step": 33575 + }, + { + "epoch": 0.79, + "grad_norm": 1.8049337308900717, + "learning_rate": 2.2043003878792156e-06, + "loss": 1.0336, + "step": 33576 + }, + { + "epoch": 0.79, + "grad_norm": 1.9386051495660337, + "learning_rate": 2.2038225063928887e-06, + "loss": 0.8802, + "step": 33577 + }, + { + "epoch": 0.79, + "grad_norm": 2.1104685944230015, + "learning_rate": 2.203344670298375e-06, + "loss": 1.0755, + "step": 33578 + }, + { + "epoch": 0.79, + "grad_norm": 2.0968453167967254, + "learning_rate": 2.202866879598454e-06, + "loss": 0.975, + "step": 33579 + }, + { + "epoch": 0.79, + "grad_norm": 2.10810632038139, + "learning_rate": 2.2023891342959114e-06, + "loss": 1.0184, + "step": 33580 + }, + { + "epoch": 0.79, + "grad_norm": 1.8518499545348333, + "learning_rate": 2.20191143439353e-06, + "loss": 0.9441, + "step": 33581 + }, + { + "epoch": 0.79, + "grad_norm": 1.9414981053006037, + "learning_rate": 2.20143377989409e-06, + "loss": 1.0554, + "step": 33582 + }, + { + "epoch": 0.79, + "grad_norm": 1.9051892298875328, + "learning_rate": 2.200956170800368e-06, + "loss": 0.8843, + "step": 33583 + }, + { + "epoch": 0.79, + "grad_norm": 1.9509669743189733, + "learning_rate": 2.2004786071151486e-06, + "loss": 0.9752, + "step": 33584 + }, + { + "epoch": 0.79, + "grad_norm": 2.0320250793034815, + "learning_rate": 2.2000010888412137e-06, + "loss": 0.8724, + "step": 33585 + }, + { + "epoch": 0.79, + "grad_norm": 1.8287491008831664, + "learning_rate": 2.1995236159813437e-06, + "loss": 1.0961, + "step": 33586 + }, + { + "epoch": 0.79, + "grad_norm": 1.9156819998759598, + "learning_rate": 2.199046188538313e-06, + "loss": 1.0352, + "step": 33587 + }, + { + "epoch": 0.79, + "grad_norm": 2.233460018022476, + "learning_rate": 2.198568806514906e-06, + "loss": 0.9447, + "step": 33588 + }, + { + "epoch": 0.79, + "grad_norm": 2.1554316700052905, + "learning_rate": 2.198091469913902e-06, + "loss": 0.9489, + "step": 33589 + }, + { + "epoch": 0.79, + "grad_norm": 1.865375270627439, + "learning_rate": 2.197614178738081e-06, + "loss": 1.078, + "step": 33590 + }, + { + "epoch": 0.79, + "grad_norm": 2.2937495282868436, + "learning_rate": 2.197136932990218e-06, + "loss": 0.9943, + "step": 33591 + }, + { + "epoch": 0.79, + "grad_norm": 1.9628940696116726, + "learning_rate": 2.1966597326730944e-06, + "loss": 1.1193, + "step": 33592 + }, + { + "epoch": 0.79, + "grad_norm": 1.8825892501807784, + "learning_rate": 2.19618257778949e-06, + "loss": 0.9247, + "step": 33593 + }, + { + "epoch": 0.79, + "grad_norm": 1.977729702461234, + "learning_rate": 2.1957054683421818e-06, + "loss": 0.9843, + "step": 33594 + }, + { + "epoch": 0.79, + "grad_norm": 1.8464268066681584, + "learning_rate": 2.1952284043339443e-06, + "loss": 1.0428, + "step": 33595 + }, + { + "epoch": 0.79, + "grad_norm": 1.8620549013445609, + "learning_rate": 2.194751385767561e-06, + "loss": 0.922, + "step": 33596 + }, + { + "epoch": 0.79, + "grad_norm": 2.056656564907059, + "learning_rate": 2.1942744126458027e-06, + "loss": 0.9808, + "step": 33597 + }, + { + "epoch": 0.79, + "grad_norm": 2.157037979623748, + "learning_rate": 2.1937974849714505e-06, + "loss": 0.9389, + "step": 33598 + }, + { + "epoch": 0.79, + "grad_norm": 2.001382692525841, + "learning_rate": 2.193320602747283e-06, + "loss": 0.9135, + "step": 33599 + }, + { + "epoch": 0.79, + "grad_norm": 1.9423862406780472, + "learning_rate": 2.192843765976075e-06, + "loss": 1.0483, + "step": 33600 + }, + { + "epoch": 0.79, + "grad_norm": 2.1119635409437794, + "learning_rate": 2.1923669746606004e-06, + "loss": 1.0038, + "step": 33601 + }, + { + "epoch": 0.79, + "grad_norm": 1.8378295429692522, + "learning_rate": 2.191890228803636e-06, + "loss": 1.0103, + "step": 33602 + }, + { + "epoch": 0.79, + "grad_norm": 1.946422250607672, + "learning_rate": 2.191413528407964e-06, + "loss": 1.0139, + "step": 33603 + }, + { + "epoch": 0.79, + "grad_norm": 2.0896959885384465, + "learning_rate": 2.1909368734763505e-06, + "loss": 1.1448, + "step": 33604 + }, + { + "epoch": 0.79, + "grad_norm": 1.6958631101215145, + "learning_rate": 2.1904602640115733e-06, + "loss": 1.0444, + "step": 33605 + }, + { + "epoch": 0.79, + "grad_norm": 1.9259316085137377, + "learning_rate": 2.189983700016409e-06, + "loss": 1.0444, + "step": 33606 + }, + { + "epoch": 0.79, + "grad_norm": 1.8195734581826266, + "learning_rate": 2.1895071814936386e-06, + "loss": 1.0422, + "step": 33607 + }, + { + "epoch": 0.79, + "grad_norm": 3.4247652036736573, + "learning_rate": 2.189030708446024e-06, + "loss": 0.9508, + "step": 33608 + }, + { + "epoch": 0.79, + "grad_norm": 1.962307953646719, + "learning_rate": 2.188554280876346e-06, + "loss": 1.0042, + "step": 33609 + }, + { + "epoch": 0.79, + "grad_norm": 1.9016209336544014, + "learning_rate": 2.1880778987873806e-06, + "loss": 0.9953, + "step": 33610 + }, + { + "epoch": 0.79, + "grad_norm": 1.8103772064473753, + "learning_rate": 2.1876015621818956e-06, + "loss": 0.9833, + "step": 33611 + }, + { + "epoch": 0.79, + "grad_norm": 1.9161191546810539, + "learning_rate": 2.187125271062671e-06, + "loss": 1.0942, + "step": 33612 + }, + { + "epoch": 0.79, + "grad_norm": 1.0470268697368434, + "learning_rate": 2.1866490254324745e-06, + "loss": 0.8628, + "step": 33613 + }, + { + "epoch": 0.79, + "grad_norm": 2.0597482307561608, + "learning_rate": 2.186172825294083e-06, + "loss": 1.0118, + "step": 33614 + }, + { + "epoch": 0.79, + "grad_norm": 2.125074744224839, + "learning_rate": 2.1856966706502647e-06, + "loss": 0.936, + "step": 33615 + }, + { + "epoch": 0.79, + "grad_norm": 2.006834968859028, + "learning_rate": 2.1852205615037968e-06, + "loss": 0.9814, + "step": 33616 + }, + { + "epoch": 0.79, + "grad_norm": 1.8380639466277549, + "learning_rate": 2.1847444978574485e-06, + "loss": 1.019, + "step": 33617 + }, + { + "epoch": 0.79, + "grad_norm": 1.0625650444264085, + "learning_rate": 2.18426847971399e-06, + "loss": 1.0243, + "step": 33618 + }, + { + "epoch": 0.79, + "grad_norm": 1.1199111041123888, + "learning_rate": 2.1837925070761946e-06, + "loss": 0.9656, + "step": 33619 + }, + { + "epoch": 0.79, + "grad_norm": 1.0051920963281933, + "learning_rate": 2.183316579946836e-06, + "loss": 0.8955, + "step": 33620 + }, + { + "epoch": 0.79, + "grad_norm": 1.8398364333571189, + "learning_rate": 2.1828406983286832e-06, + "loss": 0.9003, + "step": 33621 + }, + { + "epoch": 0.79, + "grad_norm": 2.0979800602610235, + "learning_rate": 2.1823648622245043e-06, + "loss": 0.939, + "step": 33622 + }, + { + "epoch": 0.79, + "grad_norm": 2.0621727502392107, + "learning_rate": 2.181889071637072e-06, + "loss": 0.9944, + "step": 33623 + }, + { + "epoch": 0.79, + "grad_norm": 2.0801974358615065, + "learning_rate": 2.1814133265691594e-06, + "loss": 0.9982, + "step": 33624 + }, + { + "epoch": 0.79, + "grad_norm": 2.1479633043152098, + "learning_rate": 2.1809376270235348e-06, + "loss": 0.9946, + "step": 33625 + }, + { + "epoch": 0.79, + "grad_norm": 2.0268748780890156, + "learning_rate": 2.180461973002963e-06, + "loss": 1.0272, + "step": 33626 + }, + { + "epoch": 0.79, + "grad_norm": 2.1564047177950547, + "learning_rate": 2.179986364510217e-06, + "loss": 0.9342, + "step": 33627 + }, + { + "epoch": 0.79, + "grad_norm": 1.09929479978653, + "learning_rate": 2.1795108015480694e-06, + "loss": 0.8945, + "step": 33628 + }, + { + "epoch": 0.79, + "grad_norm": 2.101977524494368, + "learning_rate": 2.1790352841192863e-06, + "loss": 1.0036, + "step": 33629 + }, + { + "epoch": 0.79, + "grad_norm": 2.164512320167543, + "learning_rate": 2.1785598122266327e-06, + "loss": 0.8419, + "step": 33630 + }, + { + "epoch": 0.79, + "grad_norm": 1.8379942464957175, + "learning_rate": 2.178084385872882e-06, + "loss": 0.8897, + "step": 33631 + }, + { + "epoch": 0.79, + "grad_norm": 3.365163101896122, + "learning_rate": 2.177609005060799e-06, + "loss": 0.9819, + "step": 33632 + }, + { + "epoch": 0.79, + "grad_norm": 2.0564138045357714, + "learning_rate": 2.1771336697931544e-06, + "loss": 0.9572, + "step": 33633 + }, + { + "epoch": 0.79, + "grad_norm": 1.902378353914507, + "learning_rate": 2.1766583800727128e-06, + "loss": 0.9446, + "step": 33634 + }, + { + "epoch": 0.79, + "grad_norm": 1.9407892897673509, + "learning_rate": 2.1761831359022446e-06, + "loss": 0.9511, + "step": 33635 + }, + { + "epoch": 0.79, + "grad_norm": 1.1004838699410153, + "learning_rate": 2.1757079372845125e-06, + "loss": 0.9104, + "step": 33636 + }, + { + "epoch": 0.79, + "grad_norm": 3.444871559297656, + "learning_rate": 2.175232784222289e-06, + "loss": 0.9065, + "step": 33637 + }, + { + "epoch": 0.79, + "grad_norm": 2.003061407167002, + "learning_rate": 2.1747576767183374e-06, + "loss": 0.9488, + "step": 33638 + }, + { + "epoch": 0.79, + "grad_norm": 1.9171141639328102, + "learning_rate": 2.1742826147754213e-06, + "loss": 0.8495, + "step": 33639 + }, + { + "epoch": 0.79, + "grad_norm": 2.021645962607804, + "learning_rate": 2.173807598396309e-06, + "loss": 0.932, + "step": 33640 + }, + { + "epoch": 0.79, + "grad_norm": 2.006325312550602, + "learning_rate": 2.1733326275837708e-06, + "loss": 0.937, + "step": 33641 + }, + { + "epoch": 0.79, + "grad_norm": 2.021290949595189, + "learning_rate": 2.172857702340566e-06, + "loss": 1.0028, + "step": 33642 + }, + { + "epoch": 0.79, + "grad_norm": 1.9299001592834786, + "learning_rate": 2.17238282266946e-06, + "loss": 1.0853, + "step": 33643 + }, + { + "epoch": 0.79, + "grad_norm": 1.991137931258507, + "learning_rate": 2.17190798857322e-06, + "loss": 0.9143, + "step": 33644 + }, + { + "epoch": 0.79, + "grad_norm": 2.0740885051597413, + "learning_rate": 2.1714332000546114e-06, + "loss": 0.8631, + "step": 33645 + }, + { + "epoch": 0.79, + "grad_norm": 2.163574078206252, + "learning_rate": 2.1709584571163976e-06, + "loss": 0.9227, + "step": 33646 + }, + { + "epoch": 0.79, + "grad_norm": 2.204784812311346, + "learning_rate": 2.1704837597613406e-06, + "loss": 1.0493, + "step": 33647 + }, + { + "epoch": 0.79, + "grad_norm": 1.9971233338147172, + "learning_rate": 2.1700091079922047e-06, + "loss": 1.0644, + "step": 33648 + }, + { + "epoch": 0.79, + "grad_norm": 1.9389593030167867, + "learning_rate": 2.1695345018117577e-06, + "loss": 1.0593, + "step": 33649 + }, + { + "epoch": 0.79, + "grad_norm": 2.03430242220836, + "learning_rate": 2.16905994122276e-06, + "loss": 1.0193, + "step": 33650 + }, + { + "epoch": 0.79, + "grad_norm": 2.0068385163056264, + "learning_rate": 2.1685854262279714e-06, + "loss": 1.043, + "step": 33651 + }, + { + "epoch": 0.79, + "grad_norm": 2.274772031445179, + "learning_rate": 2.1681109568301617e-06, + "loss": 0.8141, + "step": 33652 + }, + { + "epoch": 0.79, + "grad_norm": 1.9563138550133983, + "learning_rate": 2.1676365330320857e-06, + "loss": 1.0063, + "step": 33653 + }, + { + "epoch": 0.79, + "grad_norm": 1.0772823110479555, + "learning_rate": 2.167162154836513e-06, + "loss": 0.9356, + "step": 33654 + }, + { + "epoch": 0.79, + "grad_norm": 2.2241720910826035, + "learning_rate": 2.1666878222461986e-06, + "loss": 0.9257, + "step": 33655 + }, + { + "epoch": 0.79, + "grad_norm": 2.0253818272893414, + "learning_rate": 2.1662135352639115e-06, + "loss": 1.0377, + "step": 33656 + }, + { + "epoch": 0.79, + "grad_norm": 2.6624852516940165, + "learning_rate": 2.1657392938924047e-06, + "loss": 1.006, + "step": 33657 + }, + { + "epoch": 0.79, + "grad_norm": 2.0687452953379126, + "learning_rate": 2.165265098134448e-06, + "loss": 0.8951, + "step": 33658 + }, + { + "epoch": 0.79, + "grad_norm": 1.95706065842524, + "learning_rate": 2.1647909479927987e-06, + "loss": 0.9634, + "step": 33659 + }, + { + "epoch": 0.79, + "grad_norm": 1.9298060464377416, + "learning_rate": 2.1643168434702142e-06, + "loss": 1.0325, + "step": 33660 + }, + { + "epoch": 0.79, + "grad_norm": 1.103846517784815, + "learning_rate": 2.1638427845694574e-06, + "loss": 0.9001, + "step": 33661 + }, + { + "epoch": 0.79, + "grad_norm": 2.9801775834929716, + "learning_rate": 2.1633687712932917e-06, + "loss": 0.8745, + "step": 33662 + }, + { + "epoch": 0.79, + "grad_norm": 1.1164439330304874, + "learning_rate": 2.1628948036444743e-06, + "loss": 0.9046, + "step": 33663 + }, + { + "epoch": 0.79, + "grad_norm": 1.7973859221284567, + "learning_rate": 2.162420881625761e-06, + "loss": 0.9039, + "step": 33664 + }, + { + "epoch": 0.79, + "grad_norm": 2.094142246890565, + "learning_rate": 2.161947005239915e-06, + "loss": 0.9598, + "step": 33665 + }, + { + "epoch": 0.79, + "grad_norm": 1.775816182242097, + "learning_rate": 2.1614731744896976e-06, + "loss": 0.8673, + "step": 33666 + }, + { + "epoch": 0.79, + "grad_norm": 1.8816176484236329, + "learning_rate": 2.160999389377866e-06, + "loss": 0.9675, + "step": 33667 + }, + { + "epoch": 0.79, + "grad_norm": 1.876522173949992, + "learning_rate": 2.160525649907175e-06, + "loss": 0.9527, + "step": 33668 + }, + { + "epoch": 0.79, + "grad_norm": 2.0100352684033154, + "learning_rate": 2.1600519560803844e-06, + "loss": 1.1763, + "step": 33669 + }, + { + "epoch": 0.79, + "grad_norm": 1.9320921719061788, + "learning_rate": 2.1595783079002565e-06, + "loss": 0.9637, + "step": 33670 + }, + { + "epoch": 0.79, + "grad_norm": 1.9910234005878928, + "learning_rate": 2.159104705369546e-06, + "loss": 1.07, + "step": 33671 + }, + { + "epoch": 0.79, + "grad_norm": 2.1156525683354466, + "learning_rate": 2.158631148491007e-06, + "loss": 1.063, + "step": 33672 + }, + { + "epoch": 0.79, + "grad_norm": 1.9413535033703397, + "learning_rate": 2.158157637267404e-06, + "loss": 0.9554, + "step": 33673 + }, + { + "epoch": 0.79, + "grad_norm": 2.268619670873208, + "learning_rate": 2.1576841717014864e-06, + "loss": 0.9732, + "step": 33674 + }, + { + "epoch": 0.79, + "grad_norm": 1.9742901663850962, + "learning_rate": 2.1572107517960174e-06, + "loss": 0.9799, + "step": 33675 + }, + { + "epoch": 0.79, + "grad_norm": 1.7683634458437711, + "learning_rate": 2.156737377553748e-06, + "loss": 0.8945, + "step": 33676 + }, + { + "epoch": 0.79, + "grad_norm": 2.004197648469767, + "learning_rate": 2.1562640489774388e-06, + "loss": 0.9994, + "step": 33677 + }, + { + "epoch": 0.79, + "grad_norm": 1.878398999100291, + "learning_rate": 2.1557907660698408e-06, + "loss": 0.9926, + "step": 33678 + }, + { + "epoch": 0.79, + "grad_norm": 1.9406005751610174, + "learning_rate": 2.155317528833716e-06, + "loss": 1.0813, + "step": 33679 + }, + { + "epoch": 0.79, + "grad_norm": 2.0337416369038204, + "learning_rate": 2.1548443372718153e-06, + "loss": 0.9693, + "step": 33680 + }, + { + "epoch": 0.79, + "grad_norm": 1.8796683238641578, + "learning_rate": 2.1543711913868915e-06, + "loss": 0.9681, + "step": 33681 + }, + { + "epoch": 0.79, + "grad_norm": 3.9537888367861895, + "learning_rate": 2.153898091181703e-06, + "loss": 0.9612, + "step": 33682 + }, + { + "epoch": 0.79, + "grad_norm": 1.9690924981153104, + "learning_rate": 2.153425036659007e-06, + "loss": 0.9137, + "step": 33683 + }, + { + "epoch": 0.79, + "grad_norm": 2.0633951486782984, + "learning_rate": 2.1529520278215543e-06, + "loss": 0.9181, + "step": 33684 + }, + { + "epoch": 0.79, + "grad_norm": 2.0783479028067258, + "learning_rate": 2.152479064672096e-06, + "loss": 0.9862, + "step": 33685 + }, + { + "epoch": 0.79, + "grad_norm": 1.8364371698381887, + "learning_rate": 2.1520061472133903e-06, + "loss": 0.9181, + "step": 33686 + }, + { + "epoch": 0.79, + "grad_norm": 1.9198343759343048, + "learning_rate": 2.151533275448191e-06, + "loss": 0.9819, + "step": 33687 + }, + { + "epoch": 0.79, + "grad_norm": 2.6923347179322112, + "learning_rate": 2.151060449379251e-06, + "loss": 1.0161, + "step": 33688 + }, + { + "epoch": 0.79, + "grad_norm": 2.133758506294239, + "learning_rate": 2.1505876690093176e-06, + "loss": 1.0028, + "step": 33689 + }, + { + "epoch": 0.79, + "grad_norm": 2.1704796962303767, + "learning_rate": 2.1501149343411486e-06, + "loss": 0.9927, + "step": 33690 + }, + { + "epoch": 0.79, + "grad_norm": 1.8348894554515591, + "learning_rate": 2.149642245377499e-06, + "loss": 0.9809, + "step": 33691 + }, + { + "epoch": 0.79, + "grad_norm": 1.7087349231406876, + "learning_rate": 2.1491696021211174e-06, + "loss": 0.9281, + "step": 33692 + }, + { + "epoch": 0.79, + "grad_norm": 2.0226116633096205, + "learning_rate": 2.1486970045747525e-06, + "loss": 0.8935, + "step": 33693 + }, + { + "epoch": 0.79, + "grad_norm": 1.8870481736979787, + "learning_rate": 2.148224452741162e-06, + "loss": 0.9352, + "step": 33694 + }, + { + "epoch": 0.79, + "grad_norm": 2.2814888031041396, + "learning_rate": 2.1477519466230935e-06, + "loss": 1.0171, + "step": 33695 + }, + { + "epoch": 0.79, + "grad_norm": 1.8392936567928635, + "learning_rate": 2.147279486223297e-06, + "loss": 1.1438, + "step": 33696 + }, + { + "epoch": 0.79, + "grad_norm": 1.0247303502394138, + "learning_rate": 2.14680707154453e-06, + "loss": 0.931, + "step": 33697 + }, + { + "epoch": 0.79, + "grad_norm": 1.740983394085776, + "learning_rate": 2.1463347025895375e-06, + "loss": 1.0802, + "step": 33698 + }, + { + "epoch": 0.79, + "grad_norm": 2.5038228116847536, + "learning_rate": 2.1458623793610678e-06, + "loss": 1.1037, + "step": 33699 + }, + { + "epoch": 0.79, + "grad_norm": 1.9178595869348576, + "learning_rate": 2.145390101861875e-06, + "loss": 0.9351, + "step": 33700 + }, + { + "epoch": 0.79, + "grad_norm": 1.9868418346679981, + "learning_rate": 2.1449178700947125e-06, + "loss": 0.9679, + "step": 33701 + }, + { + "epoch": 0.79, + "grad_norm": 2.0376017543258795, + "learning_rate": 2.1444456840623195e-06, + "loss": 0.9868, + "step": 33702 + }, + { + "epoch": 0.79, + "grad_norm": 1.9645834002816316, + "learning_rate": 2.1439735437674504e-06, + "loss": 0.9129, + "step": 33703 + }, + { + "epoch": 0.79, + "grad_norm": 1.9085642927717401, + "learning_rate": 2.1435014492128547e-06, + "loss": 0.9598, + "step": 33704 + }, + { + "epoch": 0.79, + "grad_norm": 1.794528304474605, + "learning_rate": 2.1430294004012874e-06, + "loss": 0.8307, + "step": 33705 + }, + { + "epoch": 0.79, + "grad_norm": 1.8362889072760025, + "learning_rate": 2.142557397335484e-06, + "loss": 0.9932, + "step": 33706 + }, + { + "epoch": 0.79, + "grad_norm": 1.1251580228562792, + "learning_rate": 2.142085440018199e-06, + "loss": 0.9538, + "step": 33707 + }, + { + "epoch": 0.79, + "grad_norm": 2.265067252152446, + "learning_rate": 2.1416135284521812e-06, + "loss": 0.9151, + "step": 33708 + }, + { + "epoch": 0.79, + "grad_norm": 2.1326944687299445, + "learning_rate": 2.1411416626401793e-06, + "loss": 1.0103, + "step": 33709 + }, + { + "epoch": 0.79, + "grad_norm": 2.3214013679733507, + "learning_rate": 2.1406698425849383e-06, + "loss": 1.0759, + "step": 33710 + }, + { + "epoch": 0.79, + "grad_norm": 2.1255461654037267, + "learning_rate": 2.140198068289204e-06, + "loss": 0.9057, + "step": 33711 + }, + { + "epoch": 0.79, + "grad_norm": 1.9296887380628824, + "learning_rate": 2.1397263397557266e-06, + "loss": 1.0023, + "step": 33712 + }, + { + "epoch": 0.79, + "grad_norm": 1.9461445428768622, + "learning_rate": 2.1392546569872498e-06, + "loss": 1.0583, + "step": 33713 + }, + { + "epoch": 0.79, + "grad_norm": 2.149381002330483, + "learning_rate": 2.1387830199865224e-06, + "loss": 0.9602, + "step": 33714 + }, + { + "epoch": 0.79, + "grad_norm": 2.327999595150438, + "learning_rate": 2.138311428756287e-06, + "loss": 1.1684, + "step": 33715 + }, + { + "epoch": 0.79, + "grad_norm": 2.1340929042005192, + "learning_rate": 2.137839883299294e-06, + "loss": 1.0766, + "step": 33716 + }, + { + "epoch": 0.79, + "grad_norm": 2.149491718177176, + "learning_rate": 2.1373683836182835e-06, + "loss": 1.0237, + "step": 33717 + }, + { + "epoch": 0.79, + "grad_norm": 2.257031196861293, + "learning_rate": 2.136896929716006e-06, + "loss": 0.8441, + "step": 33718 + }, + { + "epoch": 0.79, + "grad_norm": 1.8455950169212092, + "learning_rate": 2.1364255215952047e-06, + "loss": 0.9611, + "step": 33719 + }, + { + "epoch": 0.79, + "grad_norm": 1.6547700502640963, + "learning_rate": 2.1359541592586197e-06, + "loss": 0.9601, + "step": 33720 + }, + { + "epoch": 0.79, + "grad_norm": 1.8769560321662293, + "learning_rate": 2.1354828427090004e-06, + "loss": 1.0231, + "step": 33721 + }, + { + "epoch": 0.79, + "grad_norm": 1.8935595668977743, + "learning_rate": 2.135011571949093e-06, + "loss": 1.0829, + "step": 33722 + }, + { + "epoch": 0.79, + "grad_norm": 2.101908935962509, + "learning_rate": 2.134540346981637e-06, + "loss": 1.0918, + "step": 33723 + }, + { + "epoch": 0.79, + "grad_norm": 2.1853076040111192, + "learning_rate": 2.134069167809375e-06, + "loss": 1.0115, + "step": 33724 + }, + { + "epoch": 0.79, + "grad_norm": 1.7242400761129362, + "learning_rate": 2.1335980344350527e-06, + "loss": 1.0378, + "step": 33725 + }, + { + "epoch": 0.79, + "grad_norm": 1.9040958731281938, + "learning_rate": 2.1331269468614156e-06, + "loss": 0.9115, + "step": 33726 + }, + { + "epoch": 0.79, + "grad_norm": 2.241013729544482, + "learning_rate": 2.1326559050912042e-06, + "loss": 0.8518, + "step": 33727 + }, + { + "epoch": 0.79, + "grad_norm": 1.9019394777667817, + "learning_rate": 2.1321849091271573e-06, + "loss": 0.961, + "step": 33728 + }, + { + "epoch": 0.79, + "grad_norm": 2.168665599735245, + "learning_rate": 2.1317139589720227e-06, + "loss": 0.9758, + "step": 33729 + }, + { + "epoch": 0.79, + "grad_norm": 2.25700097675263, + "learning_rate": 2.1312430546285414e-06, + "loss": 0.9553, + "step": 33730 + }, + { + "epoch": 0.79, + "grad_norm": 2.533808760333773, + "learning_rate": 2.1307721960994556e-06, + "loss": 1.0359, + "step": 33731 + }, + { + "epoch": 0.79, + "grad_norm": 1.9704507120512031, + "learning_rate": 2.130301383387502e-06, + "loss": 0.9559, + "step": 33732 + }, + { + "epoch": 0.79, + "grad_norm": 1.814128583780272, + "learning_rate": 2.1298306164954275e-06, + "loss": 0.9336, + "step": 33733 + }, + { + "epoch": 0.79, + "grad_norm": 2.150215930711659, + "learning_rate": 2.129359895425969e-06, + "loss": 0.9391, + "step": 33734 + }, + { + "epoch": 0.79, + "grad_norm": 1.8139651519292515, + "learning_rate": 2.128889220181871e-06, + "loss": 0.9376, + "step": 33735 + }, + { + "epoch": 0.79, + "grad_norm": 1.8679330051538803, + "learning_rate": 2.1284185907658695e-06, + "loss": 0.933, + "step": 33736 + }, + { + "epoch": 0.79, + "grad_norm": 2.246511707913214, + "learning_rate": 2.12794800718071e-06, + "loss": 0.9763, + "step": 33737 + }, + { + "epoch": 0.79, + "grad_norm": 1.942664922643041, + "learning_rate": 2.1274774694291267e-06, + "loss": 1.0017, + "step": 33738 + }, + { + "epoch": 0.79, + "grad_norm": 2.1738300324013493, + "learning_rate": 2.1270069775138647e-06, + "loss": 1.0401, + "step": 33739 + }, + { + "epoch": 0.79, + "grad_norm": 2.0300501580350674, + "learning_rate": 2.1265365314376598e-06, + "loss": 1.0514, + "step": 33740 + }, + { + "epoch": 0.79, + "grad_norm": 2.2241125446871313, + "learning_rate": 2.1260661312032506e-06, + "loss": 1.0944, + "step": 33741 + }, + { + "epoch": 0.79, + "grad_norm": 1.9324315689392026, + "learning_rate": 2.125595776813376e-06, + "loss": 0.9892, + "step": 33742 + }, + { + "epoch": 0.79, + "grad_norm": 1.9338229589002605, + "learning_rate": 2.125125468270779e-06, + "loss": 0.9693, + "step": 33743 + }, + { + "epoch": 0.79, + "grad_norm": 2.070670027945993, + "learning_rate": 2.124655205578194e-06, + "loss": 1.0056, + "step": 33744 + }, + { + "epoch": 0.8, + "grad_norm": 2.0710163044776393, + "learning_rate": 2.1241849887383583e-06, + "loss": 0.9632, + "step": 33745 + }, + { + "epoch": 0.8, + "grad_norm": 2.1384903799192974, + "learning_rate": 2.12371481775401e-06, + "loss": 0.9855, + "step": 33746 + }, + { + "epoch": 0.8, + "grad_norm": 2.290331029479017, + "learning_rate": 2.1232446926278905e-06, + "loss": 0.8544, + "step": 33747 + }, + { + "epoch": 0.8, + "grad_norm": 1.843294042847009, + "learning_rate": 2.1227746133627346e-06, + "loss": 1.0211, + "step": 33748 + }, + { + "epoch": 0.8, + "grad_norm": 1.9318744768099878, + "learning_rate": 2.1223045799612763e-06, + "loss": 0.9978, + "step": 33749 + }, + { + "epoch": 0.8, + "grad_norm": 1.7838422866527797, + "learning_rate": 2.121834592426254e-06, + "loss": 0.8421, + "step": 33750 + }, + { + "epoch": 0.8, + "grad_norm": 2.4394704603933777, + "learning_rate": 2.121364650760408e-06, + "loss": 0.9399, + "step": 33751 + }, + { + "epoch": 0.8, + "grad_norm": 1.1188698691994645, + "learning_rate": 2.1208947549664706e-06, + "loss": 0.9358, + "step": 33752 + }, + { + "epoch": 0.8, + "grad_norm": 2.062117736921936, + "learning_rate": 2.120424905047176e-06, + "loss": 1.0156, + "step": 33753 + }, + { + "epoch": 0.8, + "grad_norm": 2.047243706263411, + "learning_rate": 2.1199551010052653e-06, + "loss": 0.8884, + "step": 33754 + }, + { + "epoch": 0.8, + "grad_norm": 2.226728988161776, + "learning_rate": 2.119485342843468e-06, + "loss": 0.9991, + "step": 33755 + }, + { + "epoch": 0.8, + "grad_norm": 2.0183737899379217, + "learning_rate": 2.1190156305645247e-06, + "loss": 0.9964, + "step": 33756 + }, + { + "epoch": 0.8, + "grad_norm": 2.12317453762106, + "learning_rate": 2.1185459641711646e-06, + "loss": 1.0266, + "step": 33757 + }, + { + "epoch": 0.8, + "grad_norm": 2.256371603543352, + "learning_rate": 2.118076343666129e-06, + "loss": 1.048, + "step": 33758 + }, + { + "epoch": 0.8, + "grad_norm": 2.0104089289857683, + "learning_rate": 2.117606769052144e-06, + "loss": 0.9488, + "step": 33759 + }, + { + "epoch": 0.8, + "grad_norm": 2.2910911623013606, + "learning_rate": 2.1171372403319514e-06, + "loss": 1.0181, + "step": 33760 + }, + { + "epoch": 0.8, + "grad_norm": 2.0777348910939075, + "learning_rate": 2.116667757508282e-06, + "loss": 1.0702, + "step": 33761 + }, + { + "epoch": 0.8, + "grad_norm": 1.942599710712364, + "learning_rate": 2.1161983205838645e-06, + "loss": 1.0155, + "step": 33762 + }, + { + "epoch": 0.8, + "grad_norm": 2.214506472724554, + "learning_rate": 2.115728929561437e-06, + "loss": 1.0368, + "step": 33763 + }, + { + "epoch": 0.8, + "grad_norm": 2.4597802968501634, + "learning_rate": 2.115259584443735e-06, + "loss": 0.9668, + "step": 33764 + }, + { + "epoch": 0.8, + "grad_norm": 2.096668091574461, + "learning_rate": 2.114790285233488e-06, + "loss": 0.7908, + "step": 33765 + }, + { + "epoch": 0.8, + "grad_norm": 2.1307341966521673, + "learning_rate": 2.114321031933425e-06, + "loss": 0.9056, + "step": 33766 + }, + { + "epoch": 0.8, + "grad_norm": 2.1884427756348805, + "learning_rate": 2.1138518245462816e-06, + "loss": 0.913, + "step": 33767 + }, + { + "epoch": 0.8, + "grad_norm": 2.137913939925042, + "learning_rate": 2.113382663074792e-06, + "loss": 0.9911, + "step": 33768 + }, + { + "epoch": 0.8, + "grad_norm": 2.2015050410198533, + "learning_rate": 2.1129135475216866e-06, + "loss": 0.8692, + "step": 33769 + }, + { + "epoch": 0.8, + "grad_norm": 2.810118171829221, + "learning_rate": 2.1124444778896914e-06, + "loss": 0.8693, + "step": 33770 + }, + { + "epoch": 0.8, + "grad_norm": 2.1474663225655477, + "learning_rate": 2.111975454181543e-06, + "loss": 1.0413, + "step": 33771 + }, + { + "epoch": 0.8, + "grad_norm": 1.1735018432992135, + "learning_rate": 2.1115064763999726e-06, + "loss": 0.9267, + "step": 33772 + }, + { + "epoch": 0.8, + "grad_norm": 2.729666614247859, + "learning_rate": 2.1110375445477093e-06, + "loss": 1.0454, + "step": 33773 + }, + { + "epoch": 0.8, + "grad_norm": 1.945756132865226, + "learning_rate": 2.1105686586274797e-06, + "loss": 1.0118, + "step": 33774 + }, + { + "epoch": 0.8, + "grad_norm": 2.025500414239857, + "learning_rate": 2.110099818642021e-06, + "loss": 0.9021, + "step": 33775 + }, + { + "epoch": 0.8, + "grad_norm": 2.182796591441005, + "learning_rate": 2.1096310245940555e-06, + "loss": 0.9773, + "step": 33776 + }, + { + "epoch": 0.8, + "grad_norm": 2.0349262150102234, + "learning_rate": 2.1091622764863185e-06, + "loss": 0.9916, + "step": 33777 + }, + { + "epoch": 0.8, + "grad_norm": 1.9275651935910227, + "learning_rate": 2.1086935743215354e-06, + "loss": 0.9852, + "step": 33778 + }, + { + "epoch": 0.8, + "grad_norm": 2.1967882472791724, + "learning_rate": 2.1082249181024384e-06, + "loss": 0.9241, + "step": 33779 + }, + { + "epoch": 0.8, + "grad_norm": 1.8876617424215352, + "learning_rate": 2.107756307831752e-06, + "loss": 0.8712, + "step": 33780 + }, + { + "epoch": 0.8, + "grad_norm": 1.117362027107475, + "learning_rate": 2.1072877435122095e-06, + "loss": 0.9458, + "step": 33781 + }, + { + "epoch": 0.8, + "grad_norm": 1.8492028364633792, + "learning_rate": 2.106819225146538e-06, + "loss": 0.9397, + "step": 33782 + }, + { + "epoch": 0.8, + "grad_norm": 2.097126389247831, + "learning_rate": 2.1063507527374594e-06, + "loss": 0.9943, + "step": 33783 + }, + { + "epoch": 0.8, + "grad_norm": 2.003293653108825, + "learning_rate": 2.105882326287707e-06, + "loss": 1.0616, + "step": 33784 + }, + { + "epoch": 0.8, + "grad_norm": 1.9572834151765977, + "learning_rate": 2.105413945800007e-06, + "loss": 0.9847, + "step": 33785 + }, + { + "epoch": 0.8, + "grad_norm": 1.1457727695754956, + "learning_rate": 2.1049456112770907e-06, + "loss": 0.9252, + "step": 33786 + }, + { + "epoch": 0.8, + "grad_norm": 2.2552009311034023, + "learning_rate": 2.104477322721675e-06, + "loss": 0.9997, + "step": 33787 + }, + { + "epoch": 0.8, + "grad_norm": 2.0054018524714525, + "learning_rate": 2.104009080136493e-06, + "loss": 0.9151, + "step": 33788 + }, + { + "epoch": 0.8, + "grad_norm": 1.1094636617241702, + "learning_rate": 2.103540883524272e-06, + "loss": 0.917, + "step": 33789 + }, + { + "epoch": 0.8, + "grad_norm": 1.955298340208445, + "learning_rate": 2.1030727328877367e-06, + "loss": 0.9799, + "step": 33790 + }, + { + "epoch": 0.8, + "grad_norm": 1.938275081183328, + "learning_rate": 2.1026046282296085e-06, + "loss": 0.9496, + "step": 33791 + }, + { + "epoch": 0.8, + "grad_norm": 2.0489651860974787, + "learning_rate": 2.1021365695526176e-06, + "loss": 0.982, + "step": 33792 + }, + { + "epoch": 0.8, + "grad_norm": 2.1483594775120967, + "learning_rate": 2.10166855685949e-06, + "loss": 1.0607, + "step": 33793 + }, + { + "epoch": 0.8, + "grad_norm": 2.1379593166736717, + "learning_rate": 2.1012005901529463e-06, + "loss": 1.0361, + "step": 33794 + }, + { + "epoch": 0.8, + "grad_norm": 2.0093286143884574, + "learning_rate": 2.1007326694357153e-06, + "loss": 0.8419, + "step": 33795 + }, + { + "epoch": 0.8, + "grad_norm": 1.129553714992969, + "learning_rate": 2.1002647947105205e-06, + "loss": 0.9325, + "step": 33796 + }, + { + "epoch": 0.8, + "grad_norm": 1.9200900339855487, + "learning_rate": 2.099796965980082e-06, + "loss": 0.9156, + "step": 33797 + }, + { + "epoch": 0.8, + "grad_norm": 1.9634633538243842, + "learning_rate": 2.099329183247126e-06, + "loss": 0.8847, + "step": 33798 + }, + { + "epoch": 0.8, + "grad_norm": 1.737745025924161, + "learning_rate": 2.0988614465143807e-06, + "loss": 0.9214, + "step": 33799 + }, + { + "epoch": 0.8, + "grad_norm": 2.0788546741935594, + "learning_rate": 2.0983937557845647e-06, + "loss": 1.0222, + "step": 33800 + }, + { + "epoch": 0.8, + "grad_norm": 1.1090679945439141, + "learning_rate": 2.0979261110604e-06, + "loss": 0.9783, + "step": 33801 + }, + { + "epoch": 0.8, + "grad_norm": 1.9614944656089504, + "learning_rate": 2.09745851234461e-06, + "loss": 1.0116, + "step": 33802 + }, + { + "epoch": 0.8, + "grad_norm": 3.5216710593364504, + "learning_rate": 2.0969909596399253e-06, + "loss": 0.9731, + "step": 33803 + }, + { + "epoch": 0.8, + "grad_norm": 1.1280019858750399, + "learning_rate": 2.0965234529490552e-06, + "loss": 0.9446, + "step": 33804 + }, + { + "epoch": 0.8, + "grad_norm": 1.9330999946902152, + "learning_rate": 2.096055992274727e-06, + "loss": 0.9929, + "step": 33805 + }, + { + "epoch": 0.8, + "grad_norm": 2.0681386801523574, + "learning_rate": 2.0955885776196637e-06, + "loss": 1.0256, + "step": 33806 + }, + { + "epoch": 0.8, + "grad_norm": 1.9378920738116374, + "learning_rate": 2.09512120898659e-06, + "loss": 0.9904, + "step": 33807 + }, + { + "epoch": 0.8, + "grad_norm": 2.294605151651059, + "learning_rate": 2.0946538863782216e-06, + "loss": 0.9829, + "step": 33808 + }, + { + "epoch": 0.8, + "grad_norm": 1.9760351743321356, + "learning_rate": 2.0941866097972797e-06, + "loss": 0.8924, + "step": 33809 + }, + { + "epoch": 0.8, + "grad_norm": 2.1216719163876823, + "learning_rate": 2.0937193792464883e-06, + "loss": 0.9149, + "step": 33810 + }, + { + "epoch": 0.8, + "grad_norm": 2.17888478569108, + "learning_rate": 2.0932521947285635e-06, + "loss": 1.1216, + "step": 33811 + }, + { + "epoch": 0.8, + "grad_norm": 2.043484442081529, + "learning_rate": 2.092785056246229e-06, + "loss": 0.9854, + "step": 33812 + }, + { + "epoch": 0.8, + "grad_norm": 3.0163741123643897, + "learning_rate": 2.092317963802202e-06, + "loss": 0.927, + "step": 33813 + }, + { + "epoch": 0.8, + "grad_norm": 2.102650637994938, + "learning_rate": 2.091850917399206e-06, + "loss": 1.0933, + "step": 33814 + }, + { + "epoch": 0.8, + "grad_norm": 1.9209622293496809, + "learning_rate": 2.091383917039954e-06, + "loss": 0.9447, + "step": 33815 + }, + { + "epoch": 0.8, + "grad_norm": 1.0762904927398558, + "learning_rate": 2.090916962727172e-06, + "loss": 0.9111, + "step": 33816 + }, + { + "epoch": 0.8, + "grad_norm": 1.8905906118711788, + "learning_rate": 2.0904500544635752e-06, + "loss": 1.0796, + "step": 33817 + }, + { + "epoch": 0.8, + "grad_norm": 2.1281793111161833, + "learning_rate": 2.08998319225188e-06, + "loss": 0.8943, + "step": 33818 + }, + { + "epoch": 0.8, + "grad_norm": 1.8762905988908722, + "learning_rate": 2.089516376094807e-06, + "loss": 0.9286, + "step": 33819 + }, + { + "epoch": 0.8, + "grad_norm": 1.8625692564443668, + "learning_rate": 2.089049605995076e-06, + "loss": 0.9755, + "step": 33820 + }, + { + "epoch": 0.8, + "grad_norm": 1.8725704050997132, + "learning_rate": 2.0885828819554035e-06, + "loss": 0.9546, + "step": 33821 + }, + { + "epoch": 0.8, + "grad_norm": 2.1228425669867774, + "learning_rate": 2.088116203978503e-06, + "loss": 1.0507, + "step": 33822 + }, + { + "epoch": 0.8, + "grad_norm": 2.4526116376802314, + "learning_rate": 2.087649572067095e-06, + "loss": 0.8901, + "step": 33823 + }, + { + "epoch": 0.8, + "grad_norm": 1.9061225778661541, + "learning_rate": 2.087182986223899e-06, + "loss": 0.9171, + "step": 33824 + }, + { + "epoch": 0.8, + "grad_norm": 2.1381796206114854, + "learning_rate": 2.0867164464516287e-06, + "loss": 0.9212, + "step": 33825 + }, + { + "epoch": 0.8, + "grad_norm": 2.242245854680052, + "learning_rate": 2.086249952752998e-06, + "loss": 0.94, + "step": 33826 + }, + { + "epoch": 0.8, + "grad_norm": 1.8074920242323296, + "learning_rate": 2.0857835051307253e-06, + "loss": 0.9102, + "step": 33827 + }, + { + "epoch": 0.8, + "grad_norm": 2.1062882606477724, + "learning_rate": 2.0853171035875296e-06, + "loss": 1.0244, + "step": 33828 + }, + { + "epoch": 0.8, + "grad_norm": 2.2360125458533773, + "learning_rate": 2.084850748126124e-06, + "loss": 0.7951, + "step": 33829 + }, + { + "epoch": 0.8, + "grad_norm": 2.086724671178602, + "learning_rate": 2.0843844387492197e-06, + "loss": 1.0143, + "step": 33830 + }, + { + "epoch": 0.8, + "grad_norm": 2.226953946283059, + "learning_rate": 2.083918175459535e-06, + "loss": 1.07, + "step": 33831 + }, + { + "epoch": 0.8, + "grad_norm": 1.9501278921239302, + "learning_rate": 2.083451958259788e-06, + "loss": 0.8969, + "step": 33832 + }, + { + "epoch": 0.8, + "grad_norm": 1.9314833986643407, + "learning_rate": 2.0829857871526904e-06, + "loss": 0.8967, + "step": 33833 + }, + { + "epoch": 0.8, + "grad_norm": 1.0566758508680643, + "learning_rate": 2.0825196621409526e-06, + "loss": 0.9056, + "step": 33834 + }, + { + "epoch": 0.8, + "grad_norm": 2.206782523419595, + "learning_rate": 2.0820535832272947e-06, + "loss": 0.8948, + "step": 33835 + }, + { + "epoch": 0.8, + "grad_norm": 1.0362545429767973, + "learning_rate": 2.0815875504144257e-06, + "loss": 0.9467, + "step": 33836 + }, + { + "epoch": 0.8, + "grad_norm": 2.1117839697182466, + "learning_rate": 2.081121563705063e-06, + "loss": 0.9705, + "step": 33837 + }, + { + "epoch": 0.8, + "grad_norm": 2.2572668693538587, + "learning_rate": 2.080655623101915e-06, + "loss": 1.097, + "step": 33838 + }, + { + "epoch": 0.8, + "grad_norm": 1.9534425176973456, + "learning_rate": 2.0801897286077e-06, + "loss": 1.0493, + "step": 33839 + }, + { + "epoch": 0.8, + "grad_norm": 1.7470086702812806, + "learning_rate": 2.0797238802251252e-06, + "loss": 0.933, + "step": 33840 + }, + { + "epoch": 0.8, + "grad_norm": 1.8782897624843227, + "learning_rate": 2.079258077956907e-06, + "loss": 1.0665, + "step": 33841 + }, + { + "epoch": 0.8, + "grad_norm": 2.043429312067929, + "learning_rate": 2.078792321805757e-06, + "loss": 0.9777, + "step": 33842 + }, + { + "epoch": 0.8, + "grad_norm": 4.988937646768678, + "learning_rate": 2.078326611774384e-06, + "loss": 0.9507, + "step": 33843 + }, + { + "epoch": 0.8, + "grad_norm": 2.2772186605432965, + "learning_rate": 2.0778609478655e-06, + "loss": 0.858, + "step": 33844 + }, + { + "epoch": 0.8, + "grad_norm": 1.1682325699323561, + "learning_rate": 2.0773953300818204e-06, + "loss": 0.9877, + "step": 33845 + }, + { + "epoch": 0.8, + "grad_norm": 2.972103493822949, + "learning_rate": 2.076929758426055e-06, + "loss": 1.0629, + "step": 33846 + }, + { + "epoch": 0.8, + "grad_norm": 1.0425738750662712, + "learning_rate": 2.0764642329009087e-06, + "loss": 0.9922, + "step": 33847 + }, + { + "epoch": 0.8, + "grad_norm": 2.097389629915373, + "learning_rate": 2.0759987535090966e-06, + "loss": 1.0247, + "step": 33848 + }, + { + "epoch": 0.8, + "grad_norm": 1.8754198352472262, + "learning_rate": 2.075533320253331e-06, + "loss": 1.1136, + "step": 33849 + }, + { + "epoch": 0.8, + "grad_norm": 1.9020965604545905, + "learning_rate": 2.07506793313632e-06, + "loss": 0.9642, + "step": 33850 + }, + { + "epoch": 0.8, + "grad_norm": 1.8949695844488958, + "learning_rate": 2.0746025921607695e-06, + "loss": 1.0705, + "step": 33851 + }, + { + "epoch": 0.8, + "grad_norm": 2.5846511557381686, + "learning_rate": 2.0741372973293915e-06, + "loss": 1.121, + "step": 33852 + }, + { + "epoch": 0.8, + "grad_norm": 1.9435957896018312, + "learning_rate": 2.073672048644898e-06, + "loss": 0.9355, + "step": 33853 + }, + { + "epoch": 0.8, + "grad_norm": 1.8875266359682823, + "learning_rate": 2.0732068461099952e-06, + "loss": 0.9549, + "step": 33854 + }, + { + "epoch": 0.8, + "grad_norm": 2.123786269627171, + "learning_rate": 2.07274168972739e-06, + "loss": 0.9633, + "step": 33855 + }, + { + "epoch": 0.8, + "grad_norm": 1.8153268311446744, + "learning_rate": 2.0722765794997947e-06, + "loss": 0.8946, + "step": 33856 + }, + { + "epoch": 0.8, + "grad_norm": 2.1158162301440644, + "learning_rate": 2.0718115154299113e-06, + "loss": 0.991, + "step": 33857 + }, + { + "epoch": 0.8, + "grad_norm": 2.0634834603791457, + "learning_rate": 2.0713464975204556e-06, + "loss": 1.0023, + "step": 33858 + }, + { + "epoch": 0.8, + "grad_norm": 2.040259408600939, + "learning_rate": 2.0708815257741264e-06, + "loss": 0.9725, + "step": 33859 + }, + { + "epoch": 0.8, + "grad_norm": 2.1752495815411397, + "learning_rate": 2.0704166001936387e-06, + "loss": 1.0557, + "step": 33860 + }, + { + "epoch": 0.8, + "grad_norm": 2.17429608869451, + "learning_rate": 2.0699517207816944e-06, + "loss": 0.9604, + "step": 33861 + }, + { + "epoch": 0.8, + "grad_norm": 1.9102618510313767, + "learning_rate": 2.0694868875410036e-06, + "loss": 1.1231, + "step": 33862 + }, + { + "epoch": 0.8, + "grad_norm": 1.9691270574084507, + "learning_rate": 2.0690221004742704e-06, + "loss": 0.924, + "step": 33863 + }, + { + "epoch": 0.8, + "grad_norm": 2.267688872346086, + "learning_rate": 2.068557359584199e-06, + "loss": 0.9756, + "step": 33864 + }, + { + "epoch": 0.8, + "grad_norm": 1.9619840503369101, + "learning_rate": 2.068092664873498e-06, + "loss": 1.051, + "step": 33865 + }, + { + "epoch": 0.8, + "grad_norm": 1.080828158631702, + "learning_rate": 2.067628016344875e-06, + "loss": 0.9351, + "step": 33866 + }, + { + "epoch": 0.8, + "grad_norm": 1.9140716728895668, + "learning_rate": 2.0671634140010333e-06, + "loss": 0.8757, + "step": 33867 + }, + { + "epoch": 0.8, + "grad_norm": 2.0553249956153223, + "learning_rate": 2.0666988578446757e-06, + "loss": 0.9806, + "step": 33868 + }, + { + "epoch": 0.8, + "grad_norm": 1.8786201862146437, + "learning_rate": 2.0662343478785074e-06, + "loss": 1.0121, + "step": 33869 + }, + { + "epoch": 0.8, + "grad_norm": 2.079735315532338, + "learning_rate": 2.0657698841052386e-06, + "loss": 1.0483, + "step": 33870 + }, + { + "epoch": 0.8, + "grad_norm": 1.8809210636197073, + "learning_rate": 2.0653054665275686e-06, + "loss": 0.971, + "step": 33871 + }, + { + "epoch": 0.8, + "grad_norm": 2.0697041456684815, + "learning_rate": 2.0648410951482e-06, + "loss": 0.9209, + "step": 33872 + }, + { + "epoch": 0.8, + "grad_norm": 1.9504005066777748, + "learning_rate": 2.064376769969839e-06, + "loss": 1.0806, + "step": 33873 + }, + { + "epoch": 0.8, + "grad_norm": 1.0257990947672555, + "learning_rate": 2.0639124909951913e-06, + "loss": 0.9821, + "step": 33874 + }, + { + "epoch": 0.8, + "grad_norm": 2.033542428851062, + "learning_rate": 2.0634482582269578e-06, + "loss": 1.0861, + "step": 33875 + }, + { + "epoch": 0.8, + "grad_norm": 2.230267545185512, + "learning_rate": 2.0629840716678384e-06, + "loss": 0.9838, + "step": 33876 + }, + { + "epoch": 0.8, + "grad_norm": 1.8648212741840584, + "learning_rate": 2.0625199313205403e-06, + "loss": 0.8489, + "step": 33877 + }, + { + "epoch": 0.8, + "grad_norm": 1.9133922950229658, + "learning_rate": 2.062055837187763e-06, + "loss": 0.9588, + "step": 33878 + }, + { + "epoch": 0.8, + "grad_norm": 1.7992016082151867, + "learning_rate": 2.061591789272208e-06, + "loss": 0.9423, + "step": 33879 + }, + { + "epoch": 0.8, + "grad_norm": 1.9288927013121082, + "learning_rate": 2.0611277875765835e-06, + "loss": 0.8936, + "step": 33880 + }, + { + "epoch": 0.8, + "grad_norm": 1.8230367606346105, + "learning_rate": 2.060663832103584e-06, + "loss": 1.1331, + "step": 33881 + }, + { + "epoch": 0.8, + "grad_norm": 2.4776999531659314, + "learning_rate": 2.0601999228559134e-06, + "loss": 0.968, + "step": 33882 + }, + { + "epoch": 0.8, + "grad_norm": 1.8712274818671704, + "learning_rate": 2.0597360598362705e-06, + "loss": 1.0257, + "step": 33883 + }, + { + "epoch": 0.8, + "grad_norm": 2.2362192842700344, + "learning_rate": 2.0592722430473645e-06, + "loss": 0.8517, + "step": 33884 + }, + { + "epoch": 0.8, + "grad_norm": 1.8056677509567278, + "learning_rate": 2.0588084724918834e-06, + "loss": 0.9851, + "step": 33885 + }, + { + "epoch": 0.8, + "grad_norm": 1.1224857080997792, + "learning_rate": 2.058344748172534e-06, + "loss": 0.9709, + "step": 33886 + }, + { + "epoch": 0.8, + "grad_norm": 3.003946243752824, + "learning_rate": 2.057881070092015e-06, + "loss": 0.9698, + "step": 33887 + }, + { + "epoch": 0.8, + "grad_norm": 2.19743317113627, + "learning_rate": 2.0574174382530332e-06, + "loss": 1.0109, + "step": 33888 + }, + { + "epoch": 0.8, + "grad_norm": 1.7320086301208162, + "learning_rate": 2.0569538526582753e-06, + "loss": 0.921, + "step": 33889 + }, + { + "epoch": 0.8, + "grad_norm": 2.07634409404843, + "learning_rate": 2.0564903133104474e-06, + "loss": 0.9284, + "step": 33890 + }, + { + "epoch": 0.8, + "grad_norm": 1.1189640509822047, + "learning_rate": 2.0560268202122514e-06, + "loss": 0.9117, + "step": 33891 + }, + { + "epoch": 0.8, + "grad_norm": 2.3100648574324896, + "learning_rate": 2.055563373366378e-06, + "loss": 0.9761, + "step": 33892 + }, + { + "epoch": 0.8, + "grad_norm": 1.0496684631411937, + "learning_rate": 2.0550999727755327e-06, + "loss": 0.8888, + "step": 33893 + }, + { + "epoch": 0.8, + "grad_norm": 2.0958202234689467, + "learning_rate": 2.0546366184424094e-06, + "loss": 1.0481, + "step": 33894 + }, + { + "epoch": 0.8, + "grad_norm": 1.9341946265487224, + "learning_rate": 2.054173310369708e-06, + "loss": 0.9841, + "step": 33895 + }, + { + "epoch": 0.8, + "grad_norm": 1.968191723819863, + "learning_rate": 2.0537100485601238e-06, + "loss": 1.0484, + "step": 33896 + }, + { + "epoch": 0.8, + "grad_norm": 2.066114357156803, + "learning_rate": 2.053246833016358e-06, + "loss": 1.0598, + "step": 33897 + }, + { + "epoch": 0.8, + "grad_norm": 1.9694276074779902, + "learning_rate": 2.052783663741105e-06, + "loss": 0.9222, + "step": 33898 + }, + { + "epoch": 0.8, + "grad_norm": 1.8163033459468263, + "learning_rate": 2.052320540737058e-06, + "loss": 1.0618, + "step": 33899 + }, + { + "epoch": 0.8, + "grad_norm": 1.9679821601806784, + "learning_rate": 2.0518574640069178e-06, + "loss": 0.9228, + "step": 33900 + }, + { + "epoch": 0.8, + "grad_norm": 1.1594929010857924, + "learning_rate": 2.051394433553382e-06, + "loss": 0.9573, + "step": 33901 + }, + { + "epoch": 0.8, + "grad_norm": 2.0122713584587184, + "learning_rate": 2.0509314493791444e-06, + "loss": 1.0688, + "step": 33902 + }, + { + "epoch": 0.8, + "grad_norm": 2.1974464632428465, + "learning_rate": 2.050468511486898e-06, + "loss": 1.0687, + "step": 33903 + }, + { + "epoch": 0.8, + "grad_norm": 2.3758525010491476, + "learning_rate": 2.0500056198793405e-06, + "loss": 0.9493, + "step": 33904 + }, + { + "epoch": 0.8, + "grad_norm": 1.0808637004704742, + "learning_rate": 2.04954277455917e-06, + "loss": 0.958, + "step": 33905 + }, + { + "epoch": 0.8, + "grad_norm": 1.9856825881305853, + "learning_rate": 2.049079975529079e-06, + "loss": 0.9768, + "step": 33906 + }, + { + "epoch": 0.8, + "grad_norm": 2.1381045425643475, + "learning_rate": 2.0486172227917587e-06, + "loss": 0.8723, + "step": 33907 + }, + { + "epoch": 0.8, + "grad_norm": 1.8379491199154974, + "learning_rate": 2.0481545163499062e-06, + "loss": 1.0021, + "step": 33908 + }, + { + "epoch": 0.8, + "grad_norm": 2.0368895311494213, + "learning_rate": 2.0476918562062186e-06, + "loss": 0.9546, + "step": 33909 + }, + { + "epoch": 0.8, + "grad_norm": 1.8558834859722644, + "learning_rate": 2.0472292423633857e-06, + "loss": 0.8561, + "step": 33910 + }, + { + "epoch": 0.8, + "grad_norm": 1.9811755141678695, + "learning_rate": 2.0467666748241e-06, + "loss": 0.8738, + "step": 33911 + }, + { + "epoch": 0.8, + "grad_norm": 2.375608139405569, + "learning_rate": 2.0463041535910587e-06, + "loss": 1.0061, + "step": 33912 + }, + { + "epoch": 0.8, + "grad_norm": 1.9572522360797848, + "learning_rate": 2.045841678666951e-06, + "loss": 0.9692, + "step": 33913 + }, + { + "epoch": 0.8, + "grad_norm": 2.5193546682401275, + "learning_rate": 2.0453792500544734e-06, + "loss": 1.0164, + "step": 33914 + }, + { + "epoch": 0.8, + "grad_norm": 1.1387058178568472, + "learning_rate": 2.0449168677563136e-06, + "loss": 0.9593, + "step": 33915 + }, + { + "epoch": 0.8, + "grad_norm": 2.165856936143661, + "learning_rate": 2.04445453177517e-06, + "loss": 1.1189, + "step": 33916 + }, + { + "epoch": 0.8, + "grad_norm": 2.08043971332864, + "learning_rate": 2.043992242113727e-06, + "loss": 0.9269, + "step": 33917 + }, + { + "epoch": 0.8, + "grad_norm": 2.0274899331601497, + "learning_rate": 2.0435299987746825e-06, + "loss": 0.9945, + "step": 33918 + }, + { + "epoch": 0.8, + "grad_norm": 1.9406173355666059, + "learning_rate": 2.0430678017607266e-06, + "loss": 0.9061, + "step": 33919 + }, + { + "epoch": 0.8, + "grad_norm": 2.4036500956213236, + "learning_rate": 2.0426056510745453e-06, + "loss": 0.9981, + "step": 33920 + }, + { + "epoch": 0.8, + "grad_norm": 1.9825122463834144, + "learning_rate": 2.0421435467188333e-06, + "loss": 1.0313, + "step": 33921 + }, + { + "epoch": 0.8, + "grad_norm": 1.9454645490972489, + "learning_rate": 2.041681488696283e-06, + "loss": 1.1196, + "step": 33922 + }, + { + "epoch": 0.8, + "grad_norm": 1.8731145023651452, + "learning_rate": 2.041219477009584e-06, + "loss": 1.0246, + "step": 33923 + }, + { + "epoch": 0.8, + "grad_norm": 2.2450173433775964, + "learning_rate": 2.040757511661421e-06, + "loss": 1.0334, + "step": 33924 + }, + { + "epoch": 0.8, + "grad_norm": 1.7799596838287746, + "learning_rate": 2.0402955926544887e-06, + "loss": 1.0756, + "step": 33925 + }, + { + "epoch": 0.8, + "grad_norm": 1.9083499167478455, + "learning_rate": 2.0398337199914774e-06, + "loss": 0.9985, + "step": 33926 + }, + { + "epoch": 0.8, + "grad_norm": 1.9165121492585204, + "learning_rate": 2.039371893675075e-06, + "loss": 0.9689, + "step": 33927 + }, + { + "epoch": 0.8, + "grad_norm": 1.9913290321096677, + "learning_rate": 2.038910113707966e-06, + "loss": 0.9999, + "step": 33928 + }, + { + "epoch": 0.8, + "grad_norm": 1.8979311664459808, + "learning_rate": 2.0384483800928433e-06, + "loss": 0.9488, + "step": 33929 + }, + { + "epoch": 0.8, + "grad_norm": 1.797500200582675, + "learning_rate": 2.0379866928323977e-06, + "loss": 1.0252, + "step": 33930 + }, + { + "epoch": 0.8, + "grad_norm": 2.279654524300672, + "learning_rate": 2.0375250519293145e-06, + "loss": 1.0249, + "step": 33931 + }, + { + "epoch": 0.8, + "grad_norm": 2.0576495649861894, + "learning_rate": 2.0370634573862778e-06, + "loss": 0.8148, + "step": 33932 + }, + { + "epoch": 0.8, + "grad_norm": 2.339685889484969, + "learning_rate": 2.036601909205982e-06, + "loss": 0.984, + "step": 33933 + }, + { + "epoch": 0.8, + "grad_norm": 2.0361541717195717, + "learning_rate": 2.0361404073911086e-06, + "loss": 0.9778, + "step": 33934 + }, + { + "epoch": 0.8, + "grad_norm": 2.208319279080367, + "learning_rate": 2.035678951944349e-06, + "loss": 0.8753, + "step": 33935 + }, + { + "epoch": 0.8, + "grad_norm": 2.1941339989717985, + "learning_rate": 2.0352175428683853e-06, + "loss": 1.049, + "step": 33936 + }, + { + "epoch": 0.8, + "grad_norm": 2.2894514639529, + "learning_rate": 2.03475618016591e-06, + "loss": 0.937, + "step": 33937 + }, + { + "epoch": 0.8, + "grad_norm": 2.0599961469145356, + "learning_rate": 2.034294863839603e-06, + "loss": 1.0288, + "step": 33938 + }, + { + "epoch": 0.8, + "grad_norm": 4.007102105773854, + "learning_rate": 2.033833593892156e-06, + "loss": 1.0561, + "step": 33939 + }, + { + "epoch": 0.8, + "grad_norm": 2.2491061003505206, + "learning_rate": 2.033372370326252e-06, + "loss": 1.0138, + "step": 33940 + }, + { + "epoch": 0.8, + "grad_norm": 3.6302808556055304, + "learning_rate": 2.032911193144573e-06, + "loss": 0.9525, + "step": 33941 + }, + { + "epoch": 0.8, + "grad_norm": 2.1789327728241106, + "learning_rate": 2.0324500623498076e-06, + "loss": 1.0097, + "step": 33942 + }, + { + "epoch": 0.8, + "grad_norm": 1.825405320780427, + "learning_rate": 2.0319889779446432e-06, + "loss": 0.9853, + "step": 33943 + }, + { + "epoch": 0.8, + "grad_norm": 1.8580030830342757, + "learning_rate": 2.031527939931761e-06, + "loss": 1.0317, + "step": 33944 + }, + { + "epoch": 0.8, + "grad_norm": 1.9393686418293872, + "learning_rate": 2.0310669483138446e-06, + "loss": 0.8954, + "step": 33945 + }, + { + "epoch": 0.8, + "grad_norm": 2.217691161407129, + "learning_rate": 2.030606003093578e-06, + "loss": 0.9753, + "step": 33946 + }, + { + "epoch": 0.8, + "grad_norm": 1.923783760963163, + "learning_rate": 2.0301451042736496e-06, + "loss": 0.8409, + "step": 33947 + }, + { + "epoch": 0.8, + "grad_norm": 2.3339052273685876, + "learning_rate": 2.0296842518567395e-06, + "loss": 0.9615, + "step": 33948 + }, + { + "epoch": 0.8, + "grad_norm": 1.1130825307698218, + "learning_rate": 2.0292234458455284e-06, + "loss": 0.9024, + "step": 33949 + }, + { + "epoch": 0.8, + "grad_norm": 2.4588694201396155, + "learning_rate": 2.0287626862427024e-06, + "loss": 0.9452, + "step": 33950 + }, + { + "epoch": 0.8, + "grad_norm": 2.100260100438074, + "learning_rate": 2.0283019730509456e-06, + "loss": 0.9073, + "step": 33951 + }, + { + "epoch": 0.8, + "grad_norm": 2.1890499855542505, + "learning_rate": 2.0278413062729394e-06, + "loss": 0.9773, + "step": 33952 + }, + { + "epoch": 0.8, + "grad_norm": 1.0759126873362124, + "learning_rate": 2.027380685911361e-06, + "loss": 0.9445, + "step": 33953 + }, + { + "epoch": 0.8, + "grad_norm": 2.0811970274743254, + "learning_rate": 2.0269201119688975e-06, + "loss": 1.1378, + "step": 33954 + }, + { + "epoch": 0.8, + "grad_norm": 1.0847753862694571, + "learning_rate": 2.0264595844482315e-06, + "loss": 0.9361, + "step": 33955 + }, + { + "epoch": 0.8, + "grad_norm": 1.8677647546381675, + "learning_rate": 2.025999103352042e-06, + "loss": 0.9404, + "step": 33956 + }, + { + "epoch": 0.8, + "grad_norm": 2.120181936758309, + "learning_rate": 2.025538668683008e-06, + "loss": 0.9473, + "step": 33957 + }, + { + "epoch": 0.8, + "grad_norm": 1.0752974543098124, + "learning_rate": 2.025078280443815e-06, + "loss": 0.9337, + "step": 33958 + }, + { + "epoch": 0.8, + "grad_norm": 1.94582675045398, + "learning_rate": 2.024617938637139e-06, + "loss": 0.9638, + "step": 33959 + }, + { + "epoch": 0.8, + "grad_norm": 2.178182272973909, + "learning_rate": 2.024157643265665e-06, + "loss": 1.0216, + "step": 33960 + }, + { + "epoch": 0.8, + "grad_norm": 2.00991860501188, + "learning_rate": 2.023697394332067e-06, + "loss": 1.0741, + "step": 33961 + }, + { + "epoch": 0.8, + "grad_norm": 1.8292343031145946, + "learning_rate": 2.0232371918390305e-06, + "loss": 0.8401, + "step": 33962 + }, + { + "epoch": 0.8, + "grad_norm": 2.3529804287991167, + "learning_rate": 2.0227770357892307e-06, + "loss": 0.8766, + "step": 33963 + }, + { + "epoch": 0.8, + "grad_norm": 1.0666099452485829, + "learning_rate": 2.02231692618535e-06, + "loss": 0.8816, + "step": 33964 + }, + { + "epoch": 0.8, + "grad_norm": 2.084163650574864, + "learning_rate": 2.021856863030067e-06, + "loss": 1.028, + "step": 33965 + }, + { + "epoch": 0.8, + "grad_norm": 2.109544821587748, + "learning_rate": 2.0213968463260556e-06, + "loss": 0.863, + "step": 33966 + }, + { + "epoch": 0.8, + "grad_norm": 1.994030840964016, + "learning_rate": 2.0209368760759983e-06, + "loss": 1.0309, + "step": 33967 + }, + { + "epoch": 0.8, + "grad_norm": 1.0866832663771513, + "learning_rate": 2.020476952282573e-06, + "loss": 1.0062, + "step": 33968 + }, + { + "epoch": 0.8, + "grad_norm": 2.1350025978426084, + "learning_rate": 2.0200170749484614e-06, + "loss": 0.8972, + "step": 33969 + }, + { + "epoch": 0.8, + "grad_norm": 2.2331828939930394, + "learning_rate": 2.019557244076332e-06, + "loss": 1.0119, + "step": 33970 + }, + { + "epoch": 0.8, + "grad_norm": 2.723945460219201, + "learning_rate": 2.0190974596688673e-06, + "loss": 0.9509, + "step": 33971 + }, + { + "epoch": 0.8, + "grad_norm": 1.9728568392335497, + "learning_rate": 2.018637721728747e-06, + "loss": 0.9601, + "step": 33972 + }, + { + "epoch": 0.8, + "grad_norm": 2.0356720873101533, + "learning_rate": 2.018178030258643e-06, + "loss": 1.05, + "step": 33973 + }, + { + "epoch": 0.8, + "grad_norm": 2.0101400858086116, + "learning_rate": 2.0177183852612316e-06, + "loss": 1.1016, + "step": 33974 + }, + { + "epoch": 0.8, + "grad_norm": 1.022614113379097, + "learning_rate": 2.017258786739191e-06, + "loss": 0.9565, + "step": 33975 + }, + { + "epoch": 0.8, + "grad_norm": 2.0772387521821263, + "learning_rate": 2.0167992346952005e-06, + "loss": 0.9661, + "step": 33976 + }, + { + "epoch": 0.8, + "grad_norm": 2.0350031283731878, + "learning_rate": 2.016339729131929e-06, + "loss": 1.024, + "step": 33977 + }, + { + "epoch": 0.8, + "grad_norm": 1.8754025556144711, + "learning_rate": 2.0158802700520576e-06, + "loss": 0.8114, + "step": 33978 + }, + { + "epoch": 0.8, + "grad_norm": 2.08267807949526, + "learning_rate": 2.015420857458259e-06, + "loss": 0.855, + "step": 33979 + }, + { + "epoch": 0.8, + "grad_norm": 1.7749191887461524, + "learning_rate": 2.0149614913532055e-06, + "loss": 0.8428, + "step": 33980 + }, + { + "epoch": 0.8, + "grad_norm": 2.279131625985543, + "learning_rate": 2.014502171739574e-06, + "loss": 0.9326, + "step": 33981 + }, + { + "epoch": 0.8, + "grad_norm": 2.2232909385820716, + "learning_rate": 2.014042898620042e-06, + "loss": 0.935, + "step": 33982 + }, + { + "epoch": 0.8, + "grad_norm": 2.174908585629583, + "learning_rate": 2.013583671997281e-06, + "loss": 1.0872, + "step": 33983 + }, + { + "epoch": 0.8, + "grad_norm": 1.199036508056454, + "learning_rate": 2.0131244918739612e-06, + "loss": 1.0023, + "step": 33984 + }, + { + "epoch": 0.8, + "grad_norm": 1.930020732862086, + "learning_rate": 2.012665358252759e-06, + "loss": 1.0075, + "step": 33985 + }, + { + "epoch": 0.8, + "grad_norm": 2.3376906684106196, + "learning_rate": 2.012206271136353e-06, + "loss": 1.0325, + "step": 33986 + }, + { + "epoch": 0.8, + "grad_norm": 1.9074890207123725, + "learning_rate": 2.0117472305274065e-06, + "loss": 1.0722, + "step": 33987 + }, + { + "epoch": 0.8, + "grad_norm": 1.8476956524715253, + "learning_rate": 2.0112882364285968e-06, + "loss": 0.9265, + "step": 33988 + }, + { + "epoch": 0.8, + "grad_norm": 2.0064275568539727, + "learning_rate": 2.010829288842595e-06, + "loss": 0.8101, + "step": 33989 + }, + { + "epoch": 0.8, + "grad_norm": 1.9777131967052943, + "learning_rate": 2.0103703877720783e-06, + "loss": 0.868, + "step": 33990 + }, + { + "epoch": 0.8, + "grad_norm": 1.1311027453572453, + "learning_rate": 2.009911533219714e-06, + "loss": 0.967, + "step": 33991 + }, + { + "epoch": 0.8, + "grad_norm": 1.8081047098409044, + "learning_rate": 2.0094527251881713e-06, + "loss": 1.0657, + "step": 33992 + }, + { + "epoch": 0.8, + "grad_norm": 1.956377424163022, + "learning_rate": 2.008993963680127e-06, + "loss": 1.0113, + "step": 33993 + }, + { + "epoch": 0.8, + "grad_norm": 2.0083584104774856, + "learning_rate": 2.008535248698248e-06, + "loss": 0.9869, + "step": 33994 + }, + { + "epoch": 0.8, + "grad_norm": 1.991220510585124, + "learning_rate": 2.0080765802452097e-06, + "loss": 0.9554, + "step": 33995 + }, + { + "epoch": 0.8, + "grad_norm": 1.1171908219119682, + "learning_rate": 2.0076179583236765e-06, + "loss": 0.9414, + "step": 33996 + }, + { + "epoch": 0.8, + "grad_norm": 1.0741586895971917, + "learning_rate": 2.0071593829363245e-06, + "loss": 0.9609, + "step": 33997 + }, + { + "epoch": 0.8, + "grad_norm": 1.9396891047004157, + "learning_rate": 2.006700854085819e-06, + "loss": 0.9647, + "step": 33998 + }, + { + "epoch": 0.8, + "grad_norm": 2.1949036228468572, + "learning_rate": 2.0062423717748336e-06, + "loss": 0.95, + "step": 33999 + }, + { + "epoch": 0.8, + "grad_norm": 2.637106137560828, + "learning_rate": 2.005783936006036e-06, + "loss": 0.9871, + "step": 34000 + }, + { + "epoch": 0.8, + "grad_norm": 1.7104918901068553, + "learning_rate": 2.0053255467820933e-06, + "loss": 0.8516, + "step": 34001 + }, + { + "epoch": 0.8, + "grad_norm": 1.976316522711689, + "learning_rate": 2.004867204105676e-06, + "loss": 1.0898, + "step": 34002 + }, + { + "epoch": 0.8, + "grad_norm": 1.933963887834839, + "learning_rate": 2.004408907979456e-06, + "loss": 0.9833, + "step": 34003 + }, + { + "epoch": 0.8, + "grad_norm": 1.8593323577721457, + "learning_rate": 2.0039506584060987e-06, + "loss": 0.9535, + "step": 34004 + }, + { + "epoch": 0.8, + "grad_norm": 2.1643408108341275, + "learning_rate": 2.0034924553882683e-06, + "loss": 1.0077, + "step": 34005 + }, + { + "epoch": 0.8, + "grad_norm": 2.0240631476792244, + "learning_rate": 2.0030342989286376e-06, + "loss": 0.9905, + "step": 34006 + }, + { + "epoch": 0.8, + "grad_norm": 1.903936097925543, + "learning_rate": 2.0025761890298757e-06, + "loss": 0.9597, + "step": 34007 + }, + { + "epoch": 0.8, + "grad_norm": 1.8747776765195974, + "learning_rate": 2.0021181256946465e-06, + "loss": 1.0612, + "step": 34008 + }, + { + "epoch": 0.8, + "grad_norm": 1.6141926722153959, + "learning_rate": 2.001660108925616e-06, + "loss": 0.9901, + "step": 34009 + }, + { + "epoch": 0.8, + "grad_norm": 2.2759011631114476, + "learning_rate": 2.001202138725451e-06, + "loss": 0.9128, + "step": 34010 + }, + { + "epoch": 0.8, + "grad_norm": 2.0049222719543063, + "learning_rate": 2.0007442150968236e-06, + "loss": 0.9305, + "step": 34011 + }, + { + "epoch": 0.8, + "grad_norm": 2.2077810670031752, + "learning_rate": 2.000286338042395e-06, + "loss": 0.939, + "step": 34012 + }, + { + "epoch": 0.8, + "grad_norm": 1.849249338603931, + "learning_rate": 1.9998285075648304e-06, + "loss": 0.9463, + "step": 34013 + }, + { + "epoch": 0.8, + "grad_norm": 2.78128069768765, + "learning_rate": 1.9993707236667992e-06, + "loss": 0.8832, + "step": 34014 + }, + { + "epoch": 0.8, + "grad_norm": 2.040626924806957, + "learning_rate": 1.9989129863509617e-06, + "loss": 0.9603, + "step": 34015 + }, + { + "epoch": 0.8, + "grad_norm": 1.7455414729625283, + "learning_rate": 1.998455295619989e-06, + "loss": 0.8955, + "step": 34016 + }, + { + "epoch": 0.8, + "grad_norm": 1.1256887498383241, + "learning_rate": 1.9979976514765397e-06, + "loss": 1.0048, + "step": 34017 + }, + { + "epoch": 0.8, + "grad_norm": 1.958219628399572, + "learning_rate": 1.9975400539232836e-06, + "loss": 1.0398, + "step": 34018 + }, + { + "epoch": 0.8, + "grad_norm": 1.909335170326913, + "learning_rate": 1.9970825029628814e-06, + "loss": 1.0838, + "step": 34019 + }, + { + "epoch": 0.8, + "grad_norm": 1.9000118603216583, + "learning_rate": 1.996624998598e-06, + "loss": 1.0518, + "step": 34020 + }, + { + "epoch": 0.8, + "grad_norm": 2.094769240769306, + "learning_rate": 1.996167540831303e-06, + "loss": 0.9817, + "step": 34021 + }, + { + "epoch": 0.8, + "grad_norm": 1.0903035180224012, + "learning_rate": 1.995710129665449e-06, + "loss": 0.8635, + "step": 34022 + }, + { + "epoch": 0.8, + "grad_norm": 2.2962745468366768, + "learning_rate": 1.995252765103105e-06, + "loss": 0.9223, + "step": 34023 + }, + { + "epoch": 0.8, + "grad_norm": 1.8578132528932358, + "learning_rate": 1.9947954471469365e-06, + "loss": 1.0167, + "step": 34024 + }, + { + "epoch": 0.8, + "grad_norm": 2.1937470097806715, + "learning_rate": 1.9943381757996027e-06, + "loss": 0.9836, + "step": 34025 + }, + { + "epoch": 0.8, + "grad_norm": 1.8214071483528855, + "learning_rate": 1.9938809510637648e-06, + "loss": 0.9208, + "step": 34026 + }, + { + "epoch": 0.8, + "grad_norm": 2.298091325114396, + "learning_rate": 1.993423772942086e-06, + "loss": 0.9671, + "step": 34027 + }, + { + "epoch": 0.8, + "grad_norm": 2.007571893522132, + "learning_rate": 1.992966641437232e-06, + "loss": 1.0704, + "step": 34028 + }, + { + "epoch": 0.8, + "grad_norm": 1.8098660335886225, + "learning_rate": 1.9925095565518605e-06, + "loss": 1.0062, + "step": 34029 + }, + { + "epoch": 0.8, + "grad_norm": 3.1213078207597262, + "learning_rate": 1.992052518288632e-06, + "loss": 1.0065, + "step": 34030 + }, + { + "epoch": 0.8, + "grad_norm": 1.1260542719717468, + "learning_rate": 1.9915955266502073e-06, + "loss": 0.9321, + "step": 34031 + }, + { + "epoch": 0.8, + "grad_norm": 1.991983308098141, + "learning_rate": 1.9911385816392536e-06, + "loss": 0.9234, + "step": 34032 + }, + { + "epoch": 0.8, + "grad_norm": 1.9095161482979572, + "learning_rate": 1.990681683258425e-06, + "loss": 1.0537, + "step": 34033 + }, + { + "epoch": 0.8, + "grad_norm": 1.856406198711108, + "learning_rate": 1.990224831510381e-06, + "loss": 0.9243, + "step": 34034 + }, + { + "epoch": 0.8, + "grad_norm": 1.959902054591653, + "learning_rate": 1.989768026397787e-06, + "loss": 1.0828, + "step": 34035 + }, + { + "epoch": 0.8, + "grad_norm": 2.0760781668699906, + "learning_rate": 1.989311267923296e-06, + "loss": 1.016, + "step": 34036 + }, + { + "epoch": 0.8, + "grad_norm": 2.0887591201181763, + "learning_rate": 1.9888545560895746e-06, + "loss": 0.9014, + "step": 34037 + }, + { + "epoch": 0.8, + "grad_norm": 1.9501509896601252, + "learning_rate": 1.988397890899275e-06, + "loss": 0.9747, + "step": 34038 + }, + { + "epoch": 0.8, + "grad_norm": 1.9490676214021037, + "learning_rate": 1.987941272355063e-06, + "loss": 0.993, + "step": 34039 + }, + { + "epoch": 0.8, + "grad_norm": 1.896995913793792, + "learning_rate": 1.98748470045959e-06, + "loss": 0.8661, + "step": 34040 + }, + { + "epoch": 0.8, + "grad_norm": 2.496754695074027, + "learning_rate": 1.9870281752155208e-06, + "loss": 1.1331, + "step": 34041 + }, + { + "epoch": 0.8, + "grad_norm": 1.983149852064412, + "learning_rate": 1.98657169662551e-06, + "loss": 1.0233, + "step": 34042 + }, + { + "epoch": 0.8, + "grad_norm": 2.035943904410658, + "learning_rate": 1.9861152646922133e-06, + "loss": 0.8244, + "step": 34043 + }, + { + "epoch": 0.8, + "grad_norm": 1.9799184564881858, + "learning_rate": 1.9856588794182906e-06, + "loss": 0.9851, + "step": 34044 + }, + { + "epoch": 0.8, + "grad_norm": 1.0766847729842215, + "learning_rate": 1.985202540806401e-06, + "loss": 0.9346, + "step": 34045 + }, + { + "epoch": 0.8, + "grad_norm": 2.2924340639506005, + "learning_rate": 1.9847462488592017e-06, + "loss": 1.0328, + "step": 34046 + }, + { + "epoch": 0.8, + "grad_norm": 2.414935515660078, + "learning_rate": 1.9842900035793435e-06, + "loss": 0.9833, + "step": 34047 + }, + { + "epoch": 0.8, + "grad_norm": 1.9483730447892598, + "learning_rate": 1.983833804969486e-06, + "loss": 0.9021, + "step": 34048 + }, + { + "epoch": 0.8, + "grad_norm": 1.9478853078103362, + "learning_rate": 1.983377653032289e-06, + "loss": 1.0537, + "step": 34049 + }, + { + "epoch": 0.8, + "grad_norm": 1.2127039849457648, + "learning_rate": 1.9829215477704066e-06, + "loss": 0.973, + "step": 34050 + }, + { + "epoch": 0.8, + "grad_norm": 1.9698595754812713, + "learning_rate": 1.9824654891864893e-06, + "loss": 1.0755, + "step": 34051 + }, + { + "epoch": 0.8, + "grad_norm": 1.8252327261657355, + "learning_rate": 1.982009477283197e-06, + "loss": 0.9794, + "step": 34052 + }, + { + "epoch": 0.8, + "grad_norm": 1.0736806117630608, + "learning_rate": 1.9815535120631846e-06, + "loss": 0.9398, + "step": 34053 + }, + { + "epoch": 0.8, + "grad_norm": 2.1122967862736703, + "learning_rate": 1.9810975935291076e-06, + "loss": 0.8391, + "step": 34054 + }, + { + "epoch": 0.8, + "grad_norm": 2.2775637817251755, + "learning_rate": 1.980641721683617e-06, + "loss": 0.9097, + "step": 34055 + }, + { + "epoch": 0.8, + "grad_norm": 2.525387795414265, + "learning_rate": 1.980185896529372e-06, + "loss": 0.996, + "step": 34056 + }, + { + "epoch": 0.8, + "grad_norm": 1.884078143548867, + "learning_rate": 1.9797301180690197e-06, + "loss": 0.9931, + "step": 34057 + }, + { + "epoch": 0.8, + "grad_norm": 2.407166439160589, + "learning_rate": 1.9792743863052223e-06, + "loss": 0.8883, + "step": 34058 + }, + { + "epoch": 0.8, + "grad_norm": 2.023506226900531, + "learning_rate": 1.9788187012406247e-06, + "loss": 1.0522, + "step": 34059 + }, + { + "epoch": 0.8, + "grad_norm": 1.9849014662134719, + "learning_rate": 1.978363062877887e-06, + "loss": 0.9646, + "step": 34060 + }, + { + "epoch": 0.8, + "grad_norm": 1.792626099761235, + "learning_rate": 1.9779074712196567e-06, + "loss": 0.9588, + "step": 34061 + }, + { + "epoch": 0.8, + "grad_norm": 1.863936590663195, + "learning_rate": 1.9774519262685908e-06, + "loss": 1.0071, + "step": 34062 + }, + { + "epoch": 0.8, + "grad_norm": 2.120525686013559, + "learning_rate": 1.97699642802734e-06, + "loss": 1.027, + "step": 34063 + }, + { + "epoch": 0.8, + "grad_norm": 1.1077254244524297, + "learning_rate": 1.9765409764985543e-06, + "loss": 0.8761, + "step": 34064 + }, + { + "epoch": 0.8, + "grad_norm": 1.0512679971975214, + "learning_rate": 1.976085571684887e-06, + "loss": 0.913, + "step": 34065 + }, + { + "epoch": 0.8, + "grad_norm": 1.8415681702933986, + "learning_rate": 1.9756302135889894e-06, + "loss": 0.9031, + "step": 34066 + }, + { + "epoch": 0.8, + "grad_norm": 1.8184327146355659, + "learning_rate": 1.975174902213518e-06, + "loss": 1.076, + "step": 34067 + }, + { + "epoch": 0.8, + "grad_norm": 1.8512283840598145, + "learning_rate": 1.9747196375611143e-06, + "loss": 0.9942, + "step": 34068 + }, + { + "epoch": 0.8, + "grad_norm": 1.9808451990841045, + "learning_rate": 1.974264419634433e-06, + "loss": 0.9886, + "step": 34069 + }, + { + "epoch": 0.8, + "grad_norm": 2.095313263694928, + "learning_rate": 1.973809248436128e-06, + "loss": 1.1443, + "step": 34070 + }, + { + "epoch": 0.8, + "grad_norm": 1.8521573890639473, + "learning_rate": 1.973354123968847e-06, + "loss": 0.8457, + "step": 34071 + }, + { + "epoch": 0.8, + "grad_norm": 1.077370510916957, + "learning_rate": 1.972899046235237e-06, + "loss": 0.913, + "step": 34072 + }, + { + "epoch": 0.8, + "grad_norm": 1.7957957805232214, + "learning_rate": 1.97244401523795e-06, + "loss": 0.8924, + "step": 34073 + }, + { + "epoch": 0.8, + "grad_norm": 1.9676929806478187, + "learning_rate": 1.9719890309796374e-06, + "loss": 1.0995, + "step": 34074 + }, + { + "epoch": 0.8, + "grad_norm": 2.272470018084276, + "learning_rate": 1.9715340934629446e-06, + "loss": 0.9635, + "step": 34075 + }, + { + "epoch": 0.8, + "grad_norm": 1.945508565223762, + "learning_rate": 1.9710792026905243e-06, + "loss": 0.852, + "step": 34076 + }, + { + "epoch": 0.8, + "grad_norm": 1.9300008885641442, + "learning_rate": 1.97062435866502e-06, + "loss": 1.0693, + "step": 34077 + }, + { + "epoch": 0.8, + "grad_norm": 1.929201137683699, + "learning_rate": 1.970169561389086e-06, + "loss": 0.8024, + "step": 34078 + }, + { + "epoch": 0.8, + "grad_norm": 1.9673536269334082, + "learning_rate": 1.969714810865363e-06, + "loss": 0.9725, + "step": 34079 + }, + { + "epoch": 0.8, + "grad_norm": 2.1321526628293435, + "learning_rate": 1.969260107096507e-06, + "loss": 0.9052, + "step": 34080 + }, + { + "epoch": 0.8, + "grad_norm": 1.078770938449036, + "learning_rate": 1.9688054500851596e-06, + "loss": 0.9412, + "step": 34081 + }, + { + "epoch": 0.8, + "grad_norm": 1.8814205672360322, + "learning_rate": 1.968350839833968e-06, + "loss": 0.8166, + "step": 34082 + }, + { + "epoch": 0.8, + "grad_norm": 1.929976658293048, + "learning_rate": 1.96789627634558e-06, + "loss": 0.9948, + "step": 34083 + }, + { + "epoch": 0.8, + "grad_norm": 1.824655789739647, + "learning_rate": 1.967441759622646e-06, + "loss": 1.0065, + "step": 34084 + }, + { + "epoch": 0.8, + "grad_norm": 1.897530632609113, + "learning_rate": 1.9669872896678087e-06, + "loss": 0.9583, + "step": 34085 + }, + { + "epoch": 0.8, + "grad_norm": 11.292629591334158, + "learning_rate": 1.9665328664837114e-06, + "loss": 0.9358, + "step": 34086 + }, + { + "epoch": 0.8, + "grad_norm": 1.8620740935774525, + "learning_rate": 1.9660784900730047e-06, + "loss": 0.8975, + "step": 34087 + }, + { + "epoch": 0.8, + "grad_norm": 1.848570128636071, + "learning_rate": 1.9656241604383343e-06, + "loss": 0.8807, + "step": 34088 + }, + { + "epoch": 0.8, + "grad_norm": 1.9155414571298068, + "learning_rate": 1.9651698775823425e-06, + "loss": 0.8909, + "step": 34089 + }, + { + "epoch": 0.8, + "grad_norm": 1.9041661172381672, + "learning_rate": 1.964715641507674e-06, + "loss": 0.9392, + "step": 34090 + }, + { + "epoch": 0.8, + "grad_norm": 2.033936340401476, + "learning_rate": 1.9642614522169754e-06, + "loss": 0.9283, + "step": 34091 + }, + { + "epoch": 0.8, + "grad_norm": 1.0581042495908872, + "learning_rate": 1.963807309712893e-06, + "loss": 0.9369, + "step": 34092 + }, + { + "epoch": 0.8, + "grad_norm": 2.232688425763987, + "learning_rate": 1.9633532139980684e-06, + "loss": 0.8752, + "step": 34093 + }, + { + "epoch": 0.8, + "grad_norm": 2.0092479339801894, + "learning_rate": 1.962899165075143e-06, + "loss": 0.8944, + "step": 34094 + }, + { + "epoch": 0.8, + "grad_norm": 2.2050768419951297, + "learning_rate": 1.9624451629467658e-06, + "loss": 1.0368, + "step": 34095 + }, + { + "epoch": 0.8, + "grad_norm": 1.8305640071813492, + "learning_rate": 1.9619912076155757e-06, + "loss": 1.0749, + "step": 34096 + }, + { + "epoch": 0.8, + "grad_norm": 1.1137184765196537, + "learning_rate": 1.9615372990842184e-06, + "loss": 0.9642, + "step": 34097 + }, + { + "epoch": 0.8, + "grad_norm": 2.040582759257148, + "learning_rate": 1.9610834373553347e-06, + "loss": 0.9467, + "step": 34098 + }, + { + "epoch": 0.8, + "grad_norm": 1.9806503765985497, + "learning_rate": 1.9606296224315713e-06, + "loss": 0.9496, + "step": 34099 + }, + { + "epoch": 0.8, + "grad_norm": 2.039603313998296, + "learning_rate": 1.9601758543155636e-06, + "loss": 0.9137, + "step": 34100 + }, + { + "epoch": 0.8, + "grad_norm": 2.20578191102128, + "learning_rate": 1.9597221330099613e-06, + "loss": 1.1344, + "step": 34101 + }, + { + "epoch": 0.8, + "grad_norm": 2.002280315434547, + "learning_rate": 1.9592684585174014e-06, + "loss": 0.9782, + "step": 34102 + }, + { + "epoch": 0.8, + "grad_norm": 3.811856026711468, + "learning_rate": 1.958814830840523e-06, + "loss": 0.9594, + "step": 34103 + }, + { + "epoch": 0.8, + "grad_norm": 1.8447612151601425, + "learning_rate": 1.9583612499819713e-06, + "loss": 0.9841, + "step": 34104 + }, + { + "epoch": 0.8, + "grad_norm": 1.1382865740527843, + "learning_rate": 1.9579077159443893e-06, + "loss": 0.9378, + "step": 34105 + }, + { + "epoch": 0.8, + "grad_norm": 1.9174372135948752, + "learning_rate": 1.957454228730414e-06, + "loss": 0.9755, + "step": 34106 + }, + { + "epoch": 0.8, + "grad_norm": 2.2980418548994397, + "learning_rate": 1.9570007883426845e-06, + "loss": 1.0782, + "step": 34107 + }, + { + "epoch": 0.8, + "grad_norm": 1.900080753107521, + "learning_rate": 1.956547394783842e-06, + "loss": 0.8267, + "step": 34108 + }, + { + "epoch": 0.8, + "grad_norm": 1.8514294869544774, + "learning_rate": 1.9560940480565304e-06, + "loss": 0.9138, + "step": 34109 + }, + { + "epoch": 0.8, + "grad_norm": 2.019364276306737, + "learning_rate": 1.9556407481633854e-06, + "loss": 0.947, + "step": 34110 + }, + { + "epoch": 0.8, + "grad_norm": 2.1325936779403687, + "learning_rate": 1.9551874951070438e-06, + "loss": 0.8288, + "step": 34111 + }, + { + "epoch": 0.8, + "grad_norm": 3.7100088192228933, + "learning_rate": 1.954734288890149e-06, + "loss": 1.0565, + "step": 34112 + }, + { + "epoch": 0.8, + "grad_norm": 2.4952880546120975, + "learning_rate": 1.9542811295153396e-06, + "loss": 1.0192, + "step": 34113 + }, + { + "epoch": 0.8, + "grad_norm": 1.7853759752610094, + "learning_rate": 1.953828016985253e-06, + "loss": 0.9603, + "step": 34114 + }, + { + "epoch": 0.8, + "grad_norm": 1.9418749181204438, + "learning_rate": 1.9533749513025234e-06, + "loss": 0.8734, + "step": 34115 + }, + { + "epoch": 0.8, + "grad_norm": 1.8899276774736697, + "learning_rate": 1.9529219324697966e-06, + "loss": 1.0908, + "step": 34116 + }, + { + "epoch": 0.8, + "grad_norm": 1.9774093631314866, + "learning_rate": 1.9524689604897028e-06, + "loss": 1.01, + "step": 34117 + }, + { + "epoch": 0.8, + "grad_norm": 1.8889547335179548, + "learning_rate": 1.9520160353648852e-06, + "loss": 0.8805, + "step": 34118 + }, + { + "epoch": 0.8, + "grad_norm": 2.0785805989976507, + "learning_rate": 1.951563157097975e-06, + "loss": 0.9897, + "step": 34119 + }, + { + "epoch": 0.8, + "grad_norm": 1.9407159089528918, + "learning_rate": 1.9511103256916154e-06, + "loss": 1.0457, + "step": 34120 + }, + { + "epoch": 0.8, + "grad_norm": 2.1391154233194993, + "learning_rate": 1.9506575411484373e-06, + "loss": 1.0442, + "step": 34121 + }, + { + "epoch": 0.8, + "grad_norm": 2.020664894737698, + "learning_rate": 1.9502048034710818e-06, + "loss": 1.0048, + "step": 34122 + }, + { + "epoch": 0.8, + "grad_norm": 1.8005503850544455, + "learning_rate": 1.9497521126621823e-06, + "loss": 0.9462, + "step": 34123 + }, + { + "epoch": 0.8, + "grad_norm": 1.889432411758551, + "learning_rate": 1.9492994687243715e-06, + "loss": 0.9089, + "step": 34124 + }, + { + "epoch": 0.8, + "grad_norm": 1.0680865047405752, + "learning_rate": 1.9488468716602895e-06, + "loss": 0.9881, + "step": 34125 + }, + { + "epoch": 0.8, + "grad_norm": 2.0936646808292094, + "learning_rate": 1.948394321472571e-06, + "loss": 1.0365, + "step": 34126 + }, + { + "epoch": 0.8, + "grad_norm": 2.0866356489018707, + "learning_rate": 1.9479418181638508e-06, + "loss": 0.849, + "step": 34127 + }, + { + "epoch": 0.8, + "grad_norm": 1.7642258981728698, + "learning_rate": 1.9474893617367597e-06, + "loss": 0.9395, + "step": 34128 + }, + { + "epoch": 0.8, + "grad_norm": 1.77789010333123, + "learning_rate": 1.947036952193935e-06, + "loss": 1.0015, + "step": 34129 + }, + { + "epoch": 0.8, + "grad_norm": 2.0180409883503962, + "learning_rate": 1.946584589538013e-06, + "loss": 1.0186, + "step": 34130 + }, + { + "epoch": 0.8, + "grad_norm": 1.9130106267812372, + "learning_rate": 1.946132273771625e-06, + "loss": 0.9011, + "step": 34131 + }, + { + "epoch": 0.8, + "grad_norm": 2.0114250620330307, + "learning_rate": 1.9456800048974023e-06, + "loss": 0.8546, + "step": 34132 + }, + { + "epoch": 0.8, + "grad_norm": 1.0966304650601746, + "learning_rate": 1.945227782917981e-06, + "loss": 0.993, + "step": 34133 + }, + { + "epoch": 0.8, + "grad_norm": 1.9151436487440814, + "learning_rate": 1.944775607835996e-06, + "loss": 0.8255, + "step": 34134 + }, + { + "epoch": 0.8, + "grad_norm": 1.9100124291213856, + "learning_rate": 1.944323479654078e-06, + "loss": 1.0232, + "step": 34135 + }, + { + "epoch": 0.8, + "grad_norm": 1.975674695926436, + "learning_rate": 1.9438713983748556e-06, + "loss": 0.9383, + "step": 34136 + }, + { + "epoch": 0.8, + "grad_norm": 1.9746470981304602, + "learning_rate": 1.943419364000968e-06, + "loss": 0.9805, + "step": 34137 + }, + { + "epoch": 0.8, + "grad_norm": 2.0538762450520447, + "learning_rate": 1.94296737653504e-06, + "loss": 0.924, + "step": 34138 + }, + { + "epoch": 0.8, + "grad_norm": 2.065842567258875, + "learning_rate": 1.9425154359797103e-06, + "loss": 0.8762, + "step": 34139 + }, + { + "epoch": 0.8, + "grad_norm": 2.023530965518768, + "learning_rate": 1.9420635423376033e-06, + "loss": 1.128, + "step": 34140 + }, + { + "epoch": 0.8, + "grad_norm": 2.2407580351784913, + "learning_rate": 1.941611695611356e-06, + "loss": 0.9175, + "step": 34141 + }, + { + "epoch": 0.8, + "grad_norm": 2.6685830880499104, + "learning_rate": 1.941159895803595e-06, + "loss": 1.0258, + "step": 34142 + }, + { + "epoch": 0.8, + "grad_norm": 2.840225776795232, + "learning_rate": 1.940708142916955e-06, + "loss": 0.8871, + "step": 34143 + }, + { + "epoch": 0.8, + "grad_norm": 2.1258798792131195, + "learning_rate": 1.9402564369540634e-06, + "loss": 1.0875, + "step": 34144 + }, + { + "epoch": 0.8, + "grad_norm": 2.0273880903176384, + "learning_rate": 1.939804777917548e-06, + "loss": 1.013, + "step": 34145 + }, + { + "epoch": 0.8, + "grad_norm": 1.8287969713426877, + "learning_rate": 1.9393531658100395e-06, + "loss": 1.0783, + "step": 34146 + }, + { + "epoch": 0.8, + "grad_norm": 2.3249460642088664, + "learning_rate": 1.938901600634173e-06, + "loss": 0.8134, + "step": 34147 + }, + { + "epoch": 0.8, + "grad_norm": 1.9914059566861995, + "learning_rate": 1.9384500823925732e-06, + "loss": 0.9861, + "step": 34148 + }, + { + "epoch": 0.8, + "grad_norm": 1.9563260294338152, + "learning_rate": 1.9379986110878666e-06, + "loss": 1.1423, + "step": 34149 + }, + { + "epoch": 0.8, + "grad_norm": 1.1414118626304457, + "learning_rate": 1.9375471867226846e-06, + "loss": 0.8673, + "step": 34150 + }, + { + "epoch": 0.8, + "grad_norm": 2.2698129567938348, + "learning_rate": 1.9370958092996582e-06, + "loss": 1.0334, + "step": 34151 + }, + { + "epoch": 0.8, + "grad_norm": 2.147850802899183, + "learning_rate": 1.9366444788214123e-06, + "loss": 0.988, + "step": 34152 + }, + { + "epoch": 0.8, + "grad_norm": 1.1616237236673057, + "learning_rate": 1.9361931952905723e-06, + "loss": 0.9503, + "step": 34153 + }, + { + "epoch": 0.8, + "grad_norm": 1.0796698968020935, + "learning_rate": 1.935741958709768e-06, + "loss": 0.9498, + "step": 34154 + }, + { + "epoch": 0.8, + "grad_norm": 2.1176100227687003, + "learning_rate": 1.9352907690816313e-06, + "loss": 0.9299, + "step": 34155 + }, + { + "epoch": 0.8, + "grad_norm": 4.807469439832063, + "learning_rate": 1.9348396264087845e-06, + "loss": 0.9168, + "step": 34156 + }, + { + "epoch": 0.8, + "grad_norm": 2.4457517750552693, + "learning_rate": 1.9343885306938514e-06, + "loss": 0.984, + "step": 34157 + }, + { + "epoch": 0.8, + "grad_norm": 1.9291724198569613, + "learning_rate": 1.933937481939464e-06, + "loss": 1.0054, + "step": 34158 + }, + { + "epoch": 0.8, + "grad_norm": 1.929196301725746, + "learning_rate": 1.9334864801482446e-06, + "loss": 1.0091, + "step": 34159 + }, + { + "epoch": 0.8, + "grad_norm": 2.1846304761789845, + "learning_rate": 1.9330355253228237e-06, + "loss": 0.9545, + "step": 34160 + }, + { + "epoch": 0.8, + "grad_norm": 1.9404124954884627, + "learning_rate": 1.93258461746582e-06, + "loss": 0.923, + "step": 34161 + }, + { + "epoch": 0.8, + "grad_norm": 2.2994816358816914, + "learning_rate": 1.932133756579867e-06, + "loss": 0.8584, + "step": 34162 + }, + { + "epoch": 0.8, + "grad_norm": 2.2254671449664842, + "learning_rate": 1.931682942667582e-06, + "loss": 0.8649, + "step": 34163 + }, + { + "epoch": 0.8, + "grad_norm": 2.100290018971233, + "learning_rate": 1.931232175731593e-06, + "loss": 0.9935, + "step": 34164 + }, + { + "epoch": 0.8, + "grad_norm": 2.086084050841788, + "learning_rate": 1.9307814557745307e-06, + "loss": 0.9804, + "step": 34165 + }, + { + "epoch": 0.8, + "grad_norm": 1.8113162661280016, + "learning_rate": 1.930330782799008e-06, + "loss": 0.6805, + "step": 34166 + }, + { + "epoch": 0.8, + "grad_norm": 1.9641612302037244, + "learning_rate": 1.9298801568076543e-06, + "loss": 1.0705, + "step": 34167 + }, + { + "epoch": 0.8, + "grad_norm": 1.1690444474023312, + "learning_rate": 1.9294295778030924e-06, + "loss": 0.9475, + "step": 34168 + }, + { + "epoch": 0.8, + "grad_norm": 1.8768943912394862, + "learning_rate": 1.928979045787953e-06, + "loss": 0.9411, + "step": 34169 + }, + { + "epoch": 0.81, + "grad_norm": 1.8480314067072474, + "learning_rate": 1.9285285607648476e-06, + "loss": 1.0066, + "step": 34170 + }, + { + "epoch": 0.81, + "grad_norm": 2.184335827020002, + "learning_rate": 1.928078122736403e-06, + "loss": 0.9589, + "step": 34171 + }, + { + "epoch": 0.81, + "grad_norm": 2.1855300818727943, + "learning_rate": 1.927627731705247e-06, + "loss": 0.9058, + "step": 34172 + }, + { + "epoch": 0.81, + "grad_norm": 1.762948419040712, + "learning_rate": 1.9271773876739953e-06, + "loss": 1.0021, + "step": 34173 + }, + { + "epoch": 0.81, + "grad_norm": 2.067293387662931, + "learning_rate": 1.926727090645275e-06, + "loss": 0.9096, + "step": 34174 + }, + { + "epoch": 0.81, + "grad_norm": 1.9580824264022723, + "learning_rate": 1.9262768406217026e-06, + "loss": 0.887, + "step": 34175 + }, + { + "epoch": 0.81, + "grad_norm": 1.9621769470016346, + "learning_rate": 1.9258266376059053e-06, + "loss": 0.9699, + "step": 34176 + }, + { + "epoch": 0.81, + "grad_norm": 2.023871784617296, + "learning_rate": 1.9253764816004993e-06, + "loss": 0.8583, + "step": 34177 + }, + { + "epoch": 0.81, + "grad_norm": 1.9709638128031381, + "learning_rate": 1.9249263726081103e-06, + "loss": 1.0198, + "step": 34178 + }, + { + "epoch": 0.81, + "grad_norm": 2.1792099195671057, + "learning_rate": 1.924476310631357e-06, + "loss": 0.9171, + "step": 34179 + }, + { + "epoch": 0.81, + "grad_norm": 1.9153481570304915, + "learning_rate": 1.924026295672856e-06, + "loss": 1.0512, + "step": 34180 + }, + { + "epoch": 0.81, + "grad_norm": 2.035000430171582, + "learning_rate": 1.9235763277352304e-06, + "loss": 1.0036, + "step": 34181 + }, + { + "epoch": 0.81, + "grad_norm": 1.796729114690051, + "learning_rate": 1.9231264068211033e-06, + "loss": 0.8728, + "step": 34182 + }, + { + "epoch": 0.81, + "grad_norm": 1.8635968378654642, + "learning_rate": 1.922676532933092e-06, + "loss": 0.9624, + "step": 34183 + }, + { + "epoch": 0.81, + "grad_norm": 1.8803655107475217, + "learning_rate": 1.922226706073812e-06, + "loss": 0.8973, + "step": 34184 + }, + { + "epoch": 0.81, + "grad_norm": 1.961049372530464, + "learning_rate": 1.921776926245885e-06, + "loss": 0.8761, + "step": 34185 + }, + { + "epoch": 0.81, + "grad_norm": 1.8376859702739334, + "learning_rate": 1.921327193451934e-06, + "loss": 0.9007, + "step": 34186 + }, + { + "epoch": 0.81, + "grad_norm": 2.1103768535839067, + "learning_rate": 1.920877507694573e-06, + "loss": 0.9153, + "step": 34187 + }, + { + "epoch": 0.81, + "grad_norm": 2.443687061555139, + "learning_rate": 1.920427868976418e-06, + "loss": 0.9657, + "step": 34188 + }, + { + "epoch": 0.81, + "grad_norm": 1.982826294660304, + "learning_rate": 1.9199782773000908e-06, + "loss": 0.9315, + "step": 34189 + }, + { + "epoch": 0.81, + "grad_norm": 2.011072529659603, + "learning_rate": 1.91952873266821e-06, + "loss": 1.1585, + "step": 34190 + }, + { + "epoch": 0.81, + "grad_norm": 1.7192898605516211, + "learning_rate": 1.919079235083391e-06, + "loss": 0.9484, + "step": 34191 + }, + { + "epoch": 0.81, + "grad_norm": 2.0082891162697702, + "learning_rate": 1.9186297845482495e-06, + "loss": 1.0065, + "step": 34192 + }, + { + "epoch": 0.81, + "grad_norm": 2.3218808418815895, + "learning_rate": 1.9181803810654053e-06, + "loss": 0.8155, + "step": 34193 + }, + { + "epoch": 0.81, + "grad_norm": 1.1083268519530194, + "learning_rate": 1.9177310246374725e-06, + "loss": 0.8931, + "step": 34194 + }, + { + "epoch": 0.81, + "grad_norm": 2.202145611788677, + "learning_rate": 1.9172817152670696e-06, + "loss": 1.0179, + "step": 34195 + }, + { + "epoch": 0.81, + "grad_norm": 2.338189030129408, + "learning_rate": 1.9168324529568093e-06, + "loss": 0.9184, + "step": 34196 + }, + { + "epoch": 0.81, + "grad_norm": 1.9840328519274135, + "learning_rate": 1.9163832377093126e-06, + "loss": 1.052, + "step": 34197 + }, + { + "epoch": 0.81, + "grad_norm": 2.263843960388222, + "learning_rate": 1.915934069527189e-06, + "loss": 0.9847, + "step": 34198 + }, + { + "epoch": 0.81, + "grad_norm": 2.009047489646248, + "learning_rate": 1.91548494841306e-06, + "loss": 0.9322, + "step": 34199 + }, + { + "epoch": 0.81, + "grad_norm": 1.8876326926655178, + "learning_rate": 1.915035874369533e-06, + "loss": 0.9195, + "step": 34200 + }, + { + "epoch": 0.81, + "grad_norm": 2.416166685141782, + "learning_rate": 1.91458684739923e-06, + "loss": 0.948, + "step": 34201 + }, + { + "epoch": 0.81, + "grad_norm": 2.102027153032978, + "learning_rate": 1.9141378675047608e-06, + "loss": 0.8446, + "step": 34202 + }, + { + "epoch": 0.81, + "grad_norm": 1.062939054723088, + "learning_rate": 1.9136889346887423e-06, + "loss": 0.8859, + "step": 34203 + }, + { + "epoch": 0.81, + "grad_norm": 1.8561053056883654, + "learning_rate": 1.9132400489537872e-06, + "loss": 1.1367, + "step": 34204 + }, + { + "epoch": 0.81, + "grad_norm": 1.8586122800220803, + "learning_rate": 1.9127912103025058e-06, + "loss": 0.9787, + "step": 34205 + }, + { + "epoch": 0.81, + "grad_norm": 1.9307651175557368, + "learning_rate": 1.912342418737515e-06, + "loss": 0.906, + "step": 34206 + }, + { + "epoch": 0.81, + "grad_norm": 4.358443172886815, + "learning_rate": 1.911893674261429e-06, + "loss": 1.1407, + "step": 34207 + }, + { + "epoch": 0.81, + "grad_norm": 2.021560040499862, + "learning_rate": 1.9114449768768584e-06, + "loss": 1.0258, + "step": 34208 + }, + { + "epoch": 0.81, + "grad_norm": 2.1221939707879836, + "learning_rate": 1.910996326586414e-06, + "loss": 0.7855, + "step": 34209 + }, + { + "epoch": 0.81, + "grad_norm": 1.8985375782211567, + "learning_rate": 1.9105477233927105e-06, + "loss": 0.9668, + "step": 34210 + }, + { + "epoch": 0.81, + "grad_norm": 1.0745836858320401, + "learning_rate": 1.9100991672983616e-06, + "loss": 0.8671, + "step": 34211 + }, + { + "epoch": 0.81, + "grad_norm": 2.319772252352076, + "learning_rate": 1.909650658305976e-06, + "loss": 0.882, + "step": 34212 + }, + { + "epoch": 0.81, + "grad_norm": 3.0756661924374695, + "learning_rate": 1.909202196418163e-06, + "loss": 1.0553, + "step": 34213 + }, + { + "epoch": 0.81, + "grad_norm": 2.0094030400030207, + "learning_rate": 1.908753781637537e-06, + "loss": 1.0495, + "step": 34214 + }, + { + "epoch": 0.81, + "grad_norm": 2.1129896586691737, + "learning_rate": 1.9083054139667103e-06, + "loss": 0.8253, + "step": 34215 + }, + { + "epoch": 0.81, + "grad_norm": 1.7703625440705149, + "learning_rate": 1.9078570934082907e-06, + "loss": 0.88, + "step": 34216 + }, + { + "epoch": 0.81, + "grad_norm": 2.0948152288264708, + "learning_rate": 1.9074088199648866e-06, + "loss": 1.0097, + "step": 34217 + }, + { + "epoch": 0.81, + "grad_norm": 2.382610393966032, + "learning_rate": 1.9069605936391133e-06, + "loss": 0.8829, + "step": 34218 + }, + { + "epoch": 0.81, + "grad_norm": 2.085236190498583, + "learning_rate": 1.9065124144335756e-06, + "loss": 0.9733, + "step": 34219 + }, + { + "epoch": 0.81, + "grad_norm": 2.59417955992475, + "learning_rate": 1.9060642823508869e-06, + "loss": 0.9372, + "step": 34220 + }, + { + "epoch": 0.81, + "grad_norm": 1.0860572604608987, + "learning_rate": 1.9056161973936516e-06, + "loss": 0.9454, + "step": 34221 + }, + { + "epoch": 0.81, + "grad_norm": 1.7659745646376657, + "learning_rate": 1.9051681595644844e-06, + "loss": 0.9299, + "step": 34222 + }, + { + "epoch": 0.81, + "grad_norm": 1.951389316260214, + "learning_rate": 1.904720168865989e-06, + "loss": 1.0117, + "step": 34223 + }, + { + "epoch": 0.81, + "grad_norm": 1.836161050223308, + "learning_rate": 1.9042722253007772e-06, + "loss": 0.961, + "step": 34224 + }, + { + "epoch": 0.81, + "grad_norm": 1.8227378422723888, + "learning_rate": 1.9038243288714563e-06, + "loss": 1.0509, + "step": 34225 + }, + { + "epoch": 0.81, + "grad_norm": 1.1429419151320444, + "learning_rate": 1.903376479580631e-06, + "loss": 0.8907, + "step": 34226 + }, + { + "epoch": 0.81, + "grad_norm": 2.211695329848857, + "learning_rate": 1.9029286774309108e-06, + "loss": 0.9775, + "step": 34227 + }, + { + "epoch": 0.81, + "grad_norm": 1.0537296065699797, + "learning_rate": 1.9024809224249052e-06, + "loss": 0.9325, + "step": 34228 + }, + { + "epoch": 0.81, + "grad_norm": 1.9660713553795994, + "learning_rate": 1.9020332145652197e-06, + "loss": 0.9551, + "step": 34229 + }, + { + "epoch": 0.81, + "grad_norm": 1.8763795417482372, + "learning_rate": 1.9015855538544592e-06, + "loss": 0.9576, + "step": 34230 + }, + { + "epoch": 0.81, + "grad_norm": 1.861337412293632, + "learning_rate": 1.90113794029523e-06, + "loss": 0.9641, + "step": 34231 + }, + { + "epoch": 0.81, + "grad_norm": 2.2027721888725735, + "learning_rate": 1.9006903738901427e-06, + "loss": 1.1616, + "step": 34232 + }, + { + "epoch": 0.81, + "grad_norm": 1.0768478553240266, + "learning_rate": 1.9002428546418005e-06, + "loss": 0.9895, + "step": 34233 + }, + { + "epoch": 0.81, + "grad_norm": 2.0943943882930296, + "learning_rate": 1.8997953825528071e-06, + "loss": 0.9392, + "step": 34234 + }, + { + "epoch": 0.81, + "grad_norm": 2.0148953986439944, + "learning_rate": 1.8993479576257678e-06, + "loss": 0.8428, + "step": 34235 + }, + { + "epoch": 0.81, + "grad_norm": 2.0128630420624196, + "learning_rate": 1.898900579863293e-06, + "loss": 0.9347, + "step": 34236 + }, + { + "epoch": 0.81, + "grad_norm": 2.0123746994632956, + "learning_rate": 1.898453249267983e-06, + "loss": 1.0604, + "step": 34237 + }, + { + "epoch": 0.81, + "grad_norm": 2.035665331949499, + "learning_rate": 1.898005965842441e-06, + "loss": 1.0015, + "step": 34238 + }, + { + "epoch": 0.81, + "grad_norm": 2.0066592238464485, + "learning_rate": 1.8975587295892761e-06, + "loss": 0.8831, + "step": 34239 + }, + { + "epoch": 0.81, + "grad_norm": 2.4424648992760516, + "learning_rate": 1.8971115405110863e-06, + "loss": 0.9235, + "step": 34240 + }, + { + "epoch": 0.81, + "grad_norm": 2.2294762682252403, + "learning_rate": 1.8966643986104816e-06, + "loss": 1.0638, + "step": 34241 + }, + { + "epoch": 0.81, + "grad_norm": 1.0768172869560133, + "learning_rate": 1.896217303890059e-06, + "loss": 0.8454, + "step": 34242 + }, + { + "epoch": 0.81, + "grad_norm": 1.159078456616994, + "learning_rate": 1.8957702563524282e-06, + "loss": 0.8876, + "step": 34243 + }, + { + "epoch": 0.81, + "grad_norm": 1.9208454614217914, + "learning_rate": 1.8953232560001855e-06, + "loss": 1.0691, + "step": 34244 + }, + { + "epoch": 0.81, + "grad_norm": 1.1153153421001558, + "learning_rate": 1.8948763028359396e-06, + "loss": 0.9628, + "step": 34245 + }, + { + "epoch": 0.81, + "grad_norm": 1.8622890000665275, + "learning_rate": 1.8944293968622896e-06, + "loss": 0.8075, + "step": 34246 + }, + { + "epoch": 0.81, + "grad_norm": 2.095168065513702, + "learning_rate": 1.8939825380818354e-06, + "loss": 0.9378, + "step": 34247 + }, + { + "epoch": 0.81, + "grad_norm": 1.971748690580946, + "learning_rate": 1.893535726497182e-06, + "loss": 0.8496, + "step": 34248 + }, + { + "epoch": 0.81, + "grad_norm": 2.2583547690871737, + "learning_rate": 1.8930889621109295e-06, + "loss": 0.9823, + "step": 34249 + }, + { + "epoch": 0.81, + "grad_norm": 2.01164990811934, + "learning_rate": 1.8926422449256842e-06, + "loss": 1.1044, + "step": 34250 + }, + { + "epoch": 0.81, + "grad_norm": 1.892872803969251, + "learning_rate": 1.892195574944038e-06, + "loss": 0.8894, + "step": 34251 + }, + { + "epoch": 0.81, + "grad_norm": 1.88044301702538, + "learning_rate": 1.8917489521685961e-06, + "loss": 0.9967, + "step": 34252 + }, + { + "epoch": 0.81, + "grad_norm": 1.8449778259491183, + "learning_rate": 1.891302376601961e-06, + "loss": 1.0675, + "step": 34253 + }, + { + "epoch": 0.81, + "grad_norm": 1.1694862692323982, + "learning_rate": 1.890855848246731e-06, + "loss": 1.0444, + "step": 34254 + }, + { + "epoch": 0.81, + "grad_norm": 1.9599592296049302, + "learning_rate": 1.8904093671055023e-06, + "loss": 1.0212, + "step": 34255 + }, + { + "epoch": 0.81, + "grad_norm": 1.8539421863947103, + "learning_rate": 1.8899629331808788e-06, + "loss": 0.8802, + "step": 34256 + }, + { + "epoch": 0.81, + "grad_norm": 2.1473517424298443, + "learning_rate": 1.8895165464754605e-06, + "loss": 0.8202, + "step": 34257 + }, + { + "epoch": 0.81, + "grad_norm": 1.783220235955578, + "learning_rate": 1.8890702069918442e-06, + "loss": 1.0152, + "step": 34258 + }, + { + "epoch": 0.81, + "grad_norm": 1.9499014839384061, + "learning_rate": 1.888623914732628e-06, + "loss": 1.0201, + "step": 34259 + }, + { + "epoch": 0.81, + "grad_norm": 1.979220935494216, + "learning_rate": 1.8881776697004128e-06, + "loss": 0.9158, + "step": 34260 + }, + { + "epoch": 0.81, + "grad_norm": 1.0562423135563124, + "learning_rate": 1.8877314718977923e-06, + "loss": 0.9267, + "step": 34261 + }, + { + "epoch": 0.81, + "grad_norm": 1.8340042989662186, + "learning_rate": 1.887285321327369e-06, + "loss": 1.0143, + "step": 34262 + }, + { + "epoch": 0.81, + "grad_norm": 1.0553543226842623, + "learning_rate": 1.8868392179917404e-06, + "loss": 0.9222, + "step": 34263 + }, + { + "epoch": 0.81, + "grad_norm": 1.958705978053586, + "learning_rate": 1.8863931618935028e-06, + "loss": 1.2044, + "step": 34264 + }, + { + "epoch": 0.81, + "grad_norm": 1.9738549026597594, + "learning_rate": 1.8859471530352502e-06, + "loss": 1.0673, + "step": 34265 + }, + { + "epoch": 0.81, + "grad_norm": 2.5399574569534686, + "learning_rate": 1.885501191419582e-06, + "loss": 1.059, + "step": 34266 + }, + { + "epoch": 0.81, + "grad_norm": 2.002117356681588, + "learning_rate": 1.8850552770491004e-06, + "loss": 0.9931, + "step": 34267 + }, + { + "epoch": 0.81, + "grad_norm": 2.148261871311874, + "learning_rate": 1.8846094099263911e-06, + "loss": 0.9059, + "step": 34268 + }, + { + "epoch": 0.81, + "grad_norm": 2.0615361982191884, + "learning_rate": 1.8841635900540544e-06, + "loss": 1.0657, + "step": 34269 + }, + { + "epoch": 0.81, + "grad_norm": 1.814461350807293, + "learning_rate": 1.8837178174346882e-06, + "loss": 1.0245, + "step": 34270 + }, + { + "epoch": 0.81, + "grad_norm": 1.9133707144168643, + "learning_rate": 1.883272092070888e-06, + "loss": 1.0504, + "step": 34271 + }, + { + "epoch": 0.81, + "grad_norm": 1.916312556585147, + "learning_rate": 1.8828264139652474e-06, + "loss": 0.9048, + "step": 34272 + }, + { + "epoch": 0.81, + "grad_norm": 2.0680626349818496, + "learning_rate": 1.882380783120359e-06, + "loss": 0.9806, + "step": 34273 + }, + { + "epoch": 0.81, + "grad_norm": 2.0772078702429115, + "learning_rate": 1.8819351995388214e-06, + "loss": 0.9279, + "step": 34274 + }, + { + "epoch": 0.81, + "grad_norm": 2.0790359722791183, + "learning_rate": 1.881489663223225e-06, + "loss": 0.9869, + "step": 34275 + }, + { + "epoch": 0.81, + "grad_norm": 1.9973259592581656, + "learning_rate": 1.8810441741761687e-06, + "loss": 0.8946, + "step": 34276 + }, + { + "epoch": 0.81, + "grad_norm": 4.944151227743702, + "learning_rate": 1.8805987324002418e-06, + "loss": 0.7623, + "step": 34277 + }, + { + "epoch": 0.81, + "grad_norm": 2.354857511847981, + "learning_rate": 1.8801533378980407e-06, + "loss": 0.9973, + "step": 34278 + }, + { + "epoch": 0.81, + "grad_norm": 1.0304015879908663, + "learning_rate": 1.879707990672156e-06, + "loss": 0.9411, + "step": 34279 + }, + { + "epoch": 0.81, + "grad_norm": 2.007646804542494, + "learning_rate": 1.8792626907251855e-06, + "loss": 0.9381, + "step": 34280 + }, + { + "epoch": 0.81, + "grad_norm": 1.8233797553213988, + "learning_rate": 1.878817438059718e-06, + "loss": 0.8727, + "step": 34281 + }, + { + "epoch": 0.81, + "grad_norm": 2.6472948817109376, + "learning_rate": 1.8783722326783427e-06, + "loss": 1.0235, + "step": 34282 + }, + { + "epoch": 0.81, + "grad_norm": 2.0097636121113283, + "learning_rate": 1.8779270745836576e-06, + "loss": 0.8362, + "step": 34283 + }, + { + "epoch": 0.81, + "grad_norm": 2.0682704996947203, + "learning_rate": 1.8774819637782538e-06, + "loss": 1.0027, + "step": 34284 + }, + { + "epoch": 0.81, + "grad_norm": 2.041965202124045, + "learning_rate": 1.8770369002647215e-06, + "loss": 1.0105, + "step": 34285 + }, + { + "epoch": 0.81, + "grad_norm": 2.0991795740901456, + "learning_rate": 1.87659188404565e-06, + "loss": 1.0039, + "step": 34286 + }, + { + "epoch": 0.81, + "grad_norm": 2.0839773084579685, + "learning_rate": 1.8761469151236322e-06, + "loss": 0.9679, + "step": 34287 + }, + { + "epoch": 0.81, + "grad_norm": 1.9953247449338773, + "learning_rate": 1.875701993501261e-06, + "loss": 1.0602, + "step": 34288 + }, + { + "epoch": 0.81, + "grad_norm": 2.0261762654317796, + "learning_rate": 1.8752571191811253e-06, + "loss": 0.9626, + "step": 34289 + }, + { + "epoch": 0.81, + "grad_norm": 2.306559599717885, + "learning_rate": 1.874812292165813e-06, + "loss": 1.0037, + "step": 34290 + }, + { + "epoch": 0.81, + "grad_norm": 2.1907557098389043, + "learning_rate": 1.8743675124579152e-06, + "loss": 1.0226, + "step": 34291 + }, + { + "epoch": 0.81, + "grad_norm": 1.9783835063346564, + "learning_rate": 1.8739227800600246e-06, + "loss": 0.8459, + "step": 34292 + }, + { + "epoch": 0.81, + "grad_norm": 1.8222173989017774, + "learning_rate": 1.873478094974729e-06, + "loss": 1.0662, + "step": 34293 + }, + { + "epoch": 0.81, + "grad_norm": 1.8215191937347441, + "learning_rate": 1.8730334572046138e-06, + "loss": 1.11, + "step": 34294 + }, + { + "epoch": 0.81, + "grad_norm": 1.9141781160167832, + "learning_rate": 1.872588866752273e-06, + "loss": 1.0479, + "step": 34295 + }, + { + "epoch": 0.81, + "grad_norm": 2.3583525400454124, + "learning_rate": 1.8721443236202907e-06, + "loss": 1.0161, + "step": 34296 + }, + { + "epoch": 0.81, + "grad_norm": 2.011693848298401, + "learning_rate": 1.8716998278112597e-06, + "loss": 1.027, + "step": 34297 + }, + { + "epoch": 0.81, + "grad_norm": 2.403520960145482, + "learning_rate": 1.8712553793277643e-06, + "loss": 0.9772, + "step": 34298 + }, + { + "epoch": 0.81, + "grad_norm": 2.092334894281274, + "learning_rate": 1.8708109781723949e-06, + "loss": 0.8998, + "step": 34299 + }, + { + "epoch": 0.81, + "grad_norm": 1.9259430794831973, + "learning_rate": 1.8703666243477347e-06, + "loss": 0.9605, + "step": 34300 + }, + { + "epoch": 0.81, + "grad_norm": 1.9031465260467848, + "learning_rate": 1.8699223178563775e-06, + "loss": 1.0811, + "step": 34301 + }, + { + "epoch": 0.81, + "grad_norm": 2.232631494120683, + "learning_rate": 1.8694780587009065e-06, + "loss": 1.0945, + "step": 34302 + }, + { + "epoch": 0.81, + "grad_norm": 1.8962204025850307, + "learning_rate": 1.8690338468839053e-06, + "loss": 0.9934, + "step": 34303 + }, + { + "epoch": 0.81, + "grad_norm": 2.317290235419235, + "learning_rate": 1.868589682407963e-06, + "loss": 1.0286, + "step": 34304 + }, + { + "epoch": 0.81, + "grad_norm": 2.082490708676033, + "learning_rate": 1.8681455652756685e-06, + "loss": 0.9569, + "step": 34305 + }, + { + "epoch": 0.81, + "grad_norm": 1.165178933264922, + "learning_rate": 1.8677014954896055e-06, + "loss": 0.9364, + "step": 34306 + }, + { + "epoch": 0.81, + "grad_norm": 1.947308229001391, + "learning_rate": 1.8672574730523563e-06, + "loss": 0.941, + "step": 34307 + }, + { + "epoch": 0.81, + "grad_norm": 2.12527180351865, + "learning_rate": 1.8668134979665086e-06, + "loss": 0.9267, + "step": 34308 + }, + { + "epoch": 0.81, + "grad_norm": 3.229745901583665, + "learning_rate": 1.8663695702346508e-06, + "loss": 1.0807, + "step": 34309 + }, + { + "epoch": 0.81, + "grad_norm": 2.378000820552928, + "learning_rate": 1.8659256898593648e-06, + "loss": 0.8999, + "step": 34310 + }, + { + "epoch": 0.81, + "grad_norm": 1.835076305372343, + "learning_rate": 1.8654818568432308e-06, + "loss": 0.7654, + "step": 34311 + }, + { + "epoch": 0.81, + "grad_norm": 1.9045033888834135, + "learning_rate": 1.8650380711888372e-06, + "loss": 0.976, + "step": 34312 + }, + { + "epoch": 0.81, + "grad_norm": 1.1149190825088295, + "learning_rate": 1.8645943328987704e-06, + "loss": 0.9366, + "step": 34313 + }, + { + "epoch": 0.81, + "grad_norm": 2.03485184084243, + "learning_rate": 1.8641506419756105e-06, + "loss": 0.9962, + "step": 34314 + }, + { + "epoch": 0.81, + "grad_norm": 2.0169296928850735, + "learning_rate": 1.8637069984219391e-06, + "loss": 0.9724, + "step": 34315 + }, + { + "epoch": 0.81, + "grad_norm": 2.350855004260368, + "learning_rate": 1.863263402240344e-06, + "loss": 0.9829, + "step": 34316 + }, + { + "epoch": 0.81, + "grad_norm": 2.1666940842477076, + "learning_rate": 1.862819853433403e-06, + "loss": 0.8769, + "step": 34317 + }, + { + "epoch": 0.81, + "grad_norm": 2.3013912676019475, + "learning_rate": 1.862376352003703e-06, + "loss": 0.9156, + "step": 34318 + }, + { + "epoch": 0.81, + "grad_norm": 4.406722663180712, + "learning_rate": 1.8619328979538221e-06, + "loss": 1.0133, + "step": 34319 + }, + { + "epoch": 0.81, + "grad_norm": 1.953869427552685, + "learning_rate": 1.8614894912863469e-06, + "loss": 1.0403, + "step": 34320 + }, + { + "epoch": 0.81, + "grad_norm": 2.0372680244999626, + "learning_rate": 1.8610461320038541e-06, + "loss": 1.0802, + "step": 34321 + }, + { + "epoch": 0.81, + "grad_norm": 1.9099166147890116, + "learning_rate": 1.8606028201089298e-06, + "loss": 1.1472, + "step": 34322 + }, + { + "epoch": 0.81, + "grad_norm": 1.9802300924362333, + "learning_rate": 1.860159555604153e-06, + "loss": 1.06, + "step": 34323 + }, + { + "epoch": 0.81, + "grad_norm": 1.887996516591813, + "learning_rate": 1.8597163384921013e-06, + "loss": 1.0054, + "step": 34324 + }, + { + "epoch": 0.81, + "grad_norm": 2.0069395406924575, + "learning_rate": 1.8592731687753585e-06, + "loss": 1.0404, + "step": 34325 + }, + { + "epoch": 0.81, + "grad_norm": 2.4509900154798174, + "learning_rate": 1.8588300464565068e-06, + "loss": 1.0468, + "step": 34326 + }, + { + "epoch": 0.81, + "grad_norm": 2.600264158335353, + "learning_rate": 1.8583869715381242e-06, + "loss": 0.9082, + "step": 34327 + }, + { + "epoch": 0.81, + "grad_norm": 1.8286866958279102, + "learning_rate": 1.8579439440227886e-06, + "loss": 1.0682, + "step": 34328 + }, + { + "epoch": 0.81, + "grad_norm": 1.7567337546072097, + "learning_rate": 1.8575009639130803e-06, + "loss": 1.053, + "step": 34329 + }, + { + "epoch": 0.81, + "grad_norm": 1.1020560617282364, + "learning_rate": 1.8570580312115815e-06, + "loss": 0.9651, + "step": 34330 + }, + { + "epoch": 0.81, + "grad_norm": 2.288100327282885, + "learning_rate": 1.8566151459208692e-06, + "loss": 0.9232, + "step": 34331 + }, + { + "epoch": 0.81, + "grad_norm": 1.935261605033724, + "learning_rate": 1.856172308043519e-06, + "loss": 0.923, + "step": 34332 + }, + { + "epoch": 0.81, + "grad_norm": 2.134867350963254, + "learning_rate": 1.855729517582112e-06, + "loss": 0.997, + "step": 34333 + }, + { + "epoch": 0.81, + "grad_norm": 2.7005720451738693, + "learning_rate": 1.855286774539229e-06, + "loss": 0.996, + "step": 34334 + }, + { + "epoch": 0.81, + "grad_norm": 1.8583882979799453, + "learning_rate": 1.854844078917445e-06, + "loss": 1.0462, + "step": 34335 + }, + { + "epoch": 0.81, + "grad_norm": 2.1472783540161395, + "learning_rate": 1.8544014307193348e-06, + "loss": 0.8745, + "step": 34336 + }, + { + "epoch": 0.81, + "grad_norm": 2.277155471182814, + "learning_rate": 1.8539588299474775e-06, + "loss": 0.931, + "step": 34337 + }, + { + "epoch": 0.81, + "grad_norm": 1.923262857158707, + "learning_rate": 1.8535162766044544e-06, + "loss": 0.9719, + "step": 34338 + }, + { + "epoch": 0.81, + "grad_norm": 1.9811630643163525, + "learning_rate": 1.853073770692838e-06, + "loss": 1.0516, + "step": 34339 + }, + { + "epoch": 0.81, + "grad_norm": 2.277158634066605, + "learning_rate": 1.852631312215204e-06, + "loss": 1.1496, + "step": 34340 + }, + { + "epoch": 0.81, + "grad_norm": 1.9105788619798045, + "learning_rate": 1.8521889011741312e-06, + "loss": 0.9056, + "step": 34341 + }, + { + "epoch": 0.81, + "grad_norm": 1.99644107372338, + "learning_rate": 1.851746537572192e-06, + "loss": 1.0644, + "step": 34342 + }, + { + "epoch": 0.81, + "grad_norm": 1.9773142484339075, + "learning_rate": 1.8513042214119669e-06, + "loss": 1.0146, + "step": 34343 + }, + { + "epoch": 0.81, + "grad_norm": 2.018546070385918, + "learning_rate": 1.8508619526960258e-06, + "loss": 1.019, + "step": 34344 + }, + { + "epoch": 0.81, + "grad_norm": 1.994113890882473, + "learning_rate": 1.850419731426949e-06, + "loss": 0.9007, + "step": 34345 + }, + { + "epoch": 0.81, + "grad_norm": 2.057208179211382, + "learning_rate": 1.8499775576073064e-06, + "loss": 1.0398, + "step": 34346 + }, + { + "epoch": 0.81, + "grad_norm": 1.940041830741031, + "learning_rate": 1.849535431239674e-06, + "loss": 1.0106, + "step": 34347 + }, + { + "epoch": 0.81, + "grad_norm": 2.306035040340148, + "learning_rate": 1.8490933523266318e-06, + "loss": 0.9157, + "step": 34348 + }, + { + "epoch": 0.81, + "grad_norm": 1.802527875247612, + "learning_rate": 1.8486513208707446e-06, + "loss": 0.9298, + "step": 34349 + }, + { + "epoch": 0.81, + "grad_norm": 2.203346477930915, + "learning_rate": 1.848209336874589e-06, + "loss": 1.0906, + "step": 34350 + }, + { + "epoch": 0.81, + "grad_norm": 2.035497967802868, + "learning_rate": 1.8477674003407398e-06, + "loss": 1.0598, + "step": 34351 + }, + { + "epoch": 0.81, + "grad_norm": 2.085392516894199, + "learning_rate": 1.8473255112717747e-06, + "loss": 0.9002, + "step": 34352 + }, + { + "epoch": 0.81, + "grad_norm": 1.8177996764088133, + "learning_rate": 1.8468836696702574e-06, + "loss": 1.0098, + "step": 34353 + }, + { + "epoch": 0.81, + "grad_norm": 2.570354198092938, + "learning_rate": 1.8464418755387635e-06, + "loss": 0.8688, + "step": 34354 + }, + { + "epoch": 0.81, + "grad_norm": 1.8258289255464974, + "learning_rate": 1.846000128879869e-06, + "loss": 0.9654, + "step": 34355 + }, + { + "epoch": 0.81, + "grad_norm": 1.906860311635286, + "learning_rate": 1.8455584296961415e-06, + "loss": 1.0185, + "step": 34356 + }, + { + "epoch": 0.81, + "grad_norm": 1.9133928700441178, + "learning_rate": 1.8451167779901568e-06, + "loss": 0.8227, + "step": 34357 + }, + { + "epoch": 0.81, + "grad_norm": 1.9909158338202337, + "learning_rate": 1.8446751737644808e-06, + "loss": 0.8778, + "step": 34358 + }, + { + "epoch": 0.81, + "grad_norm": 1.8521440797649127, + "learning_rate": 1.8442336170216912e-06, + "loss": 0.98, + "step": 34359 + }, + { + "epoch": 0.81, + "grad_norm": 1.9300362661764103, + "learning_rate": 1.8437921077643516e-06, + "loss": 0.9872, + "step": 34360 + }, + { + "epoch": 0.81, + "grad_norm": 1.9615243153872564, + "learning_rate": 1.8433506459950412e-06, + "loss": 0.9537, + "step": 34361 + }, + { + "epoch": 0.81, + "grad_norm": 2.0340158148827108, + "learning_rate": 1.8429092317163244e-06, + "loss": 0.7817, + "step": 34362 + }, + { + "epoch": 0.81, + "grad_norm": 2.1047070957189704, + "learning_rate": 1.8424678649307702e-06, + "loss": 1.0579, + "step": 34363 + }, + { + "epoch": 0.81, + "grad_norm": 1.7984768017547708, + "learning_rate": 1.8420265456409502e-06, + "loss": 0.9044, + "step": 34364 + }, + { + "epoch": 0.81, + "grad_norm": 2.4737892272733637, + "learning_rate": 1.8415852738494378e-06, + "loss": 1.0737, + "step": 34365 + }, + { + "epoch": 0.81, + "grad_norm": 1.864336050494028, + "learning_rate": 1.8411440495587985e-06, + "loss": 1.1208, + "step": 34366 + }, + { + "epoch": 0.81, + "grad_norm": 2.163933859239684, + "learning_rate": 1.8407028727715992e-06, + "loss": 0.9209, + "step": 34367 + }, + { + "epoch": 0.81, + "grad_norm": 1.1874523451302972, + "learning_rate": 1.8402617434904103e-06, + "loss": 0.9889, + "step": 34368 + }, + { + "epoch": 0.81, + "grad_norm": 1.073156548316573, + "learning_rate": 1.839820661717804e-06, + "loss": 0.9612, + "step": 34369 + }, + { + "epoch": 0.81, + "grad_norm": 1.0748645836184867, + "learning_rate": 1.8393796274563458e-06, + "loss": 0.9006, + "step": 34370 + }, + { + "epoch": 0.81, + "grad_norm": 1.9328817566677101, + "learning_rate": 1.8389386407085997e-06, + "loss": 1.0813, + "step": 34371 + }, + { + "epoch": 0.81, + "grad_norm": 1.9003337607055393, + "learning_rate": 1.8384977014771365e-06, + "loss": 0.8664, + "step": 34372 + }, + { + "epoch": 0.81, + "grad_norm": 2.572429416064553, + "learning_rate": 1.8380568097645268e-06, + "loss": 0.9521, + "step": 34373 + }, + { + "epoch": 0.81, + "grad_norm": 2.115167733535745, + "learning_rate": 1.837615965573334e-06, + "loss": 0.9671, + "step": 34374 + }, + { + "epoch": 0.81, + "grad_norm": 2.072373851063202, + "learning_rate": 1.8371751689061223e-06, + "loss": 0.932, + "step": 34375 + }, + { + "epoch": 0.81, + "grad_norm": 1.8874083939593154, + "learning_rate": 1.8367344197654647e-06, + "loss": 1.0948, + "step": 34376 + }, + { + "epoch": 0.81, + "grad_norm": 2.082650531346966, + "learning_rate": 1.8362937181539198e-06, + "loss": 0.7789, + "step": 34377 + }, + { + "epoch": 0.81, + "grad_norm": 2.0298326542729215, + "learning_rate": 1.8358530640740612e-06, + "loss": 0.9512, + "step": 34378 + }, + { + "epoch": 0.81, + "grad_norm": 1.978224379035044, + "learning_rate": 1.8354124575284471e-06, + "loss": 0.9341, + "step": 34379 + }, + { + "epoch": 0.81, + "grad_norm": 1.9752396369846716, + "learning_rate": 1.8349718985196496e-06, + "loss": 0.9076, + "step": 34380 + }, + { + "epoch": 0.81, + "grad_norm": 1.96603305030811, + "learning_rate": 1.834531387050229e-06, + "loss": 0.8636, + "step": 34381 + }, + { + "epoch": 0.81, + "grad_norm": 1.0980804788508693, + "learning_rate": 1.8340909231227532e-06, + "loss": 0.9394, + "step": 34382 + }, + { + "epoch": 0.81, + "grad_norm": 2.146952507209925, + "learning_rate": 1.8336505067397858e-06, + "loss": 1.0308, + "step": 34383 + }, + { + "epoch": 0.81, + "grad_norm": 2.1135602145979613, + "learning_rate": 1.833210137903888e-06, + "loss": 1.1227, + "step": 34384 + }, + { + "epoch": 0.81, + "grad_norm": 1.9580931250139522, + "learning_rate": 1.8327698166176266e-06, + "loss": 1.0261, + "step": 34385 + }, + { + "epoch": 0.81, + "grad_norm": 1.9820613830459466, + "learning_rate": 1.8323295428835674e-06, + "loss": 0.9506, + "step": 34386 + }, + { + "epoch": 0.81, + "grad_norm": 1.9705100157473314, + "learning_rate": 1.8318893167042717e-06, + "loss": 1.0678, + "step": 34387 + }, + { + "epoch": 0.81, + "grad_norm": 1.884799111765517, + "learning_rate": 1.8314491380822985e-06, + "loss": 1.0436, + "step": 34388 + }, + { + "epoch": 0.81, + "grad_norm": 2.385146311594927, + "learning_rate": 1.8310090070202158e-06, + "loss": 1.0071, + "step": 34389 + }, + { + "epoch": 0.81, + "grad_norm": 1.9827415135804216, + "learning_rate": 1.830568923520587e-06, + "loss": 0.9562, + "step": 34390 + }, + { + "epoch": 0.81, + "grad_norm": 1.9684546613132514, + "learning_rate": 1.8301288875859724e-06, + "loss": 0.9795, + "step": 34391 + }, + { + "epoch": 0.81, + "grad_norm": 1.756332816569526, + "learning_rate": 1.829688899218932e-06, + "loss": 0.9567, + "step": 34392 + }, + { + "epoch": 0.81, + "grad_norm": 2.9519181173510742, + "learning_rate": 1.8292489584220286e-06, + "loss": 0.9696, + "step": 34393 + }, + { + "epoch": 0.81, + "grad_norm": 2.1503966495540396, + "learning_rate": 1.8288090651978274e-06, + "loss": 0.9611, + "step": 34394 + }, + { + "epoch": 0.81, + "grad_norm": 2.550452212507951, + "learning_rate": 1.8283692195488877e-06, + "loss": 1.0533, + "step": 34395 + }, + { + "epoch": 0.81, + "grad_norm": 1.0414981237171388, + "learning_rate": 1.8279294214777653e-06, + "loss": 0.9078, + "step": 34396 + }, + { + "epoch": 0.81, + "grad_norm": 1.7756732395351507, + "learning_rate": 1.8274896709870294e-06, + "loss": 0.8107, + "step": 34397 + }, + { + "epoch": 0.81, + "grad_norm": 1.9162876489442662, + "learning_rate": 1.827049968079232e-06, + "loss": 0.9936, + "step": 34398 + }, + { + "epoch": 0.81, + "grad_norm": 1.9218637476519955, + "learning_rate": 1.8266103127569401e-06, + "loss": 1.013, + "step": 34399 + }, + { + "epoch": 0.81, + "grad_norm": 2.187617716591475, + "learning_rate": 1.8261707050227084e-06, + "loss": 0.9207, + "step": 34400 + }, + { + "epoch": 0.81, + "grad_norm": 2.0295972022451023, + "learning_rate": 1.8257311448791015e-06, + "loss": 0.8979, + "step": 34401 + }, + { + "epoch": 0.81, + "grad_norm": 1.7714874079997964, + "learning_rate": 1.825291632328673e-06, + "loss": 0.9774, + "step": 34402 + }, + { + "epoch": 0.81, + "grad_norm": 2.343792796489278, + "learning_rate": 1.8248521673739872e-06, + "loss": 1.0117, + "step": 34403 + }, + { + "epoch": 0.81, + "grad_norm": 1.9375404014936743, + "learning_rate": 1.8244127500176e-06, + "loss": 0.8823, + "step": 34404 + }, + { + "epoch": 0.81, + "grad_norm": 2.0089553127656674, + "learning_rate": 1.8239733802620674e-06, + "loss": 1.0849, + "step": 34405 + }, + { + "epoch": 0.81, + "grad_norm": 1.9665382460493208, + "learning_rate": 1.8235340581099503e-06, + "loss": 1.035, + "step": 34406 + }, + { + "epoch": 0.81, + "grad_norm": 2.0885423884376664, + "learning_rate": 1.823094783563809e-06, + "loss": 1.0721, + "step": 34407 + }, + { + "epoch": 0.81, + "grad_norm": 1.8611018885452233, + "learning_rate": 1.8226555566261995e-06, + "loss": 1.0039, + "step": 34408 + }, + { + "epoch": 0.81, + "grad_norm": 1.8411211632453781, + "learning_rate": 1.8222163772996749e-06, + "loss": 0.9256, + "step": 34409 + }, + { + "epoch": 0.81, + "grad_norm": 1.8754481701383998, + "learning_rate": 1.8217772455867965e-06, + "loss": 0.9834, + "step": 34410 + }, + { + "epoch": 0.81, + "grad_norm": 1.9943049139698101, + "learning_rate": 1.8213381614901215e-06, + "loss": 0.9962, + "step": 34411 + }, + { + "epoch": 0.81, + "grad_norm": 2.2530872577066052, + "learning_rate": 1.8208991250122054e-06, + "loss": 0.9869, + "step": 34412 + }, + { + "epoch": 0.81, + "grad_norm": 1.958440620908317, + "learning_rate": 1.8204601361556019e-06, + "loss": 0.9756, + "step": 34413 + }, + { + "epoch": 0.81, + "grad_norm": 1.8431070390435669, + "learning_rate": 1.8200211949228686e-06, + "loss": 1.002, + "step": 34414 + }, + { + "epoch": 0.81, + "grad_norm": 1.7490134037897218, + "learning_rate": 1.8195823013165648e-06, + "loss": 1.0854, + "step": 34415 + }, + { + "epoch": 0.81, + "grad_norm": 1.9914165996777926, + "learning_rate": 1.819143455339243e-06, + "loss": 0.9471, + "step": 34416 + }, + { + "epoch": 0.81, + "grad_norm": 1.9342460943149395, + "learning_rate": 1.818704656993454e-06, + "loss": 0.9495, + "step": 34417 + }, + { + "epoch": 0.81, + "grad_norm": 3.079962268097513, + "learning_rate": 1.8182659062817609e-06, + "loss": 1.0431, + "step": 34418 + }, + { + "epoch": 0.81, + "grad_norm": 2.086613625766446, + "learning_rate": 1.8178272032067102e-06, + "loss": 1.0209, + "step": 34419 + }, + { + "epoch": 0.81, + "grad_norm": 1.8843507767242487, + "learning_rate": 1.8173885477708631e-06, + "loss": 1.1148, + "step": 34420 + }, + { + "epoch": 0.81, + "grad_norm": 1.0782619592110914, + "learning_rate": 1.8169499399767676e-06, + "loss": 0.9356, + "step": 34421 + }, + { + "epoch": 0.81, + "grad_norm": 2.0094712234531262, + "learning_rate": 1.8165113798269828e-06, + "loss": 0.9754, + "step": 34422 + }, + { + "epoch": 0.81, + "grad_norm": 1.83122519608381, + "learning_rate": 1.8160728673240568e-06, + "loss": 0.896, + "step": 34423 + }, + { + "epoch": 0.81, + "grad_norm": 1.9680772440357388, + "learning_rate": 1.8156344024705485e-06, + "loss": 1.0576, + "step": 34424 + }, + { + "epoch": 0.81, + "grad_norm": 2.0681570860800633, + "learning_rate": 1.8151959852690083e-06, + "loss": 1.0332, + "step": 34425 + }, + { + "epoch": 0.81, + "grad_norm": 2.012512919401803, + "learning_rate": 1.814757615721985e-06, + "loss": 0.9601, + "step": 34426 + }, + { + "epoch": 0.81, + "grad_norm": 1.871670922445728, + "learning_rate": 1.8143192938320353e-06, + "loss": 0.9728, + "step": 34427 + }, + { + "epoch": 0.81, + "grad_norm": 1.9385562869524318, + "learning_rate": 1.813881019601712e-06, + "loss": 0.863, + "step": 34428 + }, + { + "epoch": 0.81, + "grad_norm": 1.8299105563947835, + "learning_rate": 1.8134427930335651e-06, + "loss": 0.9895, + "step": 34429 + }, + { + "epoch": 0.81, + "grad_norm": 2.4743980990345897, + "learning_rate": 1.8130046141301426e-06, + "loss": 0.8076, + "step": 34430 + }, + { + "epoch": 0.81, + "grad_norm": 2.1183488202671805, + "learning_rate": 1.8125664828940005e-06, + "loss": 0.9552, + "step": 34431 + }, + { + "epoch": 0.81, + "grad_norm": 2.1886852867890654, + "learning_rate": 1.8121283993276906e-06, + "loss": 1.0174, + "step": 34432 + }, + { + "epoch": 0.81, + "grad_norm": 2.6339571962007304, + "learning_rate": 1.8116903634337612e-06, + "loss": 0.9491, + "step": 34433 + }, + { + "epoch": 0.81, + "grad_norm": 2.2378832489645535, + "learning_rate": 1.8112523752147615e-06, + "loss": 0.9097, + "step": 34434 + }, + { + "epoch": 0.81, + "grad_norm": 1.8569498541017482, + "learning_rate": 1.8108144346732414e-06, + "loss": 0.9395, + "step": 34435 + }, + { + "epoch": 0.81, + "grad_norm": 2.019144544617931, + "learning_rate": 1.810376541811756e-06, + "loss": 1.0313, + "step": 34436 + }, + { + "epoch": 0.81, + "grad_norm": 3.50331340992642, + "learning_rate": 1.8099386966328503e-06, + "loss": 0.8493, + "step": 34437 + }, + { + "epoch": 0.81, + "grad_norm": 2.617152965552669, + "learning_rate": 1.8095008991390728e-06, + "loss": 0.9788, + "step": 34438 + }, + { + "epoch": 0.81, + "grad_norm": 2.142815272291684, + "learning_rate": 1.8090631493329759e-06, + "loss": 0.8938, + "step": 34439 + }, + { + "epoch": 0.81, + "grad_norm": 1.0699414296102419, + "learning_rate": 1.8086254472171039e-06, + "loss": 0.9738, + "step": 34440 + }, + { + "epoch": 0.81, + "grad_norm": 1.8814984756666093, + "learning_rate": 1.8081877927940107e-06, + "loss": 1.1017, + "step": 34441 + }, + { + "epoch": 0.81, + "grad_norm": 1.8001620284102742, + "learning_rate": 1.8077501860662395e-06, + "loss": 0.8861, + "step": 34442 + }, + { + "epoch": 0.81, + "grad_norm": 2.118976019364227, + "learning_rate": 1.8073126270363428e-06, + "loss": 0.9766, + "step": 34443 + }, + { + "epoch": 0.81, + "grad_norm": 2.7826672271487314, + "learning_rate": 1.8068751157068632e-06, + "loss": 0.8376, + "step": 34444 + }, + { + "epoch": 0.81, + "grad_norm": 1.1002069126730138, + "learning_rate": 1.8064376520803495e-06, + "loss": 0.9153, + "step": 34445 + }, + { + "epoch": 0.81, + "grad_norm": 2.2240317281697637, + "learning_rate": 1.8060002361593565e-06, + "loss": 0.952, + "step": 34446 + }, + { + "epoch": 0.81, + "grad_norm": 1.1341500389430785, + "learning_rate": 1.8055628679464177e-06, + "loss": 0.9024, + "step": 34447 + }, + { + "epoch": 0.81, + "grad_norm": 1.0926451082490378, + "learning_rate": 1.8051255474440876e-06, + "loss": 0.8503, + "step": 34448 + }, + { + "epoch": 0.81, + "grad_norm": 1.944930826045579, + "learning_rate": 1.8046882746549099e-06, + "loss": 0.9988, + "step": 34449 + }, + { + "epoch": 0.81, + "grad_norm": 1.209256627991989, + "learning_rate": 1.8042510495814369e-06, + "loss": 0.9352, + "step": 34450 + }, + { + "epoch": 0.81, + "grad_norm": 1.9021956315258666, + "learning_rate": 1.8038138722262045e-06, + "loss": 1.006, + "step": 34451 + }, + { + "epoch": 0.81, + "grad_norm": 2.4719758631092015, + "learning_rate": 1.8033767425917625e-06, + "loss": 0.8567, + "step": 34452 + }, + { + "epoch": 0.81, + "grad_norm": 1.897709969855712, + "learning_rate": 1.802939660680656e-06, + "loss": 1.0104, + "step": 34453 + }, + { + "epoch": 0.81, + "grad_norm": 1.0567779721679051, + "learning_rate": 1.8025026264954326e-06, + "loss": 0.8942, + "step": 34454 + }, + { + "epoch": 0.81, + "grad_norm": 2.2380714480450963, + "learning_rate": 1.8020656400386338e-06, + "loss": 1.0058, + "step": 34455 + }, + { + "epoch": 0.81, + "grad_norm": 1.9030351674750103, + "learning_rate": 1.801628701312802e-06, + "loss": 1.0764, + "step": 34456 + }, + { + "epoch": 0.81, + "grad_norm": 1.8944634411907437, + "learning_rate": 1.801191810320486e-06, + "loss": 1.0658, + "step": 34457 + }, + { + "epoch": 0.81, + "grad_norm": 2.176691092139913, + "learning_rate": 1.8007549670642244e-06, + "loss": 1.0475, + "step": 34458 + }, + { + "epoch": 0.81, + "grad_norm": 1.9929153697485145, + "learning_rate": 1.8003181715465657e-06, + "loss": 1.0378, + "step": 34459 + }, + { + "epoch": 0.81, + "grad_norm": 1.8870789680605415, + "learning_rate": 1.799881423770049e-06, + "loss": 1.0075, + "step": 34460 + }, + { + "epoch": 0.81, + "grad_norm": 1.9239050182960764, + "learning_rate": 1.7994447237372204e-06, + "loss": 0.9564, + "step": 34461 + }, + { + "epoch": 0.81, + "grad_norm": 1.0626583230124669, + "learning_rate": 1.7990080714506175e-06, + "loss": 0.9599, + "step": 34462 + }, + { + "epoch": 0.81, + "grad_norm": 2.0190056668241096, + "learning_rate": 1.7985714669127897e-06, + "loss": 1.0877, + "step": 34463 + }, + { + "epoch": 0.81, + "grad_norm": 2.159173088978071, + "learning_rate": 1.7981349101262746e-06, + "loss": 0.9519, + "step": 34464 + }, + { + "epoch": 0.81, + "grad_norm": 1.9712774693441943, + "learning_rate": 1.7976984010936128e-06, + "loss": 0.9563, + "step": 34465 + }, + { + "epoch": 0.81, + "grad_norm": 2.2218907033875728, + "learning_rate": 1.7972619398173474e-06, + "loss": 1.1036, + "step": 34466 + }, + { + "epoch": 0.81, + "grad_norm": 2.4772128632649393, + "learning_rate": 1.7968255263000235e-06, + "loss": 1.0574, + "step": 34467 + }, + { + "epoch": 0.81, + "grad_norm": 2.295393301450142, + "learning_rate": 1.7963891605441775e-06, + "loss": 1.0245, + "step": 34468 + }, + { + "epoch": 0.81, + "grad_norm": 1.9876164703165615, + "learning_rate": 1.7959528425523488e-06, + "loss": 0.9501, + "step": 34469 + }, + { + "epoch": 0.81, + "grad_norm": 4.492548806701643, + "learning_rate": 1.7955165723270796e-06, + "loss": 0.9448, + "step": 34470 + }, + { + "epoch": 0.81, + "grad_norm": 1.0752158848382123, + "learning_rate": 1.7950803498709124e-06, + "loss": 0.9487, + "step": 34471 + }, + { + "epoch": 0.81, + "grad_norm": 1.8846853218117814, + "learning_rate": 1.7946441751863864e-06, + "loss": 0.8546, + "step": 34472 + }, + { + "epoch": 0.81, + "grad_norm": 2.173900274234453, + "learning_rate": 1.794208048276036e-06, + "loss": 0.9083, + "step": 34473 + }, + { + "epoch": 0.81, + "grad_norm": 1.0731379322312884, + "learning_rate": 1.793771969142405e-06, + "loss": 0.9718, + "step": 34474 + }, + { + "epoch": 0.81, + "grad_norm": 2.197600608722058, + "learning_rate": 1.7933359377880344e-06, + "loss": 1.0401, + "step": 34475 + }, + { + "epoch": 0.81, + "grad_norm": 2.0005476708031082, + "learning_rate": 1.79289995421546e-06, + "loss": 0.9945, + "step": 34476 + }, + { + "epoch": 0.81, + "grad_norm": 2.018347604862894, + "learning_rate": 1.7924640184272169e-06, + "loss": 0.88, + "step": 34477 + }, + { + "epoch": 0.81, + "grad_norm": 2.2284050878739863, + "learning_rate": 1.7920281304258502e-06, + "loss": 0.9448, + "step": 34478 + }, + { + "epoch": 0.81, + "grad_norm": 1.966780636721802, + "learning_rate": 1.7915922902138916e-06, + "loss": 0.9373, + "step": 34479 + }, + { + "epoch": 0.81, + "grad_norm": 2.365703400119167, + "learning_rate": 1.7911564977938845e-06, + "loss": 0.9516, + "step": 34480 + }, + { + "epoch": 0.81, + "grad_norm": 1.1071146334044164, + "learning_rate": 1.7907207531683601e-06, + "loss": 0.9449, + "step": 34481 + }, + { + "epoch": 0.81, + "grad_norm": 1.9226396141990134, + "learning_rate": 1.790285056339861e-06, + "loss": 0.8921, + "step": 34482 + }, + { + "epoch": 0.81, + "grad_norm": 1.7718829307402846, + "learning_rate": 1.7898494073109184e-06, + "loss": 0.9014, + "step": 34483 + }, + { + "epoch": 0.81, + "grad_norm": 1.1878613230893125, + "learning_rate": 1.7894138060840749e-06, + "loss": 0.9277, + "step": 34484 + }, + { + "epoch": 0.81, + "grad_norm": 2.2107084620279536, + "learning_rate": 1.7889782526618637e-06, + "loss": 0.9338, + "step": 34485 + }, + { + "epoch": 0.81, + "grad_norm": 2.578547582166812, + "learning_rate": 1.7885427470468185e-06, + "loss": 1.009, + "step": 34486 + }, + { + "epoch": 0.81, + "grad_norm": 1.8803400820535128, + "learning_rate": 1.7881072892414763e-06, + "loss": 0.9134, + "step": 34487 + }, + { + "epoch": 0.81, + "grad_norm": 2.0508656700106402, + "learning_rate": 1.7876718792483772e-06, + "loss": 0.8226, + "step": 34488 + }, + { + "epoch": 0.81, + "grad_norm": 2.1371218638437814, + "learning_rate": 1.7872365170700511e-06, + "loss": 1.0401, + "step": 34489 + }, + { + "epoch": 0.81, + "grad_norm": 1.9219186215034456, + "learning_rate": 1.786801202709032e-06, + "loss": 1.0473, + "step": 34490 + }, + { + "epoch": 0.81, + "grad_norm": 2.1898603276539808, + "learning_rate": 1.7863659361678566e-06, + "loss": 0.9882, + "step": 34491 + }, + { + "epoch": 0.81, + "grad_norm": 1.0154072693208085, + "learning_rate": 1.7859307174490614e-06, + "loss": 0.9207, + "step": 34492 + }, + { + "epoch": 0.81, + "grad_norm": 1.9538931851091317, + "learning_rate": 1.7854955465551783e-06, + "loss": 1.0108, + "step": 34493 + }, + { + "epoch": 0.81, + "grad_norm": 1.8748285047512718, + "learning_rate": 1.7850604234887382e-06, + "loss": 1.0552, + "step": 34494 + }, + { + "epoch": 0.81, + "grad_norm": 2.096974651943464, + "learning_rate": 1.7846253482522768e-06, + "loss": 1.0048, + "step": 34495 + }, + { + "epoch": 0.81, + "grad_norm": 1.9202692216321604, + "learning_rate": 1.7841903208483302e-06, + "loss": 0.9394, + "step": 34496 + }, + { + "epoch": 0.81, + "grad_norm": 2.0397395445488, + "learning_rate": 1.7837553412794295e-06, + "loss": 1.16, + "step": 34497 + }, + { + "epoch": 0.81, + "grad_norm": 2.218123017987141, + "learning_rate": 1.7833204095481027e-06, + "loss": 0.9433, + "step": 34498 + }, + { + "epoch": 0.81, + "grad_norm": 2.438766466985531, + "learning_rate": 1.7828855256568889e-06, + "loss": 0.9365, + "step": 34499 + }, + { + "epoch": 0.81, + "grad_norm": 2.0897903689581288, + "learning_rate": 1.782450689608315e-06, + "loss": 0.8822, + "step": 34500 + }, + { + "epoch": 0.81, + "grad_norm": 2.7914355616454536, + "learning_rate": 1.7820159014049165e-06, + "loss": 0.9955, + "step": 34501 + }, + { + "epoch": 0.81, + "grad_norm": 2.056937092644659, + "learning_rate": 1.7815811610492217e-06, + "loss": 1.0435, + "step": 34502 + }, + { + "epoch": 0.81, + "grad_norm": 1.0604544734569956, + "learning_rate": 1.781146468543765e-06, + "loss": 0.9186, + "step": 34503 + }, + { + "epoch": 0.81, + "grad_norm": 2.386865576307105, + "learning_rate": 1.7807118238910725e-06, + "loss": 0.8859, + "step": 34504 + }, + { + "epoch": 0.81, + "grad_norm": 1.9305626875630124, + "learning_rate": 1.7802772270936818e-06, + "loss": 1.0341, + "step": 34505 + }, + { + "epoch": 0.81, + "grad_norm": 1.8259495658502785, + "learning_rate": 1.779842678154119e-06, + "loss": 0.8158, + "step": 34506 + }, + { + "epoch": 0.81, + "grad_norm": 1.8771875820041617, + "learning_rate": 1.7794081770749116e-06, + "loss": 1.0849, + "step": 34507 + }, + { + "epoch": 0.81, + "grad_norm": 4.008233421641224, + "learning_rate": 1.7789737238585925e-06, + "loss": 0.9601, + "step": 34508 + }, + { + "epoch": 0.81, + "grad_norm": 1.931914615292031, + "learning_rate": 1.7785393185076939e-06, + "loss": 0.9496, + "step": 34509 + }, + { + "epoch": 0.81, + "grad_norm": 1.7666162137815242, + "learning_rate": 1.7781049610247425e-06, + "loss": 1.0117, + "step": 34510 + }, + { + "epoch": 0.81, + "grad_norm": 2.1547452367330324, + "learning_rate": 1.7776706514122644e-06, + "loss": 1.0306, + "step": 34511 + }, + { + "epoch": 0.81, + "grad_norm": 2.06218939148294, + "learning_rate": 1.7772363896727907e-06, + "loss": 1.0518, + "step": 34512 + }, + { + "epoch": 0.81, + "grad_norm": 2.435881520809702, + "learning_rate": 1.7768021758088517e-06, + "loss": 0.8138, + "step": 34513 + }, + { + "epoch": 0.81, + "grad_norm": 1.8038050834676749, + "learning_rate": 1.7763680098229752e-06, + "loss": 1.1512, + "step": 34514 + }, + { + "epoch": 0.81, + "grad_norm": 1.9898953953995124, + "learning_rate": 1.775933891717685e-06, + "loss": 0.9273, + "step": 34515 + }, + { + "epoch": 0.81, + "grad_norm": 2.0811496639775937, + "learning_rate": 1.775499821495511e-06, + "loss": 0.9637, + "step": 34516 + }, + { + "epoch": 0.81, + "grad_norm": 1.8662974764472982, + "learning_rate": 1.7750657991589827e-06, + "loss": 1.0505, + "step": 34517 + }, + { + "epoch": 0.81, + "grad_norm": 2.0068828071448617, + "learning_rate": 1.7746318247106265e-06, + "loss": 1.006, + "step": 34518 + }, + { + "epoch": 0.81, + "grad_norm": 1.0899597741992249, + "learning_rate": 1.7741978981529651e-06, + "loss": 0.9425, + "step": 34519 + }, + { + "epoch": 0.81, + "grad_norm": 1.9376004237733022, + "learning_rate": 1.7737640194885297e-06, + "loss": 0.9696, + "step": 34520 + }, + { + "epoch": 0.81, + "grad_norm": 2.2449749696317123, + "learning_rate": 1.773330188719843e-06, + "loss": 1.1123, + "step": 34521 + }, + { + "epoch": 0.81, + "grad_norm": 2.1680319952975666, + "learning_rate": 1.7728964058494336e-06, + "loss": 0.9143, + "step": 34522 + }, + { + "epoch": 0.81, + "grad_norm": 1.9129039989209964, + "learning_rate": 1.7724626708798253e-06, + "loss": 0.9738, + "step": 34523 + }, + { + "epoch": 0.81, + "grad_norm": 1.9312319229164945, + "learning_rate": 1.7720289838135452e-06, + "loss": 1.0323, + "step": 34524 + }, + { + "epoch": 0.81, + "grad_norm": 1.9265895350462834, + "learning_rate": 1.7715953446531153e-06, + "loss": 1.0355, + "step": 34525 + }, + { + "epoch": 0.81, + "grad_norm": 2.2205136183885013, + "learning_rate": 1.771161753401065e-06, + "loss": 0.9278, + "step": 34526 + }, + { + "epoch": 0.81, + "grad_norm": 1.7312306295494302, + "learning_rate": 1.7707282100599154e-06, + "loss": 0.8328, + "step": 34527 + }, + { + "epoch": 0.81, + "grad_norm": 2.240634067735187, + "learning_rate": 1.7702947146321904e-06, + "loss": 1.0139, + "step": 34528 + }, + { + "epoch": 0.81, + "grad_norm": 2.551110260842487, + "learning_rate": 1.769861267120414e-06, + "loss": 0.9022, + "step": 34529 + }, + { + "epoch": 0.81, + "grad_norm": 1.9331880207602865, + "learning_rate": 1.7694278675271115e-06, + "loss": 0.9736, + "step": 34530 + }, + { + "epoch": 0.81, + "grad_norm": 2.0007340485320073, + "learning_rate": 1.7689945158548094e-06, + "loss": 0.9045, + "step": 34531 + }, + { + "epoch": 0.81, + "grad_norm": 2.281673668032872, + "learning_rate": 1.7685612121060235e-06, + "loss": 1.0378, + "step": 34532 + }, + { + "epoch": 0.81, + "grad_norm": 1.1327287615253105, + "learning_rate": 1.7681279562832798e-06, + "loss": 0.9713, + "step": 34533 + }, + { + "epoch": 0.81, + "grad_norm": 2.752499524828117, + "learning_rate": 1.7676947483891039e-06, + "loss": 0.9607, + "step": 34534 + }, + { + "epoch": 0.81, + "grad_norm": 1.090670490285402, + "learning_rate": 1.7672615884260159e-06, + "loss": 0.9476, + "step": 34535 + }, + { + "epoch": 0.81, + "grad_norm": 2.107554523144295, + "learning_rate": 1.7668284763965337e-06, + "loss": 1.0981, + "step": 34536 + }, + { + "epoch": 0.81, + "grad_norm": 2.0890571638037723, + "learning_rate": 1.7663954123031834e-06, + "loss": 0.9698, + "step": 34537 + }, + { + "epoch": 0.81, + "grad_norm": 1.899924611583911, + "learning_rate": 1.7659623961484883e-06, + "loss": 0.923, + "step": 34538 + }, + { + "epoch": 0.81, + "grad_norm": 2.027072737875304, + "learning_rate": 1.7655294279349666e-06, + "loss": 1.0371, + "step": 34539 + }, + { + "epoch": 0.81, + "grad_norm": 1.994805108065343, + "learning_rate": 1.7650965076651383e-06, + "loss": 0.9922, + "step": 34540 + }, + { + "epoch": 0.81, + "grad_norm": 2.140852023793737, + "learning_rate": 1.764663635341527e-06, + "loss": 1.0109, + "step": 34541 + }, + { + "epoch": 0.81, + "grad_norm": 2.3314237411332295, + "learning_rate": 1.7642308109666483e-06, + "loss": 0.9381, + "step": 34542 + }, + { + "epoch": 0.81, + "grad_norm": 1.8995547154281687, + "learning_rate": 1.763798034543025e-06, + "loss": 1.0158, + "step": 34543 + }, + { + "epoch": 0.81, + "grad_norm": 2.146212334744982, + "learning_rate": 1.7633653060731803e-06, + "loss": 0.9822, + "step": 34544 + }, + { + "epoch": 0.81, + "grad_norm": 1.8600785304132805, + "learning_rate": 1.7629326255596302e-06, + "loss": 1.0268, + "step": 34545 + }, + { + "epoch": 0.81, + "grad_norm": 2.1599057020581953, + "learning_rate": 1.7624999930048913e-06, + "loss": 0.8257, + "step": 34546 + }, + { + "epoch": 0.81, + "grad_norm": 2.429295818718056, + "learning_rate": 1.7620674084114865e-06, + "loss": 1.0745, + "step": 34547 + }, + { + "epoch": 0.81, + "grad_norm": 1.8454395457935995, + "learning_rate": 1.7616348717819376e-06, + "loss": 1.0068, + "step": 34548 + }, + { + "epoch": 0.81, + "grad_norm": 1.0950395704916034, + "learning_rate": 1.761202383118754e-06, + "loss": 0.9583, + "step": 34549 + }, + { + "epoch": 0.81, + "grad_norm": 2.152605155554559, + "learning_rate": 1.7607699424244583e-06, + "loss": 0.9563, + "step": 34550 + }, + { + "epoch": 0.81, + "grad_norm": 1.844683229319633, + "learning_rate": 1.7603375497015684e-06, + "loss": 0.9975, + "step": 34551 + }, + { + "epoch": 0.81, + "grad_norm": 2.4029649297747864, + "learning_rate": 1.7599052049526045e-06, + "loss": 0.9355, + "step": 34552 + }, + { + "epoch": 0.81, + "grad_norm": 2.036532355176784, + "learning_rate": 1.759472908180081e-06, + "loss": 0.9485, + "step": 34553 + }, + { + "epoch": 0.81, + "grad_norm": 1.8162862915630817, + "learning_rate": 1.7590406593865128e-06, + "loss": 0.9309, + "step": 34554 + }, + { + "epoch": 0.81, + "grad_norm": 1.9621140943911324, + "learning_rate": 1.7586084585744213e-06, + "loss": 0.9628, + "step": 34555 + }, + { + "epoch": 0.81, + "grad_norm": 2.1354472183837117, + "learning_rate": 1.7581763057463187e-06, + "loss": 0.8702, + "step": 34556 + }, + { + "epoch": 0.81, + "grad_norm": 1.968535001639109, + "learning_rate": 1.757744200904724e-06, + "loss": 1.1094, + "step": 34557 + }, + { + "epoch": 0.81, + "grad_norm": 1.985717044125969, + "learning_rate": 1.7573121440521513e-06, + "loss": 0.9295, + "step": 34558 + }, + { + "epoch": 0.81, + "grad_norm": 1.9946850977118318, + "learning_rate": 1.756880135191118e-06, + "loss": 0.9332, + "step": 34559 + }, + { + "epoch": 0.81, + "grad_norm": 1.975569252961022, + "learning_rate": 1.756448174324137e-06, + "loss": 0.9898, + "step": 34560 + }, + { + "epoch": 0.81, + "grad_norm": 2.973000969414591, + "learning_rate": 1.7560162614537257e-06, + "loss": 1.0284, + "step": 34561 + }, + { + "epoch": 0.81, + "grad_norm": 2.0050614864308174, + "learning_rate": 1.7555843965823992e-06, + "loss": 0.9091, + "step": 34562 + }, + { + "epoch": 0.81, + "grad_norm": 2.0587040598601556, + "learning_rate": 1.7551525797126668e-06, + "loss": 0.9337, + "step": 34563 + }, + { + "epoch": 0.81, + "grad_norm": 2.179421969915914, + "learning_rate": 1.7547208108470471e-06, + "loss": 0.9749, + "step": 34564 + }, + { + "epoch": 0.81, + "grad_norm": 1.9252003284169183, + "learning_rate": 1.7542890899880548e-06, + "loss": 0.9942, + "step": 34565 + }, + { + "epoch": 0.81, + "grad_norm": 1.7506382000142313, + "learning_rate": 1.7538574171382028e-06, + "loss": 0.9704, + "step": 34566 + }, + { + "epoch": 0.81, + "grad_norm": 2.1836942595110993, + "learning_rate": 1.7534257923000008e-06, + "loss": 1.048, + "step": 34567 + }, + { + "epoch": 0.81, + "grad_norm": 1.6852914765457345, + "learning_rate": 1.752994215475965e-06, + "loss": 1.0198, + "step": 34568 + }, + { + "epoch": 0.81, + "grad_norm": 1.9896460458842786, + "learning_rate": 1.7525626866686107e-06, + "loss": 0.9484, + "step": 34569 + }, + { + "epoch": 0.81, + "grad_norm": 1.9184206559608332, + "learning_rate": 1.7521312058804462e-06, + "loss": 0.8351, + "step": 34570 + }, + { + "epoch": 0.81, + "grad_norm": 1.0891910858127012, + "learning_rate": 1.751699773113984e-06, + "loss": 0.9516, + "step": 34571 + }, + { + "epoch": 0.81, + "grad_norm": 1.038506913593566, + "learning_rate": 1.7512683883717363e-06, + "loss": 0.9059, + "step": 34572 + }, + { + "epoch": 0.81, + "grad_norm": 3.183997782713032, + "learning_rate": 1.7508370516562179e-06, + "loss": 1.0561, + "step": 34573 + }, + { + "epoch": 0.81, + "grad_norm": 2.0191074233047637, + "learning_rate": 1.7504057629699389e-06, + "loss": 0.9506, + "step": 34574 + }, + { + "epoch": 0.81, + "grad_norm": 1.9000738881302786, + "learning_rate": 1.7499745223154053e-06, + "loss": 0.976, + "step": 34575 + }, + { + "epoch": 0.81, + "grad_norm": 1.8320153042266065, + "learning_rate": 1.7495433296951325e-06, + "loss": 1.0802, + "step": 34576 + }, + { + "epoch": 0.81, + "grad_norm": 2.1084526966851285, + "learning_rate": 1.7491121851116332e-06, + "loss": 0.7973, + "step": 34577 + }, + { + "epoch": 0.81, + "grad_norm": 1.085960727874615, + "learning_rate": 1.7486810885674154e-06, + "loss": 0.9834, + "step": 34578 + }, + { + "epoch": 0.81, + "grad_norm": 2.3107205904643866, + "learning_rate": 1.7482500400649849e-06, + "loss": 0.9947, + "step": 34579 + }, + { + "epoch": 0.81, + "grad_norm": 1.8041289970788428, + "learning_rate": 1.7478190396068584e-06, + "loss": 0.8166, + "step": 34580 + }, + { + "epoch": 0.81, + "grad_norm": 1.9817060081229378, + "learning_rate": 1.7473880871955395e-06, + "loss": 1.0434, + "step": 34581 + }, + { + "epoch": 0.81, + "grad_norm": 1.8614143415284348, + "learning_rate": 1.746957182833543e-06, + "loss": 0.908, + "step": 34582 + }, + { + "epoch": 0.81, + "grad_norm": 1.9792932801884977, + "learning_rate": 1.7465263265233712e-06, + "loss": 1.0566, + "step": 34583 + }, + { + "epoch": 0.81, + "grad_norm": 2.44903403061668, + "learning_rate": 1.7460955182675387e-06, + "loss": 1.0774, + "step": 34584 + }, + { + "epoch": 0.81, + "grad_norm": 1.9932259920195736, + "learning_rate": 1.7456647580685492e-06, + "loss": 1.0429, + "step": 34585 + }, + { + "epoch": 0.81, + "grad_norm": 2.1031880338367657, + "learning_rate": 1.7452340459289152e-06, + "loss": 0.9706, + "step": 34586 + }, + { + "epoch": 0.81, + "grad_norm": 2.9244018245393315, + "learning_rate": 1.7448033818511422e-06, + "loss": 0.9845, + "step": 34587 + }, + { + "epoch": 0.81, + "grad_norm": 2.2415456257040005, + "learning_rate": 1.744372765837734e-06, + "loss": 0.95, + "step": 34588 + }, + { + "epoch": 0.81, + "grad_norm": 2.200641002056666, + "learning_rate": 1.743942197891202e-06, + "loss": 1.0179, + "step": 34589 + }, + { + "epoch": 0.81, + "grad_norm": 2.25544360822224, + "learning_rate": 1.743511678014055e-06, + "loss": 0.8816, + "step": 34590 + }, + { + "epoch": 0.81, + "grad_norm": 3.110390120421497, + "learning_rate": 1.7430812062087965e-06, + "loss": 0.9427, + "step": 34591 + }, + { + "epoch": 0.81, + "grad_norm": 1.8394071865000865, + "learning_rate": 1.7426507824779303e-06, + "loss": 0.9796, + "step": 34592 + }, + { + "epoch": 0.81, + "grad_norm": 1.993464649099734, + "learning_rate": 1.7422204068239667e-06, + "loss": 0.9212, + "step": 34593 + }, + { + "epoch": 0.82, + "grad_norm": 1.965562836972917, + "learning_rate": 1.7417900792494126e-06, + "loss": 1.0252, + "step": 34594 + }, + { + "epoch": 0.82, + "grad_norm": 2.5653643369246986, + "learning_rate": 1.7413597997567699e-06, + "loss": 0.9274, + "step": 34595 + }, + { + "epoch": 0.82, + "grad_norm": 2.0819528381659413, + "learning_rate": 1.740929568348544e-06, + "loss": 0.8662, + "step": 34596 + }, + { + "epoch": 0.82, + "grad_norm": 2.0001811718699307, + "learning_rate": 1.74049938502724e-06, + "loss": 0.9798, + "step": 34597 + }, + { + "epoch": 0.82, + "grad_norm": 1.9131942874000514, + "learning_rate": 1.7400692497953665e-06, + "loss": 1.0295, + "step": 34598 + }, + { + "epoch": 0.82, + "grad_norm": 2.3515920721338044, + "learning_rate": 1.7396391626554254e-06, + "loss": 1.0352, + "step": 34599 + }, + { + "epoch": 0.82, + "grad_norm": 1.1330214352965022, + "learning_rate": 1.7392091236099173e-06, + "loss": 0.9667, + "step": 34600 + }, + { + "epoch": 0.82, + "grad_norm": 2.080271224788169, + "learning_rate": 1.7387791326613511e-06, + "loss": 1.005, + "step": 34601 + }, + { + "epoch": 0.82, + "grad_norm": 1.9640331116412675, + "learning_rate": 1.7383491898122262e-06, + "loss": 0.9492, + "step": 34602 + }, + { + "epoch": 0.82, + "grad_norm": 1.9664795191802706, + "learning_rate": 1.7379192950650504e-06, + "loss": 1.0267, + "step": 34603 + }, + { + "epoch": 0.82, + "grad_norm": 1.963024349900377, + "learning_rate": 1.7374894484223215e-06, + "loss": 0.9793, + "step": 34604 + }, + { + "epoch": 0.82, + "grad_norm": 2.0703550037483476, + "learning_rate": 1.7370596498865467e-06, + "loss": 1.0226, + "step": 34605 + }, + { + "epoch": 0.82, + "grad_norm": 1.994902800042945, + "learning_rate": 1.736629899460225e-06, + "loss": 0.9987, + "step": 34606 + }, + { + "epoch": 0.82, + "grad_norm": 2.072718322210324, + "learning_rate": 1.736200197145863e-06, + "loss": 1.0002, + "step": 34607 + }, + { + "epoch": 0.82, + "grad_norm": 2.0455316991020536, + "learning_rate": 1.7357705429459592e-06, + "loss": 0.9558, + "step": 34608 + }, + { + "epoch": 0.82, + "grad_norm": 1.1140604480643255, + "learning_rate": 1.7353409368630126e-06, + "loss": 0.9871, + "step": 34609 + }, + { + "epoch": 0.82, + "grad_norm": 1.942884117029129, + "learning_rate": 1.7349113788995288e-06, + "loss": 0.9871, + "step": 34610 + }, + { + "epoch": 0.82, + "grad_norm": 1.0873405071108915, + "learning_rate": 1.734481869058009e-06, + "loss": 0.9607, + "step": 34611 + }, + { + "epoch": 0.82, + "grad_norm": 2.544053126574745, + "learning_rate": 1.7340524073409537e-06, + "loss": 0.9246, + "step": 34612 + }, + { + "epoch": 0.82, + "grad_norm": 1.1356556818245165, + "learning_rate": 1.7336229937508586e-06, + "loss": 0.9562, + "step": 34613 + }, + { + "epoch": 0.82, + "grad_norm": 1.1379924406635085, + "learning_rate": 1.7331936282902284e-06, + "loss": 0.9747, + "step": 34614 + }, + { + "epoch": 0.82, + "grad_norm": 1.9168378121163836, + "learning_rate": 1.7327643109615644e-06, + "loss": 0.9319, + "step": 34615 + }, + { + "epoch": 0.82, + "grad_norm": 1.9945746877858417, + "learning_rate": 1.7323350417673634e-06, + "loss": 0.9378, + "step": 34616 + }, + { + "epoch": 0.82, + "grad_norm": 2.6737670938929528, + "learning_rate": 1.7319058207101237e-06, + "loss": 1.0301, + "step": 34617 + }, + { + "epoch": 0.82, + "grad_norm": 2.4590555213698186, + "learning_rate": 1.7314766477923461e-06, + "loss": 0.9316, + "step": 34618 + }, + { + "epoch": 0.82, + "grad_norm": 1.9040726140213788, + "learning_rate": 1.7310475230165302e-06, + "loss": 0.9491, + "step": 34619 + }, + { + "epoch": 0.82, + "grad_norm": 1.909985345599064, + "learning_rate": 1.7306184463851749e-06, + "loss": 0.9373, + "step": 34620 + }, + { + "epoch": 0.82, + "grad_norm": 2.0569363008602766, + "learning_rate": 1.7301894179007738e-06, + "loss": 0.9077, + "step": 34621 + }, + { + "epoch": 0.82, + "grad_norm": 1.9778148076228483, + "learning_rate": 1.7297604375658317e-06, + "loss": 0.8903, + "step": 34622 + }, + { + "epoch": 0.82, + "grad_norm": 1.7469077218544726, + "learning_rate": 1.7293315053828385e-06, + "loss": 0.9716, + "step": 34623 + }, + { + "epoch": 0.82, + "grad_norm": 1.966215665381869, + "learning_rate": 1.7289026213542992e-06, + "loss": 1.0315, + "step": 34624 + }, + { + "epoch": 0.82, + "grad_norm": 1.9363356563310254, + "learning_rate": 1.728473785482705e-06, + "loss": 0.8857, + "step": 34625 + }, + { + "epoch": 0.82, + "grad_norm": 2.1537054128412536, + "learning_rate": 1.7280449977705572e-06, + "loss": 0.9376, + "step": 34626 + }, + { + "epoch": 0.82, + "grad_norm": 1.9310661329830359, + "learning_rate": 1.727616258220348e-06, + "loss": 0.9773, + "step": 34627 + }, + { + "epoch": 0.82, + "grad_norm": 1.819111356116826, + "learning_rate": 1.727187566834576e-06, + "loss": 0.9339, + "step": 34628 + }, + { + "epoch": 0.82, + "grad_norm": 2.0565443070726728, + "learning_rate": 1.7267589236157422e-06, + "loss": 0.9611, + "step": 34629 + }, + { + "epoch": 0.82, + "grad_norm": 1.8213525033006013, + "learning_rate": 1.7263303285663324e-06, + "loss": 0.9447, + "step": 34630 + }, + { + "epoch": 0.82, + "grad_norm": 1.0900540005925343, + "learning_rate": 1.7259017816888469e-06, + "loss": 0.9116, + "step": 34631 + }, + { + "epoch": 0.82, + "grad_norm": 1.9237705552496456, + "learning_rate": 1.72547328298578e-06, + "loss": 0.9488, + "step": 34632 + }, + { + "epoch": 0.82, + "grad_norm": 2.0331543275059953, + "learning_rate": 1.7250448324596325e-06, + "loss": 0.9462, + "step": 34633 + }, + { + "epoch": 0.82, + "grad_norm": 2.0110676310087348, + "learning_rate": 1.7246164301128888e-06, + "loss": 1.1155, + "step": 34634 + }, + { + "epoch": 0.82, + "grad_norm": 1.9437616627373377, + "learning_rate": 1.724188075948049e-06, + "loss": 1.1534, + "step": 34635 + }, + { + "epoch": 0.82, + "grad_norm": 2.35125421417342, + "learning_rate": 1.7237597699676079e-06, + "loss": 0.9508, + "step": 34636 + }, + { + "epoch": 0.82, + "grad_norm": 2.0688064844348633, + "learning_rate": 1.723331512174058e-06, + "loss": 0.8973, + "step": 34637 + }, + { + "epoch": 0.82, + "grad_norm": 1.078979070106421, + "learning_rate": 1.7229033025698894e-06, + "loss": 0.8883, + "step": 34638 + }, + { + "epoch": 0.82, + "grad_norm": 1.9780057176364962, + "learning_rate": 1.7224751411575991e-06, + "loss": 0.9601, + "step": 34639 + }, + { + "epoch": 0.82, + "grad_norm": 1.9022769498895302, + "learning_rate": 1.7220470279396806e-06, + "loss": 1.0696, + "step": 34640 + }, + { + "epoch": 0.82, + "grad_norm": 2.3310034245446527, + "learning_rate": 1.721618962918623e-06, + "loss": 0.9696, + "step": 34641 + }, + { + "epoch": 0.82, + "grad_norm": 1.8782499269549506, + "learning_rate": 1.7211909460969233e-06, + "loss": 1.0934, + "step": 34642 + }, + { + "epoch": 0.82, + "grad_norm": 2.128651536897116, + "learning_rate": 1.7207629774770718e-06, + "loss": 0.9422, + "step": 34643 + }, + { + "epoch": 0.82, + "grad_norm": 2.1745657244627417, + "learning_rate": 1.7203350570615563e-06, + "loss": 0.9783, + "step": 34644 + }, + { + "epoch": 0.82, + "grad_norm": 1.0718981418443503, + "learning_rate": 1.7199071848528703e-06, + "loss": 0.9182, + "step": 34645 + }, + { + "epoch": 0.82, + "grad_norm": 1.817278587859445, + "learning_rate": 1.7194793608535099e-06, + "loss": 1.0763, + "step": 34646 + }, + { + "epoch": 0.82, + "grad_norm": 2.29117572525634, + "learning_rate": 1.7190515850659616e-06, + "loss": 1.0472, + "step": 34647 + }, + { + "epoch": 0.82, + "grad_norm": 1.9264579817351468, + "learning_rate": 1.718623857492715e-06, + "loss": 1.0798, + "step": 34648 + }, + { + "epoch": 0.82, + "grad_norm": 2.017909826200864, + "learning_rate": 1.718196178136261e-06, + "loss": 1.0112, + "step": 34649 + }, + { + "epoch": 0.82, + "grad_norm": 2.088486564988289, + "learning_rate": 1.7177685469990946e-06, + "loss": 0.9731, + "step": 34650 + }, + { + "epoch": 0.82, + "grad_norm": 1.13377410425378, + "learning_rate": 1.7173409640837002e-06, + "loss": 0.9298, + "step": 34651 + }, + { + "epoch": 0.82, + "grad_norm": 3.701467174825389, + "learning_rate": 1.7169134293925682e-06, + "loss": 0.8456, + "step": 34652 + }, + { + "epoch": 0.82, + "grad_norm": 1.9236097158586372, + "learning_rate": 1.7164859429281876e-06, + "loss": 0.9972, + "step": 34653 + }, + { + "epoch": 0.82, + "grad_norm": 1.8452413910482626, + "learning_rate": 1.716058504693051e-06, + "loss": 0.9551, + "step": 34654 + }, + { + "epoch": 0.82, + "grad_norm": 1.8932472985806053, + "learning_rate": 1.7156311146896442e-06, + "loss": 0.9848, + "step": 34655 + }, + { + "epoch": 0.82, + "grad_norm": 2.06707895795726, + "learning_rate": 1.715203772920453e-06, + "loss": 0.967, + "step": 34656 + }, + { + "epoch": 0.82, + "grad_norm": 2.0850607822976315, + "learning_rate": 1.7147764793879718e-06, + "loss": 0.8804, + "step": 34657 + }, + { + "epoch": 0.82, + "grad_norm": 2.036560587577535, + "learning_rate": 1.7143492340946821e-06, + "loss": 0.8888, + "step": 34658 + }, + { + "epoch": 0.82, + "grad_norm": 1.7656386322680977, + "learning_rate": 1.7139220370430765e-06, + "loss": 1.0127, + "step": 34659 + }, + { + "epoch": 0.82, + "grad_norm": 1.7951105113886403, + "learning_rate": 1.7134948882356383e-06, + "loss": 1.085, + "step": 34660 + }, + { + "epoch": 0.82, + "grad_norm": 1.804041138822683, + "learning_rate": 1.7130677876748592e-06, + "loss": 0.9958, + "step": 34661 + }, + { + "epoch": 0.82, + "grad_norm": 2.05415426102613, + "learning_rate": 1.7126407353632191e-06, + "loss": 0.8882, + "step": 34662 + }, + { + "epoch": 0.82, + "grad_norm": 2.5170209298935218, + "learning_rate": 1.7122137313032116e-06, + "loss": 1.0647, + "step": 34663 + }, + { + "epoch": 0.82, + "grad_norm": 1.79352879568892, + "learning_rate": 1.7117867754973206e-06, + "loss": 0.9444, + "step": 34664 + }, + { + "epoch": 0.82, + "grad_norm": 2.1097605302170344, + "learning_rate": 1.711359867948027e-06, + "loss": 1.0655, + "step": 34665 + }, + { + "epoch": 0.82, + "grad_norm": 2.332682601661814, + "learning_rate": 1.7109330086578213e-06, + "loss": 0.9587, + "step": 34666 + }, + { + "epoch": 0.82, + "grad_norm": 2.0843810843027897, + "learning_rate": 1.7105061976291893e-06, + "loss": 0.9195, + "step": 34667 + }, + { + "epoch": 0.82, + "grad_norm": 1.9369620366572906, + "learning_rate": 1.7100794348646155e-06, + "loss": 0.8833, + "step": 34668 + }, + { + "epoch": 0.82, + "grad_norm": 1.072990451521133, + "learning_rate": 1.7096527203665814e-06, + "loss": 0.9684, + "step": 34669 + }, + { + "epoch": 0.82, + "grad_norm": 2.026339423322633, + "learning_rate": 1.7092260541375738e-06, + "loss": 0.9494, + "step": 34670 + }, + { + "epoch": 0.82, + "grad_norm": 2.2237043420539124, + "learning_rate": 1.7087994361800787e-06, + "loss": 0.8483, + "step": 34671 + }, + { + "epoch": 0.82, + "grad_norm": 1.920828291339969, + "learning_rate": 1.7083728664965793e-06, + "loss": 1.1262, + "step": 34672 + }, + { + "epoch": 0.82, + "grad_norm": 2.190061359299245, + "learning_rate": 1.707946345089555e-06, + "loss": 0.9392, + "step": 34673 + }, + { + "epoch": 0.82, + "grad_norm": 2.43923228156035, + "learning_rate": 1.7075198719614927e-06, + "loss": 1.0531, + "step": 34674 + }, + { + "epoch": 0.82, + "grad_norm": 2.092989761530788, + "learning_rate": 1.707093447114877e-06, + "loss": 0.9508, + "step": 34675 + }, + { + "epoch": 0.82, + "grad_norm": 2.0165088011071526, + "learning_rate": 1.7066670705521893e-06, + "loss": 1.0616, + "step": 34676 + }, + { + "epoch": 0.82, + "grad_norm": 2.0152778785176237, + "learning_rate": 1.7062407422759085e-06, + "loss": 1.0373, + "step": 34677 + }, + { + "epoch": 0.82, + "grad_norm": 1.0628536585509851, + "learning_rate": 1.705814462288522e-06, + "loss": 0.9164, + "step": 34678 + }, + { + "epoch": 0.82, + "grad_norm": 1.9412575345530645, + "learning_rate": 1.7053882305925085e-06, + "loss": 0.8931, + "step": 34679 + }, + { + "epoch": 0.82, + "grad_norm": 1.8326263890822603, + "learning_rate": 1.7049620471903517e-06, + "loss": 0.8109, + "step": 34680 + }, + { + "epoch": 0.82, + "grad_norm": 1.8175830796059658, + "learning_rate": 1.7045359120845307e-06, + "loss": 0.9189, + "step": 34681 + }, + { + "epoch": 0.82, + "grad_norm": 2.0651963489925054, + "learning_rate": 1.70410982527753e-06, + "loss": 0.9668, + "step": 34682 + }, + { + "epoch": 0.82, + "grad_norm": 2.2776511105139683, + "learning_rate": 1.7036837867718259e-06, + "loss": 1.083, + "step": 34683 + }, + { + "epoch": 0.82, + "grad_norm": 1.8984048455720715, + "learning_rate": 1.7032577965699037e-06, + "loss": 1.1091, + "step": 34684 + }, + { + "epoch": 0.82, + "grad_norm": 2.0791691433820176, + "learning_rate": 1.7028318546742418e-06, + "loss": 0.9356, + "step": 34685 + }, + { + "epoch": 0.82, + "grad_norm": 2.5515010809886567, + "learning_rate": 1.7024059610873167e-06, + "loss": 1.0286, + "step": 34686 + }, + { + "epoch": 0.82, + "grad_norm": 1.8071447319305654, + "learning_rate": 1.7019801158116112e-06, + "loss": 1.0053, + "step": 34687 + }, + { + "epoch": 0.82, + "grad_norm": 1.8864613184257666, + "learning_rate": 1.7015543188496065e-06, + "loss": 0.9922, + "step": 34688 + }, + { + "epoch": 0.82, + "grad_norm": 1.8279866998935934, + "learning_rate": 1.7011285702037804e-06, + "loss": 0.8934, + "step": 34689 + }, + { + "epoch": 0.82, + "grad_norm": 1.759193301145544, + "learning_rate": 1.7007028698766082e-06, + "loss": 0.8773, + "step": 34690 + }, + { + "epoch": 0.82, + "grad_norm": 2.4268798930778943, + "learning_rate": 1.7002772178705717e-06, + "loss": 1.0181, + "step": 34691 + }, + { + "epoch": 0.82, + "grad_norm": 2.589614688659997, + "learning_rate": 1.6998516141881517e-06, + "loss": 0.9856, + "step": 34692 + }, + { + "epoch": 0.82, + "grad_norm": 1.8887673449278057, + "learning_rate": 1.6994260588318235e-06, + "loss": 0.9547, + "step": 34693 + }, + { + "epoch": 0.82, + "grad_norm": 2.6351596365877765, + "learning_rate": 1.699000551804062e-06, + "loss": 0.993, + "step": 34694 + }, + { + "epoch": 0.82, + "grad_norm": 1.8874557596461028, + "learning_rate": 1.6985750931073476e-06, + "loss": 0.9451, + "step": 34695 + }, + { + "epoch": 0.82, + "grad_norm": 1.9997031092308375, + "learning_rate": 1.698149682744159e-06, + "loss": 0.9968, + "step": 34696 + }, + { + "epoch": 0.82, + "grad_norm": 2.112461515965671, + "learning_rate": 1.6977243207169714e-06, + "loss": 0.9935, + "step": 34697 + }, + { + "epoch": 0.82, + "grad_norm": 2.212640703026009, + "learning_rate": 1.697299007028259e-06, + "loss": 0.8494, + "step": 34698 + }, + { + "epoch": 0.82, + "grad_norm": 2.128654547672555, + "learning_rate": 1.6968737416805003e-06, + "loss": 0.8272, + "step": 34699 + }, + { + "epoch": 0.82, + "grad_norm": 1.9717698064234639, + "learning_rate": 1.696448524676173e-06, + "loss": 1.0274, + "step": 34700 + }, + { + "epoch": 0.82, + "grad_norm": 2.162568745987239, + "learning_rate": 1.6960233560177531e-06, + "loss": 1.0149, + "step": 34701 + }, + { + "epoch": 0.82, + "grad_norm": 2.1340936923425238, + "learning_rate": 1.6955982357077106e-06, + "loss": 1.0265, + "step": 34702 + }, + { + "epoch": 0.82, + "grad_norm": 1.9516050179685454, + "learning_rate": 1.6951731637485269e-06, + "loss": 0.9468, + "step": 34703 + }, + { + "epoch": 0.82, + "grad_norm": 1.9245402466190775, + "learning_rate": 1.6947481401426712e-06, + "loss": 1.0834, + "step": 34704 + }, + { + "epoch": 0.82, + "grad_norm": 1.0805190658197086, + "learning_rate": 1.6943231648926251e-06, + "loss": 0.9576, + "step": 34705 + }, + { + "epoch": 0.82, + "grad_norm": 2.26773483544375, + "learning_rate": 1.693898238000856e-06, + "loss": 0.8391, + "step": 34706 + }, + { + "epoch": 0.82, + "grad_norm": 2.061481136855263, + "learning_rate": 1.6934733594698426e-06, + "loss": 0.9684, + "step": 34707 + }, + { + "epoch": 0.82, + "grad_norm": 1.829121439220375, + "learning_rate": 1.693048529302056e-06, + "loss": 1.0266, + "step": 34708 + }, + { + "epoch": 0.82, + "grad_norm": 1.913509076319604, + "learning_rate": 1.692623747499973e-06, + "loss": 0.9009, + "step": 34709 + }, + { + "epoch": 0.82, + "grad_norm": 2.34429875158587, + "learning_rate": 1.6921990140660638e-06, + "loss": 0.9834, + "step": 34710 + }, + { + "epoch": 0.82, + "grad_norm": 2.0238171799101834, + "learning_rate": 1.6917743290027999e-06, + "loss": 1.1166, + "step": 34711 + }, + { + "epoch": 0.82, + "grad_norm": 2.037316181533334, + "learning_rate": 1.691349692312657e-06, + "loss": 0.9191, + "step": 34712 + }, + { + "epoch": 0.82, + "grad_norm": 1.8202271921787971, + "learning_rate": 1.6909251039981056e-06, + "loss": 0.924, + "step": 34713 + }, + { + "epoch": 0.82, + "grad_norm": 2.8911925571391865, + "learning_rate": 1.6905005640616245e-06, + "loss": 0.9463, + "step": 34714 + }, + { + "epoch": 0.82, + "grad_norm": 1.9629761030198838, + "learning_rate": 1.6900760725056742e-06, + "loss": 0.9841, + "step": 34715 + }, + { + "epoch": 0.82, + "grad_norm": 1.9969579769988322, + "learning_rate": 1.6896516293327326e-06, + "loss": 0.8612, + "step": 34716 + }, + { + "epoch": 0.82, + "grad_norm": 1.9230489215437006, + "learning_rate": 1.6892272345452732e-06, + "loss": 0.9044, + "step": 34717 + }, + { + "epoch": 0.82, + "grad_norm": 3.024363732935367, + "learning_rate": 1.688802888145763e-06, + "loss": 0.9252, + "step": 34718 + }, + { + "epoch": 0.82, + "grad_norm": 1.8902001917806484, + "learning_rate": 1.6883785901366712e-06, + "loss": 1.1338, + "step": 34719 + }, + { + "epoch": 0.82, + "grad_norm": 2.0103342412241134, + "learning_rate": 1.6879543405204712e-06, + "loss": 0.8517, + "step": 34720 + }, + { + "epoch": 0.82, + "grad_norm": 2.165117862654235, + "learning_rate": 1.6875301392996346e-06, + "loss": 1.003, + "step": 34721 + }, + { + "epoch": 0.82, + "grad_norm": 2.635315972806467, + "learning_rate": 1.6871059864766293e-06, + "loss": 0.9247, + "step": 34722 + }, + { + "epoch": 0.82, + "grad_norm": 2.146256581403565, + "learning_rate": 1.6866818820539232e-06, + "loss": 1.0276, + "step": 34723 + }, + { + "epoch": 0.82, + "grad_norm": 2.004933803930003, + "learning_rate": 1.686257826033988e-06, + "loss": 1.0198, + "step": 34724 + }, + { + "epoch": 0.82, + "grad_norm": 2.1778587857448657, + "learning_rate": 1.68583381841929e-06, + "loss": 0.8872, + "step": 34725 + }, + { + "epoch": 0.82, + "grad_norm": 1.9568403309362412, + "learning_rate": 1.6854098592123002e-06, + "loss": 0.9751, + "step": 34726 + }, + { + "epoch": 0.82, + "grad_norm": 1.9735772883260267, + "learning_rate": 1.6849859484154885e-06, + "loss": 0.9071, + "step": 34727 + }, + { + "epoch": 0.82, + "grad_norm": 1.9055378913407615, + "learning_rate": 1.6845620860313216e-06, + "loss": 0.8409, + "step": 34728 + }, + { + "epoch": 0.82, + "grad_norm": 2.024635714145347, + "learning_rate": 1.6841382720622646e-06, + "loss": 0.9041, + "step": 34729 + }, + { + "epoch": 0.82, + "grad_norm": 2.273450518149697, + "learning_rate": 1.6837145065107862e-06, + "loss": 1.0306, + "step": 34730 + }, + { + "epoch": 0.82, + "grad_norm": 2.0736719552783356, + "learning_rate": 1.6832907893793604e-06, + "loss": 1.0774, + "step": 34731 + }, + { + "epoch": 0.82, + "grad_norm": 1.9035143191869144, + "learning_rate": 1.682867120670444e-06, + "loss": 0.9984, + "step": 34732 + }, + { + "epoch": 0.82, + "grad_norm": 1.084831339497835, + "learning_rate": 1.6824435003865081e-06, + "loss": 0.9029, + "step": 34733 + }, + { + "epoch": 0.82, + "grad_norm": 2.085941591341403, + "learning_rate": 1.6820199285300199e-06, + "loss": 0.7982, + "step": 34734 + }, + { + "epoch": 0.82, + "grad_norm": 2.103206596547, + "learning_rate": 1.6815964051034472e-06, + "loss": 0.96, + "step": 34735 + }, + { + "epoch": 0.82, + "grad_norm": 1.0826798842050283, + "learning_rate": 1.6811729301092528e-06, + "loss": 0.921, + "step": 34736 + }, + { + "epoch": 0.82, + "grad_norm": 1.8867052872001722, + "learning_rate": 1.6807495035499022e-06, + "loss": 1.0049, + "step": 34737 + }, + { + "epoch": 0.82, + "grad_norm": 1.8750014855830943, + "learning_rate": 1.6803261254278635e-06, + "loss": 1.1357, + "step": 34738 + }, + { + "epoch": 0.82, + "grad_norm": 2.019238361569974, + "learning_rate": 1.6799027957455983e-06, + "loss": 0.9836, + "step": 34739 + }, + { + "epoch": 0.82, + "grad_norm": 2.0394190906031295, + "learning_rate": 1.6794795145055753e-06, + "loss": 0.9453, + "step": 34740 + }, + { + "epoch": 0.82, + "grad_norm": 2.21062460133648, + "learning_rate": 1.6790562817102541e-06, + "loss": 0.9195, + "step": 34741 + }, + { + "epoch": 0.82, + "grad_norm": 1.8593731030436464, + "learning_rate": 1.6786330973621045e-06, + "loss": 0.9717, + "step": 34742 + }, + { + "epoch": 0.82, + "grad_norm": 1.9820752692352794, + "learning_rate": 1.6782099614635848e-06, + "loss": 0.8483, + "step": 34743 + }, + { + "epoch": 0.82, + "grad_norm": 1.8910416053448222, + "learning_rate": 1.6777868740171633e-06, + "loss": 0.9382, + "step": 34744 + }, + { + "epoch": 0.82, + "grad_norm": 2.002712269864802, + "learning_rate": 1.677363835025302e-06, + "loss": 0.8003, + "step": 34745 + }, + { + "epoch": 0.82, + "grad_norm": 2.0511819534293894, + "learning_rate": 1.6769408444904612e-06, + "loss": 0.8804, + "step": 34746 + }, + { + "epoch": 0.82, + "grad_norm": 2.2697637881220123, + "learning_rate": 1.6765179024151047e-06, + "loss": 1.0178, + "step": 34747 + }, + { + "epoch": 0.82, + "grad_norm": 2.0660117739345645, + "learning_rate": 1.6760950088016991e-06, + "loss": 1.13, + "step": 34748 + }, + { + "epoch": 0.82, + "grad_norm": 2.163006917035403, + "learning_rate": 1.675672163652704e-06, + "loss": 1.1341, + "step": 34749 + }, + { + "epoch": 0.82, + "grad_norm": 1.8413926905534164, + "learning_rate": 1.6752493669705782e-06, + "loss": 0.9718, + "step": 34750 + }, + { + "epoch": 0.82, + "grad_norm": 1.9178020774885365, + "learning_rate": 1.6748266187577866e-06, + "loss": 0.9616, + "step": 34751 + }, + { + "epoch": 0.82, + "grad_norm": 1.9854487691097005, + "learning_rate": 1.6744039190167926e-06, + "loss": 1.0065, + "step": 34752 + }, + { + "epoch": 0.82, + "grad_norm": 1.899111812012467, + "learning_rate": 1.6739812677500543e-06, + "loss": 1.0092, + "step": 34753 + }, + { + "epoch": 0.82, + "grad_norm": 1.8757713863213956, + "learning_rate": 1.673558664960031e-06, + "loss": 0.7797, + "step": 34754 + }, + { + "epoch": 0.82, + "grad_norm": 1.7953872194840292, + "learning_rate": 1.673136110649185e-06, + "loss": 1.0248, + "step": 34755 + }, + { + "epoch": 0.82, + "grad_norm": 1.8821632584229835, + "learning_rate": 1.6727136048199788e-06, + "loss": 0.8398, + "step": 34756 + }, + { + "epoch": 0.82, + "grad_norm": 2.457218788525969, + "learning_rate": 1.6722911474748715e-06, + "loss": 1.0758, + "step": 34757 + }, + { + "epoch": 0.82, + "grad_norm": 2.068858601867769, + "learning_rate": 1.671868738616319e-06, + "loss": 0.8981, + "step": 34758 + }, + { + "epoch": 0.82, + "grad_norm": 2.0851205419093644, + "learning_rate": 1.6714463782467848e-06, + "loss": 0.8591, + "step": 34759 + }, + { + "epoch": 0.82, + "grad_norm": 1.7883379250572349, + "learning_rate": 1.671024066368725e-06, + "loss": 0.9695, + "step": 34760 + }, + { + "epoch": 0.82, + "grad_norm": 2.0906308613134232, + "learning_rate": 1.6706018029846016e-06, + "loss": 0.872, + "step": 34761 + }, + { + "epoch": 0.82, + "grad_norm": 2.2157350547296017, + "learning_rate": 1.6701795880968696e-06, + "loss": 0.8432, + "step": 34762 + }, + { + "epoch": 0.82, + "grad_norm": 2.127101027169031, + "learning_rate": 1.6697574217079926e-06, + "loss": 0.936, + "step": 34763 + }, + { + "epoch": 0.82, + "grad_norm": 2.179353309164923, + "learning_rate": 1.6693353038204218e-06, + "loss": 0.8148, + "step": 34764 + }, + { + "epoch": 0.82, + "grad_norm": 1.988472126410729, + "learning_rate": 1.668913234436621e-06, + "loss": 0.9888, + "step": 34765 + }, + { + "epoch": 0.82, + "grad_norm": 1.9445538650153982, + "learning_rate": 1.6684912135590447e-06, + "loss": 0.9054, + "step": 34766 + }, + { + "epoch": 0.82, + "grad_norm": 2.030469972511766, + "learning_rate": 1.6680692411901478e-06, + "loss": 1.0472, + "step": 34767 + }, + { + "epoch": 0.82, + "grad_norm": 1.9192052062169003, + "learning_rate": 1.6676473173323892e-06, + "loss": 0.9341, + "step": 34768 + }, + { + "epoch": 0.82, + "grad_norm": 1.9470276903633537, + "learning_rate": 1.6672254419882295e-06, + "loss": 1.0369, + "step": 34769 + }, + { + "epoch": 0.82, + "grad_norm": 1.9224897203293458, + "learning_rate": 1.6668036151601197e-06, + "loss": 0.9839, + "step": 34770 + }, + { + "epoch": 0.82, + "grad_norm": 2.309475217367592, + "learning_rate": 1.6663818368505168e-06, + "loss": 0.9895, + "step": 34771 + }, + { + "epoch": 0.82, + "grad_norm": 2.066024010848244, + "learning_rate": 1.6659601070618758e-06, + "loss": 0.9812, + "step": 34772 + }, + { + "epoch": 0.82, + "grad_norm": 2.0535894597545923, + "learning_rate": 1.6655384257966566e-06, + "loss": 0.9352, + "step": 34773 + }, + { + "epoch": 0.82, + "grad_norm": 2.188970735919807, + "learning_rate": 1.665116793057312e-06, + "loss": 0.9691, + "step": 34774 + }, + { + "epoch": 0.82, + "grad_norm": 1.9283488923706564, + "learning_rate": 1.6646952088462932e-06, + "loss": 0.9466, + "step": 34775 + }, + { + "epoch": 0.82, + "grad_norm": 2.5326802964921997, + "learning_rate": 1.664273673166058e-06, + "loss": 0.984, + "step": 34776 + }, + { + "epoch": 0.82, + "grad_norm": 2.7627350021661936, + "learning_rate": 1.6638521860190625e-06, + "loss": 1.022, + "step": 34777 + }, + { + "epoch": 0.82, + "grad_norm": 2.0965093828119405, + "learning_rate": 1.6634307474077594e-06, + "loss": 0.8988, + "step": 34778 + }, + { + "epoch": 0.82, + "grad_norm": 1.9567920181441296, + "learning_rate": 1.6630093573345985e-06, + "loss": 0.8872, + "step": 34779 + }, + { + "epoch": 0.82, + "grad_norm": 1.923001196242431, + "learning_rate": 1.6625880158020403e-06, + "loss": 1.0033, + "step": 34780 + }, + { + "epoch": 0.82, + "grad_norm": 1.849764071725415, + "learning_rate": 1.6621667228125305e-06, + "loss": 0.988, + "step": 34781 + }, + { + "epoch": 0.82, + "grad_norm": 1.8720624377677553, + "learning_rate": 1.6617454783685283e-06, + "loss": 0.9795, + "step": 34782 + }, + { + "epoch": 0.82, + "grad_norm": 1.9129084274235961, + "learning_rate": 1.6613242824724818e-06, + "loss": 0.8385, + "step": 34783 + }, + { + "epoch": 0.82, + "grad_norm": 1.831728204326364, + "learning_rate": 1.660903135126848e-06, + "loss": 0.9484, + "step": 34784 + }, + { + "epoch": 0.82, + "grad_norm": 1.9832195315631926, + "learning_rate": 1.6604820363340735e-06, + "loss": 1.0826, + "step": 34785 + }, + { + "epoch": 0.82, + "grad_norm": 1.9914327167510275, + "learning_rate": 1.6600609860966155e-06, + "loss": 0.9836, + "step": 34786 + }, + { + "epoch": 0.82, + "grad_norm": 2.0515745857117755, + "learning_rate": 1.659639984416922e-06, + "loss": 0.924, + "step": 34787 + }, + { + "epoch": 0.82, + "grad_norm": 2.08756963449892, + "learning_rate": 1.6592190312974433e-06, + "loss": 0.8575, + "step": 34788 + }, + { + "epoch": 0.82, + "grad_norm": 2.1691678013362314, + "learning_rate": 1.6587981267406328e-06, + "loss": 0.9927, + "step": 34789 + }, + { + "epoch": 0.82, + "grad_norm": 1.8019841154120353, + "learning_rate": 1.658377270748941e-06, + "loss": 0.8932, + "step": 34790 + }, + { + "epoch": 0.82, + "grad_norm": 1.8902701315017572, + "learning_rate": 1.6579564633248191e-06, + "loss": 1.004, + "step": 34791 + }, + { + "epoch": 0.82, + "grad_norm": 2.106088939425542, + "learning_rate": 1.657535704470713e-06, + "loss": 0.8935, + "step": 34792 + }, + { + "epoch": 0.82, + "grad_norm": 2.057944254935948, + "learning_rate": 1.6571149941890753e-06, + "loss": 0.9472, + "step": 34793 + }, + { + "epoch": 0.82, + "grad_norm": 1.97604837817807, + "learning_rate": 1.656694332482357e-06, + "loss": 1.0954, + "step": 34794 + }, + { + "epoch": 0.82, + "grad_norm": 1.0956561793335766, + "learning_rate": 1.6562737193530065e-06, + "loss": 0.9842, + "step": 34795 + }, + { + "epoch": 0.82, + "grad_norm": 2.0731445176405208, + "learning_rate": 1.6558531548034696e-06, + "loss": 0.9013, + "step": 34796 + }, + { + "epoch": 0.82, + "grad_norm": 2.1939176680513492, + "learning_rate": 1.6554326388361963e-06, + "loss": 0.8597, + "step": 34797 + }, + { + "epoch": 0.82, + "grad_norm": 1.8566692968173566, + "learning_rate": 1.6550121714536393e-06, + "loss": 0.9622, + "step": 34798 + }, + { + "epoch": 0.82, + "grad_norm": 2.026565824677263, + "learning_rate": 1.6545917526582434e-06, + "loss": 0.9657, + "step": 34799 + }, + { + "epoch": 0.82, + "grad_norm": 1.9747772117169489, + "learning_rate": 1.6541713824524542e-06, + "loss": 0.9003, + "step": 34800 + }, + { + "epoch": 0.82, + "grad_norm": 1.0138160859893492, + "learning_rate": 1.653751060838723e-06, + "loss": 0.9143, + "step": 34801 + }, + { + "epoch": 0.82, + "grad_norm": 2.178255622753687, + "learning_rate": 1.653330787819495e-06, + "loss": 0.9468, + "step": 34802 + }, + { + "epoch": 0.82, + "grad_norm": 1.814602764919746, + "learning_rate": 1.6529105633972176e-06, + "loss": 0.7823, + "step": 34803 + }, + { + "epoch": 0.82, + "grad_norm": 1.9471251180925206, + "learning_rate": 1.652490387574337e-06, + "loss": 1.0394, + "step": 34804 + }, + { + "epoch": 0.82, + "grad_norm": 1.846205363489485, + "learning_rate": 1.6520702603533012e-06, + "loss": 1.0727, + "step": 34805 + }, + { + "epoch": 0.82, + "grad_norm": 2.037737193472932, + "learning_rate": 1.6516501817365538e-06, + "loss": 0.9693, + "step": 34806 + }, + { + "epoch": 0.82, + "grad_norm": 1.9000060547782, + "learning_rate": 1.651230151726545e-06, + "loss": 0.9864, + "step": 34807 + }, + { + "epoch": 0.82, + "grad_norm": 1.8930490599736114, + "learning_rate": 1.6508101703257161e-06, + "loss": 0.9402, + "step": 34808 + }, + { + "epoch": 0.82, + "grad_norm": 3.2277809642509587, + "learning_rate": 1.6503902375365121e-06, + "loss": 0.8885, + "step": 34809 + }, + { + "epoch": 0.82, + "grad_norm": 1.9880865069359577, + "learning_rate": 1.6499703533613788e-06, + "loss": 0.8394, + "step": 34810 + }, + { + "epoch": 0.82, + "grad_norm": 1.968052768522791, + "learning_rate": 1.6495505178027626e-06, + "loss": 0.9562, + "step": 34811 + }, + { + "epoch": 0.82, + "grad_norm": 1.9789305880316292, + "learning_rate": 1.649130730863111e-06, + "loss": 0.8587, + "step": 34812 + }, + { + "epoch": 0.82, + "grad_norm": 1.9239430107629751, + "learning_rate": 1.6487109925448597e-06, + "loss": 0.8768, + "step": 34813 + }, + { + "epoch": 0.82, + "grad_norm": 1.9466828620406726, + "learning_rate": 1.6482913028504577e-06, + "loss": 1.1896, + "step": 34814 + }, + { + "epoch": 0.82, + "grad_norm": 1.8409719998499363, + "learning_rate": 1.6478716617823488e-06, + "loss": 0.9736, + "step": 34815 + }, + { + "epoch": 0.82, + "grad_norm": 2.5511455241095153, + "learning_rate": 1.6474520693429763e-06, + "loss": 0.8787, + "step": 34816 + }, + { + "epoch": 0.82, + "grad_norm": 1.7592087687833822, + "learning_rate": 1.6470325255347798e-06, + "loss": 0.9815, + "step": 34817 + }, + { + "epoch": 0.82, + "grad_norm": 2.1121824830348466, + "learning_rate": 1.6466130303602047e-06, + "loss": 1.0093, + "step": 34818 + }, + { + "epoch": 0.82, + "grad_norm": 2.071552805563367, + "learning_rate": 1.6461935838216947e-06, + "loss": 0.8254, + "step": 34819 + }, + { + "epoch": 0.82, + "grad_norm": 2.89482452619196, + "learning_rate": 1.6457741859216903e-06, + "loss": 1.1545, + "step": 34820 + }, + { + "epoch": 0.82, + "grad_norm": 2.9620048727244606, + "learning_rate": 1.6453548366626316e-06, + "loss": 0.9663, + "step": 34821 + }, + { + "epoch": 0.82, + "grad_norm": 2.1244800439124694, + "learning_rate": 1.6449355360469621e-06, + "loss": 0.9171, + "step": 34822 + }, + { + "epoch": 0.82, + "grad_norm": 1.069273196735695, + "learning_rate": 1.6445162840771256e-06, + "loss": 0.9513, + "step": 34823 + }, + { + "epoch": 0.82, + "grad_norm": 2.301650704330413, + "learning_rate": 1.644097080755558e-06, + "loss": 1.0012, + "step": 34824 + }, + { + "epoch": 0.82, + "grad_norm": 2.0507260344656135, + "learning_rate": 1.6436779260847058e-06, + "loss": 0.9979, + "step": 34825 + }, + { + "epoch": 0.82, + "grad_norm": 2.7633657071866424, + "learning_rate": 1.6432588200670053e-06, + "loss": 1.057, + "step": 34826 + }, + { + "epoch": 0.82, + "grad_norm": 2.1720248740314085, + "learning_rate": 1.6428397627048953e-06, + "loss": 1.0348, + "step": 34827 + }, + { + "epoch": 0.82, + "grad_norm": 2.0469185403886727, + "learning_rate": 1.6424207540008185e-06, + "loss": 0.9874, + "step": 34828 + }, + { + "epoch": 0.82, + "grad_norm": 2.253039161546471, + "learning_rate": 1.6420017939572164e-06, + "loss": 0.9124, + "step": 34829 + }, + { + "epoch": 0.82, + "grad_norm": 1.9736167988700841, + "learning_rate": 1.6415828825765256e-06, + "loss": 0.9965, + "step": 34830 + }, + { + "epoch": 0.82, + "grad_norm": 1.9363085405822642, + "learning_rate": 1.6411640198611834e-06, + "loss": 0.8946, + "step": 34831 + }, + { + "epoch": 0.82, + "grad_norm": 2.300444132701106, + "learning_rate": 1.6407452058136298e-06, + "loss": 0.9135, + "step": 34832 + }, + { + "epoch": 0.82, + "grad_norm": 2.3864858637785846, + "learning_rate": 1.6403264404363073e-06, + "loss": 0.9824, + "step": 34833 + }, + { + "epoch": 0.82, + "grad_norm": 2.00143240948669, + "learning_rate": 1.6399077237316507e-06, + "loss": 0.9235, + "step": 34834 + }, + { + "epoch": 0.82, + "grad_norm": 2.720580675220347, + "learning_rate": 1.639489055702096e-06, + "loss": 1.0803, + "step": 34835 + }, + { + "epoch": 0.82, + "grad_norm": 1.788535660003421, + "learning_rate": 1.6390704363500832e-06, + "loss": 1.0072, + "step": 34836 + }, + { + "epoch": 0.82, + "grad_norm": 2.105446991427633, + "learning_rate": 1.6386518656780504e-06, + "loss": 0.8257, + "step": 34837 + }, + { + "epoch": 0.82, + "grad_norm": 1.9570203367874102, + "learning_rate": 1.6382333436884345e-06, + "loss": 1.023, + "step": 34838 + }, + { + "epoch": 0.82, + "grad_norm": 2.7305925550791685, + "learning_rate": 1.6378148703836695e-06, + "loss": 0.7357, + "step": 34839 + }, + { + "epoch": 0.82, + "grad_norm": 1.9380035258626096, + "learning_rate": 1.637396445766195e-06, + "loss": 0.9715, + "step": 34840 + }, + { + "epoch": 0.82, + "grad_norm": 1.8572634768310126, + "learning_rate": 1.6369780698384442e-06, + "loss": 0.9616, + "step": 34841 + }, + { + "epoch": 0.82, + "grad_norm": 1.9815611227911054, + "learning_rate": 1.6365597426028569e-06, + "loss": 0.9306, + "step": 34842 + }, + { + "epoch": 0.82, + "grad_norm": 1.8559015773084955, + "learning_rate": 1.6361414640618645e-06, + "loss": 1.0407, + "step": 34843 + }, + { + "epoch": 0.82, + "grad_norm": 2.8967586792260747, + "learning_rate": 1.6357232342179063e-06, + "loss": 0.918, + "step": 34844 + }, + { + "epoch": 0.82, + "grad_norm": 2.199807451286737, + "learning_rate": 1.6353050530734139e-06, + "loss": 0.997, + "step": 34845 + }, + { + "epoch": 0.82, + "grad_norm": 2.2541131769812686, + "learning_rate": 1.634886920630826e-06, + "loss": 0.9972, + "step": 34846 + }, + { + "epoch": 0.82, + "grad_norm": 2.1251434987607296, + "learning_rate": 1.6344688368925743e-06, + "loss": 0.9436, + "step": 34847 + }, + { + "epoch": 0.82, + "grad_norm": 1.9779126779810492, + "learning_rate": 1.6340508018610913e-06, + "loss": 0.817, + "step": 34848 + }, + { + "epoch": 0.82, + "grad_norm": 2.103579517331017, + "learning_rate": 1.6336328155388138e-06, + "loss": 1.0508, + "step": 34849 + }, + { + "epoch": 0.82, + "grad_norm": 1.0816912916403156, + "learning_rate": 1.6332148779281765e-06, + "loss": 0.8569, + "step": 34850 + }, + { + "epoch": 0.82, + "grad_norm": 1.8320895961720518, + "learning_rate": 1.6327969890316109e-06, + "loss": 0.923, + "step": 34851 + }, + { + "epoch": 0.82, + "grad_norm": 2.0165555358742004, + "learning_rate": 1.6323791488515472e-06, + "loss": 1.0246, + "step": 34852 + }, + { + "epoch": 0.82, + "grad_norm": 1.0404685919942418, + "learning_rate": 1.6319613573904226e-06, + "loss": 0.9073, + "step": 34853 + }, + { + "epoch": 0.82, + "grad_norm": 1.1000862721007798, + "learning_rate": 1.6315436146506702e-06, + "loss": 1.0065, + "step": 34854 + }, + { + "epoch": 0.82, + "grad_norm": 1.8383621558278638, + "learning_rate": 1.6311259206347207e-06, + "loss": 0.8565, + "step": 34855 + }, + { + "epoch": 0.82, + "grad_norm": 2.315153773503203, + "learning_rate": 1.6307082753450022e-06, + "loss": 0.9401, + "step": 34856 + }, + { + "epoch": 0.82, + "grad_norm": 2.2338364098308903, + "learning_rate": 1.6302906787839501e-06, + "loss": 0.9759, + "step": 34857 + }, + { + "epoch": 0.82, + "grad_norm": 2.060209549732277, + "learning_rate": 1.6298731309539984e-06, + "loss": 0.8165, + "step": 34858 + }, + { + "epoch": 0.82, + "grad_norm": 1.7848397802371208, + "learning_rate": 1.629455631857575e-06, + "loss": 1.0817, + "step": 34859 + }, + { + "epoch": 0.82, + "grad_norm": 2.2214577817658308, + "learning_rate": 1.629038181497109e-06, + "loss": 0.8228, + "step": 34860 + }, + { + "epoch": 0.82, + "grad_norm": 2.026315437138664, + "learning_rate": 1.6286207798750342e-06, + "loss": 0.9873, + "step": 34861 + }, + { + "epoch": 0.82, + "grad_norm": 1.8945461298417219, + "learning_rate": 1.6282034269937785e-06, + "loss": 1.0773, + "step": 34862 + }, + { + "epoch": 0.82, + "grad_norm": 2.112687518876082, + "learning_rate": 1.6277861228557745e-06, + "loss": 1.1566, + "step": 34863 + }, + { + "epoch": 0.82, + "grad_norm": 1.1265563574228663, + "learning_rate": 1.627368867463448e-06, + "loss": 0.9564, + "step": 34864 + }, + { + "epoch": 0.82, + "grad_norm": 1.0754271387484173, + "learning_rate": 1.6269516608192338e-06, + "loss": 0.9391, + "step": 34865 + }, + { + "epoch": 0.82, + "grad_norm": 2.0307346964985284, + "learning_rate": 1.6265345029255542e-06, + "loss": 0.938, + "step": 34866 + }, + { + "epoch": 0.82, + "grad_norm": 1.8810941063449413, + "learning_rate": 1.6261173937848451e-06, + "loss": 0.9834, + "step": 34867 + }, + { + "epoch": 0.82, + "grad_norm": 1.9062238692912006, + "learning_rate": 1.6257003333995315e-06, + "loss": 0.8512, + "step": 34868 + }, + { + "epoch": 0.82, + "grad_norm": 2.2126544529025205, + "learning_rate": 1.6252833217720386e-06, + "loss": 1.0533, + "step": 34869 + }, + { + "epoch": 0.82, + "grad_norm": 1.8674799070929446, + "learning_rate": 1.6248663589047985e-06, + "loss": 1.0744, + "step": 34870 + }, + { + "epoch": 0.82, + "grad_norm": 1.8836716544384264, + "learning_rate": 1.6244494448002402e-06, + "loss": 0.8561, + "step": 34871 + }, + { + "epoch": 0.82, + "grad_norm": 2.151874554270084, + "learning_rate": 1.624032579460788e-06, + "loss": 0.8018, + "step": 34872 + }, + { + "epoch": 0.82, + "grad_norm": 1.1488814637212612, + "learning_rate": 1.6236157628888694e-06, + "loss": 0.9284, + "step": 34873 + }, + { + "epoch": 0.82, + "grad_norm": 1.9576334953437609, + "learning_rate": 1.6231989950869099e-06, + "loss": 0.9874, + "step": 34874 + }, + { + "epoch": 0.82, + "grad_norm": 2.605818629811183, + "learning_rate": 1.6227822760573408e-06, + "loss": 0.9393, + "step": 34875 + }, + { + "epoch": 0.82, + "grad_norm": 1.7105277927122247, + "learning_rate": 1.6223656058025861e-06, + "loss": 0.8407, + "step": 34876 + }, + { + "epoch": 0.82, + "grad_norm": 1.909669465586486, + "learning_rate": 1.621948984325068e-06, + "loss": 1.0473, + "step": 34877 + }, + { + "epoch": 0.82, + "grad_norm": 2.0326669444750274, + "learning_rate": 1.6215324116272158e-06, + "loss": 1.0684, + "step": 34878 + }, + { + "epoch": 0.82, + "grad_norm": 1.9238966513304203, + "learning_rate": 1.6211158877114574e-06, + "loss": 0.9276, + "step": 34879 + }, + { + "epoch": 0.82, + "grad_norm": 2.218460650532806, + "learning_rate": 1.6206994125802144e-06, + "loss": 0.967, + "step": 34880 + }, + { + "epoch": 0.82, + "grad_norm": 1.8208184709403172, + "learning_rate": 1.6202829862359093e-06, + "loss": 0.921, + "step": 34881 + }, + { + "epoch": 0.82, + "grad_norm": 1.904146903441495, + "learning_rate": 1.6198666086809733e-06, + "loss": 0.9187, + "step": 34882 + }, + { + "epoch": 0.82, + "grad_norm": 1.9997912929927217, + "learning_rate": 1.6194502799178236e-06, + "loss": 0.9525, + "step": 34883 + }, + { + "epoch": 0.82, + "grad_norm": 1.7318792462997037, + "learning_rate": 1.6190339999488902e-06, + "loss": 0.9113, + "step": 34884 + }, + { + "epoch": 0.82, + "grad_norm": 1.9411530556849659, + "learning_rate": 1.6186177687765926e-06, + "loss": 1.0543, + "step": 34885 + }, + { + "epoch": 0.82, + "grad_norm": 2.1882605218365248, + "learning_rate": 1.6182015864033574e-06, + "loss": 0.9357, + "step": 34886 + }, + { + "epoch": 0.82, + "grad_norm": 1.9670166629265273, + "learning_rate": 1.617785452831604e-06, + "loss": 0.9501, + "step": 34887 + }, + { + "epoch": 0.82, + "grad_norm": 2.128027676811681, + "learning_rate": 1.6173693680637592e-06, + "loss": 0.9593, + "step": 34888 + }, + { + "epoch": 0.82, + "grad_norm": 1.8907188359482978, + "learning_rate": 1.6169533321022446e-06, + "loss": 0.97, + "step": 34889 + }, + { + "epoch": 0.82, + "grad_norm": 2.9463355779769143, + "learning_rate": 1.6165373449494793e-06, + "loss": 1.0169, + "step": 34890 + }, + { + "epoch": 0.82, + "grad_norm": 2.047249545207251, + "learning_rate": 1.6161214066078878e-06, + "loss": 0.931, + "step": 34891 + }, + { + "epoch": 0.82, + "grad_norm": 1.9162919998859886, + "learning_rate": 1.615705517079894e-06, + "loss": 0.9553, + "step": 34892 + }, + { + "epoch": 0.82, + "grad_norm": 1.9984692032120928, + "learning_rate": 1.615289676367917e-06, + "loss": 0.9712, + "step": 34893 + }, + { + "epoch": 0.82, + "grad_norm": 2.274367021894804, + "learning_rate": 1.614873884474376e-06, + "loss": 0.9566, + "step": 34894 + }, + { + "epoch": 0.82, + "grad_norm": 2.443983108547129, + "learning_rate": 1.6144581414016936e-06, + "loss": 0.9392, + "step": 34895 + }, + { + "epoch": 0.82, + "grad_norm": 2.221879361475481, + "learning_rate": 1.614042447152293e-06, + "loss": 1.0493, + "step": 34896 + }, + { + "epoch": 0.82, + "grad_norm": 1.794464729713664, + "learning_rate": 1.6136268017285928e-06, + "loss": 0.8881, + "step": 34897 + }, + { + "epoch": 0.82, + "grad_norm": 1.0682644115361402, + "learning_rate": 1.6132112051330095e-06, + "loss": 0.9342, + "step": 34898 + }, + { + "epoch": 0.82, + "grad_norm": 2.3058470322596714, + "learning_rate": 1.612795657367966e-06, + "loss": 0.9103, + "step": 34899 + }, + { + "epoch": 0.82, + "grad_norm": 2.8851071433537303, + "learning_rate": 1.6123801584358845e-06, + "loss": 0.8435, + "step": 34900 + }, + { + "epoch": 0.82, + "grad_norm": 2.037397034364855, + "learning_rate": 1.6119647083391798e-06, + "loss": 1.0034, + "step": 34901 + }, + { + "epoch": 0.82, + "grad_norm": 1.9209826614104575, + "learning_rate": 1.6115493070802712e-06, + "loss": 1.0426, + "step": 34902 + }, + { + "epoch": 0.82, + "grad_norm": 1.8410454038214459, + "learning_rate": 1.61113395466158e-06, + "loss": 1.0455, + "step": 34903 + }, + { + "epoch": 0.82, + "grad_norm": 2.472537803317663, + "learning_rate": 1.61071865108552e-06, + "loss": 1.0982, + "step": 34904 + }, + { + "epoch": 0.82, + "grad_norm": 1.9543800213985711, + "learning_rate": 1.6103033963545156e-06, + "loss": 1.0733, + "step": 34905 + }, + { + "epoch": 0.82, + "grad_norm": 1.9030132589163022, + "learning_rate": 1.6098881904709774e-06, + "loss": 0.8461, + "step": 34906 + }, + { + "epoch": 0.82, + "grad_norm": 2.03270330661759, + "learning_rate": 1.6094730334373287e-06, + "loss": 0.9785, + "step": 34907 + }, + { + "epoch": 0.82, + "grad_norm": 1.9148203844330036, + "learning_rate": 1.6090579252559834e-06, + "loss": 0.9927, + "step": 34908 + }, + { + "epoch": 0.82, + "grad_norm": 1.815395611535022, + "learning_rate": 1.6086428659293585e-06, + "loss": 1.0137, + "step": 34909 + }, + { + "epoch": 0.82, + "grad_norm": 1.946791607311993, + "learning_rate": 1.6082278554598763e-06, + "loss": 0.9549, + "step": 34910 + }, + { + "epoch": 0.82, + "grad_norm": 1.9301614257805562, + "learning_rate": 1.6078128938499427e-06, + "loss": 1.0348, + "step": 34911 + }, + { + "epoch": 0.82, + "grad_norm": 2.1894571040982544, + "learning_rate": 1.6073979811019803e-06, + "loss": 0.8967, + "step": 34912 + }, + { + "epoch": 0.82, + "grad_norm": 1.9410185890986964, + "learning_rate": 1.6069831172184035e-06, + "loss": 0.9913, + "step": 34913 + }, + { + "epoch": 0.82, + "grad_norm": 2.5357568077082275, + "learning_rate": 1.6065683022016343e-06, + "loss": 0.9014, + "step": 34914 + }, + { + "epoch": 0.82, + "grad_norm": 1.979599325699224, + "learning_rate": 1.6061535360540758e-06, + "loss": 0.9008, + "step": 34915 + }, + { + "epoch": 0.82, + "grad_norm": 3.0612402439706727, + "learning_rate": 1.6057388187781486e-06, + "loss": 0.9321, + "step": 34916 + }, + { + "epoch": 0.82, + "grad_norm": 2.0506210631865507, + "learning_rate": 1.6053241503762707e-06, + "loss": 0.9294, + "step": 34917 + }, + { + "epoch": 0.82, + "grad_norm": 2.0182676481480186, + "learning_rate": 1.6049095308508521e-06, + "loss": 0.8089, + "step": 34918 + }, + { + "epoch": 0.82, + "grad_norm": 2.1895782984071293, + "learning_rate": 1.6044949602043069e-06, + "loss": 1.0283, + "step": 34919 + }, + { + "epoch": 0.82, + "grad_norm": 2.349269774819689, + "learning_rate": 1.6040804384390496e-06, + "loss": 0.9853, + "step": 34920 + }, + { + "epoch": 0.82, + "grad_norm": 2.0405275263721196, + "learning_rate": 1.603665965557497e-06, + "loss": 0.8891, + "step": 34921 + }, + { + "epoch": 0.82, + "grad_norm": 2.108541837147212, + "learning_rate": 1.6032515415620564e-06, + "loss": 1.0667, + "step": 34922 + }, + { + "epoch": 0.82, + "grad_norm": 2.0621587381110893, + "learning_rate": 1.6028371664551456e-06, + "loss": 0.9533, + "step": 34923 + }, + { + "epoch": 0.82, + "grad_norm": 2.620412851266197, + "learning_rate": 1.6024228402391762e-06, + "loss": 1.0198, + "step": 34924 + }, + { + "epoch": 0.82, + "grad_norm": 2.19883157295181, + "learning_rate": 1.6020085629165572e-06, + "loss": 1.0095, + "step": 34925 + }, + { + "epoch": 0.82, + "grad_norm": 1.8910065855258562, + "learning_rate": 1.6015943344897022e-06, + "loss": 0.9237, + "step": 34926 + }, + { + "epoch": 0.82, + "grad_norm": 2.1855096026034357, + "learning_rate": 1.6011801549610273e-06, + "loss": 1.0478, + "step": 34927 + }, + { + "epoch": 0.82, + "grad_norm": 2.06902796148892, + "learning_rate": 1.6007660243329405e-06, + "loss": 0.9499, + "step": 34928 + }, + { + "epoch": 0.82, + "grad_norm": 2.3289989420837784, + "learning_rate": 1.6003519426078507e-06, + "loss": 0.9848, + "step": 34929 + }, + { + "epoch": 0.82, + "grad_norm": 1.7704957635099208, + "learning_rate": 1.5999379097881706e-06, + "loss": 0.9138, + "step": 34930 + }, + { + "epoch": 0.82, + "grad_norm": 1.0494034371822334, + "learning_rate": 1.599523925876314e-06, + "loss": 0.8922, + "step": 34931 + }, + { + "epoch": 0.82, + "grad_norm": 2.1161613220340656, + "learning_rate": 1.5991099908746887e-06, + "loss": 0.9272, + "step": 34932 + }, + { + "epoch": 0.82, + "grad_norm": 2.2490684608917078, + "learning_rate": 1.598696104785702e-06, + "loss": 0.9316, + "step": 34933 + }, + { + "epoch": 0.82, + "grad_norm": 1.9961032863121333, + "learning_rate": 1.5982822676117672e-06, + "loss": 0.913, + "step": 34934 + }, + { + "epoch": 0.82, + "grad_norm": 1.6969942876321036, + "learning_rate": 1.5978684793552946e-06, + "loss": 1.0095, + "step": 34935 + }, + { + "epoch": 0.82, + "grad_norm": 1.9233522333654975, + "learning_rate": 1.5974547400186914e-06, + "loss": 1.0764, + "step": 34936 + }, + { + "epoch": 0.82, + "grad_norm": 2.2380629613212806, + "learning_rate": 1.5970410496043653e-06, + "loss": 1.0808, + "step": 34937 + }, + { + "epoch": 0.82, + "grad_norm": 2.1615341963749715, + "learning_rate": 1.5966274081147293e-06, + "loss": 0.9984, + "step": 34938 + }, + { + "epoch": 0.82, + "grad_norm": 2.1764517053757086, + "learning_rate": 1.5962138155521855e-06, + "loss": 1.0046, + "step": 34939 + }, + { + "epoch": 0.82, + "grad_norm": 2.23143259692795, + "learning_rate": 1.595800271919149e-06, + "loss": 0.8809, + "step": 34940 + }, + { + "epoch": 0.82, + "grad_norm": 4.016118324241835, + "learning_rate": 1.595386777218021e-06, + "loss": 0.9758, + "step": 34941 + }, + { + "epoch": 0.82, + "grad_norm": 1.2035547162909912, + "learning_rate": 1.594973331451214e-06, + "loss": 0.9752, + "step": 34942 + }, + { + "epoch": 0.82, + "grad_norm": 2.0589015270745166, + "learning_rate": 1.5945599346211315e-06, + "loss": 0.9275, + "step": 34943 + }, + { + "epoch": 0.82, + "grad_norm": 1.801299821917444, + "learning_rate": 1.5941465867301842e-06, + "loss": 0.8466, + "step": 34944 + }, + { + "epoch": 0.82, + "grad_norm": 2.1084517849423, + "learning_rate": 1.5937332877807753e-06, + "loss": 0.9665, + "step": 34945 + }, + { + "epoch": 0.82, + "grad_norm": 1.0484320763674717, + "learning_rate": 1.5933200377753145e-06, + "loss": 0.9321, + "step": 34946 + }, + { + "epoch": 0.82, + "grad_norm": 1.9444431700946518, + "learning_rate": 1.5929068367162026e-06, + "loss": 0.9812, + "step": 34947 + }, + { + "epoch": 0.82, + "grad_norm": 2.3040783032041774, + "learning_rate": 1.5924936846058515e-06, + "loss": 1.0445, + "step": 34948 + }, + { + "epoch": 0.82, + "grad_norm": 2.188505687034691, + "learning_rate": 1.5920805814466643e-06, + "loss": 0.9043, + "step": 34949 + }, + { + "epoch": 0.82, + "grad_norm": 2.971321871595864, + "learning_rate": 1.5916675272410442e-06, + "loss": 0.9363, + "step": 34950 + }, + { + "epoch": 0.82, + "grad_norm": 2.664173079746632, + "learning_rate": 1.5912545219913966e-06, + "loss": 0.966, + "step": 34951 + }, + { + "epoch": 0.82, + "grad_norm": 1.8911047176144535, + "learning_rate": 1.5908415657001297e-06, + "loss": 0.9406, + "step": 34952 + }, + { + "epoch": 0.82, + "grad_norm": 1.7946119003198004, + "learning_rate": 1.590428658369646e-06, + "loss": 0.893, + "step": 34953 + }, + { + "epoch": 0.82, + "grad_norm": 2.179690818470049, + "learning_rate": 1.5900158000023469e-06, + "loss": 0.8735, + "step": 34954 + }, + { + "epoch": 0.82, + "grad_norm": 1.9148303343214603, + "learning_rate": 1.5896029906006383e-06, + "loss": 1.1102, + "step": 34955 + }, + { + "epoch": 0.82, + "grad_norm": 2.827894614630074, + "learning_rate": 1.589190230166926e-06, + "loss": 0.8412, + "step": 34956 + }, + { + "epoch": 0.82, + "grad_norm": 1.9426953381722918, + "learning_rate": 1.5887775187036113e-06, + "loss": 0.9367, + "step": 34957 + }, + { + "epoch": 0.82, + "grad_norm": 1.8428746111907348, + "learning_rate": 1.5883648562130937e-06, + "loss": 1.0268, + "step": 34958 + }, + { + "epoch": 0.82, + "grad_norm": 1.9845887686401809, + "learning_rate": 1.587952242697779e-06, + "loss": 0.8761, + "step": 34959 + }, + { + "epoch": 0.82, + "grad_norm": 2.011790963261035, + "learning_rate": 1.5875396781600727e-06, + "loss": 0.8573, + "step": 34960 + }, + { + "epoch": 0.82, + "grad_norm": 2.054283081892603, + "learning_rate": 1.5871271626023733e-06, + "loss": 0.9681, + "step": 34961 + }, + { + "epoch": 0.82, + "grad_norm": 2.220944803284145, + "learning_rate": 1.5867146960270806e-06, + "loss": 0.9004, + "step": 34962 + }, + { + "epoch": 0.82, + "grad_norm": 2.1164711993289176, + "learning_rate": 1.5863022784366012e-06, + "loss": 1.0432, + "step": 34963 + }, + { + "epoch": 0.82, + "grad_norm": 2.8490658848291326, + "learning_rate": 1.5858899098333314e-06, + "loss": 0.9968, + "step": 34964 + }, + { + "epoch": 0.82, + "grad_norm": 1.133554679381658, + "learning_rate": 1.5854775902196762e-06, + "loss": 0.9096, + "step": 34965 + }, + { + "epoch": 0.82, + "grad_norm": 2.0851532956799543, + "learning_rate": 1.5850653195980325e-06, + "loss": 0.864, + "step": 34966 + }, + { + "epoch": 0.82, + "grad_norm": 1.884795448704694, + "learning_rate": 1.5846530979708053e-06, + "loss": 0.8856, + "step": 34967 + }, + { + "epoch": 0.82, + "grad_norm": 1.939957097285558, + "learning_rate": 1.584240925340389e-06, + "loss": 0.8948, + "step": 34968 + }, + { + "epoch": 0.82, + "grad_norm": 1.9296088669007794, + "learning_rate": 1.5838288017091885e-06, + "loss": 1.0786, + "step": 34969 + }, + { + "epoch": 0.82, + "grad_norm": 2.048310474121543, + "learning_rate": 1.583416727079602e-06, + "loss": 0.8862, + "step": 34970 + }, + { + "epoch": 0.82, + "grad_norm": 2.1878586566617764, + "learning_rate": 1.583004701454025e-06, + "loss": 0.9444, + "step": 34971 + }, + { + "epoch": 0.82, + "grad_norm": 1.9958338570629357, + "learning_rate": 1.5825927248348604e-06, + "loss": 0.8995, + "step": 34972 + }, + { + "epoch": 0.82, + "grad_norm": 1.163649662127709, + "learning_rate": 1.5821807972245073e-06, + "loss": 0.9683, + "step": 34973 + }, + { + "epoch": 0.82, + "grad_norm": 1.927033253880341, + "learning_rate": 1.5817689186253627e-06, + "loss": 0.9958, + "step": 34974 + }, + { + "epoch": 0.82, + "grad_norm": 1.0460220717081092, + "learning_rate": 1.5813570890398234e-06, + "loss": 0.9589, + "step": 34975 + }, + { + "epoch": 0.82, + "grad_norm": 2.014121606168875, + "learning_rate": 1.5809453084702875e-06, + "loss": 1.1341, + "step": 34976 + }, + { + "epoch": 0.82, + "grad_norm": 2.0549668857098546, + "learning_rate": 1.5805335769191555e-06, + "loss": 1.002, + "step": 34977 + }, + { + "epoch": 0.82, + "grad_norm": 2.43682948265944, + "learning_rate": 1.580121894388823e-06, + "loss": 1.0941, + "step": 34978 + }, + { + "epoch": 0.82, + "grad_norm": 1.9690496642034796, + "learning_rate": 1.5797102608816838e-06, + "loss": 0.9132, + "step": 34979 + }, + { + "epoch": 0.82, + "grad_norm": 1.7789468899982033, + "learning_rate": 1.5792986764001372e-06, + "loss": 0.9755, + "step": 34980 + }, + { + "epoch": 0.82, + "grad_norm": 1.1022256671023, + "learning_rate": 1.5788871409465822e-06, + "loss": 0.9753, + "step": 34981 + }, + { + "epoch": 0.82, + "grad_norm": 2.79876488022181, + "learning_rate": 1.5784756545234126e-06, + "loss": 0.7713, + "step": 34982 + }, + { + "epoch": 0.82, + "grad_norm": 2.2662996212624735, + "learning_rate": 1.5780642171330217e-06, + "loss": 1.0083, + "step": 34983 + }, + { + "epoch": 0.82, + "grad_norm": 1.8343137634708189, + "learning_rate": 1.5776528287778093e-06, + "loss": 1.106, + "step": 34984 + }, + { + "epoch": 0.82, + "grad_norm": 1.9336453096783677, + "learning_rate": 1.5772414894601663e-06, + "loss": 1.0743, + "step": 34985 + }, + { + "epoch": 0.82, + "grad_norm": 2.0939318438729178, + "learning_rate": 1.576830199182493e-06, + "loss": 0.9453, + "step": 34986 + }, + { + "epoch": 0.82, + "grad_norm": 1.1241303984610123, + "learning_rate": 1.5764189579471778e-06, + "loss": 0.9086, + "step": 34987 + }, + { + "epoch": 0.82, + "grad_norm": 2.1443243158689858, + "learning_rate": 1.5760077657566208e-06, + "loss": 0.8762, + "step": 34988 + }, + { + "epoch": 0.82, + "grad_norm": 1.9995426404981733, + "learning_rate": 1.5755966226132114e-06, + "loss": 0.9717, + "step": 34989 + }, + { + "epoch": 0.82, + "grad_norm": 2.1183977632396127, + "learning_rate": 1.5751855285193474e-06, + "loss": 0.9825, + "step": 34990 + }, + { + "epoch": 0.82, + "grad_norm": 1.905020184087511, + "learning_rate": 1.5747744834774215e-06, + "loss": 1.0872, + "step": 34991 + }, + { + "epoch": 0.82, + "grad_norm": 1.9367810710127782, + "learning_rate": 1.5743634874898228e-06, + "loss": 0.945, + "step": 34992 + }, + { + "epoch": 0.82, + "grad_norm": 2.319470270583288, + "learning_rate": 1.5739525405589474e-06, + "loss": 0.8749, + "step": 34993 + }, + { + "epoch": 0.82, + "grad_norm": 1.8251803174533119, + "learning_rate": 1.5735416426871886e-06, + "loss": 0.8808, + "step": 34994 + }, + { + "epoch": 0.82, + "grad_norm": 2.0372440559577027, + "learning_rate": 1.5731307938769426e-06, + "loss": 0.8847, + "step": 34995 + }, + { + "epoch": 0.82, + "grad_norm": 2.070810720711753, + "learning_rate": 1.5727199941305914e-06, + "loss": 0.9547, + "step": 34996 + }, + { + "epoch": 0.82, + "grad_norm": 2.0475797545580003, + "learning_rate": 1.5723092434505327e-06, + "loss": 0.8609, + "step": 34997 + }, + { + "epoch": 0.82, + "grad_norm": 1.1527838503016448, + "learning_rate": 1.5718985418391609e-06, + "loss": 0.9751, + "step": 34998 + }, + { + "epoch": 0.82, + "grad_norm": 1.8484426003568744, + "learning_rate": 1.5714878892988638e-06, + "loss": 0.9619, + "step": 34999 + }, + { + "epoch": 0.82, + "grad_norm": 2.276369482853001, + "learning_rate": 1.57107728583203e-06, + "loss": 0.8946, + "step": 35000 + }, + { + "epoch": 0.82, + "grad_norm": 1.8396040127884399, + "learning_rate": 1.5706667314410528e-06, + "loss": 0.86, + "step": 35001 + }, + { + "epoch": 0.82, + "grad_norm": 2.0246486615294574, + "learning_rate": 1.570256226128325e-06, + "loss": 0.9609, + "step": 35002 + }, + { + "epoch": 0.82, + "grad_norm": 1.083603978963942, + "learning_rate": 1.5698457698962344e-06, + "loss": 1.0097, + "step": 35003 + }, + { + "epoch": 0.82, + "grad_norm": 1.987860128331257, + "learning_rate": 1.569435362747168e-06, + "loss": 1.0914, + "step": 35004 + }, + { + "epoch": 0.82, + "grad_norm": 1.9280762627719679, + "learning_rate": 1.5690250046835209e-06, + "loss": 1.11, + "step": 35005 + }, + { + "epoch": 0.82, + "grad_norm": 1.868521155536167, + "learning_rate": 1.5686146957076776e-06, + "loss": 0.8801, + "step": 35006 + }, + { + "epoch": 0.82, + "grad_norm": 1.0409058792361672, + "learning_rate": 1.5682044358220272e-06, + "loss": 0.9798, + "step": 35007 + }, + { + "epoch": 0.82, + "grad_norm": 2.1311020400459215, + "learning_rate": 1.5677942250289634e-06, + "loss": 1.0092, + "step": 35008 + }, + { + "epoch": 0.82, + "grad_norm": 2.2796744033552976, + "learning_rate": 1.5673840633308713e-06, + "loss": 0.9714, + "step": 35009 + }, + { + "epoch": 0.82, + "grad_norm": 1.0793909683779612, + "learning_rate": 1.5669739507301362e-06, + "loss": 0.8357, + "step": 35010 + }, + { + "epoch": 0.82, + "grad_norm": 2.0663956267123464, + "learning_rate": 1.5665638872291488e-06, + "loss": 1.0636, + "step": 35011 + }, + { + "epoch": 0.82, + "grad_norm": 1.0329606413010877, + "learning_rate": 1.5661538728303016e-06, + "loss": 0.8772, + "step": 35012 + }, + { + "epoch": 0.82, + "grad_norm": 1.9487417116082186, + "learning_rate": 1.5657439075359716e-06, + "loss": 0.9232, + "step": 35013 + }, + { + "epoch": 0.82, + "grad_norm": 2.1366599658266936, + "learning_rate": 1.5653339913485498e-06, + "loss": 0.9854, + "step": 35014 + }, + { + "epoch": 0.82, + "grad_norm": 2.056565257932598, + "learning_rate": 1.564924124270425e-06, + "loss": 0.9444, + "step": 35015 + }, + { + "epoch": 0.82, + "grad_norm": 1.9232468989329943, + "learning_rate": 1.564514306303987e-06, + "loss": 0.9671, + "step": 35016 + }, + { + "epoch": 0.82, + "grad_norm": 1.9515631817214438, + "learning_rate": 1.5641045374516117e-06, + "loss": 1.0137, + "step": 35017 + }, + { + "epoch": 0.82, + "grad_norm": 2.3073705715686446, + "learning_rate": 1.563694817715692e-06, + "loss": 1.0907, + "step": 35018 + }, + { + "epoch": 0.83, + "grad_norm": 1.1620334469209397, + "learning_rate": 1.5632851470986132e-06, + "loss": 0.8723, + "step": 35019 + }, + { + "epoch": 0.83, + "grad_norm": 3.4095191603518313, + "learning_rate": 1.5628755256027573e-06, + "loss": 0.811, + "step": 35020 + }, + { + "epoch": 0.83, + "grad_norm": 2.131480266164195, + "learning_rate": 1.5624659532305142e-06, + "loss": 1.0462, + "step": 35021 + }, + { + "epoch": 0.83, + "grad_norm": 1.8380026777462786, + "learning_rate": 1.5620564299842623e-06, + "loss": 0.969, + "step": 35022 + }, + { + "epoch": 0.83, + "grad_norm": 1.8118219050045006, + "learning_rate": 1.5616469558663927e-06, + "loss": 0.9723, + "step": 35023 + }, + { + "epoch": 0.83, + "grad_norm": 1.6846364164512844, + "learning_rate": 1.5612375308792826e-06, + "loss": 0.9318, + "step": 35024 + }, + { + "epoch": 0.83, + "grad_norm": 1.7986943820902117, + "learning_rate": 1.5608281550253224e-06, + "loss": 1.0044, + "step": 35025 + }, + { + "epoch": 0.83, + "grad_norm": 1.9785537541886913, + "learning_rate": 1.5604188283068921e-06, + "loss": 0.9006, + "step": 35026 + }, + { + "epoch": 0.83, + "grad_norm": 2.288146300360149, + "learning_rate": 1.5600095507263725e-06, + "loss": 0.8316, + "step": 35027 + }, + { + "epoch": 0.83, + "grad_norm": 1.9904528848632446, + "learning_rate": 1.55960032228615e-06, + "loss": 0.9959, + "step": 35028 + }, + { + "epoch": 0.83, + "grad_norm": 1.9368763942004157, + "learning_rate": 1.5591911429886087e-06, + "loss": 0.9816, + "step": 35029 + }, + { + "epoch": 0.83, + "grad_norm": 1.9385133599264734, + "learning_rate": 1.5587820128361297e-06, + "loss": 0.9791, + "step": 35030 + }, + { + "epoch": 0.83, + "grad_norm": 2.3788685680170483, + "learning_rate": 1.5583729318310902e-06, + "loss": 0.9991, + "step": 35031 + }, + { + "epoch": 0.83, + "grad_norm": 2.2309360298609295, + "learning_rate": 1.5579638999758772e-06, + "loss": 1.0661, + "step": 35032 + }, + { + "epoch": 0.83, + "grad_norm": 1.916502408069584, + "learning_rate": 1.5575549172728733e-06, + "loss": 1.0267, + "step": 35033 + }, + { + "epoch": 0.83, + "grad_norm": 2.1267119116309203, + "learning_rate": 1.5571459837244585e-06, + "loss": 0.9876, + "step": 35034 + }, + { + "epoch": 0.83, + "grad_norm": 2.0611699310354004, + "learning_rate": 1.556737099333009e-06, + "loss": 0.986, + "step": 35035 + }, + { + "epoch": 0.83, + "grad_norm": 2.0953062086914076, + "learning_rate": 1.5563282641009092e-06, + "loss": 0.8249, + "step": 35036 + }, + { + "epoch": 0.83, + "grad_norm": 2.0362044067397185, + "learning_rate": 1.555919478030542e-06, + "loss": 0.9489, + "step": 35037 + }, + { + "epoch": 0.83, + "grad_norm": 1.7916165549346388, + "learning_rate": 1.5555107411242854e-06, + "loss": 1.0357, + "step": 35038 + }, + { + "epoch": 0.83, + "grad_norm": 1.165839260188278, + "learning_rate": 1.5551020533845162e-06, + "loss": 0.9034, + "step": 35039 + }, + { + "epoch": 0.83, + "grad_norm": 2.223277643903416, + "learning_rate": 1.5546934148136196e-06, + "loss": 0.9432, + "step": 35040 + }, + { + "epoch": 0.83, + "grad_norm": 1.919278302797114, + "learning_rate": 1.5542848254139674e-06, + "loss": 1.0669, + "step": 35041 + }, + { + "epoch": 0.83, + "grad_norm": 2.0096839764472496, + "learning_rate": 1.5538762851879464e-06, + "loss": 0.9535, + "step": 35042 + }, + { + "epoch": 0.83, + "grad_norm": 1.8433869029476602, + "learning_rate": 1.5534677941379295e-06, + "loss": 1.0432, + "step": 35043 + }, + { + "epoch": 0.83, + "grad_norm": 1.9759582539586216, + "learning_rate": 1.5530593522662995e-06, + "loss": 0.976, + "step": 35044 + }, + { + "epoch": 0.83, + "grad_norm": 1.9898464902461857, + "learning_rate": 1.5526509595754291e-06, + "loss": 0.9731, + "step": 35045 + }, + { + "epoch": 0.83, + "grad_norm": 2.058772555942733, + "learning_rate": 1.5522426160677029e-06, + "loss": 1.1083, + "step": 35046 + }, + { + "epoch": 0.83, + "grad_norm": 2.611697173378926, + "learning_rate": 1.5518343217454934e-06, + "loss": 0.941, + "step": 35047 + }, + { + "epoch": 0.83, + "grad_norm": 1.9440831455218819, + "learning_rate": 1.5514260766111777e-06, + "loss": 1.013, + "step": 35048 + }, + { + "epoch": 0.83, + "grad_norm": 2.4991175041107363, + "learning_rate": 1.5510178806671328e-06, + "loss": 0.9922, + "step": 35049 + }, + { + "epoch": 0.83, + "grad_norm": 2.122995599910107, + "learning_rate": 1.5506097339157399e-06, + "loss": 0.9428, + "step": 35050 + }, + { + "epoch": 0.83, + "grad_norm": 2.0692622883640106, + "learning_rate": 1.550201636359373e-06, + "loss": 0.9484, + "step": 35051 + }, + { + "epoch": 0.83, + "grad_norm": 2.075139897448085, + "learning_rate": 1.5497935880004034e-06, + "loss": 1.0952, + "step": 35052 + }, + { + "epoch": 0.83, + "grad_norm": 2.4741883770331423, + "learning_rate": 1.5493855888412123e-06, + "loss": 0.9458, + "step": 35053 + }, + { + "epoch": 0.83, + "grad_norm": 1.7832603917400485, + "learning_rate": 1.5489776388841748e-06, + "loss": 0.9753, + "step": 35054 + }, + { + "epoch": 0.83, + "grad_norm": 2.000738878280435, + "learning_rate": 1.5485697381316656e-06, + "loss": 1.0082, + "step": 35055 + }, + { + "epoch": 0.83, + "grad_norm": 2.1467320500629663, + "learning_rate": 1.548161886586057e-06, + "loss": 0.9162, + "step": 35056 + }, + { + "epoch": 0.83, + "grad_norm": 2.031663372773887, + "learning_rate": 1.5477540842497251e-06, + "loss": 0.9753, + "step": 35057 + }, + { + "epoch": 0.83, + "grad_norm": 1.8900236645588022, + "learning_rate": 1.5473463311250481e-06, + "loss": 0.969, + "step": 35058 + }, + { + "epoch": 0.83, + "grad_norm": 2.2254098182385054, + "learning_rate": 1.5469386272143961e-06, + "loss": 0.8964, + "step": 35059 + }, + { + "epoch": 0.83, + "grad_norm": 1.7478040642432888, + "learning_rate": 1.5465309725201416e-06, + "loss": 0.9315, + "step": 35060 + }, + { + "epoch": 0.83, + "grad_norm": 2.036389516410495, + "learning_rate": 1.5461233670446618e-06, + "loss": 0.9588, + "step": 35061 + }, + { + "epoch": 0.83, + "grad_norm": 2.2154027585663276, + "learning_rate": 1.5457158107903269e-06, + "loss": 0.8996, + "step": 35062 + }, + { + "epoch": 0.83, + "grad_norm": 2.004560289951637, + "learning_rate": 1.5453083037595128e-06, + "loss": 1.0439, + "step": 35063 + }, + { + "epoch": 0.83, + "grad_norm": 1.8908218017080105, + "learning_rate": 1.5449008459545877e-06, + "loss": 0.9917, + "step": 35064 + }, + { + "epoch": 0.83, + "grad_norm": 1.0760496207684442, + "learning_rate": 1.5444934373779296e-06, + "loss": 0.8993, + "step": 35065 + }, + { + "epoch": 0.83, + "grad_norm": 2.025614670204032, + "learning_rate": 1.5440860780319055e-06, + "loss": 0.99, + "step": 35066 + }, + { + "epoch": 0.83, + "grad_norm": 2.0542365148525783, + "learning_rate": 1.5436787679188913e-06, + "loss": 1.0991, + "step": 35067 + }, + { + "epoch": 0.83, + "grad_norm": 2.0020317877457483, + "learning_rate": 1.5432715070412552e-06, + "loss": 0.9536, + "step": 35068 + }, + { + "epoch": 0.83, + "grad_norm": 2.517230878517752, + "learning_rate": 1.5428642954013706e-06, + "loss": 0.8201, + "step": 35069 + }, + { + "epoch": 0.83, + "grad_norm": 1.1324880544369176, + "learning_rate": 1.5424571330016059e-06, + "loss": 0.9974, + "step": 35070 + }, + { + "epoch": 0.83, + "grad_norm": 1.8011400931662342, + "learning_rate": 1.5420500198443367e-06, + "loss": 0.9884, + "step": 35071 + }, + { + "epoch": 0.83, + "grad_norm": 2.021896835762723, + "learning_rate": 1.541642955931929e-06, + "loss": 0.9438, + "step": 35072 + }, + { + "epoch": 0.83, + "grad_norm": 1.9344643174335994, + "learning_rate": 1.541235941266751e-06, + "loss": 1.0366, + "step": 35073 + }, + { + "epoch": 0.83, + "grad_norm": 1.772464886180168, + "learning_rate": 1.5408289758511763e-06, + "loss": 0.851, + "step": 35074 + }, + { + "epoch": 0.83, + "grad_norm": 1.0563946310229406, + "learning_rate": 1.5404220596875752e-06, + "loss": 0.9531, + "step": 35075 + }, + { + "epoch": 0.83, + "grad_norm": 1.8588689966561929, + "learning_rate": 1.5400151927783158e-06, + "loss": 0.9335, + "step": 35076 + }, + { + "epoch": 0.83, + "grad_norm": 1.82144760127029, + "learning_rate": 1.5396083751257628e-06, + "loss": 1.0477, + "step": 35077 + }, + { + "epoch": 0.83, + "grad_norm": 3.1056194997206727, + "learning_rate": 1.5392016067322891e-06, + "loss": 0.9381, + "step": 35078 + }, + { + "epoch": 0.83, + "grad_norm": 1.9889592086819676, + "learning_rate": 1.5387948876002645e-06, + "loss": 0.9027, + "step": 35079 + }, + { + "epoch": 0.83, + "grad_norm": 2.0527290972316052, + "learning_rate": 1.5383882177320553e-06, + "loss": 0.8869, + "step": 35080 + }, + { + "epoch": 0.83, + "grad_norm": 2.387059059870632, + "learning_rate": 1.537981597130026e-06, + "loss": 0.9849, + "step": 35081 + }, + { + "epoch": 0.83, + "grad_norm": 2.476392221820697, + "learning_rate": 1.537575025796547e-06, + "loss": 0.9072, + "step": 35082 + }, + { + "epoch": 0.83, + "grad_norm": 2.245161255304027, + "learning_rate": 1.5371685037339878e-06, + "loss": 0.9564, + "step": 35083 + }, + { + "epoch": 0.83, + "grad_norm": 1.113229455495928, + "learning_rate": 1.536762030944713e-06, + "loss": 1.0097, + "step": 35084 + }, + { + "epoch": 0.83, + "grad_norm": 1.8526987278221614, + "learning_rate": 1.536355607431086e-06, + "loss": 1.0103, + "step": 35085 + }, + { + "epoch": 0.83, + "grad_norm": 2.248271578873438, + "learning_rate": 1.5359492331954795e-06, + "loss": 0.9863, + "step": 35086 + }, + { + "epoch": 0.83, + "grad_norm": 2.508159653530943, + "learning_rate": 1.5355429082402528e-06, + "loss": 0.9798, + "step": 35087 + }, + { + "epoch": 0.83, + "grad_norm": 2.019581324082898, + "learning_rate": 1.5351366325677785e-06, + "loss": 0.9259, + "step": 35088 + }, + { + "epoch": 0.83, + "grad_norm": 1.9041933233276567, + "learning_rate": 1.5347304061804158e-06, + "loss": 0.8408, + "step": 35089 + }, + { + "epoch": 0.83, + "grad_norm": 2.4281489534428533, + "learning_rate": 1.5343242290805348e-06, + "loss": 0.9922, + "step": 35090 + }, + { + "epoch": 0.83, + "grad_norm": 1.0187979318124396, + "learning_rate": 1.5339181012704974e-06, + "loss": 0.8488, + "step": 35091 + }, + { + "epoch": 0.83, + "grad_norm": 2.0040859765514725, + "learning_rate": 1.5335120227526678e-06, + "loss": 0.9204, + "step": 35092 + }, + { + "epoch": 0.83, + "grad_norm": 1.9621824949710907, + "learning_rate": 1.5331059935294168e-06, + "loss": 0.9167, + "step": 35093 + }, + { + "epoch": 0.83, + "grad_norm": 1.8890011617721698, + "learning_rate": 1.532700013603098e-06, + "loss": 1.0243, + "step": 35094 + }, + { + "epoch": 0.83, + "grad_norm": 1.9556194843513601, + "learning_rate": 1.5322940829760802e-06, + "loss": 1.1278, + "step": 35095 + }, + { + "epoch": 0.83, + "grad_norm": 2.270824177553847, + "learning_rate": 1.5318882016507286e-06, + "loss": 1.0594, + "step": 35096 + }, + { + "epoch": 0.83, + "grad_norm": 2.061569389768976, + "learning_rate": 1.531482369629408e-06, + "loss": 0.9215, + "step": 35097 + }, + { + "epoch": 0.83, + "grad_norm": 2.3260602904268346, + "learning_rate": 1.531076586914474e-06, + "loss": 1.0251, + "step": 35098 + }, + { + "epoch": 0.83, + "grad_norm": 1.8431953706231803, + "learning_rate": 1.5306708535082937e-06, + "loss": 0.8988, + "step": 35099 + }, + { + "epoch": 0.83, + "grad_norm": 1.04919635701498, + "learning_rate": 1.530265169413231e-06, + "loss": 0.8853, + "step": 35100 + }, + { + "epoch": 0.83, + "grad_norm": 2.0185532775771557, + "learning_rate": 1.5298595346316457e-06, + "loss": 0.957, + "step": 35101 + }, + { + "epoch": 0.83, + "grad_norm": 2.122538276462798, + "learning_rate": 1.5294539491658978e-06, + "loss": 0.9203, + "step": 35102 + }, + { + "epoch": 0.83, + "grad_norm": 1.8625380534116542, + "learning_rate": 1.5290484130183515e-06, + "loss": 0.8948, + "step": 35103 + }, + { + "epoch": 0.83, + "grad_norm": 1.7282326772487344, + "learning_rate": 1.5286429261913682e-06, + "loss": 0.9243, + "step": 35104 + }, + { + "epoch": 0.83, + "grad_norm": 1.9224628939120154, + "learning_rate": 1.5282374886873063e-06, + "loss": 0.8914, + "step": 35105 + }, + { + "epoch": 0.83, + "grad_norm": 2.2500655131914393, + "learning_rate": 1.5278321005085296e-06, + "loss": 0.9477, + "step": 35106 + }, + { + "epoch": 0.83, + "grad_norm": 2.138396932073266, + "learning_rate": 1.5274267616573968e-06, + "loss": 1.0709, + "step": 35107 + }, + { + "epoch": 0.83, + "grad_norm": 3.9030781886269206, + "learning_rate": 1.5270214721362653e-06, + "loss": 0.9693, + "step": 35108 + }, + { + "epoch": 0.83, + "grad_norm": 2.4746519600117756, + "learning_rate": 1.5266162319474974e-06, + "loss": 0.9455, + "step": 35109 + }, + { + "epoch": 0.83, + "grad_norm": 1.9125478120570465, + "learning_rate": 1.5262110410934561e-06, + "loss": 0.9155, + "step": 35110 + }, + { + "epoch": 0.83, + "grad_norm": 2.1214665531824495, + "learning_rate": 1.5258058995764946e-06, + "loss": 0.9428, + "step": 35111 + }, + { + "epoch": 0.83, + "grad_norm": 1.9176374123003253, + "learning_rate": 1.5254008073989734e-06, + "loss": 1.0778, + "step": 35112 + }, + { + "epoch": 0.83, + "grad_norm": 2.646662101813156, + "learning_rate": 1.5249957645632508e-06, + "loss": 0.9643, + "step": 35113 + }, + { + "epoch": 0.83, + "grad_norm": 2.1511739222167168, + "learning_rate": 1.5245907710716912e-06, + "loss": 0.9269, + "step": 35114 + }, + { + "epoch": 0.83, + "grad_norm": 2.063773338551909, + "learning_rate": 1.524185826926643e-06, + "loss": 0.9785, + "step": 35115 + }, + { + "epoch": 0.83, + "grad_norm": 2.673261699318235, + "learning_rate": 1.5237809321304674e-06, + "loss": 1.0679, + "step": 35116 + }, + { + "epoch": 0.83, + "grad_norm": 1.8872644208445581, + "learning_rate": 1.523376086685523e-06, + "loss": 1.031, + "step": 35117 + }, + { + "epoch": 0.83, + "grad_norm": 1.8624234607355061, + "learning_rate": 1.5229712905941696e-06, + "loss": 0.8083, + "step": 35118 + }, + { + "epoch": 0.83, + "grad_norm": 1.8821678199311986, + "learning_rate": 1.52256654385876e-06, + "loss": 1.0177, + "step": 35119 + }, + { + "epoch": 0.83, + "grad_norm": 2.072328936974134, + "learning_rate": 1.5221618464816501e-06, + "loss": 0.942, + "step": 35120 + }, + { + "epoch": 0.83, + "grad_norm": 2.424725652325371, + "learning_rate": 1.5217571984652012e-06, + "loss": 0.8653, + "step": 35121 + }, + { + "epoch": 0.83, + "grad_norm": 2.2803769276024717, + "learning_rate": 1.5213525998117628e-06, + "loss": 1.0923, + "step": 35122 + }, + { + "epoch": 0.83, + "grad_norm": 1.9905228003379998, + "learning_rate": 1.5209480505236963e-06, + "loss": 0.9995, + "step": 35123 + }, + { + "epoch": 0.83, + "grad_norm": 2.4660419441935377, + "learning_rate": 1.5205435506033529e-06, + "loss": 0.9241, + "step": 35124 + }, + { + "epoch": 0.83, + "grad_norm": 1.8028817660304626, + "learning_rate": 1.5201391000530908e-06, + "loss": 0.9565, + "step": 35125 + }, + { + "epoch": 0.83, + "grad_norm": 1.918802381305398, + "learning_rate": 1.5197346988752615e-06, + "loss": 0.8818, + "step": 35126 + }, + { + "epoch": 0.83, + "grad_norm": 3.9090881450101804, + "learning_rate": 1.5193303470722243e-06, + "loss": 1.0769, + "step": 35127 + }, + { + "epoch": 0.83, + "grad_norm": 2.145973351155181, + "learning_rate": 1.5189260446463306e-06, + "loss": 0.9701, + "step": 35128 + }, + { + "epoch": 0.83, + "grad_norm": 2.0072061011100257, + "learning_rate": 1.5185217915999328e-06, + "loss": 0.946, + "step": 35129 + }, + { + "epoch": 0.83, + "grad_norm": 1.8994256160208185, + "learning_rate": 1.518117587935385e-06, + "loss": 0.9433, + "step": 35130 + }, + { + "epoch": 0.83, + "grad_norm": 1.9192453117137023, + "learning_rate": 1.517713433655046e-06, + "loss": 0.9023, + "step": 35131 + }, + { + "epoch": 0.83, + "grad_norm": 1.7880507928758091, + "learning_rate": 1.5173093287612628e-06, + "loss": 0.9664, + "step": 35132 + }, + { + "epoch": 0.83, + "grad_norm": 2.1938759990638323, + "learning_rate": 1.5169052732563895e-06, + "loss": 0.9043, + "step": 35133 + }, + { + "epoch": 0.83, + "grad_norm": 5.028828861487601, + "learning_rate": 1.5165012671427793e-06, + "loss": 1.0804, + "step": 35134 + }, + { + "epoch": 0.83, + "grad_norm": 1.9002259657017806, + "learning_rate": 1.516097310422786e-06, + "loss": 0.9851, + "step": 35135 + }, + { + "epoch": 0.83, + "grad_norm": 1.9076421489246358, + "learning_rate": 1.5156934030987603e-06, + "loss": 0.9344, + "step": 35136 + }, + { + "epoch": 0.83, + "grad_norm": 2.1231938888513704, + "learning_rate": 1.515289545173052e-06, + "loss": 0.925, + "step": 35137 + }, + { + "epoch": 0.83, + "grad_norm": 1.8886904053582783, + "learning_rate": 1.5148857366480141e-06, + "loss": 1.025, + "step": 35138 + }, + { + "epoch": 0.83, + "grad_norm": 1.8438604475082476, + "learning_rate": 1.514481977525999e-06, + "loss": 1.0168, + "step": 35139 + }, + { + "epoch": 0.83, + "grad_norm": 2.0170907283975863, + "learning_rate": 1.514078267809358e-06, + "loss": 0.9761, + "step": 35140 + }, + { + "epoch": 0.83, + "grad_norm": 1.9927070168185526, + "learning_rate": 1.513674607500436e-06, + "loss": 0.9168, + "step": 35141 + }, + { + "epoch": 0.83, + "grad_norm": 1.0483537664363989, + "learning_rate": 1.5132709966015901e-06, + "loss": 0.9035, + "step": 35142 + }, + { + "epoch": 0.83, + "grad_norm": 1.7799116352650142, + "learning_rate": 1.512867435115165e-06, + "loss": 0.9312, + "step": 35143 + }, + { + "epoch": 0.83, + "grad_norm": 2.034633821739176, + "learning_rate": 1.5124639230435157e-06, + "loss": 0.9206, + "step": 35144 + }, + { + "epoch": 0.83, + "grad_norm": 1.9362258465457238, + "learning_rate": 1.5120604603889855e-06, + "loss": 1.0114, + "step": 35145 + }, + { + "epoch": 0.83, + "grad_norm": 2.0090298573814587, + "learning_rate": 1.5116570471539294e-06, + "loss": 1.0379, + "step": 35146 + }, + { + "epoch": 0.83, + "grad_norm": 1.091643225805927, + "learning_rate": 1.5112536833406898e-06, + "loss": 0.9489, + "step": 35147 + }, + { + "epoch": 0.83, + "grad_norm": 1.8776598782161873, + "learning_rate": 1.5108503689516218e-06, + "loss": 0.9772, + "step": 35148 + }, + { + "epoch": 0.83, + "grad_norm": 2.1542701393138883, + "learning_rate": 1.5104471039890711e-06, + "loss": 0.9054, + "step": 35149 + }, + { + "epoch": 0.83, + "grad_norm": 1.8570235190734135, + "learning_rate": 1.5100438884553826e-06, + "loss": 0.9324, + "step": 35150 + }, + { + "epoch": 0.83, + "grad_norm": 2.0736817922778266, + "learning_rate": 1.5096407223529053e-06, + "loss": 0.9643, + "step": 35151 + }, + { + "epoch": 0.83, + "grad_norm": 1.9001857077628042, + "learning_rate": 1.5092376056839908e-06, + "loss": 1.0197, + "step": 35152 + }, + { + "epoch": 0.83, + "grad_norm": 2.1501145997211264, + "learning_rate": 1.5088345384509838e-06, + "loss": 0.9261, + "step": 35153 + }, + { + "epoch": 0.83, + "grad_norm": 1.8407673135635974, + "learning_rate": 1.5084315206562272e-06, + "loss": 0.8173, + "step": 35154 + }, + { + "epoch": 0.83, + "grad_norm": 1.855986947428462, + "learning_rate": 1.508028552302071e-06, + "loss": 1.0094, + "step": 35155 + }, + { + "epoch": 0.83, + "grad_norm": 1.7648310340565307, + "learning_rate": 1.5076256333908635e-06, + "loss": 0.8516, + "step": 35156 + }, + { + "epoch": 0.83, + "grad_norm": 1.8650076009294096, + "learning_rate": 1.5072227639249482e-06, + "loss": 1.0453, + "step": 35157 + }, + { + "epoch": 0.83, + "grad_norm": 2.020580856801481, + "learning_rate": 1.5068199439066678e-06, + "loss": 0.8986, + "step": 35158 + }, + { + "epoch": 0.83, + "grad_norm": 2.2490984312640565, + "learning_rate": 1.5064171733383715e-06, + "loss": 0.9144, + "step": 35159 + }, + { + "epoch": 0.83, + "grad_norm": 2.115148721391243, + "learning_rate": 1.5060144522224052e-06, + "loss": 1.0679, + "step": 35160 + }, + { + "epoch": 0.83, + "grad_norm": 1.1660641617393424, + "learning_rate": 1.5056117805611115e-06, + "loss": 0.9262, + "step": 35161 + }, + { + "epoch": 0.83, + "grad_norm": 1.9047810112703465, + "learning_rate": 1.505209158356834e-06, + "loss": 0.8436, + "step": 35162 + }, + { + "epoch": 0.83, + "grad_norm": 1.886521972465519, + "learning_rate": 1.50480658561192e-06, + "loss": 0.8559, + "step": 35163 + }, + { + "epoch": 0.83, + "grad_norm": 1.8811763801300554, + "learning_rate": 1.5044040623287092e-06, + "loss": 0.9756, + "step": 35164 + }, + { + "epoch": 0.83, + "grad_norm": 2.0606688196699583, + "learning_rate": 1.5040015885095494e-06, + "loss": 0.8391, + "step": 35165 + }, + { + "epoch": 0.83, + "grad_norm": 2.0995067624589376, + "learning_rate": 1.5035991641567803e-06, + "loss": 0.9881, + "step": 35166 + }, + { + "epoch": 0.83, + "grad_norm": 2.103870373001057, + "learning_rate": 1.5031967892727494e-06, + "loss": 0.8237, + "step": 35167 + }, + { + "epoch": 0.83, + "grad_norm": 1.8743855219515784, + "learning_rate": 1.5027944638597947e-06, + "loss": 0.914, + "step": 35168 + }, + { + "epoch": 0.83, + "grad_norm": 1.9314237459351757, + "learning_rate": 1.502392187920263e-06, + "loss": 1.1179, + "step": 35169 + }, + { + "epoch": 0.83, + "grad_norm": 1.933793256081042, + "learning_rate": 1.5019899614564937e-06, + "loss": 0.9961, + "step": 35170 + }, + { + "epoch": 0.83, + "grad_norm": 2.2139071244077484, + "learning_rate": 1.5015877844708272e-06, + "loss": 0.9387, + "step": 35171 + }, + { + "epoch": 0.83, + "grad_norm": 1.9987493519527746, + "learning_rate": 1.501185656965607e-06, + "loss": 1.0764, + "step": 35172 + }, + { + "epoch": 0.83, + "grad_norm": 1.0669267765754704, + "learning_rate": 1.500783578943177e-06, + "loss": 0.9024, + "step": 35173 + }, + { + "epoch": 0.83, + "grad_norm": 2.2521883580308213, + "learning_rate": 1.500381550405876e-06, + "loss": 0.9359, + "step": 35174 + }, + { + "epoch": 0.83, + "grad_norm": 2.406909726688877, + "learning_rate": 1.4999795713560428e-06, + "loss": 1.1377, + "step": 35175 + }, + { + "epoch": 0.83, + "grad_norm": 2.1617731928819777, + "learning_rate": 1.4995776417960184e-06, + "loss": 0.8301, + "step": 35176 + }, + { + "epoch": 0.83, + "grad_norm": 2.655303644146674, + "learning_rate": 1.4991757617281466e-06, + "loss": 1.0492, + "step": 35177 + }, + { + "epoch": 0.83, + "grad_norm": 2.192722219732211, + "learning_rate": 1.4987739311547656e-06, + "loss": 1.0661, + "step": 35178 + }, + { + "epoch": 0.83, + "grad_norm": 2.1306579546657716, + "learning_rate": 1.4983721500782122e-06, + "loss": 0.9389, + "step": 35179 + }, + { + "epoch": 0.83, + "grad_norm": 1.8123518245037642, + "learning_rate": 1.497970418500827e-06, + "loss": 0.9424, + "step": 35180 + }, + { + "epoch": 0.83, + "grad_norm": 1.850376148461182, + "learning_rate": 1.4975687364249513e-06, + "loss": 1.048, + "step": 35181 + }, + { + "epoch": 0.83, + "grad_norm": 2.5069988324679344, + "learning_rate": 1.4971671038529235e-06, + "loss": 1.0325, + "step": 35182 + }, + { + "epoch": 0.83, + "grad_norm": 2.048096529532682, + "learning_rate": 1.496765520787078e-06, + "loss": 0.9064, + "step": 35183 + }, + { + "epoch": 0.83, + "grad_norm": 1.94310138945255, + "learning_rate": 1.4963639872297586e-06, + "loss": 1.0044, + "step": 35184 + }, + { + "epoch": 0.83, + "grad_norm": 2.0734429475609253, + "learning_rate": 1.4959625031832969e-06, + "loss": 1.0247, + "step": 35185 + }, + { + "epoch": 0.83, + "grad_norm": 1.8540189084130998, + "learning_rate": 1.4955610686500365e-06, + "loss": 0.8927, + "step": 35186 + }, + { + "epoch": 0.83, + "grad_norm": 2.22407536846982, + "learning_rate": 1.4951596836323102e-06, + "loss": 0.8872, + "step": 35187 + }, + { + "epoch": 0.83, + "grad_norm": 2.125786066227276, + "learning_rate": 1.494758348132459e-06, + "loss": 0.9904, + "step": 35188 + }, + { + "epoch": 0.83, + "grad_norm": 1.8413340686400477, + "learning_rate": 1.4943570621528148e-06, + "loss": 1.0575, + "step": 35189 + }, + { + "epoch": 0.83, + "grad_norm": 2.068547727682806, + "learning_rate": 1.4939558256957166e-06, + "loss": 0.8873, + "step": 35190 + }, + { + "epoch": 0.83, + "grad_norm": 1.9628175444307356, + "learning_rate": 1.4935546387635047e-06, + "loss": 1.032, + "step": 35191 + }, + { + "epoch": 0.83, + "grad_norm": 2.786399609252943, + "learning_rate": 1.4931535013585075e-06, + "loss": 0.9023, + "step": 35192 + }, + { + "epoch": 0.83, + "grad_norm": 2.122726250400911, + "learning_rate": 1.4927524134830629e-06, + "loss": 0.946, + "step": 35193 + }, + { + "epoch": 0.83, + "grad_norm": 2.131681479259498, + "learning_rate": 1.492351375139507e-06, + "loss": 0.9024, + "step": 35194 + }, + { + "epoch": 0.83, + "grad_norm": 2.0641190870799444, + "learning_rate": 1.4919503863301799e-06, + "loss": 1.0102, + "step": 35195 + }, + { + "epoch": 0.83, + "grad_norm": 1.8797048584819493, + "learning_rate": 1.4915494470574065e-06, + "loss": 0.955, + "step": 35196 + }, + { + "epoch": 0.83, + "grad_norm": 2.1249865602057763, + "learning_rate": 1.4911485573235262e-06, + "loss": 1.0645, + "step": 35197 + }, + { + "epoch": 0.83, + "grad_norm": 2.059562458350742, + "learning_rate": 1.4907477171308727e-06, + "loss": 1.0794, + "step": 35198 + }, + { + "epoch": 0.83, + "grad_norm": 2.0462319772505357, + "learning_rate": 1.4903469264817838e-06, + "loss": 1.1098, + "step": 35199 + }, + { + "epoch": 0.83, + "grad_norm": 1.9449373368454836, + "learning_rate": 1.4899461853785857e-06, + "loss": 0.9272, + "step": 35200 + }, + { + "epoch": 0.83, + "grad_norm": 2.0094032657188476, + "learning_rate": 1.4895454938236142e-06, + "loss": 1.0352, + "step": 35201 + }, + { + "epoch": 0.83, + "grad_norm": 2.413135624732555, + "learning_rate": 1.4891448518192065e-06, + "loss": 1.0168, + "step": 35202 + }, + { + "epoch": 0.83, + "grad_norm": 1.8864454869702778, + "learning_rate": 1.4887442593676893e-06, + "loss": 0.8731, + "step": 35203 + }, + { + "epoch": 0.83, + "grad_norm": 1.8630218789551998, + "learning_rate": 1.4883437164713999e-06, + "loss": 0.9348, + "step": 35204 + }, + { + "epoch": 0.83, + "grad_norm": 2.1279832998212216, + "learning_rate": 1.4879432231326663e-06, + "loss": 0.9842, + "step": 35205 + }, + { + "epoch": 0.83, + "grad_norm": 1.8509013777760355, + "learning_rate": 1.4875427793538233e-06, + "loss": 0.9837, + "step": 35206 + }, + { + "epoch": 0.83, + "grad_norm": 2.3859666671106816, + "learning_rate": 1.487142385137199e-06, + "loss": 0.9469, + "step": 35207 + }, + { + "epoch": 0.83, + "grad_norm": 1.819799442688103, + "learning_rate": 1.4867420404851306e-06, + "loss": 1.0356, + "step": 35208 + }, + { + "epoch": 0.83, + "grad_norm": 1.92029319512976, + "learning_rate": 1.486341745399944e-06, + "loss": 1.0093, + "step": 35209 + }, + { + "epoch": 0.83, + "grad_norm": 2.061774910732531, + "learning_rate": 1.4859414998839694e-06, + "loss": 1.0244, + "step": 35210 + }, + { + "epoch": 0.83, + "grad_norm": 2.0807577291743033, + "learning_rate": 1.4855413039395383e-06, + "loss": 0.9058, + "step": 35211 + }, + { + "epoch": 0.83, + "grad_norm": 2.160040415722247, + "learning_rate": 1.4851411575689845e-06, + "loss": 0.9227, + "step": 35212 + }, + { + "epoch": 0.83, + "grad_norm": 2.1697957085250197, + "learning_rate": 1.4847410607746337e-06, + "loss": 0.9531, + "step": 35213 + }, + { + "epoch": 0.83, + "grad_norm": 3.2322997500868627, + "learning_rate": 1.4843410135588155e-06, + "loss": 0.8204, + "step": 35214 + }, + { + "epoch": 0.83, + "grad_norm": 1.9015292361028286, + "learning_rate": 1.4839410159238587e-06, + "loss": 0.9121, + "step": 35215 + }, + { + "epoch": 0.83, + "grad_norm": 2.074553932593127, + "learning_rate": 1.4835410678720963e-06, + "loss": 1.0262, + "step": 35216 + }, + { + "epoch": 0.83, + "grad_norm": 1.9948257108153253, + "learning_rate": 1.483141169405854e-06, + "loss": 0.9301, + "step": 35217 + }, + { + "epoch": 0.83, + "grad_norm": 1.8135353921326012, + "learning_rate": 1.4827413205274565e-06, + "loss": 1.066, + "step": 35218 + }, + { + "epoch": 0.83, + "grad_norm": 1.0497366253280351, + "learning_rate": 1.4823415212392378e-06, + "loss": 0.9161, + "step": 35219 + }, + { + "epoch": 0.83, + "grad_norm": 2.6377492692238373, + "learning_rate": 1.4819417715435247e-06, + "loss": 0.9518, + "step": 35220 + }, + { + "epoch": 0.83, + "grad_norm": 1.8607833309946333, + "learning_rate": 1.481542071442643e-06, + "loss": 0.9072, + "step": 35221 + }, + { + "epoch": 0.83, + "grad_norm": 1.966368858317596, + "learning_rate": 1.481142420938918e-06, + "loss": 1.0534, + "step": 35222 + }, + { + "epoch": 0.83, + "grad_norm": 2.144802199941397, + "learning_rate": 1.4807428200346817e-06, + "loss": 0.9912, + "step": 35223 + }, + { + "epoch": 0.83, + "grad_norm": 1.9550459117884136, + "learning_rate": 1.480343268732255e-06, + "loss": 1.0634, + "step": 35224 + }, + { + "epoch": 0.83, + "grad_norm": 1.9207625005249993, + "learning_rate": 1.4799437670339701e-06, + "loss": 0.8335, + "step": 35225 + }, + { + "epoch": 0.83, + "grad_norm": 2.0612903898180788, + "learning_rate": 1.4795443149421473e-06, + "loss": 1.0312, + "step": 35226 + }, + { + "epoch": 0.83, + "grad_norm": 2.0080711487127134, + "learning_rate": 1.4791449124591173e-06, + "loss": 0.9029, + "step": 35227 + }, + { + "epoch": 0.83, + "grad_norm": 2.1028597573199366, + "learning_rate": 1.4787455595872002e-06, + "loss": 1.029, + "step": 35228 + }, + { + "epoch": 0.83, + "grad_norm": 2.0598594475914256, + "learning_rate": 1.4783462563287277e-06, + "loss": 0.743, + "step": 35229 + }, + { + "epoch": 0.83, + "grad_norm": 1.9628235960898264, + "learning_rate": 1.477947002686021e-06, + "loss": 1.0161, + "step": 35230 + }, + { + "epoch": 0.83, + "grad_norm": 2.047166150257912, + "learning_rate": 1.477547798661403e-06, + "loss": 0.9876, + "step": 35231 + }, + { + "epoch": 0.83, + "grad_norm": 1.9618515185777672, + "learning_rate": 1.477148644257198e-06, + "loss": 0.9674, + "step": 35232 + }, + { + "epoch": 0.83, + "grad_norm": 2.329439289470796, + "learning_rate": 1.4767495394757359e-06, + "loss": 0.9487, + "step": 35233 + }, + { + "epoch": 0.83, + "grad_norm": 1.9187531682633474, + "learning_rate": 1.4763504843193366e-06, + "loss": 0.9309, + "step": 35234 + }, + { + "epoch": 0.83, + "grad_norm": 1.840656942484445, + "learning_rate": 1.4759514787903195e-06, + "loss": 0.8145, + "step": 35235 + }, + { + "epoch": 0.83, + "grad_norm": 2.1773619912888695, + "learning_rate": 1.4755525228910128e-06, + "loss": 0.8843, + "step": 35236 + }, + { + "epoch": 0.83, + "grad_norm": 1.7996056431631648, + "learning_rate": 1.4751536166237402e-06, + "loss": 0.9563, + "step": 35237 + }, + { + "epoch": 0.83, + "grad_norm": 2.161463666805424, + "learning_rate": 1.4747547599908218e-06, + "loss": 0.9906, + "step": 35238 + }, + { + "epoch": 0.83, + "grad_norm": 1.9665069894425018, + "learning_rate": 1.4743559529945784e-06, + "loss": 0.9559, + "step": 35239 + }, + { + "epoch": 0.83, + "grad_norm": 2.091667694556132, + "learning_rate": 1.4739571956373333e-06, + "loss": 1.0311, + "step": 35240 + }, + { + "epoch": 0.83, + "grad_norm": 2.0715124346130906, + "learning_rate": 1.4735584879214116e-06, + "loss": 0.9289, + "step": 35241 + }, + { + "epoch": 0.83, + "grad_norm": 2.0262322591188298, + "learning_rate": 1.4731598298491313e-06, + "loss": 1.0558, + "step": 35242 + }, + { + "epoch": 0.83, + "grad_norm": 1.9053710479163461, + "learning_rate": 1.4727612214228115e-06, + "loss": 1.0196, + "step": 35243 + }, + { + "epoch": 0.83, + "grad_norm": 1.9655307111362057, + "learning_rate": 1.4723626626447784e-06, + "loss": 1.0145, + "step": 35244 + }, + { + "epoch": 0.83, + "grad_norm": 1.8237807796744219, + "learning_rate": 1.471964153517348e-06, + "loss": 0.9877, + "step": 35245 + }, + { + "epoch": 0.83, + "grad_norm": 1.8155433489414265, + "learning_rate": 1.4715656940428437e-06, + "loss": 0.938, + "step": 35246 + }, + { + "epoch": 0.83, + "grad_norm": 2.5427195007338828, + "learning_rate": 1.4711672842235814e-06, + "loss": 0.9442, + "step": 35247 + }, + { + "epoch": 0.83, + "grad_norm": 1.9849746839408975, + "learning_rate": 1.470768924061886e-06, + "loss": 1.0371, + "step": 35248 + }, + { + "epoch": 0.83, + "grad_norm": 2.083536613898178, + "learning_rate": 1.4703706135600715e-06, + "loss": 0.9373, + "step": 35249 + }, + { + "epoch": 0.83, + "grad_norm": 2.028738068595415, + "learning_rate": 1.4699723527204623e-06, + "loss": 1.0176, + "step": 35250 + }, + { + "epoch": 0.83, + "grad_norm": 1.8625771250608505, + "learning_rate": 1.4695741415453757e-06, + "loss": 1.0779, + "step": 35251 + }, + { + "epoch": 0.83, + "grad_norm": 1.9701607536019312, + "learning_rate": 1.4691759800371252e-06, + "loss": 1.1177, + "step": 35252 + }, + { + "epoch": 0.83, + "grad_norm": 2.230170761673526, + "learning_rate": 1.4687778681980335e-06, + "loss": 0.7797, + "step": 35253 + }, + { + "epoch": 0.83, + "grad_norm": 2.1314560128553968, + "learning_rate": 1.468379806030421e-06, + "loss": 1.0233, + "step": 35254 + }, + { + "epoch": 0.83, + "grad_norm": 2.077242384177321, + "learning_rate": 1.4679817935366014e-06, + "loss": 0.9407, + "step": 35255 + }, + { + "epoch": 0.83, + "grad_norm": 2.083158179671316, + "learning_rate": 1.4675838307188916e-06, + "loss": 0.9909, + "step": 35256 + }, + { + "epoch": 0.83, + "grad_norm": 2.120323853938472, + "learning_rate": 1.4671859175796087e-06, + "loss": 1.0248, + "step": 35257 + }, + { + "epoch": 0.83, + "grad_norm": 1.9458807661974624, + "learning_rate": 1.4667880541210744e-06, + "loss": 1.1224, + "step": 35258 + }, + { + "epoch": 0.83, + "grad_norm": 2.0507357580952137, + "learning_rate": 1.4663902403456009e-06, + "loss": 0.9056, + "step": 35259 + }, + { + "epoch": 0.83, + "grad_norm": 1.8968382773063521, + "learning_rate": 1.4659924762555033e-06, + "loss": 1.0517, + "step": 35260 + }, + { + "epoch": 0.83, + "grad_norm": 2.0078877326400075, + "learning_rate": 1.4655947618530998e-06, + "loss": 0.8975, + "step": 35261 + }, + { + "epoch": 0.83, + "grad_norm": 1.8236963107396809, + "learning_rate": 1.465197097140707e-06, + "loss": 0.9, + "step": 35262 + }, + { + "epoch": 0.83, + "grad_norm": 2.171826561252133, + "learning_rate": 1.464799482120639e-06, + "loss": 0.9349, + "step": 35263 + }, + { + "epoch": 0.83, + "grad_norm": 2.176656722493475, + "learning_rate": 1.4644019167952084e-06, + "loss": 0.9595, + "step": 35264 + }, + { + "epoch": 0.83, + "grad_norm": 2.6935530935025103, + "learning_rate": 1.4640044011667343e-06, + "loss": 0.9409, + "step": 35265 + }, + { + "epoch": 0.83, + "grad_norm": 1.0727431975452228, + "learning_rate": 1.463606935237526e-06, + "loss": 0.9009, + "step": 35266 + }, + { + "epoch": 0.83, + "grad_norm": 1.9780289561204594, + "learning_rate": 1.4632095190099039e-06, + "loss": 0.9373, + "step": 35267 + }, + { + "epoch": 0.83, + "grad_norm": 1.8183353813152565, + "learning_rate": 1.462812152486176e-06, + "loss": 0.9562, + "step": 35268 + }, + { + "epoch": 0.83, + "grad_norm": 2.02732814598372, + "learning_rate": 1.4624148356686619e-06, + "loss": 0.973, + "step": 35269 + }, + { + "epoch": 0.83, + "grad_norm": 1.7983133568885985, + "learning_rate": 1.4620175685596683e-06, + "loss": 0.883, + "step": 35270 + }, + { + "epoch": 0.83, + "grad_norm": 2.243778645278131, + "learning_rate": 1.4616203511615134e-06, + "loss": 1.0701, + "step": 35271 + }, + { + "epoch": 0.83, + "grad_norm": 2.0975909716866554, + "learning_rate": 1.461223183476509e-06, + "loss": 1.1041, + "step": 35272 + }, + { + "epoch": 0.83, + "grad_norm": 2.066278045558653, + "learning_rate": 1.4608260655069639e-06, + "loss": 0.9983, + "step": 35273 + }, + { + "epoch": 0.83, + "grad_norm": 2.0401395468938532, + "learning_rate": 1.4604289972551933e-06, + "loss": 1.0851, + "step": 35274 + }, + { + "epoch": 0.83, + "grad_norm": 1.850541439125312, + "learning_rate": 1.4600319787235073e-06, + "loss": 0.9088, + "step": 35275 + }, + { + "epoch": 0.83, + "grad_norm": 2.124055586275614, + "learning_rate": 1.4596350099142243e-06, + "loss": 1.1764, + "step": 35276 + }, + { + "epoch": 0.83, + "grad_norm": 2.036295961228988, + "learning_rate": 1.4592380908296455e-06, + "loss": 0.9757, + "step": 35277 + }, + { + "epoch": 0.83, + "grad_norm": 1.8754053189362088, + "learning_rate": 1.458841221472087e-06, + "loss": 0.8491, + "step": 35278 + }, + { + "epoch": 0.83, + "grad_norm": 3.0688851751008155, + "learning_rate": 1.4584444018438616e-06, + "loss": 0.9755, + "step": 35279 + }, + { + "epoch": 0.83, + "grad_norm": 2.767457357465791, + "learning_rate": 1.4580476319472758e-06, + "loss": 0.9218, + "step": 35280 + }, + { + "epoch": 0.83, + "grad_norm": 2.1736899248549677, + "learning_rate": 1.4576509117846394e-06, + "loss": 1.033, + "step": 35281 + }, + { + "epoch": 0.83, + "grad_norm": 2.0310638958012133, + "learning_rate": 1.4572542413582635e-06, + "loss": 1.059, + "step": 35282 + }, + { + "epoch": 0.83, + "grad_norm": 1.9982015926460954, + "learning_rate": 1.456857620670461e-06, + "loss": 1.0169, + "step": 35283 + }, + { + "epoch": 0.83, + "grad_norm": 2.160397855395281, + "learning_rate": 1.4564610497235377e-06, + "loss": 0.9194, + "step": 35284 + }, + { + "epoch": 0.83, + "grad_norm": 1.9788836646899584, + "learning_rate": 1.4560645285198006e-06, + "loss": 0.9628, + "step": 35285 + }, + { + "epoch": 0.83, + "grad_norm": 1.9671504201075907, + "learning_rate": 1.4556680570615623e-06, + "loss": 0.975, + "step": 35286 + }, + { + "epoch": 0.83, + "grad_norm": 2.2920731942408854, + "learning_rate": 1.4552716353511286e-06, + "loss": 0.9608, + "step": 35287 + }, + { + "epoch": 0.83, + "grad_norm": 1.0638071835551066, + "learning_rate": 1.454875263390808e-06, + "loss": 0.872, + "step": 35288 + }, + { + "epoch": 0.83, + "grad_norm": 1.9895263948435986, + "learning_rate": 1.4544789411829118e-06, + "loss": 0.9782, + "step": 35289 + }, + { + "epoch": 0.83, + "grad_norm": 1.8359052662860507, + "learning_rate": 1.454082668729745e-06, + "loss": 1.0144, + "step": 35290 + }, + { + "epoch": 0.83, + "grad_norm": 2.6538249773767366, + "learning_rate": 1.4536864460336108e-06, + "loss": 0.9504, + "step": 35291 + }, + { + "epoch": 0.83, + "grad_norm": 2.1030556067070387, + "learning_rate": 1.4532902730968213e-06, + "loss": 0.94, + "step": 35292 + }, + { + "epoch": 0.83, + "grad_norm": 2.328815955618591, + "learning_rate": 1.4528941499216853e-06, + "loss": 0.9985, + "step": 35293 + }, + { + "epoch": 0.83, + "grad_norm": 2.1918206720831344, + "learning_rate": 1.4524980765105013e-06, + "loss": 0.882, + "step": 35294 + }, + { + "epoch": 0.83, + "grad_norm": 2.0105675648828516, + "learning_rate": 1.4521020528655794e-06, + "loss": 0.9926, + "step": 35295 + }, + { + "epoch": 0.83, + "grad_norm": 1.8640431828339288, + "learning_rate": 1.4517060789892257e-06, + "loss": 0.9342, + "step": 35296 + }, + { + "epoch": 0.83, + "grad_norm": 2.3220892411895617, + "learning_rate": 1.4513101548837505e-06, + "loss": 0.909, + "step": 35297 + }, + { + "epoch": 0.83, + "grad_norm": 2.091931202637714, + "learning_rate": 1.4509142805514486e-06, + "loss": 0.8692, + "step": 35298 + }, + { + "epoch": 0.83, + "grad_norm": 2.285676435161535, + "learning_rate": 1.4505184559946306e-06, + "loss": 1.0438, + "step": 35299 + }, + { + "epoch": 0.83, + "grad_norm": 1.053034215100056, + "learning_rate": 1.4501226812156032e-06, + "loss": 0.9742, + "step": 35300 + }, + { + "epoch": 0.83, + "grad_norm": 2.284627948421884, + "learning_rate": 1.4497269562166672e-06, + "loss": 0.9132, + "step": 35301 + }, + { + "epoch": 0.83, + "grad_norm": 2.121161407876597, + "learning_rate": 1.4493312810001293e-06, + "loss": 1.0523, + "step": 35302 + }, + { + "epoch": 0.83, + "grad_norm": 1.9178243491559215, + "learning_rate": 1.4489356555682887e-06, + "loss": 0.9609, + "step": 35303 + }, + { + "epoch": 0.83, + "grad_norm": 2.207756120666107, + "learning_rate": 1.448540079923455e-06, + "loss": 1.0421, + "step": 35304 + }, + { + "epoch": 0.83, + "grad_norm": 1.180405987299318, + "learning_rate": 1.4481445540679261e-06, + "loss": 0.8734, + "step": 35305 + }, + { + "epoch": 0.83, + "grad_norm": 2.037619706621206, + "learning_rate": 1.4477490780040093e-06, + "loss": 0.9886, + "step": 35306 + }, + { + "epoch": 0.83, + "grad_norm": 1.2052230774832204, + "learning_rate": 1.4473536517340047e-06, + "loss": 0.9702, + "step": 35307 + }, + { + "epoch": 0.83, + "grad_norm": 1.9015019134699824, + "learning_rate": 1.446958275260213e-06, + "loss": 0.9168, + "step": 35308 + }, + { + "epoch": 0.83, + "grad_norm": 2.180252338380975, + "learning_rate": 1.4465629485849376e-06, + "loss": 0.978, + "step": 35309 + }, + { + "epoch": 0.83, + "grad_norm": 1.8299770060753366, + "learning_rate": 1.4461676717104834e-06, + "loss": 1.007, + "step": 35310 + }, + { + "epoch": 0.83, + "grad_norm": 2.7915098403098915, + "learning_rate": 1.4457724446391486e-06, + "loss": 0.9172, + "step": 35311 + }, + { + "epoch": 0.83, + "grad_norm": 2.0212958394232885, + "learning_rate": 1.4453772673732336e-06, + "loss": 0.9133, + "step": 35312 + }, + { + "epoch": 0.83, + "grad_norm": 1.7469008369136658, + "learning_rate": 1.4449821399150398e-06, + "loss": 0.9598, + "step": 35313 + }, + { + "epoch": 0.83, + "grad_norm": 2.0627055908390783, + "learning_rate": 1.44458706226687e-06, + "loss": 1.0192, + "step": 35314 + }, + { + "epoch": 0.83, + "grad_norm": 3.534529866659952, + "learning_rate": 1.4441920344310235e-06, + "loss": 1.0266, + "step": 35315 + }, + { + "epoch": 0.83, + "grad_norm": 12.852436974885718, + "learning_rate": 1.443797056409798e-06, + "loss": 0.9482, + "step": 35316 + }, + { + "epoch": 0.83, + "grad_norm": 1.8872255874649273, + "learning_rate": 1.4434021282054943e-06, + "loss": 1.06, + "step": 35317 + }, + { + "epoch": 0.83, + "grad_norm": 2.6625924928847002, + "learning_rate": 1.443007249820415e-06, + "loss": 1.0142, + "step": 35318 + }, + { + "epoch": 0.83, + "grad_norm": 2.1548802589064957, + "learning_rate": 1.442612421256856e-06, + "loss": 0.8958, + "step": 35319 + }, + { + "epoch": 0.83, + "grad_norm": 2.0153696107232766, + "learning_rate": 1.4422176425171153e-06, + "loss": 0.9808, + "step": 35320 + }, + { + "epoch": 0.83, + "grad_norm": 1.8197086971556904, + "learning_rate": 1.4418229136034922e-06, + "loss": 0.8731, + "step": 35321 + }, + { + "epoch": 0.83, + "grad_norm": 2.055344666542982, + "learning_rate": 1.4414282345182885e-06, + "loss": 1.1159, + "step": 35322 + }, + { + "epoch": 0.83, + "grad_norm": 2.1821126222376326, + "learning_rate": 1.4410336052637985e-06, + "loss": 0.7806, + "step": 35323 + }, + { + "epoch": 0.83, + "grad_norm": 1.8819294180266253, + "learning_rate": 1.4406390258423186e-06, + "loss": 0.8678, + "step": 35324 + }, + { + "epoch": 0.83, + "grad_norm": 1.9396208498998662, + "learning_rate": 1.4402444962561502e-06, + "loss": 0.9564, + "step": 35325 + }, + { + "epoch": 0.83, + "grad_norm": 2.0267458584301417, + "learning_rate": 1.4398500165075857e-06, + "loss": 0.9863, + "step": 35326 + }, + { + "epoch": 0.83, + "grad_norm": 2.1698853970263334, + "learning_rate": 1.4394555865989279e-06, + "loss": 0.9738, + "step": 35327 + }, + { + "epoch": 0.83, + "grad_norm": 1.0405861060610895, + "learning_rate": 1.439061206532466e-06, + "loss": 0.9002, + "step": 35328 + }, + { + "epoch": 0.83, + "grad_norm": 2.2671482211970457, + "learning_rate": 1.4386668763105039e-06, + "loss": 0.9016, + "step": 35329 + }, + { + "epoch": 0.83, + "grad_norm": 1.8825175430974819, + "learning_rate": 1.4382725959353305e-06, + "loss": 0.9391, + "step": 35330 + }, + { + "epoch": 0.83, + "grad_norm": 1.9767142215696414, + "learning_rate": 1.4378783654092476e-06, + "loss": 1.0522, + "step": 35331 + }, + { + "epoch": 0.83, + "grad_norm": 1.997603256259659, + "learning_rate": 1.4374841847345478e-06, + "loss": 0.9318, + "step": 35332 + }, + { + "epoch": 0.83, + "grad_norm": 2.3990487166773753, + "learning_rate": 1.4370900539135225e-06, + "loss": 0.9964, + "step": 35333 + }, + { + "epoch": 0.83, + "grad_norm": 1.7999932450503644, + "learning_rate": 1.4366959729484708e-06, + "loss": 0.9688, + "step": 35334 + }, + { + "epoch": 0.83, + "grad_norm": 1.1014607871642361, + "learning_rate": 1.4363019418416878e-06, + "loss": 0.9953, + "step": 35335 + }, + { + "epoch": 0.83, + "grad_norm": 2.578196488639172, + "learning_rate": 1.4359079605954673e-06, + "loss": 0.941, + "step": 35336 + }, + { + "epoch": 0.83, + "grad_norm": 2.4617029139804765, + "learning_rate": 1.4355140292120983e-06, + "loss": 1.043, + "step": 35337 + }, + { + "epoch": 0.83, + "grad_norm": 1.0494540546064606, + "learning_rate": 1.4351201476938792e-06, + "loss": 0.9365, + "step": 35338 + }, + { + "epoch": 0.83, + "grad_norm": 2.2542266029225226, + "learning_rate": 1.4347263160431036e-06, + "loss": 1.074, + "step": 35339 + }, + { + "epoch": 0.83, + "grad_norm": 1.076635381625496, + "learning_rate": 1.434332534262064e-06, + "loss": 0.8867, + "step": 35340 + }, + { + "epoch": 0.83, + "grad_norm": 1.777847365538721, + "learning_rate": 1.4339388023530498e-06, + "loss": 1.0494, + "step": 35341 + }, + { + "epoch": 0.83, + "grad_norm": 2.046045785573332, + "learning_rate": 1.433545120318355e-06, + "loss": 0.9655, + "step": 35342 + }, + { + "epoch": 0.83, + "grad_norm": 1.9929968977827963, + "learning_rate": 1.4331514881602759e-06, + "loss": 0.8888, + "step": 35343 + }, + { + "epoch": 0.83, + "grad_norm": 2.0191517786737108, + "learning_rate": 1.4327579058811003e-06, + "loss": 1.0276, + "step": 35344 + }, + { + "epoch": 0.83, + "grad_norm": 2.0040801182686927, + "learning_rate": 1.4323643734831182e-06, + "loss": 1.0894, + "step": 35345 + }, + { + "epoch": 0.83, + "grad_norm": 1.9542968772493219, + "learning_rate": 1.4319708909686259e-06, + "loss": 1.0669, + "step": 35346 + }, + { + "epoch": 0.83, + "grad_norm": 2.090646499912939, + "learning_rate": 1.43157745833991e-06, + "loss": 1.0012, + "step": 35347 + }, + { + "epoch": 0.83, + "grad_norm": 2.266843150850748, + "learning_rate": 1.4311840755992634e-06, + "loss": 0.9471, + "step": 35348 + }, + { + "epoch": 0.83, + "grad_norm": 1.860295033910175, + "learning_rate": 1.4307907427489753e-06, + "loss": 1.1165, + "step": 35349 + }, + { + "epoch": 0.83, + "grad_norm": 2.0343542930579046, + "learning_rate": 1.4303974597913383e-06, + "loss": 0.9017, + "step": 35350 + }, + { + "epoch": 0.83, + "grad_norm": 2.393256543472559, + "learning_rate": 1.4300042267286374e-06, + "loss": 0.9495, + "step": 35351 + }, + { + "epoch": 0.83, + "grad_norm": 1.972114870700588, + "learning_rate": 1.4296110435631683e-06, + "loss": 0.9392, + "step": 35352 + }, + { + "epoch": 0.83, + "grad_norm": 1.160514712445742, + "learning_rate": 1.429217910297217e-06, + "loss": 0.9836, + "step": 35353 + }, + { + "epoch": 0.83, + "grad_norm": 1.795725643133153, + "learning_rate": 1.4288248269330695e-06, + "loss": 0.9706, + "step": 35354 + }, + { + "epoch": 0.83, + "grad_norm": 2.02960714009702, + "learning_rate": 1.4284317934730175e-06, + "loss": 0.8682, + "step": 35355 + }, + { + "epoch": 0.83, + "grad_norm": 2.194875323217251, + "learning_rate": 1.428038809919352e-06, + "loss": 1.0472, + "step": 35356 + }, + { + "epoch": 0.83, + "grad_norm": 1.8155267826974293, + "learning_rate": 1.4276458762743583e-06, + "loss": 0.9886, + "step": 35357 + }, + { + "epoch": 0.83, + "grad_norm": 1.939893615488287, + "learning_rate": 1.4272529925403221e-06, + "loss": 1.0195, + "step": 35358 + }, + { + "epoch": 0.83, + "grad_norm": 2.0687321572591353, + "learning_rate": 1.426860158719533e-06, + "loss": 0.9987, + "step": 35359 + }, + { + "epoch": 0.83, + "grad_norm": 1.946281490153288, + "learning_rate": 1.426467374814281e-06, + "loss": 0.8984, + "step": 35360 + }, + { + "epoch": 0.83, + "grad_norm": 2.1683298030474734, + "learning_rate": 1.426074640826849e-06, + "loss": 0.9702, + "step": 35361 + }, + { + "epoch": 0.83, + "grad_norm": 1.9873910077372352, + "learning_rate": 1.4256819567595238e-06, + "loss": 0.946, + "step": 35362 + }, + { + "epoch": 0.83, + "grad_norm": 3.066468772812654, + "learning_rate": 1.4252893226145924e-06, + "loss": 0.8099, + "step": 35363 + }, + { + "epoch": 0.83, + "grad_norm": 1.966502319876046, + "learning_rate": 1.4248967383943435e-06, + "loss": 1.0619, + "step": 35364 + }, + { + "epoch": 0.83, + "grad_norm": 1.9889373354944102, + "learning_rate": 1.4245042041010604e-06, + "loss": 1.0112, + "step": 35365 + }, + { + "epoch": 0.83, + "grad_norm": 2.08844317537336, + "learning_rate": 1.424111719737027e-06, + "loss": 1.0241, + "step": 35366 + }, + { + "epoch": 0.83, + "grad_norm": 1.1044574033469208, + "learning_rate": 1.4237192853045323e-06, + "loss": 0.9171, + "step": 35367 + }, + { + "epoch": 0.83, + "grad_norm": 1.8536644594849698, + "learning_rate": 1.423326900805857e-06, + "loss": 1.0911, + "step": 35368 + }, + { + "epoch": 0.83, + "grad_norm": 1.829858537729345, + "learning_rate": 1.4229345662432903e-06, + "loss": 0.8536, + "step": 35369 + }, + { + "epoch": 0.83, + "grad_norm": 2.1060918047999495, + "learning_rate": 1.4225422816191115e-06, + "loss": 1.0363, + "step": 35370 + }, + { + "epoch": 0.83, + "grad_norm": 2.1054771168056874, + "learning_rate": 1.4221500469356086e-06, + "loss": 0.9216, + "step": 35371 + }, + { + "epoch": 0.83, + "grad_norm": 2.04802246169692, + "learning_rate": 1.4217578621950623e-06, + "loss": 1.0447, + "step": 35372 + }, + { + "epoch": 0.83, + "grad_norm": 2.4561716322046507, + "learning_rate": 1.4213657273997571e-06, + "loss": 1.0221, + "step": 35373 + }, + { + "epoch": 0.83, + "grad_norm": 2.0111706937365903, + "learning_rate": 1.4209736425519827e-06, + "loss": 1.0251, + "step": 35374 + }, + { + "epoch": 0.83, + "grad_norm": 2.529058303407072, + "learning_rate": 1.420581607654009e-06, + "loss": 0.9289, + "step": 35375 + }, + { + "epoch": 0.83, + "grad_norm": 1.9111789310145975, + "learning_rate": 1.420189622708127e-06, + "loss": 1.0622, + "step": 35376 + }, + { + "epoch": 0.83, + "grad_norm": 2.1961869816742094, + "learning_rate": 1.4197976877166165e-06, + "loss": 1.0075, + "step": 35377 + }, + { + "epoch": 0.83, + "grad_norm": 2.3449358468391712, + "learning_rate": 1.4194058026817658e-06, + "loss": 0.964, + "step": 35378 + }, + { + "epoch": 0.83, + "grad_norm": 2.8155764389527502, + "learning_rate": 1.4190139676058457e-06, + "loss": 0.9764, + "step": 35379 + }, + { + "epoch": 0.83, + "grad_norm": 2.148564688158092, + "learning_rate": 1.4186221824911428e-06, + "loss": 0.9483, + "step": 35380 + }, + { + "epoch": 0.83, + "grad_norm": 2.0141665882817072, + "learning_rate": 1.418230447339941e-06, + "loss": 0.9652, + "step": 35381 + }, + { + "epoch": 0.83, + "grad_norm": 2.0274520898019057, + "learning_rate": 1.4178387621545187e-06, + "loss": 0.8665, + "step": 35382 + }, + { + "epoch": 0.83, + "grad_norm": 1.7745898483269733, + "learning_rate": 1.4174471269371537e-06, + "loss": 0.9538, + "step": 35383 + }, + { + "epoch": 0.83, + "grad_norm": 2.0580835013161853, + "learning_rate": 1.417055541690129e-06, + "loss": 0.928, + "step": 35384 + }, + { + "epoch": 0.83, + "grad_norm": 1.9395396059900434, + "learning_rate": 1.4166640064157267e-06, + "loss": 1.0423, + "step": 35385 + }, + { + "epoch": 0.83, + "grad_norm": 1.8094320753315019, + "learning_rate": 1.416272521116221e-06, + "loss": 0.8857, + "step": 35386 + }, + { + "epoch": 0.83, + "grad_norm": 2.3321270738296334, + "learning_rate": 1.4158810857938964e-06, + "loss": 0.9861, + "step": 35387 + }, + { + "epoch": 0.83, + "grad_norm": 1.9636354262234368, + "learning_rate": 1.4154897004510304e-06, + "loss": 1.052, + "step": 35388 + }, + { + "epoch": 0.83, + "grad_norm": 1.052560084221159, + "learning_rate": 1.4150983650898987e-06, + "loss": 0.9348, + "step": 35389 + }, + { + "epoch": 0.83, + "grad_norm": 1.9383135667611717, + "learning_rate": 1.4147070797127826e-06, + "loss": 1.0048, + "step": 35390 + }, + { + "epoch": 0.83, + "grad_norm": 2.0144850167207156, + "learning_rate": 1.4143158443219617e-06, + "loss": 0.9693, + "step": 35391 + }, + { + "epoch": 0.83, + "grad_norm": 2.5991614732036887, + "learning_rate": 1.413924658919712e-06, + "loss": 0.8527, + "step": 35392 + }, + { + "epoch": 0.83, + "grad_norm": 2.4312018441912624, + "learning_rate": 1.4135335235083092e-06, + "loss": 0.9708, + "step": 35393 + }, + { + "epoch": 0.83, + "grad_norm": 1.9619289253321275, + "learning_rate": 1.4131424380900327e-06, + "loss": 1.0799, + "step": 35394 + }, + { + "epoch": 0.83, + "grad_norm": 2.4719499774067804, + "learning_rate": 1.412751402667164e-06, + "loss": 1.0744, + "step": 35395 + }, + { + "epoch": 0.83, + "grad_norm": 2.0487241101521994, + "learning_rate": 1.4123604172419714e-06, + "loss": 0.9759, + "step": 35396 + }, + { + "epoch": 0.83, + "grad_norm": 1.9341954661712528, + "learning_rate": 1.4119694818167351e-06, + "loss": 0.9539, + "step": 35397 + }, + { + "epoch": 0.83, + "grad_norm": 2.2980869400034027, + "learning_rate": 1.4115785963937302e-06, + "loss": 0.9383, + "step": 35398 + }, + { + "epoch": 0.83, + "grad_norm": 2.2033986203237226, + "learning_rate": 1.4111877609752367e-06, + "loss": 1.0178, + "step": 35399 + }, + { + "epoch": 0.83, + "grad_norm": 2.102369325761156, + "learning_rate": 1.4107969755635277e-06, + "loss": 0.9887, + "step": 35400 + }, + { + "epoch": 0.83, + "grad_norm": 1.0922090414775059, + "learning_rate": 1.4104062401608753e-06, + "loss": 0.9247, + "step": 35401 + }, + { + "epoch": 0.83, + "grad_norm": 2.0412757100939256, + "learning_rate": 1.4100155547695592e-06, + "loss": 0.9631, + "step": 35402 + }, + { + "epoch": 0.83, + "grad_norm": 2.0584067337417498, + "learning_rate": 1.4096249193918509e-06, + "loss": 0.9778, + "step": 35403 + }, + { + "epoch": 0.83, + "grad_norm": 1.9678050810113084, + "learning_rate": 1.4092343340300273e-06, + "loss": 0.9577, + "step": 35404 + }, + { + "epoch": 0.83, + "grad_norm": 1.9589377813227364, + "learning_rate": 1.408843798686359e-06, + "loss": 0.9757, + "step": 35405 + }, + { + "epoch": 0.83, + "grad_norm": 1.8576359800932938, + "learning_rate": 1.408453313363124e-06, + "loss": 0.8811, + "step": 35406 + }, + { + "epoch": 0.83, + "grad_norm": 1.850429421980068, + "learning_rate": 1.4080628780625927e-06, + "loss": 0.9226, + "step": 35407 + }, + { + "epoch": 0.83, + "grad_norm": 2.0618085695644566, + "learning_rate": 1.407672492787041e-06, + "loss": 0.8645, + "step": 35408 + }, + { + "epoch": 0.83, + "grad_norm": 1.8949980829170692, + "learning_rate": 1.4072821575387407e-06, + "loss": 0.8934, + "step": 35409 + }, + { + "epoch": 0.83, + "grad_norm": 1.9186639528539302, + "learning_rate": 1.406891872319961e-06, + "loss": 1.0031, + "step": 35410 + }, + { + "epoch": 0.83, + "grad_norm": 1.8660434201569889, + "learning_rate": 1.4065016371329775e-06, + "loss": 0.9503, + "step": 35411 + }, + { + "epoch": 0.83, + "grad_norm": 2.052791842827655, + "learning_rate": 1.4061114519800655e-06, + "loss": 0.9406, + "step": 35412 + }, + { + "epoch": 0.83, + "grad_norm": 1.816024373722788, + "learning_rate": 1.405721316863492e-06, + "loss": 0.9433, + "step": 35413 + }, + { + "epoch": 0.83, + "grad_norm": 2.257938100095709, + "learning_rate": 1.4053312317855284e-06, + "loss": 0.9619, + "step": 35414 + }, + { + "epoch": 0.83, + "grad_norm": 2.169911504625421, + "learning_rate": 1.4049411967484472e-06, + "loss": 0.9523, + "step": 35415 + }, + { + "epoch": 0.83, + "grad_norm": 3.0862000274502783, + "learning_rate": 1.4045512117545224e-06, + "loss": 0.9476, + "step": 35416 + }, + { + "epoch": 0.83, + "grad_norm": 1.8749493081720787, + "learning_rate": 1.4041612768060208e-06, + "loss": 1.0649, + "step": 35417 + }, + { + "epoch": 0.83, + "grad_norm": 1.910098235942635, + "learning_rate": 1.4037713919052121e-06, + "loss": 1.0251, + "step": 35418 + }, + { + "epoch": 0.83, + "grad_norm": 1.8665943668851859, + "learning_rate": 1.4033815570543673e-06, + "loss": 0.9268, + "step": 35419 + }, + { + "epoch": 0.83, + "grad_norm": 1.1326360993706688, + "learning_rate": 1.4029917722557597e-06, + "loss": 0.9385, + "step": 35420 + }, + { + "epoch": 0.83, + "grad_norm": 2.4896404327540114, + "learning_rate": 1.4026020375116545e-06, + "loss": 0.8942, + "step": 35421 + }, + { + "epoch": 0.83, + "grad_norm": 2.0089971104174094, + "learning_rate": 1.4022123528243204e-06, + "loss": 1.0555, + "step": 35422 + }, + { + "epoch": 0.83, + "grad_norm": 1.8943923209193807, + "learning_rate": 1.4018227181960308e-06, + "loss": 0.8665, + "step": 35423 + }, + { + "epoch": 0.83, + "grad_norm": 1.1099668092436425, + "learning_rate": 1.4014331336290487e-06, + "loss": 0.9411, + "step": 35424 + }, + { + "epoch": 0.83, + "grad_norm": 2.1703774776937967, + "learning_rate": 1.4010435991256477e-06, + "loss": 0.9216, + "step": 35425 + }, + { + "epoch": 0.83, + "grad_norm": 1.9741289826588617, + "learning_rate": 1.4006541146880914e-06, + "loss": 1.008, + "step": 35426 + }, + { + "epoch": 0.83, + "grad_norm": 1.899538097891702, + "learning_rate": 1.4002646803186514e-06, + "loss": 1.0777, + "step": 35427 + }, + { + "epoch": 0.83, + "grad_norm": 2.1280078391061426, + "learning_rate": 1.3998752960195905e-06, + "loss": 0.9743, + "step": 35428 + }, + { + "epoch": 0.83, + "grad_norm": 2.0232573111632, + "learning_rate": 1.39948596179318e-06, + "loss": 1.0395, + "step": 35429 + }, + { + "epoch": 0.83, + "grad_norm": 1.926665250214699, + "learning_rate": 1.3990966776416859e-06, + "loss": 0.8916, + "step": 35430 + }, + { + "epoch": 0.83, + "grad_norm": 1.8894130767212105, + "learning_rate": 1.398707443567372e-06, + "loss": 1.0236, + "step": 35431 + }, + { + "epoch": 0.83, + "grad_norm": 1.0905609573864044, + "learning_rate": 1.3983182595725054e-06, + "loss": 0.9147, + "step": 35432 + }, + { + "epoch": 0.83, + "grad_norm": 1.1030957489746838, + "learning_rate": 1.3979291256593564e-06, + "loss": 1.0087, + "step": 35433 + }, + { + "epoch": 0.83, + "grad_norm": 2.095236318536434, + "learning_rate": 1.3975400418301866e-06, + "loss": 0.9405, + "step": 35434 + }, + { + "epoch": 0.83, + "grad_norm": 3.188784948773246, + "learning_rate": 1.3971510080872608e-06, + "loss": 0.9368, + "step": 35435 + }, + { + "epoch": 0.83, + "grad_norm": 1.9173505250678422, + "learning_rate": 1.396762024432845e-06, + "loss": 1.0869, + "step": 35436 + }, + { + "epoch": 0.83, + "grad_norm": 1.8552306636722058, + "learning_rate": 1.3963730908692065e-06, + "loss": 1.0228, + "step": 35437 + }, + { + "epoch": 0.83, + "grad_norm": 1.838569952497407, + "learning_rate": 1.3959842073986085e-06, + "loss": 1.0621, + "step": 35438 + }, + { + "epoch": 0.83, + "grad_norm": 3.9164692944727353, + "learning_rate": 1.3955953740233107e-06, + "loss": 1.0954, + "step": 35439 + }, + { + "epoch": 0.83, + "grad_norm": 2.119722132235296, + "learning_rate": 1.3952065907455825e-06, + "loss": 1.1637, + "step": 35440 + }, + { + "epoch": 0.83, + "grad_norm": 2.1730607557856296, + "learning_rate": 1.3948178575676862e-06, + "loss": 0.9632, + "step": 35441 + }, + { + "epoch": 0.83, + "grad_norm": 1.05911522598269, + "learning_rate": 1.394429174491886e-06, + "loss": 0.8908, + "step": 35442 + }, + { + "epoch": 0.84, + "grad_norm": 2.1352942020444416, + "learning_rate": 1.3940405415204417e-06, + "loss": 1.1178, + "step": 35443 + }, + { + "epoch": 0.84, + "grad_norm": 1.0400641000671085, + "learning_rate": 1.3936519586556163e-06, + "loss": 0.9673, + "step": 35444 + }, + { + "epoch": 0.84, + "grad_norm": 2.631804629902239, + "learning_rate": 1.393263425899678e-06, + "loss": 0.9882, + "step": 35445 + }, + { + "epoch": 0.84, + "grad_norm": 1.9024056449285314, + "learning_rate": 1.3928749432548837e-06, + "loss": 1.0345, + "step": 35446 + }, + { + "epoch": 0.84, + "grad_norm": 1.7940136087133285, + "learning_rate": 1.3924865107234942e-06, + "loss": 1.0904, + "step": 35447 + }, + { + "epoch": 0.84, + "grad_norm": 2.0314714766651183, + "learning_rate": 1.3920981283077762e-06, + "loss": 0.9454, + "step": 35448 + }, + { + "epoch": 0.84, + "grad_norm": 2.0428238505286584, + "learning_rate": 1.3917097960099856e-06, + "loss": 0.9771, + "step": 35449 + }, + { + "epoch": 0.84, + "grad_norm": 2.1899967519385832, + "learning_rate": 1.3913215138323877e-06, + "loss": 1.0708, + "step": 35450 + }, + { + "epoch": 0.84, + "grad_norm": 1.7548581478782266, + "learning_rate": 1.3909332817772402e-06, + "loss": 0.9509, + "step": 35451 + }, + { + "epoch": 0.84, + "grad_norm": 1.8817216371270045, + "learning_rate": 1.3905450998468062e-06, + "loss": 1.0338, + "step": 35452 + }, + { + "epoch": 0.84, + "grad_norm": 2.154350186182892, + "learning_rate": 1.3901569680433425e-06, + "loss": 1.0006, + "step": 35453 + }, + { + "epoch": 0.84, + "grad_norm": 2.1739909245729807, + "learning_rate": 1.3897688863691128e-06, + "loss": 0.9989, + "step": 35454 + }, + { + "epoch": 0.84, + "grad_norm": 2.060556899233928, + "learning_rate": 1.3893808548263742e-06, + "loss": 0.9496, + "step": 35455 + }, + { + "epoch": 0.84, + "grad_norm": 1.9239827097591131, + "learning_rate": 1.388992873417384e-06, + "loss": 0.9057, + "step": 35456 + }, + { + "epoch": 0.84, + "grad_norm": 1.7730773483146571, + "learning_rate": 1.3886049421444048e-06, + "loss": 0.9636, + "step": 35457 + }, + { + "epoch": 0.84, + "grad_norm": 2.191411463115904, + "learning_rate": 1.3882170610096946e-06, + "loss": 0.8394, + "step": 35458 + }, + { + "epoch": 0.84, + "grad_norm": 2.3165123217260866, + "learning_rate": 1.3878292300155115e-06, + "loss": 1.0268, + "step": 35459 + }, + { + "epoch": 0.84, + "grad_norm": 1.1052010900035443, + "learning_rate": 1.387441449164112e-06, + "loss": 0.9022, + "step": 35460 + }, + { + "epoch": 0.84, + "grad_norm": 2.0455129213532497, + "learning_rate": 1.3870537184577538e-06, + "loss": 1.0884, + "step": 35461 + }, + { + "epoch": 0.84, + "grad_norm": 2.1209080549618022, + "learning_rate": 1.3866660378986985e-06, + "loss": 0.8167, + "step": 35462 + }, + { + "epoch": 0.84, + "grad_norm": 1.1318844841591869, + "learning_rate": 1.386278407489201e-06, + "loss": 0.9746, + "step": 35463 + }, + { + "epoch": 0.84, + "grad_norm": 1.9215635892026979, + "learning_rate": 1.3858908272315153e-06, + "loss": 0.883, + "step": 35464 + }, + { + "epoch": 0.84, + "grad_norm": 1.8798655467045546, + "learning_rate": 1.3855032971278991e-06, + "loss": 1.0276, + "step": 35465 + }, + { + "epoch": 0.84, + "grad_norm": 2.0059932519665895, + "learning_rate": 1.3851158171806134e-06, + "loss": 0.9618, + "step": 35466 + }, + { + "epoch": 0.84, + "grad_norm": 2.3623714197107613, + "learning_rate": 1.3847283873919103e-06, + "loss": 1.0385, + "step": 35467 + }, + { + "epoch": 0.84, + "grad_norm": 2.0830304684012715, + "learning_rate": 1.384341007764044e-06, + "loss": 0.9487, + "step": 35468 + }, + { + "epoch": 0.84, + "grad_norm": 1.8858023428446484, + "learning_rate": 1.3839536782992746e-06, + "loss": 1.0262, + "step": 35469 + }, + { + "epoch": 0.84, + "grad_norm": 1.9884398042117528, + "learning_rate": 1.3835663989998528e-06, + "loss": 0.9559, + "step": 35470 + }, + { + "epoch": 0.84, + "grad_norm": 2.197352112445477, + "learning_rate": 1.3831791698680342e-06, + "loss": 0.9183, + "step": 35471 + }, + { + "epoch": 0.84, + "grad_norm": 1.034095673507583, + "learning_rate": 1.3827919909060772e-06, + "loss": 0.9379, + "step": 35472 + }, + { + "epoch": 0.84, + "grad_norm": 2.238336304666582, + "learning_rate": 1.3824048621162334e-06, + "loss": 0.8878, + "step": 35473 + }, + { + "epoch": 0.84, + "grad_norm": 2.2071634776247375, + "learning_rate": 1.3820177835007541e-06, + "loss": 0.9151, + "step": 35474 + }, + { + "epoch": 0.84, + "grad_norm": 2.8269276748535277, + "learning_rate": 1.3816307550618968e-06, + "loss": 0.8933, + "step": 35475 + }, + { + "epoch": 0.84, + "grad_norm": 7.086373153286869, + "learning_rate": 1.3812437768019172e-06, + "loss": 1.0174, + "step": 35476 + }, + { + "epoch": 0.84, + "grad_norm": 2.0142222132515055, + "learning_rate": 1.3808568487230611e-06, + "loss": 1.1048, + "step": 35477 + }, + { + "epoch": 0.84, + "grad_norm": 1.8566426303860921, + "learning_rate": 1.3804699708275847e-06, + "loss": 0.9203, + "step": 35478 + }, + { + "epoch": 0.84, + "grad_norm": 1.8670341563902522, + "learning_rate": 1.3800831431177419e-06, + "loss": 0.7713, + "step": 35479 + }, + { + "epoch": 0.84, + "grad_norm": 2.007668982135121, + "learning_rate": 1.3796963655957872e-06, + "loss": 0.8863, + "step": 35480 + }, + { + "epoch": 0.84, + "grad_norm": 2.3012453324210957, + "learning_rate": 1.3793096382639658e-06, + "loss": 0.9855, + "step": 35481 + }, + { + "epoch": 0.84, + "grad_norm": 2.489135370464483, + "learning_rate": 1.3789229611245335e-06, + "loss": 0.8817, + "step": 35482 + }, + { + "epoch": 0.84, + "grad_norm": 1.9583887156616804, + "learning_rate": 1.378536334179743e-06, + "loss": 0.9988, + "step": 35483 + }, + { + "epoch": 0.84, + "grad_norm": 2.109436184163545, + "learning_rate": 1.3781497574318402e-06, + "loss": 0.9158, + "step": 35484 + }, + { + "epoch": 0.84, + "grad_norm": 2.24944080696841, + "learning_rate": 1.3777632308830824e-06, + "loss": 0.9969, + "step": 35485 + }, + { + "epoch": 0.84, + "grad_norm": 1.9479613187726503, + "learning_rate": 1.3773767545357141e-06, + "loss": 1.0057, + "step": 35486 + }, + { + "epoch": 0.84, + "grad_norm": 1.8301477964781732, + "learning_rate": 1.3769903283919905e-06, + "loss": 0.9328, + "step": 35487 + }, + { + "epoch": 0.84, + "grad_norm": 2.5562453620111927, + "learning_rate": 1.3766039524541575e-06, + "loss": 0.9492, + "step": 35488 + }, + { + "epoch": 0.84, + "grad_norm": 3.3208683124418337, + "learning_rate": 1.3762176267244686e-06, + "loss": 0.9588, + "step": 35489 + }, + { + "epoch": 0.84, + "grad_norm": 2.219781525955554, + "learning_rate": 1.3758313512051702e-06, + "loss": 0.92, + "step": 35490 + }, + { + "epoch": 0.84, + "grad_norm": 1.0620061355750992, + "learning_rate": 1.3754451258985102e-06, + "loss": 0.9753, + "step": 35491 + }, + { + "epoch": 0.84, + "grad_norm": 1.8534045780318626, + "learning_rate": 1.375058950806739e-06, + "loss": 0.906, + "step": 35492 + }, + { + "epoch": 0.84, + "grad_norm": 2.0162633707545794, + "learning_rate": 1.3746728259321085e-06, + "loss": 0.9497, + "step": 35493 + }, + { + "epoch": 0.84, + "grad_norm": 1.1231766946021744, + "learning_rate": 1.374286751276862e-06, + "loss": 0.9807, + "step": 35494 + }, + { + "epoch": 0.84, + "grad_norm": 2.056856079580704, + "learning_rate": 1.373900726843247e-06, + "loss": 0.9416, + "step": 35495 + }, + { + "epoch": 0.84, + "grad_norm": 2.5611124763417394, + "learning_rate": 1.3735147526335135e-06, + "loss": 1.0659, + "step": 35496 + }, + { + "epoch": 0.84, + "grad_norm": 1.9167567366977738, + "learning_rate": 1.37312882864991e-06, + "loss": 1.0329, + "step": 35497 + }, + { + "epoch": 0.84, + "grad_norm": 2.0608812240553496, + "learning_rate": 1.3727429548946825e-06, + "loss": 1.0196, + "step": 35498 + }, + { + "epoch": 0.84, + "grad_norm": 2.0407654137765823, + "learning_rate": 1.3723571313700734e-06, + "loss": 0.8812, + "step": 35499 + }, + { + "epoch": 0.84, + "grad_norm": 1.084647632461299, + "learning_rate": 1.3719713580783334e-06, + "loss": 0.9932, + "step": 35500 + }, + { + "epoch": 0.84, + "grad_norm": 1.994726341954048, + "learning_rate": 1.3715856350217095e-06, + "loss": 0.9122, + "step": 35501 + }, + { + "epoch": 0.84, + "grad_norm": 2.850635869051829, + "learning_rate": 1.3711999622024463e-06, + "loss": 0.9832, + "step": 35502 + }, + { + "epoch": 0.84, + "grad_norm": 1.9723216987652243, + "learning_rate": 1.3708143396227857e-06, + "loss": 1.006, + "step": 35503 + }, + { + "epoch": 0.84, + "grad_norm": 1.1500157817641719, + "learning_rate": 1.3704287672849793e-06, + "loss": 0.9249, + "step": 35504 + }, + { + "epoch": 0.84, + "grad_norm": 2.0681596484282614, + "learning_rate": 1.370043245191266e-06, + "loss": 0.9092, + "step": 35505 + }, + { + "epoch": 0.84, + "grad_norm": 1.9283610013200279, + "learning_rate": 1.3696577733438943e-06, + "loss": 0.944, + "step": 35506 + }, + { + "epoch": 0.84, + "grad_norm": 1.9283830498590806, + "learning_rate": 1.3692723517451067e-06, + "loss": 0.9672, + "step": 35507 + }, + { + "epoch": 0.84, + "grad_norm": 1.9282471864013737, + "learning_rate": 1.3688869803971493e-06, + "loss": 0.9005, + "step": 35508 + }, + { + "epoch": 0.84, + "grad_norm": 2.748407092629001, + "learning_rate": 1.3685016593022627e-06, + "loss": 0.9344, + "step": 35509 + }, + { + "epoch": 0.84, + "grad_norm": 1.0751286282635941, + "learning_rate": 1.3681163884626935e-06, + "loss": 0.9269, + "step": 35510 + }, + { + "epoch": 0.84, + "grad_norm": 2.3075826560991177, + "learning_rate": 1.3677311678806838e-06, + "loss": 0.932, + "step": 35511 + }, + { + "epoch": 0.84, + "grad_norm": 2.1306391502988506, + "learning_rate": 1.3673459975584746e-06, + "loss": 0.9421, + "step": 35512 + }, + { + "epoch": 0.84, + "grad_norm": 2.0442088528392217, + "learning_rate": 1.3669608774983102e-06, + "loss": 1.0318, + "step": 35513 + }, + { + "epoch": 0.84, + "grad_norm": 2.248438101795585, + "learning_rate": 1.3665758077024338e-06, + "loss": 0.9902, + "step": 35514 + }, + { + "epoch": 0.84, + "grad_norm": 2.0143066099848053, + "learning_rate": 1.3661907881730874e-06, + "loss": 0.9871, + "step": 35515 + }, + { + "epoch": 0.84, + "grad_norm": 2.1234806969984255, + "learning_rate": 1.365805818912509e-06, + "loss": 0.9415, + "step": 35516 + }, + { + "epoch": 0.84, + "grad_norm": 2.0567551267503776, + "learning_rate": 1.3654208999229423e-06, + "loss": 0.9675, + "step": 35517 + }, + { + "epoch": 0.84, + "grad_norm": 2.0575477271990876, + "learning_rate": 1.3650360312066314e-06, + "loss": 0.8788, + "step": 35518 + }, + { + "epoch": 0.84, + "grad_norm": 1.9996895600390416, + "learning_rate": 1.364651212765814e-06, + "loss": 0.9303, + "step": 35519 + }, + { + "epoch": 0.84, + "grad_norm": 2.127233571706248, + "learning_rate": 1.3642664446027298e-06, + "loss": 0.94, + "step": 35520 + }, + { + "epoch": 0.84, + "grad_norm": 2.072011246649741, + "learning_rate": 1.3638817267196203e-06, + "loss": 1.0969, + "step": 35521 + }, + { + "epoch": 0.84, + "grad_norm": 1.7546224331293436, + "learning_rate": 1.3634970591187269e-06, + "loss": 0.8375, + "step": 35522 + }, + { + "epoch": 0.84, + "grad_norm": 1.0844969981809616, + "learning_rate": 1.363112441802289e-06, + "loss": 0.9775, + "step": 35523 + }, + { + "epoch": 0.84, + "grad_norm": 1.8546808206121401, + "learning_rate": 1.3627278747725415e-06, + "loss": 1.0616, + "step": 35524 + }, + { + "epoch": 0.84, + "grad_norm": 2.2927181691447007, + "learning_rate": 1.3623433580317303e-06, + "loss": 0.9026, + "step": 35525 + }, + { + "epoch": 0.84, + "grad_norm": 2.0988266619267733, + "learning_rate": 1.3619588915820869e-06, + "loss": 1.0455, + "step": 35526 + }, + { + "epoch": 0.84, + "grad_norm": 1.9668145472538534, + "learning_rate": 1.3615744754258575e-06, + "loss": 0.9715, + "step": 35527 + }, + { + "epoch": 0.84, + "grad_norm": 1.8865970436325388, + "learning_rate": 1.3611901095652736e-06, + "loss": 0.8972, + "step": 35528 + }, + { + "epoch": 0.84, + "grad_norm": 1.8955446804240517, + "learning_rate": 1.3608057940025775e-06, + "loss": 0.8845, + "step": 35529 + }, + { + "epoch": 0.84, + "grad_norm": 1.9282284597926298, + "learning_rate": 1.3604215287400036e-06, + "loss": 0.9069, + "step": 35530 + }, + { + "epoch": 0.84, + "grad_norm": 2.092389357040815, + "learning_rate": 1.3600373137797917e-06, + "loss": 0.9663, + "step": 35531 + }, + { + "epoch": 0.84, + "grad_norm": 2.050554522061053, + "learning_rate": 1.3596531491241793e-06, + "loss": 0.8608, + "step": 35532 + }, + { + "epoch": 0.84, + "grad_norm": 1.0392381149508656, + "learning_rate": 1.3592690347754e-06, + "loss": 0.8998, + "step": 35533 + }, + { + "epoch": 0.84, + "grad_norm": 2.3086078447913736, + "learning_rate": 1.358884970735691e-06, + "loss": 0.9588, + "step": 35534 + }, + { + "epoch": 0.84, + "grad_norm": 4.261933108534539, + "learning_rate": 1.3585009570072917e-06, + "loss": 1.1361, + "step": 35535 + }, + { + "epoch": 0.84, + "grad_norm": 2.3293624865230447, + "learning_rate": 1.3581169935924355e-06, + "loss": 1.0536, + "step": 35536 + }, + { + "epoch": 0.84, + "grad_norm": 1.8776296529286272, + "learning_rate": 1.3577330804933563e-06, + "loss": 0.878, + "step": 35537 + }, + { + "epoch": 0.84, + "grad_norm": 2.06805302012198, + "learning_rate": 1.3573492177122916e-06, + "loss": 0.946, + "step": 35538 + }, + { + "epoch": 0.84, + "grad_norm": 2.2080874397061767, + "learning_rate": 1.356965405251478e-06, + "loss": 0.9909, + "step": 35539 + }, + { + "epoch": 0.84, + "grad_norm": 1.7774683891599012, + "learning_rate": 1.3565816431131474e-06, + "loss": 0.9748, + "step": 35540 + }, + { + "epoch": 0.84, + "grad_norm": 1.821717883064729, + "learning_rate": 1.3561979312995333e-06, + "loss": 0.9266, + "step": 35541 + }, + { + "epoch": 0.84, + "grad_norm": 2.0500509533059645, + "learning_rate": 1.3558142698128719e-06, + "loss": 0.9889, + "step": 35542 + }, + { + "epoch": 0.84, + "grad_norm": 1.8329591951695554, + "learning_rate": 1.355430658655398e-06, + "loss": 0.78, + "step": 35543 + }, + { + "epoch": 0.84, + "grad_norm": 1.9983577405454116, + "learning_rate": 1.355047097829344e-06, + "loss": 0.9233, + "step": 35544 + }, + { + "epoch": 0.84, + "grad_norm": 1.1419438462067342, + "learning_rate": 1.3546635873369396e-06, + "loss": 0.9732, + "step": 35545 + }, + { + "epoch": 0.84, + "grad_norm": 1.9229228949501915, + "learning_rate": 1.354280127180424e-06, + "loss": 0.9615, + "step": 35546 + }, + { + "epoch": 0.84, + "grad_norm": 2.146620744142913, + "learning_rate": 1.3538967173620243e-06, + "loss": 0.8988, + "step": 35547 + }, + { + "epoch": 0.84, + "grad_norm": 1.9795971675797732, + "learning_rate": 1.3535133578839777e-06, + "loss": 0.9953, + "step": 35548 + }, + { + "epoch": 0.84, + "grad_norm": 2.0670230506954517, + "learning_rate": 1.3531300487485111e-06, + "loss": 1.0229, + "step": 35549 + }, + { + "epoch": 0.84, + "grad_norm": 1.1326821909967766, + "learning_rate": 1.3527467899578617e-06, + "loss": 0.9948, + "step": 35550 + }, + { + "epoch": 0.84, + "grad_norm": 2.0052617133557553, + "learning_rate": 1.3523635815142556e-06, + "loss": 0.9902, + "step": 35551 + }, + { + "epoch": 0.84, + "grad_norm": 1.6387760849020285, + "learning_rate": 1.3519804234199286e-06, + "loss": 0.9962, + "step": 35552 + }, + { + "epoch": 0.84, + "grad_norm": 2.340100600377741, + "learning_rate": 1.351597315677109e-06, + "loss": 0.8807, + "step": 35553 + }, + { + "epoch": 0.84, + "grad_norm": 1.9545029662105606, + "learning_rate": 1.351214258288026e-06, + "loss": 0.9383, + "step": 35554 + }, + { + "epoch": 0.84, + "grad_norm": 1.2076790330608103, + "learning_rate": 1.3508312512549127e-06, + "loss": 0.9686, + "step": 35555 + }, + { + "epoch": 0.84, + "grad_norm": 2.129858991716221, + "learning_rate": 1.3504482945799991e-06, + "loss": 1.0439, + "step": 35556 + }, + { + "epoch": 0.84, + "grad_norm": 2.2760890681762413, + "learning_rate": 1.3500653882655146e-06, + "loss": 0.9157, + "step": 35557 + }, + { + "epoch": 0.84, + "grad_norm": 2.118814480532337, + "learning_rate": 1.3496825323136852e-06, + "loss": 1.0504, + "step": 35558 + }, + { + "epoch": 0.84, + "grad_norm": 1.0914804353193048, + "learning_rate": 1.349299726726744e-06, + "loss": 0.9107, + "step": 35559 + }, + { + "epoch": 0.84, + "grad_norm": 2.129405554130801, + "learning_rate": 1.3489169715069195e-06, + "loss": 0.9646, + "step": 35560 + }, + { + "epoch": 0.84, + "grad_norm": 1.9328639008274957, + "learning_rate": 1.3485342666564394e-06, + "loss": 0.9723, + "step": 35561 + }, + { + "epoch": 0.84, + "grad_norm": 2.4653956343110686, + "learning_rate": 1.3481516121775295e-06, + "loss": 0.9693, + "step": 35562 + }, + { + "epoch": 0.84, + "grad_norm": 1.9643299078362835, + "learning_rate": 1.3477690080724203e-06, + "loss": 1.0957, + "step": 35563 + }, + { + "epoch": 0.84, + "grad_norm": 1.0354028539621545, + "learning_rate": 1.3473864543433423e-06, + "loss": 0.9554, + "step": 35564 + }, + { + "epoch": 0.84, + "grad_norm": 1.8625226979654734, + "learning_rate": 1.3470039509925192e-06, + "loss": 0.8318, + "step": 35565 + }, + { + "epoch": 0.84, + "grad_norm": 2.271821474369761, + "learning_rate": 1.346621498022176e-06, + "loss": 1.0326, + "step": 35566 + }, + { + "epoch": 0.84, + "grad_norm": 2.1409624671887904, + "learning_rate": 1.3462390954345416e-06, + "loss": 0.9261, + "step": 35567 + }, + { + "epoch": 0.84, + "grad_norm": 2.6023873434342115, + "learning_rate": 1.3458567432318458e-06, + "loss": 1.0191, + "step": 35568 + }, + { + "epoch": 0.84, + "grad_norm": 1.9238566315422088, + "learning_rate": 1.3454744414163101e-06, + "loss": 1.0186, + "step": 35569 + }, + { + "epoch": 0.84, + "grad_norm": 2.829731556533894, + "learning_rate": 1.3450921899901637e-06, + "loss": 0.9505, + "step": 35570 + }, + { + "epoch": 0.84, + "grad_norm": 2.0203829832082967, + "learning_rate": 1.3447099889556302e-06, + "loss": 1.0455, + "step": 35571 + }, + { + "epoch": 0.84, + "grad_norm": 1.7850116667534024, + "learning_rate": 1.3443278383149338e-06, + "loss": 0.9881, + "step": 35572 + }, + { + "epoch": 0.84, + "grad_norm": 1.024587547782919, + "learning_rate": 1.3439457380703015e-06, + "loss": 0.9036, + "step": 35573 + }, + { + "epoch": 0.84, + "grad_norm": 2.0878220157484235, + "learning_rate": 1.3435636882239588e-06, + "loss": 0.9162, + "step": 35574 + }, + { + "epoch": 0.84, + "grad_norm": 6.683602415346206, + "learning_rate": 1.3431816887781302e-06, + "loss": 0.8555, + "step": 35575 + }, + { + "epoch": 0.84, + "grad_norm": 1.041873643485897, + "learning_rate": 1.3427997397350356e-06, + "loss": 0.9601, + "step": 35576 + }, + { + "epoch": 0.84, + "grad_norm": 1.8883291202113948, + "learning_rate": 1.3424178410969013e-06, + "loss": 0.9124, + "step": 35577 + }, + { + "epoch": 0.84, + "grad_norm": 1.0641505089026981, + "learning_rate": 1.3420359928659564e-06, + "loss": 0.9538, + "step": 35578 + }, + { + "epoch": 0.84, + "grad_norm": 1.8260230259360166, + "learning_rate": 1.341654195044414e-06, + "loss": 0.9875, + "step": 35579 + }, + { + "epoch": 0.84, + "grad_norm": 2.0782514899790816, + "learning_rate": 1.341272447634503e-06, + "loss": 0.8584, + "step": 35580 + }, + { + "epoch": 0.84, + "grad_norm": 2.0161850683893148, + "learning_rate": 1.340890750638444e-06, + "loss": 1.0263, + "step": 35581 + }, + { + "epoch": 0.84, + "grad_norm": 1.070514044572131, + "learning_rate": 1.3405091040584628e-06, + "loss": 0.9453, + "step": 35582 + }, + { + "epoch": 0.84, + "grad_norm": 1.8483550474010477, + "learning_rate": 1.34012750789678e-06, + "loss": 1.0899, + "step": 35583 + }, + { + "epoch": 0.84, + "grad_norm": 3.2162214778165197, + "learning_rate": 1.339745962155613e-06, + "loss": 1.0169, + "step": 35584 + }, + { + "epoch": 0.84, + "grad_norm": 1.1072865765958655, + "learning_rate": 1.3393644668371897e-06, + "loss": 0.8878, + "step": 35585 + }, + { + "epoch": 0.84, + "grad_norm": 2.083934668825873, + "learning_rate": 1.3389830219437272e-06, + "loss": 1.0408, + "step": 35586 + }, + { + "epoch": 0.84, + "grad_norm": 1.8477968886745924, + "learning_rate": 1.3386016274774482e-06, + "loss": 0.9748, + "step": 35587 + }, + { + "epoch": 0.84, + "grad_norm": 1.9663740240857175, + "learning_rate": 1.338220283440571e-06, + "loss": 0.8904, + "step": 35588 + }, + { + "epoch": 0.84, + "grad_norm": 1.9773571225052882, + "learning_rate": 1.3378389898353206e-06, + "loss": 1.0559, + "step": 35589 + }, + { + "epoch": 0.84, + "grad_norm": 1.7630086957051425, + "learning_rate": 1.3374577466639105e-06, + "loss": 0.8707, + "step": 35590 + }, + { + "epoch": 0.84, + "grad_norm": 1.065765821718264, + "learning_rate": 1.337076553928567e-06, + "loss": 0.9014, + "step": 35591 + }, + { + "epoch": 0.84, + "grad_norm": 1.0976506923300662, + "learning_rate": 1.3366954116315057e-06, + "loss": 0.9659, + "step": 35592 + }, + { + "epoch": 0.84, + "grad_norm": 1.714884237776003, + "learning_rate": 1.3363143197749451e-06, + "loss": 0.9085, + "step": 35593 + }, + { + "epoch": 0.84, + "grad_norm": 2.052563743289103, + "learning_rate": 1.3359332783611045e-06, + "loss": 0.9797, + "step": 35594 + }, + { + "epoch": 0.84, + "grad_norm": 2.163636452045199, + "learning_rate": 1.3355522873922068e-06, + "loss": 1.0015, + "step": 35595 + }, + { + "epoch": 0.84, + "grad_norm": 1.9674005580499616, + "learning_rate": 1.3351713468704663e-06, + "loss": 0.9127, + "step": 35596 + }, + { + "epoch": 0.84, + "grad_norm": 1.9939955380468466, + "learning_rate": 1.3347904567980984e-06, + "loss": 0.9518, + "step": 35597 + }, + { + "epoch": 0.84, + "grad_norm": 2.280985556371973, + "learning_rate": 1.3344096171773246e-06, + "loss": 0.9051, + "step": 35598 + }, + { + "epoch": 0.84, + "grad_norm": 2.0577341364669905, + "learning_rate": 1.3340288280103641e-06, + "loss": 0.9435, + "step": 35599 + }, + { + "epoch": 0.84, + "grad_norm": 1.9987574999338211, + "learning_rate": 1.333648089299431e-06, + "loss": 0.8672, + "step": 35600 + }, + { + "epoch": 0.84, + "grad_norm": 2.0830529287873936, + "learning_rate": 1.3332674010467395e-06, + "loss": 0.9252, + "step": 35601 + }, + { + "epoch": 0.84, + "grad_norm": 1.8916951876636705, + "learning_rate": 1.3328867632545094e-06, + "loss": 0.956, + "step": 35602 + }, + { + "epoch": 0.84, + "grad_norm": 1.952179944755934, + "learning_rate": 1.3325061759249591e-06, + "loss": 0.9229, + "step": 35603 + }, + { + "epoch": 0.84, + "grad_norm": 1.8672828474150727, + "learning_rate": 1.332125639060302e-06, + "loss": 1.0558, + "step": 35604 + }, + { + "epoch": 0.84, + "grad_norm": 1.9906281518293638, + "learning_rate": 1.3317451526627512e-06, + "loss": 1.049, + "step": 35605 + }, + { + "epoch": 0.84, + "grad_norm": 2.51007757010895, + "learning_rate": 1.3313647167345266e-06, + "loss": 0.9462, + "step": 35606 + }, + { + "epoch": 0.84, + "grad_norm": 2.284254776192182, + "learning_rate": 1.330984331277838e-06, + "loss": 1.07, + "step": 35607 + }, + { + "epoch": 0.84, + "grad_norm": 1.9395465238897154, + "learning_rate": 1.330603996294907e-06, + "loss": 0.9576, + "step": 35608 + }, + { + "epoch": 0.84, + "grad_norm": 1.896472243114054, + "learning_rate": 1.3302237117879413e-06, + "loss": 0.8123, + "step": 35609 + }, + { + "epoch": 0.84, + "grad_norm": 2.126955203006279, + "learning_rate": 1.3298434777591596e-06, + "loss": 1.0282, + "step": 35610 + }, + { + "epoch": 0.84, + "grad_norm": 2.0627528423627393, + "learning_rate": 1.3294632942107733e-06, + "loss": 0.8295, + "step": 35611 + }, + { + "epoch": 0.84, + "grad_norm": 2.4662952337769792, + "learning_rate": 1.3290831611449973e-06, + "loss": 0.9385, + "step": 35612 + }, + { + "epoch": 0.84, + "grad_norm": 1.9654639389363355, + "learning_rate": 1.3287030785640454e-06, + "loss": 0.8333, + "step": 35613 + }, + { + "epoch": 0.84, + "grad_norm": 2.1551055937913235, + "learning_rate": 1.3283230464701268e-06, + "loss": 0.9378, + "step": 35614 + }, + { + "epoch": 0.84, + "grad_norm": 1.8241030431214074, + "learning_rate": 1.3279430648654567e-06, + "loss": 1.0148, + "step": 35615 + }, + { + "epoch": 0.84, + "grad_norm": 1.8073636824079136, + "learning_rate": 1.327563133752251e-06, + "loss": 0.8854, + "step": 35616 + }, + { + "epoch": 0.84, + "grad_norm": 2.041720768163467, + "learning_rate": 1.3271832531327166e-06, + "loss": 1.0338, + "step": 35617 + }, + { + "epoch": 0.84, + "grad_norm": 2.4988353352116928, + "learning_rate": 1.3268034230090664e-06, + "loss": 1.0015, + "step": 35618 + }, + { + "epoch": 0.84, + "grad_norm": 2.101899582877972, + "learning_rate": 1.326423643383512e-06, + "loss": 0.9939, + "step": 35619 + }, + { + "epoch": 0.84, + "grad_norm": 3.2004847812519963, + "learning_rate": 1.326043914258267e-06, + "loss": 0.9552, + "step": 35620 + }, + { + "epoch": 0.84, + "grad_norm": 1.789238925344251, + "learning_rate": 1.325664235635541e-06, + "loss": 0.8409, + "step": 35621 + }, + { + "epoch": 0.84, + "grad_norm": 2.1903608099150516, + "learning_rate": 1.3252846075175407e-06, + "loss": 1.052, + "step": 35622 + }, + { + "epoch": 0.84, + "grad_norm": 2.291791242824524, + "learning_rate": 1.3249050299064815e-06, + "loss": 1.0345, + "step": 35623 + }, + { + "epoch": 0.84, + "grad_norm": 1.9368421647676632, + "learning_rate": 1.3245255028045722e-06, + "loss": 0.9154, + "step": 35624 + }, + { + "epoch": 0.84, + "grad_norm": 2.0146434292012185, + "learning_rate": 1.3241460262140237e-06, + "loss": 0.9153, + "step": 35625 + }, + { + "epoch": 0.84, + "grad_norm": 2.320344076098301, + "learning_rate": 1.323766600137041e-06, + "loss": 1.0422, + "step": 35626 + }, + { + "epoch": 0.84, + "grad_norm": 2.2111419162633594, + "learning_rate": 1.3233872245758373e-06, + "loss": 0.9883, + "step": 35627 + }, + { + "epoch": 0.84, + "grad_norm": 1.830612911209562, + "learning_rate": 1.323007899532619e-06, + "loss": 1.02, + "step": 35628 + }, + { + "epoch": 0.84, + "grad_norm": 2.111869739785118, + "learning_rate": 1.3226286250095978e-06, + "loss": 1.0037, + "step": 35629 + }, + { + "epoch": 0.84, + "grad_norm": 2.51455859023755, + "learning_rate": 1.322249401008977e-06, + "loss": 0.9028, + "step": 35630 + }, + { + "epoch": 0.84, + "grad_norm": 2.1438042229325065, + "learning_rate": 1.3218702275329708e-06, + "loss": 1.0146, + "step": 35631 + }, + { + "epoch": 0.84, + "grad_norm": 1.8464177760149576, + "learning_rate": 1.321491104583781e-06, + "loss": 1.0161, + "step": 35632 + }, + { + "epoch": 0.84, + "grad_norm": 2.099409214725375, + "learning_rate": 1.3211120321636194e-06, + "loss": 1.0844, + "step": 35633 + }, + { + "epoch": 0.84, + "grad_norm": 1.8717017993827012, + "learning_rate": 1.3207330102746907e-06, + "loss": 1.1475, + "step": 35634 + }, + { + "epoch": 0.84, + "grad_norm": 2.1446103100326304, + "learning_rate": 1.3203540389192014e-06, + "loss": 1.0062, + "step": 35635 + }, + { + "epoch": 0.84, + "grad_norm": 2.3341123945847575, + "learning_rate": 1.319975118099358e-06, + "loss": 0.9927, + "step": 35636 + }, + { + "epoch": 0.84, + "grad_norm": 2.1875151086609055, + "learning_rate": 1.3195962478173695e-06, + "loss": 1.102, + "step": 35637 + }, + { + "epoch": 0.84, + "grad_norm": 2.000242307306635, + "learning_rate": 1.3192174280754411e-06, + "loss": 1.0341, + "step": 35638 + }, + { + "epoch": 0.84, + "grad_norm": 1.2953931452917193, + "learning_rate": 1.318838658875774e-06, + "loss": 0.9925, + "step": 35639 + }, + { + "epoch": 0.84, + "grad_norm": 2.2208339502487764, + "learning_rate": 1.3184599402205766e-06, + "loss": 1.0361, + "step": 35640 + }, + { + "epoch": 0.84, + "grad_norm": 2.411160175922194, + "learning_rate": 1.318081272112056e-06, + "loss": 1.0177, + "step": 35641 + }, + { + "epoch": 0.84, + "grad_norm": 1.0783250757394456, + "learning_rate": 1.3177026545524152e-06, + "loss": 0.9678, + "step": 35642 + }, + { + "epoch": 0.84, + "grad_norm": 2.089725340068933, + "learning_rate": 1.3173240875438565e-06, + "loss": 0.8912, + "step": 35643 + }, + { + "epoch": 0.84, + "grad_norm": 2.2045750354510303, + "learning_rate": 1.3169455710885847e-06, + "loss": 0.9382, + "step": 35644 + }, + { + "epoch": 0.84, + "grad_norm": 1.9052099148759882, + "learning_rate": 1.3165671051888073e-06, + "loss": 0.8703, + "step": 35645 + }, + { + "epoch": 0.84, + "grad_norm": 2.016631257979264, + "learning_rate": 1.316188689846727e-06, + "loss": 0.9051, + "step": 35646 + }, + { + "epoch": 0.84, + "grad_norm": 1.8674725697486039, + "learning_rate": 1.3158103250645416e-06, + "loss": 0.9824, + "step": 35647 + }, + { + "epoch": 0.84, + "grad_norm": 2.0552849559897086, + "learning_rate": 1.3154320108444597e-06, + "loss": 0.9665, + "step": 35648 + }, + { + "epoch": 0.84, + "grad_norm": 1.0590744317267065, + "learning_rate": 1.3150537471886814e-06, + "loss": 0.9531, + "step": 35649 + }, + { + "epoch": 0.84, + "grad_norm": 1.9693804543446356, + "learning_rate": 1.314675534099411e-06, + "loss": 0.8991, + "step": 35650 + }, + { + "epoch": 0.84, + "grad_norm": 2.0349217860284123, + "learning_rate": 1.3142973715788477e-06, + "loss": 0.9549, + "step": 35651 + }, + { + "epoch": 0.84, + "grad_norm": 2.21816332337709, + "learning_rate": 1.313919259629196e-06, + "loss": 0.9561, + "step": 35652 + }, + { + "epoch": 0.84, + "grad_norm": 2.042756027441916, + "learning_rate": 1.3135411982526546e-06, + "loss": 1.0395, + "step": 35653 + }, + { + "epoch": 0.84, + "grad_norm": 2.879446879036068, + "learning_rate": 1.3131631874514272e-06, + "loss": 0.9077, + "step": 35654 + }, + { + "epoch": 0.84, + "grad_norm": 2.0856284526954125, + "learning_rate": 1.3127852272277187e-06, + "loss": 1.0202, + "step": 35655 + }, + { + "epoch": 0.84, + "grad_norm": 1.1306115487915278, + "learning_rate": 1.3124073175837193e-06, + "loss": 0.9368, + "step": 35656 + }, + { + "epoch": 0.84, + "grad_norm": 1.047909104232924, + "learning_rate": 1.3120294585216353e-06, + "loss": 0.9427, + "step": 35657 + }, + { + "epoch": 0.84, + "grad_norm": 2.0071748667171785, + "learning_rate": 1.311651650043666e-06, + "loss": 0.874, + "step": 35658 + }, + { + "epoch": 0.84, + "grad_norm": 1.8810842356476662, + "learning_rate": 1.3112738921520164e-06, + "loss": 0.7731, + "step": 35659 + }, + { + "epoch": 0.84, + "grad_norm": 1.9009959598603179, + "learning_rate": 1.3108961848488767e-06, + "loss": 0.9586, + "step": 35660 + }, + { + "epoch": 0.84, + "grad_norm": 1.7435394221568397, + "learning_rate": 1.3105185281364507e-06, + "loss": 0.998, + "step": 35661 + }, + { + "epoch": 0.84, + "grad_norm": 1.9403328655432543, + "learning_rate": 1.310140922016938e-06, + "loss": 0.9089, + "step": 35662 + }, + { + "epoch": 0.84, + "grad_norm": 1.9671443350664939, + "learning_rate": 1.309763366492537e-06, + "loss": 0.9161, + "step": 35663 + }, + { + "epoch": 0.84, + "grad_norm": 2.328191051010227, + "learning_rate": 1.3093858615654421e-06, + "loss": 0.9393, + "step": 35664 + }, + { + "epoch": 0.84, + "grad_norm": 1.8357806829936958, + "learning_rate": 1.309008407237854e-06, + "loss": 0.9702, + "step": 35665 + }, + { + "epoch": 0.84, + "grad_norm": 2.0738626037852805, + "learning_rate": 1.3086310035119732e-06, + "loss": 0.979, + "step": 35666 + }, + { + "epoch": 0.84, + "grad_norm": 1.9087978155522278, + "learning_rate": 1.3082536503899922e-06, + "loss": 0.9436, + "step": 35667 + }, + { + "epoch": 0.84, + "grad_norm": 1.797346857302116, + "learning_rate": 1.3078763478741118e-06, + "loss": 0.9032, + "step": 35668 + }, + { + "epoch": 0.84, + "grad_norm": 1.7660289760893684, + "learning_rate": 1.3074990959665279e-06, + "loss": 0.9524, + "step": 35669 + }, + { + "epoch": 0.84, + "grad_norm": 2.7320839468939666, + "learning_rate": 1.307121894669433e-06, + "loss": 0.887, + "step": 35670 + }, + { + "epoch": 0.84, + "grad_norm": 4.454896083939794, + "learning_rate": 1.3067447439850279e-06, + "loss": 1.0905, + "step": 35671 + }, + { + "epoch": 0.84, + "grad_norm": 1.9404569774161986, + "learning_rate": 1.306367643915508e-06, + "loss": 0.9322, + "step": 35672 + }, + { + "epoch": 0.84, + "grad_norm": 2.1298519833662874, + "learning_rate": 1.3059905944630691e-06, + "loss": 0.9475, + "step": 35673 + }, + { + "epoch": 0.84, + "grad_norm": 2.395914239420052, + "learning_rate": 1.3056135956299032e-06, + "loss": 1.0356, + "step": 35674 + }, + { + "epoch": 0.84, + "grad_norm": 1.7647836181634935, + "learning_rate": 1.3052366474182065e-06, + "loss": 0.9292, + "step": 35675 + }, + { + "epoch": 0.84, + "grad_norm": 1.8810021899818097, + "learning_rate": 1.3048597498301808e-06, + "loss": 1.0545, + "step": 35676 + }, + { + "epoch": 0.84, + "grad_norm": 1.806841603583001, + "learning_rate": 1.3044829028680084e-06, + "loss": 0.9593, + "step": 35677 + }, + { + "epoch": 0.84, + "grad_norm": 1.8414758926415693, + "learning_rate": 1.30410610653389e-06, + "loss": 0.9864, + "step": 35678 + }, + { + "epoch": 0.84, + "grad_norm": 2.1716060141674802, + "learning_rate": 1.3037293608300195e-06, + "loss": 0.9662, + "step": 35679 + }, + { + "epoch": 0.84, + "grad_norm": 1.8535883029289588, + "learning_rate": 1.3033526657585915e-06, + "loss": 1.0865, + "step": 35680 + }, + { + "epoch": 0.84, + "grad_norm": 1.8191426003457474, + "learning_rate": 1.302976021321798e-06, + "loss": 0.9342, + "step": 35681 + }, + { + "epoch": 0.84, + "grad_norm": 2.044222754687055, + "learning_rate": 1.302599427521829e-06, + "loss": 0.9936, + "step": 35682 + }, + { + "epoch": 0.84, + "grad_norm": 2.2307585663245697, + "learning_rate": 1.302222884360882e-06, + "loss": 0.976, + "step": 35683 + }, + { + "epoch": 0.84, + "grad_norm": 1.9616966732957075, + "learning_rate": 1.301846391841145e-06, + "loss": 1.0537, + "step": 35684 + }, + { + "epoch": 0.84, + "grad_norm": 2.328296503048158, + "learning_rate": 1.3014699499648142e-06, + "loss": 0.854, + "step": 35685 + }, + { + "epoch": 0.84, + "grad_norm": 2.2242679128750646, + "learning_rate": 1.3010935587340767e-06, + "loss": 1.0005, + "step": 35686 + }, + { + "epoch": 0.84, + "grad_norm": 2.0232814653632754, + "learning_rate": 1.3007172181511296e-06, + "loss": 0.9277, + "step": 35687 + }, + { + "epoch": 0.84, + "grad_norm": 1.991893870088064, + "learning_rate": 1.3003409282181579e-06, + "loss": 1.1076, + "step": 35688 + }, + { + "epoch": 0.84, + "grad_norm": 1.0886723690490367, + "learning_rate": 1.2999646889373585e-06, + "loss": 0.9829, + "step": 35689 + }, + { + "epoch": 0.84, + "grad_norm": 2.247578235874374, + "learning_rate": 1.2995885003109166e-06, + "loss": 0.9629, + "step": 35690 + }, + { + "epoch": 0.84, + "grad_norm": 1.7991538913191853, + "learning_rate": 1.2992123623410268e-06, + "loss": 0.9442, + "step": 35691 + }, + { + "epoch": 0.84, + "grad_norm": 1.0866395916730915, + "learning_rate": 1.2988362750298754e-06, + "loss": 0.9401, + "step": 35692 + }, + { + "epoch": 0.84, + "grad_norm": 1.927760291251445, + "learning_rate": 1.2984602383796551e-06, + "loss": 1.0119, + "step": 35693 + }, + { + "epoch": 0.84, + "grad_norm": 2.033829567098438, + "learning_rate": 1.2980842523925553e-06, + "loss": 0.8189, + "step": 35694 + }, + { + "epoch": 0.84, + "grad_norm": 1.7590733851807623, + "learning_rate": 1.2977083170707616e-06, + "loss": 0.9342, + "step": 35695 + }, + { + "epoch": 0.84, + "grad_norm": 2.022465641570227, + "learning_rate": 1.2973324324164649e-06, + "loss": 0.9868, + "step": 35696 + }, + { + "epoch": 0.84, + "grad_norm": 2.3577148694730963, + "learning_rate": 1.2969565984318567e-06, + "loss": 0.9323, + "step": 35697 + }, + { + "epoch": 0.84, + "grad_norm": 2.159433146187629, + "learning_rate": 1.2965808151191229e-06, + "loss": 0.9926, + "step": 35698 + }, + { + "epoch": 0.84, + "grad_norm": 1.9652335039166056, + "learning_rate": 1.2962050824804472e-06, + "loss": 1.116, + "step": 35699 + }, + { + "epoch": 0.84, + "grad_norm": 2.0116902084316575, + "learning_rate": 1.2958294005180228e-06, + "loss": 1.0396, + "step": 35700 + }, + { + "epoch": 0.84, + "grad_norm": 1.9539592935148844, + "learning_rate": 1.2954537692340363e-06, + "loss": 0.9558, + "step": 35701 + }, + { + "epoch": 0.84, + "grad_norm": 2.8846003321858027, + "learning_rate": 1.295078188630674e-06, + "loss": 0.8452, + "step": 35702 + }, + { + "epoch": 0.84, + "grad_norm": 2.4420841543307015, + "learning_rate": 1.2947026587101208e-06, + "loss": 1.0905, + "step": 35703 + }, + { + "epoch": 0.84, + "grad_norm": 2.393655826395964, + "learning_rate": 1.2943271794745649e-06, + "loss": 0.8872, + "step": 35704 + }, + { + "epoch": 0.84, + "grad_norm": 2.3560513972620174, + "learning_rate": 1.2939517509261934e-06, + "loss": 0.9492, + "step": 35705 + }, + { + "epoch": 0.84, + "grad_norm": 2.0621061530406903, + "learning_rate": 1.2935763730671923e-06, + "loss": 1.0065, + "step": 35706 + }, + { + "epoch": 0.84, + "grad_norm": 1.883371213986007, + "learning_rate": 1.2932010458997434e-06, + "loss": 0.9784, + "step": 35707 + }, + { + "epoch": 0.84, + "grad_norm": 2.102598565808189, + "learning_rate": 1.2928257694260359e-06, + "loss": 0.9221, + "step": 35708 + }, + { + "epoch": 0.84, + "grad_norm": 1.8598479472855456, + "learning_rate": 1.2924505436482527e-06, + "loss": 1.0157, + "step": 35709 + }, + { + "epoch": 0.84, + "grad_norm": 1.87384567428965, + "learning_rate": 1.2920753685685805e-06, + "loss": 1.1185, + "step": 35710 + }, + { + "epoch": 0.84, + "grad_norm": 1.9673887760702893, + "learning_rate": 1.2917002441892002e-06, + "loss": 0.9837, + "step": 35711 + }, + { + "epoch": 0.84, + "grad_norm": 2.0245186754082964, + "learning_rate": 1.2913251705122998e-06, + "loss": 1.1187, + "step": 35712 + }, + { + "epoch": 0.84, + "grad_norm": 1.9820066287259794, + "learning_rate": 1.2909501475400598e-06, + "loss": 0.9669, + "step": 35713 + }, + { + "epoch": 0.84, + "grad_norm": 2.101178050772333, + "learning_rate": 1.2905751752746676e-06, + "loss": 1.0441, + "step": 35714 + }, + { + "epoch": 0.84, + "grad_norm": 1.913043053987598, + "learning_rate": 1.2902002537183044e-06, + "loss": 1.0075, + "step": 35715 + }, + { + "epoch": 0.84, + "grad_norm": 1.943870656170397, + "learning_rate": 1.28982538287315e-06, + "loss": 0.8884, + "step": 35716 + }, + { + "epoch": 0.84, + "grad_norm": 1.9313255791782173, + "learning_rate": 1.28945056274139e-06, + "loss": 0.949, + "step": 35717 + }, + { + "epoch": 0.84, + "grad_norm": 2.0421084805754095, + "learning_rate": 1.2890757933252084e-06, + "loss": 0.9068, + "step": 35718 + }, + { + "epoch": 0.84, + "grad_norm": 2.367505390355649, + "learning_rate": 1.2887010746267859e-06, + "loss": 1.0242, + "step": 35719 + }, + { + "epoch": 0.84, + "grad_norm": 1.019014431005944, + "learning_rate": 1.2883264066483015e-06, + "loss": 0.9396, + "step": 35720 + }, + { + "epoch": 0.84, + "grad_norm": 1.9761148544725728, + "learning_rate": 1.287951789391938e-06, + "loss": 0.919, + "step": 35721 + }, + { + "epoch": 0.84, + "grad_norm": 1.8701347234466528, + "learning_rate": 1.2875772228598805e-06, + "loss": 0.959, + "step": 35722 + }, + { + "epoch": 0.84, + "grad_norm": 1.8786534485793926, + "learning_rate": 1.2872027070543057e-06, + "loss": 1.0841, + "step": 35723 + }, + { + "epoch": 0.84, + "grad_norm": 2.1510491543007197, + "learning_rate": 1.2868282419773936e-06, + "loss": 1.0427, + "step": 35724 + }, + { + "epoch": 0.84, + "grad_norm": 2.3883976824781565, + "learning_rate": 1.2864538276313265e-06, + "loss": 1.005, + "step": 35725 + }, + { + "epoch": 0.84, + "grad_norm": 2.060016385035041, + "learning_rate": 1.286079464018285e-06, + "loss": 0.9839, + "step": 35726 + }, + { + "epoch": 0.84, + "grad_norm": 2.2380119697671077, + "learning_rate": 1.2857051511404484e-06, + "loss": 0.9417, + "step": 35727 + }, + { + "epoch": 0.84, + "grad_norm": 1.9461535203471387, + "learning_rate": 1.2853308889999927e-06, + "loss": 0.8018, + "step": 35728 + }, + { + "epoch": 0.84, + "grad_norm": 1.997961678278007, + "learning_rate": 1.2849566775991029e-06, + "loss": 0.953, + "step": 35729 + }, + { + "epoch": 0.84, + "grad_norm": 2.086182568518096, + "learning_rate": 1.2845825169399506e-06, + "loss": 0.9947, + "step": 35730 + }, + { + "epoch": 0.84, + "grad_norm": 1.8653426409177443, + "learning_rate": 1.2842084070247217e-06, + "loss": 0.8425, + "step": 35731 + }, + { + "epoch": 0.84, + "grad_norm": 2.4408880512558317, + "learning_rate": 1.283834347855588e-06, + "loss": 1.0509, + "step": 35732 + }, + { + "epoch": 0.84, + "grad_norm": 2.066151829206821, + "learning_rate": 1.2834603394347323e-06, + "loss": 1.0119, + "step": 35733 + }, + { + "epoch": 0.84, + "grad_norm": 1.1143478951411165, + "learning_rate": 1.2830863817643279e-06, + "loss": 0.9715, + "step": 35734 + }, + { + "epoch": 0.84, + "grad_norm": 2.300051003728303, + "learning_rate": 1.2827124748465569e-06, + "loss": 0.9097, + "step": 35735 + }, + { + "epoch": 0.84, + "grad_norm": 2.052611771559951, + "learning_rate": 1.282338618683594e-06, + "loss": 1.0623, + "step": 35736 + }, + { + "epoch": 0.84, + "grad_norm": 2.3015779942703474, + "learning_rate": 1.281964813277612e-06, + "loss": 0.9884, + "step": 35737 + }, + { + "epoch": 0.84, + "grad_norm": 1.9112072865804874, + "learning_rate": 1.2815910586307923e-06, + "loss": 0.9415, + "step": 35738 + }, + { + "epoch": 0.84, + "grad_norm": 2.071083613282388, + "learning_rate": 1.281217354745311e-06, + "loss": 1.0038, + "step": 35739 + }, + { + "epoch": 0.84, + "grad_norm": 1.8747279605610288, + "learning_rate": 1.2808437016233433e-06, + "loss": 0.9888, + "step": 35740 + }, + { + "epoch": 0.84, + "grad_norm": 6.8036076657788485, + "learning_rate": 1.2804700992670616e-06, + "loss": 0.9768, + "step": 35741 + }, + { + "epoch": 0.84, + "grad_norm": 1.9829113620218022, + "learning_rate": 1.2800965476786431e-06, + "loss": 0.9383, + "step": 35742 + }, + { + "epoch": 0.84, + "grad_norm": 2.2880008449577467, + "learning_rate": 1.279723046860266e-06, + "loss": 0.9389, + "step": 35743 + }, + { + "epoch": 0.84, + "grad_norm": 1.9320038032423579, + "learning_rate": 1.2793495968141023e-06, + "loss": 0.9269, + "step": 35744 + }, + { + "epoch": 0.84, + "grad_norm": 2.1155256957831465, + "learning_rate": 1.2789761975423231e-06, + "loss": 1.0342, + "step": 35745 + }, + { + "epoch": 0.84, + "grad_norm": 1.9263790955828617, + "learning_rate": 1.2786028490471058e-06, + "loss": 0.8693, + "step": 35746 + }, + { + "epoch": 0.84, + "grad_norm": 2.4997366677583357, + "learning_rate": 1.2782295513306264e-06, + "loss": 0.8806, + "step": 35747 + }, + { + "epoch": 0.84, + "grad_norm": 1.914242140091134, + "learning_rate": 1.2778563043950553e-06, + "loss": 0.9677, + "step": 35748 + }, + { + "epoch": 0.84, + "grad_norm": 2.124323229186265, + "learning_rate": 1.2774831082425644e-06, + "loss": 1.058, + "step": 35749 + }, + { + "epoch": 0.84, + "grad_norm": 1.9077556062399816, + "learning_rate": 1.2771099628753304e-06, + "loss": 0.8943, + "step": 35750 + }, + { + "epoch": 0.84, + "grad_norm": 1.8726511127936982, + "learning_rate": 1.2767368682955216e-06, + "loss": 0.8407, + "step": 35751 + }, + { + "epoch": 0.84, + "grad_norm": 2.318868407459631, + "learning_rate": 1.2763638245053134e-06, + "loss": 0.8415, + "step": 35752 + }, + { + "epoch": 0.84, + "grad_norm": 1.9203389139052365, + "learning_rate": 1.275990831506878e-06, + "loss": 0.9131, + "step": 35753 + }, + { + "epoch": 0.84, + "grad_norm": 2.1003410332037737, + "learning_rate": 1.2756178893023862e-06, + "loss": 0.9809, + "step": 35754 + }, + { + "epoch": 0.84, + "grad_norm": 3.892268953807191, + "learning_rate": 1.2752449978940062e-06, + "loss": 1.0916, + "step": 35755 + }, + { + "epoch": 0.84, + "grad_norm": 2.0958375539931304, + "learning_rate": 1.274872157283913e-06, + "loss": 1.0249, + "step": 35756 + }, + { + "epoch": 0.84, + "grad_norm": 2.0180720790235975, + "learning_rate": 1.2744993674742812e-06, + "loss": 0.9951, + "step": 35757 + }, + { + "epoch": 0.84, + "grad_norm": 2.06302180708781, + "learning_rate": 1.2741266284672704e-06, + "loss": 0.987, + "step": 35758 + }, + { + "epoch": 0.84, + "grad_norm": 2.713760654943271, + "learning_rate": 1.273753940265059e-06, + "loss": 0.9226, + "step": 35759 + }, + { + "epoch": 0.84, + "grad_norm": 2.2238613884649503, + "learning_rate": 1.2733813028698138e-06, + "loss": 1.1542, + "step": 35760 + }, + { + "epoch": 0.84, + "grad_norm": 1.97901232835515, + "learning_rate": 1.273008716283709e-06, + "loss": 0.851, + "step": 35761 + }, + { + "epoch": 0.84, + "grad_norm": 1.9545140411340358, + "learning_rate": 1.2726361805089072e-06, + "loss": 0.8866, + "step": 35762 + }, + { + "epoch": 0.84, + "grad_norm": 2.4463871229051954, + "learning_rate": 1.272263695547581e-06, + "loss": 0.9393, + "step": 35763 + }, + { + "epoch": 0.84, + "grad_norm": 2.0038462297800046, + "learning_rate": 1.2718912614018996e-06, + "loss": 1.0692, + "step": 35764 + }, + { + "epoch": 0.84, + "grad_norm": 1.7182209766714445, + "learning_rate": 1.2715188780740295e-06, + "loss": 1.112, + "step": 35765 + }, + { + "epoch": 0.84, + "grad_norm": 3.0384936388681876, + "learning_rate": 1.271146545566141e-06, + "loss": 0.891, + "step": 35766 + }, + { + "epoch": 0.84, + "grad_norm": 2.1426303422162603, + "learning_rate": 1.2707742638803987e-06, + "loss": 0.9174, + "step": 35767 + }, + { + "epoch": 0.84, + "grad_norm": 2.1089826215054726, + "learning_rate": 1.2704020330189759e-06, + "loss": 1.0078, + "step": 35768 + }, + { + "epoch": 0.84, + "grad_norm": 2.105519327472287, + "learning_rate": 1.2700298529840327e-06, + "loss": 1.1151, + "step": 35769 + }, + { + "epoch": 0.84, + "grad_norm": 4.3849501996284586, + "learning_rate": 1.2696577237777418e-06, + "loss": 1.1001, + "step": 35770 + }, + { + "epoch": 0.84, + "grad_norm": 2.024090184343681, + "learning_rate": 1.2692856454022684e-06, + "loss": 1.002, + "step": 35771 + }, + { + "epoch": 0.84, + "grad_norm": 1.856332071884989, + "learning_rate": 1.268913617859775e-06, + "loss": 0.969, + "step": 35772 + }, + { + "epoch": 0.84, + "grad_norm": 2.2414837389660187, + "learning_rate": 1.268541641152431e-06, + "loss": 0.9853, + "step": 35773 + }, + { + "epoch": 0.84, + "grad_norm": 1.8852868868895416, + "learning_rate": 1.2681697152824045e-06, + "loss": 0.9018, + "step": 35774 + }, + { + "epoch": 0.84, + "grad_norm": 1.8225513560179958, + "learning_rate": 1.2677978402518576e-06, + "loss": 0.888, + "step": 35775 + }, + { + "epoch": 0.84, + "grad_norm": 1.9802205101349826, + "learning_rate": 1.2674260160629547e-06, + "loss": 1.0341, + "step": 35776 + }, + { + "epoch": 0.84, + "grad_norm": 2.034200218177656, + "learning_rate": 1.2670542427178623e-06, + "loss": 0.9443, + "step": 35777 + }, + { + "epoch": 0.84, + "grad_norm": 2.011876040739972, + "learning_rate": 1.266682520218747e-06, + "loss": 0.9649, + "step": 35778 + }, + { + "epoch": 0.84, + "grad_norm": 2.3975543948873987, + "learning_rate": 1.266310848567771e-06, + "loss": 1.0093, + "step": 35779 + }, + { + "epoch": 0.84, + "grad_norm": 1.8407335720065443, + "learning_rate": 1.2659392277670956e-06, + "loss": 0.9543, + "step": 35780 + }, + { + "epoch": 0.84, + "grad_norm": 2.0121597010473544, + "learning_rate": 1.265567657818888e-06, + "loss": 0.9717, + "step": 35781 + }, + { + "epoch": 0.84, + "grad_norm": 1.6977951913843623, + "learning_rate": 1.265196138725312e-06, + "loss": 0.9618, + "step": 35782 + }, + { + "epoch": 0.84, + "grad_norm": 1.7135450865401376, + "learning_rate": 1.2648246704885303e-06, + "loss": 0.8686, + "step": 35783 + }, + { + "epoch": 0.84, + "grad_norm": 1.7454144565405723, + "learning_rate": 1.2644532531107024e-06, + "loss": 0.9163, + "step": 35784 + }, + { + "epoch": 0.84, + "grad_norm": 1.8964827832913618, + "learning_rate": 1.2640818865939952e-06, + "loss": 0.9551, + "step": 35785 + }, + { + "epoch": 0.84, + "grad_norm": 1.9159016931366428, + "learning_rate": 1.2637105709405662e-06, + "loss": 0.8433, + "step": 35786 + }, + { + "epoch": 0.84, + "grad_norm": 2.6337709138950047, + "learning_rate": 1.2633393061525834e-06, + "loss": 0.8642, + "step": 35787 + }, + { + "epoch": 0.84, + "grad_norm": 1.91825877689819, + "learning_rate": 1.262968092232203e-06, + "loss": 1.0643, + "step": 35788 + }, + { + "epoch": 0.84, + "grad_norm": 1.8983805710690764, + "learning_rate": 1.2625969291815899e-06, + "loss": 0.9425, + "step": 35789 + }, + { + "epoch": 0.84, + "grad_norm": 1.878326848189681, + "learning_rate": 1.2622258170029023e-06, + "loss": 1.0597, + "step": 35790 + }, + { + "epoch": 0.84, + "grad_norm": 1.91543810766866, + "learning_rate": 1.2618547556983042e-06, + "loss": 1.0378, + "step": 35791 + }, + { + "epoch": 0.84, + "grad_norm": 2.2832956065912247, + "learning_rate": 1.2614837452699536e-06, + "loss": 0.9853, + "step": 35792 + }, + { + "epoch": 0.84, + "grad_norm": 2.066296859342179, + "learning_rate": 1.26111278572001e-06, + "loss": 0.9279, + "step": 35793 + }, + { + "epoch": 0.84, + "grad_norm": 1.1383879922748181, + "learning_rate": 1.2607418770506342e-06, + "loss": 0.9518, + "step": 35794 + }, + { + "epoch": 0.84, + "grad_norm": 2.018849153107091, + "learning_rate": 1.2603710192639874e-06, + "loss": 1.0427, + "step": 35795 + }, + { + "epoch": 0.84, + "grad_norm": 2.1193774404136607, + "learning_rate": 1.2600002123622279e-06, + "loss": 1.123, + "step": 35796 + }, + { + "epoch": 0.84, + "grad_norm": 2.0290293023779804, + "learning_rate": 1.259629456347512e-06, + "loss": 1.0463, + "step": 35797 + }, + { + "epoch": 0.84, + "grad_norm": 2.4569787801058554, + "learning_rate": 1.2592587512220011e-06, + "loss": 0.923, + "step": 35798 + }, + { + "epoch": 0.84, + "grad_norm": 1.7423841976550032, + "learning_rate": 1.2588880969878547e-06, + "loss": 0.8808, + "step": 35799 + }, + { + "epoch": 0.84, + "grad_norm": 2.0314992914454297, + "learning_rate": 1.2585174936472288e-06, + "loss": 1.0099, + "step": 35800 + }, + { + "epoch": 0.84, + "grad_norm": 2.1614263250011336, + "learning_rate": 1.2581469412022806e-06, + "loss": 1.0409, + "step": 35801 + }, + { + "epoch": 0.84, + "grad_norm": 1.8698453485260595, + "learning_rate": 1.2577764396551673e-06, + "loss": 1.0242, + "step": 35802 + }, + { + "epoch": 0.84, + "grad_norm": 1.0939730767978553, + "learning_rate": 1.2574059890080502e-06, + "loss": 0.921, + "step": 35803 + }, + { + "epoch": 0.84, + "grad_norm": 3.9573251615031824, + "learning_rate": 1.2570355892630825e-06, + "loss": 1.1055, + "step": 35804 + }, + { + "epoch": 0.84, + "grad_norm": 1.9217701491775576, + "learning_rate": 1.25666524042242e-06, + "loss": 1.0281, + "step": 35805 + }, + { + "epoch": 0.84, + "grad_norm": 1.8816672242329497, + "learning_rate": 1.2562949424882232e-06, + "loss": 0.8892, + "step": 35806 + }, + { + "epoch": 0.84, + "grad_norm": 2.574531290412723, + "learning_rate": 1.2559246954626424e-06, + "loss": 1.0035, + "step": 35807 + }, + { + "epoch": 0.84, + "grad_norm": 3.3502081725758233, + "learning_rate": 1.2555544993478386e-06, + "loss": 0.8882, + "step": 35808 + }, + { + "epoch": 0.84, + "grad_norm": 1.8343956976621318, + "learning_rate": 1.255184354145963e-06, + "loss": 0.9225, + "step": 35809 + }, + { + "epoch": 0.84, + "grad_norm": 1.8367935622287088, + "learning_rate": 1.254814259859175e-06, + "loss": 1.0508, + "step": 35810 + }, + { + "epoch": 0.84, + "grad_norm": 1.9632269556344812, + "learning_rate": 1.2544442164896254e-06, + "loss": 1.0093, + "step": 35811 + }, + { + "epoch": 0.84, + "grad_norm": 1.9958943542807701, + "learning_rate": 1.254074224039471e-06, + "loss": 1.0086, + "step": 35812 + }, + { + "epoch": 0.84, + "grad_norm": 2.148435335266583, + "learning_rate": 1.2537042825108647e-06, + "loss": 0.9424, + "step": 35813 + }, + { + "epoch": 0.84, + "grad_norm": 1.7871526670553881, + "learning_rate": 1.2533343919059626e-06, + "loss": 0.9238, + "step": 35814 + }, + { + "epoch": 0.84, + "grad_norm": 2.1765204256758466, + "learning_rate": 1.2529645522269151e-06, + "loss": 1.0374, + "step": 35815 + }, + { + "epoch": 0.84, + "grad_norm": 2.2272574335765394, + "learning_rate": 1.2525947634758784e-06, + "loss": 0.941, + "step": 35816 + }, + { + "epoch": 0.84, + "grad_norm": 1.815622087638804, + "learning_rate": 1.252225025655005e-06, + "loss": 0.8763, + "step": 35817 + }, + { + "epoch": 0.84, + "grad_norm": 1.8684518810975423, + "learning_rate": 1.2518553387664456e-06, + "loss": 0.9105, + "step": 35818 + }, + { + "epoch": 0.84, + "grad_norm": 1.8494889365267018, + "learning_rate": 1.251485702812353e-06, + "loss": 0.9354, + "step": 35819 + }, + { + "epoch": 0.84, + "grad_norm": 2.5620338013862183, + "learning_rate": 1.2511161177948827e-06, + "loss": 0.8816, + "step": 35820 + }, + { + "epoch": 0.84, + "grad_norm": 2.204959781210432, + "learning_rate": 1.2507465837161848e-06, + "loss": 1.027, + "step": 35821 + }, + { + "epoch": 0.84, + "grad_norm": 2.064079390574203, + "learning_rate": 1.2503771005784071e-06, + "loss": 0.966, + "step": 35822 + }, + { + "epoch": 0.84, + "grad_norm": 1.8817245496705877, + "learning_rate": 1.2500076683837048e-06, + "loss": 1.0343, + "step": 35823 + }, + { + "epoch": 0.84, + "grad_norm": 1.9567128520161754, + "learning_rate": 1.2496382871342295e-06, + "loss": 1.0209, + "step": 35824 + }, + { + "epoch": 0.84, + "grad_norm": 1.9533232954214377, + "learning_rate": 1.2492689568321314e-06, + "loss": 0.9137, + "step": 35825 + }, + { + "epoch": 0.84, + "grad_norm": 1.0770549869809767, + "learning_rate": 1.248899677479557e-06, + "loss": 0.9451, + "step": 35826 + }, + { + "epoch": 0.84, + "grad_norm": 2.6558369213663804, + "learning_rate": 1.2485304490786587e-06, + "loss": 0.9097, + "step": 35827 + }, + { + "epoch": 0.84, + "grad_norm": 2.0605426438003414, + "learning_rate": 1.2481612716315905e-06, + "loss": 1.0391, + "step": 35828 + }, + { + "epoch": 0.84, + "grad_norm": 2.027666603802021, + "learning_rate": 1.2477921451404973e-06, + "loss": 1.068, + "step": 35829 + }, + { + "epoch": 0.84, + "grad_norm": 2.010424130834899, + "learning_rate": 1.2474230696075262e-06, + "loss": 1.0593, + "step": 35830 + }, + { + "epoch": 0.84, + "grad_norm": 2.0011827058757774, + "learning_rate": 1.2470540450348323e-06, + "loss": 0.9013, + "step": 35831 + }, + { + "epoch": 0.84, + "grad_norm": 1.944421164346604, + "learning_rate": 1.2466850714245581e-06, + "loss": 0.9093, + "step": 35832 + }, + { + "epoch": 0.84, + "grad_norm": 1.9307354727294361, + "learning_rate": 1.2463161487788578e-06, + "loss": 0.9723, + "step": 35833 + }, + { + "epoch": 0.84, + "grad_norm": 2.0891082532837104, + "learning_rate": 1.2459472770998727e-06, + "loss": 0.9211, + "step": 35834 + }, + { + "epoch": 0.84, + "grad_norm": 1.924292079052577, + "learning_rate": 1.2455784563897577e-06, + "loss": 0.9963, + "step": 35835 + }, + { + "epoch": 0.84, + "grad_norm": 1.7895760086852293, + "learning_rate": 1.2452096866506536e-06, + "loss": 1.0258, + "step": 35836 + }, + { + "epoch": 0.84, + "grad_norm": 2.405550865125771, + "learning_rate": 1.2448409678847117e-06, + "loss": 0.9737, + "step": 35837 + }, + { + "epoch": 0.84, + "grad_norm": 2.2025912567043946, + "learning_rate": 1.2444723000940785e-06, + "loss": 0.8077, + "step": 35838 + }, + { + "epoch": 0.84, + "grad_norm": 1.934737464504222, + "learning_rate": 1.2441036832808973e-06, + "loss": 0.9119, + "step": 35839 + }, + { + "epoch": 0.84, + "grad_norm": 2.1072251041971963, + "learning_rate": 1.2437351174473167e-06, + "loss": 1.0527, + "step": 35840 + }, + { + "epoch": 0.84, + "grad_norm": 2.316210206751664, + "learning_rate": 1.2433666025954815e-06, + "loss": 0.9106, + "step": 35841 + }, + { + "epoch": 0.84, + "grad_norm": 2.125336713927449, + "learning_rate": 1.2429981387275435e-06, + "loss": 0.9466, + "step": 35842 + }, + { + "epoch": 0.84, + "grad_norm": 2.1646557704367995, + "learning_rate": 1.2426297258456388e-06, + "loss": 0.9387, + "step": 35843 + }, + { + "epoch": 0.84, + "grad_norm": 1.8726378323377784, + "learning_rate": 1.2422613639519154e-06, + "loss": 0.9256, + "step": 35844 + }, + { + "epoch": 0.84, + "grad_norm": 1.9398675355905797, + "learning_rate": 1.2418930530485219e-06, + "loss": 0.8825, + "step": 35845 + }, + { + "epoch": 0.84, + "grad_norm": 1.8049133816619407, + "learning_rate": 1.2415247931375996e-06, + "loss": 0.9453, + "step": 35846 + }, + { + "epoch": 0.84, + "grad_norm": 1.1270021951857865, + "learning_rate": 1.2411565842212902e-06, + "loss": 0.9373, + "step": 35847 + }, + { + "epoch": 0.84, + "grad_norm": 2.34907600618845, + "learning_rate": 1.24078842630174e-06, + "loss": 1.0925, + "step": 35848 + }, + { + "epoch": 0.84, + "grad_norm": 2.068481204656754, + "learning_rate": 1.240420319381096e-06, + "loss": 1.0972, + "step": 35849 + }, + { + "epoch": 0.84, + "grad_norm": 2.010308591211432, + "learning_rate": 1.2400522634614953e-06, + "loss": 0.9527, + "step": 35850 + }, + { + "epoch": 0.84, + "grad_norm": 1.06757860098213, + "learning_rate": 1.2396842585450852e-06, + "loss": 0.9593, + "step": 35851 + }, + { + "epoch": 0.84, + "grad_norm": 2.106955508364578, + "learning_rate": 1.2393163046340073e-06, + "loss": 0.8458, + "step": 35852 + }, + { + "epoch": 0.84, + "grad_norm": 1.9354179691974733, + "learning_rate": 1.238948401730402e-06, + "loss": 0.9665, + "step": 35853 + }, + { + "epoch": 0.84, + "grad_norm": 1.7642971430267826, + "learning_rate": 1.2385805498364124e-06, + "loss": 1.0255, + "step": 35854 + }, + { + "epoch": 0.84, + "grad_norm": 1.9783315534697887, + "learning_rate": 1.238212748954183e-06, + "loss": 0.9895, + "step": 35855 + }, + { + "epoch": 0.84, + "grad_norm": 1.0956970585903318, + "learning_rate": 1.2378449990858522e-06, + "loss": 0.9449, + "step": 35856 + }, + { + "epoch": 0.84, + "grad_norm": 1.9331775953178028, + "learning_rate": 1.2374773002335605e-06, + "loss": 0.882, + "step": 35857 + }, + { + "epoch": 0.84, + "grad_norm": 2.0260455318032937, + "learning_rate": 1.2371096523994498e-06, + "loss": 1.0112, + "step": 35858 + }, + { + "epoch": 0.84, + "grad_norm": 1.9392593164862675, + "learning_rate": 1.2367420555856646e-06, + "loss": 0.9173, + "step": 35859 + }, + { + "epoch": 0.84, + "grad_norm": 1.9247266452631275, + "learning_rate": 1.236374509794338e-06, + "loss": 0.872, + "step": 35860 + }, + { + "epoch": 0.84, + "grad_norm": 1.9587598204985608, + "learning_rate": 1.2360070150276137e-06, + "loss": 0.8634, + "step": 35861 + }, + { + "epoch": 0.84, + "grad_norm": 2.4146701998730062, + "learning_rate": 1.2356395712876312e-06, + "loss": 0.9254, + "step": 35862 + }, + { + "epoch": 0.84, + "grad_norm": 1.5603657448463462, + "learning_rate": 1.235272178576531e-06, + "loss": 0.9016, + "step": 35863 + }, + { + "epoch": 0.84, + "grad_norm": 1.9374051827929522, + "learning_rate": 1.2349048368964522e-06, + "loss": 1.038, + "step": 35864 + }, + { + "epoch": 0.84, + "grad_norm": 1.760125737676064, + "learning_rate": 1.234537546249529e-06, + "loss": 1.0891, + "step": 35865 + }, + { + "epoch": 0.84, + "grad_norm": 1.9090492163709358, + "learning_rate": 1.2341703066379073e-06, + "loss": 0.8718, + "step": 35866 + }, + { + "epoch": 0.84, + "grad_norm": 2.0766486053898987, + "learning_rate": 1.2338031180637179e-06, + "loss": 0.9455, + "step": 35867 + }, + { + "epoch": 0.85, + "grad_norm": 2.003751132261485, + "learning_rate": 1.233435980529104e-06, + "loss": 0.9984, + "step": 35868 + }, + { + "epoch": 0.85, + "grad_norm": 1.9482106234926317, + "learning_rate": 1.2330688940362002e-06, + "loss": 1.0359, + "step": 35869 + }, + { + "epoch": 0.85, + "grad_norm": 1.771165643247944, + "learning_rate": 1.2327018585871463e-06, + "loss": 0.9944, + "step": 35870 + }, + { + "epoch": 0.85, + "grad_norm": 2.4123009322818785, + "learning_rate": 1.2323348741840756e-06, + "loss": 0.9301, + "step": 35871 + }, + { + "epoch": 0.85, + "grad_norm": 1.985068049548841, + "learning_rate": 1.2319679408291296e-06, + "loss": 0.8852, + "step": 35872 + }, + { + "epoch": 0.85, + "grad_norm": 2.0136790942600094, + "learning_rate": 1.231601058524442e-06, + "loss": 0.8885, + "step": 35873 + }, + { + "epoch": 0.85, + "grad_norm": 2.094000521293544, + "learning_rate": 1.2312342272721467e-06, + "loss": 0.8973, + "step": 35874 + }, + { + "epoch": 0.85, + "grad_norm": 1.902883405752975, + "learning_rate": 1.2308674470743819e-06, + "loss": 1.0235, + "step": 35875 + }, + { + "epoch": 0.85, + "grad_norm": 1.9625340170416838, + "learning_rate": 1.2305007179332851e-06, + "loss": 1.0044, + "step": 35876 + }, + { + "epoch": 0.85, + "grad_norm": 1.854358787318695, + "learning_rate": 1.2301340398509899e-06, + "loss": 1.0601, + "step": 35877 + }, + { + "epoch": 0.85, + "grad_norm": 1.9613181310313403, + "learning_rate": 1.2297674128296288e-06, + "loss": 0.9025, + "step": 35878 + }, + { + "epoch": 0.85, + "grad_norm": 2.234277877226487, + "learning_rate": 1.2294008368713385e-06, + "loss": 0.8575, + "step": 35879 + }, + { + "epoch": 0.85, + "grad_norm": 1.0946019019382016, + "learning_rate": 1.2290343119782545e-06, + "loss": 0.9594, + "step": 35880 + }, + { + "epoch": 0.85, + "grad_norm": 2.006550443603529, + "learning_rate": 1.2286678381525108e-06, + "loss": 0.8867, + "step": 35881 + }, + { + "epoch": 0.85, + "grad_norm": 2.017370867446841, + "learning_rate": 1.2283014153962368e-06, + "loss": 0.9252, + "step": 35882 + }, + { + "epoch": 0.85, + "grad_norm": 2.853045804788095, + "learning_rate": 1.227935043711569e-06, + "loss": 0.9811, + "step": 35883 + }, + { + "epoch": 0.85, + "grad_norm": 1.737586069666046, + "learning_rate": 1.2275687231006428e-06, + "loss": 0.9192, + "step": 35884 + }, + { + "epoch": 0.85, + "grad_norm": 1.9428983027577493, + "learning_rate": 1.227202453565589e-06, + "loss": 1.0196, + "step": 35885 + }, + { + "epoch": 0.85, + "grad_norm": 2.0173711849477334, + "learning_rate": 1.2268362351085383e-06, + "loss": 1.0388, + "step": 35886 + }, + { + "epoch": 0.85, + "grad_norm": 1.942435024148539, + "learning_rate": 1.2264700677316266e-06, + "loss": 0.9322, + "step": 35887 + }, + { + "epoch": 0.85, + "grad_norm": 2.0773698550205553, + "learning_rate": 1.226103951436981e-06, + "loss": 1.0853, + "step": 35888 + }, + { + "epoch": 0.85, + "grad_norm": 1.7594104358378417, + "learning_rate": 1.2257378862267388e-06, + "loss": 0.9102, + "step": 35889 + }, + { + "epoch": 0.85, + "grad_norm": 3.782106107120399, + "learning_rate": 1.225371872103026e-06, + "loss": 1.0357, + "step": 35890 + }, + { + "epoch": 0.85, + "grad_norm": 1.8791224145659136, + "learning_rate": 1.2250059090679779e-06, + "loss": 0.9839, + "step": 35891 + }, + { + "epoch": 0.85, + "grad_norm": 2.1821936385215404, + "learning_rate": 1.2246399971237221e-06, + "loss": 1.0221, + "step": 35892 + }, + { + "epoch": 0.85, + "grad_norm": 2.364023806920457, + "learning_rate": 1.224274136272392e-06, + "loss": 0.9171, + "step": 35893 + }, + { + "epoch": 0.85, + "grad_norm": 1.9299457251549363, + "learning_rate": 1.2239083265161166e-06, + "loss": 0.9456, + "step": 35894 + }, + { + "epoch": 0.85, + "grad_norm": 2.139017767946569, + "learning_rate": 1.2235425678570234e-06, + "loss": 0.9295, + "step": 35895 + }, + { + "epoch": 0.85, + "grad_norm": 2.0635003129383924, + "learning_rate": 1.2231768602972439e-06, + "loss": 0.9931, + "step": 35896 + }, + { + "epoch": 0.85, + "grad_norm": 1.7328686629933414, + "learning_rate": 1.2228112038389085e-06, + "loss": 0.9307, + "step": 35897 + }, + { + "epoch": 0.85, + "grad_norm": 2.0319841644525303, + "learning_rate": 1.2224455984841466e-06, + "loss": 0.9752, + "step": 35898 + }, + { + "epoch": 0.85, + "grad_norm": 2.0891031269778697, + "learning_rate": 1.2220800442350823e-06, + "loss": 0.9361, + "step": 35899 + }, + { + "epoch": 0.85, + "grad_norm": 2.188101674040077, + "learning_rate": 1.2217145410938468e-06, + "loss": 1.0103, + "step": 35900 + }, + { + "epoch": 0.85, + "grad_norm": 2.0605309033791817, + "learning_rate": 1.2213490890625711e-06, + "loss": 0.9144, + "step": 35901 + }, + { + "epoch": 0.85, + "grad_norm": 1.9949137557244323, + "learning_rate": 1.2209836881433802e-06, + "loss": 0.9441, + "step": 35902 + }, + { + "epoch": 0.85, + "grad_norm": 1.129091032787768, + "learning_rate": 1.2206183383383997e-06, + "loss": 0.9473, + "step": 35903 + }, + { + "epoch": 0.85, + "grad_norm": 1.774201753209259, + "learning_rate": 1.2202530396497592e-06, + "loss": 0.9292, + "step": 35904 + }, + { + "epoch": 0.85, + "grad_norm": 2.3527224910804243, + "learning_rate": 1.2198877920795859e-06, + "loss": 0.9485, + "step": 35905 + }, + { + "epoch": 0.85, + "grad_norm": 1.78354111560941, + "learning_rate": 1.2195225956300073e-06, + "loss": 0.9499, + "step": 35906 + }, + { + "epoch": 0.85, + "grad_norm": 1.9849628679224274, + "learning_rate": 1.2191574503031445e-06, + "loss": 0.8643, + "step": 35907 + }, + { + "epoch": 0.85, + "grad_norm": 2.3775818210001245, + "learning_rate": 1.2187923561011295e-06, + "loss": 0.9428, + "step": 35908 + }, + { + "epoch": 0.85, + "grad_norm": 1.9681045742277623, + "learning_rate": 1.2184273130260837e-06, + "loss": 0.8385, + "step": 35909 + }, + { + "epoch": 0.85, + "grad_norm": 2.017350333885171, + "learning_rate": 1.2180623210801357e-06, + "loss": 1.131, + "step": 35910 + }, + { + "epoch": 0.85, + "grad_norm": 2.294502150016117, + "learning_rate": 1.2176973802654068e-06, + "loss": 0.9849, + "step": 35911 + }, + { + "epoch": 0.85, + "grad_norm": 1.9061373622146656, + "learning_rate": 1.2173324905840278e-06, + "loss": 0.9619, + "step": 35912 + }, + { + "epoch": 0.85, + "grad_norm": 1.9124426971303483, + "learning_rate": 1.2169676520381168e-06, + "loss": 0.9027, + "step": 35913 + }, + { + "epoch": 0.85, + "grad_norm": 1.9006628600992133, + "learning_rate": 1.2166028646298022e-06, + "loss": 0.8997, + "step": 35914 + }, + { + "epoch": 0.85, + "grad_norm": 2.2728195758850984, + "learning_rate": 1.2162381283612067e-06, + "loss": 1.0287, + "step": 35915 + }, + { + "epoch": 0.85, + "grad_norm": 2.1933696247457495, + "learning_rate": 1.2158734432344521e-06, + "loss": 1.0212, + "step": 35916 + }, + { + "epoch": 0.85, + "grad_norm": 1.0023082491654456, + "learning_rate": 1.2155088092516632e-06, + "loss": 0.9782, + "step": 35917 + }, + { + "epoch": 0.85, + "grad_norm": 2.6888129540536174, + "learning_rate": 1.215144226414965e-06, + "loss": 0.9635, + "step": 35918 + }, + { + "epoch": 0.85, + "grad_norm": 2.1671366744417613, + "learning_rate": 1.214779694726479e-06, + "loss": 1.0909, + "step": 35919 + }, + { + "epoch": 0.85, + "grad_norm": 1.9139671537921674, + "learning_rate": 1.2144152141883247e-06, + "loss": 0.9384, + "step": 35920 + }, + { + "epoch": 0.85, + "grad_norm": 2.0049183875345196, + "learning_rate": 1.2140507848026261e-06, + "loss": 0.9971, + "step": 35921 + }, + { + "epoch": 0.85, + "grad_norm": 2.119002208770832, + "learning_rate": 1.213686406571507e-06, + "loss": 0.998, + "step": 35922 + }, + { + "epoch": 0.85, + "grad_norm": 1.9814624169910082, + "learning_rate": 1.2133220794970879e-06, + "loss": 0.9868, + "step": 35923 + }, + { + "epoch": 0.85, + "grad_norm": 2.0408441020907166, + "learning_rate": 1.2129578035814881e-06, + "loss": 0.9726, + "step": 35924 + }, + { + "epoch": 0.85, + "grad_norm": 2.822040111118244, + "learning_rate": 1.2125935788268296e-06, + "loss": 1.0042, + "step": 35925 + }, + { + "epoch": 0.85, + "grad_norm": 1.039910121769201, + "learning_rate": 1.2122294052352347e-06, + "loss": 0.9232, + "step": 35926 + }, + { + "epoch": 0.85, + "grad_norm": 1.9549016950316398, + "learning_rate": 1.211865282808823e-06, + "loss": 0.9894, + "step": 35927 + }, + { + "epoch": 0.85, + "grad_norm": 2.050167670081306, + "learning_rate": 1.2115012115497115e-06, + "loss": 1.1179, + "step": 35928 + }, + { + "epoch": 0.85, + "grad_norm": 1.9857042260217022, + "learning_rate": 1.2111371914600245e-06, + "loss": 1.0176, + "step": 35929 + }, + { + "epoch": 0.85, + "grad_norm": 2.0768521382952723, + "learning_rate": 1.2107732225418766e-06, + "loss": 0.9039, + "step": 35930 + }, + { + "epoch": 0.85, + "grad_norm": 1.9194198557930486, + "learning_rate": 1.210409304797392e-06, + "loss": 0.9592, + "step": 35931 + }, + { + "epoch": 0.85, + "grad_norm": 1.0929748518497717, + "learning_rate": 1.2100454382286853e-06, + "loss": 0.9644, + "step": 35932 + }, + { + "epoch": 0.85, + "grad_norm": 1.9280162191454069, + "learning_rate": 1.2096816228378794e-06, + "loss": 0.9176, + "step": 35933 + }, + { + "epoch": 0.85, + "grad_norm": 1.9754582903977222, + "learning_rate": 1.209317858627087e-06, + "loss": 0.9758, + "step": 35934 + }, + { + "epoch": 0.85, + "grad_norm": 5.237711835570489, + "learning_rate": 1.2089541455984321e-06, + "loss": 0.938, + "step": 35935 + }, + { + "epoch": 0.85, + "grad_norm": 1.8013525887064956, + "learning_rate": 1.2085904837540275e-06, + "loss": 0.8642, + "step": 35936 + }, + { + "epoch": 0.85, + "grad_norm": 1.9668272268478144, + "learning_rate": 1.2082268730959935e-06, + "loss": 1.0102, + "step": 35937 + }, + { + "epoch": 0.85, + "grad_norm": 1.676137861102518, + "learning_rate": 1.207863313626445e-06, + "loss": 1.0133, + "step": 35938 + }, + { + "epoch": 0.85, + "grad_norm": 2.1633077546476716, + "learning_rate": 1.2074998053474984e-06, + "loss": 0.9912, + "step": 35939 + }, + { + "epoch": 0.85, + "grad_norm": 2.1421017349727003, + "learning_rate": 1.2071363482612775e-06, + "loss": 0.9326, + "step": 35940 + }, + { + "epoch": 0.85, + "grad_norm": 1.0876522943874554, + "learning_rate": 1.2067729423698871e-06, + "loss": 0.9575, + "step": 35941 + }, + { + "epoch": 0.85, + "grad_norm": 1.9814578099888367, + "learning_rate": 1.206409587675449e-06, + "loss": 1.0304, + "step": 35942 + }, + { + "epoch": 0.85, + "grad_norm": 1.8030764890200413, + "learning_rate": 1.2060462841800779e-06, + "loss": 0.9337, + "step": 35943 + }, + { + "epoch": 0.85, + "grad_norm": 2.088164018680077, + "learning_rate": 1.2056830318858947e-06, + "loss": 0.8651, + "step": 35944 + }, + { + "epoch": 0.85, + "grad_norm": 2.0276656428155775, + "learning_rate": 1.2053198307950042e-06, + "loss": 0.9716, + "step": 35945 + }, + { + "epoch": 0.85, + "grad_norm": 2.0910882142337583, + "learning_rate": 1.2049566809095259e-06, + "loss": 0.853, + "step": 35946 + }, + { + "epoch": 0.85, + "grad_norm": 7.432751166632629, + "learning_rate": 1.2045935822315768e-06, + "loss": 0.851, + "step": 35947 + }, + { + "epoch": 0.85, + "grad_norm": 2.033373604256791, + "learning_rate": 1.2042305347632655e-06, + "loss": 0.8664, + "step": 35948 + }, + { + "epoch": 0.85, + "grad_norm": 2.0326506703977416, + "learning_rate": 1.2038675385067112e-06, + "loss": 1.0159, + "step": 35949 + }, + { + "epoch": 0.85, + "grad_norm": 1.0579930013782204, + "learning_rate": 1.2035045934640233e-06, + "loss": 0.9818, + "step": 35950 + }, + { + "epoch": 0.85, + "grad_norm": 1.8846956720613415, + "learning_rate": 1.2031416996373179e-06, + "loss": 0.9867, + "step": 35951 + }, + { + "epoch": 0.85, + "grad_norm": 1.9025440414155224, + "learning_rate": 1.2027788570287047e-06, + "loss": 1.149, + "step": 35952 + }, + { + "epoch": 0.85, + "grad_norm": 1.9888547378717178, + "learning_rate": 1.2024160656403006e-06, + "loss": 1.0062, + "step": 35953 + }, + { + "epoch": 0.85, + "grad_norm": 2.1463315713458213, + "learning_rate": 1.202053325474215e-06, + "loss": 1.0018, + "step": 35954 + }, + { + "epoch": 0.85, + "grad_norm": 2.204867903651573, + "learning_rate": 1.2016906365325586e-06, + "loss": 0.9062, + "step": 35955 + }, + { + "epoch": 0.85, + "grad_norm": 1.9476643088899328, + "learning_rate": 1.2013279988174454e-06, + "loss": 0.9486, + "step": 35956 + }, + { + "epoch": 0.85, + "grad_norm": 1.8662152575997295, + "learning_rate": 1.2009654123309878e-06, + "loss": 1.0755, + "step": 35957 + }, + { + "epoch": 0.85, + "grad_norm": 2.167330465886074, + "learning_rate": 1.2006028770752942e-06, + "loss": 0.9412, + "step": 35958 + }, + { + "epoch": 0.85, + "grad_norm": 2.1086995001306197, + "learning_rate": 1.2002403930524752e-06, + "loss": 0.9463, + "step": 35959 + }, + { + "epoch": 0.85, + "grad_norm": 2.185018472889227, + "learning_rate": 1.1998779602646438e-06, + "loss": 1.0112, + "step": 35960 + }, + { + "epoch": 0.85, + "grad_norm": 1.1006605883314462, + "learning_rate": 1.199515578713909e-06, + "loss": 0.946, + "step": 35961 + }, + { + "epoch": 0.85, + "grad_norm": 1.9858584229599343, + "learning_rate": 1.1991532484023816e-06, + "loss": 0.911, + "step": 35962 + }, + { + "epoch": 0.85, + "grad_norm": 1.9526343561450565, + "learning_rate": 1.1987909693321687e-06, + "loss": 1.0119, + "step": 35963 + }, + { + "epoch": 0.85, + "grad_norm": 2.20508156052669, + "learning_rate": 1.1984287415053807e-06, + "loss": 0.9766, + "step": 35964 + }, + { + "epoch": 0.85, + "grad_norm": 2.161012673022928, + "learning_rate": 1.1980665649241297e-06, + "loss": 1.0834, + "step": 35965 + }, + { + "epoch": 0.85, + "grad_norm": 2.003629659201922, + "learning_rate": 1.1977044395905214e-06, + "loss": 1.1209, + "step": 35966 + }, + { + "epoch": 0.85, + "grad_norm": 2.059480580147287, + "learning_rate": 1.1973423655066619e-06, + "loss": 0.9054, + "step": 35967 + }, + { + "epoch": 0.85, + "grad_norm": 1.019953007558026, + "learning_rate": 1.1969803426746651e-06, + "loss": 0.9473, + "step": 35968 + }, + { + "epoch": 0.85, + "grad_norm": 2.268966453206986, + "learning_rate": 1.1966183710966329e-06, + "loss": 1.1019, + "step": 35969 + }, + { + "epoch": 0.85, + "grad_norm": 2.1634584879974668, + "learning_rate": 1.1962564507746776e-06, + "loss": 1.0669, + "step": 35970 + }, + { + "epoch": 0.85, + "grad_norm": 1.7681526043498632, + "learning_rate": 1.1958945817109035e-06, + "loss": 0.8959, + "step": 35971 + }, + { + "epoch": 0.85, + "grad_norm": 1.81988992385631, + "learning_rate": 1.1955327639074198e-06, + "loss": 0.926, + "step": 35972 + }, + { + "epoch": 0.85, + "grad_norm": 2.1166762280893057, + "learning_rate": 1.1951709973663294e-06, + "loss": 0.9611, + "step": 35973 + }, + { + "epoch": 0.85, + "grad_norm": 1.9737499239829375, + "learning_rate": 1.194809282089744e-06, + "loss": 1.0437, + "step": 35974 + }, + { + "epoch": 0.85, + "grad_norm": 1.9327505285199102, + "learning_rate": 1.194447618079766e-06, + "loss": 1.0539, + "step": 35975 + }, + { + "epoch": 0.85, + "grad_norm": 2.291060906297696, + "learning_rate": 1.1940860053384994e-06, + "loss": 0.9806, + "step": 35976 + }, + { + "epoch": 0.85, + "grad_norm": 1.8659170520454988, + "learning_rate": 1.1937244438680529e-06, + "loss": 0.9969, + "step": 35977 + }, + { + "epoch": 0.85, + "grad_norm": 2.4736733499150363, + "learning_rate": 1.1933629336705322e-06, + "loss": 0.8875, + "step": 35978 + }, + { + "epoch": 0.85, + "grad_norm": 1.9430250744129451, + "learning_rate": 1.19300147474804e-06, + "loss": 0.8925, + "step": 35979 + }, + { + "epoch": 0.85, + "grad_norm": 2.0703793093185086, + "learning_rate": 1.1926400671026805e-06, + "loss": 0.8999, + "step": 35980 + }, + { + "epoch": 0.85, + "grad_norm": 1.7833512856115998, + "learning_rate": 1.1922787107365585e-06, + "loss": 0.8437, + "step": 35981 + }, + { + "epoch": 0.85, + "grad_norm": 2.454887064526363, + "learning_rate": 1.19191740565178e-06, + "loss": 0.9485, + "step": 35982 + }, + { + "epoch": 0.85, + "grad_norm": 1.9441543457899122, + "learning_rate": 1.1915561518504482e-06, + "loss": 0.9708, + "step": 35983 + }, + { + "epoch": 0.85, + "grad_norm": 1.9030389630422067, + "learning_rate": 1.191194949334662e-06, + "loss": 0.9635, + "step": 35984 + }, + { + "epoch": 0.85, + "grad_norm": 2.1005527814272797, + "learning_rate": 1.1908337981065276e-06, + "loss": 1.0355, + "step": 35985 + }, + { + "epoch": 0.85, + "grad_norm": 1.8648047004109312, + "learning_rate": 1.1904726981681502e-06, + "loss": 1.0503, + "step": 35986 + }, + { + "epoch": 0.85, + "grad_norm": 2.0844466101001418, + "learning_rate": 1.1901116495216303e-06, + "loss": 0.9553, + "step": 35987 + }, + { + "epoch": 0.85, + "grad_norm": 2.014914998706638, + "learning_rate": 1.189750652169066e-06, + "loss": 0.9251, + "step": 35988 + }, + { + "epoch": 0.85, + "grad_norm": 2.2060665663913976, + "learning_rate": 1.189389706112566e-06, + "loss": 1.0229, + "step": 35989 + }, + { + "epoch": 0.85, + "grad_norm": 2.2039359373903635, + "learning_rate": 1.189028811354226e-06, + "loss": 1.0497, + "step": 35990 + }, + { + "epoch": 0.85, + "grad_norm": 2.076468828999681, + "learning_rate": 1.188667967896152e-06, + "loss": 0.9856, + "step": 35991 + }, + { + "epoch": 0.85, + "grad_norm": 2.2393888238877566, + "learning_rate": 1.1883071757404408e-06, + "loss": 1.0707, + "step": 35992 + }, + { + "epoch": 0.85, + "grad_norm": 1.9471749431200713, + "learning_rate": 1.1879464348891967e-06, + "loss": 0.8262, + "step": 35993 + }, + { + "epoch": 0.85, + "grad_norm": 1.901710738084954, + "learning_rate": 1.1875857453445162e-06, + "loss": 1.0319, + "step": 35994 + }, + { + "epoch": 0.85, + "grad_norm": 1.9555408604699966, + "learning_rate": 1.1872251071085038e-06, + "loss": 0.8763, + "step": 35995 + }, + { + "epoch": 0.85, + "grad_norm": 1.9486713976355345, + "learning_rate": 1.1868645201832563e-06, + "loss": 0.8819, + "step": 35996 + }, + { + "epoch": 0.85, + "grad_norm": 2.103338918639926, + "learning_rate": 1.1865039845708725e-06, + "loss": 0.905, + "step": 35997 + }, + { + "epoch": 0.85, + "grad_norm": 2.093984570113882, + "learning_rate": 1.1861435002734523e-06, + "loss": 1.0122, + "step": 35998 + }, + { + "epoch": 0.85, + "grad_norm": 1.9210254336791677, + "learning_rate": 1.1857830672930982e-06, + "loss": 1.0131, + "step": 35999 + }, + { + "epoch": 0.85, + "grad_norm": 1.7123838097494173, + "learning_rate": 1.185422685631904e-06, + "loss": 1.0055, + "step": 36000 + }, + { + "epoch": 0.85, + "grad_norm": 1.0319017432212452, + "learning_rate": 1.1850623552919683e-06, + "loss": 1.0052, + "step": 36001 + }, + { + "epoch": 0.85, + "grad_norm": 2.0928401025228394, + "learning_rate": 1.1847020762753903e-06, + "loss": 0.9961, + "step": 36002 + }, + { + "epoch": 0.85, + "grad_norm": 2.207550563575152, + "learning_rate": 1.1843418485842696e-06, + "loss": 1.0261, + "step": 36003 + }, + { + "epoch": 0.85, + "grad_norm": 1.9863785563947938, + "learning_rate": 1.183981672220702e-06, + "loss": 0.9817, + "step": 36004 + }, + { + "epoch": 0.85, + "grad_norm": 2.672920901423886, + "learning_rate": 1.1836215471867819e-06, + "loss": 0.8954, + "step": 36005 + }, + { + "epoch": 0.85, + "grad_norm": 1.9345134384223657, + "learning_rate": 1.1832614734846082e-06, + "loss": 1.0492, + "step": 36006 + }, + { + "epoch": 0.85, + "grad_norm": 2.143091191752323, + "learning_rate": 1.1829014511162794e-06, + "loss": 1.0158, + "step": 36007 + }, + { + "epoch": 0.85, + "grad_norm": 2.1901723179658465, + "learning_rate": 1.1825414800838908e-06, + "loss": 0.8384, + "step": 36008 + }, + { + "epoch": 0.85, + "grad_norm": 1.0720520537510854, + "learning_rate": 1.1821815603895336e-06, + "loss": 0.8901, + "step": 36009 + }, + { + "epoch": 0.85, + "grad_norm": 1.969448264858932, + "learning_rate": 1.1818216920353109e-06, + "loss": 1.0445, + "step": 36010 + }, + { + "epoch": 0.85, + "grad_norm": 1.8631002263987764, + "learning_rate": 1.181461875023311e-06, + "loss": 0.9044, + "step": 36011 + }, + { + "epoch": 0.85, + "grad_norm": 1.9709562076747085, + "learning_rate": 1.1811021093556341e-06, + "loss": 0.9833, + "step": 36012 + }, + { + "epoch": 0.85, + "grad_norm": 2.2429105275352774, + "learning_rate": 1.180742395034371e-06, + "loss": 1.0404, + "step": 36013 + }, + { + "epoch": 0.85, + "grad_norm": 2.0954449533265884, + "learning_rate": 1.180382732061619e-06, + "loss": 1.0129, + "step": 36014 + }, + { + "epoch": 0.85, + "grad_norm": 1.132295272986858, + "learning_rate": 1.1800231204394696e-06, + "loss": 0.9232, + "step": 36015 + }, + { + "epoch": 0.85, + "grad_norm": 2.002561199485742, + "learning_rate": 1.1796635601700201e-06, + "loss": 1.0524, + "step": 36016 + }, + { + "epoch": 0.85, + "grad_norm": 1.8431822112916885, + "learning_rate": 1.1793040512553623e-06, + "loss": 1.0202, + "step": 36017 + }, + { + "epoch": 0.85, + "grad_norm": 1.937956002145121, + "learning_rate": 1.1789445936975862e-06, + "loss": 1.0415, + "step": 36018 + }, + { + "epoch": 0.85, + "grad_norm": 2.116805606586186, + "learning_rate": 1.1785851874987874e-06, + "loss": 1.0808, + "step": 36019 + }, + { + "epoch": 0.85, + "grad_norm": 2.3124356667649173, + "learning_rate": 1.1782258326610606e-06, + "loss": 1.0089, + "step": 36020 + }, + { + "epoch": 0.85, + "grad_norm": 2.2372610869669507, + "learning_rate": 1.1778665291864955e-06, + "loss": 0.9799, + "step": 36021 + }, + { + "epoch": 0.85, + "grad_norm": 2.5078417683093712, + "learning_rate": 1.1775072770771833e-06, + "loss": 1.0367, + "step": 36022 + }, + { + "epoch": 0.85, + "grad_norm": 2.052895026029588, + "learning_rate": 1.177148076335216e-06, + "loss": 0.8696, + "step": 36023 + }, + { + "epoch": 0.85, + "grad_norm": 2.9560171358626928, + "learning_rate": 1.1767889269626887e-06, + "loss": 0.9907, + "step": 36024 + }, + { + "epoch": 0.85, + "grad_norm": 1.7118731611205351, + "learning_rate": 1.1764298289616893e-06, + "loss": 0.914, + "step": 36025 + }, + { + "epoch": 0.85, + "grad_norm": 1.1461393136191813, + "learning_rate": 1.1760707823343065e-06, + "loss": 0.9282, + "step": 36026 + }, + { + "epoch": 0.85, + "grad_norm": 1.9654387062454926, + "learning_rate": 1.1757117870826328e-06, + "loss": 1.0122, + "step": 36027 + }, + { + "epoch": 0.85, + "grad_norm": 2.325638732181141, + "learning_rate": 1.1753528432087624e-06, + "loss": 0.9307, + "step": 36028 + }, + { + "epoch": 0.85, + "grad_norm": 2.0672489462696984, + "learning_rate": 1.174993950714781e-06, + "loss": 0.9511, + "step": 36029 + }, + { + "epoch": 0.85, + "grad_norm": 2.339228076209137, + "learning_rate": 1.1746351096027763e-06, + "loss": 0.8962, + "step": 36030 + }, + { + "epoch": 0.85, + "grad_norm": 2.2697378003434974, + "learning_rate": 1.174276319874842e-06, + "loss": 0.9653, + "step": 36031 + }, + { + "epoch": 0.85, + "grad_norm": 1.893500918244177, + "learning_rate": 1.173917581533064e-06, + "loss": 0.9421, + "step": 36032 + }, + { + "epoch": 0.85, + "grad_norm": 1.919611003848241, + "learning_rate": 1.1735588945795307e-06, + "loss": 0.9547, + "step": 36033 + }, + { + "epoch": 0.85, + "grad_norm": 1.0784959234923774, + "learning_rate": 1.1732002590163349e-06, + "loss": 0.9494, + "step": 36034 + }, + { + "epoch": 0.85, + "grad_norm": 2.4008320226236473, + "learning_rate": 1.1728416748455619e-06, + "loss": 1.0845, + "step": 36035 + }, + { + "epoch": 0.85, + "grad_norm": 1.066912647852946, + "learning_rate": 1.1724831420692973e-06, + "loss": 0.9656, + "step": 36036 + }, + { + "epoch": 0.85, + "grad_norm": 2.111489068279025, + "learning_rate": 1.1721246606896297e-06, + "loss": 0.87, + "step": 36037 + }, + { + "epoch": 0.85, + "grad_norm": 1.9470754393512852, + "learning_rate": 1.171766230708652e-06, + "loss": 1.0177, + "step": 36038 + }, + { + "epoch": 0.85, + "grad_norm": 2.051075374480984, + "learning_rate": 1.1714078521284422e-06, + "loss": 1.0567, + "step": 36039 + }, + { + "epoch": 0.85, + "grad_norm": 1.985527063771299, + "learning_rate": 1.1710495249510911e-06, + "loss": 1.014, + "step": 36040 + }, + { + "epoch": 0.85, + "grad_norm": 2.04374365226408, + "learning_rate": 1.1706912491786837e-06, + "loss": 1.0388, + "step": 36041 + }, + { + "epoch": 0.85, + "grad_norm": 2.03115457232687, + "learning_rate": 1.1703330248133128e-06, + "loss": 1.0856, + "step": 36042 + }, + { + "epoch": 0.85, + "grad_norm": 1.9870967560555486, + "learning_rate": 1.169974851857053e-06, + "loss": 0.9225, + "step": 36043 + }, + { + "epoch": 0.85, + "grad_norm": 2.401691882539302, + "learning_rate": 1.1696167303119964e-06, + "loss": 0.9699, + "step": 36044 + }, + { + "epoch": 0.85, + "grad_norm": 2.2361641242765202, + "learning_rate": 1.169258660180228e-06, + "loss": 1.0222, + "step": 36045 + }, + { + "epoch": 0.85, + "grad_norm": 1.8793439938843386, + "learning_rate": 1.1689006414638294e-06, + "loss": 0.8727, + "step": 36046 + }, + { + "epoch": 0.85, + "grad_norm": 2.023053119163283, + "learning_rate": 1.1685426741648897e-06, + "loss": 0.9121, + "step": 36047 + }, + { + "epoch": 0.85, + "grad_norm": 1.8114437990599956, + "learning_rate": 1.168184758285489e-06, + "loss": 0.8386, + "step": 36048 + }, + { + "epoch": 0.85, + "grad_norm": 2.0883199950144267, + "learning_rate": 1.1678268938277137e-06, + "loss": 0.7935, + "step": 36049 + }, + { + "epoch": 0.85, + "grad_norm": 2.101491490335667, + "learning_rate": 1.167469080793645e-06, + "loss": 1.0441, + "step": 36050 + }, + { + "epoch": 0.85, + "grad_norm": 2.2856202827142003, + "learning_rate": 1.1671113191853688e-06, + "loss": 0.9468, + "step": 36051 + }, + { + "epoch": 0.85, + "grad_norm": 2.0059118342364988, + "learning_rate": 1.166753609004968e-06, + "loss": 0.9428, + "step": 36052 + }, + { + "epoch": 0.85, + "grad_norm": 2.174895150481178, + "learning_rate": 1.1663959502545218e-06, + "loss": 1.007, + "step": 36053 + }, + { + "epoch": 0.85, + "grad_norm": 1.099789884866829, + "learning_rate": 1.1660383429361155e-06, + "loss": 0.9491, + "step": 36054 + }, + { + "epoch": 0.85, + "grad_norm": 1.9311263434110204, + "learning_rate": 1.1656807870518316e-06, + "loss": 0.8642, + "step": 36055 + }, + { + "epoch": 0.85, + "grad_norm": 1.8433027295104, + "learning_rate": 1.165323282603752e-06, + "loss": 0.9729, + "step": 36056 + }, + { + "epoch": 0.85, + "grad_norm": 1.9227625069815055, + "learning_rate": 1.1649658295939548e-06, + "loss": 1.0109, + "step": 36057 + }, + { + "epoch": 0.85, + "grad_norm": 2.036235879672531, + "learning_rate": 1.164608428024524e-06, + "loss": 0.7898, + "step": 36058 + }, + { + "epoch": 0.85, + "grad_norm": 2.0712544320523985, + "learning_rate": 1.1642510778975425e-06, + "loss": 1.0687, + "step": 36059 + }, + { + "epoch": 0.85, + "grad_norm": 2.1383703897868176, + "learning_rate": 1.1638937792150883e-06, + "loss": 1.1159, + "step": 36060 + }, + { + "epoch": 0.85, + "grad_norm": 2.1623520869435224, + "learning_rate": 1.1635365319792402e-06, + "loss": 0.9102, + "step": 36061 + }, + { + "epoch": 0.85, + "grad_norm": 1.9702860310018413, + "learning_rate": 1.1631793361920795e-06, + "loss": 1.0229, + "step": 36062 + }, + { + "epoch": 0.85, + "grad_norm": 2.140331753341184, + "learning_rate": 1.1628221918556882e-06, + "loss": 0.9562, + "step": 36063 + }, + { + "epoch": 0.85, + "grad_norm": 2.254652195473239, + "learning_rate": 1.1624650989721443e-06, + "loss": 0.9428, + "step": 36064 + }, + { + "epoch": 0.85, + "grad_norm": 2.0915571122887884, + "learning_rate": 1.162108057543524e-06, + "loss": 0.9892, + "step": 36065 + }, + { + "epoch": 0.85, + "grad_norm": 1.8765686514175528, + "learning_rate": 1.1617510675719091e-06, + "loss": 0.8901, + "step": 36066 + }, + { + "epoch": 0.85, + "grad_norm": 2.058244713725805, + "learning_rate": 1.16139412905938e-06, + "loss": 0.9215, + "step": 36067 + }, + { + "epoch": 0.85, + "grad_norm": 2.2279664534358243, + "learning_rate": 1.1610372420080118e-06, + "loss": 1.1074, + "step": 36068 + }, + { + "epoch": 0.85, + "grad_norm": 2.7018756252245244, + "learning_rate": 1.1606804064198817e-06, + "loss": 0.906, + "step": 36069 + }, + { + "epoch": 0.85, + "grad_norm": 1.9885761647916707, + "learning_rate": 1.1603236222970704e-06, + "loss": 1.0016, + "step": 36070 + }, + { + "epoch": 0.85, + "grad_norm": 1.068696946893786, + "learning_rate": 1.1599668896416516e-06, + "loss": 0.8524, + "step": 36071 + }, + { + "epoch": 0.85, + "grad_norm": 1.0927824417592138, + "learning_rate": 1.159610208455707e-06, + "loss": 0.9828, + "step": 36072 + }, + { + "epoch": 0.85, + "grad_norm": 3.533311712121006, + "learning_rate": 1.1592535787413074e-06, + "loss": 0.8891, + "step": 36073 + }, + { + "epoch": 0.85, + "grad_norm": 2.5607191155368754, + "learning_rate": 1.158897000500535e-06, + "loss": 0.9459, + "step": 36074 + }, + { + "epoch": 0.85, + "grad_norm": 1.9903880518345891, + "learning_rate": 1.158540473735461e-06, + "loss": 0.9269, + "step": 36075 + }, + { + "epoch": 0.85, + "grad_norm": 2.05516313624644, + "learning_rate": 1.1581839984481657e-06, + "loss": 1.1625, + "step": 36076 + }, + { + "epoch": 0.85, + "grad_norm": 1.9629795534714078, + "learning_rate": 1.157827574640723e-06, + "loss": 1.0341, + "step": 36077 + }, + { + "epoch": 0.85, + "grad_norm": 1.747772565755283, + "learning_rate": 1.1574712023152058e-06, + "loss": 0.9095, + "step": 36078 + }, + { + "epoch": 0.85, + "grad_norm": 2.739283421765599, + "learning_rate": 1.1571148814736898e-06, + "loss": 0.9609, + "step": 36079 + }, + { + "epoch": 0.85, + "grad_norm": 2.7748889478910215, + "learning_rate": 1.1567586121182529e-06, + "loss": 0.9609, + "step": 36080 + }, + { + "epoch": 0.85, + "grad_norm": 2.0877999484754133, + "learning_rate": 1.1564023942509662e-06, + "loss": 0.8292, + "step": 36081 + }, + { + "epoch": 0.85, + "grad_norm": 1.9695027864461856, + "learning_rate": 1.1560462278739037e-06, + "loss": 1.0147, + "step": 36082 + }, + { + "epoch": 0.85, + "grad_norm": 1.9792134613802301, + "learning_rate": 1.1556901129891396e-06, + "loss": 0.9199, + "step": 36083 + }, + { + "epoch": 0.85, + "grad_norm": 1.892335884709463, + "learning_rate": 1.1553340495987497e-06, + "loss": 1.0459, + "step": 36084 + }, + { + "epoch": 0.85, + "grad_norm": 1.9304688312273073, + "learning_rate": 1.154978037704806e-06, + "loss": 1.0134, + "step": 36085 + }, + { + "epoch": 0.85, + "grad_norm": 2.2916679494305114, + "learning_rate": 1.1546220773093774e-06, + "loss": 1.0066, + "step": 36086 + }, + { + "epoch": 0.85, + "grad_norm": 2.071876932331338, + "learning_rate": 1.1542661684145396e-06, + "loss": 1.0548, + "step": 36087 + }, + { + "epoch": 0.85, + "grad_norm": 1.798865824769276, + "learning_rate": 1.153910311022367e-06, + "loss": 0.846, + "step": 36088 + }, + { + "epoch": 0.85, + "grad_norm": 2.7847107017438213, + "learning_rate": 1.1535545051349284e-06, + "loss": 0.969, + "step": 36089 + }, + { + "epoch": 0.85, + "grad_norm": 2.3596994723811773, + "learning_rate": 1.1531987507542953e-06, + "loss": 1.0284, + "step": 36090 + }, + { + "epoch": 0.85, + "grad_norm": 1.1855744039742404, + "learning_rate": 1.1528430478825414e-06, + "loss": 0.8911, + "step": 36091 + }, + { + "epoch": 0.85, + "grad_norm": 1.0148563152558403, + "learning_rate": 1.1524873965217342e-06, + "loss": 0.8799, + "step": 36092 + }, + { + "epoch": 0.85, + "grad_norm": 1.1002268143936313, + "learning_rate": 1.1521317966739486e-06, + "loss": 0.8619, + "step": 36093 + }, + { + "epoch": 0.85, + "grad_norm": 1.8773318814970261, + "learning_rate": 1.1517762483412509e-06, + "loss": 0.8916, + "step": 36094 + }, + { + "epoch": 0.85, + "grad_norm": 2.404198156060856, + "learning_rate": 1.1514207515257147e-06, + "loss": 0.9535, + "step": 36095 + }, + { + "epoch": 0.85, + "grad_norm": 2.159992032099341, + "learning_rate": 1.1510653062294075e-06, + "loss": 0.9477, + "step": 36096 + }, + { + "epoch": 0.85, + "grad_norm": 2.0814905415052647, + "learning_rate": 1.1507099124544008e-06, + "loss": 1.0332, + "step": 36097 + }, + { + "epoch": 0.85, + "grad_norm": 2.3642430177979654, + "learning_rate": 1.1503545702027619e-06, + "loss": 0.9891, + "step": 36098 + }, + { + "epoch": 0.85, + "grad_norm": 1.1363716748800625, + "learning_rate": 1.1499992794765592e-06, + "loss": 0.9871, + "step": 36099 + }, + { + "epoch": 0.85, + "grad_norm": 2.021091786561682, + "learning_rate": 1.149644040277863e-06, + "loss": 0.9879, + "step": 36100 + }, + { + "epoch": 0.85, + "grad_norm": 2.0007372409792774, + "learning_rate": 1.149288852608743e-06, + "loss": 0.9625, + "step": 36101 + }, + { + "epoch": 0.85, + "grad_norm": 1.9860513377891513, + "learning_rate": 1.1489337164712644e-06, + "loss": 0.997, + "step": 36102 + }, + { + "epoch": 0.85, + "grad_norm": 2.1268383135034314, + "learning_rate": 1.1485786318674951e-06, + "loss": 1.0102, + "step": 36103 + }, + { + "epoch": 0.85, + "grad_norm": 2.2406614387289556, + "learning_rate": 1.1482235987995027e-06, + "loss": 0.8842, + "step": 36104 + }, + { + "epoch": 0.85, + "grad_norm": 2.332951133457891, + "learning_rate": 1.1478686172693576e-06, + "loss": 1.0967, + "step": 36105 + }, + { + "epoch": 0.85, + "grad_norm": 1.9904841378856748, + "learning_rate": 1.1475136872791226e-06, + "loss": 0.9918, + "step": 36106 + }, + { + "epoch": 0.85, + "grad_norm": 3.905958127919277, + "learning_rate": 1.1471588088308649e-06, + "loss": 0.8672, + "step": 36107 + }, + { + "epoch": 0.85, + "grad_norm": 2.0634810502481473, + "learning_rate": 1.1468039819266508e-06, + "loss": 1.0349, + "step": 36108 + }, + { + "epoch": 0.85, + "grad_norm": 2.0456791026711016, + "learning_rate": 1.1464492065685483e-06, + "loss": 0.9731, + "step": 36109 + }, + { + "epoch": 0.85, + "grad_norm": 2.0449440826478074, + "learning_rate": 1.1460944827586228e-06, + "loss": 0.9561, + "step": 36110 + }, + { + "epoch": 0.85, + "grad_norm": 1.9546189306002253, + "learning_rate": 1.1457398104989347e-06, + "loss": 1.0458, + "step": 36111 + }, + { + "epoch": 0.85, + "grad_norm": 1.8673494780732864, + "learning_rate": 1.1453851897915557e-06, + "loss": 1.0272, + "step": 36112 + }, + { + "epoch": 0.85, + "grad_norm": 2.1079537595030398, + "learning_rate": 1.1450306206385453e-06, + "loss": 0.9868, + "step": 36113 + }, + { + "epoch": 0.85, + "grad_norm": 1.0588595255294384, + "learning_rate": 1.1446761030419718e-06, + "loss": 0.9892, + "step": 36114 + }, + { + "epoch": 0.85, + "grad_norm": 1.9497026220835931, + "learning_rate": 1.1443216370038968e-06, + "loss": 0.9047, + "step": 36115 + }, + { + "epoch": 0.85, + "grad_norm": 2.184049663222836, + "learning_rate": 1.1439672225263853e-06, + "loss": 0.9322, + "step": 36116 + }, + { + "epoch": 0.85, + "grad_norm": 2.009481464993656, + "learning_rate": 1.1436128596114992e-06, + "loss": 1.016, + "step": 36117 + }, + { + "epoch": 0.85, + "grad_norm": 2.2956163048655798, + "learning_rate": 1.1432585482613045e-06, + "loss": 1.0624, + "step": 36118 + }, + { + "epoch": 0.85, + "grad_norm": 2.020025281065756, + "learning_rate": 1.142904288477863e-06, + "loss": 0.8564, + "step": 36119 + }, + { + "epoch": 0.85, + "grad_norm": 1.1383465008403308, + "learning_rate": 1.142550080263234e-06, + "loss": 0.9455, + "step": 36120 + }, + { + "epoch": 0.85, + "grad_norm": 1.9454104686583809, + "learning_rate": 1.1421959236194835e-06, + "loss": 0.9557, + "step": 36121 + }, + { + "epoch": 0.85, + "grad_norm": 2.3045171388805694, + "learning_rate": 1.1418418185486725e-06, + "loss": 1.0808, + "step": 36122 + }, + { + "epoch": 0.85, + "grad_norm": 2.1829038939460634, + "learning_rate": 1.1414877650528667e-06, + "loss": 1.126, + "step": 36123 + }, + { + "epoch": 0.85, + "grad_norm": 1.9704906792868984, + "learning_rate": 1.1411337631341202e-06, + "loss": 0.8905, + "step": 36124 + }, + { + "epoch": 0.85, + "grad_norm": 2.1104198510517276, + "learning_rate": 1.140779812794497e-06, + "loss": 1.0049, + "step": 36125 + }, + { + "epoch": 0.85, + "grad_norm": 2.0497366969812374, + "learning_rate": 1.1404259140360606e-06, + "loss": 0.9525, + "step": 36126 + }, + { + "epoch": 0.85, + "grad_norm": 2.2035086201402767, + "learning_rate": 1.140072066860869e-06, + "loss": 0.9766, + "step": 36127 + }, + { + "epoch": 0.85, + "grad_norm": 2.0370863810784043, + "learning_rate": 1.139718271270982e-06, + "loss": 1.0403, + "step": 36128 + }, + { + "epoch": 0.85, + "grad_norm": 1.9582709630874096, + "learning_rate": 1.1393645272684595e-06, + "loss": 1.059, + "step": 36129 + }, + { + "epoch": 0.85, + "grad_norm": 1.8769523552874157, + "learning_rate": 1.139010834855364e-06, + "loss": 1.0041, + "step": 36130 + }, + { + "epoch": 0.85, + "grad_norm": 1.7383163743274412, + "learning_rate": 1.138657194033751e-06, + "loss": 0.9547, + "step": 36131 + }, + { + "epoch": 0.85, + "grad_norm": 2.0986928028161222, + "learning_rate": 1.1383036048056828e-06, + "loss": 0.9605, + "step": 36132 + }, + { + "epoch": 0.85, + "grad_norm": 3.5765733066468903, + "learning_rate": 1.137950067173217e-06, + "loss": 0.9836, + "step": 36133 + }, + { + "epoch": 0.85, + "grad_norm": 3.656237678629337, + "learning_rate": 1.1375965811384093e-06, + "loss": 1.1264, + "step": 36134 + }, + { + "epoch": 0.85, + "grad_norm": 1.9315096718685691, + "learning_rate": 1.1372431467033208e-06, + "loss": 1.0619, + "step": 36135 + }, + { + "epoch": 0.85, + "grad_norm": 1.8900051178297994, + "learning_rate": 1.1368897638700105e-06, + "loss": 0.8516, + "step": 36136 + }, + { + "epoch": 0.85, + "grad_norm": 2.323428846616029, + "learning_rate": 1.1365364326405325e-06, + "loss": 0.9297, + "step": 36137 + }, + { + "epoch": 0.85, + "grad_norm": 1.7869721886999859, + "learning_rate": 1.1361831530169453e-06, + "loss": 0.954, + "step": 36138 + }, + { + "epoch": 0.85, + "grad_norm": 1.155035467419604, + "learning_rate": 1.1358299250013049e-06, + "loss": 0.9238, + "step": 36139 + }, + { + "epoch": 0.85, + "grad_norm": 2.151027046821945, + "learning_rate": 1.1354767485956741e-06, + "loss": 1.0283, + "step": 36140 + }, + { + "epoch": 0.85, + "grad_norm": 1.8142260664996948, + "learning_rate": 1.135123623802099e-06, + "loss": 0.8975, + "step": 36141 + }, + { + "epoch": 0.85, + "grad_norm": 2.2824233789053823, + "learning_rate": 1.1347705506226415e-06, + "loss": 0.8322, + "step": 36142 + }, + { + "epoch": 0.85, + "grad_norm": 1.092274664183434, + "learning_rate": 1.1344175290593563e-06, + "loss": 0.8932, + "step": 36143 + }, + { + "epoch": 0.85, + "grad_norm": 1.216333832741174, + "learning_rate": 1.1340645591143007e-06, + "loss": 0.9881, + "step": 36144 + }, + { + "epoch": 0.85, + "grad_norm": 1.9244278163112563, + "learning_rate": 1.1337116407895289e-06, + "loss": 0.8571, + "step": 36145 + }, + { + "epoch": 0.85, + "grad_norm": 1.9226527048258084, + "learning_rate": 1.1333587740870921e-06, + "loss": 1.0585, + "step": 36146 + }, + { + "epoch": 0.85, + "grad_norm": 2.0231688365506533, + "learning_rate": 1.1330059590090502e-06, + "loss": 1.0617, + "step": 36147 + }, + { + "epoch": 0.85, + "grad_norm": 2.1263551022387728, + "learning_rate": 1.1326531955574526e-06, + "loss": 0.9538, + "step": 36148 + }, + { + "epoch": 0.85, + "grad_norm": 1.8797006190554457, + "learning_rate": 1.1323004837343576e-06, + "loss": 1.029, + "step": 36149 + }, + { + "epoch": 0.85, + "grad_norm": 2.054868418987855, + "learning_rate": 1.1319478235418146e-06, + "loss": 1.0289, + "step": 36150 + }, + { + "epoch": 0.85, + "grad_norm": 1.8879961070409241, + "learning_rate": 1.131595214981881e-06, + "loss": 0.8774, + "step": 36151 + }, + { + "epoch": 0.85, + "grad_norm": 1.0940279891548035, + "learning_rate": 1.131242658056606e-06, + "loss": 0.9663, + "step": 36152 + }, + { + "epoch": 0.85, + "grad_norm": 1.8692019046705435, + "learning_rate": 1.130890152768046e-06, + "loss": 0.8396, + "step": 36153 + }, + { + "epoch": 0.85, + "grad_norm": 2.3132218361161927, + "learning_rate": 1.1305376991182516e-06, + "loss": 0.9938, + "step": 36154 + }, + { + "epoch": 0.85, + "grad_norm": 1.9048767190071303, + "learning_rate": 1.130185297109272e-06, + "loss": 1.0483, + "step": 36155 + }, + { + "epoch": 0.85, + "grad_norm": 1.1570371521085103, + "learning_rate": 1.1298329467431624e-06, + "loss": 0.9905, + "step": 36156 + }, + { + "epoch": 0.85, + "grad_norm": 1.8938034134893267, + "learning_rate": 1.1294806480219744e-06, + "loss": 0.9882, + "step": 36157 + }, + { + "epoch": 0.85, + "grad_norm": 1.9290036850890848, + "learning_rate": 1.1291284009477598e-06, + "loss": 0.8838, + "step": 36158 + }, + { + "epoch": 0.85, + "grad_norm": 2.0061218387504653, + "learning_rate": 1.128776205522566e-06, + "loss": 0.9418, + "step": 36159 + }, + { + "epoch": 0.85, + "grad_norm": 1.8649858496402552, + "learning_rate": 1.1284240617484455e-06, + "loss": 1.0211, + "step": 36160 + }, + { + "epoch": 0.85, + "grad_norm": 1.9802321277029358, + "learning_rate": 1.12807196962745e-06, + "loss": 0.9205, + "step": 36161 + }, + { + "epoch": 0.85, + "grad_norm": 2.4625796817361514, + "learning_rate": 1.1277199291616291e-06, + "loss": 0.9069, + "step": 36162 + }, + { + "epoch": 0.85, + "grad_norm": 2.1449110004539813, + "learning_rate": 1.127367940353029e-06, + "loss": 1.0552, + "step": 36163 + }, + { + "epoch": 0.85, + "grad_norm": 2.1528571921310387, + "learning_rate": 1.1270160032037024e-06, + "loss": 1.1596, + "step": 36164 + }, + { + "epoch": 0.85, + "grad_norm": 2.1438870087666433, + "learning_rate": 1.1266641177157e-06, + "loss": 1.0254, + "step": 36165 + }, + { + "epoch": 0.85, + "grad_norm": 1.1152870572677696, + "learning_rate": 1.126312283891069e-06, + "loss": 0.9253, + "step": 36166 + }, + { + "epoch": 0.85, + "grad_norm": 2.0607380322967725, + "learning_rate": 1.1259605017318543e-06, + "loss": 1.0727, + "step": 36167 + }, + { + "epoch": 0.85, + "grad_norm": 2.079087693301197, + "learning_rate": 1.1256087712401087e-06, + "loss": 0.9885, + "step": 36168 + }, + { + "epoch": 0.85, + "grad_norm": 1.8934237924669832, + "learning_rate": 1.1252570924178763e-06, + "loss": 0.8823, + "step": 36169 + }, + { + "epoch": 0.85, + "grad_norm": 1.9880596526417353, + "learning_rate": 1.1249054652672097e-06, + "loss": 0.9067, + "step": 36170 + }, + { + "epoch": 0.85, + "grad_norm": 1.9461777964387008, + "learning_rate": 1.124553889790151e-06, + "loss": 0.9003, + "step": 36171 + }, + { + "epoch": 0.85, + "grad_norm": 1.7945608069941232, + "learning_rate": 1.1242023659887513e-06, + "loss": 0.906, + "step": 36172 + }, + { + "epoch": 0.85, + "grad_norm": 1.9213475933478037, + "learning_rate": 1.1238508938650527e-06, + "loss": 0.8636, + "step": 36173 + }, + { + "epoch": 0.85, + "grad_norm": 1.9133113245271929, + "learning_rate": 1.123499473421108e-06, + "loss": 0.9399, + "step": 36174 + }, + { + "epoch": 0.85, + "grad_norm": 2.0629082075074727, + "learning_rate": 1.1231481046589587e-06, + "loss": 1.0943, + "step": 36175 + }, + { + "epoch": 0.85, + "grad_norm": 1.8347006722843844, + "learning_rate": 1.1227967875806489e-06, + "loss": 0.932, + "step": 36176 + }, + { + "epoch": 0.85, + "grad_norm": 1.772470069564855, + "learning_rate": 1.122445522188228e-06, + "loss": 0.9018, + "step": 36177 + }, + { + "epoch": 0.85, + "grad_norm": 3.390258008349514, + "learning_rate": 1.122094308483741e-06, + "loss": 1.0522, + "step": 36178 + }, + { + "epoch": 0.85, + "grad_norm": 2.147826243164878, + "learning_rate": 1.121743146469232e-06, + "loss": 0.9011, + "step": 36179 + }, + { + "epoch": 0.85, + "grad_norm": 1.7398744710279201, + "learning_rate": 1.1213920361467422e-06, + "loss": 1.0313, + "step": 36180 + }, + { + "epoch": 0.85, + "grad_norm": 2.455878932970416, + "learning_rate": 1.1210409775183207e-06, + "loss": 0.9005, + "step": 36181 + }, + { + "epoch": 0.85, + "grad_norm": 1.8468547698458444, + "learning_rate": 1.1206899705860097e-06, + "loss": 0.8848, + "step": 36182 + }, + { + "epoch": 0.85, + "grad_norm": 1.9347182118159116, + "learning_rate": 1.1203390153518546e-06, + "loss": 0.9508, + "step": 36183 + }, + { + "epoch": 0.85, + "grad_norm": 1.983178030461625, + "learning_rate": 1.1199881118178934e-06, + "loss": 1.1008, + "step": 36184 + }, + { + "epoch": 0.85, + "grad_norm": 2.048045975152698, + "learning_rate": 1.1196372599861738e-06, + "loss": 1.0663, + "step": 36185 + }, + { + "epoch": 0.85, + "grad_norm": 1.8863535248898182, + "learning_rate": 1.1192864598587395e-06, + "loss": 0.9694, + "step": 36186 + }, + { + "epoch": 0.85, + "grad_norm": 2.154089178628558, + "learning_rate": 1.1189357114376308e-06, + "loss": 0.8042, + "step": 36187 + }, + { + "epoch": 0.85, + "grad_norm": 1.9322462616858789, + "learning_rate": 1.1185850147248879e-06, + "loss": 0.9548, + "step": 36188 + }, + { + "epoch": 0.85, + "grad_norm": 3.8976381425758073, + "learning_rate": 1.118234369722555e-06, + "loss": 0.9722, + "step": 36189 + }, + { + "epoch": 0.85, + "grad_norm": 1.772751874795932, + "learning_rate": 1.1178837764326766e-06, + "loss": 0.9295, + "step": 36190 + }, + { + "epoch": 0.85, + "grad_norm": 1.9706473684079238, + "learning_rate": 1.1175332348572898e-06, + "loss": 0.982, + "step": 36191 + }, + { + "epoch": 0.85, + "grad_norm": 1.7132709662952454, + "learning_rate": 1.1171827449984341e-06, + "loss": 0.9691, + "step": 36192 + }, + { + "epoch": 0.85, + "grad_norm": 3.074268315671826, + "learning_rate": 1.1168323068581555e-06, + "loss": 1.0108, + "step": 36193 + }, + { + "epoch": 0.85, + "grad_norm": 1.9312894750776626, + "learning_rate": 1.1164819204384891e-06, + "loss": 1.0216, + "step": 36194 + }, + { + "epoch": 0.85, + "grad_norm": 2.4021976684334874, + "learning_rate": 1.11613158574148e-06, + "loss": 1.0322, + "step": 36195 + }, + { + "epoch": 0.85, + "grad_norm": 1.9828203511373874, + "learning_rate": 1.1157813027691632e-06, + "loss": 1.078, + "step": 36196 + }, + { + "epoch": 0.85, + "grad_norm": 3.1528471911938625, + "learning_rate": 1.1154310715235827e-06, + "loss": 0.9865, + "step": 36197 + }, + { + "epoch": 0.85, + "grad_norm": 2.10058733819483, + "learning_rate": 1.1150808920067723e-06, + "loss": 0.9904, + "step": 36198 + }, + { + "epoch": 0.85, + "grad_norm": 1.9586455043383528, + "learning_rate": 1.1147307642207771e-06, + "loss": 0.9586, + "step": 36199 + }, + { + "epoch": 0.85, + "grad_norm": 2.215240315814621, + "learning_rate": 1.1143806881676311e-06, + "loss": 1.0239, + "step": 36200 + }, + { + "epoch": 0.85, + "grad_norm": 2.17335518529667, + "learning_rate": 1.1140306638493725e-06, + "loss": 0.9825, + "step": 36201 + }, + { + "epoch": 0.85, + "grad_norm": 2.2130910971152744, + "learning_rate": 1.113680691268041e-06, + "loss": 1.0395, + "step": 36202 + }, + { + "epoch": 0.85, + "grad_norm": 1.9616055745678722, + "learning_rate": 1.1133307704256758e-06, + "loss": 1.08, + "step": 36203 + }, + { + "epoch": 0.85, + "grad_norm": 1.0375389115028721, + "learning_rate": 1.112980901324312e-06, + "loss": 0.9512, + "step": 36204 + }, + { + "epoch": 0.85, + "grad_norm": 2.0847821340453145, + "learning_rate": 1.112631083965985e-06, + "loss": 1.0123, + "step": 36205 + }, + { + "epoch": 0.85, + "grad_norm": 1.9213761509916139, + "learning_rate": 1.1122813183527337e-06, + "loss": 0.9897, + "step": 36206 + }, + { + "epoch": 0.85, + "grad_norm": 2.2073481714101493, + "learning_rate": 1.111931604486597e-06, + "loss": 0.9476, + "step": 36207 + }, + { + "epoch": 0.85, + "grad_norm": 1.9095578163115006, + "learning_rate": 1.1115819423696073e-06, + "loss": 0.9075, + "step": 36208 + }, + { + "epoch": 0.85, + "grad_norm": 2.0429529403481435, + "learning_rate": 1.1112323320037998e-06, + "loss": 1.0066, + "step": 36209 + }, + { + "epoch": 0.85, + "grad_norm": 2.3336805396252633, + "learning_rate": 1.110882773391212e-06, + "loss": 1.1317, + "step": 36210 + }, + { + "epoch": 0.85, + "grad_norm": 1.8296066008661453, + "learning_rate": 1.1105332665338808e-06, + "loss": 0.7908, + "step": 36211 + }, + { + "epoch": 0.85, + "grad_norm": 10.347267175155379, + "learning_rate": 1.1101838114338393e-06, + "loss": 1.0953, + "step": 36212 + }, + { + "epoch": 0.85, + "grad_norm": 1.0629919230136335, + "learning_rate": 1.10983440809312e-06, + "loss": 0.9172, + "step": 36213 + }, + { + "epoch": 0.85, + "grad_norm": 2.117902643699415, + "learning_rate": 1.1094850565137616e-06, + "loss": 1.0408, + "step": 36214 + }, + { + "epoch": 0.85, + "grad_norm": 1.7709834042092678, + "learning_rate": 1.1091357566977934e-06, + "loss": 0.9115, + "step": 36215 + }, + { + "epoch": 0.85, + "grad_norm": 1.9301214087442171, + "learning_rate": 1.108786508647255e-06, + "loss": 1.0375, + "step": 36216 + }, + { + "epoch": 0.85, + "grad_norm": 2.3381867754981926, + "learning_rate": 1.1084373123641734e-06, + "loss": 1.0389, + "step": 36217 + }, + { + "epoch": 0.85, + "grad_norm": 1.955957739437211, + "learning_rate": 1.1080881678505862e-06, + "loss": 1.0198, + "step": 36218 + }, + { + "epoch": 0.85, + "grad_norm": 2.2027936687440297, + "learning_rate": 1.1077390751085237e-06, + "loss": 0.9671, + "step": 36219 + }, + { + "epoch": 0.85, + "grad_norm": 2.2083949442886697, + "learning_rate": 1.107390034140019e-06, + "loss": 1.0455, + "step": 36220 + }, + { + "epoch": 0.85, + "grad_norm": 2.5259265056352787, + "learning_rate": 1.1070410449471103e-06, + "loss": 0.8458, + "step": 36221 + }, + { + "epoch": 0.85, + "grad_norm": 1.9799180175019808, + "learning_rate": 1.1066921075318183e-06, + "loss": 0.9117, + "step": 36222 + }, + { + "epoch": 0.85, + "grad_norm": 2.270256577815114, + "learning_rate": 1.106343221896181e-06, + "loss": 1.087, + "step": 36223 + }, + { + "epoch": 0.85, + "grad_norm": 1.953957909847169, + "learning_rate": 1.1059943880422297e-06, + "loss": 1.0122, + "step": 36224 + }, + { + "epoch": 0.85, + "grad_norm": 1.025190250936925, + "learning_rate": 1.1056456059719989e-06, + "loss": 0.9473, + "step": 36225 + }, + { + "epoch": 0.85, + "grad_norm": 3.2256736855318486, + "learning_rate": 1.1052968756875104e-06, + "loss": 1.0576, + "step": 36226 + }, + { + "epoch": 0.85, + "grad_norm": 2.2400514092264343, + "learning_rate": 1.1049481971907993e-06, + "loss": 1.0238, + "step": 36227 + }, + { + "epoch": 0.85, + "grad_norm": 2.0347000687234162, + "learning_rate": 1.1045995704838997e-06, + "loss": 0.9928, + "step": 36228 + }, + { + "epoch": 0.85, + "grad_norm": 1.8601510580652094, + "learning_rate": 1.104250995568834e-06, + "loss": 0.9194, + "step": 36229 + }, + { + "epoch": 0.85, + "grad_norm": 1.146019012078161, + "learning_rate": 1.1039024724476376e-06, + "loss": 0.9771, + "step": 36230 + }, + { + "epoch": 0.85, + "grad_norm": 2.0272168154660144, + "learning_rate": 1.1035540011223355e-06, + "loss": 0.9524, + "step": 36231 + }, + { + "epoch": 0.85, + "grad_norm": 1.824823015746272, + "learning_rate": 1.1032055815949616e-06, + "loss": 1.0214, + "step": 36232 + }, + { + "epoch": 0.85, + "grad_norm": 1.738815261889694, + "learning_rate": 1.1028572138675387e-06, + "loss": 1.0439, + "step": 36233 + }, + { + "epoch": 0.85, + "grad_norm": 1.1407402581286534, + "learning_rate": 1.1025088979420995e-06, + "loss": 0.9703, + "step": 36234 + }, + { + "epoch": 0.85, + "grad_norm": 2.208053631625581, + "learning_rate": 1.1021606338206714e-06, + "loss": 1.0109, + "step": 36235 + }, + { + "epoch": 0.85, + "grad_norm": 2.294736847824825, + "learning_rate": 1.1018124215052783e-06, + "loss": 0.9467, + "step": 36236 + }, + { + "epoch": 0.85, + "grad_norm": 1.945059026583468, + "learning_rate": 1.1014642609979498e-06, + "loss": 0.9591, + "step": 36237 + }, + { + "epoch": 0.85, + "grad_norm": 1.9273296331571717, + "learning_rate": 1.1011161523007163e-06, + "loss": 0.9102, + "step": 36238 + }, + { + "epoch": 0.85, + "grad_norm": 2.0371567705991986, + "learning_rate": 1.100768095415603e-06, + "loss": 0.991, + "step": 36239 + }, + { + "epoch": 0.85, + "grad_norm": 2.447279763485345, + "learning_rate": 1.1004200903446316e-06, + "loss": 1.0485, + "step": 36240 + }, + { + "epoch": 0.85, + "grad_norm": 2.726949408140274, + "learning_rate": 1.1000721370898326e-06, + "loss": 0.8616, + "step": 36241 + }, + { + "epoch": 0.85, + "grad_norm": 1.9412091312703255, + "learning_rate": 1.0997242356532335e-06, + "loss": 1.0782, + "step": 36242 + }, + { + "epoch": 0.85, + "grad_norm": 2.756799906161759, + "learning_rate": 1.099376386036858e-06, + "loss": 0.9595, + "step": 36243 + }, + { + "epoch": 0.85, + "grad_norm": 1.8997072285345311, + "learning_rate": 1.0990285882427286e-06, + "loss": 0.9025, + "step": 36244 + }, + { + "epoch": 0.85, + "grad_norm": 1.9948679903288675, + "learning_rate": 1.0986808422728722e-06, + "loss": 1.0265, + "step": 36245 + }, + { + "epoch": 0.85, + "grad_norm": 2.215429713555159, + "learning_rate": 1.0983331481293168e-06, + "loss": 1.0118, + "step": 36246 + }, + { + "epoch": 0.85, + "grad_norm": 1.974132812138758, + "learning_rate": 1.0979855058140842e-06, + "loss": 0.9546, + "step": 36247 + }, + { + "epoch": 0.85, + "grad_norm": 2.1464987708564403, + "learning_rate": 1.097637915329196e-06, + "loss": 0.9208, + "step": 36248 + }, + { + "epoch": 0.85, + "grad_norm": 2.3129927487377517, + "learning_rate": 1.0972903766766797e-06, + "loss": 0.9613, + "step": 36249 + }, + { + "epoch": 0.85, + "grad_norm": 1.914804379071285, + "learning_rate": 1.0969428898585565e-06, + "loss": 1.0298, + "step": 36250 + }, + { + "epoch": 0.85, + "grad_norm": 2.064303420674036, + "learning_rate": 1.0965954548768509e-06, + "loss": 1.0005, + "step": 36251 + }, + { + "epoch": 0.85, + "grad_norm": 1.6489805117728498, + "learning_rate": 1.0962480717335843e-06, + "loss": 0.8268, + "step": 36252 + }, + { + "epoch": 0.85, + "grad_norm": 1.9021517106643107, + "learning_rate": 1.095900740430783e-06, + "loss": 0.991, + "step": 36253 + }, + { + "epoch": 0.85, + "grad_norm": 1.0700831292578061, + "learning_rate": 1.0955534609704644e-06, + "loss": 0.9984, + "step": 36254 + }, + { + "epoch": 0.85, + "grad_norm": 2.07216081075548, + "learning_rate": 1.095206233354653e-06, + "loss": 0.9525, + "step": 36255 + }, + { + "epoch": 0.85, + "grad_norm": 2.24107886748039, + "learning_rate": 1.0948590575853723e-06, + "loss": 1.0045, + "step": 36256 + }, + { + "epoch": 0.85, + "grad_norm": 2.3507928955262547, + "learning_rate": 1.0945119336646382e-06, + "loss": 0.932, + "step": 36257 + }, + { + "epoch": 0.85, + "grad_norm": 2.0751712447622026, + "learning_rate": 1.0941648615944756e-06, + "loss": 0.9235, + "step": 36258 + }, + { + "epoch": 0.85, + "grad_norm": 2.2627873767794227, + "learning_rate": 1.0938178413769062e-06, + "loss": 0.99, + "step": 36259 + }, + { + "epoch": 0.85, + "grad_norm": 1.0735498580456857, + "learning_rate": 1.0934708730139486e-06, + "loss": 0.944, + "step": 36260 + }, + { + "epoch": 0.85, + "grad_norm": 2.1273054508384974, + "learning_rate": 1.0931239565076224e-06, + "loss": 0.9533, + "step": 36261 + }, + { + "epoch": 0.85, + "grad_norm": 1.977281574275339, + "learning_rate": 1.0927770918599478e-06, + "loss": 1.044, + "step": 36262 + }, + { + "epoch": 0.85, + "grad_norm": 1.8127892940690318, + "learning_rate": 1.0924302790729468e-06, + "loss": 0.9347, + "step": 36263 + }, + { + "epoch": 0.85, + "grad_norm": 1.9739119136783478, + "learning_rate": 1.0920835181486378e-06, + "loss": 1.1028, + "step": 36264 + }, + { + "epoch": 0.85, + "grad_norm": 1.9081554689357139, + "learning_rate": 1.0917368090890357e-06, + "loss": 1.0011, + "step": 36265 + }, + { + "epoch": 0.85, + "grad_norm": 1.0779268410488367, + "learning_rate": 1.0913901518961623e-06, + "loss": 0.9485, + "step": 36266 + }, + { + "epoch": 0.85, + "grad_norm": 1.9241885804865453, + "learning_rate": 1.0910435465720382e-06, + "loss": 0.8635, + "step": 36267 + }, + { + "epoch": 0.85, + "grad_norm": 2.0397436439145866, + "learning_rate": 1.0906969931186795e-06, + "loss": 0.8867, + "step": 36268 + }, + { + "epoch": 0.85, + "grad_norm": 2.2854400521704052, + "learning_rate": 1.0903504915381002e-06, + "loss": 0.9654, + "step": 36269 + }, + { + "epoch": 0.85, + "grad_norm": 2.0887603597579196, + "learning_rate": 1.0900040418323243e-06, + "loss": 1.0482, + "step": 36270 + }, + { + "epoch": 0.85, + "grad_norm": 2.020912105186515, + "learning_rate": 1.0896576440033636e-06, + "loss": 0.9959, + "step": 36271 + }, + { + "epoch": 0.85, + "grad_norm": 1.8806869380998563, + "learning_rate": 1.0893112980532384e-06, + "loss": 1.0918, + "step": 36272 + }, + { + "epoch": 0.85, + "grad_norm": 2.4855208552846757, + "learning_rate": 1.0889650039839627e-06, + "loss": 0.9248, + "step": 36273 + }, + { + "epoch": 0.85, + "grad_norm": 2.025459631288749, + "learning_rate": 1.0886187617975552e-06, + "loss": 0.9753, + "step": 36274 + }, + { + "epoch": 0.85, + "grad_norm": 2.039501947837084, + "learning_rate": 1.0882725714960296e-06, + "loss": 0.9501, + "step": 36275 + }, + { + "epoch": 0.85, + "grad_norm": 2.1217128586021867, + "learning_rate": 1.0879264330814033e-06, + "loss": 1.0175, + "step": 36276 + }, + { + "epoch": 0.85, + "grad_norm": 1.902895547909435, + "learning_rate": 1.0875803465556911e-06, + "loss": 1.0116, + "step": 36277 + }, + { + "epoch": 0.85, + "grad_norm": 2.296678823309028, + "learning_rate": 1.087234311920906e-06, + "loss": 0.8548, + "step": 36278 + }, + { + "epoch": 0.85, + "grad_norm": 2.2511515461954343, + "learning_rate": 1.0868883291790645e-06, + "loss": 0.8066, + "step": 36279 + }, + { + "epoch": 0.85, + "grad_norm": 1.859842963169048, + "learning_rate": 1.0865423983321832e-06, + "loss": 0.9901, + "step": 36280 + }, + { + "epoch": 0.85, + "grad_norm": 2.231274629937048, + "learning_rate": 1.0861965193822743e-06, + "loss": 0.9346, + "step": 36281 + }, + { + "epoch": 0.85, + "grad_norm": 2.140999463447871, + "learning_rate": 1.0858506923313484e-06, + "loss": 0.8275, + "step": 36282 + }, + { + "epoch": 0.85, + "grad_norm": 2.2646020072933033, + "learning_rate": 1.0855049171814226e-06, + "loss": 1.0206, + "step": 36283 + }, + { + "epoch": 0.85, + "grad_norm": 2.0345231315459094, + "learning_rate": 1.0851591939345108e-06, + "loss": 0.8822, + "step": 36284 + }, + { + "epoch": 0.85, + "grad_norm": 2.4419017159292014, + "learning_rate": 1.0848135225926248e-06, + "loss": 0.9533, + "step": 36285 + }, + { + "epoch": 0.85, + "grad_norm": 2.0596150789736076, + "learning_rate": 1.0844679031577754e-06, + "loss": 0.898, + "step": 36286 + }, + { + "epoch": 0.85, + "grad_norm": 1.9865905742825931, + "learning_rate": 1.0841223356319763e-06, + "loss": 0.8976, + "step": 36287 + }, + { + "epoch": 0.85, + "grad_norm": 1.895829477831349, + "learning_rate": 1.0837768200172417e-06, + "loss": 0.8936, + "step": 36288 + }, + { + "epoch": 0.85, + "grad_norm": 2.13929218605542, + "learning_rate": 1.0834313563155807e-06, + "loss": 1.0245, + "step": 36289 + }, + { + "epoch": 0.85, + "grad_norm": 1.9507849652038605, + "learning_rate": 1.0830859445290044e-06, + "loss": 1.0005, + "step": 36290 + }, + { + "epoch": 0.85, + "grad_norm": 2.1614242852523424, + "learning_rate": 1.0827405846595251e-06, + "loss": 0.9555, + "step": 36291 + }, + { + "epoch": 0.86, + "grad_norm": 2.001348992815327, + "learning_rate": 1.0823952767091527e-06, + "loss": 1.0267, + "step": 36292 + }, + { + "epoch": 0.86, + "grad_norm": 1.13399592975836, + "learning_rate": 1.0820500206799002e-06, + "loss": 0.9286, + "step": 36293 + }, + { + "epoch": 0.86, + "grad_norm": 1.951499510811123, + "learning_rate": 1.0817048165737732e-06, + "loss": 0.9959, + "step": 36294 + }, + { + "epoch": 0.86, + "grad_norm": 1.1627054787311022, + "learning_rate": 1.0813596643927871e-06, + "loss": 1.0479, + "step": 36295 + }, + { + "epoch": 0.86, + "grad_norm": 2.1099999537913043, + "learning_rate": 1.0810145641389458e-06, + "loss": 0.8521, + "step": 36296 + }, + { + "epoch": 0.86, + "grad_norm": 1.9321411819058405, + "learning_rate": 1.0806695158142633e-06, + "loss": 0.9404, + "step": 36297 + }, + { + "epoch": 0.86, + "grad_norm": 1.8859858030176513, + "learning_rate": 1.0803245194207468e-06, + "loss": 1.0684, + "step": 36298 + }, + { + "epoch": 0.86, + "grad_norm": 2.3865510666836407, + "learning_rate": 1.0799795749604037e-06, + "loss": 0.9655, + "step": 36299 + }, + { + "epoch": 0.86, + "grad_norm": 1.871178385437062, + "learning_rate": 1.0796346824352433e-06, + "loss": 0.8547, + "step": 36300 + }, + { + "epoch": 0.86, + "grad_norm": 1.7420656040366114, + "learning_rate": 1.0792898418472753e-06, + "loss": 0.9731, + "step": 36301 + }, + { + "epoch": 0.86, + "grad_norm": 1.8515764774879382, + "learning_rate": 1.0789450531985068e-06, + "loss": 1.0175, + "step": 36302 + }, + { + "epoch": 0.86, + "grad_norm": 2.0640401541631395, + "learning_rate": 1.078600316490942e-06, + "loss": 0.9103, + "step": 36303 + }, + { + "epoch": 0.86, + "grad_norm": 2.0827057244267584, + "learning_rate": 1.0782556317265902e-06, + "loss": 1.0244, + "step": 36304 + }, + { + "epoch": 0.86, + "grad_norm": 1.9300814187443107, + "learning_rate": 1.0779109989074621e-06, + "loss": 1.0046, + "step": 36305 + }, + { + "epoch": 0.86, + "grad_norm": 2.231138727403941, + "learning_rate": 1.0775664180355593e-06, + "loss": 0.9475, + "step": 36306 + }, + { + "epoch": 0.86, + "grad_norm": 1.993104673077417, + "learning_rate": 1.0772218891128882e-06, + "loss": 1.0542, + "step": 36307 + }, + { + "epoch": 0.86, + "grad_norm": 1.8974917696923066, + "learning_rate": 1.076877412141457e-06, + "loss": 0.9646, + "step": 36308 + }, + { + "epoch": 0.86, + "grad_norm": 1.9320588461894552, + "learning_rate": 1.076532987123271e-06, + "loss": 0.916, + "step": 36309 + }, + { + "epoch": 0.86, + "grad_norm": 1.8941696646569297, + "learning_rate": 1.0761886140603362e-06, + "loss": 0.9812, + "step": 36310 + }, + { + "epoch": 0.86, + "grad_norm": 1.9589876156358927, + "learning_rate": 1.0758442929546543e-06, + "loss": 0.9672, + "step": 36311 + }, + { + "epoch": 0.86, + "grad_norm": 1.9274237925702704, + "learning_rate": 1.0755000238082326e-06, + "loss": 1.0275, + "step": 36312 + }, + { + "epoch": 0.86, + "grad_norm": 1.9392930732019262, + "learning_rate": 1.075155806623076e-06, + "loss": 1.1043, + "step": 36313 + }, + { + "epoch": 0.86, + "grad_norm": 1.0393601122186733, + "learning_rate": 1.074811641401189e-06, + "loss": 0.8747, + "step": 36314 + }, + { + "epoch": 0.86, + "grad_norm": 1.8722405071481816, + "learning_rate": 1.0744675281445726e-06, + "loss": 0.9437, + "step": 36315 + }, + { + "epoch": 0.86, + "grad_norm": 1.1556859106384336, + "learning_rate": 1.0741234668552324e-06, + "loss": 0.9032, + "step": 36316 + }, + { + "epoch": 0.86, + "grad_norm": 2.033938274306164, + "learning_rate": 1.0737794575351712e-06, + "loss": 1.1336, + "step": 36317 + }, + { + "epoch": 0.86, + "grad_norm": 1.7861237912213535, + "learning_rate": 1.0734355001863904e-06, + "loss": 1.0147, + "step": 36318 + }, + { + "epoch": 0.86, + "grad_norm": 2.036629226602482, + "learning_rate": 1.0730915948108966e-06, + "loss": 1.0076, + "step": 36319 + }, + { + "epoch": 0.86, + "grad_norm": 1.8024288105926296, + "learning_rate": 1.0727477414106903e-06, + "loss": 1.0588, + "step": 36320 + }, + { + "epoch": 0.86, + "grad_norm": 2.1036606429167772, + "learning_rate": 1.0724039399877705e-06, + "loss": 0.9736, + "step": 36321 + }, + { + "epoch": 0.86, + "grad_norm": 2.0081995918385314, + "learning_rate": 1.0720601905441419e-06, + "loss": 0.9972, + "step": 36322 + }, + { + "epoch": 0.86, + "grad_norm": 1.7624380493256802, + "learning_rate": 1.071716493081809e-06, + "loss": 0.8836, + "step": 36323 + }, + { + "epoch": 0.86, + "grad_norm": 1.9925654152791261, + "learning_rate": 1.0713728476027653e-06, + "loss": 0.9854, + "step": 36324 + }, + { + "epoch": 0.86, + "grad_norm": 1.7816880231591927, + "learning_rate": 1.0710292541090161e-06, + "loss": 1.0749, + "step": 36325 + }, + { + "epoch": 0.86, + "grad_norm": 2.1699664817778097, + "learning_rate": 1.0706857126025604e-06, + "loss": 0.9609, + "step": 36326 + }, + { + "epoch": 0.86, + "grad_norm": 2.4014647517041965, + "learning_rate": 1.0703422230854022e-06, + "loss": 0.8538, + "step": 36327 + }, + { + "epoch": 0.86, + "grad_norm": 1.9352177533543193, + "learning_rate": 1.0699987855595384e-06, + "loss": 0.9062, + "step": 36328 + }, + { + "epoch": 0.86, + "grad_norm": 1.9695890232751874, + "learning_rate": 1.0696554000269677e-06, + "loss": 0.8962, + "step": 36329 + }, + { + "epoch": 0.86, + "grad_norm": 2.1556970775628845, + "learning_rate": 1.0693120664896916e-06, + "loss": 0.9312, + "step": 36330 + }, + { + "epoch": 0.86, + "grad_norm": 2.01748323344344, + "learning_rate": 1.0689687849497066e-06, + "loss": 0.9767, + "step": 36331 + }, + { + "epoch": 0.86, + "grad_norm": 2.264450233795534, + "learning_rate": 1.068625555409014e-06, + "loss": 0.9727, + "step": 36332 + }, + { + "epoch": 0.86, + "grad_norm": 1.0369596057760244, + "learning_rate": 1.0682823778696105e-06, + "loss": 0.9064, + "step": 36333 + }, + { + "epoch": 0.86, + "grad_norm": 1.2039129555386743, + "learning_rate": 1.0679392523334954e-06, + "loss": 0.9559, + "step": 36334 + }, + { + "epoch": 0.86, + "grad_norm": 2.3044816349006, + "learning_rate": 1.0675961788026646e-06, + "loss": 1.1322, + "step": 36335 + }, + { + "epoch": 0.86, + "grad_norm": 1.8572203565585714, + "learning_rate": 1.0672531572791178e-06, + "loss": 0.9675, + "step": 36336 + }, + { + "epoch": 0.86, + "grad_norm": 2.18674585826292, + "learning_rate": 1.0669101877648524e-06, + "loss": 1.1047, + "step": 36337 + }, + { + "epoch": 0.86, + "grad_norm": 1.9605796706429781, + "learning_rate": 1.0665672702618612e-06, + "loss": 1.0458, + "step": 36338 + }, + { + "epoch": 0.86, + "grad_norm": 1.884964994401077, + "learning_rate": 1.0662244047721437e-06, + "loss": 0.8534, + "step": 36339 + }, + { + "epoch": 0.86, + "grad_norm": 2.033590577477909, + "learning_rate": 1.0658815912976983e-06, + "loss": 0.9425, + "step": 36340 + }, + { + "epoch": 0.86, + "grad_norm": 2.1313919218454576, + "learning_rate": 1.065538829840519e-06, + "loss": 0.9902, + "step": 36341 + }, + { + "epoch": 0.86, + "grad_norm": 1.9028232643149645, + "learning_rate": 1.0651961204025995e-06, + "loss": 0.9759, + "step": 36342 + }, + { + "epoch": 0.86, + "grad_norm": 2.9922616458588087, + "learning_rate": 1.0648534629859363e-06, + "loss": 1.0204, + "step": 36343 + }, + { + "epoch": 0.86, + "grad_norm": 1.8338067826461002, + "learning_rate": 1.0645108575925267e-06, + "loss": 0.9157, + "step": 36344 + }, + { + "epoch": 0.86, + "grad_norm": 1.9645578544221238, + "learning_rate": 1.0641683042243645e-06, + "loss": 0.9166, + "step": 36345 + }, + { + "epoch": 0.86, + "grad_norm": 1.9737898038287167, + "learning_rate": 1.0638258028834415e-06, + "loss": 0.982, + "step": 36346 + }, + { + "epoch": 0.86, + "grad_norm": 1.9963566214261341, + "learning_rate": 1.0634833535717537e-06, + "loss": 0.9994, + "step": 36347 + }, + { + "epoch": 0.86, + "grad_norm": 1.929089445082065, + "learning_rate": 1.0631409562912975e-06, + "loss": 1.0424, + "step": 36348 + }, + { + "epoch": 0.86, + "grad_norm": 1.9545777457713596, + "learning_rate": 1.0627986110440637e-06, + "loss": 1.0505, + "step": 36349 + }, + { + "epoch": 0.86, + "grad_norm": 1.9015013915453145, + "learning_rate": 1.0624563178320435e-06, + "loss": 0.9522, + "step": 36350 + }, + { + "epoch": 0.86, + "grad_norm": 2.0611738078337445, + "learning_rate": 1.0621140766572336e-06, + "loss": 1.0559, + "step": 36351 + }, + { + "epoch": 0.86, + "grad_norm": 1.7723480231992281, + "learning_rate": 1.0617718875216242e-06, + "loss": 1.1207, + "step": 36352 + }, + { + "epoch": 0.86, + "grad_norm": 2.1764764365382447, + "learning_rate": 1.0614297504272108e-06, + "loss": 0.8122, + "step": 36353 + }, + { + "epoch": 0.86, + "grad_norm": 1.9030706154752588, + "learning_rate": 1.0610876653759805e-06, + "loss": 0.9916, + "step": 36354 + }, + { + "epoch": 0.86, + "grad_norm": 2.0729729072228897, + "learning_rate": 1.0607456323699306e-06, + "loss": 0.9887, + "step": 36355 + }, + { + "epoch": 0.86, + "grad_norm": 2.6377760522766094, + "learning_rate": 1.0604036514110471e-06, + "loss": 0.9839, + "step": 36356 + }, + { + "epoch": 0.86, + "grad_norm": 2.217899690244402, + "learning_rate": 1.0600617225013265e-06, + "loss": 0.9957, + "step": 36357 + }, + { + "epoch": 0.86, + "grad_norm": 1.7916302245244946, + "learning_rate": 1.059719845642756e-06, + "loss": 0.9373, + "step": 36358 + }, + { + "epoch": 0.86, + "grad_norm": 2.0692412957052446, + "learning_rate": 1.059378020837326e-06, + "loss": 1.0315, + "step": 36359 + }, + { + "epoch": 0.86, + "grad_norm": 1.910405033459369, + "learning_rate": 1.0590362480870275e-06, + "loss": 0.9444, + "step": 36360 + }, + { + "epoch": 0.86, + "grad_norm": 2.1690098201368495, + "learning_rate": 1.058694527393852e-06, + "loss": 0.9111, + "step": 36361 + }, + { + "epoch": 0.86, + "grad_norm": 2.9607417933976325, + "learning_rate": 1.058352858759788e-06, + "loss": 0.9264, + "step": 36362 + }, + { + "epoch": 0.86, + "grad_norm": 2.0542653987031363, + "learning_rate": 1.0580112421868226e-06, + "loss": 0.9693, + "step": 36363 + }, + { + "epoch": 0.86, + "grad_norm": 2.187432689461365, + "learning_rate": 1.0576696776769469e-06, + "loss": 0.986, + "step": 36364 + }, + { + "epoch": 0.86, + "grad_norm": 1.9389376675039693, + "learning_rate": 1.057328165232151e-06, + "loss": 0.9659, + "step": 36365 + }, + { + "epoch": 0.86, + "grad_norm": 1.8871562311853634, + "learning_rate": 1.0569867048544225e-06, + "loss": 1.0459, + "step": 36366 + }, + { + "epoch": 0.86, + "grad_norm": 2.93167484533185, + "learning_rate": 1.0566452965457452e-06, + "loss": 1.0018, + "step": 36367 + }, + { + "epoch": 0.86, + "grad_norm": 1.1073967366922743, + "learning_rate": 1.056303940308112e-06, + "loss": 0.9495, + "step": 36368 + }, + { + "epoch": 0.86, + "grad_norm": 3.0794028361775907, + "learning_rate": 1.0559626361435104e-06, + "loss": 0.9619, + "step": 36369 + }, + { + "epoch": 0.86, + "grad_norm": 2.59276109439261, + "learning_rate": 1.0556213840539253e-06, + "loss": 0.9226, + "step": 36370 + }, + { + "epoch": 0.86, + "grad_norm": 1.9380156509569675, + "learning_rate": 1.0552801840413428e-06, + "loss": 0.9478, + "step": 36371 + }, + { + "epoch": 0.86, + "grad_norm": 1.8667593330654175, + "learning_rate": 1.0549390361077527e-06, + "loss": 0.954, + "step": 36372 + }, + { + "epoch": 0.86, + "grad_norm": 2.103594341948108, + "learning_rate": 1.0545979402551375e-06, + "loss": 1.0173, + "step": 36373 + }, + { + "epoch": 0.86, + "grad_norm": 1.9981872029448882, + "learning_rate": 1.054256896485487e-06, + "loss": 0.9562, + "step": 36374 + }, + { + "epoch": 0.86, + "grad_norm": 1.9057720120763117, + "learning_rate": 1.0539159048007842e-06, + "loss": 0.9743, + "step": 36375 + }, + { + "epoch": 0.86, + "grad_norm": 2.105502775241119, + "learning_rate": 1.0535749652030159e-06, + "loss": 0.9937, + "step": 36376 + }, + { + "epoch": 0.86, + "grad_norm": 1.8729636487828099, + "learning_rate": 1.0532340776941653e-06, + "loss": 1.0103, + "step": 36377 + }, + { + "epoch": 0.86, + "grad_norm": 2.6926223253491726, + "learning_rate": 1.0528932422762195e-06, + "loss": 0.8976, + "step": 36378 + }, + { + "epoch": 0.86, + "grad_norm": 1.7836751267488862, + "learning_rate": 1.0525524589511627e-06, + "loss": 1.1163, + "step": 36379 + }, + { + "epoch": 0.86, + "grad_norm": 1.9742834044275404, + "learning_rate": 1.0522117277209764e-06, + "loss": 0.9113, + "step": 36380 + }, + { + "epoch": 0.86, + "grad_norm": 1.986690125819842, + "learning_rate": 1.051871048587646e-06, + "loss": 0.8646, + "step": 36381 + }, + { + "epoch": 0.86, + "grad_norm": 1.8045609724496425, + "learning_rate": 1.0515304215531563e-06, + "loss": 0.9232, + "step": 36382 + }, + { + "epoch": 0.86, + "grad_norm": 2.4306325703531466, + "learning_rate": 1.0511898466194903e-06, + "loss": 0.9869, + "step": 36383 + }, + { + "epoch": 0.86, + "grad_norm": 2.1506472848239206, + "learning_rate": 1.0508493237886286e-06, + "loss": 0.9917, + "step": 36384 + }, + { + "epoch": 0.86, + "grad_norm": 1.9990791643581365, + "learning_rate": 1.050508853062555e-06, + "loss": 1.0441, + "step": 36385 + }, + { + "epoch": 0.86, + "grad_norm": 2.055786451062559, + "learning_rate": 1.0501684344432538e-06, + "loss": 0.8518, + "step": 36386 + }, + { + "epoch": 0.86, + "grad_norm": 1.1143224038836619, + "learning_rate": 1.0498280679327067e-06, + "loss": 0.9613, + "step": 36387 + }, + { + "epoch": 0.86, + "grad_norm": 1.9277884339035305, + "learning_rate": 1.0494877535328907e-06, + "loss": 0.9844, + "step": 36388 + }, + { + "epoch": 0.86, + "grad_norm": 2.005215776636606, + "learning_rate": 1.049147491245791e-06, + "loss": 1.1015, + "step": 36389 + }, + { + "epoch": 0.86, + "grad_norm": 2.124329807339796, + "learning_rate": 1.0488072810733906e-06, + "loss": 1.0249, + "step": 36390 + }, + { + "epoch": 0.86, + "grad_norm": 1.0685265137267252, + "learning_rate": 1.048467123017669e-06, + "loss": 0.9119, + "step": 36391 + }, + { + "epoch": 0.86, + "grad_norm": 2.0967185095029954, + "learning_rate": 1.0481270170806024e-06, + "loss": 1.0576, + "step": 36392 + }, + { + "epoch": 0.86, + "grad_norm": 2.012418596628114, + "learning_rate": 1.0477869632641769e-06, + "loss": 0.9588, + "step": 36393 + }, + { + "epoch": 0.86, + "grad_norm": 2.2028108455883704, + "learning_rate": 1.0474469615703686e-06, + "loss": 0.978, + "step": 36394 + }, + { + "epoch": 0.86, + "grad_norm": 2.018618453889062, + "learning_rate": 1.0471070120011595e-06, + "loss": 0.8706, + "step": 36395 + }, + { + "epoch": 0.86, + "grad_norm": 2.895541588295181, + "learning_rate": 1.0467671145585268e-06, + "loss": 0.9848, + "step": 36396 + }, + { + "epoch": 0.86, + "grad_norm": 1.1885848254635625, + "learning_rate": 1.0464272692444511e-06, + "loss": 0.9988, + "step": 36397 + }, + { + "epoch": 0.86, + "grad_norm": 1.9206637381911598, + "learning_rate": 1.0460874760609096e-06, + "loss": 0.9395, + "step": 36398 + }, + { + "epoch": 0.86, + "grad_norm": 1.0917421250016583, + "learning_rate": 1.0457477350098832e-06, + "loss": 0.8992, + "step": 36399 + }, + { + "epoch": 0.86, + "grad_norm": 1.9262127956162414, + "learning_rate": 1.045408046093348e-06, + "loss": 0.9852, + "step": 36400 + }, + { + "epoch": 0.86, + "grad_norm": 2.0148953348000185, + "learning_rate": 1.0450684093132813e-06, + "loss": 0.9844, + "step": 36401 + }, + { + "epoch": 0.86, + "grad_norm": 2.40967591438027, + "learning_rate": 1.0447288246716602e-06, + "loss": 1.0082, + "step": 36402 + }, + { + "epoch": 0.86, + "grad_norm": 1.0336321967857507, + "learning_rate": 1.0443892921704635e-06, + "loss": 1.0365, + "step": 36403 + }, + { + "epoch": 0.86, + "grad_norm": 2.0816728125030957, + "learning_rate": 1.0440498118116715e-06, + "loss": 0.8839, + "step": 36404 + }, + { + "epoch": 0.86, + "grad_norm": 2.0192878581550584, + "learning_rate": 1.0437103835972528e-06, + "loss": 0.7904, + "step": 36405 + }, + { + "epoch": 0.86, + "grad_norm": 2.016950856339696, + "learning_rate": 1.043371007529188e-06, + "loss": 0.9689, + "step": 36406 + }, + { + "epoch": 0.86, + "grad_norm": 1.928682917379301, + "learning_rate": 1.0430316836094555e-06, + "loss": 1.0839, + "step": 36407 + }, + { + "epoch": 0.86, + "grad_norm": 1.9269569845545649, + "learning_rate": 1.0426924118400272e-06, + "loss": 1.0407, + "step": 36408 + }, + { + "epoch": 0.86, + "grad_norm": 1.9611608298047691, + "learning_rate": 1.042353192222878e-06, + "loss": 0.9503, + "step": 36409 + }, + { + "epoch": 0.86, + "grad_norm": 2.0382797873759695, + "learning_rate": 1.0420140247599842e-06, + "loss": 0.964, + "step": 36410 + }, + { + "epoch": 0.86, + "grad_norm": 2.0877081896821688, + "learning_rate": 1.041674909453323e-06, + "loss": 0.9551, + "step": 36411 + }, + { + "epoch": 0.86, + "grad_norm": 1.9881922739660975, + "learning_rate": 1.0413358463048674e-06, + "loss": 0.9529, + "step": 36412 + }, + { + "epoch": 0.86, + "grad_norm": 2.0837692637508134, + "learning_rate": 1.0409968353165878e-06, + "loss": 1.001, + "step": 36413 + }, + { + "epoch": 0.86, + "grad_norm": 2.129784220364222, + "learning_rate": 1.040657876490463e-06, + "loss": 1.062, + "step": 36414 + }, + { + "epoch": 0.86, + "grad_norm": 2.4452102544732264, + "learning_rate": 1.040318969828462e-06, + "loss": 0.8466, + "step": 36415 + }, + { + "epoch": 0.86, + "grad_norm": 2.2024556111731974, + "learning_rate": 1.0399801153325616e-06, + "loss": 1.0057, + "step": 36416 + }, + { + "epoch": 0.86, + "grad_norm": 1.987421392135167, + "learning_rate": 1.0396413130047356e-06, + "loss": 0.9394, + "step": 36417 + }, + { + "epoch": 0.86, + "grad_norm": 1.9483055887207623, + "learning_rate": 1.0393025628469545e-06, + "loss": 0.9473, + "step": 36418 + }, + { + "epoch": 0.86, + "grad_norm": 2.0530137071418357, + "learning_rate": 1.038963864861189e-06, + "loss": 1.0258, + "step": 36419 + }, + { + "epoch": 0.86, + "grad_norm": 1.910783743435691, + "learning_rate": 1.0386252190494117e-06, + "loss": 1.0465, + "step": 36420 + }, + { + "epoch": 0.86, + "grad_norm": 2.1110952996182, + "learning_rate": 1.0382866254136004e-06, + "loss": 0.8194, + "step": 36421 + }, + { + "epoch": 0.86, + "grad_norm": 2.09554054877518, + "learning_rate": 1.0379480839557177e-06, + "loss": 0.9886, + "step": 36422 + }, + { + "epoch": 0.86, + "grad_norm": 1.9653376934609679, + "learning_rate": 1.0376095946777387e-06, + "loss": 0.9843, + "step": 36423 + }, + { + "epoch": 0.86, + "grad_norm": 2.414042791614821, + "learning_rate": 1.0372711575816341e-06, + "loss": 1.0285, + "step": 36424 + }, + { + "epoch": 0.86, + "grad_norm": 1.0569808409164418, + "learning_rate": 1.0369327726693757e-06, + "loss": 0.9147, + "step": 36425 + }, + { + "epoch": 0.86, + "grad_norm": 2.1956815102961733, + "learning_rate": 1.0365944399429327e-06, + "loss": 0.9064, + "step": 36426 + }, + { + "epoch": 0.86, + "grad_norm": 2.229288684038843, + "learning_rate": 1.036256159404273e-06, + "loss": 0.9096, + "step": 36427 + }, + { + "epoch": 0.86, + "grad_norm": 2.0495292920199737, + "learning_rate": 1.035917931055369e-06, + "loss": 0.9411, + "step": 36428 + }, + { + "epoch": 0.86, + "grad_norm": 1.7676673118473132, + "learning_rate": 1.035579754898187e-06, + "loss": 0.8963, + "step": 36429 + }, + { + "epoch": 0.86, + "grad_norm": 1.9930957529463553, + "learning_rate": 1.0352416309347003e-06, + "loss": 0.8407, + "step": 36430 + }, + { + "epoch": 0.86, + "grad_norm": 1.835423499934631, + "learning_rate": 1.0349035591668732e-06, + "loss": 0.9634, + "step": 36431 + }, + { + "epoch": 0.86, + "grad_norm": 1.8849198546333361, + "learning_rate": 1.034565539596677e-06, + "loss": 0.9288, + "step": 36432 + }, + { + "epoch": 0.86, + "grad_norm": 1.9685722168623478, + "learning_rate": 1.0342275722260776e-06, + "loss": 0.9029, + "step": 36433 + }, + { + "epoch": 0.86, + "grad_norm": 2.1481089928204447, + "learning_rate": 1.0338896570570456e-06, + "loss": 0.9855, + "step": 36434 + }, + { + "epoch": 0.86, + "grad_norm": 2.194346539939049, + "learning_rate": 1.0335517940915441e-06, + "loss": 0.9428, + "step": 36435 + }, + { + "epoch": 0.86, + "grad_norm": 1.927191942192101, + "learning_rate": 1.0332139833315447e-06, + "loss": 1.0459, + "step": 36436 + }, + { + "epoch": 0.86, + "grad_norm": 1.859828219555225, + "learning_rate": 1.0328762247790103e-06, + "loss": 0.9983, + "step": 36437 + }, + { + "epoch": 0.86, + "grad_norm": 2.2854630686116946, + "learning_rate": 1.0325385184359126e-06, + "loss": 0.951, + "step": 36438 + }, + { + "epoch": 0.86, + "grad_norm": 2.0259544725546004, + "learning_rate": 1.0322008643042136e-06, + "loss": 1.189, + "step": 36439 + }, + { + "epoch": 0.86, + "grad_norm": 2.255855011935783, + "learning_rate": 1.031863262385878e-06, + "loss": 0.9968, + "step": 36440 + }, + { + "epoch": 0.86, + "grad_norm": 1.8861489078357985, + "learning_rate": 1.0315257126828749e-06, + "loss": 1.1097, + "step": 36441 + }, + { + "epoch": 0.86, + "grad_norm": 1.9928322466909543, + "learning_rate": 1.0311882151971697e-06, + "loss": 0.8942, + "step": 36442 + }, + { + "epoch": 0.86, + "grad_norm": 1.900257744081115, + "learning_rate": 1.0308507699307268e-06, + "loss": 0.9059, + "step": 36443 + }, + { + "epoch": 0.86, + "grad_norm": 2.0429939581645127, + "learning_rate": 1.030513376885508e-06, + "loss": 0.8479, + "step": 36444 + }, + { + "epoch": 0.86, + "grad_norm": 1.0534466740442283, + "learning_rate": 1.0301760360634794e-06, + "loss": 0.8317, + "step": 36445 + }, + { + "epoch": 0.86, + "grad_norm": 2.909593326893443, + "learning_rate": 1.0298387474666083e-06, + "loss": 0.9528, + "step": 36446 + }, + { + "epoch": 0.86, + "grad_norm": 2.08112888145745, + "learning_rate": 1.0295015110968564e-06, + "loss": 0.8871, + "step": 36447 + }, + { + "epoch": 0.86, + "grad_norm": 1.7910896060961266, + "learning_rate": 1.0291643269561835e-06, + "loss": 0.9615, + "step": 36448 + }, + { + "epoch": 0.86, + "grad_norm": 1.1699464003128894, + "learning_rate": 1.0288271950465566e-06, + "loss": 0.9453, + "step": 36449 + }, + { + "epoch": 0.86, + "grad_norm": 1.9457751182673049, + "learning_rate": 1.0284901153699412e-06, + "loss": 1.0211, + "step": 36450 + }, + { + "epoch": 0.86, + "grad_norm": 1.9654752297684683, + "learning_rate": 1.0281530879282953e-06, + "loss": 0.9862, + "step": 36451 + }, + { + "epoch": 0.86, + "grad_norm": 3.6537176213529854, + "learning_rate": 1.0278161127235808e-06, + "loss": 0.9567, + "step": 36452 + }, + { + "epoch": 0.86, + "grad_norm": 2.0639817490076013, + "learning_rate": 1.027479189757763e-06, + "loss": 1.0172, + "step": 36453 + }, + { + "epoch": 0.86, + "grad_norm": 2.0333532533085563, + "learning_rate": 1.027142319032801e-06, + "loss": 0.9618, + "step": 36454 + }, + { + "epoch": 0.86, + "grad_norm": 1.9349434966607317, + "learning_rate": 1.0268055005506583e-06, + "loss": 1.0139, + "step": 36455 + }, + { + "epoch": 0.86, + "grad_norm": 1.0599019424867149, + "learning_rate": 1.0264687343132928e-06, + "loss": 0.9239, + "step": 36456 + }, + { + "epoch": 0.86, + "grad_norm": 1.9576156207969644, + "learning_rate": 1.02613202032267e-06, + "loss": 0.8505, + "step": 36457 + }, + { + "epoch": 0.86, + "grad_norm": 2.0534584458865823, + "learning_rate": 1.0257953585807456e-06, + "loss": 0.9184, + "step": 36458 + }, + { + "epoch": 0.86, + "grad_norm": 2.0284568580413525, + "learning_rate": 1.025458749089484e-06, + "loss": 0.8045, + "step": 36459 + }, + { + "epoch": 0.86, + "grad_norm": 1.0876838369409751, + "learning_rate": 1.0251221918508425e-06, + "loss": 0.9338, + "step": 36460 + }, + { + "epoch": 0.86, + "grad_norm": 2.574188753946035, + "learning_rate": 1.0247856868667782e-06, + "loss": 0.9312, + "step": 36461 + }, + { + "epoch": 0.86, + "grad_norm": 2.373629098275809, + "learning_rate": 1.0244492341392543e-06, + "loss": 1.0732, + "step": 36462 + }, + { + "epoch": 0.86, + "grad_norm": 1.9784065689126786, + "learning_rate": 1.02411283367023e-06, + "loss": 0.8372, + "step": 36463 + }, + { + "epoch": 0.86, + "grad_norm": 2.2011745727035117, + "learning_rate": 1.0237764854616628e-06, + "loss": 0.9409, + "step": 36464 + }, + { + "epoch": 0.86, + "grad_norm": 1.93802628151912, + "learning_rate": 1.0234401895155076e-06, + "loss": 1.0073, + "step": 36465 + }, + { + "epoch": 0.86, + "grad_norm": 2.0501846418632717, + "learning_rate": 1.0231039458337267e-06, + "loss": 0.9781, + "step": 36466 + }, + { + "epoch": 0.86, + "grad_norm": 1.0848671117621052, + "learning_rate": 1.022767754418279e-06, + "loss": 0.8611, + "step": 36467 + }, + { + "epoch": 0.86, + "grad_norm": 1.9172385374348186, + "learning_rate": 1.022431615271119e-06, + "loss": 1.0571, + "step": 36468 + }, + { + "epoch": 0.86, + "grad_norm": 1.8222592836425036, + "learning_rate": 1.0220955283942024e-06, + "loss": 0.9417, + "step": 36469 + }, + { + "epoch": 0.86, + "grad_norm": 1.9809437006858448, + "learning_rate": 1.021759493789488e-06, + "loss": 1.0306, + "step": 36470 + }, + { + "epoch": 0.86, + "grad_norm": 1.9087294756268567, + "learning_rate": 1.0214235114589344e-06, + "loss": 1.0303, + "step": 36471 + }, + { + "epoch": 0.86, + "grad_norm": 16.644369130046144, + "learning_rate": 1.021087581404495e-06, + "loss": 1.0769, + "step": 36472 + }, + { + "epoch": 0.86, + "grad_norm": 2.180672483261393, + "learning_rate": 1.0207517036281256e-06, + "loss": 0.9262, + "step": 36473 + }, + { + "epoch": 0.86, + "grad_norm": 1.136524586853971, + "learning_rate": 1.0204158781317841e-06, + "loss": 0.9269, + "step": 36474 + }, + { + "epoch": 0.86, + "grad_norm": 2.31116145298853, + "learning_rate": 1.0200801049174214e-06, + "loss": 0.9674, + "step": 36475 + }, + { + "epoch": 0.86, + "grad_norm": 2.2946412892982995, + "learning_rate": 1.019744383986998e-06, + "loss": 0.9302, + "step": 36476 + }, + { + "epoch": 0.86, + "grad_norm": 1.1506970699519603, + "learning_rate": 1.0194087153424648e-06, + "loss": 0.9916, + "step": 36477 + }, + { + "epoch": 0.86, + "grad_norm": 1.9906863400382044, + "learning_rate": 1.0190730989857777e-06, + "loss": 0.9155, + "step": 36478 + }, + { + "epoch": 0.86, + "grad_norm": 2.0107246457678407, + "learning_rate": 1.0187375349188887e-06, + "loss": 0.9768, + "step": 36479 + }, + { + "epoch": 0.86, + "grad_norm": 1.9341058270955926, + "learning_rate": 1.0184020231437564e-06, + "loss": 0.9798, + "step": 36480 + }, + { + "epoch": 0.86, + "grad_norm": 1.9055833388521322, + "learning_rate": 1.01806656366233e-06, + "loss": 1.1085, + "step": 36481 + }, + { + "epoch": 0.86, + "grad_norm": 2.047609086000188, + "learning_rate": 1.0177311564765612e-06, + "loss": 0.9429, + "step": 36482 + }, + { + "epoch": 0.86, + "grad_norm": 1.9102956467088374, + "learning_rate": 1.0173958015884066e-06, + "loss": 1.099, + "step": 36483 + }, + { + "epoch": 0.86, + "grad_norm": 2.064078456072404, + "learning_rate": 1.0170604989998189e-06, + "loss": 0.9741, + "step": 36484 + }, + { + "epoch": 0.86, + "grad_norm": 2.0086893538856287, + "learning_rate": 1.0167252487127488e-06, + "loss": 0.8788, + "step": 36485 + }, + { + "epoch": 0.86, + "grad_norm": 2.21701251506182, + "learning_rate": 1.0163900507291457e-06, + "loss": 1.0309, + "step": 36486 + }, + { + "epoch": 0.86, + "grad_norm": 2.2847673229338237, + "learning_rate": 1.016054905050965e-06, + "loss": 0.9742, + "step": 36487 + }, + { + "epoch": 0.86, + "grad_norm": 1.8310553689955398, + "learning_rate": 1.0157198116801592e-06, + "loss": 0.9818, + "step": 36488 + }, + { + "epoch": 0.86, + "grad_norm": 2.971396813819151, + "learning_rate": 1.015384770618676e-06, + "loss": 1.0615, + "step": 36489 + }, + { + "epoch": 0.86, + "grad_norm": 2.0647338040977594, + "learning_rate": 1.0150497818684657e-06, + "loss": 0.8965, + "step": 36490 + }, + { + "epoch": 0.86, + "grad_norm": 2.015270274424926, + "learning_rate": 1.0147148454314792e-06, + "loss": 0.9628, + "step": 36491 + }, + { + "epoch": 0.86, + "grad_norm": 1.982032026753858, + "learning_rate": 1.0143799613096706e-06, + "loss": 0.8901, + "step": 36492 + }, + { + "epoch": 0.86, + "grad_norm": 2.629785663678066, + "learning_rate": 1.014045129504987e-06, + "loss": 1.018, + "step": 36493 + }, + { + "epoch": 0.86, + "grad_norm": 2.2702741863689457, + "learning_rate": 1.0137103500193746e-06, + "loss": 1.0566, + "step": 36494 + }, + { + "epoch": 0.86, + "grad_norm": 1.9007426187200245, + "learning_rate": 1.0133756228547875e-06, + "loss": 0.8919, + "step": 36495 + }, + { + "epoch": 0.86, + "grad_norm": 2.0894375308245245, + "learning_rate": 1.013040948013172e-06, + "loss": 0.95, + "step": 36496 + }, + { + "epoch": 0.86, + "grad_norm": 1.9084232441538522, + "learning_rate": 1.0127063254964776e-06, + "loss": 0.9249, + "step": 36497 + }, + { + "epoch": 0.86, + "grad_norm": 1.9422923375544094, + "learning_rate": 1.0123717553066515e-06, + "loss": 1.0136, + "step": 36498 + }, + { + "epoch": 0.86, + "grad_norm": 2.0922957074000568, + "learning_rate": 1.0120372374456445e-06, + "loss": 0.8863, + "step": 36499 + }, + { + "epoch": 0.86, + "grad_norm": 2.0423149807429115, + "learning_rate": 1.0117027719153993e-06, + "loss": 1.0075, + "step": 36500 + }, + { + "epoch": 0.86, + "grad_norm": 2.402672745499241, + "learning_rate": 1.011368358717867e-06, + "loss": 0.9145, + "step": 36501 + }, + { + "epoch": 0.86, + "grad_norm": 2.0549171903178522, + "learning_rate": 1.0110339978549977e-06, + "loss": 1.0171, + "step": 36502 + }, + { + "epoch": 0.86, + "grad_norm": 2.1721183599560896, + "learning_rate": 1.0106996893287303e-06, + "loss": 0.9404, + "step": 36503 + }, + { + "epoch": 0.86, + "grad_norm": 1.9180457909158237, + "learning_rate": 1.010365433141015e-06, + "loss": 1.0338, + "step": 36504 + }, + { + "epoch": 0.86, + "grad_norm": 1.0162375354456992, + "learning_rate": 1.0100312292937986e-06, + "loss": 0.9661, + "step": 36505 + }, + { + "epoch": 0.86, + "grad_norm": 1.9529694811178084, + "learning_rate": 1.0096970777890302e-06, + "loss": 1.0252, + "step": 36506 + }, + { + "epoch": 0.86, + "grad_norm": 1.9865109630636928, + "learning_rate": 1.0093629786286475e-06, + "loss": 0.9256, + "step": 36507 + }, + { + "epoch": 0.86, + "grad_norm": 2.001500773787896, + "learning_rate": 1.0090289318146007e-06, + "loss": 1.002, + "step": 36508 + }, + { + "epoch": 0.86, + "grad_norm": 1.9711685460456265, + "learning_rate": 1.0086949373488352e-06, + "loss": 0.965, + "step": 36509 + }, + { + "epoch": 0.86, + "grad_norm": 1.9739120275628521, + "learning_rate": 1.0083609952332928e-06, + "loss": 0.8971, + "step": 36510 + }, + { + "epoch": 0.86, + "grad_norm": 2.094237949387321, + "learning_rate": 1.0080271054699208e-06, + "loss": 1.0757, + "step": 36511 + }, + { + "epoch": 0.86, + "grad_norm": 2.0887447735838154, + "learning_rate": 1.007693268060661e-06, + "loss": 1.0136, + "step": 36512 + }, + { + "epoch": 0.86, + "grad_norm": 3.221156961465914, + "learning_rate": 1.0073594830074596e-06, + "loss": 1.0281, + "step": 36513 + }, + { + "epoch": 0.86, + "grad_norm": 1.9724035135977342, + "learning_rate": 1.007025750312256e-06, + "loss": 0.9205, + "step": 36514 + }, + { + "epoch": 0.86, + "grad_norm": 2.125295803399639, + "learning_rate": 1.0066920699769966e-06, + "loss": 0.8192, + "step": 36515 + }, + { + "epoch": 0.86, + "grad_norm": 1.9805056144830966, + "learning_rate": 1.0063584420036243e-06, + "loss": 0.9347, + "step": 36516 + }, + { + "epoch": 0.86, + "grad_norm": 1.977710800131164, + "learning_rate": 1.0060248663940787e-06, + "loss": 1.1007, + "step": 36517 + }, + { + "epoch": 0.86, + "grad_norm": 2.5252500219804177, + "learning_rate": 1.0056913431503035e-06, + "loss": 1.0534, + "step": 36518 + }, + { + "epoch": 0.86, + "grad_norm": 1.9216415729449414, + "learning_rate": 1.005357872274243e-06, + "loss": 0.9861, + "step": 36519 + }, + { + "epoch": 0.86, + "grad_norm": 2.093064013528584, + "learning_rate": 1.0050244537678356e-06, + "loss": 1.0959, + "step": 36520 + }, + { + "epoch": 0.86, + "grad_norm": 2.2071132235368416, + "learning_rate": 1.004691087633023e-06, + "loss": 0.9681, + "step": 36521 + }, + { + "epoch": 0.86, + "grad_norm": 2.137529606036373, + "learning_rate": 1.0043577738717459e-06, + "loss": 0.9063, + "step": 36522 + }, + { + "epoch": 0.86, + "grad_norm": 2.712065972176351, + "learning_rate": 1.0040245124859483e-06, + "loss": 0.8989, + "step": 36523 + }, + { + "epoch": 0.86, + "grad_norm": 1.9351343688259088, + "learning_rate": 1.0036913034775675e-06, + "loss": 0.896, + "step": 36524 + }, + { + "epoch": 0.86, + "grad_norm": 1.82680129485115, + "learning_rate": 1.0033581468485422e-06, + "loss": 1.0532, + "step": 36525 + }, + { + "epoch": 0.86, + "grad_norm": 2.1942824186140264, + "learning_rate": 1.0030250426008148e-06, + "loss": 0.9385, + "step": 36526 + }, + { + "epoch": 0.86, + "grad_norm": 1.8835227741266931, + "learning_rate": 1.0026919907363252e-06, + "loss": 1.0263, + "step": 36527 + }, + { + "epoch": 0.86, + "grad_norm": 1.0038385477411151, + "learning_rate": 1.002358991257012e-06, + "loss": 0.9379, + "step": 36528 + }, + { + "epoch": 0.86, + "grad_norm": 1.0547264140508923, + "learning_rate": 1.0020260441648111e-06, + "loss": 0.9626, + "step": 36529 + }, + { + "epoch": 0.86, + "grad_norm": 1.0665613335024091, + "learning_rate": 1.0016931494616644e-06, + "loss": 0.9691, + "step": 36530 + }, + { + "epoch": 0.86, + "grad_norm": 1.859743194028162, + "learning_rate": 1.001360307149508e-06, + "loss": 0.9857, + "step": 36531 + }, + { + "epoch": 0.86, + "grad_norm": 2.1805057198017326, + "learning_rate": 1.001027517230283e-06, + "loss": 0.8709, + "step": 36532 + }, + { + "epoch": 0.86, + "grad_norm": 1.8001093522491056, + "learning_rate": 1.000694779705922e-06, + "loss": 0.9296, + "step": 36533 + }, + { + "epoch": 0.86, + "grad_norm": 1.9396352412328337, + "learning_rate": 1.0003620945783676e-06, + "loss": 0.9758, + "step": 36534 + }, + { + "epoch": 0.86, + "grad_norm": 2.067603324748821, + "learning_rate": 1.000029461849552e-06, + "loss": 0.9378, + "step": 36535 + }, + { + "epoch": 0.86, + "grad_norm": 1.9900313674874577, + "learning_rate": 9.99696881521417e-07, + "loss": 0.9942, + "step": 36536 + }, + { + "epoch": 0.86, + "grad_norm": 1.816069190608104, + "learning_rate": 9.993643535958952e-07, + "loss": 0.8897, + "step": 36537 + }, + { + "epoch": 0.86, + "grad_norm": 1.9602897095782605, + "learning_rate": 9.990318780749219e-07, + "loss": 0.9474, + "step": 36538 + }, + { + "epoch": 0.86, + "grad_norm": 2.21615869146892, + "learning_rate": 9.986994549604346e-07, + "loss": 0.9374, + "step": 36539 + }, + { + "epoch": 0.86, + "grad_norm": 1.0446724514873356, + "learning_rate": 9.983670842543713e-07, + "loss": 0.9056, + "step": 36540 + }, + { + "epoch": 0.86, + "grad_norm": 1.8605129788925356, + "learning_rate": 9.980347659586643e-07, + "loss": 0.9768, + "step": 36541 + }, + { + "epoch": 0.86, + "grad_norm": 2.2304437543233395, + "learning_rate": 9.977025000752472e-07, + "loss": 0.9466, + "step": 36542 + }, + { + "epoch": 0.86, + "grad_norm": 4.7221614193715595, + "learning_rate": 9.973702866060565e-07, + "loss": 1.0879, + "step": 36543 + }, + { + "epoch": 0.86, + "grad_norm": 1.9130529103999052, + "learning_rate": 9.97038125553028e-07, + "loss": 1.0566, + "step": 36544 + }, + { + "epoch": 0.86, + "grad_norm": 1.9786072202735299, + "learning_rate": 9.96706016918093e-07, + "loss": 0.9904, + "step": 36545 + }, + { + "epoch": 0.86, + "grad_norm": 1.8767377317843816, + "learning_rate": 9.963739607031852e-07, + "loss": 1.0571, + "step": 36546 + }, + { + "epoch": 0.86, + "grad_norm": 1.914354350625768, + "learning_rate": 9.960419569102375e-07, + "loss": 0.9099, + "step": 36547 + }, + { + "epoch": 0.86, + "grad_norm": 1.8915608644907738, + "learning_rate": 9.957100055411872e-07, + "loss": 1.0282, + "step": 36548 + }, + { + "epoch": 0.86, + "grad_norm": 2.1029875001239575, + "learning_rate": 9.95378106597964e-07, + "loss": 0.9823, + "step": 36549 + }, + { + "epoch": 0.86, + "grad_norm": 1.1711788879612322, + "learning_rate": 9.950462600824973e-07, + "loss": 1.0334, + "step": 36550 + }, + { + "epoch": 0.86, + "grad_norm": 1.9758760536950788, + "learning_rate": 9.947144659967245e-07, + "loss": 0.9014, + "step": 36551 + }, + { + "epoch": 0.86, + "grad_norm": 2.2232650695462057, + "learning_rate": 9.94382724342573e-07, + "loss": 0.908, + "step": 36552 + }, + { + "epoch": 0.86, + "grad_norm": 2.0367480059901393, + "learning_rate": 9.94051035121979e-07, + "loss": 0.9285, + "step": 36553 + }, + { + "epoch": 0.86, + "grad_norm": 2.1296439055431304, + "learning_rate": 9.937193983368688e-07, + "loss": 0.9109, + "step": 36554 + }, + { + "epoch": 0.86, + "grad_norm": 2.056780593905582, + "learning_rate": 9.933878139891772e-07, + "loss": 1.0715, + "step": 36555 + }, + { + "epoch": 0.86, + "grad_norm": 1.8967393872846439, + "learning_rate": 9.93056282080832e-07, + "loss": 1.0487, + "step": 36556 + }, + { + "epoch": 0.86, + "grad_norm": 2.06443178293065, + "learning_rate": 9.92724802613766e-07, + "loss": 0.9898, + "step": 36557 + }, + { + "epoch": 0.86, + "grad_norm": 2.1521524181583604, + "learning_rate": 9.923933755899062e-07, + "loss": 1.0282, + "step": 36558 + }, + { + "epoch": 0.86, + "grad_norm": 2.1899893807522246, + "learning_rate": 9.920620010111858e-07, + "loss": 0.896, + "step": 36559 + }, + { + "epoch": 0.86, + "grad_norm": 2.0668807081079765, + "learning_rate": 9.9173067887953e-07, + "loss": 1.031, + "step": 36560 + }, + { + "epoch": 0.86, + "grad_norm": 1.0349518153462958, + "learning_rate": 9.913994091968725e-07, + "loss": 0.9809, + "step": 36561 + }, + { + "epoch": 0.86, + "grad_norm": 1.8860228609046656, + "learning_rate": 9.910681919651399e-07, + "loss": 0.9852, + "step": 36562 + }, + { + "epoch": 0.86, + "grad_norm": 1.966910464736629, + "learning_rate": 9.907370271862582e-07, + "loss": 1.0178, + "step": 36563 + }, + { + "epoch": 0.86, + "grad_norm": 2.09736429401307, + "learning_rate": 9.904059148621581e-07, + "loss": 1.0055, + "step": 36564 + }, + { + "epoch": 0.86, + "grad_norm": 3.4627889119619386, + "learning_rate": 9.900748549947692e-07, + "loss": 1.0189, + "step": 36565 + }, + { + "epoch": 0.86, + "grad_norm": 1.8196247586192953, + "learning_rate": 9.897438475860165e-07, + "loss": 0.9674, + "step": 36566 + }, + { + "epoch": 0.86, + "grad_norm": 1.1088249253575782, + "learning_rate": 9.894128926378255e-07, + "loss": 0.9056, + "step": 36567 + }, + { + "epoch": 0.86, + "grad_norm": 1.9494833561175968, + "learning_rate": 9.890819901521266e-07, + "loss": 0.8827, + "step": 36568 + }, + { + "epoch": 0.86, + "grad_norm": 2.425569284318797, + "learning_rate": 9.887511401308458e-07, + "loss": 1.003, + "step": 36569 + }, + { + "epoch": 0.86, + "grad_norm": 1.9589456892424901, + "learning_rate": 9.884203425759098e-07, + "loss": 0.928, + "step": 36570 + }, + { + "epoch": 0.86, + "grad_norm": 2.2454622447579786, + "learning_rate": 9.880895974892413e-07, + "loss": 0.9511, + "step": 36571 + }, + { + "epoch": 0.86, + "grad_norm": 2.0337444829712097, + "learning_rate": 9.877589048727676e-07, + "loss": 0.9457, + "step": 36572 + }, + { + "epoch": 0.86, + "grad_norm": 2.1530097610979673, + "learning_rate": 9.874282647284173e-07, + "loss": 0.9852, + "step": 36573 + }, + { + "epoch": 0.86, + "grad_norm": 1.867323575126607, + "learning_rate": 9.87097677058113e-07, + "loss": 1.0131, + "step": 36574 + }, + { + "epoch": 0.86, + "grad_norm": 1.950871010634651, + "learning_rate": 9.86767141863777e-07, + "loss": 1.0151, + "step": 36575 + }, + { + "epoch": 0.86, + "grad_norm": 2.075306561405188, + "learning_rate": 9.864366591473385e-07, + "loss": 1.1221, + "step": 36576 + }, + { + "epoch": 0.86, + "grad_norm": 1.9505264301868583, + "learning_rate": 9.86106228910718e-07, + "loss": 0.9413, + "step": 36577 + }, + { + "epoch": 0.86, + "grad_norm": 1.95800973970599, + "learning_rate": 9.857758511558423e-07, + "loss": 0.9784, + "step": 36578 + }, + { + "epoch": 0.86, + "grad_norm": 1.8712513335030467, + "learning_rate": 9.854455258846319e-07, + "loss": 0.9459, + "step": 36579 + }, + { + "epoch": 0.86, + "grad_norm": 2.635505071544959, + "learning_rate": 9.851152530990127e-07, + "loss": 0.9542, + "step": 36580 + }, + { + "epoch": 0.86, + "grad_norm": 2.1958914327284953, + "learning_rate": 9.847850328009056e-07, + "loss": 0.9756, + "step": 36581 + }, + { + "epoch": 0.86, + "grad_norm": 1.6954884209857093, + "learning_rate": 9.844548649922358e-07, + "loss": 0.8939, + "step": 36582 + }, + { + "epoch": 0.86, + "grad_norm": 2.081261385359925, + "learning_rate": 9.84124749674923e-07, + "loss": 0.9232, + "step": 36583 + }, + { + "epoch": 0.86, + "grad_norm": 1.9862276609112, + "learning_rate": 9.837946868508897e-07, + "loss": 0.9075, + "step": 36584 + }, + { + "epoch": 0.86, + "grad_norm": 1.9231748818881238, + "learning_rate": 9.83464676522058e-07, + "loss": 1.0091, + "step": 36585 + }, + { + "epoch": 0.86, + "grad_norm": 2.078623326953631, + "learning_rate": 9.831347186903495e-07, + "loss": 1.04, + "step": 36586 + }, + { + "epoch": 0.86, + "grad_norm": 1.0812516780331027, + "learning_rate": 9.828048133576884e-07, + "loss": 0.9245, + "step": 36587 + }, + { + "epoch": 0.86, + "grad_norm": 1.8497111761764584, + "learning_rate": 9.824749605259898e-07, + "loss": 0.9785, + "step": 36588 + }, + { + "epoch": 0.86, + "grad_norm": 1.0806375074846932, + "learning_rate": 9.821451601971755e-07, + "loss": 0.8947, + "step": 36589 + }, + { + "epoch": 0.86, + "grad_norm": 2.0108213319945745, + "learning_rate": 9.818154123731704e-07, + "loss": 1.0002, + "step": 36590 + }, + { + "epoch": 0.86, + "grad_norm": 2.291447672203722, + "learning_rate": 9.8148571705589e-07, + "loss": 0.9947, + "step": 36591 + }, + { + "epoch": 0.86, + "grad_norm": 2.6025895976229196, + "learning_rate": 9.811560742472537e-07, + "loss": 0.8513, + "step": 36592 + }, + { + "epoch": 0.86, + "grad_norm": 1.0855446898773673, + "learning_rate": 9.808264839491822e-07, + "loss": 0.9749, + "step": 36593 + }, + { + "epoch": 0.86, + "grad_norm": 2.0121484901615725, + "learning_rate": 9.80496946163596e-07, + "loss": 1.0106, + "step": 36594 + }, + { + "epoch": 0.86, + "grad_norm": 2.0541927906758226, + "learning_rate": 9.801674608924118e-07, + "loss": 1.0101, + "step": 36595 + }, + { + "epoch": 0.86, + "grad_norm": 1.1368153170978645, + "learning_rate": 9.798380281375463e-07, + "loss": 0.9947, + "step": 36596 + }, + { + "epoch": 0.86, + "grad_norm": 2.059980111233182, + "learning_rate": 9.795086479009218e-07, + "loss": 1.0065, + "step": 36597 + }, + { + "epoch": 0.86, + "grad_norm": 2.1035928665526353, + "learning_rate": 9.791793201844525e-07, + "loss": 0.8476, + "step": 36598 + }, + { + "epoch": 0.86, + "grad_norm": 1.9763080308607774, + "learning_rate": 9.788500449900563e-07, + "loss": 0.9336, + "step": 36599 + }, + { + "epoch": 0.86, + "grad_norm": 1.9688573939970635, + "learning_rate": 9.78520822319653e-07, + "loss": 0.9742, + "step": 36600 + }, + { + "epoch": 0.86, + "grad_norm": 2.13421330320977, + "learning_rate": 9.781916521751577e-07, + "loss": 0.8957, + "step": 36601 + }, + { + "epoch": 0.86, + "grad_norm": 2.0055997992448087, + "learning_rate": 9.778625345584857e-07, + "loss": 0.8665, + "step": 36602 + }, + { + "epoch": 0.86, + "grad_norm": 2.209296422958699, + "learning_rate": 9.775334694715533e-07, + "loss": 1.0245, + "step": 36603 + }, + { + "epoch": 0.86, + "grad_norm": 1.901332553916441, + "learning_rate": 9.77204456916282e-07, + "loss": 1.0724, + "step": 36604 + }, + { + "epoch": 0.86, + "grad_norm": 1.1528428646525435, + "learning_rate": 9.768754968945792e-07, + "loss": 0.8406, + "step": 36605 + }, + { + "epoch": 0.86, + "grad_norm": 2.347030891572455, + "learning_rate": 9.765465894083637e-07, + "loss": 0.9153, + "step": 36606 + }, + { + "epoch": 0.86, + "grad_norm": 2.1211263694612814, + "learning_rate": 9.762177344595514e-07, + "loss": 1.1604, + "step": 36607 + }, + { + "epoch": 0.86, + "grad_norm": 2.290322905357884, + "learning_rate": 9.758889320500575e-07, + "loss": 0.9493, + "step": 36608 + }, + { + "epoch": 0.86, + "grad_norm": 1.9768698832847522, + "learning_rate": 9.75560182181795e-07, + "loss": 0.904, + "step": 36609 + }, + { + "epoch": 0.86, + "grad_norm": 1.8952360092826352, + "learning_rate": 9.75231484856677e-07, + "loss": 0.9553, + "step": 36610 + }, + { + "epoch": 0.86, + "grad_norm": 2.0415239668423686, + "learning_rate": 9.749028400766191e-07, + "loss": 0.9853, + "step": 36611 + }, + { + "epoch": 0.86, + "grad_norm": 2.0436032797208736, + "learning_rate": 9.745742478435339e-07, + "loss": 0.8001, + "step": 36612 + }, + { + "epoch": 0.86, + "grad_norm": 2.291322221448412, + "learning_rate": 9.74245708159336e-07, + "loss": 0.9287, + "step": 36613 + }, + { + "epoch": 0.86, + "grad_norm": 2.6857613520375074, + "learning_rate": 9.73917221025935e-07, + "loss": 0.9696, + "step": 36614 + }, + { + "epoch": 0.86, + "grad_norm": 1.7608425043173883, + "learning_rate": 9.735887864452476e-07, + "loss": 0.9335, + "step": 36615 + }, + { + "epoch": 0.86, + "grad_norm": 2.236029978934112, + "learning_rate": 9.732604044191818e-07, + "loss": 1.0874, + "step": 36616 + }, + { + "epoch": 0.86, + "grad_norm": 2.2421916107437454, + "learning_rate": 9.72932074949654e-07, + "loss": 0.9739, + "step": 36617 + }, + { + "epoch": 0.86, + "grad_norm": 2.0903570130906433, + "learning_rate": 9.72603798038574e-07, + "loss": 1.0251, + "step": 36618 + }, + { + "epoch": 0.86, + "grad_norm": 2.4855232087689805, + "learning_rate": 9.722755736878498e-07, + "loss": 0.974, + "step": 36619 + }, + { + "epoch": 0.86, + "grad_norm": 1.9698150383923196, + "learning_rate": 9.719474018993957e-07, + "loss": 1.0268, + "step": 36620 + }, + { + "epoch": 0.86, + "grad_norm": 1.7447179291958985, + "learning_rate": 9.716192826751246e-07, + "loss": 0.8387, + "step": 36621 + }, + { + "epoch": 0.86, + "grad_norm": 1.9074846609208644, + "learning_rate": 9.71291216016944e-07, + "loss": 1.0136, + "step": 36622 + }, + { + "epoch": 0.86, + "grad_norm": 1.9017302876532793, + "learning_rate": 9.709632019267624e-07, + "loss": 0.9328, + "step": 36623 + }, + { + "epoch": 0.86, + "grad_norm": 1.8822284730415864, + "learning_rate": 9.706352404064911e-07, + "loss": 0.9768, + "step": 36624 + }, + { + "epoch": 0.86, + "grad_norm": 2.0929293443153414, + "learning_rate": 9.703073314580425e-07, + "loss": 0.9767, + "step": 36625 + }, + { + "epoch": 0.86, + "grad_norm": 2.2468025314980005, + "learning_rate": 9.699794750833235e-07, + "loss": 0.9963, + "step": 36626 + }, + { + "epoch": 0.86, + "grad_norm": 1.919912548881729, + "learning_rate": 9.696516712842407e-07, + "loss": 0.9802, + "step": 36627 + }, + { + "epoch": 0.86, + "grad_norm": 2.0263853789282806, + "learning_rate": 9.693239200627047e-07, + "loss": 0.9086, + "step": 36628 + }, + { + "epoch": 0.86, + "grad_norm": 2.028756585699718, + "learning_rate": 9.68996221420626e-07, + "loss": 0.8827, + "step": 36629 + }, + { + "epoch": 0.86, + "grad_norm": 2.0110233149959993, + "learning_rate": 9.686685753599113e-07, + "loss": 1.0167, + "step": 36630 + }, + { + "epoch": 0.86, + "grad_norm": 2.976229552047938, + "learning_rate": 9.683409818824641e-07, + "loss": 0.8879, + "step": 36631 + }, + { + "epoch": 0.86, + "grad_norm": 2.175502848962767, + "learning_rate": 9.680134409901975e-07, + "loss": 1.0085, + "step": 36632 + }, + { + "epoch": 0.86, + "grad_norm": 2.054523256009669, + "learning_rate": 9.676859526850145e-07, + "loss": 0.8692, + "step": 36633 + }, + { + "epoch": 0.86, + "grad_norm": 2.11295207770945, + "learning_rate": 9.67358516968826e-07, + "loss": 1.0458, + "step": 36634 + }, + { + "epoch": 0.86, + "grad_norm": 2.6637818443502392, + "learning_rate": 9.670311338435324e-07, + "loss": 0.9192, + "step": 36635 + }, + { + "epoch": 0.86, + "grad_norm": 3.434851630665366, + "learning_rate": 9.667038033110465e-07, + "loss": 0.907, + "step": 36636 + }, + { + "epoch": 0.86, + "grad_norm": 1.8861163598822674, + "learning_rate": 9.663765253732682e-07, + "loss": 1.0303, + "step": 36637 + }, + { + "epoch": 0.86, + "grad_norm": 2.002790943075645, + "learning_rate": 9.660493000321081e-07, + "loss": 0.9405, + "step": 36638 + }, + { + "epoch": 0.86, + "grad_norm": 2.1858928138601126, + "learning_rate": 9.65722127289469e-07, + "loss": 0.9484, + "step": 36639 + }, + { + "epoch": 0.86, + "grad_norm": 1.850289943143151, + "learning_rate": 9.65395007147254e-07, + "loss": 0.8326, + "step": 36640 + }, + { + "epoch": 0.86, + "grad_norm": 2.2154889856213473, + "learning_rate": 9.650679396073691e-07, + "loss": 0.921, + "step": 36641 + }, + { + "epoch": 0.86, + "grad_norm": 2.015791356051825, + "learning_rate": 9.647409246717222e-07, + "loss": 0.9392, + "step": 36642 + }, + { + "epoch": 0.86, + "grad_norm": 2.0329990831957874, + "learning_rate": 9.644139623422132e-07, + "loss": 1.1036, + "step": 36643 + }, + { + "epoch": 0.86, + "grad_norm": 1.987065380740634, + "learning_rate": 9.640870526207445e-07, + "loss": 0.8959, + "step": 36644 + }, + { + "epoch": 0.86, + "grad_norm": 2.208155169245472, + "learning_rate": 9.637601955092224e-07, + "loss": 0.931, + "step": 36645 + }, + { + "epoch": 0.86, + "grad_norm": 1.958431564647979, + "learning_rate": 9.634333910095516e-07, + "loss": 1.0342, + "step": 36646 + }, + { + "epoch": 0.86, + "grad_norm": 2.3067260262259186, + "learning_rate": 9.63106639123632e-07, + "loss": 1.0513, + "step": 36647 + }, + { + "epoch": 0.86, + "grad_norm": 2.109273624798816, + "learning_rate": 9.627799398533644e-07, + "loss": 0.9941, + "step": 36648 + }, + { + "epoch": 0.86, + "grad_norm": 2.0853994386341173, + "learning_rate": 9.624532932006536e-07, + "loss": 1.0934, + "step": 36649 + }, + { + "epoch": 0.86, + "grad_norm": 2.1548697203045957, + "learning_rate": 9.621266991674017e-07, + "loss": 0.8922, + "step": 36650 + }, + { + "epoch": 0.86, + "grad_norm": 1.894492231168762, + "learning_rate": 9.618001577555113e-07, + "loss": 0.929, + "step": 36651 + }, + { + "epoch": 0.86, + "grad_norm": 1.8924577590343175, + "learning_rate": 9.61473668966879e-07, + "loss": 1.0077, + "step": 36652 + }, + { + "epoch": 0.86, + "grad_norm": 2.199952114648165, + "learning_rate": 9.6114723280341e-07, + "loss": 0.9216, + "step": 36653 + }, + { + "epoch": 0.86, + "grad_norm": 1.8216105252849695, + "learning_rate": 9.608208492670023e-07, + "loss": 0.9456, + "step": 36654 + }, + { + "epoch": 0.86, + "grad_norm": 2.149763878039636, + "learning_rate": 9.604945183595593e-07, + "loss": 0.9393, + "step": 36655 + }, + { + "epoch": 0.86, + "grad_norm": 1.9811162030185125, + "learning_rate": 9.60168240082976e-07, + "loss": 0.9234, + "step": 36656 + }, + { + "epoch": 0.86, + "grad_norm": 1.812129617880893, + "learning_rate": 9.598420144391585e-07, + "loss": 0.8595, + "step": 36657 + }, + { + "epoch": 0.86, + "grad_norm": 1.9104680989609784, + "learning_rate": 9.5951584143e-07, + "loss": 0.976, + "step": 36658 + }, + { + "epoch": 0.86, + "grad_norm": 2.0132617686607346, + "learning_rate": 9.591897210574053e-07, + "loss": 1.0163, + "step": 36659 + }, + { + "epoch": 0.86, + "grad_norm": 2.141492808251164, + "learning_rate": 9.588636533232697e-07, + "loss": 0.96, + "step": 36660 + }, + { + "epoch": 0.86, + "grad_norm": 1.808074518624416, + "learning_rate": 9.585376382294898e-07, + "loss": 0.9696, + "step": 36661 + }, + { + "epoch": 0.86, + "grad_norm": 2.4903442492274523, + "learning_rate": 9.582116757779668e-07, + "loss": 0.9863, + "step": 36662 + }, + { + "epoch": 0.86, + "grad_norm": 2.1238682264860893, + "learning_rate": 9.57885765970601e-07, + "loss": 0.9448, + "step": 36663 + }, + { + "epoch": 0.86, + "grad_norm": 1.049507413736219, + "learning_rate": 9.575599088092858e-07, + "loss": 0.9018, + "step": 36664 + }, + { + "epoch": 0.86, + "grad_norm": 2.032991260044675, + "learning_rate": 9.572341042959177e-07, + "loss": 0.8866, + "step": 36665 + }, + { + "epoch": 0.86, + "grad_norm": 2.242748392313744, + "learning_rate": 9.569083524323952e-07, + "loss": 0.8752, + "step": 36666 + }, + { + "epoch": 0.86, + "grad_norm": 2.124192992765275, + "learning_rate": 9.56582653220618e-07, + "loss": 1.0281, + "step": 36667 + }, + { + "epoch": 0.86, + "grad_norm": 1.8655438593935278, + "learning_rate": 9.562570066624788e-07, + "loss": 0.9789, + "step": 36668 + }, + { + "epoch": 0.86, + "grad_norm": 1.930418900477114, + "learning_rate": 9.55931412759874e-07, + "loss": 0.9492, + "step": 36669 + }, + { + "epoch": 0.86, + "grad_norm": 2.3571875230239234, + "learning_rate": 9.556058715146987e-07, + "loss": 0.8883, + "step": 36670 + }, + { + "epoch": 0.86, + "grad_norm": 2.0579313756634137, + "learning_rate": 9.552803829288514e-07, + "loss": 0.8128, + "step": 36671 + }, + { + "epoch": 0.86, + "grad_norm": 2.455517471119436, + "learning_rate": 9.549549470042251e-07, + "loss": 1.0967, + "step": 36672 + }, + { + "epoch": 0.86, + "grad_norm": 2.2638885206253754, + "learning_rate": 9.546295637427138e-07, + "loss": 1.0454, + "step": 36673 + }, + { + "epoch": 0.86, + "grad_norm": 1.0470509970547934, + "learning_rate": 9.543042331462139e-07, + "loss": 0.8849, + "step": 36674 + }, + { + "epoch": 0.86, + "grad_norm": 2.1119825230571583, + "learning_rate": 9.539789552166167e-07, + "loss": 0.9436, + "step": 36675 + }, + { + "epoch": 0.86, + "grad_norm": 2.0420330214328555, + "learning_rate": 9.536537299558202e-07, + "loss": 0.8645, + "step": 36676 + }, + { + "epoch": 0.86, + "grad_norm": 2.821425836185052, + "learning_rate": 9.533285573657137e-07, + "loss": 1.0048, + "step": 36677 + }, + { + "epoch": 0.86, + "grad_norm": 1.1614960730069686, + "learning_rate": 9.530034374481945e-07, + "loss": 0.9029, + "step": 36678 + }, + { + "epoch": 0.86, + "grad_norm": 1.959528237468768, + "learning_rate": 9.526783702051512e-07, + "loss": 0.9168, + "step": 36679 + }, + { + "epoch": 0.86, + "grad_norm": 1.864514830493501, + "learning_rate": 9.523533556384812e-07, + "loss": 1.0405, + "step": 36680 + }, + { + "epoch": 0.86, + "grad_norm": 1.9194079038093932, + "learning_rate": 9.520283937500718e-07, + "loss": 0.8996, + "step": 36681 + }, + { + "epoch": 0.86, + "grad_norm": 1.9091430237364324, + "learning_rate": 9.517034845418194e-07, + "loss": 1.014, + "step": 36682 + }, + { + "epoch": 0.86, + "grad_norm": 2.347763729888114, + "learning_rate": 9.513786280156135e-07, + "loss": 1.0385, + "step": 36683 + }, + { + "epoch": 0.86, + "grad_norm": 1.9960182658804788, + "learning_rate": 9.510538241733446e-07, + "loss": 1.0687, + "step": 36684 + }, + { + "epoch": 0.86, + "grad_norm": 1.9358725276845092, + "learning_rate": 9.507290730169094e-07, + "loss": 0.9707, + "step": 36685 + }, + { + "epoch": 0.86, + "grad_norm": 1.9551768948446218, + "learning_rate": 9.504043745481905e-07, + "loss": 0.8625, + "step": 36686 + }, + { + "epoch": 0.86, + "grad_norm": 2.095201391809433, + "learning_rate": 9.500797287690833e-07, + "loss": 1.0108, + "step": 36687 + }, + { + "epoch": 0.86, + "grad_norm": 2.5166905077114112, + "learning_rate": 9.49755135681476e-07, + "loss": 0.8906, + "step": 36688 + }, + { + "epoch": 0.86, + "grad_norm": 3.9899420566961186, + "learning_rate": 9.49430595287264e-07, + "loss": 1.0354, + "step": 36689 + }, + { + "epoch": 0.86, + "grad_norm": 2.157423728676676, + "learning_rate": 9.491061075883279e-07, + "loss": 0.9472, + "step": 36690 + }, + { + "epoch": 0.86, + "grad_norm": 1.988347582305476, + "learning_rate": 9.487816725865618e-07, + "loss": 1.0661, + "step": 36691 + }, + { + "epoch": 0.86, + "grad_norm": 2.0879179442672937, + "learning_rate": 9.484572902838563e-07, + "loss": 0.995, + "step": 36692 + }, + { + "epoch": 0.86, + "grad_norm": 1.848288873354755, + "learning_rate": 9.481329606820988e-07, + "loss": 0.9597, + "step": 36693 + }, + { + "epoch": 0.86, + "grad_norm": 2.063144881946585, + "learning_rate": 9.478086837831746e-07, + "loss": 0.941, + "step": 36694 + }, + { + "epoch": 0.86, + "grad_norm": 1.8714755538301782, + "learning_rate": 9.474844595889743e-07, + "loss": 0.9985, + "step": 36695 + }, + { + "epoch": 0.86, + "grad_norm": 2.1524513358584643, + "learning_rate": 9.471602881013864e-07, + "loss": 0.8789, + "step": 36696 + }, + { + "epoch": 0.86, + "grad_norm": 1.00553533655052, + "learning_rate": 9.468361693222972e-07, + "loss": 0.9135, + "step": 36697 + }, + { + "epoch": 0.86, + "grad_norm": 1.8479027293872485, + "learning_rate": 9.46512103253594e-07, + "loss": 1.0088, + "step": 36698 + }, + { + "epoch": 0.86, + "grad_norm": 2.2067790045527667, + "learning_rate": 9.461880898971654e-07, + "loss": 0.9493, + "step": 36699 + }, + { + "epoch": 0.86, + "grad_norm": 1.8301225567653414, + "learning_rate": 9.458641292548931e-07, + "loss": 0.9754, + "step": 36700 + }, + { + "epoch": 0.86, + "grad_norm": 1.8486596235642427, + "learning_rate": 9.455402213286669e-07, + "loss": 1.0021, + "step": 36701 + }, + { + "epoch": 0.86, + "grad_norm": 2.052747798307758, + "learning_rate": 9.452163661203728e-07, + "loss": 0.9056, + "step": 36702 + }, + { + "epoch": 0.86, + "grad_norm": 1.8755086950081516, + "learning_rate": 9.448925636318973e-07, + "loss": 1.0766, + "step": 36703 + }, + { + "epoch": 0.86, + "grad_norm": 1.7775456925570188, + "learning_rate": 9.44568813865121e-07, + "loss": 0.8831, + "step": 36704 + }, + { + "epoch": 0.86, + "grad_norm": 1.0979712234171044, + "learning_rate": 9.442451168219325e-07, + "loss": 0.9766, + "step": 36705 + }, + { + "epoch": 0.86, + "grad_norm": 2.6312431094622064, + "learning_rate": 9.439214725042178e-07, + "loss": 1.0487, + "step": 36706 + }, + { + "epoch": 0.86, + "grad_norm": 2.053620623962229, + "learning_rate": 9.435978809138601e-07, + "loss": 1.0166, + "step": 36707 + }, + { + "epoch": 0.86, + "grad_norm": 2.0124005982711717, + "learning_rate": 9.432743420527401e-07, + "loss": 0.9009, + "step": 36708 + }, + { + "epoch": 0.86, + "grad_norm": 2.34587721095281, + "learning_rate": 9.429508559227441e-07, + "loss": 1.1015, + "step": 36709 + }, + { + "epoch": 0.86, + "grad_norm": 1.9473200827964106, + "learning_rate": 9.426274225257582e-07, + "loss": 0.8619, + "step": 36710 + }, + { + "epoch": 0.86, + "grad_norm": 1.912128229579178, + "learning_rate": 9.42304041863662e-07, + "loss": 0.9232, + "step": 36711 + }, + { + "epoch": 0.86, + "grad_norm": 2.1029030920592997, + "learning_rate": 9.419807139383374e-07, + "loss": 1.0428, + "step": 36712 + }, + { + "epoch": 0.86, + "grad_norm": 2.688647338824371, + "learning_rate": 9.41657438751672e-07, + "loss": 0.9993, + "step": 36713 + }, + { + "epoch": 0.86, + "grad_norm": 1.0973795504084114, + "learning_rate": 9.413342163055416e-07, + "loss": 0.9002, + "step": 36714 + }, + { + "epoch": 0.86, + "grad_norm": 1.8827577700218585, + "learning_rate": 9.410110466018341e-07, + "loss": 1.0245, + "step": 36715 + }, + { + "epoch": 0.86, + "grad_norm": 1.982023703645253, + "learning_rate": 9.406879296424254e-07, + "loss": 0.9366, + "step": 36716 + }, + { + "epoch": 0.87, + "grad_norm": 1.9435423313214248, + "learning_rate": 9.403648654292019e-07, + "loss": 0.9119, + "step": 36717 + }, + { + "epoch": 0.87, + "grad_norm": 1.6949721488587122, + "learning_rate": 9.40041853964041e-07, + "loss": 0.902, + "step": 36718 + }, + { + "epoch": 0.87, + "grad_norm": 1.9571491186805536, + "learning_rate": 9.397188952488267e-07, + "loss": 0.8442, + "step": 36719 + }, + { + "epoch": 0.87, + "grad_norm": 2.0392814020654515, + "learning_rate": 9.393959892854366e-07, + "loss": 0.925, + "step": 36720 + }, + { + "epoch": 0.87, + "grad_norm": 2.044203447703823, + "learning_rate": 9.39073136075751e-07, + "loss": 0.9261, + "step": 36721 + }, + { + "epoch": 0.87, + "grad_norm": 1.0893933653170818, + "learning_rate": 9.387503356216488e-07, + "loss": 0.9313, + "step": 36722 + }, + { + "epoch": 0.87, + "grad_norm": 2.0204371231508067, + "learning_rate": 9.384275879250138e-07, + "loss": 0.8163, + "step": 36723 + }, + { + "epoch": 0.87, + "grad_norm": 1.9466039025636708, + "learning_rate": 9.381048929877223e-07, + "loss": 0.9945, + "step": 36724 + }, + { + "epoch": 0.87, + "grad_norm": 1.8917990747805655, + "learning_rate": 9.377822508116507e-07, + "loss": 0.7971, + "step": 36725 + }, + { + "epoch": 0.87, + "grad_norm": 1.0409327372480226, + "learning_rate": 9.374596613986797e-07, + "loss": 0.8791, + "step": 36726 + }, + { + "epoch": 0.87, + "grad_norm": 2.0234326448046307, + "learning_rate": 9.371371247506899e-07, + "loss": 0.8841, + "step": 36727 + }, + { + "epoch": 0.87, + "grad_norm": 2.013953404536748, + "learning_rate": 9.368146408695566e-07, + "loss": 0.986, + "step": 36728 + }, + { + "epoch": 0.87, + "grad_norm": 2.1480240495536993, + "learning_rate": 9.364922097571572e-07, + "loss": 0.9655, + "step": 36729 + }, + { + "epoch": 0.87, + "grad_norm": 2.016154663973597, + "learning_rate": 9.361698314153678e-07, + "loss": 0.9317, + "step": 36730 + }, + { + "epoch": 0.87, + "grad_norm": 1.936811456744659, + "learning_rate": 9.358475058460704e-07, + "loss": 1.0681, + "step": 36731 + }, + { + "epoch": 0.87, + "grad_norm": 1.7844222469742972, + "learning_rate": 9.355252330511377e-07, + "loss": 0.9479, + "step": 36732 + }, + { + "epoch": 0.87, + "grad_norm": 1.9098203641410711, + "learning_rate": 9.352030130324463e-07, + "loss": 1.1687, + "step": 36733 + }, + { + "epoch": 0.87, + "grad_norm": 1.1386568680260167, + "learning_rate": 9.348808457918734e-07, + "loss": 0.9846, + "step": 36734 + }, + { + "epoch": 0.87, + "grad_norm": 1.0999301872033154, + "learning_rate": 9.345587313312921e-07, + "loss": 0.9124, + "step": 36735 + }, + { + "epoch": 0.87, + "grad_norm": 1.0207235712824045, + "learning_rate": 9.342366696525829e-07, + "loss": 0.856, + "step": 36736 + }, + { + "epoch": 0.87, + "grad_norm": 2.9654256619653045, + "learning_rate": 9.339146607576155e-07, + "loss": 1.0497, + "step": 36737 + }, + { + "epoch": 0.87, + "grad_norm": 2.0743549788454336, + "learning_rate": 9.335927046482695e-07, + "loss": 0.9859, + "step": 36738 + }, + { + "epoch": 0.87, + "grad_norm": 1.9114072537575464, + "learning_rate": 9.332708013264147e-07, + "loss": 0.9427, + "step": 36739 + }, + { + "epoch": 0.87, + "grad_norm": 2.205583017158747, + "learning_rate": 9.329489507939304e-07, + "loss": 0.9412, + "step": 36740 + }, + { + "epoch": 0.87, + "grad_norm": 2.041699898127089, + "learning_rate": 9.326271530526865e-07, + "loss": 1.0723, + "step": 36741 + }, + { + "epoch": 0.87, + "grad_norm": 2.8888318457294373, + "learning_rate": 9.323054081045579e-07, + "loss": 0.951, + "step": 36742 + }, + { + "epoch": 0.87, + "grad_norm": 2.0684803333206316, + "learning_rate": 9.319837159514167e-07, + "loss": 1.1269, + "step": 36743 + }, + { + "epoch": 0.87, + "grad_norm": 1.961970736471608, + "learning_rate": 9.31662076595139e-07, + "loss": 0.971, + "step": 36744 + }, + { + "epoch": 0.87, + "grad_norm": 2.0493029093394486, + "learning_rate": 9.313404900375966e-07, + "loss": 0.8973, + "step": 36745 + }, + { + "epoch": 0.87, + "grad_norm": 2.081400733645593, + "learning_rate": 9.310189562806582e-07, + "loss": 0.9103, + "step": 36746 + }, + { + "epoch": 0.87, + "grad_norm": 2.1835465989114624, + "learning_rate": 9.306974753261999e-07, + "loss": 1.0141, + "step": 36747 + }, + { + "epoch": 0.87, + "grad_norm": 2.0402980075612667, + "learning_rate": 9.303760471760925e-07, + "loss": 1.0254, + "step": 36748 + }, + { + "epoch": 0.87, + "grad_norm": 2.072893635496174, + "learning_rate": 9.300546718322079e-07, + "loss": 0.9158, + "step": 36749 + }, + { + "epoch": 0.87, + "grad_norm": 1.9073466509389116, + "learning_rate": 9.297333492964144e-07, + "loss": 1.0193, + "step": 36750 + }, + { + "epoch": 0.87, + "grad_norm": 2.040541742720854, + "learning_rate": 9.294120795705852e-07, + "loss": 0.9412, + "step": 36751 + }, + { + "epoch": 0.87, + "grad_norm": 2.1362629555548054, + "learning_rate": 9.290908626565931e-07, + "loss": 0.9309, + "step": 36752 + }, + { + "epoch": 0.87, + "grad_norm": 2.145791447400028, + "learning_rate": 9.287696985563044e-07, + "loss": 0.8982, + "step": 36753 + }, + { + "epoch": 0.87, + "grad_norm": 2.021096633729179, + "learning_rate": 9.284485872715898e-07, + "loss": 0.8912, + "step": 36754 + }, + { + "epoch": 0.87, + "grad_norm": 2.0975744613982528, + "learning_rate": 9.281275288043212e-07, + "loss": 0.9932, + "step": 36755 + }, + { + "epoch": 0.87, + "grad_norm": 1.0754364896943465, + "learning_rate": 9.278065231563637e-07, + "loss": 0.9288, + "step": 36756 + }, + { + "epoch": 0.87, + "grad_norm": 1.8264203201572626, + "learning_rate": 9.274855703295904e-07, + "loss": 0.9215, + "step": 36757 + }, + { + "epoch": 0.87, + "grad_norm": 2.135937609734459, + "learning_rate": 9.271646703258674e-07, + "loss": 1.1116, + "step": 36758 + }, + { + "epoch": 0.87, + "grad_norm": 2.2156813684295646, + "learning_rate": 9.268438231470656e-07, + "loss": 1.0127, + "step": 36759 + }, + { + "epoch": 0.87, + "grad_norm": 1.8898879074435169, + "learning_rate": 9.265230287950499e-07, + "loss": 0.9791, + "step": 36760 + }, + { + "epoch": 0.87, + "grad_norm": 2.059663474741967, + "learning_rate": 9.262022872716914e-07, + "loss": 0.9525, + "step": 36761 + }, + { + "epoch": 0.87, + "grad_norm": 1.8912615511098911, + "learning_rate": 9.25881598578856e-07, + "loss": 1.0059, + "step": 36762 + }, + { + "epoch": 0.87, + "grad_norm": 1.7688325803707425, + "learning_rate": 9.255609627184092e-07, + "loss": 0.9732, + "step": 36763 + }, + { + "epoch": 0.87, + "grad_norm": 1.9178530591759433, + "learning_rate": 9.252403796922193e-07, + "loss": 0.8932, + "step": 36764 + }, + { + "epoch": 0.87, + "grad_norm": 1.1426197792269135, + "learning_rate": 9.249198495021549e-07, + "loss": 0.9859, + "step": 36765 + }, + { + "epoch": 0.87, + "grad_norm": 2.135749307340625, + "learning_rate": 9.245993721500801e-07, + "loss": 0.8119, + "step": 36766 + }, + { + "epoch": 0.87, + "grad_norm": 2.0152592983357276, + "learning_rate": 9.242789476378589e-07, + "loss": 0.9805, + "step": 36767 + }, + { + "epoch": 0.87, + "grad_norm": 1.8935124321615455, + "learning_rate": 9.239585759673586e-07, + "loss": 1.0491, + "step": 36768 + }, + { + "epoch": 0.87, + "grad_norm": 2.1053404993468727, + "learning_rate": 9.236382571404478e-07, + "loss": 0.9614, + "step": 36769 + }, + { + "epoch": 0.87, + "grad_norm": 1.9895436498125345, + "learning_rate": 9.233179911589874e-07, + "loss": 1.0354, + "step": 36770 + }, + { + "epoch": 0.87, + "grad_norm": 2.0785509740422037, + "learning_rate": 9.229977780248423e-07, + "loss": 0.8278, + "step": 36771 + }, + { + "epoch": 0.87, + "grad_norm": 2.0080686342176874, + "learning_rate": 9.226776177398778e-07, + "loss": 0.9131, + "step": 36772 + }, + { + "epoch": 0.87, + "grad_norm": 2.204155721092002, + "learning_rate": 9.223575103059601e-07, + "loss": 1.0323, + "step": 36773 + }, + { + "epoch": 0.87, + "grad_norm": 2.186738032441339, + "learning_rate": 9.2203745572495e-07, + "loss": 1.0693, + "step": 36774 + }, + { + "epoch": 0.87, + "grad_norm": 1.8621459058345602, + "learning_rate": 9.217174539987106e-07, + "loss": 0.8888, + "step": 36775 + }, + { + "epoch": 0.87, + "grad_norm": 2.1578404146599715, + "learning_rate": 9.21397505129108e-07, + "loss": 0.9213, + "step": 36776 + }, + { + "epoch": 0.87, + "grad_norm": 1.9559192100507745, + "learning_rate": 9.210776091180018e-07, + "loss": 0.9526, + "step": 36777 + }, + { + "epoch": 0.87, + "grad_norm": 1.8767194741428481, + "learning_rate": 9.207577659672573e-07, + "loss": 0.8577, + "step": 36778 + }, + { + "epoch": 0.87, + "grad_norm": 1.9251674440181306, + "learning_rate": 9.204379756787341e-07, + "loss": 1.0733, + "step": 36779 + }, + { + "epoch": 0.87, + "grad_norm": 1.7620273003229898, + "learning_rate": 9.201182382542983e-07, + "loss": 0.9176, + "step": 36780 + }, + { + "epoch": 0.87, + "grad_norm": 2.029182551103617, + "learning_rate": 9.197985536958054e-07, + "loss": 1.0633, + "step": 36781 + }, + { + "epoch": 0.87, + "grad_norm": 3.5253347508660138, + "learning_rate": 9.194789220051214e-07, + "loss": 1.0152, + "step": 36782 + }, + { + "epoch": 0.87, + "grad_norm": 1.8590036624219302, + "learning_rate": 9.191593431841084e-07, + "loss": 0.9082, + "step": 36783 + }, + { + "epoch": 0.87, + "grad_norm": 1.9983676169830702, + "learning_rate": 9.188398172346214e-07, + "loss": 0.9713, + "step": 36784 + }, + { + "epoch": 0.87, + "grad_norm": 1.874129197444692, + "learning_rate": 9.185203441585244e-07, + "loss": 0.9215, + "step": 36785 + }, + { + "epoch": 0.87, + "grad_norm": 1.8861014320531768, + "learning_rate": 9.182009239576773e-07, + "loss": 1.1015, + "step": 36786 + }, + { + "epoch": 0.87, + "grad_norm": 2.103113069175425, + "learning_rate": 9.178815566339427e-07, + "loss": 0.9062, + "step": 36787 + }, + { + "epoch": 0.87, + "grad_norm": 1.0603636197802733, + "learning_rate": 9.175622421891738e-07, + "loss": 0.9493, + "step": 36788 + }, + { + "epoch": 0.87, + "grad_norm": 2.112234340282009, + "learning_rate": 9.172429806252325e-07, + "loss": 1.0346, + "step": 36789 + }, + { + "epoch": 0.87, + "grad_norm": 1.8882808286639177, + "learning_rate": 9.169237719439805e-07, + "loss": 0.9351, + "step": 36790 + }, + { + "epoch": 0.87, + "grad_norm": 2.171514467676599, + "learning_rate": 9.16604616147273e-07, + "loss": 0.8506, + "step": 36791 + }, + { + "epoch": 0.87, + "grad_norm": 1.877740954584883, + "learning_rate": 9.162855132369686e-07, + "loss": 0.9349, + "step": 36792 + }, + { + "epoch": 0.87, + "grad_norm": 2.0579469272416815, + "learning_rate": 9.159664632149245e-07, + "loss": 1.0742, + "step": 36793 + }, + { + "epoch": 0.87, + "grad_norm": 2.085273191310773, + "learning_rate": 9.156474660830028e-07, + "loss": 0.888, + "step": 36794 + }, + { + "epoch": 0.87, + "grad_norm": 1.9553421940897169, + "learning_rate": 9.153285218430552e-07, + "loss": 1.0639, + "step": 36795 + }, + { + "epoch": 0.87, + "grad_norm": 2.2898814212166037, + "learning_rate": 9.150096304969425e-07, + "loss": 1.0032, + "step": 36796 + }, + { + "epoch": 0.87, + "grad_norm": 1.9403475569807966, + "learning_rate": 9.146907920465197e-07, + "loss": 0.8831, + "step": 36797 + }, + { + "epoch": 0.87, + "grad_norm": 2.234711163404711, + "learning_rate": 9.143720064936423e-07, + "loss": 0.8859, + "step": 36798 + }, + { + "epoch": 0.87, + "grad_norm": 2.280611164213447, + "learning_rate": 9.140532738401675e-07, + "loss": 0.9507, + "step": 36799 + }, + { + "epoch": 0.87, + "grad_norm": 1.9828648064158778, + "learning_rate": 9.137345940879538e-07, + "loss": 1.0504, + "step": 36800 + }, + { + "epoch": 0.87, + "grad_norm": 1.12395066113519, + "learning_rate": 9.134159672388531e-07, + "loss": 0.9289, + "step": 36801 + }, + { + "epoch": 0.87, + "grad_norm": 1.9068592014874606, + "learning_rate": 9.130973932947196e-07, + "loss": 1.0891, + "step": 36802 + }, + { + "epoch": 0.87, + "grad_norm": 3.0674872373010946, + "learning_rate": 9.127788722574094e-07, + "loss": 1.1092, + "step": 36803 + }, + { + "epoch": 0.87, + "grad_norm": 2.083652406550029, + "learning_rate": 9.124604041287799e-07, + "loss": 1.029, + "step": 36804 + }, + { + "epoch": 0.87, + "grad_norm": 1.9279876680749857, + "learning_rate": 9.121419889106831e-07, + "loss": 0.9131, + "step": 36805 + }, + { + "epoch": 0.87, + "grad_norm": 1.9074578773646402, + "learning_rate": 9.118236266049707e-07, + "loss": 0.8852, + "step": 36806 + }, + { + "epoch": 0.87, + "grad_norm": 2.157521057771545, + "learning_rate": 9.115053172134991e-07, + "loss": 1.0349, + "step": 36807 + }, + { + "epoch": 0.87, + "grad_norm": 1.9904804465425938, + "learning_rate": 9.111870607381223e-07, + "loss": 0.8172, + "step": 36808 + }, + { + "epoch": 0.87, + "grad_norm": 2.0232508853611604, + "learning_rate": 9.108688571806912e-07, + "loss": 1.0922, + "step": 36809 + }, + { + "epoch": 0.87, + "grad_norm": 1.8151307792969686, + "learning_rate": 9.105507065430586e-07, + "loss": 0.865, + "step": 36810 + }, + { + "epoch": 0.87, + "grad_norm": 1.9265764405333239, + "learning_rate": 9.102326088270763e-07, + "loss": 1.0101, + "step": 36811 + }, + { + "epoch": 0.87, + "grad_norm": 2.330318696912028, + "learning_rate": 9.099145640346007e-07, + "loss": 1.0099, + "step": 36812 + }, + { + "epoch": 0.87, + "grad_norm": 2.111008568090849, + "learning_rate": 9.095965721674793e-07, + "loss": 0.9732, + "step": 36813 + }, + { + "epoch": 0.87, + "grad_norm": 3.23989587263669, + "learning_rate": 9.092786332275627e-07, + "loss": 0.9743, + "step": 36814 + }, + { + "epoch": 0.87, + "grad_norm": 2.1340576365117037, + "learning_rate": 9.089607472167061e-07, + "loss": 1.092, + "step": 36815 + }, + { + "epoch": 0.87, + "grad_norm": 1.098589708542311, + "learning_rate": 9.08642914136757e-07, + "loss": 0.9421, + "step": 36816 + }, + { + "epoch": 0.87, + "grad_norm": 2.0486717599697375, + "learning_rate": 9.083251339895682e-07, + "loss": 0.9631, + "step": 36817 + }, + { + "epoch": 0.87, + "grad_norm": 1.8142055082281447, + "learning_rate": 9.080074067769872e-07, + "loss": 0.9039, + "step": 36818 + }, + { + "epoch": 0.87, + "grad_norm": 2.3061461638897462, + "learning_rate": 9.076897325008671e-07, + "loss": 1.0731, + "step": 36819 + }, + { + "epoch": 0.87, + "grad_norm": 5.818662163678916, + "learning_rate": 9.073721111630552e-07, + "loss": 0.8684, + "step": 36820 + }, + { + "epoch": 0.87, + "grad_norm": 2.105246579271142, + "learning_rate": 9.070545427654021e-07, + "loss": 0.9173, + "step": 36821 + }, + { + "epoch": 0.87, + "grad_norm": 1.0776240217785729, + "learning_rate": 9.067370273097575e-07, + "loss": 0.9451, + "step": 36822 + }, + { + "epoch": 0.87, + "grad_norm": 2.051185518222848, + "learning_rate": 9.064195647979656e-07, + "loss": 1.0095, + "step": 36823 + }, + { + "epoch": 0.87, + "grad_norm": 2.2867950982759733, + "learning_rate": 9.061021552318783e-07, + "loss": 1.0128, + "step": 36824 + }, + { + "epoch": 0.87, + "grad_norm": 2.012199245286007, + "learning_rate": 9.057847986133461e-07, + "loss": 0.8774, + "step": 36825 + }, + { + "epoch": 0.87, + "grad_norm": 2.120750848265142, + "learning_rate": 9.054674949442132e-07, + "loss": 0.9865, + "step": 36826 + }, + { + "epoch": 0.87, + "grad_norm": 2.094611037676675, + "learning_rate": 9.05150244226326e-07, + "loss": 0.9753, + "step": 36827 + }, + { + "epoch": 0.87, + "grad_norm": 1.932590471182309, + "learning_rate": 9.04833046461534e-07, + "loss": 0.8604, + "step": 36828 + }, + { + "epoch": 0.87, + "grad_norm": 1.9719711525676646, + "learning_rate": 9.045159016516847e-07, + "loss": 0.967, + "step": 36829 + }, + { + "epoch": 0.87, + "grad_norm": 1.8884881062341718, + "learning_rate": 9.041988097986232e-07, + "loss": 0.9291, + "step": 36830 + }, + { + "epoch": 0.87, + "grad_norm": 2.0402781725067496, + "learning_rate": 9.038817709041947e-07, + "loss": 0.9931, + "step": 36831 + }, + { + "epoch": 0.87, + "grad_norm": 2.02893636487065, + "learning_rate": 9.035647849702467e-07, + "loss": 0.9655, + "step": 36832 + }, + { + "epoch": 0.87, + "grad_norm": 2.3232371147608686, + "learning_rate": 9.032478519986254e-07, + "loss": 0.8814, + "step": 36833 + }, + { + "epoch": 0.87, + "grad_norm": 2.0185077635944966, + "learning_rate": 9.02930971991175e-07, + "loss": 1.0213, + "step": 36834 + }, + { + "epoch": 0.87, + "grad_norm": 2.0546612312945736, + "learning_rate": 9.026141449497394e-07, + "loss": 0.9685, + "step": 36835 + }, + { + "epoch": 0.87, + "grad_norm": 1.9475830307514677, + "learning_rate": 9.022973708761662e-07, + "loss": 0.9885, + "step": 36836 + }, + { + "epoch": 0.87, + "grad_norm": 1.84651366833595, + "learning_rate": 9.019806497722961e-07, + "loss": 0.9581, + "step": 36837 + }, + { + "epoch": 0.87, + "grad_norm": 1.9263894150176866, + "learning_rate": 9.016639816399764e-07, + "loss": 1.0558, + "step": 36838 + }, + { + "epoch": 0.87, + "grad_norm": 1.9053708677979462, + "learning_rate": 9.013473664810479e-07, + "loss": 0.9343, + "step": 36839 + }, + { + "epoch": 0.87, + "grad_norm": 1.849639218937044, + "learning_rate": 9.01030804297357e-07, + "loss": 0.7852, + "step": 36840 + }, + { + "epoch": 0.87, + "grad_norm": 2.1861033259659126, + "learning_rate": 9.007142950907444e-07, + "loss": 0.9579, + "step": 36841 + }, + { + "epoch": 0.87, + "grad_norm": 2.3216481292179845, + "learning_rate": 9.003978388630552e-07, + "loss": 1.1454, + "step": 36842 + }, + { + "epoch": 0.87, + "grad_norm": 1.1011434851510518, + "learning_rate": 9.000814356161314e-07, + "loss": 0.9647, + "step": 36843 + }, + { + "epoch": 0.87, + "grad_norm": 1.9050632590742826, + "learning_rate": 8.997650853518125e-07, + "loss": 1.0064, + "step": 36844 + }, + { + "epoch": 0.87, + "grad_norm": 1.0774888182150324, + "learning_rate": 8.994487880719415e-07, + "loss": 0.8912, + "step": 36845 + }, + { + "epoch": 0.87, + "grad_norm": 2.094288874517777, + "learning_rate": 8.991325437783627e-07, + "loss": 0.9296, + "step": 36846 + }, + { + "epoch": 0.87, + "grad_norm": 2.1524380291776035, + "learning_rate": 8.988163524729165e-07, + "loss": 0.9124, + "step": 36847 + }, + { + "epoch": 0.87, + "grad_norm": 1.908953005841012, + "learning_rate": 8.985002141574406e-07, + "loss": 0.8406, + "step": 36848 + }, + { + "epoch": 0.87, + "grad_norm": 2.1441742155995533, + "learning_rate": 8.981841288337778e-07, + "loss": 0.9692, + "step": 36849 + }, + { + "epoch": 0.87, + "grad_norm": 2.1543197221011368, + "learning_rate": 8.978680965037712e-07, + "loss": 0.9779, + "step": 36850 + }, + { + "epoch": 0.87, + "grad_norm": 2.0556581142593187, + "learning_rate": 8.975521171692569e-07, + "loss": 0.9452, + "step": 36851 + }, + { + "epoch": 0.87, + "grad_norm": 1.0865154195088873, + "learning_rate": 8.972361908320748e-07, + "loss": 0.9579, + "step": 36852 + }, + { + "epoch": 0.87, + "grad_norm": 1.0406924519071405, + "learning_rate": 8.969203174940655e-07, + "loss": 0.9479, + "step": 36853 + }, + { + "epoch": 0.87, + "grad_norm": 1.9915397731723905, + "learning_rate": 8.966044971570698e-07, + "loss": 0.954, + "step": 36854 + }, + { + "epoch": 0.87, + "grad_norm": 1.070571023068006, + "learning_rate": 8.96288729822925e-07, + "loss": 0.9844, + "step": 36855 + }, + { + "epoch": 0.87, + "grad_norm": 2.252287724650261, + "learning_rate": 8.959730154934676e-07, + "loss": 1.0932, + "step": 36856 + }, + { + "epoch": 0.87, + "grad_norm": 2.120301413032708, + "learning_rate": 8.956573541705393e-07, + "loss": 0.9465, + "step": 36857 + }, + { + "epoch": 0.87, + "grad_norm": 1.9514287549446223, + "learning_rate": 8.953417458559743e-07, + "loss": 0.9482, + "step": 36858 + }, + { + "epoch": 0.87, + "grad_norm": 1.8821950704277104, + "learning_rate": 8.950261905516145e-07, + "loss": 0.9735, + "step": 36859 + }, + { + "epoch": 0.87, + "grad_norm": 1.9268596476524587, + "learning_rate": 8.947106882592927e-07, + "loss": 1.0048, + "step": 36860 + }, + { + "epoch": 0.87, + "grad_norm": 2.324277764060021, + "learning_rate": 8.943952389808497e-07, + "loss": 1.0108, + "step": 36861 + }, + { + "epoch": 0.87, + "grad_norm": 2.0367411251360843, + "learning_rate": 8.940798427181185e-07, + "loss": 0.853, + "step": 36862 + }, + { + "epoch": 0.87, + "grad_norm": 1.8906674230068121, + "learning_rate": 8.937644994729388e-07, + "loss": 1.0127, + "step": 36863 + }, + { + "epoch": 0.87, + "grad_norm": 1.9277393143889157, + "learning_rate": 8.934492092471459e-07, + "loss": 0.9987, + "step": 36864 + }, + { + "epoch": 0.87, + "grad_norm": 2.0104695807400628, + "learning_rate": 8.931339720425725e-07, + "loss": 0.8984, + "step": 36865 + }, + { + "epoch": 0.87, + "grad_norm": 1.8287798799273278, + "learning_rate": 8.928187878610561e-07, + "loss": 0.9746, + "step": 36866 + }, + { + "epoch": 0.87, + "grad_norm": 1.8329126553587274, + "learning_rate": 8.925036567044332e-07, + "loss": 0.9282, + "step": 36867 + }, + { + "epoch": 0.87, + "grad_norm": 1.070035471792675, + "learning_rate": 8.921885785745399e-07, + "loss": 0.9153, + "step": 36868 + }, + { + "epoch": 0.87, + "grad_norm": 1.9455192668597898, + "learning_rate": 8.918735534732048e-07, + "loss": 0.8699, + "step": 36869 + }, + { + "epoch": 0.87, + "grad_norm": 2.182630885164431, + "learning_rate": 8.915585814022653e-07, + "loss": 0.8836, + "step": 36870 + }, + { + "epoch": 0.87, + "grad_norm": 1.775171979863527, + "learning_rate": 8.912436623635579e-07, + "loss": 1.0161, + "step": 36871 + }, + { + "epoch": 0.87, + "grad_norm": 2.039170248306159, + "learning_rate": 8.90928796358913e-07, + "loss": 1.2648, + "step": 36872 + }, + { + "epoch": 0.87, + "grad_norm": 1.8668629117773243, + "learning_rate": 8.906139833901629e-07, + "loss": 0.8401, + "step": 36873 + }, + { + "epoch": 0.87, + "grad_norm": 2.1406721777529025, + "learning_rate": 8.902992234591424e-07, + "loss": 1.116, + "step": 36874 + }, + { + "epoch": 0.87, + "grad_norm": 1.8981528308503655, + "learning_rate": 8.899845165676857e-07, + "loss": 0.8592, + "step": 36875 + }, + { + "epoch": 0.87, + "grad_norm": 2.0665007483557067, + "learning_rate": 8.896698627176226e-07, + "loss": 1.0413, + "step": 36876 + }, + { + "epoch": 0.87, + "grad_norm": 2.162213222642451, + "learning_rate": 8.893552619107848e-07, + "loss": 0.9775, + "step": 36877 + }, + { + "epoch": 0.87, + "grad_norm": 4.143197035430127, + "learning_rate": 8.890407141490065e-07, + "loss": 0.8631, + "step": 36878 + }, + { + "epoch": 0.87, + "grad_norm": 1.8782417715733553, + "learning_rate": 8.887262194341162e-07, + "loss": 1.1158, + "step": 36879 + }, + { + "epoch": 0.87, + "grad_norm": 1.0605621903193565, + "learning_rate": 8.884117777679458e-07, + "loss": 0.8561, + "step": 36880 + }, + { + "epoch": 0.87, + "grad_norm": 2.053810176977538, + "learning_rate": 8.880973891523292e-07, + "loss": 1.0362, + "step": 36881 + }, + { + "epoch": 0.87, + "grad_norm": 2.034560389978383, + "learning_rate": 8.877830535890941e-07, + "loss": 1.0778, + "step": 36882 + }, + { + "epoch": 0.87, + "grad_norm": 2.173550683822319, + "learning_rate": 8.874687710800689e-07, + "loss": 1.0738, + "step": 36883 + }, + { + "epoch": 0.87, + "grad_norm": 1.7612948206288572, + "learning_rate": 8.871545416270855e-07, + "loss": 0.8962, + "step": 36884 + }, + { + "epoch": 0.87, + "grad_norm": 2.3603556906790693, + "learning_rate": 8.868403652319779e-07, + "loss": 0.8846, + "step": 36885 + }, + { + "epoch": 0.87, + "grad_norm": 2.2558969500743977, + "learning_rate": 8.865262418965659e-07, + "loss": 0.9016, + "step": 36886 + }, + { + "epoch": 0.87, + "grad_norm": 2.9221101430537217, + "learning_rate": 8.862121716226846e-07, + "loss": 0.8994, + "step": 36887 + }, + { + "epoch": 0.87, + "grad_norm": 1.994028775644627, + "learning_rate": 8.858981544121614e-07, + "loss": 1.1292, + "step": 36888 + }, + { + "epoch": 0.87, + "grad_norm": 2.169970825902437, + "learning_rate": 8.85584190266825e-07, + "loss": 1.0598, + "step": 36889 + }, + { + "epoch": 0.87, + "grad_norm": 2.1962404721712683, + "learning_rate": 8.852702791885048e-07, + "loss": 0.8906, + "step": 36890 + }, + { + "epoch": 0.87, + "grad_norm": 2.085008275948603, + "learning_rate": 8.84956421179024e-07, + "loss": 1.0544, + "step": 36891 + }, + { + "epoch": 0.87, + "grad_norm": 2.2104129147861045, + "learning_rate": 8.846426162402144e-07, + "loss": 0.846, + "step": 36892 + }, + { + "epoch": 0.87, + "grad_norm": 2.4445947660749527, + "learning_rate": 8.843288643739001e-07, + "loss": 0.8344, + "step": 36893 + }, + { + "epoch": 0.87, + "grad_norm": 2.4844147128703815, + "learning_rate": 8.840151655819107e-07, + "loss": 0.8008, + "step": 36894 + }, + { + "epoch": 0.87, + "grad_norm": 1.8172227458091117, + "learning_rate": 8.837015198660681e-07, + "loss": 0.9419, + "step": 36895 + }, + { + "epoch": 0.87, + "grad_norm": 2.1597558397062624, + "learning_rate": 8.833879272282053e-07, + "loss": 0.9892, + "step": 36896 + }, + { + "epoch": 0.87, + "grad_norm": 1.9437568314421145, + "learning_rate": 8.830743876701408e-07, + "loss": 0.8917, + "step": 36897 + }, + { + "epoch": 0.87, + "grad_norm": 2.098861459160369, + "learning_rate": 8.827609011937066e-07, + "loss": 0.945, + "step": 36898 + }, + { + "epoch": 0.87, + "grad_norm": 1.9546891814267713, + "learning_rate": 8.824474678007244e-07, + "loss": 0.9733, + "step": 36899 + }, + { + "epoch": 0.87, + "grad_norm": 1.8477362891764066, + "learning_rate": 8.821340874930173e-07, + "loss": 0.9741, + "step": 36900 + }, + { + "epoch": 0.87, + "grad_norm": 1.8422986405943522, + "learning_rate": 8.818207602724127e-07, + "loss": 0.9779, + "step": 36901 + }, + { + "epoch": 0.87, + "grad_norm": 2.1165078348318613, + "learning_rate": 8.815074861407358e-07, + "loss": 0.8948, + "step": 36902 + }, + { + "epoch": 0.87, + "grad_norm": 2.5860695886226566, + "learning_rate": 8.811942650998096e-07, + "loss": 0.8729, + "step": 36903 + }, + { + "epoch": 0.87, + "grad_norm": 1.0360380098123994, + "learning_rate": 8.808810971514559e-07, + "loss": 0.9406, + "step": 36904 + }, + { + "epoch": 0.87, + "grad_norm": 2.0845061236054123, + "learning_rate": 8.805679822974988e-07, + "loss": 0.8325, + "step": 36905 + }, + { + "epoch": 0.87, + "grad_norm": 1.7321802746147257, + "learning_rate": 8.802549205397637e-07, + "loss": 0.7594, + "step": 36906 + }, + { + "epoch": 0.87, + "grad_norm": 1.0471049412542626, + "learning_rate": 8.799419118800722e-07, + "loss": 0.9425, + "step": 36907 + }, + { + "epoch": 0.87, + "grad_norm": 2.023656207826972, + "learning_rate": 8.796289563202454e-07, + "loss": 1.0303, + "step": 36908 + }, + { + "epoch": 0.87, + "grad_norm": 2.077906770895339, + "learning_rate": 8.793160538621048e-07, + "loss": 1.0005, + "step": 36909 + }, + { + "epoch": 0.87, + "grad_norm": 1.8908636325167374, + "learning_rate": 8.790032045074759e-07, + "loss": 0.9913, + "step": 36910 + }, + { + "epoch": 0.87, + "grad_norm": 1.933612465406645, + "learning_rate": 8.786904082581782e-07, + "loss": 1.071, + "step": 36911 + }, + { + "epoch": 0.87, + "grad_norm": 1.9820584994163843, + "learning_rate": 8.783776651160303e-07, + "loss": 1.062, + "step": 36912 + }, + { + "epoch": 0.87, + "grad_norm": 2.14094670505724, + "learning_rate": 8.780649750828585e-07, + "loss": 0.9921, + "step": 36913 + }, + { + "epoch": 0.87, + "grad_norm": 1.9441799565156195, + "learning_rate": 8.77752338160478e-07, + "loss": 1.0728, + "step": 36914 + }, + { + "epoch": 0.87, + "grad_norm": 1.9872090555655293, + "learning_rate": 8.774397543507129e-07, + "loss": 0.9132, + "step": 36915 + }, + { + "epoch": 0.87, + "grad_norm": 2.009556611052668, + "learning_rate": 8.771272236553796e-07, + "loss": 0.8178, + "step": 36916 + }, + { + "epoch": 0.87, + "grad_norm": 2.1998872542882992, + "learning_rate": 8.768147460763021e-07, + "loss": 1.0156, + "step": 36917 + }, + { + "epoch": 0.87, + "grad_norm": 2.0317919401349886, + "learning_rate": 8.765023216152969e-07, + "loss": 1.0114, + "step": 36918 + }, + { + "epoch": 0.87, + "grad_norm": 1.8467232998076113, + "learning_rate": 8.761899502741833e-07, + "loss": 0.9878, + "step": 36919 + }, + { + "epoch": 0.87, + "grad_norm": 2.2380283333069606, + "learning_rate": 8.758776320547812e-07, + "loss": 0.9715, + "step": 36920 + }, + { + "epoch": 0.87, + "grad_norm": 2.722552402789231, + "learning_rate": 8.75565366958907e-07, + "loss": 1.0565, + "step": 36921 + }, + { + "epoch": 0.87, + "grad_norm": 1.9481266482231963, + "learning_rate": 8.752531549883791e-07, + "loss": 0.9369, + "step": 36922 + }, + { + "epoch": 0.87, + "grad_norm": 1.085293727167529, + "learning_rate": 8.749409961450183e-07, + "loss": 0.9251, + "step": 36923 + }, + { + "epoch": 0.87, + "grad_norm": 1.9219751982475601, + "learning_rate": 8.74628890430641e-07, + "loss": 1.0709, + "step": 36924 + }, + { + "epoch": 0.87, + "grad_norm": 2.2361143593047985, + "learning_rate": 8.743168378470601e-07, + "loss": 0.9359, + "step": 36925 + }, + { + "epoch": 0.87, + "grad_norm": 1.9535465456712888, + "learning_rate": 8.740048383960964e-07, + "loss": 1.1379, + "step": 36926 + }, + { + "epoch": 0.87, + "grad_norm": 2.102078216763581, + "learning_rate": 8.736928920795674e-07, + "loss": 0.9486, + "step": 36927 + }, + { + "epoch": 0.87, + "grad_norm": 1.9310130929957308, + "learning_rate": 8.733809988992881e-07, + "loss": 1.0625, + "step": 36928 + }, + { + "epoch": 0.87, + "grad_norm": 2.13776738579873, + "learning_rate": 8.730691588570716e-07, + "loss": 1.0024, + "step": 36929 + }, + { + "epoch": 0.87, + "grad_norm": 1.1309821003779712, + "learning_rate": 8.727573719547367e-07, + "loss": 0.9764, + "step": 36930 + }, + { + "epoch": 0.87, + "grad_norm": 2.290399269596683, + "learning_rate": 8.724456381940994e-07, + "loss": 0.991, + "step": 36931 + }, + { + "epoch": 0.87, + "grad_norm": 1.8285803532150404, + "learning_rate": 8.721339575769727e-07, + "loss": 0.9812, + "step": 36932 + }, + { + "epoch": 0.87, + "grad_norm": 1.9394942689355752, + "learning_rate": 8.71822330105171e-07, + "loss": 0.9475, + "step": 36933 + }, + { + "epoch": 0.87, + "grad_norm": 2.2709285037682534, + "learning_rate": 8.715107557805092e-07, + "loss": 0.9685, + "step": 36934 + }, + { + "epoch": 0.87, + "grad_norm": 3.4799811031166485, + "learning_rate": 8.711992346048038e-07, + "loss": 0.797, + "step": 36935 + }, + { + "epoch": 0.87, + "grad_norm": 1.90667005629353, + "learning_rate": 8.708877665798666e-07, + "loss": 0.8719, + "step": 36936 + }, + { + "epoch": 0.87, + "grad_norm": 2.249304604461074, + "learning_rate": 8.705763517075094e-07, + "loss": 1.0669, + "step": 36937 + }, + { + "epoch": 0.87, + "grad_norm": 1.962229095017031, + "learning_rate": 8.702649899895488e-07, + "loss": 0.9021, + "step": 36938 + }, + { + "epoch": 0.87, + "grad_norm": 1.846955565793186, + "learning_rate": 8.699536814277931e-07, + "loss": 0.9804, + "step": 36939 + }, + { + "epoch": 0.87, + "grad_norm": 2.035074337040275, + "learning_rate": 8.69642426024061e-07, + "loss": 0.9175, + "step": 36940 + }, + { + "epoch": 0.87, + "grad_norm": 1.7980562037410974, + "learning_rate": 8.693312237801588e-07, + "loss": 1.0235, + "step": 36941 + }, + { + "epoch": 0.87, + "grad_norm": 1.7772298587736894, + "learning_rate": 8.690200746979016e-07, + "loss": 0.9694, + "step": 36942 + }, + { + "epoch": 0.87, + "grad_norm": 2.2103730591600597, + "learning_rate": 8.687089787791003e-07, + "loss": 1.0967, + "step": 36943 + }, + { + "epoch": 0.87, + "grad_norm": 2.0922108095062146, + "learning_rate": 8.683979360255668e-07, + "loss": 1.0434, + "step": 36944 + }, + { + "epoch": 0.87, + "grad_norm": 1.8538117626910446, + "learning_rate": 8.680869464391129e-07, + "loss": 0.9268, + "step": 36945 + }, + { + "epoch": 0.87, + "grad_norm": 2.040939605588525, + "learning_rate": 8.677760100215449e-07, + "loss": 1.0201, + "step": 36946 + }, + { + "epoch": 0.87, + "grad_norm": 1.9104555050148293, + "learning_rate": 8.674651267746769e-07, + "loss": 1.0308, + "step": 36947 + }, + { + "epoch": 0.87, + "grad_norm": 1.0593592057678844, + "learning_rate": 8.671542967003188e-07, + "loss": 0.9788, + "step": 36948 + }, + { + "epoch": 0.87, + "grad_norm": 1.8752931772318453, + "learning_rate": 8.668435198002812e-07, + "loss": 1.0402, + "step": 36949 + }, + { + "epoch": 0.87, + "grad_norm": 2.0026058228586754, + "learning_rate": 8.665327960763692e-07, + "loss": 0.9797, + "step": 36950 + }, + { + "epoch": 0.87, + "grad_norm": 3.128518759427267, + "learning_rate": 8.66222125530396e-07, + "loss": 0.9606, + "step": 36951 + }, + { + "epoch": 0.87, + "grad_norm": 2.2436518321488124, + "learning_rate": 8.659115081641712e-07, + "loss": 0.8713, + "step": 36952 + }, + { + "epoch": 0.87, + "grad_norm": 1.8142254993183442, + "learning_rate": 8.656009439794998e-07, + "loss": 1.0323, + "step": 36953 + }, + { + "epoch": 0.87, + "grad_norm": 1.9277025557006071, + "learning_rate": 8.652904329781908e-07, + "loss": 0.8707, + "step": 36954 + }, + { + "epoch": 0.87, + "grad_norm": 1.8756787394088759, + "learning_rate": 8.649799751620536e-07, + "loss": 0.9032, + "step": 36955 + }, + { + "epoch": 0.87, + "grad_norm": 2.2236878381885132, + "learning_rate": 8.646695705328955e-07, + "loss": 0.9697, + "step": 36956 + }, + { + "epoch": 0.87, + "grad_norm": 2.2125073497245706, + "learning_rate": 8.643592190925243e-07, + "loss": 0.9324, + "step": 36957 + }, + { + "epoch": 0.87, + "grad_norm": 1.0639001698421062, + "learning_rate": 8.640489208427438e-07, + "loss": 0.8972, + "step": 36958 + }, + { + "epoch": 0.87, + "grad_norm": 2.6568312063005375, + "learning_rate": 8.637386757853649e-07, + "loss": 0.9626, + "step": 36959 + }, + { + "epoch": 0.87, + "grad_norm": 1.8944346325241876, + "learning_rate": 8.634284839221907e-07, + "loss": 0.8971, + "step": 36960 + }, + { + "epoch": 0.87, + "grad_norm": 1.9372942701352498, + "learning_rate": 8.631183452550295e-07, + "loss": 1.0387, + "step": 36961 + }, + { + "epoch": 0.87, + "grad_norm": 1.9240714324757688, + "learning_rate": 8.628082597856835e-07, + "loss": 0.9367, + "step": 36962 + }, + { + "epoch": 0.87, + "grad_norm": 2.101549098440075, + "learning_rate": 8.624982275159633e-07, + "loss": 1.0716, + "step": 36963 + }, + { + "epoch": 0.87, + "grad_norm": 1.941994773933425, + "learning_rate": 8.621882484476695e-07, + "loss": 1.0784, + "step": 36964 + }, + { + "epoch": 0.87, + "grad_norm": 2.0458910582325123, + "learning_rate": 8.618783225826088e-07, + "loss": 0.9922, + "step": 36965 + }, + { + "epoch": 0.87, + "grad_norm": 1.9649787005907444, + "learning_rate": 8.615684499225884e-07, + "loss": 0.9902, + "step": 36966 + }, + { + "epoch": 0.87, + "grad_norm": 1.894407283645095, + "learning_rate": 8.61258630469407e-07, + "loss": 1.053, + "step": 36967 + }, + { + "epoch": 0.87, + "grad_norm": 1.8879511391448105, + "learning_rate": 8.609488642248709e-07, + "loss": 1.0515, + "step": 36968 + }, + { + "epoch": 0.87, + "grad_norm": 1.76200217808026, + "learning_rate": 8.606391511907841e-07, + "loss": 0.8767, + "step": 36969 + }, + { + "epoch": 0.87, + "grad_norm": 1.081578039931093, + "learning_rate": 8.60329491368953e-07, + "loss": 0.9696, + "step": 36970 + }, + { + "epoch": 0.87, + "grad_norm": 1.902126973486207, + "learning_rate": 8.60019884761173e-07, + "loss": 0.9914, + "step": 36971 + }, + { + "epoch": 0.87, + "grad_norm": 1.981532410685682, + "learning_rate": 8.597103313692512e-07, + "loss": 1.0127, + "step": 36972 + }, + { + "epoch": 0.87, + "grad_norm": 1.954400662227814, + "learning_rate": 8.59400831194992e-07, + "loss": 1.0352, + "step": 36973 + }, + { + "epoch": 0.87, + "grad_norm": 1.1282902673805517, + "learning_rate": 8.590913842401949e-07, + "loss": 0.9101, + "step": 36974 + }, + { + "epoch": 0.87, + "grad_norm": 2.081578241369764, + "learning_rate": 8.587819905066597e-07, + "loss": 0.9647, + "step": 36975 + }, + { + "epoch": 0.87, + "grad_norm": 2.177408520200037, + "learning_rate": 8.584726499961893e-07, + "loss": 1.0506, + "step": 36976 + }, + { + "epoch": 0.87, + "grad_norm": 2.0421198456556824, + "learning_rate": 8.581633627105879e-07, + "loss": 0.9761, + "step": 36977 + }, + { + "epoch": 0.87, + "grad_norm": 2.078939774361028, + "learning_rate": 8.578541286516506e-07, + "loss": 0.9507, + "step": 36978 + }, + { + "epoch": 0.87, + "grad_norm": 1.9412222685479843, + "learning_rate": 8.575449478211839e-07, + "loss": 0.9231, + "step": 36979 + }, + { + "epoch": 0.87, + "grad_norm": 1.9867240025756157, + "learning_rate": 8.572358202209851e-07, + "loss": 0.7536, + "step": 36980 + }, + { + "epoch": 0.87, + "grad_norm": 1.9733271329063284, + "learning_rate": 8.569267458528519e-07, + "loss": 0.8752, + "step": 36981 + }, + { + "epoch": 0.87, + "grad_norm": 2.033746962343726, + "learning_rate": 8.566177247185847e-07, + "loss": 0.9132, + "step": 36982 + }, + { + "epoch": 0.87, + "grad_norm": 1.9392205653060324, + "learning_rate": 8.563087568199868e-07, + "loss": 0.9831, + "step": 36983 + }, + { + "epoch": 0.87, + "grad_norm": 2.0438132233138595, + "learning_rate": 8.559998421588545e-07, + "loss": 1.0555, + "step": 36984 + }, + { + "epoch": 0.87, + "grad_norm": 1.8420157648862336, + "learning_rate": 8.55690980736984e-07, + "loss": 1.0746, + "step": 36985 + }, + { + "epoch": 0.87, + "grad_norm": 1.9200951305481015, + "learning_rate": 8.553821725561751e-07, + "loss": 1.0047, + "step": 36986 + }, + { + "epoch": 0.87, + "grad_norm": 2.1127650303378647, + "learning_rate": 8.550734176182285e-07, + "loss": 0.9183, + "step": 36987 + }, + { + "epoch": 0.87, + "grad_norm": 1.9537975524229616, + "learning_rate": 8.547647159249395e-07, + "loss": 1.014, + "step": 36988 + }, + { + "epoch": 0.87, + "grad_norm": 1.9654518741296283, + "learning_rate": 8.544560674781033e-07, + "loss": 1.015, + "step": 36989 + }, + { + "epoch": 0.87, + "grad_norm": 2.4694410042488077, + "learning_rate": 8.541474722795207e-07, + "loss": 0.9048, + "step": 36990 + }, + { + "epoch": 0.87, + "grad_norm": 2.0591600710639937, + "learning_rate": 8.538389303309879e-07, + "loss": 0.9349, + "step": 36991 + }, + { + "epoch": 0.87, + "grad_norm": 1.8975695913557702, + "learning_rate": 8.535304416343005e-07, + "loss": 0.9768, + "step": 36992 + }, + { + "epoch": 0.87, + "grad_norm": 1.8945722963247003, + "learning_rate": 8.532220061912533e-07, + "loss": 0.9485, + "step": 36993 + }, + { + "epoch": 0.87, + "grad_norm": 1.9211644493677908, + "learning_rate": 8.529136240036439e-07, + "loss": 0.8809, + "step": 36994 + }, + { + "epoch": 0.87, + "grad_norm": 1.8040439109778517, + "learning_rate": 8.526052950732666e-07, + "loss": 0.9482, + "step": 36995 + }, + { + "epoch": 0.87, + "grad_norm": 2.6329981040018664, + "learning_rate": 8.522970194019187e-07, + "loss": 1.0569, + "step": 36996 + }, + { + "epoch": 0.87, + "grad_norm": 1.8391237809197527, + "learning_rate": 8.51988796991392e-07, + "loss": 0.8704, + "step": 36997 + }, + { + "epoch": 0.87, + "grad_norm": 2.089563503815871, + "learning_rate": 8.516806278434852e-07, + "loss": 0.8616, + "step": 36998 + }, + { + "epoch": 0.87, + "grad_norm": 2.0295185853977595, + "learning_rate": 8.51372511959988e-07, + "loss": 0.9927, + "step": 36999 + }, + { + "epoch": 0.87, + "grad_norm": 2.128066842367005, + "learning_rate": 8.510644493426979e-07, + "loss": 1.0025, + "step": 37000 + }, + { + "epoch": 0.87, + "grad_norm": 2.1983060105174754, + "learning_rate": 8.507564399934065e-07, + "loss": 0.9932, + "step": 37001 + }, + { + "epoch": 0.87, + "grad_norm": 1.09546835274513, + "learning_rate": 8.504484839139071e-07, + "loss": 0.9141, + "step": 37002 + }, + { + "epoch": 0.87, + "grad_norm": 2.1705914868964133, + "learning_rate": 8.501405811059926e-07, + "loss": 0.9012, + "step": 37003 + }, + { + "epoch": 0.87, + "grad_norm": 2.0795248434141995, + "learning_rate": 8.498327315714583e-07, + "loss": 0.8795, + "step": 37004 + }, + { + "epoch": 0.87, + "grad_norm": 1.860548620850887, + "learning_rate": 8.495249353120949e-07, + "loss": 0.9912, + "step": 37005 + }, + { + "epoch": 0.87, + "grad_norm": 1.1098741179665985, + "learning_rate": 8.492171923296922e-07, + "loss": 0.9983, + "step": 37006 + }, + { + "epoch": 0.87, + "grad_norm": 1.9279275262830382, + "learning_rate": 8.489095026260452e-07, + "loss": 0.8483, + "step": 37007 + }, + { + "epoch": 0.87, + "grad_norm": 1.9487439639953523, + "learning_rate": 8.48601866202945e-07, + "loss": 1.0103, + "step": 37008 + }, + { + "epoch": 0.87, + "grad_norm": 2.091042122440498, + "learning_rate": 8.482942830621821e-07, + "loss": 1.0097, + "step": 37009 + }, + { + "epoch": 0.87, + "grad_norm": 2.4628094468627553, + "learning_rate": 8.479867532055452e-07, + "loss": 1.0641, + "step": 37010 + }, + { + "epoch": 0.87, + "grad_norm": 1.7620587375065089, + "learning_rate": 8.476792766348274e-07, + "loss": 0.9469, + "step": 37011 + }, + { + "epoch": 0.87, + "grad_norm": 1.1372471250957223, + "learning_rate": 8.473718533518205e-07, + "loss": 0.9058, + "step": 37012 + }, + { + "epoch": 0.87, + "grad_norm": 1.9812451807235745, + "learning_rate": 8.470644833583119e-07, + "loss": 1.022, + "step": 37013 + }, + { + "epoch": 0.87, + "grad_norm": 1.1011230365943816, + "learning_rate": 8.467571666560903e-07, + "loss": 0.9258, + "step": 37014 + }, + { + "epoch": 0.87, + "grad_norm": 1.854336841370825, + "learning_rate": 8.464499032469475e-07, + "loss": 0.9934, + "step": 37015 + }, + { + "epoch": 0.87, + "grad_norm": 1.8586204224461913, + "learning_rate": 8.461426931326699e-07, + "loss": 0.9935, + "step": 37016 + }, + { + "epoch": 0.87, + "grad_norm": 2.104943874857789, + "learning_rate": 8.458355363150494e-07, + "loss": 0.9668, + "step": 37017 + }, + { + "epoch": 0.87, + "grad_norm": 2.241854598194891, + "learning_rate": 8.455284327958702e-07, + "loss": 1.0013, + "step": 37018 + }, + { + "epoch": 0.87, + "grad_norm": 1.8179906470342113, + "learning_rate": 8.452213825769251e-07, + "loss": 0.8275, + "step": 37019 + }, + { + "epoch": 0.87, + "grad_norm": 2.325109043828064, + "learning_rate": 8.449143856599973e-07, + "loss": 0.9112, + "step": 37020 + }, + { + "epoch": 0.87, + "grad_norm": 1.9096005426513836, + "learning_rate": 8.446074420468786e-07, + "loss": 1.0025, + "step": 37021 + }, + { + "epoch": 0.87, + "grad_norm": 1.1119226102150321, + "learning_rate": 8.443005517393543e-07, + "loss": 1.0183, + "step": 37022 + }, + { + "epoch": 0.87, + "grad_norm": 2.7864736123943628, + "learning_rate": 8.439937147392086e-07, + "loss": 1.0035, + "step": 37023 + }, + { + "epoch": 0.87, + "grad_norm": 1.8518280808220389, + "learning_rate": 8.436869310482298e-07, + "loss": 0.812, + "step": 37024 + }, + { + "epoch": 0.87, + "grad_norm": 1.8908476530323153, + "learning_rate": 8.433802006682068e-07, + "loss": 0.9602, + "step": 37025 + }, + { + "epoch": 0.87, + "grad_norm": 2.500396861394266, + "learning_rate": 8.430735236009225e-07, + "loss": 0.9433, + "step": 37026 + }, + { + "epoch": 0.87, + "grad_norm": 2.1911796307726594, + "learning_rate": 8.427668998481619e-07, + "loss": 1.0186, + "step": 37027 + }, + { + "epoch": 0.87, + "grad_norm": 1.0975446587054773, + "learning_rate": 8.424603294117118e-07, + "loss": 0.9459, + "step": 37028 + }, + { + "epoch": 0.87, + "grad_norm": 1.9202785374095628, + "learning_rate": 8.421538122933593e-07, + "loss": 0.9624, + "step": 37029 + }, + { + "epoch": 0.87, + "grad_norm": 2.128395364207691, + "learning_rate": 8.418473484948853e-07, + "loss": 1.0104, + "step": 37030 + }, + { + "epoch": 0.87, + "grad_norm": 2.4441199096718154, + "learning_rate": 8.41540938018074e-07, + "loss": 0.9887, + "step": 37031 + }, + { + "epoch": 0.87, + "grad_norm": 1.9453700655056885, + "learning_rate": 8.412345808647115e-07, + "loss": 0.9448, + "step": 37032 + }, + { + "epoch": 0.87, + "grad_norm": 2.2805555563238054, + "learning_rate": 8.409282770365823e-07, + "loss": 1.0075, + "step": 37033 + }, + { + "epoch": 0.87, + "grad_norm": 2.0022293971329823, + "learning_rate": 8.406220265354681e-07, + "loss": 0.8165, + "step": 37034 + }, + { + "epoch": 0.87, + "grad_norm": 1.9344003996063528, + "learning_rate": 8.403158293631508e-07, + "loss": 1.0453, + "step": 37035 + }, + { + "epoch": 0.87, + "grad_norm": 2.188757430926363, + "learning_rate": 8.400096855214157e-07, + "loss": 1.0455, + "step": 37036 + }, + { + "epoch": 0.87, + "grad_norm": 1.9085802894060824, + "learning_rate": 8.397035950120435e-07, + "loss": 0.835, + "step": 37037 + }, + { + "epoch": 0.87, + "grad_norm": 2.156445425704015, + "learning_rate": 8.393975578368186e-07, + "loss": 0.9783, + "step": 37038 + }, + { + "epoch": 0.87, + "grad_norm": 2.024139243029042, + "learning_rate": 8.390915739975192e-07, + "loss": 1.0352, + "step": 37039 + }, + { + "epoch": 0.87, + "grad_norm": 1.9533228479551434, + "learning_rate": 8.387856434959307e-07, + "loss": 0.9744, + "step": 37040 + }, + { + "epoch": 0.87, + "grad_norm": 1.9305231372034308, + "learning_rate": 8.384797663338307e-07, + "loss": 1.0359, + "step": 37041 + }, + { + "epoch": 0.87, + "grad_norm": 3.1621715087158306, + "learning_rate": 8.381739425130042e-07, + "loss": 0.9833, + "step": 37042 + }, + { + "epoch": 0.87, + "grad_norm": 1.1279122045739112, + "learning_rate": 8.3786817203523e-07, + "loss": 0.8635, + "step": 37043 + }, + { + "epoch": 0.87, + "grad_norm": 1.8909097516923963, + "learning_rate": 8.375624549022854e-07, + "loss": 1.0233, + "step": 37044 + }, + { + "epoch": 0.87, + "grad_norm": 2.0916895748715594, + "learning_rate": 8.372567911159535e-07, + "loss": 0.9385, + "step": 37045 + }, + { + "epoch": 0.87, + "grad_norm": 2.0647911207417735, + "learning_rate": 8.369511806780151e-07, + "loss": 1.0639, + "step": 37046 + }, + { + "epoch": 0.87, + "grad_norm": 1.8751361082266893, + "learning_rate": 8.366456235902476e-07, + "loss": 1.0333, + "step": 37047 + }, + { + "epoch": 0.87, + "grad_norm": 3.5307713822892133, + "learning_rate": 8.363401198544285e-07, + "loss": 0.9469, + "step": 37048 + }, + { + "epoch": 0.87, + "grad_norm": 2.0067399223590474, + "learning_rate": 8.360346694723387e-07, + "loss": 0.876, + "step": 37049 + }, + { + "epoch": 0.87, + "grad_norm": 2.1916710322065516, + "learning_rate": 8.357292724457577e-07, + "loss": 1.0651, + "step": 37050 + }, + { + "epoch": 0.87, + "grad_norm": 2.072960249421791, + "learning_rate": 8.35423928776462e-07, + "loss": 1.0627, + "step": 37051 + }, + { + "epoch": 0.87, + "grad_norm": 1.964000599088826, + "learning_rate": 8.351186384662291e-07, + "loss": 1.0507, + "step": 37052 + }, + { + "epoch": 0.87, + "grad_norm": 1.8894313780000573, + "learning_rate": 8.348134015168364e-07, + "loss": 1.0132, + "step": 37053 + }, + { + "epoch": 0.87, + "grad_norm": 1.7811820833299188, + "learning_rate": 8.345082179300634e-07, + "loss": 0.9265, + "step": 37054 + }, + { + "epoch": 0.87, + "grad_norm": 2.116296634426407, + "learning_rate": 8.342030877076856e-07, + "loss": 1.0286, + "step": 37055 + }, + { + "epoch": 0.87, + "grad_norm": 1.1953994931248018, + "learning_rate": 8.338980108514772e-07, + "loss": 0.9711, + "step": 37056 + }, + { + "epoch": 0.87, + "grad_norm": 1.917038796976121, + "learning_rate": 8.335929873632164e-07, + "loss": 1.011, + "step": 37057 + }, + { + "epoch": 0.87, + "grad_norm": 2.6650065458165453, + "learning_rate": 8.332880172446811e-07, + "loss": 1.0234, + "step": 37058 + }, + { + "epoch": 0.87, + "grad_norm": 2.1075675612826994, + "learning_rate": 8.329831004976452e-07, + "loss": 0.9247, + "step": 37059 + }, + { + "epoch": 0.87, + "grad_norm": 1.8068453172120267, + "learning_rate": 8.326782371238818e-07, + "loss": 0.9353, + "step": 37060 + }, + { + "epoch": 0.87, + "grad_norm": 2.639128874047261, + "learning_rate": 8.323734271251704e-07, + "loss": 0.9339, + "step": 37061 + }, + { + "epoch": 0.87, + "grad_norm": 1.8976155160185009, + "learning_rate": 8.32068670503281e-07, + "loss": 0.9498, + "step": 37062 + }, + { + "epoch": 0.87, + "grad_norm": 2.3259934850204163, + "learning_rate": 8.317639672599909e-07, + "loss": 0.9615, + "step": 37063 + }, + { + "epoch": 0.87, + "grad_norm": 1.9938959916568324, + "learning_rate": 8.314593173970753e-07, + "loss": 1.0779, + "step": 37064 + }, + { + "epoch": 0.87, + "grad_norm": 1.821132468581538, + "learning_rate": 8.311547209163052e-07, + "loss": 0.8846, + "step": 37065 + }, + { + "epoch": 0.87, + "grad_norm": 1.0816075567287553, + "learning_rate": 8.308501778194545e-07, + "loss": 0.9669, + "step": 37066 + }, + { + "epoch": 0.87, + "grad_norm": 2.00629360674378, + "learning_rate": 8.305456881082963e-07, + "loss": 0.962, + "step": 37067 + }, + { + "epoch": 0.87, + "grad_norm": 2.1961865326440035, + "learning_rate": 8.302412517846081e-07, + "loss": 0.9726, + "step": 37068 + }, + { + "epoch": 0.87, + "grad_norm": 1.1288945420608454, + "learning_rate": 8.299368688501552e-07, + "loss": 0.9506, + "step": 37069 + }, + { + "epoch": 0.87, + "grad_norm": 1.91862358364017, + "learning_rate": 8.296325393067128e-07, + "loss": 0.9472, + "step": 37070 + }, + { + "epoch": 0.87, + "grad_norm": 2.0865666115179184, + "learning_rate": 8.293282631560529e-07, + "loss": 0.9761, + "step": 37071 + }, + { + "epoch": 0.87, + "grad_norm": 1.9304674319891288, + "learning_rate": 8.290240403999505e-07, + "loss": 0.9027, + "step": 37072 + }, + { + "epoch": 0.87, + "grad_norm": 2.0363554152468044, + "learning_rate": 8.287198710401701e-07, + "loss": 1.0219, + "step": 37073 + }, + { + "epoch": 0.87, + "grad_norm": 1.1196891519539047, + "learning_rate": 8.284157550784866e-07, + "loss": 0.9361, + "step": 37074 + }, + { + "epoch": 0.87, + "grad_norm": 1.9738175547172567, + "learning_rate": 8.281116925166721e-07, + "loss": 0.9207, + "step": 37075 + }, + { + "epoch": 0.87, + "grad_norm": 1.9623861296772434, + "learning_rate": 8.278076833564929e-07, + "loss": 0.9618, + "step": 37076 + }, + { + "epoch": 0.87, + "grad_norm": 1.972197812816276, + "learning_rate": 8.275037275997233e-07, + "loss": 1.0841, + "step": 37077 + }, + { + "epoch": 0.87, + "grad_norm": 2.0669326303342417, + "learning_rate": 8.271998252481283e-07, + "loss": 1.0178, + "step": 37078 + }, + { + "epoch": 0.87, + "grad_norm": 2.0427383884235564, + "learning_rate": 8.268959763034834e-07, + "loss": 0.9794, + "step": 37079 + }, + { + "epoch": 0.87, + "grad_norm": 2.080148041640355, + "learning_rate": 8.265921807675515e-07, + "loss": 1.0392, + "step": 37080 + }, + { + "epoch": 0.87, + "grad_norm": 2.1570223668042297, + "learning_rate": 8.262884386421055e-07, + "loss": 0.9933, + "step": 37081 + }, + { + "epoch": 0.87, + "grad_norm": 1.9841219299967883, + "learning_rate": 8.25984749928913e-07, + "loss": 1.0275, + "step": 37082 + }, + { + "epoch": 0.87, + "grad_norm": 1.965055515431618, + "learning_rate": 8.256811146297405e-07, + "loss": 0.9841, + "step": 37083 + }, + { + "epoch": 0.87, + "grad_norm": 2.036153672017707, + "learning_rate": 8.253775327463565e-07, + "loss": 0.9732, + "step": 37084 + }, + { + "epoch": 0.87, + "grad_norm": 1.9537926111050472, + "learning_rate": 8.250740042805317e-07, + "loss": 0.8782, + "step": 37085 + }, + { + "epoch": 0.87, + "grad_norm": 3.0802471591343017, + "learning_rate": 8.247705292340302e-07, + "loss": 0.9527, + "step": 37086 + }, + { + "epoch": 0.87, + "grad_norm": 1.918434952676576, + "learning_rate": 8.244671076086174e-07, + "loss": 1.0089, + "step": 37087 + }, + { + "epoch": 0.87, + "grad_norm": 2.292354473840117, + "learning_rate": 8.241637394060619e-07, + "loss": 0.9703, + "step": 37088 + }, + { + "epoch": 0.87, + "grad_norm": 1.0998186832039445, + "learning_rate": 8.238604246281323e-07, + "loss": 0.8984, + "step": 37089 + }, + { + "epoch": 0.87, + "grad_norm": 1.9163188755105798, + "learning_rate": 8.235571632765927e-07, + "loss": 1.0066, + "step": 37090 + }, + { + "epoch": 0.87, + "grad_norm": 2.0318653318619706, + "learning_rate": 8.232539553532071e-07, + "loss": 0.9473, + "step": 37091 + }, + { + "epoch": 0.87, + "grad_norm": 2.041787529955552, + "learning_rate": 8.229508008597409e-07, + "loss": 1.0605, + "step": 37092 + }, + { + "epoch": 0.87, + "grad_norm": 2.055710202036823, + "learning_rate": 8.226476997979638e-07, + "loss": 0.9358, + "step": 37093 + }, + { + "epoch": 0.87, + "grad_norm": 2.0778638072362408, + "learning_rate": 8.223446521696366e-07, + "loss": 1.0288, + "step": 37094 + }, + { + "epoch": 0.87, + "grad_norm": 1.9779940117160697, + "learning_rate": 8.220416579765222e-07, + "loss": 0.9006, + "step": 37095 + }, + { + "epoch": 0.87, + "grad_norm": 2.5198326253105816, + "learning_rate": 8.217387172203895e-07, + "loss": 1.0244, + "step": 37096 + }, + { + "epoch": 0.87, + "grad_norm": 1.8123686048907959, + "learning_rate": 8.214358299029968e-07, + "loss": 1.1294, + "step": 37097 + }, + { + "epoch": 0.87, + "grad_norm": 1.075912724990319, + "learning_rate": 8.211329960261138e-07, + "loss": 0.9517, + "step": 37098 + }, + { + "epoch": 0.87, + "grad_norm": 1.9682615628995954, + "learning_rate": 8.208302155914982e-07, + "loss": 0.9812, + "step": 37099 + }, + { + "epoch": 0.87, + "grad_norm": 2.2410169336192305, + "learning_rate": 8.205274886009162e-07, + "loss": 0.9983, + "step": 37100 + }, + { + "epoch": 0.87, + "grad_norm": 2.5089683723852514, + "learning_rate": 8.202248150561288e-07, + "loss": 0.9806, + "step": 37101 + }, + { + "epoch": 0.87, + "grad_norm": 1.0657741168440507, + "learning_rate": 8.199221949588998e-07, + "loss": 0.9165, + "step": 37102 + }, + { + "epoch": 0.87, + "grad_norm": 2.635310517088782, + "learning_rate": 8.196196283109903e-07, + "loss": 1.1077, + "step": 37103 + }, + { + "epoch": 0.87, + "grad_norm": 1.1205512250807388, + "learning_rate": 8.193171151141599e-07, + "loss": 0.9561, + "step": 37104 + }, + { + "epoch": 0.87, + "grad_norm": 1.8448752336534802, + "learning_rate": 8.190146553701716e-07, + "loss": 0.8172, + "step": 37105 + }, + { + "epoch": 0.87, + "grad_norm": 1.99488882579826, + "learning_rate": 8.187122490807897e-07, + "loss": 0.9683, + "step": 37106 + }, + { + "epoch": 0.87, + "grad_norm": 1.134887388191785, + "learning_rate": 8.184098962477705e-07, + "loss": 0.8932, + "step": 37107 + }, + { + "epoch": 0.87, + "grad_norm": 2.086797400816485, + "learning_rate": 8.181075968728746e-07, + "loss": 1.0463, + "step": 37108 + }, + { + "epoch": 0.87, + "grad_norm": 1.9122458829069908, + "learning_rate": 8.178053509578631e-07, + "loss": 1.0701, + "step": 37109 + }, + { + "epoch": 0.87, + "grad_norm": 2.002628252585672, + "learning_rate": 8.175031585044968e-07, + "loss": 0.9158, + "step": 37110 + }, + { + "epoch": 0.87, + "grad_norm": 2.848029609500507, + "learning_rate": 8.172010195145352e-07, + "loss": 1.041, + "step": 37111 + }, + { + "epoch": 0.87, + "grad_norm": 1.0364503214446987, + "learning_rate": 8.168989339897348e-07, + "loss": 0.9404, + "step": 37112 + }, + { + "epoch": 0.87, + "grad_norm": 2.146692990556052, + "learning_rate": 8.165969019318554e-07, + "loss": 0.9741, + "step": 37113 + }, + { + "epoch": 0.87, + "grad_norm": 2.3162426672672156, + "learning_rate": 8.162949233426587e-07, + "loss": 1.073, + "step": 37114 + }, + { + "epoch": 0.87, + "grad_norm": 1.7729841700044955, + "learning_rate": 8.159929982239001e-07, + "loss": 0.9052, + "step": 37115 + }, + { + "epoch": 0.87, + "grad_norm": 2.0846322182414965, + "learning_rate": 8.156911265773359e-07, + "loss": 0.953, + "step": 37116 + }, + { + "epoch": 0.87, + "grad_norm": 2.4541769071170965, + "learning_rate": 8.153893084047281e-07, + "loss": 0.8802, + "step": 37117 + }, + { + "epoch": 0.87, + "grad_norm": 2.049869231924847, + "learning_rate": 8.150875437078298e-07, + "loss": 0.9968, + "step": 37118 + }, + { + "epoch": 0.87, + "grad_norm": 2.1351307077324706, + "learning_rate": 8.147858324884005e-07, + "loss": 1.005, + "step": 37119 + }, + { + "epoch": 0.87, + "grad_norm": 2.73111992549176, + "learning_rate": 8.144841747481947e-07, + "loss": 1.1075, + "step": 37120 + }, + { + "epoch": 0.87, + "grad_norm": 1.8489456167785572, + "learning_rate": 8.141825704889728e-07, + "loss": 0.87, + "step": 37121 + }, + { + "epoch": 0.87, + "grad_norm": 1.852318573840059, + "learning_rate": 8.138810197124846e-07, + "loss": 0.8501, + "step": 37122 + }, + { + "epoch": 0.87, + "grad_norm": 2.0036178869031906, + "learning_rate": 8.135795224204923e-07, + "loss": 1.0276, + "step": 37123 + }, + { + "epoch": 0.87, + "grad_norm": 2.127942378573936, + "learning_rate": 8.132780786147487e-07, + "loss": 0.9807, + "step": 37124 + }, + { + "epoch": 0.87, + "grad_norm": 1.0970023075347455, + "learning_rate": 8.129766882970069e-07, + "loss": 0.8876, + "step": 37125 + }, + { + "epoch": 0.87, + "grad_norm": 2.4217904414520315, + "learning_rate": 8.126753514690222e-07, + "loss": 1.0052, + "step": 37126 + }, + { + "epoch": 0.87, + "grad_norm": 1.1019759296405738, + "learning_rate": 8.123740681325531e-07, + "loss": 0.9668, + "step": 37127 + }, + { + "epoch": 0.87, + "grad_norm": 1.8555162047290767, + "learning_rate": 8.120728382893505e-07, + "loss": 0.924, + "step": 37128 + }, + { + "epoch": 0.87, + "grad_norm": 2.0943250050046394, + "learning_rate": 8.117716619411675e-07, + "loss": 0.9647, + "step": 37129 + }, + { + "epoch": 0.87, + "grad_norm": 2.089234809306652, + "learning_rate": 8.114705390897581e-07, + "loss": 0.9472, + "step": 37130 + }, + { + "epoch": 0.87, + "grad_norm": 1.8881834172326937, + "learning_rate": 8.111694697368778e-07, + "loss": 0.99, + "step": 37131 + }, + { + "epoch": 0.87, + "grad_norm": 1.9745074918207297, + "learning_rate": 8.108684538842781e-07, + "loss": 0.9461, + "step": 37132 + }, + { + "epoch": 0.87, + "grad_norm": 2.6402605356496993, + "learning_rate": 8.105674915337103e-07, + "loss": 0.9518, + "step": 37133 + }, + { + "epoch": 0.87, + "grad_norm": 2.1398135336570134, + "learning_rate": 8.102665826869283e-07, + "loss": 0.9106, + "step": 37134 + }, + { + "epoch": 0.87, + "grad_norm": 2.269881602347546, + "learning_rate": 8.09965727345684e-07, + "loss": 0.9607, + "step": 37135 + }, + { + "epoch": 0.87, + "grad_norm": 1.9643158903930404, + "learning_rate": 8.096649255117294e-07, + "loss": 1.0586, + "step": 37136 + }, + { + "epoch": 0.87, + "grad_norm": 2.007509844345984, + "learning_rate": 8.093641771868144e-07, + "loss": 0.9682, + "step": 37137 + }, + { + "epoch": 0.87, + "grad_norm": 1.8600421655120392, + "learning_rate": 8.090634823726917e-07, + "loss": 0.8869, + "step": 37138 + }, + { + "epoch": 0.87, + "grad_norm": 1.8075226899320946, + "learning_rate": 8.08762841071109e-07, + "loss": 0.8296, + "step": 37139 + }, + { + "epoch": 0.87, + "grad_norm": 3.0924126584318508, + "learning_rate": 8.084622532838215e-07, + "loss": 0.9386, + "step": 37140 + }, + { + "epoch": 0.88, + "grad_norm": 1.8789938154852204, + "learning_rate": 8.081617190125745e-07, + "loss": 0.9628, + "step": 37141 + }, + { + "epoch": 0.88, + "grad_norm": 1.11883159653644, + "learning_rate": 8.078612382591222e-07, + "loss": 0.9644, + "step": 37142 + }, + { + "epoch": 0.88, + "grad_norm": 1.7509342873704208, + "learning_rate": 8.075608110252098e-07, + "loss": 0.8827, + "step": 37143 + }, + { + "epoch": 0.88, + "grad_norm": 2.0468297174864793, + "learning_rate": 8.072604373125914e-07, + "loss": 0.9455, + "step": 37144 + }, + { + "epoch": 0.88, + "grad_norm": 1.99913470879162, + "learning_rate": 8.069601171230124e-07, + "loss": 0.9876, + "step": 37145 + }, + { + "epoch": 0.88, + "grad_norm": 2.129651240703451, + "learning_rate": 8.066598504582201e-07, + "loss": 0.9789, + "step": 37146 + }, + { + "epoch": 0.88, + "grad_norm": 1.0533244499848184, + "learning_rate": 8.063596373199645e-07, + "loss": 0.9161, + "step": 37147 + }, + { + "epoch": 0.88, + "grad_norm": 1.921352983709339, + "learning_rate": 8.060594777099939e-07, + "loss": 1.0663, + "step": 37148 + }, + { + "epoch": 0.88, + "grad_norm": 1.9401160331284235, + "learning_rate": 8.057593716300605e-07, + "loss": 1.0305, + "step": 37149 + }, + { + "epoch": 0.88, + "grad_norm": 2.3148571804505447, + "learning_rate": 8.054593190819016e-07, + "loss": 0.9024, + "step": 37150 + }, + { + "epoch": 0.88, + "grad_norm": 3.0490473359006915, + "learning_rate": 8.051593200672703e-07, + "loss": 1.0258, + "step": 37151 + }, + { + "epoch": 0.88, + "grad_norm": 2.0883402984721537, + "learning_rate": 8.048593745879141e-07, + "loss": 0.9703, + "step": 37152 + }, + { + "epoch": 0.88, + "grad_norm": 1.8236590187599866, + "learning_rate": 8.045594826455783e-07, + "loss": 1.0565, + "step": 37153 + }, + { + "epoch": 0.88, + "grad_norm": 1.9630792434060373, + "learning_rate": 8.042596442420059e-07, + "loss": 0.9627, + "step": 37154 + }, + { + "epoch": 0.88, + "grad_norm": 2.0922742124469362, + "learning_rate": 8.039598593789455e-07, + "loss": 0.9906, + "step": 37155 + }, + { + "epoch": 0.88, + "grad_norm": 1.889147107702148, + "learning_rate": 8.036601280581446e-07, + "loss": 1.0539, + "step": 37156 + }, + { + "epoch": 0.88, + "grad_norm": 1.8804462956942964, + "learning_rate": 8.033604502813453e-07, + "loss": 0.9847, + "step": 37157 + }, + { + "epoch": 0.88, + "grad_norm": 2.4095734052065425, + "learning_rate": 8.030608260502915e-07, + "loss": 0.9147, + "step": 37158 + }, + { + "epoch": 0.88, + "grad_norm": 2.205722508998418, + "learning_rate": 8.027612553667308e-07, + "loss": 0.8485, + "step": 37159 + }, + { + "epoch": 0.88, + "grad_norm": 2.09563590588094, + "learning_rate": 8.02461738232404e-07, + "loss": 1.0631, + "step": 37160 + }, + { + "epoch": 0.88, + "grad_norm": 1.8886242079415019, + "learning_rate": 8.021622746490565e-07, + "loss": 0.9114, + "step": 37161 + }, + { + "epoch": 0.88, + "grad_norm": 1.931561046519409, + "learning_rate": 8.018628646184345e-07, + "loss": 0.885, + "step": 37162 + }, + { + "epoch": 0.88, + "grad_norm": 2.1738799818148125, + "learning_rate": 8.01563508142279e-07, + "loss": 0.9508, + "step": 37163 + }, + { + "epoch": 0.88, + "grad_norm": 2.15736907964836, + "learning_rate": 8.012642052223319e-07, + "loss": 0.9445, + "step": 37164 + }, + { + "epoch": 0.88, + "grad_norm": 1.0859791402286034, + "learning_rate": 8.009649558603361e-07, + "loss": 0.921, + "step": 37165 + }, + { + "epoch": 0.88, + "grad_norm": 1.8304787591544749, + "learning_rate": 8.006657600580381e-07, + "loss": 1.0328, + "step": 37166 + }, + { + "epoch": 0.88, + "grad_norm": 1.882820181563841, + "learning_rate": 8.003666178171732e-07, + "loss": 0.9207, + "step": 37167 + }, + { + "epoch": 0.88, + "grad_norm": 1.8250207725699739, + "learning_rate": 8.000675291394855e-07, + "loss": 0.9804, + "step": 37168 + }, + { + "epoch": 0.88, + "grad_norm": 1.9759802233918564, + "learning_rate": 7.997684940267181e-07, + "loss": 0.8515, + "step": 37169 + }, + { + "epoch": 0.88, + "grad_norm": 2.161159874940127, + "learning_rate": 7.994695124806152e-07, + "loss": 1.0332, + "step": 37170 + }, + { + "epoch": 0.88, + "grad_norm": 1.0287847722105978, + "learning_rate": 7.991705845029096e-07, + "loss": 0.9029, + "step": 37171 + }, + { + "epoch": 0.88, + "grad_norm": 1.963083125515587, + "learning_rate": 7.988717100953458e-07, + "loss": 0.9655, + "step": 37172 + }, + { + "epoch": 0.88, + "grad_norm": 2.2024083687203846, + "learning_rate": 7.985728892596656e-07, + "loss": 1.1249, + "step": 37173 + }, + { + "epoch": 0.88, + "grad_norm": 1.891437749042004, + "learning_rate": 7.982741219976065e-07, + "loss": 1.0459, + "step": 37174 + }, + { + "epoch": 0.88, + "grad_norm": 2.116365444702463, + "learning_rate": 7.979754083109092e-07, + "loss": 0.8859, + "step": 37175 + }, + { + "epoch": 0.88, + "grad_norm": 2.117217954785206, + "learning_rate": 7.976767482013115e-07, + "loss": 1.0405, + "step": 37176 + }, + { + "epoch": 0.88, + "grad_norm": 2.1454016903844755, + "learning_rate": 7.973781416705539e-07, + "loss": 0.9938, + "step": 37177 + }, + { + "epoch": 0.88, + "grad_norm": 2.364944253673195, + "learning_rate": 7.970795887203731e-07, + "loss": 1.0074, + "step": 37178 + }, + { + "epoch": 0.88, + "grad_norm": 2.0738314711115677, + "learning_rate": 7.967810893525096e-07, + "loss": 0.9381, + "step": 37179 + }, + { + "epoch": 0.88, + "grad_norm": 2.12588718355218, + "learning_rate": 7.964826435687001e-07, + "loss": 0.8187, + "step": 37180 + }, + { + "epoch": 0.88, + "grad_norm": 1.900369002419162, + "learning_rate": 7.96184251370683e-07, + "loss": 0.899, + "step": 37181 + }, + { + "epoch": 0.88, + "grad_norm": 2.1177392123958305, + "learning_rate": 7.958859127601937e-07, + "loss": 0.9704, + "step": 37182 + }, + { + "epoch": 0.88, + "grad_norm": 2.0324233087392196, + "learning_rate": 7.955876277389718e-07, + "loss": 1.0293, + "step": 37183 + }, + { + "epoch": 0.88, + "grad_norm": 1.102202163164569, + "learning_rate": 7.952893963087527e-07, + "loss": 0.9269, + "step": 37184 + }, + { + "epoch": 0.88, + "grad_norm": 2.0056914351621207, + "learning_rate": 7.949912184712716e-07, + "loss": 0.9385, + "step": 37185 + }, + { + "epoch": 0.88, + "grad_norm": 2.2841815968864716, + "learning_rate": 7.94693094228266e-07, + "loss": 0.9538, + "step": 37186 + }, + { + "epoch": 0.88, + "grad_norm": 1.8833327422967039, + "learning_rate": 7.943950235814734e-07, + "loss": 1.0402, + "step": 37187 + }, + { + "epoch": 0.88, + "grad_norm": 1.0811874531896002, + "learning_rate": 7.94097006532627e-07, + "loss": 0.9537, + "step": 37188 + }, + { + "epoch": 0.88, + "grad_norm": 2.1862439972750347, + "learning_rate": 7.937990430834607e-07, + "loss": 1.0215, + "step": 37189 + }, + { + "epoch": 0.88, + "grad_norm": 2.257329331257514, + "learning_rate": 7.935011332357113e-07, + "loss": 0.9755, + "step": 37190 + }, + { + "epoch": 0.88, + "grad_norm": 2.0132503950063336, + "learning_rate": 7.932032769911146e-07, + "loss": 0.9373, + "step": 37191 + }, + { + "epoch": 0.88, + "grad_norm": 2.016142305052982, + "learning_rate": 7.929054743514042e-07, + "loss": 0.8154, + "step": 37192 + }, + { + "epoch": 0.88, + "grad_norm": 1.9645353586208094, + "learning_rate": 7.926077253183096e-07, + "loss": 0.8861, + "step": 37193 + }, + { + "epoch": 0.88, + "grad_norm": 2.497637537764808, + "learning_rate": 7.923100298935693e-07, + "loss": 1.0203, + "step": 37194 + }, + { + "epoch": 0.88, + "grad_norm": 1.8596776684133138, + "learning_rate": 7.920123880789155e-07, + "loss": 1.0181, + "step": 37195 + }, + { + "epoch": 0.88, + "grad_norm": 2.2425218773733095, + "learning_rate": 7.917147998760821e-07, + "loss": 0.9143, + "step": 37196 + }, + { + "epoch": 0.88, + "grad_norm": 2.0987926317346957, + "learning_rate": 7.91417265286798e-07, + "loss": 0.8824, + "step": 37197 + }, + { + "epoch": 0.88, + "grad_norm": 2.032914043450651, + "learning_rate": 7.911197843127993e-07, + "loss": 0.8965, + "step": 37198 + }, + { + "epoch": 0.88, + "grad_norm": 2.0868819971303245, + "learning_rate": 7.90822356955816e-07, + "loss": 1.0131, + "step": 37199 + }, + { + "epoch": 0.88, + "grad_norm": 2.2784614475330427, + "learning_rate": 7.905249832175809e-07, + "loss": 0.9496, + "step": 37200 + }, + { + "epoch": 0.88, + "grad_norm": 2.222430887973739, + "learning_rate": 7.90227663099824e-07, + "loss": 0.9361, + "step": 37201 + }, + { + "epoch": 0.88, + "grad_norm": 2.0232541887730355, + "learning_rate": 7.899303966042793e-07, + "loss": 1.0401, + "step": 37202 + }, + { + "epoch": 0.88, + "grad_norm": 2.0272463683717996, + "learning_rate": 7.896331837326743e-07, + "loss": 0.9978, + "step": 37203 + }, + { + "epoch": 0.88, + "grad_norm": 1.8704843615095206, + "learning_rate": 7.893360244867421e-07, + "loss": 0.9541, + "step": 37204 + }, + { + "epoch": 0.88, + "grad_norm": 2.1572611940046773, + "learning_rate": 7.890389188682112e-07, + "loss": 0.8578, + "step": 37205 + }, + { + "epoch": 0.88, + "grad_norm": 1.9542456549459772, + "learning_rate": 7.887418668788104e-07, + "loss": 0.8646, + "step": 37206 + }, + { + "epoch": 0.88, + "grad_norm": 2.0902146118420646, + "learning_rate": 7.884448685202717e-07, + "loss": 0.9099, + "step": 37207 + }, + { + "epoch": 0.88, + "grad_norm": 2.058150895727168, + "learning_rate": 7.881479237943235e-07, + "loss": 1.1267, + "step": 37208 + }, + { + "epoch": 0.88, + "grad_norm": 2.0036837739448354, + "learning_rate": 7.878510327026956e-07, + "loss": 0.9018, + "step": 37209 + }, + { + "epoch": 0.88, + "grad_norm": 1.9784834365824726, + "learning_rate": 7.875541952471133e-07, + "loss": 0.9841, + "step": 37210 + }, + { + "epoch": 0.88, + "grad_norm": 1.0936130280389054, + "learning_rate": 7.872574114293075e-07, + "loss": 0.9563, + "step": 37211 + }, + { + "epoch": 0.88, + "grad_norm": 1.8383288436192862, + "learning_rate": 7.869606812510067e-07, + "loss": 0.9669, + "step": 37212 + }, + { + "epoch": 0.88, + "grad_norm": 2.0533399325022788, + "learning_rate": 7.866640047139384e-07, + "loss": 1.0303, + "step": 37213 + }, + { + "epoch": 0.88, + "grad_norm": 1.9206123326858233, + "learning_rate": 7.863673818198269e-07, + "loss": 1.0332, + "step": 37214 + }, + { + "epoch": 0.88, + "grad_norm": 2.276203921275605, + "learning_rate": 7.860708125704019e-07, + "loss": 0.7939, + "step": 37215 + }, + { + "epoch": 0.88, + "grad_norm": 2.181911028592639, + "learning_rate": 7.857742969673909e-07, + "loss": 0.9584, + "step": 37216 + }, + { + "epoch": 0.88, + "grad_norm": 2.2760053702629404, + "learning_rate": 7.854778350125191e-07, + "loss": 0.9155, + "step": 37217 + }, + { + "epoch": 0.88, + "grad_norm": 1.9928349208450988, + "learning_rate": 7.85181426707512e-07, + "loss": 1.1093, + "step": 37218 + }, + { + "epoch": 0.88, + "grad_norm": 1.817362732445805, + "learning_rate": 7.848850720540968e-07, + "loss": 0.9611, + "step": 37219 + }, + { + "epoch": 0.88, + "grad_norm": 2.021164937626367, + "learning_rate": 7.845887710539957e-07, + "loss": 0.8888, + "step": 37220 + }, + { + "epoch": 0.88, + "grad_norm": 1.9321621783999392, + "learning_rate": 7.842925237089394e-07, + "loss": 0.963, + "step": 37221 + }, + { + "epoch": 0.88, + "grad_norm": 2.65941000493312, + "learning_rate": 7.839963300206477e-07, + "loss": 0.8434, + "step": 37222 + }, + { + "epoch": 0.88, + "grad_norm": 1.7817290173344893, + "learning_rate": 7.837001899908481e-07, + "loss": 0.7677, + "step": 37223 + }, + { + "epoch": 0.88, + "grad_norm": 2.0294039762704963, + "learning_rate": 7.834041036212625e-07, + "loss": 1.0864, + "step": 37224 + }, + { + "epoch": 0.88, + "grad_norm": 2.602007609635769, + "learning_rate": 7.831080709136185e-07, + "loss": 0.8218, + "step": 37225 + }, + { + "epoch": 0.88, + "grad_norm": 1.8758792603739793, + "learning_rate": 7.828120918696369e-07, + "loss": 0.8597, + "step": 37226 + }, + { + "epoch": 0.88, + "grad_norm": 2.046443514489723, + "learning_rate": 7.825161664910397e-07, + "loss": 0.8758, + "step": 37227 + }, + { + "epoch": 0.88, + "grad_norm": 2.003915753606937, + "learning_rate": 7.822202947795509e-07, + "loss": 0.9135, + "step": 37228 + }, + { + "epoch": 0.88, + "grad_norm": 1.0658438975075382, + "learning_rate": 7.81924476736896e-07, + "loss": 0.9059, + "step": 37229 + }, + { + "epoch": 0.88, + "grad_norm": 1.9939853505480927, + "learning_rate": 7.816287123647958e-07, + "loss": 0.9819, + "step": 37230 + }, + { + "epoch": 0.88, + "grad_norm": 2.3973202718501776, + "learning_rate": 7.813330016649689e-07, + "loss": 1.1173, + "step": 37231 + }, + { + "epoch": 0.88, + "grad_norm": 2.36002353551885, + "learning_rate": 7.810373446391406e-07, + "loss": 0.9872, + "step": 37232 + }, + { + "epoch": 0.88, + "grad_norm": 1.791887446507531, + "learning_rate": 7.807417412890339e-07, + "loss": 0.9545, + "step": 37233 + }, + { + "epoch": 0.88, + "grad_norm": 1.0797011950118738, + "learning_rate": 7.804461916163663e-07, + "loss": 0.9372, + "step": 37234 + }, + { + "epoch": 0.88, + "grad_norm": 2.150115856776791, + "learning_rate": 7.801506956228588e-07, + "loss": 1.0787, + "step": 37235 + }, + { + "epoch": 0.88, + "grad_norm": 1.9982072095477428, + "learning_rate": 7.798552533102322e-07, + "loss": 0.9048, + "step": 37236 + }, + { + "epoch": 0.88, + "grad_norm": 2.3468535729332043, + "learning_rate": 7.795598646802105e-07, + "loss": 1.0041, + "step": 37237 + }, + { + "epoch": 0.88, + "grad_norm": 1.9219913382034375, + "learning_rate": 7.792645297345091e-07, + "loss": 0.9151, + "step": 37238 + }, + { + "epoch": 0.88, + "grad_norm": 1.8993460785851972, + "learning_rate": 7.789692484748479e-07, + "loss": 1.0159, + "step": 37239 + }, + { + "epoch": 0.88, + "grad_norm": 2.0737338273784904, + "learning_rate": 7.786740209029475e-07, + "loss": 1.0132, + "step": 37240 + }, + { + "epoch": 0.88, + "grad_norm": 1.9700009519434336, + "learning_rate": 7.783788470205256e-07, + "loss": 1.0006, + "step": 37241 + }, + { + "epoch": 0.88, + "grad_norm": 1.7804587431758678, + "learning_rate": 7.78083726829304e-07, + "loss": 0.8612, + "step": 37242 + }, + { + "epoch": 0.88, + "grad_norm": 1.9799931648073124, + "learning_rate": 7.777886603309958e-07, + "loss": 0.8927, + "step": 37243 + }, + { + "epoch": 0.88, + "grad_norm": 2.2013320973488026, + "learning_rate": 7.774936475273243e-07, + "loss": 0.8529, + "step": 37244 + }, + { + "epoch": 0.88, + "grad_norm": 1.6890183665288712, + "learning_rate": 7.771986884200022e-07, + "loss": 1.0005, + "step": 37245 + }, + { + "epoch": 0.88, + "grad_norm": 2.2086155821352342, + "learning_rate": 7.769037830107484e-07, + "loss": 0.9424, + "step": 37246 + }, + { + "epoch": 0.88, + "grad_norm": 1.9339779626227247, + "learning_rate": 7.766089313012859e-07, + "loss": 1.0158, + "step": 37247 + }, + { + "epoch": 0.88, + "grad_norm": 2.0370152841236226, + "learning_rate": 7.763141332933222e-07, + "loss": 1.0573, + "step": 37248 + }, + { + "epoch": 0.88, + "grad_norm": 1.9629314086245164, + "learning_rate": 7.76019388988577e-07, + "loss": 0.9536, + "step": 37249 + }, + { + "epoch": 0.88, + "grad_norm": 1.847901640140788, + "learning_rate": 7.757246983887679e-07, + "loss": 0.8905, + "step": 37250 + }, + { + "epoch": 0.88, + "grad_norm": 1.7913985396846037, + "learning_rate": 7.754300614956134e-07, + "loss": 0.9547, + "step": 37251 + }, + { + "epoch": 0.88, + "grad_norm": 2.2478374254995424, + "learning_rate": 7.751354783108223e-07, + "loss": 0.9846, + "step": 37252 + }, + { + "epoch": 0.88, + "grad_norm": 2.061074205126701, + "learning_rate": 7.748409488361131e-07, + "loss": 1.028, + "step": 37253 + }, + { + "epoch": 0.88, + "grad_norm": 2.02833473212379, + "learning_rate": 7.745464730732011e-07, + "loss": 0.9778, + "step": 37254 + }, + { + "epoch": 0.88, + "grad_norm": 2.0551755808401704, + "learning_rate": 7.742520510238016e-07, + "loss": 0.9584, + "step": 37255 + }, + { + "epoch": 0.88, + "grad_norm": 2.0233133223970006, + "learning_rate": 7.739576826896244e-07, + "loss": 0.9857, + "step": 37256 + }, + { + "epoch": 0.88, + "grad_norm": 2.1045059667211414, + "learning_rate": 7.736633680723882e-07, + "loss": 0.9785, + "step": 37257 + }, + { + "epoch": 0.88, + "grad_norm": 1.982652613055866, + "learning_rate": 7.733691071738047e-07, + "loss": 1.0446, + "step": 37258 + }, + { + "epoch": 0.88, + "grad_norm": 1.982049023865618, + "learning_rate": 7.730748999955872e-07, + "loss": 0.9682, + "step": 37259 + }, + { + "epoch": 0.88, + "grad_norm": 1.070211143158311, + "learning_rate": 7.7278074653945e-07, + "loss": 0.9779, + "step": 37260 + }, + { + "epoch": 0.88, + "grad_norm": 2.0213952301705373, + "learning_rate": 7.724866468071046e-07, + "loss": 0.9574, + "step": 37261 + }, + { + "epoch": 0.88, + "grad_norm": 1.7887971179008686, + "learning_rate": 7.721926008002611e-07, + "loss": 0.9591, + "step": 37262 + }, + { + "epoch": 0.88, + "grad_norm": 1.990780891519174, + "learning_rate": 7.718986085206348e-07, + "loss": 0.979, + "step": 37263 + }, + { + "epoch": 0.88, + "grad_norm": 1.9017982886879587, + "learning_rate": 7.716046699699376e-07, + "loss": 1.0038, + "step": 37264 + }, + { + "epoch": 0.88, + "grad_norm": 1.9323972618576297, + "learning_rate": 7.713107851498791e-07, + "loss": 1.0151, + "step": 37265 + }, + { + "epoch": 0.88, + "grad_norm": 2.1687035715729093, + "learning_rate": 7.710169540621692e-07, + "loss": 0.8909, + "step": 37266 + }, + { + "epoch": 0.88, + "grad_norm": 2.057993249051079, + "learning_rate": 7.70723176708521e-07, + "loss": 0.9477, + "step": 37267 + }, + { + "epoch": 0.88, + "grad_norm": 1.910349029118709, + "learning_rate": 7.704294530906487e-07, + "loss": 0.9876, + "step": 37268 + }, + { + "epoch": 0.88, + "grad_norm": 2.3149056982035394, + "learning_rate": 7.70135783210253e-07, + "loss": 1.0001, + "step": 37269 + }, + { + "epoch": 0.88, + "grad_norm": 1.9931326592783007, + "learning_rate": 7.698421670690503e-07, + "loss": 0.9973, + "step": 37270 + }, + { + "epoch": 0.88, + "grad_norm": 2.235206591282737, + "learning_rate": 7.695486046687473e-07, + "loss": 0.9085, + "step": 37271 + }, + { + "epoch": 0.88, + "grad_norm": 1.9087701760738198, + "learning_rate": 7.692550960110578e-07, + "loss": 1.0523, + "step": 37272 + }, + { + "epoch": 0.88, + "grad_norm": 1.9552636826024845, + "learning_rate": 7.689616410976863e-07, + "loss": 1.0753, + "step": 37273 + }, + { + "epoch": 0.88, + "grad_norm": 1.083244060296739, + "learning_rate": 7.686682399303425e-07, + "loss": 0.9151, + "step": 37274 + }, + { + "epoch": 0.88, + "grad_norm": 1.9235503807090928, + "learning_rate": 7.683748925107348e-07, + "loss": 0.9693, + "step": 37275 + }, + { + "epoch": 0.88, + "grad_norm": 1.8495062652798342, + "learning_rate": 7.680815988405698e-07, + "loss": 0.8847, + "step": 37276 + }, + { + "epoch": 0.88, + "grad_norm": 2.3088305179655313, + "learning_rate": 7.677883589215596e-07, + "loss": 0.9937, + "step": 37277 + }, + { + "epoch": 0.88, + "grad_norm": 1.87642420457464, + "learning_rate": 7.674951727554058e-07, + "loss": 1.002, + "step": 37278 + }, + { + "epoch": 0.88, + "grad_norm": 1.9483335244129492, + "learning_rate": 7.672020403438196e-07, + "loss": 0.8883, + "step": 37279 + }, + { + "epoch": 0.88, + "grad_norm": 2.104156163842224, + "learning_rate": 7.66908961688504e-07, + "loss": 0.939, + "step": 37280 + }, + { + "epoch": 0.88, + "grad_norm": 2.137696776884056, + "learning_rate": 7.666159367911696e-07, + "loss": 0.8861, + "step": 37281 + }, + { + "epoch": 0.88, + "grad_norm": 2.211353364197996, + "learning_rate": 7.663229656535209e-07, + "loss": 1.0553, + "step": 37282 + }, + { + "epoch": 0.88, + "grad_norm": 2.0796416742619726, + "learning_rate": 7.66030048277262e-07, + "loss": 0.951, + "step": 37283 + }, + { + "epoch": 0.88, + "grad_norm": 2.270857185638052, + "learning_rate": 7.657371846640982e-07, + "loss": 0.9253, + "step": 37284 + }, + { + "epoch": 0.88, + "grad_norm": 1.0946812647976942, + "learning_rate": 7.654443748157391e-07, + "loss": 0.8984, + "step": 37285 + }, + { + "epoch": 0.88, + "grad_norm": 2.029523608384176, + "learning_rate": 7.651516187338859e-07, + "loss": 1.0409, + "step": 37286 + }, + { + "epoch": 0.88, + "grad_norm": 2.313819836588614, + "learning_rate": 7.648589164202413e-07, + "loss": 1.0526, + "step": 37287 + }, + { + "epoch": 0.88, + "grad_norm": 2.057113839362993, + "learning_rate": 7.645662678765131e-07, + "loss": 0.8763, + "step": 37288 + }, + { + "epoch": 0.88, + "grad_norm": 1.9854106154293119, + "learning_rate": 7.642736731044043e-07, + "loss": 1.0244, + "step": 37289 + }, + { + "epoch": 0.88, + "grad_norm": 1.8472157340311253, + "learning_rate": 7.639811321056189e-07, + "loss": 0.9302, + "step": 37290 + }, + { + "epoch": 0.88, + "grad_norm": 1.9602381971523513, + "learning_rate": 7.636886448818582e-07, + "loss": 1.1142, + "step": 37291 + }, + { + "epoch": 0.88, + "grad_norm": 1.8427104639746488, + "learning_rate": 7.633962114348259e-07, + "loss": 1.0263, + "step": 37292 + }, + { + "epoch": 0.88, + "grad_norm": 1.9046443079408015, + "learning_rate": 7.631038317662265e-07, + "loss": 0.9743, + "step": 37293 + }, + { + "epoch": 0.88, + "grad_norm": 1.861220856814927, + "learning_rate": 7.628115058777619e-07, + "loss": 0.9131, + "step": 37294 + }, + { + "epoch": 0.88, + "grad_norm": 1.9580428754135697, + "learning_rate": 7.625192337711306e-07, + "loss": 1.013, + "step": 37295 + }, + { + "epoch": 0.88, + "grad_norm": 1.888224716838629, + "learning_rate": 7.622270154480382e-07, + "loss": 0.9918, + "step": 37296 + }, + { + "epoch": 0.88, + "grad_norm": 1.8818988535894563, + "learning_rate": 7.619348509101843e-07, + "loss": 1.0435, + "step": 37297 + }, + { + "epoch": 0.88, + "grad_norm": 2.170980898839247, + "learning_rate": 7.616427401592708e-07, + "loss": 0.9354, + "step": 37298 + }, + { + "epoch": 0.88, + "grad_norm": 1.9007225022495793, + "learning_rate": 7.613506831969975e-07, + "loss": 0.8694, + "step": 37299 + }, + { + "epoch": 0.88, + "grad_norm": 2.586907349554736, + "learning_rate": 7.610586800250663e-07, + "loss": 0.9145, + "step": 37300 + }, + { + "epoch": 0.88, + "grad_norm": 2.473440369971394, + "learning_rate": 7.607667306451749e-07, + "loss": 0.8579, + "step": 37301 + }, + { + "epoch": 0.88, + "grad_norm": 2.907808722382633, + "learning_rate": 7.604748350590274e-07, + "loss": 1.0155, + "step": 37302 + }, + { + "epoch": 0.88, + "grad_norm": 2.318637016092931, + "learning_rate": 7.60182993268318e-07, + "loss": 0.9866, + "step": 37303 + }, + { + "epoch": 0.88, + "grad_norm": 2.31899674214057, + "learning_rate": 7.598912052747498e-07, + "loss": 1.0771, + "step": 37304 + }, + { + "epoch": 0.88, + "grad_norm": 1.8611648545878183, + "learning_rate": 7.595994710800192e-07, + "loss": 0.9051, + "step": 37305 + }, + { + "epoch": 0.88, + "grad_norm": 3.041943392966918, + "learning_rate": 7.593077906858281e-07, + "loss": 1.0489, + "step": 37306 + }, + { + "epoch": 0.88, + "grad_norm": 1.1176171118660345, + "learning_rate": 7.59016164093872e-07, + "loss": 0.9289, + "step": 37307 + }, + { + "epoch": 0.88, + "grad_norm": 2.2206255269416526, + "learning_rate": 7.587245913058472e-07, + "loss": 0.9712, + "step": 37308 + }, + { + "epoch": 0.88, + "grad_norm": 1.9810049878879379, + "learning_rate": 7.584330723234545e-07, + "loss": 1.1285, + "step": 37309 + }, + { + "epoch": 0.88, + "grad_norm": 1.8194686127935524, + "learning_rate": 7.581416071483915e-07, + "loss": 1.0189, + "step": 37310 + }, + { + "epoch": 0.88, + "grad_norm": 1.7662763496569263, + "learning_rate": 7.578501957823547e-07, + "loss": 0.9717, + "step": 37311 + }, + { + "epoch": 0.88, + "grad_norm": 1.904152614924698, + "learning_rate": 7.575588382270372e-07, + "loss": 0.9756, + "step": 37312 + }, + { + "epoch": 0.88, + "grad_norm": 1.9173770477282674, + "learning_rate": 7.572675344841396e-07, + "loss": 1.0396, + "step": 37313 + }, + { + "epoch": 0.88, + "grad_norm": 2.264905598306851, + "learning_rate": 7.569762845553574e-07, + "loss": 1.0075, + "step": 37314 + }, + { + "epoch": 0.88, + "grad_norm": 1.0748758725025702, + "learning_rate": 7.566850884423859e-07, + "loss": 0.9502, + "step": 37315 + }, + { + "epoch": 0.88, + "grad_norm": 3.6571802335613532, + "learning_rate": 7.563939461469183e-07, + "loss": 0.9275, + "step": 37316 + }, + { + "epoch": 0.88, + "grad_norm": 1.999767926095541, + "learning_rate": 7.561028576706519e-07, + "loss": 0.9318, + "step": 37317 + }, + { + "epoch": 0.88, + "grad_norm": 1.968772976746904, + "learning_rate": 7.558118230152822e-07, + "loss": 1.0485, + "step": 37318 + }, + { + "epoch": 0.88, + "grad_norm": 3.2328473616382483, + "learning_rate": 7.555208421825033e-07, + "loss": 0.9331, + "step": 37319 + }, + { + "epoch": 0.88, + "grad_norm": 1.111450563375212, + "learning_rate": 7.552299151740072e-07, + "loss": 0.953, + "step": 37320 + }, + { + "epoch": 0.88, + "grad_norm": 2.276499014070596, + "learning_rate": 7.549390419914915e-07, + "loss": 1.0138, + "step": 37321 + }, + { + "epoch": 0.88, + "grad_norm": 1.9220710903680152, + "learning_rate": 7.546482226366447e-07, + "loss": 1.0146, + "step": 37322 + }, + { + "epoch": 0.88, + "grad_norm": 2.004092482913943, + "learning_rate": 7.543574571111656e-07, + "loss": 1.0091, + "step": 37323 + }, + { + "epoch": 0.88, + "grad_norm": 1.1293317923331714, + "learning_rate": 7.540667454167427e-07, + "loss": 0.9121, + "step": 37324 + }, + { + "epoch": 0.88, + "grad_norm": 1.855793627418106, + "learning_rate": 7.537760875550714e-07, + "loss": 1.0852, + "step": 37325 + }, + { + "epoch": 0.88, + "grad_norm": 2.090120904249807, + "learning_rate": 7.534854835278427e-07, + "loss": 0.9047, + "step": 37326 + }, + { + "epoch": 0.88, + "grad_norm": 1.9207604824896667, + "learning_rate": 7.531949333367495e-07, + "loss": 0.9462, + "step": 37327 + }, + { + "epoch": 0.88, + "grad_norm": 2.06433628869545, + "learning_rate": 7.529044369834826e-07, + "loss": 1.0647, + "step": 37328 + }, + { + "epoch": 0.88, + "grad_norm": 2.0878161404471705, + "learning_rate": 7.526139944697331e-07, + "loss": 0.927, + "step": 37329 + }, + { + "epoch": 0.88, + "grad_norm": 2.1084955222034654, + "learning_rate": 7.523236057971917e-07, + "loss": 0.9206, + "step": 37330 + }, + { + "epoch": 0.88, + "grad_norm": 2.0300657469206196, + "learning_rate": 7.520332709675504e-07, + "loss": 0.9337, + "step": 37331 + }, + { + "epoch": 0.88, + "grad_norm": 2.012632304111812, + "learning_rate": 7.517429899825035e-07, + "loss": 0.8648, + "step": 37332 + }, + { + "epoch": 0.88, + "grad_norm": 2.0375998409240332, + "learning_rate": 7.514527628437318e-07, + "loss": 1.0126, + "step": 37333 + }, + { + "epoch": 0.88, + "grad_norm": 1.968715092774334, + "learning_rate": 7.511625895529317e-07, + "loss": 1.007, + "step": 37334 + }, + { + "epoch": 0.88, + "grad_norm": 2.0988159759873803, + "learning_rate": 7.50872470111792e-07, + "loss": 1.1124, + "step": 37335 + }, + { + "epoch": 0.88, + "grad_norm": 2.650304945430192, + "learning_rate": 7.505824045220012e-07, + "loss": 1.0496, + "step": 37336 + }, + { + "epoch": 0.88, + "grad_norm": 1.858786655154708, + "learning_rate": 7.502923927852457e-07, + "loss": 0.7897, + "step": 37337 + }, + { + "epoch": 0.88, + "grad_norm": 2.2694963171005464, + "learning_rate": 7.500024349032165e-07, + "loss": 1.0763, + "step": 37338 + }, + { + "epoch": 0.88, + "grad_norm": 1.9208450372383037, + "learning_rate": 7.497125308776043e-07, + "loss": 1.0297, + "step": 37339 + }, + { + "epoch": 0.88, + "grad_norm": 2.088883973148275, + "learning_rate": 7.494226807100935e-07, + "loss": 1.0006, + "step": 37340 + }, + { + "epoch": 0.88, + "grad_norm": 2.0718763620196263, + "learning_rate": 7.491328844023704e-07, + "loss": 0.9189, + "step": 37341 + }, + { + "epoch": 0.88, + "grad_norm": 1.9975057534096858, + "learning_rate": 7.48843141956127e-07, + "loss": 1.0226, + "step": 37342 + }, + { + "epoch": 0.88, + "grad_norm": 2.203497625710067, + "learning_rate": 7.485534533730454e-07, + "loss": 1.0454, + "step": 37343 + }, + { + "epoch": 0.88, + "grad_norm": 2.6877858081338735, + "learning_rate": 7.482638186548153e-07, + "loss": 1.0114, + "step": 37344 + }, + { + "epoch": 0.88, + "grad_norm": 2.038130310861427, + "learning_rate": 7.47974237803123e-07, + "loss": 0.9314, + "step": 37345 + }, + { + "epoch": 0.88, + "grad_norm": 2.2223508895656447, + "learning_rate": 7.47684710819655e-07, + "loss": 0.9777, + "step": 37346 + }, + { + "epoch": 0.88, + "grad_norm": 2.0027287461233527, + "learning_rate": 7.473952377060934e-07, + "loss": 1.0573, + "step": 37347 + }, + { + "epoch": 0.88, + "grad_norm": 2.01454900895167, + "learning_rate": 7.471058184641255e-07, + "loss": 1.0769, + "step": 37348 + }, + { + "epoch": 0.88, + "grad_norm": 2.194517914208504, + "learning_rate": 7.468164530954414e-07, + "loss": 1.0167, + "step": 37349 + }, + { + "epoch": 0.88, + "grad_norm": 2.0440418545941954, + "learning_rate": 7.465271416017172e-07, + "loss": 0.8737, + "step": 37350 + }, + { + "epoch": 0.88, + "grad_norm": 1.7727102291540369, + "learning_rate": 7.462378839846419e-07, + "loss": 0.8884, + "step": 37351 + }, + { + "epoch": 0.88, + "grad_norm": 1.87891057820922, + "learning_rate": 7.459486802458993e-07, + "loss": 0.9307, + "step": 37352 + }, + { + "epoch": 0.88, + "grad_norm": 2.0903907010007883, + "learning_rate": 7.456595303871772e-07, + "loss": 0.9119, + "step": 37353 + }, + { + "epoch": 0.88, + "grad_norm": 1.8139005136185575, + "learning_rate": 7.453704344101509e-07, + "loss": 0.964, + "step": 37354 + }, + { + "epoch": 0.88, + "grad_norm": 1.1487512153935178, + "learning_rate": 7.45081392316509e-07, + "loss": 0.9457, + "step": 37355 + }, + { + "epoch": 0.88, + "grad_norm": 1.989336045144953, + "learning_rate": 7.447924041079347e-07, + "loss": 0.9623, + "step": 37356 + }, + { + "epoch": 0.88, + "grad_norm": 2.8409371322513164, + "learning_rate": 7.445034697861064e-07, + "loss": 1.0611, + "step": 37357 + }, + { + "epoch": 0.88, + "grad_norm": 2.0367210373835456, + "learning_rate": 7.442145893527119e-07, + "loss": 0.9288, + "step": 37358 + }, + { + "epoch": 0.88, + "grad_norm": 1.9494021926392768, + "learning_rate": 7.439257628094288e-07, + "loss": 0.9825, + "step": 37359 + }, + { + "epoch": 0.88, + "grad_norm": 1.1395540982893702, + "learning_rate": 7.43636990157941e-07, + "loss": 1.0238, + "step": 37360 + }, + { + "epoch": 0.88, + "grad_norm": 1.0760812481090543, + "learning_rate": 7.433482713999285e-07, + "loss": 0.9262, + "step": 37361 + }, + { + "epoch": 0.88, + "grad_norm": 2.1691588580496823, + "learning_rate": 7.430596065370743e-07, + "loss": 0.8184, + "step": 37362 + }, + { + "epoch": 0.88, + "grad_norm": 2.0534712144311853, + "learning_rate": 7.427709955710572e-07, + "loss": 0.8894, + "step": 37363 + }, + { + "epoch": 0.88, + "grad_norm": 2.044609904858007, + "learning_rate": 7.424824385035568e-07, + "loss": 1.0065, + "step": 37364 + }, + { + "epoch": 0.88, + "grad_norm": 2.116495962074792, + "learning_rate": 7.42193935336255e-07, + "loss": 1.017, + "step": 37365 + }, + { + "epoch": 0.88, + "grad_norm": 1.913377917247667, + "learning_rate": 7.419054860708319e-07, + "loss": 0.9809, + "step": 37366 + }, + { + "epoch": 0.88, + "grad_norm": 1.985563651529418, + "learning_rate": 7.416170907089671e-07, + "loss": 0.9533, + "step": 37367 + }, + { + "epoch": 0.88, + "grad_norm": 1.0564251447976947, + "learning_rate": 7.413287492523369e-07, + "loss": 0.9327, + "step": 37368 + }, + { + "epoch": 0.88, + "grad_norm": 2.051495561970362, + "learning_rate": 7.410404617026212e-07, + "loss": 1.05, + "step": 37369 + }, + { + "epoch": 0.88, + "grad_norm": 2.168002545892357, + "learning_rate": 7.40752228061502e-07, + "loss": 0.9471, + "step": 37370 + }, + { + "epoch": 0.88, + "grad_norm": 1.9998194475185305, + "learning_rate": 7.404640483306536e-07, + "loss": 0.9637, + "step": 37371 + }, + { + "epoch": 0.88, + "grad_norm": 2.028529725378362, + "learning_rate": 7.401759225117544e-07, + "loss": 0.9429, + "step": 37372 + }, + { + "epoch": 0.88, + "grad_norm": 2.2828832899763767, + "learning_rate": 7.398878506064821e-07, + "loss": 0.878, + "step": 37373 + }, + { + "epoch": 0.88, + "grad_norm": 1.85905737697232, + "learning_rate": 7.395998326165165e-07, + "loss": 0.8934, + "step": 37374 + }, + { + "epoch": 0.88, + "grad_norm": 1.8975596076477124, + "learning_rate": 7.393118685435319e-07, + "loss": 1.0807, + "step": 37375 + }, + { + "epoch": 0.88, + "grad_norm": 2.184986195350509, + "learning_rate": 7.390239583892045e-07, + "loss": 0.9587, + "step": 37376 + }, + { + "epoch": 0.88, + "grad_norm": 1.1952549680564901, + "learning_rate": 7.38736102155212e-07, + "loss": 0.9642, + "step": 37377 + }, + { + "epoch": 0.88, + "grad_norm": 2.00178674785739, + "learning_rate": 7.384482998432296e-07, + "loss": 0.92, + "step": 37378 + }, + { + "epoch": 0.88, + "grad_norm": 2.0503945010946754, + "learning_rate": 7.381605514549339e-07, + "loss": 0.8376, + "step": 37379 + }, + { + "epoch": 0.88, + "grad_norm": 2.063646045987467, + "learning_rate": 7.378728569919991e-07, + "loss": 0.9968, + "step": 37380 + }, + { + "epoch": 0.88, + "grad_norm": 2.1388073497260947, + "learning_rate": 7.375852164561026e-07, + "loss": 0.9496, + "step": 37381 + }, + { + "epoch": 0.88, + "grad_norm": 2.211033139902478, + "learning_rate": 7.372976298489154e-07, + "loss": 0.9498, + "step": 37382 + }, + { + "epoch": 0.88, + "grad_norm": 2.011930595164672, + "learning_rate": 7.370100971721151e-07, + "loss": 0.9797, + "step": 37383 + }, + { + "epoch": 0.88, + "grad_norm": 1.8854417120729319, + "learning_rate": 7.367226184273746e-07, + "loss": 0.9364, + "step": 37384 + }, + { + "epoch": 0.88, + "grad_norm": 2.017818384345649, + "learning_rate": 7.364351936163672e-07, + "loss": 1.0939, + "step": 37385 + }, + { + "epoch": 0.88, + "grad_norm": 2.5923148377785585, + "learning_rate": 7.361478227407659e-07, + "loss": 0.9469, + "step": 37386 + }, + { + "epoch": 0.88, + "grad_norm": 1.1720620534542614, + "learning_rate": 7.358605058022461e-07, + "loss": 0.9336, + "step": 37387 + }, + { + "epoch": 0.88, + "grad_norm": 1.064416897008506, + "learning_rate": 7.355732428024808e-07, + "loss": 0.9168, + "step": 37388 + }, + { + "epoch": 0.88, + "grad_norm": 1.1275598714468764, + "learning_rate": 7.352860337431389e-07, + "loss": 0.9915, + "step": 37389 + }, + { + "epoch": 0.88, + "grad_norm": 2.076805413056797, + "learning_rate": 7.349988786258944e-07, + "loss": 0.9677, + "step": 37390 + }, + { + "epoch": 0.88, + "grad_norm": 2.134931393096035, + "learning_rate": 7.347117774524215e-07, + "loss": 0.9294, + "step": 37391 + }, + { + "epoch": 0.88, + "grad_norm": 1.7714416909884783, + "learning_rate": 7.344247302243901e-07, + "loss": 0.9809, + "step": 37392 + }, + { + "epoch": 0.88, + "grad_norm": 1.7674854906319482, + "learning_rate": 7.3413773694347e-07, + "loss": 0.9334, + "step": 37393 + }, + { + "epoch": 0.88, + "grad_norm": 1.0747120036903164, + "learning_rate": 7.338507976113329e-07, + "loss": 0.8955, + "step": 37394 + }, + { + "epoch": 0.88, + "grad_norm": 2.4019070546286625, + "learning_rate": 7.335639122296522e-07, + "loss": 1.0141, + "step": 37395 + }, + { + "epoch": 0.88, + "grad_norm": 1.9486109216132816, + "learning_rate": 7.332770808000955e-07, + "loss": 1.0233, + "step": 37396 + }, + { + "epoch": 0.88, + "grad_norm": 2.1271961804959303, + "learning_rate": 7.329903033243324e-07, + "loss": 0.9807, + "step": 37397 + }, + { + "epoch": 0.88, + "grad_norm": 2.0322213132745572, + "learning_rate": 7.32703579804036e-07, + "loss": 1.0723, + "step": 37398 + }, + { + "epoch": 0.88, + "grad_norm": 2.1479914001052776, + "learning_rate": 7.324169102408707e-07, + "loss": 0.9467, + "step": 37399 + }, + { + "epoch": 0.88, + "grad_norm": 1.987338017552066, + "learning_rate": 7.321302946365094e-07, + "loss": 0.9656, + "step": 37400 + }, + { + "epoch": 0.88, + "grad_norm": 1.9820431053824266, + "learning_rate": 7.318437329926187e-07, + "loss": 0.9368, + "step": 37401 + }, + { + "epoch": 0.88, + "grad_norm": 1.1017253521585635, + "learning_rate": 7.315572253108694e-07, + "loss": 0.9771, + "step": 37402 + }, + { + "epoch": 0.88, + "grad_norm": 1.9254529900046327, + "learning_rate": 7.312707715929268e-07, + "loss": 1.0774, + "step": 37403 + }, + { + "epoch": 0.88, + "grad_norm": 1.8085407218152068, + "learning_rate": 7.309843718404619e-07, + "loss": 0.9427, + "step": 37404 + }, + { + "epoch": 0.88, + "grad_norm": 2.1225247733635952, + "learning_rate": 7.306980260551411e-07, + "loss": 0.9632, + "step": 37405 + }, + { + "epoch": 0.88, + "grad_norm": 1.8603486828268598, + "learning_rate": 7.304117342386286e-07, + "loss": 0.9815, + "step": 37406 + }, + { + "epoch": 0.88, + "grad_norm": 2.0547172721910507, + "learning_rate": 7.301254963925941e-07, + "loss": 0.9804, + "step": 37407 + }, + { + "epoch": 0.88, + "grad_norm": 1.9611432537005682, + "learning_rate": 7.298393125187054e-07, + "loss": 0.9285, + "step": 37408 + }, + { + "epoch": 0.88, + "grad_norm": 1.9053268479511376, + "learning_rate": 7.295531826186264e-07, + "loss": 0.9741, + "step": 37409 + }, + { + "epoch": 0.88, + "grad_norm": 2.080500684607485, + "learning_rate": 7.292671066940226e-07, + "loss": 0.868, + "step": 37410 + }, + { + "epoch": 0.88, + "grad_norm": 1.8844403176409972, + "learning_rate": 7.289810847465606e-07, + "loss": 0.9032, + "step": 37411 + }, + { + "epoch": 0.88, + "grad_norm": 2.075837396582737, + "learning_rate": 7.286951167779077e-07, + "loss": 0.9705, + "step": 37412 + }, + { + "epoch": 0.88, + "grad_norm": 2.0220427748537606, + "learning_rate": 7.284092027897272e-07, + "loss": 0.8444, + "step": 37413 + }, + { + "epoch": 0.88, + "grad_norm": 1.9427785487063172, + "learning_rate": 7.28123342783682e-07, + "loss": 1.1693, + "step": 37414 + }, + { + "epoch": 0.88, + "grad_norm": 1.9033127432761239, + "learning_rate": 7.278375367614377e-07, + "loss": 0.97, + "step": 37415 + }, + { + "epoch": 0.88, + "grad_norm": 1.9122825086090791, + "learning_rate": 7.275517847246604e-07, + "loss": 1.0044, + "step": 37416 + }, + { + "epoch": 0.88, + "grad_norm": 1.9271848505522917, + "learning_rate": 7.272660866750125e-07, + "loss": 0.9414, + "step": 37417 + }, + { + "epoch": 0.88, + "grad_norm": 2.3467174588305664, + "learning_rate": 7.269804426141558e-07, + "loss": 1.1207, + "step": 37418 + }, + { + "epoch": 0.88, + "grad_norm": 3.0149650192794075, + "learning_rate": 7.266948525437556e-07, + "loss": 0.9721, + "step": 37419 + }, + { + "epoch": 0.88, + "grad_norm": 2.0462941654203957, + "learning_rate": 7.264093164654728e-07, + "loss": 0.9897, + "step": 37420 + }, + { + "epoch": 0.88, + "grad_norm": 2.632057835596661, + "learning_rate": 7.261238343809728e-07, + "loss": 0.9096, + "step": 37421 + }, + { + "epoch": 0.88, + "grad_norm": 2.0491535863075923, + "learning_rate": 7.258384062919144e-07, + "loss": 0.975, + "step": 37422 + }, + { + "epoch": 0.88, + "grad_norm": 2.072123735470719, + "learning_rate": 7.255530321999616e-07, + "loss": 0.97, + "step": 37423 + }, + { + "epoch": 0.88, + "grad_norm": 1.9018519026290701, + "learning_rate": 7.252677121067742e-07, + "loss": 0.8962, + "step": 37424 + }, + { + "epoch": 0.88, + "grad_norm": 3.459030925544119, + "learning_rate": 7.249824460140165e-07, + "loss": 0.9583, + "step": 37425 + }, + { + "epoch": 0.88, + "grad_norm": 1.792988460134801, + "learning_rate": 7.246972339233449e-07, + "loss": 1.016, + "step": 37426 + }, + { + "epoch": 0.88, + "grad_norm": 2.0499121331450976, + "learning_rate": 7.244120758364248e-07, + "loss": 1.0063, + "step": 37427 + }, + { + "epoch": 0.88, + "grad_norm": 2.262539994700687, + "learning_rate": 7.241269717549126e-07, + "loss": 1.0096, + "step": 37428 + }, + { + "epoch": 0.88, + "grad_norm": 1.949321635054702, + "learning_rate": 7.238419216804693e-07, + "loss": 0.9748, + "step": 37429 + }, + { + "epoch": 0.88, + "grad_norm": 2.127075528291214, + "learning_rate": 7.23556925614759e-07, + "loss": 0.9836, + "step": 37430 + }, + { + "epoch": 0.88, + "grad_norm": 2.0792789402601586, + "learning_rate": 7.232719835594337e-07, + "loss": 0.9755, + "step": 37431 + }, + { + "epoch": 0.88, + "grad_norm": 2.2923734985273505, + "learning_rate": 7.229870955161555e-07, + "loss": 1.0226, + "step": 37432 + }, + { + "epoch": 0.88, + "grad_norm": 1.8750225170094759, + "learning_rate": 7.227022614865841e-07, + "loss": 0.9569, + "step": 37433 + }, + { + "epoch": 0.88, + "grad_norm": 1.9487402204098943, + "learning_rate": 7.224174814723794e-07, + "loss": 0.9948, + "step": 37434 + }, + { + "epoch": 0.88, + "grad_norm": 1.0661507047493533, + "learning_rate": 7.221327554751956e-07, + "loss": 0.9395, + "step": 37435 + }, + { + "epoch": 0.88, + "grad_norm": 2.2417668656645287, + "learning_rate": 7.218480834966912e-07, + "loss": 1.0113, + "step": 37436 + }, + { + "epoch": 0.88, + "grad_norm": 1.8321004569294628, + "learning_rate": 7.215634655385273e-07, + "loss": 0.9656, + "step": 37437 + }, + { + "epoch": 0.88, + "grad_norm": 1.7930371410648671, + "learning_rate": 7.212789016023569e-07, + "loss": 0.9331, + "step": 37438 + }, + { + "epoch": 0.88, + "grad_norm": 1.0903438790835462, + "learning_rate": 7.209943916898377e-07, + "loss": 0.8682, + "step": 37439 + }, + { + "epoch": 0.88, + "grad_norm": 1.7620332339007683, + "learning_rate": 7.20709935802627e-07, + "loss": 0.8754, + "step": 37440 + }, + { + "epoch": 0.88, + "grad_norm": 2.087851641971728, + "learning_rate": 7.204255339423815e-07, + "loss": 1.0315, + "step": 37441 + }, + { + "epoch": 0.88, + "grad_norm": 2.0034984626609496, + "learning_rate": 7.201411861107555e-07, + "loss": 0.9765, + "step": 37442 + }, + { + "epoch": 0.88, + "grad_norm": 2.3005772300419087, + "learning_rate": 7.198568923094074e-07, + "loss": 1.0125, + "step": 37443 + }, + { + "epoch": 0.88, + "grad_norm": 1.9900243400783924, + "learning_rate": 7.195726525399905e-07, + "loss": 1.0257, + "step": 37444 + }, + { + "epoch": 0.88, + "grad_norm": 1.7579668126984758, + "learning_rate": 7.192884668041578e-07, + "loss": 0.9522, + "step": 37445 + }, + { + "epoch": 0.88, + "grad_norm": 1.8901376608449825, + "learning_rate": 7.190043351035658e-07, + "loss": 0.9366, + "step": 37446 + }, + { + "epoch": 0.88, + "grad_norm": 1.7579894850959796, + "learning_rate": 7.187202574398699e-07, + "loss": 0.8871, + "step": 37447 + }, + { + "epoch": 0.88, + "grad_norm": 1.065438551345793, + "learning_rate": 7.184362338147244e-07, + "loss": 0.981, + "step": 37448 + }, + { + "epoch": 0.88, + "grad_norm": 1.9617703301132987, + "learning_rate": 7.181522642297789e-07, + "loss": 0.9143, + "step": 37449 + }, + { + "epoch": 0.88, + "grad_norm": 1.7963524758387286, + "learning_rate": 7.178683486866899e-07, + "loss": 0.961, + "step": 37450 + }, + { + "epoch": 0.88, + "grad_norm": 2.4159259322701443, + "learning_rate": 7.175844871871129e-07, + "loss": 0.8862, + "step": 37451 + }, + { + "epoch": 0.88, + "grad_norm": 2.474545268550915, + "learning_rate": 7.173006797326954e-07, + "loss": 0.9427, + "step": 37452 + }, + { + "epoch": 0.88, + "grad_norm": 1.9613177862343116, + "learning_rate": 7.170169263250914e-07, + "loss": 1.1877, + "step": 37453 + }, + { + "epoch": 0.88, + "grad_norm": 1.00589583340097, + "learning_rate": 7.167332269659544e-07, + "loss": 0.8667, + "step": 37454 + }, + { + "epoch": 0.88, + "grad_norm": 2.0738226765871643, + "learning_rate": 7.164495816569373e-07, + "loss": 0.926, + "step": 37455 + }, + { + "epoch": 0.88, + "grad_norm": 2.1620899354526855, + "learning_rate": 7.161659903996887e-07, + "loss": 0.9961, + "step": 37456 + }, + { + "epoch": 0.88, + "grad_norm": 2.047549108982792, + "learning_rate": 7.158824531958608e-07, + "loss": 0.8082, + "step": 37457 + }, + { + "epoch": 0.88, + "grad_norm": 1.9411734075820786, + "learning_rate": 7.155989700471056e-07, + "loss": 0.9163, + "step": 37458 + }, + { + "epoch": 0.88, + "grad_norm": 2.1638891972900343, + "learning_rate": 7.153155409550705e-07, + "loss": 1.0182, + "step": 37459 + }, + { + "epoch": 0.88, + "grad_norm": 2.050857011326599, + "learning_rate": 7.150321659214099e-07, + "loss": 0.9659, + "step": 37460 + }, + { + "epoch": 0.88, + "grad_norm": 2.0896472878113785, + "learning_rate": 7.14748844947769e-07, + "loss": 1.0071, + "step": 37461 + }, + { + "epoch": 0.88, + "grad_norm": 1.8731111080653633, + "learning_rate": 7.144655780358034e-07, + "loss": 0.8842, + "step": 37462 + }, + { + "epoch": 0.88, + "grad_norm": 2.8866213070208517, + "learning_rate": 7.14182365187156e-07, + "loss": 0.9724, + "step": 37463 + }, + { + "epoch": 0.88, + "grad_norm": 1.9834737999345766, + "learning_rate": 7.138992064034811e-07, + "loss": 0.9886, + "step": 37464 + }, + { + "epoch": 0.88, + "grad_norm": 2.9538601272221277, + "learning_rate": 7.136161016864241e-07, + "loss": 0.9025, + "step": 37465 + }, + { + "epoch": 0.88, + "grad_norm": 1.9127649772618671, + "learning_rate": 7.133330510376335e-07, + "loss": 0.941, + "step": 37466 + }, + { + "epoch": 0.88, + "grad_norm": 1.961266523044145, + "learning_rate": 7.13050054458757e-07, + "loss": 1.0118, + "step": 37467 + }, + { + "epoch": 0.88, + "grad_norm": 2.4839901826658353, + "learning_rate": 7.127671119514456e-07, + "loss": 0.9185, + "step": 37468 + }, + { + "epoch": 0.88, + "grad_norm": 2.0005929191300504, + "learning_rate": 7.124842235173445e-07, + "loss": 0.8811, + "step": 37469 + }, + { + "epoch": 0.88, + "grad_norm": 2.410168866154652, + "learning_rate": 7.122013891580981e-07, + "loss": 0.9071, + "step": 37470 + }, + { + "epoch": 0.88, + "grad_norm": 2.0229001491597773, + "learning_rate": 7.11918608875356e-07, + "loss": 1.0529, + "step": 37471 + }, + { + "epoch": 0.88, + "grad_norm": 1.8972230320122248, + "learning_rate": 7.11635882670767e-07, + "loss": 0.8824, + "step": 37472 + }, + { + "epoch": 0.88, + "grad_norm": 2.1716590157967737, + "learning_rate": 7.113532105459731e-07, + "loss": 0.9094, + "step": 37473 + }, + { + "epoch": 0.88, + "grad_norm": 2.415136368484536, + "learning_rate": 7.110705925026207e-07, + "loss": 0.8748, + "step": 37474 + }, + { + "epoch": 0.88, + "grad_norm": 1.0741434836028791, + "learning_rate": 7.107880285423563e-07, + "loss": 0.9973, + "step": 37475 + }, + { + "epoch": 0.88, + "grad_norm": 2.2190105930985533, + "learning_rate": 7.105055186668264e-07, + "loss": 1.0494, + "step": 37476 + }, + { + "epoch": 0.88, + "grad_norm": 2.1393048468902607, + "learning_rate": 7.10223062877674e-07, + "loss": 0.9495, + "step": 37477 + }, + { + "epoch": 0.88, + "grad_norm": 2.1346010160371507, + "learning_rate": 7.099406611765425e-07, + "loss": 0.9678, + "step": 37478 + }, + { + "epoch": 0.88, + "grad_norm": 2.0199868564023924, + "learning_rate": 7.096583135650792e-07, + "loss": 1.0007, + "step": 37479 + }, + { + "epoch": 0.88, + "grad_norm": 1.814475968127798, + "learning_rate": 7.093760200449251e-07, + "loss": 0.9534, + "step": 37480 + }, + { + "epoch": 0.88, + "grad_norm": 2.0920385795193557, + "learning_rate": 7.090937806177267e-07, + "loss": 1.0094, + "step": 37481 + }, + { + "epoch": 0.88, + "grad_norm": 2.1889113643681664, + "learning_rate": 7.088115952851238e-07, + "loss": 1.0103, + "step": 37482 + }, + { + "epoch": 0.88, + "grad_norm": 2.2095073260964586, + "learning_rate": 7.085294640487639e-07, + "loss": 0.8599, + "step": 37483 + }, + { + "epoch": 0.88, + "grad_norm": 2.0481511220619972, + "learning_rate": 7.082473869102846e-07, + "loss": 0.9816, + "step": 37484 + }, + { + "epoch": 0.88, + "grad_norm": 2.145123137766506, + "learning_rate": 7.079653638713324e-07, + "loss": 0.9075, + "step": 37485 + }, + { + "epoch": 0.88, + "grad_norm": 2.0696209474528557, + "learning_rate": 7.076833949335482e-07, + "loss": 0.9479, + "step": 37486 + }, + { + "epoch": 0.88, + "grad_norm": 1.8276136009829225, + "learning_rate": 7.074014800985718e-07, + "loss": 1.0345, + "step": 37487 + }, + { + "epoch": 0.88, + "grad_norm": 2.1016953266921203, + "learning_rate": 7.071196193680452e-07, + "loss": 0.9341, + "step": 37488 + }, + { + "epoch": 0.88, + "grad_norm": 1.8596855423160121, + "learning_rate": 7.068378127436115e-07, + "loss": 0.9101, + "step": 37489 + }, + { + "epoch": 0.88, + "grad_norm": 1.9360577549523428, + "learning_rate": 7.065560602269106e-07, + "loss": 1.0209, + "step": 37490 + }, + { + "epoch": 0.88, + "grad_norm": 1.8925734123519706, + "learning_rate": 7.062743618195822e-07, + "loss": 0.9258, + "step": 37491 + }, + { + "epoch": 0.88, + "grad_norm": 2.0719646842958883, + "learning_rate": 7.05992717523265e-07, + "loss": 1.0977, + "step": 37492 + }, + { + "epoch": 0.88, + "grad_norm": 2.0488885950368805, + "learning_rate": 7.057111273396034e-07, + "loss": 0.947, + "step": 37493 + }, + { + "epoch": 0.88, + "grad_norm": 2.24056177129043, + "learning_rate": 7.054295912702336e-07, + "loss": 0.9962, + "step": 37494 + }, + { + "epoch": 0.88, + "grad_norm": 1.8508514310729436, + "learning_rate": 7.051481093167933e-07, + "loss": 0.9073, + "step": 37495 + }, + { + "epoch": 0.88, + "grad_norm": 1.8314223346375305, + "learning_rate": 7.048666814809246e-07, + "loss": 0.9646, + "step": 37496 + }, + { + "epoch": 0.88, + "grad_norm": 1.0620634292202844, + "learning_rate": 7.045853077642661e-07, + "loss": 0.9613, + "step": 37497 + }, + { + "epoch": 0.88, + "grad_norm": 1.9005250994698768, + "learning_rate": 7.043039881684544e-07, + "loss": 1.0148, + "step": 37498 + }, + { + "epoch": 0.88, + "grad_norm": 1.9475649769902257, + "learning_rate": 7.040227226951268e-07, + "loss": 0.9683, + "step": 37499 + }, + { + "epoch": 0.88, + "grad_norm": 1.8978715796817007, + "learning_rate": 7.037415113459223e-07, + "loss": 0.9353, + "step": 37500 + }, + { + "epoch": 0.88, + "grad_norm": 1.8064356358968368, + "learning_rate": 7.034603541224771e-07, + "loss": 0.7656, + "step": 37501 + }, + { + "epoch": 0.88, + "grad_norm": 2.007228188803113, + "learning_rate": 7.031792510264312e-07, + "loss": 1.0229, + "step": 37502 + }, + { + "epoch": 0.88, + "grad_norm": 2.0925666704824795, + "learning_rate": 7.028982020594166e-07, + "loss": 1.0963, + "step": 37503 + }, + { + "epoch": 0.88, + "grad_norm": 1.9229700232187203, + "learning_rate": 7.026172072230742e-07, + "loss": 0.9676, + "step": 37504 + }, + { + "epoch": 0.88, + "grad_norm": 1.1573002727388415, + "learning_rate": 7.023362665190358e-07, + "loss": 0.9809, + "step": 37505 + }, + { + "epoch": 0.88, + "grad_norm": 1.0612798322882926, + "learning_rate": 7.020553799489415e-07, + "loss": 0.9254, + "step": 37506 + }, + { + "epoch": 0.88, + "grad_norm": 2.288218982595935, + "learning_rate": 7.017745475144233e-07, + "loss": 0.8495, + "step": 37507 + }, + { + "epoch": 0.88, + "grad_norm": 1.8520148665025655, + "learning_rate": 7.014937692171176e-07, + "loss": 1.0101, + "step": 37508 + }, + { + "epoch": 0.88, + "grad_norm": 1.9079399275728004, + "learning_rate": 7.012130450586574e-07, + "loss": 0.9091, + "step": 37509 + }, + { + "epoch": 0.88, + "grad_norm": 2.1936498536770532, + "learning_rate": 7.009323750406816e-07, + "loss": 0.9769, + "step": 37510 + }, + { + "epoch": 0.88, + "grad_norm": 2.06700051006915, + "learning_rate": 7.00651759164821e-07, + "loss": 0.9816, + "step": 37511 + }, + { + "epoch": 0.88, + "grad_norm": 2.0534448651965924, + "learning_rate": 7.003711974327076e-07, + "loss": 1.0547, + "step": 37512 + }, + { + "epoch": 0.88, + "grad_norm": 1.9974301855700922, + "learning_rate": 7.00090689845978e-07, + "loss": 0.9426, + "step": 37513 + }, + { + "epoch": 0.88, + "grad_norm": 2.9716758621446933, + "learning_rate": 6.998102364062665e-07, + "loss": 0.9945, + "step": 37514 + }, + { + "epoch": 0.88, + "grad_norm": 1.9959968054734833, + "learning_rate": 6.995298371152037e-07, + "loss": 0.9576, + "step": 37515 + }, + { + "epoch": 0.88, + "grad_norm": 2.0912704820564776, + "learning_rate": 6.992494919744219e-07, + "loss": 0.9787, + "step": 37516 + }, + { + "epoch": 0.88, + "grad_norm": 2.637736550294403, + "learning_rate": 6.989692009855543e-07, + "loss": 0.9703, + "step": 37517 + }, + { + "epoch": 0.88, + "grad_norm": 2.491027189534193, + "learning_rate": 6.986889641502326e-07, + "loss": 1.0512, + "step": 37518 + }, + { + "epoch": 0.88, + "grad_norm": 1.884483374328782, + "learning_rate": 6.984087814700902e-07, + "loss": 0.9758, + "step": 37519 + }, + { + "epoch": 0.88, + "grad_norm": 2.013824672907362, + "learning_rate": 6.981286529467546e-07, + "loss": 0.9684, + "step": 37520 + }, + { + "epoch": 0.88, + "grad_norm": 1.1381600846784274, + "learning_rate": 6.9784857858186e-07, + "loss": 0.9271, + "step": 37521 + }, + { + "epoch": 0.88, + "grad_norm": 1.9373524707243641, + "learning_rate": 6.975685583770341e-07, + "loss": 0.9184, + "step": 37522 + }, + { + "epoch": 0.88, + "grad_norm": 2.2992311941037573, + "learning_rate": 6.972885923339123e-07, + "loss": 0.8785, + "step": 37523 + }, + { + "epoch": 0.88, + "grad_norm": 1.934166337135409, + "learning_rate": 6.970086804541187e-07, + "loss": 1.0371, + "step": 37524 + }, + { + "epoch": 0.88, + "grad_norm": 2.1968049788278927, + "learning_rate": 6.967288227392877e-07, + "loss": 0.9089, + "step": 37525 + }, + { + "epoch": 0.88, + "grad_norm": 1.9788765779265018, + "learning_rate": 6.964490191910456e-07, + "loss": 0.9597, + "step": 37526 + }, + { + "epoch": 0.88, + "grad_norm": 2.659236065541499, + "learning_rate": 6.961692698110223e-07, + "loss": 0.9234, + "step": 37527 + }, + { + "epoch": 0.88, + "grad_norm": 1.036322070803969, + "learning_rate": 6.958895746008499e-07, + "loss": 0.9184, + "step": 37528 + }, + { + "epoch": 0.88, + "grad_norm": 2.02586582628525, + "learning_rate": 6.956099335621513e-07, + "loss": 0.9211, + "step": 37529 + }, + { + "epoch": 0.88, + "grad_norm": 2.1490935156363165, + "learning_rate": 6.953303466965578e-07, + "loss": 0.9863, + "step": 37530 + }, + { + "epoch": 0.88, + "grad_norm": 2.302254681151506, + "learning_rate": 6.950508140056966e-07, + "loss": 0.9779, + "step": 37531 + }, + { + "epoch": 0.88, + "grad_norm": 2.159662547267497, + "learning_rate": 6.947713354911978e-07, + "loss": 1.0559, + "step": 37532 + }, + { + "epoch": 0.88, + "grad_norm": 2.1697775950951685, + "learning_rate": 6.944919111546844e-07, + "loss": 0.8502, + "step": 37533 + }, + { + "epoch": 0.88, + "grad_norm": 2.031826918657237, + "learning_rate": 6.94212540997784e-07, + "loss": 0.9103, + "step": 37534 + }, + { + "epoch": 0.88, + "grad_norm": 1.9284562881981213, + "learning_rate": 6.939332250221265e-07, + "loss": 0.9741, + "step": 37535 + }, + { + "epoch": 0.88, + "grad_norm": 2.056673256331286, + "learning_rate": 6.93653963229336e-07, + "loss": 0.9076, + "step": 37536 + }, + { + "epoch": 0.88, + "grad_norm": 2.030600568798344, + "learning_rate": 6.933747556210368e-07, + "loss": 0.8559, + "step": 37537 + }, + { + "epoch": 0.88, + "grad_norm": 1.9526508803687554, + "learning_rate": 6.930956021988567e-07, + "loss": 0.9767, + "step": 37538 + }, + { + "epoch": 0.88, + "grad_norm": 2.0134358487428834, + "learning_rate": 6.928165029644219e-07, + "loss": 0.862, + "step": 37539 + }, + { + "epoch": 0.88, + "grad_norm": 1.9231697460292652, + "learning_rate": 6.925374579193533e-07, + "loss": 1.0928, + "step": 37540 + }, + { + "epoch": 0.88, + "grad_norm": 2.3099718849145754, + "learning_rate": 6.92258467065281e-07, + "loss": 1.0548, + "step": 37541 + }, + { + "epoch": 0.88, + "grad_norm": 1.0519984256971195, + "learning_rate": 6.919795304038268e-07, + "loss": 0.9003, + "step": 37542 + }, + { + "epoch": 0.88, + "grad_norm": 2.094112698322175, + "learning_rate": 6.917006479366128e-07, + "loss": 0.9717, + "step": 37543 + }, + { + "epoch": 0.88, + "grad_norm": 2.0072316663632073, + "learning_rate": 6.914218196652645e-07, + "loss": 1.0157, + "step": 37544 + }, + { + "epoch": 0.88, + "grad_norm": 1.9114435363147206, + "learning_rate": 6.91143045591407e-07, + "loss": 0.9032, + "step": 37545 + }, + { + "epoch": 0.88, + "grad_norm": 2.0381957233856336, + "learning_rate": 6.908643257166625e-07, + "loss": 0.9131, + "step": 37546 + }, + { + "epoch": 0.88, + "grad_norm": 1.9703723380093376, + "learning_rate": 6.905856600426508e-07, + "loss": 1.0229, + "step": 37547 + }, + { + "epoch": 0.88, + "grad_norm": 1.797004487461776, + "learning_rate": 6.903070485709962e-07, + "loss": 0.7597, + "step": 37548 + }, + { + "epoch": 0.88, + "grad_norm": 2.2416059876167527, + "learning_rate": 6.90028491303324e-07, + "loss": 0.9337, + "step": 37549 + }, + { + "epoch": 0.88, + "grad_norm": 2.287094386546003, + "learning_rate": 6.897499882412529e-07, + "loss": 0.9718, + "step": 37550 + }, + { + "epoch": 0.88, + "grad_norm": 1.1021086652678684, + "learning_rate": 6.894715393864027e-07, + "loss": 0.9304, + "step": 37551 + }, + { + "epoch": 0.88, + "grad_norm": 1.888833705957204, + "learning_rate": 6.891931447403977e-07, + "loss": 0.9833, + "step": 37552 + }, + { + "epoch": 0.88, + "grad_norm": 2.068787784691447, + "learning_rate": 6.889148043048599e-07, + "loss": 0.9331, + "step": 37553 + }, + { + "epoch": 0.88, + "grad_norm": 2.2362415936233, + "learning_rate": 6.886365180814069e-07, + "loss": 0.902, + "step": 37554 + }, + { + "epoch": 0.88, + "grad_norm": 2.2056437526946038, + "learning_rate": 6.883582860716598e-07, + "loss": 1.026, + "step": 37555 + }, + { + "epoch": 0.88, + "grad_norm": 1.9396446279702109, + "learning_rate": 6.880801082772382e-07, + "loss": 1.0182, + "step": 37556 + }, + { + "epoch": 0.88, + "grad_norm": 1.9745679612467306, + "learning_rate": 6.87801984699763e-07, + "loss": 0.992, + "step": 37557 + }, + { + "epoch": 0.88, + "grad_norm": 2.049805572567731, + "learning_rate": 6.875239153408541e-07, + "loss": 0.9434, + "step": 37558 + }, + { + "epoch": 0.88, + "grad_norm": 2.0919142383175897, + "learning_rate": 6.87245900202127e-07, + "loss": 0.9275, + "step": 37559 + }, + { + "epoch": 0.88, + "grad_norm": 2.2250042061334003, + "learning_rate": 6.869679392852035e-07, + "loss": 1.0395, + "step": 37560 + }, + { + "epoch": 0.88, + "grad_norm": 1.10488547461703, + "learning_rate": 6.866900325917003e-07, + "loss": 0.8882, + "step": 37561 + }, + { + "epoch": 0.88, + "grad_norm": 2.2950320890014977, + "learning_rate": 6.864121801232381e-07, + "loss": 0.9918, + "step": 37562 + }, + { + "epoch": 0.88, + "grad_norm": 1.9832361414108806, + "learning_rate": 6.861343818814315e-07, + "loss": 1.0683, + "step": 37563 + }, + { + "epoch": 0.88, + "grad_norm": 2.209558963038375, + "learning_rate": 6.858566378678999e-07, + "loss": 0.9209, + "step": 37564 + }, + { + "epoch": 0.89, + "grad_norm": 1.892311086209757, + "learning_rate": 6.855789480842589e-07, + "loss": 1.0402, + "step": 37565 + }, + { + "epoch": 0.89, + "grad_norm": 2.0056198452079834, + "learning_rate": 6.853013125321273e-07, + "loss": 1.0188, + "step": 37566 + }, + { + "epoch": 0.89, + "grad_norm": 1.908647733431708, + "learning_rate": 6.850237312131213e-07, + "loss": 0.9619, + "step": 37567 + }, + { + "epoch": 0.89, + "grad_norm": 1.7791293457400537, + "learning_rate": 6.847462041288533e-07, + "loss": 0.9215, + "step": 37568 + }, + { + "epoch": 0.89, + "grad_norm": 1.875418284010516, + "learning_rate": 6.844687312809429e-07, + "loss": 0.9606, + "step": 37569 + }, + { + "epoch": 0.89, + "grad_norm": 1.0811134973241543, + "learning_rate": 6.841913126710065e-07, + "loss": 0.9491, + "step": 37570 + }, + { + "epoch": 0.89, + "grad_norm": 1.9727693609701402, + "learning_rate": 6.839139483006574e-07, + "loss": 0.9427, + "step": 37571 + }, + { + "epoch": 0.89, + "grad_norm": 1.8485961931594737, + "learning_rate": 6.836366381715087e-07, + "loss": 1.0445, + "step": 37572 + }, + { + "epoch": 0.89, + "grad_norm": 1.1019628949883342, + "learning_rate": 6.83359382285177e-07, + "loss": 0.9627, + "step": 37573 + }, + { + "epoch": 0.89, + "grad_norm": 2.3903510433032693, + "learning_rate": 6.830821806432786e-07, + "loss": 0.9532, + "step": 37574 + }, + { + "epoch": 0.89, + "grad_norm": 1.0961134414516738, + "learning_rate": 6.828050332474256e-07, + "loss": 0.9355, + "step": 37575 + }, + { + "epoch": 0.89, + "grad_norm": 1.0599325783618903, + "learning_rate": 6.825279400992291e-07, + "loss": 0.8959, + "step": 37576 + }, + { + "epoch": 0.89, + "grad_norm": 2.0515246490308314, + "learning_rate": 6.822509012003054e-07, + "loss": 0.9647, + "step": 37577 + }, + { + "epoch": 0.89, + "grad_norm": 1.9622253032359096, + "learning_rate": 6.819739165522688e-07, + "loss": 0.9312, + "step": 37578 + }, + { + "epoch": 0.89, + "grad_norm": 1.9398082422927436, + "learning_rate": 6.816969861567291e-07, + "loss": 1.0568, + "step": 37579 + }, + { + "epoch": 0.89, + "grad_norm": 1.0533708420390815, + "learning_rate": 6.814201100152995e-07, + "loss": 0.9367, + "step": 37580 + }, + { + "epoch": 0.89, + "grad_norm": 2.003618548052074, + "learning_rate": 6.811432881295921e-07, + "loss": 0.9417, + "step": 37581 + }, + { + "epoch": 0.89, + "grad_norm": 1.9972349276236263, + "learning_rate": 6.808665205012188e-07, + "loss": 1.0167, + "step": 37582 + }, + { + "epoch": 0.89, + "grad_norm": 2.1940131162189136, + "learning_rate": 6.805898071317918e-07, + "loss": 1.0485, + "step": 37583 + }, + { + "epoch": 0.89, + "grad_norm": 1.0998042875698029, + "learning_rate": 6.803131480229197e-07, + "loss": 0.9748, + "step": 37584 + }, + { + "epoch": 0.89, + "grad_norm": 2.184022421634235, + "learning_rate": 6.800365431762168e-07, + "loss": 1.0269, + "step": 37585 + }, + { + "epoch": 0.89, + "grad_norm": 2.553058697427318, + "learning_rate": 6.797599925932907e-07, + "loss": 1.0618, + "step": 37586 + }, + { + "epoch": 0.89, + "grad_norm": 2.1644403441085696, + "learning_rate": 6.794834962757535e-07, + "loss": 0.8923, + "step": 37587 + }, + { + "epoch": 0.89, + "grad_norm": 1.820800418219475, + "learning_rate": 6.792070542252138e-07, + "loss": 0.8749, + "step": 37588 + }, + { + "epoch": 0.89, + "grad_norm": 2.0455323808509154, + "learning_rate": 6.789306664432815e-07, + "loss": 0.9764, + "step": 37589 + }, + { + "epoch": 0.89, + "grad_norm": 2.0195406566171146, + "learning_rate": 6.786543329315643e-07, + "loss": 1.0219, + "step": 37590 + }, + { + "epoch": 0.89, + "grad_norm": 2.18115628707338, + "learning_rate": 6.783780536916751e-07, + "loss": 0.945, + "step": 37591 + }, + { + "epoch": 0.89, + "grad_norm": 2.5499581849974176, + "learning_rate": 6.781018287252206e-07, + "loss": 0.8985, + "step": 37592 + }, + { + "epoch": 0.89, + "grad_norm": 2.580047986531725, + "learning_rate": 6.77825658033806e-07, + "loss": 0.9889, + "step": 37593 + }, + { + "epoch": 0.89, + "grad_norm": 2.1168312717943913, + "learning_rate": 6.775495416190414e-07, + "loss": 0.9718, + "step": 37594 + }, + { + "epoch": 0.89, + "grad_norm": 2.3245164650487697, + "learning_rate": 6.772734794825375e-07, + "loss": 0.9591, + "step": 37595 + }, + { + "epoch": 0.89, + "grad_norm": 2.0244582689393122, + "learning_rate": 6.769974716258987e-07, + "loss": 0.9519, + "step": 37596 + }, + { + "epoch": 0.89, + "grad_norm": 1.9049370114027802, + "learning_rate": 6.767215180507303e-07, + "loss": 1.0262, + "step": 37597 + }, + { + "epoch": 0.89, + "grad_norm": 1.926438727278927, + "learning_rate": 6.76445618758641e-07, + "loss": 0.9579, + "step": 37598 + }, + { + "epoch": 0.89, + "grad_norm": 1.9164236128442735, + "learning_rate": 6.761697737512385e-07, + "loss": 0.8186, + "step": 37599 + }, + { + "epoch": 0.89, + "grad_norm": 1.9668501035359296, + "learning_rate": 6.75893983030127e-07, + "loss": 1.0052, + "step": 37600 + }, + { + "epoch": 0.89, + "grad_norm": 2.1866670361069587, + "learning_rate": 6.756182465969119e-07, + "loss": 0.9697, + "step": 37601 + }, + { + "epoch": 0.89, + "grad_norm": 2.1206269972400342, + "learning_rate": 6.753425644532008e-07, + "loss": 1.0978, + "step": 37602 + }, + { + "epoch": 0.89, + "grad_norm": 1.0903282275434636, + "learning_rate": 6.750669366005957e-07, + "loss": 0.9311, + "step": 37603 + }, + { + "epoch": 0.89, + "grad_norm": 2.0043989474193324, + "learning_rate": 6.747913630407054e-07, + "loss": 0.993, + "step": 37604 + }, + { + "epoch": 0.89, + "grad_norm": 1.102941640709228, + "learning_rate": 6.745158437751298e-07, + "loss": 0.9284, + "step": 37605 + }, + { + "epoch": 0.89, + "grad_norm": 1.9485621829503672, + "learning_rate": 6.742403788054775e-07, + "loss": 1.1006, + "step": 37606 + }, + { + "epoch": 0.89, + "grad_norm": 1.9108467748714826, + "learning_rate": 6.739649681333482e-07, + "loss": 1.0845, + "step": 37607 + }, + { + "epoch": 0.89, + "grad_norm": 2.0871519075709917, + "learning_rate": 6.736896117603487e-07, + "loss": 1.0081, + "step": 37608 + }, + { + "epoch": 0.89, + "grad_norm": 1.155943044065606, + "learning_rate": 6.734143096880808e-07, + "loss": 0.9655, + "step": 37609 + }, + { + "epoch": 0.89, + "grad_norm": 2.195236750204545, + "learning_rate": 6.731390619181466e-07, + "loss": 0.984, + "step": 37610 + }, + { + "epoch": 0.89, + "grad_norm": 2.066415617485303, + "learning_rate": 6.728638684521493e-07, + "loss": 1.0227, + "step": 37611 + }, + { + "epoch": 0.89, + "grad_norm": 1.997225213397413, + "learning_rate": 6.725887292916933e-07, + "loss": 0.977, + "step": 37612 + }, + { + "epoch": 0.89, + "grad_norm": 2.1201817547926076, + "learning_rate": 6.723136444383793e-07, + "loss": 1.067, + "step": 37613 + }, + { + "epoch": 0.89, + "grad_norm": 1.8632677042641685, + "learning_rate": 6.720386138938062e-07, + "loss": 0.9887, + "step": 37614 + }, + { + "epoch": 0.89, + "grad_norm": 1.7221053039472767, + "learning_rate": 6.717636376595771e-07, + "loss": 0.8813, + "step": 37615 + }, + { + "epoch": 0.89, + "grad_norm": 1.855271345353827, + "learning_rate": 6.714887157372952e-07, + "loss": 0.9646, + "step": 37616 + }, + { + "epoch": 0.89, + "grad_norm": 1.1230678935779246, + "learning_rate": 6.71213848128559e-07, + "loss": 0.9546, + "step": 37617 + }, + { + "epoch": 0.89, + "grad_norm": 1.9909958089448965, + "learning_rate": 6.709390348349687e-07, + "loss": 1.0173, + "step": 37618 + }, + { + "epoch": 0.89, + "grad_norm": 2.5883445001649537, + "learning_rate": 6.706642758581239e-07, + "loss": 0.9277, + "step": 37619 + }, + { + "epoch": 0.89, + "grad_norm": 1.9788870112213972, + "learning_rate": 6.703895711996278e-07, + "loss": 0.7837, + "step": 37620 + }, + { + "epoch": 0.89, + "grad_norm": 2.0111870484320042, + "learning_rate": 6.701149208610758e-07, + "loss": 0.8552, + "step": 37621 + }, + { + "epoch": 0.89, + "grad_norm": 1.88784401754885, + "learning_rate": 6.698403248440688e-07, + "loss": 0.906, + "step": 37622 + }, + { + "epoch": 0.89, + "grad_norm": 1.845906780627896, + "learning_rate": 6.695657831502056e-07, + "loss": 1.0496, + "step": 37623 + }, + { + "epoch": 0.89, + "grad_norm": 1.9915311949511139, + "learning_rate": 6.692912957810826e-07, + "loss": 1.0271, + "step": 37624 + }, + { + "epoch": 0.89, + "grad_norm": 2.0949025236375314, + "learning_rate": 6.69016862738301e-07, + "loss": 0.9496, + "step": 37625 + }, + { + "epoch": 0.89, + "grad_norm": 1.9737089601084368, + "learning_rate": 6.687424840234579e-07, + "loss": 0.9123, + "step": 37626 + }, + { + "epoch": 0.89, + "grad_norm": 3.24405590915077, + "learning_rate": 6.684681596381503e-07, + "loss": 0.9733, + "step": 37627 + }, + { + "epoch": 0.89, + "grad_norm": 2.0347987447190032, + "learning_rate": 6.681938895839746e-07, + "loss": 0.8619, + "step": 37628 + }, + { + "epoch": 0.89, + "grad_norm": 2.008585336301075, + "learning_rate": 6.679196738625282e-07, + "loss": 0.8563, + "step": 37629 + }, + { + "epoch": 0.89, + "grad_norm": 2.008915596886491, + "learning_rate": 6.676455124754099e-07, + "loss": 0.9065, + "step": 37630 + }, + { + "epoch": 0.89, + "grad_norm": 2.0682904612017716, + "learning_rate": 6.67371405424212e-07, + "loss": 0.9894, + "step": 37631 + }, + { + "epoch": 0.89, + "grad_norm": 3.216489840911448, + "learning_rate": 6.670973527105329e-07, + "loss": 1.0202, + "step": 37632 + }, + { + "epoch": 0.89, + "grad_norm": 2.0738990926549894, + "learning_rate": 6.66823354335967e-07, + "loss": 0.9046, + "step": 37633 + }, + { + "epoch": 0.89, + "grad_norm": 1.9006093432049862, + "learning_rate": 6.66549410302113e-07, + "loss": 1.0151, + "step": 37634 + }, + { + "epoch": 0.89, + "grad_norm": 2.1841447450429627, + "learning_rate": 6.662755206105608e-07, + "loss": 0.8795, + "step": 37635 + }, + { + "epoch": 0.89, + "grad_norm": 1.9761853428374363, + "learning_rate": 6.66001685262908e-07, + "loss": 0.8967, + "step": 37636 + }, + { + "epoch": 0.89, + "grad_norm": 1.9094777716013727, + "learning_rate": 6.65727904260749e-07, + "loss": 1.0211, + "step": 37637 + }, + { + "epoch": 0.89, + "grad_norm": 2.05922565520158, + "learning_rate": 6.654541776056767e-07, + "loss": 0.9612, + "step": 37638 + }, + { + "epoch": 0.89, + "grad_norm": 2.055385974798989, + "learning_rate": 6.651805052992866e-07, + "loss": 0.9161, + "step": 37639 + }, + { + "epoch": 0.89, + "grad_norm": 2.0599449636215796, + "learning_rate": 6.649068873431697e-07, + "loss": 0.9044, + "step": 37640 + }, + { + "epoch": 0.89, + "grad_norm": 1.0668063014885782, + "learning_rate": 6.646333237389213e-07, + "loss": 0.9499, + "step": 37641 + }, + { + "epoch": 0.89, + "grad_norm": 1.0994480937540272, + "learning_rate": 6.643598144881324e-07, + "loss": 0.9608, + "step": 37642 + }, + { + "epoch": 0.89, + "grad_norm": 1.8724880675223259, + "learning_rate": 6.640863595923975e-07, + "loss": 1.0421, + "step": 37643 + }, + { + "epoch": 0.89, + "grad_norm": 2.0877358748461066, + "learning_rate": 6.638129590533071e-07, + "loss": 0.9897, + "step": 37644 + }, + { + "epoch": 0.89, + "grad_norm": 2.1243703702974632, + "learning_rate": 6.635396128724513e-07, + "loss": 0.9841, + "step": 37645 + }, + { + "epoch": 0.89, + "grad_norm": 1.8889395889955032, + "learning_rate": 6.632663210514257e-07, + "loss": 0.8943, + "step": 37646 + }, + { + "epoch": 0.89, + "grad_norm": 1.795923195977342, + "learning_rate": 6.629930835918196e-07, + "loss": 1.1881, + "step": 37647 + }, + { + "epoch": 0.89, + "grad_norm": 1.893716118031641, + "learning_rate": 6.627199004952245e-07, + "loss": 0.9073, + "step": 37648 + }, + { + "epoch": 0.89, + "grad_norm": 1.9254039188359668, + "learning_rate": 6.624467717632277e-07, + "loss": 0.9684, + "step": 37649 + }, + { + "epoch": 0.89, + "grad_norm": 2.284022693204953, + "learning_rate": 6.621736973974236e-07, + "loss": 0.9288, + "step": 37650 + }, + { + "epoch": 0.89, + "grad_norm": 2.0883809434262273, + "learning_rate": 6.619006773994008e-07, + "loss": 0.903, + "step": 37651 + }, + { + "epoch": 0.89, + "grad_norm": 1.8779789665355326, + "learning_rate": 6.616277117707493e-07, + "loss": 0.9671, + "step": 37652 + }, + { + "epoch": 0.89, + "grad_norm": 2.2911919956214484, + "learning_rate": 6.613548005130555e-07, + "loss": 1.0538, + "step": 37653 + }, + { + "epoch": 0.89, + "grad_norm": 2.00185480742823, + "learning_rate": 6.610819436279115e-07, + "loss": 0.9558, + "step": 37654 + }, + { + "epoch": 0.89, + "grad_norm": 1.960535141319688, + "learning_rate": 6.60809141116906e-07, + "loss": 0.9272, + "step": 37655 + }, + { + "epoch": 0.89, + "grad_norm": 2.03358638613621, + "learning_rate": 6.605363929816266e-07, + "loss": 1.0836, + "step": 37656 + }, + { + "epoch": 0.89, + "grad_norm": 2.4331080848937003, + "learning_rate": 6.6026369922366e-07, + "loss": 0.9714, + "step": 37657 + }, + { + "epoch": 0.89, + "grad_norm": 1.9352302425205534, + "learning_rate": 6.599910598445958e-07, + "loss": 0.9297, + "step": 37658 + }, + { + "epoch": 0.89, + "grad_norm": 1.0526417908713093, + "learning_rate": 6.597184748460195e-07, + "loss": 0.9402, + "step": 37659 + }, + { + "epoch": 0.89, + "grad_norm": 1.1466790052149411, + "learning_rate": 6.594459442295209e-07, + "loss": 0.9419, + "step": 37660 + }, + { + "epoch": 0.89, + "grad_norm": 1.8904319382104442, + "learning_rate": 6.591734679966844e-07, + "loss": 0.8993, + "step": 37661 + }, + { + "epoch": 0.89, + "grad_norm": 1.9795089571825413, + "learning_rate": 6.589010461490974e-07, + "loss": 0.9942, + "step": 37662 + }, + { + "epoch": 0.89, + "grad_norm": 1.8615669358599298, + "learning_rate": 6.586286786883455e-07, + "loss": 0.9019, + "step": 37663 + }, + { + "epoch": 0.89, + "grad_norm": 2.57525771254052, + "learning_rate": 6.583563656160152e-07, + "loss": 0.9243, + "step": 37664 + }, + { + "epoch": 0.89, + "grad_norm": 2.372583379143011, + "learning_rate": 6.580841069336929e-07, + "loss": 0.9179, + "step": 37665 + }, + { + "epoch": 0.89, + "grad_norm": 1.976271991174012, + "learning_rate": 6.578119026429597e-07, + "loss": 0.9665, + "step": 37666 + }, + { + "epoch": 0.89, + "grad_norm": 1.1268269405841573, + "learning_rate": 6.575397527454042e-07, + "loss": 0.9927, + "step": 37667 + }, + { + "epoch": 0.89, + "grad_norm": 1.8581974717195138, + "learning_rate": 6.572676572426118e-07, + "loss": 0.9196, + "step": 37668 + }, + { + "epoch": 0.89, + "grad_norm": 1.9961137422425492, + "learning_rate": 6.569956161361646e-07, + "loss": 0.814, + "step": 37669 + }, + { + "epoch": 0.89, + "grad_norm": 1.163880575808502, + "learning_rate": 6.567236294276446e-07, + "loss": 0.9842, + "step": 37670 + }, + { + "epoch": 0.89, + "grad_norm": 2.157627382769738, + "learning_rate": 6.564516971186385e-07, + "loss": 0.9605, + "step": 37671 + }, + { + "epoch": 0.89, + "grad_norm": 1.9083915591664988, + "learning_rate": 6.561798192107305e-07, + "loss": 1.0334, + "step": 37672 + }, + { + "epoch": 0.89, + "grad_norm": 2.1704269591969427, + "learning_rate": 6.559079957055015e-07, + "loss": 0.9635, + "step": 37673 + }, + { + "epoch": 0.89, + "grad_norm": 2.3843729485134855, + "learning_rate": 6.556362266045324e-07, + "loss": 1.0316, + "step": 37674 + }, + { + "epoch": 0.89, + "grad_norm": 2.138683577509874, + "learning_rate": 6.553645119094076e-07, + "loss": 1.0145, + "step": 37675 + }, + { + "epoch": 0.89, + "grad_norm": 1.8103126315967206, + "learning_rate": 6.550928516217115e-07, + "loss": 0.9538, + "step": 37676 + }, + { + "epoch": 0.89, + "grad_norm": 1.8637726344991448, + "learning_rate": 6.548212457430225e-07, + "loss": 0.9624, + "step": 37677 + }, + { + "epoch": 0.89, + "grad_norm": 1.9092617220949497, + "learning_rate": 6.545496942749219e-07, + "loss": 0.9236, + "step": 37678 + }, + { + "epoch": 0.89, + "grad_norm": 1.8860846619139777, + "learning_rate": 6.542781972189915e-07, + "loss": 0.9447, + "step": 37679 + }, + { + "epoch": 0.89, + "grad_norm": 1.8939379984814837, + "learning_rate": 6.540067545768136e-07, + "loss": 1.0388, + "step": 37680 + }, + { + "epoch": 0.89, + "grad_norm": 2.750136680171389, + "learning_rate": 6.537353663499669e-07, + "loss": 0.8635, + "step": 37681 + }, + { + "epoch": 0.89, + "grad_norm": 2.144575133009312, + "learning_rate": 6.53464032540031e-07, + "loss": 1.0764, + "step": 37682 + }, + { + "epoch": 0.89, + "grad_norm": 1.8535546490502492, + "learning_rate": 6.531927531485882e-07, + "loss": 0.9613, + "step": 37683 + }, + { + "epoch": 0.89, + "grad_norm": 1.922688844863991, + "learning_rate": 6.529215281772139e-07, + "loss": 0.9433, + "step": 37684 + }, + { + "epoch": 0.89, + "grad_norm": 2.112017766691006, + "learning_rate": 6.526503576274923e-07, + "loss": 0.9832, + "step": 37685 + }, + { + "epoch": 0.89, + "grad_norm": 1.9007915419916492, + "learning_rate": 6.523792415009977e-07, + "loss": 0.9993, + "step": 37686 + }, + { + "epoch": 0.89, + "grad_norm": 1.8260699610908122, + "learning_rate": 6.521081797993123e-07, + "loss": 1.0079, + "step": 37687 + }, + { + "epoch": 0.89, + "grad_norm": 2.0156904057376672, + "learning_rate": 6.518371725240103e-07, + "loss": 1.0761, + "step": 37688 + }, + { + "epoch": 0.89, + "grad_norm": 2.092427372188673, + "learning_rate": 6.515662196766737e-07, + "loss": 0.9819, + "step": 37689 + }, + { + "epoch": 0.89, + "grad_norm": 2.331985026838764, + "learning_rate": 6.512953212588791e-07, + "loss": 0.9414, + "step": 37690 + }, + { + "epoch": 0.89, + "grad_norm": 2.485859950068034, + "learning_rate": 6.510244772721996e-07, + "loss": 0.9658, + "step": 37691 + }, + { + "epoch": 0.89, + "grad_norm": 2.0489664323133296, + "learning_rate": 6.507536877182175e-07, + "loss": 1.0902, + "step": 37692 + }, + { + "epoch": 0.89, + "grad_norm": 2.2653830455584893, + "learning_rate": 6.504829525985079e-07, + "loss": 0.8933, + "step": 37693 + }, + { + "epoch": 0.89, + "grad_norm": 5.198599456974512, + "learning_rate": 6.502122719146464e-07, + "loss": 0.973, + "step": 37694 + }, + { + "epoch": 0.89, + "grad_norm": 1.690809316955754, + "learning_rate": 6.499416456682072e-07, + "loss": 0.9099, + "step": 37695 + }, + { + "epoch": 0.89, + "grad_norm": 1.0978572839965517, + "learning_rate": 6.496710738607681e-07, + "loss": 0.8852, + "step": 37696 + }, + { + "epoch": 0.89, + "grad_norm": 2.295750048373907, + "learning_rate": 6.494005564939066e-07, + "loss": 0.9325, + "step": 37697 + }, + { + "epoch": 0.89, + "grad_norm": 1.8584979724613957, + "learning_rate": 6.491300935691947e-07, + "loss": 0.8131, + "step": 37698 + }, + { + "epoch": 0.89, + "grad_norm": 1.9765514158609079, + "learning_rate": 6.488596850882068e-07, + "loss": 0.9381, + "step": 37699 + }, + { + "epoch": 0.89, + "grad_norm": 2.0480460560959077, + "learning_rate": 6.485893310525182e-07, + "loss": 0.946, + "step": 37700 + }, + { + "epoch": 0.89, + "grad_norm": 1.0899844651320965, + "learning_rate": 6.483190314637045e-07, + "loss": 0.9363, + "step": 37701 + }, + { + "epoch": 0.89, + "grad_norm": 2.01613457886304, + "learning_rate": 6.480487863233387e-07, + "loss": 1.0222, + "step": 37702 + }, + { + "epoch": 0.89, + "grad_norm": 1.8583111862166162, + "learning_rate": 6.477785956329907e-07, + "loss": 0.8834, + "step": 37703 + }, + { + "epoch": 0.89, + "grad_norm": 2.201777866949486, + "learning_rate": 6.475084593942393e-07, + "loss": 1.0774, + "step": 37704 + }, + { + "epoch": 0.89, + "grad_norm": 1.928491243754878, + "learning_rate": 6.472383776086533e-07, + "loss": 0.8738, + "step": 37705 + }, + { + "epoch": 0.89, + "grad_norm": 1.9461228242968398, + "learning_rate": 6.469683502778068e-07, + "loss": 1.1598, + "step": 37706 + }, + { + "epoch": 0.89, + "grad_norm": 2.113308128328791, + "learning_rate": 6.466983774032709e-07, + "loss": 0.9442, + "step": 37707 + }, + { + "epoch": 0.89, + "grad_norm": 1.7657202642458456, + "learning_rate": 6.464284589866199e-07, + "loss": 0.9619, + "step": 37708 + }, + { + "epoch": 0.89, + "grad_norm": 1.9472185239672053, + "learning_rate": 6.461585950294214e-07, + "loss": 0.9673, + "step": 37709 + }, + { + "epoch": 0.89, + "grad_norm": 2.0413568170455205, + "learning_rate": 6.458887855332518e-07, + "loss": 1.0242, + "step": 37710 + }, + { + "epoch": 0.89, + "grad_norm": 1.8709408488794397, + "learning_rate": 6.456190304996779e-07, + "loss": 0.9734, + "step": 37711 + }, + { + "epoch": 0.89, + "grad_norm": 1.9367224977652453, + "learning_rate": 6.453493299302704e-07, + "loss": 0.9654, + "step": 37712 + }, + { + "epoch": 0.89, + "grad_norm": 1.978895964221149, + "learning_rate": 6.450796838266015e-07, + "loss": 0.9694, + "step": 37713 + }, + { + "epoch": 0.89, + "grad_norm": 1.995434179204829, + "learning_rate": 6.44810092190239e-07, + "loss": 0.8382, + "step": 37714 + }, + { + "epoch": 0.89, + "grad_norm": 2.8830660931218013, + "learning_rate": 6.44540555022758e-07, + "loss": 0.787, + "step": 37715 + }, + { + "epoch": 0.89, + "grad_norm": 2.0162487798087625, + "learning_rate": 6.442710723257206e-07, + "loss": 0.9363, + "step": 37716 + }, + { + "epoch": 0.89, + "grad_norm": 2.037311071133169, + "learning_rate": 6.440016441006991e-07, + "loss": 1.0457, + "step": 37717 + }, + { + "epoch": 0.89, + "grad_norm": 2.628976632489671, + "learning_rate": 6.437322703492633e-07, + "loss": 0.8731, + "step": 37718 + }, + { + "epoch": 0.89, + "grad_norm": 1.8578530423111614, + "learning_rate": 6.434629510729806e-07, + "loss": 0.8188, + "step": 37719 + }, + { + "epoch": 0.89, + "grad_norm": 1.995466093960375, + "learning_rate": 6.431936862734167e-07, + "loss": 0.9353, + "step": 37720 + }, + { + "epoch": 0.89, + "grad_norm": 1.07905410580259, + "learning_rate": 6.429244759521425e-07, + "loss": 0.8997, + "step": 37721 + }, + { + "epoch": 0.89, + "grad_norm": 1.871383195081078, + "learning_rate": 6.426553201107256e-07, + "loss": 0.8421, + "step": 37722 + }, + { + "epoch": 0.89, + "grad_norm": 2.030697511995717, + "learning_rate": 6.423862187507291e-07, + "loss": 0.9744, + "step": 37723 + }, + { + "epoch": 0.89, + "grad_norm": 1.0321082076192492, + "learning_rate": 6.421171718737252e-07, + "loss": 0.8983, + "step": 37724 + }, + { + "epoch": 0.89, + "grad_norm": 1.1122096611823862, + "learning_rate": 6.418481794812781e-07, + "loss": 0.9319, + "step": 37725 + }, + { + "epoch": 0.89, + "grad_norm": 1.8189181149039793, + "learning_rate": 6.415792415749522e-07, + "loss": 0.9695, + "step": 37726 + }, + { + "epoch": 0.89, + "grad_norm": 1.8333093348067706, + "learning_rate": 6.41310358156314e-07, + "loss": 0.8907, + "step": 37727 + }, + { + "epoch": 0.89, + "grad_norm": 2.0787111438083374, + "learning_rate": 6.410415292269312e-07, + "loss": 0.9418, + "step": 37728 + }, + { + "epoch": 0.89, + "grad_norm": 1.8761802894841215, + "learning_rate": 6.407727547883691e-07, + "loss": 1.0364, + "step": 37729 + }, + { + "epoch": 0.89, + "grad_norm": 2.071366017046046, + "learning_rate": 6.405040348421876e-07, + "loss": 0.9427, + "step": 37730 + }, + { + "epoch": 0.89, + "grad_norm": 2.146270345364471, + "learning_rate": 6.402353693899566e-07, + "loss": 0.908, + "step": 37731 + }, + { + "epoch": 0.89, + "grad_norm": 3.0354539601396007, + "learning_rate": 6.399667584332403e-07, + "loss": 0.9238, + "step": 37732 + }, + { + "epoch": 0.89, + "grad_norm": 2.038530928557565, + "learning_rate": 6.396982019735987e-07, + "loss": 1.0583, + "step": 37733 + }, + { + "epoch": 0.89, + "grad_norm": 2.0058401093166, + "learning_rate": 6.394297000125971e-07, + "loss": 1.0443, + "step": 37734 + }, + { + "epoch": 0.89, + "grad_norm": 2.0764338600786836, + "learning_rate": 6.391612525517988e-07, + "loss": 0.9343, + "step": 37735 + }, + { + "epoch": 0.89, + "grad_norm": 2.048882754400314, + "learning_rate": 6.388928595927691e-07, + "loss": 0.8041, + "step": 37736 + }, + { + "epoch": 0.89, + "grad_norm": 1.8731992990688815, + "learning_rate": 6.38624521137069e-07, + "loss": 0.9385, + "step": 37737 + }, + { + "epoch": 0.89, + "grad_norm": 3.0825857263077565, + "learning_rate": 6.383562371862595e-07, + "loss": 1.0562, + "step": 37738 + }, + { + "epoch": 0.89, + "grad_norm": 2.254944293710941, + "learning_rate": 6.380880077419049e-07, + "loss": 0.9091, + "step": 37739 + }, + { + "epoch": 0.89, + "grad_norm": 2.0105073122599686, + "learning_rate": 6.378198328055652e-07, + "loss": 0.9373, + "step": 37740 + }, + { + "epoch": 0.89, + "grad_norm": 2.1502781639566573, + "learning_rate": 6.375517123788033e-07, + "loss": 0.929, + "step": 37741 + }, + { + "epoch": 0.89, + "grad_norm": 2.0150260116137506, + "learning_rate": 6.372836464631771e-07, + "loss": 0.9553, + "step": 37742 + }, + { + "epoch": 0.89, + "grad_norm": 2.1910413756619644, + "learning_rate": 6.37015635060253e-07, + "loss": 0.9964, + "step": 37743 + }, + { + "epoch": 0.89, + "grad_norm": 2.6576896669266543, + "learning_rate": 6.367476781715853e-07, + "loss": 0.8688, + "step": 37744 + }, + { + "epoch": 0.89, + "grad_norm": 1.9891096492061398, + "learning_rate": 6.364797757987384e-07, + "loss": 0.9054, + "step": 37745 + }, + { + "epoch": 0.89, + "grad_norm": 1.8357155313128195, + "learning_rate": 6.36211927943271e-07, + "loss": 0.9705, + "step": 37746 + }, + { + "epoch": 0.89, + "grad_norm": 1.8849123841427728, + "learning_rate": 6.359441346067419e-07, + "loss": 0.9623, + "step": 37747 + }, + { + "epoch": 0.89, + "grad_norm": 2.0510643937287374, + "learning_rate": 6.356763957907097e-07, + "loss": 0.9372, + "step": 37748 + }, + { + "epoch": 0.89, + "grad_norm": 2.3038727506201964, + "learning_rate": 6.354087114967355e-07, + "loss": 0.9859, + "step": 37749 + }, + { + "epoch": 0.89, + "grad_norm": 2.1203736520649543, + "learning_rate": 6.35141081726377e-07, + "loss": 0.8194, + "step": 37750 + }, + { + "epoch": 0.89, + "grad_norm": 2.1161155023430696, + "learning_rate": 6.348735064811895e-07, + "loss": 0.9269, + "step": 37751 + }, + { + "epoch": 0.89, + "grad_norm": 1.9583801014299478, + "learning_rate": 6.34605985762734e-07, + "loss": 0.9297, + "step": 37752 + }, + { + "epoch": 0.89, + "grad_norm": 1.9271573898491785, + "learning_rate": 6.343385195725693e-07, + "loss": 0.9445, + "step": 37753 + }, + { + "epoch": 0.89, + "grad_norm": 2.0775490858869587, + "learning_rate": 6.340711079122508e-07, + "loss": 0.9809, + "step": 37754 + }, + { + "epoch": 0.89, + "grad_norm": 2.1214572975729795, + "learning_rate": 6.338037507833339e-07, + "loss": 0.9074, + "step": 37755 + }, + { + "epoch": 0.89, + "grad_norm": 1.9966862218818549, + "learning_rate": 6.335364481873762e-07, + "loss": 0.9279, + "step": 37756 + }, + { + "epoch": 0.89, + "grad_norm": 1.884422842414995, + "learning_rate": 6.332692001259377e-07, + "loss": 1.0413, + "step": 37757 + }, + { + "epoch": 0.89, + "grad_norm": 2.3282474999734157, + "learning_rate": 6.330020066005704e-07, + "loss": 1.009, + "step": 37758 + }, + { + "epoch": 0.89, + "grad_norm": 2.0366691190653956, + "learning_rate": 6.327348676128298e-07, + "loss": 0.9145, + "step": 37759 + }, + { + "epoch": 0.89, + "grad_norm": 1.9434880952042644, + "learning_rate": 6.324677831642734e-07, + "loss": 0.9075, + "step": 37760 + }, + { + "epoch": 0.89, + "grad_norm": 2.0923857166201048, + "learning_rate": 6.322007532564533e-07, + "loss": 1.0773, + "step": 37761 + }, + { + "epoch": 0.89, + "grad_norm": 1.8342545983622902, + "learning_rate": 6.319337778909295e-07, + "loss": 1.0488, + "step": 37762 + }, + { + "epoch": 0.89, + "grad_norm": 2.1196626917772408, + "learning_rate": 6.316668570692497e-07, + "loss": 0.8693, + "step": 37763 + }, + { + "epoch": 0.89, + "grad_norm": 1.9758263363078639, + "learning_rate": 6.313999907929746e-07, + "loss": 0.947, + "step": 37764 + }, + { + "epoch": 0.89, + "grad_norm": 2.1430945831318637, + "learning_rate": 6.311331790636521e-07, + "loss": 1.0412, + "step": 37765 + }, + { + "epoch": 0.89, + "grad_norm": 1.7813984260473217, + "learning_rate": 6.308664218828397e-07, + "loss": 1.0304, + "step": 37766 + }, + { + "epoch": 0.89, + "grad_norm": 1.876950198105744, + "learning_rate": 6.305997192520908e-07, + "loss": 0.8674, + "step": 37767 + }, + { + "epoch": 0.89, + "grad_norm": 1.9030118263994218, + "learning_rate": 6.303330711729539e-07, + "loss": 0.9286, + "step": 37768 + }, + { + "epoch": 0.89, + "grad_norm": 1.8444641628661806, + "learning_rate": 6.300664776469845e-07, + "loss": 0.8822, + "step": 37769 + }, + { + "epoch": 0.89, + "grad_norm": 1.8703994560363846, + "learning_rate": 6.297999386757359e-07, + "loss": 0.9576, + "step": 37770 + }, + { + "epoch": 0.89, + "grad_norm": 1.9514151676624278, + "learning_rate": 6.29533454260759e-07, + "loss": 0.8269, + "step": 37771 + }, + { + "epoch": 0.89, + "grad_norm": 2.6303303271321665, + "learning_rate": 6.292670244036026e-07, + "loss": 0.9885, + "step": 37772 + }, + { + "epoch": 0.89, + "grad_norm": 1.8537810221538145, + "learning_rate": 6.290006491058209e-07, + "loss": 0.9076, + "step": 37773 + }, + { + "epoch": 0.89, + "grad_norm": 2.2018483328546083, + "learning_rate": 6.287343283689662e-07, + "loss": 1.0007, + "step": 37774 + }, + { + "epoch": 0.89, + "grad_norm": 2.0906647818551605, + "learning_rate": 6.28468062194586e-07, + "loss": 0.8443, + "step": 37775 + }, + { + "epoch": 0.89, + "grad_norm": 2.106626946742895, + "learning_rate": 6.282018505842314e-07, + "loss": 0.9563, + "step": 37776 + }, + { + "epoch": 0.89, + "grad_norm": 1.9637488533914498, + "learning_rate": 6.27935693539452e-07, + "loss": 0.8727, + "step": 37777 + }, + { + "epoch": 0.89, + "grad_norm": 1.9353411657329564, + "learning_rate": 6.276695910618002e-07, + "loss": 1.0657, + "step": 37778 + }, + { + "epoch": 0.89, + "grad_norm": 2.332719713612327, + "learning_rate": 6.274035431528225e-07, + "loss": 1.0527, + "step": 37779 + }, + { + "epoch": 0.89, + "grad_norm": 1.9918847862662097, + "learning_rate": 6.271375498140675e-07, + "loss": 1.0927, + "step": 37780 + }, + { + "epoch": 0.89, + "grad_norm": 2.3796224528987207, + "learning_rate": 6.268716110470863e-07, + "loss": 1.0129, + "step": 37781 + }, + { + "epoch": 0.89, + "grad_norm": 1.8466978299174166, + "learning_rate": 6.266057268534254e-07, + "loss": 0.9247, + "step": 37782 + }, + { + "epoch": 0.89, + "grad_norm": 1.9248459625028211, + "learning_rate": 6.263398972346346e-07, + "loss": 0.8802, + "step": 37783 + }, + { + "epoch": 0.89, + "grad_norm": 2.3015702062526033, + "learning_rate": 6.260741221922583e-07, + "loss": 1.0411, + "step": 37784 + }, + { + "epoch": 0.89, + "grad_norm": 2.065671126816217, + "learning_rate": 6.258084017278487e-07, + "loss": 0.9844, + "step": 37785 + }, + { + "epoch": 0.89, + "grad_norm": 1.8908560175529576, + "learning_rate": 6.255427358429489e-07, + "loss": 0.9339, + "step": 37786 + }, + { + "epoch": 0.89, + "grad_norm": 2.2258097370595262, + "learning_rate": 6.252771245391087e-07, + "loss": 1.1316, + "step": 37787 + }, + { + "epoch": 0.89, + "grad_norm": 1.8831588353128, + "learning_rate": 6.250115678178725e-07, + "loss": 0.8517, + "step": 37788 + }, + { + "epoch": 0.89, + "grad_norm": 2.20961523765678, + "learning_rate": 6.247460656807857e-07, + "loss": 0.9174, + "step": 37789 + }, + { + "epoch": 0.89, + "grad_norm": 2.218468602059586, + "learning_rate": 6.24480618129395e-07, + "loss": 0.9356, + "step": 37790 + }, + { + "epoch": 0.89, + "grad_norm": 1.8760242303444703, + "learning_rate": 6.242152251652489e-07, + "loss": 0.8774, + "step": 37791 + }, + { + "epoch": 0.89, + "grad_norm": 2.1619230260507236, + "learning_rate": 6.239498867898897e-07, + "loss": 1.0609, + "step": 37792 + }, + { + "epoch": 0.89, + "grad_norm": 1.866354100805967, + "learning_rate": 6.236846030048605e-07, + "loss": 1.0682, + "step": 37793 + }, + { + "epoch": 0.89, + "grad_norm": 1.8718360388393616, + "learning_rate": 6.234193738117089e-07, + "loss": 1.0061, + "step": 37794 + }, + { + "epoch": 0.89, + "grad_norm": 2.291132388184237, + "learning_rate": 6.231541992119794e-07, + "loss": 1.0498, + "step": 37795 + }, + { + "epoch": 0.89, + "grad_norm": 1.9032156967758305, + "learning_rate": 6.22889079207215e-07, + "loss": 0.9055, + "step": 37796 + }, + { + "epoch": 0.89, + "grad_norm": 1.9122037346783582, + "learning_rate": 6.226240137989581e-07, + "loss": 0.8967, + "step": 37797 + }, + { + "epoch": 0.89, + "grad_norm": 1.9753710029497629, + "learning_rate": 6.223590029887516e-07, + "loss": 1.0025, + "step": 37798 + }, + { + "epoch": 0.89, + "grad_norm": 1.8752168873923603, + "learning_rate": 6.220940467781422e-07, + "loss": 0.9132, + "step": 37799 + }, + { + "epoch": 0.89, + "grad_norm": 2.408460521323166, + "learning_rate": 6.218291451686709e-07, + "loss": 0.9159, + "step": 37800 + }, + { + "epoch": 0.89, + "grad_norm": 2.5710993652259253, + "learning_rate": 6.215642981618775e-07, + "loss": 0.9533, + "step": 37801 + }, + { + "epoch": 0.89, + "grad_norm": 1.8730428587365275, + "learning_rate": 6.212995057593052e-07, + "loss": 1.0169, + "step": 37802 + }, + { + "epoch": 0.89, + "grad_norm": 1.1269982856151026, + "learning_rate": 6.210347679624984e-07, + "loss": 0.9871, + "step": 37803 + }, + { + "epoch": 0.89, + "grad_norm": 1.95056616414167, + "learning_rate": 6.207700847729969e-07, + "loss": 1.0675, + "step": 37804 + }, + { + "epoch": 0.89, + "grad_norm": 2.428519960879784, + "learning_rate": 6.205054561923396e-07, + "loss": 1.0303, + "step": 37805 + }, + { + "epoch": 0.89, + "grad_norm": 1.7855842849815764, + "learning_rate": 6.202408822220696e-07, + "loss": 0.9043, + "step": 37806 + }, + { + "epoch": 0.89, + "grad_norm": 1.9248980806019105, + "learning_rate": 6.199763628637267e-07, + "loss": 1.0935, + "step": 37807 + }, + { + "epoch": 0.89, + "grad_norm": 1.7865362872455757, + "learning_rate": 6.197118981188499e-07, + "loss": 0.8611, + "step": 37808 + }, + { + "epoch": 0.89, + "grad_norm": 2.143937439190348, + "learning_rate": 6.194474879889823e-07, + "loss": 0.963, + "step": 37809 + }, + { + "epoch": 0.89, + "grad_norm": 2.284640549990279, + "learning_rate": 6.191831324756615e-07, + "loss": 0.8932, + "step": 37810 + }, + { + "epoch": 0.89, + "grad_norm": 2.524188694403078, + "learning_rate": 6.189188315804251e-07, + "loss": 0.8226, + "step": 37811 + }, + { + "epoch": 0.89, + "grad_norm": 2.3824209468541557, + "learning_rate": 6.18654585304812e-07, + "loss": 1.0345, + "step": 37812 + }, + { + "epoch": 0.89, + "grad_norm": 1.7910669441239102, + "learning_rate": 6.183903936503666e-07, + "loss": 0.8874, + "step": 37813 + }, + { + "epoch": 0.89, + "grad_norm": 1.162976301424469, + "learning_rate": 6.181262566186186e-07, + "loss": 0.9977, + "step": 37814 + }, + { + "epoch": 0.89, + "grad_norm": 2.204228650199494, + "learning_rate": 6.178621742111102e-07, + "loss": 1.0649, + "step": 37815 + }, + { + "epoch": 0.89, + "grad_norm": 1.90432618693265, + "learning_rate": 6.175981464293779e-07, + "loss": 0.9883, + "step": 37816 + }, + { + "epoch": 0.89, + "grad_norm": 1.866289721013227, + "learning_rate": 6.173341732749627e-07, + "loss": 1.0612, + "step": 37817 + }, + { + "epoch": 0.89, + "grad_norm": 2.0673654244258843, + "learning_rate": 6.170702547493956e-07, + "loss": 0.9939, + "step": 37818 + }, + { + "epoch": 0.89, + "grad_norm": 2.1198805972866235, + "learning_rate": 6.168063908542166e-07, + "loss": 1.0105, + "step": 37819 + }, + { + "epoch": 0.89, + "grad_norm": 1.8966342285384026, + "learning_rate": 6.16542581590962e-07, + "loss": 0.9841, + "step": 37820 + }, + { + "epoch": 0.89, + "grad_norm": 2.0387255299543265, + "learning_rate": 6.162788269611664e-07, + "loss": 1.0398, + "step": 37821 + }, + { + "epoch": 0.89, + "grad_norm": 3.3793055341857894, + "learning_rate": 6.160151269663684e-07, + "loss": 0.8982, + "step": 37822 + }, + { + "epoch": 0.89, + "grad_norm": 2.112834282809003, + "learning_rate": 6.157514816080989e-07, + "loss": 0.9221, + "step": 37823 + }, + { + "epoch": 0.89, + "grad_norm": 1.757284836448385, + "learning_rate": 6.15487890887897e-07, + "loss": 0.9668, + "step": 37824 + }, + { + "epoch": 0.89, + "grad_norm": 2.1445664811584324, + "learning_rate": 6.152243548072933e-07, + "loss": 0.9822, + "step": 37825 + }, + { + "epoch": 0.89, + "grad_norm": 2.049910554428271, + "learning_rate": 6.149608733678269e-07, + "loss": 0.9545, + "step": 37826 + }, + { + "epoch": 0.89, + "grad_norm": 2.0793130542198046, + "learning_rate": 6.146974465710298e-07, + "loss": 1.0485, + "step": 37827 + }, + { + "epoch": 0.89, + "grad_norm": 2.201147680934485, + "learning_rate": 6.144340744184341e-07, + "loss": 0.8192, + "step": 37828 + }, + { + "epoch": 0.89, + "grad_norm": 2.468970434919418, + "learning_rate": 6.14170756911574e-07, + "loss": 0.9773, + "step": 37829 + }, + { + "epoch": 0.89, + "grad_norm": 1.8955288149600222, + "learning_rate": 6.139074940519851e-07, + "loss": 1.022, + "step": 37830 + }, + { + "epoch": 0.89, + "grad_norm": 2.0951940081628746, + "learning_rate": 6.136442858411984e-07, + "loss": 1.1298, + "step": 37831 + }, + { + "epoch": 0.89, + "grad_norm": 2.101666502546165, + "learning_rate": 6.133811322807437e-07, + "loss": 0.9999, + "step": 37832 + }, + { + "epoch": 0.89, + "grad_norm": 2.0682892465544915, + "learning_rate": 6.131180333721576e-07, + "loss": 0.8802, + "step": 37833 + }, + { + "epoch": 0.89, + "grad_norm": 1.9248630832593907, + "learning_rate": 6.128549891169711e-07, + "loss": 0.9487, + "step": 37834 + }, + { + "epoch": 0.89, + "grad_norm": 2.042761483936166, + "learning_rate": 6.12591999516714e-07, + "loss": 1.1049, + "step": 37835 + }, + { + "epoch": 0.89, + "grad_norm": 1.0516815313952452, + "learning_rate": 6.123290645729174e-07, + "loss": 0.9824, + "step": 37836 + }, + { + "epoch": 0.89, + "grad_norm": 1.9209770609324053, + "learning_rate": 6.120661842871134e-07, + "loss": 0.9414, + "step": 37837 + }, + { + "epoch": 0.89, + "grad_norm": 1.9330525886045884, + "learning_rate": 6.11803358660833e-07, + "loss": 0.8942, + "step": 37838 + }, + { + "epoch": 0.89, + "grad_norm": 1.938632827689246, + "learning_rate": 6.11540587695606e-07, + "loss": 0.9225, + "step": 37839 + }, + { + "epoch": 0.89, + "grad_norm": 1.910716832677152, + "learning_rate": 6.112778713929601e-07, + "loss": 0.861, + "step": 37840 + }, + { + "epoch": 0.89, + "grad_norm": 1.8994586972432337, + "learning_rate": 6.110152097544286e-07, + "loss": 0.8595, + "step": 37841 + }, + { + "epoch": 0.89, + "grad_norm": 2.012720123727598, + "learning_rate": 6.10752602781538e-07, + "loss": 0.9194, + "step": 37842 + }, + { + "epoch": 0.89, + "grad_norm": 1.9164779433538888, + "learning_rate": 6.104900504758193e-07, + "loss": 1.0641, + "step": 37843 + }, + { + "epoch": 0.89, + "grad_norm": 2.5394866419457878, + "learning_rate": 6.10227552838799e-07, + "loss": 0.9152, + "step": 37844 + }, + { + "epoch": 0.89, + "grad_norm": 2.393261800419574, + "learning_rate": 6.099651098720083e-07, + "loss": 1.0208, + "step": 37845 + }, + { + "epoch": 0.89, + "grad_norm": 2.000311091096552, + "learning_rate": 6.097027215769725e-07, + "loss": 0.9747, + "step": 37846 + }, + { + "epoch": 0.89, + "grad_norm": 2.605680589163847, + "learning_rate": 6.094403879552213e-07, + "loss": 0.98, + "step": 37847 + }, + { + "epoch": 0.89, + "grad_norm": 1.9485446885702227, + "learning_rate": 6.091781090082804e-07, + "loss": 0.8812, + "step": 37848 + }, + { + "epoch": 0.89, + "grad_norm": 0.9980889937160051, + "learning_rate": 6.089158847376775e-07, + "loss": 0.8982, + "step": 37849 + }, + { + "epoch": 0.89, + "grad_norm": 2.0196098109677525, + "learning_rate": 6.08653715144939e-07, + "loss": 1.054, + "step": 37850 + }, + { + "epoch": 0.89, + "grad_norm": 2.5662973783101286, + "learning_rate": 6.083916002315926e-07, + "loss": 0.9653, + "step": 37851 + }, + { + "epoch": 0.89, + "grad_norm": 1.848397476367793, + "learning_rate": 6.081295399991649e-07, + "loss": 0.9401, + "step": 37852 + }, + { + "epoch": 0.89, + "grad_norm": 2.4509860862179385, + "learning_rate": 6.07867534449178e-07, + "loss": 1.0464, + "step": 37853 + }, + { + "epoch": 0.89, + "grad_norm": 5.740698990424667, + "learning_rate": 6.076055835831596e-07, + "loss": 0.9863, + "step": 37854 + }, + { + "epoch": 0.89, + "grad_norm": 2.0595440789410944, + "learning_rate": 6.073436874026373e-07, + "loss": 0.9844, + "step": 37855 + }, + { + "epoch": 0.89, + "grad_norm": 2.040404705283736, + "learning_rate": 6.070818459091333e-07, + "loss": 0.9128, + "step": 37856 + }, + { + "epoch": 0.89, + "grad_norm": 1.8965016346095982, + "learning_rate": 6.068200591041717e-07, + "loss": 0.9593, + "step": 37857 + }, + { + "epoch": 0.89, + "grad_norm": 2.247236073249674, + "learning_rate": 6.065583269892761e-07, + "loss": 0.9876, + "step": 37858 + }, + { + "epoch": 0.89, + "grad_norm": 1.140098473420003, + "learning_rate": 6.06296649565975e-07, + "loss": 0.9689, + "step": 37859 + }, + { + "epoch": 0.89, + "grad_norm": 1.9410669684648691, + "learning_rate": 6.060350268357873e-07, + "loss": 0.9375, + "step": 37860 + }, + { + "epoch": 0.89, + "grad_norm": 1.9629844650213242, + "learning_rate": 6.057734588002373e-07, + "loss": 1.0248, + "step": 37861 + }, + { + "epoch": 0.89, + "grad_norm": 1.9006852020157299, + "learning_rate": 6.055119454608493e-07, + "loss": 0.8152, + "step": 37862 + }, + { + "epoch": 0.89, + "grad_norm": 2.7547138770234745, + "learning_rate": 6.052504868191444e-07, + "loss": 1.0365, + "step": 37863 + }, + { + "epoch": 0.89, + "grad_norm": 2.0076665815995582, + "learning_rate": 6.049890828766458e-07, + "loss": 0.9144, + "step": 37864 + }, + { + "epoch": 0.89, + "grad_norm": 2.1419584056470917, + "learning_rate": 6.047277336348745e-07, + "loss": 1.1204, + "step": 37865 + }, + { + "epoch": 0.89, + "grad_norm": 1.8640949017975335, + "learning_rate": 6.044664390953547e-07, + "loss": 0.9342, + "step": 37866 + }, + { + "epoch": 0.89, + "grad_norm": 1.8038377482330077, + "learning_rate": 6.042051992596032e-07, + "loss": 0.9487, + "step": 37867 + }, + { + "epoch": 0.89, + "grad_norm": 2.4720004283198085, + "learning_rate": 6.039440141291464e-07, + "loss": 0.7981, + "step": 37868 + }, + { + "epoch": 0.89, + "grad_norm": 1.1093637824121019, + "learning_rate": 6.036828837055008e-07, + "loss": 0.9501, + "step": 37869 + }, + { + "epoch": 0.89, + "grad_norm": 1.0497522481862165, + "learning_rate": 6.034218079901877e-07, + "loss": 0.9361, + "step": 37870 + }, + { + "epoch": 0.89, + "grad_norm": 1.102929643426127, + "learning_rate": 6.031607869847278e-07, + "loss": 0.9227, + "step": 37871 + }, + { + "epoch": 0.89, + "grad_norm": 2.0324154000304704, + "learning_rate": 6.028998206906423e-07, + "loss": 0.9448, + "step": 37872 + }, + { + "epoch": 0.89, + "grad_norm": 1.8499917332628242, + "learning_rate": 6.026389091094487e-07, + "loss": 1.2081, + "step": 37873 + }, + { + "epoch": 0.89, + "grad_norm": 1.1043359895191192, + "learning_rate": 6.023780522426659e-07, + "loss": 0.9085, + "step": 37874 + }, + { + "epoch": 0.89, + "grad_norm": 1.780629348698943, + "learning_rate": 6.021172500918127e-07, + "loss": 0.9612, + "step": 37875 + }, + { + "epoch": 0.89, + "grad_norm": 1.9773354896503246, + "learning_rate": 6.018565026584089e-07, + "loss": 0.8272, + "step": 37876 + }, + { + "epoch": 0.89, + "grad_norm": 2.1728767494284895, + "learning_rate": 6.015958099439734e-07, + "loss": 1.0119, + "step": 37877 + }, + { + "epoch": 0.89, + "grad_norm": 1.9566652455759055, + "learning_rate": 6.013351719500205e-07, + "loss": 0.9505, + "step": 37878 + }, + { + "epoch": 0.89, + "grad_norm": 2.0341422549808317, + "learning_rate": 6.010745886780689e-07, + "loss": 1.0026, + "step": 37879 + }, + { + "epoch": 0.89, + "grad_norm": 1.1192815065337287, + "learning_rate": 6.008140601296398e-07, + "loss": 0.978, + "step": 37880 + }, + { + "epoch": 0.89, + "grad_norm": 1.8989788191100994, + "learning_rate": 6.005535863062461e-07, + "loss": 1.0356, + "step": 37881 + }, + { + "epoch": 0.89, + "grad_norm": 1.974165932858239, + "learning_rate": 6.002931672094037e-07, + "loss": 1.0486, + "step": 37882 + }, + { + "epoch": 0.89, + "grad_norm": 2.9750287810090974, + "learning_rate": 6.000328028406322e-07, + "loss": 0.9361, + "step": 37883 + }, + { + "epoch": 0.89, + "grad_norm": 2.019606772514965, + "learning_rate": 5.997724932014437e-07, + "loss": 1.0311, + "step": 37884 + }, + { + "epoch": 0.89, + "grad_norm": 3.467286980264128, + "learning_rate": 5.995122382933571e-07, + "loss": 0.7778, + "step": 37885 + }, + { + "epoch": 0.89, + "grad_norm": 2.1487698477831305, + "learning_rate": 5.992520381178857e-07, + "loss": 1.038, + "step": 37886 + }, + { + "epoch": 0.89, + "grad_norm": 1.9667395556866156, + "learning_rate": 5.98991892676547e-07, + "loss": 1.0251, + "step": 37887 + }, + { + "epoch": 0.89, + "grad_norm": 1.9817411319178486, + "learning_rate": 5.98731801970851e-07, + "loss": 0.938, + "step": 37888 + }, + { + "epoch": 0.89, + "grad_norm": 1.9694165193590576, + "learning_rate": 5.984717660023176e-07, + "loss": 0.92, + "step": 37889 + }, + { + "epoch": 0.89, + "grad_norm": 1.9681951503999424, + "learning_rate": 5.982117847724567e-07, + "loss": 0.8655, + "step": 37890 + }, + { + "epoch": 0.89, + "grad_norm": 2.0404427079062555, + "learning_rate": 5.979518582827826e-07, + "loss": 1.1286, + "step": 37891 + }, + { + "epoch": 0.89, + "grad_norm": 2.138188917898661, + "learning_rate": 5.976919865348097e-07, + "loss": 0.8247, + "step": 37892 + }, + { + "epoch": 0.89, + "grad_norm": 1.015519289099783, + "learning_rate": 5.974321695300523e-07, + "loss": 0.9271, + "step": 37893 + }, + { + "epoch": 0.89, + "grad_norm": 2.0903279162560002, + "learning_rate": 5.971724072700213e-07, + "loss": 1.0654, + "step": 37894 + }, + { + "epoch": 0.89, + "grad_norm": 2.038538007928559, + "learning_rate": 5.96912699756228e-07, + "loss": 1.0607, + "step": 37895 + }, + { + "epoch": 0.89, + "grad_norm": 2.2244136541466517, + "learning_rate": 5.966530469901854e-07, + "loss": 0.9858, + "step": 37896 + }, + { + "epoch": 0.89, + "grad_norm": 2.2865156978357417, + "learning_rate": 5.96393448973408e-07, + "loss": 0.8733, + "step": 37897 + }, + { + "epoch": 0.89, + "grad_norm": 2.3072108269331433, + "learning_rate": 5.961339057074045e-07, + "loss": 0.8953, + "step": 37898 + }, + { + "epoch": 0.89, + "grad_norm": 1.8816331227413137, + "learning_rate": 5.958744171936859e-07, + "loss": 0.9642, + "step": 37899 + }, + { + "epoch": 0.89, + "grad_norm": 1.8381365953259754, + "learning_rate": 5.956149834337632e-07, + "loss": 0.9006, + "step": 37900 + }, + { + "epoch": 0.89, + "grad_norm": 1.9076441768707606, + "learning_rate": 5.953556044291487e-07, + "loss": 0.8436, + "step": 37901 + }, + { + "epoch": 0.89, + "grad_norm": 2.159160890523284, + "learning_rate": 5.950962801813521e-07, + "loss": 0.8954, + "step": 37902 + }, + { + "epoch": 0.89, + "grad_norm": 2.1693086238277366, + "learning_rate": 5.948370106918799e-07, + "loss": 0.9246, + "step": 37903 + }, + { + "epoch": 0.89, + "grad_norm": 1.9598452953160737, + "learning_rate": 5.945777959622467e-07, + "loss": 0.9934, + "step": 37904 + }, + { + "epoch": 0.89, + "grad_norm": 2.086236176467128, + "learning_rate": 5.943186359939579e-07, + "loss": 0.9443, + "step": 37905 + }, + { + "epoch": 0.89, + "grad_norm": 1.047073353362013, + "learning_rate": 5.940595307885233e-07, + "loss": 0.9343, + "step": 37906 + }, + { + "epoch": 0.89, + "grad_norm": 1.1084204055062767, + "learning_rate": 5.93800480347454e-07, + "loss": 0.9533, + "step": 37907 + }, + { + "epoch": 0.89, + "grad_norm": 2.4682362747578876, + "learning_rate": 5.935414846722554e-07, + "loss": 1.0569, + "step": 37908 + }, + { + "epoch": 0.89, + "grad_norm": 2.0769513565315667, + "learning_rate": 5.932825437644352e-07, + "loss": 0.9975, + "step": 37909 + }, + { + "epoch": 0.89, + "grad_norm": 1.951227170382164, + "learning_rate": 5.930236576255022e-07, + "loss": 1.0475, + "step": 37910 + }, + { + "epoch": 0.89, + "grad_norm": 1.9497755471750438, + "learning_rate": 5.927648262569663e-07, + "loss": 0.9061, + "step": 37911 + }, + { + "epoch": 0.89, + "grad_norm": 2.108641407866803, + "learning_rate": 5.925060496603297e-07, + "loss": 1.0267, + "step": 37912 + }, + { + "epoch": 0.89, + "grad_norm": 1.8640063847772221, + "learning_rate": 5.92247327837101e-07, + "loss": 0.9288, + "step": 37913 + }, + { + "epoch": 0.89, + "grad_norm": 2.3120817627009633, + "learning_rate": 5.919886607887859e-07, + "loss": 0.9224, + "step": 37914 + }, + { + "epoch": 0.89, + "grad_norm": 1.8979632921411067, + "learning_rate": 5.917300485168953e-07, + "loss": 0.8781, + "step": 37915 + }, + { + "epoch": 0.89, + "grad_norm": 2.228143597440965, + "learning_rate": 5.91471491022928e-07, + "loss": 0.94, + "step": 37916 + }, + { + "epoch": 0.89, + "grad_norm": 1.049230550747916, + "learning_rate": 5.912129883083917e-07, + "loss": 0.9863, + "step": 37917 + }, + { + "epoch": 0.89, + "grad_norm": 2.155431212106663, + "learning_rate": 5.909545403747941e-07, + "loss": 1.0219, + "step": 37918 + }, + { + "epoch": 0.89, + "grad_norm": 1.882992267312017, + "learning_rate": 5.906961472236361e-07, + "loss": 1.0688, + "step": 37919 + }, + { + "epoch": 0.89, + "grad_norm": 2.079115143915686, + "learning_rate": 5.904378088564255e-07, + "loss": 0.9247, + "step": 37920 + }, + { + "epoch": 0.89, + "grad_norm": 2.434853066635753, + "learning_rate": 5.901795252746644e-07, + "loss": 0.9956, + "step": 37921 + }, + { + "epoch": 0.89, + "grad_norm": 1.7644292926113065, + "learning_rate": 5.899212964798584e-07, + "loss": 1.0761, + "step": 37922 + }, + { + "epoch": 0.89, + "grad_norm": 2.0078429707300494, + "learning_rate": 5.896631224735072e-07, + "loss": 1.0554, + "step": 37923 + }, + { + "epoch": 0.89, + "grad_norm": 1.9714018860178362, + "learning_rate": 5.894050032571197e-07, + "loss": 1.043, + "step": 37924 + }, + { + "epoch": 0.89, + "grad_norm": 2.023466473748467, + "learning_rate": 5.891469388321936e-07, + "loss": 0.9122, + "step": 37925 + }, + { + "epoch": 0.89, + "grad_norm": 2.235296143815171, + "learning_rate": 5.888889292002342e-07, + "loss": 1.0056, + "step": 37926 + }, + { + "epoch": 0.89, + "grad_norm": 1.9807214975375769, + "learning_rate": 5.886309743627416e-07, + "loss": 1.1514, + "step": 37927 + }, + { + "epoch": 0.89, + "grad_norm": 2.0796929243984854, + "learning_rate": 5.883730743212213e-07, + "loss": 1.0492, + "step": 37928 + }, + { + "epoch": 0.89, + "grad_norm": 1.9576250008867966, + "learning_rate": 5.881152290771719e-07, + "loss": 0.9679, + "step": 37929 + }, + { + "epoch": 0.89, + "grad_norm": 2.3454992951579876, + "learning_rate": 5.878574386320945e-07, + "loss": 1.0054, + "step": 37930 + }, + { + "epoch": 0.89, + "grad_norm": 2.0465098429107456, + "learning_rate": 5.875997029874903e-07, + "loss": 0.9809, + "step": 37931 + }, + { + "epoch": 0.89, + "grad_norm": 1.9752947869846234, + "learning_rate": 5.873420221448623e-07, + "loss": 0.9603, + "step": 37932 + }, + { + "epoch": 0.89, + "grad_norm": 3.1233732032219894, + "learning_rate": 5.870843961057093e-07, + "loss": 0.8491, + "step": 37933 + }, + { + "epoch": 0.89, + "grad_norm": 1.9171229425140677, + "learning_rate": 5.868268248715292e-07, + "loss": 0.9851, + "step": 37934 + }, + { + "epoch": 0.89, + "grad_norm": 1.9045048190558553, + "learning_rate": 5.86569308443824e-07, + "loss": 1.0037, + "step": 37935 + }, + { + "epoch": 0.89, + "grad_norm": 2.1489454635589, + "learning_rate": 5.863118468240936e-07, + "loss": 0.9918, + "step": 37936 + }, + { + "epoch": 0.89, + "grad_norm": 1.9401530069512416, + "learning_rate": 5.860544400138357e-07, + "loss": 0.9835, + "step": 37937 + }, + { + "epoch": 0.89, + "grad_norm": 1.8628834231252034, + "learning_rate": 5.857970880145492e-07, + "loss": 1.0002, + "step": 37938 + }, + { + "epoch": 0.89, + "grad_norm": 2.108601440926331, + "learning_rate": 5.855397908277305e-07, + "loss": 0.9481, + "step": 37939 + }, + { + "epoch": 0.89, + "grad_norm": 2.2789024767306625, + "learning_rate": 5.85282548454883e-07, + "loss": 1.0397, + "step": 37940 + }, + { + "epoch": 0.89, + "grad_norm": 1.9222136733716546, + "learning_rate": 5.850253608974998e-07, + "loss": 0.7687, + "step": 37941 + }, + { + "epoch": 0.89, + "grad_norm": 1.7809894360340017, + "learning_rate": 5.847682281570788e-07, + "loss": 0.9354, + "step": 37942 + }, + { + "epoch": 0.89, + "grad_norm": 2.4416596480473376, + "learning_rate": 5.845111502351197e-07, + "loss": 0.9556, + "step": 37943 + }, + { + "epoch": 0.89, + "grad_norm": 2.0553011147947324, + "learning_rate": 5.842541271331148e-07, + "loss": 0.9757, + "step": 37944 + }, + { + "epoch": 0.89, + "grad_norm": 2.304697834816083, + "learning_rate": 5.839971588525661e-07, + "loss": 1.0744, + "step": 37945 + }, + { + "epoch": 0.89, + "grad_norm": 2.410175031379372, + "learning_rate": 5.837402453949648e-07, + "loss": 0.9804, + "step": 37946 + }, + { + "epoch": 0.89, + "grad_norm": 2.0053110041693762, + "learning_rate": 5.834833867618117e-07, + "loss": 1.0984, + "step": 37947 + }, + { + "epoch": 0.89, + "grad_norm": 1.0598989221040098, + "learning_rate": 5.832265829545969e-07, + "loss": 0.9567, + "step": 37948 + }, + { + "epoch": 0.89, + "grad_norm": 1.0802216593993457, + "learning_rate": 5.829698339748202e-07, + "loss": 0.915, + "step": 37949 + }, + { + "epoch": 0.89, + "grad_norm": 1.9302692800211907, + "learning_rate": 5.827131398239738e-07, + "loss": 1.0092, + "step": 37950 + }, + { + "epoch": 0.89, + "grad_norm": 2.042941953490614, + "learning_rate": 5.824565005035521e-07, + "loss": 1.016, + "step": 37951 + }, + { + "epoch": 0.89, + "grad_norm": 1.0731402345259426, + "learning_rate": 5.821999160150504e-07, + "loss": 0.8745, + "step": 37952 + }, + { + "epoch": 0.89, + "grad_norm": 2.1274820253400697, + "learning_rate": 5.819433863599633e-07, + "loss": 0.9828, + "step": 37953 + }, + { + "epoch": 0.89, + "grad_norm": 1.0591584967695693, + "learning_rate": 5.816869115397838e-07, + "loss": 0.8799, + "step": 37954 + }, + { + "epoch": 0.89, + "grad_norm": 1.933555120144876, + "learning_rate": 5.814304915560031e-07, + "loss": 0.8964, + "step": 37955 + }, + { + "epoch": 0.89, + "grad_norm": 1.9198266232021082, + "learning_rate": 5.811741264101167e-07, + "loss": 0.9448, + "step": 37956 + }, + { + "epoch": 0.89, + "grad_norm": 2.065320440766953, + "learning_rate": 5.809178161036166e-07, + "loss": 0.9444, + "step": 37957 + }, + { + "epoch": 0.89, + "grad_norm": 2.0565178298790414, + "learning_rate": 5.806615606379961e-07, + "loss": 1.0598, + "step": 37958 + }, + { + "epoch": 0.89, + "grad_norm": 1.0457577621908483, + "learning_rate": 5.804053600147452e-07, + "loss": 0.9069, + "step": 37959 + }, + { + "epoch": 0.89, + "grad_norm": 1.8482996423251068, + "learning_rate": 5.801492142353549e-07, + "loss": 1.0111, + "step": 37960 + }, + { + "epoch": 0.89, + "grad_norm": 2.212095597956556, + "learning_rate": 5.798931233013205e-07, + "loss": 1.1068, + "step": 37961 + }, + { + "epoch": 0.89, + "grad_norm": 1.9445434444734038, + "learning_rate": 5.796370872141311e-07, + "loss": 0.8381, + "step": 37962 + }, + { + "epoch": 0.89, + "grad_norm": 2.441067770477892, + "learning_rate": 5.793811059752752e-07, + "loss": 0.9877, + "step": 37963 + }, + { + "epoch": 0.89, + "grad_norm": 2.08579223485072, + "learning_rate": 5.791251795862463e-07, + "loss": 0.949, + "step": 37964 + }, + { + "epoch": 0.89, + "grad_norm": 2.058469236768671, + "learning_rate": 5.788693080485319e-07, + "loss": 0.9284, + "step": 37965 + }, + { + "epoch": 0.89, + "grad_norm": 1.8552958365229852, + "learning_rate": 5.786134913636243e-07, + "loss": 0.8346, + "step": 37966 + }, + { + "epoch": 0.89, + "grad_norm": 1.8766248897426436, + "learning_rate": 5.7835772953301e-07, + "loss": 0.9464, + "step": 37967 + }, + { + "epoch": 0.89, + "grad_norm": 2.1367445138161707, + "learning_rate": 5.781020225581813e-07, + "loss": 1.0129, + "step": 37968 + }, + { + "epoch": 0.89, + "grad_norm": 2.154674301995707, + "learning_rate": 5.778463704406245e-07, + "loss": 0.9549, + "step": 37969 + }, + { + "epoch": 0.89, + "grad_norm": 1.9514113605958703, + "learning_rate": 5.775907731818308e-07, + "loss": 1.0215, + "step": 37970 + }, + { + "epoch": 0.89, + "grad_norm": 1.1038482732204302, + "learning_rate": 5.773352307832869e-07, + "loss": 0.8778, + "step": 37971 + }, + { + "epoch": 0.89, + "grad_norm": 2.0158137959167144, + "learning_rate": 5.770797432464781e-07, + "loss": 0.9622, + "step": 37972 + }, + { + "epoch": 0.89, + "grad_norm": 1.8660298306505572, + "learning_rate": 5.768243105728954e-07, + "loss": 0.9849, + "step": 37973 + }, + { + "epoch": 0.89, + "grad_norm": 2.132947246616674, + "learning_rate": 5.765689327640256e-07, + "loss": 0.9355, + "step": 37974 + }, + { + "epoch": 0.89, + "grad_norm": 1.973286803661593, + "learning_rate": 5.763136098213551e-07, + "loss": 1.0087, + "step": 37975 + }, + { + "epoch": 0.89, + "grad_norm": 1.9828716234175052, + "learning_rate": 5.760583417463683e-07, + "loss": 1.0448, + "step": 37976 + }, + { + "epoch": 0.89, + "grad_norm": 2.085047544963136, + "learning_rate": 5.758031285405541e-07, + "loss": 0.9944, + "step": 37977 + }, + { + "epoch": 0.89, + "grad_norm": 1.8503604054474736, + "learning_rate": 5.755479702053979e-07, + "loss": 0.9603, + "step": 37978 + }, + { + "epoch": 0.89, + "grad_norm": 1.958186718800953, + "learning_rate": 5.752928667423862e-07, + "loss": 0.8906, + "step": 37979 + }, + { + "epoch": 0.89, + "grad_norm": 1.821364237458582, + "learning_rate": 5.750378181530025e-07, + "loss": 0.9922, + "step": 37980 + }, + { + "epoch": 0.89, + "grad_norm": 1.970769419845173, + "learning_rate": 5.74782824438731e-07, + "loss": 1.0527, + "step": 37981 + }, + { + "epoch": 0.89, + "grad_norm": 2.0028782776452063, + "learning_rate": 5.745278856010605e-07, + "loss": 0.9651, + "step": 37982 + }, + { + "epoch": 0.89, + "grad_norm": 1.8359004814646536, + "learning_rate": 5.742730016414722e-07, + "loss": 0.9048, + "step": 37983 + }, + { + "epoch": 0.89, + "grad_norm": 1.9415312899818962, + "learning_rate": 5.740181725614491e-07, + "loss": 0.861, + "step": 37984 + }, + { + "epoch": 0.89, + "grad_norm": 1.8650063028253836, + "learning_rate": 5.737633983624779e-07, + "loss": 0.8296, + "step": 37985 + }, + { + "epoch": 0.89, + "grad_norm": 1.863994104398836, + "learning_rate": 5.735086790460387e-07, + "loss": 0.9003, + "step": 37986 + }, + { + "epoch": 0.89, + "grad_norm": 1.8524129997600034, + "learning_rate": 5.73254014613619e-07, + "loss": 0.8083, + "step": 37987 + }, + { + "epoch": 0.89, + "grad_norm": 1.0805563019091755, + "learning_rate": 5.729994050666965e-07, + "loss": 0.9172, + "step": 37988 + }, + { + "epoch": 0.89, + "grad_norm": 2.157141267295602, + "learning_rate": 5.727448504067578e-07, + "loss": 0.9742, + "step": 37989 + }, + { + "epoch": 0.9, + "grad_norm": 2.1239596360588515, + "learning_rate": 5.724903506352808e-07, + "loss": 1.1357, + "step": 37990 + }, + { + "epoch": 0.9, + "grad_norm": 2.1753481115122937, + "learning_rate": 5.722359057537519e-07, + "loss": 1.0531, + "step": 37991 + }, + { + "epoch": 0.9, + "grad_norm": 1.0540578437278791, + "learning_rate": 5.71981515763651e-07, + "loss": 0.9874, + "step": 37992 + }, + { + "epoch": 0.9, + "grad_norm": 1.953828843134284, + "learning_rate": 5.71727180666456e-07, + "loss": 1.0424, + "step": 37993 + }, + { + "epoch": 0.9, + "grad_norm": 2.03298167819251, + "learning_rate": 5.71472900463651e-07, + "loss": 0.8529, + "step": 37994 + }, + { + "epoch": 0.9, + "grad_norm": 1.9556768807627625, + "learning_rate": 5.71218675156715e-07, + "loss": 0.8286, + "step": 37995 + }, + { + "epoch": 0.9, + "grad_norm": 1.839137281480071, + "learning_rate": 5.709645047471324e-07, + "loss": 0.9396, + "step": 37996 + }, + { + "epoch": 0.9, + "grad_norm": 2.299132123327735, + "learning_rate": 5.707103892363774e-07, + "loss": 0.9289, + "step": 37997 + }, + { + "epoch": 0.9, + "grad_norm": 1.9588699302636643, + "learning_rate": 5.704563286259313e-07, + "loss": 0.988, + "step": 37998 + }, + { + "epoch": 0.9, + "grad_norm": 1.9571685099245775, + "learning_rate": 5.702023229172749e-07, + "loss": 0.8107, + "step": 37999 + }, + { + "epoch": 0.9, + "grad_norm": 2.083717389811172, + "learning_rate": 5.699483721118859e-07, + "loss": 1.0259, + "step": 38000 + }, + { + "epoch": 0.9, + "grad_norm": 1.9722816704153847, + "learning_rate": 5.696944762112422e-07, + "loss": 0.9695, + "step": 38001 + }, + { + "epoch": 0.9, + "grad_norm": 2.208834462281546, + "learning_rate": 5.694406352168214e-07, + "loss": 0.9989, + "step": 38002 + }, + { + "epoch": 0.9, + "grad_norm": 1.93261011926169, + "learning_rate": 5.691868491301056e-07, + "loss": 0.9043, + "step": 38003 + }, + { + "epoch": 0.9, + "grad_norm": 2.1705516227343393, + "learning_rate": 5.689331179525681e-07, + "loss": 1.1073, + "step": 38004 + }, + { + "epoch": 0.9, + "grad_norm": 2.0902639974764945, + "learning_rate": 5.686794416856889e-07, + "loss": 0.8921, + "step": 38005 + }, + { + "epoch": 0.9, + "grad_norm": 2.0416941103664015, + "learning_rate": 5.684258203309446e-07, + "loss": 1.1083, + "step": 38006 + }, + { + "epoch": 0.9, + "grad_norm": 2.207468301300549, + "learning_rate": 5.681722538898093e-07, + "loss": 0.919, + "step": 38007 + }, + { + "epoch": 0.9, + "grad_norm": 2.049831059673781, + "learning_rate": 5.679187423637611e-07, + "loss": 1.0463, + "step": 38008 + }, + { + "epoch": 0.9, + "grad_norm": 1.9736666176291853, + "learning_rate": 5.676652857542775e-07, + "loss": 0.9196, + "step": 38009 + }, + { + "epoch": 0.9, + "grad_norm": 1.0125882488516051, + "learning_rate": 5.674118840628329e-07, + "loss": 0.8452, + "step": 38010 + }, + { + "epoch": 0.9, + "grad_norm": 1.8145993472024362, + "learning_rate": 5.671585372909016e-07, + "loss": 0.9669, + "step": 38011 + }, + { + "epoch": 0.9, + "grad_norm": 2.0146363976147015, + "learning_rate": 5.669052454399593e-07, + "loss": 0.9584, + "step": 38012 + }, + { + "epoch": 0.9, + "grad_norm": 1.9231161772909444, + "learning_rate": 5.666520085114846e-07, + "loss": 0.9782, + "step": 38013 + }, + { + "epoch": 0.9, + "grad_norm": 2.0241533544909363, + "learning_rate": 5.663988265069454e-07, + "loss": 0.9842, + "step": 38014 + }, + { + "epoch": 0.9, + "grad_norm": 2.463999191880251, + "learning_rate": 5.661456994278181e-07, + "loss": 0.8451, + "step": 38015 + }, + { + "epoch": 0.9, + "grad_norm": 1.973454933252789, + "learning_rate": 5.658926272755772e-07, + "loss": 1.0481, + "step": 38016 + }, + { + "epoch": 0.9, + "grad_norm": 2.0255002649202054, + "learning_rate": 5.656396100516981e-07, + "loss": 1.0318, + "step": 38017 + }, + { + "epoch": 0.9, + "grad_norm": 2.097641454898437, + "learning_rate": 5.65386647757652e-07, + "loss": 1.0648, + "step": 38018 + }, + { + "epoch": 0.9, + "grad_norm": 1.1085901454841063, + "learning_rate": 5.651337403949109e-07, + "loss": 0.9087, + "step": 38019 + }, + { + "epoch": 0.9, + "grad_norm": 2.3717877458020804, + "learning_rate": 5.648808879649493e-07, + "loss": 0.9165, + "step": 38020 + }, + { + "epoch": 0.9, + "grad_norm": 1.8746893276436438, + "learning_rate": 5.64628090469237e-07, + "loss": 0.9149, + "step": 38021 + }, + { + "epoch": 0.9, + "grad_norm": 1.8560805183061526, + "learning_rate": 5.643753479092473e-07, + "loss": 0.9432, + "step": 38022 + }, + { + "epoch": 0.9, + "grad_norm": 2.0930330604524348, + "learning_rate": 5.641226602864514e-07, + "loss": 0.8224, + "step": 38023 + }, + { + "epoch": 0.9, + "grad_norm": 1.8800756273841555, + "learning_rate": 5.638700276023223e-07, + "loss": 1.1161, + "step": 38024 + }, + { + "epoch": 0.9, + "grad_norm": 2.1410643772077473, + "learning_rate": 5.636174498583269e-07, + "loss": 0.9102, + "step": 38025 + }, + { + "epoch": 0.9, + "grad_norm": 2.1574918190782983, + "learning_rate": 5.633649270559405e-07, + "loss": 1.0272, + "step": 38026 + }, + { + "epoch": 0.9, + "grad_norm": 1.7982732268891817, + "learning_rate": 5.631124591966308e-07, + "loss": 0.9019, + "step": 38027 + }, + { + "epoch": 0.9, + "grad_norm": 1.133001298011092, + "learning_rate": 5.628600462818668e-07, + "loss": 0.9363, + "step": 38028 + }, + { + "epoch": 0.9, + "grad_norm": 1.8908608121793196, + "learning_rate": 5.626076883131193e-07, + "loss": 0.8629, + "step": 38029 + }, + { + "epoch": 0.9, + "grad_norm": 1.8957090129152652, + "learning_rate": 5.623553852918584e-07, + "loss": 0.805, + "step": 38030 + }, + { + "epoch": 0.9, + "grad_norm": 2.030156117662454, + "learning_rate": 5.621031372195529e-07, + "loss": 1.1112, + "step": 38031 + }, + { + "epoch": 0.9, + "grad_norm": 1.8303647702316537, + "learning_rate": 5.618509440976694e-07, + "loss": 1.0261, + "step": 38032 + }, + { + "epoch": 0.9, + "grad_norm": 1.876751603902698, + "learning_rate": 5.615988059276777e-07, + "loss": 1.0097, + "step": 38033 + }, + { + "epoch": 0.9, + "grad_norm": 2.087615251446912, + "learning_rate": 5.613467227110481e-07, + "loss": 0.9566, + "step": 38034 + }, + { + "epoch": 0.9, + "grad_norm": 2.2203317889943452, + "learning_rate": 5.610946944492457e-07, + "loss": 0.8406, + "step": 38035 + }, + { + "epoch": 0.9, + "grad_norm": 2.00491552086993, + "learning_rate": 5.608427211437362e-07, + "loss": 0.9273, + "step": 38036 + }, + { + "epoch": 0.9, + "grad_norm": 2.0243793233613374, + "learning_rate": 5.605908027959894e-07, + "loss": 0.8746, + "step": 38037 + }, + { + "epoch": 0.9, + "grad_norm": 1.787634797454123, + "learning_rate": 5.603389394074732e-07, + "loss": 0.864, + "step": 38038 + }, + { + "epoch": 0.9, + "grad_norm": 2.1162320125192315, + "learning_rate": 5.600871309796519e-07, + "loss": 0.9499, + "step": 38039 + }, + { + "epoch": 0.9, + "grad_norm": 2.058028418228816, + "learning_rate": 5.598353775139898e-07, + "loss": 1.004, + "step": 38040 + }, + { + "epoch": 0.9, + "grad_norm": 1.7687782832677572, + "learning_rate": 5.595836790119569e-07, + "loss": 0.9526, + "step": 38041 + }, + { + "epoch": 0.9, + "grad_norm": 1.8979719865756752, + "learning_rate": 5.593320354750154e-07, + "loss": 0.968, + "step": 38042 + }, + { + "epoch": 0.9, + "grad_norm": 1.9232975545840443, + "learning_rate": 5.59080446904633e-07, + "loss": 0.9779, + "step": 38043 + }, + { + "epoch": 0.9, + "grad_norm": 1.9226655522595781, + "learning_rate": 5.588289133022706e-07, + "loss": 0.9891, + "step": 38044 + }, + { + "epoch": 0.9, + "grad_norm": 2.2848802149165075, + "learning_rate": 5.585774346693984e-07, + "loss": 0.9853, + "step": 38045 + }, + { + "epoch": 0.9, + "grad_norm": 1.0846860829971274, + "learning_rate": 5.583260110074751e-07, + "loss": 0.8758, + "step": 38046 + }, + { + "epoch": 0.9, + "grad_norm": 2.0474517711823403, + "learning_rate": 5.580746423179684e-07, + "loss": 0.898, + "step": 38047 + }, + { + "epoch": 0.9, + "grad_norm": 2.05177770497857, + "learning_rate": 5.578233286023393e-07, + "loss": 0.9081, + "step": 38048 + }, + { + "epoch": 0.9, + "grad_norm": 1.9546355894445187, + "learning_rate": 5.575720698620535e-07, + "loss": 0.8623, + "step": 38049 + }, + { + "epoch": 0.9, + "grad_norm": 1.9265300312901845, + "learning_rate": 5.573208660985707e-07, + "loss": 0.8709, + "step": 38050 + }, + { + "epoch": 0.9, + "grad_norm": 1.9248668978976273, + "learning_rate": 5.570697173133565e-07, + "loss": 0.8288, + "step": 38051 + }, + { + "epoch": 0.9, + "grad_norm": 1.9238858011599496, + "learning_rate": 5.568186235078732e-07, + "loss": 1.0432, + "step": 38052 + }, + { + "epoch": 0.9, + "grad_norm": 2.1406839049480992, + "learning_rate": 5.565675846835794e-07, + "loss": 0.7907, + "step": 38053 + }, + { + "epoch": 0.9, + "grad_norm": 1.7909057248981104, + "learning_rate": 5.563166008419385e-07, + "loss": 0.9606, + "step": 38054 + }, + { + "epoch": 0.9, + "grad_norm": 2.2854233730824443, + "learning_rate": 5.560656719844149e-07, + "loss": 0.9747, + "step": 38055 + }, + { + "epoch": 0.9, + "grad_norm": 2.1203264905607493, + "learning_rate": 5.55814798112465e-07, + "loss": 0.9757, + "step": 38056 + }, + { + "epoch": 0.9, + "grad_norm": 1.972817621735975, + "learning_rate": 5.555639792275514e-07, + "loss": 0.8716, + "step": 38057 + }, + { + "epoch": 0.9, + "grad_norm": 2.009145057235527, + "learning_rate": 5.553132153311335e-07, + "loss": 0.9193, + "step": 38058 + }, + { + "epoch": 0.9, + "grad_norm": 2.146731443673216, + "learning_rate": 5.55062506424674e-07, + "loss": 0.924, + "step": 38059 + }, + { + "epoch": 0.9, + "grad_norm": 1.917076394291072, + "learning_rate": 5.548118525096313e-07, + "loss": 1.0435, + "step": 38060 + }, + { + "epoch": 0.9, + "grad_norm": 1.9290324841229975, + "learning_rate": 5.545612535874622e-07, + "loss": 0.9683, + "step": 38061 + }, + { + "epoch": 0.9, + "grad_norm": 1.9166536267227774, + "learning_rate": 5.543107096596279e-07, + "loss": 0.9734, + "step": 38062 + }, + { + "epoch": 0.9, + "grad_norm": 1.9698169254423081, + "learning_rate": 5.54060220727588e-07, + "loss": 1.0733, + "step": 38063 + }, + { + "epoch": 0.9, + "grad_norm": 1.887153608200132, + "learning_rate": 5.538097867928016e-07, + "loss": 1.0261, + "step": 38064 + }, + { + "epoch": 0.9, + "grad_norm": 1.9005480987864818, + "learning_rate": 5.53559407856722e-07, + "loss": 0.9926, + "step": 38065 + }, + { + "epoch": 0.9, + "grad_norm": 2.553311236995762, + "learning_rate": 5.533090839208133e-07, + "loss": 0.8586, + "step": 38066 + }, + { + "epoch": 0.9, + "grad_norm": 2.0625247928996844, + "learning_rate": 5.530588149865268e-07, + "loss": 1.0541, + "step": 38067 + }, + { + "epoch": 0.9, + "grad_norm": 2.062259242336262, + "learning_rate": 5.528086010553257e-07, + "loss": 0.9361, + "step": 38068 + }, + { + "epoch": 0.9, + "grad_norm": 5.6602761794449865, + "learning_rate": 5.525584421286612e-07, + "loss": 0.8306, + "step": 38069 + }, + { + "epoch": 0.9, + "grad_norm": 2.201926810516093, + "learning_rate": 5.523083382079952e-07, + "loss": 0.9134, + "step": 38070 + }, + { + "epoch": 0.9, + "grad_norm": 1.7760761581864835, + "learning_rate": 5.520582892947779e-07, + "loss": 0.8001, + "step": 38071 + }, + { + "epoch": 0.9, + "grad_norm": 1.861198336833635, + "learning_rate": 5.518082953904714e-07, + "loss": 1.0378, + "step": 38072 + }, + { + "epoch": 0.9, + "grad_norm": 2.2846934762873157, + "learning_rate": 5.515583564965277e-07, + "loss": 0.9745, + "step": 38073 + }, + { + "epoch": 0.9, + "grad_norm": 2.1056126810637834, + "learning_rate": 5.513084726144013e-07, + "loss": 1.0457, + "step": 38074 + }, + { + "epoch": 0.9, + "grad_norm": 2.3072601144810614, + "learning_rate": 5.510586437455478e-07, + "loss": 0.9701, + "step": 38075 + }, + { + "epoch": 0.9, + "grad_norm": 1.8410396619786071, + "learning_rate": 5.508088698914249e-07, + "loss": 0.9655, + "step": 38076 + }, + { + "epoch": 0.9, + "grad_norm": 2.0317614824940926, + "learning_rate": 5.505591510534836e-07, + "loss": 1.0597, + "step": 38077 + }, + { + "epoch": 0.9, + "grad_norm": 1.955527348431932, + "learning_rate": 5.503094872331771e-07, + "loss": 0.9107, + "step": 38078 + }, + { + "epoch": 0.9, + "grad_norm": 1.8541076245354993, + "learning_rate": 5.5005987843196e-07, + "loss": 0.9575, + "step": 38079 + }, + { + "epoch": 0.9, + "grad_norm": 1.8621337559759163, + "learning_rate": 5.498103246512887e-07, + "loss": 0.9857, + "step": 38080 + }, + { + "epoch": 0.9, + "grad_norm": 1.9277845751657336, + "learning_rate": 5.495608258926133e-07, + "loss": 1.0539, + "step": 38081 + }, + { + "epoch": 0.9, + "grad_norm": 2.0964074134604864, + "learning_rate": 5.493113821573847e-07, + "loss": 0.9397, + "step": 38082 + }, + { + "epoch": 0.9, + "grad_norm": 2.104311773351172, + "learning_rate": 5.490619934470576e-07, + "loss": 1.0332, + "step": 38083 + }, + { + "epoch": 0.9, + "grad_norm": 2.0395104047181913, + "learning_rate": 5.48812659763085e-07, + "loss": 1.0104, + "step": 38084 + }, + { + "epoch": 0.9, + "grad_norm": 2.31032838235756, + "learning_rate": 5.485633811069169e-07, + "loss": 0.839, + "step": 38085 + }, + { + "epoch": 0.9, + "grad_norm": 1.194584315662988, + "learning_rate": 5.483141574800044e-07, + "loss": 0.88, + "step": 38086 + }, + { + "epoch": 0.9, + "grad_norm": 1.9867842967677103, + "learning_rate": 5.480649888838008e-07, + "loss": 1.0898, + "step": 38087 + }, + { + "epoch": 0.9, + "grad_norm": 2.294951038073844, + "learning_rate": 5.478158753197527e-07, + "loss": 1.0335, + "step": 38088 + }, + { + "epoch": 0.9, + "grad_norm": 1.8788217847881414, + "learning_rate": 5.475668167893156e-07, + "loss": 0.9467, + "step": 38089 + }, + { + "epoch": 0.9, + "grad_norm": 2.152598008072232, + "learning_rate": 5.47317813293935e-07, + "loss": 1.1195, + "step": 38090 + }, + { + "epoch": 0.9, + "grad_norm": 1.0687537536525549, + "learning_rate": 5.470688648350642e-07, + "loss": 0.985, + "step": 38091 + }, + { + "epoch": 0.9, + "grad_norm": 1.8812648837654835, + "learning_rate": 5.468199714141509e-07, + "loss": 1.0709, + "step": 38092 + }, + { + "epoch": 0.9, + "grad_norm": 2.1139529470931144, + "learning_rate": 5.465711330326428e-07, + "loss": 1.0026, + "step": 38093 + }, + { + "epoch": 0.9, + "grad_norm": 2.0140720306537956, + "learning_rate": 5.463223496919945e-07, + "loss": 1.0755, + "step": 38094 + }, + { + "epoch": 0.9, + "grad_norm": 1.8890120970089146, + "learning_rate": 5.460736213936479e-07, + "loss": 1.0319, + "step": 38095 + }, + { + "epoch": 0.9, + "grad_norm": 1.0484790256537113, + "learning_rate": 5.458249481390532e-07, + "loss": 0.9202, + "step": 38096 + }, + { + "epoch": 0.9, + "grad_norm": 2.1416937907790556, + "learning_rate": 5.45576329929658e-07, + "loss": 1.0136, + "step": 38097 + }, + { + "epoch": 0.9, + "grad_norm": 1.0542126937860175, + "learning_rate": 5.453277667669143e-07, + "loss": 0.9209, + "step": 38098 + }, + { + "epoch": 0.9, + "grad_norm": 1.929563072619685, + "learning_rate": 5.450792586522624e-07, + "loss": 0.8924, + "step": 38099 + }, + { + "epoch": 0.9, + "grad_norm": 1.9554989230433826, + "learning_rate": 5.448308055871532e-07, + "loss": 0.8898, + "step": 38100 + }, + { + "epoch": 0.9, + "grad_norm": 2.138388164418517, + "learning_rate": 5.445824075730333e-07, + "loss": 1.1344, + "step": 38101 + }, + { + "epoch": 0.9, + "grad_norm": 1.1003339437928596, + "learning_rate": 5.443340646113471e-07, + "loss": 0.9569, + "step": 38102 + }, + { + "epoch": 0.9, + "grad_norm": 1.9288922326675417, + "learning_rate": 5.440857767035434e-07, + "loss": 0.9603, + "step": 38103 + }, + { + "epoch": 0.9, + "grad_norm": 1.7828753779022295, + "learning_rate": 5.438375438510635e-07, + "loss": 0.9713, + "step": 38104 + }, + { + "epoch": 0.9, + "grad_norm": 1.9456976546141902, + "learning_rate": 5.43589366055357e-07, + "loss": 0.895, + "step": 38105 + }, + { + "epoch": 0.9, + "grad_norm": 2.0039513569629506, + "learning_rate": 5.433412433178664e-07, + "loss": 1.0668, + "step": 38106 + }, + { + "epoch": 0.9, + "grad_norm": 1.9044106165222898, + "learning_rate": 5.43093175640037e-07, + "loss": 0.9906, + "step": 38107 + }, + { + "epoch": 0.9, + "grad_norm": 2.074589668531203, + "learning_rate": 5.428451630233144e-07, + "loss": 0.9637, + "step": 38108 + }, + { + "epoch": 0.9, + "grad_norm": 2.2969614287073985, + "learning_rate": 5.425972054691397e-07, + "loss": 1.0564, + "step": 38109 + }, + { + "epoch": 0.9, + "grad_norm": 2.0369413516634127, + "learning_rate": 5.423493029789573e-07, + "loss": 1.0558, + "step": 38110 + }, + { + "epoch": 0.9, + "grad_norm": 2.2154506438664097, + "learning_rate": 5.421014555542137e-07, + "loss": 1.0085, + "step": 38111 + }, + { + "epoch": 0.9, + "grad_norm": 1.0566853050632443, + "learning_rate": 5.418536631963501e-07, + "loss": 0.8662, + "step": 38112 + }, + { + "epoch": 0.9, + "grad_norm": 2.0681845860836834, + "learning_rate": 5.416059259068063e-07, + "loss": 1.0157, + "step": 38113 + }, + { + "epoch": 0.9, + "grad_norm": 2.3685166773833224, + "learning_rate": 5.41358243687028e-07, + "loss": 1.0296, + "step": 38114 + }, + { + "epoch": 0.9, + "grad_norm": 2.173664058121727, + "learning_rate": 5.411106165384583e-07, + "loss": 0.9668, + "step": 38115 + }, + { + "epoch": 0.9, + "grad_norm": 1.9132991156458516, + "learning_rate": 5.408630444625373e-07, + "loss": 0.9593, + "step": 38116 + }, + { + "epoch": 0.9, + "grad_norm": 1.9680571503015307, + "learning_rate": 5.406155274607039e-07, + "loss": 1.0152, + "step": 38117 + }, + { + "epoch": 0.9, + "grad_norm": 2.431349116588975, + "learning_rate": 5.403680655344023e-07, + "loss": 0.9409, + "step": 38118 + }, + { + "epoch": 0.9, + "grad_norm": 1.9565273605711346, + "learning_rate": 5.40120658685075e-07, + "loss": 1.1267, + "step": 38119 + }, + { + "epoch": 0.9, + "grad_norm": 2.053490006484956, + "learning_rate": 5.398733069141593e-07, + "loss": 0.9707, + "step": 38120 + }, + { + "epoch": 0.9, + "grad_norm": 1.0812585299508914, + "learning_rate": 5.396260102230944e-07, + "loss": 0.9578, + "step": 38121 + }, + { + "epoch": 0.9, + "grad_norm": 1.0633502345284005, + "learning_rate": 5.393787686133234e-07, + "loss": 0.9089, + "step": 38122 + }, + { + "epoch": 0.9, + "grad_norm": 2.1869638064706893, + "learning_rate": 5.391315820862841e-07, + "loss": 1.0771, + "step": 38123 + }, + { + "epoch": 0.9, + "grad_norm": 2.159555984688283, + "learning_rate": 5.388844506434166e-07, + "loss": 0.8864, + "step": 38124 + }, + { + "epoch": 0.9, + "grad_norm": 1.9394895096628408, + "learning_rate": 5.386373742861572e-07, + "loss": 1.0088, + "step": 38125 + }, + { + "epoch": 0.9, + "grad_norm": 2.0608818753633593, + "learning_rate": 5.383903530159484e-07, + "loss": 0.8421, + "step": 38126 + }, + { + "epoch": 0.9, + "grad_norm": 2.267463065299572, + "learning_rate": 5.381433868342256e-07, + "loss": 0.9051, + "step": 38127 + }, + { + "epoch": 0.9, + "grad_norm": 1.0688744870979283, + "learning_rate": 5.378964757424287e-07, + "loss": 0.9099, + "step": 38128 + }, + { + "epoch": 0.9, + "grad_norm": 1.7542533857547995, + "learning_rate": 5.376496197419934e-07, + "loss": 1.1091, + "step": 38129 + }, + { + "epoch": 0.9, + "grad_norm": 1.9627006100734856, + "learning_rate": 5.374028188343572e-07, + "loss": 1.0458, + "step": 38130 + }, + { + "epoch": 0.9, + "grad_norm": 2.001225649380921, + "learning_rate": 5.371560730209579e-07, + "loss": 0.9551, + "step": 38131 + }, + { + "epoch": 0.9, + "grad_norm": 1.9800864786303736, + "learning_rate": 5.369093823032323e-07, + "loss": 0.8965, + "step": 38132 + }, + { + "epoch": 0.9, + "grad_norm": 1.893419276029338, + "learning_rate": 5.366627466826169e-07, + "loss": 0.857, + "step": 38133 + }, + { + "epoch": 0.9, + "grad_norm": 2.0229527241848886, + "learning_rate": 5.364161661605449e-07, + "loss": 1.0486, + "step": 38134 + }, + { + "epoch": 0.9, + "grad_norm": 1.9534993619366778, + "learning_rate": 5.361696407384543e-07, + "loss": 1.1143, + "step": 38135 + }, + { + "epoch": 0.9, + "grad_norm": 1.9679330699528659, + "learning_rate": 5.359231704177825e-07, + "loss": 0.944, + "step": 38136 + }, + { + "epoch": 0.9, + "grad_norm": 2.108809720784625, + "learning_rate": 5.356767551999608e-07, + "loss": 0.902, + "step": 38137 + }, + { + "epoch": 0.9, + "grad_norm": 2.0990684453427266, + "learning_rate": 5.354303950864248e-07, + "loss": 1.0215, + "step": 38138 + }, + { + "epoch": 0.9, + "grad_norm": 1.8834247552015404, + "learning_rate": 5.351840900786098e-07, + "loss": 0.8949, + "step": 38139 + }, + { + "epoch": 0.9, + "grad_norm": 2.0186391039057003, + "learning_rate": 5.349378401779504e-07, + "loss": 0.9088, + "step": 38140 + }, + { + "epoch": 0.9, + "grad_norm": 2.1299208116661643, + "learning_rate": 5.346916453858797e-07, + "loss": 1.0021, + "step": 38141 + }, + { + "epoch": 0.9, + "grad_norm": 2.000685634380176, + "learning_rate": 5.3444550570383e-07, + "loss": 0.8483, + "step": 38142 + }, + { + "epoch": 0.9, + "grad_norm": 2.0839703190936394, + "learning_rate": 5.341994211332358e-07, + "loss": 1.1292, + "step": 38143 + }, + { + "epoch": 0.9, + "grad_norm": 1.0649649318149954, + "learning_rate": 5.339533916755291e-07, + "loss": 0.9103, + "step": 38144 + }, + { + "epoch": 0.9, + "grad_norm": 2.066812805118625, + "learning_rate": 5.337074173321433e-07, + "loss": 0.9814, + "step": 38145 + }, + { + "epoch": 0.9, + "grad_norm": 1.156969165444322, + "learning_rate": 5.334614981045083e-07, + "loss": 0.9128, + "step": 38146 + }, + { + "epoch": 0.9, + "grad_norm": 1.8769649119210474, + "learning_rate": 5.332156339940598e-07, + "loss": 0.9061, + "step": 38147 + }, + { + "epoch": 0.9, + "grad_norm": 2.082090432545289, + "learning_rate": 5.329698250022241e-07, + "loss": 1.015, + "step": 38148 + }, + { + "epoch": 0.9, + "grad_norm": 2.121925601971215, + "learning_rate": 5.327240711304383e-07, + "loss": 0.9372, + "step": 38149 + }, + { + "epoch": 0.9, + "grad_norm": 1.8844191747925292, + "learning_rate": 5.324783723801296e-07, + "loss": 1.0564, + "step": 38150 + }, + { + "epoch": 0.9, + "grad_norm": 1.7406903710655017, + "learning_rate": 5.322327287527274e-07, + "loss": 1.0591, + "step": 38151 + }, + { + "epoch": 0.9, + "grad_norm": 2.035428569178146, + "learning_rate": 5.319871402496645e-07, + "loss": 0.9722, + "step": 38152 + }, + { + "epoch": 0.9, + "grad_norm": 2.037831972874751, + "learning_rate": 5.317416068723702e-07, + "loss": 0.9884, + "step": 38153 + }, + { + "epoch": 0.9, + "grad_norm": 1.8419836232551838, + "learning_rate": 5.314961286222752e-07, + "loss": 0.9469, + "step": 38154 + }, + { + "epoch": 0.9, + "grad_norm": 1.9848147840311208, + "learning_rate": 5.312507055008054e-07, + "loss": 0.9618, + "step": 38155 + }, + { + "epoch": 0.9, + "grad_norm": 1.9050579540654555, + "learning_rate": 5.310053375093915e-07, + "loss": 0.9435, + "step": 38156 + }, + { + "epoch": 0.9, + "grad_norm": 2.0273285251479414, + "learning_rate": 5.307600246494638e-07, + "loss": 1.0405, + "step": 38157 + }, + { + "epoch": 0.9, + "grad_norm": 1.8964645927229338, + "learning_rate": 5.305147669224497e-07, + "loss": 1.141, + "step": 38158 + }, + { + "epoch": 0.9, + "grad_norm": 1.7973456936322656, + "learning_rate": 5.302695643297751e-07, + "loss": 1.016, + "step": 38159 + }, + { + "epoch": 0.9, + "grad_norm": 2.1702596781758685, + "learning_rate": 5.300244168728686e-07, + "loss": 0.9558, + "step": 38160 + }, + { + "epoch": 0.9, + "grad_norm": 1.9928056667110379, + "learning_rate": 5.297793245531602e-07, + "loss": 1.0199, + "step": 38161 + }, + { + "epoch": 0.9, + "grad_norm": 1.8545166128768755, + "learning_rate": 5.295342873720743e-07, + "loss": 0.9319, + "step": 38162 + }, + { + "epoch": 0.9, + "grad_norm": 1.9626407633291265, + "learning_rate": 5.292893053310366e-07, + "loss": 0.8839, + "step": 38163 + }, + { + "epoch": 0.9, + "grad_norm": 1.9398738671458324, + "learning_rate": 5.29044378431477e-07, + "loss": 1.0038, + "step": 38164 + }, + { + "epoch": 0.9, + "grad_norm": 1.977389253899474, + "learning_rate": 5.287995066748175e-07, + "loss": 0.9138, + "step": 38165 + }, + { + "epoch": 0.9, + "grad_norm": 2.0628051659614832, + "learning_rate": 5.285546900624872e-07, + "loss": 0.8475, + "step": 38166 + }, + { + "epoch": 0.9, + "grad_norm": 2.0620923727721663, + "learning_rate": 5.283099285959081e-07, + "loss": 1.0913, + "step": 38167 + }, + { + "epoch": 0.9, + "grad_norm": 1.9368403239084435, + "learning_rate": 5.280652222765103e-07, + "loss": 0.9233, + "step": 38168 + }, + { + "epoch": 0.9, + "grad_norm": 2.4307232349913046, + "learning_rate": 5.278205711057127e-07, + "loss": 0.9921, + "step": 38169 + }, + { + "epoch": 0.9, + "grad_norm": 1.8826707714528386, + "learning_rate": 5.275759750849441e-07, + "loss": 0.8987, + "step": 38170 + }, + { + "epoch": 0.9, + "grad_norm": 1.8882901625527466, + "learning_rate": 5.273314342156266e-07, + "loss": 1.0039, + "step": 38171 + }, + { + "epoch": 0.9, + "grad_norm": 2.7064686299516603, + "learning_rate": 5.270869484991858e-07, + "loss": 1.0004, + "step": 38172 + }, + { + "epoch": 0.9, + "grad_norm": 2.2073150668540764, + "learning_rate": 5.268425179370418e-07, + "loss": 0.9998, + "step": 38173 + }, + { + "epoch": 0.9, + "grad_norm": 1.9426938681004287, + "learning_rate": 5.26598142530621e-07, + "loss": 0.9821, + "step": 38174 + }, + { + "epoch": 0.9, + "grad_norm": 1.0095111550029843, + "learning_rate": 5.263538222813457e-07, + "loss": 1.0108, + "step": 38175 + }, + { + "epoch": 0.9, + "grad_norm": 2.0493032907434836, + "learning_rate": 5.261095571906361e-07, + "loss": 0.9418, + "step": 38176 + }, + { + "epoch": 0.9, + "grad_norm": 1.1035027318597932, + "learning_rate": 5.258653472599162e-07, + "loss": 0.8838, + "step": 38177 + }, + { + "epoch": 0.9, + "grad_norm": 1.8173394487548349, + "learning_rate": 5.256211924906074e-07, + "loss": 1.0604, + "step": 38178 + }, + { + "epoch": 0.9, + "grad_norm": 2.046755293765534, + "learning_rate": 5.253770928841351e-07, + "loss": 0.836, + "step": 38179 + }, + { + "epoch": 0.9, + "grad_norm": 1.993489209016648, + "learning_rate": 5.251330484419137e-07, + "loss": 1.1008, + "step": 38180 + }, + { + "epoch": 0.9, + "grad_norm": 2.3689712967470893, + "learning_rate": 5.248890591653666e-07, + "loss": 0.9684, + "step": 38181 + }, + { + "epoch": 0.9, + "grad_norm": 1.8140687972403262, + "learning_rate": 5.246451250559182e-07, + "loss": 0.8947, + "step": 38182 + }, + { + "epoch": 0.9, + "grad_norm": 2.1780518110137854, + "learning_rate": 5.24401246114985e-07, + "loss": 1.1677, + "step": 38183 + }, + { + "epoch": 0.9, + "grad_norm": 2.0706720044536184, + "learning_rate": 5.24157422343986e-07, + "loss": 1.0083, + "step": 38184 + }, + { + "epoch": 0.9, + "grad_norm": 2.35974192460211, + "learning_rate": 5.239136537443423e-07, + "loss": 0.9524, + "step": 38185 + }, + { + "epoch": 0.9, + "grad_norm": 1.8956269776317556, + "learning_rate": 5.23669940317475e-07, + "loss": 0.9594, + "step": 38186 + }, + { + "epoch": 0.9, + "grad_norm": 1.8571151906155081, + "learning_rate": 5.234262820648006e-07, + "loss": 1.0606, + "step": 38187 + }, + { + "epoch": 0.9, + "grad_norm": 1.9141276063234434, + "learning_rate": 5.231826789877392e-07, + "loss": 0.9391, + "step": 38188 + }, + { + "epoch": 0.9, + "grad_norm": 1.8832316965151494, + "learning_rate": 5.229391310877085e-07, + "loss": 0.876, + "step": 38189 + }, + { + "epoch": 0.9, + "grad_norm": 2.2470125540538866, + "learning_rate": 5.226956383661263e-07, + "loss": 1.0157, + "step": 38190 + }, + { + "epoch": 0.9, + "grad_norm": 5.018684347628203, + "learning_rate": 5.224522008244092e-07, + "loss": 0.7502, + "step": 38191 + }, + { + "epoch": 0.9, + "grad_norm": 2.063872060118398, + "learning_rate": 5.222088184639785e-07, + "loss": 0.8388, + "step": 38192 + }, + { + "epoch": 0.9, + "grad_norm": 1.8898001308077133, + "learning_rate": 5.219654912862482e-07, + "loss": 1.0006, + "step": 38193 + }, + { + "epoch": 0.9, + "grad_norm": 2.2895322691343325, + "learning_rate": 5.217222192926342e-07, + "loss": 0.9041, + "step": 38194 + }, + { + "epoch": 0.9, + "grad_norm": 1.7897532875423086, + "learning_rate": 5.214790024845539e-07, + "loss": 0.97, + "step": 38195 + }, + { + "epoch": 0.9, + "grad_norm": 1.8695868780428948, + "learning_rate": 5.212358408634266e-07, + "loss": 0.9422, + "step": 38196 + }, + { + "epoch": 0.9, + "grad_norm": 2.256738444582558, + "learning_rate": 5.20992734430662e-07, + "loss": 0.9716, + "step": 38197 + }, + { + "epoch": 0.9, + "grad_norm": 2.194824014754917, + "learning_rate": 5.207496831876791e-07, + "loss": 0.9853, + "step": 38198 + }, + { + "epoch": 0.9, + "grad_norm": 1.9142817434334556, + "learning_rate": 5.205066871358932e-07, + "loss": 0.8283, + "step": 38199 + }, + { + "epoch": 0.9, + "grad_norm": 2.425150018990864, + "learning_rate": 5.20263746276719e-07, + "loss": 0.9528, + "step": 38200 + }, + { + "epoch": 0.9, + "grad_norm": 1.9344687738494681, + "learning_rate": 5.200208606115709e-07, + "loss": 1.0153, + "step": 38201 + }, + { + "epoch": 0.9, + "grad_norm": 2.129288851410216, + "learning_rate": 5.197780301418609e-07, + "loss": 1.0382, + "step": 38202 + }, + { + "epoch": 0.9, + "grad_norm": 2.4606794309027915, + "learning_rate": 5.195352548690069e-07, + "loss": 0.9627, + "step": 38203 + }, + { + "epoch": 0.9, + "grad_norm": 2.1899498330680296, + "learning_rate": 5.192925347944178e-07, + "loss": 0.9963, + "step": 38204 + }, + { + "epoch": 0.9, + "grad_norm": 2.097563248265136, + "learning_rate": 5.190498699195112e-07, + "loss": 1.0272, + "step": 38205 + }, + { + "epoch": 0.9, + "grad_norm": 2.3060130497858315, + "learning_rate": 5.188072602456962e-07, + "loss": 0.8434, + "step": 38206 + }, + { + "epoch": 0.9, + "grad_norm": 1.9474437631185093, + "learning_rate": 5.185647057743881e-07, + "loss": 1.1242, + "step": 38207 + }, + { + "epoch": 0.9, + "grad_norm": 2.6469263071141005, + "learning_rate": 5.18322206506997e-07, + "loss": 0.8442, + "step": 38208 + }, + { + "epoch": 0.9, + "grad_norm": 1.9079147314210199, + "learning_rate": 5.180797624449363e-07, + "loss": 1.0469, + "step": 38209 + }, + { + "epoch": 0.9, + "grad_norm": 1.8322235739266461, + "learning_rate": 5.17837373589618e-07, + "loss": 0.9628, + "step": 38210 + }, + { + "epoch": 0.9, + "grad_norm": 1.8735567979703505, + "learning_rate": 5.175950399424512e-07, + "loss": 1.031, + "step": 38211 + }, + { + "epoch": 0.9, + "grad_norm": 2.015630881964895, + "learning_rate": 5.173527615048491e-07, + "loss": 0.9528, + "step": 38212 + }, + { + "epoch": 0.9, + "grad_norm": 1.0982946440205545, + "learning_rate": 5.171105382782215e-07, + "loss": 0.9603, + "step": 38213 + }, + { + "epoch": 0.9, + "grad_norm": 2.0042130956181916, + "learning_rate": 5.168683702639788e-07, + "loss": 0.8936, + "step": 38214 + }, + { + "epoch": 0.9, + "grad_norm": 2.0918926115593948, + "learning_rate": 5.166262574635295e-07, + "loss": 0.9757, + "step": 38215 + }, + { + "epoch": 0.9, + "grad_norm": 1.9015104085467494, + "learning_rate": 5.163841998782837e-07, + "loss": 1.0005, + "step": 38216 + }, + { + "epoch": 0.9, + "grad_norm": 2.0686445306425925, + "learning_rate": 5.161421975096537e-07, + "loss": 0.9692, + "step": 38217 + }, + { + "epoch": 0.9, + "grad_norm": 2.3036194772461953, + "learning_rate": 5.159002503590472e-07, + "loss": 0.9531, + "step": 38218 + }, + { + "epoch": 0.9, + "grad_norm": 2.2819118954624398, + "learning_rate": 5.156583584278696e-07, + "loss": 1.0528, + "step": 38219 + }, + { + "epoch": 0.9, + "grad_norm": 2.024488674984184, + "learning_rate": 5.154165217175311e-07, + "loss": 0.9388, + "step": 38220 + }, + { + "epoch": 0.9, + "grad_norm": 2.098183951785445, + "learning_rate": 5.151747402294427e-07, + "loss": 0.846, + "step": 38221 + }, + { + "epoch": 0.9, + "grad_norm": 2.1949715227548103, + "learning_rate": 5.1493301396501e-07, + "loss": 0.9097, + "step": 38222 + }, + { + "epoch": 0.9, + "grad_norm": 1.9506105650407717, + "learning_rate": 5.146913429256384e-07, + "loss": 1.0117, + "step": 38223 + }, + { + "epoch": 0.9, + "grad_norm": 2.2263273581352134, + "learning_rate": 5.14449727112738e-07, + "loss": 1.0365, + "step": 38224 + }, + { + "epoch": 0.9, + "grad_norm": 2.1311388056526566, + "learning_rate": 5.142081665277132e-07, + "loss": 0.8351, + "step": 38225 + }, + { + "epoch": 0.9, + "grad_norm": 1.9158088069861572, + "learning_rate": 5.13966661171974e-07, + "loss": 0.9719, + "step": 38226 + }, + { + "epoch": 0.9, + "grad_norm": 1.7910390328975294, + "learning_rate": 5.137252110469215e-07, + "loss": 1.0309, + "step": 38227 + }, + { + "epoch": 0.9, + "grad_norm": 2.005056307988145, + "learning_rate": 5.134838161539657e-07, + "loss": 1.0222, + "step": 38228 + }, + { + "epoch": 0.9, + "grad_norm": 1.9153688370918147, + "learning_rate": 5.132424764945088e-07, + "loss": 0.8767, + "step": 38229 + }, + { + "epoch": 0.9, + "grad_norm": 2.1477699394661394, + "learning_rate": 5.130011920699596e-07, + "loss": 0.8529, + "step": 38230 + }, + { + "epoch": 0.9, + "grad_norm": 2.331095765650428, + "learning_rate": 5.127599628817215e-07, + "loss": 1.0074, + "step": 38231 + }, + { + "epoch": 0.9, + "grad_norm": 1.9494511891194806, + "learning_rate": 5.125187889311966e-07, + "loss": 0.9124, + "step": 38232 + }, + { + "epoch": 0.9, + "grad_norm": 1.947745939024732, + "learning_rate": 5.122776702197907e-07, + "loss": 1.067, + "step": 38233 + }, + { + "epoch": 0.9, + "grad_norm": 2.1132237431851197, + "learning_rate": 5.120366067489091e-07, + "loss": 0.8719, + "step": 38234 + }, + { + "epoch": 0.9, + "grad_norm": 2.3332198801298034, + "learning_rate": 5.117955985199552e-07, + "loss": 1.0718, + "step": 38235 + }, + { + "epoch": 0.9, + "grad_norm": 1.8633879097714572, + "learning_rate": 5.115546455343289e-07, + "loss": 0.9033, + "step": 38236 + }, + { + "epoch": 0.9, + "grad_norm": 1.9273363250058635, + "learning_rate": 5.11313747793436e-07, + "loss": 0.9308, + "step": 38237 + }, + { + "epoch": 0.9, + "grad_norm": 2.1745379350291127, + "learning_rate": 5.110729052986796e-07, + "loss": 1.0258, + "step": 38238 + }, + { + "epoch": 0.9, + "grad_norm": 1.9185520373966594, + "learning_rate": 5.108321180514608e-07, + "loss": 0.8946, + "step": 38239 + }, + { + "epoch": 0.9, + "grad_norm": 1.9238809148085867, + "learning_rate": 5.105913860531808e-07, + "loss": 0.9373, + "step": 38240 + }, + { + "epoch": 0.9, + "grad_norm": 2.0189882414018787, + "learning_rate": 5.103507093052418e-07, + "loss": 0.9349, + "step": 38241 + }, + { + "epoch": 0.9, + "grad_norm": 2.1716694560917666, + "learning_rate": 5.101100878090471e-07, + "loss": 0.916, + "step": 38242 + }, + { + "epoch": 0.9, + "grad_norm": 1.953372736895235, + "learning_rate": 5.098695215659954e-07, + "loss": 0.923, + "step": 38243 + }, + { + "epoch": 0.9, + "grad_norm": 1.8403350165088113, + "learning_rate": 5.096290105774859e-07, + "loss": 0.9965, + "step": 38244 + }, + { + "epoch": 0.9, + "grad_norm": 2.0758042524782185, + "learning_rate": 5.093885548449229e-07, + "loss": 1.0522, + "step": 38245 + }, + { + "epoch": 0.9, + "grad_norm": 1.8657590761575993, + "learning_rate": 5.091481543697042e-07, + "loss": 1.0348, + "step": 38246 + }, + { + "epoch": 0.9, + "grad_norm": 2.174354689843792, + "learning_rate": 5.089078091532296e-07, + "loss": 1.075, + "step": 38247 + }, + { + "epoch": 0.9, + "grad_norm": 1.8374519133641707, + "learning_rate": 5.086675191968971e-07, + "loss": 1.0402, + "step": 38248 + }, + { + "epoch": 0.9, + "grad_norm": 2.1537208003050683, + "learning_rate": 5.084272845021088e-07, + "loss": 0.9367, + "step": 38249 + }, + { + "epoch": 0.9, + "grad_norm": 2.355996662755288, + "learning_rate": 5.081871050702614e-07, + "loss": 0.9262, + "step": 38250 + }, + { + "epoch": 0.9, + "grad_norm": 2.359905736937148, + "learning_rate": 5.079469809027537e-07, + "loss": 0.9991, + "step": 38251 + }, + { + "epoch": 0.9, + "grad_norm": 1.9755159756703364, + "learning_rate": 5.077069120009848e-07, + "loss": 1.0059, + "step": 38252 + }, + { + "epoch": 0.9, + "grad_norm": 1.9828027922419533, + "learning_rate": 5.0746689836635e-07, + "loss": 0.9906, + "step": 38253 + }, + { + "epoch": 0.9, + "grad_norm": 2.2231389429376995, + "learning_rate": 5.072269400002483e-07, + "loss": 0.9162, + "step": 38254 + }, + { + "epoch": 0.9, + "grad_norm": 2.15977361077005, + "learning_rate": 5.069870369040775e-07, + "loss": 1.0777, + "step": 38255 + }, + { + "epoch": 0.9, + "grad_norm": 1.9944427793850978, + "learning_rate": 5.067471890792341e-07, + "loss": 0.8832, + "step": 38256 + }, + { + "epoch": 0.9, + "grad_norm": 1.9873484231789835, + "learning_rate": 5.065073965271128e-07, + "loss": 1.0212, + "step": 38257 + }, + { + "epoch": 0.9, + "grad_norm": 1.8733120853949887, + "learning_rate": 5.062676592491101e-07, + "loss": 1.0183, + "step": 38258 + }, + { + "epoch": 0.9, + "grad_norm": 2.3387031242789447, + "learning_rate": 5.060279772466248e-07, + "loss": 1.0395, + "step": 38259 + }, + { + "epoch": 0.9, + "grad_norm": 2.0062395912356994, + "learning_rate": 5.057883505210504e-07, + "loss": 0.9657, + "step": 38260 + }, + { + "epoch": 0.9, + "grad_norm": 2.0886557451070886, + "learning_rate": 5.055487790737812e-07, + "loss": 0.9289, + "step": 38261 + }, + { + "epoch": 0.9, + "grad_norm": 1.9420877536306482, + "learning_rate": 5.053092629062117e-07, + "loss": 1.0347, + "step": 38262 + }, + { + "epoch": 0.9, + "grad_norm": 1.8357760704625998, + "learning_rate": 5.050698020197398e-07, + "loss": 1.0706, + "step": 38263 + }, + { + "epoch": 0.9, + "grad_norm": 1.9452119704586823, + "learning_rate": 5.048303964157575e-07, + "loss": 1.0022, + "step": 38264 + }, + { + "epoch": 0.9, + "grad_norm": 2.267889857016413, + "learning_rate": 5.045910460956572e-07, + "loss": 0.8967, + "step": 38265 + }, + { + "epoch": 0.9, + "grad_norm": 2.0620097738564396, + "learning_rate": 5.043517510608354e-07, + "loss": 0.8168, + "step": 38266 + }, + { + "epoch": 0.9, + "grad_norm": 2.048052432409943, + "learning_rate": 5.041125113126821e-07, + "loss": 0.8513, + "step": 38267 + }, + { + "epoch": 0.9, + "grad_norm": 1.1807426934448293, + "learning_rate": 5.038733268525941e-07, + "loss": 0.9171, + "step": 38268 + }, + { + "epoch": 0.9, + "grad_norm": 5.002498868451191, + "learning_rate": 5.036341976819615e-07, + "loss": 1.1462, + "step": 38269 + }, + { + "epoch": 0.9, + "grad_norm": 1.912933698867477, + "learning_rate": 5.033951238021773e-07, + "loss": 0.9785, + "step": 38270 + }, + { + "epoch": 0.9, + "grad_norm": 2.1294108011647164, + "learning_rate": 5.031561052146316e-07, + "loss": 0.8871, + "step": 38271 + }, + { + "epoch": 0.9, + "grad_norm": 2.0516376123120788, + "learning_rate": 5.029171419207202e-07, + "loss": 1.0092, + "step": 38272 + }, + { + "epoch": 0.9, + "grad_norm": 1.9282395553379639, + "learning_rate": 5.026782339218306e-07, + "loss": 1.0939, + "step": 38273 + }, + { + "epoch": 0.9, + "grad_norm": 1.8960710794846483, + "learning_rate": 5.024393812193551e-07, + "loss": 0.9851, + "step": 38274 + }, + { + "epoch": 0.9, + "grad_norm": 1.9572637463542428, + "learning_rate": 5.022005838146837e-07, + "loss": 0.8811, + "step": 38275 + }, + { + "epoch": 0.9, + "grad_norm": 2.1385437110488748, + "learning_rate": 5.019618417092075e-07, + "loss": 1.0524, + "step": 38276 + }, + { + "epoch": 0.9, + "grad_norm": 1.8171883240199331, + "learning_rate": 5.017231549043189e-07, + "loss": 0.991, + "step": 38277 + }, + { + "epoch": 0.9, + "grad_norm": 1.156160712620647, + "learning_rate": 5.014845234014032e-07, + "loss": 0.9681, + "step": 38278 + }, + { + "epoch": 0.9, + "grad_norm": 2.0884383667853164, + "learning_rate": 5.012459472018516e-07, + "loss": 1.0329, + "step": 38279 + }, + { + "epoch": 0.9, + "grad_norm": 2.1051741358083498, + "learning_rate": 5.010074263070541e-07, + "loss": 0.9459, + "step": 38280 + }, + { + "epoch": 0.9, + "grad_norm": 2.241737156220624, + "learning_rate": 5.007689607183997e-07, + "loss": 1.0846, + "step": 38281 + }, + { + "epoch": 0.9, + "grad_norm": 2.0559595623923963, + "learning_rate": 5.005305504372738e-07, + "loss": 0.9344, + "step": 38282 + }, + { + "epoch": 0.9, + "grad_norm": 1.0059418154931223, + "learning_rate": 5.002921954650664e-07, + "loss": 0.9386, + "step": 38283 + }, + { + "epoch": 0.9, + "grad_norm": 1.02738253795355, + "learning_rate": 5.000538958031664e-07, + "loss": 0.9431, + "step": 38284 + }, + { + "epoch": 0.9, + "grad_norm": 2.333296598366756, + "learning_rate": 4.998156514529595e-07, + "loss": 0.8413, + "step": 38285 + }, + { + "epoch": 0.9, + "grad_norm": 1.873355987274247, + "learning_rate": 4.995774624158356e-07, + "loss": 0.9171, + "step": 38286 + }, + { + "epoch": 0.9, + "grad_norm": 2.1311299733405167, + "learning_rate": 4.993393286931781e-07, + "loss": 1.0113, + "step": 38287 + }, + { + "epoch": 0.9, + "grad_norm": 1.8194305426806607, + "learning_rate": 4.991012502863745e-07, + "loss": 0.9266, + "step": 38288 + }, + { + "epoch": 0.9, + "grad_norm": 2.269973291318774, + "learning_rate": 4.988632271968108e-07, + "loss": 0.9207, + "step": 38289 + }, + { + "epoch": 0.9, + "grad_norm": 2.0070384200709763, + "learning_rate": 4.986252594258756e-07, + "loss": 0.8736, + "step": 38290 + }, + { + "epoch": 0.9, + "grad_norm": 2.0333839143971444, + "learning_rate": 4.983873469749511e-07, + "loss": 0.9184, + "step": 38291 + }, + { + "epoch": 0.9, + "grad_norm": 2.153717712248296, + "learning_rate": 4.981494898454232e-07, + "loss": 0.9103, + "step": 38292 + }, + { + "epoch": 0.9, + "grad_norm": 2.11130164515346, + "learning_rate": 4.97911688038677e-07, + "loss": 1.0345, + "step": 38293 + }, + { + "epoch": 0.9, + "grad_norm": 1.9794860131313756, + "learning_rate": 4.976739415560982e-07, + "loss": 0.8799, + "step": 38294 + }, + { + "epoch": 0.9, + "grad_norm": 1.0651080939230027, + "learning_rate": 4.974362503990704e-07, + "loss": 0.8859, + "step": 38295 + }, + { + "epoch": 0.9, + "grad_norm": 1.9795838193311073, + "learning_rate": 4.971986145689756e-07, + "loss": 1.0503, + "step": 38296 + }, + { + "epoch": 0.9, + "grad_norm": 2.0870878646382196, + "learning_rate": 4.969610340671993e-07, + "loss": 0.9842, + "step": 38297 + }, + { + "epoch": 0.9, + "grad_norm": 1.962113157723316, + "learning_rate": 4.967235088951261e-07, + "loss": 1.0535, + "step": 38298 + }, + { + "epoch": 0.9, + "grad_norm": 2.4698683959874614, + "learning_rate": 4.96486039054136e-07, + "loss": 0.9936, + "step": 38299 + }, + { + "epoch": 0.9, + "grad_norm": 1.9754476845356657, + "learning_rate": 4.962486245456133e-07, + "loss": 1.0005, + "step": 38300 + }, + { + "epoch": 0.9, + "grad_norm": 2.223464667224816, + "learning_rate": 4.960112653709392e-07, + "loss": 0.8356, + "step": 38301 + }, + { + "epoch": 0.9, + "grad_norm": 1.0638669302862844, + "learning_rate": 4.957739615314972e-07, + "loss": 0.9318, + "step": 38302 + }, + { + "epoch": 0.9, + "grad_norm": 1.8703306458514353, + "learning_rate": 4.955367130286692e-07, + "loss": 0.7973, + "step": 38303 + }, + { + "epoch": 0.9, + "grad_norm": 1.118046958280068, + "learning_rate": 4.952995198638333e-07, + "loss": 0.9256, + "step": 38304 + }, + { + "epoch": 0.9, + "grad_norm": 1.8872219639047947, + "learning_rate": 4.950623820383749e-07, + "loss": 0.9226, + "step": 38305 + }, + { + "epoch": 0.9, + "grad_norm": 1.8794351781739524, + "learning_rate": 4.948252995536707e-07, + "loss": 0.9381, + "step": 38306 + }, + { + "epoch": 0.9, + "grad_norm": 1.7673905031989832, + "learning_rate": 4.94588272411104e-07, + "loss": 0.8999, + "step": 38307 + }, + { + "epoch": 0.9, + "grad_norm": 1.8956555287340169, + "learning_rate": 4.943513006120526e-07, + "loss": 0.7939, + "step": 38308 + }, + { + "epoch": 0.9, + "grad_norm": 2.075065994384682, + "learning_rate": 4.941143841578988e-07, + "loss": 0.9359, + "step": 38309 + }, + { + "epoch": 0.9, + "grad_norm": 2.0343055862404467, + "learning_rate": 4.938775230500192e-07, + "loss": 0.8005, + "step": 38310 + }, + { + "epoch": 0.9, + "grad_norm": 2.0217691371822846, + "learning_rate": 4.936407172897961e-07, + "loss": 1.012, + "step": 38311 + }, + { + "epoch": 0.9, + "grad_norm": 1.1079412889288438, + "learning_rate": 4.934039668786061e-07, + "loss": 0.957, + "step": 38312 + }, + { + "epoch": 0.9, + "grad_norm": 2.0468822335909835, + "learning_rate": 4.931672718178259e-07, + "loss": 0.9548, + "step": 38313 + }, + { + "epoch": 0.9, + "grad_norm": 1.0873348399699791, + "learning_rate": 4.929306321088368e-07, + "loss": 0.9447, + "step": 38314 + }, + { + "epoch": 0.9, + "grad_norm": 1.9441376091104192, + "learning_rate": 4.926940477530151e-07, + "loss": 0.8678, + "step": 38315 + }, + { + "epoch": 0.9, + "grad_norm": 2.0227069373574054, + "learning_rate": 4.924575187517399e-07, + "loss": 0.9082, + "step": 38316 + }, + { + "epoch": 0.9, + "grad_norm": 1.8631598536888752, + "learning_rate": 4.922210451063858e-07, + "loss": 0.8508, + "step": 38317 + }, + { + "epoch": 0.9, + "grad_norm": 2.0713959992401247, + "learning_rate": 4.919846268183303e-07, + "loss": 1.0017, + "step": 38318 + }, + { + "epoch": 0.9, + "grad_norm": 1.1131369271340854, + "learning_rate": 4.917482638889525e-07, + "loss": 0.9225, + "step": 38319 + }, + { + "epoch": 0.9, + "grad_norm": 1.0162793745065257, + "learning_rate": 4.915119563196258e-07, + "loss": 0.8845, + "step": 38320 + }, + { + "epoch": 0.9, + "grad_norm": 2.006622054311327, + "learning_rate": 4.912757041117266e-07, + "loss": 0.9944, + "step": 38321 + }, + { + "epoch": 0.9, + "grad_norm": 1.7729411530790558, + "learning_rate": 4.910395072666297e-07, + "loss": 0.9125, + "step": 38322 + }, + { + "epoch": 0.9, + "grad_norm": 2.1237298348724467, + "learning_rate": 4.908033657857125e-07, + "loss": 0.9642, + "step": 38323 + }, + { + "epoch": 0.9, + "grad_norm": 1.8033845703809293, + "learning_rate": 4.905672796703498e-07, + "loss": 0.9833, + "step": 38324 + }, + { + "epoch": 0.9, + "grad_norm": 2.2662668483321733, + "learning_rate": 4.903312489219125e-07, + "loss": 0.8467, + "step": 38325 + }, + { + "epoch": 0.9, + "grad_norm": 2.147302850622136, + "learning_rate": 4.900952735417786e-07, + "loss": 0.9149, + "step": 38326 + }, + { + "epoch": 0.9, + "grad_norm": 1.7871114162136366, + "learning_rate": 4.898593535313201e-07, + "loss": 0.9239, + "step": 38327 + }, + { + "epoch": 0.9, + "grad_norm": 2.090028068066002, + "learning_rate": 4.896234888919127e-07, + "loss": 0.7721, + "step": 38328 + }, + { + "epoch": 0.9, + "grad_norm": 1.920338700757291, + "learning_rate": 4.893876796249264e-07, + "loss": 1.0236, + "step": 38329 + }, + { + "epoch": 0.9, + "grad_norm": 1.902161687773858, + "learning_rate": 4.891519257317379e-07, + "loss": 1.1209, + "step": 38330 + }, + { + "epoch": 0.9, + "grad_norm": 2.101151240280382, + "learning_rate": 4.889162272137171e-07, + "loss": 0.9172, + "step": 38331 + }, + { + "epoch": 0.9, + "grad_norm": 1.9823413957035514, + "learning_rate": 4.886805840722375e-07, + "loss": 0.883, + "step": 38332 + }, + { + "epoch": 0.9, + "grad_norm": 2.0367453589697635, + "learning_rate": 4.884449963086712e-07, + "loss": 1.0451, + "step": 38333 + }, + { + "epoch": 0.9, + "grad_norm": 1.8647074083579167, + "learning_rate": 4.882094639243884e-07, + "loss": 0.9594, + "step": 38334 + }, + { + "epoch": 0.9, + "grad_norm": 2.3229247997741433, + "learning_rate": 4.879739869207611e-07, + "loss": 1.0346, + "step": 38335 + }, + { + "epoch": 0.9, + "grad_norm": 1.9194752735473435, + "learning_rate": 4.877385652991618e-07, + "loss": 1.0022, + "step": 38336 + }, + { + "epoch": 0.9, + "grad_norm": 2.0476061151349567, + "learning_rate": 4.875031990609602e-07, + "loss": 0.8963, + "step": 38337 + }, + { + "epoch": 0.9, + "grad_norm": 1.9632150684361744, + "learning_rate": 4.872678882075255e-07, + "loss": 1.0665, + "step": 38338 + }, + { + "epoch": 0.9, + "grad_norm": 1.0922187549238611, + "learning_rate": 4.870326327402297e-07, + "loss": 0.9651, + "step": 38339 + }, + { + "epoch": 0.9, + "grad_norm": 1.1183098524118371, + "learning_rate": 4.867974326604419e-07, + "loss": 0.9193, + "step": 38340 + }, + { + "epoch": 0.9, + "grad_norm": 1.8665846982347585, + "learning_rate": 4.865622879695309e-07, + "loss": 0.9497, + "step": 38341 + }, + { + "epoch": 0.9, + "grad_norm": 2.324801644793813, + "learning_rate": 4.863271986688656e-07, + "loss": 0.9763, + "step": 38342 + }, + { + "epoch": 0.9, + "grad_norm": 2.1220848287734553, + "learning_rate": 4.860921647598149e-07, + "loss": 0.9408, + "step": 38343 + }, + { + "epoch": 0.9, + "grad_norm": 1.9560246042717861, + "learning_rate": 4.8585718624375e-07, + "loss": 1.0087, + "step": 38344 + }, + { + "epoch": 0.9, + "grad_norm": 2.0932005530075206, + "learning_rate": 4.856222631220354e-07, + "loss": 0.9121, + "step": 38345 + }, + { + "epoch": 0.9, + "grad_norm": 1.9115064442870142, + "learning_rate": 4.853873953960397e-07, + "loss": 0.9411, + "step": 38346 + }, + { + "epoch": 0.9, + "grad_norm": 1.8738593526134362, + "learning_rate": 4.851525830671322e-07, + "loss": 0.9681, + "step": 38347 + }, + { + "epoch": 0.9, + "grad_norm": 2.993335595124203, + "learning_rate": 4.849178261366771e-07, + "loss": 1.0613, + "step": 38348 + }, + { + "epoch": 0.9, + "grad_norm": 2.049463962894287, + "learning_rate": 4.846831246060435e-07, + "loss": 0.864, + "step": 38349 + }, + { + "epoch": 0.9, + "grad_norm": 2.2169066780052975, + "learning_rate": 4.844484784765968e-07, + "loss": 1.0052, + "step": 38350 + }, + { + "epoch": 0.9, + "grad_norm": 1.0587294962635374, + "learning_rate": 4.842138877497049e-07, + "loss": 0.8718, + "step": 38351 + }, + { + "epoch": 0.9, + "grad_norm": 2.6253625534762195, + "learning_rate": 4.839793524267311e-07, + "loss": 0.9742, + "step": 38352 + }, + { + "epoch": 0.9, + "grad_norm": 2.39692121090352, + "learning_rate": 4.837448725090433e-07, + "loss": 1.036, + "step": 38353 + }, + { + "epoch": 0.9, + "grad_norm": 2.2499394772919694, + "learning_rate": 4.835104479980058e-07, + "loss": 1.0009, + "step": 38354 + }, + { + "epoch": 0.9, + "grad_norm": 2.4159826805124434, + "learning_rate": 4.832760788949809e-07, + "loss": 1.0055, + "step": 38355 + }, + { + "epoch": 0.9, + "grad_norm": 1.9356875973529524, + "learning_rate": 4.830417652013364e-07, + "loss": 1.0582, + "step": 38356 + }, + { + "epoch": 0.9, + "grad_norm": 2.4393319944738696, + "learning_rate": 4.828075069184379e-07, + "loss": 0.8205, + "step": 38357 + }, + { + "epoch": 0.9, + "grad_norm": 1.9287218385702214, + "learning_rate": 4.825733040476465e-07, + "loss": 0.8874, + "step": 38358 + }, + { + "epoch": 0.9, + "grad_norm": 2.569383517016526, + "learning_rate": 4.823391565903257e-07, + "loss": 1.0475, + "step": 38359 + }, + { + "epoch": 0.9, + "grad_norm": 1.9854857533809356, + "learning_rate": 4.821050645478386e-07, + "loss": 0.9637, + "step": 38360 + }, + { + "epoch": 0.9, + "grad_norm": 1.8425815213976595, + "learning_rate": 4.81871027921551e-07, + "loss": 0.9496, + "step": 38361 + }, + { + "epoch": 0.9, + "grad_norm": 2.5842675850734462, + "learning_rate": 4.81637046712824e-07, + "loss": 1.0928, + "step": 38362 + }, + { + "epoch": 0.9, + "grad_norm": 2.1149608733759124, + "learning_rate": 4.814031209230175e-07, + "loss": 0.925, + "step": 38363 + }, + { + "epoch": 0.9, + "grad_norm": 1.9536741813205085, + "learning_rate": 4.811692505534959e-07, + "loss": 1.0163, + "step": 38364 + }, + { + "epoch": 0.9, + "grad_norm": 1.7659651490087593, + "learning_rate": 4.809354356056217e-07, + "loss": 1.0299, + "step": 38365 + }, + { + "epoch": 0.9, + "grad_norm": 3.1551618751873063, + "learning_rate": 4.807016760807559e-07, + "loss": 0.9748, + "step": 38366 + }, + { + "epoch": 0.9, + "grad_norm": 2.009653914797432, + "learning_rate": 4.804679719802563e-07, + "loss": 0.9865, + "step": 38367 + }, + { + "epoch": 0.9, + "grad_norm": 2.1471550186014103, + "learning_rate": 4.802343233054884e-07, + "loss": 0.9093, + "step": 38368 + }, + { + "epoch": 0.9, + "grad_norm": 1.9430816254223702, + "learning_rate": 4.80000730057808e-07, + "loss": 1.1184, + "step": 38369 + }, + { + "epoch": 0.9, + "grad_norm": 1.0448675091919073, + "learning_rate": 4.797671922385783e-07, + "loss": 0.9234, + "step": 38370 + }, + { + "epoch": 0.9, + "grad_norm": 2.2955515199138024, + "learning_rate": 4.795337098491582e-07, + "loss": 0.9297, + "step": 38371 + }, + { + "epoch": 0.9, + "grad_norm": 2.0174557654749377, + "learning_rate": 4.793002828909066e-07, + "loss": 1.0162, + "step": 38372 + }, + { + "epoch": 0.9, + "grad_norm": 2.330617446148708, + "learning_rate": 4.790669113651825e-07, + "loss": 0.9917, + "step": 38373 + }, + { + "epoch": 0.9, + "grad_norm": 1.7367187984130208, + "learning_rate": 4.788335952733459e-07, + "loss": 0.9075, + "step": 38374 + }, + { + "epoch": 0.9, + "grad_norm": 1.8774977630917076, + "learning_rate": 4.786003346167556e-07, + "loss": 0.7818, + "step": 38375 + }, + { + "epoch": 0.9, + "grad_norm": 1.9843225512087508, + "learning_rate": 4.783671293967673e-07, + "loss": 1.0643, + "step": 38376 + }, + { + "epoch": 0.9, + "grad_norm": 2.0669605123837234, + "learning_rate": 4.781339796147389e-07, + "loss": 0.8519, + "step": 38377 + }, + { + "epoch": 0.9, + "grad_norm": 2.39973818226263, + "learning_rate": 4.77900885272029e-07, + "loss": 0.8736, + "step": 38378 + }, + { + "epoch": 0.9, + "grad_norm": 1.9103594231402368, + "learning_rate": 4.77667846369998e-07, + "loss": 0.9564, + "step": 38379 + }, + { + "epoch": 0.9, + "grad_norm": 2.008532018076384, + "learning_rate": 4.774348629099967e-07, + "loss": 0.8985, + "step": 38380 + }, + { + "epoch": 0.9, + "grad_norm": 2.0221161790833047, + "learning_rate": 4.772019348933854e-07, + "loss": 1.0508, + "step": 38381 + }, + { + "epoch": 0.9, + "grad_norm": 1.9431159954929886, + "learning_rate": 4.769690623215195e-07, + "loss": 0.8989, + "step": 38382 + }, + { + "epoch": 0.9, + "grad_norm": 2.0922161333501457, + "learning_rate": 4.7673624519575355e-07, + "loss": 0.8857, + "step": 38383 + }, + { + "epoch": 0.9, + "grad_norm": 1.8248648104439202, + "learning_rate": 4.7650348351744533e-07, + "loss": 0.973, + "step": 38384 + }, + { + "epoch": 0.9, + "grad_norm": 1.9576840220969016, + "learning_rate": 4.7627077728794824e-07, + "loss": 0.8851, + "step": 38385 + }, + { + "epoch": 0.9, + "grad_norm": 1.9089995228474927, + "learning_rate": 4.7603812650861893e-07, + "loss": 1.0034, + "step": 38386 + }, + { + "epoch": 0.9, + "grad_norm": 1.8888715682335406, + "learning_rate": 4.758055311808096e-07, + "loss": 1.0604, + "step": 38387 + }, + { + "epoch": 0.9, + "grad_norm": 1.9490137719285399, + "learning_rate": 4.755729913058771e-07, + "loss": 0.8514, + "step": 38388 + }, + { + "epoch": 0.9, + "grad_norm": 2.5118633459182083, + "learning_rate": 4.753405068851735e-07, + "loss": 1.1631, + "step": 38389 + }, + { + "epoch": 0.9, + "grad_norm": 1.965753819536855, + "learning_rate": 4.751080779200512e-07, + "loss": 1.0187, + "step": 38390 + }, + { + "epoch": 0.9, + "grad_norm": 2.1009331027918603, + "learning_rate": 4.748757044118668e-07, + "loss": 1.021, + "step": 38391 + }, + { + "epoch": 0.9, + "grad_norm": 2.0257928241085743, + "learning_rate": 4.746433863619715e-07, + "loss": 1.0719, + "step": 38392 + }, + { + "epoch": 0.9, + "grad_norm": 2.001186989691884, + "learning_rate": 4.744111237717186e-07, + "loss": 0.8874, + "step": 38393 + }, + { + "epoch": 0.9, + "grad_norm": 1.8546075955760064, + "learning_rate": 4.741789166424593e-07, + "loss": 0.955, + "step": 38394 + }, + { + "epoch": 0.9, + "grad_norm": 2.0483270121867, + "learning_rate": 4.739467649755447e-07, + "loss": 0.9424, + "step": 38395 + }, + { + "epoch": 0.9, + "grad_norm": 2.203089733689088, + "learning_rate": 4.7371466877233043e-07, + "loss": 0.9719, + "step": 38396 + }, + { + "epoch": 0.9, + "grad_norm": 1.811837362069443, + "learning_rate": 4.734826280341653e-07, + "loss": 1.0742, + "step": 38397 + }, + { + "epoch": 0.9, + "grad_norm": 2.040274261595608, + "learning_rate": 4.732506427623984e-07, + "loss": 0.9639, + "step": 38398 + }, + { + "epoch": 0.9, + "grad_norm": 1.954556354378274, + "learning_rate": 4.7301871295838297e-07, + "loss": 0.9352, + "step": 38399 + }, + { + "epoch": 0.9, + "grad_norm": 2.1722498427045647, + "learning_rate": 4.727868386234702e-07, + "loss": 0.9803, + "step": 38400 + }, + { + "epoch": 0.9, + "grad_norm": 2.057460151415991, + "learning_rate": 4.725550197590079e-07, + "loss": 1.0986, + "step": 38401 + }, + { + "epoch": 0.9, + "grad_norm": 2.2405054097132204, + "learning_rate": 4.7232325636634603e-07, + "loss": 0.9044, + "step": 38402 + }, + { + "epoch": 0.9, + "grad_norm": 2.0332800498375647, + "learning_rate": 4.720915484468358e-07, + "loss": 0.9666, + "step": 38403 + }, + { + "epoch": 0.9, + "grad_norm": 1.11157316263985, + "learning_rate": 4.718598960018228e-07, + "loss": 0.9174, + "step": 38404 + }, + { + "epoch": 0.9, + "grad_norm": 1.0967950624358171, + "learning_rate": 4.716282990326604e-07, + "loss": 0.9902, + "step": 38405 + }, + { + "epoch": 0.9, + "grad_norm": 2.0459436613060107, + "learning_rate": 4.7139675754069304e-07, + "loss": 0.9598, + "step": 38406 + }, + { + "epoch": 0.9, + "grad_norm": 1.9950112110696236, + "learning_rate": 4.7116527152727187e-07, + "loss": 1.0365, + "step": 38407 + }, + { + "epoch": 0.9, + "grad_norm": 2.26994460380816, + "learning_rate": 4.709338409937414e-07, + "loss": 1.0049, + "step": 38408 + }, + { + "epoch": 0.9, + "grad_norm": 1.9741071986686678, + "learning_rate": 4.7070246594145276e-07, + "loss": 0.992, + "step": 38409 + }, + { + "epoch": 0.9, + "grad_norm": 2.0023985393980257, + "learning_rate": 4.704711463717515e-07, + "loss": 0.9487, + "step": 38410 + }, + { + "epoch": 0.9, + "grad_norm": 1.9822762648423162, + "learning_rate": 4.7023988228598326e-07, + "loss": 1.0736, + "step": 38411 + }, + { + "epoch": 0.9, + "grad_norm": 1.9442029151409237, + "learning_rate": 4.7000867368549473e-07, + "loss": 0.9765, + "step": 38412 + }, + { + "epoch": 0.9, + "grad_norm": 1.8159506012224955, + "learning_rate": 4.697775205716348e-07, + "loss": 0.8919, + "step": 38413 + }, + { + "epoch": 0.91, + "grad_norm": 1.0499610512744149, + "learning_rate": 4.695464229457469e-07, + "loss": 0.8535, + "step": 38414 + }, + { + "epoch": 0.91, + "grad_norm": 2.574742421674368, + "learning_rate": 4.693153808091755e-07, + "loss": 0.9415, + "step": 38415 + }, + { + "epoch": 0.91, + "grad_norm": 1.7786718652896292, + "learning_rate": 4.690843941632672e-07, + "loss": 0.8333, + "step": 38416 + }, + { + "epoch": 0.91, + "grad_norm": 2.0321143215342965, + "learning_rate": 4.6885346300936885e-07, + "loss": 0.8409, + "step": 38417 + }, + { + "epoch": 0.91, + "grad_norm": 1.7980867407006307, + "learning_rate": 4.686225873488226e-07, + "loss": 0.9807, + "step": 38418 + }, + { + "epoch": 0.91, + "grad_norm": 1.90786312567808, + "learning_rate": 4.6839176718297297e-07, + "loss": 0.9409, + "step": 38419 + }, + { + "epoch": 0.91, + "grad_norm": 1.0740267796645648, + "learning_rate": 4.6816100251316333e-07, + "loss": 0.9524, + "step": 38420 + }, + { + "epoch": 0.91, + "grad_norm": 1.1511605712745339, + "learning_rate": 4.6793029334073926e-07, + "loss": 1.0508, + "step": 38421 + }, + { + "epoch": 0.91, + "grad_norm": 2.049419531421368, + "learning_rate": 4.67699639667043e-07, + "loss": 0.9784, + "step": 38422 + }, + { + "epoch": 0.91, + "grad_norm": 1.9586432310635882, + "learning_rate": 4.6746904149341576e-07, + "loss": 0.9974, + "step": 38423 + }, + { + "epoch": 0.91, + "grad_norm": 1.0995084797169177, + "learning_rate": 4.67238498821202e-07, + "loss": 0.9677, + "step": 38424 + }, + { + "epoch": 0.91, + "grad_norm": 1.8459493938903486, + "learning_rate": 4.6700801165174504e-07, + "loss": 0.9617, + "step": 38425 + }, + { + "epoch": 0.91, + "grad_norm": 2.0379546310064702, + "learning_rate": 4.667775799863861e-07, + "loss": 1.0471, + "step": 38426 + }, + { + "epoch": 0.91, + "grad_norm": 2.5313234105179143, + "learning_rate": 4.66547203826464e-07, + "loss": 0.9306, + "step": 38427 + }, + { + "epoch": 0.91, + "grad_norm": 2.0414479616844, + "learning_rate": 4.6631688317332447e-07, + "loss": 0.9978, + "step": 38428 + }, + { + "epoch": 0.91, + "grad_norm": 2.0456749379753005, + "learning_rate": 4.660866180283041e-07, + "loss": 0.9968, + "step": 38429 + }, + { + "epoch": 0.91, + "grad_norm": 2.454723841862849, + "learning_rate": 4.658564083927475e-07, + "loss": 1.0329, + "step": 38430 + }, + { + "epoch": 0.91, + "grad_norm": 2.07347256970195, + "learning_rate": 4.6562625426799237e-07, + "loss": 1.1116, + "step": 38431 + }, + { + "epoch": 0.91, + "grad_norm": 1.8083394786009663, + "learning_rate": 4.65396155655381e-07, + "loss": 0.9917, + "step": 38432 + }, + { + "epoch": 0.91, + "grad_norm": 1.0644117083716482, + "learning_rate": 4.651661125562501e-07, + "loss": 0.9517, + "step": 38433 + }, + { + "epoch": 0.91, + "grad_norm": 2.157696809971713, + "learning_rate": 4.64936124971942e-07, + "loss": 0.8437, + "step": 38434 + }, + { + "epoch": 0.91, + "grad_norm": 1.974294991657053, + "learning_rate": 4.6470619290379547e-07, + "loss": 0.9944, + "step": 38435 + }, + { + "epoch": 0.91, + "grad_norm": 1.934043598587724, + "learning_rate": 4.6447631635314626e-07, + "loss": 1.0721, + "step": 38436 + }, + { + "epoch": 0.91, + "grad_norm": 1.0487122140889347, + "learning_rate": 4.6424649532133437e-07, + "loss": 0.8669, + "step": 38437 + }, + { + "epoch": 0.91, + "grad_norm": 1.7469112468824828, + "learning_rate": 4.640167298097009e-07, + "loss": 0.9364, + "step": 38438 + }, + { + "epoch": 0.91, + "grad_norm": 1.8788797683743108, + "learning_rate": 4.637870198195804e-07, + "loss": 0.9899, + "step": 38439 + }, + { + "epoch": 0.91, + "grad_norm": 1.9876799661417455, + "learning_rate": 4.6355736535230956e-07, + "loss": 1.1193, + "step": 38440 + }, + { + "epoch": 0.91, + "grad_norm": 2.2328729053954546, + "learning_rate": 4.633277664092262e-07, + "loss": 0.9743, + "step": 38441 + }, + { + "epoch": 0.91, + "grad_norm": 2.1553753787469057, + "learning_rate": 4.630982229916703e-07, + "loss": 0.8832, + "step": 38442 + }, + { + "epoch": 0.91, + "grad_norm": 2.2336053072752797, + "learning_rate": 4.628687351009753e-07, + "loss": 0.9085, + "step": 38443 + }, + { + "epoch": 0.91, + "grad_norm": 1.9183645574439114, + "learning_rate": 4.626393027384768e-07, + "loss": 0.9556, + "step": 38444 + }, + { + "epoch": 0.91, + "grad_norm": 1.8336260224464669, + "learning_rate": 4.624099259055126e-07, + "loss": 0.9186, + "step": 38445 + }, + { + "epoch": 0.91, + "grad_norm": 2.0790847058237745, + "learning_rate": 4.6218060460341716e-07, + "loss": 0.9704, + "step": 38446 + }, + { + "epoch": 0.91, + "grad_norm": 2.029743521756381, + "learning_rate": 4.619513388335273e-07, + "loss": 0.9532, + "step": 38447 + }, + { + "epoch": 0.91, + "grad_norm": 1.8060599086479119, + "learning_rate": 4.6172212859717403e-07, + "loss": 0.8883, + "step": 38448 + }, + { + "epoch": 0.91, + "grad_norm": 1.1003703145838086, + "learning_rate": 4.6149297389569634e-07, + "loss": 0.9326, + "step": 38449 + }, + { + "epoch": 0.91, + "grad_norm": 2.119100505090902, + "learning_rate": 4.612638747304243e-07, + "loss": 1.0471, + "step": 38450 + }, + { + "epoch": 0.91, + "grad_norm": 2.09367489613336, + "learning_rate": 4.6103483110269575e-07, + "loss": 0.9398, + "step": 38451 + }, + { + "epoch": 0.91, + "grad_norm": 2.1268078810909374, + "learning_rate": 4.608058430138407e-07, + "loss": 1.0384, + "step": 38452 + }, + { + "epoch": 0.91, + "grad_norm": 1.830751146842632, + "learning_rate": 4.605769104651958e-07, + "loss": 0.9945, + "step": 38453 + }, + { + "epoch": 0.91, + "grad_norm": 2.1009521461102487, + "learning_rate": 4.603480334580912e-07, + "loss": 0.8703, + "step": 38454 + }, + { + "epoch": 0.91, + "grad_norm": 3.851026036735093, + "learning_rate": 4.6011921199386136e-07, + "loss": 0.9245, + "step": 38455 + }, + { + "epoch": 0.91, + "grad_norm": 1.802543502298497, + "learning_rate": 4.5989044607383846e-07, + "loss": 0.9205, + "step": 38456 + }, + { + "epoch": 0.91, + "grad_norm": 2.5882747625957054, + "learning_rate": 4.5966173569935156e-07, + "loss": 1.0176, + "step": 38457 + }, + { + "epoch": 0.91, + "grad_norm": 2.2757582505650467, + "learning_rate": 4.594330808717351e-07, + "loss": 1.0487, + "step": 38458 + }, + { + "epoch": 0.91, + "grad_norm": 1.927444559115392, + "learning_rate": 4.592044815923191e-07, + "loss": 1.0353, + "step": 38459 + }, + { + "epoch": 0.91, + "grad_norm": 2.062435159117802, + "learning_rate": 4.589759378624381e-07, + "loss": 0.9452, + "step": 38460 + }, + { + "epoch": 0.91, + "grad_norm": 2.071613025869499, + "learning_rate": 4.587474496834166e-07, + "loss": 1.1129, + "step": 38461 + }, + { + "epoch": 0.91, + "grad_norm": 2.0047114977734495, + "learning_rate": 4.58519017056589e-07, + "loss": 1.0409, + "step": 38462 + }, + { + "epoch": 0.91, + "grad_norm": 2.235772623799888, + "learning_rate": 4.582906399832854e-07, + "loss": 0.9752, + "step": 38463 + }, + { + "epoch": 0.91, + "grad_norm": 1.9879427504723548, + "learning_rate": 4.5806231846483475e-07, + "loss": 1.1045, + "step": 38464 + }, + { + "epoch": 0.91, + "grad_norm": 2.064039118571726, + "learning_rate": 4.5783405250256487e-07, + "loss": 0.9594, + "step": 38465 + }, + { + "epoch": 0.91, + "grad_norm": 2.0760373550143134, + "learning_rate": 4.5760584209780577e-07, + "loss": 0.973, + "step": 38466 + }, + { + "epoch": 0.91, + "grad_norm": 2.219516572615054, + "learning_rate": 4.573776872518887e-07, + "loss": 1.0111, + "step": 38467 + }, + { + "epoch": 0.91, + "grad_norm": 1.9767383785317696, + "learning_rate": 4.571495879661392e-07, + "loss": 0.8506, + "step": 38468 + }, + { + "epoch": 0.91, + "grad_norm": 2.0137884303885327, + "learning_rate": 4.5692154424188506e-07, + "loss": 0.9657, + "step": 38469 + }, + { + "epoch": 0.91, + "grad_norm": 1.8712944986248299, + "learning_rate": 4.566935560804564e-07, + "loss": 1.0995, + "step": 38470 + }, + { + "epoch": 0.91, + "grad_norm": 2.3918227260537663, + "learning_rate": 4.5646562348317657e-07, + "loss": 0.8051, + "step": 38471 + }, + { + "epoch": 0.91, + "grad_norm": 1.8949511803068564, + "learning_rate": 4.5623774645137675e-07, + "loss": 0.9825, + "step": 38472 + }, + { + "epoch": 0.91, + "grad_norm": 1.1336126214868052, + "learning_rate": 4.5600992498638364e-07, + "loss": 0.9078, + "step": 38473 + }, + { + "epoch": 0.91, + "grad_norm": 1.9810515757413778, + "learning_rate": 4.557821590895217e-07, + "loss": 0.8785, + "step": 38474 + }, + { + "epoch": 0.91, + "grad_norm": 1.9527080008480233, + "learning_rate": 4.5555444876211664e-07, + "loss": 0.9441, + "step": 38475 + }, + { + "epoch": 0.91, + "grad_norm": 1.0954398584364669, + "learning_rate": 4.553267940054951e-07, + "loss": 0.9597, + "step": 38476 + }, + { + "epoch": 0.91, + "grad_norm": 1.8466242067514813, + "learning_rate": 4.5509919482098486e-07, + "loss": 0.8314, + "step": 38477 + }, + { + "epoch": 0.91, + "grad_norm": 2.074041685666581, + "learning_rate": 4.5487165120990716e-07, + "loss": 1.0175, + "step": 38478 + }, + { + "epoch": 0.91, + "grad_norm": 1.9036432464342745, + "learning_rate": 4.546441631735887e-07, + "loss": 1.0279, + "step": 38479 + }, + { + "epoch": 0.91, + "grad_norm": 1.9193549140063109, + "learning_rate": 4.544167307133529e-07, + "loss": 1.1049, + "step": 38480 + }, + { + "epoch": 0.91, + "grad_norm": 1.7891820893391155, + "learning_rate": 4.5418935383052756e-07, + "loss": 1.0503, + "step": 38481 + }, + { + "epoch": 0.91, + "grad_norm": 2.0114679714944823, + "learning_rate": 4.5396203252643377e-07, + "loss": 1.0482, + "step": 38482 + }, + { + "epoch": 0.91, + "grad_norm": 1.956394352467715, + "learning_rate": 4.5373476680239283e-07, + "loss": 1.0392, + "step": 38483 + }, + { + "epoch": 0.91, + "grad_norm": 2.0012440827267395, + "learning_rate": 4.5350755665973243e-07, + "loss": 0.9913, + "step": 38484 + }, + { + "epoch": 0.91, + "grad_norm": 1.8516452486373522, + "learning_rate": 4.532804020997728e-07, + "loss": 0.9983, + "step": 38485 + }, + { + "epoch": 0.91, + "grad_norm": 1.0866542331657476, + "learning_rate": 4.5305330312383713e-07, + "loss": 0.9355, + "step": 38486 + }, + { + "epoch": 0.91, + "grad_norm": 2.3059985887776855, + "learning_rate": 4.5282625973324666e-07, + "loss": 0.9373, + "step": 38487 + }, + { + "epoch": 0.91, + "grad_norm": 2.0356643407694572, + "learning_rate": 4.52599271929326e-07, + "loss": 0.9886, + "step": 38488 + }, + { + "epoch": 0.91, + "grad_norm": 1.8512465825152358, + "learning_rate": 4.523723397133939e-07, + "loss": 0.9893, + "step": 38489 + }, + { + "epoch": 0.91, + "grad_norm": 1.7969179351098223, + "learning_rate": 4.5214546308677386e-07, + "loss": 0.86, + "step": 38490 + }, + { + "epoch": 0.91, + "grad_norm": 2.126606923009146, + "learning_rate": 4.5191864205078596e-07, + "loss": 0.9139, + "step": 38491 + }, + { + "epoch": 0.91, + "grad_norm": 2.0982782313746773, + "learning_rate": 4.516918766067491e-07, + "loss": 1.0202, + "step": 38492 + }, + { + "epoch": 0.91, + "grad_norm": 1.9431096055435417, + "learning_rate": 4.514651667559844e-07, + "loss": 1.0146, + "step": 38493 + }, + { + "epoch": 0.91, + "grad_norm": 2.0704492526053615, + "learning_rate": 4.512385124998142e-07, + "loss": 1.0095, + "step": 38494 + }, + { + "epoch": 0.91, + "grad_norm": 1.1856623670720903, + "learning_rate": 4.510119138395563e-07, + "loss": 0.9073, + "step": 38495 + }, + { + "epoch": 0.91, + "grad_norm": 2.000069928064354, + "learning_rate": 4.5078537077652975e-07, + "loss": 1.0179, + "step": 38496 + }, + { + "epoch": 0.91, + "grad_norm": 2.304844525612957, + "learning_rate": 4.5055888331205335e-07, + "loss": 1.0582, + "step": 38497 + }, + { + "epoch": 0.91, + "grad_norm": 1.0809488872550754, + "learning_rate": 4.503324514474483e-07, + "loss": 0.9759, + "step": 38498 + }, + { + "epoch": 0.91, + "grad_norm": 1.8865951989771876, + "learning_rate": 4.5010607518403024e-07, + "loss": 0.9968, + "step": 38499 + }, + { + "epoch": 0.91, + "grad_norm": 1.9122466289162192, + "learning_rate": 4.498797545231171e-07, + "loss": 0.8244, + "step": 38500 + }, + { + "epoch": 0.91, + "grad_norm": 2.0549336901683177, + "learning_rate": 4.496534894660276e-07, + "loss": 0.9892, + "step": 38501 + }, + { + "epoch": 0.91, + "grad_norm": 1.851287433809163, + "learning_rate": 4.4942728001408087e-07, + "loss": 0.9993, + "step": 38502 + }, + { + "epoch": 0.91, + "grad_norm": 1.1411082094624168, + "learning_rate": 4.4920112616859135e-07, + "loss": 0.9281, + "step": 38503 + }, + { + "epoch": 0.91, + "grad_norm": 2.187381695524276, + "learning_rate": 4.4897502793087576e-07, + "loss": 0.9897, + "step": 38504 + }, + { + "epoch": 0.91, + "grad_norm": 1.809173061227344, + "learning_rate": 4.487489853022531e-07, + "loss": 0.9463, + "step": 38505 + }, + { + "epoch": 0.91, + "grad_norm": 2.1991473813317195, + "learning_rate": 4.485229982840356e-07, + "loss": 0.9555, + "step": 38506 + }, + { + "epoch": 0.91, + "grad_norm": 1.9205125651198431, + "learning_rate": 4.4829706687754213e-07, + "loss": 0.9541, + "step": 38507 + }, + { + "epoch": 0.91, + "grad_norm": 1.9337611490213853, + "learning_rate": 4.480711910840874e-07, + "loss": 0.8405, + "step": 38508 + }, + { + "epoch": 0.91, + "grad_norm": 2.0980195457561206, + "learning_rate": 4.478453709049857e-07, + "loss": 0.993, + "step": 38509 + }, + { + "epoch": 0.91, + "grad_norm": 1.8935952669770957, + "learning_rate": 4.476196063415528e-07, + "loss": 1.0884, + "step": 38510 + }, + { + "epoch": 0.91, + "grad_norm": 1.8461963850361325, + "learning_rate": 4.4739389739510306e-07, + "loss": 0.9136, + "step": 38511 + }, + { + "epoch": 0.91, + "grad_norm": 1.9913512586689022, + "learning_rate": 4.4716824406695005e-07, + "loss": 1.0034, + "step": 38512 + }, + { + "epoch": 0.91, + "grad_norm": 2.08445608282571, + "learning_rate": 4.4694264635840814e-07, + "loss": 0.9078, + "step": 38513 + }, + { + "epoch": 0.91, + "grad_norm": 5.803776117074683, + "learning_rate": 4.467171042707896e-07, + "loss": 1.0045, + "step": 38514 + }, + { + "epoch": 0.91, + "grad_norm": 1.1177818028071829, + "learning_rate": 4.4649161780540907e-07, + "loss": 0.9531, + "step": 38515 + }, + { + "epoch": 0.91, + "grad_norm": 2.23916648062116, + "learning_rate": 4.4626618696357983e-07, + "loss": 1.0078, + "step": 38516 + }, + { + "epoch": 0.91, + "grad_norm": 1.8470930608781737, + "learning_rate": 4.460408117466131e-07, + "loss": 0.9636, + "step": 38517 + }, + { + "epoch": 0.91, + "grad_norm": 2.114607355899589, + "learning_rate": 4.4581549215582e-07, + "loss": 0.9403, + "step": 38518 + }, + { + "epoch": 0.91, + "grad_norm": 2.099738754899328, + "learning_rate": 4.455902281925162e-07, + "loss": 0.8583, + "step": 38519 + }, + { + "epoch": 0.91, + "grad_norm": 2.072646564561844, + "learning_rate": 4.453650198580106e-07, + "loss": 0.8288, + "step": 38520 + }, + { + "epoch": 0.91, + "grad_norm": 2.502669601828748, + "learning_rate": 4.4513986715361336e-07, + "loss": 0.955, + "step": 38521 + }, + { + "epoch": 0.91, + "grad_norm": 1.9459406432713282, + "learning_rate": 4.4491477008063666e-07, + "loss": 1.0165, + "step": 38522 + }, + { + "epoch": 0.91, + "grad_norm": 2.4632796250517024, + "learning_rate": 4.446897286403928e-07, + "loss": 1.0109, + "step": 38523 + }, + { + "epoch": 0.91, + "grad_norm": 2.109509244373442, + "learning_rate": 4.444647428341908e-07, + "loss": 1.0377, + "step": 38524 + }, + { + "epoch": 0.91, + "grad_norm": 2.0479611947061134, + "learning_rate": 4.4423981266333847e-07, + "loss": 0.9525, + "step": 38525 + }, + { + "epoch": 0.91, + "grad_norm": 2.0026695711679157, + "learning_rate": 4.440149381291492e-07, + "loss": 0.9987, + "step": 38526 + }, + { + "epoch": 0.91, + "grad_norm": 2.051036067413123, + "learning_rate": 4.437901192329286e-07, + "loss": 0.996, + "step": 38527 + }, + { + "epoch": 0.91, + "grad_norm": 2.1431453846293382, + "learning_rate": 4.435653559759878e-07, + "loss": 0.9207, + "step": 38528 + }, + { + "epoch": 0.91, + "grad_norm": 2.020040190168988, + "learning_rate": 4.4334064835963475e-07, + "loss": 0.9322, + "step": 38529 + }, + { + "epoch": 0.91, + "grad_norm": 1.8484103654135862, + "learning_rate": 4.4311599638517944e-07, + "loss": 0.9181, + "step": 38530 + }, + { + "epoch": 0.91, + "grad_norm": 1.0634089950678802, + "learning_rate": 4.4289140005392637e-07, + "loss": 0.9447, + "step": 38531 + }, + { + "epoch": 0.91, + "grad_norm": 2.1353142244807466, + "learning_rate": 4.426668593671879e-07, + "loss": 1.0108, + "step": 38532 + }, + { + "epoch": 0.91, + "grad_norm": 1.9938528562461992, + "learning_rate": 4.4244237432626737e-07, + "loss": 1.0185, + "step": 38533 + }, + { + "epoch": 0.91, + "grad_norm": 2.4170656116969824, + "learning_rate": 4.422179449324726e-07, + "loss": 0.8921, + "step": 38534 + }, + { + "epoch": 0.91, + "grad_norm": 1.1133175889994382, + "learning_rate": 4.419935711871115e-07, + "loss": 1.0145, + "step": 38535 + }, + { + "epoch": 0.91, + "grad_norm": 1.9611213307490492, + "learning_rate": 4.417692530914908e-07, + "loss": 1.0218, + "step": 38536 + }, + { + "epoch": 0.91, + "grad_norm": 1.867803338999323, + "learning_rate": 4.415449906469149e-07, + "loss": 1.0253, + "step": 38537 + }, + { + "epoch": 0.91, + "grad_norm": 1.686642606202232, + "learning_rate": 4.4132078385468956e-07, + "loss": 0.9286, + "step": 38538 + }, + { + "epoch": 0.91, + "grad_norm": 2.0264033935943173, + "learning_rate": 4.4109663271612036e-07, + "loss": 0.928, + "step": 38539 + }, + { + "epoch": 0.91, + "grad_norm": 1.9909210828986081, + "learning_rate": 4.40872537232514e-07, + "loss": 1.0787, + "step": 38540 + }, + { + "epoch": 0.91, + "grad_norm": 1.838689587280136, + "learning_rate": 4.40648497405175e-07, + "loss": 0.9383, + "step": 38541 + }, + { + "epoch": 0.91, + "grad_norm": 1.8862561568654599, + "learning_rate": 4.4042451323540457e-07, + "loss": 0.8234, + "step": 38542 + }, + { + "epoch": 0.91, + "grad_norm": 2.0263831056406247, + "learning_rate": 4.4020058472450946e-07, + "loss": 0.9738, + "step": 38543 + }, + { + "epoch": 0.91, + "grad_norm": 1.9195403068425467, + "learning_rate": 4.399767118737952e-07, + "loss": 0.926, + "step": 38544 + }, + { + "epoch": 0.91, + "grad_norm": 1.8885889323601464, + "learning_rate": 4.3975289468456194e-07, + "loss": 1.0315, + "step": 38545 + }, + { + "epoch": 0.91, + "grad_norm": 1.8013282273450266, + "learning_rate": 4.395291331581142e-07, + "loss": 0.985, + "step": 38546 + }, + { + "epoch": 0.91, + "grad_norm": 1.7821658288719031, + "learning_rate": 4.393054272957531e-07, + "loss": 0.9799, + "step": 38547 + }, + { + "epoch": 0.91, + "grad_norm": 2.124691633266089, + "learning_rate": 4.390817770987854e-07, + "loss": 0.9209, + "step": 38548 + }, + { + "epoch": 0.91, + "grad_norm": 2.256493604821059, + "learning_rate": 4.38858182568509e-07, + "loss": 0.9404, + "step": 38549 + }, + { + "epoch": 0.91, + "grad_norm": 2.1957082175402927, + "learning_rate": 4.3863464370622723e-07, + "loss": 0.8966, + "step": 38550 + }, + { + "epoch": 0.91, + "grad_norm": 1.9418272818981965, + "learning_rate": 4.384111605132424e-07, + "loss": 0.9733, + "step": 38551 + }, + { + "epoch": 0.91, + "grad_norm": 1.819742076737939, + "learning_rate": 4.381877329908535e-07, + "loss": 1.0369, + "step": 38552 + }, + { + "epoch": 0.91, + "grad_norm": 1.8975209253903311, + "learning_rate": 4.37964361140365e-07, + "loss": 0.8608, + "step": 38553 + }, + { + "epoch": 0.91, + "grad_norm": 1.9115066744371536, + "learning_rate": 4.377410449630737e-07, + "loss": 0.9502, + "step": 38554 + }, + { + "epoch": 0.91, + "grad_norm": 2.205296912924384, + "learning_rate": 4.375177844602818e-07, + "loss": 1.0543, + "step": 38555 + }, + { + "epoch": 0.91, + "grad_norm": 3.887512094194835, + "learning_rate": 4.3729457963328837e-07, + "loss": 0.9983, + "step": 38556 + }, + { + "epoch": 0.91, + "grad_norm": 2.1236063371193996, + "learning_rate": 4.370714304833923e-07, + "loss": 1.0292, + "step": 38557 + }, + { + "epoch": 0.91, + "grad_norm": 1.0442491862800478, + "learning_rate": 4.368483370118981e-07, + "loss": 0.8978, + "step": 38558 + }, + { + "epoch": 0.91, + "grad_norm": 1.8585787937188694, + "learning_rate": 4.36625299220097e-07, + "loss": 1.0681, + "step": 38559 + }, + { + "epoch": 0.91, + "grad_norm": 1.9546673599965827, + "learning_rate": 4.3640231710929124e-07, + "loss": 1.0118, + "step": 38560 + }, + { + "epoch": 0.91, + "grad_norm": 2.0829045681017977, + "learning_rate": 4.3617939068077874e-07, + "loss": 1.0222, + "step": 38561 + }, + { + "epoch": 0.91, + "grad_norm": 1.9377511741555713, + "learning_rate": 4.359565199358606e-07, + "loss": 0.9984, + "step": 38562 + }, + { + "epoch": 0.91, + "grad_norm": 1.853915341019796, + "learning_rate": 4.357337048758292e-07, + "loss": 0.9589, + "step": 38563 + }, + { + "epoch": 0.91, + "grad_norm": 2.0180895855022642, + "learning_rate": 4.355109455019835e-07, + "loss": 0.9541, + "step": 38564 + }, + { + "epoch": 0.91, + "grad_norm": 1.990700929105383, + "learning_rate": 4.3528824181562237e-07, + "loss": 1.0124, + "step": 38565 + }, + { + "epoch": 0.91, + "grad_norm": 2.2025876598945344, + "learning_rate": 4.350655938180426e-07, + "loss": 0.9947, + "step": 38566 + }, + { + "epoch": 0.91, + "grad_norm": 2.059820733553065, + "learning_rate": 4.348430015105365e-07, + "loss": 1.0441, + "step": 38567 + }, + { + "epoch": 0.91, + "grad_norm": 1.9995842597946876, + "learning_rate": 4.34620464894403e-07, + "loss": 1.1013, + "step": 38568 + }, + { + "epoch": 0.91, + "grad_norm": 2.0446653110639157, + "learning_rate": 4.343979839709389e-07, + "loss": 0.9175, + "step": 38569 + }, + { + "epoch": 0.91, + "grad_norm": 1.9189144442650228, + "learning_rate": 4.3417555874143644e-07, + "loss": 1.0, + "step": 38570 + }, + { + "epoch": 0.91, + "grad_norm": 1.8533934162498875, + "learning_rate": 4.3395318920719465e-07, + "loss": 1.0984, + "step": 38571 + }, + { + "epoch": 0.91, + "grad_norm": 1.9205829904852427, + "learning_rate": 4.3373087536950574e-07, + "loss": 0.9507, + "step": 38572 + }, + { + "epoch": 0.91, + "grad_norm": 2.0605305711724937, + "learning_rate": 4.335086172296621e-07, + "loss": 0.8174, + "step": 38573 + }, + { + "epoch": 0.91, + "grad_norm": 2.1447753443796653, + "learning_rate": 4.3328641478896037e-07, + "loss": 1.0135, + "step": 38574 + }, + { + "epoch": 0.91, + "grad_norm": 1.9891178340989473, + "learning_rate": 4.330642680486952e-07, + "loss": 0.9787, + "step": 38575 + }, + { + "epoch": 0.91, + "grad_norm": 1.8300231501170376, + "learning_rate": 4.3284217701015985e-07, + "loss": 0.8912, + "step": 38576 + }, + { + "epoch": 0.91, + "grad_norm": 1.1467481692284511, + "learning_rate": 4.3262014167464337e-07, + "loss": 0.9822, + "step": 38577 + }, + { + "epoch": 0.91, + "grad_norm": 1.8560482272208614, + "learning_rate": 4.3239816204344256e-07, + "loss": 0.93, + "step": 38578 + }, + { + "epoch": 0.91, + "grad_norm": 1.8515686817709824, + "learning_rate": 4.321762381178496e-07, + "loss": 0.9687, + "step": 38579 + }, + { + "epoch": 0.91, + "grad_norm": 2.2162568402934464, + "learning_rate": 4.319543698991546e-07, + "loss": 1.0076, + "step": 38580 + }, + { + "epoch": 0.91, + "grad_norm": 1.178744552657616, + "learning_rate": 4.3173255738865104e-07, + "loss": 0.9463, + "step": 38581 + }, + { + "epoch": 0.91, + "grad_norm": 2.0800798263901035, + "learning_rate": 4.3151080058762897e-07, + "loss": 0.8885, + "step": 38582 + }, + { + "epoch": 0.91, + "grad_norm": 2.26835031954909, + "learning_rate": 4.312890994973806e-07, + "loss": 0.9539, + "step": 38583 + }, + { + "epoch": 0.91, + "grad_norm": 2.2658184723213193, + "learning_rate": 4.310674541191984e-07, + "loss": 1.053, + "step": 38584 + }, + { + "epoch": 0.91, + "grad_norm": 2.208684017947681, + "learning_rate": 4.308458644543678e-07, + "loss": 0.9488, + "step": 38585 + }, + { + "epoch": 0.91, + "grad_norm": 1.0050200108251717, + "learning_rate": 4.306243305041846e-07, + "loss": 0.8998, + "step": 38586 + }, + { + "epoch": 0.91, + "grad_norm": 1.8174714214681267, + "learning_rate": 4.304028522699344e-07, + "loss": 0.9959, + "step": 38587 + }, + { + "epoch": 0.91, + "grad_norm": 2.004753019750862, + "learning_rate": 4.3018142975291054e-07, + "loss": 1.0754, + "step": 38588 + }, + { + "epoch": 0.91, + "grad_norm": 2.006380269952263, + "learning_rate": 4.2996006295439764e-07, + "loss": 1.0169, + "step": 38589 + }, + { + "epoch": 0.91, + "grad_norm": 1.936158833155222, + "learning_rate": 4.2973875187568903e-07, + "loss": 0.9878, + "step": 38590 + }, + { + "epoch": 0.91, + "grad_norm": 2.1004929991946724, + "learning_rate": 4.295174965180704e-07, + "loss": 0.9061, + "step": 38591 + }, + { + "epoch": 0.91, + "grad_norm": 2.4024496979727545, + "learning_rate": 4.292962968828318e-07, + "loss": 1.0122, + "step": 38592 + }, + { + "epoch": 0.91, + "grad_norm": 2.1180063476847826, + "learning_rate": 4.2907515297126003e-07, + "loss": 0.9073, + "step": 38593 + }, + { + "epoch": 0.91, + "grad_norm": 2.4126225629253177, + "learning_rate": 4.288540647846418e-07, + "loss": 0.9044, + "step": 38594 + }, + { + "epoch": 0.91, + "grad_norm": 1.89864035614598, + "learning_rate": 4.286330323242649e-07, + "loss": 0.8858, + "step": 38595 + }, + { + "epoch": 0.91, + "grad_norm": 2.7695883377147608, + "learning_rate": 4.284120555914184e-07, + "loss": 0.899, + "step": 38596 + }, + { + "epoch": 0.91, + "grad_norm": 2.377693741859973, + "learning_rate": 4.281911345873868e-07, + "loss": 1.0057, + "step": 38597 + }, + { + "epoch": 0.91, + "grad_norm": 1.9484181986681028, + "learning_rate": 4.279702693134569e-07, + "loss": 0.8863, + "step": 38598 + }, + { + "epoch": 0.91, + "grad_norm": 1.9154241991387473, + "learning_rate": 4.277494597709131e-07, + "loss": 0.999, + "step": 38599 + }, + { + "epoch": 0.91, + "grad_norm": 2.0218798163075102, + "learning_rate": 4.2752870596104446e-07, + "loss": 1.0524, + "step": 38600 + }, + { + "epoch": 0.91, + "grad_norm": 1.9183063793814363, + "learning_rate": 4.273080078851333e-07, + "loss": 0.885, + "step": 38601 + }, + { + "epoch": 0.91, + "grad_norm": 1.0507344549910949, + "learning_rate": 4.270873655444652e-07, + "loss": 0.8872, + "step": 38602 + }, + { + "epoch": 0.91, + "grad_norm": 1.9936382280121423, + "learning_rate": 4.268667789403247e-07, + "loss": 1.0089, + "step": 38603 + }, + { + "epoch": 0.91, + "grad_norm": 1.1237842049824045, + "learning_rate": 4.2664624807399857e-07, + "loss": 0.9198, + "step": 38604 + }, + { + "epoch": 0.91, + "grad_norm": 2.079607519901818, + "learning_rate": 4.264257729467691e-07, + "loss": 1.1259, + "step": 38605 + }, + { + "epoch": 0.91, + "grad_norm": 2.2967222532544653, + "learning_rate": 4.2620535355991756e-07, + "loss": 0.9753, + "step": 38606 + }, + { + "epoch": 0.91, + "grad_norm": 2.0454427917310145, + "learning_rate": 4.2598498991473057e-07, + "loss": 0.9418, + "step": 38607 + }, + { + "epoch": 0.91, + "grad_norm": 1.843219989152676, + "learning_rate": 4.2576468201248946e-07, + "loss": 1.1064, + "step": 38608 + }, + { + "epoch": 0.91, + "grad_norm": 2.5477165853145927, + "learning_rate": 4.2554442985447864e-07, + "loss": 0.8161, + "step": 38609 + }, + { + "epoch": 0.91, + "grad_norm": 1.997954891554216, + "learning_rate": 4.2532423344197825e-07, + "loss": 0.9706, + "step": 38610 + }, + { + "epoch": 0.91, + "grad_norm": 2.154031808564079, + "learning_rate": 4.2510409277627174e-07, + "loss": 1.0269, + "step": 38611 + }, + { + "epoch": 0.91, + "grad_norm": 2.1810151143026806, + "learning_rate": 4.2488400785864027e-07, + "loss": 0.9227, + "step": 38612 + }, + { + "epoch": 0.91, + "grad_norm": 2.060009714931453, + "learning_rate": 4.2466397869036724e-07, + "loss": 0.8385, + "step": 38613 + }, + { + "epoch": 0.91, + "grad_norm": 2.1561003693872713, + "learning_rate": 4.2444400527273056e-07, + "loss": 1.0324, + "step": 38614 + }, + { + "epoch": 0.91, + "grad_norm": 2.1137529868965257, + "learning_rate": 4.2422408760701253e-07, + "loss": 1.061, + "step": 38615 + }, + { + "epoch": 0.91, + "grad_norm": 1.0189722229208402, + "learning_rate": 4.240042256944932e-07, + "loss": 0.9144, + "step": 38616 + }, + { + "epoch": 0.91, + "grad_norm": 1.8276610422515918, + "learning_rate": 4.2378441953645375e-07, + "loss": 0.9431, + "step": 38617 + }, + { + "epoch": 0.91, + "grad_norm": 1.1300769330465767, + "learning_rate": 4.2356466913417326e-07, + "loss": 0.9743, + "step": 38618 + }, + { + "epoch": 0.91, + "grad_norm": 3.229950339276994, + "learning_rate": 4.233449744889295e-07, + "loss": 0.8992, + "step": 38619 + }, + { + "epoch": 0.91, + "grad_norm": 1.8975156775179967, + "learning_rate": 4.2312533560200376e-07, + "loss": 0.9272, + "step": 38620 + }, + { + "epoch": 0.91, + "grad_norm": 1.885848069910109, + "learning_rate": 4.22905752474676e-07, + "loss": 0.9481, + "step": 38621 + }, + { + "epoch": 0.91, + "grad_norm": 2.363182722355007, + "learning_rate": 4.226862251082231e-07, + "loss": 0.9234, + "step": 38622 + }, + { + "epoch": 0.91, + "grad_norm": 2.0387938660941147, + "learning_rate": 4.224667535039206e-07, + "loss": 1.0349, + "step": 38623 + }, + { + "epoch": 0.91, + "grad_norm": 2.231797048906429, + "learning_rate": 4.222473376630498e-07, + "loss": 0.9295, + "step": 38624 + }, + { + "epoch": 0.91, + "grad_norm": 1.9468027245489044, + "learning_rate": 4.2202797758688853e-07, + "loss": 0.9233, + "step": 38625 + }, + { + "epoch": 0.91, + "grad_norm": 1.7629120728448329, + "learning_rate": 4.2180867327671237e-07, + "loss": 0.914, + "step": 38626 + }, + { + "epoch": 0.91, + "grad_norm": 1.9147662478971093, + "learning_rate": 4.2158942473379706e-07, + "loss": 1.0421, + "step": 38627 + }, + { + "epoch": 0.91, + "grad_norm": 1.8201710993720375, + "learning_rate": 4.2137023195942264e-07, + "loss": 0.9915, + "step": 38628 + }, + { + "epoch": 0.91, + "grad_norm": 1.9535734778727836, + "learning_rate": 4.211510949548614e-07, + "loss": 0.8966, + "step": 38629 + }, + { + "epoch": 0.91, + "grad_norm": 2.267880216538357, + "learning_rate": 4.209320137213924e-07, + "loss": 0.9343, + "step": 38630 + }, + { + "epoch": 0.91, + "grad_norm": 2.3430824357262217, + "learning_rate": 4.2071298826028894e-07, + "loss": 0.9411, + "step": 38631 + }, + { + "epoch": 0.91, + "grad_norm": 1.1818079810207929, + "learning_rate": 4.204940185728279e-07, + "loss": 0.8737, + "step": 38632 + }, + { + "epoch": 0.91, + "grad_norm": 1.8155261796934532, + "learning_rate": 4.2027510466028265e-07, + "loss": 1.037, + "step": 38633 + }, + { + "epoch": 0.91, + "grad_norm": 2.344721055927984, + "learning_rate": 4.2005624652392996e-07, + "loss": 0.9828, + "step": 38634 + }, + { + "epoch": 0.91, + "grad_norm": 3.027047759168204, + "learning_rate": 4.1983744416504323e-07, + "loss": 0.9013, + "step": 38635 + }, + { + "epoch": 0.91, + "grad_norm": 2.5185672300979736, + "learning_rate": 4.196186975848937e-07, + "loss": 1.0852, + "step": 38636 + }, + { + "epoch": 0.91, + "grad_norm": 2.127042478312524, + "learning_rate": 4.19400006784757e-07, + "loss": 1.0029, + "step": 38637 + }, + { + "epoch": 0.91, + "grad_norm": 1.9875109252131768, + "learning_rate": 4.1918137176590876e-07, + "loss": 1.0033, + "step": 38638 + }, + { + "epoch": 0.91, + "grad_norm": 1.8574642104243995, + "learning_rate": 4.189627925296202e-07, + "loss": 0.975, + "step": 38639 + }, + { + "epoch": 0.91, + "grad_norm": 2.0556358689415783, + "learning_rate": 4.187442690771615e-07, + "loss": 1.0071, + "step": 38640 + }, + { + "epoch": 0.91, + "grad_norm": 1.976102567204989, + "learning_rate": 4.18525801409807e-07, + "loss": 1.0446, + "step": 38641 + }, + { + "epoch": 0.91, + "grad_norm": 2.0300017757580324, + "learning_rate": 4.183073895288303e-07, + "loss": 0.9677, + "step": 38642 + }, + { + "epoch": 0.91, + "grad_norm": 1.9953302516125202, + "learning_rate": 4.1808903343550145e-07, + "loss": 1.0482, + "step": 38643 + }, + { + "epoch": 0.91, + "grad_norm": 2.31315497256141, + "learning_rate": 4.178707331310905e-07, + "loss": 0.9263, + "step": 38644 + }, + { + "epoch": 0.91, + "grad_norm": 1.863801388559467, + "learning_rate": 4.176524886168698e-07, + "loss": 0.9134, + "step": 38645 + }, + { + "epoch": 0.91, + "grad_norm": 2.907089331469427, + "learning_rate": 4.1743429989411054e-07, + "loss": 0.9577, + "step": 38646 + }, + { + "epoch": 0.91, + "grad_norm": 1.9219839023539982, + "learning_rate": 4.1721616696408394e-07, + "loss": 0.9731, + "step": 38647 + }, + { + "epoch": 0.91, + "grad_norm": 1.6885544994452186, + "learning_rate": 4.169980898280568e-07, + "loss": 0.927, + "step": 38648 + }, + { + "epoch": 0.91, + "grad_norm": 1.8193761492348273, + "learning_rate": 4.167800684873013e-07, + "loss": 0.8819, + "step": 38649 + }, + { + "epoch": 0.91, + "grad_norm": 1.892848388039446, + "learning_rate": 4.165621029430855e-07, + "loss": 0.8819, + "step": 38650 + }, + { + "epoch": 0.91, + "grad_norm": 2.647526299001517, + "learning_rate": 4.1634419319668163e-07, + "loss": 0.9057, + "step": 38651 + }, + { + "epoch": 0.91, + "grad_norm": 1.8624254792871568, + "learning_rate": 4.1612633924935307e-07, + "loss": 0.9087, + "step": 38652 + }, + { + "epoch": 0.91, + "grad_norm": 1.921922967935336, + "learning_rate": 4.1590854110237333e-07, + "loss": 1.0341, + "step": 38653 + }, + { + "epoch": 0.91, + "grad_norm": 2.0342007564739175, + "learning_rate": 4.156907987570069e-07, + "loss": 0.9866, + "step": 38654 + }, + { + "epoch": 0.91, + "grad_norm": 1.032117816809711, + "learning_rate": 4.154731122145228e-07, + "loss": 0.9124, + "step": 38655 + }, + { + "epoch": 0.91, + "grad_norm": 2.461491617666625, + "learning_rate": 4.152554814761911e-07, + "loss": 1.0162, + "step": 38656 + }, + { + "epoch": 0.91, + "grad_norm": 1.9935097295365822, + "learning_rate": 4.1503790654327414e-07, + "loss": 0.8659, + "step": 38657 + }, + { + "epoch": 0.91, + "grad_norm": 2.551212266525731, + "learning_rate": 4.148203874170409e-07, + "loss": 0.9505, + "step": 38658 + }, + { + "epoch": 0.91, + "grad_norm": 1.7455212151261599, + "learning_rate": 4.1460292409875815e-07, + "loss": 0.8584, + "step": 38659 + }, + { + "epoch": 0.91, + "grad_norm": 2.0472053875892855, + "learning_rate": 4.1438551658969484e-07, + "loss": 0.8754, + "step": 38660 + }, + { + "epoch": 0.91, + "grad_norm": 2.6379305045908583, + "learning_rate": 4.1416816489111e-07, + "loss": 0.9896, + "step": 38661 + }, + { + "epoch": 0.91, + "grad_norm": 2.076654296459072, + "learning_rate": 4.139508690042737e-07, + "loss": 0.8674, + "step": 38662 + }, + { + "epoch": 0.91, + "grad_norm": 2.0031780783348396, + "learning_rate": 4.137336289304517e-07, + "loss": 0.9519, + "step": 38663 + }, + { + "epoch": 0.91, + "grad_norm": 2.2074468531517626, + "learning_rate": 4.1351644467090614e-07, + "loss": 0.9406, + "step": 38664 + }, + { + "epoch": 0.91, + "grad_norm": 1.9437851910758501, + "learning_rate": 4.132993162269039e-07, + "loss": 0.8403, + "step": 38665 + }, + { + "epoch": 0.91, + "grad_norm": 2.014882430071035, + "learning_rate": 4.130822435997073e-07, + "loss": 0.8916, + "step": 38666 + }, + { + "epoch": 0.91, + "grad_norm": 1.8844719722256502, + "learning_rate": 4.128652267905819e-07, + "loss": 0.9905, + "step": 38667 + }, + { + "epoch": 0.91, + "grad_norm": 2.4136606849587587, + "learning_rate": 4.126482658007891e-07, + "loss": 0.9942, + "step": 38668 + }, + { + "epoch": 0.91, + "grad_norm": 1.9549922611052988, + "learning_rate": 4.124313606315955e-07, + "loss": 1.0025, + "step": 38669 + }, + { + "epoch": 0.91, + "grad_norm": 1.0568273325590765, + "learning_rate": 4.1221451128426014e-07, + "loss": 0.9247, + "step": 38670 + }, + { + "epoch": 0.91, + "grad_norm": 1.9268912500814068, + "learning_rate": 4.119977177600476e-07, + "loss": 0.936, + "step": 38671 + }, + { + "epoch": 0.91, + "grad_norm": 2.008355091433242, + "learning_rate": 4.117809800602202e-07, + "loss": 0.8756, + "step": 38672 + }, + { + "epoch": 0.91, + "grad_norm": 2.021666427951545, + "learning_rate": 4.11564298186039e-07, + "loss": 1.0097, + "step": 38673 + }, + { + "epoch": 0.91, + "grad_norm": 2.3222544342385776, + "learning_rate": 4.113476721387677e-07, + "loss": 0.9498, + "step": 38674 + }, + { + "epoch": 0.91, + "grad_norm": 1.8335483332784561, + "learning_rate": 4.11131101919664e-07, + "loss": 1.0123, + "step": 38675 + }, + { + "epoch": 0.91, + "grad_norm": 2.1560200674124275, + "learning_rate": 4.1091458752999136e-07, + "loss": 0.9936, + "step": 38676 + }, + { + "epoch": 0.91, + "grad_norm": 1.9881375764197864, + "learning_rate": 4.106981289710099e-07, + "loss": 0.874, + "step": 38677 + }, + { + "epoch": 0.91, + "grad_norm": 2.0659427614286408, + "learning_rate": 4.104817262439809e-07, + "loss": 0.9944, + "step": 38678 + }, + { + "epoch": 0.91, + "grad_norm": 2.264744091296235, + "learning_rate": 4.102653793501621e-07, + "loss": 1.0325, + "step": 38679 + }, + { + "epoch": 0.91, + "grad_norm": 1.81147697991953, + "learning_rate": 4.100490882908137e-07, + "loss": 0.9274, + "step": 38680 + }, + { + "epoch": 0.91, + "grad_norm": 2.1889602484304884, + "learning_rate": 4.098328530671969e-07, + "loss": 1.0637, + "step": 38681 + }, + { + "epoch": 0.91, + "grad_norm": 3.276262940853258, + "learning_rate": 4.0961667368056957e-07, + "loss": 0.9119, + "step": 38682 + }, + { + "epoch": 0.91, + "grad_norm": 2.0650655186433275, + "learning_rate": 4.094005501321896e-07, + "loss": 1.0209, + "step": 38683 + }, + { + "epoch": 0.91, + "grad_norm": 1.0837497662423483, + "learning_rate": 4.09184482423316e-07, + "loss": 0.9455, + "step": 38684 + }, + { + "epoch": 0.91, + "grad_norm": 1.091265327227447, + "learning_rate": 4.089684705552066e-07, + "loss": 0.9821, + "step": 38685 + }, + { + "epoch": 0.91, + "grad_norm": 1.8402228018089668, + "learning_rate": 4.087525145291205e-07, + "loss": 0.9545, + "step": 38686 + }, + { + "epoch": 0.91, + "grad_norm": 1.7194045996532847, + "learning_rate": 4.085366143463121e-07, + "loss": 0.7478, + "step": 38687 + }, + { + "epoch": 0.91, + "grad_norm": 1.9294123825705753, + "learning_rate": 4.083207700080416e-07, + "loss": 0.9642, + "step": 38688 + }, + { + "epoch": 0.91, + "grad_norm": 1.9895891892405138, + "learning_rate": 4.0810498151556246e-07, + "loss": 0.9289, + "step": 38689 + }, + { + "epoch": 0.91, + "grad_norm": 2.4055587279535167, + "learning_rate": 4.078892488701347e-07, + "loss": 0.8859, + "step": 38690 + }, + { + "epoch": 0.91, + "grad_norm": 2.353037586969584, + "learning_rate": 4.0767357207301073e-07, + "loss": 0.9449, + "step": 38691 + }, + { + "epoch": 0.91, + "grad_norm": 2.109282481021136, + "learning_rate": 4.074579511254495e-07, + "loss": 1.0148, + "step": 38692 + }, + { + "epoch": 0.91, + "grad_norm": 1.9310468527944673, + "learning_rate": 4.072423860287045e-07, + "loss": 0.8513, + "step": 38693 + }, + { + "epoch": 0.91, + "grad_norm": 1.9148806818198543, + "learning_rate": 4.070268767840324e-07, + "loss": 0.9133, + "step": 38694 + }, + { + "epoch": 0.91, + "grad_norm": 2.7351964189350912, + "learning_rate": 4.0681142339268564e-07, + "loss": 0.8909, + "step": 38695 + }, + { + "epoch": 0.91, + "grad_norm": 1.914697055280618, + "learning_rate": 4.0659602585591984e-07, + "loss": 1.0683, + "step": 38696 + }, + { + "epoch": 0.91, + "grad_norm": 2.105960035111894, + "learning_rate": 4.0638068417498955e-07, + "loss": 0.8954, + "step": 38697 + }, + { + "epoch": 0.91, + "grad_norm": 1.9115791558577528, + "learning_rate": 4.0616539835114823e-07, + "loss": 0.9693, + "step": 38698 + }, + { + "epoch": 0.91, + "grad_norm": 2.106413882296959, + "learning_rate": 4.059501683856504e-07, + "loss": 1.0752, + "step": 38699 + }, + { + "epoch": 0.91, + "grad_norm": 2.1588495230476212, + "learning_rate": 4.0573499427974616e-07, + "loss": 1.047, + "step": 38700 + }, + { + "epoch": 0.91, + "grad_norm": 1.904774544912947, + "learning_rate": 4.055198760346901e-07, + "loss": 0.9142, + "step": 38701 + }, + { + "epoch": 0.91, + "grad_norm": 2.0317372556086535, + "learning_rate": 4.053048136517368e-07, + "loss": 0.9995, + "step": 38702 + }, + { + "epoch": 0.91, + "grad_norm": 1.745411828778275, + "learning_rate": 4.050898071321363e-07, + "loss": 0.8458, + "step": 38703 + }, + { + "epoch": 0.91, + "grad_norm": 1.9585046177157637, + "learning_rate": 4.048748564771399e-07, + "loss": 1.0596, + "step": 38704 + }, + { + "epoch": 0.91, + "grad_norm": 2.317874056326221, + "learning_rate": 4.046599616879987e-07, + "loss": 0.9728, + "step": 38705 + }, + { + "epoch": 0.91, + "grad_norm": 1.8936766362133342, + "learning_rate": 4.044451227659674e-07, + "loss": 0.9346, + "step": 38706 + }, + { + "epoch": 0.91, + "grad_norm": 2.027066749440574, + "learning_rate": 4.0423033971229375e-07, + "loss": 0.9921, + "step": 38707 + }, + { + "epoch": 0.91, + "grad_norm": 2.0793303721128003, + "learning_rate": 4.0401561252822796e-07, + "loss": 0.9349, + "step": 38708 + }, + { + "epoch": 0.91, + "grad_norm": 1.986810527749692, + "learning_rate": 4.0380094121502236e-07, + "loss": 0.9282, + "step": 38709 + }, + { + "epoch": 0.91, + "grad_norm": 2.29112891055739, + "learning_rate": 4.035863257739248e-07, + "loss": 0.9445, + "step": 38710 + }, + { + "epoch": 0.91, + "grad_norm": 1.973755541755396, + "learning_rate": 4.0337176620618756e-07, + "loss": 0.9706, + "step": 38711 + }, + { + "epoch": 0.91, + "grad_norm": 1.9341727293051731, + "learning_rate": 4.031572625130564e-07, + "loss": 0.926, + "step": 38712 + }, + { + "epoch": 0.91, + "grad_norm": 2.1329197347934254, + "learning_rate": 4.0294281469578364e-07, + "loss": 0.9463, + "step": 38713 + }, + { + "epoch": 0.91, + "grad_norm": 1.071490323209784, + "learning_rate": 4.0272842275561385e-07, + "loss": 0.9137, + "step": 38714 + }, + { + "epoch": 0.91, + "grad_norm": 2.0063293130216677, + "learning_rate": 4.025140866937993e-07, + "loss": 1.1087, + "step": 38715 + }, + { + "epoch": 0.91, + "grad_norm": 1.9929122262236236, + "learning_rate": 4.0229980651158686e-07, + "loss": 1.0096, + "step": 38716 + }, + { + "epoch": 0.91, + "grad_norm": 1.8565024074160594, + "learning_rate": 4.020855822102221e-07, + "loss": 0.9776, + "step": 38717 + }, + { + "epoch": 0.91, + "grad_norm": 2.037123640404564, + "learning_rate": 4.0187141379095294e-07, + "loss": 0.9569, + "step": 38718 + }, + { + "epoch": 0.91, + "grad_norm": 1.9210963255152167, + "learning_rate": 4.0165730125502956e-07, + "loss": 0.9782, + "step": 38719 + }, + { + "epoch": 0.91, + "grad_norm": 2.447125511161903, + "learning_rate": 4.0144324460369535e-07, + "loss": 0.8813, + "step": 38720 + }, + { + "epoch": 0.91, + "grad_norm": 1.971675814817989, + "learning_rate": 4.012292438381971e-07, + "loss": 0.9951, + "step": 38721 + }, + { + "epoch": 0.91, + "grad_norm": 2.462473193376124, + "learning_rate": 4.010152989597804e-07, + "loss": 0.8609, + "step": 38722 + }, + { + "epoch": 0.91, + "grad_norm": 1.975799069502423, + "learning_rate": 4.008014099696922e-07, + "loss": 0.9951, + "step": 38723 + }, + { + "epoch": 0.91, + "grad_norm": 1.853515660926087, + "learning_rate": 4.0058757686917805e-07, + "loss": 0.9403, + "step": 38724 + }, + { + "epoch": 0.91, + "grad_norm": 2.0123784892530248, + "learning_rate": 4.003737996594803e-07, + "loss": 0.9036, + "step": 38725 + }, + { + "epoch": 0.91, + "grad_norm": 1.9091492989504582, + "learning_rate": 4.001600783418469e-07, + "loss": 1.0103, + "step": 38726 + }, + { + "epoch": 0.91, + "grad_norm": 2.3712106336441168, + "learning_rate": 3.999464129175201e-07, + "loss": 0.9273, + "step": 38727 + }, + { + "epoch": 0.91, + "grad_norm": 2.8090612225729084, + "learning_rate": 3.997328033877457e-07, + "loss": 0.8827, + "step": 38728 + }, + { + "epoch": 0.91, + "grad_norm": 1.0728158671378405, + "learning_rate": 3.995192497537659e-07, + "loss": 0.9067, + "step": 38729 + }, + { + "epoch": 0.91, + "grad_norm": 2.003362410349325, + "learning_rate": 3.993057520168242e-07, + "loss": 1.0722, + "step": 38730 + }, + { + "epoch": 0.91, + "grad_norm": 2.3065695298460454, + "learning_rate": 3.99092310178163e-07, + "loss": 0.9157, + "step": 38731 + }, + { + "epoch": 0.91, + "grad_norm": 2.121996527414698, + "learning_rate": 3.988789242390279e-07, + "loss": 0.9174, + "step": 38732 + }, + { + "epoch": 0.91, + "grad_norm": 2.287732562175722, + "learning_rate": 3.986655942006579e-07, + "loss": 0.9423, + "step": 38733 + }, + { + "epoch": 0.91, + "grad_norm": 1.9518387308136431, + "learning_rate": 3.9845232006429755e-07, + "loss": 1.0107, + "step": 38734 + }, + { + "epoch": 0.91, + "grad_norm": 2.150915118831662, + "learning_rate": 3.98239101831186e-07, + "loss": 1.0151, + "step": 38735 + }, + { + "epoch": 0.91, + "grad_norm": 1.9795968035608098, + "learning_rate": 3.9802593950256875e-07, + "loss": 1.1122, + "step": 38736 + }, + { + "epoch": 0.91, + "grad_norm": 1.9409390682297192, + "learning_rate": 3.9781283307968266e-07, + "loss": 0.9571, + "step": 38737 + }, + { + "epoch": 0.91, + "grad_norm": 1.8560324294529842, + "learning_rate": 3.9759978256377006e-07, + "loss": 0.9608, + "step": 38738 + }, + { + "epoch": 0.91, + "grad_norm": 2.020364761589325, + "learning_rate": 3.973867879560722e-07, + "loss": 0.9429, + "step": 38739 + }, + { + "epoch": 0.91, + "grad_norm": 1.105205679820498, + "learning_rate": 3.9717384925782697e-07, + "loss": 0.9352, + "step": 38740 + }, + { + "epoch": 0.91, + "grad_norm": 2.0139313774803966, + "learning_rate": 3.9696096647028004e-07, + "loss": 0.9694, + "step": 38741 + }, + { + "epoch": 0.91, + "grad_norm": 2.116148869258211, + "learning_rate": 3.9674813959466263e-07, + "loss": 1.0414, + "step": 38742 + }, + { + "epoch": 0.91, + "grad_norm": 2.041944303457395, + "learning_rate": 3.965353686322193e-07, + "loss": 1.0045, + "step": 38743 + }, + { + "epoch": 0.91, + "grad_norm": 2.085926546780621, + "learning_rate": 3.9632265358418796e-07, + "loss": 0.9146, + "step": 38744 + }, + { + "epoch": 0.91, + "grad_norm": 1.8650512190964, + "learning_rate": 3.961099944518065e-07, + "loss": 0.8816, + "step": 38745 + }, + { + "epoch": 0.91, + "grad_norm": 1.0699781521907068, + "learning_rate": 3.9589739123631177e-07, + "loss": 0.8989, + "step": 38746 + }, + { + "epoch": 0.91, + "grad_norm": 2.457307441311006, + "learning_rate": 3.956848439389427e-07, + "loss": 0.8626, + "step": 38747 + }, + { + "epoch": 0.91, + "grad_norm": 2.733526158049098, + "learning_rate": 3.954723525609394e-07, + "loss": 0.984, + "step": 38748 + }, + { + "epoch": 0.91, + "grad_norm": 2.1364147661001605, + "learning_rate": 3.952599171035365e-07, + "loss": 0.9753, + "step": 38749 + }, + { + "epoch": 0.91, + "grad_norm": 2.0537502509076053, + "learning_rate": 3.950475375679697e-07, + "loss": 1.0152, + "step": 38750 + }, + { + "epoch": 0.91, + "grad_norm": 2.3947414203668544, + "learning_rate": 3.9483521395547895e-07, + "loss": 0.8419, + "step": 38751 + }, + { + "epoch": 0.91, + "grad_norm": 2.3959516104647687, + "learning_rate": 3.946229462672957e-07, + "loss": 0.979, + "step": 38752 + }, + { + "epoch": 0.91, + "grad_norm": 2.1703981373457366, + "learning_rate": 3.9441073450465997e-07, + "loss": 1.0822, + "step": 38753 + }, + { + "epoch": 0.91, + "grad_norm": 1.1590838021067111, + "learning_rate": 3.941985786688074e-07, + "loss": 0.9342, + "step": 38754 + }, + { + "epoch": 0.91, + "grad_norm": 2.096034729898344, + "learning_rate": 3.939864787609715e-07, + "loss": 0.9833, + "step": 38755 + }, + { + "epoch": 0.91, + "grad_norm": 1.7691688001926067, + "learning_rate": 3.9377443478238575e-07, + "loss": 0.8454, + "step": 38756 + }, + { + "epoch": 0.91, + "grad_norm": 2.4991578826761884, + "learning_rate": 3.935624467342869e-07, + "loss": 0.8853, + "step": 38757 + }, + { + "epoch": 0.91, + "grad_norm": 1.828127712264046, + "learning_rate": 3.933505146179117e-07, + "loss": 0.8933, + "step": 38758 + }, + { + "epoch": 0.91, + "grad_norm": 2.1468573640801063, + "learning_rate": 3.9313863843448817e-07, + "loss": 0.8928, + "step": 38759 + }, + { + "epoch": 0.91, + "grad_norm": 2.037105516467093, + "learning_rate": 3.92926818185253e-07, + "loss": 1.0986, + "step": 38760 + }, + { + "epoch": 0.91, + "grad_norm": 1.9571894940531032, + "learning_rate": 3.927150538714397e-07, + "loss": 0.8965, + "step": 38761 + }, + { + "epoch": 0.91, + "grad_norm": 1.940212758803354, + "learning_rate": 3.925033454942828e-07, + "loss": 0.8377, + "step": 38762 + }, + { + "epoch": 0.91, + "grad_norm": 2.9405648211698496, + "learning_rate": 3.922916930550125e-07, + "loss": 0.9135, + "step": 38763 + }, + { + "epoch": 0.91, + "grad_norm": 2.0187785042300117, + "learning_rate": 3.920800965548599e-07, + "loss": 0.941, + "step": 38764 + }, + { + "epoch": 0.91, + "grad_norm": 2.009731986556274, + "learning_rate": 3.9186855599506077e-07, + "loss": 0.9383, + "step": 38765 + }, + { + "epoch": 0.91, + "grad_norm": 2.728312723586305, + "learning_rate": 3.916570713768442e-07, + "loss": 0.9066, + "step": 38766 + }, + { + "epoch": 0.91, + "grad_norm": 1.9946966095573606, + "learning_rate": 3.914456427014424e-07, + "loss": 1.066, + "step": 38767 + }, + { + "epoch": 0.91, + "grad_norm": 2.1046296320417235, + "learning_rate": 3.912342699700844e-07, + "loss": 0.8514, + "step": 38768 + }, + { + "epoch": 0.91, + "grad_norm": 1.8079422710465336, + "learning_rate": 3.9102295318400487e-07, + "loss": 1.0156, + "step": 38769 + }, + { + "epoch": 0.91, + "grad_norm": 2.090280940158368, + "learning_rate": 3.908116923444294e-07, + "loss": 0.9424, + "step": 38770 + }, + { + "epoch": 0.91, + "grad_norm": 2.1231550141538036, + "learning_rate": 3.906004874525926e-07, + "loss": 0.8173, + "step": 38771 + }, + { + "epoch": 0.91, + "grad_norm": 2.347571791178511, + "learning_rate": 3.903893385097224e-07, + "loss": 0.9958, + "step": 38772 + }, + { + "epoch": 0.91, + "grad_norm": 2.025885884851502, + "learning_rate": 3.9017824551704554e-07, + "loss": 0.9005, + "step": 38773 + }, + { + "epoch": 0.91, + "grad_norm": 2.304157856418207, + "learning_rate": 3.8996720847579437e-07, + "loss": 1.0159, + "step": 38774 + }, + { + "epoch": 0.91, + "grad_norm": 2.3554847771698735, + "learning_rate": 3.8975622738719685e-07, + "loss": 0.9925, + "step": 38775 + }, + { + "epoch": 0.91, + "grad_norm": 1.9815659787082531, + "learning_rate": 3.89545302252482e-07, + "loss": 0.9924, + "step": 38776 + }, + { + "epoch": 0.91, + "grad_norm": 2.2056469841350546, + "learning_rate": 3.893344330728754e-07, + "loss": 0.9451, + "step": 38777 + }, + { + "epoch": 0.91, + "grad_norm": 1.8844435490765465, + "learning_rate": 3.891236198496062e-07, + "loss": 0.8791, + "step": 38778 + }, + { + "epoch": 0.91, + "grad_norm": 1.978214473766762, + "learning_rate": 3.889128625839034e-07, + "loss": 0.9627, + "step": 38779 + }, + { + "epoch": 0.91, + "grad_norm": 2.058720893299447, + "learning_rate": 3.887021612769937e-07, + "loss": 0.8731, + "step": 38780 + }, + { + "epoch": 0.91, + "grad_norm": 2.0843804541887145, + "learning_rate": 3.884915159301006e-07, + "loss": 1.0908, + "step": 38781 + }, + { + "epoch": 0.91, + "grad_norm": 1.125908627698049, + "learning_rate": 3.8828092654445314e-07, + "loss": 0.9332, + "step": 38782 + }, + { + "epoch": 0.91, + "grad_norm": 1.9947168100894255, + "learning_rate": 3.8807039312127813e-07, + "loss": 0.842, + "step": 38783 + }, + { + "epoch": 0.91, + "grad_norm": 2.1368588650292106, + "learning_rate": 3.8785991566180017e-07, + "loss": 1.073, + "step": 38784 + }, + { + "epoch": 0.91, + "grad_norm": 2.509129227299148, + "learning_rate": 3.876494941672437e-07, + "loss": 0.9437, + "step": 38785 + }, + { + "epoch": 0.91, + "grad_norm": 1.959910708999314, + "learning_rate": 3.874391286388368e-07, + "loss": 1.0036, + "step": 38786 + }, + { + "epoch": 0.91, + "grad_norm": 2.3228273334007734, + "learning_rate": 3.8722881907780173e-07, + "loss": 0.9353, + "step": 38787 + }, + { + "epoch": 0.91, + "grad_norm": 1.0546664934582997, + "learning_rate": 3.870185654853642e-07, + "loss": 0.8763, + "step": 38788 + }, + { + "epoch": 0.91, + "grad_norm": 1.878397704920732, + "learning_rate": 3.8680836786274764e-07, + "loss": 0.9726, + "step": 38789 + }, + { + "epoch": 0.91, + "grad_norm": 1.9312159941319718, + "learning_rate": 3.865982262111767e-07, + "loss": 0.9442, + "step": 38790 + }, + { + "epoch": 0.91, + "grad_norm": 2.028421754141584, + "learning_rate": 3.863881405318737e-07, + "loss": 0.9437, + "step": 38791 + }, + { + "epoch": 0.91, + "grad_norm": 2.354859568889395, + "learning_rate": 3.8617811082606317e-07, + "loss": 0.8447, + "step": 38792 + }, + { + "epoch": 0.91, + "grad_norm": 2.1464132608707547, + "learning_rate": 3.859681370949664e-07, + "loss": 0.9166, + "step": 38793 + }, + { + "epoch": 0.91, + "grad_norm": 2.1919958552639414, + "learning_rate": 3.8575821933980906e-07, + "loss": 0.9947, + "step": 38794 + }, + { + "epoch": 0.91, + "grad_norm": 2.4287596031623444, + "learning_rate": 3.855483575618091e-07, + "loss": 0.9908, + "step": 38795 + }, + { + "epoch": 0.91, + "grad_norm": 1.8427516308521692, + "learning_rate": 3.85338551762191e-07, + "loss": 1.0111, + "step": 38796 + }, + { + "epoch": 0.91, + "grad_norm": 1.1031884648413217, + "learning_rate": 3.851288019421773e-07, + "loss": 0.9186, + "step": 38797 + }, + { + "epoch": 0.91, + "grad_norm": 2.132900016599532, + "learning_rate": 3.8491910810298574e-07, + "loss": 0.8216, + "step": 38798 + }, + { + "epoch": 0.91, + "grad_norm": 2.400125901955916, + "learning_rate": 3.847094702458387e-07, + "loss": 1.0252, + "step": 38799 + }, + { + "epoch": 0.91, + "grad_norm": 1.080090801637148, + "learning_rate": 3.8449988837195975e-07, + "loss": 0.9117, + "step": 38800 + }, + { + "epoch": 0.91, + "grad_norm": 2.538239581010285, + "learning_rate": 3.842903624825656e-07, + "loss": 1.0165, + "step": 38801 + }, + { + "epoch": 0.91, + "grad_norm": 1.872578878553951, + "learning_rate": 3.840808925788764e-07, + "loss": 0.9349, + "step": 38802 + }, + { + "epoch": 0.91, + "grad_norm": 2.210039143204711, + "learning_rate": 3.838714786621123e-07, + "loss": 0.9723, + "step": 38803 + }, + { + "epoch": 0.91, + "grad_norm": 1.0784856865206527, + "learning_rate": 3.8366212073349453e-07, + "loss": 0.9355, + "step": 38804 + }, + { + "epoch": 0.91, + "grad_norm": 2.0424998245079045, + "learning_rate": 3.834528187942399e-07, + "loss": 1.051, + "step": 38805 + }, + { + "epoch": 0.91, + "grad_norm": 1.8943943255633826, + "learning_rate": 3.8324357284556744e-07, + "loss": 1.017, + "step": 38806 + }, + { + "epoch": 0.91, + "grad_norm": 1.0853868635828465, + "learning_rate": 3.8303438288869397e-07, + "loss": 0.9482, + "step": 38807 + }, + { + "epoch": 0.91, + "grad_norm": 2.25377656324507, + "learning_rate": 3.828252489248407e-07, + "loss": 0.911, + "step": 38808 + }, + { + "epoch": 0.91, + "grad_norm": 2.6278483821400322, + "learning_rate": 3.826161709552245e-07, + "loss": 0.9288, + "step": 38809 + }, + { + "epoch": 0.91, + "grad_norm": 2.3784415294402015, + "learning_rate": 3.824071489810599e-07, + "loss": 0.8874, + "step": 38810 + }, + { + "epoch": 0.91, + "grad_norm": 1.895449375279172, + "learning_rate": 3.8219818300356706e-07, + "loss": 0.9352, + "step": 38811 + }, + { + "epoch": 0.91, + "grad_norm": 2.7537624806049266, + "learning_rate": 3.819892730239594e-07, + "loss": 0.9439, + "step": 38812 + }, + { + "epoch": 0.91, + "grad_norm": 1.7336164902826177, + "learning_rate": 3.817804190434571e-07, + "loss": 0.8161, + "step": 38813 + }, + { + "epoch": 0.91, + "grad_norm": 1.9034499706028267, + "learning_rate": 3.815716210632736e-07, + "loss": 0.8821, + "step": 38814 + }, + { + "epoch": 0.91, + "grad_norm": 2.036472989214155, + "learning_rate": 3.813628790846258e-07, + "loss": 0.9325, + "step": 38815 + }, + { + "epoch": 0.91, + "grad_norm": 2.0456585761695987, + "learning_rate": 3.81154193108727e-07, + "loss": 1.0376, + "step": 38816 + }, + { + "epoch": 0.91, + "grad_norm": 1.9627099697167616, + "learning_rate": 3.809455631367953e-07, + "loss": 0.8876, + "step": 38817 + }, + { + "epoch": 0.91, + "grad_norm": 1.168123679705984, + "learning_rate": 3.807369891700441e-07, + "loss": 0.9915, + "step": 38818 + }, + { + "epoch": 0.91, + "grad_norm": 1.7811467621556507, + "learning_rate": 3.805284712096857e-07, + "loss": 0.9272, + "step": 38819 + }, + { + "epoch": 0.91, + "grad_norm": 1.1421216192346855, + "learning_rate": 3.80320009256937e-07, + "loss": 0.9501, + "step": 38820 + }, + { + "epoch": 0.91, + "grad_norm": 1.9902569139681185, + "learning_rate": 3.8011160331301147e-07, + "loss": 0.967, + "step": 38821 + }, + { + "epoch": 0.91, + "grad_norm": 1.9513969931122144, + "learning_rate": 3.7990325337912136e-07, + "loss": 1.0098, + "step": 38822 + }, + { + "epoch": 0.91, + "grad_norm": 1.840548356859676, + "learning_rate": 3.7969495945647915e-07, + "loss": 0.915, + "step": 38823 + }, + { + "epoch": 0.91, + "grad_norm": 2.1431694143983373, + "learning_rate": 3.794867215462994e-07, + "loss": 0.9861, + "step": 38824 + }, + { + "epoch": 0.91, + "grad_norm": 1.8176644095307626, + "learning_rate": 3.7927853964979443e-07, + "loss": 1.0233, + "step": 38825 + }, + { + "epoch": 0.91, + "grad_norm": 2.198238724178787, + "learning_rate": 3.7907041376817554e-07, + "loss": 1.0008, + "step": 38826 + }, + { + "epoch": 0.91, + "grad_norm": 3.438259819570071, + "learning_rate": 3.7886234390265286e-07, + "loss": 0.861, + "step": 38827 + }, + { + "epoch": 0.91, + "grad_norm": 2.069548727574104, + "learning_rate": 3.78654330054441e-07, + "loss": 0.9123, + "step": 38828 + }, + { + "epoch": 0.91, + "grad_norm": 1.9885098082632657, + "learning_rate": 3.7844637222475e-07, + "loss": 1.0053, + "step": 38829 + }, + { + "epoch": 0.91, + "grad_norm": 1.9569624493263458, + "learning_rate": 3.782384704147901e-07, + "loss": 0.9772, + "step": 38830 + }, + { + "epoch": 0.91, + "grad_norm": 2.5658290228776055, + "learning_rate": 3.7803062462577147e-07, + "loss": 0.8862, + "step": 38831 + }, + { + "epoch": 0.91, + "grad_norm": 1.9337643125474215, + "learning_rate": 3.778228348589064e-07, + "loss": 0.8844, + "step": 38832 + }, + { + "epoch": 0.91, + "grad_norm": 2.145162505233213, + "learning_rate": 3.7761510111540054e-07, + "loss": 1.0751, + "step": 38833 + }, + { + "epoch": 0.91, + "grad_norm": 2.160747810955742, + "learning_rate": 3.774074233964686e-07, + "loss": 0.8741, + "step": 38834 + }, + { + "epoch": 0.91, + "grad_norm": 1.9987896667603795, + "learning_rate": 3.7719980170331516e-07, + "loss": 0.8319, + "step": 38835 + }, + { + "epoch": 0.91, + "grad_norm": 1.056381590905498, + "learning_rate": 3.769922360371525e-07, + "loss": 0.9757, + "step": 38836 + }, + { + "epoch": 0.91, + "grad_norm": 2.0010573437513357, + "learning_rate": 3.767847263991853e-07, + "loss": 0.9536, + "step": 38837 + }, + { + "epoch": 0.91, + "grad_norm": 1.8645375996241946, + "learning_rate": 3.7657727279062585e-07, + "loss": 1.0264, + "step": 38838 + }, + { + "epoch": 0.92, + "grad_norm": 1.902878992033717, + "learning_rate": 3.7636987521268097e-07, + "loss": 0.8665, + "step": 38839 + }, + { + "epoch": 0.92, + "grad_norm": 1.084915563414001, + "learning_rate": 3.761625336665564e-07, + "loss": 0.9507, + "step": 38840 + }, + { + "epoch": 0.92, + "grad_norm": 1.6237093611881483, + "learning_rate": 3.759552481534601e-07, + "loss": 0.8527, + "step": 38841 + }, + { + "epoch": 0.92, + "grad_norm": 2.1855377799727815, + "learning_rate": 3.7574801867459877e-07, + "loss": 1.0266, + "step": 38842 + }, + { + "epoch": 0.92, + "grad_norm": 1.903067529449512, + "learning_rate": 3.755408452311826e-07, + "loss": 1.0253, + "step": 38843 + }, + { + "epoch": 0.92, + "grad_norm": 1.9367260059298284, + "learning_rate": 3.753337278244118e-07, + "loss": 0.9566, + "step": 38844 + }, + { + "epoch": 0.92, + "grad_norm": 1.9313807858747825, + "learning_rate": 3.7512666645549533e-07, + "loss": 0.9588, + "step": 38845 + }, + { + "epoch": 0.92, + "grad_norm": 1.8503624635199867, + "learning_rate": 3.7491966112563893e-07, + "loss": 0.9955, + "step": 38846 + }, + { + "epoch": 0.92, + "grad_norm": 1.1026956879467409, + "learning_rate": 3.747127118360483e-07, + "loss": 0.9734, + "step": 38847 + }, + { + "epoch": 0.92, + "grad_norm": 1.9709473622564344, + "learning_rate": 3.745058185879269e-07, + "loss": 1.1078, + "step": 38848 + }, + { + "epoch": 0.92, + "grad_norm": 2.0924787901365436, + "learning_rate": 3.7429898138247933e-07, + "loss": 0.9145, + "step": 38849 + }, + { + "epoch": 0.92, + "grad_norm": 2.018989376385483, + "learning_rate": 3.740922002209113e-07, + "loss": 1.0282, + "step": 38850 + }, + { + "epoch": 0.92, + "grad_norm": 1.839225990182762, + "learning_rate": 3.738854751044252e-07, + "loss": 0.9863, + "step": 38851 + }, + { + "epoch": 0.92, + "grad_norm": 2.300979286460504, + "learning_rate": 3.736788060342267e-07, + "loss": 1.0042, + "step": 38852 + }, + { + "epoch": 0.92, + "grad_norm": 1.0416011482113128, + "learning_rate": 3.7347219301151815e-07, + "loss": 0.9198, + "step": 38853 + }, + { + "epoch": 0.92, + "grad_norm": 1.8708214968660322, + "learning_rate": 3.7326563603750085e-07, + "loss": 0.9688, + "step": 38854 + }, + { + "epoch": 0.92, + "grad_norm": 1.9990762919989216, + "learning_rate": 3.7305913511337834e-07, + "loss": 0.9972, + "step": 38855 + }, + { + "epoch": 0.92, + "grad_norm": 2.128739045952699, + "learning_rate": 3.728526902403551e-07, + "loss": 0.9478, + "step": 38856 + }, + { + "epoch": 0.92, + "grad_norm": 1.934700746233562, + "learning_rate": 3.726463014196313e-07, + "loss": 0.8883, + "step": 38857 + }, + { + "epoch": 0.92, + "grad_norm": 1.995038968334719, + "learning_rate": 3.724399686524072e-07, + "loss": 1.1431, + "step": 38858 + }, + { + "epoch": 0.92, + "grad_norm": 1.8881904575044597, + "learning_rate": 3.722336919398861e-07, + "loss": 0.9997, + "step": 38859 + }, + { + "epoch": 0.92, + "grad_norm": 2.092966689587227, + "learning_rate": 3.7202747128326943e-07, + "loss": 1.0028, + "step": 38860 + }, + { + "epoch": 0.92, + "grad_norm": 1.8020523051592057, + "learning_rate": 3.7182130668375726e-07, + "loss": 1.0414, + "step": 38861 + }, + { + "epoch": 0.92, + "grad_norm": 2.1086706099127097, + "learning_rate": 3.7161519814254864e-07, + "loss": 0.9624, + "step": 38862 + }, + { + "epoch": 0.92, + "grad_norm": 1.886286979216867, + "learning_rate": 3.7140914566084376e-07, + "loss": 0.9079, + "step": 38863 + }, + { + "epoch": 0.92, + "grad_norm": 2.101907511501808, + "learning_rate": 3.712031492398449e-07, + "loss": 1.0224, + "step": 38864 + }, + { + "epoch": 0.92, + "grad_norm": 1.089165924985267, + "learning_rate": 3.7099720888075006e-07, + "loss": 1.0082, + "step": 38865 + }, + { + "epoch": 0.92, + "grad_norm": 1.8492234910964422, + "learning_rate": 3.7079132458475606e-07, + "loss": 0.8028, + "step": 38866 + }, + { + "epoch": 0.92, + "grad_norm": 2.7467350834906026, + "learning_rate": 3.705854963530653e-07, + "loss": 1.0079, + "step": 38867 + }, + { + "epoch": 0.92, + "grad_norm": 1.9722159007500772, + "learning_rate": 3.703797241868734e-07, + "loss": 0.9841, + "step": 38868 + }, + { + "epoch": 0.92, + "grad_norm": 1.8653999944126387, + "learning_rate": 3.7017400808737946e-07, + "loss": 0.9733, + "step": 38869 + }, + { + "epoch": 0.92, + "grad_norm": 2.0209155767115314, + "learning_rate": 3.699683480557814e-07, + "loss": 0.9836, + "step": 38870 + }, + { + "epoch": 0.92, + "grad_norm": 1.0863690420400478, + "learning_rate": 3.6976274409327717e-07, + "loss": 0.8931, + "step": 38871 + }, + { + "epoch": 0.92, + "grad_norm": 1.9070467397802273, + "learning_rate": 3.6955719620106135e-07, + "loss": 0.9819, + "step": 38872 + }, + { + "epoch": 0.92, + "grad_norm": 2.07157538151329, + "learning_rate": 3.693517043803341e-07, + "loss": 1.0929, + "step": 38873 + }, + { + "epoch": 0.92, + "grad_norm": 2.063124440005529, + "learning_rate": 3.6914626863229e-07, + "loss": 0.964, + "step": 38874 + }, + { + "epoch": 0.92, + "grad_norm": 1.1247200261529808, + "learning_rate": 3.689408889581247e-07, + "loss": 0.9381, + "step": 38875 + }, + { + "epoch": 0.92, + "grad_norm": 1.925480370177116, + "learning_rate": 3.687355653590341e-07, + "loss": 0.9732, + "step": 38876 + }, + { + "epoch": 0.92, + "grad_norm": 1.913179374003015, + "learning_rate": 3.6853029783621596e-07, + "loss": 1.0043, + "step": 38877 + }, + { + "epoch": 0.92, + "grad_norm": 1.9613746528856064, + "learning_rate": 3.6832508639086386e-07, + "loss": 0.9457, + "step": 38878 + }, + { + "epoch": 0.92, + "grad_norm": 2.374671017611802, + "learning_rate": 3.681199310241701e-07, + "loss": 1.0452, + "step": 38879 + }, + { + "epoch": 0.92, + "grad_norm": 2.0648819562373633, + "learning_rate": 3.6791483173733267e-07, + "loss": 1.0779, + "step": 38880 + }, + { + "epoch": 0.92, + "grad_norm": 2.1275499018306596, + "learning_rate": 3.67709788531545e-07, + "loss": 1.0028, + "step": 38881 + }, + { + "epoch": 0.92, + "grad_norm": 4.051120925905156, + "learning_rate": 3.675048014080007e-07, + "loss": 1.0726, + "step": 38882 + }, + { + "epoch": 0.92, + "grad_norm": 2.04193851160044, + "learning_rate": 3.6729987036789205e-07, + "loss": 1.0789, + "step": 38883 + }, + { + "epoch": 0.92, + "grad_norm": 2.1590317176597886, + "learning_rate": 3.6709499541241367e-07, + "loss": 1.0188, + "step": 38884 + }, + { + "epoch": 0.92, + "grad_norm": 1.1781666641985096, + "learning_rate": 3.6689017654275905e-07, + "loss": 0.9003, + "step": 38885 + }, + { + "epoch": 0.92, + "grad_norm": 1.8967088767557254, + "learning_rate": 3.6668541376011836e-07, + "loss": 0.9628, + "step": 38886 + }, + { + "epoch": 0.92, + "grad_norm": 2.134474242886494, + "learning_rate": 3.664807070656851e-07, + "loss": 0.9649, + "step": 38887 + }, + { + "epoch": 0.92, + "grad_norm": 2.0090033146174804, + "learning_rate": 3.6627605646065157e-07, + "loss": 0.996, + "step": 38888 + }, + { + "epoch": 0.92, + "grad_norm": 1.9311959719883875, + "learning_rate": 3.66071461946208e-07, + "loss": 0.9249, + "step": 38889 + }, + { + "epoch": 0.92, + "grad_norm": 2.311939269383888, + "learning_rate": 3.6586692352354793e-07, + "loss": 0.9275, + "step": 38890 + }, + { + "epoch": 0.92, + "grad_norm": 2.293176893477449, + "learning_rate": 3.656624411938592e-07, + "loss": 0.979, + "step": 38891 + }, + { + "epoch": 0.92, + "grad_norm": 1.9463997961666943, + "learning_rate": 3.6545801495833534e-07, + "loss": 0.9297, + "step": 38892 + }, + { + "epoch": 0.92, + "grad_norm": 1.1302493254326098, + "learning_rate": 3.6525364481816315e-07, + "loss": 0.9932, + "step": 38893 + }, + { + "epoch": 0.92, + "grad_norm": 2.083853310263995, + "learning_rate": 3.650493307745362e-07, + "loss": 1.0992, + "step": 38894 + }, + { + "epoch": 0.92, + "grad_norm": 2.0417511570619538, + "learning_rate": 3.6484507282864233e-07, + "loss": 0.9259, + "step": 38895 + }, + { + "epoch": 0.92, + "grad_norm": 2.062692453628761, + "learning_rate": 3.646408709816707e-07, + "loss": 1.0018, + "step": 38896 + }, + { + "epoch": 0.92, + "grad_norm": 2.366620145294883, + "learning_rate": 3.644367252348091e-07, + "loss": 1.0199, + "step": 38897 + }, + { + "epoch": 0.92, + "grad_norm": 1.9283164660942742, + "learning_rate": 3.6423263558924895e-07, + "loss": 0.9215, + "step": 38898 + }, + { + "epoch": 0.92, + "grad_norm": 2.70862345729577, + "learning_rate": 3.64028602046177e-07, + "loss": 0.9962, + "step": 38899 + }, + { + "epoch": 0.92, + "grad_norm": 1.9695483811752046, + "learning_rate": 3.6382462460678113e-07, + "loss": 0.865, + "step": 38900 + }, + { + "epoch": 0.92, + "grad_norm": 2.2870671692284055, + "learning_rate": 3.636207032722483e-07, + "loss": 0.9893, + "step": 38901 + }, + { + "epoch": 0.92, + "grad_norm": 2.018882675334949, + "learning_rate": 3.634168380437675e-07, + "loss": 1.0172, + "step": 38902 + }, + { + "epoch": 0.92, + "grad_norm": 2.695137600821781, + "learning_rate": 3.632130289225244e-07, + "loss": 0.9961, + "step": 38903 + }, + { + "epoch": 0.92, + "grad_norm": 1.9740400291200226, + "learning_rate": 3.630092759097059e-07, + "loss": 0.9971, + "step": 38904 + }, + { + "epoch": 0.92, + "grad_norm": 1.951723261889742, + "learning_rate": 3.628055790064977e-07, + "loss": 1.0746, + "step": 38905 + }, + { + "epoch": 0.92, + "grad_norm": 2.101567598517519, + "learning_rate": 3.6260193821408776e-07, + "loss": 0.9567, + "step": 38906 + }, + { + "epoch": 0.92, + "grad_norm": 1.8764640986530177, + "learning_rate": 3.623983535336606e-07, + "loss": 1.0307, + "step": 38907 + }, + { + "epoch": 0.92, + "grad_norm": 2.1093675458741052, + "learning_rate": 3.621948249663998e-07, + "loss": 0.9097, + "step": 38908 + }, + { + "epoch": 0.92, + "grad_norm": 2.2296379470297714, + "learning_rate": 3.6199135251349324e-07, + "loss": 0.8443, + "step": 38909 + }, + { + "epoch": 0.92, + "grad_norm": 1.9808901703119597, + "learning_rate": 3.617879361761245e-07, + "loss": 0.9402, + "step": 38910 + }, + { + "epoch": 0.92, + "grad_norm": 2.0172852609893552, + "learning_rate": 3.615845759554781e-07, + "loss": 1.1537, + "step": 38911 + }, + { + "epoch": 0.92, + "grad_norm": 1.9225710032813375, + "learning_rate": 3.613812718527365e-07, + "loss": 0.9512, + "step": 38912 + }, + { + "epoch": 0.92, + "grad_norm": 1.9241605562220336, + "learning_rate": 3.611780238690854e-07, + "loss": 1.024, + "step": 38913 + }, + { + "epoch": 0.92, + "grad_norm": 2.125765479187749, + "learning_rate": 3.6097483200570714e-07, + "loss": 1.0819, + "step": 38914 + }, + { + "epoch": 0.92, + "grad_norm": 1.9899750981453137, + "learning_rate": 3.6077169626378526e-07, + "loss": 0.9675, + "step": 38915 + }, + { + "epoch": 0.92, + "grad_norm": 1.9517498967413787, + "learning_rate": 3.6056861664450214e-07, + "loss": 0.988, + "step": 38916 + }, + { + "epoch": 0.92, + "grad_norm": 2.0136016497914375, + "learning_rate": 3.603655931490413e-07, + "loss": 1.096, + "step": 38917 + }, + { + "epoch": 0.92, + "grad_norm": 1.9968180371179534, + "learning_rate": 3.6016262577858174e-07, + "loss": 1.0691, + "step": 38918 + }, + { + "epoch": 0.92, + "grad_norm": 1.9355483556726376, + "learning_rate": 3.599597145343092e-07, + "loss": 1.0654, + "step": 38919 + }, + { + "epoch": 0.92, + "grad_norm": 1.850248990736974, + "learning_rate": 3.597568594174028e-07, + "loss": 0.9857, + "step": 38920 + }, + { + "epoch": 0.92, + "grad_norm": 1.046539657263894, + "learning_rate": 3.595540604290437e-07, + "loss": 0.8574, + "step": 38921 + }, + { + "epoch": 0.92, + "grad_norm": 1.9713601728031416, + "learning_rate": 3.593513175704122e-07, + "loss": 1.0104, + "step": 38922 + }, + { + "epoch": 0.92, + "grad_norm": 1.0472714477239888, + "learning_rate": 3.5914863084268946e-07, + "loss": 0.9855, + "step": 38923 + }, + { + "epoch": 0.92, + "grad_norm": 1.955507825421686, + "learning_rate": 3.5894600024705904e-07, + "loss": 0.9817, + "step": 38924 + }, + { + "epoch": 0.92, + "grad_norm": 2.1787155635208406, + "learning_rate": 3.5874342578469337e-07, + "loss": 0.9742, + "step": 38925 + }, + { + "epoch": 0.92, + "grad_norm": 2.0388200609366556, + "learning_rate": 3.5854090745677696e-07, + "loss": 0.9041, + "step": 38926 + }, + { + "epoch": 0.92, + "grad_norm": 2.0683362116835937, + "learning_rate": 3.583384452644889e-07, + "loss": 0.9849, + "step": 38927 + }, + { + "epoch": 0.92, + "grad_norm": 1.04501433687583, + "learning_rate": 3.581360392090072e-07, + "loss": 0.9403, + "step": 38928 + }, + { + "epoch": 0.92, + "grad_norm": 2.0058351011952573, + "learning_rate": 3.579336892915075e-07, + "loss": 0.8792, + "step": 38929 + }, + { + "epoch": 0.92, + "grad_norm": 1.9042095763419111, + "learning_rate": 3.5773139551317226e-07, + "loss": 1.0061, + "step": 38930 + }, + { + "epoch": 0.92, + "grad_norm": 1.9288886989691727, + "learning_rate": 3.5752915787517716e-07, + "loss": 0.9604, + "step": 38931 + }, + { + "epoch": 0.92, + "grad_norm": 2.037534562849678, + "learning_rate": 3.573269763787013e-07, + "loss": 1.073, + "step": 38932 + }, + { + "epoch": 0.92, + "grad_norm": 1.9632318325878175, + "learning_rate": 3.571248510249181e-07, + "loss": 1.0061, + "step": 38933 + }, + { + "epoch": 0.92, + "grad_norm": 4.08138137109121, + "learning_rate": 3.5692278181500893e-07, + "loss": 1.0584, + "step": 38934 + }, + { + "epoch": 0.92, + "grad_norm": 1.7927376024423967, + "learning_rate": 3.5672076875014727e-07, + "loss": 0.9196, + "step": 38935 + }, + { + "epoch": 0.92, + "grad_norm": 1.9940843873706102, + "learning_rate": 3.5651881183150995e-07, + "loss": 1.1504, + "step": 38936 + }, + { + "epoch": 0.92, + "grad_norm": 2.0248110272026882, + "learning_rate": 3.563169110602749e-07, + "loss": 1.0233, + "step": 38937 + }, + { + "epoch": 0.92, + "grad_norm": 2.060281364568765, + "learning_rate": 3.5611506643761453e-07, + "loss": 0.9623, + "step": 38938 + }, + { + "epoch": 0.92, + "grad_norm": 1.147205003745574, + "learning_rate": 3.5591327796470567e-07, + "loss": 0.9079, + "step": 38939 + }, + { + "epoch": 0.92, + "grad_norm": 2.183159207348366, + "learning_rate": 3.5571154564272293e-07, + "loss": 1.0263, + "step": 38940 + }, + { + "epoch": 0.92, + "grad_norm": 1.8099491535481143, + "learning_rate": 3.5550986947284205e-07, + "loss": 0.9722, + "step": 38941 + }, + { + "epoch": 0.92, + "grad_norm": 3.2953670352353415, + "learning_rate": 3.553082494562354e-07, + "loss": 1.0275, + "step": 38942 + }, + { + "epoch": 0.92, + "grad_norm": 1.936631070818681, + "learning_rate": 3.5510668559407656e-07, + "loss": 0.9567, + "step": 38943 + }, + { + "epoch": 0.92, + "grad_norm": 2.3938647214933066, + "learning_rate": 3.549051778875401e-07, + "loss": 0.9917, + "step": 38944 + }, + { + "epoch": 0.92, + "grad_norm": 1.812021061958766, + "learning_rate": 3.547037263378017e-07, + "loss": 0.9489, + "step": 38945 + }, + { + "epoch": 0.92, + "grad_norm": 2.0309295565449013, + "learning_rate": 3.545023309460294e-07, + "loss": 0.9648, + "step": 38946 + }, + { + "epoch": 0.92, + "grad_norm": 1.055202659651187, + "learning_rate": 3.543009917133988e-07, + "loss": 0.9061, + "step": 38947 + }, + { + "epoch": 0.92, + "grad_norm": 2.150716271892898, + "learning_rate": 3.5409970864108247e-07, + "loss": 0.8955, + "step": 38948 + }, + { + "epoch": 0.92, + "grad_norm": 1.957699044775102, + "learning_rate": 3.538984817302504e-07, + "loss": 0.9419, + "step": 38949 + }, + { + "epoch": 0.92, + "grad_norm": 2.0354815084063844, + "learning_rate": 3.536973109820763e-07, + "loss": 0.9562, + "step": 38950 + }, + { + "epoch": 0.92, + "grad_norm": 2.1494625734715607, + "learning_rate": 3.5349619639772903e-07, + "loss": 0.9234, + "step": 38951 + }, + { + "epoch": 0.92, + "grad_norm": 1.96457272350383, + "learning_rate": 3.532951379783822e-07, + "loss": 0.9839, + "step": 38952 + }, + { + "epoch": 0.92, + "grad_norm": 1.9561637956543902, + "learning_rate": 3.530941357252038e-07, + "loss": 0.9519, + "step": 38953 + }, + { + "epoch": 0.92, + "grad_norm": 1.1019282616212605, + "learning_rate": 3.5289318963936726e-07, + "loss": 0.8515, + "step": 38954 + }, + { + "epoch": 0.92, + "grad_norm": 1.8065815038286508, + "learning_rate": 3.526922997220406e-07, + "loss": 0.9648, + "step": 38955 + }, + { + "epoch": 0.92, + "grad_norm": 2.2791067001581697, + "learning_rate": 3.524914659743928e-07, + "loss": 0.983, + "step": 38956 + }, + { + "epoch": 0.92, + "grad_norm": 1.713576596640445, + "learning_rate": 3.5229068839759297e-07, + "loss": 0.9395, + "step": 38957 + }, + { + "epoch": 0.92, + "grad_norm": 1.8681869578965087, + "learning_rate": 3.520899669928135e-07, + "loss": 0.961, + "step": 38958 + }, + { + "epoch": 0.92, + "grad_norm": 1.94165198128068, + "learning_rate": 3.5188930176122016e-07, + "loss": 0.9471, + "step": 38959 + }, + { + "epoch": 0.92, + "grad_norm": 2.3881847361700537, + "learning_rate": 3.5168869270398085e-07, + "loss": 0.9561, + "step": 38960 + }, + { + "epoch": 0.92, + "grad_norm": 3.5392111200085394, + "learning_rate": 3.514881398222647e-07, + "loss": 0.8102, + "step": 38961 + }, + { + "epoch": 0.92, + "grad_norm": 2.6107878233922857, + "learning_rate": 3.512876431172396e-07, + "loss": 1.0061, + "step": 38962 + }, + { + "epoch": 0.92, + "grad_norm": 1.7780525656273258, + "learning_rate": 3.5108720259007356e-07, + "loss": 1.0368, + "step": 38963 + }, + { + "epoch": 0.92, + "grad_norm": 2.0173997362795117, + "learning_rate": 3.508868182419312e-07, + "loss": 0.9331, + "step": 38964 + }, + { + "epoch": 0.92, + "grad_norm": 1.1566827866610425, + "learning_rate": 3.506864900739804e-07, + "loss": 0.8812, + "step": 38965 + }, + { + "epoch": 0.92, + "grad_norm": 2.2439179732283945, + "learning_rate": 3.5048621808738913e-07, + "loss": 0.9736, + "step": 38966 + }, + { + "epoch": 0.92, + "grad_norm": 2.212940752359365, + "learning_rate": 3.502860022833221e-07, + "loss": 0.9683, + "step": 38967 + }, + { + "epoch": 0.92, + "grad_norm": 1.9150107211783463, + "learning_rate": 3.500858426629439e-07, + "loss": 1.0211, + "step": 38968 + }, + { + "epoch": 0.92, + "grad_norm": 1.0274506773211767, + "learning_rate": 3.498857392274213e-07, + "loss": 0.9736, + "step": 38969 + }, + { + "epoch": 0.92, + "grad_norm": 1.8699920105063235, + "learning_rate": 3.49685691977919e-07, + "loss": 1.0998, + "step": 38970 + }, + { + "epoch": 0.92, + "grad_norm": 1.7850005291386948, + "learning_rate": 3.4948570091560276e-07, + "loss": 1.0171, + "step": 38971 + }, + { + "epoch": 0.92, + "grad_norm": 2.3659939292506893, + "learning_rate": 3.4928576604163487e-07, + "loss": 0.9246, + "step": 38972 + }, + { + "epoch": 0.92, + "grad_norm": 1.8229205760175928, + "learning_rate": 3.4908588735718116e-07, + "loss": 0.9171, + "step": 38973 + }, + { + "epoch": 0.92, + "grad_norm": 1.8869816581364316, + "learning_rate": 3.4888606486340515e-07, + "loss": 0.9566, + "step": 38974 + }, + { + "epoch": 0.92, + "grad_norm": 2.027986745399976, + "learning_rate": 3.486862985614703e-07, + "loss": 1.0227, + "step": 38975 + }, + { + "epoch": 0.92, + "grad_norm": 1.7459184182015162, + "learning_rate": 3.4848658845253903e-07, + "loss": 0.9057, + "step": 38976 + }, + { + "epoch": 0.92, + "grad_norm": 1.1206637937663528, + "learning_rate": 3.4828693453777375e-07, + "loss": 0.8596, + "step": 38977 + }, + { + "epoch": 0.92, + "grad_norm": 2.025475423448447, + "learning_rate": 3.4808733681833686e-07, + "loss": 0.983, + "step": 38978 + }, + { + "epoch": 0.92, + "grad_norm": 2.1120326600125976, + "learning_rate": 3.4788779529539297e-07, + "loss": 0.8584, + "step": 38979 + }, + { + "epoch": 0.92, + "grad_norm": 2.383642627776937, + "learning_rate": 3.476883099701023e-07, + "loss": 0.8524, + "step": 38980 + }, + { + "epoch": 0.92, + "grad_norm": 2.067469323692519, + "learning_rate": 3.4748888084362497e-07, + "loss": 0.9151, + "step": 38981 + }, + { + "epoch": 0.92, + "grad_norm": 1.956583316468827, + "learning_rate": 3.4728950791712343e-07, + "loss": 0.9512, + "step": 38982 + }, + { + "epoch": 0.92, + "grad_norm": 2.0470987757953356, + "learning_rate": 3.47090191191759e-07, + "loss": 0.9673, + "step": 38983 + }, + { + "epoch": 0.92, + "grad_norm": 2.1225848657023336, + "learning_rate": 3.4689093066869184e-07, + "loss": 1.077, + "step": 38984 + }, + { + "epoch": 0.92, + "grad_norm": 1.9118959366933594, + "learning_rate": 3.4669172634908097e-07, + "loss": 0.9785, + "step": 38985 + }, + { + "epoch": 0.92, + "grad_norm": 1.8934212425381651, + "learning_rate": 3.4649257823408667e-07, + "loss": 0.9631, + "step": 38986 + }, + { + "epoch": 0.92, + "grad_norm": 2.167316789692292, + "learning_rate": 3.462934863248701e-07, + "loss": 1.0073, + "step": 38987 + }, + { + "epoch": 0.92, + "grad_norm": 2.0583050573371278, + "learning_rate": 3.460944506225894e-07, + "loss": 0.9398, + "step": 38988 + }, + { + "epoch": 0.92, + "grad_norm": 1.9391602692313707, + "learning_rate": 3.458954711284013e-07, + "loss": 0.9221, + "step": 38989 + }, + { + "epoch": 0.92, + "grad_norm": 2.179054965767034, + "learning_rate": 3.4569654784346816e-07, + "loss": 0.998, + "step": 38990 + }, + { + "epoch": 0.92, + "grad_norm": 1.05125540842562, + "learning_rate": 3.4549768076894473e-07, + "loss": 0.9048, + "step": 38991 + }, + { + "epoch": 0.92, + "grad_norm": 2.4551551539514773, + "learning_rate": 3.4529886990599113e-07, + "loss": 0.9741, + "step": 38992 + }, + { + "epoch": 0.92, + "grad_norm": 2.0084651132708338, + "learning_rate": 3.4510011525576317e-07, + "loss": 0.9312, + "step": 38993 + }, + { + "epoch": 0.92, + "grad_norm": 1.038022370984987, + "learning_rate": 3.449014168194209e-07, + "loss": 0.9183, + "step": 38994 + }, + { + "epoch": 0.92, + "grad_norm": 1.95463450273835, + "learning_rate": 3.44702774598118e-07, + "loss": 0.953, + "step": 38995 + }, + { + "epoch": 0.92, + "grad_norm": 2.0141746242465954, + "learning_rate": 3.4450418859301226e-07, + "loss": 1.0522, + "step": 38996 + }, + { + "epoch": 0.92, + "grad_norm": 2.0272468130736216, + "learning_rate": 3.443056588052618e-07, + "loss": 1.146, + "step": 38997 + }, + { + "epoch": 0.92, + "grad_norm": 1.8714275913426244, + "learning_rate": 3.441071852360178e-07, + "loss": 0.9721, + "step": 38998 + }, + { + "epoch": 0.92, + "grad_norm": 3.10043559791254, + "learning_rate": 3.4390876788644056e-07, + "loss": 0.9638, + "step": 38999 + }, + { + "epoch": 0.92, + "grad_norm": 4.914622725572726, + "learning_rate": 3.437104067576835e-07, + "loss": 0.9872, + "step": 39000 + }, + { + "epoch": 0.92, + "grad_norm": 2.107198109382964, + "learning_rate": 3.4351210185090133e-07, + "loss": 0.9242, + "step": 39001 + }, + { + "epoch": 0.92, + "grad_norm": 2.0371934040437236, + "learning_rate": 3.4331385316724864e-07, + "loss": 0.9067, + "step": 39002 + }, + { + "epoch": 0.92, + "grad_norm": 1.8325364515853417, + "learning_rate": 3.4311566070788004e-07, + "loss": 0.9015, + "step": 39003 + }, + { + "epoch": 0.92, + "grad_norm": 1.9559686722081195, + "learning_rate": 3.429175244739502e-07, + "loss": 1.1219, + "step": 39004 + }, + { + "epoch": 0.92, + "grad_norm": 1.275366130267958, + "learning_rate": 3.4271944446661153e-07, + "loss": 0.9362, + "step": 39005 + }, + { + "epoch": 0.92, + "grad_norm": 1.9322547176324145, + "learning_rate": 3.4252142068701645e-07, + "loss": 1.035, + "step": 39006 + }, + { + "epoch": 0.92, + "grad_norm": 2.0294625266995774, + "learning_rate": 3.423234531363195e-07, + "loss": 0.8375, + "step": 39007 + }, + { + "epoch": 0.92, + "grad_norm": 1.982914654847119, + "learning_rate": 3.421255418156744e-07, + "loss": 1.0183, + "step": 39008 + }, + { + "epoch": 0.92, + "grad_norm": 2.160353002331327, + "learning_rate": 3.4192768672623225e-07, + "loss": 0.959, + "step": 39009 + }, + { + "epoch": 0.92, + "grad_norm": 1.760499420318893, + "learning_rate": 3.417298878691433e-07, + "loss": 1.102, + "step": 39010 + }, + { + "epoch": 0.92, + "grad_norm": 1.121775073484772, + "learning_rate": 3.415321452455611e-07, + "loss": 0.8884, + "step": 39011 + }, + { + "epoch": 0.92, + "grad_norm": 1.046033144449507, + "learning_rate": 3.4133445885663584e-07, + "loss": 0.9892, + "step": 39012 + }, + { + "epoch": 0.92, + "grad_norm": 1.9246991508181654, + "learning_rate": 3.4113682870352103e-07, + "loss": 0.9063, + "step": 39013 + }, + { + "epoch": 0.92, + "grad_norm": 2.9630660100715267, + "learning_rate": 3.409392547873636e-07, + "loss": 1.102, + "step": 39014 + }, + { + "epoch": 0.92, + "grad_norm": 2.2976679267026423, + "learning_rate": 3.4074173710931804e-07, + "loss": 0.9554, + "step": 39015 + }, + { + "epoch": 0.92, + "grad_norm": 1.9430083674901149, + "learning_rate": 3.405442756705302e-07, + "loss": 0.8668, + "step": 39016 + }, + { + "epoch": 0.92, + "grad_norm": 1.9312123232664666, + "learning_rate": 3.4034687047215245e-07, + "loss": 0.975, + "step": 39017 + }, + { + "epoch": 0.92, + "grad_norm": 2.0752842393221846, + "learning_rate": 3.40149521515335e-07, + "loss": 1.0118, + "step": 39018 + }, + { + "epoch": 0.92, + "grad_norm": 1.9011449644795384, + "learning_rate": 3.3995222880122245e-07, + "loss": 0.9154, + "step": 39019 + }, + { + "epoch": 0.92, + "grad_norm": 2.6995528054268245, + "learning_rate": 3.3975499233096734e-07, + "loss": 0.9916, + "step": 39020 + }, + { + "epoch": 0.92, + "grad_norm": 2.1898930077536534, + "learning_rate": 3.395578121057164e-07, + "loss": 0.9647, + "step": 39021 + }, + { + "epoch": 0.92, + "grad_norm": 2.0572801145498856, + "learning_rate": 3.3936068812662094e-07, + "loss": 0.8771, + "step": 39022 + }, + { + "epoch": 0.92, + "grad_norm": 1.900134168053414, + "learning_rate": 3.391636203948234e-07, + "loss": 0.9075, + "step": 39023 + }, + { + "epoch": 0.92, + "grad_norm": 1.8949171931630249, + "learning_rate": 3.3896660891147516e-07, + "loss": 0.9588, + "step": 39024 + }, + { + "epoch": 0.92, + "grad_norm": 1.104092217135393, + "learning_rate": 3.3876965367772183e-07, + "loss": 0.8948, + "step": 39025 + }, + { + "epoch": 0.92, + "grad_norm": 2.1815542072257856, + "learning_rate": 3.385727546947104e-07, + "loss": 1.0667, + "step": 39026 + }, + { + "epoch": 0.92, + "grad_norm": 2.0381309085512527, + "learning_rate": 3.3837591196358657e-07, + "loss": 0.8126, + "step": 39027 + }, + { + "epoch": 0.92, + "grad_norm": 1.816887394376689, + "learning_rate": 3.38179125485496e-07, + "loss": 1.0204, + "step": 39028 + }, + { + "epoch": 0.92, + "grad_norm": 2.379517817811279, + "learning_rate": 3.3798239526158793e-07, + "loss": 0.9815, + "step": 39029 + }, + { + "epoch": 0.92, + "grad_norm": 2.3223221858606427, + "learning_rate": 3.377857212930047e-07, + "loss": 1.0235, + "step": 39030 + }, + { + "epoch": 0.92, + "grad_norm": 1.832141642609846, + "learning_rate": 3.375891035808898e-07, + "loss": 1.0285, + "step": 39031 + }, + { + "epoch": 0.92, + "grad_norm": 2.0061788320823704, + "learning_rate": 3.373925421263924e-07, + "loss": 0.8839, + "step": 39032 + }, + { + "epoch": 0.92, + "grad_norm": 2.0881094345078095, + "learning_rate": 3.3719603693065373e-07, + "loss": 0.784, + "step": 39033 + }, + { + "epoch": 0.92, + "grad_norm": 2.0952330388812856, + "learning_rate": 3.369995879948185e-07, + "loss": 0.8491, + "step": 39034 + }, + { + "epoch": 0.92, + "grad_norm": 2.364680456691618, + "learning_rate": 3.3680319532003234e-07, + "loss": 0.894, + "step": 39035 + }, + { + "epoch": 0.92, + "grad_norm": 2.0380445740271815, + "learning_rate": 3.366068589074378e-07, + "loss": 1.0242, + "step": 39036 + }, + { + "epoch": 0.92, + "grad_norm": 2.0010511872291317, + "learning_rate": 3.3641057875817507e-07, + "loss": 0.9146, + "step": 39037 + }, + { + "epoch": 0.92, + "grad_norm": 1.9721152599489873, + "learning_rate": 3.3621435487339095e-07, + "loss": 0.8642, + "step": 39038 + }, + { + "epoch": 0.92, + "grad_norm": 1.9582395474723773, + "learning_rate": 3.360181872542256e-07, + "loss": 0.9484, + "step": 39039 + }, + { + "epoch": 0.92, + "grad_norm": 2.0037872516609694, + "learning_rate": 3.358220759018238e-07, + "loss": 1.0031, + "step": 39040 + }, + { + "epoch": 0.92, + "grad_norm": 1.9933537631709939, + "learning_rate": 3.3562602081732344e-07, + "loss": 0.983, + "step": 39041 + }, + { + "epoch": 0.92, + "grad_norm": 3.1960222926645274, + "learning_rate": 3.3543002200186803e-07, + "loss": 0.9847, + "step": 39042 + }, + { + "epoch": 0.92, + "grad_norm": 2.0282518704051746, + "learning_rate": 3.3523407945660004e-07, + "loss": 1.0107, + "step": 39043 + }, + { + "epoch": 0.92, + "grad_norm": 2.0241477093428517, + "learning_rate": 3.350381931826585e-07, + "loss": 1.0847, + "step": 39044 + }, + { + "epoch": 0.92, + "grad_norm": 1.9817023251604102, + "learning_rate": 3.348423631811848e-07, + "loss": 1.0192, + "step": 39045 + }, + { + "epoch": 0.92, + "grad_norm": 1.8065151522212757, + "learning_rate": 3.3464658945331796e-07, + "loss": 1.071, + "step": 39046 + }, + { + "epoch": 0.92, + "grad_norm": 2.8427263997706453, + "learning_rate": 3.3445087200019934e-07, + "loss": 1.0288, + "step": 39047 + }, + { + "epoch": 0.92, + "grad_norm": 2.0641925875408997, + "learning_rate": 3.3425521082296797e-07, + "loss": 0.8721, + "step": 39048 + }, + { + "epoch": 0.92, + "grad_norm": 1.887393133159038, + "learning_rate": 3.340596059227619e-07, + "loss": 0.96, + "step": 39049 + }, + { + "epoch": 0.92, + "grad_norm": 1.0716126211099362, + "learning_rate": 3.3386405730072237e-07, + "loss": 1.0121, + "step": 39050 + }, + { + "epoch": 0.92, + "grad_norm": 2.0374688947221604, + "learning_rate": 3.336685649579852e-07, + "loss": 0.9461, + "step": 39051 + }, + { + "epoch": 0.92, + "grad_norm": 2.502000864747509, + "learning_rate": 3.334731288956916e-07, + "loss": 1.0066, + "step": 39052 + }, + { + "epoch": 0.92, + "grad_norm": 1.9984149449345452, + "learning_rate": 3.332777491149763e-07, + "loss": 1.0444, + "step": 39053 + }, + { + "epoch": 0.92, + "grad_norm": 5.927101311234059, + "learning_rate": 3.3308242561697956e-07, + "loss": 0.9529, + "step": 39054 + }, + { + "epoch": 0.92, + "grad_norm": 1.0819850739204218, + "learning_rate": 3.328871584028359e-07, + "loss": 0.9226, + "step": 39055 + }, + { + "epoch": 0.92, + "grad_norm": 1.1166055739719223, + "learning_rate": 3.326919474736856e-07, + "loss": 0.9458, + "step": 39056 + }, + { + "epoch": 0.92, + "grad_norm": 2.2069026622615846, + "learning_rate": 3.324967928306622e-07, + "loss": 0.9405, + "step": 39057 + }, + { + "epoch": 0.92, + "grad_norm": 1.8794879128231992, + "learning_rate": 3.3230169447490356e-07, + "loss": 0.8615, + "step": 39058 + }, + { + "epoch": 0.92, + "grad_norm": 2.0061873893943063, + "learning_rate": 3.3210665240754446e-07, + "loss": 0.9759, + "step": 39059 + }, + { + "epoch": 0.92, + "grad_norm": 1.0564473775537386, + "learning_rate": 3.3191166662972174e-07, + "loss": 0.9661, + "step": 39060 + }, + { + "epoch": 0.92, + "grad_norm": 4.314207783633513, + "learning_rate": 3.3171673714257113e-07, + "loss": 0.9876, + "step": 39061 + }, + { + "epoch": 0.92, + "grad_norm": 1.073348910411698, + "learning_rate": 3.3152186394722506e-07, + "loss": 0.9464, + "step": 39062 + }, + { + "epoch": 0.92, + "grad_norm": 2.0015731213613166, + "learning_rate": 3.3132704704481934e-07, + "loss": 0.9658, + "step": 39063 + }, + { + "epoch": 0.92, + "grad_norm": 1.759863606486728, + "learning_rate": 3.3113228643649075e-07, + "loss": 0.9775, + "step": 39064 + }, + { + "epoch": 0.92, + "grad_norm": 1.9368873441589538, + "learning_rate": 3.309375821233696e-07, + "loss": 0.8031, + "step": 39065 + }, + { + "epoch": 0.92, + "grad_norm": 2.075501736400909, + "learning_rate": 3.3074293410659154e-07, + "loss": 0.8934, + "step": 39066 + }, + { + "epoch": 0.92, + "grad_norm": 1.0641939498415207, + "learning_rate": 3.3054834238728795e-07, + "loss": 0.8655, + "step": 39067 + }, + { + "epoch": 0.92, + "grad_norm": 1.9031433395806148, + "learning_rate": 3.303538069665957e-07, + "loss": 0.9023, + "step": 39068 + }, + { + "epoch": 0.92, + "grad_norm": 1.0440058439733588, + "learning_rate": 3.3015932784564385e-07, + "loss": 0.9241, + "step": 39069 + }, + { + "epoch": 0.92, + "grad_norm": 2.1106198398275597, + "learning_rate": 3.2996490502556486e-07, + "loss": 0.9478, + "step": 39070 + }, + { + "epoch": 0.92, + "grad_norm": 1.8364530252359375, + "learning_rate": 3.297705385074934e-07, + "loss": 1.1039, + "step": 39071 + }, + { + "epoch": 0.92, + "grad_norm": 1.101015331486778, + "learning_rate": 3.2957622829255744e-07, + "loss": 0.9345, + "step": 39072 + }, + { + "epoch": 0.92, + "grad_norm": 1.0684324649020684, + "learning_rate": 3.2938197438189154e-07, + "loss": 0.9055, + "step": 39073 + }, + { + "epoch": 0.92, + "grad_norm": 2.2467869904722604, + "learning_rate": 3.2918777677662495e-07, + "loss": 1.0041, + "step": 39074 + }, + { + "epoch": 0.92, + "grad_norm": 2.122209175642452, + "learning_rate": 3.2899363547789e-07, + "loss": 0.8363, + "step": 39075 + }, + { + "epoch": 0.92, + "grad_norm": 2.1329333350602036, + "learning_rate": 3.2879955048681465e-07, + "loss": 1.0854, + "step": 39076 + }, + { + "epoch": 0.92, + "grad_norm": 1.0542057971055596, + "learning_rate": 3.286055218045303e-07, + "loss": 0.9538, + "step": 39077 + }, + { + "epoch": 0.92, + "grad_norm": 1.1177199433310858, + "learning_rate": 3.2841154943216714e-07, + "loss": 0.936, + "step": 39078 + }, + { + "epoch": 0.92, + "grad_norm": 1.9031552621924492, + "learning_rate": 3.282176333708531e-07, + "loss": 0.9842, + "step": 39079 + }, + { + "epoch": 0.92, + "grad_norm": 2.0342880530114615, + "learning_rate": 3.2802377362171845e-07, + "loss": 0.9665, + "step": 39080 + }, + { + "epoch": 0.92, + "grad_norm": 11.038679106543865, + "learning_rate": 3.278299701858922e-07, + "loss": 1.0597, + "step": 39081 + }, + { + "epoch": 0.92, + "grad_norm": 1.1136157391851522, + "learning_rate": 3.276362230645014e-07, + "loss": 0.9478, + "step": 39082 + }, + { + "epoch": 0.92, + "grad_norm": 1.8006507434871044, + "learning_rate": 3.2744253225867386e-07, + "loss": 0.9793, + "step": 39083 + }, + { + "epoch": 0.92, + "grad_norm": 3.966294233487986, + "learning_rate": 3.2724889776953874e-07, + "loss": 1.0237, + "step": 39084 + }, + { + "epoch": 0.92, + "grad_norm": 1.920394468025183, + "learning_rate": 3.270553195982229e-07, + "loss": 0.9655, + "step": 39085 + }, + { + "epoch": 0.92, + "grad_norm": 2.0541574582941293, + "learning_rate": 3.2686179774585434e-07, + "loss": 0.9809, + "step": 39086 + }, + { + "epoch": 0.92, + "grad_norm": 2.094426627905023, + "learning_rate": 3.2666833221355774e-07, + "loss": 0.9656, + "step": 39087 + }, + { + "epoch": 0.92, + "grad_norm": 2.139733437312732, + "learning_rate": 3.2647492300246e-07, + "loss": 1.0028, + "step": 39088 + }, + { + "epoch": 0.92, + "grad_norm": 2.2144360141762625, + "learning_rate": 3.262815701136901e-07, + "loss": 1.001, + "step": 39089 + }, + { + "epoch": 0.92, + "grad_norm": 1.9414880045638754, + "learning_rate": 3.2608827354837057e-07, + "loss": 0.9217, + "step": 39090 + }, + { + "epoch": 0.92, + "grad_norm": 1.9319207833373528, + "learning_rate": 3.258950333076272e-07, + "loss": 0.9399, + "step": 39091 + }, + { + "epoch": 0.92, + "grad_norm": 1.747110320585228, + "learning_rate": 3.2570184939258674e-07, + "loss": 0.9297, + "step": 39092 + }, + { + "epoch": 0.92, + "grad_norm": 2.1206665402471563, + "learning_rate": 3.255087218043718e-07, + "loss": 1.0073, + "step": 39093 + }, + { + "epoch": 0.92, + "grad_norm": 2.2090542332795, + "learning_rate": 3.2531565054410797e-07, + "loss": 1.0894, + "step": 39094 + }, + { + "epoch": 0.92, + "grad_norm": 1.7990586351097533, + "learning_rate": 3.2512263561292e-07, + "loss": 0.9503, + "step": 39095 + }, + { + "epoch": 0.92, + "grad_norm": 2.0868926952550773, + "learning_rate": 3.2492967701193144e-07, + "loss": 1.0159, + "step": 39096 + }, + { + "epoch": 0.92, + "grad_norm": 2.124408401593242, + "learning_rate": 3.2473677474226363e-07, + "loss": 0.9123, + "step": 39097 + }, + { + "epoch": 0.92, + "grad_norm": 1.788118320921347, + "learning_rate": 3.245439288050434e-07, + "loss": 0.9078, + "step": 39098 + }, + { + "epoch": 0.92, + "grad_norm": 1.9526597093438722, + "learning_rate": 3.2435113920139094e-07, + "loss": 1.0202, + "step": 39099 + }, + { + "epoch": 0.92, + "grad_norm": 2.0257671550178733, + "learning_rate": 3.2415840593242874e-07, + "loss": 1.0013, + "step": 39100 + }, + { + "epoch": 0.92, + "grad_norm": 1.9591261156601496, + "learning_rate": 3.2396572899928037e-07, + "loss": 0.8765, + "step": 39101 + }, + { + "epoch": 0.92, + "grad_norm": 1.8860140464498867, + "learning_rate": 3.237731084030671e-07, + "loss": 1.0095, + "step": 39102 + }, + { + "epoch": 0.92, + "grad_norm": 2.0984764088545957, + "learning_rate": 3.2358054414491024e-07, + "loss": 1.0396, + "step": 39103 + }, + { + "epoch": 0.92, + "grad_norm": 2.15119831125416, + "learning_rate": 3.2338803622593006e-07, + "loss": 0.9255, + "step": 39104 + }, + { + "epoch": 0.92, + "grad_norm": 2.2695675444245085, + "learning_rate": 3.231955846472479e-07, + "loss": 1.0014, + "step": 39105 + }, + { + "epoch": 0.92, + "grad_norm": 1.7788147570199344, + "learning_rate": 3.2300318940998607e-07, + "loss": 1.0015, + "step": 39106 + }, + { + "epoch": 0.92, + "grad_norm": 1.9133729950566223, + "learning_rate": 3.2281085051526383e-07, + "loss": 1.1327, + "step": 39107 + }, + { + "epoch": 0.92, + "grad_norm": 2.2935639185542245, + "learning_rate": 3.226185679641991e-07, + "loss": 1.0214, + "step": 39108 + }, + { + "epoch": 0.92, + "grad_norm": 2.1904730201530316, + "learning_rate": 3.224263417579132e-07, + "loss": 0.9012, + "step": 39109 + }, + { + "epoch": 0.92, + "grad_norm": 1.9093445160390414, + "learning_rate": 3.2223417189752525e-07, + "loss": 0.769, + "step": 39110 + }, + { + "epoch": 0.92, + "grad_norm": 2.3525539793006374, + "learning_rate": 3.2204205838415437e-07, + "loss": 0.9914, + "step": 39111 + }, + { + "epoch": 0.92, + "grad_norm": 1.9059324933260098, + "learning_rate": 3.218500012189174e-07, + "loss": 0.8889, + "step": 39112 + }, + { + "epoch": 0.92, + "grad_norm": 2.3622999967896727, + "learning_rate": 3.216580004029357e-07, + "loss": 1.0292, + "step": 39113 + }, + { + "epoch": 0.92, + "grad_norm": 1.8354998596417682, + "learning_rate": 3.214660559373228e-07, + "loss": 0.9213, + "step": 39114 + }, + { + "epoch": 0.92, + "grad_norm": 1.9293938826770198, + "learning_rate": 3.212741678232012e-07, + "loss": 1.0238, + "step": 39115 + }, + { + "epoch": 0.92, + "grad_norm": 2.144258412793796, + "learning_rate": 3.210823360616833e-07, + "loss": 0.8746, + "step": 39116 + }, + { + "epoch": 0.92, + "grad_norm": 1.9447537108188435, + "learning_rate": 3.2089056065388924e-07, + "loss": 1.0119, + "step": 39117 + }, + { + "epoch": 0.92, + "grad_norm": 2.495056377330533, + "learning_rate": 3.206988416009349e-07, + "loss": 1.081, + "step": 39118 + }, + { + "epoch": 0.92, + "grad_norm": 1.9602883522332193, + "learning_rate": 3.20507178903936e-07, + "loss": 0.8593, + "step": 39119 + }, + { + "epoch": 0.92, + "grad_norm": 2.155981158949923, + "learning_rate": 3.203155725640106e-07, + "loss": 0.9952, + "step": 39120 + }, + { + "epoch": 0.92, + "grad_norm": 2.8308057570324405, + "learning_rate": 3.2012402258226995e-07, + "loss": 1.0095, + "step": 39121 + }, + { + "epoch": 0.92, + "grad_norm": 2.04436660022041, + "learning_rate": 3.1993252895983205e-07, + "loss": 0.9648, + "step": 39122 + }, + { + "epoch": 0.92, + "grad_norm": 1.8707968152957941, + "learning_rate": 3.1974109169781165e-07, + "loss": 0.9556, + "step": 39123 + }, + { + "epoch": 0.92, + "grad_norm": 1.7456038654426735, + "learning_rate": 3.1954971079732554e-07, + "loss": 1.0053, + "step": 39124 + }, + { + "epoch": 0.92, + "grad_norm": 2.009924100319681, + "learning_rate": 3.1935838625948403e-07, + "loss": 0.8587, + "step": 39125 + }, + { + "epoch": 0.92, + "grad_norm": 1.9985237785311865, + "learning_rate": 3.191671180854028e-07, + "loss": 0.9235, + "step": 39126 + }, + { + "epoch": 0.92, + "grad_norm": 2.322923411013755, + "learning_rate": 3.1897590627619546e-07, + "loss": 1.097, + "step": 39127 + }, + { + "epoch": 0.92, + "grad_norm": 1.9757269184620507, + "learning_rate": 3.1878475083297667e-07, + "loss": 0.9334, + "step": 39128 + }, + { + "epoch": 0.92, + "grad_norm": 1.9150685568760226, + "learning_rate": 3.1859365175685666e-07, + "loss": 0.9799, + "step": 39129 + }, + { + "epoch": 0.92, + "grad_norm": 2.3960262737468523, + "learning_rate": 3.1840260904894896e-07, + "loss": 0.9915, + "step": 39130 + }, + { + "epoch": 0.92, + "grad_norm": 1.9658560216647913, + "learning_rate": 3.182116227103682e-07, + "loss": 1.0606, + "step": 39131 + }, + { + "epoch": 0.92, + "grad_norm": 2.2554038340766183, + "learning_rate": 3.180206927422236e-07, + "loss": 1.0124, + "step": 39132 + }, + { + "epoch": 0.92, + "grad_norm": 1.9752839169186651, + "learning_rate": 3.178298191456275e-07, + "loss": 0.8971, + "step": 39133 + }, + { + "epoch": 0.92, + "grad_norm": 1.9722303765262381, + "learning_rate": 3.176390019216924e-07, + "loss": 0.9643, + "step": 39134 + }, + { + "epoch": 0.92, + "grad_norm": 1.9780403629741794, + "learning_rate": 3.174482410715274e-07, + "loss": 0.8397, + "step": 39135 + }, + { + "epoch": 0.92, + "grad_norm": 5.6342863531096015, + "learning_rate": 3.172575365962449e-07, + "loss": 0.9374, + "step": 39136 + }, + { + "epoch": 0.92, + "grad_norm": 1.9276812897202322, + "learning_rate": 3.170668884969552e-07, + "loss": 0.9161, + "step": 39137 + }, + { + "epoch": 0.92, + "grad_norm": 2.0947890125649002, + "learning_rate": 3.168762967747674e-07, + "loss": 0.9754, + "step": 39138 + }, + { + "epoch": 0.92, + "grad_norm": 2.253052453084734, + "learning_rate": 3.1668576143078946e-07, + "loss": 0.9557, + "step": 39139 + }, + { + "epoch": 0.92, + "grad_norm": 2.011338409503385, + "learning_rate": 3.164952824661338e-07, + "loss": 0.9415, + "step": 39140 + }, + { + "epoch": 0.92, + "grad_norm": 1.9969157042500059, + "learning_rate": 3.163048598819096e-07, + "loss": 0.9264, + "step": 39141 + }, + { + "epoch": 0.92, + "grad_norm": 2.111251491996191, + "learning_rate": 3.1611449367922374e-07, + "loss": 0.9588, + "step": 39142 + }, + { + "epoch": 0.92, + "grad_norm": 1.0963091761529449, + "learning_rate": 3.1592418385918424e-07, + "loss": 0.9367, + "step": 39143 + }, + { + "epoch": 0.92, + "grad_norm": 1.036697390053495, + "learning_rate": 3.1573393042290015e-07, + "loss": 0.8601, + "step": 39144 + }, + { + "epoch": 0.92, + "grad_norm": 2.1429267621820483, + "learning_rate": 3.155437333714795e-07, + "loss": 1.0166, + "step": 39145 + }, + { + "epoch": 0.92, + "grad_norm": 1.9886860320678132, + "learning_rate": 3.1535359270603027e-07, + "loss": 0.9794, + "step": 39146 + }, + { + "epoch": 0.92, + "grad_norm": 1.8658055115117402, + "learning_rate": 3.151635084276572e-07, + "loss": 0.8653, + "step": 39147 + }, + { + "epoch": 0.92, + "grad_norm": 1.0726256240668328, + "learning_rate": 3.1497348053746934e-07, + "loss": 0.9314, + "step": 39148 + }, + { + "epoch": 0.92, + "grad_norm": 2.0803985549257833, + "learning_rate": 3.1478350903657026e-07, + "loss": 0.9763, + "step": 39149 + }, + { + "epoch": 0.92, + "grad_norm": 1.1142508791157004, + "learning_rate": 3.1459359392607025e-07, + "loss": 0.9413, + "step": 39150 + }, + { + "epoch": 0.92, + "grad_norm": 2.1479263102282555, + "learning_rate": 3.1440373520707055e-07, + "loss": 1.0526, + "step": 39151 + }, + { + "epoch": 0.92, + "grad_norm": 1.8928054483838377, + "learning_rate": 3.1421393288068035e-07, + "loss": 1.011, + "step": 39152 + }, + { + "epoch": 0.92, + "grad_norm": 1.9349171740583828, + "learning_rate": 3.1402418694800207e-07, + "loss": 0.9159, + "step": 39153 + }, + { + "epoch": 0.92, + "grad_norm": 1.9028468239732699, + "learning_rate": 3.138344974101426e-07, + "loss": 0.8664, + "step": 39154 + }, + { + "epoch": 0.92, + "grad_norm": 1.927929915964934, + "learning_rate": 3.136448642682044e-07, + "loss": 0.9463, + "step": 39155 + }, + { + "epoch": 0.92, + "grad_norm": 1.1205755580907832, + "learning_rate": 3.1345528752329214e-07, + "loss": 0.8771, + "step": 39156 + }, + { + "epoch": 0.92, + "grad_norm": 1.8859523117059394, + "learning_rate": 3.132657671765105e-07, + "loss": 1.0283, + "step": 39157 + }, + { + "epoch": 0.92, + "grad_norm": 1.907022360338423, + "learning_rate": 3.130763032289619e-07, + "loss": 0.858, + "step": 39158 + }, + { + "epoch": 0.92, + "grad_norm": 1.8730033606798917, + "learning_rate": 3.12886895681751e-07, + "loss": 0.9368, + "step": 39159 + }, + { + "epoch": 0.92, + "grad_norm": 1.918610210679824, + "learning_rate": 3.126975445359781e-07, + "loss": 0.9525, + "step": 39160 + }, + { + "epoch": 0.92, + "grad_norm": 2.1839593536903186, + "learning_rate": 3.1250824979274675e-07, + "loss": 0.9205, + "step": 39161 + }, + { + "epoch": 0.92, + "grad_norm": 2.0213864116151523, + "learning_rate": 3.1231901145316156e-07, + "loss": 1.1434, + "step": 39162 + }, + { + "epoch": 0.92, + "grad_norm": 2.2947102222847886, + "learning_rate": 3.1212982951832063e-07, + "loss": 1.0064, + "step": 39163 + }, + { + "epoch": 0.92, + "grad_norm": 1.9228709065165153, + "learning_rate": 3.119407039893274e-07, + "loss": 0.929, + "step": 39164 + }, + { + "epoch": 0.92, + "grad_norm": 1.9794448996034717, + "learning_rate": 3.1175163486728113e-07, + "loss": 1.0727, + "step": 39165 + }, + { + "epoch": 0.92, + "grad_norm": 2.0551570774567716, + "learning_rate": 3.115626221532864e-07, + "loss": 1.0507, + "step": 39166 + }, + { + "epoch": 0.92, + "grad_norm": 2.2043868012568, + "learning_rate": 3.113736658484412e-07, + "loss": 0.9891, + "step": 39167 + }, + { + "epoch": 0.92, + "grad_norm": 1.8533762799930074, + "learning_rate": 3.1118476595384363e-07, + "loss": 0.9991, + "step": 39168 + }, + { + "epoch": 0.92, + "grad_norm": 2.335494391929429, + "learning_rate": 3.1099592247059715e-07, + "loss": 0.8857, + "step": 39169 + }, + { + "epoch": 0.92, + "grad_norm": 1.9778537271487149, + "learning_rate": 3.108071353997999e-07, + "loss": 1.009, + "step": 39170 + }, + { + "epoch": 0.92, + "grad_norm": 2.9992457137127952, + "learning_rate": 3.10618404742552e-07, + "loss": 0.9079, + "step": 39171 + }, + { + "epoch": 0.92, + "grad_norm": 1.1717598839684704, + "learning_rate": 3.1042973049994927e-07, + "loss": 0.8798, + "step": 39172 + }, + { + "epoch": 0.92, + "grad_norm": 1.7927183421665815, + "learning_rate": 3.102411126730942e-07, + "loss": 0.8466, + "step": 39173 + }, + { + "epoch": 0.92, + "grad_norm": 1.9600741537131352, + "learning_rate": 3.1005255126308255e-07, + "loss": 0.9397, + "step": 39174 + }, + { + "epoch": 0.92, + "grad_norm": 1.7968587489792653, + "learning_rate": 3.098640462710123e-07, + "loss": 0.9161, + "step": 39175 + }, + { + "epoch": 0.92, + "grad_norm": 1.7993546202340005, + "learning_rate": 3.096755976979815e-07, + "loss": 1.0653, + "step": 39176 + }, + { + "epoch": 0.92, + "grad_norm": 2.0591406393224827, + "learning_rate": 3.094872055450882e-07, + "loss": 0.9687, + "step": 39177 + }, + { + "epoch": 0.92, + "grad_norm": 1.849417136191077, + "learning_rate": 3.0929886981342694e-07, + "loss": 0.9698, + "step": 39178 + }, + { + "epoch": 0.92, + "grad_norm": 2.088121773012475, + "learning_rate": 3.0911059050409696e-07, + "loss": 1.033, + "step": 39179 + }, + { + "epoch": 0.92, + "grad_norm": 1.1236713435474066, + "learning_rate": 3.0892236761819297e-07, + "loss": 0.9195, + "step": 39180 + }, + { + "epoch": 0.92, + "grad_norm": 1.0928995652615066, + "learning_rate": 3.0873420115680955e-07, + "loss": 0.8892, + "step": 39181 + }, + { + "epoch": 0.92, + "grad_norm": 2.1576672355827795, + "learning_rate": 3.085460911210447e-07, + "loss": 0.9854, + "step": 39182 + }, + { + "epoch": 0.92, + "grad_norm": 1.8483994762264606, + "learning_rate": 3.083580375119932e-07, + "loss": 0.9066, + "step": 39183 + }, + { + "epoch": 0.92, + "grad_norm": 1.8728778406577749, + "learning_rate": 3.0817004033075083e-07, + "loss": 0.8952, + "step": 39184 + }, + { + "epoch": 0.92, + "grad_norm": 2.050493511949606, + "learning_rate": 3.0798209957840886e-07, + "loss": 0.8913, + "step": 39185 + }, + { + "epoch": 0.92, + "grad_norm": 2.228918229239537, + "learning_rate": 3.077942152560631e-07, + "loss": 0.994, + "step": 39186 + }, + { + "epoch": 0.92, + "grad_norm": 1.0539358359076545, + "learning_rate": 3.076063873648094e-07, + "loss": 0.8687, + "step": 39187 + }, + { + "epoch": 0.92, + "grad_norm": 2.2312588647207505, + "learning_rate": 3.074186159057402e-07, + "loss": 0.9855, + "step": 39188 + }, + { + "epoch": 0.92, + "grad_norm": 2.176640662715678, + "learning_rate": 3.0723090087994676e-07, + "loss": 0.9392, + "step": 39189 + }, + { + "epoch": 0.92, + "grad_norm": 1.9926736010128472, + "learning_rate": 3.070432422885239e-07, + "loss": 1.025, + "step": 39190 + }, + { + "epoch": 0.92, + "grad_norm": 1.8721014943309404, + "learning_rate": 3.068556401325651e-07, + "loss": 0.9662, + "step": 39191 + }, + { + "epoch": 0.92, + "grad_norm": 2.870436741260945, + "learning_rate": 3.066680944131617e-07, + "loss": 1.0565, + "step": 39192 + }, + { + "epoch": 0.92, + "grad_norm": 2.0284647460850733, + "learning_rate": 3.06480605131404e-07, + "loss": 0.9625, + "step": 39193 + }, + { + "epoch": 0.92, + "grad_norm": 1.853046546617907, + "learning_rate": 3.062931722883866e-07, + "loss": 0.9262, + "step": 39194 + }, + { + "epoch": 0.92, + "grad_norm": 1.9921557492876878, + "learning_rate": 3.0610579588519874e-07, + "loss": 0.9687, + "step": 39195 + }, + { + "epoch": 0.92, + "grad_norm": 2.0680112903814534, + "learning_rate": 3.0591847592293277e-07, + "loss": 1.0355, + "step": 39196 + }, + { + "epoch": 0.92, + "grad_norm": 2.0670188930153865, + "learning_rate": 3.057312124026768e-07, + "loss": 0.9364, + "step": 39197 + }, + { + "epoch": 0.92, + "grad_norm": 2.0894983825116444, + "learning_rate": 3.055440053255243e-07, + "loss": 0.856, + "step": 39198 + }, + { + "epoch": 0.92, + "grad_norm": 2.0308607473532034, + "learning_rate": 3.0535685469256336e-07, + "loss": 0.9456, + "step": 39199 + }, + { + "epoch": 0.92, + "grad_norm": 2.469101833313527, + "learning_rate": 3.051697605048842e-07, + "loss": 0.9605, + "step": 39200 + }, + { + "epoch": 0.92, + "grad_norm": 2.0694968220438112, + "learning_rate": 3.049827227635771e-07, + "loss": 0.9815, + "step": 39201 + }, + { + "epoch": 0.92, + "grad_norm": 1.0301774073994845, + "learning_rate": 3.0479574146972777e-07, + "loss": 0.9352, + "step": 39202 + }, + { + "epoch": 0.92, + "grad_norm": 1.7875124304846055, + "learning_rate": 3.046088166244276e-07, + "loss": 0.9663, + "step": 39203 + }, + { + "epoch": 0.92, + "grad_norm": 1.2132714832272655, + "learning_rate": 3.044219482287647e-07, + "loss": 0.9167, + "step": 39204 + }, + { + "epoch": 0.92, + "grad_norm": 1.932164898263115, + "learning_rate": 3.0423513628382805e-07, + "loss": 0.9812, + "step": 39205 + }, + { + "epoch": 0.92, + "grad_norm": 1.967691830985134, + "learning_rate": 3.040483807907024e-07, + "loss": 0.8525, + "step": 39206 + }, + { + "epoch": 0.92, + "grad_norm": 2.0202197530530714, + "learning_rate": 3.0386168175047694e-07, + "loss": 1.0727, + "step": 39207 + }, + { + "epoch": 0.92, + "grad_norm": 1.8370905747750472, + "learning_rate": 3.0367503916423956e-07, + "loss": 0.9509, + "step": 39208 + }, + { + "epoch": 0.92, + "grad_norm": 2.438771210583855, + "learning_rate": 3.034884530330762e-07, + "loss": 0.8939, + "step": 39209 + }, + { + "epoch": 0.92, + "grad_norm": 1.965314515906527, + "learning_rate": 3.033019233580714e-07, + "loss": 0.9901, + "step": 39210 + }, + { + "epoch": 0.92, + "grad_norm": 1.9243215511478775, + "learning_rate": 3.0311545014031327e-07, + "loss": 0.976, + "step": 39211 + }, + { + "epoch": 0.92, + "grad_norm": 2.619278170551517, + "learning_rate": 3.0292903338088654e-07, + "loss": 0.93, + "step": 39212 + }, + { + "epoch": 0.92, + "grad_norm": 2.319639091957556, + "learning_rate": 3.0274267308087913e-07, + "loss": 1.0075, + "step": 39213 + }, + { + "epoch": 0.92, + "grad_norm": 1.947464354321458, + "learning_rate": 3.0255636924137134e-07, + "loss": 0.9207, + "step": 39214 + }, + { + "epoch": 0.92, + "grad_norm": 2.307647502783182, + "learning_rate": 3.0237012186345225e-07, + "loss": 0.9228, + "step": 39215 + }, + { + "epoch": 0.92, + "grad_norm": 1.1585368008276102, + "learning_rate": 3.021839309482033e-07, + "loss": 0.9856, + "step": 39216 + }, + { + "epoch": 0.92, + "grad_norm": 1.9836566048346282, + "learning_rate": 3.0199779649670915e-07, + "loss": 1.0853, + "step": 39217 + }, + { + "epoch": 0.92, + "grad_norm": 1.810663769244661, + "learning_rate": 3.018117185100566e-07, + "loss": 1.022, + "step": 39218 + }, + { + "epoch": 0.92, + "grad_norm": 1.8275527222664987, + "learning_rate": 3.016256969893261e-07, + "loss": 1.0884, + "step": 39219 + }, + { + "epoch": 0.92, + "grad_norm": 1.9694455908105712, + "learning_rate": 3.0143973193559994e-07, + "loss": 0.9928, + "step": 39220 + }, + { + "epoch": 0.92, + "grad_norm": 1.0630601583526869, + "learning_rate": 3.012538233499618e-07, + "loss": 0.8975, + "step": 39221 + }, + { + "epoch": 0.92, + "grad_norm": 1.8942906187555724, + "learning_rate": 3.010679712334974e-07, + "loss": 0.9027, + "step": 39222 + }, + { + "epoch": 0.92, + "grad_norm": 2.679726203292798, + "learning_rate": 3.008821755872826e-07, + "loss": 0.9744, + "step": 39223 + }, + { + "epoch": 0.92, + "grad_norm": 2.0309333238053955, + "learning_rate": 3.006964364124032e-07, + "loss": 0.9922, + "step": 39224 + }, + { + "epoch": 0.92, + "grad_norm": 2.986402764597636, + "learning_rate": 3.005107537099405e-07, + "loss": 0.8912, + "step": 39225 + }, + { + "epoch": 0.92, + "grad_norm": 1.0421809555726484, + "learning_rate": 3.0032512748097595e-07, + "loss": 0.9688, + "step": 39226 + }, + { + "epoch": 0.92, + "grad_norm": 2.1060332869988665, + "learning_rate": 3.0013955772658864e-07, + "loss": 0.9577, + "step": 39227 + }, + { + "epoch": 0.92, + "grad_norm": 2.0265221580167943, + "learning_rate": 2.999540444478588e-07, + "loss": 0.9762, + "step": 39228 + }, + { + "epoch": 0.92, + "grad_norm": 2.08676449583805, + "learning_rate": 2.997685876458689e-07, + "loss": 1.0959, + "step": 39229 + }, + { + "epoch": 0.92, + "grad_norm": 2.0054006879706665, + "learning_rate": 2.9958318732169587e-07, + "loss": 0.9231, + "step": 39230 + }, + { + "epoch": 0.92, + "grad_norm": 2.1582082465702235, + "learning_rate": 2.993978434764222e-07, + "loss": 1.0424, + "step": 39231 + }, + { + "epoch": 0.92, + "grad_norm": 1.0599861867064795, + "learning_rate": 2.9921255611112363e-07, + "loss": 0.993, + "step": 39232 + }, + { + "epoch": 0.92, + "grad_norm": 2.2467358572160565, + "learning_rate": 2.9902732522688273e-07, + "loss": 0.9717, + "step": 39233 + }, + { + "epoch": 0.92, + "grad_norm": 1.9470741507603062, + "learning_rate": 2.988421508247741e-07, + "loss": 1.0664, + "step": 39234 + }, + { + "epoch": 0.92, + "grad_norm": 2.007660853968909, + "learning_rate": 2.986570329058791e-07, + "loss": 1.0748, + "step": 39235 + }, + { + "epoch": 0.92, + "grad_norm": 2.171462803235293, + "learning_rate": 2.9847197147127473e-07, + "loss": 0.9636, + "step": 39236 + }, + { + "epoch": 0.92, + "grad_norm": 1.104667143578517, + "learning_rate": 2.982869665220367e-07, + "loss": 0.9132, + "step": 39237 + }, + { + "epoch": 0.92, + "grad_norm": 2.4165026183364837, + "learning_rate": 2.981020180592431e-07, + "loss": 0.9699, + "step": 39238 + }, + { + "epoch": 0.92, + "grad_norm": 1.0192057759848354, + "learning_rate": 2.9791712608397306e-07, + "loss": 0.9061, + "step": 39239 + }, + { + "epoch": 0.92, + "grad_norm": 0.985163746373442, + "learning_rate": 2.977322905973001e-07, + "loss": 0.9135, + "step": 39240 + }, + { + "epoch": 0.92, + "grad_norm": 1.9758309961526828, + "learning_rate": 2.975475116003013e-07, + "loss": 1.1292, + "step": 39241 + }, + { + "epoch": 0.92, + "grad_norm": 1.8332693715372328, + "learning_rate": 2.973627890940522e-07, + "loss": 0.8474, + "step": 39242 + }, + { + "epoch": 0.92, + "grad_norm": 2.071757889534234, + "learning_rate": 2.971781230796289e-07, + "loss": 0.904, + "step": 39243 + }, + { + "epoch": 0.92, + "grad_norm": 2.0033210313259597, + "learning_rate": 2.969935135581081e-07, + "loss": 0.9748, + "step": 39244 + }, + { + "epoch": 0.92, + "grad_norm": 2.1100928432857393, + "learning_rate": 2.968089605305613e-07, + "loss": 0.9787, + "step": 39245 + }, + { + "epoch": 0.92, + "grad_norm": 1.7857600303308392, + "learning_rate": 2.966244639980642e-07, + "loss": 1.0419, + "step": 39246 + }, + { + "epoch": 0.92, + "grad_norm": 2.119013941401777, + "learning_rate": 2.964400239616927e-07, + "loss": 1.0311, + "step": 39247 + }, + { + "epoch": 0.92, + "grad_norm": 1.8718732415228434, + "learning_rate": 2.962556404225192e-07, + "loss": 0.9642, + "step": 39248 + }, + { + "epoch": 0.92, + "grad_norm": 1.862618381548142, + "learning_rate": 2.9607131338161733e-07, + "loss": 0.9246, + "step": 39249 + }, + { + "epoch": 0.92, + "grad_norm": 2.354205125353961, + "learning_rate": 2.9588704284006176e-07, + "loss": 0.9616, + "step": 39250 + }, + { + "epoch": 0.92, + "grad_norm": 2.11619048285842, + "learning_rate": 2.9570282879892166e-07, + "loss": 0.9044, + "step": 39251 + }, + { + "epoch": 0.92, + "grad_norm": 2.2040457846285775, + "learning_rate": 2.955186712592739e-07, + "loss": 0.9732, + "step": 39252 + }, + { + "epoch": 0.92, + "grad_norm": 2.2168604276596646, + "learning_rate": 2.953345702221877e-07, + "loss": 1.0253, + "step": 39253 + }, + { + "epoch": 0.92, + "grad_norm": 1.8049621195046994, + "learning_rate": 2.951505256887377e-07, + "loss": 1.0563, + "step": 39254 + }, + { + "epoch": 0.92, + "grad_norm": 2.123711073696483, + "learning_rate": 2.9496653765999196e-07, + "loss": 1.0239, + "step": 39255 + }, + { + "epoch": 0.92, + "grad_norm": 2.105712207572937, + "learning_rate": 2.9478260613702513e-07, + "loss": 1.0567, + "step": 39256 + }, + { + "epoch": 0.92, + "grad_norm": 2.0788229663388047, + "learning_rate": 2.9459873112090644e-07, + "loss": 1.0423, + "step": 39257 + }, + { + "epoch": 0.92, + "grad_norm": 2.3455891036055094, + "learning_rate": 2.944149126127049e-07, + "loss": 0.9856, + "step": 39258 + }, + { + "epoch": 0.92, + "grad_norm": 2.065754906434126, + "learning_rate": 2.9423115061349315e-07, + "loss": 0.9573, + "step": 39259 + }, + { + "epoch": 0.92, + "grad_norm": 2.0543960850875633, + "learning_rate": 2.940474451243414e-07, + "loss": 0.925, + "step": 39260 + }, + { + "epoch": 0.92, + "grad_norm": 2.0290734851479724, + "learning_rate": 2.9386379614631866e-07, + "loss": 0.839, + "step": 39261 + }, + { + "epoch": 0.92, + "grad_norm": 1.9592466462862101, + "learning_rate": 2.93680203680492e-07, + "loss": 1.0576, + "step": 39262 + }, + { + "epoch": 0.93, + "grad_norm": 1.8905339786839332, + "learning_rate": 2.934966677279327e-07, + "loss": 0.9138, + "step": 39263 + }, + { + "epoch": 0.93, + "grad_norm": 2.0525933270196726, + "learning_rate": 2.9331318828971e-07, + "loss": 0.8011, + "step": 39264 + }, + { + "epoch": 0.93, + "grad_norm": 1.172998530901436, + "learning_rate": 2.9312976536689076e-07, + "loss": 0.946, + "step": 39265 + }, + { + "epoch": 0.93, + "grad_norm": 2.112166550224699, + "learning_rate": 2.9294639896054187e-07, + "loss": 0.9263, + "step": 39266 + }, + { + "epoch": 0.93, + "grad_norm": 2.027944950677144, + "learning_rate": 2.927630890717326e-07, + "loss": 0.991, + "step": 39267 + }, + { + "epoch": 0.93, + "grad_norm": 1.8878286496408025, + "learning_rate": 2.925798357015308e-07, + "loss": 0.8815, + "step": 39268 + }, + { + "epoch": 0.93, + "grad_norm": 1.9957695041411423, + "learning_rate": 2.9239663885100244e-07, + "loss": 1.0196, + "step": 39269 + }, + { + "epoch": 0.93, + "grad_norm": 2.0693659939997686, + "learning_rate": 2.922134985212133e-07, + "loss": 1.0283, + "step": 39270 + }, + { + "epoch": 0.93, + "grad_norm": 1.77920172800203, + "learning_rate": 2.9203041471323136e-07, + "loss": 0.9773, + "step": 39271 + }, + { + "epoch": 0.93, + "grad_norm": 2.0514804163703246, + "learning_rate": 2.918473874281214e-07, + "loss": 0.9747, + "step": 39272 + }, + { + "epoch": 0.93, + "grad_norm": 1.120564218771676, + "learning_rate": 2.9166441666695023e-07, + "loss": 0.9653, + "step": 39273 + }, + { + "epoch": 0.93, + "grad_norm": 1.9930245104960127, + "learning_rate": 2.9148150243078154e-07, + "loss": 0.9661, + "step": 39274 + }, + { + "epoch": 0.93, + "grad_norm": 1.9320681469377372, + "learning_rate": 2.912986447206811e-07, + "loss": 0.9377, + "step": 39275 + }, + { + "epoch": 0.93, + "grad_norm": 2.0105395551935152, + "learning_rate": 2.911158435377137e-07, + "loss": 0.9893, + "step": 39276 + }, + { + "epoch": 0.93, + "grad_norm": 2.0769057467240217, + "learning_rate": 2.9093309888294507e-07, + "loss": 0.9734, + "step": 39277 + }, + { + "epoch": 0.93, + "grad_norm": 2.328723228664191, + "learning_rate": 2.907504107574377e-07, + "loss": 1.0247, + "step": 39278 + }, + { + "epoch": 0.93, + "grad_norm": 1.934463769738827, + "learning_rate": 2.90567779162253e-07, + "loss": 0.9522, + "step": 39279 + }, + { + "epoch": 0.93, + "grad_norm": 2.0381324402448664, + "learning_rate": 2.9038520409845785e-07, + "loss": 0.995, + "step": 39280 + }, + { + "epoch": 0.93, + "grad_norm": 2.0866313868060025, + "learning_rate": 2.9020268556711475e-07, + "loss": 0.8961, + "step": 39281 + }, + { + "epoch": 0.93, + "grad_norm": 1.6867406639760059, + "learning_rate": 2.9002022356928615e-07, + "loss": 0.9081, + "step": 39282 + }, + { + "epoch": 0.93, + "grad_norm": 2.9751668853052164, + "learning_rate": 2.898378181060324e-07, + "loss": 0.9828, + "step": 39283 + }, + { + "epoch": 0.93, + "grad_norm": 2.4036697582582156, + "learning_rate": 2.8965546917841705e-07, + "loss": 0.9483, + "step": 39284 + }, + { + "epoch": 0.93, + "grad_norm": 2.0295900456859264, + "learning_rate": 2.894731767875025e-07, + "loss": 0.9815, + "step": 39285 + }, + { + "epoch": 0.93, + "grad_norm": 1.006358850639897, + "learning_rate": 2.892909409343503e-07, + "loss": 0.9434, + "step": 39286 + }, + { + "epoch": 0.93, + "grad_norm": 2.0407066819236213, + "learning_rate": 2.8910876162001835e-07, + "loss": 1.0804, + "step": 39287 + }, + { + "epoch": 0.93, + "grad_norm": 1.8966039776778338, + "learning_rate": 2.8892663884557024e-07, + "loss": 1.0377, + "step": 39288 + }, + { + "epoch": 0.93, + "grad_norm": 2.1684362577574157, + "learning_rate": 2.8874457261206743e-07, + "loss": 0.9712, + "step": 39289 + }, + { + "epoch": 0.93, + "grad_norm": 1.9453600830746676, + "learning_rate": 2.8856256292056797e-07, + "loss": 1.0035, + "step": 39290 + }, + { + "epoch": 0.93, + "grad_norm": 2.1849832881972007, + "learning_rate": 2.883806097721298e-07, + "loss": 0.9383, + "step": 39291 + }, + { + "epoch": 0.93, + "grad_norm": 2.1950998417511536, + "learning_rate": 2.8819871316781543e-07, + "loss": 0.91, + "step": 39292 + }, + { + "epoch": 0.93, + "grad_norm": 1.9343244920208365, + "learning_rate": 2.88016873108683e-07, + "loss": 0.9147, + "step": 39293 + }, + { + "epoch": 0.93, + "grad_norm": 1.81843467747522, + "learning_rate": 2.8783508959579153e-07, + "loss": 1.0526, + "step": 39294 + }, + { + "epoch": 0.93, + "grad_norm": 2.089864406082533, + "learning_rate": 2.8765336263019805e-07, + "loss": 0.9205, + "step": 39295 + }, + { + "epoch": 0.93, + "grad_norm": 1.9733064650412506, + "learning_rate": 2.874716922129628e-07, + "loss": 0.944, + "step": 39296 + }, + { + "epoch": 0.93, + "grad_norm": 1.1029346181388768, + "learning_rate": 2.8729007834514157e-07, + "loss": 0.9622, + "step": 39297 + }, + { + "epoch": 0.93, + "grad_norm": 2.282454502131046, + "learning_rate": 2.871085210277924e-07, + "loss": 1.1306, + "step": 39298 + }, + { + "epoch": 0.93, + "grad_norm": 1.921066033843856, + "learning_rate": 2.869270202619723e-07, + "loss": 1.1329, + "step": 39299 + }, + { + "epoch": 0.93, + "grad_norm": 1.9447145930363703, + "learning_rate": 2.867455760487392e-07, + "loss": 1.1546, + "step": 39300 + }, + { + "epoch": 0.93, + "grad_norm": 2.706484463475913, + "learning_rate": 2.865641883891468e-07, + "loss": 0.9252, + "step": 39301 + }, + { + "epoch": 0.93, + "grad_norm": 2.0181777981276205, + "learning_rate": 2.863828572842542e-07, + "loss": 0.899, + "step": 39302 + }, + { + "epoch": 0.93, + "grad_norm": 2.0166322204716334, + "learning_rate": 2.862015827351172e-07, + "loss": 1.1404, + "step": 39303 + }, + { + "epoch": 0.93, + "grad_norm": 2.1500667246663956, + "learning_rate": 2.860203647427873e-07, + "loss": 1.0771, + "step": 39304 + }, + { + "epoch": 0.93, + "grad_norm": 1.9739915242441028, + "learning_rate": 2.8583920330832237e-07, + "loss": 0.9382, + "step": 39305 + }, + { + "epoch": 0.93, + "grad_norm": 2.065776517056134, + "learning_rate": 2.8565809843277726e-07, + "loss": 0.9254, + "step": 39306 + }, + { + "epoch": 0.93, + "grad_norm": 1.0611118023667276, + "learning_rate": 2.8547705011720885e-07, + "loss": 0.9522, + "step": 39307 + }, + { + "epoch": 0.93, + "grad_norm": 3.0433952851842885, + "learning_rate": 2.8529605836266625e-07, + "loss": 1.0221, + "step": 39308 + }, + { + "epoch": 0.93, + "grad_norm": 2.1405062224855267, + "learning_rate": 2.851151231702054e-07, + "loss": 0.8945, + "step": 39309 + }, + { + "epoch": 0.93, + "grad_norm": 1.923997655469158, + "learning_rate": 2.849342445408809e-07, + "loss": 1.1041, + "step": 39310 + }, + { + "epoch": 0.93, + "grad_norm": 1.8684710045078035, + "learning_rate": 2.8475342247574535e-07, + "loss": 0.9999, + "step": 39311 + }, + { + "epoch": 0.93, + "grad_norm": 2.602011403902376, + "learning_rate": 2.8457265697585e-07, + "loss": 0.8421, + "step": 39312 + }, + { + "epoch": 0.93, + "grad_norm": 2.0598456168001706, + "learning_rate": 2.8439194804224857e-07, + "loss": 0.9442, + "step": 39313 + }, + { + "epoch": 0.93, + "grad_norm": 2.0412884622260132, + "learning_rate": 2.842112956759946e-07, + "loss": 0.9514, + "step": 39314 + }, + { + "epoch": 0.93, + "grad_norm": 1.7583339050317086, + "learning_rate": 2.8403069987813613e-07, + "loss": 1.0505, + "step": 39315 + }, + { + "epoch": 0.93, + "grad_norm": 2.2195184020966634, + "learning_rate": 2.8385016064972905e-07, + "loss": 1.0714, + "step": 39316 + }, + { + "epoch": 0.93, + "grad_norm": 2.466554706902703, + "learning_rate": 2.836696779918224e-07, + "loss": 0.8795, + "step": 39317 + }, + { + "epoch": 0.93, + "grad_norm": 2.0958485231818202, + "learning_rate": 2.834892519054655e-07, + "loss": 0.9949, + "step": 39318 + }, + { + "epoch": 0.93, + "grad_norm": 1.0732251596358262, + "learning_rate": 2.8330888239171074e-07, + "loss": 0.9534, + "step": 39319 + }, + { + "epoch": 0.93, + "grad_norm": 2.022081180139102, + "learning_rate": 2.831285694516084e-07, + "loss": 0.8877, + "step": 39320 + }, + { + "epoch": 0.93, + "grad_norm": 1.972951052609676, + "learning_rate": 2.829483130862076e-07, + "loss": 0.8487, + "step": 39321 + }, + { + "epoch": 0.93, + "grad_norm": 1.152065407749955, + "learning_rate": 2.8276811329655764e-07, + "loss": 0.9267, + "step": 39322 + }, + { + "epoch": 0.93, + "grad_norm": 2.098968246402929, + "learning_rate": 2.8258797008370865e-07, + "loss": 1.1795, + "step": 39323 + }, + { + "epoch": 0.93, + "grad_norm": 1.089600039521991, + "learning_rate": 2.8240788344871094e-07, + "loss": 0.9292, + "step": 39324 + }, + { + "epoch": 0.93, + "grad_norm": 2.0062066690930087, + "learning_rate": 2.822278533926082e-07, + "loss": 0.9585, + "step": 39325 + }, + { + "epoch": 0.93, + "grad_norm": 2.3173256717638813, + "learning_rate": 2.8204787991645163e-07, + "loss": 1.0159, + "step": 39326 + }, + { + "epoch": 0.93, + "grad_norm": 1.8974844514020066, + "learning_rate": 2.818679630212895e-07, + "loss": 0.9489, + "step": 39327 + }, + { + "epoch": 0.93, + "grad_norm": 1.8950958541287393, + "learning_rate": 2.816881027081697e-07, + "loss": 0.8226, + "step": 39328 + }, + { + "epoch": 0.93, + "grad_norm": 2.2340066243390435, + "learning_rate": 2.8150829897813814e-07, + "loss": 0.8929, + "step": 39329 + }, + { + "epoch": 0.93, + "grad_norm": 1.0887800619097, + "learning_rate": 2.813285518322417e-07, + "loss": 0.9304, + "step": 39330 + }, + { + "epoch": 0.93, + "grad_norm": 2.0474703418688924, + "learning_rate": 2.811488612715274e-07, + "loss": 0.9722, + "step": 39331 + }, + { + "epoch": 0.93, + "grad_norm": 1.060026907655932, + "learning_rate": 2.80969227297041e-07, + "loss": 0.9068, + "step": 39332 + }, + { + "epoch": 0.93, + "grad_norm": 2.190857840693395, + "learning_rate": 2.807896499098295e-07, + "loss": 1.0326, + "step": 39333 + }, + { + "epoch": 0.93, + "grad_norm": 2.119130821798652, + "learning_rate": 2.8061012911093756e-07, + "loss": 0.9038, + "step": 39334 + }, + { + "epoch": 0.93, + "grad_norm": 1.8663454306749978, + "learning_rate": 2.804306649014099e-07, + "loss": 0.9277, + "step": 39335 + }, + { + "epoch": 0.93, + "grad_norm": 1.9904187050878477, + "learning_rate": 2.802512572822924e-07, + "loss": 1.0243, + "step": 39336 + }, + { + "epoch": 0.93, + "grad_norm": 1.062684079573412, + "learning_rate": 2.800719062546298e-07, + "loss": 0.9233, + "step": 39337 + }, + { + "epoch": 0.93, + "grad_norm": 2.133812513920912, + "learning_rate": 2.798926118194667e-07, + "loss": 1.1164, + "step": 39338 + }, + { + "epoch": 0.93, + "grad_norm": 1.8437159552678937, + "learning_rate": 2.797133739778435e-07, + "loss": 0.8771, + "step": 39339 + }, + { + "epoch": 0.93, + "grad_norm": 1.0845455173541312, + "learning_rate": 2.795341927308082e-07, + "loss": 0.9096, + "step": 39340 + }, + { + "epoch": 0.93, + "grad_norm": 2.3362114544924553, + "learning_rate": 2.793550680794022e-07, + "loss": 1.0187, + "step": 39341 + }, + { + "epoch": 0.93, + "grad_norm": 1.9045284598090317, + "learning_rate": 2.79176000024669e-07, + "loss": 1.0187, + "step": 39342 + }, + { + "epoch": 0.93, + "grad_norm": 2.756878270112374, + "learning_rate": 2.789969885676491e-07, + "loss": 1.0412, + "step": 39343 + }, + { + "epoch": 0.93, + "grad_norm": 2.2007287596153087, + "learning_rate": 2.78818033709386e-07, + "loss": 0.8646, + "step": 39344 + }, + { + "epoch": 0.93, + "grad_norm": 1.875820483308204, + "learning_rate": 2.7863913545092434e-07, + "loss": 1.0326, + "step": 39345 + }, + { + "epoch": 0.93, + "grad_norm": 1.9267649543497403, + "learning_rate": 2.784602937933023e-07, + "loss": 1.0033, + "step": 39346 + }, + { + "epoch": 0.93, + "grad_norm": 2.3713875536301954, + "learning_rate": 2.7828150873756124e-07, + "loss": 0.8457, + "step": 39347 + }, + { + "epoch": 0.93, + "grad_norm": 2.2459679612927643, + "learning_rate": 2.781027802847436e-07, + "loss": 0.9454, + "step": 39348 + }, + { + "epoch": 0.93, + "grad_norm": 2.039812999944289, + "learning_rate": 2.779241084358908e-07, + "loss": 0.8899, + "step": 39349 + }, + { + "epoch": 0.93, + "grad_norm": 2.138654141483205, + "learning_rate": 2.777454931920409e-07, + "loss": 1.0525, + "step": 39350 + }, + { + "epoch": 0.93, + "grad_norm": 2.035714947408364, + "learning_rate": 2.775669345542342e-07, + "loss": 0.8652, + "step": 39351 + }, + { + "epoch": 0.93, + "grad_norm": 1.9675241605883567, + "learning_rate": 2.773884325235121e-07, + "loss": 1.0084, + "step": 39352 + }, + { + "epoch": 0.93, + "grad_norm": 2.004300034417737, + "learning_rate": 2.772099871009115e-07, + "loss": 1.14, + "step": 39353 + }, + { + "epoch": 0.93, + "grad_norm": 1.1663599212839448, + "learning_rate": 2.770315982874727e-07, + "loss": 0.9257, + "step": 39354 + }, + { + "epoch": 0.93, + "grad_norm": 2.1062974663491776, + "learning_rate": 2.768532660842349e-07, + "loss": 1.0093, + "step": 39355 + }, + { + "epoch": 0.93, + "grad_norm": 2.1299589888799737, + "learning_rate": 2.76674990492235e-07, + "loss": 0.9187, + "step": 39356 + }, + { + "epoch": 0.93, + "grad_norm": 2.0638618305342007, + "learning_rate": 2.764967715125111e-07, + "loss": 0.9175, + "step": 39357 + }, + { + "epoch": 0.93, + "grad_norm": 2.0629432466771798, + "learning_rate": 2.763186091461034e-07, + "loss": 0.9008, + "step": 39358 + }, + { + "epoch": 0.93, + "grad_norm": 2.0615535823781084, + "learning_rate": 2.761405033940467e-07, + "loss": 0.9965, + "step": 39359 + }, + { + "epoch": 0.93, + "grad_norm": 2.2301154319252263, + "learning_rate": 2.759624542573769e-07, + "loss": 1.111, + "step": 39360 + }, + { + "epoch": 0.93, + "grad_norm": 1.9224340771837496, + "learning_rate": 2.7578446173713305e-07, + "loss": 0.92, + "step": 39361 + }, + { + "epoch": 0.93, + "grad_norm": 2.145465899580278, + "learning_rate": 2.7560652583435097e-07, + "loss": 0.9445, + "step": 39362 + }, + { + "epoch": 0.93, + "grad_norm": 2.539309146528875, + "learning_rate": 2.754286465500666e-07, + "loss": 1.0736, + "step": 39363 + }, + { + "epoch": 0.93, + "grad_norm": 1.7763796149758124, + "learning_rate": 2.7525082388531465e-07, + "loss": 0.9211, + "step": 39364 + }, + { + "epoch": 0.93, + "grad_norm": 2.265552628631459, + "learning_rate": 2.7507305784113094e-07, + "loss": 0.9339, + "step": 39365 + }, + { + "epoch": 0.93, + "grad_norm": 2.0016417846259067, + "learning_rate": 2.7489534841855235e-07, + "loss": 0.7764, + "step": 39366 + }, + { + "epoch": 0.93, + "grad_norm": 2.189716054810463, + "learning_rate": 2.747176956186115e-07, + "loss": 0.8953, + "step": 39367 + }, + { + "epoch": 0.93, + "grad_norm": 1.8512079972407225, + "learning_rate": 2.7454009944234195e-07, + "loss": 1.0276, + "step": 39368 + }, + { + "epoch": 0.93, + "grad_norm": 1.7710642660843614, + "learning_rate": 2.7436255989077844e-07, + "loss": 0.9294, + "step": 39369 + }, + { + "epoch": 0.93, + "grad_norm": 2.169060730750855, + "learning_rate": 2.741850769649579e-07, + "loss": 0.9394, + "step": 39370 + }, + { + "epoch": 0.93, + "grad_norm": 1.1101827132408488, + "learning_rate": 2.740076506659095e-07, + "loss": 1.0009, + "step": 39371 + }, + { + "epoch": 0.93, + "grad_norm": 2.0275363639976742, + "learning_rate": 2.738302809946669e-07, + "loss": 0.9931, + "step": 39372 + }, + { + "epoch": 0.93, + "grad_norm": 2.098345712429487, + "learning_rate": 2.736529679522659e-07, + "loss": 0.9763, + "step": 39373 + }, + { + "epoch": 0.93, + "grad_norm": 1.9289684736962112, + "learning_rate": 2.734757115397346e-07, + "loss": 0.9887, + "step": 39374 + }, + { + "epoch": 0.93, + "grad_norm": 1.9291553914743143, + "learning_rate": 2.732985117581077e-07, + "loss": 1.0328, + "step": 39375 + }, + { + "epoch": 0.93, + "grad_norm": 2.0399623741620156, + "learning_rate": 2.7312136860841666e-07, + "loss": 1.0946, + "step": 39376 + }, + { + "epoch": 0.93, + "grad_norm": 2.061008513634721, + "learning_rate": 2.729442820916928e-07, + "loss": 1.0121, + "step": 39377 + }, + { + "epoch": 0.93, + "grad_norm": 2.115902464969129, + "learning_rate": 2.7276725220896526e-07, + "loss": 0.9017, + "step": 39378 + }, + { + "epoch": 0.93, + "grad_norm": 2.2320624886779514, + "learning_rate": 2.7259027896126776e-07, + "loss": 0.8321, + "step": 39379 + }, + { + "epoch": 0.93, + "grad_norm": 2.1501991785396535, + "learning_rate": 2.7241336234962943e-07, + "loss": 0.8954, + "step": 39380 + }, + { + "epoch": 0.93, + "grad_norm": 2.2116192865475917, + "learning_rate": 2.722365023750795e-07, + "loss": 0.939, + "step": 39381 + }, + { + "epoch": 0.93, + "grad_norm": 1.9052543075819277, + "learning_rate": 2.7205969903864814e-07, + "loss": 1.0172, + "step": 39382 + }, + { + "epoch": 0.93, + "grad_norm": 1.90655920833832, + "learning_rate": 2.7188295234136575e-07, + "loss": 1.0242, + "step": 39383 + }, + { + "epoch": 0.93, + "grad_norm": 2.001977842690068, + "learning_rate": 2.717062622842603e-07, + "loss": 1.0285, + "step": 39384 + }, + { + "epoch": 0.93, + "grad_norm": 1.9635538897441671, + "learning_rate": 2.715296288683611e-07, + "loss": 0.8385, + "step": 39385 + }, + { + "epoch": 0.93, + "grad_norm": 1.9757675042914897, + "learning_rate": 2.7135305209469607e-07, + "loss": 0.9816, + "step": 39386 + }, + { + "epoch": 0.93, + "grad_norm": 1.7617706018417012, + "learning_rate": 2.7117653196429337e-07, + "loss": 0.9254, + "step": 39387 + }, + { + "epoch": 0.93, + "grad_norm": 1.9729345368502398, + "learning_rate": 2.7100006847818214e-07, + "loss": 1.0437, + "step": 39388 + }, + { + "epoch": 0.93, + "grad_norm": 1.7769983211401916, + "learning_rate": 2.7082366163738825e-07, + "loss": 0.8238, + "step": 39389 + }, + { + "epoch": 0.93, + "grad_norm": 2.1916089852369445, + "learning_rate": 2.7064731144293755e-07, + "loss": 0.9952, + "step": 39390 + }, + { + "epoch": 0.93, + "grad_norm": 1.7847407448146526, + "learning_rate": 2.7047101789586026e-07, + "loss": 0.9048, + "step": 39391 + }, + { + "epoch": 0.93, + "grad_norm": 1.999887261738263, + "learning_rate": 2.702947809971812e-07, + "loss": 1.0388, + "step": 39392 + }, + { + "epoch": 0.93, + "grad_norm": 1.0402158723450523, + "learning_rate": 2.7011860074792616e-07, + "loss": 0.945, + "step": 39393 + }, + { + "epoch": 0.93, + "grad_norm": 2.0911927753920856, + "learning_rate": 2.6994247714912103e-07, + "loss": 0.9857, + "step": 39394 + }, + { + "epoch": 0.93, + "grad_norm": 1.980373181227654, + "learning_rate": 2.697664102017905e-07, + "loss": 0.8613, + "step": 39395 + }, + { + "epoch": 0.93, + "grad_norm": 1.1482292347526382, + "learning_rate": 2.6959039990696046e-07, + "loss": 0.9433, + "step": 39396 + }, + { + "epoch": 0.93, + "grad_norm": 1.0743572289984713, + "learning_rate": 2.694144462656556e-07, + "loss": 0.8717, + "step": 39397 + }, + { + "epoch": 0.93, + "grad_norm": 2.00509101247213, + "learning_rate": 2.692385492789018e-07, + "loss": 0.9391, + "step": 39398 + }, + { + "epoch": 0.93, + "grad_norm": 2.023344521061659, + "learning_rate": 2.6906270894772047e-07, + "loss": 0.9932, + "step": 39399 + }, + { + "epoch": 0.93, + "grad_norm": 2.203969091276243, + "learning_rate": 2.688869252731374e-07, + "loss": 0.9416, + "step": 39400 + }, + { + "epoch": 0.93, + "grad_norm": 2.153229163093983, + "learning_rate": 2.687111982561763e-07, + "loss": 0.9393, + "step": 39401 + }, + { + "epoch": 0.93, + "grad_norm": 2.4466175983727423, + "learning_rate": 2.6853552789785855e-07, + "loss": 0.9271, + "step": 39402 + }, + { + "epoch": 0.93, + "grad_norm": 2.9840640408009547, + "learning_rate": 2.683599141992077e-07, + "loss": 0.9179, + "step": 39403 + }, + { + "epoch": 0.93, + "grad_norm": 1.0939188438396403, + "learning_rate": 2.6818435716124525e-07, + "loss": 0.9537, + "step": 39404 + }, + { + "epoch": 0.93, + "grad_norm": 1.9695727726120207, + "learning_rate": 2.680088567849981e-07, + "loss": 0.9424, + "step": 39405 + }, + { + "epoch": 0.93, + "grad_norm": 1.8374558293254495, + "learning_rate": 2.6783341307148103e-07, + "loss": 0.9029, + "step": 39406 + }, + { + "epoch": 0.93, + "grad_norm": 2.1148256757797577, + "learning_rate": 2.6765802602171987e-07, + "loss": 0.9333, + "step": 39407 + }, + { + "epoch": 0.93, + "grad_norm": 1.975442646744501, + "learning_rate": 2.674826956367349e-07, + "loss": 0.8373, + "step": 39408 + }, + { + "epoch": 0.93, + "grad_norm": 2.20371454379412, + "learning_rate": 2.6730742191754864e-07, + "loss": 1.0442, + "step": 39409 + }, + { + "epoch": 0.93, + "grad_norm": 2.193139601611051, + "learning_rate": 2.671322048651781e-07, + "loss": 0.9324, + "step": 39410 + }, + { + "epoch": 0.93, + "grad_norm": 2.046871235348974, + "learning_rate": 2.6695704448064575e-07, + "loss": 0.9986, + "step": 39411 + }, + { + "epoch": 0.93, + "grad_norm": 2.1545319250633668, + "learning_rate": 2.6678194076497187e-07, + "loss": 0.9632, + "step": 39412 + }, + { + "epoch": 0.93, + "grad_norm": 2.3311735873189057, + "learning_rate": 2.6660689371917346e-07, + "loss": 0.8118, + "step": 39413 + }, + { + "epoch": 0.93, + "grad_norm": 2.727303251807472, + "learning_rate": 2.6643190334427194e-07, + "loss": 0.8787, + "step": 39414 + }, + { + "epoch": 0.93, + "grad_norm": 2.0660725635802746, + "learning_rate": 2.662569696412842e-07, + "loss": 0.9921, + "step": 39415 + }, + { + "epoch": 0.93, + "grad_norm": 2.0986443885831876, + "learning_rate": 2.6608209261123176e-07, + "loss": 0.9319, + "step": 39416 + }, + { + "epoch": 0.93, + "grad_norm": 1.9867207641718858, + "learning_rate": 2.659072722551304e-07, + "loss": 0.8829, + "step": 39417 + }, + { + "epoch": 0.93, + "grad_norm": 1.1331467438536214, + "learning_rate": 2.657325085739992e-07, + "loss": 0.9508, + "step": 39418 + }, + { + "epoch": 0.93, + "grad_norm": 1.2111335620668648, + "learning_rate": 2.6555780156885533e-07, + "loss": 1.0056, + "step": 39419 + }, + { + "epoch": 0.93, + "grad_norm": 1.1468512994419902, + "learning_rate": 2.6538315124071454e-07, + "loss": 0.9096, + "step": 39420 + }, + { + "epoch": 0.93, + "grad_norm": 2.069539065897984, + "learning_rate": 2.652085575905949e-07, + "loss": 1.002, + "step": 39421 + }, + { + "epoch": 0.93, + "grad_norm": 1.8200137116091017, + "learning_rate": 2.6503402061951456e-07, + "loss": 0.9366, + "step": 39422 + }, + { + "epoch": 0.93, + "grad_norm": 1.971270383287881, + "learning_rate": 2.6485954032848814e-07, + "loss": 0.9553, + "step": 39423 + }, + { + "epoch": 0.93, + "grad_norm": 2.0224479143468477, + "learning_rate": 2.6468511671853047e-07, + "loss": 0.9121, + "step": 39424 + }, + { + "epoch": 0.93, + "grad_norm": 2.1096487238562025, + "learning_rate": 2.645107497906574e-07, + "loss": 0.9424, + "step": 39425 + }, + { + "epoch": 0.93, + "grad_norm": 1.8934614647704147, + "learning_rate": 2.6433643954588695e-07, + "loss": 1.0415, + "step": 39426 + }, + { + "epoch": 0.93, + "grad_norm": 2.05464265585415, + "learning_rate": 2.641621859852317e-07, + "loss": 1.0027, + "step": 39427 + }, + { + "epoch": 0.93, + "grad_norm": 1.9474065589983507, + "learning_rate": 2.6398798910970525e-07, + "loss": 0.9649, + "step": 39428 + }, + { + "epoch": 0.93, + "grad_norm": 1.8394147980065623, + "learning_rate": 2.638138489203235e-07, + "loss": 0.9694, + "step": 39429 + }, + { + "epoch": 0.93, + "grad_norm": 1.9504440181982916, + "learning_rate": 2.636397654181011e-07, + "loss": 1.1294, + "step": 39430 + }, + { + "epoch": 0.93, + "grad_norm": 2.0096624870870614, + "learning_rate": 2.6346573860404954e-07, + "loss": 0.968, + "step": 39431 + }, + { + "epoch": 0.93, + "grad_norm": 1.880217272061882, + "learning_rate": 2.6329176847918245e-07, + "loss": 0.927, + "step": 39432 + }, + { + "epoch": 0.93, + "grad_norm": 1.8781676083432886, + "learning_rate": 2.631178550445146e-07, + "loss": 0.9694, + "step": 39433 + }, + { + "epoch": 0.93, + "grad_norm": 1.028716426413381, + "learning_rate": 2.629439983010562e-07, + "loss": 0.8632, + "step": 39434 + }, + { + "epoch": 0.93, + "grad_norm": 2.224500891168754, + "learning_rate": 2.627701982498221e-07, + "loss": 0.8922, + "step": 39435 + }, + { + "epoch": 0.93, + "grad_norm": 2.0918980576426844, + "learning_rate": 2.625964548918203e-07, + "loss": 0.962, + "step": 39436 + }, + { + "epoch": 0.93, + "grad_norm": 1.0787103436808736, + "learning_rate": 2.6242276822806666e-07, + "loss": 0.9771, + "step": 39437 + }, + { + "epoch": 0.93, + "grad_norm": 1.8728592226577123, + "learning_rate": 2.6224913825956933e-07, + "loss": 0.9971, + "step": 39438 + }, + { + "epoch": 0.93, + "grad_norm": 2.1492272637710643, + "learning_rate": 2.6207556498734076e-07, + "loss": 1.0998, + "step": 39439 + }, + { + "epoch": 0.93, + "grad_norm": 1.0777058708825897, + "learning_rate": 2.6190204841239244e-07, + "loss": 0.9911, + "step": 39440 + }, + { + "epoch": 0.93, + "grad_norm": 1.1134634780650319, + "learning_rate": 2.6172858853573125e-07, + "loss": 0.8744, + "step": 39441 + }, + { + "epoch": 0.93, + "grad_norm": 2.0335328276700473, + "learning_rate": 2.6155518535836976e-07, + "loss": 0.9706, + "step": 39442 + }, + { + "epoch": 0.93, + "grad_norm": 2.85136922318468, + "learning_rate": 2.6138183888131827e-07, + "loss": 1.0428, + "step": 39443 + }, + { + "epoch": 0.93, + "grad_norm": 2.1445682976580835, + "learning_rate": 2.612085491055849e-07, + "loss": 1.0363, + "step": 39444 + }, + { + "epoch": 0.93, + "grad_norm": 1.985036102972976, + "learning_rate": 2.610353160321777e-07, + "loss": 0.8652, + "step": 39445 + }, + { + "epoch": 0.93, + "grad_norm": 2.0452746931870993, + "learning_rate": 2.6086213966210697e-07, + "loss": 1.0621, + "step": 39446 + }, + { + "epoch": 0.93, + "grad_norm": 2.11675590254444, + "learning_rate": 2.6068901999637965e-07, + "loss": 0.8647, + "step": 39447 + }, + { + "epoch": 0.93, + "grad_norm": 2.179367811825758, + "learning_rate": 2.6051595703600494e-07, + "loss": 0.9327, + "step": 39448 + }, + { + "epoch": 0.93, + "grad_norm": 1.9341777979686396, + "learning_rate": 2.603429507819888e-07, + "loss": 0.9049, + "step": 39449 + }, + { + "epoch": 0.93, + "grad_norm": 2.0615720682177274, + "learning_rate": 2.601700012353392e-07, + "loss": 0.9212, + "step": 39450 + }, + { + "epoch": 0.93, + "grad_norm": 5.509996234284417, + "learning_rate": 2.5999710839706537e-07, + "loss": 0.9616, + "step": 39451 + }, + { + "epoch": 0.93, + "grad_norm": 2.289812576221007, + "learning_rate": 2.59824272268171e-07, + "loss": 0.9982, + "step": 39452 + }, + { + "epoch": 0.93, + "grad_norm": 2.56265041417596, + "learning_rate": 2.5965149284966297e-07, + "loss": 0.9224, + "step": 39453 + }, + { + "epoch": 0.93, + "grad_norm": 2.20728614556834, + "learning_rate": 2.594787701425483e-07, + "loss": 0.9055, + "step": 39454 + }, + { + "epoch": 0.93, + "grad_norm": 1.9365904671283085, + "learning_rate": 2.5930610414783177e-07, + "loss": 0.8834, + "step": 39455 + }, + { + "epoch": 0.93, + "grad_norm": 1.9573830025034544, + "learning_rate": 2.591334948665192e-07, + "loss": 1.0027, + "step": 39456 + }, + { + "epoch": 0.93, + "grad_norm": 1.7998908548017922, + "learning_rate": 2.589609422996142e-07, + "loss": 1.016, + "step": 39457 + }, + { + "epoch": 0.93, + "grad_norm": 2.3106165220533073, + "learning_rate": 2.587884464481238e-07, + "loss": 1.0511, + "step": 39458 + }, + { + "epoch": 0.93, + "grad_norm": 1.8228097292565155, + "learning_rate": 2.586160073130506e-07, + "loss": 1.0079, + "step": 39459 + }, + { + "epoch": 0.93, + "grad_norm": 2.1946470547673718, + "learning_rate": 2.584436248954003e-07, + "loss": 0.8725, + "step": 39460 + }, + { + "epoch": 0.93, + "grad_norm": 1.0609779905367047, + "learning_rate": 2.582712991961744e-07, + "loss": 0.8385, + "step": 39461 + }, + { + "epoch": 0.93, + "grad_norm": 1.8768771702132894, + "learning_rate": 2.580990302163766e-07, + "loss": 0.8975, + "step": 39462 + }, + { + "epoch": 0.93, + "grad_norm": 2.087475985962474, + "learning_rate": 2.5792681795701046e-07, + "loss": 0.976, + "step": 39463 + }, + { + "epoch": 0.93, + "grad_norm": 1.9285161995056275, + "learning_rate": 2.577546624190797e-07, + "loss": 1.0231, + "step": 39464 + }, + { + "epoch": 0.93, + "grad_norm": 2.2976460172905417, + "learning_rate": 2.5758256360358444e-07, + "loss": 0.9267, + "step": 39465 + }, + { + "epoch": 0.93, + "grad_norm": 1.990856087119098, + "learning_rate": 2.5741052151152855e-07, + "loss": 1.0342, + "step": 39466 + }, + { + "epoch": 0.93, + "grad_norm": 2.151342246477695, + "learning_rate": 2.5723853614391225e-07, + "loss": 0.9468, + "step": 39467 + }, + { + "epoch": 0.93, + "grad_norm": 2.109733360256939, + "learning_rate": 2.5706660750173916e-07, + "loss": 0.9848, + "step": 39468 + }, + { + "epoch": 0.93, + "grad_norm": 2.0577055553865535, + "learning_rate": 2.568947355860085e-07, + "loss": 0.9856, + "step": 39469 + }, + { + "epoch": 0.93, + "grad_norm": 1.999958114492761, + "learning_rate": 2.567229203977206e-07, + "loss": 0.8787, + "step": 39470 + }, + { + "epoch": 0.93, + "grad_norm": 2.1336484170909706, + "learning_rate": 2.565511619378758e-07, + "loss": 0.998, + "step": 39471 + }, + { + "epoch": 0.93, + "grad_norm": 1.9252166264297383, + "learning_rate": 2.563794602074765e-07, + "loss": 1.048, + "step": 39472 + }, + { + "epoch": 0.93, + "grad_norm": 1.921301184680755, + "learning_rate": 2.562078152075209e-07, + "loss": 0.9033, + "step": 39473 + }, + { + "epoch": 0.93, + "grad_norm": 1.9210332062982642, + "learning_rate": 2.560362269390071e-07, + "loss": 0.874, + "step": 39474 + }, + { + "epoch": 0.93, + "grad_norm": 1.9726811389722114, + "learning_rate": 2.5586469540293643e-07, + "loss": 0.8973, + "step": 39475 + }, + { + "epoch": 0.93, + "grad_norm": 2.136518845248969, + "learning_rate": 2.5569322060030487e-07, + "loss": 0.9426, + "step": 39476 + }, + { + "epoch": 0.93, + "grad_norm": 2.0413464717816296, + "learning_rate": 2.555218025321138e-07, + "loss": 0.9052, + "step": 39477 + }, + { + "epoch": 0.93, + "grad_norm": 1.9725995185455962, + "learning_rate": 2.5535044119935903e-07, + "loss": 0.9539, + "step": 39478 + }, + { + "epoch": 0.93, + "grad_norm": 1.8678403656565479, + "learning_rate": 2.5517913660303984e-07, + "loss": 0.9208, + "step": 39479 + }, + { + "epoch": 0.93, + "grad_norm": 2.565833882833307, + "learning_rate": 2.5500788874415097e-07, + "loss": 0.8791, + "step": 39480 + }, + { + "epoch": 0.93, + "grad_norm": 2.0303727505721034, + "learning_rate": 2.5483669762369377e-07, + "loss": 1.0204, + "step": 39481 + }, + { + "epoch": 0.93, + "grad_norm": 2.1936617025725864, + "learning_rate": 2.54665563242662e-07, + "loss": 0.8836, + "step": 39482 + }, + { + "epoch": 0.93, + "grad_norm": 1.821603418455863, + "learning_rate": 2.544944856020515e-07, + "loss": 1.0386, + "step": 39483 + }, + { + "epoch": 0.93, + "grad_norm": 2.648379145794506, + "learning_rate": 2.543234647028603e-07, + "loss": 0.9567, + "step": 39484 + }, + { + "epoch": 0.93, + "grad_norm": 2.002306086082367, + "learning_rate": 2.541525005460821e-07, + "loss": 0.7806, + "step": 39485 + }, + { + "epoch": 0.93, + "grad_norm": 1.8393755248520374, + "learning_rate": 2.539815931327161e-07, + "loss": 0.8923, + "step": 39486 + }, + { + "epoch": 0.93, + "grad_norm": 1.9950229851385042, + "learning_rate": 2.538107424637526e-07, + "loss": 1.0537, + "step": 39487 + }, + { + "epoch": 0.93, + "grad_norm": 2.018445951511759, + "learning_rate": 2.536399485401897e-07, + "loss": 1.1064, + "step": 39488 + }, + { + "epoch": 0.93, + "grad_norm": 1.0489825333962441, + "learning_rate": 2.5346921136301995e-07, + "loss": 0.9504, + "step": 39489 + }, + { + "epoch": 0.93, + "grad_norm": 2.015250504285071, + "learning_rate": 2.532985309332392e-07, + "loss": 0.8253, + "step": 39490 + }, + { + "epoch": 0.93, + "grad_norm": 2.032764040882854, + "learning_rate": 2.531279072518389e-07, + "loss": 0.8732, + "step": 39491 + }, + { + "epoch": 0.93, + "grad_norm": 2.1544976762297607, + "learning_rate": 2.529573403198138e-07, + "loss": 1.0111, + "step": 39492 + }, + { + "epoch": 0.93, + "grad_norm": 2.2138357238590824, + "learning_rate": 2.5278683013815755e-07, + "loss": 0.9523, + "step": 39493 + }, + { + "epoch": 0.93, + "grad_norm": 2.0727924611804682, + "learning_rate": 2.5261637670786266e-07, + "loss": 0.9722, + "step": 39494 + }, + { + "epoch": 0.93, + "grad_norm": 2.0553153234904746, + "learning_rate": 2.524459800299206e-07, + "loss": 0.8777, + "step": 39495 + }, + { + "epoch": 0.93, + "grad_norm": 1.939541220825082, + "learning_rate": 2.522756401053239e-07, + "loss": 0.9362, + "step": 39496 + }, + { + "epoch": 0.93, + "grad_norm": 2.085231812992089, + "learning_rate": 2.521053569350651e-07, + "loss": 0.8804, + "step": 39497 + }, + { + "epoch": 0.93, + "grad_norm": 1.1300416320037217, + "learning_rate": 2.5193513052013343e-07, + "loss": 0.9804, + "step": 39498 + }, + { + "epoch": 0.93, + "grad_norm": 2.28988270581654, + "learning_rate": 2.517649608615236e-07, + "loss": 1.1213, + "step": 39499 + }, + { + "epoch": 0.93, + "grad_norm": 1.9133124376851331, + "learning_rate": 2.515948479602237e-07, + "loss": 0.905, + "step": 39500 + }, + { + "epoch": 0.93, + "grad_norm": 1.1630169746330927, + "learning_rate": 2.5142479181722525e-07, + "loss": 0.9902, + "step": 39501 + }, + { + "epoch": 0.93, + "grad_norm": 1.835893032430796, + "learning_rate": 2.512547924335162e-07, + "loss": 1.0216, + "step": 39502 + }, + { + "epoch": 0.93, + "grad_norm": 1.7847380236249715, + "learning_rate": 2.510848498100915e-07, + "loss": 0.9885, + "step": 39503 + }, + { + "epoch": 0.93, + "grad_norm": 2.425752823792864, + "learning_rate": 2.509149639479347e-07, + "loss": 0.897, + "step": 39504 + }, + { + "epoch": 0.93, + "grad_norm": 2.195898937686981, + "learning_rate": 2.507451348480372e-07, + "loss": 0.8716, + "step": 39505 + }, + { + "epoch": 0.93, + "grad_norm": 2.143898946711279, + "learning_rate": 2.505753625113894e-07, + "loss": 1.0575, + "step": 39506 + }, + { + "epoch": 0.93, + "grad_norm": 1.0785186051676747, + "learning_rate": 2.504056469389793e-07, + "loss": 0.9064, + "step": 39507 + }, + { + "epoch": 0.93, + "grad_norm": 2.1376563576097984, + "learning_rate": 2.502359881317917e-07, + "loss": 0.9095, + "step": 39508 + }, + { + "epoch": 0.93, + "grad_norm": 2.1418488638469837, + "learning_rate": 2.500663860908181e-07, + "loss": 0.9899, + "step": 39509 + }, + { + "epoch": 0.93, + "grad_norm": 2.33953905101854, + "learning_rate": 2.4989684081704545e-07, + "loss": 0.9746, + "step": 39510 + }, + { + "epoch": 0.93, + "grad_norm": 2.1357487811207867, + "learning_rate": 2.497273523114585e-07, + "loss": 1.071, + "step": 39511 + }, + { + "epoch": 0.93, + "grad_norm": 2.231676777389579, + "learning_rate": 2.495579205750476e-07, + "loss": 1.0483, + "step": 39512 + }, + { + "epoch": 0.93, + "grad_norm": 2.08945873899672, + "learning_rate": 2.493885456087963e-07, + "loss": 0.9448, + "step": 39513 + }, + { + "epoch": 0.93, + "grad_norm": 1.789664095508587, + "learning_rate": 2.4921922741369286e-07, + "loss": 0.9429, + "step": 39514 + }, + { + "epoch": 0.93, + "grad_norm": 1.9300447066530528, + "learning_rate": 2.4904996599072087e-07, + "loss": 0.8324, + "step": 39515 + }, + { + "epoch": 0.93, + "grad_norm": 1.8734125647628337, + "learning_rate": 2.4888076134086834e-07, + "loss": 0.9911, + "step": 39516 + }, + { + "epoch": 0.93, + "grad_norm": 1.9738260529772613, + "learning_rate": 2.4871161346511906e-07, + "loss": 0.8692, + "step": 39517 + }, + { + "epoch": 0.93, + "grad_norm": 2.0820187900199008, + "learning_rate": 2.4854252236445664e-07, + "loss": 1.0111, + "step": 39518 + }, + { + "epoch": 0.93, + "grad_norm": 1.765895033358704, + "learning_rate": 2.4837348803986803e-07, + "loss": 0.7879, + "step": 39519 + }, + { + "epoch": 0.93, + "grad_norm": 1.836679237955145, + "learning_rate": 2.4820451049233585e-07, + "loss": 0.9666, + "step": 39520 + }, + { + "epoch": 0.93, + "grad_norm": 2.1280663850901185, + "learning_rate": 2.480355897228448e-07, + "loss": 0.9581, + "step": 39521 + }, + { + "epoch": 0.93, + "grad_norm": 1.862837043855733, + "learning_rate": 2.4786672573237744e-07, + "loss": 1.0855, + "step": 39522 + }, + { + "epoch": 0.93, + "grad_norm": 1.9537137226693908, + "learning_rate": 2.476979185219175e-07, + "loss": 0.9788, + "step": 39523 + }, + { + "epoch": 0.93, + "grad_norm": 1.954269841596387, + "learning_rate": 2.4752916809244856e-07, + "loss": 1.0689, + "step": 39524 + }, + { + "epoch": 0.93, + "grad_norm": 1.1474047342699178, + "learning_rate": 2.4736047444495206e-07, + "loss": 0.9071, + "step": 39525 + }, + { + "epoch": 0.93, + "grad_norm": 1.9781190207788006, + "learning_rate": 2.4719183758041056e-07, + "loss": 0.9707, + "step": 39526 + }, + { + "epoch": 0.93, + "grad_norm": 1.862695515684517, + "learning_rate": 2.470232574998044e-07, + "loss": 0.9372, + "step": 39527 + }, + { + "epoch": 0.93, + "grad_norm": 1.9645078313368511, + "learning_rate": 2.468547342041183e-07, + "loss": 1.0259, + "step": 39528 + }, + { + "epoch": 0.93, + "grad_norm": 2.5746567937542557, + "learning_rate": 2.4668626769433155e-07, + "loss": 1.0121, + "step": 39529 + }, + { + "epoch": 0.93, + "grad_norm": 1.9044007330474095, + "learning_rate": 2.4651785797142447e-07, + "loss": 0.8969, + "step": 39530 + }, + { + "epoch": 0.93, + "grad_norm": 2.151612791803023, + "learning_rate": 2.4634950503637957e-07, + "loss": 0.9938, + "step": 39531 + }, + { + "epoch": 0.93, + "grad_norm": 2.320655944441257, + "learning_rate": 2.461812088901749e-07, + "loss": 1.087, + "step": 39532 + }, + { + "epoch": 0.93, + "grad_norm": 2.1957769797682714, + "learning_rate": 2.4601296953379094e-07, + "loss": 0.9585, + "step": 39533 + }, + { + "epoch": 0.93, + "grad_norm": 1.9041774665857443, + "learning_rate": 2.458447869682079e-07, + "loss": 0.9531, + "step": 39534 + }, + { + "epoch": 0.93, + "grad_norm": 2.000044897393225, + "learning_rate": 2.4567666119440504e-07, + "loss": 1.0146, + "step": 39535 + }, + { + "epoch": 0.93, + "grad_norm": 2.0059351495862594, + "learning_rate": 2.455085922133604e-07, + "loss": 0.8211, + "step": 39536 + }, + { + "epoch": 0.93, + "grad_norm": 2.047360855440449, + "learning_rate": 2.4534058002605334e-07, + "loss": 1.0756, + "step": 39537 + }, + { + "epoch": 0.93, + "grad_norm": 2.17965853974799, + "learning_rate": 2.451726246334607e-07, + "loss": 0.9555, + "step": 39538 + }, + { + "epoch": 0.93, + "grad_norm": 1.9206362063141562, + "learning_rate": 2.450047260365618e-07, + "loss": 0.9037, + "step": 39539 + }, + { + "epoch": 0.93, + "grad_norm": 1.8984384147961695, + "learning_rate": 2.4483688423633356e-07, + "loss": 1.0718, + "step": 39540 + }, + { + "epoch": 0.93, + "grad_norm": 1.076724492094976, + "learning_rate": 2.4466909923375414e-07, + "loss": 0.9602, + "step": 39541 + }, + { + "epoch": 0.93, + "grad_norm": 1.9469722551442135, + "learning_rate": 2.445013710298005e-07, + "loss": 1.0027, + "step": 39542 + }, + { + "epoch": 0.93, + "grad_norm": 2.0604809648810996, + "learning_rate": 2.443336996254464e-07, + "loss": 0.9588, + "step": 39543 + }, + { + "epoch": 0.93, + "grad_norm": 1.072944588680282, + "learning_rate": 2.4416608502167095e-07, + "loss": 0.8692, + "step": 39544 + }, + { + "epoch": 0.93, + "grad_norm": 2.082251916899096, + "learning_rate": 2.4399852721944894e-07, + "loss": 0.9773, + "step": 39545 + }, + { + "epoch": 0.93, + "grad_norm": 2.395662414305603, + "learning_rate": 2.438310262197574e-07, + "loss": 1.0337, + "step": 39546 + }, + { + "epoch": 0.93, + "grad_norm": 1.998538971526127, + "learning_rate": 2.4366358202356887e-07, + "loss": 0.944, + "step": 39547 + }, + { + "epoch": 0.93, + "grad_norm": 1.8599327117131574, + "learning_rate": 2.4349619463186036e-07, + "loss": 0.9655, + "step": 39548 + }, + { + "epoch": 0.93, + "grad_norm": 1.9979743422886378, + "learning_rate": 2.4332886404560553e-07, + "loss": 0.8977, + "step": 39549 + }, + { + "epoch": 0.93, + "grad_norm": 1.0748504337617872, + "learning_rate": 2.431615902657802e-07, + "loss": 0.9407, + "step": 39550 + }, + { + "epoch": 0.93, + "grad_norm": 2.645148171252062, + "learning_rate": 2.4299437329335595e-07, + "loss": 1.0719, + "step": 39551 + }, + { + "epoch": 0.93, + "grad_norm": 1.9219507041007198, + "learning_rate": 2.428272131293075e-07, + "loss": 1.0429, + "step": 39552 + }, + { + "epoch": 0.93, + "grad_norm": 1.9610253411495675, + "learning_rate": 2.426601097746084e-07, + "loss": 1.0432, + "step": 39553 + }, + { + "epoch": 0.93, + "grad_norm": 1.0727804896464548, + "learning_rate": 2.424930632302314e-07, + "loss": 0.9697, + "step": 39554 + }, + { + "epoch": 0.93, + "grad_norm": 2.92382915042268, + "learning_rate": 2.4232607349714775e-07, + "loss": 1.0257, + "step": 39555 + }, + { + "epoch": 0.93, + "grad_norm": 1.89867897322905, + "learning_rate": 2.421591405763324e-07, + "loss": 0.9599, + "step": 39556 + }, + { + "epoch": 0.93, + "grad_norm": 2.051615334579137, + "learning_rate": 2.4199226446875557e-07, + "loss": 0.9706, + "step": 39557 + }, + { + "epoch": 0.93, + "grad_norm": 2.00089087030499, + "learning_rate": 2.4182544517538874e-07, + "loss": 1.0817, + "step": 39558 + }, + { + "epoch": 0.93, + "grad_norm": 1.8649440956425927, + "learning_rate": 2.416586826972034e-07, + "loss": 0.9802, + "step": 39559 + }, + { + "epoch": 0.93, + "grad_norm": 1.965145264215841, + "learning_rate": 2.414919770351709e-07, + "loss": 0.9148, + "step": 39560 + }, + { + "epoch": 0.93, + "grad_norm": 1.1304348621754505, + "learning_rate": 2.4132532819026165e-07, + "loss": 0.951, + "step": 39561 + }, + { + "epoch": 0.93, + "grad_norm": 2.198549242300285, + "learning_rate": 2.41158736163446e-07, + "loss": 1.0782, + "step": 39562 + }, + { + "epoch": 0.93, + "grad_norm": 1.9976616086671648, + "learning_rate": 2.409922009556942e-07, + "loss": 1.0718, + "step": 39563 + }, + { + "epoch": 0.93, + "grad_norm": 1.987856745955932, + "learning_rate": 2.408257225679744e-07, + "loss": 1.1196, + "step": 39564 + }, + { + "epoch": 0.93, + "grad_norm": 1.7997227004301237, + "learning_rate": 2.40659301001257e-07, + "loss": 0.9314, + "step": 39565 + }, + { + "epoch": 0.93, + "grad_norm": 1.8265082465800575, + "learning_rate": 2.4049293625651224e-07, + "loss": 0.9208, + "step": 39566 + }, + { + "epoch": 0.93, + "grad_norm": 1.814967997135264, + "learning_rate": 2.403266283347061e-07, + "loss": 1.057, + "step": 39567 + }, + { + "epoch": 0.93, + "grad_norm": 1.9742444375147905, + "learning_rate": 2.4016037723680886e-07, + "loss": 1.1312, + "step": 39568 + }, + { + "epoch": 0.93, + "grad_norm": 1.0723848985517201, + "learning_rate": 2.3999418296378643e-07, + "loss": 0.9926, + "step": 39569 + }, + { + "epoch": 0.93, + "grad_norm": 1.8903365138613093, + "learning_rate": 2.398280455166091e-07, + "loss": 0.9638, + "step": 39570 + }, + { + "epoch": 0.93, + "grad_norm": 1.9640741214345783, + "learning_rate": 2.396619648962439e-07, + "loss": 0.9385, + "step": 39571 + }, + { + "epoch": 0.93, + "grad_norm": 1.8405633587055372, + "learning_rate": 2.394959411036546e-07, + "loss": 1.0489, + "step": 39572 + }, + { + "epoch": 0.93, + "grad_norm": 2.005383958791516, + "learning_rate": 2.393299741398114e-07, + "loss": 0.9408, + "step": 39573 + }, + { + "epoch": 0.93, + "grad_norm": 1.8091002355913182, + "learning_rate": 2.3916406400567915e-07, + "loss": 0.9104, + "step": 39574 + }, + { + "epoch": 0.93, + "grad_norm": 1.0988709547705042, + "learning_rate": 2.3899821070222375e-07, + "loss": 0.928, + "step": 39575 + }, + { + "epoch": 0.93, + "grad_norm": 1.91903634085978, + "learning_rate": 2.38832414230411e-07, + "loss": 0.9052, + "step": 39576 + }, + { + "epoch": 0.93, + "grad_norm": 2.161425198941115, + "learning_rate": 2.38666674591207e-07, + "loss": 1.0461, + "step": 39577 + }, + { + "epoch": 0.93, + "grad_norm": 2.038173423131428, + "learning_rate": 2.3850099178557517e-07, + "loss": 0.9483, + "step": 39578 + }, + { + "epoch": 0.93, + "grad_norm": 2.0578639911476433, + "learning_rate": 2.3833536581448157e-07, + "loss": 0.9629, + "step": 39579 + }, + { + "epoch": 0.93, + "grad_norm": 1.9602094542843744, + "learning_rate": 2.3816979667888872e-07, + "loss": 1.033, + "step": 39580 + }, + { + "epoch": 0.93, + "grad_norm": 2.022695974514814, + "learning_rate": 2.3800428437976364e-07, + "loss": 0.9369, + "step": 39581 + }, + { + "epoch": 0.93, + "grad_norm": 2.2399933267203043, + "learning_rate": 2.3783882891806664e-07, + "loss": 1.0277, + "step": 39582 + }, + { + "epoch": 0.93, + "grad_norm": 2.375993322492849, + "learning_rate": 2.3767343029476253e-07, + "loss": 1.0171, + "step": 39583 + }, + { + "epoch": 0.93, + "grad_norm": 1.8433476949686711, + "learning_rate": 2.3750808851081607e-07, + "loss": 0.9917, + "step": 39584 + }, + { + "epoch": 0.93, + "grad_norm": 2.293679365670583, + "learning_rate": 2.3734280356718764e-07, + "loss": 0.9479, + "step": 39585 + }, + { + "epoch": 0.93, + "grad_norm": 2.1158525095605953, + "learning_rate": 2.3717757546483866e-07, + "loss": 0.8826, + "step": 39586 + }, + { + "epoch": 0.93, + "grad_norm": 2.2106409317321294, + "learning_rate": 2.370124042047339e-07, + "loss": 1.0847, + "step": 39587 + }, + { + "epoch": 0.93, + "grad_norm": 1.8127176378541185, + "learning_rate": 2.368472897878349e-07, + "loss": 0.947, + "step": 39588 + }, + { + "epoch": 0.93, + "grad_norm": 1.9154356101418961, + "learning_rate": 2.3668223221510078e-07, + "loss": 0.8859, + "step": 39589 + }, + { + "epoch": 0.93, + "grad_norm": 2.3135561536863225, + "learning_rate": 2.3651723148749305e-07, + "loss": 0.9638, + "step": 39590 + }, + { + "epoch": 0.93, + "grad_norm": 2.053602896803838, + "learning_rate": 2.3635228760597429e-07, + "loss": 0.9479, + "step": 39591 + }, + { + "epoch": 0.93, + "grad_norm": 2.5831376524906133, + "learning_rate": 2.3618740057150368e-07, + "loss": 0.9304, + "step": 39592 + }, + { + "epoch": 0.93, + "grad_norm": 1.979172302134024, + "learning_rate": 2.360225703850405e-07, + "loss": 0.8821, + "step": 39593 + }, + { + "epoch": 0.93, + "grad_norm": 2.2287427035750573, + "learning_rate": 2.3585779704754507e-07, + "loss": 0.9608, + "step": 39594 + }, + { + "epoch": 0.93, + "grad_norm": 1.9979346266872355, + "learning_rate": 2.3569308055997776e-07, + "loss": 0.8689, + "step": 39595 + }, + { + "epoch": 0.93, + "grad_norm": 2.8580626428802534, + "learning_rate": 2.3552842092329554e-07, + "loss": 0.9738, + "step": 39596 + }, + { + "epoch": 0.93, + "grad_norm": 1.941462242779265, + "learning_rate": 2.3536381813845878e-07, + "loss": 0.9969, + "step": 39597 + }, + { + "epoch": 0.93, + "grad_norm": 2.094309597764468, + "learning_rate": 2.3519927220642558e-07, + "loss": 0.9789, + "step": 39598 + }, + { + "epoch": 0.93, + "grad_norm": 2.1776076550457857, + "learning_rate": 2.3503478312815298e-07, + "loss": 0.9337, + "step": 39599 + }, + { + "epoch": 0.93, + "grad_norm": 1.1061655098582912, + "learning_rate": 2.348703509045991e-07, + "loss": 0.8815, + "step": 39600 + }, + { + "epoch": 0.93, + "grad_norm": 1.8934988908898434, + "learning_rate": 2.3470597553672314e-07, + "loss": 0.969, + "step": 39601 + }, + { + "epoch": 0.93, + "grad_norm": 1.9356897503002455, + "learning_rate": 2.3454165702547993e-07, + "loss": 0.9782, + "step": 39602 + }, + { + "epoch": 0.93, + "grad_norm": 1.1171845863502696, + "learning_rate": 2.3437739537182646e-07, + "loss": 0.9841, + "step": 39603 + }, + { + "epoch": 0.93, + "grad_norm": 1.9698323617742945, + "learning_rate": 2.3421319057671975e-07, + "loss": 0.9642, + "step": 39604 + }, + { + "epoch": 0.93, + "grad_norm": 1.1811621611535272, + "learning_rate": 2.3404904264111683e-07, + "loss": 0.9363, + "step": 39605 + }, + { + "epoch": 0.93, + "grad_norm": 1.9578471940769067, + "learning_rate": 2.3388495156597136e-07, + "loss": 0.8548, + "step": 39606 + }, + { + "epoch": 0.93, + "grad_norm": 2.046137357374458, + "learning_rate": 2.3372091735223812e-07, + "loss": 0.7808, + "step": 39607 + }, + { + "epoch": 0.93, + "grad_norm": 1.1479417268450975, + "learning_rate": 2.3355694000087525e-07, + "loss": 0.9638, + "step": 39608 + }, + { + "epoch": 0.93, + "grad_norm": 1.9289303655988324, + "learning_rate": 2.3339301951283532e-07, + "loss": 1.0824, + "step": 39609 + }, + { + "epoch": 0.93, + "grad_norm": 2.680135057970104, + "learning_rate": 2.3322915588907425e-07, + "loss": 0.9686, + "step": 39610 + }, + { + "epoch": 0.93, + "grad_norm": 2.0861865423601396, + "learning_rate": 2.3306534913054347e-07, + "loss": 0.9876, + "step": 39611 + }, + { + "epoch": 0.93, + "grad_norm": 2.303839198021149, + "learning_rate": 2.329015992382e-07, + "loss": 0.8867, + "step": 39612 + }, + { + "epoch": 0.93, + "grad_norm": 1.0308390225390367, + "learning_rate": 2.327379062129942e-07, + "loss": 0.9034, + "step": 39613 + }, + { + "epoch": 0.93, + "grad_norm": 2.0468975774096787, + "learning_rate": 2.3257427005588197e-07, + "loss": 0.8863, + "step": 39614 + }, + { + "epoch": 0.93, + "grad_norm": 2.051992647148784, + "learning_rate": 2.3241069076781365e-07, + "loss": 1.012, + "step": 39615 + }, + { + "epoch": 0.93, + "grad_norm": 1.0485171711988597, + "learning_rate": 2.3224716834974404e-07, + "loss": 0.9851, + "step": 39616 + }, + { + "epoch": 0.93, + "grad_norm": 2.3599113162590437, + "learning_rate": 2.3208370280262127e-07, + "loss": 1.0459, + "step": 39617 + }, + { + "epoch": 0.93, + "grad_norm": 1.9673842615580273, + "learning_rate": 2.3192029412740237e-07, + "loss": 1.0202, + "step": 39618 + }, + { + "epoch": 0.93, + "grad_norm": 2.7277861764738507, + "learning_rate": 2.3175694232503542e-07, + "loss": 0.9272, + "step": 39619 + }, + { + "epoch": 0.93, + "grad_norm": 1.1185578097549154, + "learning_rate": 2.3159364739647084e-07, + "loss": 0.9322, + "step": 39620 + }, + { + "epoch": 0.93, + "grad_norm": 2.463329675275121, + "learning_rate": 2.3143040934266115e-07, + "loss": 0.9505, + "step": 39621 + }, + { + "epoch": 0.93, + "grad_norm": 2.055949702544485, + "learning_rate": 2.312672281645567e-07, + "loss": 0.9027, + "step": 39622 + }, + { + "epoch": 0.93, + "grad_norm": 1.071221179515353, + "learning_rate": 2.311041038631079e-07, + "loss": 0.9921, + "step": 39623 + }, + { + "epoch": 0.93, + "grad_norm": 2.075677708263215, + "learning_rate": 2.3094103643926168e-07, + "loss": 0.8454, + "step": 39624 + }, + { + "epoch": 0.93, + "grad_norm": 2.048102172048304, + "learning_rate": 2.3077802589397068e-07, + "loss": 0.8792, + "step": 39625 + }, + { + "epoch": 0.93, + "grad_norm": 2.284631119100928, + "learning_rate": 2.3061507222818303e-07, + "loss": 0.9692, + "step": 39626 + }, + { + "epoch": 0.93, + "grad_norm": 1.9589143545144068, + "learning_rate": 2.3045217544284791e-07, + "loss": 0.8867, + "step": 39627 + }, + { + "epoch": 0.93, + "grad_norm": 1.7912120615167724, + "learning_rate": 2.302893355389124e-07, + "loss": 0.8857, + "step": 39628 + }, + { + "epoch": 0.93, + "grad_norm": 1.9357011070305226, + "learning_rate": 2.3012655251732574e-07, + "loss": 0.9615, + "step": 39629 + }, + { + "epoch": 0.93, + "grad_norm": 2.033634773464184, + "learning_rate": 2.2996382637903492e-07, + "loss": 1.0752, + "step": 39630 + }, + { + "epoch": 0.93, + "grad_norm": 1.9940676666460955, + "learning_rate": 2.2980115712498918e-07, + "loss": 0.9309, + "step": 39631 + }, + { + "epoch": 0.93, + "grad_norm": 1.8847273160569293, + "learning_rate": 2.2963854475613223e-07, + "loss": 0.9541, + "step": 39632 + }, + { + "epoch": 0.93, + "grad_norm": 2.287872813316494, + "learning_rate": 2.2947598927341442e-07, + "loss": 0.9303, + "step": 39633 + }, + { + "epoch": 0.93, + "grad_norm": 1.9676883371471203, + "learning_rate": 2.2931349067778053e-07, + "loss": 0.9875, + "step": 39634 + }, + { + "epoch": 0.93, + "grad_norm": 2.158353106381517, + "learning_rate": 2.2915104897017648e-07, + "loss": 0.9835, + "step": 39635 + }, + { + "epoch": 0.93, + "grad_norm": 1.9711083473100874, + "learning_rate": 2.2898866415154818e-07, + "loss": 0.9651, + "step": 39636 + }, + { + "epoch": 0.93, + "grad_norm": 2.078848342405283, + "learning_rate": 2.2882633622284268e-07, + "loss": 0.9856, + "step": 39637 + }, + { + "epoch": 0.93, + "grad_norm": 1.908044019233145, + "learning_rate": 2.2866406518500252e-07, + "loss": 1.0362, + "step": 39638 + }, + { + "epoch": 0.93, + "grad_norm": 2.140810662701059, + "learning_rate": 2.285018510389736e-07, + "loss": 1.0199, + "step": 39639 + }, + { + "epoch": 0.93, + "grad_norm": 1.8168244368860127, + "learning_rate": 2.2833969378570188e-07, + "loss": 1.0291, + "step": 39640 + }, + { + "epoch": 0.93, + "grad_norm": 2.3555094789942346, + "learning_rate": 2.2817759342612878e-07, + "loss": 0.8062, + "step": 39641 + }, + { + "epoch": 0.93, + "grad_norm": 1.9896801245466231, + "learning_rate": 2.2801554996119912e-07, + "loss": 0.9902, + "step": 39642 + }, + { + "epoch": 0.93, + "grad_norm": 2.064768760415825, + "learning_rate": 2.2785356339185772e-07, + "loss": 1.0001, + "step": 39643 + }, + { + "epoch": 0.93, + "grad_norm": 2.242062221036102, + "learning_rate": 2.2769163371904713e-07, + "loss": 1.049, + "step": 39644 + }, + { + "epoch": 0.93, + "grad_norm": 2.0320620907903737, + "learning_rate": 2.275297609437077e-07, + "loss": 1.0234, + "step": 39645 + }, + { + "epoch": 0.93, + "grad_norm": 1.9026687803246047, + "learning_rate": 2.2736794506678539e-07, + "loss": 0.8704, + "step": 39646 + }, + { + "epoch": 0.93, + "grad_norm": 1.894079552587495, + "learning_rate": 2.272061860892205e-07, + "loss": 0.7927, + "step": 39647 + }, + { + "epoch": 0.93, + "grad_norm": 2.3652039851969784, + "learning_rate": 2.2704448401195567e-07, + "loss": 1.0405, + "step": 39648 + }, + { + "epoch": 0.93, + "grad_norm": 2.115834437550896, + "learning_rate": 2.268828388359312e-07, + "loss": 0.9527, + "step": 39649 + }, + { + "epoch": 0.93, + "grad_norm": 2.119984078215181, + "learning_rate": 2.267212505620886e-07, + "loss": 1.0356, + "step": 39650 + }, + { + "epoch": 0.93, + "grad_norm": 1.0745948386743143, + "learning_rate": 2.2655971919137044e-07, + "loss": 0.9283, + "step": 39651 + }, + { + "epoch": 0.93, + "grad_norm": 1.977975792873418, + "learning_rate": 2.2639824472471594e-07, + "loss": 1.0409, + "step": 39652 + }, + { + "epoch": 0.93, + "grad_norm": 1.9540878336046217, + "learning_rate": 2.262368271630633e-07, + "loss": 0.8675, + "step": 39653 + }, + { + "epoch": 0.93, + "grad_norm": 1.9401384760502616, + "learning_rate": 2.2607546650735613e-07, + "loss": 1.0968, + "step": 39654 + }, + { + "epoch": 0.93, + "grad_norm": 2.030761877772148, + "learning_rate": 2.259141627585304e-07, + "loss": 1.0116, + "step": 39655 + }, + { + "epoch": 0.93, + "grad_norm": 1.8752064724081428, + "learning_rate": 2.2575291591752756e-07, + "loss": 0.8355, + "step": 39656 + }, + { + "epoch": 0.93, + "grad_norm": 1.0254671322572808, + "learning_rate": 2.2559172598528578e-07, + "loss": 0.9872, + "step": 39657 + }, + { + "epoch": 0.93, + "grad_norm": 2.0318190988424845, + "learning_rate": 2.2543059296274316e-07, + "loss": 0.9242, + "step": 39658 + }, + { + "epoch": 0.93, + "grad_norm": 1.7641739198045905, + "learning_rate": 2.2526951685083787e-07, + "loss": 0.9407, + "step": 39659 + }, + { + "epoch": 0.93, + "grad_norm": 1.8549857204154057, + "learning_rate": 2.2510849765050914e-07, + "loss": 0.9737, + "step": 39660 + }, + { + "epoch": 0.93, + "grad_norm": 2.0281804262217356, + "learning_rate": 2.2494753536269288e-07, + "loss": 0.913, + "step": 39661 + }, + { + "epoch": 0.93, + "grad_norm": 2.84731366367495, + "learning_rate": 2.2478662998832725e-07, + "loss": 1.1281, + "step": 39662 + }, + { + "epoch": 0.93, + "grad_norm": 1.970437847100176, + "learning_rate": 2.2462578152834812e-07, + "loss": 1.0875, + "step": 39663 + }, + { + "epoch": 0.93, + "grad_norm": 1.9947256098666262, + "learning_rate": 2.2446498998369259e-07, + "loss": 0.9848, + "step": 39664 + }, + { + "epoch": 0.93, + "grad_norm": 2.182901526177938, + "learning_rate": 2.2430425535529765e-07, + "loss": 0.9688, + "step": 39665 + }, + { + "epoch": 0.93, + "grad_norm": 1.8995114911326232, + "learning_rate": 2.24143577644097e-07, + "loss": 0.8086, + "step": 39666 + }, + { + "epoch": 0.93, + "grad_norm": 2.189938231904994, + "learning_rate": 2.2398295685102655e-07, + "loss": 1.0765, + "step": 39667 + }, + { + "epoch": 0.93, + "grad_norm": 1.9800549338504323, + "learning_rate": 2.2382239297702447e-07, + "loss": 1.1052, + "step": 39668 + }, + { + "epoch": 0.93, + "grad_norm": 1.81333687066575, + "learning_rate": 2.236618860230233e-07, + "loss": 1.0132, + "step": 39669 + }, + { + "epoch": 0.93, + "grad_norm": 2.314084294563119, + "learning_rate": 2.2350143598995676e-07, + "loss": 0.9496, + "step": 39670 + }, + { + "epoch": 0.93, + "grad_norm": 2.2952857179552923, + "learning_rate": 2.2334104287875969e-07, + "loss": 0.9809, + "step": 39671 + }, + { + "epoch": 0.93, + "grad_norm": 1.8119151558815685, + "learning_rate": 2.2318070669036686e-07, + "loss": 0.8663, + "step": 39672 + }, + { + "epoch": 0.93, + "grad_norm": 1.9973047432490345, + "learning_rate": 2.2302042742571194e-07, + "loss": 0.8648, + "step": 39673 + }, + { + "epoch": 0.93, + "grad_norm": 2.1158771895346495, + "learning_rate": 2.2286020508572648e-07, + "loss": 0.9704, + "step": 39674 + }, + { + "epoch": 0.93, + "grad_norm": 1.9397304562081084, + "learning_rate": 2.2270003967134414e-07, + "loss": 1.0174, + "step": 39675 + }, + { + "epoch": 0.93, + "grad_norm": 2.1471723989458424, + "learning_rate": 2.225399311834986e-07, + "loss": 1.1223, + "step": 39676 + }, + { + "epoch": 0.93, + "grad_norm": 1.94305712767222, + "learning_rate": 2.2237987962312134e-07, + "loss": 1.0835, + "step": 39677 + }, + { + "epoch": 0.93, + "grad_norm": 1.9840924948232266, + "learning_rate": 2.2221988499114278e-07, + "loss": 0.9614, + "step": 39678 + }, + { + "epoch": 0.93, + "grad_norm": 2.2151756119993524, + "learning_rate": 2.2205994728849655e-07, + "loss": 0.8806, + "step": 39679 + }, + { + "epoch": 0.93, + "grad_norm": 1.7857796477325558, + "learning_rate": 2.2190006651611197e-07, + "loss": 0.9092, + "step": 39680 + }, + { + "epoch": 0.93, + "grad_norm": 2.030884394158646, + "learning_rate": 2.2174024267492155e-07, + "loss": 0.9919, + "step": 39681 + }, + { + "epoch": 0.93, + "grad_norm": 2.1689782244805813, + "learning_rate": 2.2158047576585572e-07, + "loss": 0.9082, + "step": 39682 + }, + { + "epoch": 0.93, + "grad_norm": 1.869239509479579, + "learning_rate": 2.2142076578984372e-07, + "loss": 1.0085, + "step": 39683 + }, + { + "epoch": 0.93, + "grad_norm": 2.3264433482067, + "learning_rate": 2.212611127478148e-07, + "loss": 0.8513, + "step": 39684 + }, + { + "epoch": 0.93, + "grad_norm": 2.0593748583236438, + "learning_rate": 2.2110151664070044e-07, + "loss": 0.9476, + "step": 39685 + }, + { + "epoch": 0.93, + "grad_norm": 1.9039728357255274, + "learning_rate": 2.209419774694299e-07, + "loss": 0.9434, + "step": 39686 + }, + { + "epoch": 0.93, + "grad_norm": 2.3374482271469517, + "learning_rate": 2.2078249523493022e-07, + "loss": 0.9053, + "step": 39687 + }, + { + "epoch": 0.94, + "grad_norm": 2.135138361107764, + "learning_rate": 2.2062306993813064e-07, + "loss": 0.9341, + "step": 39688 + }, + { + "epoch": 0.94, + "grad_norm": 1.0715975977821137, + "learning_rate": 2.2046370157995934e-07, + "loss": 0.9024, + "step": 39689 + }, + { + "epoch": 0.94, + "grad_norm": 1.9044239305022992, + "learning_rate": 2.2030439016134664e-07, + "loss": 0.8804, + "step": 39690 + }, + { + "epoch": 0.94, + "grad_norm": 1.9293043152141824, + "learning_rate": 2.2014513568321628e-07, + "loss": 0.937, + "step": 39691 + }, + { + "epoch": 0.94, + "grad_norm": 1.8653129253399547, + "learning_rate": 2.1998593814649637e-07, + "loss": 0.9076, + "step": 39692 + }, + { + "epoch": 0.94, + "grad_norm": 1.819849996304881, + "learning_rate": 2.1982679755211622e-07, + "loss": 0.862, + "step": 39693 + }, + { + "epoch": 0.94, + "grad_norm": 2.1662289320497834, + "learning_rate": 2.196677139009995e-07, + "loss": 0.9654, + "step": 39694 + }, + { + "epoch": 0.94, + "grad_norm": 1.0573272481568232, + "learning_rate": 2.1950868719407548e-07, + "loss": 0.9061, + "step": 39695 + }, + { + "epoch": 0.94, + "grad_norm": 1.9363986831542552, + "learning_rate": 2.1934971743226674e-07, + "loss": 1.0445, + "step": 39696 + }, + { + "epoch": 0.94, + "grad_norm": 2.009014033356951, + "learning_rate": 2.1919080461650144e-07, + "loss": 0.9859, + "step": 39697 + }, + { + "epoch": 0.94, + "grad_norm": 2.0408244904772808, + "learning_rate": 2.190319487477033e-07, + "loss": 0.9049, + "step": 39698 + }, + { + "epoch": 0.94, + "grad_norm": 2.082665979618221, + "learning_rate": 2.188731498267982e-07, + "loss": 0.9607, + "step": 39699 + }, + { + "epoch": 0.94, + "grad_norm": 1.965065439399071, + "learning_rate": 2.1871440785470987e-07, + "loss": 0.9453, + "step": 39700 + }, + { + "epoch": 0.94, + "grad_norm": 2.729000275408626, + "learning_rate": 2.1855572283236314e-07, + "loss": 0.9075, + "step": 39701 + }, + { + "epoch": 0.94, + "grad_norm": 2.1310043817747566, + "learning_rate": 2.183970947606806e-07, + "loss": 0.9904, + "step": 39702 + }, + { + "epoch": 0.94, + "grad_norm": 1.8512011699316602, + "learning_rate": 2.1823852364058928e-07, + "loss": 0.8013, + "step": 39703 + }, + { + "epoch": 0.94, + "grad_norm": 2.012524433160744, + "learning_rate": 2.1808000947300843e-07, + "loss": 1.02, + "step": 39704 + }, + { + "epoch": 0.94, + "grad_norm": 1.1393063730787163, + "learning_rate": 2.179215522588629e-07, + "loss": 0.888, + "step": 39705 + }, + { + "epoch": 0.94, + "grad_norm": 1.9740428261546854, + "learning_rate": 2.1776315199907526e-07, + "loss": 0.9139, + "step": 39706 + }, + { + "epoch": 0.94, + "grad_norm": 1.9857976015314671, + "learning_rate": 2.1760480869456812e-07, + "loss": 0.9279, + "step": 39707 + }, + { + "epoch": 0.94, + "grad_norm": 2.209869330610862, + "learning_rate": 2.1744652234626295e-07, + "loss": 1.0325, + "step": 39708 + }, + { + "epoch": 0.94, + "grad_norm": 1.1168650117088235, + "learning_rate": 2.1728829295508013e-07, + "loss": 0.952, + "step": 39709 + }, + { + "epoch": 0.94, + "grad_norm": 2.0872487788013685, + "learning_rate": 2.1713012052194114e-07, + "loss": 0.9172, + "step": 39710 + }, + { + "epoch": 0.94, + "grad_norm": 2.1668404583633576, + "learning_rate": 2.1697200504776973e-07, + "loss": 0.9515, + "step": 39711 + }, + { + "epoch": 0.94, + "grad_norm": 1.0500983207930772, + "learning_rate": 2.1681394653348508e-07, + "loss": 0.9657, + "step": 39712 + }, + { + "epoch": 0.94, + "grad_norm": 2.208648559145265, + "learning_rate": 2.166559449800043e-07, + "loss": 1.1043, + "step": 39713 + }, + { + "epoch": 0.94, + "grad_norm": 1.8722798894800028, + "learning_rate": 2.164980003882522e-07, + "loss": 1.0544, + "step": 39714 + }, + { + "epoch": 0.94, + "grad_norm": 1.0363369492537249, + "learning_rate": 2.1634011275914468e-07, + "loss": 0.8967, + "step": 39715 + }, + { + "epoch": 0.94, + "grad_norm": 2.7293762184527672, + "learning_rate": 2.1618228209360438e-07, + "loss": 1.0281, + "step": 39716 + }, + { + "epoch": 0.94, + "grad_norm": 2.0814693941495293, + "learning_rate": 2.1602450839254608e-07, + "loss": 0.9673, + "step": 39717 + }, + { + "epoch": 0.94, + "grad_norm": 2.2546744760459667, + "learning_rate": 2.158667916568924e-07, + "loss": 1.0438, + "step": 39718 + }, + { + "epoch": 0.94, + "grad_norm": 2.0289208698420738, + "learning_rate": 2.1570913188755925e-07, + "loss": 0.9729, + "step": 39719 + }, + { + "epoch": 0.94, + "grad_norm": 1.8088927058588475, + "learning_rate": 2.1555152908546484e-07, + "loss": 0.9528, + "step": 39720 + }, + { + "epoch": 0.94, + "grad_norm": 2.0030407992166714, + "learning_rate": 2.1539398325152839e-07, + "loss": 0.8911, + "step": 39721 + }, + { + "epoch": 0.94, + "grad_norm": 2.0791580687271574, + "learning_rate": 2.1523649438666472e-07, + "loss": 0.961, + "step": 39722 + }, + { + "epoch": 0.94, + "grad_norm": 2.629326776288587, + "learning_rate": 2.15079062491792e-07, + "loss": 0.9569, + "step": 39723 + }, + { + "epoch": 0.94, + "grad_norm": 1.859216860399457, + "learning_rate": 2.1492168756782728e-07, + "loss": 1.0543, + "step": 39724 + }, + { + "epoch": 0.94, + "grad_norm": 1.1167746922033477, + "learning_rate": 2.1476436961568648e-07, + "loss": 0.9923, + "step": 39725 + }, + { + "epoch": 0.94, + "grad_norm": 1.9438314453959207, + "learning_rate": 2.1460710863628443e-07, + "loss": 1.0365, + "step": 39726 + }, + { + "epoch": 0.94, + "grad_norm": 1.773294033300378, + "learning_rate": 2.1444990463053817e-07, + "loss": 0.9658, + "step": 39727 + }, + { + "epoch": 0.94, + "grad_norm": 1.912271240543534, + "learning_rate": 2.1429275759936363e-07, + "loss": 0.8556, + "step": 39728 + }, + { + "epoch": 0.94, + "grad_norm": 2.0861695963365117, + "learning_rate": 2.1413566754367454e-07, + "loss": 1.1085, + "step": 39729 + }, + { + "epoch": 0.94, + "grad_norm": 2.2120028375869447, + "learning_rate": 2.139786344643846e-07, + "loss": 0.9358, + "step": 39730 + }, + { + "epoch": 0.94, + "grad_norm": 1.8634496963582603, + "learning_rate": 2.1382165836240977e-07, + "loss": 0.9931, + "step": 39731 + }, + { + "epoch": 0.94, + "grad_norm": 2.086018711477707, + "learning_rate": 2.1366473923866372e-07, + "loss": 0.9083, + "step": 39732 + }, + { + "epoch": 0.94, + "grad_norm": 2.120323447260799, + "learning_rate": 2.135078770940602e-07, + "loss": 0.9402, + "step": 39733 + }, + { + "epoch": 0.94, + "grad_norm": 1.891432390245092, + "learning_rate": 2.1335107192951176e-07, + "loss": 1.0407, + "step": 39734 + }, + { + "epoch": 0.94, + "grad_norm": 2.2227528436702726, + "learning_rate": 2.1319432374593218e-07, + "loss": 1.0445, + "step": 39735 + }, + { + "epoch": 0.94, + "grad_norm": 3.9882932728984084, + "learning_rate": 2.1303763254423293e-07, + "loss": 1.0235, + "step": 39736 + }, + { + "epoch": 0.94, + "grad_norm": 1.0608947378355524, + "learning_rate": 2.128809983253277e-07, + "loss": 0.8829, + "step": 39737 + }, + { + "epoch": 0.94, + "grad_norm": 1.8868094103120627, + "learning_rate": 2.127244210901269e-07, + "loss": 1.0804, + "step": 39738 + }, + { + "epoch": 0.94, + "grad_norm": 2.0309752745033007, + "learning_rate": 2.1256790083954537e-07, + "loss": 1.0531, + "step": 39739 + }, + { + "epoch": 0.94, + "grad_norm": 1.8722676379371852, + "learning_rate": 2.1241143757449012e-07, + "loss": 0.9852, + "step": 39740 + }, + { + "epoch": 0.94, + "grad_norm": 1.9259017017795579, + "learning_rate": 2.1225503129587488e-07, + "loss": 0.9331, + "step": 39741 + }, + { + "epoch": 0.94, + "grad_norm": 2.129768237156902, + "learning_rate": 2.1209868200461004e-07, + "loss": 1.0108, + "step": 39742 + }, + { + "epoch": 0.94, + "grad_norm": 1.980244176474241, + "learning_rate": 2.1194238970160376e-07, + "loss": 0.8964, + "step": 39743 + }, + { + "epoch": 0.94, + "grad_norm": 1.9500744648005754, + "learning_rate": 2.117861543877686e-07, + "loss": 0.9045, + "step": 39744 + }, + { + "epoch": 0.94, + "grad_norm": 1.9050149632293192, + "learning_rate": 2.1162997606401392e-07, + "loss": 1.0528, + "step": 39745 + }, + { + "epoch": 0.94, + "grad_norm": 1.7832270120052844, + "learning_rate": 2.1147385473124892e-07, + "loss": 0.9663, + "step": 39746 + }, + { + "epoch": 0.94, + "grad_norm": 2.4361436038480537, + "learning_rate": 2.113177903903807e-07, + "loss": 1.0534, + "step": 39747 + }, + { + "epoch": 0.94, + "grad_norm": 2.145296572681164, + "learning_rate": 2.111617830423196e-07, + "loss": 0.9174, + "step": 39748 + }, + { + "epoch": 0.94, + "grad_norm": 1.967050967452314, + "learning_rate": 2.1100583268797493e-07, + "loss": 0.841, + "step": 39749 + }, + { + "epoch": 0.94, + "grad_norm": 2.181363356242123, + "learning_rate": 2.1084993932825259e-07, + "loss": 0.921, + "step": 39750 + }, + { + "epoch": 0.94, + "grad_norm": 1.0676931982862439, + "learning_rate": 2.1069410296406079e-07, + "loss": 0.9032, + "step": 39751 + }, + { + "epoch": 0.94, + "grad_norm": 2.2541728393538354, + "learning_rate": 2.1053832359630655e-07, + "loss": 0.8419, + "step": 39752 + }, + { + "epoch": 0.94, + "grad_norm": 1.7425209308220018, + "learning_rate": 2.1038260122589915e-07, + "loss": 1.0627, + "step": 39753 + }, + { + "epoch": 0.94, + "grad_norm": 1.9339395784569156, + "learning_rate": 2.1022693585374343e-07, + "loss": 0.9882, + "step": 39754 + }, + { + "epoch": 0.94, + "grad_norm": 2.060867585064753, + "learning_rate": 2.1007132748074533e-07, + "loss": 0.9752, + "step": 39755 + }, + { + "epoch": 0.94, + "grad_norm": 2.1752900672662054, + "learning_rate": 2.0991577610781188e-07, + "loss": 1.0489, + "step": 39756 + }, + { + "epoch": 0.94, + "grad_norm": 2.030967507525879, + "learning_rate": 2.0976028173584683e-07, + "loss": 0.8714, + "step": 39757 + }, + { + "epoch": 0.94, + "grad_norm": 1.738137624397845, + "learning_rate": 2.0960484436575835e-07, + "loss": 0.9883, + "step": 39758 + }, + { + "epoch": 0.94, + "grad_norm": 1.8593762332272046, + "learning_rate": 2.09449463998449e-07, + "loss": 1.034, + "step": 39759 + }, + { + "epoch": 0.94, + "grad_norm": 2.0841588895342893, + "learning_rate": 2.092941406348259e-07, + "loss": 0.9791, + "step": 39760 + }, + { + "epoch": 0.94, + "grad_norm": 1.9887453490317668, + "learning_rate": 2.0913887427579048e-07, + "loss": 0.9587, + "step": 39761 + }, + { + "epoch": 0.94, + "grad_norm": 2.0325342666148836, + "learning_rate": 2.0898366492224986e-07, + "loss": 0.9158, + "step": 39762 + }, + { + "epoch": 0.94, + "grad_norm": 1.8516833312710526, + "learning_rate": 2.0882851257510551e-07, + "loss": 0.8798, + "step": 39763 + }, + { + "epoch": 0.94, + "grad_norm": 1.841777501681955, + "learning_rate": 2.0867341723526002e-07, + "loss": 0.9195, + "step": 39764 + }, + { + "epoch": 0.94, + "grad_norm": 2.6488266142727768, + "learning_rate": 2.0851837890361936e-07, + "loss": 0.9406, + "step": 39765 + }, + { + "epoch": 0.94, + "grad_norm": 1.9852023562352556, + "learning_rate": 2.0836339758108393e-07, + "loss": 0.9076, + "step": 39766 + }, + { + "epoch": 0.94, + "grad_norm": 2.2644370250958703, + "learning_rate": 2.082084732685574e-07, + "loss": 0.9249, + "step": 39767 + }, + { + "epoch": 0.94, + "grad_norm": 1.1339052891937456, + "learning_rate": 2.0805360596694024e-07, + "loss": 0.9771, + "step": 39768 + }, + { + "epoch": 0.94, + "grad_norm": 2.2334637637565344, + "learning_rate": 2.078987956771361e-07, + "loss": 0.7798, + "step": 39769 + }, + { + "epoch": 0.94, + "grad_norm": 2.0206245978148076, + "learning_rate": 2.0774404240004432e-07, + "loss": 1.0095, + "step": 39770 + }, + { + "epoch": 0.94, + "grad_norm": 1.824833792161715, + "learning_rate": 2.0758934613656856e-07, + "loss": 0.9167, + "step": 39771 + }, + { + "epoch": 0.94, + "grad_norm": 1.8381599359819458, + "learning_rate": 2.0743470688760593e-07, + "loss": 0.8764, + "step": 39772 + }, + { + "epoch": 0.94, + "grad_norm": 1.9743622755884875, + "learning_rate": 2.07280124654059e-07, + "loss": 1.0025, + "step": 39773 + }, + { + "epoch": 0.94, + "grad_norm": 2.1922761303873353, + "learning_rate": 2.071255994368282e-07, + "loss": 1.1194, + "step": 39774 + }, + { + "epoch": 0.94, + "grad_norm": 3.1931543473969684, + "learning_rate": 2.069711312368128e-07, + "loss": 1.1003, + "step": 39775 + }, + { + "epoch": 0.94, + "grad_norm": 1.9488698044038113, + "learning_rate": 2.0681672005491093e-07, + "loss": 0.9609, + "step": 39776 + }, + { + "epoch": 0.94, + "grad_norm": 1.9370413777460946, + "learning_rate": 2.0666236589202305e-07, + "loss": 0.9961, + "step": 39777 + }, + { + "epoch": 0.94, + "grad_norm": 2.2502945211117367, + "learning_rate": 2.0650806874904727e-07, + "loss": 0.8839, + "step": 39778 + }, + { + "epoch": 0.94, + "grad_norm": 1.9474522376628929, + "learning_rate": 2.063538286268818e-07, + "loss": 0.9263, + "step": 39779 + }, + { + "epoch": 0.94, + "grad_norm": 2.0039571493964456, + "learning_rate": 2.061996455264259e-07, + "loss": 0.9973, + "step": 39780 + }, + { + "epoch": 0.94, + "grad_norm": 2.003351028724696, + "learning_rate": 2.0604551944857665e-07, + "loss": 1.0101, + "step": 39781 + }, + { + "epoch": 0.94, + "grad_norm": 1.0706907634167646, + "learning_rate": 2.0589145039422887e-07, + "loss": 0.9475, + "step": 39782 + }, + { + "epoch": 0.94, + "grad_norm": 1.0980058131246744, + "learning_rate": 2.0573743836428294e-07, + "loss": 0.9914, + "step": 39783 + }, + { + "epoch": 0.94, + "grad_norm": 2.1677939980330345, + "learning_rate": 2.0558348335963595e-07, + "loss": 0.919, + "step": 39784 + }, + { + "epoch": 0.94, + "grad_norm": 1.9432383663156139, + "learning_rate": 2.0542958538118162e-07, + "loss": 0.9854, + "step": 39785 + }, + { + "epoch": 0.94, + "grad_norm": 1.1338003173001725, + "learning_rate": 2.05275744429817e-07, + "loss": 1.0085, + "step": 39786 + }, + { + "epoch": 0.94, + "grad_norm": 2.1405468265371312, + "learning_rate": 2.0512196050643697e-07, + "loss": 1.1061, + "step": 39787 + }, + { + "epoch": 0.94, + "grad_norm": 2.1878805749080383, + "learning_rate": 2.0496823361194074e-07, + "loss": 1.0475, + "step": 39788 + }, + { + "epoch": 0.94, + "grad_norm": 3.0752556054615137, + "learning_rate": 2.0481456374721763e-07, + "loss": 0.9454, + "step": 39789 + }, + { + "epoch": 0.94, + "grad_norm": 2.5023910622533627, + "learning_rate": 2.0466095091316585e-07, + "loss": 1.0424, + "step": 39790 + }, + { + "epoch": 0.94, + "grad_norm": 2.6776124991179255, + "learning_rate": 2.0450739511067907e-07, + "loss": 0.8703, + "step": 39791 + }, + { + "epoch": 0.94, + "grad_norm": 2.3895959513192393, + "learning_rate": 2.0435389634065328e-07, + "loss": 0.9218, + "step": 39792 + }, + { + "epoch": 0.94, + "grad_norm": 1.9355263771488387, + "learning_rate": 2.0420045460397885e-07, + "loss": 1.0123, + "step": 39793 + }, + { + "epoch": 0.94, + "grad_norm": 2.4770396287999024, + "learning_rate": 2.0404706990155065e-07, + "loss": 0.8682, + "step": 39794 + }, + { + "epoch": 0.94, + "grad_norm": 1.9829993741034293, + "learning_rate": 2.038937422342624e-07, + "loss": 0.9407, + "step": 39795 + }, + { + "epoch": 0.94, + "grad_norm": 2.015362956211878, + "learning_rate": 2.0374047160300557e-07, + "loss": 0.9449, + "step": 39796 + }, + { + "epoch": 0.94, + "grad_norm": 2.112634838252977, + "learning_rate": 2.0358725800867395e-07, + "loss": 0.851, + "step": 39797 + }, + { + "epoch": 0.94, + "grad_norm": 2.313952384572276, + "learning_rate": 2.0343410145215793e-07, + "loss": 0.9416, + "step": 39798 + }, + { + "epoch": 0.94, + "grad_norm": 1.8740186316635785, + "learning_rate": 2.0328100193435006e-07, + "loss": 0.8656, + "step": 39799 + }, + { + "epoch": 0.94, + "grad_norm": 1.8948432315231734, + "learning_rate": 2.0312795945614193e-07, + "loss": 0.9514, + "step": 39800 + }, + { + "epoch": 0.94, + "grad_norm": 1.9151529222845256, + "learning_rate": 2.0297497401842502e-07, + "loss": 0.9423, + "step": 39801 + }, + { + "epoch": 0.94, + "grad_norm": 1.9721739594873127, + "learning_rate": 2.0282204562208974e-07, + "loss": 0.8985, + "step": 39802 + }, + { + "epoch": 0.94, + "grad_norm": 7.651456367548694, + "learning_rate": 2.0266917426802536e-07, + "loss": 0.8679, + "step": 39803 + }, + { + "epoch": 0.94, + "grad_norm": 1.853177356302004, + "learning_rate": 2.0251635995712226e-07, + "loss": 0.9636, + "step": 39804 + }, + { + "epoch": 0.94, + "grad_norm": 2.0376651674116015, + "learning_rate": 2.02363602690272e-07, + "loss": 0.9605, + "step": 39805 + }, + { + "epoch": 0.94, + "grad_norm": 2.133454088546714, + "learning_rate": 2.0221090246836274e-07, + "loss": 0.8522, + "step": 39806 + }, + { + "epoch": 0.94, + "grad_norm": 2.250531037069942, + "learning_rate": 2.0205825929228262e-07, + "loss": 0.8824, + "step": 39807 + }, + { + "epoch": 0.94, + "grad_norm": 1.862526417268769, + "learning_rate": 2.019056731629221e-07, + "loss": 1.0481, + "step": 39808 + }, + { + "epoch": 0.94, + "grad_norm": 1.8200324218088, + "learning_rate": 2.017531440811682e-07, + "loss": 0.9642, + "step": 39809 + }, + { + "epoch": 0.94, + "grad_norm": 2.0366480607648407, + "learning_rate": 2.0160067204791023e-07, + "loss": 0.9642, + "step": 39810 + }, + { + "epoch": 0.94, + "grad_norm": 3.084126973569832, + "learning_rate": 2.0144825706403526e-07, + "loss": 0.8502, + "step": 39811 + }, + { + "epoch": 0.94, + "grad_norm": 1.9222947536721586, + "learning_rate": 2.0129589913042924e-07, + "loss": 0.9518, + "step": 39812 + }, + { + "epoch": 0.94, + "grad_norm": 2.294215861541264, + "learning_rate": 2.0114359824798257e-07, + "loss": 0.9335, + "step": 39813 + }, + { + "epoch": 0.94, + "grad_norm": 2.1107017084509407, + "learning_rate": 2.00991354417579e-07, + "loss": 0.9676, + "step": 39814 + }, + { + "epoch": 0.94, + "grad_norm": 1.8355962277586884, + "learning_rate": 2.008391676401067e-07, + "loss": 0.9612, + "step": 39815 + }, + { + "epoch": 0.94, + "grad_norm": 1.8816693974132979, + "learning_rate": 2.0068703791645162e-07, + "loss": 0.9184, + "step": 39816 + }, + { + "epoch": 0.94, + "grad_norm": 2.9907824216826606, + "learning_rate": 2.005349652474975e-07, + "loss": 0.9859, + "step": 39817 + }, + { + "epoch": 0.94, + "grad_norm": 2.175537037777185, + "learning_rate": 2.0038294963413251e-07, + "loss": 0.895, + "step": 39818 + }, + { + "epoch": 0.94, + "grad_norm": 2.033864118379325, + "learning_rate": 2.002309910772393e-07, + "loss": 0.9769, + "step": 39819 + }, + { + "epoch": 0.94, + "grad_norm": 2.188809123808941, + "learning_rate": 2.0007908957770493e-07, + "loss": 0.9351, + "step": 39820 + }, + { + "epoch": 0.94, + "grad_norm": 1.0917197262704128, + "learning_rate": 1.9992724513641204e-07, + "loss": 0.886, + "step": 39821 + }, + { + "epoch": 0.94, + "grad_norm": 2.0752864276293206, + "learning_rate": 1.9977545775424545e-07, + "loss": 0.9795, + "step": 39822 + }, + { + "epoch": 0.94, + "grad_norm": 1.0654708193742324, + "learning_rate": 1.9962372743209e-07, + "loss": 0.9709, + "step": 39823 + }, + { + "epoch": 0.94, + "grad_norm": 1.2627772207312908, + "learning_rate": 1.9947205417082615e-07, + "loss": 0.9057, + "step": 39824 + }, + { + "epoch": 0.94, + "grad_norm": 1.8580508894384367, + "learning_rate": 1.993204379713398e-07, + "loss": 0.9587, + "step": 39825 + }, + { + "epoch": 0.94, + "grad_norm": 4.65706609483297, + "learning_rate": 1.991688788345125e-07, + "loss": 1.0007, + "step": 39826 + }, + { + "epoch": 0.94, + "grad_norm": 1.9437659637400877, + "learning_rate": 1.99017376761228e-07, + "loss": 0.9673, + "step": 39827 + }, + { + "epoch": 0.94, + "grad_norm": 1.0732164598961675, + "learning_rate": 1.9886593175236557e-07, + "loss": 0.9011, + "step": 39828 + }, + { + "epoch": 0.94, + "grad_norm": 1.963415136171401, + "learning_rate": 1.9871454380880894e-07, + "loss": 1.095, + "step": 39829 + }, + { + "epoch": 0.94, + "grad_norm": 2.4904609518195615, + "learning_rate": 1.9856321293143965e-07, + "loss": 0.9457, + "step": 39830 + }, + { + "epoch": 0.94, + "grad_norm": 1.96733480148467, + "learning_rate": 1.9841193912113922e-07, + "loss": 0.9341, + "step": 39831 + }, + { + "epoch": 0.94, + "grad_norm": 1.9758729702017543, + "learning_rate": 1.982607223787858e-07, + "loss": 0.9979, + "step": 39832 + }, + { + "epoch": 0.94, + "grad_norm": 3.8540326609397284, + "learning_rate": 1.9810956270526204e-07, + "loss": 1.0497, + "step": 39833 + }, + { + "epoch": 0.94, + "grad_norm": 1.1056143869889472, + "learning_rate": 1.9795846010144837e-07, + "loss": 0.9809, + "step": 39834 + }, + { + "epoch": 0.94, + "grad_norm": 1.9478311418189467, + "learning_rate": 1.9780741456822404e-07, + "loss": 0.991, + "step": 39835 + }, + { + "epoch": 0.94, + "grad_norm": 2.955096500901209, + "learning_rate": 1.976564261064684e-07, + "loss": 1.0293, + "step": 39836 + }, + { + "epoch": 0.94, + "grad_norm": 1.9720037953478862, + "learning_rate": 1.975054947170596e-07, + "loss": 0.988, + "step": 39837 + }, + { + "epoch": 0.94, + "grad_norm": 1.9981822683936667, + "learning_rate": 1.9735462040087693e-07, + "loss": 1.099, + "step": 39838 + }, + { + "epoch": 0.94, + "grad_norm": 1.0906721779089745, + "learning_rate": 1.9720380315880083e-07, + "loss": 0.9286, + "step": 39839 + }, + { + "epoch": 0.94, + "grad_norm": 1.8394258876553207, + "learning_rate": 1.9705304299170613e-07, + "loss": 0.8634, + "step": 39840 + }, + { + "epoch": 0.94, + "grad_norm": 1.9412473638997017, + "learning_rate": 1.9690233990047325e-07, + "loss": 1.0443, + "step": 39841 + }, + { + "epoch": 0.94, + "grad_norm": 2.044199903537303, + "learning_rate": 1.9675169388597705e-07, + "loss": 0.9211, + "step": 39842 + }, + { + "epoch": 0.94, + "grad_norm": 2.7796376237929774, + "learning_rate": 1.966011049490979e-07, + "loss": 0.9674, + "step": 39843 + }, + { + "epoch": 0.94, + "grad_norm": 2.253809650907174, + "learning_rate": 1.9645057309070958e-07, + "loss": 1.0333, + "step": 39844 + }, + { + "epoch": 0.94, + "grad_norm": 2.115307124437551, + "learning_rate": 1.9630009831169027e-07, + "loss": 1.0534, + "step": 39845 + }, + { + "epoch": 0.94, + "grad_norm": 2.6446399469167337, + "learning_rate": 1.9614968061291484e-07, + "loss": 1.0165, + "step": 39846 + }, + { + "epoch": 0.94, + "grad_norm": 2.0425534932371567, + "learning_rate": 1.9599931999526035e-07, + "loss": 0.9283, + "step": 39847 + }, + { + "epoch": 0.94, + "grad_norm": 2.0377424603106276, + "learning_rate": 1.9584901645960275e-07, + "loss": 0.9504, + "step": 39848 + }, + { + "epoch": 0.94, + "grad_norm": 2.086139467314213, + "learning_rate": 1.956987700068147e-07, + "loss": 0.9485, + "step": 39849 + }, + { + "epoch": 0.94, + "grad_norm": 2.1775867098836246, + "learning_rate": 1.9554858063777215e-07, + "loss": 1.0413, + "step": 39850 + }, + { + "epoch": 0.94, + "grad_norm": 2.0167283102618216, + "learning_rate": 1.953984483533511e-07, + "loss": 0.9717, + "step": 39851 + }, + { + "epoch": 0.94, + "grad_norm": 2.0639699153418194, + "learning_rate": 1.9524837315442414e-07, + "loss": 1.1154, + "step": 39852 + }, + { + "epoch": 0.94, + "grad_norm": 1.8791828074043686, + "learning_rate": 1.9509835504186504e-07, + "loss": 0.8467, + "step": 39853 + }, + { + "epoch": 0.94, + "grad_norm": 1.8058739258340653, + "learning_rate": 1.9494839401654642e-07, + "loss": 0.9983, + "step": 39854 + }, + { + "epoch": 0.94, + "grad_norm": 2.3292528069370766, + "learning_rate": 1.9479849007934425e-07, + "loss": 1.0096, + "step": 39855 + }, + { + "epoch": 0.94, + "grad_norm": 1.8396603161667981, + "learning_rate": 1.9464864323112897e-07, + "loss": 0.9, + "step": 39856 + }, + { + "epoch": 0.94, + "grad_norm": 1.9528420127660868, + "learning_rate": 1.944988534727732e-07, + "loss": 0.9987, + "step": 39857 + }, + { + "epoch": 0.94, + "grad_norm": 1.9425808964138447, + "learning_rate": 1.9434912080514955e-07, + "loss": 0.9821, + "step": 39858 + }, + { + "epoch": 0.94, + "grad_norm": 1.9589665890834227, + "learning_rate": 1.941994452291296e-07, + "loss": 1.0008, + "step": 39859 + }, + { + "epoch": 0.94, + "grad_norm": 2.531458478895874, + "learning_rate": 1.9404982674558593e-07, + "loss": 0.9971, + "step": 39860 + }, + { + "epoch": 0.94, + "grad_norm": 2.142928900102191, + "learning_rate": 1.9390026535538675e-07, + "loss": 0.9191, + "step": 39861 + }, + { + "epoch": 0.94, + "grad_norm": 1.899592909446424, + "learning_rate": 1.9375076105940693e-07, + "loss": 0.9543, + "step": 39862 + }, + { + "epoch": 0.94, + "grad_norm": 3.1528002716823242, + "learning_rate": 1.9360131385851356e-07, + "loss": 0.9802, + "step": 39863 + }, + { + "epoch": 0.94, + "grad_norm": 1.730571351308771, + "learning_rate": 1.9345192375357812e-07, + "loss": 0.9819, + "step": 39864 + }, + { + "epoch": 0.94, + "grad_norm": 2.0576585428633543, + "learning_rate": 1.933025907454711e-07, + "loss": 0.9851, + "step": 39865 + }, + { + "epoch": 0.94, + "grad_norm": 1.0958439614056459, + "learning_rate": 1.9315331483506062e-07, + "loss": 0.8912, + "step": 39866 + }, + { + "epoch": 0.94, + "grad_norm": 2.4007445681446615, + "learning_rate": 1.930040960232149e-07, + "loss": 1.0541, + "step": 39867 + }, + { + "epoch": 0.94, + "grad_norm": 2.0688050796106237, + "learning_rate": 1.928549343108055e-07, + "loss": 0.9024, + "step": 39868 + }, + { + "epoch": 0.94, + "grad_norm": 2.0008592506041647, + "learning_rate": 1.9270582969870056e-07, + "loss": 0.7479, + "step": 39869 + }, + { + "epoch": 0.94, + "grad_norm": 1.9676839821173318, + "learning_rate": 1.9255678218776608e-07, + "loss": 1.0854, + "step": 39870 + }, + { + "epoch": 0.94, + "grad_norm": 1.9737537470025182, + "learning_rate": 1.9240779177887027e-07, + "loss": 0.9511, + "step": 39871 + }, + { + "epoch": 0.94, + "grad_norm": 2.1462504290977695, + "learning_rate": 1.922588584728824e-07, + "loss": 1.0175, + "step": 39872 + }, + { + "epoch": 0.94, + "grad_norm": 1.9161742752417454, + "learning_rate": 1.9210998227066846e-07, + "loss": 1.0362, + "step": 39873 + }, + { + "epoch": 0.94, + "grad_norm": 2.396557177203013, + "learning_rate": 1.9196116317309554e-07, + "loss": 0.8809, + "step": 39874 + }, + { + "epoch": 0.94, + "grad_norm": 2.0165668272116544, + "learning_rate": 1.918124011810285e-07, + "loss": 0.8086, + "step": 39875 + }, + { + "epoch": 0.94, + "grad_norm": 2.184684986162301, + "learning_rate": 1.9166369629533664e-07, + "loss": 1.0848, + "step": 39876 + }, + { + "epoch": 0.94, + "grad_norm": 2.2310502267033363, + "learning_rate": 1.9151504851688263e-07, + "loss": 1.0127, + "step": 39877 + }, + { + "epoch": 0.94, + "grad_norm": 1.1147136678294762, + "learning_rate": 1.9136645784653464e-07, + "loss": 0.924, + "step": 39878 + }, + { + "epoch": 0.94, + "grad_norm": 2.1700370401351017, + "learning_rate": 1.912179242851564e-07, + "loss": 1.0567, + "step": 39879 + }, + { + "epoch": 0.94, + "grad_norm": 1.9803653839136615, + "learning_rate": 1.910694478336117e-07, + "loss": 0.9537, + "step": 39880 + }, + { + "epoch": 0.94, + "grad_norm": 2.0196168742345546, + "learning_rate": 1.9092102849276651e-07, + "loss": 0.9142, + "step": 39881 + }, + { + "epoch": 0.94, + "grad_norm": 2.0999396427575237, + "learning_rate": 1.9077266626348457e-07, + "loss": 0.8468, + "step": 39882 + }, + { + "epoch": 0.94, + "grad_norm": 2.202105722837789, + "learning_rate": 1.9062436114663073e-07, + "loss": 0.9822, + "step": 39883 + }, + { + "epoch": 0.94, + "grad_norm": 2.0633940141421667, + "learning_rate": 1.9047611314306546e-07, + "loss": 0.8275, + "step": 39884 + }, + { + "epoch": 0.94, + "grad_norm": 1.8770810433729739, + "learning_rate": 1.903279222536547e-07, + "loss": 0.8911, + "step": 39885 + }, + { + "epoch": 0.94, + "grad_norm": 2.070924334268351, + "learning_rate": 1.9017978847926222e-07, + "loss": 1.0444, + "step": 39886 + }, + { + "epoch": 0.94, + "grad_norm": 2.0569817355472164, + "learning_rate": 1.9003171182074732e-07, + "loss": 0.8955, + "step": 39887 + }, + { + "epoch": 0.94, + "grad_norm": 2.340897820231524, + "learning_rate": 1.8988369227897263e-07, + "loss": 1.0336, + "step": 39888 + }, + { + "epoch": 0.94, + "grad_norm": 2.178699898082682, + "learning_rate": 1.8973572985480193e-07, + "loss": 1.0289, + "step": 39889 + }, + { + "epoch": 0.94, + "grad_norm": 2.1903997644352544, + "learning_rate": 1.8958782454909563e-07, + "loss": 1.1366, + "step": 39890 + }, + { + "epoch": 0.94, + "grad_norm": 1.066825068654573, + "learning_rate": 1.8943997636271528e-07, + "loss": 0.9509, + "step": 39891 + }, + { + "epoch": 0.94, + "grad_norm": 2.2773302797193944, + "learning_rate": 1.8929218529652128e-07, + "loss": 0.9639, + "step": 39892 + }, + { + "epoch": 0.94, + "grad_norm": 2.438605506686019, + "learning_rate": 1.8914445135137405e-07, + "loss": 0.8583, + "step": 39893 + }, + { + "epoch": 0.94, + "grad_norm": 2.2463678879440496, + "learning_rate": 1.889967745281329e-07, + "loss": 0.8611, + "step": 39894 + }, + { + "epoch": 0.94, + "grad_norm": 2.0283305574708455, + "learning_rate": 1.8884915482766054e-07, + "loss": 1.0209, + "step": 39895 + }, + { + "epoch": 0.94, + "grad_norm": 1.8514068889340147, + "learning_rate": 1.8870159225081285e-07, + "loss": 0.8814, + "step": 39896 + }, + { + "epoch": 0.94, + "grad_norm": 2.212837280777747, + "learning_rate": 1.8855408679845143e-07, + "loss": 0.9113, + "step": 39897 + }, + { + "epoch": 0.94, + "grad_norm": 2.284375228234409, + "learning_rate": 1.884066384714345e-07, + "loss": 1.0051, + "step": 39898 + }, + { + "epoch": 0.94, + "grad_norm": 1.9490502738337905, + "learning_rate": 1.8825924727062016e-07, + "loss": 0.9291, + "step": 39899 + }, + { + "epoch": 0.94, + "grad_norm": 2.0766776646657017, + "learning_rate": 1.8811191319686783e-07, + "loss": 0.8993, + "step": 39900 + }, + { + "epoch": 0.94, + "grad_norm": 1.927843187183063, + "learning_rate": 1.8796463625103233e-07, + "loss": 1.0075, + "step": 39901 + }, + { + "epoch": 0.94, + "grad_norm": 1.9419820634865192, + "learning_rate": 1.8781741643397412e-07, + "loss": 0.9484, + "step": 39902 + }, + { + "epoch": 0.94, + "grad_norm": 1.1736665243949727, + "learning_rate": 1.8767025374655023e-07, + "loss": 0.8948, + "step": 39903 + }, + { + "epoch": 0.94, + "grad_norm": 1.791692726673231, + "learning_rate": 1.875231481896167e-07, + "loss": 0.8983, + "step": 39904 + }, + { + "epoch": 0.94, + "grad_norm": 2.128483987301997, + "learning_rate": 1.8737609976402836e-07, + "loss": 1.0434, + "step": 39905 + }, + { + "epoch": 0.94, + "grad_norm": 2.1595463471619465, + "learning_rate": 1.8722910847064346e-07, + "loss": 0.8331, + "step": 39906 + }, + { + "epoch": 0.94, + "grad_norm": 1.973763776163778, + "learning_rate": 1.8708217431031793e-07, + "loss": 1.0588, + "step": 39907 + }, + { + "epoch": 0.94, + "grad_norm": 1.976069403713912, + "learning_rate": 1.869352972839067e-07, + "loss": 1.0331, + "step": 39908 + }, + { + "epoch": 0.94, + "grad_norm": 1.9578455613993162, + "learning_rate": 1.8678847739226459e-07, + "loss": 1.0089, + "step": 39909 + }, + { + "epoch": 0.94, + "grad_norm": 1.854612630796827, + "learning_rate": 1.8664171463624648e-07, + "loss": 0.8869, + "step": 39910 + }, + { + "epoch": 0.94, + "grad_norm": 1.9112424876253407, + "learning_rate": 1.8649500901670837e-07, + "loss": 0.9354, + "step": 39911 + }, + { + "epoch": 0.94, + "grad_norm": 1.8645291951441985, + "learning_rate": 1.863483605345029e-07, + "loss": 1.0749, + "step": 39912 + }, + { + "epoch": 0.94, + "grad_norm": 2.031686617962893, + "learning_rate": 1.8620176919048273e-07, + "loss": 0.9065, + "step": 39913 + }, + { + "epoch": 0.94, + "grad_norm": 1.909358574457407, + "learning_rate": 1.860552349855038e-07, + "loss": 1.074, + "step": 39914 + }, + { + "epoch": 0.94, + "grad_norm": 2.0388677011173675, + "learning_rate": 1.8590875792041884e-07, + "loss": 0.9829, + "step": 39915 + }, + { + "epoch": 0.94, + "grad_norm": 2.1624651938502835, + "learning_rate": 1.8576233799608046e-07, + "loss": 0.9343, + "step": 39916 + }, + { + "epoch": 0.94, + "grad_norm": 2.4056760192210667, + "learning_rate": 1.8561597521334018e-07, + "loss": 0.9788, + "step": 39917 + }, + { + "epoch": 0.94, + "grad_norm": 1.8710391405847973, + "learning_rate": 1.854696695730507e-07, + "loss": 0.8587, + "step": 39918 + }, + { + "epoch": 0.94, + "grad_norm": 2.2915902211929438, + "learning_rate": 1.853234210760635e-07, + "loss": 0.979, + "step": 39919 + }, + { + "epoch": 0.94, + "grad_norm": 1.9661725459931525, + "learning_rate": 1.8517722972323239e-07, + "loss": 1.0536, + "step": 39920 + }, + { + "epoch": 0.94, + "grad_norm": 1.8765628134851686, + "learning_rate": 1.8503109551540556e-07, + "loss": 0.9063, + "step": 39921 + }, + { + "epoch": 0.94, + "grad_norm": 2.116697002583524, + "learning_rate": 1.8488501845343565e-07, + "loss": 0.978, + "step": 39922 + }, + { + "epoch": 0.94, + "grad_norm": 2.7077248393988786, + "learning_rate": 1.847389985381709e-07, + "loss": 0.8643, + "step": 39923 + }, + { + "epoch": 0.94, + "grad_norm": 1.9931694171568188, + "learning_rate": 1.8459303577046505e-07, + "loss": 0.9242, + "step": 39924 + }, + { + "epoch": 0.94, + "grad_norm": 2.0580844678972547, + "learning_rate": 1.8444713015116632e-07, + "loss": 0.7194, + "step": 39925 + }, + { + "epoch": 0.94, + "grad_norm": 1.9995535567847218, + "learning_rate": 1.8430128168112293e-07, + "loss": 0.951, + "step": 39926 + }, + { + "epoch": 0.94, + "grad_norm": 2.1740623088353788, + "learning_rate": 1.841554903611842e-07, + "loss": 0.8746, + "step": 39927 + }, + { + "epoch": 0.94, + "grad_norm": 2.2157345524445473, + "learning_rate": 1.8400975619220163e-07, + "loss": 1.0275, + "step": 39928 + }, + { + "epoch": 0.94, + "grad_norm": 2.6592775719352457, + "learning_rate": 1.8386407917502126e-07, + "loss": 0.9552, + "step": 39929 + }, + { + "epoch": 0.94, + "grad_norm": 1.8987561098837662, + "learning_rate": 1.8371845931049126e-07, + "loss": 1.0647, + "step": 39930 + }, + { + "epoch": 0.94, + "grad_norm": 2.3269466956220928, + "learning_rate": 1.8357289659946098e-07, + "loss": 1.1251, + "step": 39931 + }, + { + "epoch": 0.94, + "grad_norm": 2.2947303810781055, + "learning_rate": 1.8342739104277753e-07, + "loss": 1.0139, + "step": 39932 + }, + { + "epoch": 0.94, + "grad_norm": 1.1156562908560124, + "learning_rate": 1.8328194264128686e-07, + "loss": 0.9331, + "step": 39933 + }, + { + "epoch": 0.94, + "grad_norm": 1.9157009890816166, + "learning_rate": 1.831365513958361e-07, + "loss": 0.9392, + "step": 39934 + }, + { + "epoch": 0.94, + "grad_norm": 1.7855299084096599, + "learning_rate": 1.8299121730727342e-07, + "loss": 0.8491, + "step": 39935 + }, + { + "epoch": 0.94, + "grad_norm": 1.949788530009188, + "learning_rate": 1.8284594037644377e-07, + "loss": 0.9748, + "step": 39936 + }, + { + "epoch": 0.94, + "grad_norm": 2.1898813691303474, + "learning_rate": 1.8270072060419308e-07, + "loss": 0.9858, + "step": 39937 + }, + { + "epoch": 0.94, + "grad_norm": 2.4728645548453954, + "learning_rate": 1.8255555799136627e-07, + "loss": 0.8755, + "step": 39938 + }, + { + "epoch": 0.94, + "grad_norm": 1.8747275658713036, + "learning_rate": 1.824104525388104e-07, + "loss": 0.9223, + "step": 39939 + }, + { + "epoch": 0.94, + "grad_norm": 3.0178911792838607, + "learning_rate": 1.8226540424736706e-07, + "loss": 0.9427, + "step": 39940 + }, + { + "epoch": 0.94, + "grad_norm": 1.9070489041840788, + "learning_rate": 1.8212041311788442e-07, + "loss": 0.9781, + "step": 39941 + }, + { + "epoch": 0.94, + "grad_norm": 1.0691654830448074, + "learning_rate": 1.8197547915120407e-07, + "loss": 0.9298, + "step": 39942 + }, + { + "epoch": 0.94, + "grad_norm": 1.0524574775527076, + "learning_rate": 1.8183060234817196e-07, + "loss": 0.9383, + "step": 39943 + }, + { + "epoch": 0.94, + "grad_norm": 2.6924410879596476, + "learning_rate": 1.8168578270962968e-07, + "loss": 0.9908, + "step": 39944 + }, + { + "epoch": 0.94, + "grad_norm": 2.484074704053357, + "learning_rate": 1.8154102023642205e-07, + "loss": 0.9638, + "step": 39945 + }, + { + "epoch": 0.94, + "grad_norm": 1.9334560627032351, + "learning_rate": 1.813963149293907e-07, + "loss": 0.9019, + "step": 39946 + }, + { + "epoch": 0.94, + "grad_norm": 2.0035577843103995, + "learning_rate": 1.812516667893771e-07, + "loss": 1.0164, + "step": 39947 + }, + { + "epoch": 0.94, + "grad_norm": 1.950352505311954, + "learning_rate": 1.811070758172262e-07, + "loss": 1.0852, + "step": 39948 + }, + { + "epoch": 0.94, + "grad_norm": 2.0617442539086044, + "learning_rate": 1.8096254201377838e-07, + "loss": 0.8912, + "step": 39949 + }, + { + "epoch": 0.94, + "grad_norm": 2.1376201629281617, + "learning_rate": 1.8081806537987524e-07, + "loss": 1.0643, + "step": 39950 + }, + { + "epoch": 0.94, + "grad_norm": 2.0687055199067994, + "learning_rate": 1.8067364591635828e-07, + "loss": 0.9899, + "step": 39951 + }, + { + "epoch": 0.94, + "grad_norm": 2.1246695446719612, + "learning_rate": 1.8052928362406686e-07, + "loss": 0.9125, + "step": 39952 + }, + { + "epoch": 0.94, + "grad_norm": 2.1448229611309775, + "learning_rate": 1.803849785038436e-07, + "loss": 0.8823, + "step": 39953 + }, + { + "epoch": 0.94, + "grad_norm": 1.953869386560313, + "learning_rate": 1.8024073055652903e-07, + "loss": 0.8991, + "step": 39954 + }, + { + "epoch": 0.94, + "grad_norm": 2.412752778328462, + "learning_rate": 1.8009653978296017e-07, + "loss": 1.0184, + "step": 39955 + }, + { + "epoch": 0.94, + "grad_norm": 1.9341383128053764, + "learning_rate": 1.7995240618397859e-07, + "loss": 0.9626, + "step": 39956 + }, + { + "epoch": 0.94, + "grad_norm": 2.7909012697906834, + "learning_rate": 1.798083297604236e-07, + "loss": 0.9302, + "step": 39957 + }, + { + "epoch": 0.94, + "grad_norm": 2.2540580488279933, + "learning_rate": 1.7966431051313349e-07, + "loss": 0.9159, + "step": 39958 + }, + { + "epoch": 0.94, + "grad_norm": 2.031401666494748, + "learning_rate": 1.795203484429453e-07, + "loss": 1.0348, + "step": 39959 + }, + { + "epoch": 0.94, + "grad_norm": 2.0337644410499887, + "learning_rate": 1.793764435507006e-07, + "loss": 0.981, + "step": 39960 + }, + { + "epoch": 0.94, + "grad_norm": 2.234311795570345, + "learning_rate": 1.792325958372343e-07, + "loss": 1.0547, + "step": 39961 + }, + { + "epoch": 0.94, + "grad_norm": 2.134251371839606, + "learning_rate": 1.7908880530338457e-07, + "loss": 0.9525, + "step": 39962 + }, + { + "epoch": 0.94, + "grad_norm": 1.7245088721602382, + "learning_rate": 1.7894507194999077e-07, + "loss": 0.9612, + "step": 39963 + }, + { + "epoch": 0.94, + "grad_norm": 2.257456795184223, + "learning_rate": 1.7880139577788779e-07, + "loss": 0.9812, + "step": 39964 + }, + { + "epoch": 0.94, + "grad_norm": 1.0883156182357423, + "learning_rate": 1.7865777678791163e-07, + "loss": 0.9681, + "step": 39965 + }, + { + "epoch": 0.94, + "grad_norm": 1.925286958210053, + "learning_rate": 1.7851421498089826e-07, + "loss": 0.9148, + "step": 39966 + }, + { + "epoch": 0.94, + "grad_norm": 2.0586592790047398, + "learning_rate": 1.7837071035768704e-07, + "loss": 1.1731, + "step": 39967 + }, + { + "epoch": 0.94, + "grad_norm": 2.204896621756318, + "learning_rate": 1.7822726291910952e-07, + "loss": 0.8787, + "step": 39968 + }, + { + "epoch": 0.94, + "grad_norm": 2.259274555173249, + "learning_rate": 1.780838726660028e-07, + "loss": 1.0086, + "step": 39969 + }, + { + "epoch": 0.94, + "grad_norm": 2.20076402011924, + "learning_rate": 1.7794053959920065e-07, + "loss": 1.011, + "step": 39970 + }, + { + "epoch": 0.94, + "grad_norm": 1.9241940886101605, + "learning_rate": 1.777972637195402e-07, + "loss": 0.9042, + "step": 39971 + }, + { + "epoch": 0.94, + "grad_norm": 2.1234037166080637, + "learning_rate": 1.776540450278519e-07, + "loss": 1.0311, + "step": 39972 + }, + { + "epoch": 0.94, + "grad_norm": 3.2723311950932907, + "learning_rate": 1.775108835249717e-07, + "loss": 0.9235, + "step": 39973 + }, + { + "epoch": 0.94, + "grad_norm": 2.0483074596013857, + "learning_rate": 1.7736777921173343e-07, + "loss": 0.9742, + "step": 39974 + }, + { + "epoch": 0.94, + "grad_norm": 2.3555311089278868, + "learning_rate": 1.7722473208896862e-07, + "loss": 1.0059, + "step": 39975 + }, + { + "epoch": 0.94, + "grad_norm": 1.105815965190547, + "learning_rate": 1.7708174215751217e-07, + "loss": 0.9666, + "step": 39976 + }, + { + "epoch": 0.94, + "grad_norm": 1.9972963983950502, + "learning_rate": 1.7693880941819564e-07, + "loss": 0.9793, + "step": 39977 + }, + { + "epoch": 0.94, + "grad_norm": 1.084577619109641, + "learning_rate": 1.7679593387185169e-07, + "loss": 0.9639, + "step": 39978 + }, + { + "epoch": 0.94, + "grad_norm": 2.1812676304703125, + "learning_rate": 1.7665311551931075e-07, + "loss": 0.9096, + "step": 39979 + }, + { + "epoch": 0.94, + "grad_norm": 2.279004168286684, + "learning_rate": 1.7651035436140662e-07, + "loss": 1.1105, + "step": 39980 + }, + { + "epoch": 0.94, + "grad_norm": 1.7611177781600176, + "learning_rate": 1.7636765039896865e-07, + "loss": 0.946, + "step": 39981 + }, + { + "epoch": 0.94, + "grad_norm": 1.9011051915144577, + "learning_rate": 1.7622500363282836e-07, + "loss": 0.9829, + "step": 39982 + }, + { + "epoch": 0.94, + "grad_norm": 1.1726629394765185, + "learning_rate": 1.7608241406381509e-07, + "loss": 0.9315, + "step": 39983 + }, + { + "epoch": 0.94, + "grad_norm": 1.9585219245772658, + "learning_rate": 1.7593988169276156e-07, + "loss": 0.9965, + "step": 39984 + }, + { + "epoch": 0.94, + "grad_norm": 1.681631385631633, + "learning_rate": 1.7579740652049705e-07, + "loss": 0.9546, + "step": 39985 + }, + { + "epoch": 0.94, + "grad_norm": 1.9826510633386996, + "learning_rate": 1.7565498854784868e-07, + "loss": 0.9812, + "step": 39986 + }, + { + "epoch": 0.94, + "grad_norm": 1.0608540596352294, + "learning_rate": 1.7551262777564693e-07, + "loss": 0.8851, + "step": 39987 + }, + { + "epoch": 0.94, + "grad_norm": 1.994176055589246, + "learning_rate": 1.7537032420472334e-07, + "loss": 0.9193, + "step": 39988 + }, + { + "epoch": 0.94, + "grad_norm": 1.948416798996148, + "learning_rate": 1.7522807783590278e-07, + "loss": 1.1011, + "step": 39989 + }, + { + "epoch": 0.94, + "grad_norm": 1.878386426581009, + "learning_rate": 1.7508588867001463e-07, + "loss": 0.9606, + "step": 39990 + }, + { + "epoch": 0.94, + "grad_norm": 1.696750275992883, + "learning_rate": 1.749437567078871e-07, + "loss": 0.8929, + "step": 39991 + }, + { + "epoch": 0.94, + "grad_norm": 2.025558377481115, + "learning_rate": 1.748016819503473e-07, + "loss": 1.1089, + "step": 39992 + }, + { + "epoch": 0.94, + "grad_norm": 1.984859527781411, + "learning_rate": 1.7465966439822346e-07, + "loss": 0.9312, + "step": 39993 + }, + { + "epoch": 0.94, + "grad_norm": 1.9194930880585659, + "learning_rate": 1.745177040523416e-07, + "loss": 1.0527, + "step": 39994 + }, + { + "epoch": 0.94, + "grad_norm": 2.311510030973902, + "learning_rate": 1.7437580091352878e-07, + "loss": 0.9795, + "step": 39995 + }, + { + "epoch": 0.94, + "grad_norm": 2.018489152470117, + "learning_rate": 1.7423395498260887e-07, + "loss": 1.0285, + "step": 39996 + }, + { + "epoch": 0.94, + "grad_norm": 1.809566175489025, + "learning_rate": 1.7409216626041225e-07, + "loss": 1.0858, + "step": 39997 + }, + { + "epoch": 0.94, + "grad_norm": 1.9584990240791749, + "learning_rate": 1.7395043474775942e-07, + "loss": 0.9966, + "step": 39998 + }, + { + "epoch": 0.94, + "grad_norm": 2.1150408464312043, + "learning_rate": 1.7380876044547966e-07, + "loss": 1.0232, + "step": 39999 + }, + { + "epoch": 0.94, + "grad_norm": 1.0770227548278601, + "learning_rate": 1.736671433543957e-07, + "loss": 0.9541, + "step": 40000 + }, + { + "epoch": 0.94, + "grad_norm": 1.7534143835045435, + "learning_rate": 1.7352558347533243e-07, + "loss": 0.8175, + "step": 40001 + }, + { + "epoch": 0.94, + "grad_norm": 1.0857318871766588, + "learning_rate": 1.7338408080911473e-07, + "loss": 0.8443, + "step": 40002 + }, + { + "epoch": 0.94, + "grad_norm": 2.139586345697919, + "learning_rate": 1.7324263535656528e-07, + "loss": 1.0321, + "step": 40003 + }, + { + "epoch": 0.94, + "grad_norm": 1.0802764156547413, + "learning_rate": 1.7310124711850894e-07, + "loss": 0.9844, + "step": 40004 + }, + { + "epoch": 0.94, + "grad_norm": 2.6854818444614716, + "learning_rate": 1.7295991609576845e-07, + "loss": 0.9709, + "step": 40005 + }, + { + "epoch": 0.94, + "grad_norm": 1.8615020547352563, + "learning_rate": 1.7281864228916644e-07, + "loss": 0.9566, + "step": 40006 + }, + { + "epoch": 0.94, + "grad_norm": 2.0299352366023076, + "learning_rate": 1.726774256995245e-07, + "loss": 1.0783, + "step": 40007 + }, + { + "epoch": 0.94, + "grad_norm": 1.1595145621976435, + "learning_rate": 1.725362663276653e-07, + "loss": 1.0144, + "step": 40008 + }, + { + "epoch": 0.94, + "grad_norm": 2.004258709021685, + "learning_rate": 1.723951641744137e-07, + "loss": 0.9947, + "step": 40009 + }, + { + "epoch": 0.94, + "grad_norm": 2.000553327903749, + "learning_rate": 1.72254119240588e-07, + "loss": 0.9261, + "step": 40010 + }, + { + "epoch": 0.94, + "grad_norm": 1.9000015701013746, + "learning_rate": 1.721131315270097e-07, + "loss": 0.9698, + "step": 40011 + }, + { + "epoch": 0.94, + "grad_norm": 1.9787663446136996, + "learning_rate": 1.719722010344993e-07, + "loss": 0.9936, + "step": 40012 + }, + { + "epoch": 0.94, + "grad_norm": 1.8242995929749843, + "learning_rate": 1.7183132776387945e-07, + "loss": 1.029, + "step": 40013 + }, + { + "epoch": 0.94, + "grad_norm": 9.482383621950907, + "learning_rate": 1.716905117159695e-07, + "loss": 0.952, + "step": 40014 + }, + { + "epoch": 0.94, + "grad_norm": 1.9654222990713301, + "learning_rate": 1.715497528915888e-07, + "loss": 0.9411, + "step": 40015 + }, + { + "epoch": 0.94, + "grad_norm": 1.8744918094197767, + "learning_rate": 1.714090512915567e-07, + "loss": 1.0435, + "step": 40016 + }, + { + "epoch": 0.94, + "grad_norm": 2.2567128562006746, + "learning_rate": 1.7126840691669255e-07, + "loss": 0.8973, + "step": 40017 + }, + { + "epoch": 0.94, + "grad_norm": 2.036924051900772, + "learning_rate": 1.7112781976781567e-07, + "loss": 1.082, + "step": 40018 + }, + { + "epoch": 0.94, + "grad_norm": 1.812928035669001, + "learning_rate": 1.709872898457443e-07, + "loss": 1.0178, + "step": 40019 + }, + { + "epoch": 0.94, + "grad_norm": 1.151912028657152, + "learning_rate": 1.708468171512978e-07, + "loss": 0.9427, + "step": 40020 + }, + { + "epoch": 0.94, + "grad_norm": 1.9095903990218666, + "learning_rate": 1.7070640168529107e-07, + "loss": 0.9979, + "step": 40021 + }, + { + "epoch": 0.94, + "grad_norm": 1.8552640675282925, + "learning_rate": 1.7056604344854565e-07, + "loss": 0.9299, + "step": 40022 + }, + { + "epoch": 0.94, + "grad_norm": 2.340835092316471, + "learning_rate": 1.7042574244187537e-07, + "loss": 0.9039, + "step": 40023 + }, + { + "epoch": 0.94, + "grad_norm": 2.0292399390873372, + "learning_rate": 1.7028549866609845e-07, + "loss": 0.9601, + "step": 40024 + }, + { + "epoch": 0.94, + "grad_norm": 1.8730404036895412, + "learning_rate": 1.7014531212203088e-07, + "loss": 1.0222, + "step": 40025 + }, + { + "epoch": 0.94, + "grad_norm": 1.8497640395025783, + "learning_rate": 1.7000518281049093e-07, + "loss": 0.9269, + "step": 40026 + }, + { + "epoch": 0.94, + "grad_norm": 2.054265661255459, + "learning_rate": 1.698651107322924e-07, + "loss": 1.0351, + "step": 40027 + }, + { + "epoch": 0.94, + "grad_norm": 2.2240059576790663, + "learning_rate": 1.6972509588825015e-07, + "loss": 0.9842, + "step": 40028 + }, + { + "epoch": 0.94, + "grad_norm": 2.0166376661389345, + "learning_rate": 1.6958513827918134e-07, + "loss": 0.9715, + "step": 40029 + }, + { + "epoch": 0.94, + "grad_norm": 1.9911710429720253, + "learning_rate": 1.6944523790590084e-07, + "loss": 0.9772, + "step": 40030 + }, + { + "epoch": 0.94, + "grad_norm": 2.105710221883216, + "learning_rate": 1.6930539476922248e-07, + "loss": 1.0046, + "step": 40031 + }, + { + "epoch": 0.94, + "grad_norm": 2.0046668038504074, + "learning_rate": 1.6916560886995893e-07, + "loss": 0.9661, + "step": 40032 + }, + { + "epoch": 0.94, + "grad_norm": 1.1207465117335773, + "learning_rate": 1.690258802089262e-07, + "loss": 0.891, + "step": 40033 + }, + { + "epoch": 0.94, + "grad_norm": 1.0663080646217669, + "learning_rate": 1.6888620878693806e-07, + "loss": 0.8726, + "step": 40034 + }, + { + "epoch": 0.94, + "grad_norm": 1.935897176667441, + "learning_rate": 1.6874659460480726e-07, + "loss": 0.9216, + "step": 40035 + }, + { + "epoch": 0.94, + "grad_norm": 3.721913781782775, + "learning_rate": 1.686070376633453e-07, + "loss": 1.0449, + "step": 40036 + }, + { + "epoch": 0.94, + "grad_norm": 1.8467724476488598, + "learning_rate": 1.6846753796336491e-07, + "loss": 0.9347, + "step": 40037 + }, + { + "epoch": 0.94, + "grad_norm": 2.1206606699201394, + "learning_rate": 1.6832809550568097e-07, + "loss": 1.0373, + "step": 40038 + }, + { + "epoch": 0.94, + "grad_norm": 1.8029068446393688, + "learning_rate": 1.6818871029110284e-07, + "loss": 0.9918, + "step": 40039 + }, + { + "epoch": 0.94, + "grad_norm": 1.899043132410514, + "learning_rate": 1.680493823204421e-07, + "loss": 0.8086, + "step": 40040 + }, + { + "epoch": 0.94, + "grad_norm": 2.155289169071758, + "learning_rate": 1.6791011159451141e-07, + "loss": 0.9446, + "step": 40041 + }, + { + "epoch": 0.94, + "grad_norm": 2.113519145409504, + "learning_rate": 1.6777089811412017e-07, + "loss": 0.9654, + "step": 40042 + }, + { + "epoch": 0.94, + "grad_norm": 2.2802353533111015, + "learning_rate": 1.676317418800799e-07, + "loss": 1.0332, + "step": 40043 + }, + { + "epoch": 0.94, + "grad_norm": 2.2889778366478524, + "learning_rate": 1.6749264289320111e-07, + "loss": 0.9043, + "step": 40044 + }, + { + "epoch": 0.94, + "grad_norm": 2.251295473011094, + "learning_rate": 1.67353601154292e-07, + "loss": 0.8691, + "step": 40045 + }, + { + "epoch": 0.94, + "grad_norm": 1.1156009384390135, + "learning_rate": 1.6721461666416305e-07, + "loss": 1.0008, + "step": 40046 + }, + { + "epoch": 0.94, + "grad_norm": 1.9942469463437538, + "learning_rate": 1.6707568942362473e-07, + "loss": 0.9677, + "step": 40047 + }, + { + "epoch": 0.94, + "grad_norm": 2.0307262151099796, + "learning_rate": 1.6693681943348527e-07, + "loss": 0.9658, + "step": 40048 + }, + { + "epoch": 0.94, + "grad_norm": 1.8568652647247772, + "learning_rate": 1.6679800669455072e-07, + "loss": 0.8587, + "step": 40049 + }, + { + "epoch": 0.94, + "grad_norm": 2.2150787550726156, + "learning_rate": 1.666592512076326e-07, + "loss": 0.962, + "step": 40050 + }, + { + "epoch": 0.94, + "grad_norm": 1.9998855033013476, + "learning_rate": 1.6652055297353698e-07, + "loss": 1.051, + "step": 40051 + }, + { + "epoch": 0.94, + "grad_norm": 3.9489613470594755, + "learning_rate": 1.663819119930732e-07, + "loss": 0.939, + "step": 40052 + }, + { + "epoch": 0.94, + "grad_norm": 2.0505880465779613, + "learning_rate": 1.6624332826704616e-07, + "loss": 0.8004, + "step": 40053 + }, + { + "epoch": 0.94, + "grad_norm": 1.8923335622285744, + "learning_rate": 1.66104801796263e-07, + "loss": 0.8838, + "step": 40054 + }, + { + "epoch": 0.94, + "grad_norm": 1.7954084260448244, + "learning_rate": 1.6596633258153193e-07, + "loss": 0.9403, + "step": 40055 + }, + { + "epoch": 0.94, + "grad_norm": 1.9566241891994676, + "learning_rate": 1.6582792062365905e-07, + "loss": 0.936, + "step": 40056 + }, + { + "epoch": 0.94, + "grad_norm": 4.0975930963182465, + "learning_rate": 1.6568956592344808e-07, + "loss": 0.8674, + "step": 40057 + }, + { + "epoch": 0.94, + "grad_norm": 2.8902250328949277, + "learning_rate": 1.6555126848170622e-07, + "loss": 0.9001, + "step": 40058 + }, + { + "epoch": 0.94, + "grad_norm": 1.885717379210753, + "learning_rate": 1.6541302829923945e-07, + "loss": 0.9515, + "step": 40059 + }, + { + "epoch": 0.94, + "grad_norm": 2.2703857817722395, + "learning_rate": 1.6527484537684935e-07, + "loss": 1.0428, + "step": 40060 + }, + { + "epoch": 0.94, + "grad_norm": 1.9495093649434514, + "learning_rate": 1.6513671971534528e-07, + "loss": 1.1418, + "step": 40061 + }, + { + "epoch": 0.94, + "grad_norm": 2.2714241149810817, + "learning_rate": 1.6499865131552773e-07, + "loss": 0.9659, + "step": 40062 + }, + { + "epoch": 0.94, + "grad_norm": 1.9886644555976065, + "learning_rate": 1.6486064017820157e-07, + "loss": 0.9779, + "step": 40063 + }, + { + "epoch": 0.94, + "grad_norm": 2.2503860940409206, + "learning_rate": 1.6472268630417065e-07, + "loss": 1.0908, + "step": 40064 + }, + { + "epoch": 0.94, + "grad_norm": 1.9101449419049634, + "learning_rate": 1.6458478969423876e-07, + "loss": 0.9883, + "step": 40065 + }, + { + "epoch": 0.94, + "grad_norm": 2.033919391812894, + "learning_rate": 1.6444695034920743e-07, + "loss": 1.0006, + "step": 40066 + }, + { + "epoch": 0.94, + "grad_norm": 2.0289752317958976, + "learning_rate": 1.643091682698794e-07, + "loss": 0.8846, + "step": 40067 + }, + { + "epoch": 0.94, + "grad_norm": 1.7558736213142205, + "learning_rate": 1.6417144345705738e-07, + "loss": 1.0008, + "step": 40068 + }, + { + "epoch": 0.94, + "grad_norm": 2.2420657465231435, + "learning_rate": 1.6403377591154512e-07, + "loss": 1.0363, + "step": 40069 + }, + { + "epoch": 0.94, + "grad_norm": 2.114357217055286, + "learning_rate": 1.6389616563413978e-07, + "loss": 0.939, + "step": 40070 + }, + { + "epoch": 0.94, + "grad_norm": 2.341229723804228, + "learning_rate": 1.637586126256463e-07, + "loss": 0.9127, + "step": 40071 + }, + { + "epoch": 0.94, + "grad_norm": 1.8893665635772057, + "learning_rate": 1.6362111688686288e-07, + "loss": 0.7453, + "step": 40072 + }, + { + "epoch": 0.94, + "grad_norm": 1.8267191098847644, + "learning_rate": 1.6348367841859337e-07, + "loss": 0.9602, + "step": 40073 + }, + { + "epoch": 0.94, + "grad_norm": 2.3529558750752564, + "learning_rate": 1.6334629722163597e-07, + "loss": 0.9641, + "step": 40074 + }, + { + "epoch": 0.94, + "grad_norm": 2.193187600850377, + "learning_rate": 1.6320897329678899e-07, + "loss": 1.0007, + "step": 40075 + }, + { + "epoch": 0.94, + "grad_norm": 1.1915800638124305, + "learning_rate": 1.6307170664485506e-07, + "loss": 0.8615, + "step": 40076 + }, + { + "epoch": 0.94, + "grad_norm": 1.9470851269547849, + "learning_rate": 1.6293449726663136e-07, + "loss": 0.9958, + "step": 40077 + }, + { + "epoch": 0.94, + "grad_norm": 2.2903000528341773, + "learning_rate": 1.6279734516291724e-07, + "loss": 0.9374, + "step": 40078 + }, + { + "epoch": 0.94, + "grad_norm": 2.070027412723122, + "learning_rate": 1.6266025033451206e-07, + "loss": 0.8755, + "step": 40079 + }, + { + "epoch": 0.94, + "grad_norm": 2.4687225775614112, + "learning_rate": 1.6252321278221294e-07, + "loss": 1.069, + "step": 40080 + }, + { + "epoch": 0.94, + "grad_norm": 1.9802577586568701, + "learning_rate": 1.6238623250681817e-07, + "loss": 0.9368, + "step": 40081 + }, + { + "epoch": 0.94, + "grad_norm": 1.9507144619978443, + "learning_rate": 1.6224930950912487e-07, + "loss": 1.1197, + "step": 40082 + }, + { + "epoch": 0.94, + "grad_norm": 2.483876480087436, + "learning_rate": 1.621124437899324e-07, + "loss": 0.882, + "step": 40083 + }, + { + "epoch": 0.94, + "grad_norm": 1.885074862994078, + "learning_rate": 1.6197563535003347e-07, + "loss": 0.9757, + "step": 40084 + }, + { + "epoch": 0.94, + "grad_norm": 1.9730616514116712, + "learning_rate": 1.6183888419022854e-07, + "loss": 0.8737, + "step": 40085 + }, + { + "epoch": 0.94, + "grad_norm": 1.783743319981653, + "learning_rate": 1.617021903113114e-07, + "loss": 0.903, + "step": 40086 + }, + { + "epoch": 0.94, + "grad_norm": 2.1107390400856096, + "learning_rate": 1.6156555371408033e-07, + "loss": 0.8891, + "step": 40087 + }, + { + "epoch": 0.94, + "grad_norm": 2.3056047415584957, + "learning_rate": 1.6142897439932804e-07, + "loss": 0.9778, + "step": 40088 + }, + { + "epoch": 0.94, + "grad_norm": 2.0367994945419547, + "learning_rate": 1.6129245236785162e-07, + "loss": 0.9082, + "step": 40089 + }, + { + "epoch": 0.94, + "grad_norm": 2.06158747137633, + "learning_rate": 1.6115598762044493e-07, + "loss": 0.9816, + "step": 40090 + }, + { + "epoch": 0.94, + "grad_norm": 2.1972923773090653, + "learning_rate": 1.6101958015790397e-07, + "loss": 0.891, + "step": 40091 + }, + { + "epoch": 0.94, + "grad_norm": 2.20691803245134, + "learning_rate": 1.6088322998102147e-07, + "loss": 1.0064, + "step": 40092 + }, + { + "epoch": 0.94, + "grad_norm": 2.201047671494046, + "learning_rate": 1.607469370905912e-07, + "loss": 0.9862, + "step": 40093 + }, + { + "epoch": 0.94, + "grad_norm": 1.9854621100010732, + "learning_rate": 1.606107014874081e-07, + "loss": 0.9466, + "step": 40094 + }, + { + "epoch": 0.94, + "grad_norm": 1.9120962898178782, + "learning_rate": 1.6047452317226485e-07, + "loss": 1.0571, + "step": 40095 + }, + { + "epoch": 0.94, + "grad_norm": 1.9821554435367439, + "learning_rate": 1.6033840214595309e-07, + "loss": 1.0194, + "step": 40096 + }, + { + "epoch": 0.94, + "grad_norm": 2.9495408816967683, + "learning_rate": 1.602023384092677e-07, + "loss": 1.0081, + "step": 40097 + }, + { + "epoch": 0.94, + "grad_norm": 2.0461099493677057, + "learning_rate": 1.6006633196299804e-07, + "loss": 0.8949, + "step": 40098 + }, + { + "epoch": 0.94, + "grad_norm": 2.0042899204374005, + "learning_rate": 1.5993038280793905e-07, + "loss": 1.0687, + "step": 40099 + }, + { + "epoch": 0.94, + "grad_norm": 2.0537638618880227, + "learning_rate": 1.5979449094487898e-07, + "loss": 0.9981, + "step": 40100 + }, + { + "epoch": 0.94, + "grad_norm": 1.96670867560926, + "learning_rate": 1.5965865637461163e-07, + "loss": 1.0881, + "step": 40101 + }, + { + "epoch": 0.94, + "grad_norm": 1.1483716165286677, + "learning_rate": 1.595228790979264e-07, + "loss": 0.9524, + "step": 40102 + }, + { + "epoch": 0.94, + "grad_norm": 2.3053566946630415, + "learning_rate": 1.5938715911561596e-07, + "loss": 1.0727, + "step": 40103 + }, + { + "epoch": 0.94, + "grad_norm": 2.0953621133032985, + "learning_rate": 1.592514964284686e-07, + "loss": 0.8586, + "step": 40104 + }, + { + "epoch": 0.94, + "grad_norm": 2.3223765038359745, + "learning_rate": 1.5911589103727253e-07, + "loss": 1.0637, + "step": 40105 + }, + { + "epoch": 0.94, + "grad_norm": 1.9650746347561077, + "learning_rate": 1.589803429428205e-07, + "loss": 0.9426, + "step": 40106 + }, + { + "epoch": 0.94, + "grad_norm": 1.1049656058907795, + "learning_rate": 1.5884485214590072e-07, + "loss": 0.9099, + "step": 40107 + }, + { + "epoch": 0.94, + "grad_norm": 1.856696904372513, + "learning_rate": 1.587094186473026e-07, + "loss": 0.9618, + "step": 40108 + }, + { + "epoch": 0.94, + "grad_norm": 2.161219161933424, + "learning_rate": 1.5857404244781326e-07, + "loss": 1.0522, + "step": 40109 + }, + { + "epoch": 0.94, + "grad_norm": 2.388927934136383, + "learning_rate": 1.5843872354822099e-07, + "loss": 1.025, + "step": 40110 + }, + { + "epoch": 0.94, + "grad_norm": 2.138084884165282, + "learning_rate": 1.5830346194931513e-07, + "loss": 0.947, + "step": 40111 + }, + { + "epoch": 0.95, + "grad_norm": 1.9058952450919517, + "learning_rate": 1.5816825765188171e-07, + "loss": 1.0239, + "step": 40112 + }, + { + "epoch": 0.95, + "grad_norm": 1.08282021554435, + "learning_rate": 1.58033110656709e-07, + "loss": 0.9398, + "step": 40113 + }, + { + "epoch": 0.95, + "grad_norm": 2.194857438984922, + "learning_rate": 1.5789802096458306e-07, + "loss": 0.9218, + "step": 40114 + }, + { + "epoch": 0.95, + "grad_norm": 2.342846935010994, + "learning_rate": 1.5776298857629103e-07, + "loss": 0.9917, + "step": 40115 + }, + { + "epoch": 0.95, + "grad_norm": 2.1074848429691917, + "learning_rate": 1.576280134926189e-07, + "loss": 0.9556, + "step": 40116 + }, + { + "epoch": 0.95, + "grad_norm": 1.8981352727901397, + "learning_rate": 1.5749309571435277e-07, + "loss": 1.0275, + "step": 40117 + }, + { + "epoch": 0.95, + "grad_norm": 1.8311660772138894, + "learning_rate": 1.5735823524227755e-07, + "loss": 0.8901, + "step": 40118 + }, + { + "epoch": 0.95, + "grad_norm": 1.8877497598331525, + "learning_rate": 1.5722343207717816e-07, + "loss": 0.9339, + "step": 40119 + }, + { + "epoch": 0.95, + "grad_norm": 1.0850941736516053, + "learning_rate": 1.5708868621984175e-07, + "loss": 0.9817, + "step": 40120 + }, + { + "epoch": 0.95, + "grad_norm": 2.0983198519656385, + "learning_rate": 1.569539976710499e-07, + "loss": 1.085, + "step": 40121 + }, + { + "epoch": 0.95, + "grad_norm": 2.052839347154333, + "learning_rate": 1.5681936643158867e-07, + "loss": 0.959, + "step": 40122 + }, + { + "epoch": 0.95, + "grad_norm": 2.799176337798241, + "learning_rate": 1.566847925022408e-07, + "loss": 1.0687, + "step": 40123 + }, + { + "epoch": 0.95, + "grad_norm": 1.9626411148336647, + "learning_rate": 1.5655027588379113e-07, + "loss": 0.865, + "step": 40124 + }, + { + "epoch": 0.95, + "grad_norm": 2.31999634545972, + "learning_rate": 1.5641581657702244e-07, + "loss": 1.0038, + "step": 40125 + }, + { + "epoch": 0.95, + "grad_norm": 2.1914160200401014, + "learning_rate": 1.5628141458271628e-07, + "loss": 0.9786, + "step": 40126 + }, + { + "epoch": 0.95, + "grad_norm": 2.3789452829901365, + "learning_rate": 1.561470699016565e-07, + "loss": 1.0857, + "step": 40127 + }, + { + "epoch": 0.95, + "grad_norm": 2.109900761906812, + "learning_rate": 1.5601278253462582e-07, + "loss": 0.9495, + "step": 40128 + }, + { + "epoch": 0.95, + "grad_norm": 1.9920239003876796, + "learning_rate": 1.558785524824047e-07, + "loss": 1.037, + "step": 40129 + }, + { + "epoch": 0.95, + "grad_norm": 1.8502995178792097, + "learning_rate": 1.5574437974577473e-07, + "loss": 0.8501, + "step": 40130 + }, + { + "epoch": 0.95, + "grad_norm": 2.0069593810427193, + "learning_rate": 1.5561026432551752e-07, + "loss": 1.1249, + "step": 40131 + }, + { + "epoch": 0.95, + "grad_norm": 1.1247842136988118, + "learning_rate": 1.5547620622241467e-07, + "loss": 0.8848, + "step": 40132 + }, + { + "epoch": 0.95, + "grad_norm": 1.9842398373474925, + "learning_rate": 1.5534220543724666e-07, + "loss": 0.8898, + "step": 40133 + }, + { + "epoch": 0.95, + "grad_norm": 1.09885214343453, + "learning_rate": 1.5520826197079176e-07, + "loss": 0.8891, + "step": 40134 + }, + { + "epoch": 0.95, + "grad_norm": 1.8549401240468748, + "learning_rate": 1.5507437582383156e-07, + "loss": 0.9998, + "step": 40135 + }, + { + "epoch": 0.95, + "grad_norm": 2.4810886568680774, + "learning_rate": 1.5494054699714545e-07, + "loss": 1.1015, + "step": 40136 + }, + { + "epoch": 0.95, + "grad_norm": 1.0892870183459222, + "learning_rate": 1.548067754915128e-07, + "loss": 0.9319, + "step": 40137 + }, + { + "epoch": 0.95, + "grad_norm": 2.2622689018006112, + "learning_rate": 1.5467306130771077e-07, + "loss": 0.8967, + "step": 40138 + }, + { + "epoch": 0.95, + "grad_norm": 1.989967110885256, + "learning_rate": 1.5453940444652092e-07, + "loss": 0.9515, + "step": 40139 + }, + { + "epoch": 0.95, + "grad_norm": 1.946430513626985, + "learning_rate": 1.5440580490871827e-07, + "loss": 0.961, + "step": 40140 + }, + { + "epoch": 0.95, + "grad_norm": 1.9208342066790947, + "learning_rate": 1.5427226269508322e-07, + "loss": 1.0231, + "step": 40141 + }, + { + "epoch": 0.95, + "grad_norm": 2.0930862387685476, + "learning_rate": 1.5413877780639076e-07, + "loss": 0.9, + "step": 40142 + }, + { + "epoch": 0.95, + "grad_norm": 1.8885563391113263, + "learning_rate": 1.5400535024342023e-07, + "loss": 0.9413, + "step": 40143 + }, + { + "epoch": 0.95, + "grad_norm": 2.8169002256420184, + "learning_rate": 1.538719800069477e-07, + "loss": 0.8853, + "step": 40144 + }, + { + "epoch": 0.95, + "grad_norm": 2.143898422600371, + "learning_rate": 1.537386670977503e-07, + "loss": 1.0377, + "step": 40145 + }, + { + "epoch": 0.95, + "grad_norm": 1.8516056536371797, + "learning_rate": 1.5360541151660303e-07, + "loss": 1.0638, + "step": 40146 + }, + { + "epoch": 0.95, + "grad_norm": 2.2984977366655186, + "learning_rate": 1.5347221326428187e-07, + "loss": 1.0204, + "step": 40147 + }, + { + "epoch": 0.95, + "grad_norm": 2.312286969328929, + "learning_rate": 1.5333907234156288e-07, + "loss": 0.9733, + "step": 40148 + }, + { + "epoch": 0.95, + "grad_norm": 2.1829722634628625, + "learning_rate": 1.532059887492221e-07, + "loss": 0.7873, + "step": 40149 + }, + { + "epoch": 0.95, + "grad_norm": 1.9865780370632429, + "learning_rate": 1.530729624880345e-07, + "loss": 1.0269, + "step": 40150 + }, + { + "epoch": 0.95, + "grad_norm": 2.0658971754572586, + "learning_rate": 1.5293999355877164e-07, + "loss": 1.0961, + "step": 40151 + }, + { + "epoch": 0.95, + "grad_norm": 1.901764686582235, + "learning_rate": 1.528070819622096e-07, + "loss": 0.8879, + "step": 40152 + }, + { + "epoch": 0.95, + "grad_norm": 2.2768015205770613, + "learning_rate": 1.526742276991222e-07, + "loss": 1.0254, + "step": 40153 + }, + { + "epoch": 0.95, + "grad_norm": 1.0639925675633795, + "learning_rate": 1.525414307702855e-07, + "loss": 0.9759, + "step": 40154 + }, + { + "epoch": 0.95, + "grad_norm": 1.9662356110890264, + "learning_rate": 1.5240869117646774e-07, + "loss": 0.8567, + "step": 40155 + }, + { + "epoch": 0.95, + "grad_norm": 1.901193737970664, + "learning_rate": 1.5227600891844495e-07, + "loss": 1.0319, + "step": 40156 + }, + { + "epoch": 0.95, + "grad_norm": 2.2017921509497658, + "learning_rate": 1.521433839969888e-07, + "loss": 1.0179, + "step": 40157 + }, + { + "epoch": 0.95, + "grad_norm": 1.9714671062151934, + "learning_rate": 1.5201081641287196e-07, + "loss": 1.0465, + "step": 40158 + }, + { + "epoch": 0.95, + "grad_norm": 1.115919694259056, + "learning_rate": 1.5187830616686605e-07, + "loss": 0.8687, + "step": 40159 + }, + { + "epoch": 0.95, + "grad_norm": 1.108216533774739, + "learning_rate": 1.5174585325974266e-07, + "loss": 0.9205, + "step": 40160 + }, + { + "epoch": 0.95, + "grad_norm": 1.9359299836809494, + "learning_rate": 1.516134576922723e-07, + "loss": 0.9503, + "step": 40161 + }, + { + "epoch": 0.95, + "grad_norm": 1.0448223120057574, + "learning_rate": 1.5148111946522659e-07, + "loss": 0.8572, + "step": 40162 + }, + { + "epoch": 0.95, + "grad_norm": 2.094538951799681, + "learning_rate": 1.51348838579376e-07, + "loss": 0.986, + "step": 40163 + }, + { + "epoch": 0.95, + "grad_norm": 2.556943786666194, + "learning_rate": 1.51216615035491e-07, + "loss": 0.9925, + "step": 40164 + }, + { + "epoch": 0.95, + "grad_norm": 1.8944385343654653, + "learning_rate": 1.5108444883433994e-07, + "loss": 1.0752, + "step": 40165 + }, + { + "epoch": 0.95, + "grad_norm": 2.0034286202155247, + "learning_rate": 1.5095233997669323e-07, + "loss": 0.9553, + "step": 40166 + }, + { + "epoch": 0.95, + "grad_norm": 2.265742509410973, + "learning_rate": 1.508202884633203e-07, + "loss": 0.9363, + "step": 40167 + }, + { + "epoch": 0.95, + "grad_norm": 2.0905146872101277, + "learning_rate": 1.5068829429499054e-07, + "loss": 0.9795, + "step": 40168 + }, + { + "epoch": 0.95, + "grad_norm": 2.286285709876168, + "learning_rate": 1.505563574724711e-07, + "loss": 0.9135, + "step": 40169 + }, + { + "epoch": 0.95, + "grad_norm": 1.888775508299772, + "learning_rate": 1.5042447799653136e-07, + "loss": 0.9018, + "step": 40170 + }, + { + "epoch": 0.95, + "grad_norm": 2.010143022395489, + "learning_rate": 1.5029265586793851e-07, + "loss": 0.9026, + "step": 40171 + }, + { + "epoch": 0.95, + "grad_norm": 2.523982707990648, + "learning_rate": 1.501608910874608e-07, + "loss": 0.9603, + "step": 40172 + }, + { + "epoch": 0.95, + "grad_norm": 2.460867410846221, + "learning_rate": 1.5002918365586317e-07, + "loss": 0.9856, + "step": 40173 + }, + { + "epoch": 0.95, + "grad_norm": 1.9055200653744189, + "learning_rate": 1.4989753357391502e-07, + "loss": 0.9508, + "step": 40174 + }, + { + "epoch": 0.95, + "grad_norm": 1.0756522816729681, + "learning_rate": 1.4976594084238127e-07, + "loss": 0.9039, + "step": 40175 + }, + { + "epoch": 0.95, + "grad_norm": 2.4397719697805793, + "learning_rate": 1.4963440546203023e-07, + "loss": 0.9504, + "step": 40176 + }, + { + "epoch": 0.95, + "grad_norm": 1.119580232969689, + "learning_rate": 1.495029274336246e-07, + "loss": 0.9187, + "step": 40177 + }, + { + "epoch": 0.95, + "grad_norm": 2.3769988357396716, + "learning_rate": 1.4937150675793266e-07, + "loss": 1.0129, + "step": 40178 + }, + { + "epoch": 0.95, + "grad_norm": 1.7675992433593095, + "learning_rate": 1.4924014343571824e-07, + "loss": 0.8978, + "step": 40179 + }, + { + "epoch": 0.95, + "grad_norm": 1.8417518642837745, + "learning_rate": 1.491088374677463e-07, + "loss": 1.0075, + "step": 40180 + }, + { + "epoch": 0.95, + "grad_norm": 2.1687014994287677, + "learning_rate": 1.4897758885478064e-07, + "loss": 0.8896, + "step": 40181 + }, + { + "epoch": 0.95, + "grad_norm": 1.9915436103842439, + "learning_rate": 1.4884639759758735e-07, + "loss": 0.905, + "step": 40182 + }, + { + "epoch": 0.95, + "grad_norm": 1.9146839058836311, + "learning_rate": 1.48715263696928e-07, + "loss": 1.0001, + "step": 40183 + }, + { + "epoch": 0.95, + "grad_norm": 1.9039393381560878, + "learning_rate": 1.485841871535687e-07, + "loss": 0.9872, + "step": 40184 + }, + { + "epoch": 0.95, + "grad_norm": 1.7002660723882366, + "learning_rate": 1.4845316796827103e-07, + "loss": 0.9565, + "step": 40185 + }, + { + "epoch": 0.95, + "grad_norm": 1.9373312824280209, + "learning_rate": 1.483222061417966e-07, + "loss": 1.0014, + "step": 40186 + }, + { + "epoch": 0.95, + "grad_norm": 1.9593327227773927, + "learning_rate": 1.4819130167491036e-07, + "loss": 0.9977, + "step": 40187 + }, + { + "epoch": 0.95, + "grad_norm": 1.0686729636055359, + "learning_rate": 1.4806045456837393e-07, + "loss": 0.927, + "step": 40188 + }, + { + "epoch": 0.95, + "grad_norm": 1.9334230822082938, + "learning_rate": 1.479296648229478e-07, + "loss": 0.9708, + "step": 40189 + }, + { + "epoch": 0.95, + "grad_norm": 1.8750200973367472, + "learning_rate": 1.4779893243939358e-07, + "loss": 1.0103, + "step": 40190 + }, + { + "epoch": 0.95, + "grad_norm": 2.0329984691088896, + "learning_rate": 1.47668257418474e-07, + "loss": 0.9263, + "step": 40191 + }, + { + "epoch": 0.95, + "grad_norm": 1.991811594606628, + "learning_rate": 1.4753763976094848e-07, + "loss": 0.92, + "step": 40192 + }, + { + "epoch": 0.95, + "grad_norm": 2.0004279678308508, + "learning_rate": 1.474070794675797e-07, + "loss": 1.0573, + "step": 40193 + }, + { + "epoch": 0.95, + "grad_norm": 1.9518288226204747, + "learning_rate": 1.4727657653912374e-07, + "loss": 0.9798, + "step": 40194 + }, + { + "epoch": 0.95, + "grad_norm": 2.4908590024255326, + "learning_rate": 1.4714613097634445e-07, + "loss": 1.0584, + "step": 40195 + }, + { + "epoch": 0.95, + "grad_norm": 1.928637927641201, + "learning_rate": 1.4701574277999896e-07, + "loss": 1.0175, + "step": 40196 + }, + { + "epoch": 0.95, + "grad_norm": 1.8192782760191075, + "learning_rate": 1.468854119508478e-07, + "loss": 1.0461, + "step": 40197 + }, + { + "epoch": 0.95, + "grad_norm": 2.049933496728898, + "learning_rate": 1.4675513848964928e-07, + "loss": 0.918, + "step": 40198 + }, + { + "epoch": 0.95, + "grad_norm": 2.3593268578043483, + "learning_rate": 1.4662492239716164e-07, + "loss": 1.0128, + "step": 40199 + }, + { + "epoch": 0.95, + "grad_norm": 1.8023435369424485, + "learning_rate": 1.4649476367414318e-07, + "loss": 1.0781, + "step": 40200 + }, + { + "epoch": 0.95, + "grad_norm": 2.071762339901544, + "learning_rate": 1.4636466232135217e-07, + "loss": 1.0859, + "step": 40201 + }, + { + "epoch": 0.95, + "grad_norm": 1.8610554562458526, + "learning_rate": 1.4623461833954467e-07, + "loss": 0.8475, + "step": 40202 + }, + { + "epoch": 0.95, + "grad_norm": 2.2802474489099285, + "learning_rate": 1.4610463172948008e-07, + "loss": 0.9852, + "step": 40203 + }, + { + "epoch": 0.95, + "grad_norm": 1.9826097338320918, + "learning_rate": 1.4597470249191225e-07, + "loss": 1.1719, + "step": 40204 + }, + { + "epoch": 0.95, + "grad_norm": 1.8922362505966634, + "learning_rate": 1.458448306276017e-07, + "loss": 0.9845, + "step": 40205 + }, + { + "epoch": 0.95, + "grad_norm": 1.8899132099511198, + "learning_rate": 1.457150161373011e-07, + "loss": 0.9748, + "step": 40206 + }, + { + "epoch": 0.95, + "grad_norm": 2.1655235737764555, + "learning_rate": 1.455852590217677e-07, + "loss": 1.0174, + "step": 40207 + }, + { + "epoch": 0.95, + "grad_norm": 3.5220175636384337, + "learning_rate": 1.4545555928175637e-07, + "loss": 1.0816, + "step": 40208 + }, + { + "epoch": 0.95, + "grad_norm": 1.9665688532400156, + "learning_rate": 1.4532591691802323e-07, + "loss": 0.9968, + "step": 40209 + }, + { + "epoch": 0.95, + "grad_norm": 2.1480127658189243, + "learning_rate": 1.451963319313221e-07, + "loss": 0.9296, + "step": 40210 + }, + { + "epoch": 0.95, + "grad_norm": 1.8429916547780636, + "learning_rate": 1.4506680432240794e-07, + "loss": 0.9572, + "step": 40211 + }, + { + "epoch": 0.95, + "grad_norm": 2.3982253710848966, + "learning_rate": 1.449373340920346e-07, + "loss": 1.0557, + "step": 40212 + }, + { + "epoch": 0.95, + "grad_norm": 1.8820949009662826, + "learning_rate": 1.4480792124095698e-07, + "loss": 0.8943, + "step": 40213 + }, + { + "epoch": 0.95, + "grad_norm": 1.9519975037594077, + "learning_rate": 1.446785657699279e-07, + "loss": 1.0417, + "step": 40214 + }, + { + "epoch": 0.95, + "grad_norm": 2.053139897877424, + "learning_rate": 1.445492676796989e-07, + "loss": 0.873, + "step": 40215 + }, + { + "epoch": 0.95, + "grad_norm": 2.0575496338191117, + "learning_rate": 1.4442002697102497e-07, + "loss": 0.9102, + "step": 40216 + }, + { + "epoch": 0.95, + "grad_norm": 2.0496957743407984, + "learning_rate": 1.4429084364465885e-07, + "loss": 0.8832, + "step": 40217 + }, + { + "epoch": 0.95, + "grad_norm": 1.1852392592165095, + "learning_rate": 1.44161717701351e-07, + "loss": 0.9316, + "step": 40218 + }, + { + "epoch": 0.95, + "grad_norm": 2.1727259114185196, + "learning_rate": 1.440326491418531e-07, + "loss": 0.9719, + "step": 40219 + }, + { + "epoch": 0.95, + "grad_norm": 1.8014022906708067, + "learning_rate": 1.4390363796691897e-07, + "loss": 1.0175, + "step": 40220 + }, + { + "epoch": 0.95, + "grad_norm": 1.8714813234686134, + "learning_rate": 1.437746841772969e-07, + "loss": 0.9971, + "step": 40221 + }, + { + "epoch": 0.95, + "grad_norm": 1.8618951687533951, + "learning_rate": 1.4364578777373962e-07, + "loss": 1.1578, + "step": 40222 + }, + { + "epoch": 0.95, + "grad_norm": 2.433021432476808, + "learning_rate": 1.4351694875699763e-07, + "loss": 1.043, + "step": 40223 + }, + { + "epoch": 0.95, + "grad_norm": 1.9436058350611383, + "learning_rate": 1.4338816712781921e-07, + "loss": 0.9761, + "step": 40224 + }, + { + "epoch": 0.95, + "grad_norm": 2.146283251804599, + "learning_rate": 1.4325944288695603e-07, + "loss": 0.891, + "step": 40225 + }, + { + "epoch": 0.95, + "grad_norm": 2.1154948463403933, + "learning_rate": 1.4313077603515745e-07, + "loss": 0.8854, + "step": 40226 + }, + { + "epoch": 0.95, + "grad_norm": 2.0529018305667828, + "learning_rate": 1.4300216657317178e-07, + "loss": 0.9492, + "step": 40227 + }, + { + "epoch": 0.95, + "grad_norm": 1.9758447408279385, + "learning_rate": 1.4287361450174843e-07, + "loss": 0.9735, + "step": 40228 + }, + { + "epoch": 0.95, + "grad_norm": 1.9150763293564332, + "learning_rate": 1.4274511982163454e-07, + "loss": 1.0799, + "step": 40229 + }, + { + "epoch": 0.95, + "grad_norm": 2.4702240119382397, + "learning_rate": 1.4261668253358062e-07, + "loss": 1.0035, + "step": 40230 + }, + { + "epoch": 0.95, + "grad_norm": 2.2486161987784694, + "learning_rate": 1.424883026383328e-07, + "loss": 0.8803, + "step": 40231 + }, + { + "epoch": 0.95, + "grad_norm": 1.9915418116117327, + "learning_rate": 1.423599801366382e-07, + "loss": 0.8967, + "step": 40232 + }, + { + "epoch": 0.95, + "grad_norm": 1.0089619527388154, + "learning_rate": 1.4223171502924516e-07, + "loss": 0.9308, + "step": 40233 + }, + { + "epoch": 0.95, + "grad_norm": 1.8614080775037292, + "learning_rate": 1.421035073168997e-07, + "loss": 1.0121, + "step": 40234 + }, + { + "epoch": 0.95, + "grad_norm": 1.110312971674957, + "learning_rate": 1.4197535700035014e-07, + "loss": 0.859, + "step": 40235 + }, + { + "epoch": 0.95, + "grad_norm": 2.044971222951556, + "learning_rate": 1.418472640803392e-07, + "loss": 1.0592, + "step": 40236 + }, + { + "epoch": 0.95, + "grad_norm": 2.365596484725927, + "learning_rate": 1.417192285576141e-07, + "loss": 1.0242, + "step": 40237 + }, + { + "epoch": 0.95, + "grad_norm": 1.863553325339251, + "learning_rate": 1.415912504329231e-07, + "loss": 1.0092, + "step": 40238 + }, + { + "epoch": 0.95, + "grad_norm": 1.9033468600829617, + "learning_rate": 1.4146332970700782e-07, + "loss": 0.9527, + "step": 40239 + }, + { + "epoch": 0.95, + "grad_norm": 2.0120458357254636, + "learning_rate": 1.4133546638061323e-07, + "loss": 0.8986, + "step": 40240 + }, + { + "epoch": 0.95, + "grad_norm": 1.2026492373542492, + "learning_rate": 1.4120766045448653e-07, + "loss": 0.908, + "step": 40241 + }, + { + "epoch": 0.95, + "grad_norm": 2.6869510746558656, + "learning_rate": 1.4107991192936821e-07, + "loss": 0.9504, + "step": 40242 + }, + { + "epoch": 0.95, + "grad_norm": 2.298775744124989, + "learning_rate": 1.4095222080600546e-07, + "loss": 0.9439, + "step": 40243 + }, + { + "epoch": 0.95, + "grad_norm": 1.9308337738551689, + "learning_rate": 1.408245870851388e-07, + "loss": 0.9952, + "step": 40244 + }, + { + "epoch": 0.95, + "grad_norm": 1.878110144088507, + "learning_rate": 1.406970107675143e-07, + "loss": 1.0525, + "step": 40245 + }, + { + "epoch": 0.95, + "grad_norm": 2.588848823808159, + "learning_rate": 1.4056949185387136e-07, + "loss": 0.996, + "step": 40246 + }, + { + "epoch": 0.95, + "grad_norm": 1.97525763305838, + "learning_rate": 1.4044203034495497e-07, + "loss": 1.0071, + "step": 40247 + }, + { + "epoch": 0.95, + "grad_norm": 1.8853112626755364, + "learning_rate": 1.4031462624150782e-07, + "loss": 0.8534, + "step": 40248 + }, + { + "epoch": 0.95, + "grad_norm": 2.0019018902851737, + "learning_rate": 1.4018727954426937e-07, + "loss": 1.0485, + "step": 40249 + }, + { + "epoch": 0.95, + "grad_norm": 2.123700658563828, + "learning_rate": 1.4005999025398231e-07, + "loss": 1.1123, + "step": 40250 + }, + { + "epoch": 0.95, + "grad_norm": 2.0602830975633357, + "learning_rate": 1.399327583713872e-07, + "loss": 0.9387, + "step": 40251 + }, + { + "epoch": 0.95, + "grad_norm": 1.1293185594440251, + "learning_rate": 1.3980558389722677e-07, + "loss": 0.8939, + "step": 40252 + }, + { + "epoch": 0.95, + "grad_norm": 2.2106141373727675, + "learning_rate": 1.3967846683223818e-07, + "loss": 0.8871, + "step": 40253 + }, + { + "epoch": 0.95, + "grad_norm": 2.006808518944062, + "learning_rate": 1.3955140717716308e-07, + "loss": 0.9083, + "step": 40254 + }, + { + "epoch": 0.95, + "grad_norm": 1.083021603712325, + "learning_rate": 1.3942440493274313e-07, + "loss": 0.9868, + "step": 40255 + }, + { + "epoch": 0.95, + "grad_norm": 1.8835175882458834, + "learning_rate": 1.3929746009971434e-07, + "loss": 0.9621, + "step": 40256 + }, + { + "epoch": 0.95, + "grad_norm": 2.159883281689585, + "learning_rate": 1.3917057267881951e-07, + "loss": 0.9204, + "step": 40257 + }, + { + "epoch": 0.95, + "grad_norm": 2.1270577843499434, + "learning_rate": 1.3904374267079356e-07, + "loss": 1.013, + "step": 40258 + }, + { + "epoch": 0.95, + "grad_norm": 4.433434861440781, + "learning_rate": 1.3891697007637817e-07, + "loss": 0.9331, + "step": 40259 + }, + { + "epoch": 0.95, + "grad_norm": 1.768131179526453, + "learning_rate": 1.3879025489630936e-07, + "loss": 0.9658, + "step": 40260 + }, + { + "epoch": 0.95, + "grad_norm": 2.4147913603698057, + "learning_rate": 1.386635971313266e-07, + "loss": 0.8619, + "step": 40261 + }, + { + "epoch": 0.95, + "grad_norm": 2.1305654046097873, + "learning_rate": 1.3853699678216592e-07, + "loss": 1.0032, + "step": 40262 + }, + { + "epoch": 0.95, + "grad_norm": 2.3566780874495348, + "learning_rate": 1.3841045384956453e-07, + "loss": 1.0348, + "step": 40263 + }, + { + "epoch": 0.95, + "grad_norm": 1.968630852622895, + "learning_rate": 1.3828396833425962e-07, + "loss": 0.7782, + "step": 40264 + }, + { + "epoch": 0.95, + "grad_norm": 2.9809583227325067, + "learning_rate": 1.3815754023698834e-07, + "loss": 0.9768, + "step": 40265 + }, + { + "epoch": 0.95, + "grad_norm": 1.9336676527480006, + "learning_rate": 1.3803116955848573e-07, + "loss": 1.0032, + "step": 40266 + }, + { + "epoch": 0.95, + "grad_norm": 2.047276948534226, + "learning_rate": 1.379048562994878e-07, + "loss": 1.0167, + "step": 40267 + }, + { + "epoch": 0.95, + "grad_norm": 1.9278400048826823, + "learning_rate": 1.3777860046072955e-07, + "loss": 1.0155, + "step": 40268 + }, + { + "epoch": 0.95, + "grad_norm": 1.1215419377039113, + "learning_rate": 1.3765240204294706e-07, + "loss": 0.9402, + "step": 40269 + }, + { + "epoch": 0.95, + "grad_norm": 2.190623923170829, + "learning_rate": 1.3752626104687528e-07, + "loss": 0.9524, + "step": 40270 + }, + { + "epoch": 0.95, + "grad_norm": 2.015093124362701, + "learning_rate": 1.3740017747324698e-07, + "loss": 0.9642, + "step": 40271 + }, + { + "epoch": 0.95, + "grad_norm": 1.7627100942578373, + "learning_rate": 1.372741513227982e-07, + "loss": 0.8106, + "step": 40272 + }, + { + "epoch": 0.95, + "grad_norm": 1.7022929336804975, + "learning_rate": 1.3714818259626173e-07, + "loss": 0.9052, + "step": 40273 + }, + { + "epoch": 0.95, + "grad_norm": 1.145491083544134, + "learning_rate": 1.3702227129437028e-07, + "loss": 0.8965, + "step": 40274 + }, + { + "epoch": 0.95, + "grad_norm": 2.1300668136283054, + "learning_rate": 1.3689641741785775e-07, + "loss": 0.9765, + "step": 40275 + }, + { + "epoch": 0.95, + "grad_norm": 1.8563391042691264, + "learning_rate": 1.3677062096745686e-07, + "loss": 0.9438, + "step": 40276 + }, + { + "epoch": 0.95, + "grad_norm": 2.208219587849067, + "learning_rate": 1.3664488194390147e-07, + "loss": 0.9247, + "step": 40277 + }, + { + "epoch": 0.95, + "grad_norm": 1.9279127614552927, + "learning_rate": 1.365192003479221e-07, + "loss": 0.9119, + "step": 40278 + }, + { + "epoch": 0.95, + "grad_norm": 1.9704697844618524, + "learning_rate": 1.363935761802493e-07, + "loss": 0.9106, + "step": 40279 + }, + { + "epoch": 0.95, + "grad_norm": 2.0055476384488182, + "learning_rate": 1.362680094416169e-07, + "loss": 1.013, + "step": 40280 + }, + { + "epoch": 0.95, + "grad_norm": 1.9923090974779014, + "learning_rate": 1.3614250013275433e-07, + "loss": 0.9409, + "step": 40281 + }, + { + "epoch": 0.95, + "grad_norm": 2.0934308003574498, + "learning_rate": 1.3601704825439323e-07, + "loss": 0.9869, + "step": 40282 + }, + { + "epoch": 0.95, + "grad_norm": 1.1009952970458468, + "learning_rate": 1.35891653807263e-07, + "loss": 0.9198, + "step": 40283 + }, + { + "epoch": 0.95, + "grad_norm": 2.0193939685738655, + "learning_rate": 1.357663167920953e-07, + "loss": 0.9724, + "step": 40284 + }, + { + "epoch": 0.95, + "grad_norm": 1.8680606110230806, + "learning_rate": 1.356410372096184e-07, + "loss": 0.8873, + "step": 40285 + }, + { + "epoch": 0.95, + "grad_norm": 2.7237692075930626, + "learning_rate": 1.35515815060564e-07, + "loss": 0.9545, + "step": 40286 + }, + { + "epoch": 0.95, + "grad_norm": 2.51863783851268, + "learning_rate": 1.3539065034565813e-07, + "loss": 0.9302, + "step": 40287 + }, + { + "epoch": 0.95, + "grad_norm": 2.141813625381163, + "learning_rate": 1.3526554306563134e-07, + "loss": 0.7944, + "step": 40288 + }, + { + "epoch": 0.95, + "grad_norm": 1.8904363236585238, + "learning_rate": 1.351404932212119e-07, + "loss": 0.9882, + "step": 40289 + }, + { + "epoch": 0.95, + "grad_norm": 2.130708460529246, + "learning_rate": 1.3501550081312707e-07, + "loss": 0.94, + "step": 40290 + }, + { + "epoch": 0.95, + "grad_norm": 1.9240027997067233, + "learning_rate": 1.3489056584210626e-07, + "loss": 0.9493, + "step": 40291 + }, + { + "epoch": 0.95, + "grad_norm": 1.9641790586378602, + "learning_rate": 1.3476568830887439e-07, + "loss": 0.9331, + "step": 40292 + }, + { + "epoch": 0.95, + "grad_norm": 2.083113471055117, + "learning_rate": 1.346408682141609e-07, + "loss": 0.9406, + "step": 40293 + }, + { + "epoch": 0.95, + "grad_norm": 1.9235134120282886, + "learning_rate": 1.345161055586919e-07, + "loss": 1.1474, + "step": 40294 + }, + { + "epoch": 0.95, + "grad_norm": 1.9577970112436613, + "learning_rate": 1.3439140034319342e-07, + "loss": 1.0247, + "step": 40295 + }, + { + "epoch": 0.95, + "grad_norm": 1.040317644947057, + "learning_rate": 1.342667525683905e-07, + "loss": 0.7665, + "step": 40296 + }, + { + "epoch": 0.95, + "grad_norm": 1.0959984952638093, + "learning_rate": 1.3414216223501031e-07, + "loss": 0.9367, + "step": 40297 + }, + { + "epoch": 0.95, + "grad_norm": 1.9578233649832593, + "learning_rate": 1.3401762934377894e-07, + "loss": 0.9246, + "step": 40298 + }, + { + "epoch": 0.95, + "grad_norm": 1.933421083649678, + "learning_rate": 1.3389315389541911e-07, + "loss": 1.0098, + "step": 40299 + }, + { + "epoch": 0.95, + "grad_norm": 2.140052323762888, + "learning_rate": 1.3376873589065809e-07, + "loss": 1.1234, + "step": 40300 + }, + { + "epoch": 0.95, + "grad_norm": 1.8668770285079894, + "learning_rate": 1.3364437533021856e-07, + "loss": 1.0897, + "step": 40301 + }, + { + "epoch": 0.95, + "grad_norm": 1.0447156459742923, + "learning_rate": 1.3352007221482443e-07, + "loss": 0.9525, + "step": 40302 + }, + { + "epoch": 0.95, + "grad_norm": 1.769764592457068, + "learning_rate": 1.3339582654520177e-07, + "loss": 0.9871, + "step": 40303 + }, + { + "epoch": 0.95, + "grad_norm": 1.9583850949568964, + "learning_rate": 1.3327163832207114e-07, + "loss": 0.9043, + "step": 40304 + }, + { + "epoch": 0.95, + "grad_norm": 2.0253110460053736, + "learning_rate": 1.3314750754615634e-07, + "loss": 1.0549, + "step": 40305 + }, + { + "epoch": 0.95, + "grad_norm": 1.9943874496052445, + "learning_rate": 1.330234342181813e-07, + "loss": 0.9504, + "step": 40306 + }, + { + "epoch": 0.95, + "grad_norm": 2.361068802454299, + "learning_rate": 1.3289941833886767e-07, + "loss": 0.9362, + "step": 40307 + }, + { + "epoch": 0.95, + "grad_norm": 2.0695687919543864, + "learning_rate": 1.3277545990893813e-07, + "loss": 0.9597, + "step": 40308 + }, + { + "epoch": 0.95, + "grad_norm": 2.0082198116712595, + "learning_rate": 1.326515589291122e-07, + "loss": 1.0915, + "step": 40309 + }, + { + "epoch": 0.95, + "grad_norm": 2.2156431172242876, + "learning_rate": 1.3252771540011367e-07, + "loss": 1.0958, + "step": 40310 + }, + { + "epoch": 0.95, + "grad_norm": 2.1941591265539118, + "learning_rate": 1.324039293226631e-07, + "loss": 0.9744, + "step": 40311 + }, + { + "epoch": 0.95, + "grad_norm": 2.2395640553524463, + "learning_rate": 1.3228020069748105e-07, + "loss": 0.8784, + "step": 40312 + }, + { + "epoch": 0.95, + "grad_norm": 2.1132167819109036, + "learning_rate": 1.321565295252869e-07, + "loss": 0.9973, + "step": 40313 + }, + { + "epoch": 0.95, + "grad_norm": 1.8249547456942368, + "learning_rate": 1.3203291580680122e-07, + "loss": 0.8323, + "step": 40314 + }, + { + "epoch": 0.95, + "grad_norm": 2.1167604510152485, + "learning_rate": 1.3190935954274565e-07, + "loss": 0.8444, + "step": 40315 + }, + { + "epoch": 0.95, + "grad_norm": 2.079691542937478, + "learning_rate": 1.3178586073383738e-07, + "loss": 0.9264, + "step": 40316 + }, + { + "epoch": 0.95, + "grad_norm": 2.361629421160059, + "learning_rate": 1.3166241938079472e-07, + "loss": 1.0631, + "step": 40317 + }, + { + "epoch": 0.95, + "grad_norm": 1.8563514097641272, + "learning_rate": 1.3153903548433822e-07, + "loss": 1.0769, + "step": 40318 + }, + { + "epoch": 0.95, + "grad_norm": 1.9817876414324778, + "learning_rate": 1.314157090451873e-07, + "loss": 0.9284, + "step": 40319 + }, + { + "epoch": 0.95, + "grad_norm": 1.9000033824404887, + "learning_rate": 1.3129244006405805e-07, + "loss": 0.9879, + "step": 40320 + }, + { + "epoch": 0.95, + "grad_norm": 2.0315458705804175, + "learning_rate": 1.311692285416666e-07, + "loss": 0.9734, + "step": 40321 + }, + { + "epoch": 0.95, + "grad_norm": 2.054266838291899, + "learning_rate": 1.3104607447873452e-07, + "loss": 1.1121, + "step": 40322 + }, + { + "epoch": 0.95, + "grad_norm": 2.1199656588414877, + "learning_rate": 1.3092297787597576e-07, + "loss": 0.9013, + "step": 40323 + }, + { + "epoch": 0.95, + "grad_norm": 1.9854897611298947, + "learning_rate": 1.3079993873410857e-07, + "loss": 0.8854, + "step": 40324 + }, + { + "epoch": 0.95, + "grad_norm": 1.9867407584443295, + "learning_rate": 1.3067695705384687e-07, + "loss": 0.913, + "step": 40325 + }, + { + "epoch": 0.95, + "grad_norm": 2.2143648796133917, + "learning_rate": 1.3055403283591006e-07, + "loss": 1.0307, + "step": 40326 + }, + { + "epoch": 0.95, + "grad_norm": 5.486421641959689, + "learning_rate": 1.3043116608101204e-07, + "loss": 0.948, + "step": 40327 + }, + { + "epoch": 0.95, + "grad_norm": 2.113420301959766, + "learning_rate": 1.3030835678986886e-07, + "loss": 0.9498, + "step": 40328 + }, + { + "epoch": 0.95, + "grad_norm": 2.6095445944193507, + "learning_rate": 1.3018560496319443e-07, + "loss": 1.0334, + "step": 40329 + }, + { + "epoch": 0.95, + "grad_norm": 1.9325451691698554, + "learning_rate": 1.3006291060170373e-07, + "loss": 1.0089, + "step": 40330 + }, + { + "epoch": 0.95, + "grad_norm": 2.0878267226017018, + "learning_rate": 1.2994027370611173e-07, + "loss": 1.0315, + "step": 40331 + }, + { + "epoch": 0.95, + "grad_norm": 2.229673453651311, + "learning_rate": 1.298176942771323e-07, + "loss": 1.022, + "step": 40332 + }, + { + "epoch": 0.95, + "grad_norm": 2.074343459709579, + "learning_rate": 1.2969517231548045e-07, + "loss": 0.9296, + "step": 40333 + }, + { + "epoch": 0.95, + "grad_norm": 1.8948142905777088, + "learning_rate": 1.295727078218667e-07, + "loss": 0.9633, + "step": 40334 + }, + { + "epoch": 0.95, + "grad_norm": 1.838675125100254, + "learning_rate": 1.2945030079700606e-07, + "loss": 1.0278, + "step": 40335 + }, + { + "epoch": 0.95, + "grad_norm": 2.9106764004522008, + "learning_rate": 1.2932795124161124e-07, + "loss": 1.0224, + "step": 40336 + }, + { + "epoch": 0.95, + "grad_norm": 2.26090037882048, + "learning_rate": 1.2920565915639395e-07, + "loss": 1.0651, + "step": 40337 + }, + { + "epoch": 0.95, + "grad_norm": 2.602741182210091, + "learning_rate": 1.290834245420658e-07, + "loss": 0.9928, + "step": 40338 + }, + { + "epoch": 0.95, + "grad_norm": 1.940584708519144, + "learning_rate": 1.2896124739933846e-07, + "loss": 0.9133, + "step": 40339 + }, + { + "epoch": 0.95, + "grad_norm": 2.0091517037176216, + "learning_rate": 1.288391277289247e-07, + "loss": 0.8309, + "step": 40340 + }, + { + "epoch": 0.95, + "grad_norm": 2.258614260489843, + "learning_rate": 1.2871706553153395e-07, + "loss": 1.0117, + "step": 40341 + }, + { + "epoch": 0.95, + "grad_norm": 2.1414531090107145, + "learning_rate": 1.2859506080787898e-07, + "loss": 1.059, + "step": 40342 + }, + { + "epoch": 0.95, + "grad_norm": 1.8898095741740837, + "learning_rate": 1.284731135586681e-07, + "loss": 0.9636, + "step": 40343 + }, + { + "epoch": 0.95, + "grad_norm": 2.088080271174052, + "learning_rate": 1.2835122378461186e-07, + "loss": 1.0393, + "step": 40344 + }, + { + "epoch": 0.95, + "grad_norm": 1.7702034244378424, + "learning_rate": 1.2822939148641967e-07, + "loss": 0.9797, + "step": 40345 + }, + { + "epoch": 0.95, + "grad_norm": 2.1959652267592467, + "learning_rate": 1.2810761666480321e-07, + "loss": 0.9354, + "step": 40346 + }, + { + "epoch": 0.95, + "grad_norm": 1.91666399220331, + "learning_rate": 1.2798589932046856e-07, + "loss": 0.9052, + "step": 40347 + }, + { + "epoch": 0.95, + "grad_norm": 2.0107126005093194, + "learning_rate": 1.2786423945412407e-07, + "loss": 0.8995, + "step": 40348 + }, + { + "epoch": 0.95, + "grad_norm": 2.192756502465457, + "learning_rate": 1.2774263706648027e-07, + "loss": 0.9694, + "step": 40349 + }, + { + "epoch": 0.95, + "grad_norm": 3.021951370700534, + "learning_rate": 1.2762109215824547e-07, + "loss": 1.0674, + "step": 40350 + }, + { + "epoch": 0.95, + "grad_norm": 1.9767460365022858, + "learning_rate": 1.274996047301258e-07, + "loss": 0.9047, + "step": 40351 + }, + { + "epoch": 0.95, + "grad_norm": 1.9978359900956693, + "learning_rate": 1.2737817478282845e-07, + "loss": 0.9317, + "step": 40352 + }, + { + "epoch": 0.95, + "grad_norm": 1.9371597150802913, + "learning_rate": 1.2725680231706061e-07, + "loss": 0.9551, + "step": 40353 + }, + { + "epoch": 0.95, + "grad_norm": 1.9363830559288042, + "learning_rate": 1.2713548733352955e-07, + "loss": 0.9735, + "step": 40354 + }, + { + "epoch": 0.95, + "grad_norm": 1.8689974778119125, + "learning_rate": 1.2701422983294243e-07, + "loss": 0.9678, + "step": 40355 + }, + { + "epoch": 0.95, + "grad_norm": 2.421488413268527, + "learning_rate": 1.2689302981600316e-07, + "loss": 0.9469, + "step": 40356 + }, + { + "epoch": 0.95, + "grad_norm": 2.431052147845657, + "learning_rate": 1.2677188728341895e-07, + "loss": 0.8271, + "step": 40357 + }, + { + "epoch": 0.95, + "grad_norm": 1.907296310805794, + "learning_rate": 1.2665080223589366e-07, + "loss": 1.0962, + "step": 40358 + }, + { + "epoch": 0.95, + "grad_norm": 1.818249140838893, + "learning_rate": 1.2652977467413453e-07, + "loss": 0.8641, + "step": 40359 + }, + { + "epoch": 0.95, + "grad_norm": 2.0815706715965137, + "learning_rate": 1.264088045988443e-07, + "loss": 0.9409, + "step": 40360 + }, + { + "epoch": 0.95, + "grad_norm": 1.081126475061298, + "learning_rate": 1.26287892010728e-07, + "loss": 0.8917, + "step": 40361 + }, + { + "epoch": 0.95, + "grad_norm": 2.2177779682882464, + "learning_rate": 1.2616703691048947e-07, + "loss": 1.0082, + "step": 40362 + }, + { + "epoch": 0.95, + "grad_norm": 1.9971876746484096, + "learning_rate": 1.2604623929883265e-07, + "loss": 0.8888, + "step": 40363 + }, + { + "epoch": 0.95, + "grad_norm": 1.9755024208767067, + "learning_rate": 1.2592549917646136e-07, + "loss": 0.9405, + "step": 40364 + }, + { + "epoch": 0.95, + "grad_norm": 2.382060703121772, + "learning_rate": 1.2580481654407616e-07, + "loss": 1.085, + "step": 40365 + }, + { + "epoch": 0.95, + "grad_norm": 2.096010383362095, + "learning_rate": 1.2568419140238208e-07, + "loss": 0.9841, + "step": 40366 + }, + { + "epoch": 0.95, + "grad_norm": 2.1803108017602364, + "learning_rate": 1.2556362375208186e-07, + "loss": 0.9355, + "step": 40367 + }, + { + "epoch": 0.95, + "grad_norm": 2.333252672496883, + "learning_rate": 1.2544311359387606e-07, + "loss": 0.9106, + "step": 40368 + }, + { + "epoch": 0.95, + "grad_norm": 2.159893483217114, + "learning_rate": 1.2532266092846524e-07, + "loss": 0.8766, + "step": 40369 + }, + { + "epoch": 0.95, + "grad_norm": 1.739430971939391, + "learning_rate": 1.2520226575655325e-07, + "loss": 0.8844, + "step": 40370 + }, + { + "epoch": 0.95, + "grad_norm": 2.090931419509954, + "learning_rate": 1.2508192807883956e-07, + "loss": 1.0516, + "step": 40371 + }, + { + "epoch": 0.95, + "grad_norm": 1.1167796731046358, + "learning_rate": 1.2496164789602583e-07, + "loss": 0.9355, + "step": 40372 + }, + { + "epoch": 0.95, + "grad_norm": 1.9672886625208015, + "learning_rate": 1.248414252088115e-07, + "loss": 0.934, + "step": 40373 + }, + { + "epoch": 0.95, + "grad_norm": 1.861150501938581, + "learning_rate": 1.2472126001789708e-07, + "loss": 1.009, + "step": 40374 + }, + { + "epoch": 0.95, + "grad_norm": 1.876105592158645, + "learning_rate": 1.2460115232398207e-07, + "loss": 0.9136, + "step": 40375 + }, + { + "epoch": 0.95, + "grad_norm": 3.3683816504104107, + "learning_rate": 1.2448110212776477e-07, + "loss": 0.9655, + "step": 40376 + }, + { + "epoch": 0.95, + "grad_norm": 1.7771197717620508, + "learning_rate": 1.2436110942994573e-07, + "loss": 0.8928, + "step": 40377 + }, + { + "epoch": 0.95, + "grad_norm": 2.2277227144782312, + "learning_rate": 1.242411742312233e-07, + "loss": 0.8639, + "step": 40378 + }, + { + "epoch": 0.95, + "grad_norm": 2.1432546645684694, + "learning_rate": 1.2412129653229355e-07, + "loss": 0.8055, + "step": 40379 + }, + { + "epoch": 0.95, + "grad_norm": 2.742760543963393, + "learning_rate": 1.2400147633385817e-07, + "loss": 0.9631, + "step": 40380 + }, + { + "epoch": 0.95, + "grad_norm": 2.1164627016752857, + "learning_rate": 1.2388171363661216e-07, + "loss": 1.0068, + "step": 40381 + }, + { + "epoch": 0.95, + "grad_norm": 2.1492915788040117, + "learning_rate": 1.2376200844125387e-07, + "loss": 1.0234, + "step": 40382 + }, + { + "epoch": 0.95, + "grad_norm": 2.011220885330599, + "learning_rate": 1.2364236074847934e-07, + "loss": 1.0114, + "step": 40383 + }, + { + "epoch": 0.95, + "grad_norm": 1.9276470240143513, + "learning_rate": 1.2352277055898587e-07, + "loss": 0.9182, + "step": 40384 + }, + { + "epoch": 0.95, + "grad_norm": 1.957066668341744, + "learning_rate": 1.2340323787347064e-07, + "loss": 0.9577, + "step": 40385 + }, + { + "epoch": 0.95, + "grad_norm": 2.0741778603616505, + "learning_rate": 1.2328376269262754e-07, + "loss": 0.9209, + "step": 40386 + }, + { + "epoch": 0.95, + "grad_norm": 2.5437282824790097, + "learning_rate": 1.231643450171527e-07, + "loss": 0.8663, + "step": 40387 + }, + { + "epoch": 0.95, + "grad_norm": 1.899123308708432, + "learning_rate": 1.2304498484774329e-07, + "loss": 0.9884, + "step": 40388 + }, + { + "epoch": 0.95, + "grad_norm": 2.2563855134594464, + "learning_rate": 1.2292568218509325e-07, + "loss": 1.071, + "step": 40389 + }, + { + "epoch": 0.95, + "grad_norm": 2.0271246878795424, + "learning_rate": 1.2280643702989647e-07, + "loss": 0.937, + "step": 40390 + }, + { + "epoch": 0.95, + "grad_norm": 1.9394896684855247, + "learning_rate": 1.2268724938284682e-07, + "loss": 0.9921, + "step": 40391 + }, + { + "epoch": 0.95, + "grad_norm": 1.9444812056267484, + "learning_rate": 1.225681192446404e-07, + "loss": 0.925, + "step": 40392 + }, + { + "epoch": 0.95, + "grad_norm": 1.8391630734388271, + "learning_rate": 1.224490466159689e-07, + "loss": 0.9048, + "step": 40393 + }, + { + "epoch": 0.95, + "grad_norm": 1.9067530251688807, + "learning_rate": 1.223300314975262e-07, + "loss": 1.0502, + "step": 40394 + }, + { + "epoch": 0.95, + "grad_norm": 2.265468232921857, + "learning_rate": 1.2221107389000508e-07, + "loss": 0.9742, + "step": 40395 + }, + { + "epoch": 0.95, + "grad_norm": 1.9474660542845255, + "learning_rate": 1.2209217379409834e-07, + "loss": 0.9909, + "step": 40396 + }, + { + "epoch": 0.95, + "grad_norm": 1.8895183548029495, + "learning_rate": 1.219733312104987e-07, + "loss": 0.8768, + "step": 40397 + }, + { + "epoch": 0.95, + "grad_norm": 2.32321852731703, + "learning_rate": 1.2185454613989678e-07, + "loss": 0.9934, + "step": 40398 + }, + { + "epoch": 0.95, + "grad_norm": 1.092012785818095, + "learning_rate": 1.2173581858298644e-07, + "loss": 0.8934, + "step": 40399 + }, + { + "epoch": 0.95, + "grad_norm": 2.0602025314945296, + "learning_rate": 1.2161714854045713e-07, + "loss": 1.0408, + "step": 40400 + }, + { + "epoch": 0.95, + "grad_norm": 2.098484670452561, + "learning_rate": 1.2149853601299945e-07, + "loss": 1.0837, + "step": 40401 + }, + { + "epoch": 0.95, + "grad_norm": 2.015205780328961, + "learning_rate": 1.21379981001305e-07, + "loss": 1.0761, + "step": 40402 + }, + { + "epoch": 0.95, + "grad_norm": 1.8948559583559421, + "learning_rate": 1.2126148350606438e-07, + "loss": 0.8705, + "step": 40403 + }, + { + "epoch": 0.95, + "grad_norm": 1.9743741306964424, + "learning_rate": 1.2114304352796703e-07, + "loss": 0.8496, + "step": 40404 + }, + { + "epoch": 0.95, + "grad_norm": 1.9491555501944662, + "learning_rate": 1.210246610677024e-07, + "loss": 1.0668, + "step": 40405 + }, + { + "epoch": 0.95, + "grad_norm": 1.921391072746976, + "learning_rate": 1.2090633612595881e-07, + "loss": 0.9643, + "step": 40406 + }, + { + "epoch": 0.95, + "grad_norm": 2.014117780644247, + "learning_rate": 1.2078806870342684e-07, + "loss": 0.905, + "step": 40407 + }, + { + "epoch": 0.95, + "grad_norm": 1.0831674828092392, + "learning_rate": 1.2066985880079484e-07, + "loss": 0.9862, + "step": 40408 + }, + { + "epoch": 0.95, + "grad_norm": 2.15380640166004, + "learning_rate": 1.2055170641875003e-07, + "loss": 0.8599, + "step": 40409 + }, + { + "epoch": 0.95, + "grad_norm": 2.0202216991086597, + "learning_rate": 1.2043361155798184e-07, + "loss": 0.9125, + "step": 40410 + }, + { + "epoch": 0.95, + "grad_norm": 1.8317598882089128, + "learning_rate": 1.2031557421917639e-07, + "loss": 1.024, + "step": 40411 + }, + { + "epoch": 0.95, + "grad_norm": 2.172800753353003, + "learning_rate": 1.2019759440302093e-07, + "loss": 0.9849, + "step": 40412 + }, + { + "epoch": 0.95, + "grad_norm": 2.1763335080122497, + "learning_rate": 1.2007967211020378e-07, + "loss": 1.0105, + "step": 40413 + }, + { + "epoch": 0.95, + "grad_norm": 2.1445992442505952, + "learning_rate": 1.1996180734141104e-07, + "loss": 0.9781, + "step": 40414 + }, + { + "epoch": 0.95, + "grad_norm": 2.036992997635719, + "learning_rate": 1.1984400009732776e-07, + "loss": 1.0872, + "step": 40415 + }, + { + "epoch": 0.95, + "grad_norm": 2.072406801741352, + "learning_rate": 1.1972625037864117e-07, + "loss": 0.8731, + "step": 40416 + }, + { + "epoch": 0.95, + "grad_norm": 2.6270098669454387, + "learning_rate": 1.1960855818603624e-07, + "loss": 0.9202, + "step": 40417 + }, + { + "epoch": 0.95, + "grad_norm": 1.1374723371479933, + "learning_rate": 1.194909235201991e-07, + "loss": 0.9567, + "step": 40418 + }, + { + "epoch": 0.95, + "grad_norm": 2.1290435941384613, + "learning_rate": 1.1937334638181252e-07, + "loss": 0.9216, + "step": 40419 + }, + { + "epoch": 0.95, + "grad_norm": 1.9630255089386632, + "learning_rate": 1.1925582677156268e-07, + "loss": 1.0583, + "step": 40420 + }, + { + "epoch": 0.95, + "grad_norm": 2.003100360613376, + "learning_rate": 1.1913836469013451e-07, + "loss": 1.0321, + "step": 40421 + }, + { + "epoch": 0.95, + "grad_norm": 1.9522857315985953, + "learning_rate": 1.1902096013821085e-07, + "loss": 0.9866, + "step": 40422 + }, + { + "epoch": 0.95, + "grad_norm": 2.4162113795389133, + "learning_rate": 1.1890361311647558e-07, + "loss": 1.0714, + "step": 40423 + }, + { + "epoch": 0.95, + "grad_norm": 2.030341615123906, + "learning_rate": 1.1878632362561149e-07, + "loss": 0.9563, + "step": 40424 + }, + { + "epoch": 0.95, + "grad_norm": 2.5814565970068917, + "learning_rate": 1.1866909166630136e-07, + "loss": 1.0056, + "step": 40425 + }, + { + "epoch": 0.95, + "grad_norm": 1.8883343110651, + "learning_rate": 1.185519172392291e-07, + "loss": 1.1042, + "step": 40426 + }, + { + "epoch": 0.95, + "grad_norm": 1.9311612462973082, + "learning_rate": 1.1843480034507415e-07, + "loss": 1.0911, + "step": 40427 + }, + { + "epoch": 0.95, + "grad_norm": 2.2443513665462764, + "learning_rate": 1.1831774098452153e-07, + "loss": 0.9181, + "step": 40428 + }, + { + "epoch": 0.95, + "grad_norm": 1.989774266903589, + "learning_rate": 1.1820073915825181e-07, + "loss": 1.1016, + "step": 40429 + }, + { + "epoch": 0.95, + "grad_norm": 1.8074371359248669, + "learning_rate": 1.1808379486694444e-07, + "loss": 1.009, + "step": 40430 + }, + { + "epoch": 0.95, + "grad_norm": 2.3740510070993532, + "learning_rate": 1.1796690811128442e-07, + "loss": 0.9426, + "step": 40431 + }, + { + "epoch": 0.95, + "grad_norm": 1.9446054022337964, + "learning_rate": 1.1785007889194788e-07, + "loss": 0.9077, + "step": 40432 + }, + { + "epoch": 0.95, + "grad_norm": 2.4881003172186644, + "learning_rate": 1.1773330720961652e-07, + "loss": 1.0205, + "step": 40433 + }, + { + "epoch": 0.95, + "grad_norm": 2.043179475627525, + "learning_rate": 1.1761659306497086e-07, + "loss": 1.0039, + "step": 40434 + }, + { + "epoch": 0.95, + "grad_norm": 2.092798747987907, + "learning_rate": 1.1749993645869151e-07, + "loss": 0.9955, + "step": 40435 + }, + { + "epoch": 0.95, + "grad_norm": 2.051382383162974, + "learning_rate": 1.1738333739145457e-07, + "loss": 0.9107, + "step": 40436 + }, + { + "epoch": 0.95, + "grad_norm": 1.1437262663730414, + "learning_rate": 1.1726679586394064e-07, + "loss": 0.8614, + "step": 40437 + }, + { + "epoch": 0.95, + "grad_norm": 2.032063423567148, + "learning_rate": 1.1715031187682912e-07, + "loss": 0.8659, + "step": 40438 + }, + { + "epoch": 0.95, + "grad_norm": 1.1051859098387284, + "learning_rate": 1.1703388543079619e-07, + "loss": 0.9516, + "step": 40439 + }, + { + "epoch": 0.95, + "grad_norm": 2.3831350364252994, + "learning_rate": 1.169175165265224e-07, + "loss": 0.9325, + "step": 40440 + }, + { + "epoch": 0.95, + "grad_norm": 1.119441461635296, + "learning_rate": 1.1680120516468274e-07, + "loss": 0.9428, + "step": 40441 + }, + { + "epoch": 0.95, + "grad_norm": 2.0000494493577734, + "learning_rate": 1.1668495134595558e-07, + "loss": 0.8569, + "step": 40442 + }, + { + "epoch": 0.95, + "grad_norm": 2.0264531883695773, + "learning_rate": 1.1656875507101706e-07, + "loss": 0.8751, + "step": 40443 + }, + { + "epoch": 0.95, + "grad_norm": 1.957459189307299, + "learning_rate": 1.164526163405455e-07, + "loss": 1.0355, + "step": 40444 + }, + { + "epoch": 0.95, + "grad_norm": 1.9286332597663067, + "learning_rate": 1.1633653515521593e-07, + "loss": 0.9945, + "step": 40445 + }, + { + "epoch": 0.95, + "grad_norm": 1.1066886412497436, + "learning_rate": 1.1622051151570335e-07, + "loss": 1.0479, + "step": 40446 + }, + { + "epoch": 0.95, + "grad_norm": 1.0841940595395467, + "learning_rate": 1.161045454226839e-07, + "loss": 0.9512, + "step": 40447 + }, + { + "epoch": 0.95, + "grad_norm": 2.0141342125580044, + "learning_rate": 1.1598863687683371e-07, + "loss": 0.9507, + "step": 40448 + }, + { + "epoch": 0.95, + "grad_norm": 2.0737365186336714, + "learning_rate": 1.1587278587882667e-07, + "loss": 1.0404, + "step": 40449 + }, + { + "epoch": 0.95, + "grad_norm": 1.8910058449889426, + "learning_rate": 1.157569924293367e-07, + "loss": 0.8715, + "step": 40450 + }, + { + "epoch": 0.95, + "grad_norm": 1.036948160417245, + "learning_rate": 1.156412565290399e-07, + "loss": 1.0061, + "step": 40451 + }, + { + "epoch": 0.95, + "grad_norm": 1.901484471129146, + "learning_rate": 1.155255781786091e-07, + "loss": 0.935, + "step": 40452 + }, + { + "epoch": 0.95, + "grad_norm": 1.9442042960508985, + "learning_rate": 1.1540995737871708e-07, + "loss": 0.9952, + "step": 40453 + }, + { + "epoch": 0.95, + "grad_norm": 2.0115319311098028, + "learning_rate": 1.1529439413003773e-07, + "loss": 0.8492, + "step": 40454 + }, + { + "epoch": 0.95, + "grad_norm": 1.0980376627931547, + "learning_rate": 1.1517888843324387e-07, + "loss": 0.9073, + "step": 40455 + }, + { + "epoch": 0.95, + "grad_norm": 2.516400908381878, + "learning_rate": 1.1506344028900829e-07, + "loss": 1.002, + "step": 40456 + }, + { + "epoch": 0.95, + "grad_norm": 2.626503579059205, + "learning_rate": 1.1494804969800377e-07, + "loss": 1.0202, + "step": 40457 + }, + { + "epoch": 0.95, + "grad_norm": 1.9406384969833579, + "learning_rate": 1.148327166608998e-07, + "loss": 0.9568, + "step": 40458 + }, + { + "epoch": 0.95, + "grad_norm": 1.8972789026250911, + "learning_rate": 1.1471744117837024e-07, + "loss": 0.9339, + "step": 40459 + }, + { + "epoch": 0.95, + "grad_norm": 1.8497125086157726, + "learning_rate": 1.1460222325108461e-07, + "loss": 1.0328, + "step": 40460 + }, + { + "epoch": 0.95, + "grad_norm": 2.1045642699931912, + "learning_rate": 1.1448706287971567e-07, + "loss": 0.792, + "step": 40461 + }, + { + "epoch": 0.95, + "grad_norm": 2.0217602865152497, + "learning_rate": 1.1437196006493179e-07, + "loss": 1.0302, + "step": 40462 + }, + { + "epoch": 0.95, + "grad_norm": 2.002446693542726, + "learning_rate": 1.1425691480740353e-07, + "loss": 1.0346, + "step": 40463 + }, + { + "epoch": 0.95, + "grad_norm": 2.011209971362134, + "learning_rate": 1.1414192710780259e-07, + "loss": 0.8839, + "step": 40464 + }, + { + "epoch": 0.95, + "grad_norm": 1.9352669563994949, + "learning_rate": 1.140269969667962e-07, + "loss": 0.8559, + "step": 40465 + }, + { + "epoch": 0.95, + "grad_norm": 1.819292732289872, + "learning_rate": 1.1391212438505605e-07, + "loss": 0.8642, + "step": 40466 + }, + { + "epoch": 0.95, + "grad_norm": 2.0979002650480743, + "learning_rate": 1.1379730936324717e-07, + "loss": 1.007, + "step": 40467 + }, + { + "epoch": 0.95, + "grad_norm": 2.816229969739875, + "learning_rate": 1.1368255190204125e-07, + "loss": 0.9354, + "step": 40468 + }, + { + "epoch": 0.95, + "grad_norm": 2.1598655967969056, + "learning_rate": 1.1356785200210552e-07, + "loss": 0.7992, + "step": 40469 + }, + { + "epoch": 0.95, + "grad_norm": 2.120585084549731, + "learning_rate": 1.1345320966410833e-07, + "loss": 0.9606, + "step": 40470 + }, + { + "epoch": 0.95, + "grad_norm": 2.129746384986879, + "learning_rate": 1.1333862488871583e-07, + "loss": 0.9572, + "step": 40471 + }, + { + "epoch": 0.95, + "grad_norm": 1.8426876273667254, + "learning_rate": 1.1322409767659526e-07, + "loss": 0.994, + "step": 40472 + }, + { + "epoch": 0.95, + "grad_norm": 1.0839371130709579, + "learning_rate": 1.1310962802841607e-07, + "loss": 0.9421, + "step": 40473 + }, + { + "epoch": 0.95, + "grad_norm": 2.160726024901191, + "learning_rate": 1.129952159448422e-07, + "loss": 0.9417, + "step": 40474 + }, + { + "epoch": 0.95, + "grad_norm": 2.0610806419009453, + "learning_rate": 1.1288086142653865e-07, + "loss": 0.9314, + "step": 40475 + }, + { + "epoch": 0.95, + "grad_norm": 2.090054963481573, + "learning_rate": 1.1276656447417378e-07, + "loss": 0.9831, + "step": 40476 + }, + { + "epoch": 0.95, + "grad_norm": 2.8958646107375023, + "learning_rate": 1.1265232508841262e-07, + "loss": 0.8795, + "step": 40477 + }, + { + "epoch": 0.95, + "grad_norm": 2.277996727814587, + "learning_rate": 1.1253814326992019e-07, + "loss": 1.0232, + "step": 40478 + }, + { + "epoch": 0.95, + "grad_norm": 2.075974886153597, + "learning_rate": 1.124240190193604e-07, + "loss": 0.9628, + "step": 40479 + }, + { + "epoch": 0.95, + "grad_norm": 1.8382596728298375, + "learning_rate": 1.1230995233739939e-07, + "loss": 1.0785, + "step": 40480 + }, + { + "epoch": 0.95, + "grad_norm": 1.9002463987074205, + "learning_rate": 1.1219594322469885e-07, + "loss": 1.0102, + "step": 40481 + }, + { + "epoch": 0.95, + "grad_norm": 2.05127672817545, + "learning_rate": 1.1208199168192491e-07, + "loss": 0.9258, + "step": 40482 + }, + { + "epoch": 0.95, + "grad_norm": 2.2064402834514256, + "learning_rate": 1.1196809770973926e-07, + "loss": 0.9552, + "step": 40483 + }, + { + "epoch": 0.95, + "grad_norm": 1.9529028819860057, + "learning_rate": 1.1185426130880694e-07, + "loss": 1.025, + "step": 40484 + }, + { + "epoch": 0.95, + "grad_norm": 1.053007991056724, + "learning_rate": 1.1174048247978852e-07, + "loss": 0.8474, + "step": 40485 + }, + { + "epoch": 0.95, + "grad_norm": 2.028080401812829, + "learning_rate": 1.1162676122334793e-07, + "loss": 0.9742, + "step": 40486 + }, + { + "epoch": 0.95, + "grad_norm": 2.826325290308239, + "learning_rate": 1.1151309754014683e-07, + "loss": 1.0235, + "step": 40487 + }, + { + "epoch": 0.95, + "grad_norm": 1.8749478915295195, + "learning_rate": 1.1139949143084804e-07, + "loss": 0.9176, + "step": 40488 + }, + { + "epoch": 0.95, + "grad_norm": 2.060254194617008, + "learning_rate": 1.1128594289611105e-07, + "loss": 1.0346, + "step": 40489 + }, + { + "epoch": 0.95, + "grad_norm": 2.208513715042435, + "learning_rate": 1.1117245193659864e-07, + "loss": 0.9424, + "step": 40490 + }, + { + "epoch": 0.95, + "grad_norm": 2.0600726810244003, + "learning_rate": 1.110590185529703e-07, + "loss": 1.0473, + "step": 40491 + }, + { + "epoch": 0.95, + "grad_norm": 1.1190588022032877, + "learning_rate": 1.109456427458877e-07, + "loss": 0.9063, + "step": 40492 + }, + { + "epoch": 0.95, + "grad_norm": 2.082047758072757, + "learning_rate": 1.1083232451600922e-07, + "loss": 1.0182, + "step": 40493 + }, + { + "epoch": 0.95, + "grad_norm": 2.1596514168419194, + "learning_rate": 1.1071906386399767e-07, + "loss": 1.0577, + "step": 40494 + }, + { + "epoch": 0.95, + "grad_norm": 2.000661793801045, + "learning_rate": 1.1060586079050917e-07, + "loss": 1.0914, + "step": 40495 + }, + { + "epoch": 0.95, + "grad_norm": 1.9455595636363792, + "learning_rate": 1.1049271529620542e-07, + "loss": 1.0108, + "step": 40496 + }, + { + "epoch": 0.95, + "grad_norm": 1.783478365254253, + "learning_rate": 1.1037962738174257e-07, + "loss": 0.9493, + "step": 40497 + }, + { + "epoch": 0.95, + "grad_norm": 1.052248117263512, + "learning_rate": 1.102665970477823e-07, + "loss": 0.9186, + "step": 40498 + }, + { + "epoch": 0.95, + "grad_norm": 1.9839967058846877, + "learning_rate": 1.1015362429497967e-07, + "loss": 0.8824, + "step": 40499 + }, + { + "epoch": 0.95, + "grad_norm": 2.2230663413581513, + "learning_rate": 1.1004070912399411e-07, + "loss": 0.9756, + "step": 40500 + }, + { + "epoch": 0.95, + "grad_norm": 2.0980387079070697, + "learning_rate": 1.099278515354829e-07, + "loss": 0.9578, + "step": 40501 + }, + { + "epoch": 0.95, + "grad_norm": 1.8102151243755118, + "learning_rate": 1.0981505153010218e-07, + "loss": 1.0231, + "step": 40502 + }, + { + "epoch": 0.95, + "grad_norm": 1.8034061600109543, + "learning_rate": 1.0970230910851031e-07, + "loss": 1.0924, + "step": 40503 + }, + { + "epoch": 0.95, + "grad_norm": 1.8900738276743125, + "learning_rate": 1.0958962427136232e-07, + "loss": 1.0275, + "step": 40504 + }, + { + "epoch": 0.95, + "grad_norm": 2.0959030487182053, + "learning_rate": 1.0947699701931436e-07, + "loss": 0.9274, + "step": 40505 + }, + { + "epoch": 0.95, + "grad_norm": 2.176565403144339, + "learning_rate": 1.0936442735302255e-07, + "loss": 0.9014, + "step": 40506 + }, + { + "epoch": 0.95, + "grad_norm": 2.387596156200846, + "learning_rate": 1.0925191527314416e-07, + "loss": 0.9404, + "step": 40507 + }, + { + "epoch": 0.95, + "grad_norm": 2.0769849469866384, + "learning_rate": 1.0913946078033089e-07, + "loss": 1.0457, + "step": 40508 + }, + { + "epoch": 0.95, + "grad_norm": 1.843912446881512, + "learning_rate": 1.0902706387523887e-07, + "loss": 1.141, + "step": 40509 + }, + { + "epoch": 0.95, + "grad_norm": 1.9132954301187914, + "learning_rate": 1.0891472455852315e-07, + "loss": 1.0283, + "step": 40510 + }, + { + "epoch": 0.95, + "grad_norm": 2.3801284447263353, + "learning_rate": 1.0880244283083762e-07, + "loss": 0.9595, + "step": 40511 + }, + { + "epoch": 0.95, + "grad_norm": 2.035399747864195, + "learning_rate": 1.0869021869283625e-07, + "loss": 0.9616, + "step": 40512 + }, + { + "epoch": 0.95, + "grad_norm": 1.854557423275347, + "learning_rate": 1.085780521451707e-07, + "loss": 0.9149, + "step": 40513 + }, + { + "epoch": 0.95, + "grad_norm": 2.1588944255200206, + "learning_rate": 1.0846594318849602e-07, + "loss": 0.8932, + "step": 40514 + }, + { + "epoch": 0.95, + "grad_norm": 2.0989806148080103, + "learning_rate": 1.0835389182346389e-07, + "loss": 0.9971, + "step": 40515 + }, + { + "epoch": 0.95, + "grad_norm": 1.9991362972668105, + "learning_rate": 1.0824189805072827e-07, + "loss": 0.9702, + "step": 40516 + }, + { + "epoch": 0.95, + "grad_norm": 1.182050204621441, + "learning_rate": 1.081299618709386e-07, + "loss": 0.9217, + "step": 40517 + }, + { + "epoch": 0.95, + "grad_norm": 1.9990842239956947, + "learning_rate": 1.0801808328474772e-07, + "loss": 0.9837, + "step": 40518 + }, + { + "epoch": 0.95, + "grad_norm": 1.8553803897196917, + "learning_rate": 1.0790626229280842e-07, + "loss": 0.931, + "step": 40519 + }, + { + "epoch": 0.95, + "grad_norm": 1.9801906735754353, + "learning_rate": 1.077944988957702e-07, + "loss": 0.9984, + "step": 40520 + }, + { + "epoch": 0.95, + "grad_norm": 2.0980124470835833, + "learning_rate": 1.0768279309428365e-07, + "loss": 0.8865, + "step": 40521 + }, + { + "epoch": 0.95, + "grad_norm": 1.802984630106874, + "learning_rate": 1.0757114488900156e-07, + "loss": 0.8907, + "step": 40522 + }, + { + "epoch": 0.95, + "grad_norm": 2.3169208230676372, + "learning_rate": 1.0745955428057009e-07, + "loss": 0.9643, + "step": 40523 + }, + { + "epoch": 0.95, + "grad_norm": 1.097498275929883, + "learning_rate": 1.0734802126964206e-07, + "loss": 0.9969, + "step": 40524 + }, + { + "epoch": 0.95, + "grad_norm": 2.062379345832693, + "learning_rate": 1.0723654585686582e-07, + "loss": 0.9638, + "step": 40525 + }, + { + "epoch": 0.95, + "grad_norm": 1.8730044629249558, + "learning_rate": 1.0712512804288977e-07, + "loss": 0.9755, + "step": 40526 + }, + { + "epoch": 0.95, + "grad_norm": 2.1052195914915517, + "learning_rate": 1.0701376782836337e-07, + "loss": 0.9751, + "step": 40527 + }, + { + "epoch": 0.95, + "grad_norm": 1.9198850126598677, + "learning_rate": 1.0690246521393499e-07, + "loss": 0.9335, + "step": 40528 + }, + { + "epoch": 0.95, + "grad_norm": 1.963707031499108, + "learning_rate": 1.0679122020025301e-07, + "loss": 1.0701, + "step": 40529 + }, + { + "epoch": 0.95, + "grad_norm": 2.0916420892005325, + "learning_rate": 1.0668003278796468e-07, + "loss": 1.0081, + "step": 40530 + }, + { + "epoch": 0.95, + "grad_norm": 2.022577731659589, + "learning_rate": 1.0656890297771727e-07, + "loss": 0.9536, + "step": 40531 + }, + { + "epoch": 0.95, + "grad_norm": 4.22242266639036, + "learning_rate": 1.0645783077015692e-07, + "loss": 0.9488, + "step": 40532 + }, + { + "epoch": 0.95, + "grad_norm": 1.0410474528346176, + "learning_rate": 1.0634681616593312e-07, + "loss": 0.9078, + "step": 40533 + }, + { + "epoch": 0.95, + "grad_norm": 1.1347536420435669, + "learning_rate": 1.0623585916568979e-07, + "loss": 0.9476, + "step": 40534 + }, + { + "epoch": 0.95, + "grad_norm": 2.241454687659684, + "learning_rate": 1.0612495977007197e-07, + "loss": 1.0022, + "step": 40535 + }, + { + "epoch": 0.95, + "grad_norm": 1.9388269782135967, + "learning_rate": 1.0601411797972804e-07, + "loss": 0.9991, + "step": 40536 + }, + { + "epoch": 0.96, + "grad_norm": 2.0771375514831587, + "learning_rate": 1.0590333379530304e-07, + "loss": 0.9637, + "step": 40537 + }, + { + "epoch": 0.96, + "grad_norm": 2.890004460640083, + "learning_rate": 1.0579260721744089e-07, + "loss": 0.9651, + "step": 40538 + }, + { + "epoch": 0.96, + "grad_norm": 2.120587652883894, + "learning_rate": 1.0568193824678663e-07, + "loss": 1.0097, + "step": 40539 + }, + { + "epoch": 0.96, + "grad_norm": 2.235973438721192, + "learning_rate": 1.055713268839853e-07, + "loss": 0.9188, + "step": 40540 + }, + { + "epoch": 0.96, + "grad_norm": 2.413035657816989, + "learning_rate": 1.0546077312967862e-07, + "loss": 0.9792, + "step": 40541 + }, + { + "epoch": 0.96, + "grad_norm": 2.1990235313597113, + "learning_rate": 1.0535027698451272e-07, + "loss": 0.937, + "step": 40542 + }, + { + "epoch": 0.96, + "grad_norm": 3.5141676433208384, + "learning_rate": 1.0523983844913043e-07, + "loss": 0.9202, + "step": 40543 + }, + { + "epoch": 0.96, + "grad_norm": 2.3338945672781026, + "learning_rate": 1.0512945752417347e-07, + "loss": 1.0595, + "step": 40544 + }, + { + "epoch": 0.96, + "grad_norm": 1.9663365995337097, + "learning_rate": 1.0501913421028575e-07, + "loss": 1.0796, + "step": 40545 + }, + { + "epoch": 0.96, + "grad_norm": 2.1031800757773595, + "learning_rate": 1.049088685081101e-07, + "loss": 0.9489, + "step": 40546 + }, + { + "epoch": 0.96, + "grad_norm": 2.865470378302228, + "learning_rate": 1.0479866041828824e-07, + "loss": 0.9478, + "step": 40547 + }, + { + "epoch": 0.96, + "grad_norm": 1.081857431046372, + "learning_rate": 1.0468850994145962e-07, + "loss": 1.0227, + "step": 40548 + }, + { + "epoch": 0.96, + "grad_norm": 2.1579500230114785, + "learning_rate": 1.0457841707826822e-07, + "loss": 0.9571, + "step": 40549 + }, + { + "epoch": 0.96, + "grad_norm": 2.754139589086127, + "learning_rate": 1.0446838182935349e-07, + "loss": 0.9127, + "step": 40550 + }, + { + "epoch": 0.96, + "grad_norm": 1.8541723158654648, + "learning_rate": 1.0435840419535826e-07, + "loss": 0.9488, + "step": 40551 + }, + { + "epoch": 0.96, + "grad_norm": 2.0345691887395683, + "learning_rate": 1.042484841769198e-07, + "loss": 1.031, + "step": 40552 + }, + { + "epoch": 0.96, + "grad_norm": 2.458353552052815, + "learning_rate": 1.0413862177467982e-07, + "loss": 1.0515, + "step": 40553 + }, + { + "epoch": 0.96, + "grad_norm": 1.8294273431049222, + "learning_rate": 1.040288169892778e-07, + "loss": 0.9669, + "step": 40554 + }, + { + "epoch": 0.96, + "grad_norm": 2.107509980865697, + "learning_rate": 1.0391906982135324e-07, + "loss": 0.827, + "step": 40555 + }, + { + "epoch": 0.96, + "grad_norm": 2.163045500254631, + "learning_rate": 1.0380938027154452e-07, + "loss": 1.0211, + "step": 40556 + }, + { + "epoch": 0.96, + "grad_norm": 2.435100737968693, + "learning_rate": 1.0369974834049112e-07, + "loss": 0.9361, + "step": 40557 + }, + { + "epoch": 0.96, + "grad_norm": 1.8771021784517326, + "learning_rate": 1.0359017402883143e-07, + "loss": 0.9391, + "step": 40558 + }, + { + "epoch": 0.96, + "grad_norm": 2.057810274235996, + "learning_rate": 1.0348065733720269e-07, + "loss": 0.8556, + "step": 40559 + }, + { + "epoch": 0.96, + "grad_norm": 2.0613238984853135, + "learning_rate": 1.0337119826624109e-07, + "loss": 0.9674, + "step": 40560 + }, + { + "epoch": 0.96, + "grad_norm": 2.07899553788774, + "learning_rate": 1.0326179681658721e-07, + "loss": 0.9742, + "step": 40561 + }, + { + "epoch": 0.96, + "grad_norm": 1.9781435143966293, + "learning_rate": 1.031524529888761e-07, + "loss": 0.9504, + "step": 40562 + }, + { + "epoch": 0.96, + "grad_norm": 1.9569695171144015, + "learning_rate": 1.0304316678374504e-07, + "loss": 0.9681, + "step": 40563 + }, + { + "epoch": 0.96, + "grad_norm": 2.0491429771454164, + "learning_rate": 1.0293393820182907e-07, + "loss": 0.9631, + "step": 40564 + }, + { + "epoch": 0.96, + "grad_norm": 2.143781781603975, + "learning_rate": 1.0282476724376544e-07, + "loss": 0.9654, + "step": 40565 + }, + { + "epoch": 0.96, + "grad_norm": 1.980293726099942, + "learning_rate": 1.0271565391018922e-07, + "loss": 1.0207, + "step": 40566 + }, + { + "epoch": 0.96, + "grad_norm": 1.8136332733195244, + "learning_rate": 1.0260659820173658e-07, + "loss": 1.0415, + "step": 40567 + }, + { + "epoch": 0.96, + "grad_norm": 1.9958834944482113, + "learning_rate": 1.0249760011904253e-07, + "loss": 0.9041, + "step": 40568 + }, + { + "epoch": 0.96, + "grad_norm": 2.112725758491869, + "learning_rate": 1.0238865966273881e-07, + "loss": 1.0462, + "step": 40569 + }, + { + "epoch": 0.96, + "grad_norm": 2.2196414545612404, + "learning_rate": 1.0227977683346268e-07, + "loss": 1.0137, + "step": 40570 + }, + { + "epoch": 0.96, + "grad_norm": 1.9756340232197753, + "learning_rate": 1.021709516318481e-07, + "loss": 1.0009, + "step": 40571 + }, + { + "epoch": 0.96, + "grad_norm": 2.771805621132846, + "learning_rate": 1.0206218405852786e-07, + "loss": 0.8839, + "step": 40572 + }, + { + "epoch": 0.96, + "grad_norm": 1.9817311675073228, + "learning_rate": 1.019534741141337e-07, + "loss": 1.0326, + "step": 40573 + }, + { + "epoch": 0.96, + "grad_norm": 1.8309169418657871, + "learning_rate": 1.0184482179930066e-07, + "loss": 0.8403, + "step": 40574 + }, + { + "epoch": 0.96, + "grad_norm": 1.7420197542970988, + "learning_rate": 1.0173622711466157e-07, + "loss": 0.8713, + "step": 40575 + }, + { + "epoch": 0.96, + "grad_norm": 1.9881224377962878, + "learning_rate": 1.0162769006084705e-07, + "loss": 0.9793, + "step": 40576 + }, + { + "epoch": 0.96, + "grad_norm": 2.0347738680711016, + "learning_rate": 1.0151921063848991e-07, + "loss": 1.0207, + "step": 40577 + }, + { + "epoch": 0.96, + "grad_norm": 2.176289159584841, + "learning_rate": 1.0141078884822187e-07, + "loss": 0.9615, + "step": 40578 + }, + { + "epoch": 0.96, + "grad_norm": 1.9585662026554607, + "learning_rate": 1.0130242469067464e-07, + "loss": 0.9997, + "step": 40579 + }, + { + "epoch": 0.96, + "grad_norm": 1.8084926922002622, + "learning_rate": 1.0119411816647884e-07, + "loss": 0.9793, + "step": 40580 + }, + { + "epoch": 0.96, + "grad_norm": 1.8280862708350132, + "learning_rate": 1.0108586927626285e-07, + "loss": 0.854, + "step": 40581 + }, + { + "epoch": 0.96, + "grad_norm": 1.5886081847578304, + "learning_rate": 1.009776780206606e-07, + "loss": 0.8316, + "step": 40582 + }, + { + "epoch": 0.96, + "grad_norm": 2.3271669377277147, + "learning_rate": 1.008695444002994e-07, + "loss": 1.0573, + "step": 40583 + }, + { + "epoch": 0.96, + "grad_norm": 2.088266923777143, + "learning_rate": 1.0076146841581091e-07, + "loss": 1.0562, + "step": 40584 + }, + { + "epoch": 0.96, + "grad_norm": 1.8075840911303271, + "learning_rate": 1.0065345006782246e-07, + "loss": 0.9582, + "step": 40585 + }, + { + "epoch": 0.96, + "grad_norm": 1.945267447463143, + "learning_rate": 1.0054548935696351e-07, + "loss": 1.0445, + "step": 40586 + }, + { + "epoch": 0.96, + "grad_norm": 2.2279883686855966, + "learning_rate": 1.0043758628386247e-07, + "loss": 0.9567, + "step": 40587 + }, + { + "epoch": 0.96, + "grad_norm": 2.1653945716035627, + "learning_rate": 1.0032974084914882e-07, + "loss": 0.8624, + "step": 40588 + }, + { + "epoch": 0.96, + "grad_norm": 1.9080713827178337, + "learning_rate": 1.0022195305344984e-07, + "loss": 0.8448, + "step": 40589 + }, + { + "epoch": 0.96, + "grad_norm": 1.1182133122388347, + "learning_rate": 1.0011422289739169e-07, + "loss": 0.9078, + "step": 40590 + }, + { + "epoch": 0.96, + "grad_norm": 1.7849333721027203, + "learning_rate": 1.0000655038160278e-07, + "loss": 1.0066, + "step": 40591 + }, + { + "epoch": 0.96, + "grad_norm": 2.146295709547188, + "learning_rate": 9.989893550671148e-08, + "loss": 0.9879, + "step": 40592 + }, + { + "epoch": 0.96, + "grad_norm": 1.9835254392809687, + "learning_rate": 9.979137827334173e-08, + "loss": 0.9972, + "step": 40593 + }, + { + "epoch": 0.96, + "grad_norm": 1.8964759473387507, + "learning_rate": 9.968387868212192e-08, + "loss": 0.9835, + "step": 40594 + }, + { + "epoch": 0.96, + "grad_norm": 4.222671361928377, + "learning_rate": 9.95764367336749e-08, + "loss": 0.9638, + "step": 40595 + }, + { + "epoch": 0.96, + "grad_norm": 2.2241483315057256, + "learning_rate": 9.946905242863014e-08, + "loss": 0.9177, + "step": 40596 + }, + { + "epoch": 0.96, + "grad_norm": 2.8507647662445446, + "learning_rate": 9.93617257676116e-08, + "loss": 0.8384, + "step": 40597 + }, + { + "epoch": 0.96, + "grad_norm": 1.9491628439832638, + "learning_rate": 9.925445675124212e-08, + "loss": 0.929, + "step": 40598 + }, + { + "epoch": 0.96, + "grad_norm": 1.9158976231264762, + "learning_rate": 9.914724538014787e-08, + "loss": 1.0326, + "step": 40599 + }, + { + "epoch": 0.96, + "grad_norm": 2.31525244832889, + "learning_rate": 9.904009165495276e-08, + "loss": 0.9591, + "step": 40600 + }, + { + "epoch": 0.96, + "grad_norm": 2.0883891395494723, + "learning_rate": 9.893299557628188e-08, + "loss": 0.9381, + "step": 40601 + }, + { + "epoch": 0.96, + "grad_norm": 1.9439381065970978, + "learning_rate": 9.882595714475696e-08, + "loss": 0.9409, + "step": 40602 + }, + { + "epoch": 0.96, + "grad_norm": 1.9417827590580095, + "learning_rate": 9.871897636100191e-08, + "loss": 0.9493, + "step": 40603 + }, + { + "epoch": 0.96, + "grad_norm": 2.0183662707022925, + "learning_rate": 9.86120532256396e-08, + "loss": 0.8778, + "step": 40604 + }, + { + "epoch": 0.96, + "grad_norm": 1.9379594964229803, + "learning_rate": 9.850518773929286e-08, + "loss": 0.9224, + "step": 40605 + }, + { + "epoch": 0.96, + "grad_norm": 2.0404397224989403, + "learning_rate": 9.839837990258338e-08, + "loss": 0.9069, + "step": 40606 + }, + { + "epoch": 0.96, + "grad_norm": 2.0243728951400968, + "learning_rate": 9.829162971613405e-08, + "loss": 0.913, + "step": 40607 + }, + { + "epoch": 0.96, + "grad_norm": 1.9033266600018073, + "learning_rate": 9.818493718056543e-08, + "loss": 1.0134, + "step": 40608 + }, + { + "epoch": 0.96, + "grad_norm": 2.2042059277885366, + "learning_rate": 9.807830229649929e-08, + "loss": 0.9159, + "step": 40609 + }, + { + "epoch": 0.96, + "grad_norm": 2.17778124822969, + "learning_rate": 9.79717250645551e-08, + "loss": 0.9729, + "step": 40610 + }, + { + "epoch": 0.96, + "grad_norm": 2.4384039140259413, + "learning_rate": 9.78652054853546e-08, + "loss": 0.994, + "step": 40611 + }, + { + "epoch": 0.96, + "grad_norm": 1.9358388199895153, + "learning_rate": 9.775874355951842e-08, + "loss": 0.9389, + "step": 40612 + }, + { + "epoch": 0.96, + "grad_norm": 1.9527613211833135, + "learning_rate": 9.765233928766493e-08, + "loss": 1.114, + "step": 40613 + }, + { + "epoch": 0.96, + "grad_norm": 1.9998324714518527, + "learning_rate": 9.754599267041697e-08, + "loss": 0.8699, + "step": 40614 + }, + { + "epoch": 0.96, + "grad_norm": 1.8618443586645732, + "learning_rate": 9.743970370838963e-08, + "loss": 1.0594, + "step": 40615 + }, + { + "epoch": 0.96, + "grad_norm": 2.073612709308162, + "learning_rate": 9.733347240220237e-08, + "loss": 1.0641, + "step": 40616 + }, + { + "epoch": 0.96, + "grad_norm": 2.045798392342015, + "learning_rate": 9.722729875247694e-08, + "loss": 0.9516, + "step": 40617 + }, + { + "epoch": 0.96, + "grad_norm": 1.9883047270653484, + "learning_rate": 9.71211827598284e-08, + "loss": 1.1433, + "step": 40618 + }, + { + "epoch": 0.96, + "grad_norm": 1.9154658096186896, + "learning_rate": 9.701512442487515e-08, + "loss": 0.8848, + "step": 40619 + }, + { + "epoch": 0.96, + "grad_norm": 1.99713909367659, + "learning_rate": 9.690912374823446e-08, + "loss": 1.0173, + "step": 40620 + }, + { + "epoch": 0.96, + "grad_norm": 1.8919766493131986, + "learning_rate": 9.680318073052474e-08, + "loss": 1.0576, + "step": 40621 + }, + { + "epoch": 0.96, + "grad_norm": 2.2092245035285787, + "learning_rate": 9.669729537236327e-08, + "loss": 0.8295, + "step": 40622 + }, + { + "epoch": 0.96, + "grad_norm": 1.1123828183340008, + "learning_rate": 9.6591467674364e-08, + "loss": 0.9617, + "step": 40623 + }, + { + "epoch": 0.96, + "grad_norm": 1.9864027405258633, + "learning_rate": 9.648569763714532e-08, + "loss": 0.8644, + "step": 40624 + }, + { + "epoch": 0.96, + "grad_norm": 3.076487108911738, + "learning_rate": 9.637998526132231e-08, + "loss": 1.0068, + "step": 40625 + }, + { + "epoch": 0.96, + "grad_norm": 1.9378633001110412, + "learning_rate": 9.627433054751001e-08, + "loss": 1.14, + "step": 40626 + }, + { + "epoch": 0.96, + "grad_norm": 2.0735259817552216, + "learning_rate": 9.616873349632461e-08, + "loss": 1.0498, + "step": 40627 + }, + { + "epoch": 0.96, + "grad_norm": 2.058088806306626, + "learning_rate": 9.606319410838006e-08, + "loss": 1.0262, + "step": 40628 + }, + { + "epoch": 0.96, + "grad_norm": 1.7975971931286698, + "learning_rate": 9.595771238429142e-08, + "loss": 0.8999, + "step": 40629 + }, + { + "epoch": 0.96, + "grad_norm": 2.000469767673608, + "learning_rate": 9.585228832467153e-08, + "loss": 0.9131, + "step": 40630 + }, + { + "epoch": 0.96, + "grad_norm": 1.9213115638663185, + "learning_rate": 9.574692193013769e-08, + "loss": 0.9149, + "step": 40631 + }, + { + "epoch": 0.96, + "grad_norm": 1.8499848592406825, + "learning_rate": 9.564161320129828e-08, + "loss": 0.9681, + "step": 40632 + }, + { + "epoch": 0.96, + "grad_norm": 1.0985518719209089, + "learning_rate": 9.553636213876949e-08, + "loss": 0.9549, + "step": 40633 + }, + { + "epoch": 0.96, + "grad_norm": 1.8054311606242421, + "learning_rate": 9.543116874316416e-08, + "loss": 0.9891, + "step": 40634 + }, + { + "epoch": 0.96, + "grad_norm": 2.1722797007740073, + "learning_rate": 9.532603301509514e-08, + "loss": 0.999, + "step": 40635 + }, + { + "epoch": 0.96, + "grad_norm": 1.1003116111982347, + "learning_rate": 9.522095495517303e-08, + "loss": 0.8691, + "step": 40636 + }, + { + "epoch": 0.96, + "grad_norm": 1.8043588787045108, + "learning_rate": 9.511593456400958e-08, + "loss": 0.9195, + "step": 40637 + }, + { + "epoch": 0.96, + "grad_norm": 1.8303256387184654, + "learning_rate": 9.501097184221763e-08, + "loss": 1.043, + "step": 40638 + }, + { + "epoch": 0.96, + "grad_norm": 1.0931003382444808, + "learning_rate": 9.490606679040782e-08, + "loss": 0.8935, + "step": 40639 + }, + { + "epoch": 0.96, + "grad_norm": 1.1145480396063558, + "learning_rate": 9.480121940919074e-08, + "loss": 0.9572, + "step": 40640 + }, + { + "epoch": 0.96, + "grad_norm": 1.0331510470305088, + "learning_rate": 9.469642969917592e-08, + "loss": 0.9123, + "step": 40641 + }, + { + "epoch": 0.96, + "grad_norm": 2.601243693167606, + "learning_rate": 9.45916976609762e-08, + "loss": 1.0067, + "step": 40642 + }, + { + "epoch": 0.96, + "grad_norm": 1.059643032285749, + "learning_rate": 9.448702329519777e-08, + "loss": 0.8805, + "step": 40643 + }, + { + "epoch": 0.96, + "grad_norm": 1.942049421073508, + "learning_rate": 9.438240660245346e-08, + "loss": 1.1021, + "step": 40644 + }, + { + "epoch": 0.96, + "grad_norm": 1.8396274179155263, + "learning_rate": 9.427784758335056e-08, + "loss": 1.0437, + "step": 40645 + }, + { + "epoch": 0.96, + "grad_norm": 2.1437613721059683, + "learning_rate": 9.41733462384975e-08, + "loss": 0.963, + "step": 40646 + }, + { + "epoch": 0.96, + "grad_norm": 2.186770832721179, + "learning_rate": 9.406890256850376e-08, + "loss": 1.0699, + "step": 40647 + }, + { + "epoch": 0.96, + "grad_norm": 2.0329826354062823, + "learning_rate": 9.396451657397666e-08, + "loss": 1.015, + "step": 40648 + }, + { + "epoch": 0.96, + "grad_norm": 2.1032446042035224, + "learning_rate": 9.386018825552568e-08, + "loss": 0.9953, + "step": 40649 + }, + { + "epoch": 0.96, + "grad_norm": 1.0570022344131127, + "learning_rate": 9.37559176137548e-08, + "loss": 0.945, + "step": 40650 + }, + { + "epoch": 0.96, + "grad_norm": 2.5003542669564394, + "learning_rate": 9.365170464927464e-08, + "loss": 0.9765, + "step": 40651 + }, + { + "epoch": 0.96, + "grad_norm": 2.1560691061366266, + "learning_rate": 9.354754936269029e-08, + "loss": 1.1197, + "step": 40652 + }, + { + "epoch": 0.96, + "grad_norm": 2.0347305922852907, + "learning_rate": 9.344345175460901e-08, + "loss": 1.0619, + "step": 40653 + }, + { + "epoch": 0.96, + "grad_norm": 2.0247066067216015, + "learning_rate": 9.333941182563477e-08, + "loss": 0.9641, + "step": 40654 + }, + { + "epoch": 0.96, + "grad_norm": 1.887788808626731, + "learning_rate": 9.323542957637599e-08, + "loss": 0.924, + "step": 40655 + }, + { + "epoch": 0.96, + "grad_norm": 1.9071591398790377, + "learning_rate": 9.313150500743661e-08, + "loss": 0.9529, + "step": 40656 + }, + { + "epoch": 0.96, + "grad_norm": 1.8563395467044528, + "learning_rate": 9.302763811942283e-08, + "loss": 0.9894, + "step": 40657 + }, + { + "epoch": 0.96, + "grad_norm": 2.0090980550608806, + "learning_rate": 9.29238289129375e-08, + "loss": 0.9467, + "step": 40658 + }, + { + "epoch": 0.96, + "grad_norm": 2.178479919449647, + "learning_rate": 9.282007738858679e-08, + "loss": 1.0366, + "step": 40659 + }, + { + "epoch": 0.96, + "grad_norm": 2.4871361312734606, + "learning_rate": 9.271638354697354e-08, + "loss": 0.9283, + "step": 40660 + }, + { + "epoch": 0.96, + "grad_norm": 2.0312121293774146, + "learning_rate": 9.261274738870286e-08, + "loss": 0.9451, + "step": 40661 + }, + { + "epoch": 0.96, + "grad_norm": 2.209646199824237, + "learning_rate": 9.250916891437756e-08, + "loss": 0.9313, + "step": 40662 + }, + { + "epoch": 0.96, + "grad_norm": 1.9981258636347603, + "learning_rate": 9.24056481245994e-08, + "loss": 0.9556, + "step": 40663 + }, + { + "epoch": 0.96, + "grad_norm": 1.7738838866799564, + "learning_rate": 9.230218501997346e-08, + "loss": 0.8707, + "step": 40664 + }, + { + "epoch": 0.96, + "grad_norm": 2.2051941925144507, + "learning_rate": 9.219877960110036e-08, + "loss": 0.8963, + "step": 40665 + }, + { + "epoch": 0.96, + "grad_norm": 2.007588751641904, + "learning_rate": 9.209543186858294e-08, + "loss": 0.8459, + "step": 40666 + }, + { + "epoch": 0.96, + "grad_norm": 2.0481088600709203, + "learning_rate": 9.199214182302186e-08, + "loss": 0.8488, + "step": 40667 + }, + { + "epoch": 0.96, + "grad_norm": 1.7728626724006211, + "learning_rate": 9.188890946501993e-08, + "loss": 0.9836, + "step": 40668 + }, + { + "epoch": 0.96, + "grad_norm": 1.983747602294111, + "learning_rate": 9.178573479517783e-08, + "loss": 0.9355, + "step": 40669 + }, + { + "epoch": 0.96, + "grad_norm": 2.342366210802647, + "learning_rate": 9.168261781409726e-08, + "loss": 1.0072, + "step": 40670 + }, + { + "epoch": 0.96, + "grad_norm": 1.9106842363961596, + "learning_rate": 9.157955852237554e-08, + "loss": 0.9499, + "step": 40671 + }, + { + "epoch": 0.96, + "grad_norm": 2.0490838408100602, + "learning_rate": 9.147655692061442e-08, + "loss": 1.0613, + "step": 40672 + }, + { + "epoch": 0.96, + "grad_norm": 1.9925197286897494, + "learning_rate": 9.13736130094145e-08, + "loss": 1.0061, + "step": 40673 + }, + { + "epoch": 0.96, + "grad_norm": 2.2203244106119566, + "learning_rate": 9.127072678937532e-08, + "loss": 0.8329, + "step": 40674 + }, + { + "epoch": 0.96, + "grad_norm": 1.9573796727341948, + "learning_rate": 9.116789826109417e-08, + "loss": 1.022, + "step": 40675 + }, + { + "epoch": 0.96, + "grad_norm": 2.169935455427842, + "learning_rate": 9.106512742516948e-08, + "loss": 1.0175, + "step": 40676 + }, + { + "epoch": 0.96, + "grad_norm": 1.120227485687165, + "learning_rate": 9.096241428220299e-08, + "loss": 0.9691, + "step": 40677 + }, + { + "epoch": 0.96, + "grad_norm": 2.0978571681903198, + "learning_rate": 9.085975883278864e-08, + "loss": 0.952, + "step": 40678 + }, + { + "epoch": 0.96, + "grad_norm": 1.8350492155970142, + "learning_rate": 9.075716107752708e-08, + "loss": 0.8732, + "step": 40679 + }, + { + "epoch": 0.96, + "grad_norm": 1.1039336814672935, + "learning_rate": 9.065462101701339e-08, + "loss": 0.9591, + "step": 40680 + }, + { + "epoch": 0.96, + "grad_norm": 1.9170363331126374, + "learning_rate": 9.055213865184708e-08, + "loss": 0.9449, + "step": 40681 + }, + { + "epoch": 0.96, + "grad_norm": 2.0810352743081086, + "learning_rate": 9.044971398262325e-08, + "loss": 0.8846, + "step": 40682 + }, + { + "epoch": 0.96, + "grad_norm": 1.9421296548064613, + "learning_rate": 9.034734700993696e-08, + "loss": 0.9348, + "step": 40683 + }, + { + "epoch": 0.96, + "grad_norm": 1.8404831566781314, + "learning_rate": 9.024503773438776e-08, + "loss": 0.981, + "step": 40684 + }, + { + "epoch": 0.96, + "grad_norm": 2.08965433077824, + "learning_rate": 9.014278615656735e-08, + "loss": 0.9778, + "step": 40685 + }, + { + "epoch": 0.96, + "grad_norm": 2.0748544601557857, + "learning_rate": 9.004059227707528e-08, + "loss": 0.894, + "step": 40686 + }, + { + "epoch": 0.96, + "grad_norm": 2.199872286005272, + "learning_rate": 8.99384560965022e-08, + "loss": 1.0536, + "step": 40687 + }, + { + "epoch": 0.96, + "grad_norm": 2.069904230681139, + "learning_rate": 8.98363776154465e-08, + "loss": 0.9034, + "step": 40688 + }, + { + "epoch": 0.96, + "grad_norm": 2.003236343435777, + "learning_rate": 8.973435683449994e-08, + "loss": 1.0679, + "step": 40689 + }, + { + "epoch": 0.96, + "grad_norm": 1.9731169875463792, + "learning_rate": 8.96323937542587e-08, + "loss": 0.9774, + "step": 40690 + }, + { + "epoch": 0.96, + "grad_norm": 2.167546019655374, + "learning_rate": 8.953048837531453e-08, + "loss": 0.8492, + "step": 40691 + }, + { + "epoch": 0.96, + "grad_norm": 1.9011211981273894, + "learning_rate": 8.94286406982603e-08, + "loss": 0.9082, + "step": 40692 + }, + { + "epoch": 0.96, + "grad_norm": 1.9649345955693456, + "learning_rate": 8.932685072369107e-08, + "loss": 1.0495, + "step": 40693 + }, + { + "epoch": 0.96, + "grad_norm": 2.001894753116855, + "learning_rate": 8.922511845219972e-08, + "loss": 0.9221, + "step": 40694 + }, + { + "epoch": 0.96, + "grad_norm": 1.7915380706762989, + "learning_rate": 8.912344388437577e-08, + "loss": 0.9462, + "step": 40695 + }, + { + "epoch": 0.96, + "grad_norm": 3.0414884538238076, + "learning_rate": 8.902182702081318e-08, + "loss": 0.9656, + "step": 40696 + }, + { + "epoch": 0.96, + "grad_norm": 4.353939634415366, + "learning_rate": 8.892026786210372e-08, + "loss": 0.8555, + "step": 40697 + }, + { + "epoch": 0.96, + "grad_norm": 2.28831238609059, + "learning_rate": 8.881876640883802e-08, + "loss": 0.9705, + "step": 40698 + }, + { + "epoch": 0.96, + "grad_norm": 2.2656700021902982, + "learning_rate": 8.871732266160782e-08, + "loss": 0.9978, + "step": 40699 + }, + { + "epoch": 0.96, + "grad_norm": 1.8854378851559164, + "learning_rate": 8.861593662100265e-08, + "loss": 1.0048, + "step": 40700 + }, + { + "epoch": 0.96, + "grad_norm": 3.672256825446887, + "learning_rate": 8.851460828761316e-08, + "loss": 0.8981, + "step": 40701 + }, + { + "epoch": 0.96, + "grad_norm": 2.0917837425695907, + "learning_rate": 8.841333766202997e-08, + "loss": 0.9649, + "step": 40702 + }, + { + "epoch": 0.96, + "grad_norm": 2.158917370691327, + "learning_rate": 8.831212474484263e-08, + "loss": 0.785, + "step": 40703 + }, + { + "epoch": 0.96, + "grad_norm": 1.9309650175963233, + "learning_rate": 8.821096953663955e-08, + "loss": 1.0247, + "step": 40704 + }, + { + "epoch": 0.96, + "grad_norm": 1.9685239668028338, + "learning_rate": 8.810987203801136e-08, + "loss": 0.9592, + "step": 40705 + }, + { + "epoch": 0.96, + "grad_norm": 1.9390686894017555, + "learning_rate": 8.800883224954315e-08, + "loss": 0.9671, + "step": 40706 + }, + { + "epoch": 0.96, + "grad_norm": 2.140968514901833, + "learning_rate": 8.790785017182779e-08, + "loss": 0.9831, + "step": 40707 + }, + { + "epoch": 0.96, + "grad_norm": 1.8238607685620123, + "learning_rate": 8.780692580545036e-08, + "loss": 0.9132, + "step": 40708 + }, + { + "epoch": 0.96, + "grad_norm": 1.9263239148178224, + "learning_rate": 8.770605915099927e-08, + "loss": 1.0059, + "step": 40709 + }, + { + "epoch": 0.96, + "grad_norm": 1.085369065212935, + "learning_rate": 8.760525020906074e-08, + "loss": 0.9517, + "step": 40710 + }, + { + "epoch": 0.96, + "grad_norm": 1.8308316816587054, + "learning_rate": 8.750449898022317e-08, + "loss": 0.9136, + "step": 40711 + }, + { + "epoch": 0.96, + "grad_norm": 1.7741353619008524, + "learning_rate": 8.740380546507388e-08, + "loss": 0.9433, + "step": 40712 + }, + { + "epoch": 0.96, + "grad_norm": 1.977718730405882, + "learning_rate": 8.730316966419682e-08, + "loss": 1.0528, + "step": 40713 + }, + { + "epoch": 0.96, + "grad_norm": 2.033152559248917, + "learning_rate": 8.720259157817823e-08, + "loss": 1.0234, + "step": 40714 + }, + { + "epoch": 0.96, + "grad_norm": 2.1801650173053964, + "learning_rate": 8.71020712076054e-08, + "loss": 0.9728, + "step": 40715 + }, + { + "epoch": 0.96, + "grad_norm": 2.0898336315578536, + "learning_rate": 8.70016085530634e-08, + "loss": 1.0294, + "step": 40716 + }, + { + "epoch": 0.96, + "grad_norm": 2.0996974857827104, + "learning_rate": 8.69012036151351e-08, + "loss": 1.0714, + "step": 40717 + }, + { + "epoch": 0.96, + "grad_norm": 1.7768951201541967, + "learning_rate": 8.680085639440672e-08, + "loss": 1.0449, + "step": 40718 + }, + { + "epoch": 0.96, + "grad_norm": 2.0204000418929495, + "learning_rate": 8.670056689146334e-08, + "loss": 0.9867, + "step": 40719 + }, + { + "epoch": 0.96, + "grad_norm": 2.354810343706817, + "learning_rate": 8.66003351068867e-08, + "loss": 0.9748, + "step": 40720 + }, + { + "epoch": 0.96, + "grad_norm": 1.9302849072884956, + "learning_rate": 8.650016104126192e-08, + "loss": 0.8462, + "step": 40721 + }, + { + "epoch": 0.96, + "grad_norm": 2.659108183554706, + "learning_rate": 8.640004469517072e-08, + "loss": 1.0076, + "step": 40722 + }, + { + "epoch": 0.96, + "grad_norm": 2.553856499108368, + "learning_rate": 8.62999860691982e-08, + "loss": 1.0353, + "step": 40723 + }, + { + "epoch": 0.96, + "grad_norm": 1.9263012584278114, + "learning_rate": 8.619998516392503e-08, + "loss": 1.0633, + "step": 40724 + }, + { + "epoch": 0.96, + "grad_norm": 2.1254151056971193, + "learning_rate": 8.610004197993516e-08, + "loss": 0.8625, + "step": 40725 + }, + { + "epoch": 0.96, + "grad_norm": 2.024946577875892, + "learning_rate": 8.600015651780924e-08, + "loss": 1.1142, + "step": 40726 + }, + { + "epoch": 0.96, + "grad_norm": 1.871755372376874, + "learning_rate": 8.590032877812792e-08, + "loss": 0.9724, + "step": 40727 + }, + { + "epoch": 0.96, + "grad_norm": 2.0025386822719766, + "learning_rate": 8.580055876147408e-08, + "loss": 0.9364, + "step": 40728 + }, + { + "epoch": 0.96, + "grad_norm": 1.9970581752669432, + "learning_rate": 8.570084646842946e-08, + "loss": 1.0175, + "step": 40729 + }, + { + "epoch": 0.96, + "grad_norm": 1.9726341838239863, + "learning_rate": 8.56011918995725e-08, + "loss": 1.043, + "step": 40730 + }, + { + "epoch": 0.96, + "grad_norm": 2.4208427975236244, + "learning_rate": 8.550159505548383e-08, + "loss": 0.9627, + "step": 40731 + }, + { + "epoch": 0.96, + "grad_norm": 1.9373227048421238, + "learning_rate": 8.540205593674411e-08, + "loss": 0.9244, + "step": 40732 + }, + { + "epoch": 0.96, + "grad_norm": 2.010372610875147, + "learning_rate": 8.530257454393287e-08, + "loss": 0.9815, + "step": 40733 + }, + { + "epoch": 0.96, + "grad_norm": 1.992712285481648, + "learning_rate": 8.520315087762964e-08, + "loss": 0.9221, + "step": 40734 + }, + { + "epoch": 0.96, + "grad_norm": 1.859247877417034, + "learning_rate": 8.510378493841065e-08, + "loss": 0.924, + "step": 40735 + }, + { + "epoch": 0.96, + "grad_norm": 1.9701223580965814, + "learning_rate": 8.500447672685764e-08, + "loss": 1.0598, + "step": 40736 + }, + { + "epoch": 0.96, + "grad_norm": 2.0064218376967635, + "learning_rate": 8.490522624354792e-08, + "loss": 1.0526, + "step": 40737 + }, + { + "epoch": 0.96, + "grad_norm": 1.9960984312919883, + "learning_rate": 8.48060334890588e-08, + "loss": 0.9537, + "step": 40738 + }, + { + "epoch": 0.96, + "grad_norm": 3.012175734445953, + "learning_rate": 8.470689846396762e-08, + "loss": 0.9746, + "step": 40739 + }, + { + "epoch": 0.96, + "grad_norm": 2.15756009186377, + "learning_rate": 8.460782116885169e-08, + "loss": 1.0288, + "step": 40740 + }, + { + "epoch": 0.96, + "grad_norm": 2.0845065125302598, + "learning_rate": 8.450880160428831e-08, + "loss": 0.9141, + "step": 40741 + }, + { + "epoch": 0.96, + "grad_norm": 2.168577422472333, + "learning_rate": 8.440983977085481e-08, + "loss": 0.997, + "step": 40742 + }, + { + "epoch": 0.96, + "grad_norm": 1.1024635730879913, + "learning_rate": 8.43109356691263e-08, + "loss": 0.8964, + "step": 40743 + }, + { + "epoch": 0.96, + "grad_norm": 1.948798705569218, + "learning_rate": 8.421208929967784e-08, + "loss": 0.8943, + "step": 40744 + }, + { + "epoch": 0.96, + "grad_norm": 2.710582924637262, + "learning_rate": 8.411330066308676e-08, + "loss": 1.0126, + "step": 40745 + }, + { + "epoch": 0.96, + "grad_norm": 2.144767339579496, + "learning_rate": 8.401456975992706e-08, + "loss": 1.0646, + "step": 40746 + }, + { + "epoch": 0.96, + "grad_norm": 2.064486491996956, + "learning_rate": 8.391589659077381e-08, + "loss": 0.9494, + "step": 40747 + }, + { + "epoch": 0.96, + "grad_norm": 2.0367876367512, + "learning_rate": 8.381728115620213e-08, + "loss": 0.9553, + "step": 40748 + }, + { + "epoch": 0.96, + "grad_norm": 1.8986962884885097, + "learning_rate": 8.371872345678378e-08, + "loss": 1.0075, + "step": 40749 + }, + { + "epoch": 0.96, + "grad_norm": 1.0312910271605973, + "learning_rate": 8.362022349309718e-08, + "loss": 1.0084, + "step": 40750 + }, + { + "epoch": 0.96, + "grad_norm": 6.711476451069157, + "learning_rate": 8.352178126571186e-08, + "loss": 0.9436, + "step": 40751 + }, + { + "epoch": 0.96, + "grad_norm": 1.816966506477596, + "learning_rate": 8.342339677520184e-08, + "loss": 0.8952, + "step": 40752 + }, + { + "epoch": 0.96, + "grad_norm": 1.0771181942298629, + "learning_rate": 8.332507002213996e-08, + "loss": 0.9748, + "step": 40753 + }, + { + "epoch": 0.96, + "grad_norm": 2.0430401836642673, + "learning_rate": 8.322680100710023e-08, + "loss": 0.9969, + "step": 40754 + }, + { + "epoch": 0.96, + "grad_norm": 2.014361892423089, + "learning_rate": 8.31285897306533e-08, + "loss": 0.9239, + "step": 40755 + }, + { + "epoch": 0.96, + "grad_norm": 1.1031200228363027, + "learning_rate": 8.303043619337093e-08, + "loss": 0.9496, + "step": 40756 + }, + { + "epoch": 0.96, + "grad_norm": 1.0912712581486288, + "learning_rate": 8.293234039582487e-08, + "loss": 0.8225, + "step": 40757 + }, + { + "epoch": 0.96, + "grad_norm": 1.8171080972138371, + "learning_rate": 8.283430233858691e-08, + "loss": 1.0115, + "step": 40758 + }, + { + "epoch": 0.96, + "grad_norm": 1.9602385955293955, + "learning_rate": 8.27363220222277e-08, + "loss": 0.9307, + "step": 40759 + }, + { + "epoch": 0.96, + "grad_norm": 1.8947659191076234, + "learning_rate": 8.263839944731677e-08, + "loss": 0.9706, + "step": 40760 + }, + { + "epoch": 0.96, + "grad_norm": 2.1794553380128274, + "learning_rate": 8.254053461442591e-08, + "loss": 0.9855, + "step": 40761 + }, + { + "epoch": 0.96, + "grad_norm": 2.61720663497819, + "learning_rate": 8.244272752412241e-08, + "loss": 0.9521, + "step": 40762 + }, + { + "epoch": 0.96, + "grad_norm": 2.032383902360161, + "learning_rate": 8.234497817697806e-08, + "loss": 0.9797, + "step": 40763 + }, + { + "epoch": 0.96, + "grad_norm": 2.096464487919786, + "learning_rate": 8.224728657356129e-08, + "loss": 0.943, + "step": 40764 + }, + { + "epoch": 0.96, + "grad_norm": 1.9939118343620452, + "learning_rate": 8.214965271444053e-08, + "loss": 0.8763, + "step": 40765 + }, + { + "epoch": 0.96, + "grad_norm": 1.870061355073612, + "learning_rate": 8.205207660018311e-08, + "loss": 1.1203, + "step": 40766 + }, + { + "epoch": 0.96, + "grad_norm": 1.804127966644318, + "learning_rate": 8.195455823136078e-08, + "loss": 0.8794, + "step": 40767 + }, + { + "epoch": 0.96, + "grad_norm": 2.0266720026053537, + "learning_rate": 8.185709760853754e-08, + "loss": 0.8892, + "step": 40768 + }, + { + "epoch": 0.96, + "grad_norm": 2.0732769064797028, + "learning_rate": 8.175969473228296e-08, + "loss": 0.9379, + "step": 40769 + }, + { + "epoch": 0.96, + "grad_norm": 1.087534935472474, + "learning_rate": 8.16623496031621e-08, + "loss": 0.9691, + "step": 40770 + }, + { + "epoch": 0.96, + "grad_norm": 2.406087999959407, + "learning_rate": 8.156506222174455e-08, + "loss": 0.9324, + "step": 40771 + }, + { + "epoch": 0.96, + "grad_norm": 1.9431576465972769, + "learning_rate": 8.146783258859536e-08, + "loss": 0.8845, + "step": 40772 + }, + { + "epoch": 0.96, + "grad_norm": 1.9284315221930992, + "learning_rate": 8.137066070427968e-08, + "loss": 0.9104, + "step": 40773 + }, + { + "epoch": 0.96, + "grad_norm": 2.0043144112486893, + "learning_rate": 8.127354656936481e-08, + "loss": 1.013, + "step": 40774 + }, + { + "epoch": 0.96, + "grad_norm": 1.9258771946453912, + "learning_rate": 8.117649018441587e-08, + "loss": 1.048, + "step": 40775 + }, + { + "epoch": 0.96, + "grad_norm": 3.368715071201271, + "learning_rate": 8.107949154999684e-08, + "loss": 1.0026, + "step": 40776 + }, + { + "epoch": 0.96, + "grad_norm": 1.9089259139960895, + "learning_rate": 8.098255066667393e-08, + "loss": 0.8555, + "step": 40777 + }, + { + "epoch": 0.96, + "grad_norm": 2.5919391377705, + "learning_rate": 8.088566753501004e-08, + "loss": 0.8554, + "step": 40778 + }, + { + "epoch": 0.96, + "grad_norm": 2.349303511223315, + "learning_rate": 8.078884215557137e-08, + "loss": 0.9549, + "step": 40779 + }, + { + "epoch": 0.96, + "grad_norm": 2.1190045919573075, + "learning_rate": 8.06920745289197e-08, + "loss": 1.0379, + "step": 40780 + }, + { + "epoch": 0.96, + "grad_norm": 1.995864163733561, + "learning_rate": 8.059536465561901e-08, + "loss": 0.9587, + "step": 40781 + }, + { + "epoch": 0.96, + "grad_norm": 2.1893399337575743, + "learning_rate": 8.049871253623332e-08, + "loss": 0.9549, + "step": 40782 + }, + { + "epoch": 0.96, + "grad_norm": 2.764496385512239, + "learning_rate": 8.040211817132438e-08, + "loss": 1.0164, + "step": 40783 + }, + { + "epoch": 0.96, + "grad_norm": 2.0936545664518693, + "learning_rate": 8.030558156145507e-08, + "loss": 0.9377, + "step": 40784 + }, + { + "epoch": 0.96, + "grad_norm": 1.911343856478203, + "learning_rate": 8.020910270718607e-08, + "loss": 0.8142, + "step": 40785 + }, + { + "epoch": 0.96, + "grad_norm": 2.002454489172442, + "learning_rate": 8.011268160908137e-08, + "loss": 1.0988, + "step": 40786 + }, + { + "epoch": 0.96, + "grad_norm": 2.068446693792599, + "learning_rate": 8.00163182676994e-08, + "loss": 0.9555, + "step": 40787 + }, + { + "epoch": 0.96, + "grad_norm": 1.989877702245405, + "learning_rate": 7.992001268360528e-08, + "loss": 1.0294, + "step": 40788 + }, + { + "epoch": 0.96, + "grad_norm": 1.9857639644590332, + "learning_rate": 7.982376485735632e-08, + "loss": 1.0137, + "step": 40789 + }, + { + "epoch": 0.96, + "grad_norm": 2.2652934498763106, + "learning_rate": 7.972757478951432e-08, + "loss": 0.9504, + "step": 40790 + }, + { + "epoch": 0.96, + "grad_norm": 2.0071335927976937, + "learning_rate": 7.963144248063881e-08, + "loss": 0.8767, + "step": 40791 + }, + { + "epoch": 0.96, + "grad_norm": 1.9249555657702422, + "learning_rate": 7.953536793129046e-08, + "loss": 0.9149, + "step": 40792 + }, + { + "epoch": 0.96, + "grad_norm": 1.8300238862595468, + "learning_rate": 7.943935114202772e-08, + "loss": 0.8757, + "step": 40793 + }, + { + "epoch": 0.96, + "grad_norm": 2.025130839744415, + "learning_rate": 7.934339211340902e-08, + "loss": 1.1436, + "step": 40794 + }, + { + "epoch": 0.96, + "grad_norm": 1.7926351703379302, + "learning_rate": 7.924749084599393e-08, + "loss": 0.8606, + "step": 40795 + }, + { + "epoch": 0.96, + "grad_norm": 1.8149553481349812, + "learning_rate": 7.915164734034198e-08, + "loss": 1.042, + "step": 40796 + }, + { + "epoch": 0.96, + "grad_norm": 1.120030342712261, + "learning_rate": 7.90558615970094e-08, + "loss": 0.8945, + "step": 40797 + }, + { + "epoch": 0.96, + "grad_norm": 1.7599924516348773, + "learning_rate": 7.896013361655464e-08, + "loss": 0.8956, + "step": 40798 + }, + { + "epoch": 0.96, + "grad_norm": 1.9924820343308332, + "learning_rate": 7.886446339953502e-08, + "loss": 1.0809, + "step": 40799 + }, + { + "epoch": 0.96, + "grad_norm": 2.0191348498355555, + "learning_rate": 7.876885094650677e-08, + "loss": 0.9769, + "step": 40800 + }, + { + "epoch": 0.96, + "grad_norm": 1.9952020672599178, + "learning_rate": 7.867329625802833e-08, + "loss": 1.1541, + "step": 40801 + }, + { + "epoch": 0.96, + "grad_norm": 2.0192511971238605, + "learning_rate": 7.857779933465482e-08, + "loss": 1.035, + "step": 40802 + }, + { + "epoch": 0.96, + "grad_norm": 1.982676545472045, + "learning_rate": 7.848236017694133e-08, + "loss": 0.8885, + "step": 40803 + }, + { + "epoch": 0.96, + "grad_norm": 2.0698354817052733, + "learning_rate": 7.838697878544632e-08, + "loss": 0.955, + "step": 40804 + }, + { + "epoch": 0.96, + "grad_norm": 2.3237348842238226, + "learning_rate": 7.829165516072268e-08, + "loss": 0.9416, + "step": 40805 + }, + { + "epoch": 0.96, + "grad_norm": 1.9353082399204558, + "learning_rate": 7.819638930332552e-08, + "loss": 0.9366, + "step": 40806 + }, + { + "epoch": 0.96, + "grad_norm": 1.1967570216765882, + "learning_rate": 7.810118121381105e-08, + "loss": 0.9505, + "step": 40807 + }, + { + "epoch": 0.96, + "grad_norm": 2.0823904238948194, + "learning_rate": 7.800603089273218e-08, + "loss": 1.0083, + "step": 40808 + }, + { + "epoch": 0.96, + "grad_norm": 2.5840315815375616, + "learning_rate": 7.791093834064289e-08, + "loss": 0.9042, + "step": 40809 + }, + { + "epoch": 0.96, + "grad_norm": 1.0431484688620234, + "learning_rate": 7.781590355809831e-08, + "loss": 0.8815, + "step": 40810 + }, + { + "epoch": 0.96, + "grad_norm": 1.957659393332457, + "learning_rate": 7.772092654565133e-08, + "loss": 0.9792, + "step": 40811 + }, + { + "epoch": 0.96, + "grad_norm": 2.267618652613258, + "learning_rate": 7.762600730385373e-08, + "loss": 1.0507, + "step": 40812 + }, + { + "epoch": 0.96, + "grad_norm": 2.1946638981204796, + "learning_rate": 7.753114583325838e-08, + "loss": 1.125, + "step": 40813 + }, + { + "epoch": 0.96, + "grad_norm": 2.1258074980586983, + "learning_rate": 7.74363421344193e-08, + "loss": 0.9194, + "step": 40814 + }, + { + "epoch": 0.96, + "grad_norm": 1.9193632066368875, + "learning_rate": 7.734159620788717e-08, + "loss": 0.8252, + "step": 40815 + }, + { + "epoch": 0.96, + "grad_norm": 2.363192009915782, + "learning_rate": 7.724690805421265e-08, + "loss": 0.9871, + "step": 40816 + }, + { + "epoch": 0.96, + "grad_norm": 2.0268032301123915, + "learning_rate": 7.715227767394862e-08, + "loss": 1.0969, + "step": 40817 + }, + { + "epoch": 0.96, + "grad_norm": 1.822988871679925, + "learning_rate": 7.705770506764576e-08, + "loss": 0.8866, + "step": 40818 + }, + { + "epoch": 0.96, + "grad_norm": 2.1458569090952953, + "learning_rate": 7.696319023585475e-08, + "loss": 1.1268, + "step": 40819 + }, + { + "epoch": 0.96, + "grad_norm": 2.344138136454601, + "learning_rate": 7.686873317912514e-08, + "loss": 0.9275, + "step": 40820 + }, + { + "epoch": 0.96, + "grad_norm": 2.224346468128991, + "learning_rate": 7.677433389800759e-08, + "loss": 1.1288, + "step": 40821 + }, + { + "epoch": 0.96, + "grad_norm": 1.9988444726432764, + "learning_rate": 7.667999239305057e-08, + "loss": 0.9025, + "step": 40822 + }, + { + "epoch": 0.96, + "grad_norm": 1.9399136324770732, + "learning_rate": 7.658570866480586e-08, + "loss": 0.9986, + "step": 40823 + }, + { + "epoch": 0.96, + "grad_norm": 1.84276920837661, + "learning_rate": 7.649148271381967e-08, + "loss": 1.0177, + "step": 40824 + }, + { + "epoch": 0.96, + "grad_norm": 2.262621356667597, + "learning_rate": 7.63973145406427e-08, + "loss": 0.8759, + "step": 40825 + }, + { + "epoch": 0.96, + "grad_norm": 2.2798585507141085, + "learning_rate": 7.630320414582115e-08, + "loss": 0.8885, + "step": 40826 + }, + { + "epoch": 0.96, + "grad_norm": 1.9575193405270215, + "learning_rate": 7.620915152990571e-08, + "loss": 0.9417, + "step": 40827 + }, + { + "epoch": 0.96, + "grad_norm": 1.963212276077104, + "learning_rate": 7.61151566934415e-08, + "loss": 1.0896, + "step": 40828 + }, + { + "epoch": 0.96, + "grad_norm": 2.0140384337895845, + "learning_rate": 7.602121963697695e-08, + "loss": 1.0573, + "step": 40829 + }, + { + "epoch": 0.96, + "grad_norm": 1.7943966962303575, + "learning_rate": 7.592734036105831e-08, + "loss": 0.9566, + "step": 40830 + }, + { + "epoch": 0.96, + "grad_norm": 1.9265593908974, + "learning_rate": 7.583351886623291e-08, + "loss": 0.8677, + "step": 40831 + }, + { + "epoch": 0.96, + "grad_norm": 1.9726777751852969, + "learning_rate": 7.5739755153047e-08, + "loss": 1.0393, + "step": 40832 + }, + { + "epoch": 0.96, + "grad_norm": 1.8987128124716932, + "learning_rate": 7.564604922204566e-08, + "loss": 0.9434, + "step": 40833 + }, + { + "epoch": 0.96, + "grad_norm": 2.1027759722172656, + "learning_rate": 7.555240107377404e-08, + "loss": 0.8385, + "step": 40834 + }, + { + "epoch": 0.96, + "grad_norm": 1.8591587395809859, + "learning_rate": 7.545881070878059e-08, + "loss": 1.0205, + "step": 40835 + }, + { + "epoch": 0.96, + "grad_norm": 2.18201376704469, + "learning_rate": 7.536527812760708e-08, + "loss": 0.8329, + "step": 40836 + }, + { + "epoch": 0.96, + "grad_norm": 2.01628367479329, + "learning_rate": 7.527180333079865e-08, + "loss": 1.0619, + "step": 40837 + }, + { + "epoch": 0.96, + "grad_norm": 2.4106108813982567, + "learning_rate": 7.517838631889929e-08, + "loss": 1.1343, + "step": 40838 + }, + { + "epoch": 0.96, + "grad_norm": 1.9399529421223547, + "learning_rate": 7.508502709245524e-08, + "loss": 0.9488, + "step": 40839 + }, + { + "epoch": 0.96, + "grad_norm": 2.008991698683411, + "learning_rate": 7.499172565200718e-08, + "loss": 1.0196, + "step": 40840 + }, + { + "epoch": 0.96, + "grad_norm": 2.289358556603623, + "learning_rate": 7.489848199809913e-08, + "loss": 0.9993, + "step": 40841 + }, + { + "epoch": 0.96, + "grad_norm": 2.2105197537112167, + "learning_rate": 7.480529613127507e-08, + "loss": 1.026, + "step": 40842 + }, + { + "epoch": 0.96, + "grad_norm": 2.102290782286467, + "learning_rate": 7.471216805207682e-08, + "loss": 0.8382, + "step": 40843 + }, + { + "epoch": 0.96, + "grad_norm": 1.9610908691017177, + "learning_rate": 7.461909776104725e-08, + "loss": 0.9597, + "step": 40844 + }, + { + "epoch": 0.96, + "grad_norm": 2.101503666622352, + "learning_rate": 7.452608525872595e-08, + "loss": 0.9253, + "step": 40845 + }, + { + "epoch": 0.96, + "grad_norm": 2.0434097055540272, + "learning_rate": 7.443313054565804e-08, + "loss": 1.0813, + "step": 40846 + }, + { + "epoch": 0.96, + "grad_norm": 2.0341921115537955, + "learning_rate": 7.434023362238086e-08, + "loss": 0.9628, + "step": 40847 + }, + { + "epoch": 0.96, + "grad_norm": 1.0590680750423715, + "learning_rate": 7.42473944894384e-08, + "loss": 0.9736, + "step": 40848 + }, + { + "epoch": 0.96, + "grad_norm": 2.1648508865721676, + "learning_rate": 7.415461314737026e-08, + "loss": 0.7711, + "step": 40849 + }, + { + "epoch": 0.96, + "grad_norm": 1.9312120784654956, + "learning_rate": 7.406188959671601e-08, + "loss": 0.9742, + "step": 40850 + }, + { + "epoch": 0.96, + "grad_norm": 2.0294225552749077, + "learning_rate": 7.39692238380163e-08, + "loss": 1.0104, + "step": 40851 + }, + { + "epoch": 0.96, + "grad_norm": 2.287973536711275, + "learning_rate": 7.387661587180961e-08, + "loss": 1.0375, + "step": 40852 + }, + { + "epoch": 0.96, + "grad_norm": 1.1142217231576284, + "learning_rate": 7.37840656986366e-08, + "loss": 0.9106, + "step": 40853 + }, + { + "epoch": 0.96, + "grad_norm": 2.025399541076524, + "learning_rate": 7.369157331903465e-08, + "loss": 1.0426, + "step": 40854 + }, + { + "epoch": 0.96, + "grad_norm": 2.206661130983252, + "learning_rate": 7.35991387335433e-08, + "loss": 0.9811, + "step": 40855 + }, + { + "epoch": 0.96, + "grad_norm": 2.2338181890596833, + "learning_rate": 7.350676194270101e-08, + "loss": 1.0251, + "step": 40856 + }, + { + "epoch": 0.96, + "grad_norm": 1.9918625386001918, + "learning_rate": 7.341444294704403e-08, + "loss": 1.0316, + "step": 40857 + }, + { + "epoch": 0.96, + "grad_norm": 2.0518517207170963, + "learning_rate": 7.332218174711081e-08, + "loss": 0.9276, + "step": 40858 + }, + { + "epoch": 0.96, + "grad_norm": 1.8144830319133642, + "learning_rate": 7.322997834343871e-08, + "loss": 0.8595, + "step": 40859 + }, + { + "epoch": 0.96, + "grad_norm": 2.068384487145543, + "learning_rate": 7.313783273656506e-08, + "loss": 1.0091, + "step": 40860 + }, + { + "epoch": 0.96, + "grad_norm": 1.8811768925321584, + "learning_rate": 7.304574492702609e-08, + "loss": 0.8534, + "step": 40861 + }, + { + "epoch": 0.96, + "grad_norm": 1.1282157351518378, + "learning_rate": 7.295371491535697e-08, + "loss": 0.963, + "step": 40862 + }, + { + "epoch": 0.96, + "grad_norm": 2.533991091110529, + "learning_rate": 7.28617427020939e-08, + "loss": 1.0502, + "step": 40863 + }, + { + "epoch": 0.96, + "grad_norm": 1.8584911861518245, + "learning_rate": 7.276982828777312e-08, + "loss": 0.9125, + "step": 40864 + }, + { + "epoch": 0.96, + "grad_norm": 1.9723702720924845, + "learning_rate": 7.267797167292867e-08, + "loss": 0.8371, + "step": 40865 + }, + { + "epoch": 0.96, + "grad_norm": 2.9383895416787404, + "learning_rate": 7.258617285809677e-08, + "loss": 0.9855, + "step": 40866 + }, + { + "epoch": 0.96, + "grad_norm": 2.0156939602392137, + "learning_rate": 7.249443184381033e-08, + "loss": 0.8525, + "step": 40867 + }, + { + "epoch": 0.96, + "grad_norm": 1.9582256846371637, + "learning_rate": 7.240274863060448e-08, + "loss": 1.0492, + "step": 40868 + }, + { + "epoch": 0.96, + "grad_norm": 1.9707900939962766, + "learning_rate": 7.231112321901435e-08, + "loss": 0.9422, + "step": 40869 + }, + { + "epoch": 0.96, + "grad_norm": 2.0226616835788254, + "learning_rate": 7.221955560957062e-08, + "loss": 1.0581, + "step": 40870 + }, + { + "epoch": 0.96, + "grad_norm": 2.0078036117565707, + "learning_rate": 7.21280458028073e-08, + "loss": 1.114, + "step": 40871 + }, + { + "epoch": 0.96, + "grad_norm": 1.9225485862933223, + "learning_rate": 7.203659379925843e-08, + "loss": 1.0459, + "step": 40872 + }, + { + "epoch": 0.96, + "grad_norm": 1.090489475868326, + "learning_rate": 7.194519959945467e-08, + "loss": 0.8785, + "step": 40873 + }, + { + "epoch": 0.96, + "grad_norm": 1.9416679734434368, + "learning_rate": 7.185386320393006e-08, + "loss": 0.939, + "step": 40874 + }, + { + "epoch": 0.96, + "grad_norm": 1.0992944785917764, + "learning_rate": 7.176258461321528e-08, + "loss": 0.962, + "step": 40875 + }, + { + "epoch": 0.96, + "grad_norm": 2.202542664285705, + "learning_rate": 7.167136382784212e-08, + "loss": 0.9672, + "step": 40876 + }, + { + "epoch": 0.96, + "grad_norm": 1.897270605138864, + "learning_rate": 7.158020084834128e-08, + "loss": 0.9193, + "step": 40877 + }, + { + "epoch": 0.96, + "grad_norm": 1.954737263779977, + "learning_rate": 7.148909567524343e-08, + "loss": 0.9089, + "step": 40878 + }, + { + "epoch": 0.96, + "grad_norm": 2.289987868934493, + "learning_rate": 7.139804830908037e-08, + "loss": 0.938, + "step": 40879 + }, + { + "epoch": 0.96, + "grad_norm": 1.993104437396944, + "learning_rate": 7.130705875037946e-08, + "loss": 0.9118, + "step": 40880 + }, + { + "epoch": 0.96, + "grad_norm": 2.240184004091795, + "learning_rate": 7.121612699967362e-08, + "loss": 0.8371, + "step": 40881 + }, + { + "epoch": 0.96, + "grad_norm": 1.8221775155778488, + "learning_rate": 7.112525305749129e-08, + "loss": 0.9781, + "step": 40882 + }, + { + "epoch": 0.96, + "grad_norm": 1.7457533102852072, + "learning_rate": 7.103443692435985e-08, + "loss": 0.8436, + "step": 40883 + }, + { + "epoch": 0.96, + "grad_norm": 3.3754737565506012, + "learning_rate": 7.094367860081108e-08, + "loss": 0.8924, + "step": 40884 + }, + { + "epoch": 0.96, + "grad_norm": 2.5064471058982902, + "learning_rate": 7.085297808737013e-08, + "loss": 1.0177, + "step": 40885 + }, + { + "epoch": 0.96, + "grad_norm": 1.868311785696604, + "learning_rate": 7.076233538456767e-08, + "loss": 1.0406, + "step": 40886 + }, + { + "epoch": 0.96, + "grad_norm": 2.005211249927107, + "learning_rate": 7.067175049292884e-08, + "loss": 0.9383, + "step": 40887 + }, + { + "epoch": 0.96, + "grad_norm": 1.9775700760987902, + "learning_rate": 7.058122341298435e-08, + "loss": 0.856, + "step": 40888 + }, + { + "epoch": 0.96, + "grad_norm": 1.9130957181134318, + "learning_rate": 7.049075414525818e-08, + "loss": 1.0148, + "step": 40889 + }, + { + "epoch": 0.96, + "grad_norm": 1.8720713134349043, + "learning_rate": 7.040034269027995e-08, + "loss": 0.9926, + "step": 40890 + }, + { + "epoch": 0.96, + "grad_norm": 2.2694058249223423, + "learning_rate": 7.030998904857478e-08, + "loss": 0.9958, + "step": 40891 + }, + { + "epoch": 0.96, + "grad_norm": 2.0734043078166153, + "learning_rate": 7.021969322066668e-08, + "loss": 0.983, + "step": 40892 + }, + { + "epoch": 0.96, + "grad_norm": 2.300406475865961, + "learning_rate": 7.012945520708414e-08, + "loss": 1.0094, + "step": 40893 + }, + { + "epoch": 0.96, + "grad_norm": 1.8619505097764362, + "learning_rate": 7.003927500835118e-08, + "loss": 0.9139, + "step": 40894 + }, + { + "epoch": 0.96, + "grad_norm": 2.3996962554861083, + "learning_rate": 6.994915262499513e-08, + "loss": 0.9547, + "step": 40895 + }, + { + "epoch": 0.96, + "grad_norm": 1.7598381056661623, + "learning_rate": 6.985908805753672e-08, + "loss": 1.058, + "step": 40896 + }, + { + "epoch": 0.96, + "grad_norm": 1.1342454634639347, + "learning_rate": 6.976908130650329e-08, + "loss": 0.8837, + "step": 40897 + }, + { + "epoch": 0.96, + "grad_norm": 2.0154794641144984, + "learning_rate": 6.967913237241885e-08, + "loss": 1.0442, + "step": 40898 + }, + { + "epoch": 0.96, + "grad_norm": 2.0350960160409937, + "learning_rate": 6.958924125580635e-08, + "loss": 1.009, + "step": 40899 + }, + { + "epoch": 0.96, + "grad_norm": 1.983946975240773, + "learning_rate": 6.949940795718868e-08, + "loss": 0.9853, + "step": 40900 + }, + { + "epoch": 0.96, + "grad_norm": 2.0409323471136362, + "learning_rate": 6.940963247708877e-08, + "loss": 0.9774, + "step": 40901 + }, + { + "epoch": 0.96, + "grad_norm": 2.144130651185757, + "learning_rate": 6.931991481603062e-08, + "loss": 0.8536, + "step": 40902 + }, + { + "epoch": 0.96, + "grad_norm": 1.849678061752215, + "learning_rate": 6.923025497453606e-08, + "loss": 0.9772, + "step": 40903 + }, + { + "epoch": 0.96, + "grad_norm": 2.1066046088844828, + "learning_rate": 6.914065295312689e-08, + "loss": 0.9754, + "step": 40904 + }, + { + "epoch": 0.96, + "grad_norm": 2.0992764731349522, + "learning_rate": 6.90511087523249e-08, + "loss": 1.0271, + "step": 40905 + }, + { + "epoch": 0.96, + "grad_norm": 1.8618499283632775, + "learning_rate": 6.896162237265192e-08, + "loss": 1.1236, + "step": 40906 + }, + { + "epoch": 0.96, + "grad_norm": 1.972254009948327, + "learning_rate": 6.887219381462751e-08, + "loss": 0.9411, + "step": 40907 + }, + { + "epoch": 0.96, + "grad_norm": 2.2419677042644106, + "learning_rate": 6.878282307877349e-08, + "loss": 0.9699, + "step": 40908 + }, + { + "epoch": 0.96, + "grad_norm": 1.133311087131277, + "learning_rate": 6.869351016561055e-08, + "loss": 0.9383, + "step": 40909 + }, + { + "epoch": 0.96, + "grad_norm": 1.9818547109322038, + "learning_rate": 6.860425507565826e-08, + "loss": 0.9015, + "step": 40910 + }, + { + "epoch": 0.96, + "grad_norm": 1.9662106810253472, + "learning_rate": 6.851505780943512e-08, + "loss": 1.008, + "step": 40911 + }, + { + "epoch": 0.96, + "grad_norm": 2.9317051215196352, + "learning_rate": 6.842591836746293e-08, + "loss": 0.9309, + "step": 40912 + }, + { + "epoch": 0.96, + "grad_norm": 2.3383625039998814, + "learning_rate": 6.833683675025904e-08, + "loss": 0.9214, + "step": 40913 + }, + { + "epoch": 0.96, + "grad_norm": 2.1328903099881735, + "learning_rate": 6.824781295834193e-08, + "loss": 1.0465, + "step": 40914 + }, + { + "epoch": 0.96, + "grad_norm": 1.6873276851431136, + "learning_rate": 6.815884699223007e-08, + "loss": 1.0051, + "step": 40915 + }, + { + "epoch": 0.96, + "grad_norm": 1.8332141923641465, + "learning_rate": 6.806993885244307e-08, + "loss": 1.0772, + "step": 40916 + }, + { + "epoch": 0.96, + "grad_norm": 1.9661597203112693, + "learning_rate": 6.798108853949603e-08, + "loss": 1.0312, + "step": 40917 + }, + { + "epoch": 0.96, + "grad_norm": 2.104058722443492, + "learning_rate": 6.789229605390856e-08, + "loss": 0.9498, + "step": 40918 + }, + { + "epoch": 0.96, + "grad_norm": 1.9437986350112717, + "learning_rate": 6.780356139619582e-08, + "loss": 1.0363, + "step": 40919 + }, + { + "epoch": 0.96, + "grad_norm": 1.7427261002710486, + "learning_rate": 6.771488456687514e-08, + "loss": 0.9958, + "step": 40920 + }, + { + "epoch": 0.96, + "grad_norm": 2.142640607795612, + "learning_rate": 6.76262655664639e-08, + "loss": 1.0687, + "step": 40921 + }, + { + "epoch": 0.96, + "grad_norm": 2.0534068043736795, + "learning_rate": 6.753770439547614e-08, + "loss": 0.9755, + "step": 40922 + }, + { + "epoch": 0.96, + "grad_norm": 2.0617046064265363, + "learning_rate": 6.74492010544281e-08, + "loss": 1.0555, + "step": 40923 + }, + { + "epoch": 0.96, + "grad_norm": 2.237690702582438, + "learning_rate": 6.736075554383604e-08, + "loss": 1.1008, + "step": 40924 + }, + { + "epoch": 0.96, + "grad_norm": 2.7111807582292147, + "learning_rate": 6.727236786421509e-08, + "loss": 0.9648, + "step": 40925 + }, + { + "epoch": 0.96, + "grad_norm": 1.8241423484555044, + "learning_rate": 6.718403801607709e-08, + "loss": 0.9911, + "step": 40926 + }, + { + "epoch": 0.96, + "grad_norm": 1.8452320401540419, + "learning_rate": 6.70957659999405e-08, + "loss": 0.9141, + "step": 40927 + }, + { + "epoch": 0.96, + "grad_norm": 2.1369582781616066, + "learning_rate": 6.700755181631491e-08, + "loss": 0.919, + "step": 40928 + }, + { + "epoch": 0.96, + "grad_norm": 2.0053907042933794, + "learning_rate": 6.691939546571769e-08, + "loss": 1.018, + "step": 40929 + }, + { + "epoch": 0.96, + "grad_norm": 1.1096225201310836, + "learning_rate": 6.683129694866063e-08, + "loss": 0.957, + "step": 40930 + }, + { + "epoch": 0.96, + "grad_norm": 6.130614669750862, + "learning_rate": 6.674325626565558e-08, + "loss": 0.9767, + "step": 40931 + }, + { + "epoch": 0.96, + "grad_norm": 2.0499828308010883, + "learning_rate": 6.665527341721544e-08, + "loss": 1.0191, + "step": 40932 + }, + { + "epoch": 0.96, + "grad_norm": 1.8687523134063353, + "learning_rate": 6.656734840385425e-08, + "loss": 0.969, + "step": 40933 + }, + { + "epoch": 0.96, + "grad_norm": 2.1015082394630675, + "learning_rate": 6.647948122608383e-08, + "loss": 0.9956, + "step": 40934 + }, + { + "epoch": 0.96, + "grad_norm": 2.0554187590920536, + "learning_rate": 6.639167188441264e-08, + "loss": 0.9687, + "step": 40935 + }, + { + "epoch": 0.96, + "grad_norm": 2.0438969717355757, + "learning_rate": 6.630392037935474e-08, + "loss": 0.9231, + "step": 40936 + }, + { + "epoch": 0.96, + "grad_norm": 1.9622637505673666, + "learning_rate": 6.621622671142191e-08, + "loss": 1.0702, + "step": 40937 + }, + { + "epoch": 0.96, + "grad_norm": 1.8712405988829601, + "learning_rate": 6.612859088112155e-08, + "loss": 1.0067, + "step": 40938 + }, + { + "epoch": 0.96, + "grad_norm": 1.8653140293833599, + "learning_rate": 6.604101288896548e-08, + "loss": 0.8537, + "step": 40939 + }, + { + "epoch": 0.96, + "grad_norm": 1.7827558228028264, + "learning_rate": 6.595349273546437e-08, + "loss": 0.9543, + "step": 40940 + }, + { + "epoch": 0.96, + "grad_norm": 2.200154529105503, + "learning_rate": 6.586603042112671e-08, + "loss": 1.0684, + "step": 40941 + }, + { + "epoch": 0.96, + "grad_norm": 2.0150218512164346, + "learning_rate": 6.577862594646323e-08, + "loss": 1.032, + "step": 40942 + }, + { + "epoch": 0.96, + "grad_norm": 2.082865647413157, + "learning_rate": 6.569127931198016e-08, + "loss": 0.9546, + "step": 40943 + }, + { + "epoch": 0.96, + "grad_norm": 1.983528454006547, + "learning_rate": 6.560399051818822e-08, + "loss": 1.0357, + "step": 40944 + }, + { + "epoch": 0.96, + "grad_norm": 2.0331340915721525, + "learning_rate": 6.551675956559478e-08, + "loss": 0.9936, + "step": 40945 + }, + { + "epoch": 0.96, + "grad_norm": 2.2715410456513108, + "learning_rate": 6.542958645470832e-08, + "loss": 0.9318, + "step": 40946 + }, + { + "epoch": 0.96, + "grad_norm": 1.7702450682231123, + "learning_rate": 6.534247118603621e-08, + "loss": 1.0189, + "step": 40947 + }, + { + "epoch": 0.96, + "grad_norm": 2.1176573794556903, + "learning_rate": 6.525541376008471e-08, + "loss": 1.0728, + "step": 40948 + }, + { + "epoch": 0.96, + "grad_norm": 2.267940983009334, + "learning_rate": 6.51684141773623e-08, + "loss": 0.9282, + "step": 40949 + }, + { + "epoch": 0.96, + "grad_norm": 1.8410619085163036, + "learning_rate": 6.508147243837415e-08, + "loss": 0.9919, + "step": 40950 + }, + { + "epoch": 0.96, + "grad_norm": 2.0996900873417363, + "learning_rate": 6.499458854362761e-08, + "loss": 0.7622, + "step": 40951 + }, + { + "epoch": 0.96, + "grad_norm": 1.963285319183298, + "learning_rate": 6.490776249362785e-08, + "loss": 0.919, + "step": 40952 + }, + { + "epoch": 0.96, + "grad_norm": 2.0913253191667227, + "learning_rate": 6.482099428888e-08, + "loss": 0.9583, + "step": 40953 + }, + { + "epoch": 0.96, + "grad_norm": 2.3425546295382556, + "learning_rate": 6.473428392989034e-08, + "loss": 0.9996, + "step": 40954 + }, + { + "epoch": 0.96, + "grad_norm": 1.0702559152058666, + "learning_rate": 6.46476314171629e-08, + "loss": 0.9189, + "step": 40955 + }, + { + "epoch": 0.96, + "grad_norm": 2.30986142299753, + "learning_rate": 6.456103675120173e-08, + "loss": 1.1557, + "step": 40956 + }, + { + "epoch": 0.96, + "grad_norm": 2.0352734413699376, + "learning_rate": 6.447449993251197e-08, + "loss": 0.9509, + "step": 40957 + }, + { + "epoch": 0.96, + "grad_norm": 2.1252001165686827, + "learning_rate": 6.438802096159769e-08, + "loss": 0.9726, + "step": 40958 + }, + { + "epoch": 0.96, + "grad_norm": 1.0941828525799835, + "learning_rate": 6.430159983896179e-08, + "loss": 0.95, + "step": 40959 + }, + { + "epoch": 0.96, + "grad_norm": 1.8751901316603474, + "learning_rate": 6.42152365651072e-08, + "loss": 0.95, + "step": 40960 + }, + { + "epoch": 0.97, + "grad_norm": 2.0533317849894557, + "learning_rate": 6.412893114053798e-08, + "loss": 0.9182, + "step": 40961 + }, + { + "epoch": 0.97, + "grad_norm": 2.1418645094934328, + "learning_rate": 6.404268356575483e-08, + "loss": 0.9287, + "step": 40962 + }, + { + "epoch": 0.97, + "grad_norm": 1.9942174410642917, + "learning_rate": 6.395649384126179e-08, + "loss": 0.9165, + "step": 40963 + }, + { + "epoch": 0.97, + "grad_norm": 1.9660471667102137, + "learning_rate": 6.387036196755847e-08, + "loss": 1.0673, + "step": 40964 + }, + { + "epoch": 0.97, + "grad_norm": 2.2292348848563663, + "learning_rate": 6.378428794514891e-08, + "loss": 0.9441, + "step": 40965 + }, + { + "epoch": 0.97, + "grad_norm": 2.2399704714015, + "learning_rate": 6.369827177453158e-08, + "loss": 1.0394, + "step": 40966 + }, + { + "epoch": 0.97, + "grad_norm": 1.7959473097267, + "learning_rate": 6.361231345621055e-08, + "loss": 0.8966, + "step": 40967 + }, + { + "epoch": 0.97, + "grad_norm": 2.051504593846915, + "learning_rate": 6.352641299068318e-08, + "loss": 0.9751, + "step": 40968 + }, + { + "epoch": 0.97, + "grad_norm": 1.9891903068266588, + "learning_rate": 6.344057037845131e-08, + "loss": 1.0065, + "step": 40969 + }, + { + "epoch": 0.97, + "grad_norm": 2.053583571774072, + "learning_rate": 6.33547856200134e-08, + "loss": 0.992, + "step": 40970 + }, + { + "epoch": 0.97, + "grad_norm": 1.9705819073996418, + "learning_rate": 6.32690587158713e-08, + "loss": 0.8795, + "step": 40971 + }, + { + "epoch": 0.97, + "grad_norm": 1.995258927853663, + "learning_rate": 6.318338966652127e-08, + "loss": 0.9784, + "step": 40972 + }, + { + "epoch": 0.97, + "grad_norm": 1.9783037214348937, + "learning_rate": 6.309777847246401e-08, + "loss": 1.0637, + "step": 40973 + }, + { + "epoch": 0.97, + "grad_norm": 2.180571173796336, + "learning_rate": 6.301222513419692e-08, + "loss": 1.0034, + "step": 40974 + }, + { + "epoch": 0.97, + "grad_norm": 1.750071668727277, + "learning_rate": 6.292672965221958e-08, + "loss": 1.0705, + "step": 40975 + }, + { + "epoch": 0.97, + "grad_norm": 2.0809023381573124, + "learning_rate": 6.284129202702938e-08, + "loss": 1.0194, + "step": 40976 + }, + { + "epoch": 0.97, + "grad_norm": 1.1466737253509705, + "learning_rate": 6.275591225912148e-08, + "loss": 0.9612, + "step": 40977 + }, + { + "epoch": 0.97, + "grad_norm": 1.8927173068771688, + "learning_rate": 6.267059034899547e-08, + "loss": 0.9593, + "step": 40978 + }, + { + "epoch": 0.97, + "grad_norm": 2.193649751518772, + "learning_rate": 6.258532629714765e-08, + "loss": 0.9595, + "step": 40979 + }, + { + "epoch": 0.97, + "grad_norm": 2.089857811523975, + "learning_rate": 6.250012010407425e-08, + "loss": 0.9469, + "step": 40980 + }, + { + "epoch": 0.97, + "grad_norm": 1.8337564475285262, + "learning_rate": 6.241497177027156e-08, + "loss": 0.9474, + "step": 40981 + }, + { + "epoch": 0.97, + "grad_norm": 2.0732637356883976, + "learning_rate": 6.232988129623475e-08, + "loss": 0.9035, + "step": 40982 + }, + { + "epoch": 0.97, + "grad_norm": 2.2691631925586555, + "learning_rate": 6.224484868245895e-08, + "loss": 0.9565, + "step": 40983 + }, + { + "epoch": 0.97, + "grad_norm": 1.1238925770249797, + "learning_rate": 6.215987392944045e-08, + "loss": 1.0697, + "step": 40984 + }, + { + "epoch": 0.97, + "grad_norm": 2.121727801123031, + "learning_rate": 6.20749570376733e-08, + "loss": 0.9201, + "step": 40985 + }, + { + "epoch": 0.97, + "grad_norm": 2.1198097768394337, + "learning_rate": 6.199009800765265e-08, + "loss": 0.9635, + "step": 40986 + }, + { + "epoch": 0.97, + "grad_norm": 2.253244580988974, + "learning_rate": 6.190529683987145e-08, + "loss": 1.0344, + "step": 40987 + }, + { + "epoch": 0.97, + "grad_norm": 1.7967069877075827, + "learning_rate": 6.182055353482375e-08, + "loss": 1.0257, + "step": 40988 + }, + { + "epoch": 0.97, + "grad_norm": 2.3485089041367964, + "learning_rate": 6.173586809300358e-08, + "loss": 0.9089, + "step": 40989 + }, + { + "epoch": 0.97, + "grad_norm": 1.9945210410173229, + "learning_rate": 6.16512405149039e-08, + "loss": 0.9951, + "step": 40990 + }, + { + "epoch": 0.97, + "grad_norm": 2.101377328794494, + "learning_rate": 6.156667080101652e-08, + "loss": 0.9582, + "step": 40991 + }, + { + "epoch": 0.97, + "grad_norm": 2.141184171027237, + "learning_rate": 6.14821589518344e-08, + "loss": 0.9782, + "step": 40992 + }, + { + "epoch": 0.97, + "grad_norm": 2.1505920360658175, + "learning_rate": 6.139770496785047e-08, + "loss": 1.1352, + "step": 40993 + }, + { + "epoch": 0.97, + "grad_norm": 2.107518213017234, + "learning_rate": 6.131330884955433e-08, + "loss": 0.9741, + "step": 40994 + }, + { + "epoch": 0.97, + "grad_norm": 2.057035743198312, + "learning_rate": 6.122897059743894e-08, + "loss": 0.9342, + "step": 40995 + }, + { + "epoch": 0.97, + "grad_norm": 1.8443326894978358, + "learning_rate": 6.114469021199498e-08, + "loss": 0.9063, + "step": 40996 + }, + { + "epoch": 0.97, + "grad_norm": 1.7738521013569508, + "learning_rate": 6.106046769371432e-08, + "loss": 0.9631, + "step": 40997 + }, + { + "epoch": 0.97, + "grad_norm": 2.126484952715494, + "learning_rate": 6.097630304308543e-08, + "loss": 0.9199, + "step": 40998 + }, + { + "epoch": 0.97, + "grad_norm": 2.0433024139998146, + "learning_rate": 6.089219626059905e-08, + "loss": 0.9792, + "step": 40999 + }, + { + "epoch": 0.97, + "grad_norm": 2.1528146595555757, + "learning_rate": 6.080814734674478e-08, + "loss": 0.9383, + "step": 41000 + }, + { + "epoch": 0.97, + "grad_norm": 1.9509103639631586, + "learning_rate": 6.072415630201333e-08, + "loss": 0.8685, + "step": 41001 + }, + { + "epoch": 0.97, + "grad_norm": 1.8629890613661255, + "learning_rate": 6.0640223126891e-08, + "loss": 0.9984, + "step": 41002 + }, + { + "epoch": 0.97, + "grad_norm": 2.064023381195691, + "learning_rate": 6.055634782186848e-08, + "loss": 0.8605, + "step": 41003 + }, + { + "epoch": 0.97, + "grad_norm": 1.951942382879893, + "learning_rate": 6.047253038743428e-08, + "loss": 0.8559, + "step": 41004 + }, + { + "epoch": 0.97, + "grad_norm": 1.090726387437813, + "learning_rate": 6.03887708240758e-08, + "loss": 1.0025, + "step": 41005 + }, + { + "epoch": 0.97, + "grad_norm": 2.240979322296025, + "learning_rate": 6.03050691322804e-08, + "loss": 0.97, + "step": 41006 + }, + { + "epoch": 0.97, + "grad_norm": 1.7748033091897735, + "learning_rate": 6.02214253125366e-08, + "loss": 0.9816, + "step": 41007 + }, + { + "epoch": 0.97, + "grad_norm": 2.22536067829084, + "learning_rate": 6.013783936532846e-08, + "loss": 0.9884, + "step": 41008 + }, + { + "epoch": 0.97, + "grad_norm": 1.9409133448035865, + "learning_rate": 6.00543112911467e-08, + "loss": 1.0412, + "step": 41009 + }, + { + "epoch": 0.97, + "grad_norm": 2.0271021259342317, + "learning_rate": 5.997084109047424e-08, + "loss": 0.9695, + "step": 41010 + }, + { + "epoch": 0.97, + "grad_norm": 2.119471147356834, + "learning_rate": 5.98874287637996e-08, + "loss": 0.8461, + "step": 41011 + }, + { + "epoch": 0.97, + "grad_norm": 2.1987239206213616, + "learning_rate": 5.980407431160684e-08, + "loss": 0.936, + "step": 41012 + }, + { + "epoch": 0.97, + "grad_norm": 2.1835091493242618, + "learning_rate": 5.972077773438223e-08, + "loss": 1.0453, + "step": 41013 + }, + { + "epoch": 0.97, + "grad_norm": 1.9815357872887946, + "learning_rate": 5.963753903260983e-08, + "loss": 0.9677, + "step": 41014 + }, + { + "epoch": 0.97, + "grad_norm": 2.0300430808915455, + "learning_rate": 5.95543582067748e-08, + "loss": 0.9654, + "step": 41015 + }, + { + "epoch": 0.97, + "grad_norm": 1.9904023817622676, + "learning_rate": 5.94712352573612e-08, + "loss": 0.913, + "step": 41016 + }, + { + "epoch": 0.97, + "grad_norm": 1.7523211691437224, + "learning_rate": 5.938817018485421e-08, + "loss": 1.03, + "step": 41017 + }, + { + "epoch": 0.97, + "grad_norm": 1.938096315073956, + "learning_rate": 5.930516298973565e-08, + "loss": 0.9655, + "step": 41018 + }, + { + "epoch": 0.97, + "grad_norm": 2.459469665531584, + "learning_rate": 5.922221367248959e-08, + "loss": 1.0221, + "step": 41019 + }, + { + "epoch": 0.97, + "grad_norm": 2.0456276583550492, + "learning_rate": 5.913932223359897e-08, + "loss": 0.9386, + "step": 41020 + }, + { + "epoch": 0.97, + "grad_norm": 1.9747253492626404, + "learning_rate": 5.905648867354563e-08, + "loss": 0.9954, + "step": 41021 + }, + { + "epoch": 0.97, + "grad_norm": 1.112997855958374, + "learning_rate": 5.897371299281473e-08, + "loss": 0.9819, + "step": 41022 + }, + { + "epoch": 0.97, + "grad_norm": 2.0389846537060836, + "learning_rate": 5.8890995191884795e-08, + "loss": 1.06, + "step": 41023 + }, + { + "epoch": 0.97, + "grad_norm": 1.861150987953512, + "learning_rate": 5.880833527123875e-08, + "loss": 0.8326, + "step": 41024 + }, + { + "epoch": 0.97, + "grad_norm": 1.0531765684992107, + "learning_rate": 5.872573323135844e-08, + "loss": 0.9022, + "step": 41025 + }, + { + "epoch": 0.97, + "grad_norm": 2.0860077807288, + "learning_rate": 5.8643189072723485e-08, + "loss": 1.0422, + "step": 41026 + }, + { + "epoch": 0.97, + "grad_norm": 1.894486110557521, + "learning_rate": 5.856070279581572e-08, + "loss": 1.0795, + "step": 41027 + }, + { + "epoch": 0.97, + "grad_norm": 1.958720131333134, + "learning_rate": 5.847827440111475e-08, + "loss": 1.0198, + "step": 41028 + }, + { + "epoch": 0.97, + "grad_norm": 2.2397082584758103, + "learning_rate": 5.839590388910132e-08, + "loss": 1.0761, + "step": 41029 + }, + { + "epoch": 0.97, + "grad_norm": 1.849315287730748, + "learning_rate": 5.831359126025393e-08, + "loss": 0.9628, + "step": 41030 + }, + { + "epoch": 0.97, + "grad_norm": 1.9122470345517657, + "learning_rate": 5.8231336515052196e-08, + "loss": 0.9792, + "step": 41031 + }, + { + "epoch": 0.97, + "grad_norm": 1.9647289720383947, + "learning_rate": 5.814913965397573e-08, + "loss": 1.0372, + "step": 41032 + }, + { + "epoch": 0.97, + "grad_norm": 2.163125522371875, + "learning_rate": 5.806700067750193e-08, + "loss": 0.9471, + "step": 41033 + }, + { + "epoch": 0.97, + "grad_norm": 1.9543825352168656, + "learning_rate": 5.79849195861093e-08, + "loss": 1.0431, + "step": 41034 + }, + { + "epoch": 0.97, + "grad_norm": 1.0475504437857954, + "learning_rate": 5.7902896380276354e-08, + "loss": 0.9315, + "step": 41035 + }, + { + "epoch": 0.97, + "grad_norm": 1.7955101480461557, + "learning_rate": 5.782093106048159e-08, + "loss": 0.8945, + "step": 41036 + }, + { + "epoch": 0.97, + "grad_norm": 2.6176411444626866, + "learning_rate": 5.7739023627199074e-08, + "loss": 0.9557, + "step": 41037 + }, + { + "epoch": 0.97, + "grad_norm": 1.9539798843932472, + "learning_rate": 5.765717408090843e-08, + "loss": 0.8992, + "step": 41038 + }, + { + "epoch": 0.97, + "grad_norm": 2.001935129201068, + "learning_rate": 5.7575382422085935e-08, + "loss": 0.9637, + "step": 41039 + }, + { + "epoch": 0.97, + "grad_norm": 1.040510824489455, + "learning_rate": 5.749364865120677e-08, + "loss": 0.9501, + "step": 41040 + }, + { + "epoch": 0.97, + "grad_norm": 2.172459261765914, + "learning_rate": 5.7411972768747213e-08, + "loss": 0.9464, + "step": 41041 + }, + { + "epoch": 0.97, + "grad_norm": 2.228982247083375, + "learning_rate": 5.733035477518245e-08, + "loss": 1.0425, + "step": 41042 + }, + { + "epoch": 0.97, + "grad_norm": 1.7949850527361344, + "learning_rate": 5.7248794670989874e-08, + "loss": 0.9607, + "step": 41043 + }, + { + "epoch": 0.97, + "grad_norm": 1.9826895722153688, + "learning_rate": 5.7167292456641324e-08, + "loss": 0.9084, + "step": 41044 + }, + { + "epoch": 0.97, + "grad_norm": 1.8419825340477534, + "learning_rate": 5.708584813261308e-08, + "loss": 0.8946, + "step": 41045 + }, + { + "epoch": 0.97, + "grad_norm": 1.9416714261466121, + "learning_rate": 5.700446169937812e-08, + "loss": 0.9538, + "step": 41046 + }, + { + "epoch": 0.97, + "grad_norm": 1.9788833939598707, + "learning_rate": 5.69231331574116e-08, + "loss": 0.9734, + "step": 41047 + }, + { + "epoch": 0.97, + "grad_norm": 1.9152553121117535, + "learning_rate": 5.6841862507186484e-08, + "loss": 0.9576, + "step": 41048 + }, + { + "epoch": 0.97, + "grad_norm": 1.7585480161171219, + "learning_rate": 5.6760649749175726e-08, + "loss": 0.8982, + "step": 41049 + }, + { + "epoch": 0.97, + "grad_norm": 2.4401852687398535, + "learning_rate": 5.667949488385338e-08, + "loss": 1.0319, + "step": 41050 + }, + { + "epoch": 0.97, + "grad_norm": 1.860538246332299, + "learning_rate": 5.6598397911690194e-08, + "loss": 0.9504, + "step": 41051 + }, + { + "epoch": 0.97, + "grad_norm": 2.005877717418528, + "learning_rate": 5.651735883316023e-08, + "loss": 0.8404, + "step": 41052 + }, + { + "epoch": 0.97, + "grad_norm": 1.905701907306896, + "learning_rate": 5.643637764873311e-08, + "loss": 0.939, + "step": 41053 + }, + { + "epoch": 0.97, + "grad_norm": 1.8813640454912781, + "learning_rate": 5.635545435888179e-08, + "loss": 0.991, + "step": 41054 + }, + { + "epoch": 0.97, + "grad_norm": 2.1031211210359624, + "learning_rate": 5.6274588964077006e-08, + "loss": 0.9175, + "step": 41055 + }, + { + "epoch": 0.97, + "grad_norm": 2.191231648608811, + "learning_rate": 5.619378146478949e-08, + "loss": 0.9406, + "step": 41056 + }, + { + "epoch": 0.97, + "grad_norm": 1.9411567187267398, + "learning_rate": 5.6113031861491086e-08, + "loss": 1.0117, + "step": 41057 + }, + { + "epoch": 0.97, + "grad_norm": 2.3199847876795023, + "learning_rate": 5.60323401546492e-08, + "loss": 0.9206, + "step": 41058 + }, + { + "epoch": 0.97, + "grad_norm": 1.9383615051854723, + "learning_rate": 5.595170634473568e-08, + "loss": 0.9065, + "step": 41059 + }, + { + "epoch": 0.97, + "grad_norm": 1.963165699487724, + "learning_rate": 5.587113043222015e-08, + "loss": 1.0031, + "step": 41060 + }, + { + "epoch": 0.97, + "grad_norm": 1.86135822991806, + "learning_rate": 5.579061241757111e-08, + "loss": 0.9901, + "step": 41061 + }, + { + "epoch": 0.97, + "grad_norm": 2.4009475430539577, + "learning_rate": 5.571015230125709e-08, + "loss": 0.932, + "step": 41062 + }, + { + "epoch": 0.97, + "grad_norm": 1.0993547123509877, + "learning_rate": 5.5629750083746606e-08, + "loss": 0.8797, + "step": 41063 + }, + { + "epoch": 0.97, + "grad_norm": 2.8104744667258696, + "learning_rate": 5.5549405765509265e-08, + "loss": 0.9694, + "step": 41064 + }, + { + "epoch": 0.97, + "grad_norm": 1.9045227350884082, + "learning_rate": 5.546911934701027e-08, + "loss": 0.984, + "step": 41065 + }, + { + "epoch": 0.97, + "grad_norm": 1.8423454067929315, + "learning_rate": 5.538889082871923e-08, + "loss": 0.8846, + "step": 41066 + }, + { + "epoch": 0.97, + "grad_norm": 1.7412330366091897, + "learning_rate": 5.5308720211102454e-08, + "loss": 1.001, + "step": 41067 + }, + { + "epoch": 0.97, + "grad_norm": 2.0971610396673728, + "learning_rate": 5.522860749462622e-08, + "loss": 0.9261, + "step": 41068 + }, + { + "epoch": 0.97, + "grad_norm": 2.0022719182332103, + "learning_rate": 5.514855267975794e-08, + "loss": 1.1083, + "step": 41069 + }, + { + "epoch": 0.97, + "grad_norm": 1.81073891573888, + "learning_rate": 5.506855576696279e-08, + "loss": 0.9074, + "step": 41070 + }, + { + "epoch": 0.97, + "grad_norm": 2.234693241339059, + "learning_rate": 5.498861675670819e-08, + "loss": 0.9138, + "step": 41071 + }, + { + "epoch": 0.97, + "grad_norm": 2.0158835157963164, + "learning_rate": 5.490873564945709e-08, + "loss": 0.909, + "step": 41072 + }, + { + "epoch": 0.97, + "grad_norm": 1.1019637764781385, + "learning_rate": 5.482891244567689e-08, + "loss": 0.9395, + "step": 41073 + }, + { + "epoch": 0.97, + "grad_norm": 2.087033810290443, + "learning_rate": 5.474914714583057e-08, + "loss": 0.9727, + "step": 41074 + }, + { + "epoch": 0.97, + "grad_norm": 1.1097169844500956, + "learning_rate": 5.46694397503833e-08, + "loss": 0.9138, + "step": 41075 + }, + { + "epoch": 0.97, + "grad_norm": 2.0250829199207736, + "learning_rate": 5.458979025979916e-08, + "loss": 1.0353, + "step": 41076 + }, + { + "epoch": 0.97, + "grad_norm": 1.909037764632814, + "learning_rate": 5.451019867454221e-08, + "loss": 1.0776, + "step": 41077 + }, + { + "epoch": 0.97, + "grad_norm": 2.064787153533908, + "learning_rate": 5.4430664995076543e-08, + "loss": 0.8317, + "step": 41078 + }, + { + "epoch": 0.97, + "grad_norm": 2.179772967379295, + "learning_rate": 5.435118922186289e-08, + "loss": 0.8267, + "step": 41079 + }, + { + "epoch": 0.97, + "grad_norm": 2.1036911606047175, + "learning_rate": 5.427177135536532e-08, + "loss": 0.8986, + "step": 41080 + }, + { + "epoch": 0.97, + "grad_norm": 1.8247015798332644, + "learning_rate": 5.4192411396046806e-08, + "loss": 1.0335, + "step": 41081 + }, + { + "epoch": 0.97, + "grad_norm": 2.115751298309531, + "learning_rate": 5.41131093443692e-08, + "loss": 0.8625, + "step": 41082 + }, + { + "epoch": 0.97, + "grad_norm": 1.9024646386371427, + "learning_rate": 5.4033865200793236e-08, + "loss": 0.919, + "step": 41083 + }, + { + "epoch": 0.97, + "grad_norm": 2.051331487476837, + "learning_rate": 5.395467896578077e-08, + "loss": 0.9044, + "step": 41084 + }, + { + "epoch": 0.97, + "grad_norm": 2.0143902635622988, + "learning_rate": 5.387555063979366e-08, + "loss": 0.9713, + "step": 41085 + }, + { + "epoch": 0.97, + "grad_norm": 1.148848285505449, + "learning_rate": 5.3796480223292643e-08, + "loss": 0.9585, + "step": 41086 + }, + { + "epoch": 0.97, + "grad_norm": 1.87379389070674, + "learning_rate": 5.371746771673625e-08, + "loss": 0.9754, + "step": 41087 + }, + { + "epoch": 0.97, + "grad_norm": 1.9730559968520371, + "learning_rate": 5.363851312058632e-08, + "loss": 0.9502, + "step": 41088 + }, + { + "epoch": 0.97, + "grad_norm": 1.8307728939658467, + "learning_rate": 5.355961643530139e-08, + "loss": 1.0525, + "step": 41089 + }, + { + "epoch": 0.97, + "grad_norm": 2.1582814822169736, + "learning_rate": 5.3480777661341077e-08, + "loss": 1.0317, + "step": 41090 + }, + { + "epoch": 0.97, + "grad_norm": 2.040938327776654, + "learning_rate": 5.340199679916502e-08, + "loss": 0.9298, + "step": 41091 + }, + { + "epoch": 0.97, + "grad_norm": 1.8512722905345649, + "learning_rate": 5.332327384923175e-08, + "loss": 0.9206, + "step": 41092 + }, + { + "epoch": 0.97, + "grad_norm": 1.9287532246456707, + "learning_rate": 5.3244608811999776e-08, + "loss": 0.8855, + "step": 41093 + }, + { + "epoch": 0.97, + "grad_norm": 1.9673931601562864, + "learning_rate": 5.3166001687926515e-08, + "loss": 1.0013, + "step": 41094 + }, + { + "epoch": 0.97, + "grad_norm": 1.8899554707970616, + "learning_rate": 5.308745247747049e-08, + "loss": 1.0564, + "step": 41095 + }, + { + "epoch": 0.97, + "grad_norm": 2.091045549243785, + "learning_rate": 5.3008961181086895e-08, + "loss": 0.9658, + "step": 41096 + }, + { + "epoch": 0.97, + "grad_norm": 2.1007695437038145, + "learning_rate": 5.293052779923535e-08, + "loss": 0.9741, + "step": 41097 + }, + { + "epoch": 0.97, + "grad_norm": 4.648124521652305, + "learning_rate": 5.285215233237107e-08, + "loss": 0.9111, + "step": 41098 + }, + { + "epoch": 0.97, + "grad_norm": 1.9144946064304957, + "learning_rate": 5.277383478095144e-08, + "loss": 1.1102, + "step": 41099 + }, + { + "epoch": 0.97, + "grad_norm": 1.9156325027223995, + "learning_rate": 5.2695575145431664e-08, + "loss": 0.8429, + "step": 41100 + }, + { + "epoch": 0.97, + "grad_norm": 1.7833436298089944, + "learning_rate": 5.261737342626694e-08, + "loss": 0.9648, + "step": 41101 + }, + { + "epoch": 0.97, + "grad_norm": 2.0084764146186185, + "learning_rate": 5.253922962391356e-08, + "loss": 0.9009, + "step": 41102 + }, + { + "epoch": 0.97, + "grad_norm": 2.1949913485313504, + "learning_rate": 5.2461143738825605e-08, + "loss": 0.9628, + "step": 41103 + }, + { + "epoch": 0.97, + "grad_norm": 2.0027825573155376, + "learning_rate": 5.238311577145827e-08, + "loss": 0.9227, + "step": 41104 + }, + { + "epoch": 0.97, + "grad_norm": 2.176356620599454, + "learning_rate": 5.230514572226675e-08, + "loss": 1.0659, + "step": 41105 + }, + { + "epoch": 0.97, + "grad_norm": 2.4250607774916886, + "learning_rate": 5.22272335917029e-08, + "loss": 0.8061, + "step": 41106 + }, + { + "epoch": 0.97, + "grad_norm": 2.5081429650924076, + "learning_rate": 5.214937938022191e-08, + "loss": 1.0576, + "step": 41107 + }, + { + "epoch": 0.97, + "grad_norm": 1.9239002283860525, + "learning_rate": 5.2071583088276755e-08, + "loss": 0.9236, + "step": 41108 + }, + { + "epoch": 0.97, + "grad_norm": 2.156741309169227, + "learning_rate": 5.1993844716321515e-08, + "loss": 0.973, + "step": 41109 + }, + { + "epoch": 0.97, + "grad_norm": 1.9330626551282497, + "learning_rate": 5.191616426480584e-08, + "loss": 0.8562, + "step": 41110 + }, + { + "epoch": 0.97, + "grad_norm": 2.329036173812303, + "learning_rate": 5.183854173418379e-08, + "loss": 0.9434, + "step": 41111 + }, + { + "epoch": 0.97, + "grad_norm": 1.885600557594403, + "learning_rate": 5.1760977124908354e-08, + "loss": 0.9732, + "step": 41112 + }, + { + "epoch": 0.97, + "grad_norm": 1.846728231582787, + "learning_rate": 5.168347043742916e-08, + "loss": 0.8757, + "step": 41113 + }, + { + "epoch": 0.97, + "grad_norm": 1.126475790075618, + "learning_rate": 5.160602167219919e-08, + "loss": 0.9775, + "step": 41114 + }, + { + "epoch": 0.97, + "grad_norm": 2.18121788241554, + "learning_rate": 5.152863082966808e-08, + "loss": 0.9783, + "step": 41115 + }, + { + "epoch": 0.97, + "grad_norm": 1.9137646103314052, + "learning_rate": 5.145129791028658e-08, + "loss": 1.0426, + "step": 41116 + }, + { + "epoch": 0.97, + "grad_norm": 1.9625140764832973, + "learning_rate": 5.137402291450655e-08, + "loss": 1.0002, + "step": 41117 + }, + { + "epoch": 0.97, + "grad_norm": 2.5125796719170292, + "learning_rate": 5.129680584277541e-08, + "loss": 1.0008, + "step": 41118 + }, + { + "epoch": 0.97, + "grad_norm": 1.9769976635385418, + "learning_rate": 5.1219646695543915e-08, + "loss": 0.985, + "step": 41119 + }, + { + "epoch": 0.97, + "grad_norm": 1.8660891967032514, + "learning_rate": 5.1142545473261694e-08, + "loss": 0.8935, + "step": 41120 + }, + { + "epoch": 0.97, + "grad_norm": 1.8627854598307723, + "learning_rate": 5.10655021763784e-08, + "loss": 0.9041, + "step": 41121 + }, + { + "epoch": 0.97, + "grad_norm": 2.2698086524529404, + "learning_rate": 5.0988516805340337e-08, + "loss": 1.1163, + "step": 41122 + }, + { + "epoch": 0.97, + "grad_norm": 2.0834582085983517, + "learning_rate": 5.091158936059715e-08, + "loss": 0.9996, + "step": 41123 + }, + { + "epoch": 0.97, + "grad_norm": 0.9845531296304267, + "learning_rate": 5.0834719842596246e-08, + "loss": 0.9352, + "step": 41124 + }, + { + "epoch": 0.97, + "grad_norm": 1.9385765461876252, + "learning_rate": 5.075790825178617e-08, + "loss": 0.9832, + "step": 41125 + }, + { + "epoch": 0.97, + "grad_norm": 2.0876371549411914, + "learning_rate": 5.0681154588612114e-08, + "loss": 1.0904, + "step": 41126 + }, + { + "epoch": 0.97, + "grad_norm": 1.9425501796269324, + "learning_rate": 5.0604458853523716e-08, + "loss": 1.1398, + "step": 41127 + }, + { + "epoch": 0.97, + "grad_norm": 2.2298954209710233, + "learning_rate": 5.052782104696508e-08, + "loss": 1.004, + "step": 41128 + }, + { + "epoch": 0.97, + "grad_norm": 1.9318677321046702, + "learning_rate": 5.045124116938471e-08, + "loss": 0.9954, + "step": 41129 + }, + { + "epoch": 0.97, + "grad_norm": 1.9052642300112979, + "learning_rate": 5.037471922122561e-08, + "loss": 0.9911, + "step": 41130 + }, + { + "epoch": 0.97, + "grad_norm": 2.19090683531265, + "learning_rate": 5.02982552029363e-08, + "loss": 0.9727, + "step": 41131 + }, + { + "epoch": 0.97, + "grad_norm": 1.9710415454109536, + "learning_rate": 5.022184911495864e-08, + "loss": 0.9298, + "step": 41132 + }, + { + "epoch": 0.97, + "grad_norm": 2.025465553722289, + "learning_rate": 5.0145500957741176e-08, + "loss": 1.0013, + "step": 41133 + }, + { + "epoch": 0.97, + "grad_norm": 1.1366384300058927, + "learning_rate": 5.006921073172688e-08, + "loss": 0.8691, + "step": 41134 + }, + { + "epoch": 0.97, + "grad_norm": 1.9587079886728138, + "learning_rate": 4.999297843735873e-08, + "loss": 1.0929, + "step": 41135 + }, + { + "epoch": 0.97, + "grad_norm": 2.258371157193983, + "learning_rate": 4.991680407508193e-08, + "loss": 0.9194, + "step": 41136 + }, + { + "epoch": 0.97, + "grad_norm": 2.1504563781497703, + "learning_rate": 4.9840687645339446e-08, + "loss": 0.8791, + "step": 41137 + }, + { + "epoch": 0.97, + "grad_norm": 2.1964287116104617, + "learning_rate": 4.976462914857538e-08, + "loss": 0.9104, + "step": 41138 + }, + { + "epoch": 0.97, + "grad_norm": 1.1364668140082566, + "learning_rate": 4.968862858523049e-08, + "loss": 0.937, + "step": 41139 + }, + { + "epoch": 0.97, + "grad_norm": 2.2997792716353103, + "learning_rate": 4.961268595574886e-08, + "loss": 0.9195, + "step": 41140 + }, + { + "epoch": 0.97, + "grad_norm": 1.9753829057508523, + "learning_rate": 4.9536801260573475e-08, + "loss": 1.0133, + "step": 41141 + }, + { + "epoch": 0.97, + "grad_norm": 2.2189470322966343, + "learning_rate": 4.946097450014509e-08, + "loss": 0.8399, + "step": 41142 + }, + { + "epoch": 0.97, + "grad_norm": 3.0073966073182334, + "learning_rate": 4.9385205674904455e-08, + "loss": 1.0977, + "step": 41143 + }, + { + "epoch": 0.97, + "grad_norm": 1.8300272259967298, + "learning_rate": 4.930949478529346e-08, + "loss": 0.9428, + "step": 41144 + }, + { + "epoch": 0.97, + "grad_norm": 1.9931810755379729, + "learning_rate": 4.923384183175284e-08, + "loss": 0.8316, + "step": 41145 + }, + { + "epoch": 0.97, + "grad_norm": 1.131289895739574, + "learning_rate": 4.9158246814723367e-08, + "loss": 0.9253, + "step": 41146 + }, + { + "epoch": 0.97, + "grad_norm": 2.1780790187770824, + "learning_rate": 4.90827097346458e-08, + "loss": 0.9546, + "step": 41147 + }, + { + "epoch": 0.97, + "grad_norm": 2.3256510384487745, + "learning_rate": 4.9007230591957555e-08, + "loss": 0.9746, + "step": 41148 + }, + { + "epoch": 0.97, + "grad_norm": 1.9165030551106694, + "learning_rate": 4.8931809387100515e-08, + "loss": 0.9975, + "step": 41149 + }, + { + "epoch": 0.97, + "grad_norm": 2.352441189785834, + "learning_rate": 4.8856446120513215e-08, + "loss": 1.0063, + "step": 41150 + }, + { + "epoch": 0.97, + "grad_norm": 1.0888332096494886, + "learning_rate": 4.8781140792633074e-08, + "loss": 0.8975, + "step": 41151 + }, + { + "epoch": 0.97, + "grad_norm": 2.2109359220559788, + "learning_rate": 4.870589340390086e-08, + "loss": 0.911, + "step": 41152 + }, + { + "epoch": 0.97, + "grad_norm": 2.196489514718512, + "learning_rate": 4.863070395475289e-08, + "loss": 0.9102, + "step": 41153 + }, + { + "epoch": 0.97, + "grad_norm": 1.9713563285477311, + "learning_rate": 4.85555724456277e-08, + "loss": 0.8969, + "step": 41154 + }, + { + "epoch": 0.97, + "grad_norm": 2.7688763929420457, + "learning_rate": 4.848049887696271e-08, + "loss": 0.8478, + "step": 41155 + }, + { + "epoch": 0.97, + "grad_norm": 1.1129264821540317, + "learning_rate": 4.8405483249194255e-08, + "loss": 0.9095, + "step": 41156 + }, + { + "epoch": 0.97, + "grad_norm": 1.1353226274794461, + "learning_rate": 4.8330525562759745e-08, + "loss": 0.9391, + "step": 41157 + }, + { + "epoch": 0.97, + "grad_norm": 2.078731279583672, + "learning_rate": 4.8255625818096616e-08, + "loss": 0.8997, + "step": 41158 + }, + { + "epoch": 0.97, + "grad_norm": 2.0069445740748035, + "learning_rate": 4.818078401563897e-08, + "loss": 0.8752, + "step": 41159 + }, + { + "epoch": 0.97, + "grad_norm": 1.9392331407721661, + "learning_rate": 4.810600015582312e-08, + "loss": 0.9908, + "step": 41160 + }, + { + "epoch": 0.97, + "grad_norm": 1.9618574087874154, + "learning_rate": 4.803127423908538e-08, + "loss": 1.0396, + "step": 41161 + }, + { + "epoch": 0.97, + "grad_norm": 1.0180682318371475, + "learning_rate": 4.795660626586096e-08, + "loss": 0.8946, + "step": 41162 + }, + { + "epoch": 0.97, + "grad_norm": 1.9261942309660582, + "learning_rate": 4.788199623658285e-08, + "loss": 0.8432, + "step": 41163 + }, + { + "epoch": 0.97, + "grad_norm": 1.1502246362511213, + "learning_rate": 4.7807444151687363e-08, + "loss": 1.0733, + "step": 41164 + }, + { + "epoch": 0.97, + "grad_norm": 1.9807053394448753, + "learning_rate": 4.773295001160638e-08, + "loss": 1.1039, + "step": 41165 + }, + { + "epoch": 0.97, + "grad_norm": 1.0378106677246806, + "learning_rate": 4.765851381677622e-08, + "loss": 0.8456, + "step": 41166 + }, + { + "epoch": 0.97, + "grad_norm": 1.8727327120374802, + "learning_rate": 4.758413556762986e-08, + "loss": 0.9096, + "step": 41167 + }, + { + "epoch": 0.97, + "grad_norm": 1.9354852330720926, + "learning_rate": 4.750981526459808e-08, + "loss": 1.0838, + "step": 41168 + }, + { + "epoch": 0.97, + "grad_norm": 1.864699032737296, + "learning_rate": 4.7435552908114966e-08, + "loss": 0.8671, + "step": 41169 + }, + { + "epoch": 0.97, + "grad_norm": 1.9763498798564028, + "learning_rate": 4.7361348498613516e-08, + "loss": 0.9436, + "step": 41170 + }, + { + "epoch": 0.97, + "grad_norm": 1.959976740656019, + "learning_rate": 4.728720203652559e-08, + "loss": 1.034, + "step": 41171 + }, + { + "epoch": 0.97, + "grad_norm": 1.8944477636829005, + "learning_rate": 4.721311352228197e-08, + "loss": 1.0251, + "step": 41172 + }, + { + "epoch": 0.97, + "grad_norm": 2.108839443221849, + "learning_rate": 4.713908295631564e-08, + "loss": 0.9777, + "step": 41173 + }, + { + "epoch": 0.97, + "grad_norm": 1.8754710965981893, + "learning_rate": 4.706511033905514e-08, + "loss": 0.9975, + "step": 41174 + }, + { + "epoch": 0.97, + "grad_norm": 2.0897919159882776, + "learning_rate": 4.6991195670932355e-08, + "loss": 1.0109, + "step": 41175 + }, + { + "epoch": 0.97, + "grad_norm": 2.0407171534226625, + "learning_rate": 4.6917338952380266e-08, + "loss": 1.0969, + "step": 41176 + }, + { + "epoch": 0.97, + "grad_norm": 1.9643995796625435, + "learning_rate": 4.684354018382409e-08, + "loss": 0.936, + "step": 41177 + }, + { + "epoch": 0.97, + "grad_norm": 2.387252837217781, + "learning_rate": 4.6769799365696814e-08, + "loss": 0.9247, + "step": 41178 + }, + { + "epoch": 0.97, + "grad_norm": 2.1942396279681096, + "learning_rate": 4.669611649842698e-08, + "loss": 1.0605, + "step": 41179 + }, + { + "epoch": 0.97, + "grad_norm": 1.887625404238626, + "learning_rate": 4.662249158244425e-08, + "loss": 0.8745, + "step": 41180 + }, + { + "epoch": 0.97, + "grad_norm": 1.9218008586298407, + "learning_rate": 4.6548924618176064e-08, + "loss": 0.9562, + "step": 41181 + }, + { + "epoch": 0.97, + "grad_norm": 2.0372646298683836, + "learning_rate": 4.6475415606049845e-08, + "loss": 1.1433, + "step": 41182 + }, + { + "epoch": 0.97, + "grad_norm": 2.037743890891335, + "learning_rate": 4.640196454649637e-08, + "loss": 1.0713, + "step": 41183 + }, + { + "epoch": 0.97, + "grad_norm": 1.9331663370967926, + "learning_rate": 4.632857143994196e-08, + "loss": 0.9581, + "step": 41184 + }, + { + "epoch": 0.97, + "grad_norm": 2.1458743447105464, + "learning_rate": 4.625523628681405e-08, + "loss": 0.9068, + "step": 41185 + }, + { + "epoch": 0.97, + "grad_norm": 2.162155665506876, + "learning_rate": 4.6181959087538975e-08, + "loss": 1.0156, + "step": 41186 + }, + { + "epoch": 0.97, + "grad_norm": 2.249205787089015, + "learning_rate": 4.610873984254416e-08, + "loss": 1.0157, + "step": 41187 + }, + { + "epoch": 0.97, + "grad_norm": 2.1222671481356463, + "learning_rate": 4.603557855225482e-08, + "loss": 0.9067, + "step": 41188 + }, + { + "epoch": 0.97, + "grad_norm": 1.964121058305403, + "learning_rate": 4.59624752170984e-08, + "loss": 0.9902, + "step": 41189 + }, + { + "epoch": 0.97, + "grad_norm": 1.0294164190280584, + "learning_rate": 4.588942983750011e-08, + "loss": 0.8517, + "step": 41190 + }, + { + "epoch": 0.97, + "grad_norm": 2.408068599118565, + "learning_rate": 4.5816442413884055e-08, + "loss": 0.9976, + "step": 41191 + }, + { + "epoch": 0.97, + "grad_norm": 2.172090959403654, + "learning_rate": 4.574351294667656e-08, + "loss": 1.0319, + "step": 41192 + }, + { + "epoch": 0.97, + "grad_norm": 2.014873563135265, + "learning_rate": 4.5670641436301735e-08, + "loss": 0.9137, + "step": 41193 + }, + { + "epoch": 0.97, + "grad_norm": 1.9900657269532265, + "learning_rate": 4.559782788318479e-08, + "loss": 0.9987, + "step": 41194 + }, + { + "epoch": 0.97, + "grad_norm": 2.0914525599940204, + "learning_rate": 4.552507228774761e-08, + "loss": 0.8497, + "step": 41195 + }, + { + "epoch": 0.97, + "grad_norm": 2.1119874819882116, + "learning_rate": 4.545237465041541e-08, + "loss": 1.0779, + "step": 41196 + }, + { + "epoch": 0.97, + "grad_norm": 2.105762647842551, + "learning_rate": 4.5379734971611186e-08, + "loss": 0.9327, + "step": 41197 + }, + { + "epoch": 0.97, + "grad_norm": 1.8746296195194796, + "learning_rate": 4.530715325175794e-08, + "loss": 0.9387, + "step": 41198 + }, + { + "epoch": 0.97, + "grad_norm": 2.1652687984980554, + "learning_rate": 4.5234629491277546e-08, + "loss": 0.8778, + "step": 41199 + }, + { + "epoch": 0.97, + "grad_norm": 2.8173742746227086, + "learning_rate": 4.516216369059301e-08, + "loss": 1.0304, + "step": 41200 + }, + { + "epoch": 0.97, + "grad_norm": 2.181067586353208, + "learning_rate": 4.5089755850126206e-08, + "loss": 1.1592, + "step": 41201 + }, + { + "epoch": 0.97, + "grad_norm": 1.9658415471544182, + "learning_rate": 4.501740597029791e-08, + "loss": 1.0604, + "step": 41202 + }, + { + "epoch": 0.97, + "grad_norm": 2.3026031038328694, + "learning_rate": 4.494511405153002e-08, + "loss": 0.9391, + "step": 41203 + }, + { + "epoch": 0.97, + "grad_norm": 2.218673245624408, + "learning_rate": 4.4872880094244395e-08, + "loss": 1.0191, + "step": 41204 + }, + { + "epoch": 0.97, + "grad_norm": 2.0441215757406, + "learning_rate": 4.4800704098859614e-08, + "loss": 0.9031, + "step": 41205 + }, + { + "epoch": 0.97, + "grad_norm": 1.1862607291284288, + "learning_rate": 4.472858606579755e-08, + "loss": 0.9705, + "step": 41206 + }, + { + "epoch": 0.97, + "grad_norm": 1.9395707735573882, + "learning_rate": 4.465652599547676e-08, + "loss": 0.9669, + "step": 41207 + }, + { + "epoch": 0.97, + "grad_norm": 2.254121130719619, + "learning_rate": 4.4584523888318023e-08, + "loss": 0.8285, + "step": 41208 + }, + { + "epoch": 0.97, + "grad_norm": 2.0532427795299464, + "learning_rate": 4.451257974473988e-08, + "loss": 1.1288, + "step": 41209 + }, + { + "epoch": 0.97, + "grad_norm": 2.1206626912366953, + "learning_rate": 4.4440693565160895e-08, + "loss": 1.0027, + "step": 41210 + }, + { + "epoch": 0.97, + "grad_norm": 2.040246458773732, + "learning_rate": 4.4368865350000736e-08, + "loss": 1.0443, + "step": 41211 + }, + { + "epoch": 0.97, + "grad_norm": 2.0089124681994868, + "learning_rate": 4.4297095099676836e-08, + "loss": 0.9903, + "step": 41212 + }, + { + "epoch": 0.97, + "grad_norm": 2.00324587660428, + "learning_rate": 4.4225382814606643e-08, + "loss": 1.0055, + "step": 41213 + }, + { + "epoch": 0.97, + "grad_norm": 1.9505483425264885, + "learning_rate": 4.415372849520871e-08, + "loss": 0.7837, + "step": 41214 + }, + { + "epoch": 0.97, + "grad_norm": 1.8246156765590722, + "learning_rate": 4.408213214189938e-08, + "loss": 0.9657, + "step": 41215 + }, + { + "epoch": 0.97, + "grad_norm": 2.0163830933216667, + "learning_rate": 4.4010593755096086e-08, + "loss": 1.0274, + "step": 41216 + }, + { + "epoch": 0.97, + "grad_norm": 1.8649998694694188, + "learning_rate": 4.393911333521517e-08, + "loss": 0.8476, + "step": 41217 + }, + { + "epoch": 0.97, + "grad_norm": 2.0725887707393142, + "learning_rate": 4.386769088267295e-08, + "loss": 1.0327, + "step": 41218 + }, + { + "epoch": 0.97, + "grad_norm": 2.2166416934183255, + "learning_rate": 4.379632639788467e-08, + "loss": 0.9842, + "step": 41219 + }, + { + "epoch": 0.97, + "grad_norm": 2.094049183043268, + "learning_rate": 4.372501988126665e-08, + "loss": 1.0277, + "step": 41220 + }, + { + "epoch": 0.97, + "grad_norm": 1.8113903868210228, + "learning_rate": 4.365377133323301e-08, + "loss": 0.9206, + "step": 41221 + }, + { + "epoch": 0.97, + "grad_norm": 1.1993078119525873, + "learning_rate": 4.3582580754200074e-08, + "loss": 0.984, + "step": 41222 + }, + { + "epoch": 0.97, + "grad_norm": 2.064097132661172, + "learning_rate": 4.3511448144580857e-08, + "loss": 0.8322, + "step": 41223 + }, + { + "epoch": 0.97, + "grad_norm": 1.9540694473964733, + "learning_rate": 4.3440373504790576e-08, + "loss": 0.9711, + "step": 41224 + }, + { + "epoch": 0.97, + "grad_norm": 2.042086573421152, + "learning_rate": 4.336935683524224e-08, + "loss": 1.0553, + "step": 41225 + }, + { + "epoch": 0.97, + "grad_norm": 2.495097116214177, + "learning_rate": 4.3298398136349954e-08, + "loss": 1.0285, + "step": 41226 + }, + { + "epoch": 0.97, + "grad_norm": 2.095343867511807, + "learning_rate": 4.322749740852672e-08, + "loss": 0.8728, + "step": 41227 + }, + { + "epoch": 0.97, + "grad_norm": 2.024484272966274, + "learning_rate": 4.315665465218555e-08, + "loss": 0.9594, + "step": 41228 + }, + { + "epoch": 0.97, + "grad_norm": 1.9809300139566202, + "learning_rate": 4.308586986773833e-08, + "loss": 1.0879, + "step": 41229 + }, + { + "epoch": 0.97, + "grad_norm": 1.0788636669238023, + "learning_rate": 4.3015143055596954e-08, + "loss": 0.9661, + "step": 41230 + }, + { + "epoch": 0.97, + "grad_norm": 1.9717132858707873, + "learning_rate": 4.2944474216174425e-08, + "loss": 1.0803, + "step": 41231 + }, + { + "epoch": 0.97, + "grad_norm": 2.021668588035503, + "learning_rate": 4.287386334988153e-08, + "loss": 0.9181, + "step": 41232 + }, + { + "epoch": 0.97, + "grad_norm": 1.9974644751138377, + "learning_rate": 4.2803310457129046e-08, + "loss": 1.0104, + "step": 41233 + }, + { + "epoch": 0.97, + "grad_norm": 2.113943785853613, + "learning_rate": 4.273281553832775e-08, + "loss": 0.9179, + "step": 41234 + }, + { + "epoch": 0.97, + "grad_norm": 2.119003671844043, + "learning_rate": 4.266237859388955e-08, + "loss": 0.9166, + "step": 41235 + }, + { + "epoch": 0.97, + "grad_norm": 2.187751037857955, + "learning_rate": 4.2591999624223e-08, + "loss": 0.9625, + "step": 41236 + }, + { + "epoch": 0.97, + "grad_norm": 1.8600759357111216, + "learning_rate": 4.2521678629737774e-08, + "loss": 0.8682, + "step": 41237 + }, + { + "epoch": 0.97, + "grad_norm": 1.0892923021319723, + "learning_rate": 4.245141561084354e-08, + "loss": 0.9059, + "step": 41238 + }, + { + "epoch": 0.97, + "grad_norm": 1.9614821317430249, + "learning_rate": 4.238121056794997e-08, + "loss": 0.8869, + "step": 41239 + }, + { + "epoch": 0.97, + "grad_norm": 2.7371061671913157, + "learning_rate": 4.231106350146563e-08, + "loss": 0.9927, + "step": 41240 + }, + { + "epoch": 0.97, + "grad_norm": 2.0414159069697853, + "learning_rate": 4.224097441179908e-08, + "loss": 0.9292, + "step": 41241 + }, + { + "epoch": 0.97, + "grad_norm": 2.0905401879844137, + "learning_rate": 4.217094329935778e-08, + "loss": 0.9377, + "step": 41242 + }, + { + "epoch": 0.97, + "grad_norm": 2.0494501056871752, + "learning_rate": 4.210097016455028e-08, + "loss": 0.9928, + "step": 41243 + }, + { + "epoch": 0.97, + "grad_norm": 2.1754493631470218, + "learning_rate": 4.2031055007784036e-08, + "loss": 0.8547, + "step": 41244 + }, + { + "epoch": 0.97, + "grad_norm": 1.919337363634155, + "learning_rate": 4.196119782946429e-08, + "loss": 0.9535, + "step": 41245 + }, + { + "epoch": 0.97, + "grad_norm": 2.14653250870915, + "learning_rate": 4.1891398629999583e-08, + "loss": 0.9134, + "step": 41246 + }, + { + "epoch": 0.97, + "grad_norm": 2.37780747383525, + "learning_rate": 4.182165740979627e-08, + "loss": 1.0737, + "step": 41247 + }, + { + "epoch": 0.97, + "grad_norm": 2.1300928273787525, + "learning_rate": 4.175197416925958e-08, + "loss": 0.9335, + "step": 41248 + }, + { + "epoch": 0.97, + "grad_norm": 2.0641591508784134, + "learning_rate": 4.168234890879586e-08, + "loss": 1.1098, + "step": 41249 + }, + { + "epoch": 0.97, + "grad_norm": 1.9246154698320288, + "learning_rate": 4.161278162881033e-08, + "loss": 0.8715, + "step": 41250 + }, + { + "epoch": 0.97, + "grad_norm": 1.792620435059746, + "learning_rate": 4.1543272329707116e-08, + "loss": 0.9709, + "step": 41251 + }, + { + "epoch": 0.97, + "grad_norm": 2.085076886443834, + "learning_rate": 4.1473821011892565e-08, + "loss": 1.0513, + "step": 41252 + }, + { + "epoch": 0.97, + "grad_norm": 1.004833965856692, + "learning_rate": 4.14044276757708e-08, + "loss": 0.886, + "step": 41253 + }, + { + "epoch": 0.97, + "grad_norm": 2.2282655345343705, + "learning_rate": 4.133509232174482e-08, + "loss": 1.0034, + "step": 41254 + }, + { + "epoch": 0.97, + "grad_norm": 2.0344698617026356, + "learning_rate": 4.126581495021764e-08, + "loss": 1.0738, + "step": 41255 + }, + { + "epoch": 0.97, + "grad_norm": 1.8375718519843214, + "learning_rate": 4.119659556159561e-08, + "loss": 0.822, + "step": 41256 + }, + { + "epoch": 0.97, + "grad_norm": 2.2012482305575607, + "learning_rate": 4.11274341562784e-08, + "loss": 0.8998, + "step": 41257 + }, + { + "epoch": 0.97, + "grad_norm": 1.8993691651154492, + "learning_rate": 4.105833073467125e-08, + "loss": 0.9018, + "step": 41258 + }, + { + "epoch": 0.97, + "grad_norm": 2.1139418419845537, + "learning_rate": 4.0989285297174944e-08, + "loss": 1.1114, + "step": 41259 + }, + { + "epoch": 0.97, + "grad_norm": 2.057983352278282, + "learning_rate": 4.092029784419138e-08, + "loss": 0.9211, + "step": 41260 + }, + { + "epoch": 0.97, + "grad_norm": 2.2588153146759447, + "learning_rate": 4.085136837612358e-08, + "loss": 0.9116, + "step": 41261 + }, + { + "epoch": 0.97, + "grad_norm": 2.013825963356541, + "learning_rate": 4.0782496893372327e-08, + "loss": 1.0127, + "step": 41262 + }, + { + "epoch": 0.97, + "grad_norm": 2.123526503211325, + "learning_rate": 4.071368339633841e-08, + "loss": 0.9997, + "step": 41263 + }, + { + "epoch": 0.97, + "grad_norm": 2.2092347354192206, + "learning_rate": 4.064492788542152e-08, + "loss": 0.9483, + "step": 41264 + }, + { + "epoch": 0.97, + "grad_norm": 2.273110690358475, + "learning_rate": 4.057623036102465e-08, + "loss": 0.9008, + "step": 41265 + }, + { + "epoch": 0.97, + "grad_norm": 2.174020243112796, + "learning_rate": 4.0507590823544164e-08, + "loss": 0.9687, + "step": 41266 + }, + { + "epoch": 0.97, + "grad_norm": 2.622000034285526, + "learning_rate": 4.043900927338307e-08, + "loss": 0.8943, + "step": 41267 + }, + { + "epoch": 0.97, + "grad_norm": 2.5851861881175147, + "learning_rate": 4.037048571093771e-08, + "loss": 1.0607, + "step": 41268 + }, + { + "epoch": 0.97, + "grad_norm": 1.9179921559699349, + "learning_rate": 4.0302020136609995e-08, + "loss": 1.0064, + "step": 41269 + }, + { + "epoch": 0.97, + "grad_norm": 2.1246016897479674, + "learning_rate": 4.0233612550796273e-08, + "loss": 0.8302, + "step": 41270 + }, + { + "epoch": 0.97, + "grad_norm": 1.9183143165538328, + "learning_rate": 4.0165262953896225e-08, + "loss": 1.0191, + "step": 41271 + }, + { + "epoch": 0.97, + "grad_norm": 1.1520404196212843, + "learning_rate": 4.00969713463073e-08, + "loss": 0.8648, + "step": 41272 + }, + { + "epoch": 0.97, + "grad_norm": 2.7335183612885072, + "learning_rate": 4.002873772842586e-08, + "loss": 1.1201, + "step": 41273 + }, + { + "epoch": 0.97, + "grad_norm": 1.981566332613116, + "learning_rate": 3.996056210065158e-08, + "loss": 0.9303, + "step": 41274 + }, + { + "epoch": 0.97, + "grad_norm": 2.1990643270475343, + "learning_rate": 3.989244446338081e-08, + "loss": 1.1184, + "step": 41275 + }, + { + "epoch": 0.97, + "grad_norm": 2.0671978650640566, + "learning_rate": 3.982438481700768e-08, + "loss": 1.112, + "step": 41276 + }, + { + "epoch": 0.97, + "grad_norm": 1.8806683085174083, + "learning_rate": 3.975638316193076e-08, + "loss": 0.9134, + "step": 41277 + }, + { + "epoch": 0.97, + "grad_norm": 3.156340853469642, + "learning_rate": 3.96884394985475e-08, + "loss": 0.9292, + "step": 41278 + }, + { + "epoch": 0.97, + "grad_norm": 1.9975709274879119, + "learning_rate": 3.962055382724983e-08, + "loss": 1.0665, + "step": 41279 + }, + { + "epoch": 0.97, + "grad_norm": 1.888953055469384, + "learning_rate": 3.955272614843408e-08, + "loss": 0.927, + "step": 41280 + }, + { + "epoch": 0.97, + "grad_norm": 1.952350042379939, + "learning_rate": 3.948495646249662e-08, + "loss": 0.86, + "step": 41281 + }, + { + "epoch": 0.97, + "grad_norm": 1.8810162519677782, + "learning_rate": 3.941724476983155e-08, + "loss": 0.8638, + "step": 41282 + }, + { + "epoch": 0.97, + "grad_norm": 1.7929293832971074, + "learning_rate": 3.934959107083303e-08, + "loss": 0.906, + "step": 41283 + }, + { + "epoch": 0.97, + "grad_norm": 1.07280820983452, + "learning_rate": 3.9281995365892944e-08, + "loss": 0.9555, + "step": 41284 + }, + { + "epoch": 0.97, + "grad_norm": 1.9409800105105643, + "learning_rate": 3.921445765540877e-08, + "loss": 0.8782, + "step": 41285 + }, + { + "epoch": 0.97, + "grad_norm": 2.0039484721842395, + "learning_rate": 3.914697793977018e-08, + "loss": 0.8894, + "step": 41286 + }, + { + "epoch": 0.97, + "grad_norm": 2.1320670057686852, + "learning_rate": 3.9079556219371317e-08, + "loss": 0.9909, + "step": 41287 + }, + { + "epoch": 0.97, + "grad_norm": 2.1539166583813567, + "learning_rate": 3.9012192494605197e-08, + "loss": 1.0105, + "step": 41288 + }, + { + "epoch": 0.97, + "grad_norm": 2.495713800423736, + "learning_rate": 3.894488676586372e-08, + "loss": 0.9108, + "step": 41289 + }, + { + "epoch": 0.97, + "grad_norm": 1.9118106259674985, + "learning_rate": 3.8877639033538805e-08, + "loss": 0.9151, + "step": 41290 + }, + { + "epoch": 0.97, + "grad_norm": 1.9295065670479647, + "learning_rate": 3.881044929802236e-08, + "loss": 0.9252, + "step": 41291 + }, + { + "epoch": 0.97, + "grad_norm": 2.2304203478414144, + "learning_rate": 3.8743317559705174e-08, + "loss": 1.0139, + "step": 41292 + }, + { + "epoch": 0.97, + "grad_norm": 1.9961659641533123, + "learning_rate": 3.867624381897694e-08, + "loss": 0.8408, + "step": 41293 + }, + { + "epoch": 0.97, + "grad_norm": 2.128250442338899, + "learning_rate": 3.8609228076229576e-08, + "loss": 0.9616, + "step": 41294 + }, + { + "epoch": 0.97, + "grad_norm": 1.8894869159942298, + "learning_rate": 3.854227033185387e-08, + "loss": 0.9668, + "step": 41295 + }, + { + "epoch": 0.97, + "grad_norm": 2.0116470424825366, + "learning_rate": 3.84753705862384e-08, + "loss": 0.7587, + "step": 41296 + }, + { + "epoch": 0.97, + "grad_norm": 1.9789793389247523, + "learning_rate": 3.840852883977286e-08, + "loss": 0.9681, + "step": 41297 + }, + { + "epoch": 0.97, + "grad_norm": 1.9639313189166208, + "learning_rate": 3.8341745092846936e-08, + "loss": 1.0878, + "step": 41298 + }, + { + "epoch": 0.97, + "grad_norm": 1.9795782440328955, + "learning_rate": 3.827501934584921e-08, + "loss": 0.9297, + "step": 41299 + }, + { + "epoch": 0.97, + "grad_norm": 1.8179731714348988, + "learning_rate": 3.8208351599167135e-08, + "loss": 0.9461, + "step": 41300 + }, + { + "epoch": 0.97, + "grad_norm": 1.930595102523357, + "learning_rate": 3.814174185319153e-08, + "loss": 0.9671, + "step": 41301 + }, + { + "epoch": 0.97, + "grad_norm": 1.924158245291021, + "learning_rate": 3.8075190108307624e-08, + "loss": 0.9058, + "step": 41302 + }, + { + "epoch": 0.97, + "grad_norm": 1.9662759983978755, + "learning_rate": 3.8008696364904006e-08, + "loss": 0.887, + "step": 41303 + }, + { + "epoch": 0.97, + "grad_norm": 2.0714812702149934, + "learning_rate": 3.7942260623367035e-08, + "loss": 0.9797, + "step": 41304 + }, + { + "epoch": 0.97, + "grad_norm": 1.220972872493958, + "learning_rate": 3.787588288408528e-08, + "loss": 0.9055, + "step": 41305 + }, + { + "epoch": 0.97, + "grad_norm": 2.428763617368733, + "learning_rate": 3.7809563147443993e-08, + "loss": 0.9805, + "step": 41306 + }, + { + "epoch": 0.97, + "grad_norm": 2.0212490938851584, + "learning_rate": 3.7743301413828424e-08, + "loss": 0.9196, + "step": 41307 + }, + { + "epoch": 0.97, + "grad_norm": 1.1205994041752487, + "learning_rate": 3.767709768362604e-08, + "loss": 0.9805, + "step": 41308 + }, + { + "epoch": 0.97, + "grad_norm": 1.8517542393461512, + "learning_rate": 3.7610951957220975e-08, + "loss": 0.9608, + "step": 41309 + }, + { + "epoch": 0.97, + "grad_norm": 2.012477051997628, + "learning_rate": 3.754486423499959e-08, + "loss": 1.035, + "step": 41310 + }, + { + "epoch": 0.97, + "grad_norm": 1.1017035064760583, + "learning_rate": 3.747883451734602e-08, + "loss": 0.9633, + "step": 41311 + }, + { + "epoch": 0.97, + "grad_norm": 2.001068190165823, + "learning_rate": 3.7412862804645514e-08, + "loss": 1.0545, + "step": 41312 + }, + { + "epoch": 0.97, + "grad_norm": 2.4512001072067027, + "learning_rate": 3.7346949097279984e-08, + "loss": 0.9718, + "step": 41313 + }, + { + "epoch": 0.97, + "grad_norm": 1.9261634347196355, + "learning_rate": 3.728109339563579e-08, + "loss": 1.0218, + "step": 41314 + }, + { + "epoch": 0.97, + "grad_norm": 2.2300745099758865, + "learning_rate": 3.721529570009486e-08, + "loss": 0.9957, + "step": 41315 + }, + { + "epoch": 0.97, + "grad_norm": 2.0065776515457006, + "learning_rate": 3.71495560110402e-08, + "loss": 1.0512, + "step": 41316 + }, + { + "epoch": 0.97, + "grad_norm": 2.102703940034689, + "learning_rate": 3.7083874328855965e-08, + "loss": 1.0281, + "step": 41317 + }, + { + "epoch": 0.97, + "grad_norm": 3.3241567528868283, + "learning_rate": 3.701825065392184e-08, + "loss": 1.1253, + "step": 41318 + }, + { + "epoch": 0.97, + "grad_norm": 2.1441652360411654, + "learning_rate": 3.6952684986623076e-08, + "loss": 1.012, + "step": 41319 + }, + { + "epoch": 0.97, + "grad_norm": 2.295252461464565, + "learning_rate": 3.6887177327340486e-08, + "loss": 0.9673, + "step": 41320 + }, + { + "epoch": 0.97, + "grad_norm": 1.9557360373134838, + "learning_rate": 3.682172767645375e-08, + "loss": 0.9237, + "step": 41321 + }, + { + "epoch": 0.97, + "grad_norm": 1.8517026638822114, + "learning_rate": 3.675633603434592e-08, + "loss": 0.9156, + "step": 41322 + }, + { + "epoch": 0.97, + "grad_norm": 1.8520427462813105, + "learning_rate": 3.6691002401396666e-08, + "loss": 0.914, + "step": 41323 + }, + { + "epoch": 0.97, + "grad_norm": 2.122699401604289, + "learning_rate": 3.662572677798682e-08, + "loss": 0.9291, + "step": 41324 + }, + { + "epoch": 0.97, + "grad_norm": 1.7509736905045155, + "learning_rate": 3.656050916449716e-08, + "loss": 1.1699, + "step": 41325 + }, + { + "epoch": 0.97, + "grad_norm": 1.8885001420193086, + "learning_rate": 3.649534956130518e-08, + "loss": 0.9646, + "step": 41326 + }, + { + "epoch": 0.97, + "grad_norm": 1.8497363019773871, + "learning_rate": 3.64302479687928e-08, + "loss": 0.9473, + "step": 41327 + }, + { + "epoch": 0.97, + "grad_norm": 1.122590475015459, + "learning_rate": 3.636520438733748e-08, + "loss": 0.8421, + "step": 41328 + }, + { + "epoch": 0.97, + "grad_norm": 1.8572241068522015, + "learning_rate": 3.6300218817317825e-08, + "loss": 0.9546, + "step": 41329 + }, + { + "epoch": 0.97, + "grad_norm": 1.9256770750686332, + "learning_rate": 3.6235291259113516e-08, + "loss": 0.7857, + "step": 41330 + }, + { + "epoch": 0.97, + "grad_norm": 1.9925495488397293, + "learning_rate": 3.6170421713102034e-08, + "loss": 0.9798, + "step": 41331 + }, + { + "epoch": 0.97, + "grad_norm": 2.180977438695981, + "learning_rate": 3.6105610179659745e-08, + "loss": 1.016, + "step": 41332 + }, + { + "epoch": 0.97, + "grad_norm": 2.0659415748156347, + "learning_rate": 3.6040856659166345e-08, + "loss": 0.9618, + "step": 41333 + }, + { + "epoch": 0.97, + "grad_norm": 1.8369646239077966, + "learning_rate": 3.597616115199709e-08, + "loss": 1.0058, + "step": 41334 + }, + { + "epoch": 0.97, + "grad_norm": 1.9780430408232554, + "learning_rate": 3.5911523658528347e-08, + "loss": 0.9202, + "step": 41335 + }, + { + "epoch": 0.97, + "grad_norm": 1.9484839036087842, + "learning_rate": 3.58469441791387e-08, + "loss": 0.904, + "step": 41336 + }, + { + "epoch": 0.97, + "grad_norm": 2.196674500529902, + "learning_rate": 3.578242271420229e-08, + "loss": 0.8885, + "step": 41337 + }, + { + "epoch": 0.97, + "grad_norm": 2.0130876668169644, + "learning_rate": 3.571795926409549e-08, + "loss": 1.0266, + "step": 41338 + }, + { + "epoch": 0.97, + "grad_norm": 1.0934792644411955, + "learning_rate": 3.5653553829192444e-08, + "loss": 0.9422, + "step": 41339 + }, + { + "epoch": 0.97, + "grad_norm": 2.323040191881481, + "learning_rate": 3.558920640986952e-08, + "loss": 0.9067, + "step": 41340 + }, + { + "epoch": 0.97, + "grad_norm": 1.971063062124296, + "learning_rate": 3.5524917006500844e-08, + "loss": 1.0115, + "step": 41341 + }, + { + "epoch": 0.97, + "grad_norm": 2.0241062511216086, + "learning_rate": 3.546068561946059e-08, + "loss": 0.9005, + "step": 41342 + }, + { + "epoch": 0.97, + "grad_norm": 1.9082289514506994, + "learning_rate": 3.5396512249123996e-08, + "loss": 0.9497, + "step": 41343 + }, + { + "epoch": 0.97, + "grad_norm": 1.9964845389355073, + "learning_rate": 3.533239689586188e-08, + "loss": 0.9876, + "step": 41344 + }, + { + "epoch": 0.97, + "grad_norm": 1.9972921139281847, + "learning_rate": 3.52683395600506e-08, + "loss": 1.0505, + "step": 41345 + }, + { + "epoch": 0.97, + "grad_norm": 1.8312205170879843, + "learning_rate": 3.5204340242062096e-08, + "loss": 0.9876, + "step": 41346 + }, + { + "epoch": 0.97, + "grad_norm": 1.8405879002986907, + "learning_rate": 3.5140398942268286e-08, + "loss": 0.9619, + "step": 41347 + }, + { + "epoch": 0.97, + "grad_norm": 2.0288422088271187, + "learning_rate": 3.50765156610422e-08, + "loss": 1.0067, + "step": 41348 + }, + { + "epoch": 0.97, + "grad_norm": 1.9180333239295866, + "learning_rate": 3.5012690398754654e-08, + "loss": 1.0409, + "step": 41349 + }, + { + "epoch": 0.97, + "grad_norm": 1.8127275187439937, + "learning_rate": 3.494892315577869e-08, + "loss": 0.8522, + "step": 41350 + }, + { + "epoch": 0.97, + "grad_norm": 2.866563653120206, + "learning_rate": 3.488521393248401e-08, + "loss": 1.0021, + "step": 41351 + }, + { + "epoch": 0.97, + "grad_norm": 2.0589085344806133, + "learning_rate": 3.4821562729243644e-08, + "loss": 0.8397, + "step": 41352 + }, + { + "epoch": 0.97, + "grad_norm": 1.988498172725951, + "learning_rate": 3.475796954642729e-08, + "loss": 1.0311, + "step": 41353 + }, + { + "epoch": 0.97, + "grad_norm": 2.0849728034528283, + "learning_rate": 3.469443438440467e-08, + "loss": 0.8831, + "step": 41354 + }, + { + "epoch": 0.97, + "grad_norm": 2.001899525692466, + "learning_rate": 3.463095724354548e-08, + "loss": 1.0436, + "step": 41355 + }, + { + "epoch": 0.97, + "grad_norm": 2.156044252334784, + "learning_rate": 3.456753812421943e-08, + "loss": 1.0072, + "step": 41356 + }, + { + "epoch": 0.97, + "grad_norm": 1.913475243704344, + "learning_rate": 3.450417702679731e-08, + "loss": 0.8974, + "step": 41357 + }, + { + "epoch": 0.97, + "grad_norm": 2.0944490942552854, + "learning_rate": 3.444087395164664e-08, + "loss": 0.8463, + "step": 41358 + }, + { + "epoch": 0.97, + "grad_norm": 2.424666410287431, + "learning_rate": 3.437762889913598e-08, + "loss": 0.9614, + "step": 41359 + }, + { + "epoch": 0.97, + "grad_norm": 2.045291596002393, + "learning_rate": 3.4314441869633954e-08, + "loss": 0.9786, + "step": 41360 + }, + { + "epoch": 0.97, + "grad_norm": 1.9345552097493592, + "learning_rate": 3.425131286350802e-08, + "loss": 1.1171, + "step": 41361 + }, + { + "epoch": 0.97, + "grad_norm": 2.0387384889854814, + "learning_rate": 3.418824188112679e-08, + "loss": 0.9277, + "step": 41362 + }, + { + "epoch": 0.97, + "grad_norm": 1.1170601479388063, + "learning_rate": 3.412522892285663e-08, + "loss": 0.9214, + "step": 41363 + }, + { + "epoch": 0.97, + "grad_norm": 1.0942582938650733, + "learning_rate": 3.4062273989063923e-08, + "loss": 0.9967, + "step": 41364 + }, + { + "epoch": 0.97, + "grad_norm": 1.9887069589398014, + "learning_rate": 3.3999377080115026e-08, + "loss": 0.9299, + "step": 41365 + }, + { + "epoch": 0.97, + "grad_norm": 2.7421053700225055, + "learning_rate": 3.393653819637854e-08, + "loss": 1.1082, + "step": 41366 + }, + { + "epoch": 0.97, + "grad_norm": 1.883009877360549, + "learning_rate": 3.387375733821863e-08, + "loss": 0.9118, + "step": 41367 + }, + { + "epoch": 0.97, + "grad_norm": 2.0012872193779234, + "learning_rate": 3.381103450600054e-08, + "loss": 0.7938, + "step": 41368 + }, + { + "epoch": 0.97, + "grad_norm": 1.9171476361837274, + "learning_rate": 3.3748369700089547e-08, + "loss": 0.9107, + "step": 41369 + }, + { + "epoch": 0.97, + "grad_norm": 1.9997046869502386, + "learning_rate": 3.368576292085091e-08, + "loss": 1.0155, + "step": 41370 + }, + { + "epoch": 0.97, + "grad_norm": 1.9730144999269785, + "learning_rate": 3.362321416864878e-08, + "loss": 1.0139, + "step": 41371 + }, + { + "epoch": 0.97, + "grad_norm": 1.8716958080046553, + "learning_rate": 3.356072344384842e-08, + "loss": 1.1333, + "step": 41372 + }, + { + "epoch": 0.97, + "grad_norm": 2.065555585074589, + "learning_rate": 3.3498290746812876e-08, + "loss": 0.9795, + "step": 41373 + }, + { + "epoch": 0.97, + "grad_norm": 1.83598337382469, + "learning_rate": 3.3435916077905196e-08, + "loss": 0.9836, + "step": 41374 + }, + { + "epoch": 0.97, + "grad_norm": 2.686374384920111, + "learning_rate": 3.3373599437488414e-08, + "loss": 0.8618, + "step": 41375 + }, + { + "epoch": 0.97, + "grad_norm": 2.121015713351255, + "learning_rate": 3.331134082592669e-08, + "loss": 0.9809, + "step": 41376 + }, + { + "epoch": 0.97, + "grad_norm": 2.239005877437882, + "learning_rate": 3.3249140243581947e-08, + "loss": 0.9509, + "step": 41377 + }, + { + "epoch": 0.97, + "grad_norm": 2.319072956936752, + "learning_rate": 3.318699769081501e-08, + "loss": 1.0249, + "step": 41378 + }, + { + "epoch": 0.97, + "grad_norm": 2.6774037516225486, + "learning_rate": 3.312491316798894e-08, + "loss": 1.0647, + "step": 41379 + }, + { + "epoch": 0.97, + "grad_norm": 2.1632352779579582, + "learning_rate": 3.3062886675466754e-08, + "loss": 1.0041, + "step": 41380 + }, + { + "epoch": 0.97, + "grad_norm": 2.301972651196524, + "learning_rate": 3.300091821360596e-08, + "loss": 0.9828, + "step": 41381 + }, + { + "epoch": 0.97, + "grad_norm": 1.9366150872564087, + "learning_rate": 3.29390077827707e-08, + "loss": 1.0365, + "step": 41382 + }, + { + "epoch": 0.97, + "grad_norm": 1.9784179173356606, + "learning_rate": 3.287715538331959e-08, + "loss": 0.9516, + "step": 41383 + }, + { + "epoch": 0.97, + "grad_norm": 2.0148514936754283, + "learning_rate": 3.281536101561233e-08, + "loss": 1.0081, + "step": 41384 + }, + { + "epoch": 0.97, + "grad_norm": 1.0354496533314608, + "learning_rate": 3.2753624680010864e-08, + "loss": 0.9394, + "step": 41385 + }, + { + "epoch": 0.98, + "grad_norm": 1.8330132888789368, + "learning_rate": 3.2691946376872674e-08, + "loss": 0.9269, + "step": 41386 + }, + { + "epoch": 0.98, + "grad_norm": 2.5312329883357005, + "learning_rate": 3.2630326106558584e-08, + "loss": 0.9742, + "step": 41387 + }, + { + "epoch": 0.98, + "grad_norm": 1.1759080589743063, + "learning_rate": 3.256876386942498e-08, + "loss": 0.9564, + "step": 41388 + }, + { + "epoch": 0.98, + "grad_norm": 2.0735912973533366, + "learning_rate": 3.250725966583268e-08, + "loss": 0.8704, + "step": 41389 + }, + { + "epoch": 0.98, + "grad_norm": 1.8558391146822568, + "learning_rate": 3.244581349613807e-08, + "loss": 0.9935, + "step": 41390 + }, + { + "epoch": 0.98, + "grad_norm": 2.050983842538566, + "learning_rate": 3.238442536069974e-08, + "loss": 0.9694, + "step": 41391 + }, + { + "epoch": 0.98, + "grad_norm": 1.878938926072611, + "learning_rate": 3.2323095259875205e-08, + "loss": 1.0292, + "step": 41392 + }, + { + "epoch": 0.98, + "grad_norm": 1.0902229579961926, + "learning_rate": 3.226182319402194e-08, + "loss": 0.9355, + "step": 41393 + }, + { + "epoch": 0.98, + "grad_norm": 1.7982756785531344, + "learning_rate": 3.220060916349521e-08, + "loss": 1.0098, + "step": 41394 + }, + { + "epoch": 0.98, + "grad_norm": 1.9769182048870588, + "learning_rate": 3.213945316865252e-08, + "loss": 0.8365, + "step": 41395 + }, + { + "epoch": 0.98, + "grad_norm": 1.9673482740857071, + "learning_rate": 3.2078355209849145e-08, + "loss": 0.8936, + "step": 41396 + }, + { + "epoch": 0.98, + "grad_norm": 2.2549500413035966, + "learning_rate": 3.201731528744145e-08, + "loss": 1.0216, + "step": 41397 + }, + { + "epoch": 0.98, + "grad_norm": 2.2645322317294543, + "learning_rate": 3.195633340178583e-08, + "loss": 0.9584, + "step": 41398 + }, + { + "epoch": 0.98, + "grad_norm": 2.0205625003034564, + "learning_rate": 3.189540955323534e-08, + "loss": 1.0684, + "step": 41399 + }, + { + "epoch": 0.98, + "grad_norm": 1.8240271665271746, + "learning_rate": 3.183454374214634e-08, + "loss": 0.9205, + "step": 41400 + }, + { + "epoch": 0.98, + "grad_norm": 1.9604748660674776, + "learning_rate": 3.17737359688719e-08, + "loss": 0.8319, + "step": 41401 + }, + { + "epoch": 0.98, + "grad_norm": 2.0013991183638127, + "learning_rate": 3.1712986233767286e-08, + "loss": 0.9757, + "step": 41402 + }, + { + "epoch": 0.98, + "grad_norm": 1.8706550833655204, + "learning_rate": 3.1652294537184436e-08, + "loss": 1.0542, + "step": 41403 + }, + { + "epoch": 0.98, + "grad_norm": 2.0466818768452897, + "learning_rate": 3.1591660879478624e-08, + "loss": 1.0454, + "step": 41404 + }, + { + "epoch": 0.98, + "grad_norm": 2.2583583807945535, + "learning_rate": 3.153108526100179e-08, + "loss": 1.0157, + "step": 41405 + }, + { + "epoch": 0.98, + "grad_norm": 1.1234083743433383, + "learning_rate": 3.147056768210699e-08, + "loss": 0.9635, + "step": 41406 + }, + { + "epoch": 0.98, + "grad_norm": 1.7227514984274481, + "learning_rate": 3.141010814314727e-08, + "loss": 0.8984, + "step": 41407 + }, + { + "epoch": 0.98, + "grad_norm": 2.1335246960045415, + "learning_rate": 3.1349706644473454e-08, + "loss": 0.9435, + "step": 41408 + }, + { + "epoch": 0.98, + "grad_norm": 1.9341719830856567, + "learning_rate": 3.1289363186437494e-08, + "loss": 0.9136, + "step": 41409 + }, + { + "epoch": 0.98, + "grad_norm": 1.701550558301612, + "learning_rate": 3.122907776939022e-08, + "loss": 0.8678, + "step": 41410 + }, + { + "epoch": 0.98, + "grad_norm": 2.0310285343197982, + "learning_rate": 3.1168850393684666e-08, + "loss": 1.0023, + "step": 41411 + }, + { + "epoch": 0.98, + "grad_norm": 1.100893957328221, + "learning_rate": 3.1108681059669466e-08, + "loss": 0.8694, + "step": 41412 + }, + { + "epoch": 0.98, + "grad_norm": 2.1722915493523143, + "learning_rate": 3.1048569767695435e-08, + "loss": 1.1068, + "step": 41413 + }, + { + "epoch": 0.98, + "grad_norm": 2.2651113611809626, + "learning_rate": 3.098851651811341e-08, + "loss": 1.0453, + "step": 41414 + }, + { + "epoch": 0.98, + "grad_norm": 1.8281199463342441, + "learning_rate": 3.0928521311271997e-08, + "loss": 1.0282, + "step": 41415 + }, + { + "epoch": 0.98, + "grad_norm": 2.057404415357283, + "learning_rate": 3.086858414752092e-08, + "loss": 0.9917, + "step": 41416 + }, + { + "epoch": 0.98, + "grad_norm": 1.9791946149238802, + "learning_rate": 3.0808705027208786e-08, + "loss": 1.0314, + "step": 41417 + }, + { + "epoch": 0.98, + "grad_norm": 2.0133199492932863, + "learning_rate": 3.07488839506842e-08, + "loss": 0.9025, + "step": 41418 + }, + { + "epoch": 0.98, + "grad_norm": 2.7849788542817318, + "learning_rate": 3.06891209182969e-08, + "loss": 1.0005, + "step": 41419 + }, + { + "epoch": 0.98, + "grad_norm": 2.2557554038350918, + "learning_rate": 3.062941593039326e-08, + "loss": 1.0384, + "step": 41420 + }, + { + "epoch": 0.98, + "grad_norm": 1.0256368338090276, + "learning_rate": 3.056976898732189e-08, + "loss": 0.9128, + "step": 41421 + }, + { + "epoch": 0.98, + "grad_norm": 2.2171447483054836, + "learning_rate": 3.051018008942919e-08, + "loss": 1.0399, + "step": 41422 + }, + { + "epoch": 0.98, + "grad_norm": 2.0135609354438264, + "learning_rate": 3.045064923706265e-08, + "loss": 0.8809, + "step": 41423 + }, + { + "epoch": 0.98, + "grad_norm": 1.1117869990182725, + "learning_rate": 3.039117643056755e-08, + "loss": 0.9561, + "step": 41424 + }, + { + "epoch": 0.98, + "grad_norm": 1.1371025219235424, + "learning_rate": 3.0331761670292505e-08, + "loss": 0.9765, + "step": 41425 + }, + { + "epoch": 0.98, + "grad_norm": 2.0447979076996807, + "learning_rate": 3.027240495658279e-08, + "loss": 1.0157, + "step": 41426 + }, + { + "epoch": 0.98, + "grad_norm": 1.8325369645648197, + "learning_rate": 3.021310628978258e-08, + "loss": 0.9479, + "step": 41427 + }, + { + "epoch": 0.98, + "grad_norm": 2.074179110175206, + "learning_rate": 3.015386567023826e-08, + "loss": 0.8824, + "step": 41428 + }, + { + "epoch": 0.98, + "grad_norm": 1.9246384756596313, + "learning_rate": 3.009468309829511e-08, + "loss": 0.9841, + "step": 41429 + }, + { + "epoch": 0.98, + "grad_norm": 2.1265604597683803, + "learning_rate": 3.003555857429619e-08, + "loss": 1.0016, + "step": 41430 + }, + { + "epoch": 0.98, + "grad_norm": 2.0937474326617953, + "learning_rate": 2.9976492098586775e-08, + "loss": 1.0461, + "step": 41431 + }, + { + "epoch": 0.98, + "grad_norm": 1.9898678463855506, + "learning_rate": 2.991748367151104e-08, + "loss": 1.0333, + "step": 41432 + }, + { + "epoch": 0.98, + "grad_norm": 2.8069181781195915, + "learning_rate": 2.985853329341315e-08, + "loss": 0.9107, + "step": 41433 + }, + { + "epoch": 0.98, + "grad_norm": 2.0423493004634383, + "learning_rate": 2.9799640964633946e-08, + "loss": 1.0291, + "step": 41434 + }, + { + "epoch": 0.98, + "grad_norm": 2.021205600575885, + "learning_rate": 2.9740806685517597e-08, + "loss": 0.9869, + "step": 41435 + }, + { + "epoch": 0.98, + "grad_norm": 1.8833627249263085, + "learning_rate": 2.9682030456408272e-08, + "loss": 1.1514, + "step": 41436 + }, + { + "epoch": 0.98, + "grad_norm": 1.0975169754419452, + "learning_rate": 2.96233122776457e-08, + "loss": 0.9244, + "step": 41437 + }, + { + "epoch": 0.98, + "grad_norm": 2.0086722799003778, + "learning_rate": 2.9564652149572938e-08, + "loss": 0.9472, + "step": 41438 + }, + { + "epoch": 0.98, + "grad_norm": 1.9524842050318896, + "learning_rate": 2.950605007253082e-08, + "loss": 0.9704, + "step": 41439 + }, + { + "epoch": 0.98, + "grad_norm": 1.7758967986318623, + "learning_rate": 2.9447506046862417e-08, + "loss": 0.9614, + "step": 41440 + }, + { + "epoch": 0.98, + "grad_norm": 1.84461267568078, + "learning_rate": 2.938902007290634e-08, + "loss": 0.9266, + "step": 41441 + }, + { + "epoch": 0.98, + "grad_norm": 1.838241384626083, + "learning_rate": 2.9330592151003424e-08, + "loss": 1.0319, + "step": 41442 + }, + { + "epoch": 0.98, + "grad_norm": 1.954874341044811, + "learning_rate": 2.927222228149562e-08, + "loss": 0.9495, + "step": 41443 + }, + { + "epoch": 0.98, + "grad_norm": 2.1201718947979473, + "learning_rate": 2.921391046472155e-08, + "loss": 1.0158, + "step": 41444 + }, + { + "epoch": 0.98, + "grad_norm": 1.9540090322153214, + "learning_rate": 2.915565670101983e-08, + "loss": 0.8844, + "step": 41445 + }, + { + "epoch": 0.98, + "grad_norm": 2.2155684558744557, + "learning_rate": 2.9097460990731296e-08, + "loss": 0.7934, + "step": 41446 + }, + { + "epoch": 0.98, + "grad_norm": 2.185297273354797, + "learning_rate": 2.9039323334193458e-08, + "loss": 0.9394, + "step": 41447 + }, + { + "epoch": 0.98, + "grad_norm": 2.166302095216965, + "learning_rate": 2.8981243731746044e-08, + "loss": 0.9704, + "step": 41448 + }, + { + "epoch": 0.98, + "grad_norm": 1.9686130053108888, + "learning_rate": 2.8923222183726563e-08, + "loss": 1.0778, + "step": 41449 + }, + { + "epoch": 0.98, + "grad_norm": 2.1324892303425025, + "learning_rate": 2.886525869047363e-08, + "loss": 0.8587, + "step": 41450 + }, + { + "epoch": 0.98, + "grad_norm": 2.20871220331925, + "learning_rate": 2.8807353252322533e-08, + "loss": 1.0296, + "step": 41451 + }, + { + "epoch": 0.98, + "grad_norm": 2.2952602077219972, + "learning_rate": 2.8749505869613003e-08, + "loss": 1.0304, + "step": 41452 + }, + { + "epoch": 0.98, + "grad_norm": 1.9846378924553507, + "learning_rate": 2.8691716542679215e-08, + "loss": 0.8306, + "step": 41453 + }, + { + "epoch": 0.98, + "grad_norm": 1.9133799020613438, + "learning_rate": 2.8633985271860898e-08, + "loss": 0.9303, + "step": 41454 + }, + { + "epoch": 0.98, + "grad_norm": 2.052837941213994, + "learning_rate": 2.8576312057491118e-08, + "loss": 1.0657, + "step": 41455 + }, + { + "epoch": 0.98, + "grad_norm": 1.90536961320072, + "learning_rate": 2.8518696899907382e-08, + "loss": 0.9472, + "step": 41456 + }, + { + "epoch": 0.98, + "grad_norm": 1.972329963477847, + "learning_rate": 2.8461139799444982e-08, + "loss": 1.1016, + "step": 41457 + }, + { + "epoch": 0.98, + "grad_norm": 2.750580087250703, + "learning_rate": 2.840364075643809e-08, + "loss": 0.9838, + "step": 41458 + }, + { + "epoch": 0.98, + "grad_norm": 1.8283191027921022, + "learning_rate": 2.8346199771221995e-08, + "loss": 1.0604, + "step": 41459 + }, + { + "epoch": 0.98, + "grad_norm": 1.0963183820117992, + "learning_rate": 2.8288816844131982e-08, + "loss": 0.923, + "step": 41460 + }, + { + "epoch": 0.98, + "grad_norm": 2.057991073694098, + "learning_rate": 2.8231491975502233e-08, + "loss": 0.8907, + "step": 41461 + }, + { + "epoch": 0.98, + "grad_norm": 1.9108266040561486, + "learning_rate": 2.8174225165664703e-08, + "loss": 0.9576, + "step": 41462 + }, + { + "epoch": 0.98, + "grad_norm": 2.138159235606251, + "learning_rate": 2.8117016414953567e-08, + "loss": 0.9828, + "step": 41463 + }, + { + "epoch": 0.98, + "grad_norm": 1.8739971894060545, + "learning_rate": 2.8059865723703006e-08, + "loss": 0.9514, + "step": 41464 + }, + { + "epoch": 0.98, + "grad_norm": 1.9433588710860938, + "learning_rate": 2.800277309224497e-08, + "loss": 0.9066, + "step": 41465 + }, + { + "epoch": 0.98, + "grad_norm": 3.2981524197505525, + "learning_rate": 2.7945738520911426e-08, + "loss": 0.919, + "step": 41466 + }, + { + "epoch": 0.98, + "grad_norm": 1.863841662623888, + "learning_rate": 2.788876201003432e-08, + "loss": 1.0167, + "step": 41467 + }, + { + "epoch": 0.98, + "grad_norm": 2.2821632016535256, + "learning_rate": 2.7831843559946724e-08, + "loss": 1.0616, + "step": 41468 + }, + { + "epoch": 0.98, + "grad_norm": 1.8392535949927857, + "learning_rate": 2.7774983170979487e-08, + "loss": 0.9266, + "step": 41469 + }, + { + "epoch": 0.98, + "grad_norm": 1.827109940075902, + "learning_rate": 2.771818084346345e-08, + "loss": 0.8793, + "step": 41470 + }, + { + "epoch": 0.98, + "grad_norm": 2.0699525884663035, + "learning_rate": 2.766143657772835e-08, + "loss": 0.9982, + "step": 41471 + }, + { + "epoch": 0.98, + "grad_norm": 2.986558585945208, + "learning_rate": 2.7604750374106148e-08, + "loss": 1.0498, + "step": 41472 + }, + { + "epoch": 0.98, + "grad_norm": 1.9402969298258892, + "learning_rate": 2.7548122232926578e-08, + "loss": 0.9229, + "step": 41473 + }, + { + "epoch": 0.98, + "grad_norm": 2.052311045476303, + "learning_rate": 2.7491552154519375e-08, + "loss": 0.9982, + "step": 41474 + }, + { + "epoch": 0.98, + "grad_norm": 1.9788869719826634, + "learning_rate": 2.7435040139213164e-08, + "loss": 0.8697, + "step": 41475 + }, + { + "epoch": 0.98, + "grad_norm": 1.9823566615719161, + "learning_rate": 2.7378586187336576e-08, + "loss": 1.1168, + "step": 41476 + }, + { + "epoch": 0.98, + "grad_norm": 2.009355063459093, + "learning_rate": 2.732219029921934e-08, + "loss": 0.9735, + "step": 41477 + }, + { + "epoch": 0.98, + "grad_norm": 1.8524401200435978, + "learning_rate": 2.72658524751912e-08, + "loss": 1.0862, + "step": 41478 + }, + { + "epoch": 0.98, + "grad_norm": 1.9158056175173144, + "learning_rate": 2.7209572715577447e-08, + "loss": 0.8871, + "step": 41479 + }, + { + "epoch": 0.98, + "grad_norm": 1.8572968829254528, + "learning_rate": 2.7153351020706707e-08, + "loss": 0.8202, + "step": 41480 + }, + { + "epoch": 0.98, + "grad_norm": 1.8822558374514204, + "learning_rate": 2.7097187390906497e-08, + "loss": 0.88, + "step": 41481 + }, + { + "epoch": 0.98, + "grad_norm": 2.0600862052506503, + "learning_rate": 2.7041081826504333e-08, + "loss": 0.9653, + "step": 41482 + }, + { + "epoch": 0.98, + "grad_norm": 2.0395966025372756, + "learning_rate": 2.698503432782662e-08, + "loss": 0.9429, + "step": 41483 + }, + { + "epoch": 0.98, + "grad_norm": 1.866013979075091, + "learning_rate": 2.6929044895198652e-08, + "loss": 1.0145, + "step": 41484 + }, + { + "epoch": 0.98, + "grad_norm": 1.9957767855922737, + "learning_rate": 2.6873113528947946e-08, + "loss": 0.9273, + "step": 41485 + }, + { + "epoch": 0.98, + "grad_norm": 2.162494340105209, + "learning_rate": 2.681724022939869e-08, + "loss": 1.0228, + "step": 41486 + }, + { + "epoch": 0.98, + "grad_norm": 1.1322206191332527, + "learning_rate": 2.6761424996877284e-08, + "loss": 0.9444, + "step": 41487 + }, + { + "epoch": 0.98, + "grad_norm": 2.1738659548766677, + "learning_rate": 2.6705667831707915e-08, + "loss": 0.9695, + "step": 41488 + }, + { + "epoch": 0.98, + "grad_norm": 1.9515785218584076, + "learning_rate": 2.664996873421588e-08, + "loss": 1.1592, + "step": 41489 + }, + { + "epoch": 0.98, + "grad_norm": 2.0651683343291114, + "learning_rate": 2.6594327704725365e-08, + "loss": 1.0374, + "step": 41490 + }, + { + "epoch": 0.98, + "grad_norm": 1.1553749409097993, + "learning_rate": 2.6538744743560552e-08, + "loss": 0.9792, + "step": 41491 + }, + { + "epoch": 0.98, + "grad_norm": 1.9600353897415237, + "learning_rate": 2.648321985104452e-08, + "loss": 1.0016, + "step": 41492 + }, + { + "epoch": 0.98, + "grad_norm": 2.2627852148924394, + "learning_rate": 2.6427753027500335e-08, + "loss": 0.982, + "step": 41493 + }, + { + "epoch": 0.98, + "grad_norm": 2.0457185822094663, + "learning_rate": 2.6372344273251082e-08, + "loss": 1.0053, + "step": 41494 + }, + { + "epoch": 0.98, + "grad_norm": 1.042487964111827, + "learning_rate": 2.6316993588620944e-08, + "loss": 0.9151, + "step": 41495 + }, + { + "epoch": 0.98, + "grad_norm": 2.1558395576896765, + "learning_rate": 2.626170097392966e-08, + "loss": 0.9893, + "step": 41496 + }, + { + "epoch": 0.98, + "grad_norm": 2.42034810640619, + "learning_rate": 2.620646642950142e-08, + "loss": 0.8519, + "step": 41497 + }, + { + "epoch": 0.98, + "grad_norm": 2.0015947444479014, + "learning_rate": 2.6151289955655966e-08, + "loss": 0.9929, + "step": 41498 + }, + { + "epoch": 0.98, + "grad_norm": 2.050486364167061, + "learning_rate": 2.6096171552715265e-08, + "loss": 1.0674, + "step": 41499 + }, + { + "epoch": 0.98, + "grad_norm": 2.946523776360513, + "learning_rate": 2.604111122100128e-08, + "loss": 0.9492, + "step": 41500 + }, + { + "epoch": 0.98, + "grad_norm": 1.0121587928349938, + "learning_rate": 2.5986108960832644e-08, + "loss": 0.8734, + "step": 41501 + }, + { + "epoch": 0.98, + "grad_norm": 2.193134974590742, + "learning_rate": 2.5931164772530214e-08, + "loss": 1.0398, + "step": 41502 + }, + { + "epoch": 0.98, + "grad_norm": 1.8662369953084952, + "learning_rate": 2.5876278656415954e-08, + "loss": 1.0771, + "step": 41503 + }, + { + "epoch": 0.98, + "grad_norm": 2.0926475124326966, + "learning_rate": 2.5821450612806276e-08, + "loss": 0.8958, + "step": 41504 + }, + { + "epoch": 0.98, + "grad_norm": 2.196558447100552, + "learning_rate": 2.5766680642023147e-08, + "loss": 1.0611, + "step": 41505 + }, + { + "epoch": 0.98, + "grad_norm": 2.012687862628622, + "learning_rate": 2.5711968744382975e-08, + "loss": 1.0042, + "step": 41506 + }, + { + "epoch": 0.98, + "grad_norm": 1.8201430055980434, + "learning_rate": 2.565731492020662e-08, + "loss": 0.9414, + "step": 41507 + }, + { + "epoch": 0.98, + "grad_norm": 1.7237259721435287, + "learning_rate": 2.5602719169810496e-08, + "loss": 1.1013, + "step": 41508 + }, + { + "epoch": 0.98, + "grad_norm": 1.9849259896243618, + "learning_rate": 2.5548181493513235e-08, + "loss": 1.0053, + "step": 41509 + }, + { + "epoch": 0.98, + "grad_norm": 1.8469119191967647, + "learning_rate": 2.5493701891632362e-08, + "loss": 0.9984, + "step": 41510 + }, + { + "epoch": 0.98, + "grad_norm": 2.0583862082103352, + "learning_rate": 2.5439280364485398e-08, + "loss": 1.0413, + "step": 41511 + }, + { + "epoch": 0.98, + "grad_norm": 1.960554018342989, + "learning_rate": 2.538491691238765e-08, + "loss": 1.076, + "step": 41512 + }, + { + "epoch": 0.98, + "grad_norm": 1.9226181275811418, + "learning_rate": 2.5330611535658857e-08, + "loss": 1.0483, + "step": 41513 + }, + { + "epoch": 0.98, + "grad_norm": 1.973941051442679, + "learning_rate": 2.5276364234610995e-08, + "loss": 0.9561, + "step": 41514 + }, + { + "epoch": 0.98, + "grad_norm": 1.8782236706759674, + "learning_rate": 2.5222175009563808e-08, + "loss": 0.971, + "step": 41515 + }, + { + "epoch": 0.98, + "grad_norm": 2.1925801308902386, + "learning_rate": 2.516804386083038e-08, + "loss": 0.8499, + "step": 41516 + }, + { + "epoch": 0.98, + "grad_norm": 2.009077093799472, + "learning_rate": 2.5113970788726018e-08, + "loss": 1.1134, + "step": 41517 + }, + { + "epoch": 0.98, + "grad_norm": 1.9452787730537293, + "learning_rate": 2.5059955793567125e-08, + "loss": 0.8781, + "step": 41518 + }, + { + "epoch": 0.98, + "grad_norm": 1.9644719952041851, + "learning_rate": 2.5005998875665683e-08, + "loss": 0.9672, + "step": 41519 + }, + { + "epoch": 0.98, + "grad_norm": 1.8324049427961542, + "learning_rate": 2.495210003533921e-08, + "loss": 1.0116, + "step": 41520 + }, + { + "epoch": 0.98, + "grad_norm": 2.0273901766631917, + "learning_rate": 2.4898259272898573e-08, + "loss": 0.9739, + "step": 41521 + }, + { + "epoch": 0.98, + "grad_norm": 1.9863230797563844, + "learning_rate": 2.4844476588659073e-08, + "loss": 0.9121, + "step": 41522 + }, + { + "epoch": 0.98, + "grad_norm": 2.2151228745240674, + "learning_rate": 2.4790751982931572e-08, + "loss": 1.1649, + "step": 41523 + }, + { + "epoch": 0.98, + "grad_norm": 1.9493545641269185, + "learning_rate": 2.4737085456032483e-08, + "loss": 0.9135, + "step": 41524 + }, + { + "epoch": 0.98, + "grad_norm": 1.738065387082118, + "learning_rate": 2.4683477008270453e-08, + "loss": 1.0898, + "step": 41525 + }, + { + "epoch": 0.98, + "grad_norm": 2.3343616878411293, + "learning_rate": 2.4629926639960777e-08, + "loss": 1.0147, + "step": 41526 + }, + { + "epoch": 0.98, + "grad_norm": 2.0405458208494927, + "learning_rate": 2.4576434351412103e-08, + "loss": 0.938, + "step": 41527 + }, + { + "epoch": 0.98, + "grad_norm": 2.221267061901059, + "learning_rate": 2.4523000142939735e-08, + "loss": 0.9348, + "step": 41528 + }, + { + "epoch": 0.98, + "grad_norm": 1.990033919866145, + "learning_rate": 2.44696240148512e-08, + "loss": 0.922, + "step": 41529 + }, + { + "epoch": 0.98, + "grad_norm": 2.018029033431129, + "learning_rate": 2.4416305967458476e-08, + "loss": 0.9794, + "step": 41530 + }, + { + "epoch": 0.98, + "grad_norm": 1.9250933992140156, + "learning_rate": 2.436304600107353e-08, + "loss": 0.8731, + "step": 41531 + }, + { + "epoch": 0.98, + "grad_norm": 2.1049855505272057, + "learning_rate": 2.43098441160039e-08, + "loss": 1.0124, + "step": 41532 + }, + { + "epoch": 0.98, + "grad_norm": 2.0583684806379448, + "learning_rate": 2.4256700312561554e-08, + "loss": 0.9985, + "step": 41533 + }, + { + "epoch": 0.98, + "grad_norm": 1.8800340973681047, + "learning_rate": 2.4203614591054024e-08, + "loss": 1.0174, + "step": 41534 + }, + { + "epoch": 0.98, + "grad_norm": 1.8048716396718278, + "learning_rate": 2.4150586951792175e-08, + "loss": 0.8776, + "step": 41535 + }, + { + "epoch": 0.98, + "grad_norm": 2.103917337775982, + "learning_rate": 2.409761739508354e-08, + "loss": 0.9412, + "step": 41536 + }, + { + "epoch": 0.98, + "grad_norm": 2.2837950588855698, + "learning_rate": 2.4044705921236756e-08, + "loss": 1.0447, + "step": 41537 + }, + { + "epoch": 0.98, + "grad_norm": 1.850271667453086, + "learning_rate": 2.3991852530560467e-08, + "loss": 0.9842, + "step": 41538 + }, + { + "epoch": 0.98, + "grad_norm": 1.8557046842525362, + "learning_rate": 2.3939057223362206e-08, + "loss": 0.9133, + "step": 41539 + }, + { + "epoch": 0.98, + "grad_norm": 2.2675499411202407, + "learning_rate": 2.3886319999948393e-08, + "loss": 0.8994, + "step": 41540 + }, + { + "epoch": 0.98, + "grad_norm": 1.8844930523357073, + "learning_rate": 2.383364086062656e-08, + "loss": 0.935, + "step": 41541 + }, + { + "epoch": 0.98, + "grad_norm": 1.988931113477921, + "learning_rate": 2.3781019805704243e-08, + "loss": 0.9419, + "step": 41542 + }, + { + "epoch": 0.98, + "grad_norm": 2.017603454218511, + "learning_rate": 2.3728456835487857e-08, + "loss": 1.0207, + "step": 41543 + }, + { + "epoch": 0.98, + "grad_norm": 1.9125914258273535, + "learning_rate": 2.3675951950281605e-08, + "loss": 0.8594, + "step": 41544 + }, + { + "epoch": 0.98, + "grad_norm": 2.3453364769596625, + "learning_rate": 2.3623505150393024e-08, + "loss": 0.9691, + "step": 41545 + }, + { + "epoch": 0.98, + "grad_norm": 2.0804204564513875, + "learning_rate": 2.3571116436127418e-08, + "loss": 0.7722, + "step": 41546 + }, + { + "epoch": 0.98, + "grad_norm": 1.1290421174177965, + "learning_rate": 2.351878580778788e-08, + "loss": 0.9881, + "step": 41547 + }, + { + "epoch": 0.98, + "grad_norm": 1.923468740414466, + "learning_rate": 2.3466513265680835e-08, + "loss": 0.949, + "step": 41548 + }, + { + "epoch": 0.98, + "grad_norm": 2.0259121879230664, + "learning_rate": 2.341429881011159e-08, + "loss": 0.886, + "step": 41549 + }, + { + "epoch": 0.98, + "grad_norm": 2.0958951823629017, + "learning_rate": 2.336214244138102e-08, + "loss": 0.8823, + "step": 41550 + }, + { + "epoch": 0.98, + "grad_norm": 1.8381407134448415, + "learning_rate": 2.331004415979554e-08, + "loss": 1.0104, + "step": 41551 + }, + { + "epoch": 0.98, + "grad_norm": 1.807678663284192, + "learning_rate": 2.325800396565825e-08, + "loss": 1.0334, + "step": 41552 + }, + { + "epoch": 0.98, + "grad_norm": 1.8265109733548146, + "learning_rate": 2.3206021859270012e-08, + "loss": 0.8873, + "step": 41553 + }, + { + "epoch": 0.98, + "grad_norm": 1.9477197936856132, + "learning_rate": 2.315409784093503e-08, + "loss": 0.9478, + "step": 41554 + }, + { + "epoch": 0.98, + "grad_norm": 2.336588025812684, + "learning_rate": 2.3102231910956398e-08, + "loss": 0.984, + "step": 41555 + }, + { + "epoch": 0.98, + "grad_norm": 1.9222477692768167, + "learning_rate": 2.305042406963498e-08, + "loss": 0.947, + "step": 41556 + }, + { + "epoch": 0.98, + "grad_norm": 2.2415685608338958, + "learning_rate": 2.2998674317272762e-08, + "loss": 0.8158, + "step": 41557 + }, + { + "epoch": 0.98, + "grad_norm": 2.0652998607270465, + "learning_rate": 2.2946982654170614e-08, + "loss": 1.0869, + "step": 41558 + }, + { + "epoch": 0.98, + "grad_norm": 1.8697283900909591, + "learning_rate": 2.2895349080630512e-08, + "loss": 0.9293, + "step": 41559 + }, + { + "epoch": 0.98, + "grad_norm": 1.825480720660171, + "learning_rate": 2.2843773596951115e-08, + "loss": 1.1004, + "step": 41560 + }, + { + "epoch": 0.98, + "grad_norm": 1.7612815981841936, + "learning_rate": 2.2792256203434393e-08, + "loss": 0.8574, + "step": 41561 + }, + { + "epoch": 0.98, + "grad_norm": 2.0682753265781915, + "learning_rate": 2.2740796900380114e-08, + "loss": 1.1688, + "step": 41562 + }, + { + "epoch": 0.98, + "grad_norm": 1.8925511370054433, + "learning_rate": 2.2689395688088034e-08, + "loss": 1.0253, + "step": 41563 + }, + { + "epoch": 0.98, + "grad_norm": 1.9653294902097966, + "learning_rate": 2.2638052566856806e-08, + "loss": 0.8883, + "step": 41564 + }, + { + "epoch": 0.98, + "grad_norm": 2.154175962909433, + "learning_rate": 2.2586767536985078e-08, + "loss": 0.9013, + "step": 41565 + }, + { + "epoch": 0.98, + "grad_norm": 2.075378999209928, + "learning_rate": 2.253554059877261e-08, + "loss": 0.9195, + "step": 41566 + }, + { + "epoch": 0.98, + "grad_norm": 1.9181681372029518, + "learning_rate": 2.2484371752516942e-08, + "loss": 1.023, + "step": 41567 + }, + { + "epoch": 0.98, + "grad_norm": 2.0695517435549617, + "learning_rate": 2.2433260998515615e-08, + "loss": 0.9685, + "step": 41568 + }, + { + "epoch": 0.98, + "grad_norm": 2.103188532003242, + "learning_rate": 2.238220833706728e-08, + "loss": 0.9443, + "step": 41569 + }, + { + "epoch": 0.98, + "grad_norm": 2.032152651167276, + "learning_rate": 2.2331213768468363e-08, + "loss": 1.0894, + "step": 41570 + }, + { + "epoch": 0.98, + "grad_norm": 2.0512038484621113, + "learning_rate": 2.2280277293016404e-08, + "loss": 1.0127, + "step": 41571 + }, + { + "epoch": 0.98, + "grad_norm": 1.967605440952868, + "learning_rate": 2.2229398911006726e-08, + "loss": 0.855, + "step": 41572 + }, + { + "epoch": 0.98, + "grad_norm": 2.043606384179081, + "learning_rate": 2.2178578622736867e-08, + "loss": 0.8756, + "step": 41573 + }, + { + "epoch": 0.98, + "grad_norm": 2.007356626944485, + "learning_rate": 2.2127816428502148e-08, + "loss": 0.9142, + "step": 41574 + }, + { + "epoch": 0.98, + "grad_norm": 2.0496682564184225, + "learning_rate": 2.207711232859788e-08, + "loss": 1.0781, + "step": 41575 + }, + { + "epoch": 0.98, + "grad_norm": 1.894062815225643, + "learning_rate": 2.2026466323320504e-08, + "loss": 0.9946, + "step": 41576 + }, + { + "epoch": 0.98, + "grad_norm": 1.1037115541310258, + "learning_rate": 2.1975878412964225e-08, + "loss": 0.9999, + "step": 41577 + }, + { + "epoch": 0.98, + "grad_norm": 1.9742565616286811, + "learning_rate": 2.1925348597822137e-08, + "loss": 0.7885, + "step": 41578 + }, + { + "epoch": 0.98, + "grad_norm": 8.579973101366615, + "learning_rate": 2.1874876878189564e-08, + "loss": 0.9174, + "step": 41579 + }, + { + "epoch": 0.98, + "grad_norm": 2.283235494222667, + "learning_rate": 2.1824463254361826e-08, + "loss": 0.9323, + "step": 41580 + }, + { + "epoch": 0.98, + "grad_norm": 1.9289128693970732, + "learning_rate": 2.177410772663091e-08, + "loss": 0.9521, + "step": 41581 + }, + { + "epoch": 0.98, + "grad_norm": 1.9007286518994517, + "learning_rate": 2.1723810295288807e-08, + "loss": 0.882, + "step": 41582 + }, + { + "epoch": 0.98, + "grad_norm": 2.0810767113416606, + "learning_rate": 2.167357096063083e-08, + "loss": 0.9533, + "step": 41583 + }, + { + "epoch": 0.98, + "grad_norm": 2.539023025829404, + "learning_rate": 2.1623389722947864e-08, + "loss": 1.0876, + "step": 41584 + }, + { + "epoch": 0.98, + "grad_norm": 1.9415596587989608, + "learning_rate": 2.1573266582533002e-08, + "loss": 0.9398, + "step": 41585 + }, + { + "epoch": 0.98, + "grad_norm": 2.065044785205442, + "learning_rate": 2.152320153967824e-08, + "loss": 1.0208, + "step": 41586 + }, + { + "epoch": 0.98, + "grad_norm": 1.8722874571947787, + "learning_rate": 2.147319459467334e-08, + "loss": 1.0167, + "step": 41587 + }, + { + "epoch": 0.98, + "grad_norm": 1.900113785183104, + "learning_rate": 2.1423245747811407e-08, + "loss": 0.8674, + "step": 41588 + }, + { + "epoch": 0.98, + "grad_norm": 1.8870298232098586, + "learning_rate": 2.13733549993822e-08, + "loss": 0.8184, + "step": 41589 + }, + { + "epoch": 0.98, + "grad_norm": 1.887518865871171, + "learning_rate": 2.1323522349676607e-08, + "loss": 0.8856, + "step": 41590 + }, + { + "epoch": 0.98, + "grad_norm": 1.9026552020388452, + "learning_rate": 2.12737477989855e-08, + "loss": 0.9915, + "step": 41591 + }, + { + "epoch": 0.98, + "grad_norm": 1.7797983441171672, + "learning_rate": 2.122403134759643e-08, + "loss": 0.9713, + "step": 41592 + }, + { + "epoch": 0.98, + "grad_norm": 1.9658783162907376, + "learning_rate": 2.1174372995801385e-08, + "loss": 1.0657, + "step": 41593 + }, + { + "epoch": 0.98, + "grad_norm": 1.7211354040950713, + "learning_rate": 2.112477274388791e-08, + "loss": 1.0398, + "step": 41594 + }, + { + "epoch": 0.98, + "grad_norm": 2.0482409160081168, + "learning_rate": 2.107523059214467e-08, + "loss": 1.0235, + "step": 41595 + }, + { + "epoch": 0.98, + "grad_norm": 2.315574051935036, + "learning_rate": 2.1025746540861425e-08, + "loss": 0.9848, + "step": 41596 + }, + { + "epoch": 0.98, + "grad_norm": 1.883919984728364, + "learning_rate": 2.0976320590325727e-08, + "loss": 0.9553, + "step": 41597 + }, + { + "epoch": 0.98, + "grad_norm": 1.8290633562556042, + "learning_rate": 2.0926952740825124e-08, + "loss": 0.9649, + "step": 41598 + }, + { + "epoch": 0.98, + "grad_norm": 1.1190243882420141, + "learning_rate": 2.0877642992647164e-08, + "loss": 0.9655, + "step": 41599 + }, + { + "epoch": 0.98, + "grad_norm": 1.9505094572218504, + "learning_rate": 2.082839134607828e-08, + "loss": 0.8227, + "step": 41600 + }, + { + "epoch": 0.98, + "grad_norm": 1.0888224972119778, + "learning_rate": 2.0779197801406026e-08, + "loss": 0.9373, + "step": 41601 + }, + { + "epoch": 0.98, + "grad_norm": 2.0387950850665613, + "learning_rate": 2.0730062358916835e-08, + "loss": 0.9727, + "step": 41602 + }, + { + "epoch": 0.98, + "grad_norm": 1.9202907149510462, + "learning_rate": 2.068098501889715e-08, + "loss": 0.9346, + "step": 41603 + }, + { + "epoch": 0.98, + "grad_norm": 2.192944020379713, + "learning_rate": 2.063196578163118e-08, + "loss": 1.1017, + "step": 41604 + }, + { + "epoch": 0.98, + "grad_norm": 1.9590935721822123, + "learning_rate": 2.058300464740648e-08, + "loss": 1.0935, + "step": 41605 + }, + { + "epoch": 0.98, + "grad_norm": 2.083702557250528, + "learning_rate": 2.0534101616506154e-08, + "loss": 1.0147, + "step": 41606 + }, + { + "epoch": 0.98, + "grad_norm": 1.9564661690899923, + "learning_rate": 2.0485256689215525e-08, + "loss": 0.905, + "step": 41607 + }, + { + "epoch": 0.98, + "grad_norm": 1.9539852898895926, + "learning_rate": 2.043646986581993e-08, + "loss": 1.0102, + "step": 41608 + }, + { + "epoch": 0.98, + "grad_norm": 1.757832147435483, + "learning_rate": 2.038774114660136e-08, + "loss": 0.9202, + "step": 41609 + }, + { + "epoch": 0.98, + "grad_norm": 2.255436293656186, + "learning_rate": 2.033907053184625e-08, + "loss": 0.9615, + "step": 41610 + }, + { + "epoch": 0.98, + "grad_norm": 2.2175723798985505, + "learning_rate": 2.029045802183549e-08, + "loss": 0.9652, + "step": 41611 + }, + { + "epoch": 0.98, + "grad_norm": 1.885888076617981, + "learning_rate": 2.024190361685441e-08, + "loss": 0.9109, + "step": 41612 + }, + { + "epoch": 0.98, + "grad_norm": 1.830772192271318, + "learning_rate": 2.019340731718389e-08, + "loss": 1.0591, + "step": 41613 + }, + { + "epoch": 0.98, + "grad_norm": 1.989373953661159, + "learning_rate": 2.014496912310704e-08, + "loss": 0.8691, + "step": 41614 + }, + { + "epoch": 0.98, + "grad_norm": 2.3769272705978235, + "learning_rate": 2.0096589034905856e-08, + "loss": 1.0921, + "step": 41615 + }, + { + "epoch": 0.98, + "grad_norm": 2.1219399585967076, + "learning_rate": 2.0048267052861226e-08, + "loss": 1.002, + "step": 41616 + }, + { + "epoch": 0.98, + "grad_norm": 2.1898414889903037, + "learning_rate": 2.0000003177255145e-08, + "loss": 0.8665, + "step": 41617 + }, + { + "epoch": 0.98, + "grad_norm": 2.1716195842778294, + "learning_rate": 1.9951797408369612e-08, + "loss": 0.9982, + "step": 41618 + }, + { + "epoch": 0.98, + "grad_norm": 1.8841354910328239, + "learning_rate": 1.99036497464844e-08, + "loss": 0.9403, + "step": 41619 + }, + { + "epoch": 0.98, + "grad_norm": 1.8019826305538524, + "learning_rate": 1.9855560191879285e-08, + "loss": 0.8687, + "step": 41620 + }, + { + "epoch": 0.98, + "grad_norm": 2.3850163143293366, + "learning_rate": 1.9807528744834048e-08, + "loss": 0.9645, + "step": 41621 + }, + { + "epoch": 0.98, + "grad_norm": 2.4247384404133068, + "learning_rate": 1.975955540563068e-08, + "loss": 1.0199, + "step": 41622 + }, + { + "epoch": 0.98, + "grad_norm": 1.1207456537421256, + "learning_rate": 1.9711640174545633e-08, + "loss": 0.8896, + "step": 41623 + }, + { + "epoch": 0.98, + "grad_norm": 2.717961036205722, + "learning_rate": 1.9663783051858676e-08, + "loss": 0.9177, + "step": 41624 + }, + { + "epoch": 0.98, + "grad_norm": 1.88440791945661, + "learning_rate": 1.9615984037849588e-08, + "loss": 0.8798, + "step": 41625 + }, + { + "epoch": 0.98, + "grad_norm": 2.0922911376823112, + "learning_rate": 1.9568243132797036e-08, + "loss": 0.8869, + "step": 41626 + }, + { + "epoch": 0.98, + "grad_norm": 1.7097181838080482, + "learning_rate": 1.9520560336976357e-08, + "loss": 0.8262, + "step": 41627 + }, + { + "epoch": 0.98, + "grad_norm": 1.9812016389579479, + "learning_rate": 1.9472935650667325e-08, + "loss": 0.9932, + "step": 41628 + }, + { + "epoch": 0.98, + "grad_norm": 2.106080469307798, + "learning_rate": 1.9425369074146384e-08, + "loss": 0.7931, + "step": 41629 + }, + { + "epoch": 0.98, + "grad_norm": 2.592612052688371, + "learning_rate": 1.937786060769109e-08, + "loss": 1.0575, + "step": 41630 + }, + { + "epoch": 0.98, + "grad_norm": 1.9632485369432369, + "learning_rate": 1.933041025157789e-08, + "loss": 0.9687, + "step": 41631 + }, + { + "epoch": 0.98, + "grad_norm": 2.0518424824275607, + "learning_rate": 1.928301800608212e-08, + "loss": 0.8985, + "step": 41632 + }, + { + "epoch": 0.98, + "grad_norm": 1.0261087876090849, + "learning_rate": 1.9235683871481338e-08, + "loss": 0.9525, + "step": 41633 + }, + { + "epoch": 0.98, + "grad_norm": 1.652217168347358, + "learning_rate": 1.918840784804976e-08, + "loss": 0.874, + "step": 41634 + }, + { + "epoch": 0.98, + "grad_norm": 2.2024892167993855, + "learning_rate": 1.9141189936063842e-08, + "loss": 0.9041, + "step": 41635 + }, + { + "epoch": 0.98, + "grad_norm": 2.1090620713884225, + "learning_rate": 1.9094030135796694e-08, + "loss": 0.9443, + "step": 41636 + }, + { + "epoch": 0.98, + "grad_norm": 1.7718678877324459, + "learning_rate": 1.904692844752476e-08, + "loss": 1.0023, + "step": 41637 + }, + { + "epoch": 0.98, + "grad_norm": 4.8695118987212105, + "learning_rate": 1.8999884871521157e-08, + "loss": 1.029, + "step": 41638 + }, + { + "epoch": 0.98, + "grad_norm": 1.8887999231331059, + "learning_rate": 1.895289940806011e-08, + "loss": 1.0614, + "step": 41639 + }, + { + "epoch": 0.98, + "grad_norm": 1.8777593364558298, + "learning_rate": 1.8905972057415845e-08, + "loss": 1.1369, + "step": 41640 + }, + { + "epoch": 0.98, + "grad_norm": 1.9852787014772655, + "learning_rate": 1.8859102819860365e-08, + "loss": 0.9386, + "step": 41641 + }, + { + "epoch": 0.98, + "grad_norm": 2.7608324899906447, + "learning_rate": 1.8812291695666783e-08, + "loss": 0.8855, + "step": 41642 + }, + { + "epoch": 0.98, + "grad_norm": 2.0739859061469543, + "learning_rate": 1.876553868510822e-08, + "loss": 0.9753, + "step": 41643 + }, + { + "epoch": 0.98, + "grad_norm": 1.9507527407755048, + "learning_rate": 1.8718843788457787e-08, + "loss": 1.1101, + "step": 41644 + }, + { + "epoch": 0.98, + "grad_norm": 1.9601940836985916, + "learning_rate": 1.8672207005985267e-08, + "loss": 0.8991, + "step": 41645 + }, + { + "epoch": 0.98, + "grad_norm": 2.2714789971479026, + "learning_rate": 1.8625628337963775e-08, + "loss": 0.8918, + "step": 41646 + }, + { + "epoch": 0.98, + "grad_norm": 1.9756575534660612, + "learning_rate": 1.8579107784663097e-08, + "loss": 1.0192, + "step": 41647 + }, + { + "epoch": 0.98, + "grad_norm": 1.7536613697893255, + "learning_rate": 1.853264534635635e-08, + "loss": 0.9143, + "step": 41648 + }, + { + "epoch": 0.98, + "grad_norm": 2.011826015941274, + "learning_rate": 1.8486241023311093e-08, + "loss": 0.9754, + "step": 41649 + }, + { + "epoch": 0.98, + "grad_norm": 1.0708230586415586, + "learning_rate": 1.843989481579933e-08, + "loss": 0.9428, + "step": 41650 + }, + { + "epoch": 0.98, + "grad_norm": 1.954008183828996, + "learning_rate": 1.8393606724090852e-08, + "loss": 1.0454, + "step": 41651 + }, + { + "epoch": 0.98, + "grad_norm": 1.986688781238297, + "learning_rate": 1.8347376748455438e-08, + "loss": 0.9549, + "step": 41652 + }, + { + "epoch": 0.98, + "grad_norm": 2.120143685017156, + "learning_rate": 1.8301204889160652e-08, + "loss": 0.9054, + "step": 41653 + }, + { + "epoch": 0.98, + "grad_norm": 1.862912226092964, + "learning_rate": 1.8255091146477387e-08, + "loss": 0.9656, + "step": 41654 + }, + { + "epoch": 0.98, + "grad_norm": 1.9572330398065056, + "learning_rate": 1.8209035520673213e-08, + "loss": 0.9251, + "step": 41655 + }, + { + "epoch": 0.98, + "grad_norm": 1.9874900424474662, + "learning_rate": 1.816303801201569e-08, + "loss": 1.075, + "step": 41656 + }, + { + "epoch": 0.98, + "grad_norm": 2.2927957236775347, + "learning_rate": 1.8117098620773487e-08, + "loss": 1.0115, + "step": 41657 + }, + { + "epoch": 0.98, + "grad_norm": 1.8080184291294241, + "learning_rate": 1.8071217347213066e-08, + "loss": 0.9715, + "step": 41658 + }, + { + "epoch": 0.98, + "grad_norm": 1.954542691991179, + "learning_rate": 1.8025394191603095e-08, + "loss": 0.9203, + "step": 41659 + }, + { + "epoch": 0.98, + "grad_norm": 1.937374004782825, + "learning_rate": 1.7979629154208923e-08, + "loss": 0.9124, + "step": 41660 + }, + { + "epoch": 0.98, + "grad_norm": 2.2406731044346904, + "learning_rate": 1.7933922235298108e-08, + "loss": 0.9549, + "step": 41661 + }, + { + "epoch": 0.98, + "grad_norm": 1.8468914629841677, + "learning_rate": 1.7888273435135994e-08, + "loss": 0.9501, + "step": 41662 + }, + { + "epoch": 0.98, + "grad_norm": 1.9267992138704375, + "learning_rate": 1.7842682753987927e-08, + "loss": 0.9539, + "step": 41663 + }, + { + "epoch": 0.98, + "grad_norm": 2.1788596750525624, + "learning_rate": 1.779715019212036e-08, + "loss": 0.9733, + "step": 41664 + }, + { + "epoch": 0.98, + "grad_norm": 1.8595387813417843, + "learning_rate": 1.7751675749798635e-08, + "loss": 0.9737, + "step": 41665 + }, + { + "epoch": 0.98, + "grad_norm": 2.018461206451126, + "learning_rate": 1.7706259427286988e-08, + "loss": 0.8605, + "step": 41666 + }, + { + "epoch": 0.98, + "grad_norm": 1.783500202837193, + "learning_rate": 1.7660901224849647e-08, + "loss": 0.9916, + "step": 41667 + }, + { + "epoch": 0.98, + "grad_norm": 1.942312635438538, + "learning_rate": 1.761560114275085e-08, + "loss": 0.8571, + "step": 41668 + }, + { + "epoch": 0.98, + "grad_norm": 1.9182156183112462, + "learning_rate": 1.7570359181254827e-08, + "loss": 0.9605, + "step": 41669 + }, + { + "epoch": 0.98, + "grad_norm": 2.221109318144037, + "learning_rate": 1.752517534062359e-08, + "loss": 0.8763, + "step": 41670 + }, + { + "epoch": 0.98, + "grad_norm": 1.7376694711077034, + "learning_rate": 1.7480049621122485e-08, + "loss": 0.9555, + "step": 41671 + }, + { + "epoch": 0.98, + "grad_norm": 1.0926248370118266, + "learning_rate": 1.7434982023012414e-08, + "loss": 0.9146, + "step": 41672 + }, + { + "epoch": 0.98, + "grad_norm": 1.7504918373742153, + "learning_rate": 1.738997254655539e-08, + "loss": 0.8128, + "step": 41673 + }, + { + "epoch": 0.98, + "grad_norm": 2.035854675714977, + "learning_rate": 1.7345021192015643e-08, + "loss": 0.8814, + "step": 41674 + }, + { + "epoch": 0.98, + "grad_norm": 1.9343446387396823, + "learning_rate": 1.730012795965297e-08, + "loss": 0.9803, + "step": 41675 + }, + { + "epoch": 0.98, + "grad_norm": 1.8896813304758855, + "learning_rate": 1.7255292849729376e-08, + "loss": 0.9943, + "step": 41676 + }, + { + "epoch": 0.98, + "grad_norm": 1.1893677933390532, + "learning_rate": 1.7210515862505773e-08, + "loss": 1.0095, + "step": 41677 + }, + { + "epoch": 0.98, + "grad_norm": 2.2603276593628716, + "learning_rate": 1.7165796998244168e-08, + "loss": 1.0014, + "step": 41678 + }, + { + "epoch": 0.98, + "grad_norm": 1.9652596742330004, + "learning_rate": 1.712113625720324e-08, + "loss": 0.9428, + "step": 41679 + }, + { + "epoch": 0.98, + "grad_norm": 1.0352375994962366, + "learning_rate": 1.7076533639642788e-08, + "loss": 0.913, + "step": 41680 + }, + { + "epoch": 0.98, + "grad_norm": 2.1511767970240627, + "learning_rate": 1.7031989145823713e-08, + "loss": 1.1255, + "step": 41681 + }, + { + "epoch": 0.98, + "grad_norm": 1.1235820543091, + "learning_rate": 1.6987502776005803e-08, + "loss": 0.9567, + "step": 41682 + }, + { + "epoch": 0.98, + "grad_norm": 1.8440884783570994, + "learning_rate": 1.6943074530446636e-08, + "loss": 1.0535, + "step": 41683 + }, + { + "epoch": 0.98, + "grad_norm": 2.3256144910802146, + "learning_rate": 1.6898704409404888e-08, + "loss": 0.9962, + "step": 41684 + }, + { + "epoch": 0.98, + "grad_norm": 1.955427017316205, + "learning_rate": 1.6854392413139244e-08, + "loss": 1.0227, + "step": 41685 + }, + { + "epoch": 0.98, + "grad_norm": 1.9458728164300192, + "learning_rate": 1.6810138541909492e-08, + "loss": 0.9554, + "step": 41686 + }, + { + "epoch": 0.98, + "grad_norm": 1.9743817452421617, + "learning_rate": 1.676594279597099e-08, + "loss": 1.0312, + "step": 41687 + }, + { + "epoch": 0.98, + "grad_norm": 2.0921214073005308, + "learning_rate": 1.6721805175581307e-08, + "loss": 1.1257, + "step": 41688 + }, + { + "epoch": 0.98, + "grad_norm": 1.9843539093475604, + "learning_rate": 1.6677725680999124e-08, + "loss": 1.0913, + "step": 41689 + }, + { + "epoch": 0.98, + "grad_norm": 1.9294319732930523, + "learning_rate": 1.6633704312478683e-08, + "loss": 0.9721, + "step": 41690 + }, + { + "epoch": 0.98, + "grad_norm": 1.969959653283725, + "learning_rate": 1.6589741070277555e-08, + "loss": 0.9007, + "step": 41691 + }, + { + "epoch": 0.98, + "grad_norm": 2.626671258125581, + "learning_rate": 1.65458359546522e-08, + "loss": 0.8818, + "step": 41692 + }, + { + "epoch": 0.98, + "grad_norm": 2.6308236267054834, + "learning_rate": 1.6501988965857972e-08, + "loss": 0.9314, + "step": 41693 + }, + { + "epoch": 0.98, + "grad_norm": 1.9336969524474699, + "learning_rate": 1.6458200104149115e-08, + "loss": 1.0023, + "step": 41694 + }, + { + "epoch": 0.98, + "grad_norm": 2.0510122877496393, + "learning_rate": 1.641446936978208e-08, + "loss": 1.0156, + "step": 41695 + }, + { + "epoch": 0.98, + "grad_norm": 2.1748626138786262, + "learning_rate": 1.637079676301001e-08, + "loss": 1.0076, + "step": 41696 + }, + { + "epoch": 0.98, + "grad_norm": 2.009191796589481, + "learning_rate": 1.632718228408825e-08, + "loss": 1.0417, + "step": 41697 + }, + { + "epoch": 0.98, + "grad_norm": 1.889017186580777, + "learning_rate": 1.6283625933269932e-08, + "loss": 1.0015, + "step": 41698 + }, + { + "epoch": 0.98, + "grad_norm": 2.083461509973631, + "learning_rate": 1.6240127710810405e-08, + "loss": 0.9828, + "step": 41699 + }, + { + "epoch": 0.98, + "grad_norm": 2.2696131041673326, + "learning_rate": 1.6196687616960583e-08, + "loss": 0.926, + "step": 41700 + }, + { + "epoch": 0.98, + "grad_norm": 2.229720360053319, + "learning_rate": 1.6153305651974703e-08, + "loss": 1.0761, + "step": 41701 + }, + { + "epoch": 0.98, + "grad_norm": 2.126838033034463, + "learning_rate": 1.610998181610479e-08, + "loss": 0.9609, + "step": 41702 + }, + { + "epoch": 0.98, + "grad_norm": 2.1394104478097486, + "learning_rate": 1.6066716109603975e-08, + "loss": 0.784, + "step": 41703 + }, + { + "epoch": 0.98, + "grad_norm": 2.088496944917019, + "learning_rate": 1.6023508532724273e-08, + "loss": 0.909, + "step": 41704 + }, + { + "epoch": 0.98, + "grad_norm": 2.297106340027841, + "learning_rate": 1.5980359085715493e-08, + "loss": 1.0193, + "step": 41705 + }, + { + "epoch": 0.98, + "grad_norm": 1.908148822809761, + "learning_rate": 1.5937267768829646e-08, + "loss": 1.0619, + "step": 41706 + }, + { + "epoch": 0.98, + "grad_norm": 1.8988716554149707, + "learning_rate": 1.5894234582318758e-08, + "loss": 0.9804, + "step": 41707 + }, + { + "epoch": 0.98, + "grad_norm": 2.490006166641486, + "learning_rate": 1.585125952643263e-08, + "loss": 0.9327, + "step": 41708 + }, + { + "epoch": 0.98, + "grad_norm": 1.0991891073504987, + "learning_rate": 1.5808342601421055e-08, + "loss": 0.9007, + "step": 41709 + }, + { + "epoch": 0.98, + "grad_norm": 2.0554703449594136, + "learning_rate": 1.5765483807533843e-08, + "loss": 1.0412, + "step": 41710 + }, + { + "epoch": 0.98, + "grad_norm": 1.790939660681734, + "learning_rate": 1.5722683145020788e-08, + "loss": 0.8368, + "step": 41711 + }, + { + "epoch": 0.98, + "grad_norm": 1.9029474851104293, + "learning_rate": 1.5679940614131696e-08, + "loss": 1.0281, + "step": 41712 + }, + { + "epoch": 0.98, + "grad_norm": 2.456403205518581, + "learning_rate": 1.563725621511525e-08, + "loss": 0.9612, + "step": 41713 + }, + { + "epoch": 0.98, + "grad_norm": 1.9375839244063122, + "learning_rate": 1.559462994822014e-08, + "loss": 1.0866, + "step": 41714 + }, + { + "epoch": 0.98, + "grad_norm": 1.951077054772678, + "learning_rate": 1.5552061813692843e-08, + "loss": 0.9522, + "step": 41715 + }, + { + "epoch": 0.98, + "grad_norm": 1.9348839916404479, + "learning_rate": 1.5509551811783152e-08, + "loss": 0.9329, + "step": 41716 + }, + { + "epoch": 0.98, + "grad_norm": 3.1648462747199124, + "learning_rate": 1.546709994273865e-08, + "loss": 0.9992, + "step": 41717 + }, + { + "epoch": 0.98, + "grad_norm": 1.0446815047137, + "learning_rate": 1.5424706206804695e-08, + "loss": 0.9973, + "step": 41718 + }, + { + "epoch": 0.98, + "grad_norm": 1.0470902728457188, + "learning_rate": 1.538237060422887e-08, + "loss": 0.9298, + "step": 41719 + }, + { + "epoch": 0.98, + "grad_norm": 1.8652847329426996, + "learning_rate": 1.5340093135258748e-08, + "loss": 0.9347, + "step": 41720 + }, + { + "epoch": 0.98, + "grad_norm": 1.9957836626689411, + "learning_rate": 1.5297873800139696e-08, + "loss": 0.9583, + "step": 41721 + }, + { + "epoch": 0.98, + "grad_norm": 2.000500717734296, + "learning_rate": 1.5255712599117066e-08, + "loss": 0.8331, + "step": 41722 + }, + { + "epoch": 0.98, + "grad_norm": 2.3675841041142336, + "learning_rate": 1.5213609532436223e-08, + "loss": 1.0769, + "step": 41723 + }, + { + "epoch": 0.98, + "grad_norm": 2.229286021401197, + "learning_rate": 1.5171564600343636e-08, + "loss": 0.9543, + "step": 41724 + }, + { + "epoch": 0.98, + "grad_norm": 2.1689834359505706, + "learning_rate": 1.512957780308355e-08, + "loss": 0.9955, + "step": 41725 + }, + { + "epoch": 0.98, + "grad_norm": 2.0440774573359475, + "learning_rate": 1.5087649140900218e-08, + "loss": 0.8546, + "step": 41726 + }, + { + "epoch": 0.98, + "grad_norm": 2.460784459559026, + "learning_rate": 1.5045778614036775e-08, + "loss": 0.8975, + "step": 41727 + }, + { + "epoch": 0.98, + "grad_norm": 2.493100114511052, + "learning_rate": 1.5003966222739697e-08, + "loss": 0.9112, + "step": 41728 + }, + { + "epoch": 0.98, + "grad_norm": 1.9166550983636155, + "learning_rate": 1.4962211967249895e-08, + "loss": 0.8981, + "step": 41729 + }, + { + "epoch": 0.98, + "grad_norm": 1.783007462965636, + "learning_rate": 1.4920515847810513e-08, + "loss": 0.9341, + "step": 41730 + }, + { + "epoch": 0.98, + "grad_norm": 1.8073895988391775, + "learning_rate": 1.4878877864665797e-08, + "loss": 0.9445, + "step": 41731 + }, + { + "epoch": 0.98, + "grad_norm": 2.391798244775283, + "learning_rate": 1.4837298018057778e-08, + "loss": 0.8594, + "step": 41732 + }, + { + "epoch": 0.98, + "grad_norm": 2.114897068811481, + "learning_rate": 1.4795776308228482e-08, + "loss": 1.0078, + "step": 41733 + }, + { + "epoch": 0.98, + "grad_norm": 2.304820271078188, + "learning_rate": 1.475431273541883e-08, + "loss": 0.8005, + "step": 41734 + }, + { + "epoch": 0.98, + "grad_norm": 1.9966424238209577, + "learning_rate": 1.4712907299870848e-08, + "loss": 0.967, + "step": 41735 + }, + { + "epoch": 0.98, + "grad_norm": 2.0505357425057005, + "learning_rate": 1.4671560001825458e-08, + "loss": 0.9582, + "step": 41736 + }, + { + "epoch": 0.98, + "grad_norm": 1.9511126849596894, + "learning_rate": 1.4630270841523575e-08, + "loss": 0.9687, + "step": 41737 + }, + { + "epoch": 0.98, + "grad_norm": 1.2548761029936137, + "learning_rate": 1.4589039819206118e-08, + "loss": 0.9111, + "step": 41738 + }, + { + "epoch": 0.98, + "grad_norm": 2.440833593727542, + "learning_rate": 1.4547866935111787e-08, + "loss": 1.0077, + "step": 41739 + }, + { + "epoch": 0.98, + "grad_norm": 2.298497305748561, + "learning_rate": 1.45067521894815e-08, + "loss": 0.931, + "step": 41740 + }, + { + "epoch": 0.98, + "grad_norm": 1.7981783690552715, + "learning_rate": 1.4465695582553951e-08, + "loss": 0.9225, + "step": 41741 + }, + { + "epoch": 0.98, + "grad_norm": 1.8916384712314975, + "learning_rate": 1.4424697114570063e-08, + "loss": 0.9997, + "step": 41742 + }, + { + "epoch": 0.98, + "grad_norm": 1.7215961159731465, + "learning_rate": 1.4383756785765201e-08, + "loss": 0.9233, + "step": 41743 + }, + { + "epoch": 0.98, + "grad_norm": 2.2741556933475437, + "learning_rate": 1.4342874596380285e-08, + "loss": 0.9332, + "step": 41744 + }, + { + "epoch": 0.98, + "grad_norm": 2.10499064147268, + "learning_rate": 1.4302050546651791e-08, + "loss": 0.9408, + "step": 41745 + }, + { + "epoch": 0.98, + "grad_norm": 1.8673870995508943, + "learning_rate": 1.4261284636819527e-08, + "loss": 0.9704, + "step": 41746 + }, + { + "epoch": 0.98, + "grad_norm": 1.99729257514982, + "learning_rate": 1.4220576867118862e-08, + "loss": 0.8628, + "step": 41747 + }, + { + "epoch": 0.98, + "grad_norm": 1.803661948206383, + "learning_rate": 1.4179927237786272e-08, + "loss": 0.9923, + "step": 41748 + }, + { + "epoch": 0.98, + "grad_norm": 2.275950922406653, + "learning_rate": 1.4139335749061567e-08, + "loss": 0.9683, + "step": 41749 + }, + { + "epoch": 0.98, + "grad_norm": 1.9299681021269988, + "learning_rate": 1.4098802401177891e-08, + "loss": 0.9891, + "step": 41750 + }, + { + "epoch": 0.98, + "grad_norm": 2.227060025818432, + "learning_rate": 1.4058327194372834e-08, + "loss": 0.9756, + "step": 41751 + }, + { + "epoch": 0.98, + "grad_norm": 1.0584004520844799, + "learning_rate": 1.4017910128881762e-08, + "loss": 0.9384, + "step": 41752 + }, + { + "epoch": 0.98, + "grad_norm": 2.037033940982292, + "learning_rate": 1.3977551204938932e-08, + "loss": 1.1006, + "step": 41753 + }, + { + "epoch": 0.98, + "grad_norm": 2.9275568360492774, + "learning_rate": 1.3937250422781934e-08, + "loss": 0.8457, + "step": 41754 + }, + { + "epoch": 0.98, + "grad_norm": 2.3229135580128486, + "learning_rate": 1.3897007782642802e-08, + "loss": 1.0333, + "step": 41755 + }, + { + "epoch": 0.98, + "grad_norm": 2.541697771958193, + "learning_rate": 1.3856823284756904e-08, + "loss": 0.9794, + "step": 41756 + }, + { + "epoch": 0.98, + "grad_norm": 1.825913458787607, + "learning_rate": 1.3816696929358498e-08, + "loss": 0.8535, + "step": 41757 + }, + { + "epoch": 0.98, + "grad_norm": 1.9601557871885276, + "learning_rate": 1.377662871668073e-08, + "loss": 1.049, + "step": 41758 + }, + { + "epoch": 0.98, + "grad_norm": 2.172924749020513, + "learning_rate": 1.3736618646956746e-08, + "loss": 1.0897, + "step": 41759 + }, + { + "epoch": 0.98, + "grad_norm": 1.044786102618678, + "learning_rate": 1.3696666720419693e-08, + "loss": 0.9227, + "step": 41760 + }, + { + "epoch": 0.98, + "grad_norm": 2.0026273230083325, + "learning_rate": 1.365677293730272e-08, + "loss": 1.0016, + "step": 41761 + }, + { + "epoch": 0.98, + "grad_norm": 1.1020170831155067, + "learning_rate": 1.3616937297837863e-08, + "loss": 0.9504, + "step": 41762 + }, + { + "epoch": 0.98, + "grad_norm": 2.0286938759381146, + "learning_rate": 1.3577159802256045e-08, + "loss": 0.9199, + "step": 41763 + }, + { + "epoch": 0.98, + "grad_norm": 1.8909716637450478, + "learning_rate": 1.3537440450790417e-08, + "loss": 1.0847, + "step": 41764 + }, + { + "epoch": 0.98, + "grad_norm": 3.670825900959246, + "learning_rate": 1.3497779243670794e-08, + "loss": 1.0022, + "step": 41765 + }, + { + "epoch": 0.98, + "grad_norm": 1.9784495648283946, + "learning_rate": 1.3458176181129211e-08, + "loss": 0.9431, + "step": 41766 + }, + { + "epoch": 0.98, + "grad_norm": 1.8587238256929868, + "learning_rate": 1.3418631263395487e-08, + "loss": 0.9308, + "step": 41767 + }, + { + "epoch": 0.98, + "grad_norm": 2.2087805671462744, + "learning_rate": 1.3379144490700546e-08, + "loss": 0.9587, + "step": 41768 + }, + { + "epoch": 0.98, + "grad_norm": 2.3724570469499717, + "learning_rate": 1.3339715863273094e-08, + "loss": 0.9758, + "step": 41769 + }, + { + "epoch": 0.98, + "grad_norm": 1.792512592167074, + "learning_rate": 1.330034538134517e-08, + "loss": 1.0151, + "step": 41770 + }, + { + "epoch": 0.98, + "grad_norm": 1.9573057144854642, + "learning_rate": 1.3261033045143257e-08, + "loss": 0.9662, + "step": 41771 + }, + { + "epoch": 0.98, + "grad_norm": 1.1054626663077676, + "learning_rate": 1.3221778854897171e-08, + "loss": 0.9135, + "step": 41772 + }, + { + "epoch": 0.98, + "grad_norm": 2.01030712738265, + "learning_rate": 1.318258281083562e-08, + "loss": 0.9282, + "step": 41773 + }, + { + "epoch": 0.98, + "grad_norm": 1.8758989531028571, + "learning_rate": 1.3143444913187309e-08, + "loss": 0.945, + "step": 41774 + }, + { + "epoch": 0.98, + "grad_norm": 1.9589887578084189, + "learning_rate": 1.3104365162179833e-08, + "loss": 0.852, + "step": 41775 + }, + { + "epoch": 0.98, + "grad_norm": 1.859966264458931, + "learning_rate": 1.3065343558040789e-08, + "loss": 0.9763, + "step": 41776 + }, + { + "epoch": 0.98, + "grad_norm": 1.9655810418742048, + "learning_rate": 1.302638010099666e-08, + "loss": 1.078, + "step": 41777 + }, + { + "epoch": 0.98, + "grad_norm": 2.1385597372703544, + "learning_rate": 1.2987474791273936e-08, + "loss": 0.9741, + "step": 41778 + }, + { + "epoch": 0.98, + "grad_norm": 1.0908914322366414, + "learning_rate": 1.2948627629101318e-08, + "loss": 0.9592, + "step": 41779 + }, + { + "epoch": 0.98, + "grad_norm": 1.1127239599669858, + "learning_rate": 1.2909838614703073e-08, + "loss": 0.9269, + "step": 41780 + }, + { + "epoch": 0.98, + "grad_norm": 2.118715279056074, + "learning_rate": 1.287110774830569e-08, + "loss": 1.0381, + "step": 41781 + }, + { + "epoch": 0.98, + "grad_norm": 1.8507566366379622, + "learning_rate": 1.2832435030133427e-08, + "loss": 0.8602, + "step": 41782 + }, + { + "epoch": 0.98, + "grad_norm": 2.097119205227669, + "learning_rate": 1.2793820460413886e-08, + "loss": 1.0956, + "step": 41783 + }, + { + "epoch": 0.98, + "grad_norm": 2.286632612347229, + "learning_rate": 1.275526403937022e-08, + "loss": 0.9277, + "step": 41784 + }, + { + "epoch": 0.98, + "grad_norm": 2.901431346104969, + "learning_rate": 1.2716765767227801e-08, + "loss": 0.9739, + "step": 41785 + }, + { + "epoch": 0.98, + "grad_norm": 2.1639573580801215, + "learning_rate": 1.2678325644209789e-08, + "loss": 0.878, + "step": 41786 + }, + { + "epoch": 0.98, + "grad_norm": 2.1636869184455207, + "learning_rate": 1.2639943670540445e-08, + "loss": 1.0333, + "step": 41787 + }, + { + "epoch": 0.98, + "grad_norm": 2.324878840917137, + "learning_rate": 1.2601619846444035e-08, + "loss": 0.9792, + "step": 41788 + }, + { + "epoch": 0.98, + "grad_norm": 1.731810875689038, + "learning_rate": 1.2563354172142605e-08, + "loss": 0.8773, + "step": 41789 + }, + { + "epoch": 0.98, + "grad_norm": 2.4089040593723863, + "learning_rate": 1.252514664785931e-08, + "loss": 0.8556, + "step": 41790 + }, + { + "epoch": 0.98, + "grad_norm": 1.825893525021055, + "learning_rate": 1.2486997273816192e-08, + "loss": 0.8468, + "step": 41791 + }, + { + "epoch": 0.98, + "grad_norm": 2.278797871376643, + "learning_rate": 1.2448906050236408e-08, + "loss": 1.0171, + "step": 41792 + }, + { + "epoch": 0.98, + "grad_norm": 1.9304150061100318, + "learning_rate": 1.2410872977340893e-08, + "loss": 0.8974, + "step": 41793 + }, + { + "epoch": 0.98, + "grad_norm": 1.1075678700549973, + "learning_rate": 1.2372898055350579e-08, + "loss": 0.9251, + "step": 41794 + }, + { + "epoch": 0.98, + "grad_norm": 1.1761602148642358, + "learning_rate": 1.2334981284487513e-08, + "loss": 0.9061, + "step": 41795 + }, + { + "epoch": 0.98, + "grad_norm": 1.8905947202285271, + "learning_rate": 1.2297122664972627e-08, + "loss": 0.9825, + "step": 41796 + }, + { + "epoch": 0.98, + "grad_norm": 1.9993723812623478, + "learning_rate": 1.2259322197025747e-08, + "loss": 0.9655, + "step": 41797 + }, + { + "epoch": 0.98, + "grad_norm": 1.855769391298204, + "learning_rate": 1.2221579880867807e-08, + "loss": 0.9339, + "step": 41798 + }, + { + "epoch": 0.98, + "grad_norm": 2.175152590211275, + "learning_rate": 1.218389571671752e-08, + "loss": 0.9864, + "step": 41799 + }, + { + "epoch": 0.98, + "grad_norm": 2.3625652531049637, + "learning_rate": 1.2146269704793601e-08, + "loss": 0.8995, + "step": 41800 + }, + { + "epoch": 0.98, + "grad_norm": 1.8522641955711097, + "learning_rate": 1.2108701845318094e-08, + "loss": 1.0745, + "step": 41801 + }, + { + "epoch": 0.98, + "grad_norm": 1.8754567145134464, + "learning_rate": 1.2071192138506382e-08, + "loss": 0.953, + "step": 41802 + }, + { + "epoch": 0.98, + "grad_norm": 2.203434952680293, + "learning_rate": 1.203374058457829e-08, + "loss": 1.0376, + "step": 41803 + }, + { + "epoch": 0.98, + "grad_norm": 1.0272975001739533, + "learning_rate": 1.1996347183752532e-08, + "loss": 0.9659, + "step": 41804 + }, + { + "epoch": 0.98, + "grad_norm": 2.1522800350451083, + "learning_rate": 1.1959011936246712e-08, + "loss": 0.9246, + "step": 41805 + }, + { + "epoch": 0.98, + "grad_norm": 2.120584149750863, + "learning_rate": 1.1921734842277321e-08, + "loss": 0.8497, + "step": 41806 + }, + { + "epoch": 0.98, + "grad_norm": 2.5940371503069275, + "learning_rate": 1.1884515902060856e-08, + "loss": 1.0544, + "step": 41807 + }, + { + "epoch": 0.98, + "grad_norm": 1.9596000901340536, + "learning_rate": 1.1847355115814918e-08, + "loss": 1.0703, + "step": 41808 + }, + { + "epoch": 0.98, + "grad_norm": 1.9119066610951967, + "learning_rate": 1.1810252483757112e-08, + "loss": 0.9981, + "step": 41809 + }, + { + "epoch": 0.99, + "grad_norm": 2.3214574739912095, + "learning_rate": 1.177320800610171e-08, + "loss": 0.9774, + "step": 41810 + }, + { + "epoch": 0.99, + "grad_norm": 1.0566396924788612, + "learning_rate": 1.1736221683064097e-08, + "loss": 0.8931, + "step": 41811 + }, + { + "epoch": 0.99, + "grad_norm": 1.9430159799739246, + "learning_rate": 1.1699293514861877e-08, + "loss": 0.9269, + "step": 41812 + }, + { + "epoch": 0.99, + "grad_norm": 1.7665073071508794, + "learning_rate": 1.1662423501707099e-08, + "loss": 0.89, + "step": 41813 + }, + { + "epoch": 0.99, + "grad_norm": 2.025361429779321, + "learning_rate": 1.162561164381737e-08, + "loss": 1.0549, + "step": 41814 + }, + { + "epoch": 0.99, + "grad_norm": 1.9918500214720842, + "learning_rate": 1.158885794140474e-08, + "loss": 1.0804, + "step": 41815 + }, + { + "epoch": 0.99, + "grad_norm": 1.7486416524090351, + "learning_rate": 1.1552162394684596e-08, + "loss": 0.8797, + "step": 41816 + }, + { + "epoch": 0.99, + "grad_norm": 1.9849131307523995, + "learning_rate": 1.1515525003868988e-08, + "loss": 1.0182, + "step": 41817 + }, + { + "epoch": 0.99, + "grad_norm": 1.8671988096353547, + "learning_rate": 1.14789457691733e-08, + "loss": 0.8411, + "step": 41818 + }, + { + "epoch": 0.99, + "grad_norm": 1.9229530399236694, + "learning_rate": 1.1442424690809584e-08, + "loss": 1.0088, + "step": 41819 + }, + { + "epoch": 0.99, + "grad_norm": 1.8236713279189043, + "learning_rate": 1.1405961768989893e-08, + "loss": 0.9394, + "step": 41820 + }, + { + "epoch": 0.99, + "grad_norm": 1.8448818866118244, + "learning_rate": 1.1369557003927389e-08, + "loss": 0.997, + "step": 41821 + }, + { + "epoch": 0.99, + "grad_norm": 2.1829525586971252, + "learning_rate": 1.1333210395833017e-08, + "loss": 0.9654, + "step": 41822 + }, + { + "epoch": 0.99, + "grad_norm": 1.9432931066840555, + "learning_rate": 1.1296921944919936e-08, + "loss": 0.9774, + "step": 41823 + }, + { + "epoch": 0.99, + "grad_norm": 2.2636301109412584, + "learning_rate": 1.126069165139798e-08, + "loss": 1.1005, + "step": 41824 + }, + { + "epoch": 0.99, + "grad_norm": 1.8889647896987583, + "learning_rate": 1.1224519515478095e-08, + "loss": 1.0138, + "step": 41825 + }, + { + "epoch": 0.99, + "grad_norm": 2.4350767132077493, + "learning_rate": 1.1188405537372327e-08, + "loss": 0.9916, + "step": 41826 + }, + { + "epoch": 0.99, + "grad_norm": 1.9941635073895752, + "learning_rate": 1.1152349717290512e-08, + "loss": 1.0099, + "step": 41827 + }, + { + "epoch": 0.99, + "grad_norm": 2.2217528068645818, + "learning_rate": 1.111635205544137e-08, + "loss": 0.9453, + "step": 41828 + }, + { + "epoch": 0.99, + "grad_norm": 2.7401425569034954, + "learning_rate": 1.1080412552034736e-08, + "loss": 0.9584, + "step": 41829 + }, + { + "epoch": 0.99, + "grad_norm": 1.9370887638614007, + "learning_rate": 1.104453120728155e-08, + "loss": 0.9276, + "step": 41830 + }, + { + "epoch": 0.99, + "grad_norm": 1.872502365412017, + "learning_rate": 1.1008708021389426e-08, + "loss": 1.0587, + "step": 41831 + }, + { + "epoch": 0.99, + "grad_norm": 2.5228374473792368, + "learning_rate": 1.0972942994565971e-08, + "loss": 0.9021, + "step": 41832 + }, + { + "epoch": 0.99, + "grad_norm": 1.9743920856021209, + "learning_rate": 1.0937236127022133e-08, + "loss": 0.9819, + "step": 41833 + }, + { + "epoch": 0.99, + "grad_norm": 2.1314463912077986, + "learning_rate": 1.0901587418962189e-08, + "loss": 0.9992, + "step": 41834 + }, + { + "epoch": 0.99, + "grad_norm": 1.11509786616323, + "learning_rate": 1.0865996870597084e-08, + "loss": 0.9313, + "step": 41835 + }, + { + "epoch": 0.99, + "grad_norm": 2.146051191549197, + "learning_rate": 1.0830464482132208e-08, + "loss": 0.9392, + "step": 41836 + }, + { + "epoch": 0.99, + "grad_norm": 1.9812986301160012, + "learning_rate": 1.0794990253775172e-08, + "loss": 0.8945, + "step": 41837 + }, + { + "epoch": 0.99, + "grad_norm": 1.0224921623424503, + "learning_rate": 1.075957418573248e-08, + "loss": 0.8984, + "step": 41838 + }, + { + "epoch": 0.99, + "grad_norm": 1.927134441396625, + "learning_rate": 1.072421627820952e-08, + "loss": 0.9018, + "step": 41839 + }, + { + "epoch": 0.99, + "grad_norm": 2.0967131111419706, + "learning_rate": 1.0688916531412796e-08, + "loss": 0.9676, + "step": 41840 + }, + { + "epoch": 0.99, + "grad_norm": 2.0564321356115807, + "learning_rate": 1.0653674945547699e-08, + "loss": 1.0612, + "step": 41841 + }, + { + "epoch": 0.99, + "grad_norm": 2.1731327137518415, + "learning_rate": 1.0618491520819619e-08, + "loss": 0.9951, + "step": 41842 + }, + { + "epoch": 0.99, + "grad_norm": 1.7880837910153324, + "learning_rate": 1.0583366257432836e-08, + "loss": 1.0174, + "step": 41843 + }, + { + "epoch": 0.99, + "grad_norm": 1.8039220851318405, + "learning_rate": 1.0548299155593855e-08, + "loss": 1.0125, + "step": 41844 + }, + { + "epoch": 0.99, + "grad_norm": 2.1887152214853947, + "learning_rate": 1.0513290215503624e-08, + "loss": 0.9753, + "step": 41845 + }, + { + "epoch": 0.99, + "grad_norm": 2.0254470225090397, + "learning_rate": 1.0478339437368645e-08, + "loss": 0.964, + "step": 41846 + }, + { + "epoch": 0.99, + "grad_norm": 2.224342273429943, + "learning_rate": 1.044344682139098e-08, + "loss": 0.9795, + "step": 41847 + }, + { + "epoch": 0.99, + "grad_norm": 1.8452451643631782, + "learning_rate": 1.0408612367773796e-08, + "loss": 0.8805, + "step": 41848 + }, + { + "epoch": 0.99, + "grad_norm": 1.802651998853056, + "learning_rate": 1.0373836076720267e-08, + "loss": 0.839, + "step": 41849 + }, + { + "epoch": 0.99, + "grad_norm": 2.0224716105886347, + "learning_rate": 1.0339117948433564e-08, + "loss": 0.9713, + "step": 41850 + }, + { + "epoch": 0.99, + "grad_norm": 1.931287304385716, + "learning_rate": 1.0304457983114636e-08, + "loss": 1.0055, + "step": 41851 + }, + { + "epoch": 0.99, + "grad_norm": 1.9013427311314348, + "learning_rate": 1.0269856180966653e-08, + "loss": 0.9046, + "step": 41852 + }, + { + "epoch": 0.99, + "grad_norm": 2.026860985625498, + "learning_rate": 1.0235312542189458e-08, + "loss": 0.9645, + "step": 41853 + }, + { + "epoch": 0.99, + "grad_norm": 1.9743640508609268, + "learning_rate": 1.0200827066986219e-08, + "loss": 1.0198, + "step": 41854 + }, + { + "epoch": 0.99, + "grad_norm": 2.0507408789793637, + "learning_rate": 1.0166399755554557e-08, + "loss": 0.8476, + "step": 41855 + }, + { + "epoch": 0.99, + "grad_norm": 1.0637257829553162, + "learning_rate": 1.0132030608098752e-08, + "loss": 0.9492, + "step": 41856 + }, + { + "epoch": 0.99, + "grad_norm": 1.9078764917158015, + "learning_rate": 1.0097719624815316e-08, + "loss": 1.0149, + "step": 41857 + }, + { + "epoch": 0.99, + "grad_norm": 2.0454980192074763, + "learning_rate": 1.0063466805907418e-08, + "loss": 0.9751, + "step": 41858 + }, + { + "epoch": 0.99, + "grad_norm": 1.1012453721193918, + "learning_rate": 1.0029272151571567e-08, + "loss": 0.9013, + "step": 41859 + }, + { + "epoch": 0.99, + "grad_norm": 2.3162663063294553, + "learning_rate": 9.995135662007604e-09, + "loss": 0.9491, + "step": 41860 + }, + { + "epoch": 0.99, + "grad_norm": 2.076132135479676, + "learning_rate": 9.96105733741648e-09, + "loss": 0.9948, + "step": 41861 + }, + { + "epoch": 0.99, + "grad_norm": 1.0784138606970046, + "learning_rate": 9.927037177993593e-09, + "loss": 0.9758, + "step": 41862 + }, + { + "epoch": 0.99, + "grad_norm": 2.0422527113061824, + "learning_rate": 9.893075183938783e-09, + "loss": 1.0139, + "step": 41863 + }, + { + "epoch": 0.99, + "grad_norm": 2.021976995624694, + "learning_rate": 9.859171355448561e-09, + "loss": 0.8888, + "step": 41864 + }, + { + "epoch": 0.99, + "grad_norm": 1.9652763107054025, + "learning_rate": 9.825325692721654e-09, + "loss": 0.9024, + "step": 41865 + }, + { + "epoch": 0.99, + "grad_norm": 1.920175214103982, + "learning_rate": 9.791538195954575e-09, + "loss": 1.0664, + "step": 41866 + }, + { + "epoch": 0.99, + "grad_norm": 2.1310525562059652, + "learning_rate": 9.757808865343831e-09, + "loss": 1.0836, + "step": 41867 + }, + { + "epoch": 0.99, + "grad_norm": 2.0518765406093715, + "learning_rate": 9.724137701085933e-09, + "loss": 0.9394, + "step": 41868 + }, + { + "epoch": 0.99, + "grad_norm": 2.783036908866627, + "learning_rate": 9.690524703377391e-09, + "loss": 1.1328, + "step": 41869 + }, + { + "epoch": 0.99, + "grad_norm": 2.3233942398608405, + "learning_rate": 9.656969872413601e-09, + "loss": 1.0385, + "step": 41870 + }, + { + "epoch": 0.99, + "grad_norm": 1.9323423631994823, + "learning_rate": 9.623473208388855e-09, + "loss": 1.0433, + "step": 41871 + }, + { + "epoch": 0.99, + "grad_norm": 2.329747307128538, + "learning_rate": 9.590034711499663e-09, + "loss": 1.03, + "step": 41872 + }, + { + "epoch": 0.99, + "grad_norm": 2.0896659554294437, + "learning_rate": 9.556654381940311e-09, + "loss": 0.8187, + "step": 41873 + }, + { + "epoch": 0.99, + "grad_norm": 2.5755371297079503, + "learning_rate": 9.523332219905091e-09, + "loss": 0.8228, + "step": 41874 + }, + { + "epoch": 0.99, + "grad_norm": 2.0131004648488497, + "learning_rate": 9.490068225588289e-09, + "loss": 0.9175, + "step": 41875 + }, + { + "epoch": 0.99, + "grad_norm": 2.0526327255235954, + "learning_rate": 9.456862399183087e-09, + "loss": 1.0125, + "step": 41876 + }, + { + "epoch": 0.99, + "grad_norm": 5.678953802650544, + "learning_rate": 9.423714740882661e-09, + "loss": 1.0478, + "step": 41877 + }, + { + "epoch": 0.99, + "grad_norm": 2.0748392389773294, + "learning_rate": 9.390625250880191e-09, + "loss": 1.0753, + "step": 41878 + }, + { + "epoch": 0.99, + "grad_norm": 2.230596973024704, + "learning_rate": 9.357593929368857e-09, + "loss": 0.9781, + "step": 41879 + }, + { + "epoch": 0.99, + "grad_norm": 6.17194796248206, + "learning_rate": 9.324620776540727e-09, + "loss": 1.0678, + "step": 41880 + }, + { + "epoch": 0.99, + "grad_norm": 1.9923106715866177, + "learning_rate": 9.291705792586758e-09, + "loss": 0.9612, + "step": 41881 + }, + { + "epoch": 0.99, + "grad_norm": 1.8167521298688378, + "learning_rate": 9.25884897770013e-09, + "loss": 0.8419, + "step": 41882 + }, + { + "epoch": 0.99, + "grad_norm": 1.8799773235838673, + "learning_rate": 9.226050332071801e-09, + "loss": 1.016, + "step": 41883 + }, + { + "epoch": 0.99, + "grad_norm": 1.2514057897118465, + "learning_rate": 9.193309855891619e-09, + "loss": 0.9626, + "step": 41884 + }, + { + "epoch": 0.99, + "grad_norm": 2.0074010958735564, + "learning_rate": 9.160627549351652e-09, + "loss": 0.9185, + "step": 41885 + }, + { + "epoch": 0.99, + "grad_norm": 2.0974704101358186, + "learning_rate": 9.12800341264175e-09, + "loss": 1.05, + "step": 41886 + }, + { + "epoch": 0.99, + "grad_norm": 2.2268615403155527, + "learning_rate": 9.095437445950651e-09, + "loss": 0.9529, + "step": 41887 + }, + { + "epoch": 0.99, + "grad_norm": 2.428548694996416, + "learning_rate": 9.06292964947042e-09, + "loss": 0.9714, + "step": 41888 + }, + { + "epoch": 0.99, + "grad_norm": 1.9278696130140915, + "learning_rate": 9.030480023387579e-09, + "loss": 1.1179, + "step": 41889 + }, + { + "epoch": 0.99, + "grad_norm": 2.0255862789183188, + "learning_rate": 8.998088567893081e-09, + "loss": 1.0457, + "step": 41890 + }, + { + "epoch": 0.99, + "grad_norm": 2.113392001341899, + "learning_rate": 8.96575528317456e-09, + "loss": 0.9016, + "step": 41891 + }, + { + "epoch": 0.99, + "grad_norm": 1.716687991152593, + "learning_rate": 8.93348016942075e-09, + "loss": 0.9015, + "step": 41892 + }, + { + "epoch": 0.99, + "grad_norm": 2.2008022415385886, + "learning_rate": 8.90126322681928e-09, + "loss": 0.9418, + "step": 41893 + }, + { + "epoch": 0.99, + "grad_norm": 1.7405275743804645, + "learning_rate": 8.869104455557776e-09, + "loss": 0.9232, + "step": 41894 + }, + { + "epoch": 0.99, + "grad_norm": 2.2158833339668074, + "learning_rate": 8.837003855822756e-09, + "loss": 1.0158, + "step": 41895 + }, + { + "epoch": 0.99, + "grad_norm": 2.2821251560407103, + "learning_rate": 8.804961427802961e-09, + "loss": 0.9485, + "step": 41896 + }, + { + "epoch": 0.99, + "grad_norm": 1.032421357587352, + "learning_rate": 8.772977171682684e-09, + "loss": 0.9373, + "step": 41897 + }, + { + "epoch": 0.99, + "grad_norm": 2.207814184503392, + "learning_rate": 8.741051087650665e-09, + "loss": 0.9685, + "step": 41898 + }, + { + "epoch": 0.99, + "grad_norm": 1.921226225442765, + "learning_rate": 8.70918317589009e-09, + "loss": 0.8307, + "step": 41899 + }, + { + "epoch": 0.99, + "grad_norm": 2.170428230394901, + "learning_rate": 8.677373436588588e-09, + "loss": 1.0415, + "step": 41900 + }, + { + "epoch": 0.99, + "grad_norm": 1.0751816055098038, + "learning_rate": 8.645621869930453e-09, + "loss": 0.9316, + "step": 41901 + }, + { + "epoch": 0.99, + "grad_norm": 2.296551659346676, + "learning_rate": 8.613928476101097e-09, + "loss": 0.9848, + "step": 41902 + }, + { + "epoch": 0.99, + "grad_norm": 2.1930712900703173, + "learning_rate": 8.582293255284813e-09, + "loss": 0.9504, + "step": 41903 + }, + { + "epoch": 0.99, + "grad_norm": 1.8147812910089187, + "learning_rate": 8.550716207664788e-09, + "loss": 1.0483, + "step": 41904 + }, + { + "epoch": 0.99, + "grad_norm": 2.1648287854375217, + "learning_rate": 8.519197333426431e-09, + "loss": 1.0653, + "step": 41905 + }, + { + "epoch": 0.99, + "grad_norm": 2.0557522734429168, + "learning_rate": 8.48773663275182e-09, + "loss": 1.1003, + "step": 41906 + }, + { + "epoch": 0.99, + "grad_norm": 2.819803286883457, + "learning_rate": 8.456334105825248e-09, + "loss": 0.8576, + "step": 41907 + }, + { + "epoch": 0.99, + "grad_norm": 1.1133193536039199, + "learning_rate": 8.424989752829904e-09, + "loss": 0.93, + "step": 41908 + }, + { + "epoch": 0.99, + "grad_norm": 1.8616320888725082, + "learning_rate": 8.393703573946754e-09, + "loss": 0.976, + "step": 41909 + }, + { + "epoch": 0.99, + "grad_norm": 2.4100683053647685, + "learning_rate": 8.362475569358985e-09, + "loss": 0.9266, + "step": 41910 + }, + { + "epoch": 0.99, + "grad_norm": 2.128862999227381, + "learning_rate": 8.331305739247565e-09, + "loss": 0.9196, + "step": 41911 + }, + { + "epoch": 0.99, + "grad_norm": 2.314444923939641, + "learning_rate": 8.300194083795676e-09, + "loss": 0.8889, + "step": 41912 + }, + { + "epoch": 0.99, + "grad_norm": 2.044784872578852, + "learning_rate": 8.269140603182069e-09, + "loss": 0.9641, + "step": 41913 + }, + { + "epoch": 0.99, + "grad_norm": 2.157122223516583, + "learning_rate": 8.238145297589927e-09, + "loss": 0.8296, + "step": 41914 + }, + { + "epoch": 0.99, + "grad_norm": 1.867784386761995, + "learning_rate": 8.207208167198e-09, + "loss": 0.8828, + "step": 41915 + }, + { + "epoch": 0.99, + "grad_norm": 2.1364446187717876, + "learning_rate": 8.176329212187251e-09, + "loss": 1.0453, + "step": 41916 + }, + { + "epoch": 0.99, + "grad_norm": 1.8886984503482886, + "learning_rate": 8.145508432737537e-09, + "loss": 1.0461, + "step": 41917 + }, + { + "epoch": 0.99, + "grad_norm": 2.1273897936104356, + "learning_rate": 8.114745829027604e-09, + "loss": 0.9432, + "step": 41918 + }, + { + "epoch": 0.99, + "grad_norm": 2.0713074806748795, + "learning_rate": 8.084041401237308e-09, + "loss": 1.0363, + "step": 41919 + }, + { + "epoch": 0.99, + "grad_norm": 1.9759954066121281, + "learning_rate": 8.053395149545395e-09, + "loss": 0.8951, + "step": 41920 + }, + { + "epoch": 0.99, + "grad_norm": 2.299963154617921, + "learning_rate": 8.022807074129501e-09, + "loss": 1.078, + "step": 41921 + }, + { + "epoch": 0.99, + "grad_norm": 2.2246907655112858, + "learning_rate": 7.992277175168373e-09, + "loss": 0.9315, + "step": 41922 + }, + { + "epoch": 0.99, + "grad_norm": 1.9216065007234864, + "learning_rate": 7.961805452839644e-09, + "loss": 1.0281, + "step": 41923 + }, + { + "epoch": 0.99, + "grad_norm": 2.0701633027361694, + "learning_rate": 7.931391907320951e-09, + "loss": 0.9515, + "step": 41924 + }, + { + "epoch": 0.99, + "grad_norm": 1.955910346566193, + "learning_rate": 7.901036538788819e-09, + "loss": 0.9967, + "step": 41925 + }, + { + "epoch": 0.99, + "grad_norm": 1.9911084846235059, + "learning_rate": 7.870739347420886e-09, + "loss": 0.9115, + "step": 41926 + }, + { + "epoch": 0.99, + "grad_norm": 2.225694533205667, + "learning_rate": 7.840500333392565e-09, + "loss": 0.8549, + "step": 41927 + }, + { + "epoch": 0.99, + "grad_norm": 1.8386550261548307, + "learning_rate": 7.81031949688038e-09, + "loss": 1.054, + "step": 41928 + }, + { + "epoch": 0.99, + "grad_norm": 2.2507052198033035, + "learning_rate": 7.78019683805975e-09, + "loss": 0.9396, + "step": 41929 + }, + { + "epoch": 0.99, + "grad_norm": 1.7987234994322614, + "learning_rate": 7.750132357106089e-09, + "loss": 1.0259, + "step": 41930 + }, + { + "epoch": 0.99, + "grad_norm": 2.072875442075742, + "learning_rate": 7.72012605419481e-09, + "loss": 0.9219, + "step": 41931 + }, + { + "epoch": 0.99, + "grad_norm": 1.0612302211123357, + "learning_rate": 7.69017792950022e-09, + "loss": 0.8661, + "step": 41932 + }, + { + "epoch": 0.99, + "grad_norm": 1.8777185163627768, + "learning_rate": 7.660287983197735e-09, + "loss": 1.0441, + "step": 41933 + }, + { + "epoch": 0.99, + "grad_norm": 1.9459333132955632, + "learning_rate": 7.630456215460546e-09, + "loss": 0.9012, + "step": 41934 + }, + { + "epoch": 0.99, + "grad_norm": 2.077092323242467, + "learning_rate": 7.600682626461853e-09, + "loss": 0.9621, + "step": 41935 + }, + { + "epoch": 0.99, + "grad_norm": 2.0905497773823125, + "learning_rate": 7.570967216374847e-09, + "loss": 0.8403, + "step": 41936 + }, + { + "epoch": 0.99, + "grad_norm": 2.1281075578326996, + "learning_rate": 7.541309985373834e-09, + "loss": 0.9139, + "step": 41937 + }, + { + "epoch": 0.99, + "grad_norm": 2.090559636368822, + "learning_rate": 7.51171093363201e-09, + "loss": 0.9857, + "step": 41938 + }, + { + "epoch": 0.99, + "grad_norm": 2.500371611811455, + "learning_rate": 7.482170061319238e-09, + "loss": 1.1218, + "step": 41939 + }, + { + "epoch": 0.99, + "grad_norm": 2.11550226522895, + "learning_rate": 7.452687368608713e-09, + "loss": 1.0309, + "step": 41940 + }, + { + "epoch": 0.99, + "grad_norm": 2.0438520933802353, + "learning_rate": 7.42326285567363e-09, + "loss": 1.0052, + "step": 41941 + }, + { + "epoch": 0.99, + "grad_norm": 2.0743266507952614, + "learning_rate": 7.393896522682742e-09, + "loss": 1.0294, + "step": 41942 + }, + { + "epoch": 0.99, + "grad_norm": 1.8018574150849826, + "learning_rate": 7.364588369808135e-09, + "loss": 1.0657, + "step": 41943 + }, + { + "epoch": 0.99, + "grad_norm": 1.8696010411433315, + "learning_rate": 7.335338397220782e-09, + "loss": 0.9813, + "step": 41944 + }, + { + "epoch": 0.99, + "grad_norm": 1.0928868668690868, + "learning_rate": 7.306146605090547e-09, + "loss": 0.9807, + "step": 41945 + }, + { + "epoch": 0.99, + "grad_norm": 2.074249857329823, + "learning_rate": 7.277012993588406e-09, + "loss": 0.8922, + "step": 41946 + }, + { + "epoch": 0.99, + "grad_norm": 2.3057771945408327, + "learning_rate": 7.247937562882001e-09, + "loss": 0.9242, + "step": 41947 + }, + { + "epoch": 0.99, + "grad_norm": 1.9139722672062602, + "learning_rate": 7.218920313142308e-09, + "loss": 0.967, + "step": 41948 + }, + { + "epoch": 0.99, + "grad_norm": 2.0543720721446843, + "learning_rate": 7.1899612445380795e-09, + "loss": 1.0551, + "step": 41949 + }, + { + "epoch": 0.99, + "grad_norm": 1.9900561492490383, + "learning_rate": 7.16106035723696e-09, + "loss": 1.0108, + "step": 41950 + }, + { + "epoch": 0.99, + "grad_norm": 2.4102416705882823, + "learning_rate": 7.132217651407702e-09, + "loss": 0.9173, + "step": 41951 + }, + { + "epoch": 0.99, + "grad_norm": 2.1147872820163487, + "learning_rate": 7.103433127219061e-09, + "loss": 0.9067, + "step": 41952 + }, + { + "epoch": 0.99, + "grad_norm": 1.79554432526372, + "learning_rate": 7.074706784837571e-09, + "loss": 0.9617, + "step": 41953 + }, + { + "epoch": 0.99, + "grad_norm": 2.1292885711546843, + "learning_rate": 7.0460386244308735e-09, + "loss": 1.0321, + "step": 41954 + }, + { + "epoch": 0.99, + "grad_norm": 1.8859692544380342, + "learning_rate": 7.017428646165503e-09, + "loss": 0.9917, + "step": 41955 + }, + { + "epoch": 0.99, + "grad_norm": 2.0778896196915975, + "learning_rate": 6.988876850209103e-09, + "loss": 1.0383, + "step": 41956 + }, + { + "epoch": 0.99, + "grad_norm": 2.1849893030443135, + "learning_rate": 6.960383236727098e-09, + "loss": 1.0423, + "step": 41957 + }, + { + "epoch": 0.99, + "grad_norm": 2.125183566817107, + "learning_rate": 6.931947805884909e-09, + "loss": 0.9191, + "step": 41958 + }, + { + "epoch": 0.99, + "grad_norm": 1.0914552998554294, + "learning_rate": 6.903570557849071e-09, + "loss": 0.9097, + "step": 41959 + }, + { + "epoch": 0.99, + "grad_norm": 1.9688504982574864, + "learning_rate": 6.875251492783897e-09, + "loss": 0.9809, + "step": 41960 + }, + { + "epoch": 0.99, + "grad_norm": 1.9577114548636947, + "learning_rate": 6.8469906108559195e-09, + "loss": 0.9665, + "step": 41961 + }, + { + "epoch": 0.99, + "grad_norm": 2.1990400672844195, + "learning_rate": 6.8187879122272315e-09, + "loss": 0.9106, + "step": 41962 + }, + { + "epoch": 0.99, + "grad_norm": 2.1479680446841223, + "learning_rate": 6.790643397064367e-09, + "loss": 0.9943, + "step": 41963 + }, + { + "epoch": 0.99, + "grad_norm": 1.0484492070996623, + "learning_rate": 6.762557065529418e-09, + "loss": 1.016, + "step": 41964 + }, + { + "epoch": 0.99, + "grad_norm": 1.8904523768833468, + "learning_rate": 6.734528917786698e-09, + "loss": 0.8842, + "step": 41965 + }, + { + "epoch": 0.99, + "grad_norm": 1.9002803354418252, + "learning_rate": 6.706558953999409e-09, + "loss": 0.8678, + "step": 41966 + }, + { + "epoch": 0.99, + "grad_norm": 2.051718593006507, + "learning_rate": 6.678647174330755e-09, + "loss": 1.0545, + "step": 41967 + }, + { + "epoch": 0.99, + "grad_norm": 1.0704331380056062, + "learning_rate": 6.650793578942827e-09, + "loss": 0.8718, + "step": 41968 + }, + { + "epoch": 0.99, + "grad_norm": 2.4127692429093233, + "learning_rate": 6.622998167997718e-09, + "loss": 0.8798, + "step": 41969 + }, + { + "epoch": 0.99, + "grad_norm": 2.2902706365684824, + "learning_rate": 6.595260941657522e-09, + "loss": 0.937, + "step": 41970 + }, + { + "epoch": 0.99, + "grad_norm": 2.020825904565955, + "learning_rate": 6.56758190008322e-09, + "loss": 1.065, + "step": 41971 + }, + { + "epoch": 0.99, + "grad_norm": 2.088723607320666, + "learning_rate": 6.539961043436904e-09, + "loss": 1.0317, + "step": 41972 + }, + { + "epoch": 0.99, + "grad_norm": 2.007262477537669, + "learning_rate": 6.512398371877338e-09, + "loss": 1.0297, + "step": 41973 + }, + { + "epoch": 0.99, + "grad_norm": 2.1276407455962665, + "learning_rate": 6.484893885567722e-09, + "loss": 0.8726, + "step": 41974 + }, + { + "epoch": 0.99, + "grad_norm": 1.9256464403701374, + "learning_rate": 6.457447584666821e-09, + "loss": 0.9461, + "step": 41975 + }, + { + "epoch": 0.99, + "grad_norm": 2.061474767968329, + "learning_rate": 6.430059469334504e-09, + "loss": 1.0114, + "step": 41976 + }, + { + "epoch": 0.99, + "grad_norm": 1.875511252221121, + "learning_rate": 6.4027295397306455e-09, + "loss": 0.8943, + "step": 41977 + }, + { + "epoch": 0.99, + "grad_norm": 3.1680847496862636, + "learning_rate": 6.375457796014006e-09, + "loss": 0.913, + "step": 41978 + }, + { + "epoch": 0.99, + "grad_norm": 1.9551946305670138, + "learning_rate": 6.3482442383422385e-09, + "loss": 1.0276, + "step": 41979 + }, + { + "epoch": 0.99, + "grad_norm": 2.265139133462772, + "learning_rate": 6.3210888668763235e-09, + "loss": 0.8542, + "step": 41980 + }, + { + "epoch": 0.99, + "grad_norm": 2.1132687549124713, + "learning_rate": 6.293991681771694e-09, + "loss": 0.8609, + "step": 41981 + }, + { + "epoch": 0.99, + "grad_norm": 1.9347496905152997, + "learning_rate": 6.266952683188221e-09, + "loss": 0.9625, + "step": 41982 + }, + { + "epoch": 0.99, + "grad_norm": 1.7756029197242462, + "learning_rate": 6.2399718712824465e-09, + "loss": 0.8715, + "step": 41983 + }, + { + "epoch": 0.99, + "grad_norm": 2.0360258564884623, + "learning_rate": 6.2130492462120214e-09, + "loss": 1.1021, + "step": 41984 + }, + { + "epoch": 0.99, + "grad_norm": 2.154636762398897, + "learning_rate": 6.186184808132378e-09, + "loss": 0.8598, + "step": 41985 + }, + { + "epoch": 0.99, + "grad_norm": 2.0466559896708922, + "learning_rate": 6.159378557201168e-09, + "loss": 0.99, + "step": 41986 + }, + { + "epoch": 0.99, + "grad_norm": 1.0700455910560076, + "learning_rate": 6.1326304935738215e-09, + "loss": 0.8998, + "step": 41987 + }, + { + "epoch": 0.99, + "grad_norm": 1.917710129739646, + "learning_rate": 6.10594061740577e-09, + "loss": 0.9472, + "step": 41988 + }, + { + "epoch": 0.99, + "grad_norm": 1.0594934835574765, + "learning_rate": 6.0793089288535555e-09, + "loss": 0.8742, + "step": 41989 + }, + { + "epoch": 0.99, + "grad_norm": 1.9808542602448955, + "learning_rate": 6.052735428071499e-09, + "loss": 0.9081, + "step": 41990 + }, + { + "epoch": 0.99, + "grad_norm": 2.831495711482678, + "learning_rate": 6.0262201152150315e-09, + "loss": 0.914, + "step": 41991 + }, + { + "epoch": 0.99, + "grad_norm": 1.154063104906125, + "learning_rate": 5.999762990437363e-09, + "loss": 1.0088, + "step": 41992 + }, + { + "epoch": 0.99, + "grad_norm": 1.915401888379952, + "learning_rate": 5.973364053892816e-09, + "loss": 0.9375, + "step": 41993 + }, + { + "epoch": 0.99, + "grad_norm": 2.176249101832623, + "learning_rate": 5.947023305735711e-09, + "loss": 1.0134, + "step": 41994 + }, + { + "epoch": 0.99, + "grad_norm": 2.117472569604359, + "learning_rate": 5.920740746119258e-09, + "loss": 0.8732, + "step": 41995 + }, + { + "epoch": 0.99, + "grad_norm": 1.9952047028208468, + "learning_rate": 5.894516375196668e-09, + "loss": 0.9894, + "step": 41996 + }, + { + "epoch": 0.99, + "grad_norm": 1.9259503667036588, + "learning_rate": 5.868350193120043e-09, + "loss": 1.0961, + "step": 41997 + }, + { + "epoch": 0.99, + "grad_norm": 1.9284314911640088, + "learning_rate": 5.842242200041481e-09, + "loss": 1.0161, + "step": 41998 + }, + { + "epoch": 0.99, + "grad_norm": 2.212776244360542, + "learning_rate": 5.816192396114195e-09, + "loss": 0.852, + "step": 41999 + }, + { + "epoch": 0.99, + "grad_norm": 2.0648249086806025, + "learning_rate": 5.790200781488065e-09, + "loss": 0.973, + "step": 42000 + }, + { + "epoch": 0.99, + "grad_norm": 1.9774195352977508, + "learning_rate": 5.7642673563163e-09, + "loss": 1.0209, + "step": 42001 + }, + { + "epoch": 0.99, + "grad_norm": 1.1886377109757111, + "learning_rate": 5.738392120749892e-09, + "loss": 0.9444, + "step": 42002 + }, + { + "epoch": 0.99, + "grad_norm": 2.6776952911970673, + "learning_rate": 5.712575074937609e-09, + "loss": 0.9481, + "step": 42003 + }, + { + "epoch": 0.99, + "grad_norm": 2.1348382266944848, + "learning_rate": 5.686816219030444e-09, + "loss": 0.9634, + "step": 42004 + }, + { + "epoch": 0.99, + "grad_norm": 1.9830607470019797, + "learning_rate": 5.661115553179386e-09, + "loss": 0.9508, + "step": 42005 + }, + { + "epoch": 0.99, + "grad_norm": 2.2852566126441474, + "learning_rate": 5.6354730775332046e-09, + "loss": 1.0249, + "step": 42006 + }, + { + "epoch": 0.99, + "grad_norm": 2.0648717509942767, + "learning_rate": 5.60988879224178e-09, + "loss": 0.9518, + "step": 42007 + }, + { + "epoch": 0.99, + "grad_norm": 2.0362939465523158, + "learning_rate": 5.584362697453882e-09, + "loss": 1.1096, + "step": 42008 + }, + { + "epoch": 0.99, + "grad_norm": 1.8948881901904548, + "learning_rate": 5.558894793318281e-09, + "loss": 0.938, + "step": 42009 + }, + { + "epoch": 0.99, + "grad_norm": 1.9722227761637754, + "learning_rate": 5.533485079982637e-09, + "loss": 0.9731, + "step": 42010 + }, + { + "epoch": 0.99, + "grad_norm": 1.8996259698607718, + "learning_rate": 5.508133557594608e-09, + "loss": 1.0916, + "step": 42011 + }, + { + "epoch": 0.99, + "grad_norm": 2.4061763496901363, + "learning_rate": 5.482840226302966e-09, + "loss": 1.0743, + "step": 42012 + }, + { + "epoch": 0.99, + "grad_norm": 1.9851598405467759, + "learning_rate": 5.457605086254258e-09, + "loss": 0.8997, + "step": 42013 + }, + { + "epoch": 0.99, + "grad_norm": 2.098190808246382, + "learning_rate": 5.4324281375961466e-09, + "loss": 1.0455, + "step": 42014 + }, + { + "epoch": 0.99, + "grad_norm": 2.05180362909446, + "learning_rate": 5.4073093804740685e-09, + "loss": 1.0698, + "step": 42015 + }, + { + "epoch": 0.99, + "grad_norm": 1.7937675737411483, + "learning_rate": 5.382248815034574e-09, + "loss": 0.9188, + "step": 42016 + }, + { + "epoch": 0.99, + "grad_norm": 1.9510793715296821, + "learning_rate": 5.357246441424213e-09, + "loss": 1.0938, + "step": 42017 + }, + { + "epoch": 0.99, + "grad_norm": 0.9852551164013731, + "learning_rate": 5.332302259788424e-09, + "loss": 0.9201, + "step": 42018 + }, + { + "epoch": 0.99, + "grad_norm": 1.9023326297235243, + "learning_rate": 5.307416270272647e-09, + "loss": 0.9564, + "step": 42019 + }, + { + "epoch": 0.99, + "grad_norm": 1.9522389272400187, + "learning_rate": 5.282588473021211e-09, + "loss": 0.9056, + "step": 42020 + }, + { + "epoch": 0.99, + "grad_norm": 2.1052210267063125, + "learning_rate": 5.257818868178444e-09, + "loss": 1.0506, + "step": 42021 + }, + { + "epoch": 0.99, + "grad_norm": 7.537808547714973, + "learning_rate": 5.2331074558886756e-09, + "loss": 0.9252, + "step": 42022 + }, + { + "epoch": 0.99, + "grad_norm": 1.8104819845044018, + "learning_rate": 5.208454236296234e-09, + "loss": 1.0051, + "step": 42023 + }, + { + "epoch": 0.99, + "grad_norm": 2.165633223417042, + "learning_rate": 5.18385920954434e-09, + "loss": 1.0183, + "step": 42024 + }, + { + "epoch": 0.99, + "grad_norm": 2.0795838559027096, + "learning_rate": 5.1593223757762104e-09, + "loss": 0.9953, + "step": 42025 + }, + { + "epoch": 0.99, + "grad_norm": 1.8671277133395905, + "learning_rate": 5.134843735135065e-09, + "loss": 1.0475, + "step": 42026 + }, + { + "epoch": 0.99, + "grad_norm": 1.8914866608258891, + "learning_rate": 5.110423287764121e-09, + "loss": 0.9207, + "step": 42027 + }, + { + "epoch": 0.99, + "grad_norm": 1.0858531671949103, + "learning_rate": 5.086061033803269e-09, + "loss": 0.9727, + "step": 42028 + }, + { + "epoch": 0.99, + "grad_norm": 2.245899357365507, + "learning_rate": 5.061756973396836e-09, + "loss": 1.0711, + "step": 42029 + }, + { + "epoch": 0.99, + "grad_norm": 1.9054298742981786, + "learning_rate": 5.037511106684712e-09, + "loss": 0.9741, + "step": 42030 + }, + { + "epoch": 0.99, + "grad_norm": 2.0007524772136946, + "learning_rate": 5.013323433807893e-09, + "loss": 1.0157, + "step": 42031 + }, + { + "epoch": 0.99, + "grad_norm": 1.9803249860204455, + "learning_rate": 4.989193954908489e-09, + "loss": 0.9063, + "step": 42032 + }, + { + "epoch": 0.99, + "grad_norm": 2.1075131498307966, + "learning_rate": 4.9651226701263874e-09, + "loss": 1.0441, + "step": 42033 + }, + { + "epoch": 0.99, + "grad_norm": 1.939275484698998, + "learning_rate": 4.9411095796014775e-09, + "loss": 0.8948, + "step": 42034 + }, + { + "epoch": 0.99, + "grad_norm": 2.3835578808895708, + "learning_rate": 4.917154683473646e-09, + "loss": 1.0413, + "step": 42035 + }, + { + "epoch": 0.99, + "grad_norm": 1.9351909339159201, + "learning_rate": 4.8932579818827816e-09, + "loss": 0.9672, + "step": 42036 + }, + { + "epoch": 0.99, + "grad_norm": 1.814932137984697, + "learning_rate": 4.869419474967663e-09, + "loss": 0.9782, + "step": 42037 + }, + { + "epoch": 0.99, + "grad_norm": 1.8441805620935383, + "learning_rate": 4.845639162868177e-09, + "loss": 0.9009, + "step": 42038 + }, + { + "epoch": 0.99, + "grad_norm": 2.059034464921967, + "learning_rate": 4.821917045720881e-09, + "loss": 0.9254, + "step": 42039 + }, + { + "epoch": 0.99, + "grad_norm": 1.9554013407022972, + "learning_rate": 4.7982531236645535e-09, + "loss": 1.0465, + "step": 42040 + }, + { + "epoch": 0.99, + "grad_norm": 2.032942911228608, + "learning_rate": 4.774647396837973e-09, + "loss": 0.8151, + "step": 42041 + }, + { + "epoch": 0.99, + "grad_norm": 2.0323801745051533, + "learning_rate": 4.751099865376585e-09, + "loss": 1.0265, + "step": 42042 + }, + { + "epoch": 0.99, + "grad_norm": 1.8453353849781804, + "learning_rate": 4.727610529419169e-09, + "loss": 0.8691, + "step": 42043 + }, + { + "epoch": 0.99, + "grad_norm": 2.1937208494543143, + "learning_rate": 4.704179389103391e-09, + "loss": 1.1148, + "step": 42044 + }, + { + "epoch": 0.99, + "grad_norm": 1.9972753949945714, + "learning_rate": 4.6808064445635905e-09, + "loss": 0.9657, + "step": 42045 + }, + { + "epoch": 0.99, + "grad_norm": 2.051420982465964, + "learning_rate": 4.6574916959363226e-09, + "loss": 1.018, + "step": 42046 + }, + { + "epoch": 0.99, + "grad_norm": 1.9426200050252762, + "learning_rate": 4.634235143357036e-09, + "loss": 0.9027, + "step": 42047 + }, + { + "epoch": 0.99, + "grad_norm": 1.954464798943449, + "learning_rate": 4.611036786962286e-09, + "loss": 0.9273, + "step": 42048 + }, + { + "epoch": 0.99, + "grad_norm": 1.9861152892643759, + "learning_rate": 4.587896626886412e-09, + "loss": 1.0402, + "step": 42049 + }, + { + "epoch": 0.99, + "grad_norm": 1.9411115967237111, + "learning_rate": 4.5648146632648605e-09, + "loss": 0.8148, + "step": 42050 + }, + { + "epoch": 0.99, + "grad_norm": 2.162098173174489, + "learning_rate": 4.541790896230858e-09, + "loss": 1.0609, + "step": 42051 + }, + { + "epoch": 0.99, + "grad_norm": 2.3142696613720255, + "learning_rate": 4.518825325919851e-09, + "loss": 1.042, + "step": 42052 + }, + { + "epoch": 0.99, + "grad_norm": 2.2406799412502876, + "learning_rate": 4.495917952463957e-09, + "loss": 0.9515, + "step": 42053 + }, + { + "epoch": 0.99, + "grad_norm": 2.015461893029522, + "learning_rate": 4.473068775997513e-09, + "loss": 0.9312, + "step": 42054 + }, + { + "epoch": 0.99, + "grad_norm": 1.8481478596585537, + "learning_rate": 4.450277796653746e-09, + "loss": 0.9532, + "step": 42055 + }, + { + "epoch": 0.99, + "grad_norm": 1.8828442492840323, + "learning_rate": 4.427545014564771e-09, + "loss": 1.029, + "step": 42056 + }, + { + "epoch": 0.99, + "grad_norm": 1.8919051455518634, + "learning_rate": 4.404870429862706e-09, + "loss": 0.9545, + "step": 42057 + }, + { + "epoch": 0.99, + "grad_norm": 1.013708780283902, + "learning_rate": 4.382254042680778e-09, + "loss": 0.9391, + "step": 42058 + }, + { + "epoch": 0.99, + "grad_norm": 2.017537076138743, + "learning_rate": 4.359695853149992e-09, + "loss": 0.9409, + "step": 42059 + }, + { + "epoch": 0.99, + "grad_norm": 1.153196485051076, + "learning_rate": 4.337195861400245e-09, + "loss": 0.9858, + "step": 42060 + }, + { + "epoch": 0.99, + "grad_norm": 1.927029307845587, + "learning_rate": 4.314754067565874e-09, + "loss": 0.9782, + "step": 42061 + }, + { + "epoch": 0.99, + "grad_norm": 1.923061322840415, + "learning_rate": 4.292370471774554e-09, + "loss": 1.0903, + "step": 42062 + }, + { + "epoch": 0.99, + "grad_norm": 2.10600310672565, + "learning_rate": 4.270045074157292e-09, + "loss": 0.9452, + "step": 42063 + }, + { + "epoch": 0.99, + "grad_norm": 1.8514739749264162, + "learning_rate": 4.247777874845094e-09, + "loss": 0.8877, + "step": 42064 + }, + { + "epoch": 0.99, + "grad_norm": 1.945460868734772, + "learning_rate": 4.225568873966746e-09, + "loss": 0.9899, + "step": 42065 + }, + { + "epoch": 0.99, + "grad_norm": 2.080427330606539, + "learning_rate": 4.203418071652143e-09, + "loss": 1.0364, + "step": 42066 + }, + { + "epoch": 0.99, + "grad_norm": 2.0293949740722215, + "learning_rate": 4.181325468028963e-09, + "loss": 0.9808, + "step": 42067 + }, + { + "epoch": 0.99, + "grad_norm": 1.881476629535721, + "learning_rate": 4.1592910632271e-09, + "loss": 0.9245, + "step": 42068 + }, + { + "epoch": 0.99, + "grad_norm": 1.9440302452835738, + "learning_rate": 4.13731485737423e-09, + "loss": 1.0521, + "step": 42069 + }, + { + "epoch": 0.99, + "grad_norm": 2.068945774675928, + "learning_rate": 4.11539685059914e-09, + "loss": 0.8393, + "step": 42070 + }, + { + "epoch": 0.99, + "grad_norm": 2.0769742603250787, + "learning_rate": 4.093537043028395e-09, + "loss": 0.7843, + "step": 42071 + }, + { + "epoch": 0.99, + "grad_norm": 1.0993046908022475, + "learning_rate": 4.07173543478967e-09, + "loss": 0.9827, + "step": 42072 + }, + { + "epoch": 0.99, + "grad_norm": 1.998888899902601, + "learning_rate": 4.04999202601064e-09, + "loss": 0.8581, + "step": 42073 + }, + { + "epoch": 0.99, + "grad_norm": 2.3527810221518757, + "learning_rate": 4.028306816816763e-09, + "loss": 0.9446, + "step": 42074 + }, + { + "epoch": 0.99, + "grad_norm": 2.011240380721871, + "learning_rate": 4.006679807334601e-09, + "loss": 0.8969, + "step": 42075 + }, + { + "epoch": 0.99, + "grad_norm": 1.967395488706129, + "learning_rate": 3.985110997690722e-09, + "loss": 0.9952, + "step": 42076 + }, + { + "epoch": 0.99, + "grad_norm": 2.007783286089937, + "learning_rate": 3.963600388009469e-09, + "loss": 0.963, + "step": 42077 + }, + { + "epoch": 0.99, + "grad_norm": 2.0225061758545566, + "learning_rate": 3.94214797841741e-09, + "loss": 0.9794, + "step": 42078 + }, + { + "epoch": 0.99, + "grad_norm": 1.948852047799201, + "learning_rate": 3.920753769038888e-09, + "loss": 0.9923, + "step": 42079 + }, + { + "epoch": 0.99, + "grad_norm": 2.3178593140743358, + "learning_rate": 3.899417759998248e-09, + "loss": 0.9792, + "step": 42080 + }, + { + "epoch": 0.99, + "grad_norm": 1.9455287259742025, + "learning_rate": 3.878139951419835e-09, + "loss": 0.8409, + "step": 42081 + }, + { + "epoch": 0.99, + "grad_norm": 2.252891899646859, + "learning_rate": 3.856920343427994e-09, + "loss": 0.9679, + "step": 42082 + }, + { + "epoch": 0.99, + "grad_norm": 1.9529630711902335, + "learning_rate": 3.835758936145961e-09, + "loss": 0.997, + "step": 42083 + }, + { + "epoch": 0.99, + "grad_norm": 1.891587494643743, + "learning_rate": 3.81465572969586e-09, + "loss": 0.9673, + "step": 42084 + }, + { + "epoch": 0.99, + "grad_norm": 2.01146820253467, + "learning_rate": 3.793610724202035e-09, + "loss": 1.0373, + "step": 42085 + }, + { + "epoch": 0.99, + "grad_norm": 1.8924064271088141, + "learning_rate": 3.772623919786611e-09, + "loss": 0.8414, + "step": 42086 + }, + { + "epoch": 0.99, + "grad_norm": 1.812137173719833, + "learning_rate": 3.751695316572823e-09, + "loss": 0.9553, + "step": 42087 + }, + { + "epoch": 0.99, + "grad_norm": 2.6213553384457975, + "learning_rate": 3.730824914679465e-09, + "loss": 0.9532, + "step": 42088 + }, + { + "epoch": 0.99, + "grad_norm": 2.8918398088561794, + "learning_rate": 3.7100127142308816e-09, + "loss": 0.8198, + "step": 42089 + }, + { + "epoch": 0.99, + "grad_norm": 1.9035841142236685, + "learning_rate": 3.6892587153480874e-09, + "loss": 0.9514, + "step": 42090 + }, + { + "epoch": 0.99, + "grad_norm": 1.8717315519804487, + "learning_rate": 3.6685629181509864e-09, + "loss": 1.0183, + "step": 42091 + }, + { + "epoch": 0.99, + "grad_norm": 1.8590491292550495, + "learning_rate": 3.6479253227594825e-09, + "loss": 0.8349, + "step": 42092 + }, + { + "epoch": 0.99, + "grad_norm": 2.0839336031786244, + "learning_rate": 3.62734592929459e-09, + "loss": 0.8632, + "step": 42093 + }, + { + "epoch": 0.99, + "grad_norm": 1.958962252578495, + "learning_rate": 3.6068247378773236e-09, + "loss": 0.9298, + "step": 42094 + }, + { + "epoch": 0.99, + "grad_norm": 2.0849451066695903, + "learning_rate": 3.5863617486242565e-09, + "loss": 0.9182, + "step": 42095 + }, + { + "epoch": 0.99, + "grad_norm": 2.4055128491424007, + "learning_rate": 3.565956961657513e-09, + "loss": 0.8698, + "step": 42096 + }, + { + "epoch": 0.99, + "grad_norm": 1.1421141630016003, + "learning_rate": 3.5456103770936668e-09, + "loss": 0.9129, + "step": 42097 + }, + { + "epoch": 0.99, + "grad_norm": 2.0317783440002524, + "learning_rate": 3.525321995052622e-09, + "loss": 0.8664, + "step": 42098 + }, + { + "epoch": 0.99, + "grad_norm": 2.1214252655991293, + "learning_rate": 3.5050918156520618e-09, + "loss": 1.0518, + "step": 42099 + }, + { + "epoch": 0.99, + "grad_norm": 2.0175507268725585, + "learning_rate": 3.4849198390096705e-09, + "loss": 1.1285, + "step": 42100 + }, + { + "epoch": 0.99, + "grad_norm": 1.9377424792339928, + "learning_rate": 3.4648060652431314e-09, + "loss": 0.9732, + "step": 42101 + }, + { + "epoch": 0.99, + "grad_norm": 1.9167953038855412, + "learning_rate": 3.444750494469018e-09, + "loss": 0.9365, + "step": 42102 + }, + { + "epoch": 0.99, + "grad_norm": 1.9357522534095333, + "learning_rate": 3.4247531268039034e-09, + "loss": 0.9935, + "step": 42103 + }, + { + "epoch": 0.99, + "grad_norm": 1.908362909211669, + "learning_rate": 3.404813962365472e-09, + "loss": 0.978, + "step": 42104 + }, + { + "epoch": 0.99, + "grad_norm": 1.8433367338137028, + "learning_rate": 3.384933001269186e-09, + "loss": 0.9979, + "step": 42105 + }, + { + "epoch": 0.99, + "grad_norm": 2.17460639048086, + "learning_rate": 3.3651102436293993e-09, + "loss": 0.8994, + "step": 42106 + }, + { + "epoch": 0.99, + "grad_norm": 2.8107134128577718, + "learning_rate": 3.345345689563795e-09, + "loss": 1.0177, + "step": 42107 + }, + { + "epoch": 0.99, + "grad_norm": 1.9493145519591002, + "learning_rate": 3.325639339186726e-09, + "loss": 0.9248, + "step": 42108 + }, + { + "epoch": 0.99, + "grad_norm": 1.0945918068736233, + "learning_rate": 3.3059911926125455e-09, + "loss": 0.9131, + "step": 42109 + }, + { + "epoch": 0.99, + "grad_norm": 2.1346187842072735, + "learning_rate": 3.2864012499556066e-09, + "loss": 1.0464, + "step": 42110 + }, + { + "epoch": 0.99, + "grad_norm": 2.111725012174213, + "learning_rate": 3.266869511330262e-09, + "loss": 1.0931, + "step": 42111 + }, + { + "epoch": 0.99, + "grad_norm": 2.1935085714445, + "learning_rate": 3.2473959768497542e-09, + "loss": 0.9803, + "step": 42112 + }, + { + "epoch": 0.99, + "grad_norm": 1.1163259022123138, + "learning_rate": 3.227980646628437e-09, + "loss": 0.9527, + "step": 42113 + }, + { + "epoch": 0.99, + "grad_norm": 1.896148285577106, + "learning_rate": 3.2086235207784423e-09, + "loss": 0.982, + "step": 42114 + }, + { + "epoch": 0.99, + "grad_norm": 2.123138063103437, + "learning_rate": 3.189324599413013e-09, + "loss": 0.9648, + "step": 42115 + }, + { + "epoch": 0.99, + "grad_norm": 1.8054038637077738, + "learning_rate": 3.170083882643171e-09, + "loss": 0.8768, + "step": 42116 + }, + { + "epoch": 0.99, + "grad_norm": 2.143775656787453, + "learning_rate": 3.1509013705832703e-09, + "loss": 1.0127, + "step": 42117 + }, + { + "epoch": 0.99, + "grad_norm": 1.8468455187728674, + "learning_rate": 3.131777063343222e-09, + "loss": 0.9779, + "step": 42118 + }, + { + "epoch": 0.99, + "grad_norm": 2.389243870694648, + "learning_rate": 3.1127109610351593e-09, + "loss": 0.8299, + "step": 42119 + }, + { + "epoch": 0.99, + "grad_norm": 1.1077277791866182, + "learning_rate": 3.0937030637701037e-09, + "loss": 0.9104, + "step": 42120 + }, + { + "epoch": 0.99, + "grad_norm": 2.168483980273791, + "learning_rate": 3.074753371657968e-09, + "loss": 1.0592, + "step": 42121 + }, + { + "epoch": 0.99, + "grad_norm": 1.8856557532898226, + "learning_rate": 3.0558618848097744e-09, + "loss": 0.8539, + "step": 42122 + }, + { + "epoch": 0.99, + "grad_norm": 1.0823879126440277, + "learning_rate": 3.0370286033354348e-09, + "loss": 0.9406, + "step": 42123 + }, + { + "epoch": 0.99, + "grad_norm": 1.884337503353142, + "learning_rate": 3.018253527343751e-09, + "loss": 0.8924, + "step": 42124 + }, + { + "epoch": 0.99, + "grad_norm": 2.135659576808086, + "learning_rate": 2.999536656945745e-09, + "loss": 0.9958, + "step": 42125 + }, + { + "epoch": 0.99, + "grad_norm": 2.5313407473188225, + "learning_rate": 2.980877992249109e-09, + "loss": 1.0268, + "step": 42126 + }, + { + "epoch": 0.99, + "grad_norm": 1.7907957085789288, + "learning_rate": 2.9622775333626453e-09, + "loss": 1.019, + "step": 42127 + }, + { + "epoch": 0.99, + "grad_norm": 1.8820076341006415, + "learning_rate": 2.9437352803951546e-09, + "loss": 0.8485, + "step": 42128 + }, + { + "epoch": 0.99, + "grad_norm": 1.9106561102054422, + "learning_rate": 2.9252512334543293e-09, + "loss": 0.8637, + "step": 42129 + }, + { + "epoch": 0.99, + "grad_norm": 1.8746365194865287, + "learning_rate": 2.9068253926478605e-09, + "loss": 1.119, + "step": 42130 + }, + { + "epoch": 0.99, + "grad_norm": 2.175856566281283, + "learning_rate": 2.8884577580823304e-09, + "loss": 0.9958, + "step": 42131 + }, + { + "epoch": 0.99, + "grad_norm": 1.9690392254907956, + "learning_rate": 2.87014832986543e-09, + "loss": 0.9828, + "step": 42132 + }, + { + "epoch": 0.99, + "grad_norm": 2.350896951641573, + "learning_rate": 2.8518971081037403e-09, + "loss": 0.9472, + "step": 42133 + }, + { + "epoch": 0.99, + "grad_norm": 2.275701076639062, + "learning_rate": 2.8337040929027336e-09, + "loss": 1.0506, + "step": 42134 + }, + { + "epoch": 0.99, + "grad_norm": 1.8543059270337274, + "learning_rate": 2.8155692843689907e-09, + "loss": 0.9004, + "step": 42135 + }, + { + "epoch": 0.99, + "grad_norm": 2.0793963420883146, + "learning_rate": 2.797492682609093e-09, + "loss": 1.1835, + "step": 42136 + }, + { + "epoch": 0.99, + "grad_norm": 2.270530827108254, + "learning_rate": 2.7794742877262914e-09, + "loss": 1.012, + "step": 42137 + }, + { + "epoch": 0.99, + "grad_norm": 1.9639477947997959, + "learning_rate": 2.761514099826057e-09, + "loss": 0.8961, + "step": 42138 + }, + { + "epoch": 0.99, + "grad_norm": 2.335802990621355, + "learning_rate": 2.743612119014971e-09, + "loss": 0.9987, + "step": 42139 + }, + { + "epoch": 0.99, + "grad_norm": 2.0513651419660897, + "learning_rate": 2.725768345394064e-09, + "loss": 0.9542, + "step": 42140 + }, + { + "epoch": 0.99, + "grad_norm": 2.0343869571143958, + "learning_rate": 2.7079827790699177e-09, + "loss": 1.0714, + "step": 42141 + }, + { + "epoch": 0.99, + "grad_norm": 2.032276723729322, + "learning_rate": 2.690255420144672e-09, + "loss": 1.0049, + "step": 42142 + }, + { + "epoch": 0.99, + "grad_norm": 2.1261492820235306, + "learning_rate": 2.6725862687215774e-09, + "loss": 0.9675, + "step": 42143 + }, + { + "epoch": 0.99, + "grad_norm": 2.106057605446874, + "learning_rate": 2.6549753249038858e-09, + "loss": 1.0598, + "step": 42144 + }, + { + "epoch": 0.99, + "grad_norm": 2.0841235577556603, + "learning_rate": 2.6374225887937365e-09, + "loss": 1.0161, + "step": 42145 + }, + { + "epoch": 0.99, + "grad_norm": 2.070423644159434, + "learning_rate": 2.619928060494381e-09, + "loss": 0.9092, + "step": 42146 + }, + { + "epoch": 0.99, + "grad_norm": 1.9188233107447845, + "learning_rate": 2.6024917401068495e-09, + "loss": 0.9156, + "step": 42147 + }, + { + "epoch": 0.99, + "grad_norm": 2.0475198101131826, + "learning_rate": 2.585113627732172e-09, + "loss": 0.9444, + "step": 42148 + }, + { + "epoch": 0.99, + "grad_norm": 2.328023598290021, + "learning_rate": 2.567793723471379e-09, + "loss": 0.941, + "step": 42149 + }, + { + "epoch": 0.99, + "grad_norm": 1.0926627109626161, + "learning_rate": 2.550532027427721e-09, + "loss": 0.9723, + "step": 42150 + }, + { + "epoch": 0.99, + "grad_norm": 2.0456737165010734, + "learning_rate": 2.533328539698898e-09, + "loss": 1.0015, + "step": 42151 + }, + { + "epoch": 0.99, + "grad_norm": 2.3500467660879787, + "learning_rate": 2.5161832603870506e-09, + "loss": 0.8877, + "step": 42152 + }, + { + "epoch": 0.99, + "grad_norm": 1.1477332565120273, + "learning_rate": 2.499096189589878e-09, + "loss": 0.9341, + "step": 42153 + }, + { + "epoch": 0.99, + "grad_norm": 2.023360491404746, + "learning_rate": 2.4820673274095207e-09, + "loss": 0.9146, + "step": 42154 + }, + { + "epoch": 0.99, + "grad_norm": 1.9205213500497884, + "learning_rate": 2.4650966739425686e-09, + "loss": 0.8833, + "step": 42155 + }, + { + "epoch": 0.99, + "grad_norm": 2.147751588947647, + "learning_rate": 2.4481842292900517e-09, + "loss": 0.9023, + "step": 42156 + }, + { + "epoch": 0.99, + "grad_norm": 2.4677514493971833, + "learning_rate": 2.431329993548559e-09, + "loss": 0.8783, + "step": 42157 + }, + { + "epoch": 0.99, + "grad_norm": 1.923963282044285, + "learning_rate": 2.414533966816901e-09, + "loss": 0.9241, + "step": 42158 + }, + { + "epoch": 0.99, + "grad_norm": 2.3231522959435447, + "learning_rate": 2.3977961491938874e-09, + "loss": 0.9005, + "step": 42159 + }, + { + "epoch": 0.99, + "grad_norm": 1.8843258868526132, + "learning_rate": 2.381116540774997e-09, + "loss": 0.9372, + "step": 42160 + }, + { + "epoch": 0.99, + "grad_norm": 2.0277914655850027, + "learning_rate": 2.36449514165904e-09, + "loss": 1.0135, + "step": 42161 + }, + { + "epoch": 0.99, + "grad_norm": 1.0507586843033436, + "learning_rate": 2.347931951942606e-09, + "loss": 0.9217, + "step": 42162 + }, + { + "epoch": 0.99, + "grad_norm": 1.9843112758622832, + "learning_rate": 2.3314269717211735e-09, + "loss": 0.8657, + "step": 42163 + }, + { + "epoch": 0.99, + "grad_norm": 2.129391599809273, + "learning_rate": 2.3149802010913323e-09, + "loss": 1.044, + "step": 42164 + }, + { + "epoch": 0.99, + "grad_norm": 2.10850474195406, + "learning_rate": 2.298591640149672e-09, + "loss": 1.0845, + "step": 42165 + }, + { + "epoch": 0.99, + "grad_norm": 1.9542716310701802, + "learning_rate": 2.2822612889894513e-09, + "loss": 0.9575, + "step": 42166 + }, + { + "epoch": 0.99, + "grad_norm": 2.016454934707822, + "learning_rate": 2.2659891477083694e-09, + "loss": 0.9363, + "step": 42167 + }, + { + "epoch": 0.99, + "grad_norm": 1.9215390249836903, + "learning_rate": 2.2497752163996855e-09, + "loss": 0.9631, + "step": 42168 + }, + { + "epoch": 0.99, + "grad_norm": 2.6905553662911217, + "learning_rate": 2.2336194951577685e-09, + "loss": 1.0517, + "step": 42169 + }, + { + "epoch": 0.99, + "grad_norm": 2.7014550105821495, + "learning_rate": 2.217521984076987e-09, + "loss": 0.936, + "step": 42170 + }, + { + "epoch": 0.99, + "grad_norm": 2.307864140216579, + "learning_rate": 2.2014826832506e-09, + "loss": 1.0202, + "step": 42171 + }, + { + "epoch": 0.99, + "grad_norm": 1.9849506262766938, + "learning_rate": 2.185501592772976e-09, + "loss": 0.9659, + "step": 42172 + }, + { + "epoch": 0.99, + "grad_norm": 2.1313101445116356, + "learning_rate": 2.1695787127362644e-09, + "loss": 1.0729, + "step": 42173 + }, + { + "epoch": 0.99, + "grad_norm": 1.9439219429096, + "learning_rate": 2.1537140432348335e-09, + "loss": 0.9498, + "step": 42174 + }, + { + "epoch": 0.99, + "grad_norm": 2.0586170690824503, + "learning_rate": 2.1379075843586116e-09, + "loss": 0.9268, + "step": 42175 + }, + { + "epoch": 0.99, + "grad_norm": 1.0601535172135588, + "learning_rate": 2.1221593362008575e-09, + "loss": 0.9717, + "step": 42176 + }, + { + "epoch": 0.99, + "grad_norm": 2.056179565807967, + "learning_rate": 2.1064692988526093e-09, + "loss": 1.0262, + "step": 42177 + }, + { + "epoch": 0.99, + "grad_norm": 2.3940582426619885, + "learning_rate": 2.0908374724060156e-09, + "loss": 0.94, + "step": 42178 + }, + { + "epoch": 0.99, + "grad_norm": 1.9922371598673903, + "learning_rate": 2.0752638569521144e-09, + "loss": 1.0316, + "step": 42179 + }, + { + "epoch": 0.99, + "grad_norm": 2.066800243309692, + "learning_rate": 2.0597484525819446e-09, + "loss": 0.9997, + "step": 42180 + }, + { + "epoch": 0.99, + "grad_norm": 1.888549003578713, + "learning_rate": 2.0442912593843233e-09, + "loss": 0.9386, + "step": 42181 + }, + { + "epoch": 0.99, + "grad_norm": 2.0673951139385913, + "learning_rate": 2.0288922774502896e-09, + "loss": 0.8842, + "step": 42182 + }, + { + "epoch": 0.99, + "grad_norm": 2.1382929454272124, + "learning_rate": 2.013551506869771e-09, + "loss": 1.0622, + "step": 42183 + }, + { + "epoch": 0.99, + "grad_norm": 2.684833823128356, + "learning_rate": 1.9982689477315854e-09, + "loss": 0.976, + "step": 42184 + }, + { + "epoch": 0.99, + "grad_norm": 1.1387012001076133, + "learning_rate": 1.9830446001245507e-09, + "loss": 0.9126, + "step": 42185 + }, + { + "epoch": 0.99, + "grad_norm": 2.2754824746712132, + "learning_rate": 1.967878464137485e-09, + "loss": 0.9062, + "step": 42186 + }, + { + "epoch": 0.99, + "grad_norm": 2.0510178141163955, + "learning_rate": 1.9527705398592056e-09, + "loss": 0.9618, + "step": 42187 + }, + { + "epoch": 0.99, + "grad_norm": 2.229381290443927, + "learning_rate": 1.9377208273763105e-09, + "loss": 0.962, + "step": 42188 + }, + { + "epoch": 0.99, + "grad_norm": 1.948869606323249, + "learning_rate": 1.9227293267787272e-09, + "loss": 1.0748, + "step": 42189 + }, + { + "epoch": 0.99, + "grad_norm": 2.48474342347967, + "learning_rate": 1.9077960381508333e-09, + "loss": 0.873, + "step": 42190 + }, + { + "epoch": 0.99, + "grad_norm": 1.9434534028563173, + "learning_rate": 1.892920961582556e-09, + "loss": 0.968, + "step": 42191 + }, + { + "epoch": 0.99, + "grad_norm": 2.0749187788997134, + "learning_rate": 1.878104097158273e-09, + "loss": 1.0302, + "step": 42192 + }, + { + "epoch": 0.99, + "grad_norm": 1.844467397940422, + "learning_rate": 1.8633454449645814e-09, + "loss": 0.8988, + "step": 42193 + }, + { + "epoch": 0.99, + "grad_norm": 2.2959160791901354, + "learning_rate": 1.8486450050880788e-09, + "loss": 0.9085, + "step": 42194 + }, + { + "epoch": 0.99, + "grad_norm": 1.1210766860156725, + "learning_rate": 1.8340027776142521e-09, + "loss": 0.9049, + "step": 42195 + }, + { + "epoch": 0.99, + "grad_norm": 1.8382212597858205, + "learning_rate": 1.8194187626285887e-09, + "loss": 1.0467, + "step": 42196 + }, + { + "epoch": 0.99, + "grad_norm": 1.9935496777043331, + "learning_rate": 1.8048929602154652e-09, + "loss": 0.9527, + "step": 42197 + }, + { + "epoch": 0.99, + "grad_norm": 1.9907408205400832, + "learning_rate": 1.790425370459259e-09, + "loss": 1.0466, + "step": 42198 + }, + { + "epoch": 0.99, + "grad_norm": 1.9706671281916361, + "learning_rate": 1.7760159934454568e-09, + "loss": 0.9705, + "step": 42199 + }, + { + "epoch": 0.99, + "grad_norm": 2.254932543754252, + "learning_rate": 1.7616648292562154e-09, + "loss": 1.0305, + "step": 42200 + }, + { + "epoch": 0.99, + "grad_norm": 2.149054991892529, + "learning_rate": 1.7473718779759118e-09, + "loss": 1.0852, + "step": 42201 + }, + { + "epoch": 0.99, + "grad_norm": 1.9214841431392506, + "learning_rate": 1.733137139688923e-09, + "loss": 0.967, + "step": 42202 + }, + { + "epoch": 0.99, + "grad_norm": 1.895289156662687, + "learning_rate": 1.718960614476295e-09, + "loss": 1.019, + "step": 42203 + }, + { + "epoch": 0.99, + "grad_norm": 2.0178456377797804, + "learning_rate": 1.7048423024201843e-09, + "loss": 0.9044, + "step": 42204 + }, + { + "epoch": 0.99, + "grad_norm": 1.9992763640443263, + "learning_rate": 1.6907822036049682e-09, + "loss": 0.8946, + "step": 42205 + }, + { + "epoch": 0.99, + "grad_norm": 1.0481547434101504, + "learning_rate": 1.6767803181116927e-09, + "loss": 0.9165, + "step": 42206 + }, + { + "epoch": 0.99, + "grad_norm": 1.8291646068979086, + "learning_rate": 1.6628366460214041e-09, + "loss": 0.9958, + "step": 42207 + }, + { + "epoch": 0.99, + "grad_norm": 1.9217894901465762, + "learning_rate": 1.6489511874151488e-09, + "loss": 0.8122, + "step": 42208 + }, + { + "epoch": 0.99, + "grad_norm": 1.8485288406873601, + "learning_rate": 1.6351239423739728e-09, + "loss": 0.9049, + "step": 42209 + }, + { + "epoch": 0.99, + "grad_norm": 1.9540743864058014, + "learning_rate": 1.621354910978923e-09, + "loss": 0.977, + "step": 42210 + }, + { + "epoch": 0.99, + "grad_norm": 1.7614009165665132, + "learning_rate": 1.6076440933099347e-09, + "loss": 0.9172, + "step": 42211 + }, + { + "epoch": 0.99, + "grad_norm": 1.975979799084352, + "learning_rate": 1.5939914894469444e-09, + "loss": 1.0314, + "step": 42212 + }, + { + "epoch": 0.99, + "grad_norm": 1.9394549031143835, + "learning_rate": 1.580397099468778e-09, + "loss": 1.0095, + "step": 42213 + }, + { + "epoch": 0.99, + "grad_norm": 2.0759777039300356, + "learning_rate": 1.5668609234542608e-09, + "loss": 1.0376, + "step": 42214 + }, + { + "epoch": 0.99, + "grad_norm": 2.399075592052219, + "learning_rate": 1.5533829614844397e-09, + "loss": 0.9633, + "step": 42215 + }, + { + "epoch": 0.99, + "grad_norm": 2.1284988447682314, + "learning_rate": 1.5399632136348096e-09, + "loss": 0.9778, + "step": 42216 + }, + { + "epoch": 0.99, + "grad_norm": 1.037745893323719, + "learning_rate": 1.526601679986417e-09, + "loss": 0.9308, + "step": 42217 + }, + { + "epoch": 0.99, + "grad_norm": 2.093528259980496, + "learning_rate": 1.5132983606147566e-09, + "loss": 0.9434, + "step": 42218 + }, + { + "epoch": 0.99, + "grad_norm": 2.013658416966237, + "learning_rate": 1.5000532555975444e-09, + "loss": 0.8998, + "step": 42219 + }, + { + "epoch": 0.99, + "grad_norm": 1.9613115377347736, + "learning_rate": 1.4868663650136062e-09, + "loss": 0.8758, + "step": 42220 + }, + { + "epoch": 0.99, + "grad_norm": 2.010827936810006, + "learning_rate": 1.473737688938437e-09, + "loss": 0.9857, + "step": 42221 + }, + { + "epoch": 0.99, + "grad_norm": 1.9731165887014754, + "learning_rate": 1.460667227447532e-09, + "loss": 0.9557, + "step": 42222 + }, + { + "epoch": 0.99, + "grad_norm": 2.103371790366584, + "learning_rate": 1.447654980619717e-09, + "loss": 0.9638, + "step": 42223 + }, + { + "epoch": 0.99, + "grad_norm": 1.0738885177936985, + "learning_rate": 1.434700948528267e-09, + "loss": 0.9144, + "step": 42224 + }, + { + "epoch": 0.99, + "grad_norm": 2.043023446387518, + "learning_rate": 1.421805131248677e-09, + "loss": 0.9504, + "step": 42225 + }, + { + "epoch": 0.99, + "grad_norm": 2.2494943898392394, + "learning_rate": 1.4089675288586623e-09, + "loss": 0.9457, + "step": 42226 + }, + { + "epoch": 0.99, + "grad_norm": 2.3118611531440454, + "learning_rate": 1.3961881414292776e-09, + "loss": 0.9999, + "step": 42227 + }, + { + "epoch": 0.99, + "grad_norm": 1.7524743835671102, + "learning_rate": 1.3834669690382385e-09, + "loss": 1.0957, + "step": 42228 + }, + { + "epoch": 0.99, + "grad_norm": 3.5387855274628857, + "learning_rate": 1.3708040117577093e-09, + "loss": 1.08, + "step": 42229 + }, + { + "epoch": 0.99, + "grad_norm": 2.547159600665517, + "learning_rate": 1.358199269662075e-09, + "loss": 1.101, + "step": 42230 + }, + { + "epoch": 0.99, + "grad_norm": 2.2107504569252097, + "learning_rate": 1.3456527428235e-09, + "loss": 1.0232, + "step": 42231 + }, + { + "epoch": 0.99, + "grad_norm": 2.2938614081282713, + "learning_rate": 1.3331644313174796e-09, + "loss": 0.9449, + "step": 42232 + }, + { + "epoch": 0.99, + "grad_norm": 1.983277180924457, + "learning_rate": 1.320734335213958e-09, + "loss": 1.0687, + "step": 42233 + }, + { + "epoch": 0.99, + "grad_norm": 2.083910761460803, + "learning_rate": 1.3083624545873197e-09, + "loss": 1.0543, + "step": 42234 + }, + { + "epoch": 1.0, + "grad_norm": 1.9974702670674063, + "learning_rate": 1.2960487895097295e-09, + "loss": 1.1048, + "step": 42235 + }, + { + "epoch": 1.0, + "grad_norm": 1.8554569715828741, + "learning_rate": 1.2837933400511315e-09, + "loss": 0.9389, + "step": 42236 + }, + { + "epoch": 1.0, + "grad_norm": 1.133984551653945, + "learning_rate": 1.2715961062836902e-09, + "loss": 0.9276, + "step": 42237 + }, + { + "epoch": 1.0, + "grad_norm": 1.965848518174919, + "learning_rate": 1.2594570882784595e-09, + "loss": 0.9993, + "step": 42238 + }, + { + "epoch": 1.0, + "grad_norm": 2.126325833893543, + "learning_rate": 1.2473762861064943e-09, + "loss": 0.8742, + "step": 42239 + }, + { + "epoch": 1.0, + "grad_norm": 1.1126563914028664, + "learning_rate": 1.2353536998388483e-09, + "loss": 0.8838, + "step": 42240 + }, + { + "epoch": 1.0, + "grad_norm": 2.0784709808678867, + "learning_rate": 1.2233893295432453e-09, + "loss": 1.0384, + "step": 42241 + }, + { + "epoch": 1.0, + "grad_norm": 2.487682499927064, + "learning_rate": 1.2114831752907397e-09, + "loss": 0.9285, + "step": 42242 + }, + { + "epoch": 1.0, + "grad_norm": 1.9467336811463205, + "learning_rate": 1.1996352371512755e-09, + "loss": 1.0466, + "step": 42243 + }, + { + "epoch": 1.0, + "grad_norm": 2.196277786300567, + "learning_rate": 1.187845515192576e-09, + "loss": 0.9527, + "step": 42244 + }, + { + "epoch": 1.0, + "grad_norm": 2.1265026170198382, + "learning_rate": 1.1761140094834756e-09, + "loss": 0.9905, + "step": 42245 + }, + { + "epoch": 1.0, + "grad_norm": 1.9537154252985518, + "learning_rate": 1.1644407200939177e-09, + "loss": 1.0315, + "step": 42246 + }, + { + "epoch": 1.0, + "grad_norm": 1.9880216591586122, + "learning_rate": 1.152825647089406e-09, + "loss": 1.0493, + "step": 42247 + }, + { + "epoch": 1.0, + "grad_norm": 2.0501125989496662, + "learning_rate": 1.1412687905387743e-09, + "loss": 0.8425, + "step": 42248 + }, + { + "epoch": 1.0, + "grad_norm": 1.7808465827592062, + "learning_rate": 1.1297701505097459e-09, + "loss": 0.8594, + "step": 42249 + }, + { + "epoch": 1.0, + "grad_norm": 2.041467013478491, + "learning_rate": 1.1183297270678239e-09, + "loss": 1.1585, + "step": 42250 + }, + { + "epoch": 1.0, + "grad_norm": 1.9085113420205722, + "learning_rate": 1.1069475202818425e-09, + "loss": 1.2155, + "step": 42251 + }, + { + "epoch": 1.0, + "grad_norm": 2.1986699864775163, + "learning_rate": 1.0956235302161944e-09, + "loss": 1.0358, + "step": 42252 + }, + { + "epoch": 1.0, + "grad_norm": 1.920934824169834, + "learning_rate": 1.084357756937493e-09, + "loss": 1.051, + "step": 42253 + }, + { + "epoch": 1.0, + "grad_norm": 2.0467819102154587, + "learning_rate": 1.0731502005112415e-09, + "loss": 0.9767, + "step": 42254 + }, + { + "epoch": 1.0, + "grad_norm": 2.346806612840648, + "learning_rate": 1.062000861002943e-09, + "loss": 0.8949, + "step": 42255 + }, + { + "epoch": 1.0, + "grad_norm": 1.892374039379717, + "learning_rate": 1.0509097384769906e-09, + "loss": 1.0838, + "step": 42256 + }, + { + "epoch": 1.0, + "grad_norm": 2.0605542738722895, + "learning_rate": 1.039876832997777e-09, + "loss": 0.8507, + "step": 42257 + }, + { + "epoch": 1.0, + "grad_norm": 2.167190414080727, + "learning_rate": 1.0289021446308057e-09, + "loss": 1.1654, + "step": 42258 + }, + { + "epoch": 1.0, + "grad_norm": 1.7746334515681101, + "learning_rate": 1.017985673439359e-09, + "loss": 0.8672, + "step": 42259 + }, + { + "epoch": 1.0, + "grad_norm": 1.865726007335242, + "learning_rate": 1.0071274194856096e-09, + "loss": 1.0551, + "step": 42260 + }, + { + "epoch": 1.0, + "grad_norm": 2.0703774250844496, + "learning_rate": 9.963273828350606e-10, + "loss": 1.1203, + "step": 42261 + }, + { + "epoch": 1.0, + "grad_norm": 2.8217533112694184, + "learning_rate": 9.855855635498845e-10, + "loss": 0.9705, + "step": 42262 + }, + { + "epoch": 1.0, + "grad_norm": 1.870554307194666, + "learning_rate": 9.749019616911437e-10, + "loss": 1.0159, + "step": 42263 + }, + { + "epoch": 1.0, + "grad_norm": 1.7761534296956831, + "learning_rate": 9.642765773232309e-10, + "loss": 0.9635, + "step": 42264 + }, + { + "epoch": 1.0, + "grad_norm": 2.105529448327865, + "learning_rate": 9.537094105060985e-10, + "loss": 0.9809, + "step": 42265 + }, + { + "epoch": 1.0, + "grad_norm": 1.9955023799489808, + "learning_rate": 9.432004613019186e-10, + "loss": 1.0087, + "step": 42266 + }, + { + "epoch": 1.0, + "grad_norm": 1.8860167664475889, + "learning_rate": 9.327497297728638e-10, + "loss": 1.0051, + "step": 42267 + }, + { + "epoch": 1.0, + "grad_norm": 2.4132193624389586, + "learning_rate": 9.223572159788863e-10, + "loss": 1.0243, + "step": 42268 + }, + { + "epoch": 1.0, + "grad_norm": 1.8875968167238344, + "learning_rate": 9.12022919979938e-10, + "loss": 1.073, + "step": 42269 + }, + { + "epoch": 1.0, + "grad_norm": 1.064101298179088, + "learning_rate": 9.017468418370812e-10, + "loss": 0.9295, + "step": 42270 + }, + { + "epoch": 1.0, + "grad_norm": 2.2989825969704176, + "learning_rate": 8.915289816102679e-10, + "loss": 0.8599, + "step": 42271 + }, + { + "epoch": 1.0, + "grad_norm": 1.9324080966486596, + "learning_rate": 8.813693393583401e-10, + "loss": 0.9268, + "step": 42272 + }, + { + "epoch": 1.0, + "grad_norm": 1.0738608396937201, + "learning_rate": 8.712679151412495e-10, + "loss": 0.914, + "step": 42273 + }, + { + "epoch": 1.0, + "grad_norm": 1.9269197947416816, + "learning_rate": 8.612247090167281e-10, + "loss": 0.9274, + "step": 42274 + }, + { + "epoch": 1.0, + "grad_norm": 1.8357191690273567, + "learning_rate": 8.512397210447276e-10, + "loss": 1.0001, + "step": 42275 + }, + { + "epoch": 1.0, + "grad_norm": 1.9693752378817257, + "learning_rate": 8.413129512818697e-10, + "loss": 0.9184, + "step": 42276 + }, + { + "epoch": 1.0, + "grad_norm": 2.299742688215995, + "learning_rate": 8.314443997858857e-10, + "loss": 0.9832, + "step": 42277 + }, + { + "epoch": 1.0, + "grad_norm": 1.0416170493368901, + "learning_rate": 8.216340666156175e-10, + "loss": 0.9311, + "step": 42278 + }, + { + "epoch": 1.0, + "grad_norm": 2.11941011622621, + "learning_rate": 8.118819518276866e-10, + "loss": 0.9561, + "step": 42279 + }, + { + "epoch": 1.0, + "grad_norm": 2.1163900809605045, + "learning_rate": 8.021880554787143e-10, + "loss": 1.0086, + "step": 42280 + }, + { + "epoch": 1.0, + "grad_norm": 1.9659034150746346, + "learning_rate": 7.925523776242116e-10, + "loss": 1.0466, + "step": 42281 + }, + { + "epoch": 1.0, + "grad_norm": 2.0580155354101763, + "learning_rate": 7.829749183219104e-10, + "loss": 1.0606, + "step": 42282 + }, + { + "epoch": 1.0, + "grad_norm": 2.056901782047507, + "learning_rate": 7.734556776273217e-10, + "loss": 1.0449, + "step": 42283 + }, + { + "epoch": 1.0, + "grad_norm": 2.0247278194632603, + "learning_rate": 7.639946555937361e-10, + "loss": 1.0091, + "step": 42284 + }, + { + "epoch": 1.0, + "grad_norm": 1.8706526594749016, + "learning_rate": 7.545918522788853e-10, + "loss": 0.9849, + "step": 42285 + }, + { + "epoch": 1.0, + "grad_norm": 1.981969513864839, + "learning_rate": 7.452472677371702e-10, + "loss": 1.1143, + "step": 42286 + }, + { + "epoch": 1.0, + "grad_norm": 2.0881299710450065, + "learning_rate": 7.359609020218816e-10, + "loss": 1.0261, + "step": 42287 + }, + { + "epoch": 1.0, + "grad_norm": 2.5310438169288174, + "learning_rate": 7.267327551863101e-10, + "loss": 0.9117, + "step": 42288 + }, + { + "epoch": 1.0, + "grad_norm": 2.011481873445559, + "learning_rate": 7.175628272870772e-10, + "loss": 0.928, + "step": 42289 + }, + { + "epoch": 1.0, + "grad_norm": 1.9389439035090403, + "learning_rate": 7.08451118375253e-10, + "loss": 0.8597, + "step": 42290 + }, + { + "epoch": 1.0, + "grad_norm": 2.0575418221992905, + "learning_rate": 6.993976285041282e-10, + "loss": 0.913, + "step": 42291 + }, + { + "epoch": 1.0, + "grad_norm": 1.0618255776576118, + "learning_rate": 6.904023577269936e-10, + "loss": 1.0219, + "step": 42292 + }, + { + "epoch": 1.0, + "grad_norm": 2.05579318085058, + "learning_rate": 6.814653060971399e-10, + "loss": 1.0009, + "step": 42293 + }, + { + "epoch": 1.0, + "grad_norm": 1.9545999692768479, + "learning_rate": 6.725864736645271e-10, + "loss": 1.041, + "step": 42294 + }, + { + "epoch": 1.0, + "grad_norm": 1.0843537180386287, + "learning_rate": 6.637658604813357e-10, + "loss": 0.9097, + "step": 42295 + }, + { + "epoch": 1.0, + "grad_norm": 2.100647697423113, + "learning_rate": 6.550034666008564e-10, + "loss": 1.0226, + "step": 42296 + }, + { + "epoch": 1.0, + "grad_norm": 3.533219735513509, + "learning_rate": 6.462992920719391e-10, + "loss": 0.7875, + "step": 42297 + }, + { + "epoch": 1.0, + "grad_norm": 2.055261289889093, + "learning_rate": 6.376533369456539e-10, + "loss": 1.0209, + "step": 42298 + }, + { + "epoch": 1.0, + "grad_norm": 2.059540594353661, + "learning_rate": 6.290656012730711e-10, + "loss": 0.9107, + "step": 42299 + }, + { + "epoch": 1.0, + "grad_norm": 1.9020942951823918, + "learning_rate": 6.205360851041509e-10, + "loss": 1.0243, + "step": 42300 + }, + { + "epoch": 1.0, + "grad_norm": 2.1354950339729024, + "learning_rate": 6.120647884877429e-10, + "loss": 1.1069, + "step": 42301 + }, + { + "epoch": 1.0, + "grad_norm": 1.933347555584063, + "learning_rate": 6.036517114738072e-10, + "loss": 0.9828, + "step": 42302 + }, + { + "epoch": 1.0, + "grad_norm": 1.0721001906747938, + "learning_rate": 5.952968541111937e-10, + "loss": 0.8734, + "step": 42303 + }, + { + "epoch": 1.0, + "grad_norm": 1.8180924687318343, + "learning_rate": 5.87000216448752e-10, + "loss": 1.0685, + "step": 42304 + }, + { + "epoch": 1.0, + "grad_norm": 4.1351843269648265, + "learning_rate": 5.78761798534222e-10, + "loss": 0.9883, + "step": 42305 + }, + { + "epoch": 1.0, + "grad_norm": 2.0673076367528838, + "learning_rate": 5.70581600415343e-10, + "loss": 0.943, + "step": 42306 + }, + { + "epoch": 1.0, + "grad_norm": 2.1926196506711224, + "learning_rate": 5.62459622140965e-10, + "loss": 1.0859, + "step": 42307 + }, + { + "epoch": 1.0, + "grad_norm": 2.0078056746767605, + "learning_rate": 5.543958637577174e-10, + "loss": 0.8315, + "step": 42308 + }, + { + "epoch": 1.0, + "grad_norm": 2.258005138531248, + "learning_rate": 5.463903253122294e-10, + "loss": 1.0185, + "step": 42309 + }, + { + "epoch": 1.0, + "grad_norm": 2.14634441455492, + "learning_rate": 5.384430068522406e-10, + "loss": 0.9007, + "step": 42310 + }, + { + "epoch": 1.0, + "grad_norm": 2.3696364957403753, + "learning_rate": 5.305539084221601e-10, + "loss": 0.9884, + "step": 42311 + }, + { + "epoch": 1.0, + "grad_norm": 1.937666441104709, + "learning_rate": 5.227230300697273e-10, + "loss": 0.9357, + "step": 42312 + }, + { + "epoch": 1.0, + "grad_norm": 1.9106890256220894, + "learning_rate": 5.149503718393511e-10, + "loss": 0.9741, + "step": 42313 + }, + { + "epoch": 1.0, + "grad_norm": 1.7242606027447096, + "learning_rate": 5.072359337776611e-10, + "loss": 0.9449, + "step": 42314 + }, + { + "epoch": 1.0, + "grad_norm": 4.61958560811424, + "learning_rate": 4.995797159279558e-10, + "loss": 0.9702, + "step": 42315 + }, + { + "epoch": 1.0, + "grad_norm": 1.9018035623276293, + "learning_rate": 4.919817183346443e-10, + "loss": 0.8746, + "step": 42316 + }, + { + "epoch": 1.0, + "grad_norm": 2.1520384165848374, + "learning_rate": 4.844419410443557e-10, + "loss": 1.0095, + "step": 42317 + }, + { + "epoch": 1.0, + "grad_norm": 2.0709388715426265, + "learning_rate": 4.769603840981684e-10, + "loss": 1.0382, + "step": 42318 + }, + { + "epoch": 1.0, + "grad_norm": 2.2384648926467094, + "learning_rate": 4.695370475416016e-10, + "loss": 0.8859, + "step": 42319 + }, + { + "epoch": 1.0, + "grad_norm": 2.0705849589698646, + "learning_rate": 4.621719314168438e-10, + "loss": 0.9943, + "step": 42320 + }, + { + "epoch": 1.0, + "grad_norm": 1.0730817215357495, + "learning_rate": 4.5486503576719354e-10, + "loss": 0.881, + "step": 42321 + }, + { + "epoch": 1.0, + "grad_norm": 1.9574727778133654, + "learning_rate": 4.476163606348394e-10, + "loss": 0.9497, + "step": 42322 + }, + { + "epoch": 1.0, + "grad_norm": 1.0831761967742044, + "learning_rate": 4.4042590606308e-10, + "loss": 0.8868, + "step": 42323 + }, + { + "epoch": 1.0, + "grad_norm": 1.8575399284321867, + "learning_rate": 4.3329367209188344e-10, + "loss": 0.891, + "step": 42324 + }, + { + "epoch": 1.0, + "grad_norm": 1.8929723341619007, + "learning_rate": 4.262196587645484e-10, + "loss": 1.0366, + "step": 42325 + }, + { + "epoch": 1.0, + "grad_norm": 1.1547701675377826, + "learning_rate": 4.1920386612215313e-10, + "loss": 0.9117, + "step": 42326 + }, + { + "epoch": 1.0, + "grad_norm": 1.8450215669470404, + "learning_rate": 4.1224629420466567e-10, + "loss": 0.9427, + "step": 42327 + }, + { + "epoch": 1.0, + "grad_norm": 1.9655934194334745, + "learning_rate": 4.05346943052054e-10, + "loss": 0.9316, + "step": 42328 + }, + { + "epoch": 1.0, + "grad_norm": 2.095425187609138, + "learning_rate": 3.9850581270539645e-10, + "loss": 0.9743, + "step": 42329 + }, + { + "epoch": 1.0, + "grad_norm": 1.0644736320812542, + "learning_rate": 3.917229032057712e-10, + "loss": 0.9109, + "step": 42330 + }, + { + "epoch": 1.0, + "grad_norm": 1.9129127652989468, + "learning_rate": 3.849982145898157e-10, + "loss": 0.9317, + "step": 42331 + }, + { + "epoch": 1.0, + "grad_norm": 2.059152896377199, + "learning_rate": 3.7833174689860806e-10, + "loss": 0.8614, + "step": 42332 + }, + { + "epoch": 1.0, + "grad_norm": 2.957632929958362, + "learning_rate": 3.717235001710062e-10, + "loss": 1.0603, + "step": 42333 + }, + { + "epoch": 1.0, + "grad_norm": 1.9913886899686475, + "learning_rate": 3.651734744447577e-10, + "loss": 1.0461, + "step": 42334 + }, + { + "epoch": 1.0, + "grad_norm": 2.1120762612218265, + "learning_rate": 3.586816697587203e-10, + "loss": 0.9239, + "step": 42335 + }, + { + "epoch": 1.0, + "grad_norm": 1.8625766634814362, + "learning_rate": 3.522480861506417e-10, + "loss": 0.9976, + "step": 42336 + }, + { + "epoch": 1.0, + "grad_norm": 2.0778810788287716, + "learning_rate": 3.458727236571591e-10, + "loss": 0.9096, + "step": 42337 + }, + { + "epoch": 1.0, + "grad_norm": 1.9817629769138616, + "learning_rate": 3.3955558231490994e-10, + "loss": 0.8343, + "step": 42338 + }, + { + "epoch": 1.0, + "grad_norm": 1.9378542338324047, + "learning_rate": 3.3329666216275205e-10, + "loss": 0.9357, + "step": 42339 + }, + { + "epoch": 1.0, + "grad_norm": 2.15157074794248, + "learning_rate": 3.2709596323510227e-10, + "loss": 0.91, + "step": 42340 + }, + { + "epoch": 1.0, + "grad_norm": 2.133619916090911, + "learning_rate": 3.2095348556970827e-10, + "loss": 1.0069, + "step": 42341 + }, + { + "epoch": 1.0, + "grad_norm": 1.9468306544189788, + "learning_rate": 3.1486922920098696e-10, + "loss": 1.0445, + "step": 42342 + }, + { + "epoch": 1.0, + "grad_norm": 2.534307136390292, + "learning_rate": 3.088431941655756e-10, + "loss": 0.9643, + "step": 42343 + }, + { + "epoch": 1.0, + "grad_norm": 2.615297396693216, + "learning_rate": 3.028753804978912e-10, + "loss": 1.0391, + "step": 42344 + }, + { + "epoch": 1.0, + "grad_norm": 1.8743116669999147, + "learning_rate": 2.9696578823235067e-10, + "loss": 0.7606, + "step": 42345 + }, + { + "epoch": 1.0, + "grad_norm": 1.8224253083657596, + "learning_rate": 2.9111441740337085e-10, + "loss": 0.9953, + "step": 42346 + }, + { + "epoch": 1.0, + "grad_norm": 2.421018474575888, + "learning_rate": 2.8532126804647895e-10, + "loss": 0.9463, + "step": 42347 + }, + { + "epoch": 1.0, + "grad_norm": 1.0920664095468589, + "learning_rate": 2.795863401927612e-10, + "loss": 0.9243, + "step": 42348 + }, + { + "epoch": 1.0, + "grad_norm": 2.182940366466236, + "learning_rate": 2.739096338788549e-10, + "loss": 0.9679, + "step": 42349 + }, + { + "epoch": 1.0, + "grad_norm": 1.852369896982934, + "learning_rate": 2.682911491347362e-10, + "loss": 1.0535, + "step": 42350 + }, + { + "epoch": 1.0, + "grad_norm": 2.102863697210761, + "learning_rate": 2.627308859948219e-10, + "loss": 1.0059, + "step": 42351 + }, + { + "epoch": 1.0, + "grad_norm": 3.303156101412017, + "learning_rate": 2.5722884449130845e-10, + "loss": 1.1065, + "step": 42352 + }, + { + "epoch": 1.0, + "grad_norm": 2.200655383366607, + "learning_rate": 2.517850246563924e-10, + "loss": 0.8171, + "step": 42353 + }, + { + "epoch": 1.0, + "grad_norm": 2.119855836679587, + "learning_rate": 2.4639942652116e-10, + "loss": 1.0373, + "step": 42354 + }, + { + "epoch": 1.0, + "grad_norm": 2.113281587355652, + "learning_rate": 2.4107205011780764e-10, + "loss": 1.0728, + "step": 42355 + }, + { + "epoch": 1.0, + "grad_norm": 2.1693953516738627, + "learning_rate": 2.358028954763114e-10, + "loss": 1.0777, + "step": 42356 + }, + { + "epoch": 1.0, + "grad_norm": 8.613814618989684, + "learning_rate": 2.305919626277575e-10, + "loss": 0.9989, + "step": 42357 + }, + { + "epoch": 1.0, + "grad_norm": 1.0573870725765036, + "learning_rate": 2.25439251602122e-10, + "loss": 0.9103, + "step": 42358 + }, + { + "epoch": 1.0, + "grad_norm": 1.793769288979574, + "learning_rate": 2.203447624304911e-10, + "loss": 0.942, + "step": 42359 + }, + { + "epoch": 1.0, + "grad_norm": 2.021665438660883, + "learning_rate": 2.1530849514173057e-10, + "loss": 1.0236, + "step": 42360 + }, + { + "epoch": 1.0, + "grad_norm": 2.088404065752759, + "learning_rate": 2.103304497658165e-10, + "loss": 0.89, + "step": 42361 + }, + { + "epoch": 1.0, + "grad_norm": 1.732038555105124, + "learning_rate": 2.0541062633050446e-10, + "loss": 0.9412, + "step": 42362 + }, + { + "epoch": 1.0, + "grad_norm": 1.9312518645153336, + "learning_rate": 2.0054902486466022e-10, + "loss": 1.0059, + "step": 42363 + }, + { + "epoch": 1.0, + "grad_norm": 1.8832846264520895, + "learning_rate": 1.957456453982598e-10, + "loss": 0.9079, + "step": 42364 + }, + { + "epoch": 1.0, + "grad_norm": 2.0400650364923734, + "learning_rate": 1.9100048795794858e-10, + "loss": 0.8265, + "step": 42365 + }, + { + "epoch": 1.0, + "grad_norm": 1.9530964671673399, + "learning_rate": 1.8631355257148209e-10, + "loss": 0.7181, + "step": 42366 + }, + { + "epoch": 1.0, + "grad_norm": 1.9580678552473023, + "learning_rate": 1.8168483926550574e-10, + "loss": 0.8862, + "step": 42367 + }, + { + "epoch": 1.0, + "grad_norm": 2.7635773595412423, + "learning_rate": 1.7711434806888527e-10, + "loss": 0.9291, + "step": 42368 + }, + { + "epoch": 1.0, + "grad_norm": 1.956966697937893, + "learning_rate": 1.7260207900604565e-10, + "loss": 0.9675, + "step": 42369 + }, + { + "epoch": 1.0, + "grad_norm": 1.9634450066142508, + "learning_rate": 1.6814803210363217e-10, + "loss": 0.9204, + "step": 42370 + }, + { + "epoch": 1.0, + "grad_norm": 4.086496099682254, + "learning_rate": 1.6375220738940046e-10, + "loss": 0.9409, + "step": 42371 + }, + { + "epoch": 1.0, + "grad_norm": 1.9117093733308932, + "learning_rate": 1.5941460488777538e-10, + "loss": 1.0329, + "step": 42372 + }, + { + "epoch": 1.0, + "grad_norm": 1.9657295592688715, + "learning_rate": 1.5513522462318188e-10, + "loss": 0.9693, + "step": 42373 + }, + { + "epoch": 1.0, + "grad_norm": 1.9816876686123874, + "learning_rate": 1.5091406662115505e-10, + "loss": 0.9282, + "step": 42374 + }, + { + "epoch": 1.0, + "grad_norm": 2.9457335683374786, + "learning_rate": 1.4675113090723004e-10, + "loss": 1.0001, + "step": 42375 + }, + { + "epoch": 1.0, + "grad_norm": 2.124151577975652, + "learning_rate": 1.426464175036113e-10, + "loss": 0.9283, + "step": 42376 + }, + { + "epoch": 1.0, + "grad_norm": 2.120103389487751, + "learning_rate": 1.385999264369442e-10, + "loss": 0.8852, + "step": 42377 + }, + { + "epoch": 1.0, + "grad_norm": 1.9955217090344786, + "learning_rate": 1.3461165772832297e-10, + "loss": 1.0521, + "step": 42378 + }, + { + "epoch": 1.0, + "grad_norm": 2.403487646000348, + "learning_rate": 1.3068161140217251e-10, + "loss": 0.8668, + "step": 42379 + }, + { + "epoch": 1.0, + "grad_norm": 1.9019727078246624, + "learning_rate": 1.2680978748069728e-10, + "loss": 0.9258, + "step": 42380 + }, + { + "epoch": 1.0, + "grad_norm": 1.8991299791864282, + "learning_rate": 1.2299618598721197e-10, + "loss": 1.0785, + "step": 42381 + }, + { + "epoch": 1.0, + "grad_norm": 2.3484998016007412, + "learning_rate": 1.1924080694281083e-10, + "loss": 0.9964, + "step": 42382 + }, + { + "epoch": 1.0, + "grad_norm": 1.9805938386820643, + "learning_rate": 1.1554365037080851e-10, + "loss": 0.9501, + "step": 42383 + }, + { + "epoch": 1.0, + "grad_norm": 3.9051904937824533, + "learning_rate": 1.1190471629229927e-10, + "loss": 0.8271, + "step": 42384 + }, + { + "epoch": 1.0, + "grad_norm": 2.265611667532168, + "learning_rate": 1.0832400472726711e-10, + "loss": 0.9525, + "step": 42385 + }, + { + "epoch": 1.0, + "grad_norm": 1.8344234282767742, + "learning_rate": 1.0480151569791652e-10, + "loss": 0.9427, + "step": 42386 + }, + { + "epoch": 1.0, + "grad_norm": 2.354145135324348, + "learning_rate": 1.0133724922534171e-10, + "loss": 0.9253, + "step": 42387 + }, + { + "epoch": 1.0, + "grad_norm": 1.9831172842098408, + "learning_rate": 9.793120532730627e-11, + "loss": 0.9976, + "step": 42388 + }, + { + "epoch": 1.0, + "grad_norm": 1.069617059630371, + "learning_rate": 9.458338402601464e-11, + "loss": 0.9827, + "step": 42389 + }, + { + "epoch": 1.0, + "grad_norm": 2.1400060784617265, + "learning_rate": 9.129378533923038e-11, + "loss": 0.8742, + "step": 42390 + }, + { + "epoch": 1.0, + "grad_norm": 1.9948644228207155, + "learning_rate": 8.806240928693755e-11, + "loss": 0.9121, + "step": 42391 + }, + { + "epoch": 1.0, + "grad_norm": 1.8993176095003421, + "learning_rate": 8.488925588800989e-11, + "loss": 0.9364, + "step": 42392 + }, + { + "epoch": 1.0, + "grad_norm": 1.9632225009899897, + "learning_rate": 8.177432516132122e-11, + "loss": 0.941, + "step": 42393 + }, + { + "epoch": 1.0, + "grad_norm": 1.947489865208418, + "learning_rate": 7.871761712352489e-11, + "loss": 1.0103, + "step": 42394 + }, + { + "epoch": 1.0, + "grad_norm": 1.9174978877981903, + "learning_rate": 7.571913179349466e-11, + "loss": 1.0518, + "step": 42395 + }, + { + "epoch": 1.0, + "grad_norm": 2.1379907215166165, + "learning_rate": 7.277886918899413e-11, + "loss": 0.957, + "step": 42396 + }, + { + "epoch": 1.0, + "grad_norm": 2.675058666958636, + "learning_rate": 6.989682932667663e-11, + "loss": 0.9323, + "step": 42397 + }, + { + "epoch": 1.0, + "grad_norm": 3.272032043818667, + "learning_rate": 6.707301222319551e-11, + "loss": 0.9273, + "step": 42398 + }, + { + "epoch": 1.0, + "grad_norm": 2.1009506816037704, + "learning_rate": 6.430741789520412e-11, + "loss": 1.036, + "step": 42399 + }, + { + "epoch": 1.0, + "grad_norm": 1.1511379983280905, + "learning_rate": 6.160004635824556e-11, + "loss": 0.9299, + "step": 42400 + }, + { + "epoch": 1.0, + "grad_norm": 2.024740730783111, + "learning_rate": 5.895089762897322e-11, + "loss": 1.0084, + "step": 42401 + }, + { + "epoch": 1.0, + "grad_norm": 1.7059104983511444, + "learning_rate": 5.6359971722930176e-11, + "loss": 0.9065, + "step": 42402 + }, + { + "epoch": 1.0, + "grad_norm": 2.029993468740899, + "learning_rate": 5.3827268654549346e-11, + "loss": 0.8714, + "step": 42403 + }, + { + "epoch": 1.0, + "grad_norm": 1.831249388833603, + "learning_rate": 5.135278843826363e-11, + "loss": 0.9639, + "step": 42404 + }, + { + "epoch": 1.0, + "grad_norm": 1.1205682064750913, + "learning_rate": 4.8936531088505936e-11, + "loss": 0.9397, + "step": 42405 + }, + { + "epoch": 1.0, + "grad_norm": 2.0877503226612224, + "learning_rate": 4.657849662081937e-11, + "loss": 1.0099, + "step": 42406 + }, + { + "epoch": 1.0, + "grad_norm": 1.9435965375340152, + "learning_rate": 4.4278685047416395e-11, + "loss": 0.8614, + "step": 42407 + }, + { + "epoch": 1.0, + "grad_norm": 2.2161534127525297, + "learning_rate": 4.203709638161968e-11, + "loss": 1.0031, + "step": 42408 + }, + { + "epoch": 1.0, + "grad_norm": 2.257106112716332, + "learning_rate": 3.985373063786213e-11, + "loss": 0.9501, + "step": 42409 + }, + { + "epoch": 1.0, + "grad_norm": 2.7152597284491606, + "learning_rate": 3.772858782724598e-11, + "loss": 0.9151, + "step": 42410 + }, + { + "epoch": 1.0, + "grad_norm": 2.183777416814191, + "learning_rate": 3.566166796420412e-11, + "loss": 0.9647, + "step": 42411 + }, + { + "epoch": 1.0, + "grad_norm": 1.9190977441266421, + "learning_rate": 3.3652971057618336e-11, + "loss": 0.8506, + "step": 42412 + }, + { + "epoch": 1.0, + "grad_norm": 1.0456629411410001, + "learning_rate": 3.170249712303175e-11, + "loss": 0.9475, + "step": 42413 + }, + { + "epoch": 1.0, + "grad_norm": 1.7006685798846435, + "learning_rate": 2.9810246168215926e-11, + "loss": 0.8693, + "step": 42414 + }, + { + "epoch": 1.0, + "grad_norm": 1.9798904864164475, + "learning_rate": 2.7976218206493543e-11, + "loss": 0.9709, + "step": 42415 + }, + { + "epoch": 1.0, + "grad_norm": 2.0574466240169107, + "learning_rate": 2.6200413247856605e-11, + "loss": 1.1058, + "step": 42416 + }, + { + "epoch": 1.0, + "grad_norm": 2.2898918944369866, + "learning_rate": 2.4482831302297118e-11, + "loss": 1.0851, + "step": 42417 + }, + { + "epoch": 1.0, + "grad_norm": 2.770095427085466, + "learning_rate": 2.2823472379807086e-11, + "loss": 1.0596, + "step": 42418 + }, + { + "epoch": 1.0, + "grad_norm": 2.8065301349768164, + "learning_rate": 2.1222336491488748e-11, + "loss": 1.2, + "step": 42419 + }, + { + "epoch": 1.0, + "grad_norm": 2.0608668363095286, + "learning_rate": 1.967942364511366e-11, + "loss": 0.9684, + "step": 42420 + }, + { + "epoch": 1.0, + "grad_norm": 2.123222033981212, + "learning_rate": 1.8194733849563605e-11, + "loss": 0.8825, + "step": 42421 + }, + { + "epoch": 1.0, + "grad_norm": 2.0129722223569853, + "learning_rate": 1.6768267114830594e-11, + "loss": 0.991, + "step": 42422 + }, + { + "epoch": 1.0, + "grad_norm": 1.9683680475477898, + "learning_rate": 1.5400023447575962e-11, + "loss": 1.0167, + "step": 42423 + }, + { + "epoch": 1.0, + "grad_norm": 3.4282085387576697, + "learning_rate": 1.4090002856681495e-11, + "loss": 0.986, + "step": 42424 + }, + { + "epoch": 1.0, + "grad_norm": 1.9492178048062758, + "learning_rate": 1.2838205351028976e-11, + "loss": 0.9788, + "step": 42425 + }, + { + "epoch": 1.0, + "grad_norm": 2.0081517262541713, + "learning_rate": 1.1644630935059298e-11, + "loss": 1.013, + "step": 42426 + }, + { + "epoch": 1.0, + "grad_norm": 1.7447242452529632, + "learning_rate": 1.0509279617654245e-11, + "loss": 0.8853, + "step": 42427 + }, + { + "epoch": 1.0, + "grad_norm": 2.2051343083932093, + "learning_rate": 9.432151405475153e-12, + "loss": 1.0096, + "step": 42428 + }, + { + "epoch": 1.0, + "grad_norm": 1.941197591468606, + "learning_rate": 8.413246304073142e-12, + "loss": 0.9, + "step": 42429 + }, + { + "epoch": 1.0, + "grad_norm": 2.4974925978695754, + "learning_rate": 7.452564318999322e-12, + "loss": 0.9701, + "step": 42430 + }, + { + "epoch": 1.0, + "grad_norm": 3.600367296304084, + "learning_rate": 6.550105456915035e-12, + "loss": 0.9282, + "step": 42431 + }, + { + "epoch": 1.0, + "grad_norm": 1.9499232025810442, + "learning_rate": 5.705869723371394e-12, + "loss": 0.9846, + "step": 42432 + }, + { + "epoch": 1.0, + "grad_norm": 1.8544416169618128, + "learning_rate": 4.919857121699068e-12, + "loss": 0.9846, + "step": 42433 + }, + { + "epoch": 1.0, + "grad_norm": 2.026272477154211, + "learning_rate": 4.192067657449173e-12, + "loss": 0.9577, + "step": 42434 + }, + { + "epoch": 1.0, + "grad_norm": 1.8979007171545783, + "learning_rate": 3.5225013350626004e-12, + "loss": 0.9185, + "step": 42435 + }, + { + "epoch": 1.0, + "grad_norm": 2.0148902892128966, + "learning_rate": 2.9111581578700197e-12, + "loss": 1.0188, + "step": 42436 + }, + { + "epoch": 1.0, + "grad_norm": 1.8806580426354214, + "learning_rate": 2.3580381292021006e-12, + "loss": 0.9069, + "step": 42437 + }, + { + "epoch": 1.0, + "grad_norm": 2.8998489498612554, + "learning_rate": 1.863141253499734e-12, + "loss": 0.9637, + "step": 42438 + }, + { + "epoch": 1.0, + "grad_norm": 1.8292905538140407, + "learning_rate": 1.4264675329833665e-12, + "loss": 0.9669, + "step": 42439 + }, + { + "epoch": 1.0, + "grad_norm": 1.0311048298945837, + "learning_rate": 1.0480169687632214e-12, + "loss": 0.92, + "step": 42440 + }, + { + "epoch": 1.0, + "grad_norm": 1.6913246383288112, + "learning_rate": 7.277895663904133e-13, + "loss": 0.9933, + "step": 42441 + }, + { + "epoch": 1.0, + "grad_norm": 1.918603248660248, + "learning_rate": 4.657853236444964e-13, + "loss": 0.8655, + "step": 42442 + }, + { + "epoch": 1.0, + "grad_norm": 1.9966067630157776, + "learning_rate": 2.620042460765859e-13, + "loss": 1.0174, + "step": 42443 + }, + { + "epoch": 1.0, + "grad_norm": 2.015809488031823, + "learning_rate": 1.1644633146623564e-13, + "loss": 1.0092, + "step": 42444 + }, + { + "epoch": 1.0, + "grad_norm": 1.8873075612670138, + "learning_rate": 2.9111583144114664e-14, + "loss": 0.9244, + "step": 42445 + }, + { + "epoch": 1.0, + "grad_norm": 1.9485745821659621, + "learning_rate": 0.0, + "loss": 0.8182, + "step": 42446 + }, + { + "epoch": 1.0, + "step": 42446, + "total_flos": 5.528181595673985e+18, + "train_loss": 1.0080395902070418, + "train_runtime": 136087.7129, + "train_samples_per_second": 39.924, + "train_steps_per_second": 0.312 + } + ], + "logging_steps": 1.0, + "max_steps": 42446, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 10000, + "total_flos": 5.528181595673985e+18, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}