| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9985082048731975, | |
| "eval_steps": 500, | |
| "global_step": 1004, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001989060169070114, | |
| "grad_norm": 4.783787250518799, | |
| "learning_rate": 0.0, | |
| "loss": 1.1878, | |
| "num_tokens": 1379928.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003978120338140228, | |
| "grad_norm": 4.881955623626709, | |
| "learning_rate": 9.9009900990099e-09, | |
| "loss": 1.2018, | |
| "num_tokens": 2791003.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.005967180507210343, | |
| "grad_norm": 4.82396125793457, | |
| "learning_rate": 1.98019801980198e-08, | |
| "loss": 1.1913, | |
| "num_tokens": 4194463.0, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.007956240676280457, | |
| "grad_norm": 4.832489967346191, | |
| "learning_rate": 2.97029702970297e-08, | |
| "loss": 1.1974, | |
| "num_tokens": 5584143.0, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.009945300845350571, | |
| "grad_norm": 4.7777581214904785, | |
| "learning_rate": 3.96039603960396e-08, | |
| "loss": 1.1893, | |
| "num_tokens": 7031244.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.011934361014420686, | |
| "grad_norm": 4.757320880889893, | |
| "learning_rate": 4.950495049504951e-08, | |
| "loss": 1.1874, | |
| "num_tokens": 8436181.0, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0139234211834908, | |
| "grad_norm": 4.777390480041504, | |
| "learning_rate": 5.94059405940594e-08, | |
| "loss": 1.1855, | |
| "num_tokens": 9836769.0, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.015912481352560914, | |
| "grad_norm": 4.798787593841553, | |
| "learning_rate": 6.930693069306931e-08, | |
| "loss": 1.1791, | |
| "num_tokens": 11244836.0, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.01790154152163103, | |
| "grad_norm": 4.807232856750488, | |
| "learning_rate": 7.92079207920792e-08, | |
| "loss": 1.181, | |
| "num_tokens": 12623567.0, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.019890601690701143, | |
| "grad_norm": 4.721550464630127, | |
| "learning_rate": 8.91089108910891e-08, | |
| "loss": 1.1763, | |
| "num_tokens": 14028360.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02187966185977126, | |
| "grad_norm": 4.797355651855469, | |
| "learning_rate": 9.900990099009901e-08, | |
| "loss": 1.1935, | |
| "num_tokens": 15405448.0, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.023868722028841372, | |
| "grad_norm": 4.754054069519043, | |
| "learning_rate": 1.089108910891089e-07, | |
| "loss": 1.1914, | |
| "num_tokens": 16852417.0, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.025857782197911485, | |
| "grad_norm": 4.781761169433594, | |
| "learning_rate": 1.188118811881188e-07, | |
| "loss": 1.1883, | |
| "num_tokens": 18245229.0, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.0278468423669816, | |
| "grad_norm": 4.794190406799316, | |
| "learning_rate": 1.2871287128712872e-07, | |
| "loss": 1.1829, | |
| "num_tokens": 19630250.0, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.029835902536051714, | |
| "grad_norm": 4.7866363525390625, | |
| "learning_rate": 1.3861386138613863e-07, | |
| "loss": 1.2058, | |
| "num_tokens": 21090140.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.03182496270512183, | |
| "grad_norm": 4.720057964324951, | |
| "learning_rate": 1.485148514851485e-07, | |
| "loss": 1.1942, | |
| "num_tokens": 22459305.0, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.03381402287419195, | |
| "grad_norm": 4.714616775512695, | |
| "learning_rate": 1.584158415841584e-07, | |
| "loss": 1.1765, | |
| "num_tokens": 23864152.0, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.03580308304326206, | |
| "grad_norm": 4.747438907623291, | |
| "learning_rate": 1.6831683168316832e-07, | |
| "loss": 1.1952, | |
| "num_tokens": 25269279.0, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.03779214321233217, | |
| "grad_norm": 4.7715654373168945, | |
| "learning_rate": 1.782178217821782e-07, | |
| "loss": 1.1815, | |
| "num_tokens": 26685193.0, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.039781203381402286, | |
| "grad_norm": 4.7065911293029785, | |
| "learning_rate": 1.8811881188118812e-07, | |
| "loss": 1.1685, | |
| "num_tokens": 28076540.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0417702635504724, | |
| "grad_norm": 4.74373197555542, | |
| "learning_rate": 1.9801980198019803e-07, | |
| "loss": 1.1932, | |
| "num_tokens": 29489172.0, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.04375932371954252, | |
| "grad_norm": 4.572362899780273, | |
| "learning_rate": 2.079207920792079e-07, | |
| "loss": 1.1591, | |
| "num_tokens": 30889595.0, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.04574838388861263, | |
| "grad_norm": 4.560796737670898, | |
| "learning_rate": 2.178217821782178e-07, | |
| "loss": 1.1752, | |
| "num_tokens": 32298146.0, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.047737444057682744, | |
| "grad_norm": 4.545941352844238, | |
| "learning_rate": 2.2772277227722772e-07, | |
| "loss": 1.1707, | |
| "num_tokens": 33729949.0, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.04972650422675286, | |
| "grad_norm": 4.548399448394775, | |
| "learning_rate": 2.376237623762376e-07, | |
| "loss": 1.1759, | |
| "num_tokens": 35146745.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05171556439582297, | |
| "grad_norm": 4.558365345001221, | |
| "learning_rate": 2.475247524752475e-07, | |
| "loss": 1.1859, | |
| "num_tokens": 36534224.0, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.05370462456489309, | |
| "grad_norm": 4.54498291015625, | |
| "learning_rate": 2.5742574257425743e-07, | |
| "loss": 1.1644, | |
| "num_tokens": 37942016.0, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0556936847339632, | |
| "grad_norm": 4.540126323699951, | |
| "learning_rate": 2.673267326732673e-07, | |
| "loss": 1.1689, | |
| "num_tokens": 39362015.0, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.057682744903033316, | |
| "grad_norm": 4.505973815917969, | |
| "learning_rate": 2.7722772277227726e-07, | |
| "loss": 1.158, | |
| "num_tokens": 40746995.0, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.05967180507210343, | |
| "grad_norm": 4.145939350128174, | |
| "learning_rate": 2.871287128712871e-07, | |
| "loss": 1.147, | |
| "num_tokens": 42184610.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06166086524117355, | |
| "grad_norm": 4.096794128417969, | |
| "learning_rate": 2.97029702970297e-07, | |
| "loss": 1.1433, | |
| "num_tokens": 43594850.0, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.06364992541024365, | |
| "grad_norm": 4.0818586349487305, | |
| "learning_rate": 3.069306930693069e-07, | |
| "loss": 1.1351, | |
| "num_tokens": 45000725.0, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.06563898557931377, | |
| "grad_norm": 4.103175640106201, | |
| "learning_rate": 3.168316831683168e-07, | |
| "loss": 1.1303, | |
| "num_tokens": 46389383.0, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.0676280457483839, | |
| "grad_norm": 4.075052738189697, | |
| "learning_rate": 3.2673267326732674e-07, | |
| "loss": 1.1427, | |
| "num_tokens": 47780652.0, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.069617105917454, | |
| "grad_norm": 4.112435340881348, | |
| "learning_rate": 3.3663366336633663e-07, | |
| "loss": 1.14, | |
| "num_tokens": 49201017.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.07160616608652412, | |
| "grad_norm": 4.016624450683594, | |
| "learning_rate": 3.465346534653465e-07, | |
| "loss": 1.142, | |
| "num_tokens": 50577829.0, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.07359522625559423, | |
| "grad_norm": 3.9627022743225098, | |
| "learning_rate": 3.564356435643564e-07, | |
| "loss": 1.1189, | |
| "num_tokens": 52003696.0, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.07558428642466435, | |
| "grad_norm": 4.006690502166748, | |
| "learning_rate": 3.663366336633663e-07, | |
| "loss": 1.1307, | |
| "num_tokens": 53420296.0, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.07757334659373447, | |
| "grad_norm": 3.9357409477233887, | |
| "learning_rate": 3.7623762376237623e-07, | |
| "loss": 1.1279, | |
| "num_tokens": 54812494.0, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.07956240676280457, | |
| "grad_norm": 3.903628349304199, | |
| "learning_rate": 3.861386138613861e-07, | |
| "loss": 1.1042, | |
| "num_tokens": 56180118.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08155146693187469, | |
| "grad_norm": 3.4105091094970703, | |
| "learning_rate": 3.9603960396039606e-07, | |
| "loss": 1.0558, | |
| "num_tokens": 57631209.0, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.0835405271009448, | |
| "grad_norm": 3.0491228103637695, | |
| "learning_rate": 4.0594059405940595e-07, | |
| "loss": 1.0404, | |
| "num_tokens": 59050835.0, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.08552958727001492, | |
| "grad_norm": 2.9168484210968018, | |
| "learning_rate": 4.158415841584158e-07, | |
| "loss": 1.033, | |
| "num_tokens": 60439622.0, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.08751864743908504, | |
| "grad_norm": 2.8887298107147217, | |
| "learning_rate": 4.257425742574257e-07, | |
| "loss": 1.026, | |
| "num_tokens": 61853701.0, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.08950770760815514, | |
| "grad_norm": 2.882795810699463, | |
| "learning_rate": 4.356435643564356e-07, | |
| "loss": 1.0189, | |
| "num_tokens": 63243921.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.09149676777722526, | |
| "grad_norm": 2.826184034347534, | |
| "learning_rate": 4.4554455445544555e-07, | |
| "loss": 1.0267, | |
| "num_tokens": 64692234.0, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.09348582794629537, | |
| "grad_norm": 2.837517261505127, | |
| "learning_rate": 4.5544554455445543e-07, | |
| "loss": 1.0188, | |
| "num_tokens": 66095339.0, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.09547488811536549, | |
| "grad_norm": 2.7744252681732178, | |
| "learning_rate": 4.6534653465346537e-07, | |
| "loss": 1.021, | |
| "num_tokens": 67470384.0, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.09746394828443561, | |
| "grad_norm": 2.7776126861572266, | |
| "learning_rate": 4.752475247524752e-07, | |
| "loss": 1.0072, | |
| "num_tokens": 68907983.0, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.09945300845350571, | |
| "grad_norm": 2.696685552597046, | |
| "learning_rate": 4.851485148514851e-07, | |
| "loss": 1.0018, | |
| "num_tokens": 70284975.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.10144206862257583, | |
| "grad_norm": 2.6469790935516357, | |
| "learning_rate": 4.95049504950495e-07, | |
| "loss": 0.999, | |
| "num_tokens": 71668994.0, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.10343112879164594, | |
| "grad_norm": 2.5766100883483887, | |
| "learning_rate": 5.04950495049505e-07, | |
| "loss": 0.9941, | |
| "num_tokens": 73087542.0, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.10542018896071606, | |
| "grad_norm": 2.5015177726745605, | |
| "learning_rate": 5.148514851485149e-07, | |
| "loss": 0.9808, | |
| "num_tokens": 74502903.0, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.10740924912978618, | |
| "grad_norm": 2.4191646575927734, | |
| "learning_rate": 5.247524752475247e-07, | |
| "loss": 0.9769, | |
| "num_tokens": 75902927.0, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.10939830929885629, | |
| "grad_norm": 2.2691502571105957, | |
| "learning_rate": 5.346534653465346e-07, | |
| "loss": 0.9584, | |
| "num_tokens": 77315581.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1113873694679264, | |
| "grad_norm": 2.0384151935577393, | |
| "learning_rate": 5.445544554455445e-07, | |
| "loss": 0.9216, | |
| "num_tokens": 78698236.0, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.11337642963699653, | |
| "grad_norm": 1.7994420528411865, | |
| "learning_rate": 5.544554455445545e-07, | |
| "loss": 0.8971, | |
| "num_tokens": 80129837.0, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.11536548980606663, | |
| "grad_norm": 1.6083354949951172, | |
| "learning_rate": 5.643564356435643e-07, | |
| "loss": 0.887, | |
| "num_tokens": 81529934.0, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.11735454997513675, | |
| "grad_norm": 1.4472858905792236, | |
| "learning_rate": 5.742574257425742e-07, | |
| "loss": 0.8759, | |
| "num_tokens": 82926044.0, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.11934361014420686, | |
| "grad_norm": 1.3287198543548584, | |
| "learning_rate": 5.841584158415841e-07, | |
| "loss": 0.8572, | |
| "num_tokens": 84300935.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12133267031327698, | |
| "grad_norm": 1.2509266138076782, | |
| "learning_rate": 5.94059405940594e-07, | |
| "loss": 0.8497, | |
| "num_tokens": 85709827.0, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1233217304823471, | |
| "grad_norm": 1.2169718742370605, | |
| "learning_rate": 6.03960396039604e-07, | |
| "loss": 0.8503, | |
| "num_tokens": 87094581.0, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1253107906514172, | |
| "grad_norm": 1.1747347116470337, | |
| "learning_rate": 6.138613861386138e-07, | |
| "loss": 0.8499, | |
| "num_tokens": 88486857.0, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1272998508204873, | |
| "grad_norm": 1.108124852180481, | |
| "learning_rate": 6.237623762376237e-07, | |
| "loss": 0.8385, | |
| "num_tokens": 89892391.0, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.12928891098955744, | |
| "grad_norm": 1.080589771270752, | |
| "learning_rate": 6.336633663366336e-07, | |
| "loss": 0.8415, | |
| "num_tokens": 91290866.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.13127797115862755, | |
| "grad_norm": 1.0345003604888916, | |
| "learning_rate": 6.435643564356436e-07, | |
| "loss": 0.8287, | |
| "num_tokens": 92700954.0, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.13326703132769765, | |
| "grad_norm": 0.9741297364234924, | |
| "learning_rate": 6.534653465346535e-07, | |
| "loss": 0.8193, | |
| "num_tokens": 94131765.0, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1352560914967678, | |
| "grad_norm": 0.9597648978233337, | |
| "learning_rate": 6.633663366336634e-07, | |
| "loss": 0.8245, | |
| "num_tokens": 95516370.0, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.1372451516658379, | |
| "grad_norm": 0.9024690985679626, | |
| "learning_rate": 6.732673267326733e-07, | |
| "loss": 0.8203, | |
| "num_tokens": 96921506.0, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.139234211834908, | |
| "grad_norm": 0.8436392545700073, | |
| "learning_rate": 6.831683168316831e-07, | |
| "loss": 0.8005, | |
| "num_tokens": 98292636.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.14122327200397813, | |
| "grad_norm": 0.7839356660842896, | |
| "learning_rate": 6.93069306930693e-07, | |
| "loss": 0.832, | |
| "num_tokens": 99743934.0, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.14321233217304824, | |
| "grad_norm": 0.7109583616256714, | |
| "learning_rate": 7.029702970297029e-07, | |
| "loss": 0.794, | |
| "num_tokens": 101144081.0, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.14520139234211835, | |
| "grad_norm": 0.6535574793815613, | |
| "learning_rate": 7.128712871287128e-07, | |
| "loss": 0.8081, | |
| "num_tokens": 102546821.0, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.14719045251118845, | |
| "grad_norm": 0.5921033620834351, | |
| "learning_rate": 7.227722772277227e-07, | |
| "loss": 0.7886, | |
| "num_tokens": 103957723.0, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.14917951268025859, | |
| "grad_norm": 0.5575245022773743, | |
| "learning_rate": 7.326732673267326e-07, | |
| "loss": 0.783, | |
| "num_tokens": 105353702.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1511685728493287, | |
| "grad_norm": 0.5279855132102966, | |
| "learning_rate": 7.425742574257426e-07, | |
| "loss": 0.7729, | |
| "num_tokens": 106790931.0, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.1531576330183988, | |
| "grad_norm": 0.5107793211936951, | |
| "learning_rate": 7.524752475247525e-07, | |
| "loss": 0.7583, | |
| "num_tokens": 108181851.0, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.15514669318746893, | |
| "grad_norm": 0.5129069089889526, | |
| "learning_rate": 7.623762376237624e-07, | |
| "loss": 0.7624, | |
| "num_tokens": 109558589.0, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.15713575335653904, | |
| "grad_norm": 0.479915589094162, | |
| "learning_rate": 7.722772277227722e-07, | |
| "loss": 0.7378, | |
| "num_tokens": 110915629.0, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.15912481352560914, | |
| "grad_norm": 0.4730769991874695, | |
| "learning_rate": 7.821782178217821e-07, | |
| "loss": 0.7619, | |
| "num_tokens": 112326351.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.16111387369467928, | |
| "grad_norm": 0.4519067406654358, | |
| "learning_rate": 7.920792079207921e-07, | |
| "loss": 0.7459, | |
| "num_tokens": 113740610.0, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.16310293386374938, | |
| "grad_norm": 0.4359944760799408, | |
| "learning_rate": 8.01980198019802e-07, | |
| "loss": 0.749, | |
| "num_tokens": 115148611.0, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.1650919940328195, | |
| "grad_norm": 0.4051918685436249, | |
| "learning_rate": 8.118811881188119e-07, | |
| "loss": 0.7447, | |
| "num_tokens": 116569175.0, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.1670810542018896, | |
| "grad_norm": 0.38103756308555603, | |
| "learning_rate": 8.217821782178217e-07, | |
| "loss": 0.7469, | |
| "num_tokens": 117975546.0, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.16907011437095973, | |
| "grad_norm": 0.360709011554718, | |
| "learning_rate": 8.316831683168316e-07, | |
| "loss": 0.722, | |
| "num_tokens": 119391367.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.17105917454002983, | |
| "grad_norm": 0.3431180417537689, | |
| "learning_rate": 8.415841584158416e-07, | |
| "loss": 0.7323, | |
| "num_tokens": 120769077.0, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.17304823470909994, | |
| "grad_norm": 0.32238441705703735, | |
| "learning_rate": 8.514851485148514e-07, | |
| "loss": 0.7254, | |
| "num_tokens": 122218630.0, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.17503729487817007, | |
| "grad_norm": 0.3144312798976898, | |
| "learning_rate": 8.613861386138613e-07, | |
| "loss": 0.741, | |
| "num_tokens": 123638408.0, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.17702635504724018, | |
| "grad_norm": 0.2969042658805847, | |
| "learning_rate": 8.712871287128712e-07, | |
| "loss": 0.707, | |
| "num_tokens": 125024280.0, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.1790154152163103, | |
| "grad_norm": 0.28798267245292664, | |
| "learning_rate": 8.811881188118812e-07, | |
| "loss": 0.7082, | |
| "num_tokens": 126462816.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.18100447538538042, | |
| "grad_norm": 0.2822662591934204, | |
| "learning_rate": 8.910891089108911e-07, | |
| "loss": 0.7139, | |
| "num_tokens": 127856241.0, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.18299353555445053, | |
| "grad_norm": 0.275879830121994, | |
| "learning_rate": 9.00990099009901e-07, | |
| "loss": 0.7113, | |
| "num_tokens": 129247744.0, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.18498259572352063, | |
| "grad_norm": 0.26023879647254944, | |
| "learning_rate": 9.108910891089109e-07, | |
| "loss": 0.7, | |
| "num_tokens": 130653045.0, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.18697165589259074, | |
| "grad_norm": 0.2475547045469284, | |
| "learning_rate": 9.207920792079208e-07, | |
| "loss": 0.7075, | |
| "num_tokens": 132075768.0, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.18896071606166087, | |
| "grad_norm": 0.23815281689167023, | |
| "learning_rate": 9.306930693069307e-07, | |
| "loss": 0.7034, | |
| "num_tokens": 133471787.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.19094977623073098, | |
| "grad_norm": 0.2257402390241623, | |
| "learning_rate": 9.405940594059405e-07, | |
| "loss": 0.7005, | |
| "num_tokens": 134884172.0, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.19293883639980108, | |
| "grad_norm": 0.21398675441741943, | |
| "learning_rate": 9.504950495049504e-07, | |
| "loss": 0.6954, | |
| "num_tokens": 136321122.0, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.19492789656887122, | |
| "grad_norm": 0.21331369876861572, | |
| "learning_rate": 9.603960396039604e-07, | |
| "loss": 0.6931, | |
| "num_tokens": 137714653.0, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.19691695673794132, | |
| "grad_norm": 0.21159325540065765, | |
| "learning_rate": 9.702970297029702e-07, | |
| "loss": 0.7108, | |
| "num_tokens": 139151591.0, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.19890601690701143, | |
| "grad_norm": 0.20192930102348328, | |
| "learning_rate": 9.801980198019802e-07, | |
| "loss": 0.6838, | |
| "num_tokens": 140552794.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.20089507707608156, | |
| "grad_norm": 0.20033682882785797, | |
| "learning_rate": 9.9009900990099e-07, | |
| "loss": 0.6839, | |
| "num_tokens": 141962296.0, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.20288413724515167, | |
| "grad_norm": 0.1956896036863327, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6853, | |
| "num_tokens": 143391533.0, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.20487319741422177, | |
| "grad_norm": 0.1973898708820343, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6867, | |
| "num_tokens": 144782734.0, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.20686225758329188, | |
| "grad_norm": 0.17975734174251556, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6773, | |
| "num_tokens": 146195534.0, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.20885131775236201, | |
| "grad_norm": 0.17711520195007324, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6582, | |
| "num_tokens": 147581967.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.21084037792143212, | |
| "grad_norm": 0.1741390824317932, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6852, | |
| "num_tokens": 148962092.0, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.21282943809050223, | |
| "grad_norm": 0.1642421931028366, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6722, | |
| "num_tokens": 150367148.0, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.21481849825957236, | |
| "grad_norm": 0.1667676866054535, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6628, | |
| "num_tokens": 151756604.0, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.21680755842864247, | |
| "grad_norm": 0.1586826741695404, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6622, | |
| "num_tokens": 153175572.0, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.21879661859771257, | |
| "grad_norm": 0.15848655998706818, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6775, | |
| "num_tokens": 154616863.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2207856787667827, | |
| "grad_norm": 0.15296334028244019, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6544, | |
| "num_tokens": 156003361.0, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2227747389358528, | |
| "grad_norm": 0.1545649766921997, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6646, | |
| "num_tokens": 157395330.0, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.22476379910492292, | |
| "grad_norm": 0.15351980924606323, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6646, | |
| "num_tokens": 158812399.0, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.22675285927399305, | |
| "grad_norm": 0.14907206594944, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6508, | |
| "num_tokens": 160216957.0, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.22874191944306316, | |
| "grad_norm": 0.14644992351531982, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6644, | |
| "num_tokens": 161681970.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.23073097961213326, | |
| "grad_norm": 0.15030954778194427, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6432, | |
| "num_tokens": 163031303.0, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.23272003978120337, | |
| "grad_norm": 0.1434543877840042, | |
| "learning_rate": 1e-06, | |
| "loss": 0.647, | |
| "num_tokens": 164438690.0, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.2347090999502735, | |
| "grad_norm": 0.14714758098125458, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6547, | |
| "num_tokens": 165848650.0, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.2366981601193436, | |
| "grad_norm": 0.14545480906963348, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6666, | |
| "num_tokens": 167315232.0, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.23868722028841372, | |
| "grad_norm": 0.14221689105033875, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6481, | |
| "num_tokens": 168709749.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.24067628045748385, | |
| "grad_norm": 0.14459247887134552, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6361, | |
| "num_tokens": 170093537.0, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.24266534062655395, | |
| "grad_norm": 0.14327335357666016, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6513, | |
| "num_tokens": 171464842.0, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.24465440079562406, | |
| "grad_norm": 0.13770653307437897, | |
| "learning_rate": 1e-06, | |
| "loss": 0.641, | |
| "num_tokens": 172860600.0, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.2466434609646942, | |
| "grad_norm": 0.1363484412431717, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6445, | |
| "num_tokens": 174305923.0, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.2486325211337643, | |
| "grad_norm": 0.1345747411251068, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6374, | |
| "num_tokens": 175699135.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2506215813028344, | |
| "grad_norm": 0.13693904876708984, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6491, | |
| "num_tokens": 177132442.0, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.25261064147190454, | |
| "grad_norm": 0.1340012401342392, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6349, | |
| "num_tokens": 178530726.0, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.2545997016409746, | |
| "grad_norm": 0.13670295476913452, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6439, | |
| "num_tokens": 179929069.0, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.25658876181004475, | |
| "grad_norm": 0.13612930476665497, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6382, | |
| "num_tokens": 181362595.0, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.2585778219791149, | |
| "grad_norm": 0.13262143731117249, | |
| "learning_rate": 1e-06, | |
| "loss": 0.636, | |
| "num_tokens": 182758225.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.26056688214818496, | |
| "grad_norm": 0.13065095245838165, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6339, | |
| "num_tokens": 184186457.0, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.2625559423172551, | |
| "grad_norm": 0.13491177558898926, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6424, | |
| "num_tokens": 185597538.0, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.26454500248632523, | |
| "grad_norm": 0.13115911185741425, | |
| "learning_rate": 1e-06, | |
| "loss": 0.636, | |
| "num_tokens": 186999644.0, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.2665340626553953, | |
| "grad_norm": 0.13199158012866974, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6404, | |
| "num_tokens": 188383971.0, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.26852312282446544, | |
| "grad_norm": 0.1268879771232605, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6458, | |
| "num_tokens": 189800770.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2705121829935356, | |
| "grad_norm": 0.12490399926900864, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6437, | |
| "num_tokens": 191189967.0, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.27250124316260566, | |
| "grad_norm": 0.1264086365699768, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6015, | |
| "num_tokens": 192578506.0, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.2744903033316758, | |
| "grad_norm": 0.1346651315689087, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6271, | |
| "num_tokens": 193939223.0, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.2764793635007459, | |
| "grad_norm": 0.12618686258792877, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6249, | |
| "num_tokens": 195333601.0, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.278468423669816, | |
| "grad_norm": 0.12344113737344742, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6314, | |
| "num_tokens": 196788664.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.28045748383888613, | |
| "grad_norm": 0.12506203353405, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6282, | |
| "num_tokens": 198163401.0, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.28244654400795627, | |
| "grad_norm": 0.12806549668312073, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6304, | |
| "num_tokens": 199567499.0, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.28443560417702635, | |
| "grad_norm": 0.12447736412286758, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6145, | |
| "num_tokens": 200961717.0, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.2864246643460965, | |
| "grad_norm": 0.123995341360569, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6296, | |
| "num_tokens": 202374356.0, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.28841372451516656, | |
| "grad_norm": 0.12191283702850342, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6255, | |
| "num_tokens": 203763067.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.2904027846842367, | |
| "grad_norm": 0.12908770143985748, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6134, | |
| "num_tokens": 205148329.0, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.2923918448533068, | |
| "grad_norm": 0.12782661616802216, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6366, | |
| "num_tokens": 206571749.0, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.2943809050223769, | |
| "grad_norm": 0.12458013743162155, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6274, | |
| "num_tokens": 207945787.0, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.29636996519144704, | |
| "grad_norm": 0.1228199154138565, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6126, | |
| "num_tokens": 209368496.0, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.29835902536051717, | |
| "grad_norm": 0.11965636909008026, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6222, | |
| "num_tokens": 210815947.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.30034808552958725, | |
| "grad_norm": 0.1323099434375763, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6242, | |
| "num_tokens": 212216246.0, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3023371456986574, | |
| "grad_norm": 0.12547767162322998, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6129, | |
| "num_tokens": 213565534.0, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.3043262058677275, | |
| "grad_norm": 0.12437719106674194, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6268, | |
| "num_tokens": 215022812.0, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.3063152660367976, | |
| "grad_norm": 0.12172248214483261, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6158, | |
| "num_tokens": 216418038.0, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.30830432620586773, | |
| "grad_norm": 0.12287899106740952, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6165, | |
| "num_tokens": 217835190.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.31029338637493786, | |
| "grad_norm": 0.12283938378095627, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6088, | |
| "num_tokens": 219203881.0, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.31228244654400794, | |
| "grad_norm": 0.12894928455352783, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6245, | |
| "num_tokens": 220596713.0, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.3142715067130781, | |
| "grad_norm": 0.12768110632896423, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6148, | |
| "num_tokens": 221990986.0, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.3162605668821482, | |
| "grad_norm": 0.11902462691068649, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6134, | |
| "num_tokens": 223430785.0, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.3182496270512183, | |
| "grad_norm": 0.12414086610078812, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6236, | |
| "num_tokens": 224823735.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3202386872202884, | |
| "grad_norm": 0.12295661866664886, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6222, | |
| "num_tokens": 226213815.0, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.32222774738935855, | |
| "grad_norm": 0.12154286354780197, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6154, | |
| "num_tokens": 227589479.0, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.32421680755842863, | |
| "grad_norm": 0.12159918993711472, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6112, | |
| "num_tokens": 228978021.0, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.32620586772749877, | |
| "grad_norm": 0.1207566186785698, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6101, | |
| "num_tokens": 230373569.0, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.32819492789656884, | |
| "grad_norm": 0.1222677007317543, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6089, | |
| "num_tokens": 231771815.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.330183988065639, | |
| "grad_norm": 0.1245197132229805, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6095, | |
| "num_tokens": 233172925.0, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.3321730482347091, | |
| "grad_norm": 0.1191013976931572, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6138, | |
| "num_tokens": 234578536.0, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.3341621084037792, | |
| "grad_norm": 0.12282473593950272, | |
| "learning_rate": 1e-06, | |
| "loss": 0.605, | |
| "num_tokens": 235997856.0, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.3361511685728493, | |
| "grad_norm": 0.11851899325847626, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6224, | |
| "num_tokens": 237456733.0, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.33814022874191946, | |
| "grad_norm": 0.12077327817678452, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6026, | |
| "num_tokens": 238849801.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.34012928891098954, | |
| "grad_norm": 0.11882904917001724, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6126, | |
| "num_tokens": 240265062.0, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.34211834908005967, | |
| "grad_norm": 0.12201754748821259, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5968, | |
| "num_tokens": 241660845.0, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.3441074092491298, | |
| "grad_norm": 0.11851049214601517, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6045, | |
| "num_tokens": 243073416.0, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.3460964694181999, | |
| "grad_norm": 0.1223016306757927, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6081, | |
| "num_tokens": 244457775.0, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.34808552958727, | |
| "grad_norm": 0.11766277253627777, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6115, | |
| "num_tokens": 245880299.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.35007458975634015, | |
| "grad_norm": 0.12223277240991592, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6147, | |
| "num_tokens": 247285182.0, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.3520636499254102, | |
| "grad_norm": 0.12305691093206406, | |
| "learning_rate": 1e-06, | |
| "loss": 0.621, | |
| "num_tokens": 248698180.0, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.35405271009448036, | |
| "grad_norm": 0.12199797481298447, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6028, | |
| "num_tokens": 250069197.0, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.3560417702635505, | |
| "grad_norm": 0.12232685089111328, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6219, | |
| "num_tokens": 251503045.0, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.3580308304326206, | |
| "grad_norm": 0.12115549296140671, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6139, | |
| "num_tokens": 252942250.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3600198906016907, | |
| "grad_norm": 0.12164245545864105, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6158, | |
| "num_tokens": 254386360.0, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.36200895077076084, | |
| "grad_norm": 0.12038590759038925, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6171, | |
| "num_tokens": 255791663.0, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.3639980109398309, | |
| "grad_norm": 0.12380865961313248, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6146, | |
| "num_tokens": 257180435.0, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.36598707110890105, | |
| "grad_norm": 0.12001994252204895, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5962, | |
| "num_tokens": 258575017.0, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.3679761312779712, | |
| "grad_norm": 0.12672315537929535, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5982, | |
| "num_tokens": 259971664.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.36996519144704126, | |
| "grad_norm": 0.12033303827047348, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5976, | |
| "num_tokens": 261365988.0, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.3719542516161114, | |
| "grad_norm": 0.12143931537866592, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5963, | |
| "num_tokens": 262783906.0, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.3739433117851815, | |
| "grad_norm": 0.12255272269248962, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6123, | |
| "num_tokens": 264206687.0, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.3759323719542516, | |
| "grad_norm": 0.1221979632973671, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5967, | |
| "num_tokens": 265600406.0, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.37792143212332174, | |
| "grad_norm": 0.11851304769515991, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5885, | |
| "num_tokens": 266998184.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3799104922923918, | |
| "grad_norm": 0.11754640191793442, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6109, | |
| "num_tokens": 268432018.0, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.38189955246146196, | |
| "grad_norm": 0.12079335004091263, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6018, | |
| "num_tokens": 269811989.0, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.3838886126305321, | |
| "grad_norm": 0.12268039584159851, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6076, | |
| "num_tokens": 271216935.0, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.38587767279960217, | |
| "grad_norm": 0.12338917702436447, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6127, | |
| "num_tokens": 272636458.0, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.3878667329686723, | |
| "grad_norm": 0.11832890659570694, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6117, | |
| "num_tokens": 274045339.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.38985579313774243, | |
| "grad_norm": 0.11602655798196793, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6099, | |
| "num_tokens": 275455795.0, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.3918448533068125, | |
| "grad_norm": 0.11740183085203171, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6065, | |
| "num_tokens": 276879525.0, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.39383391347588265, | |
| "grad_norm": 0.12076492607593536, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6082, | |
| "num_tokens": 278290807.0, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.3958229736449528, | |
| "grad_norm": 0.12188921868801117, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5984, | |
| "num_tokens": 279716188.0, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.39781203381402286, | |
| "grad_norm": 0.12162777781486511, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6029, | |
| "num_tokens": 281099544.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.399801093983093, | |
| "grad_norm": 0.11848218739032745, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5823, | |
| "num_tokens": 282469441.0, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.4017901541521631, | |
| "grad_norm": 0.1235736832022667, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5956, | |
| "num_tokens": 283884792.0, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.4037792143212332, | |
| "grad_norm": 0.1228078156709671, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6079, | |
| "num_tokens": 285286535.0, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.40576827449030334, | |
| "grad_norm": 0.12109891325235367, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6001, | |
| "num_tokens": 286679118.0, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.40775733465937347, | |
| "grad_norm": 0.12217779457569122, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6023, | |
| "num_tokens": 288108758.0, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.40974639482844355, | |
| "grad_norm": 0.12233875691890717, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5983, | |
| "num_tokens": 289524538.0, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.4117354549975137, | |
| "grad_norm": 0.12142330408096313, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6072, | |
| "num_tokens": 290911340.0, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.41372451516658376, | |
| "grad_norm": 0.12050288170576096, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5899, | |
| "num_tokens": 292301970.0, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.4157135753356539, | |
| "grad_norm": 0.12356843799352646, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5958, | |
| "num_tokens": 293718750.0, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.41770263550472403, | |
| "grad_norm": 0.12285276502370834, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5915, | |
| "num_tokens": 295137029.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4196916956737941, | |
| "grad_norm": 0.12178198248147964, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6015, | |
| "num_tokens": 296516937.0, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.42168075584286424, | |
| "grad_norm": 0.12042722851037979, | |
| "learning_rate": 1e-06, | |
| "loss": 0.592, | |
| "num_tokens": 297914166.0, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.4236698160119344, | |
| "grad_norm": 0.12015018612146378, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5917, | |
| "num_tokens": 299291477.0, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.42565887618100445, | |
| "grad_norm": 0.11729878187179565, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5902, | |
| "num_tokens": 300690831.0, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.4276479363500746, | |
| "grad_norm": 0.11877918988466263, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5837, | |
| "num_tokens": 302125996.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4296369965191447, | |
| "grad_norm": 0.12102558463811874, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6027, | |
| "num_tokens": 303541215.0, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.4316260566882148, | |
| "grad_norm": 0.11814971268177032, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5912, | |
| "num_tokens": 304932140.0, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.43361511685728493, | |
| "grad_norm": 0.11896595358848572, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6004, | |
| "num_tokens": 306387713.0, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.43560417702635507, | |
| "grad_norm": 0.12220481783151627, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6155, | |
| "num_tokens": 307820317.0, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.43759323719542514, | |
| "grad_norm": 0.11805558949708939, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5929, | |
| "num_tokens": 309219520.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4395822973644953, | |
| "grad_norm": 0.12132007628679276, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5837, | |
| "num_tokens": 310591459.0, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.4415713575335654, | |
| "grad_norm": 0.11762286722660065, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5861, | |
| "num_tokens": 312008252.0, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.4435604177026355, | |
| "grad_norm": 0.11986927688121796, | |
| "learning_rate": 1e-06, | |
| "loss": 0.594, | |
| "num_tokens": 313425810.0, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.4455494778717056, | |
| "grad_norm": 0.11796683073043823, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5866, | |
| "num_tokens": 314853879.0, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.44753853804077576, | |
| "grad_norm": 0.12010174244642258, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6066, | |
| "num_tokens": 316294854.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.44952759820984584, | |
| "grad_norm": 0.11594483256340027, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5822, | |
| "num_tokens": 317700098.0, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.45151665837891597, | |
| "grad_norm": 0.11875201016664505, | |
| "learning_rate": 1e-06, | |
| "loss": 0.6061, | |
| "num_tokens": 319128053.0, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.4535057185479861, | |
| "grad_norm": 0.1269533336162567, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5777, | |
| "num_tokens": 320515582.0, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.4554947787170562, | |
| "grad_norm": 0.11610081046819687, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5855, | |
| "num_tokens": 321928550.0, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.4574838388861263, | |
| "grad_norm": 0.12167216837406158, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5834, | |
| "num_tokens": 323331818.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4594728990551964, | |
| "grad_norm": 0.1215236485004425, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5922, | |
| "num_tokens": 324737162.0, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.4614619592242665, | |
| "grad_norm": 0.12563477456569672, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5898, | |
| "num_tokens": 326131119.0, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.46345101939333666, | |
| "grad_norm": 0.12811586260795593, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5749, | |
| "num_tokens": 327476246.0, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.46544007956240674, | |
| "grad_norm": 0.12052274495363235, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5946, | |
| "num_tokens": 328874811.0, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.4674291397314769, | |
| "grad_norm": 0.11987277865409851, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5886, | |
| "num_tokens": 330249034.0, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.469418199900547, | |
| "grad_norm": 0.11991394311189651, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5796, | |
| "num_tokens": 331639806.0, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.4714072600696171, | |
| "grad_norm": 0.11994659900665283, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5989, | |
| "num_tokens": 333038051.0, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.4733963202386872, | |
| "grad_norm": 0.12097202241420746, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5783, | |
| "num_tokens": 334431046.0, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.47538538040775735, | |
| "grad_norm": 0.12271056324243546, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5896, | |
| "num_tokens": 335827008.0, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.47737444057682743, | |
| "grad_norm": 0.12297134101390839, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5999, | |
| "num_tokens": 337226457.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.47936350074589756, | |
| "grad_norm": 0.1218937486410141, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5926, | |
| "num_tokens": 338666075.0, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.4813525609149677, | |
| "grad_norm": 0.12227629870176315, | |
| "learning_rate": 1e-06, | |
| "loss": 0.58, | |
| "num_tokens": 340097602.0, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.4833416210840378, | |
| "grad_norm": 0.11952868849039078, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5948, | |
| "num_tokens": 341495194.0, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.4853306812531079, | |
| "grad_norm": 0.11985652893781662, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5774, | |
| "num_tokens": 342859196.0, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.48731974142217804, | |
| "grad_norm": 0.11957862228155136, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5755, | |
| "num_tokens": 344277995.0, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.4893088015912481, | |
| "grad_norm": 0.12266214936971664, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5877, | |
| "num_tokens": 345669645.0, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.49129786176031826, | |
| "grad_norm": 0.11971444636583328, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5962, | |
| "num_tokens": 347043955.0, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.4932869219293884, | |
| "grad_norm": 0.12104374915361404, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5878, | |
| "num_tokens": 348465321.0, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.49527598209845847, | |
| "grad_norm": 0.12102899700403214, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5904, | |
| "num_tokens": 349862835.0, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.4972650422675286, | |
| "grad_norm": 0.11829105764627457, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5826, | |
| "num_tokens": 351283798.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.49925410243659873, | |
| "grad_norm": 0.11999989300966263, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5983, | |
| "num_tokens": 352684836.0, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.5012431626056688, | |
| "grad_norm": 0.11723977327346802, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5951, | |
| "num_tokens": 354084748.0, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.5032322227747389, | |
| "grad_norm": 0.12956112623214722, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5898, | |
| "num_tokens": 355478595.0, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.5052212829438091, | |
| "grad_norm": 0.11831249296665192, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5901, | |
| "num_tokens": 356886933.0, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.5072103431128792, | |
| "grad_norm": 0.11731645464897156, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5787, | |
| "num_tokens": 358295994.0, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5091994032819492, | |
| "grad_norm": 0.12075242400169373, | |
| "learning_rate": 1e-06, | |
| "loss": 0.591, | |
| "num_tokens": 359683170.0, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.5111884634510194, | |
| "grad_norm": 0.12081367522478104, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5903, | |
| "num_tokens": 361079150.0, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.5131775236200895, | |
| "grad_norm": 0.11680179834365845, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5706, | |
| "num_tokens": 362477549.0, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.5151665837891596, | |
| "grad_norm": 0.11828526854515076, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5811, | |
| "num_tokens": 363848018.0, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.5171556439582298, | |
| "grad_norm": 0.1175350472331047, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5771, | |
| "num_tokens": 365236527.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5191447041272998, | |
| "grad_norm": 0.1198112890124321, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5787, | |
| "num_tokens": 366640485.0, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.5211337642963699, | |
| "grad_norm": 0.11720699816942215, | |
| "learning_rate": 1e-06, | |
| "loss": 0.584, | |
| "num_tokens": 368033122.0, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.5231228244654401, | |
| "grad_norm": 0.12382423877716064, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5866, | |
| "num_tokens": 369417377.0, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.5251118846345102, | |
| "grad_norm": 0.12134955078363419, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5757, | |
| "num_tokens": 370794212.0, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.5271009448035803, | |
| "grad_norm": 0.12391626089811325, | |
| "learning_rate": 1e-06, | |
| "loss": 0.601, | |
| "num_tokens": 372225021.0, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5290900049726505, | |
| "grad_norm": 0.11900907754898071, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5909, | |
| "num_tokens": 373610685.0, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.5310790651417205, | |
| "grad_norm": 0.11934248358011246, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5876, | |
| "num_tokens": 375021896.0, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.5330681253107906, | |
| "grad_norm": 0.12139896303415298, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5821, | |
| "num_tokens": 376456216.0, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.5350571854798608, | |
| "grad_norm": 0.12349140644073486, | |
| "learning_rate": 1e-06, | |
| "loss": 0.58, | |
| "num_tokens": 377828715.0, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.5370462456489309, | |
| "grad_norm": 0.12981721758842468, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5782, | |
| "num_tokens": 379262595.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.539035305818001, | |
| "grad_norm": 0.12098333984613419, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5906, | |
| "num_tokens": 380657341.0, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.5410243659870712, | |
| "grad_norm": 0.1278562843799591, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5891, | |
| "num_tokens": 382061345.0, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.5430134261561412, | |
| "grad_norm": 0.11872877925634384, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5765, | |
| "num_tokens": 383439874.0, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.5450024863252113, | |
| "grad_norm": 0.11846951395273209, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5803, | |
| "num_tokens": 384849495.0, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.5469915464942815, | |
| "grad_norm": 0.11676832288503647, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5859, | |
| "num_tokens": 386272096.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5489806066633516, | |
| "grad_norm": 0.118569515645504, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5706, | |
| "num_tokens": 387673261.0, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.5509696668324217, | |
| "grad_norm": 0.11762821674346924, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5849, | |
| "num_tokens": 389067728.0, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.5529587270014918, | |
| "grad_norm": 0.11877725273370743, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5909, | |
| "num_tokens": 390485805.0, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.5549477871705619, | |
| "grad_norm": 0.11775851249694824, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5836, | |
| "num_tokens": 391908547.0, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.556936847339632, | |
| "grad_norm": 0.11834586411714554, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5788, | |
| "num_tokens": 393310396.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5589259075087022, | |
| "grad_norm": 0.1218111664056778, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5927, | |
| "num_tokens": 394698059.0, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.5609149676777723, | |
| "grad_norm": 0.12251269072294235, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5801, | |
| "num_tokens": 396077595.0, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.5629040278468423, | |
| "grad_norm": 0.11502744257450104, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5719, | |
| "num_tokens": 397494815.0, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.5648930880159125, | |
| "grad_norm": 0.11885383725166321, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5706, | |
| "num_tokens": 398915154.0, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.5668821481849826, | |
| "grad_norm": 0.1333908885717392, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5702, | |
| "num_tokens": 400321357.0, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.5688712083540527, | |
| "grad_norm": 0.12071363627910614, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5818, | |
| "num_tokens": 401692589.0, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.5708602685231228, | |
| "grad_norm": 0.12001436948776245, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5859, | |
| "num_tokens": 403117801.0, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.572849328692193, | |
| "grad_norm": 0.12118804454803467, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5738, | |
| "num_tokens": 404520032.0, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.574838388861263, | |
| "grad_norm": 0.12114690989255905, | |
| "learning_rate": 1e-06, | |
| "loss": 0.581, | |
| "num_tokens": 405931871.0, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.5768274490303331, | |
| "grad_norm": 0.11723317950963974, | |
| "learning_rate": 1e-06, | |
| "loss": 0.569, | |
| "num_tokens": 407337363.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5788165091994033, | |
| "grad_norm": 0.11783155053853989, | |
| "learning_rate": 1e-06, | |
| "loss": 0.58, | |
| "num_tokens": 408742361.0, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.5808055693684734, | |
| "grad_norm": 0.12249549478292465, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5759, | |
| "num_tokens": 410152716.0, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.5827946295375435, | |
| "grad_norm": 0.11785644292831421, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5719, | |
| "num_tokens": 411547709.0, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.5847836897066137, | |
| "grad_norm": 0.11994913220405579, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5839, | |
| "num_tokens": 412958152.0, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.5867727498756837, | |
| "grad_norm": 0.1180441826581955, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5874, | |
| "num_tokens": 414335814.0, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.5887618100447538, | |
| "grad_norm": 0.1199953630566597, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5892, | |
| "num_tokens": 415753335.0, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.590750870213824, | |
| "grad_norm": 0.12051574885845184, | |
| "learning_rate": 1e-06, | |
| "loss": 0.587, | |
| "num_tokens": 417123268.0, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.5927399303828941, | |
| "grad_norm": 0.12905830144882202, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5822, | |
| "num_tokens": 418534320.0, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.5947289905519642, | |
| "grad_norm": 0.11819034814834595, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5738, | |
| "num_tokens": 419934167.0, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.5967180507210343, | |
| "grad_norm": 0.1195630431175232, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5891, | |
| "num_tokens": 421330163.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5987071108901044, | |
| "grad_norm": 0.11934220045804977, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5681, | |
| "num_tokens": 422702201.0, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.6006961710591745, | |
| "grad_norm": 0.11755826324224472, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5843, | |
| "num_tokens": 424135521.0, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.6026852312282447, | |
| "grad_norm": 0.11706581711769104, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5748, | |
| "num_tokens": 425481108.0, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.6046742913973148, | |
| "grad_norm": 0.11658696830272675, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5702, | |
| "num_tokens": 426880860.0, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.6066633515663848, | |
| "grad_norm": 0.12044768035411835, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5723, | |
| "num_tokens": 428272765.0, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.608652411735455, | |
| "grad_norm": 0.12221231311559677, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5645, | |
| "num_tokens": 429650167.0, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.6106414719045251, | |
| "grad_norm": 0.11917625367641449, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5748, | |
| "num_tokens": 431055294.0, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.6126305320735952, | |
| "grad_norm": 0.11839272826910019, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5692, | |
| "num_tokens": 432471600.0, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.6146195922426654, | |
| "grad_norm": 0.1222674548625946, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5798, | |
| "num_tokens": 433875649.0, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.6166086524117355, | |
| "grad_norm": 0.11733808368444443, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5641, | |
| "num_tokens": 435283504.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6185977125808055, | |
| "grad_norm": 0.1265021413564682, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5758, | |
| "num_tokens": 436661567.0, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.6205867727498757, | |
| "grad_norm": 0.11938843131065369, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5749, | |
| "num_tokens": 438042305.0, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.6225758329189458, | |
| "grad_norm": 0.11977506428956985, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5715, | |
| "num_tokens": 439453603.0, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.6245648930880159, | |
| "grad_norm": 0.1187228411436081, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5738, | |
| "num_tokens": 440857261.0, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.6265539532570861, | |
| "grad_norm": 0.11981964856386185, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5798, | |
| "num_tokens": 442261964.0, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6285430134261561, | |
| "grad_norm": 0.11806244403123856, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5789, | |
| "num_tokens": 443693559.0, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.6305320735952262, | |
| "grad_norm": 0.11995609104633331, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5736, | |
| "num_tokens": 445134735.0, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.6325211337642964, | |
| "grad_norm": 0.11736578494310379, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5782, | |
| "num_tokens": 446570052.0, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.6345101939333665, | |
| "grad_norm": 0.11504673957824707, | |
| "learning_rate": 1e-06, | |
| "loss": 0.57, | |
| "num_tokens": 447974813.0, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.6364992541024366, | |
| "grad_norm": 0.11872579902410507, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5781, | |
| "num_tokens": 449374619.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6384883142715068, | |
| "grad_norm": 0.11843977123498917, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5719, | |
| "num_tokens": 450791966.0, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.6404773744405768, | |
| "grad_norm": 0.11978229135274887, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5728, | |
| "num_tokens": 452194196.0, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.6424664346096469, | |
| "grad_norm": 0.11723372340202332, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5683, | |
| "num_tokens": 453551548.0, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.6444554947787171, | |
| "grad_norm": 0.11690861731767654, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5702, | |
| "num_tokens": 454937681.0, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.6464445549477872, | |
| "grad_norm": 0.11684879660606384, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5692, | |
| "num_tokens": 456361552.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6484336151168573, | |
| "grad_norm": 0.11809241771697998, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5952, | |
| "num_tokens": 457751294.0, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.6504226752859275, | |
| "grad_norm": 0.11862190812826157, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5756, | |
| "num_tokens": 459155713.0, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.6524117354549975, | |
| "grad_norm": 0.12103772908449173, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5678, | |
| "num_tokens": 460560134.0, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.6544007956240676, | |
| "grad_norm": 0.1281164437532425, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5845, | |
| "num_tokens": 461962174.0, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.6563898557931377, | |
| "grad_norm": 0.1147463396191597, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5784, | |
| "num_tokens": 463406269.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6583789159622079, | |
| "grad_norm": 0.11639434099197388, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5762, | |
| "num_tokens": 464831720.0, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.660367976131278, | |
| "grad_norm": 0.12189288437366486, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5708, | |
| "num_tokens": 466236911.0, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.662357036300348, | |
| "grad_norm": 0.11594757437705994, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5785, | |
| "num_tokens": 467667060.0, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.6643460964694182, | |
| "grad_norm": 0.120713010430336, | |
| "learning_rate": 1e-06, | |
| "loss": 0.56, | |
| "num_tokens": 469076105.0, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.6663351566384883, | |
| "grad_norm": 0.1252930611371994, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5586, | |
| "num_tokens": 470479085.0, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.6683242168075584, | |
| "grad_norm": 0.12119864672422409, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5696, | |
| "num_tokens": 471878453.0, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.6703132769766286, | |
| "grad_norm": 0.11741020530462265, | |
| "learning_rate": 1e-06, | |
| "loss": 0.578, | |
| "num_tokens": 473317934.0, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.6723023371456986, | |
| "grad_norm": 0.11722774058580399, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5705, | |
| "num_tokens": 474718637.0, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.6742913973147687, | |
| "grad_norm": 0.1168806403875351, | |
| "learning_rate": 1e-06, | |
| "loss": 0.576, | |
| "num_tokens": 476109339.0, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.6762804574838389, | |
| "grad_norm": 0.11499282717704773, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5695, | |
| "num_tokens": 477530852.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.678269517652909, | |
| "grad_norm": 0.11914825439453125, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5647, | |
| "num_tokens": 478878070.0, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.6802585778219791, | |
| "grad_norm": 0.11614906787872314, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5749, | |
| "num_tokens": 480284827.0, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.6822476379910493, | |
| "grad_norm": 0.11405870318412781, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5728, | |
| "num_tokens": 481705922.0, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.6842366981601193, | |
| "grad_norm": 0.12021178752183914, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5757, | |
| "num_tokens": 483096669.0, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.6862257583291894, | |
| "grad_norm": 0.11697462201118469, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5778, | |
| "num_tokens": 484497174.0, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.6882148184982596, | |
| "grad_norm": 0.11658283323049545, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5527, | |
| "num_tokens": 485910302.0, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.6902038786673297, | |
| "grad_norm": 0.11884041875600815, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5774, | |
| "num_tokens": 487314373.0, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.6921929388363998, | |
| "grad_norm": 0.11939556896686554, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5808, | |
| "num_tokens": 488749735.0, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.69418199900547, | |
| "grad_norm": 0.12483149766921997, | |
| "learning_rate": 1e-06, | |
| "loss": 0.558, | |
| "num_tokens": 490118122.0, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.69617105917454, | |
| "grad_norm": 0.12257801741361618, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5624, | |
| "num_tokens": 491510538.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6981601193436101, | |
| "grad_norm": 0.11767923086881638, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5718, | |
| "num_tokens": 492926246.0, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.7001491795126803, | |
| "grad_norm": 0.11635064333677292, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5803, | |
| "num_tokens": 494380244.0, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.7021382396817504, | |
| "grad_norm": 0.12177010625600815, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5652, | |
| "num_tokens": 495730583.0, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.7041272998508205, | |
| "grad_norm": 0.11822908371686935, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5663, | |
| "num_tokens": 497125075.0, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.7061163600198906, | |
| "grad_norm": 0.12682731449604034, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5658, | |
| "num_tokens": 498549275.0, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7081054201889607, | |
| "grad_norm": 0.12045978009700775, | |
| "learning_rate": 1e-06, | |
| "loss": 0.57, | |
| "num_tokens": 499930592.0, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.7100944803580308, | |
| "grad_norm": 0.12545664608478546, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5652, | |
| "num_tokens": 501333957.0, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.712083540527101, | |
| "grad_norm": 0.11865141242742538, | |
| "learning_rate": 1e-06, | |
| "loss": 0.566, | |
| "num_tokens": 502737493.0, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.7140726006961711, | |
| "grad_norm": 0.12852944433689117, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5777, | |
| "num_tokens": 504130760.0, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.7160616608652411, | |
| "grad_norm": 0.11820497363805771, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5542, | |
| "num_tokens": 505509701.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7180507210343113, | |
| "grad_norm": 0.1146169900894165, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5568, | |
| "num_tokens": 506911298.0, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.7200397812033814, | |
| "grad_norm": 0.11726175993680954, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5609, | |
| "num_tokens": 508293436.0, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.7220288413724515, | |
| "grad_norm": 0.12022433429956436, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5605, | |
| "num_tokens": 509696103.0, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.7240179015415217, | |
| "grad_norm": 0.11907174438238144, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5652, | |
| "num_tokens": 511108327.0, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.7260069617105918, | |
| "grad_norm": 0.11890577524900436, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5618, | |
| "num_tokens": 512511723.0, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7279960218796618, | |
| "grad_norm": 0.12470176070928574, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5584, | |
| "num_tokens": 513907412.0, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.729985082048732, | |
| "grad_norm": 0.12026024609804153, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5774, | |
| "num_tokens": 515324644.0, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.7319741422178021, | |
| "grad_norm": 0.12734608352184296, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5739, | |
| "num_tokens": 516674022.0, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.7339632023868722, | |
| "grad_norm": 0.11902155727148056, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5717, | |
| "num_tokens": 518087720.0, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.7359522625559424, | |
| "grad_norm": 0.11661865562200546, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5528, | |
| "num_tokens": 519497699.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7379413227250124, | |
| "grad_norm": 0.12583503127098083, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5767, | |
| "num_tokens": 520908702.0, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.7399303828940825, | |
| "grad_norm": 0.11799920350313187, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5692, | |
| "num_tokens": 522317938.0, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.7419194430631526, | |
| "grad_norm": 0.12218224257230759, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5642, | |
| "num_tokens": 523719142.0, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.7439085032322228, | |
| "grad_norm": 0.11948370188474655, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5702, | |
| "num_tokens": 525136176.0, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.7458975634012929, | |
| "grad_norm": 0.1194443628191948, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5472, | |
| "num_tokens": 526481164.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.747886623570363, | |
| "grad_norm": 0.12138593941926956, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5667, | |
| "num_tokens": 527888658.0, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.7498756837394331, | |
| "grad_norm": 0.12166504561901093, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5552, | |
| "num_tokens": 529287967.0, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.7518647439085032, | |
| "grad_norm": 0.1202344223856926, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5527, | |
| "num_tokens": 530654844.0, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.7538538040775733, | |
| "grad_norm": 0.11958526074886322, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5647, | |
| "num_tokens": 532043235.0, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.7558428642466435, | |
| "grad_norm": 0.11844973266124725, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5721, | |
| "num_tokens": 533462609.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7578319244157136, | |
| "grad_norm": 0.12316368520259857, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5659, | |
| "num_tokens": 534847372.0, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.7598209845847836, | |
| "grad_norm": 0.1315903514623642, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5636, | |
| "num_tokens": 536262717.0, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.7618100447538538, | |
| "grad_norm": 0.11977870017290115, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5708, | |
| "num_tokens": 537686599.0, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.7637991049229239, | |
| "grad_norm": 0.12060414999723434, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5605, | |
| "num_tokens": 539059998.0, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.765788165091994, | |
| "grad_norm": 0.11461476981639862, | |
| "learning_rate": 1e-06, | |
| "loss": 0.569, | |
| "num_tokens": 540488356.0, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.7677772252610642, | |
| "grad_norm": 0.11902087926864624, | |
| "learning_rate": 1e-06, | |
| "loss": 0.563, | |
| "num_tokens": 541895951.0, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.7697662854301343, | |
| "grad_norm": 0.11860883980989456, | |
| "learning_rate": 1e-06, | |
| "loss": 0.575, | |
| "num_tokens": 543281120.0, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.7717553455992043, | |
| "grad_norm": 0.12020647525787354, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5614, | |
| "num_tokens": 544646661.0, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.7737444057682745, | |
| "grad_norm": 0.11638668179512024, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5742, | |
| "num_tokens": 546097259.0, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.7757334659373446, | |
| "grad_norm": 0.11645980924367905, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5647, | |
| "num_tokens": 547518580.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7777225261064147, | |
| "grad_norm": 0.12269024550914764, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5689, | |
| "num_tokens": 548932428.0, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.7797115862754849, | |
| "grad_norm": 0.11761284619569778, | |
| "learning_rate": 1e-06, | |
| "loss": 0.554, | |
| "num_tokens": 550354089.0, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.781700646444555, | |
| "grad_norm": 0.1194002628326416, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5691, | |
| "num_tokens": 551785808.0, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.783689706613625, | |
| "grad_norm": 0.11770683526992798, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5675, | |
| "num_tokens": 553166180.0, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.7856787667826952, | |
| "grad_norm": 0.12067251652479172, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5698, | |
| "num_tokens": 554595305.0, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.7876678269517653, | |
| "grad_norm": 0.1211480051279068, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5545, | |
| "num_tokens": 555975917.0, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.7896568871208354, | |
| "grad_norm": 0.12212938815355301, | |
| "learning_rate": 1e-06, | |
| "loss": 0.563, | |
| "num_tokens": 557357748.0, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.7916459472899056, | |
| "grad_norm": 0.12312401086091995, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5635, | |
| "num_tokens": 558758541.0, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.7936350074589756, | |
| "grad_norm": 0.11911605298519135, | |
| "learning_rate": 1e-06, | |
| "loss": 0.569, | |
| "num_tokens": 560135846.0, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.7956240676280457, | |
| "grad_norm": 0.8023229241371155, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5761, | |
| "num_tokens": 561561065.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7976131277971159, | |
| "grad_norm": 0.12671087682247162, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5718, | |
| "num_tokens": 562950768.0, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.799602187966186, | |
| "grad_norm": 0.12295536696910858, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5531, | |
| "num_tokens": 564325036.0, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.8015912481352561, | |
| "grad_norm": 0.11960072070360184, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5551, | |
| "num_tokens": 565703350.0, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.8035803083043263, | |
| "grad_norm": 0.11970996856689453, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5758, | |
| "num_tokens": 567129767.0, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.8055693684733963, | |
| "grad_norm": 0.11514374613761902, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5667, | |
| "num_tokens": 568533394.0, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8075584286424664, | |
| "grad_norm": 0.12267459183931351, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5698, | |
| "num_tokens": 569966937.0, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.8095474888115366, | |
| "grad_norm": 0.1218687891960144, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5588, | |
| "num_tokens": 571338522.0, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.8115365489806067, | |
| "grad_norm": 0.12409517914056778, | |
| "learning_rate": 1e-06, | |
| "loss": 0.564, | |
| "num_tokens": 572731805.0, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.8135256091496768, | |
| "grad_norm": 0.11474578827619553, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5704, | |
| "num_tokens": 574159917.0, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.8155146693187469, | |
| "grad_norm": 0.11738289892673492, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5615, | |
| "num_tokens": 575571352.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.817503729487817, | |
| "grad_norm": 0.12023717910051346, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5569, | |
| "num_tokens": 576984870.0, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.8194927896568871, | |
| "grad_norm": 0.1194944977760315, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5666, | |
| "num_tokens": 578414303.0, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.8214818498259573, | |
| "grad_norm": 0.12020692229270935, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5478, | |
| "num_tokens": 579812487.0, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.8234709099950274, | |
| "grad_norm": 0.11791769415140152, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5681, | |
| "num_tokens": 581203528.0, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.8254599701640974, | |
| "grad_norm": 0.11593407392501831, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5573, | |
| "num_tokens": 582588268.0, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8274490303331675, | |
| "grad_norm": 0.12016279250383377, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5777, | |
| "num_tokens": 583983078.0, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.8294380905022377, | |
| "grad_norm": 0.12653815746307373, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5839, | |
| "num_tokens": 585400140.0, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.8314271506713078, | |
| "grad_norm": 0.11608505994081497, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5575, | |
| "num_tokens": 586830320.0, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.8334162108403779, | |
| "grad_norm": 0.11865837126970291, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5744, | |
| "num_tokens": 588258999.0, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.8354052710094481, | |
| "grad_norm": 0.11725205928087234, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5693, | |
| "num_tokens": 589692364.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8373943311785181, | |
| "grad_norm": 0.11638808995485306, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5597, | |
| "num_tokens": 591094415.0, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.8393833913475882, | |
| "grad_norm": 0.12141529470682144, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5559, | |
| "num_tokens": 592500735.0, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.8413724515166584, | |
| "grad_norm": 0.11554522812366486, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5666, | |
| "num_tokens": 593903743.0, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.8433615116857285, | |
| "grad_norm": 0.11890975385904312, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5543, | |
| "num_tokens": 595314514.0, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.8453505718547986, | |
| "grad_norm": 0.12231077998876572, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5615, | |
| "num_tokens": 596671023.0, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.8473396320238687, | |
| "grad_norm": 0.12027207762002945, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5583, | |
| "num_tokens": 598031581.0, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.8493286921929388, | |
| "grad_norm": 0.11685006320476532, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5628, | |
| "num_tokens": 599477745.0, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.8513177523620089, | |
| "grad_norm": 0.17819620668888092, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5714, | |
| "num_tokens": 600848881.0, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.8533068125310791, | |
| "grad_norm": 0.11678969860076904, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5534, | |
| "num_tokens": 602266449.0, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.8552958727001492, | |
| "grad_norm": 0.11610161513090134, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5609, | |
| "num_tokens": 603675715.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8572849328692193, | |
| "grad_norm": 0.1225874274969101, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5554, | |
| "num_tokens": 605072725.0, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.8592739930382894, | |
| "grad_norm": 0.11319524794816971, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5538, | |
| "num_tokens": 606474929.0, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.8612630532073595, | |
| "grad_norm": 0.11506962776184082, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5531, | |
| "num_tokens": 607855829.0, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.8632521133764296, | |
| "grad_norm": 0.1158682256937027, | |
| "learning_rate": 1e-06, | |
| "loss": 0.567, | |
| "num_tokens": 609285755.0, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.8652411735454998, | |
| "grad_norm": 0.11761818826198578, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5671, | |
| "num_tokens": 610697282.0, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.8672302337145699, | |
| "grad_norm": 0.11838319152593613, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5534, | |
| "num_tokens": 612094535.0, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.8692192938836399, | |
| "grad_norm": 0.11570697277784348, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5632, | |
| "num_tokens": 613501408.0, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.8712083540527101, | |
| "grad_norm": 0.11999750137329102, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5594, | |
| "num_tokens": 614872296.0, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.8731974142217802, | |
| "grad_norm": 0.1152532622218132, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5573, | |
| "num_tokens": 616300717.0, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.8751864743908503, | |
| "grad_norm": 0.11520667374134064, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5441, | |
| "num_tokens": 617678025.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8771755345599205, | |
| "grad_norm": 0.11722690612077713, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5535, | |
| "num_tokens": 619082991.0, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.8791645947289906, | |
| "grad_norm": 0.11644702404737473, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5595, | |
| "num_tokens": 620470914.0, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.8811536548980606, | |
| "grad_norm": 0.11839190125465393, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5481, | |
| "num_tokens": 621862665.0, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.8831427150671308, | |
| "grad_norm": 0.1157626211643219, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5663, | |
| "num_tokens": 623310812.0, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.8851317752362009, | |
| "grad_norm": 0.11904613673686981, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5594, | |
| "num_tokens": 624718938.0, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.887120835405271, | |
| "grad_norm": 0.11730271577835083, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5515, | |
| "num_tokens": 626091716.0, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.8891098955743412, | |
| "grad_norm": 0.11457692086696625, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5506, | |
| "num_tokens": 627487215.0, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.8910989557434112, | |
| "grad_norm": 0.11875864863395691, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5526, | |
| "num_tokens": 628850853.0, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.8930880159124813, | |
| "grad_norm": 0.11865832656621933, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5526, | |
| "num_tokens": 630249653.0, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.8950770760815515, | |
| "grad_norm": 0.11921685934066772, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5651, | |
| "num_tokens": 631626896.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8970661362506216, | |
| "grad_norm": 0.11800325661897659, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5463, | |
| "num_tokens": 633018745.0, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.8990551964196917, | |
| "grad_norm": 0.11664669215679169, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5529, | |
| "num_tokens": 634434489.0, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.9010442565887619, | |
| "grad_norm": 0.11907332390546799, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5588, | |
| "num_tokens": 635836521.0, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.9030333167578319, | |
| "grad_norm": 0.119346983730793, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5605, | |
| "num_tokens": 637251420.0, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.905022376926902, | |
| "grad_norm": 0.11810777336359024, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5538, | |
| "num_tokens": 638673973.0, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9070114370959722, | |
| "grad_norm": 0.12147301435470581, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5477, | |
| "num_tokens": 640053944.0, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.9090004972650423, | |
| "grad_norm": 0.11813312768936157, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5699, | |
| "num_tokens": 641459580.0, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.9109895574341124, | |
| "grad_norm": 0.1261916309595108, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5565, | |
| "num_tokens": 642848898.0, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.9129786176031826, | |
| "grad_norm": 0.11821988970041275, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5527, | |
| "num_tokens": 644312542.0, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.9149676777722526, | |
| "grad_norm": 0.11542089283466339, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5614, | |
| "num_tokens": 645735657.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9169567379413227, | |
| "grad_norm": 0.131094828248024, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5682, | |
| "num_tokens": 647137018.0, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.9189457981103928, | |
| "grad_norm": 0.11712004989385605, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5578, | |
| "num_tokens": 648516983.0, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.920934858279463, | |
| "grad_norm": 0.11788227409124374, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5591, | |
| "num_tokens": 649953519.0, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.922923918448533, | |
| "grad_norm": 0.11795365810394287, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5463, | |
| "num_tokens": 651366264.0, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.9249129786176031, | |
| "grad_norm": 0.11824613064527512, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5664, | |
| "num_tokens": 652792598.0, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9269020387866733, | |
| "grad_norm": 0.11839722841978073, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5493, | |
| "num_tokens": 654162858.0, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.9288910989557434, | |
| "grad_norm": 0.11836480349302292, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5514, | |
| "num_tokens": 655539001.0, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.9308801591248135, | |
| "grad_norm": 0.17381928861141205, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5526, | |
| "num_tokens": 656952170.0, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.9328692192938837, | |
| "grad_norm": 0.11770905554294586, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5615, | |
| "num_tokens": 658358118.0, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.9348582794629537, | |
| "grad_norm": 0.11960657685995102, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5495, | |
| "num_tokens": 659752280.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9368473396320238, | |
| "grad_norm": 0.11758306622505188, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5592, | |
| "num_tokens": 661127510.0, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.938836399801094, | |
| "grad_norm": 0.1147281602025032, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5431, | |
| "num_tokens": 662539378.0, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.9408254599701641, | |
| "grad_norm": 0.11964991688728333, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5578, | |
| "num_tokens": 663935199.0, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.9428145201392342, | |
| "grad_norm": 0.12270357459783554, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5535, | |
| "num_tokens": 665328684.0, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.9448035803083044, | |
| "grad_norm": 0.11691749840974808, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5598, | |
| "num_tokens": 666742297.0, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.9467926404773744, | |
| "grad_norm": 0.11798378825187683, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5563, | |
| "num_tokens": 668121085.0, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.9487817006464445, | |
| "grad_norm": 0.11690951138734818, | |
| "learning_rate": 1e-06, | |
| "loss": 0.553, | |
| "num_tokens": 669537516.0, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.9507707608155147, | |
| "grad_norm": 0.12072078138589859, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5578, | |
| "num_tokens": 670932178.0, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.9527598209845848, | |
| "grad_norm": 0.11482840776443481, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5484, | |
| "num_tokens": 672306421.0, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.9547488811536549, | |
| "grad_norm": 0.12619654834270477, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5604, | |
| "num_tokens": 673688811.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.956737941322725, | |
| "grad_norm": 0.12017329037189484, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5599, | |
| "num_tokens": 675082703.0, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.9587270014917951, | |
| "grad_norm": 0.11547863483428955, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5475, | |
| "num_tokens": 676488414.0, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.9607160616608652, | |
| "grad_norm": 0.11614055931568146, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5621, | |
| "num_tokens": 677896011.0, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.9627051218299354, | |
| "grad_norm": 0.11469247937202454, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5503, | |
| "num_tokens": 679301564.0, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.9646941819990055, | |
| "grad_norm": 0.12310828268527985, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5526, | |
| "num_tokens": 680674072.0, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.9666832421680756, | |
| "grad_norm": 0.11646957695484161, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5505, | |
| "num_tokens": 682077773.0, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.9686723023371457, | |
| "grad_norm": 0.11596749722957611, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5596, | |
| "num_tokens": 683492461.0, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.9706613625062158, | |
| "grad_norm": 0.11795108765363693, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5568, | |
| "num_tokens": 684913267.0, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.9726504226752859, | |
| "grad_norm": 0.1201254203915596, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5645, | |
| "num_tokens": 686315028.0, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.9746394828443561, | |
| "grad_norm": 0.1162814274430275, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5494, | |
| "num_tokens": 687711185.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9766285430134262, | |
| "grad_norm": 0.11750290542840958, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5628, | |
| "num_tokens": 689115928.0, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.9786176031824962, | |
| "grad_norm": 0.13300985097885132, | |
| "learning_rate": 1e-06, | |
| "loss": 0.563, | |
| "num_tokens": 690534590.0, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.9806066633515664, | |
| "grad_norm": 0.12720529735088348, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5567, | |
| "num_tokens": 691981545.0, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.9825957235206365, | |
| "grad_norm": 0.11822197586297989, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5659, | |
| "num_tokens": 693402676.0, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.9845847836897066, | |
| "grad_norm": 0.11941008269786835, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5568, | |
| "num_tokens": 694809843.0, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.9865738438587768, | |
| "grad_norm": 0.11860588937997818, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5531, | |
| "num_tokens": 696211950.0, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.9885629040278469, | |
| "grad_norm": 0.12157568335533142, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5569, | |
| "num_tokens": 697624687.0, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.9905519641969169, | |
| "grad_norm": 0.11806346476078033, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5743, | |
| "num_tokens": 699049049.0, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.9925410243659871, | |
| "grad_norm": 0.1187855452299118, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5558, | |
| "num_tokens": 700440889.0, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.9945300845350572, | |
| "grad_norm": 0.11523641645908356, | |
| "learning_rate": 1e-06, | |
| "loss": 0.56, | |
| "num_tokens": 701864042.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9965191447041273, | |
| "grad_norm": 0.11771584302186966, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5536, | |
| "num_tokens": 703255455.0, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 0.9985082048731975, | |
| "grad_norm": 0.11954519152641296, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5487, | |
| "num_tokens": 704646212.0, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 1.0019890601690702, | |
| "grad_norm": 0.24864843487739563, | |
| "learning_rate": 1e-06, | |
| "loss": 1.1087, | |
| "num_tokens": 706359781.0, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 1.0039781203381402, | |
| "grad_norm": 0.15665407478809357, | |
| "learning_rate": 1e-06, | |
| "loss": 0.542, | |
| "num_tokens": 707777523.0, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 1.0059671805072103, | |
| "grad_norm": 0.12076932191848755, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5393, | |
| "num_tokens": 709165414.0, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 1.0079562406762805, | |
| "grad_norm": 0.11568966507911682, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5503, | |
| "num_tokens": 710541219.0, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 1.0099453008453505, | |
| "grad_norm": 0.1189185082912445, | |
| "learning_rate": 1e-06, | |
| "loss": 0.555, | |
| "num_tokens": 711904899.0, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.0119343610144207, | |
| "grad_norm": 0.11912715435028076, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5399, | |
| "num_tokens": 713297626.0, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 1.0139234211834909, | |
| "grad_norm": 0.11964980512857437, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5585, | |
| "num_tokens": 714712902.0, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 1.0159124813525608, | |
| "grad_norm": 0.11817710846662521, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5518, | |
| "num_tokens": 716119424.0, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.017901541521631, | |
| "grad_norm": 0.11699645221233368, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5447, | |
| "num_tokens": 717518297.0, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 1.0198906016907012, | |
| "grad_norm": 0.11419110000133514, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5532, | |
| "num_tokens": 718934437.0, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 1.0218796618597712, | |
| "grad_norm": 0.11955999583005905, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5365, | |
| "num_tokens": 720296609.0, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 1.0238687220288414, | |
| "grad_norm": 0.11673971265554428, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5549, | |
| "num_tokens": 721713131.0, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 1.0258577821979116, | |
| "grad_norm": 0.11641617864370346, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5654, | |
| "num_tokens": 723093279.0, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 1.0278468423669815, | |
| "grad_norm": 0.11750617623329163, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5391, | |
| "num_tokens": 724491829.0, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 1.0298359025360517, | |
| "grad_norm": 0.1122995987534523, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5421, | |
| "num_tokens": 725927036.0, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 1.031824962705122, | |
| "grad_norm": 0.11935453861951828, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5522, | |
| "num_tokens": 727309822.0, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 1.0338140228741919, | |
| "grad_norm": 0.1215103417634964, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5543, | |
| "num_tokens": 728731842.0, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 1.035803083043262, | |
| "grad_norm": 0.12213215231895447, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5502, | |
| "num_tokens": 730121195.0, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0377921432123323, | |
| "grad_norm": 0.11967755854129791, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5513, | |
| "num_tokens": 731512351.0, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 1.0397812033814022, | |
| "grad_norm": 0.11795569211244583, | |
| "learning_rate": 1e-06, | |
| "loss": 0.548, | |
| "num_tokens": 732931422.0, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 1.0417702635504724, | |
| "grad_norm": 0.11842501163482666, | |
| "learning_rate": 1e-06, | |
| "loss": 0.55, | |
| "num_tokens": 734353348.0, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 1.0437593237195426, | |
| "grad_norm": 0.1207478791475296, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5516, | |
| "num_tokens": 735742554.0, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 1.0457483838886126, | |
| "grad_norm": 0.11943433433771133, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5513, | |
| "num_tokens": 737158566.0, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.0477374440576828, | |
| "grad_norm": 0.12060469388961792, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5623, | |
| "num_tokens": 738552506.0, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.049726504226753, | |
| "grad_norm": 0.11873313784599304, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5502, | |
| "num_tokens": 739982404.0, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 1.051715564395823, | |
| "grad_norm": 0.11538344621658325, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5523, | |
| "num_tokens": 741379974.0, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 1.053704624564893, | |
| "grad_norm": 0.11816058307886124, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5625, | |
| "num_tokens": 742756391.0, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 1.0556936847339633, | |
| "grad_norm": 0.12244360148906708, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5433, | |
| "num_tokens": 744195716.0, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.0576827449030333, | |
| "grad_norm": 0.12584052979946136, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5553, | |
| "num_tokens": 745589558.0, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 1.0596718050721035, | |
| "grad_norm": 0.11298387497663498, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5477, | |
| "num_tokens": 746987877.0, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 1.0616608652411736, | |
| "grad_norm": 0.12312706559896469, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5701, | |
| "num_tokens": 748360690.0, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 1.0636499254102436, | |
| "grad_norm": 0.11732471734285355, | |
| "learning_rate": 1e-06, | |
| "loss": 0.545, | |
| "num_tokens": 749752632.0, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 1.0656389855793138, | |
| "grad_norm": 0.11663486808538437, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5556, | |
| "num_tokens": 751135143.0, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 1.067628045748384, | |
| "grad_norm": 0.11738405376672745, | |
| "learning_rate": 1e-06, | |
| "loss": 0.542, | |
| "num_tokens": 752504589.0, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 1.069617105917454, | |
| "grad_norm": 0.11650814861059189, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5501, | |
| "num_tokens": 753908272.0, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 1.0716061660865241, | |
| "grad_norm": 0.11730290204286575, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5525, | |
| "num_tokens": 755345567.0, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 1.0735952262555943, | |
| "grad_norm": 0.11732745170593262, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5471, | |
| "num_tokens": 756742209.0, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 1.0755842864246643, | |
| "grad_norm": 0.11920227110385895, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5453, | |
| "num_tokens": 758140909.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.0775733465937345, | |
| "grad_norm": 0.11768822371959686, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5427, | |
| "num_tokens": 759550869.0, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 1.0795624067628045, | |
| "grad_norm": 0.11679795384407043, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5615, | |
| "num_tokens": 760954672.0, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 1.0815514669318746, | |
| "grad_norm": 0.11920733749866486, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5505, | |
| "num_tokens": 762355822.0, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 1.0835405271009448, | |
| "grad_norm": 0.1188458576798439, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5678, | |
| "num_tokens": 763788308.0, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 1.085529587270015, | |
| "grad_norm": 0.11863641440868378, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5467, | |
| "num_tokens": 765177690.0, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 1.087518647439085, | |
| "grad_norm": 0.11603251844644547, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5369, | |
| "num_tokens": 766617783.0, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 1.0895077076081552, | |
| "grad_norm": 0.1217975988984108, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5431, | |
| "num_tokens": 768029229.0, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 1.0914967677772252, | |
| "grad_norm": 0.11564943194389343, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5438, | |
| "num_tokens": 769417054.0, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 1.0934858279462953, | |
| "grad_norm": 0.1208975538611412, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5502, | |
| "num_tokens": 770836479.0, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 1.0954748881153655, | |
| "grad_norm": 0.11667314916849136, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5547, | |
| "num_tokens": 772254674.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.0974639482844357, | |
| "grad_norm": 0.11516263335943222, | |
| "learning_rate": 1e-06, | |
| "loss": 0.55, | |
| "num_tokens": 773706020.0, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 1.0994530084535057, | |
| "grad_norm": 0.11822597682476044, | |
| "learning_rate": 1e-06, | |
| "loss": 0.54, | |
| "num_tokens": 775110778.0, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 1.1014420686225759, | |
| "grad_norm": 0.12346749752759933, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5471, | |
| "num_tokens": 776538512.0, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 1.1034311287916458, | |
| "grad_norm": 0.11938782781362534, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5472, | |
| "num_tokens": 777966414.0, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 1.105420188960716, | |
| "grad_norm": 0.14581921696662903, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5537, | |
| "num_tokens": 779354221.0, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 1.1074092491297862, | |
| "grad_norm": 0.11971963196992874, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5487, | |
| "num_tokens": 780779804.0, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 1.1093983092988562, | |
| "grad_norm": 0.11833789944648743, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5463, | |
| "num_tokens": 782198212.0, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 1.1113873694679264, | |
| "grad_norm": 0.12102148681879044, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5565, | |
| "num_tokens": 783630792.0, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 1.1133764296369966, | |
| "grad_norm": 0.11505177617073059, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5361, | |
| "num_tokens": 785041609.0, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 1.1153654898060665, | |
| "grad_norm": 0.12186893075704575, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5406, | |
| "num_tokens": 786417350.0, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.1173545499751367, | |
| "grad_norm": 0.12019500881433487, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5568, | |
| "num_tokens": 787852563.0, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 1.119343610144207, | |
| "grad_norm": 0.11836836487054825, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5338, | |
| "num_tokens": 789256403.0, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 1.1213326703132769, | |
| "grad_norm": 0.1224868968129158, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5363, | |
| "num_tokens": 790627959.0, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 1.123321730482347, | |
| "grad_norm": 0.1227540671825409, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5398, | |
| "num_tokens": 792047776.0, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 1.1253107906514173, | |
| "grad_norm": 0.12231338769197464, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5393, | |
| "num_tokens": 793437309.0, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 1.1272998508204872, | |
| "grad_norm": 0.11635982990264893, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5491, | |
| "num_tokens": 794832359.0, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 1.1292889109895574, | |
| "grad_norm": 0.11689839512109756, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5521, | |
| "num_tokens": 796221119.0, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 1.1312779711586276, | |
| "grad_norm": 0.11974216252565384, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5472, | |
| "num_tokens": 797660744.0, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 1.1332670313276976, | |
| "grad_norm": 0.1189328134059906, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5444, | |
| "num_tokens": 799060817.0, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 1.1352560914967678, | |
| "grad_norm": 0.1164221465587616, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5442, | |
| "num_tokens": 800454569.0, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.137245151665838, | |
| "grad_norm": 0.11653874069452286, | |
| "learning_rate": 1e-06, | |
| "loss": 0.55, | |
| "num_tokens": 801905348.0, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 1.139234211834908, | |
| "grad_norm": 0.11661481857299805, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5492, | |
| "num_tokens": 803300441.0, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 1.141223272003978, | |
| "grad_norm": 0.11541904509067535, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5299, | |
| "num_tokens": 804671703.0, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 1.1432123321730483, | |
| "grad_norm": 0.11833840608596802, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5442, | |
| "num_tokens": 806085229.0, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 1.1452013923421183, | |
| "grad_norm": 0.11650761216878891, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5363, | |
| "num_tokens": 807470114.0, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.1471904525111885, | |
| "grad_norm": 0.11970090866088867, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5517, | |
| "num_tokens": 808920938.0, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 1.1491795126802586, | |
| "grad_norm": 0.11860202997922897, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5531, | |
| "num_tokens": 810330812.0, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 1.1511685728493286, | |
| "grad_norm": 0.11822472512722015, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5386, | |
| "num_tokens": 811715751.0, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 1.1531576330183988, | |
| "grad_norm": 0.11776979267597198, | |
| "learning_rate": 1e-06, | |
| "loss": 0.538, | |
| "num_tokens": 813102121.0, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 1.155146693187469, | |
| "grad_norm": 0.11876077950000763, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5578, | |
| "num_tokens": 814501435.0, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.157135753356539, | |
| "grad_norm": 0.12163852900266647, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5301, | |
| "num_tokens": 815900526.0, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 1.1591248135256091, | |
| "grad_norm": 0.11880628764629364, | |
| "learning_rate": 1e-06, | |
| "loss": 0.553, | |
| "num_tokens": 817316569.0, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 1.1611138736946793, | |
| "grad_norm": 0.11747530102729797, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5407, | |
| "num_tokens": 818712696.0, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 1.1631029338637493, | |
| "grad_norm": 0.11508717387914658, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5525, | |
| "num_tokens": 820179550.0, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 1.1650919940328195, | |
| "grad_norm": 0.11923891305923462, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5299, | |
| "num_tokens": 821519204.0, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 1.1670810542018897, | |
| "grad_norm": 0.12130296975374222, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5584, | |
| "num_tokens": 822934109.0, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 1.1690701143709596, | |
| "grad_norm": 0.11868572235107422, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5237, | |
| "num_tokens": 824350114.0, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 1.1710591745400298, | |
| "grad_norm": 0.11723876744508743, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5535, | |
| "num_tokens": 825723811.0, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 1.1730482347091, | |
| "grad_norm": 0.11835741996765137, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5346, | |
| "num_tokens": 827152454.0, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 1.17503729487817, | |
| "grad_norm": 0.11868591606616974, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5351, | |
| "num_tokens": 828579537.0, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.1770263550472402, | |
| "grad_norm": 0.11987331509590149, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5367, | |
| "num_tokens": 829969684.0, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 1.1790154152163104, | |
| "grad_norm": 0.11743640154600143, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5548, | |
| "num_tokens": 831391395.0, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 1.1810044753853803, | |
| "grad_norm": 0.1182253286242485, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5516, | |
| "num_tokens": 832783491.0, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 1.1829935355544505, | |
| "grad_norm": 0.11603699624538422, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5356, | |
| "num_tokens": 834171588.0, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 1.1849825957235207, | |
| "grad_norm": 0.1183595210313797, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5511, | |
| "num_tokens": 835573502.0, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 1.1869716558925907, | |
| "grad_norm": 0.11468026787042618, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5509, | |
| "num_tokens": 837028240.0, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 1.1889607160616609, | |
| "grad_norm": 0.11567319929599762, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5472, | |
| "num_tokens": 838416832.0, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 1.190949776230731, | |
| "grad_norm": 0.11915028095245361, | |
| "learning_rate": 1e-06, | |
| "loss": 0.553, | |
| "num_tokens": 839812316.0, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 1.192938836399801, | |
| "grad_norm": 0.11828889697790146, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5442, | |
| "num_tokens": 841215463.0, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 1.1949278965688712, | |
| "grad_norm": 0.11770551651716232, | |
| "learning_rate": 1e-06, | |
| "loss": 0.547, | |
| "num_tokens": 842662727.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.1969169567379414, | |
| "grad_norm": 0.1193631961941719, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5452, | |
| "num_tokens": 844081793.0, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 1.1989060169070114, | |
| "grad_norm": 0.11217671632766724, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5401, | |
| "num_tokens": 845485094.0, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 1.2008950770760816, | |
| "grad_norm": 0.11600279062986374, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5565, | |
| "num_tokens": 846899288.0, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 1.2028841372451518, | |
| "grad_norm": 0.11796706914901733, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5434, | |
| "num_tokens": 848304697.0, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 1.2048731974142217, | |
| "grad_norm": 0.11813243478536606, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5535, | |
| "num_tokens": 849715462.0, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 1.206862257583292, | |
| "grad_norm": 0.1221594288945198, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5461, | |
| "num_tokens": 851059379.0, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 1.208851317752362, | |
| "grad_norm": 0.11475210636854172, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5394, | |
| "num_tokens": 852449620.0, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 1.210840377921432, | |
| "grad_norm": 0.1158720999956131, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5415, | |
| "num_tokens": 853876338.0, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 1.2128294380905023, | |
| "grad_norm": 0.11944854259490967, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5398, | |
| "num_tokens": 855326253.0, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 1.2148184982595724, | |
| "grad_norm": 0.11523836106061935, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5415, | |
| "num_tokens": 856726725.0, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.2168075584286424, | |
| "grad_norm": 0.11895252019166946, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5499, | |
| "num_tokens": 858107209.0, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 1.2187966185977126, | |
| "grad_norm": 0.11535745114088058, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5286, | |
| "num_tokens": 859540802.0, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 1.2207856787667828, | |
| "grad_norm": 0.1177186667919159, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5516, | |
| "num_tokens": 860956386.0, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 1.2227747389358528, | |
| "grad_norm": 0.11561235785484314, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5345, | |
| "num_tokens": 862351381.0, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 1.224763799104923, | |
| "grad_norm": 0.12278357893228531, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5436, | |
| "num_tokens": 863752120.0, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 1.2267528592739931, | |
| "grad_norm": 0.12000274658203125, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5409, | |
| "num_tokens": 865134261.0, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 1.228741919443063, | |
| "grad_norm": 0.11960814148187637, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5421, | |
| "num_tokens": 866500301.0, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 1.2307309796121333, | |
| "grad_norm": 0.11419054120779037, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5407, | |
| "num_tokens": 867914829.0, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 1.2327200397812033, | |
| "grad_norm": 0.11924876272678375, | |
| "learning_rate": 1e-06, | |
| "loss": 0.533, | |
| "num_tokens": 869296945.0, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 1.2347090999502734, | |
| "grad_norm": 0.12687626481056213, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5433, | |
| "num_tokens": 870720355.0, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.2366981601193436, | |
| "grad_norm": 0.11868540942668915, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5427, | |
| "num_tokens": 872090554.0, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 1.2386872202884138, | |
| "grad_norm": 0.11346638202667236, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5455, | |
| "num_tokens": 873519656.0, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 1.2406762804574838, | |
| "grad_norm": 0.12468260526657104, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5406, | |
| "num_tokens": 874897139.0, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 1.242665340626554, | |
| "grad_norm": 0.11793619394302368, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5541, | |
| "num_tokens": 876281923.0, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 1.244654400795624, | |
| "grad_norm": 0.11685628443956375, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5419, | |
| "num_tokens": 877695435.0, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.2466434609646941, | |
| "grad_norm": 0.12373646348714828, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5372, | |
| "num_tokens": 879058206.0, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 1.2486325211337643, | |
| "grad_norm": 0.11609544605016708, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5497, | |
| "num_tokens": 880465724.0, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 1.2506215813028345, | |
| "grad_norm": 0.11885792016983032, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5458, | |
| "num_tokens": 881863204.0, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 1.2526106414719045, | |
| "grad_norm": 0.1189800500869751, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5425, | |
| "num_tokens": 883235922.0, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 1.2545997016409747, | |
| "grad_norm": 0.11329221725463867, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5371, | |
| "num_tokens": 884644276.0, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.2565887618100446, | |
| "grad_norm": 0.11825796961784363, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5351, | |
| "num_tokens": 886073154.0, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 1.2585778219791148, | |
| "grad_norm": 0.114280566573143, | |
| "learning_rate": 1e-06, | |
| "loss": 0.541, | |
| "num_tokens": 887514458.0, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 1.260566882148185, | |
| "grad_norm": 0.1187988743185997, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5364, | |
| "num_tokens": 888906987.0, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 1.2625559423172552, | |
| "grad_norm": 0.11506423354148865, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5431, | |
| "num_tokens": 890319605.0, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 1.2645450024863252, | |
| "grad_norm": 0.1173451617360115, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5345, | |
| "num_tokens": 891716052.0, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 1.2665340626553954, | |
| "grad_norm": 0.11930102109909058, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5439, | |
| "num_tokens": 893119175.0, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 1.2685231228244653, | |
| "grad_norm": 0.11887087672948837, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5348, | |
| "num_tokens": 894490469.0, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 1.2705121829935355, | |
| "grad_norm": 0.11899300664663315, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5491, | |
| "num_tokens": 895868778.0, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 1.2725012431626057, | |
| "grad_norm": 0.116294264793396, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5537, | |
| "num_tokens": 897281587.0, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 1.274490303331676, | |
| "grad_norm": 0.12177430093288422, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5369, | |
| "num_tokens": 898705618.0, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.2764793635007459, | |
| "grad_norm": 0.11538566648960114, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5387, | |
| "num_tokens": 900118795.0, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 1.278468423669816, | |
| "grad_norm": 0.11888190358877182, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5448, | |
| "num_tokens": 901549973.0, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 1.280457483838886, | |
| "grad_norm": 0.11358219385147095, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5403, | |
| "num_tokens": 902947794.0, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 1.2824465440079562, | |
| "grad_norm": 0.11417380720376968, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5332, | |
| "num_tokens": 904349508.0, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 1.2844356041770264, | |
| "grad_norm": 0.11959497630596161, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5406, | |
| "num_tokens": 905747046.0, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 1.2864246643460966, | |
| "grad_norm": 0.11622175574302673, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5415, | |
| "num_tokens": 907173072.0, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 1.2884137245151666, | |
| "grad_norm": 0.11988142877817154, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5451, | |
| "num_tokens": 908584591.0, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 1.2904027846842367, | |
| "grad_norm": 0.1185469850897789, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5351, | |
| "num_tokens": 910011116.0, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 1.2923918448533067, | |
| "grad_norm": 0.11444271355867386, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5384, | |
| "num_tokens": 911427909.0, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 1.294380905022377, | |
| "grad_norm": 0.12033736705780029, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5498, | |
| "num_tokens": 912835863.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.296369965191447, | |
| "grad_norm": 0.1319677233695984, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5296, | |
| "num_tokens": 914238947.0, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 1.2983590253605173, | |
| "grad_norm": 0.12392336130142212, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5341, | |
| "num_tokens": 915655787.0, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 1.3003480855295872, | |
| "grad_norm": 0.11796288192272186, | |
| "learning_rate": 1e-06, | |
| "loss": 0.547, | |
| "num_tokens": 917089247.0, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 1.3023371456986574, | |
| "grad_norm": 0.12152981013059616, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5544, | |
| "num_tokens": 918505085.0, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 1.3043262058677274, | |
| "grad_norm": 0.11925685405731201, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5304, | |
| "num_tokens": 919892998.0, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 1.3063152660367976, | |
| "grad_norm": 0.11711208522319794, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5426, | |
| "num_tokens": 921240479.0, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 1.3083043262058678, | |
| "grad_norm": 0.12039055675268173, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5368, | |
| "num_tokens": 922632985.0, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 1.310293386374938, | |
| "grad_norm": 0.11820589005947113, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5324, | |
| "num_tokens": 924054020.0, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 1.312282446544008, | |
| "grad_norm": 0.11549760401248932, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5442, | |
| "num_tokens": 925466685.0, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 1.3142715067130781, | |
| "grad_norm": 0.12046794593334198, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5491, | |
| "num_tokens": 926874635.0, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.316260566882148, | |
| "grad_norm": 0.1153668761253357, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5371, | |
| "num_tokens": 928294778.0, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 1.3182496270512183, | |
| "grad_norm": 0.11516553908586502, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5405, | |
| "num_tokens": 929703109.0, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 1.3202386872202885, | |
| "grad_norm": 0.11781197786331177, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5342, | |
| "num_tokens": 931098455.0, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 1.3222277473893587, | |
| "grad_norm": 0.11899585276842117, | |
| "learning_rate": 1e-06, | |
| "loss": 0.546, | |
| "num_tokens": 932511314.0, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 1.3242168075584286, | |
| "grad_norm": 0.11900392174720764, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5404, | |
| "num_tokens": 933900286.0, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 1.3262058677274988, | |
| "grad_norm": 0.12140580266714096, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5355, | |
| "num_tokens": 935253652.0, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 1.3281949278965688, | |
| "grad_norm": 0.12140516191720963, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5394, | |
| "num_tokens": 936636184.0, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 1.330183988065639, | |
| "grad_norm": 0.11907900869846344, | |
| "learning_rate": 1e-06, | |
| "loss": 0.527, | |
| "num_tokens": 938023738.0, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 1.3321730482347092, | |
| "grad_norm": 0.12013056874275208, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5444, | |
| "num_tokens": 939396704.0, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 1.3341621084037791, | |
| "grad_norm": 0.11747721582651138, | |
| "learning_rate": 1e-06, | |
| "loss": 0.534, | |
| "num_tokens": 940820588.0, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.3361511685728493, | |
| "grad_norm": 0.11940892785787582, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5323, | |
| "num_tokens": 942230394.0, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.3381402287419195, | |
| "grad_norm": 0.12076081335544586, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5395, | |
| "num_tokens": 943640623.0, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 1.3401292889109895, | |
| "grad_norm": 0.11554915457963943, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5423, | |
| "num_tokens": 945055848.0, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 1.3421183490800597, | |
| "grad_norm": 0.11654532700777054, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5235, | |
| "num_tokens": 946461742.0, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 1.3441074092491299, | |
| "grad_norm": 0.11917490512132645, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5389, | |
| "num_tokens": 947841726.0, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.3460964694181998, | |
| "grad_norm": 0.1211070865392685, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5347, | |
| "num_tokens": 949278490.0, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 1.34808552958727, | |
| "grad_norm": 0.12035378068685532, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5334, | |
| "num_tokens": 950642424.0, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 1.3500745897563402, | |
| "grad_norm": 0.11970808357000351, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5304, | |
| "num_tokens": 952042727.0, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 1.3520636499254102, | |
| "grad_norm": 0.12441671639680862, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5429, | |
| "num_tokens": 953444541.0, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 1.3540527100944804, | |
| "grad_norm": 0.11760767549276352, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5433, | |
| "num_tokens": 954885585.0, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.3560417702635505, | |
| "grad_norm": 0.12732824683189392, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5339, | |
| "num_tokens": 956271933.0, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 1.3580308304326205, | |
| "grad_norm": 0.11609245091676712, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5398, | |
| "num_tokens": 957653773.0, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 1.3600198906016907, | |
| "grad_norm": 0.11394883692264557, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5357, | |
| "num_tokens": 959087783.0, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 1.362008950770761, | |
| "grad_norm": 0.13362184166908264, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5404, | |
| "num_tokens": 960495424.0, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 1.3639980109398309, | |
| "grad_norm": 0.11702502518892288, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5325, | |
| "num_tokens": 961901643.0, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 1.365987071108901, | |
| "grad_norm": 0.11658236384391785, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5322, | |
| "num_tokens": 963283706.0, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 1.3679761312779712, | |
| "grad_norm": 0.12017067521810532, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5338, | |
| "num_tokens": 964734481.0, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 1.3699651914470412, | |
| "grad_norm": 0.12005714327096939, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5423, | |
| "num_tokens": 966125592.0, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 1.3719542516161114, | |
| "grad_norm": 0.12015294283628464, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5513, | |
| "num_tokens": 967523933.0, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 1.3739433117851814, | |
| "grad_norm": 0.11920958757400513, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5319, | |
| "num_tokens": 968897975.0, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.3759323719542516, | |
| "grad_norm": 0.12052245438098907, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5333, | |
| "num_tokens": 970267864.0, | |
| "step": 691 | |
| }, | |
| { | |
| "epoch": 1.3779214321233217, | |
| "grad_norm": 0.11884114146232605, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5336, | |
| "num_tokens": 971706591.0, | |
| "step": 692 | |
| }, | |
| { | |
| "epoch": 1.379910492292392, | |
| "grad_norm": 0.11437772214412689, | |
| "learning_rate": 1e-06, | |
| "loss": 0.552, | |
| "num_tokens": 973155693.0, | |
| "step": 693 | |
| }, | |
| { | |
| "epoch": 1.381899552461462, | |
| "grad_norm": 0.12122377008199692, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5444, | |
| "num_tokens": 974547211.0, | |
| "step": 694 | |
| }, | |
| { | |
| "epoch": 1.383888612630532, | |
| "grad_norm": 0.11425941437482834, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5415, | |
| "num_tokens": 976010063.0, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 1.385877672799602, | |
| "grad_norm": 0.11765948683023453, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5284, | |
| "num_tokens": 977411227.0, | |
| "step": 696 | |
| }, | |
| { | |
| "epoch": 1.3878667329686722, | |
| "grad_norm": 0.11854742467403412, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5338, | |
| "num_tokens": 978786923.0, | |
| "step": 697 | |
| }, | |
| { | |
| "epoch": 1.3898557931377424, | |
| "grad_norm": 0.12211066484451294, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5319, | |
| "num_tokens": 980164053.0, | |
| "step": 698 | |
| }, | |
| { | |
| "epoch": 1.3918448533068126, | |
| "grad_norm": 0.1181558147072792, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5233, | |
| "num_tokens": 981562290.0, | |
| "step": 699 | |
| }, | |
| { | |
| "epoch": 1.3938339134758826, | |
| "grad_norm": 0.12071076780557632, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5394, | |
| "num_tokens": 982998912.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.3958229736449528, | |
| "grad_norm": 0.11735861748456955, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5456, | |
| "num_tokens": 984410558.0, | |
| "step": 701 | |
| }, | |
| { | |
| "epoch": 1.3978120338140227, | |
| "grad_norm": 0.11743367463350296, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5396, | |
| "num_tokens": 985836461.0, | |
| "step": 702 | |
| }, | |
| { | |
| "epoch": 1.399801093983093, | |
| "grad_norm": 0.11885622888803482, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5424, | |
| "num_tokens": 987230105.0, | |
| "step": 703 | |
| }, | |
| { | |
| "epoch": 1.4017901541521631, | |
| "grad_norm": 0.11783410608768463, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5352, | |
| "num_tokens": 988630556.0, | |
| "step": 704 | |
| }, | |
| { | |
| "epoch": 1.4037792143212333, | |
| "grad_norm": 0.11775646358728409, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5239, | |
| "num_tokens": 989996092.0, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 1.4057682744903033, | |
| "grad_norm": 0.11850979924201965, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5292, | |
| "num_tokens": 991410237.0, | |
| "step": 706 | |
| }, | |
| { | |
| "epoch": 1.4077573346593735, | |
| "grad_norm": 0.13466332852840424, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5229, | |
| "num_tokens": 992831683.0, | |
| "step": 707 | |
| }, | |
| { | |
| "epoch": 1.4097463948284434, | |
| "grad_norm": 0.1183917373418808, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5306, | |
| "num_tokens": 994220085.0, | |
| "step": 708 | |
| }, | |
| { | |
| "epoch": 1.4117354549975136, | |
| "grad_norm": 0.11719982326030731, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5347, | |
| "num_tokens": 995634427.0, | |
| "step": 709 | |
| }, | |
| { | |
| "epoch": 1.4137245151665838, | |
| "grad_norm": 0.1188247948884964, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5214, | |
| "num_tokens": 996991797.0, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.415713575335654, | |
| "grad_norm": 0.12354787439107895, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5394, | |
| "num_tokens": 998387297.0, | |
| "step": 711 | |
| }, | |
| { | |
| "epoch": 1.417702635504724, | |
| "grad_norm": 0.11752255260944366, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5289, | |
| "num_tokens": 999796509.0, | |
| "step": 712 | |
| }, | |
| { | |
| "epoch": 1.4196916956737942, | |
| "grad_norm": 0.11573730409145355, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5306, | |
| "num_tokens": 1001189936.0, | |
| "step": 713 | |
| }, | |
| { | |
| "epoch": 1.4216807558428641, | |
| "grad_norm": 0.12337598204612732, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5368, | |
| "num_tokens": 1002569771.0, | |
| "step": 714 | |
| }, | |
| { | |
| "epoch": 1.4236698160119343, | |
| "grad_norm": 0.1165643110871315, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5284, | |
| "num_tokens": 1003987924.0, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 1.4256588761810045, | |
| "grad_norm": 0.12802375853061676, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5404, | |
| "num_tokens": 1005393054.0, | |
| "step": 716 | |
| }, | |
| { | |
| "epoch": 1.4276479363500747, | |
| "grad_norm": 0.1228770762681961, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5448, | |
| "num_tokens": 1006837132.0, | |
| "step": 717 | |
| }, | |
| { | |
| "epoch": 1.4296369965191447, | |
| "grad_norm": 0.12197130918502808, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5514, | |
| "num_tokens": 1008250141.0, | |
| "step": 718 | |
| }, | |
| { | |
| "epoch": 1.4316260566882149, | |
| "grad_norm": 0.11660060286521912, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5501, | |
| "num_tokens": 1009672430.0, | |
| "step": 719 | |
| }, | |
| { | |
| "epoch": 1.4336151168572848, | |
| "grad_norm": 0.13146308064460754, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5526, | |
| "num_tokens": 1011105282.0, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.435604177026355, | |
| "grad_norm": 0.12230085581541061, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5388, | |
| "num_tokens": 1012519871.0, | |
| "step": 721 | |
| }, | |
| { | |
| "epoch": 1.4375932371954252, | |
| "grad_norm": 0.11886520683765411, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5512, | |
| "num_tokens": 1013919180.0, | |
| "step": 722 | |
| }, | |
| { | |
| "epoch": 1.4395822973644954, | |
| "grad_norm": 0.12054796516895294, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5418, | |
| "num_tokens": 1015311974.0, | |
| "step": 723 | |
| }, | |
| { | |
| "epoch": 1.4415713575335654, | |
| "grad_norm": 0.1221555769443512, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5315, | |
| "num_tokens": 1016710513.0, | |
| "step": 724 | |
| }, | |
| { | |
| "epoch": 1.4435604177026355, | |
| "grad_norm": 0.12085101753473282, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5267, | |
| "num_tokens": 1018108542.0, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.4455494778717055, | |
| "grad_norm": 0.11824904382228851, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5275, | |
| "num_tokens": 1019503680.0, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 1.4475385380407757, | |
| "grad_norm": 0.11636096239089966, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5326, | |
| "num_tokens": 1020904351.0, | |
| "step": 727 | |
| }, | |
| { | |
| "epoch": 1.449527598209846, | |
| "grad_norm": 0.12087342143058777, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5418, | |
| "num_tokens": 1022300450.0, | |
| "step": 728 | |
| }, | |
| { | |
| "epoch": 1.451516658378916, | |
| "grad_norm": 0.11914920806884766, | |
| "learning_rate": 1e-06, | |
| "loss": 0.547, | |
| "num_tokens": 1023750162.0, | |
| "step": 729 | |
| }, | |
| { | |
| "epoch": 1.453505718547986, | |
| "grad_norm": 0.11899517476558685, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5452, | |
| "num_tokens": 1025196694.0, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.4554947787170562, | |
| "grad_norm": 0.1218375712633133, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5321, | |
| "num_tokens": 1026597406.0, | |
| "step": 731 | |
| }, | |
| { | |
| "epoch": 1.4574838388861262, | |
| "grad_norm": 0.11746956408023834, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5341, | |
| "num_tokens": 1027971754.0, | |
| "step": 732 | |
| }, | |
| { | |
| "epoch": 1.4594728990551964, | |
| "grad_norm": 0.12773922085762024, | |
| "learning_rate": 1e-06, | |
| "loss": 0.547, | |
| "num_tokens": 1029364538.0, | |
| "step": 733 | |
| }, | |
| { | |
| "epoch": 1.4614619592242666, | |
| "grad_norm": 0.1228381097316742, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5403, | |
| "num_tokens": 1030782739.0, | |
| "step": 734 | |
| }, | |
| { | |
| "epoch": 1.4634510193933368, | |
| "grad_norm": 0.11855144798755646, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5345, | |
| "num_tokens": 1032154592.0, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 1.4654400795624067, | |
| "grad_norm": 0.12030474096536636, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5314, | |
| "num_tokens": 1033543123.0, | |
| "step": 736 | |
| }, | |
| { | |
| "epoch": 1.467429139731477, | |
| "grad_norm": 0.12364325672388077, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5422, | |
| "num_tokens": 1034938165.0, | |
| "step": 737 | |
| }, | |
| { | |
| "epoch": 1.469418199900547, | |
| "grad_norm": 0.12098907679319382, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5375, | |
| "num_tokens": 1036314909.0, | |
| "step": 738 | |
| }, | |
| { | |
| "epoch": 1.471407260069617, | |
| "grad_norm": 0.11932458728551865, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5347, | |
| "num_tokens": 1037727924.0, | |
| "step": 739 | |
| }, | |
| { | |
| "epoch": 1.4733963202386873, | |
| "grad_norm": 0.11801986396312714, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5248, | |
| "num_tokens": 1039111581.0, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.4753853804077575, | |
| "grad_norm": 0.11731583625078201, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5345, | |
| "num_tokens": 1040546305.0, | |
| "step": 741 | |
| }, | |
| { | |
| "epoch": 1.4773744405768274, | |
| "grad_norm": 0.11801919341087341, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5336, | |
| "num_tokens": 1041939972.0, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.4793635007458976, | |
| "grad_norm": 0.12367312610149384, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5389, | |
| "num_tokens": 1043334299.0, | |
| "step": 743 | |
| }, | |
| { | |
| "epoch": 1.4813525609149676, | |
| "grad_norm": 0.12621083855628967, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5411, | |
| "num_tokens": 1044732856.0, | |
| "step": 744 | |
| }, | |
| { | |
| "epoch": 1.4833416210840378, | |
| "grad_norm": 0.1213909313082695, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5339, | |
| "num_tokens": 1046119537.0, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 1.485330681253108, | |
| "grad_norm": 0.12064032256603241, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5501, | |
| "num_tokens": 1047526533.0, | |
| "step": 746 | |
| }, | |
| { | |
| "epoch": 1.4873197414221782, | |
| "grad_norm": 0.11744097620248795, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5323, | |
| "num_tokens": 1048909786.0, | |
| "step": 747 | |
| }, | |
| { | |
| "epoch": 1.4893088015912481, | |
| "grad_norm": 0.11546076834201813, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5396, | |
| "num_tokens": 1050351234.0, | |
| "step": 748 | |
| }, | |
| { | |
| "epoch": 1.4912978617603183, | |
| "grad_norm": 0.12047351151704788, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5439, | |
| "num_tokens": 1051747737.0, | |
| "step": 749 | |
| }, | |
| { | |
| "epoch": 1.4932869219293883, | |
| "grad_norm": 0.11532466858625412, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5267, | |
| "num_tokens": 1053183317.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.4952759820984585, | |
| "grad_norm": 0.1169387549161911, | |
| "learning_rate": 1e-06, | |
| "loss": 0.526, | |
| "num_tokens": 1054557285.0, | |
| "step": 751 | |
| }, | |
| { | |
| "epoch": 1.4972650422675287, | |
| "grad_norm": 0.1165088340640068, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5308, | |
| "num_tokens": 1056007559.0, | |
| "step": 752 | |
| }, | |
| { | |
| "epoch": 1.4992541024365988, | |
| "grad_norm": 0.11395064741373062, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5408, | |
| "num_tokens": 1057438934.0, | |
| "step": 753 | |
| }, | |
| { | |
| "epoch": 1.5012431626056688, | |
| "grad_norm": 0.12597793340682983, | |
| "learning_rate": 1e-06, | |
| "loss": 0.54, | |
| "num_tokens": 1058838684.0, | |
| "step": 754 | |
| }, | |
| { | |
| "epoch": 1.5032322227747388, | |
| "grad_norm": 0.12164228409528732, | |
| "learning_rate": 1e-06, | |
| "loss": 0.538, | |
| "num_tokens": 1060278255.0, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 1.505221282943809, | |
| "grad_norm": 0.11811868101358414, | |
| "learning_rate": 1e-06, | |
| "loss": 0.531, | |
| "num_tokens": 1061695693.0, | |
| "step": 756 | |
| }, | |
| { | |
| "epoch": 1.5072103431128792, | |
| "grad_norm": 0.1173863634467125, | |
| "learning_rate": 1e-06, | |
| "loss": 0.529, | |
| "num_tokens": 1063113004.0, | |
| "step": 757 | |
| }, | |
| { | |
| "epoch": 1.5091994032819493, | |
| "grad_norm": 0.1195053681731224, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5467, | |
| "num_tokens": 1064537749.0, | |
| "step": 758 | |
| }, | |
| { | |
| "epoch": 1.5111884634510195, | |
| "grad_norm": 0.12815718352794647, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5202, | |
| "num_tokens": 1065931962.0, | |
| "step": 759 | |
| }, | |
| { | |
| "epoch": 1.5131775236200895, | |
| "grad_norm": 0.12516118586063385, | |
| "learning_rate": 1e-06, | |
| "loss": 0.517, | |
| "num_tokens": 1067328106.0, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.5151665837891595, | |
| "grad_norm": 0.11947084218263626, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5198, | |
| "num_tokens": 1068712446.0, | |
| "step": 761 | |
| }, | |
| { | |
| "epoch": 1.5171556439582297, | |
| "grad_norm": 0.1181473359465599, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5399, | |
| "num_tokens": 1070129632.0, | |
| "step": 762 | |
| }, | |
| { | |
| "epoch": 1.5191447041272998, | |
| "grad_norm": 0.12384405732154846, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5369, | |
| "num_tokens": 1071485504.0, | |
| "step": 763 | |
| }, | |
| { | |
| "epoch": 1.52113376429637, | |
| "grad_norm": 0.12462117522954941, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5313, | |
| "num_tokens": 1072876547.0, | |
| "step": 764 | |
| }, | |
| { | |
| "epoch": 1.5231228244654402, | |
| "grad_norm": 0.11820020526647568, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5282, | |
| "num_tokens": 1074262233.0, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 1.5251118846345102, | |
| "grad_norm": 0.11712965369224548, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5286, | |
| "num_tokens": 1075681885.0, | |
| "step": 766 | |
| }, | |
| { | |
| "epoch": 1.5271009448035802, | |
| "grad_norm": 0.11938031017780304, | |
| "learning_rate": 1e-06, | |
| "loss": 0.524, | |
| "num_tokens": 1077086623.0, | |
| "step": 767 | |
| }, | |
| { | |
| "epoch": 1.5290900049726504, | |
| "grad_norm": 0.11355356127023697, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5357, | |
| "num_tokens": 1078502366.0, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 1.5310790651417205, | |
| "grad_norm": 0.12431412935256958, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5317, | |
| "num_tokens": 1079851354.0, | |
| "step": 769 | |
| }, | |
| { | |
| "epoch": 1.5330681253107907, | |
| "grad_norm": 0.12124991416931152, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5452, | |
| "num_tokens": 1081246250.0, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.535057185479861, | |
| "grad_norm": 0.11724445968866348, | |
| "learning_rate": 1e-06, | |
| "loss": 0.53, | |
| "num_tokens": 1082635196.0, | |
| "step": 771 | |
| }, | |
| { | |
| "epoch": 1.5370462456489309, | |
| "grad_norm": 0.12210683524608612, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5291, | |
| "num_tokens": 1084011995.0, | |
| "step": 772 | |
| }, | |
| { | |
| "epoch": 1.5390353058180009, | |
| "grad_norm": 0.11981118470430374, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5438, | |
| "num_tokens": 1085412000.0, | |
| "step": 773 | |
| }, | |
| { | |
| "epoch": 1.541024365987071, | |
| "grad_norm": 0.12824386358261108, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5393, | |
| "num_tokens": 1086812231.0, | |
| "step": 774 | |
| }, | |
| { | |
| "epoch": 1.5430134261561412, | |
| "grad_norm": 0.1165238469839096, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5468, | |
| "num_tokens": 1088217759.0, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.5450024863252114, | |
| "grad_norm": 0.11545863002538681, | |
| "learning_rate": 1e-06, | |
| "loss": 0.522, | |
| "num_tokens": 1089622076.0, | |
| "step": 776 | |
| }, | |
| { | |
| "epoch": 1.5469915464942816, | |
| "grad_norm": 0.11776294559240341, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5439, | |
| "num_tokens": 1091049994.0, | |
| "step": 777 | |
| }, | |
| { | |
| "epoch": 1.5489806066633516, | |
| "grad_norm": 0.11632154881954193, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5078, | |
| "num_tokens": 1092400841.0, | |
| "step": 778 | |
| }, | |
| { | |
| "epoch": 1.5509696668324215, | |
| "grad_norm": 0.1186751276254654, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5372, | |
| "num_tokens": 1093805680.0, | |
| "step": 779 | |
| }, | |
| { | |
| "epoch": 1.5529587270014917, | |
| "grad_norm": 0.11867798119783401, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5334, | |
| "num_tokens": 1095211670.0, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.554947787170562, | |
| "grad_norm": 0.11884911358356476, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5355, | |
| "num_tokens": 1096608638.0, | |
| "step": 781 | |
| }, | |
| { | |
| "epoch": 1.5569368473396321, | |
| "grad_norm": 0.12014549970626831, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5485, | |
| "num_tokens": 1098024501.0, | |
| "step": 782 | |
| }, | |
| { | |
| "epoch": 1.5589259075087023, | |
| "grad_norm": 0.115864098072052, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5372, | |
| "num_tokens": 1099418318.0, | |
| "step": 783 | |
| }, | |
| { | |
| "epoch": 1.5609149676777723, | |
| "grad_norm": 0.12087985128164291, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5331, | |
| "num_tokens": 1100818699.0, | |
| "step": 784 | |
| }, | |
| { | |
| "epoch": 1.5629040278468422, | |
| "grad_norm": 0.11292678117752075, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5345, | |
| "num_tokens": 1102255753.0, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 1.5648930880159124, | |
| "grad_norm": 0.11706893891096115, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5211, | |
| "num_tokens": 1103629573.0, | |
| "step": 786 | |
| }, | |
| { | |
| "epoch": 1.5668821481849826, | |
| "grad_norm": 0.11923506110906601, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5309, | |
| "num_tokens": 1104985171.0, | |
| "step": 787 | |
| }, | |
| { | |
| "epoch": 1.5688712083540528, | |
| "grad_norm": 0.12054122984409332, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5395, | |
| "num_tokens": 1106405658.0, | |
| "step": 788 | |
| }, | |
| { | |
| "epoch": 1.5708602685231228, | |
| "grad_norm": 0.11507654935121536, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5414, | |
| "num_tokens": 1107830805.0, | |
| "step": 789 | |
| }, | |
| { | |
| "epoch": 1.572849328692193, | |
| "grad_norm": 0.11983779072761536, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5314, | |
| "num_tokens": 1109163312.0, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.574838388861263, | |
| "grad_norm": 0.11855509877204895, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5279, | |
| "num_tokens": 1110582152.0, | |
| "step": 791 | |
| }, | |
| { | |
| "epoch": 1.5768274490303331, | |
| "grad_norm": 0.11437905579805374, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5335, | |
| "num_tokens": 1112007797.0, | |
| "step": 792 | |
| }, | |
| { | |
| "epoch": 1.5788165091994033, | |
| "grad_norm": 0.11613345891237259, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5364, | |
| "num_tokens": 1113366919.0, | |
| "step": 793 | |
| }, | |
| { | |
| "epoch": 1.5808055693684735, | |
| "grad_norm": 0.1167130321264267, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5295, | |
| "num_tokens": 1114808855.0, | |
| "step": 794 | |
| }, | |
| { | |
| "epoch": 1.5827946295375435, | |
| "grad_norm": 0.11898983269929886, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5426, | |
| "num_tokens": 1116215942.0, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 1.5847836897066137, | |
| "grad_norm": 0.11814821511507034, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5249, | |
| "num_tokens": 1117635329.0, | |
| "step": 796 | |
| }, | |
| { | |
| "epoch": 1.5867727498756836, | |
| "grad_norm": 0.11845888942480087, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5347, | |
| "num_tokens": 1119076550.0, | |
| "step": 797 | |
| }, | |
| { | |
| "epoch": 1.5887618100447538, | |
| "grad_norm": 0.11748501658439636, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5376, | |
| "num_tokens": 1120476445.0, | |
| "step": 798 | |
| }, | |
| { | |
| "epoch": 1.590750870213824, | |
| "grad_norm": 0.11534599214792252, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5204, | |
| "num_tokens": 1121859435.0, | |
| "step": 799 | |
| }, | |
| { | |
| "epoch": 1.5927399303828942, | |
| "grad_norm": 0.11943424493074417, | |
| "learning_rate": 1e-06, | |
| "loss": 0.526, | |
| "num_tokens": 1123242470.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.5947289905519642, | |
| "grad_norm": 0.12252327799797058, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5278, | |
| "num_tokens": 1124644409.0, | |
| "step": 801 | |
| }, | |
| { | |
| "epoch": 1.5967180507210343, | |
| "grad_norm": 0.11914427578449249, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5308, | |
| "num_tokens": 1126067895.0, | |
| "step": 802 | |
| }, | |
| { | |
| "epoch": 1.5987071108901043, | |
| "grad_norm": 0.12473994493484497, | |
| "learning_rate": 1e-06, | |
| "loss": 0.53, | |
| "num_tokens": 1127456815.0, | |
| "step": 803 | |
| }, | |
| { | |
| "epoch": 1.6006961710591745, | |
| "grad_norm": 0.11551981419324875, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5255, | |
| "num_tokens": 1128878563.0, | |
| "step": 804 | |
| }, | |
| { | |
| "epoch": 1.6026852312282447, | |
| "grad_norm": 0.11678687483072281, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5325, | |
| "num_tokens": 1130317088.0, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 1.6046742913973149, | |
| "grad_norm": 0.1221092939376831, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5404, | |
| "num_tokens": 1131770565.0, | |
| "step": 806 | |
| }, | |
| { | |
| "epoch": 1.6066633515663848, | |
| "grad_norm": 0.11658436805009842, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5345, | |
| "num_tokens": 1133178633.0, | |
| "step": 807 | |
| }, | |
| { | |
| "epoch": 1.608652411735455, | |
| "grad_norm": 0.11696770042181015, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5275, | |
| "num_tokens": 1134574443.0, | |
| "step": 808 | |
| }, | |
| { | |
| "epoch": 1.610641471904525, | |
| "grad_norm": 0.11893412470817566, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5349, | |
| "num_tokens": 1135948785.0, | |
| "step": 809 | |
| }, | |
| { | |
| "epoch": 1.6126305320735952, | |
| "grad_norm": 0.12174445390701294, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5292, | |
| "num_tokens": 1137342427.0, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.6146195922426654, | |
| "grad_norm": 0.12794137001037598, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5418, | |
| "num_tokens": 1138767131.0, | |
| "step": 811 | |
| }, | |
| { | |
| "epoch": 1.6166086524117356, | |
| "grad_norm": 0.11655872315168381, | |
| "learning_rate": 1e-06, | |
| "loss": 0.525, | |
| "num_tokens": 1140181396.0, | |
| "step": 812 | |
| }, | |
| { | |
| "epoch": 1.6185977125808055, | |
| "grad_norm": 0.11739625781774521, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5265, | |
| "num_tokens": 1141585956.0, | |
| "step": 813 | |
| }, | |
| { | |
| "epoch": 1.6205867727498757, | |
| "grad_norm": 0.11966431885957718, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5439, | |
| "num_tokens": 1142972552.0, | |
| "step": 814 | |
| }, | |
| { | |
| "epoch": 1.6225758329189457, | |
| "grad_norm": 0.1153174564242363, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5269, | |
| "num_tokens": 1144360432.0, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 1.6245648930880159, | |
| "grad_norm": 0.11621485650539398, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5274, | |
| "num_tokens": 1145763084.0, | |
| "step": 816 | |
| }, | |
| { | |
| "epoch": 1.626553953257086, | |
| "grad_norm": 0.12078936398029327, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5408, | |
| "num_tokens": 1147181036.0, | |
| "step": 817 | |
| }, | |
| { | |
| "epoch": 1.6285430134261563, | |
| "grad_norm": 0.11993258446455002, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5371, | |
| "num_tokens": 1148568031.0, | |
| "step": 818 | |
| }, | |
| { | |
| "epoch": 1.6305320735952262, | |
| "grad_norm": 0.12087948620319366, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5239, | |
| "num_tokens": 1149971169.0, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 1.6325211337642964, | |
| "grad_norm": 0.12193674594163895, | |
| "learning_rate": 1e-06, | |
| "loss": 0.532, | |
| "num_tokens": 1151375502.0, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.6345101939333664, | |
| "grad_norm": 0.11705011874437332, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5374, | |
| "num_tokens": 1152788267.0, | |
| "step": 821 | |
| }, | |
| { | |
| "epoch": 1.6364992541024366, | |
| "grad_norm": 0.11758929491043091, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5321, | |
| "num_tokens": 1154203978.0, | |
| "step": 822 | |
| }, | |
| { | |
| "epoch": 1.6384883142715068, | |
| "grad_norm": 0.11853373050689697, | |
| "learning_rate": 1e-06, | |
| "loss": 0.537, | |
| "num_tokens": 1155628446.0, | |
| "step": 823 | |
| }, | |
| { | |
| "epoch": 1.640477374440577, | |
| "grad_norm": 0.11777627468109131, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5272, | |
| "num_tokens": 1157039859.0, | |
| "step": 824 | |
| }, | |
| { | |
| "epoch": 1.642466434609647, | |
| "grad_norm": 0.12647390365600586, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5326, | |
| "num_tokens": 1158461070.0, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.644455494778717, | |
| "grad_norm": 0.11644790321588516, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5176, | |
| "num_tokens": 1159857529.0, | |
| "step": 826 | |
| }, | |
| { | |
| "epoch": 1.646444554947787, | |
| "grad_norm": 0.11872653663158417, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5255, | |
| "num_tokens": 1161253129.0, | |
| "step": 827 | |
| }, | |
| { | |
| "epoch": 1.6484336151168573, | |
| "grad_norm": 0.1227663904428482, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5217, | |
| "num_tokens": 1162695252.0, | |
| "step": 828 | |
| }, | |
| { | |
| "epoch": 1.6504226752859275, | |
| "grad_norm": 0.12286870181560516, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5252, | |
| "num_tokens": 1164073116.0, | |
| "step": 829 | |
| }, | |
| { | |
| "epoch": 1.6524117354549976, | |
| "grad_norm": 0.11983044445514679, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5277, | |
| "num_tokens": 1165473708.0, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.6544007956240676, | |
| "grad_norm": 0.11557100713253021, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5383, | |
| "num_tokens": 1166861777.0, | |
| "step": 831 | |
| }, | |
| { | |
| "epoch": 1.6563898557931376, | |
| "grad_norm": 0.12076076120138168, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5438, | |
| "num_tokens": 1168281147.0, | |
| "step": 832 | |
| }, | |
| { | |
| "epoch": 1.6583789159622078, | |
| "grad_norm": 0.11843711882829666, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5263, | |
| "num_tokens": 1169644876.0, | |
| "step": 833 | |
| }, | |
| { | |
| "epoch": 1.660367976131278, | |
| "grad_norm": 0.11820250749588013, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5311, | |
| "num_tokens": 1171071751.0, | |
| "step": 834 | |
| }, | |
| { | |
| "epoch": 1.6623570363003481, | |
| "grad_norm": 0.11654365807771683, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5307, | |
| "num_tokens": 1172492906.0, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 1.6643460964694183, | |
| "grad_norm": 0.12128648906946182, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5255, | |
| "num_tokens": 1173918229.0, | |
| "step": 836 | |
| }, | |
| { | |
| "epoch": 1.6663351566384883, | |
| "grad_norm": 0.1194423958659172, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5265, | |
| "num_tokens": 1175294915.0, | |
| "step": 837 | |
| }, | |
| { | |
| "epoch": 1.6683242168075583, | |
| "grad_norm": 0.11887000501155853, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5413, | |
| "num_tokens": 1176717797.0, | |
| "step": 838 | |
| }, | |
| { | |
| "epoch": 1.6703132769766285, | |
| "grad_norm": 0.1198115423321724, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5251, | |
| "num_tokens": 1178114954.0, | |
| "step": 839 | |
| }, | |
| { | |
| "epoch": 1.6723023371456986, | |
| "grad_norm": 0.12133830040693283, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5311, | |
| "num_tokens": 1179495561.0, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.6742913973147688, | |
| "grad_norm": 0.12414910644292831, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5354, | |
| "num_tokens": 1180900637.0, | |
| "step": 841 | |
| }, | |
| { | |
| "epoch": 1.676280457483839, | |
| "grad_norm": 0.11752628535032272, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5302, | |
| "num_tokens": 1182291450.0, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 1.678269517652909, | |
| "grad_norm": 0.11616742610931396, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5253, | |
| "num_tokens": 1183704362.0, | |
| "step": 843 | |
| }, | |
| { | |
| "epoch": 1.680258577821979, | |
| "grad_norm": 0.11694184690713882, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5365, | |
| "num_tokens": 1185133642.0, | |
| "step": 844 | |
| }, | |
| { | |
| "epoch": 1.6822476379910491, | |
| "grad_norm": 0.17413176596164703, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5194, | |
| "num_tokens": 1186533182.0, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 1.6842366981601193, | |
| "grad_norm": 0.11673513799905777, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5238, | |
| "num_tokens": 1187935618.0, | |
| "step": 846 | |
| }, | |
| { | |
| "epoch": 1.6862257583291895, | |
| "grad_norm": 0.11706209182739258, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5274, | |
| "num_tokens": 1189355590.0, | |
| "step": 847 | |
| }, | |
| { | |
| "epoch": 1.6882148184982597, | |
| "grad_norm": 0.12071144580841064, | |
| "learning_rate": 1e-06, | |
| "loss": 0.522, | |
| "num_tokens": 1190723818.0, | |
| "step": 848 | |
| }, | |
| { | |
| "epoch": 1.6902038786673297, | |
| "grad_norm": 0.11622277647256851, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5284, | |
| "num_tokens": 1192146190.0, | |
| "step": 849 | |
| }, | |
| { | |
| "epoch": 1.6921929388363997, | |
| "grad_norm": 0.11853674054145813, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5294, | |
| "num_tokens": 1193541029.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.6941819990054698, | |
| "grad_norm": 0.11840229481458664, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5146, | |
| "num_tokens": 1194973011.0, | |
| "step": 851 | |
| }, | |
| { | |
| "epoch": 1.69617105917454, | |
| "grad_norm": 0.11261092871427536, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5263, | |
| "num_tokens": 1196379352.0, | |
| "step": 852 | |
| }, | |
| { | |
| "epoch": 1.6981601193436102, | |
| "grad_norm": 0.11384103447198868, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5203, | |
| "num_tokens": 1197758489.0, | |
| "step": 853 | |
| }, | |
| { | |
| "epoch": 1.7001491795126804, | |
| "grad_norm": 0.12079748511314392, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5315, | |
| "num_tokens": 1199163292.0, | |
| "step": 854 | |
| }, | |
| { | |
| "epoch": 1.7021382396817504, | |
| "grad_norm": 0.11787824332714081, | |
| "learning_rate": 1e-06, | |
| "loss": 0.516, | |
| "num_tokens": 1200558367.0, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 1.7041272998508203, | |
| "grad_norm": 0.11851559579372406, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5208, | |
| "num_tokens": 1201948798.0, | |
| "step": 856 | |
| }, | |
| { | |
| "epoch": 1.7061163600198905, | |
| "grad_norm": 0.11743155121803284, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5213, | |
| "num_tokens": 1203317747.0, | |
| "step": 857 | |
| }, | |
| { | |
| "epoch": 1.7081054201889607, | |
| "grad_norm": 0.11653459817171097, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5292, | |
| "num_tokens": 1204751593.0, | |
| "step": 858 | |
| }, | |
| { | |
| "epoch": 1.710094480358031, | |
| "grad_norm": 0.11759098619222641, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5272, | |
| "num_tokens": 1206123965.0, | |
| "step": 859 | |
| }, | |
| { | |
| "epoch": 1.712083540527101, | |
| "grad_norm": 0.11782211810350418, | |
| "learning_rate": 1e-06, | |
| "loss": 0.525, | |
| "num_tokens": 1207504339.0, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.714072600696171, | |
| "grad_norm": 0.11687052994966507, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5321, | |
| "num_tokens": 1208933928.0, | |
| "step": 861 | |
| }, | |
| { | |
| "epoch": 1.716061660865241, | |
| "grad_norm": 0.11712006479501724, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5311, | |
| "num_tokens": 1210329028.0, | |
| "step": 862 | |
| }, | |
| { | |
| "epoch": 1.7180507210343112, | |
| "grad_norm": 0.11513438820838928, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5298, | |
| "num_tokens": 1211754806.0, | |
| "step": 863 | |
| }, | |
| { | |
| "epoch": 1.7200397812033814, | |
| "grad_norm": 0.11873895674943924, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5211, | |
| "num_tokens": 1213112342.0, | |
| "step": 864 | |
| }, | |
| { | |
| "epoch": 1.7220288413724516, | |
| "grad_norm": 0.11382027715444565, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5227, | |
| "num_tokens": 1214524561.0, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 1.7240179015415218, | |
| "grad_norm": 0.11664129793643951, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5255, | |
| "num_tokens": 1215928909.0, | |
| "step": 866 | |
| }, | |
| { | |
| "epoch": 1.7260069617105918, | |
| "grad_norm": 0.11998264491558075, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5195, | |
| "num_tokens": 1217291462.0, | |
| "step": 867 | |
| }, | |
| { | |
| "epoch": 1.7279960218796617, | |
| "grad_norm": 0.11994168907403946, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5359, | |
| "num_tokens": 1218671826.0, | |
| "step": 868 | |
| }, | |
| { | |
| "epoch": 1.729985082048732, | |
| "grad_norm": 0.1149284616112709, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5305, | |
| "num_tokens": 1220083806.0, | |
| "step": 869 | |
| }, | |
| { | |
| "epoch": 1.731974142217802, | |
| "grad_norm": 0.11925540119409561, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5365, | |
| "num_tokens": 1221495745.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.7339632023868723, | |
| "grad_norm": 0.11679849773645401, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5227, | |
| "num_tokens": 1222941705.0, | |
| "step": 871 | |
| }, | |
| { | |
| "epoch": 1.7359522625559425, | |
| "grad_norm": 0.1146961897611618, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5359, | |
| "num_tokens": 1224378188.0, | |
| "step": 872 | |
| }, | |
| { | |
| "epoch": 1.7379413227250124, | |
| "grad_norm": 0.11416348069906235, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5365, | |
| "num_tokens": 1225807121.0, | |
| "step": 873 | |
| }, | |
| { | |
| "epoch": 1.7399303828940824, | |
| "grad_norm": 0.12208687514066696, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5266, | |
| "num_tokens": 1227205689.0, | |
| "step": 874 | |
| }, | |
| { | |
| "epoch": 1.7419194430631526, | |
| "grad_norm": 0.11736491322517395, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5315, | |
| "num_tokens": 1228632893.0, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.7439085032322228, | |
| "grad_norm": 0.11757368594408035, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5324, | |
| "num_tokens": 1230044870.0, | |
| "step": 876 | |
| }, | |
| { | |
| "epoch": 1.745897563401293, | |
| "grad_norm": 0.1128118559718132, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5284, | |
| "num_tokens": 1231458999.0, | |
| "step": 877 | |
| }, | |
| { | |
| "epoch": 1.747886623570363, | |
| "grad_norm": 0.11604047566652298, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5246, | |
| "num_tokens": 1232886812.0, | |
| "step": 878 | |
| }, | |
| { | |
| "epoch": 1.7498756837394331, | |
| "grad_norm": 0.11722288280725479, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5205, | |
| "num_tokens": 1234263247.0, | |
| "step": 879 | |
| }, | |
| { | |
| "epoch": 1.751864743908503, | |
| "grad_norm": 0.11631619185209274, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5271, | |
| "num_tokens": 1235684949.0, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.7538538040775733, | |
| "grad_norm": 0.11910022050142288, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5174, | |
| "num_tokens": 1237092421.0, | |
| "step": 881 | |
| }, | |
| { | |
| "epoch": 1.7558428642466435, | |
| "grad_norm": 0.37577947974205017, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5305, | |
| "num_tokens": 1238488077.0, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 1.7578319244157137, | |
| "grad_norm": 0.11624370515346527, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5257, | |
| "num_tokens": 1239858200.0, | |
| "step": 883 | |
| }, | |
| { | |
| "epoch": 1.7598209845847836, | |
| "grad_norm": 0.11999819427728653, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5296, | |
| "num_tokens": 1241270454.0, | |
| "step": 884 | |
| }, | |
| { | |
| "epoch": 1.7618100447538538, | |
| "grad_norm": 0.11568225175142288, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5204, | |
| "num_tokens": 1242689791.0, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 1.7637991049229238, | |
| "grad_norm": 0.1157977357506752, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5273, | |
| "num_tokens": 1244076494.0, | |
| "step": 886 | |
| }, | |
| { | |
| "epoch": 1.765788165091994, | |
| "grad_norm": 0.12077594548463821, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5352, | |
| "num_tokens": 1245450599.0, | |
| "step": 887 | |
| }, | |
| { | |
| "epoch": 1.7677772252610642, | |
| "grad_norm": 0.11732426285743713, | |
| "learning_rate": 1e-06, | |
| "loss": 0.518, | |
| "num_tokens": 1246866283.0, | |
| "step": 888 | |
| }, | |
| { | |
| "epoch": 1.7697662854301344, | |
| "grad_norm": 0.11936353892087936, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5211, | |
| "num_tokens": 1248255137.0, | |
| "step": 889 | |
| }, | |
| { | |
| "epoch": 1.7717553455992043, | |
| "grad_norm": 0.11886170506477356, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5257, | |
| "num_tokens": 1249659564.0, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.7737444057682745, | |
| "grad_norm": 0.11828939616680145, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5405, | |
| "num_tokens": 1251058792.0, | |
| "step": 891 | |
| }, | |
| { | |
| "epoch": 1.7757334659373445, | |
| "grad_norm": 0.11856890469789505, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5306, | |
| "num_tokens": 1252485161.0, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.7777225261064147, | |
| "grad_norm": 0.11919775605201721, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5263, | |
| "num_tokens": 1253864701.0, | |
| "step": 893 | |
| }, | |
| { | |
| "epoch": 1.7797115862754849, | |
| "grad_norm": 0.11722150444984436, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5319, | |
| "num_tokens": 1255315675.0, | |
| "step": 894 | |
| }, | |
| { | |
| "epoch": 1.781700646444555, | |
| "grad_norm": 0.11736007779836655, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5212, | |
| "num_tokens": 1256727422.0, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 1.783689706613625, | |
| "grad_norm": 0.11714823544025421, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5154, | |
| "num_tokens": 1258118035.0, | |
| "step": 896 | |
| }, | |
| { | |
| "epoch": 1.7856787667826952, | |
| "grad_norm": 0.11619334667921066, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5354, | |
| "num_tokens": 1259569135.0, | |
| "step": 897 | |
| }, | |
| { | |
| "epoch": 1.7876678269517652, | |
| "grad_norm": 0.12298526614904404, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5241, | |
| "num_tokens": 1260962943.0, | |
| "step": 898 | |
| }, | |
| { | |
| "epoch": 1.7896568871208354, | |
| "grad_norm": 0.11889567226171494, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5156, | |
| "num_tokens": 1262364349.0, | |
| "step": 899 | |
| }, | |
| { | |
| "epoch": 1.7916459472899056, | |
| "grad_norm": 0.11857149749994278, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5338, | |
| "num_tokens": 1263806716.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.7936350074589757, | |
| "grad_norm": 0.11797276884317398, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5207, | |
| "num_tokens": 1265169481.0, | |
| "step": 901 | |
| }, | |
| { | |
| "epoch": 1.7956240676280457, | |
| "grad_norm": 0.11892815679311752, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5293, | |
| "num_tokens": 1266576993.0, | |
| "step": 902 | |
| }, | |
| { | |
| "epoch": 1.797613127797116, | |
| "grad_norm": 0.12395410984754562, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5366, | |
| "num_tokens": 1267993556.0, | |
| "step": 903 | |
| }, | |
| { | |
| "epoch": 1.7996021879661859, | |
| "grad_norm": 0.11953330039978027, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5368, | |
| "num_tokens": 1269389274.0, | |
| "step": 904 | |
| }, | |
| { | |
| "epoch": 1.801591248135256, | |
| "grad_norm": 0.12015491724014282, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5221, | |
| "num_tokens": 1270797211.0, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 1.8035803083043263, | |
| "grad_norm": 0.12275838851928711, | |
| "learning_rate": 1e-06, | |
| "loss": 0.516, | |
| "num_tokens": 1272228896.0, | |
| "step": 906 | |
| }, | |
| { | |
| "epoch": 1.8055693684733964, | |
| "grad_norm": 0.11699802428483963, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5261, | |
| "num_tokens": 1273621370.0, | |
| "step": 907 | |
| }, | |
| { | |
| "epoch": 1.8075584286424664, | |
| "grad_norm": 0.1161780133843422, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5274, | |
| "num_tokens": 1275017133.0, | |
| "step": 908 | |
| }, | |
| { | |
| "epoch": 1.8095474888115366, | |
| "grad_norm": 0.11965856701135635, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5337, | |
| "num_tokens": 1276448660.0, | |
| "step": 909 | |
| }, | |
| { | |
| "epoch": 1.8115365489806066, | |
| "grad_norm": 0.12168006598949432, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5175, | |
| "num_tokens": 1277862125.0, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.8135256091496768, | |
| "grad_norm": 0.11938372254371643, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5216, | |
| "num_tokens": 1279248819.0, | |
| "step": 911 | |
| }, | |
| { | |
| "epoch": 1.815514669318747, | |
| "grad_norm": 0.11735393106937408, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5233, | |
| "num_tokens": 1280646759.0, | |
| "step": 912 | |
| }, | |
| { | |
| "epoch": 1.8175037294878171, | |
| "grad_norm": 0.11623270809650421, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5233, | |
| "num_tokens": 1282057498.0, | |
| "step": 913 | |
| }, | |
| { | |
| "epoch": 1.819492789656887, | |
| "grad_norm": 0.11557810753583908, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5153, | |
| "num_tokens": 1283431495.0, | |
| "step": 914 | |
| }, | |
| { | |
| "epoch": 1.8214818498259573, | |
| "grad_norm": 0.1188741996884346, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5046, | |
| "num_tokens": 1284807817.0, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 1.8234709099950273, | |
| "grad_norm": 0.12421073764562607, | |
| "learning_rate": 1e-06, | |
| "loss": 0.529, | |
| "num_tokens": 1286222594.0, | |
| "step": 916 | |
| }, | |
| { | |
| "epoch": 1.8254599701640974, | |
| "grad_norm": 0.11472687870264053, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5189, | |
| "num_tokens": 1287660647.0, | |
| "step": 917 | |
| }, | |
| { | |
| "epoch": 1.8274490303331676, | |
| "grad_norm": 0.12024683505296707, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5327, | |
| "num_tokens": 1289025502.0, | |
| "step": 918 | |
| }, | |
| { | |
| "epoch": 1.8294380905022378, | |
| "grad_norm": 0.11754554510116577, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5266, | |
| "num_tokens": 1290429204.0, | |
| "step": 919 | |
| }, | |
| { | |
| "epoch": 1.8314271506713078, | |
| "grad_norm": 0.11749599874019623, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5346, | |
| "num_tokens": 1291835199.0, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.8334162108403778, | |
| "grad_norm": 0.12334319949150085, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5286, | |
| "num_tokens": 1293232429.0, | |
| "step": 921 | |
| }, | |
| { | |
| "epoch": 1.835405271009448, | |
| "grad_norm": 0.1195298433303833, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5342, | |
| "num_tokens": 1294636234.0, | |
| "step": 922 | |
| }, | |
| { | |
| "epoch": 1.8373943311785181, | |
| "grad_norm": 0.12667156755924225, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5329, | |
| "num_tokens": 1296037959.0, | |
| "step": 923 | |
| }, | |
| { | |
| "epoch": 1.8393833913475883, | |
| "grad_norm": 0.11679953336715698, | |
| "learning_rate": 1e-06, | |
| "loss": 0.521, | |
| "num_tokens": 1297473035.0, | |
| "step": 924 | |
| }, | |
| { | |
| "epoch": 1.8413724515166585, | |
| "grad_norm": 0.12583783268928528, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5276, | |
| "num_tokens": 1298890907.0, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.8433615116857285, | |
| "grad_norm": 0.11728479713201523, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5395, | |
| "num_tokens": 1300299527.0, | |
| "step": 926 | |
| }, | |
| { | |
| "epoch": 1.8453505718547984, | |
| "grad_norm": 0.12031624466180801, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5223, | |
| "num_tokens": 1301707639.0, | |
| "step": 927 | |
| }, | |
| { | |
| "epoch": 1.8473396320238686, | |
| "grad_norm": 0.11883748322725296, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5314, | |
| "num_tokens": 1303102646.0, | |
| "step": 928 | |
| }, | |
| { | |
| "epoch": 1.8493286921929388, | |
| "grad_norm": 0.11844377219676971, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5239, | |
| "num_tokens": 1304494162.0, | |
| "step": 929 | |
| }, | |
| { | |
| "epoch": 1.851317752362009, | |
| "grad_norm": 0.11900133639574051, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5417, | |
| "num_tokens": 1305913481.0, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.8533068125310792, | |
| "grad_norm": 0.11836715042591095, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5233, | |
| "num_tokens": 1307345252.0, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 1.8552958727001492, | |
| "grad_norm": 0.11962825059890747, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5189, | |
| "num_tokens": 1308721678.0, | |
| "step": 932 | |
| }, | |
| { | |
| "epoch": 1.8572849328692191, | |
| "grad_norm": 0.11557050049304962, | |
| "learning_rate": 1e-06, | |
| "loss": 0.534, | |
| "num_tokens": 1310185671.0, | |
| "step": 933 | |
| }, | |
| { | |
| "epoch": 1.8592739930382893, | |
| "grad_norm": 0.11940222978591919, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5201, | |
| "num_tokens": 1311578400.0, | |
| "step": 934 | |
| }, | |
| { | |
| "epoch": 1.8612630532073595, | |
| "grad_norm": 0.11809618771076202, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5302, | |
| "num_tokens": 1313012059.0, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 1.8632521133764297, | |
| "grad_norm": 0.11569247394800186, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5261, | |
| "num_tokens": 1314429380.0, | |
| "step": 936 | |
| }, | |
| { | |
| "epoch": 1.8652411735455, | |
| "grad_norm": 0.11870964616537094, | |
| "learning_rate": 1e-06, | |
| "loss": 0.531, | |
| "num_tokens": 1315859666.0, | |
| "step": 937 | |
| }, | |
| { | |
| "epoch": 1.8672302337145699, | |
| "grad_norm": 0.11804142594337463, | |
| "learning_rate": 1e-06, | |
| "loss": 0.524, | |
| "num_tokens": 1317284109.0, | |
| "step": 938 | |
| }, | |
| { | |
| "epoch": 1.8692192938836398, | |
| "grad_norm": 0.12137165665626526, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5172, | |
| "num_tokens": 1318645618.0, | |
| "step": 939 | |
| }, | |
| { | |
| "epoch": 1.87120835405271, | |
| "grad_norm": 0.1159721091389656, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5197, | |
| "num_tokens": 1320005960.0, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.8731974142217802, | |
| "grad_norm": 0.1175520122051239, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5141, | |
| "num_tokens": 1321343874.0, | |
| "step": 941 | |
| }, | |
| { | |
| "epoch": 1.8751864743908504, | |
| "grad_norm": 0.12161525338888168, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5283, | |
| "num_tokens": 1322745862.0, | |
| "step": 942 | |
| }, | |
| { | |
| "epoch": 1.8771755345599206, | |
| "grad_norm": 0.11651836335659027, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5166, | |
| "num_tokens": 1324173893.0, | |
| "step": 943 | |
| }, | |
| { | |
| "epoch": 1.8791645947289906, | |
| "grad_norm": 0.11678607016801834, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5157, | |
| "num_tokens": 1325528425.0, | |
| "step": 944 | |
| }, | |
| { | |
| "epoch": 1.8811536548980605, | |
| "grad_norm": 0.11988682299852371, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5241, | |
| "num_tokens": 1326934542.0, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 1.8831427150671307, | |
| "grad_norm": 0.1179330125451088, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5278, | |
| "num_tokens": 1328329326.0, | |
| "step": 946 | |
| }, | |
| { | |
| "epoch": 1.885131775236201, | |
| "grad_norm": 0.12124811857938766, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5121, | |
| "num_tokens": 1329690107.0, | |
| "step": 947 | |
| }, | |
| { | |
| "epoch": 1.887120835405271, | |
| "grad_norm": 0.11679516732692719, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5123, | |
| "num_tokens": 1331099443.0, | |
| "step": 948 | |
| }, | |
| { | |
| "epoch": 1.8891098955743413, | |
| "grad_norm": 0.12094994634389877, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5294, | |
| "num_tokens": 1332514974.0, | |
| "step": 949 | |
| }, | |
| { | |
| "epoch": 1.8910989557434112, | |
| "grad_norm": 0.11741626262664795, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5221, | |
| "num_tokens": 1333908900.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.8930880159124812, | |
| "grad_norm": 0.11530331522226334, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5253, | |
| "num_tokens": 1335304709.0, | |
| "step": 951 | |
| }, | |
| { | |
| "epoch": 1.8950770760815514, | |
| "grad_norm": 0.1158575564622879, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5285, | |
| "num_tokens": 1336705867.0, | |
| "step": 952 | |
| }, | |
| { | |
| "epoch": 1.8970661362506216, | |
| "grad_norm": 0.12077206373214722, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5282, | |
| "num_tokens": 1338134157.0, | |
| "step": 953 | |
| }, | |
| { | |
| "epoch": 1.8990551964196918, | |
| "grad_norm": 0.11674216389656067, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5256, | |
| "num_tokens": 1339570572.0, | |
| "step": 954 | |
| }, | |
| { | |
| "epoch": 1.901044256588762, | |
| "grad_norm": 0.11999034881591797, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5217, | |
| "num_tokens": 1340963763.0, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.903033316757832, | |
| "grad_norm": 0.1159081682562828, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5193, | |
| "num_tokens": 1342343659.0, | |
| "step": 956 | |
| }, | |
| { | |
| "epoch": 1.905022376926902, | |
| "grad_norm": 0.12077668309211731, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5258, | |
| "num_tokens": 1343740689.0, | |
| "step": 957 | |
| }, | |
| { | |
| "epoch": 1.907011437095972, | |
| "grad_norm": 0.1166224554181099, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5292, | |
| "num_tokens": 1345132334.0, | |
| "step": 958 | |
| }, | |
| { | |
| "epoch": 1.9090004972650423, | |
| "grad_norm": 0.1241859421133995, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5148, | |
| "num_tokens": 1346524317.0, | |
| "step": 959 | |
| }, | |
| { | |
| "epoch": 1.9109895574341125, | |
| "grad_norm": 0.11995749920606613, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5299, | |
| "num_tokens": 1347941197.0, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.9129786176031827, | |
| "grad_norm": 0.12070658802986145, | |
| "learning_rate": 1e-06, | |
| "loss": 0.521, | |
| "num_tokens": 1349348423.0, | |
| "step": 961 | |
| }, | |
| { | |
| "epoch": 1.9149676777722526, | |
| "grad_norm": 0.11713531613349915, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5194, | |
| "num_tokens": 1350775976.0, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 1.9169567379413226, | |
| "grad_norm": 0.11523653566837311, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5187, | |
| "num_tokens": 1352185197.0, | |
| "step": 963 | |
| }, | |
| { | |
| "epoch": 1.9189457981103928, | |
| "grad_norm": 0.1193997859954834, | |
| "learning_rate": 1e-06, | |
| "loss": 0.521, | |
| "num_tokens": 1353553440.0, | |
| "step": 964 | |
| }, | |
| { | |
| "epoch": 1.920934858279463, | |
| "grad_norm": 0.11642030626535416, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5187, | |
| "num_tokens": 1354944439.0, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 1.9229239184485332, | |
| "grad_norm": 0.12482885271310806, | |
| "learning_rate": 1e-06, | |
| "loss": 0.515, | |
| "num_tokens": 1356336096.0, | |
| "step": 966 | |
| }, | |
| { | |
| "epoch": 1.9249129786176031, | |
| "grad_norm": 0.12160996347665787, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5118, | |
| "num_tokens": 1357735474.0, | |
| "step": 967 | |
| }, | |
| { | |
| "epoch": 1.9269020387866733, | |
| "grad_norm": 0.11489357799291611, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5202, | |
| "num_tokens": 1359150923.0, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.9288910989557433, | |
| "grad_norm": 0.1207101047039032, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5185, | |
| "num_tokens": 1360546317.0, | |
| "step": 969 | |
| }, | |
| { | |
| "epoch": 1.9308801591248135, | |
| "grad_norm": 0.12328807264566422, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5359, | |
| "num_tokens": 1361956243.0, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.9328692192938837, | |
| "grad_norm": 0.11625958234071732, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5302, | |
| "num_tokens": 1363369844.0, | |
| "step": 971 | |
| }, | |
| { | |
| "epoch": 1.9348582794629539, | |
| "grad_norm": 0.11894381046295166, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5152, | |
| "num_tokens": 1364788358.0, | |
| "step": 972 | |
| }, | |
| { | |
| "epoch": 1.9368473396320238, | |
| "grad_norm": 0.11918807029724121, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5294, | |
| "num_tokens": 1366212166.0, | |
| "step": 973 | |
| }, | |
| { | |
| "epoch": 1.938836399801094, | |
| "grad_norm": 0.1214541345834732, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5279, | |
| "num_tokens": 1367620280.0, | |
| "step": 974 | |
| }, | |
| { | |
| "epoch": 1.940825459970164, | |
| "grad_norm": 0.12185929715633392, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5411, | |
| "num_tokens": 1369018460.0, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.9428145201392342, | |
| "grad_norm": 0.12074369192123413, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5169, | |
| "num_tokens": 1370404156.0, | |
| "step": 976 | |
| }, | |
| { | |
| "epoch": 1.9448035803083044, | |
| "grad_norm": 0.11974059790372849, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5189, | |
| "num_tokens": 1371779401.0, | |
| "step": 977 | |
| }, | |
| { | |
| "epoch": 1.9467926404773745, | |
| "grad_norm": 0.1180194690823555, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5255, | |
| "num_tokens": 1373167544.0, | |
| "step": 978 | |
| }, | |
| { | |
| "epoch": 1.9487817006464445, | |
| "grad_norm": 0.11891929805278778, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5181, | |
| "num_tokens": 1374625457.0, | |
| "step": 979 | |
| }, | |
| { | |
| "epoch": 1.9507707608155147, | |
| "grad_norm": 0.11925170570611954, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5386, | |
| "num_tokens": 1376040593.0, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.9527598209845847, | |
| "grad_norm": 0.11748997122049332, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5292, | |
| "num_tokens": 1377439737.0, | |
| "step": 981 | |
| }, | |
| { | |
| "epoch": 1.9547488811536549, | |
| "grad_norm": 0.12130767852067947, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5283, | |
| "num_tokens": 1378841802.0, | |
| "step": 982 | |
| }, | |
| { | |
| "epoch": 1.956737941322725, | |
| "grad_norm": 0.117047019302845, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5307, | |
| "num_tokens": 1380279333.0, | |
| "step": 983 | |
| }, | |
| { | |
| "epoch": 1.9587270014917952, | |
| "grad_norm": 0.11402452737092972, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5252, | |
| "num_tokens": 1381666440.0, | |
| "step": 984 | |
| }, | |
| { | |
| "epoch": 1.9607160616608652, | |
| "grad_norm": 0.11929241567850113, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5267, | |
| "num_tokens": 1383062810.0, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 1.9627051218299354, | |
| "grad_norm": 0.13006067276000977, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5245, | |
| "num_tokens": 1384441301.0, | |
| "step": 986 | |
| }, | |
| { | |
| "epoch": 1.9646941819990054, | |
| "grad_norm": 0.11855417490005493, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5241, | |
| "num_tokens": 1385828740.0, | |
| "step": 987 | |
| }, | |
| { | |
| "epoch": 1.9666832421680756, | |
| "grad_norm": 0.11694888770580292, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5198, | |
| "num_tokens": 1387217686.0, | |
| "step": 988 | |
| }, | |
| { | |
| "epoch": 1.9686723023371457, | |
| "grad_norm": 0.12000932544469833, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5167, | |
| "num_tokens": 1388611363.0, | |
| "step": 989 | |
| }, | |
| { | |
| "epoch": 1.970661362506216, | |
| "grad_norm": 0.11440891027450562, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5149, | |
| "num_tokens": 1390014972.0, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.972650422675286, | |
| "grad_norm": 0.11770491302013397, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5254, | |
| "num_tokens": 1391404923.0, | |
| "step": 991 | |
| }, | |
| { | |
| "epoch": 1.974639482844356, | |
| "grad_norm": 0.1166594997048378, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5316, | |
| "num_tokens": 1392816272.0, | |
| "step": 992 | |
| }, | |
| { | |
| "epoch": 1.976628543013426, | |
| "grad_norm": 0.12553872168064117, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5168, | |
| "num_tokens": 1394201525.0, | |
| "step": 993 | |
| }, | |
| { | |
| "epoch": 1.9786176031824962, | |
| "grad_norm": 0.11989990621805191, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5317, | |
| "num_tokens": 1395588841.0, | |
| "step": 994 | |
| }, | |
| { | |
| "epoch": 1.9806066633515664, | |
| "grad_norm": 0.1259329915046692, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5367, | |
| "num_tokens": 1397008721.0, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 1.9825957235206366, | |
| "grad_norm": 0.11585794389247894, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5216, | |
| "num_tokens": 1398471728.0, | |
| "step": 996 | |
| }, | |
| { | |
| "epoch": 1.9845847836897066, | |
| "grad_norm": 0.12714508175849915, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5278, | |
| "num_tokens": 1399889254.0, | |
| "step": 997 | |
| }, | |
| { | |
| "epoch": 1.9865738438587768, | |
| "grad_norm": 0.12047947198152542, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5258, | |
| "num_tokens": 1401259409.0, | |
| "step": 998 | |
| }, | |
| { | |
| "epoch": 1.9885629040278467, | |
| "grad_norm": 0.12171699851751328, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5198, | |
| "num_tokens": 1402651874.0, | |
| "step": 999 | |
| }, | |
| { | |
| "epoch": 1.990551964196917, | |
| "grad_norm": 0.11678752303123474, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5278, | |
| "num_tokens": 1404075471.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9925410243659871, | |
| "grad_norm": 0.1172628328204155, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5231, | |
| "num_tokens": 1405486757.0, | |
| "step": 1001 | |
| }, | |
| { | |
| "epoch": 1.9945300845350573, | |
| "grad_norm": 0.11984940618276596, | |
| "learning_rate": 1e-06, | |
| "loss": 0.522, | |
| "num_tokens": 1406892660.0, | |
| "step": 1002 | |
| }, | |
| { | |
| "epoch": 1.9965191447041273, | |
| "grad_norm": 0.11809879541397095, | |
| "learning_rate": 1e-06, | |
| "loss": 0.516, | |
| "num_tokens": 1408310027.0, | |
| "step": 1003 | |
| }, | |
| { | |
| "epoch": 1.9985082048731975, | |
| "grad_norm": 0.11894188821315765, | |
| "learning_rate": 1e-06, | |
| "loss": 0.5125, | |
| "num_tokens": 1409683792.0, | |
| "step": 1004 | |
| }, | |
| { | |
| "epoch": 1.9985082048731975, | |
| "step": 1004, | |
| "total_flos": 5.825582167854206e+19, | |
| "train_loss": 0.5989120793770034, | |
| "train_runtime": 31705.2652, | |
| "train_samples_per_second": 14.201, | |
| "train_steps_per_second": 0.032 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1004, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 51, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.825582167854206e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |