| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9990726429675425, | |
| "eval_steps": 500, | |
| "global_step": 202, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004945904173106646, | |
| "grad_norm": 54.500160217285156, | |
| "learning_rate": 0.0, | |
| "loss": 9.5291, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.009891808346213293, | |
| "grad_norm": 53.72835922241211, | |
| "learning_rate": 4e-05, | |
| "loss": 9.4814, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.014837712519319939, | |
| "grad_norm": 17.406553268432617, | |
| "learning_rate": 8e-05, | |
| "loss": 9.1288, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.019783616692426585, | |
| "grad_norm": 3.4949991703033447, | |
| "learning_rate": 0.00012, | |
| "loss": 8.8408, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.02472952086553323, | |
| "grad_norm": 2.9090073108673096, | |
| "learning_rate": 0.00016, | |
| "loss": 8.7705, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.029675425038639878, | |
| "grad_norm": 3.398167371749878, | |
| "learning_rate": 0.0002, | |
| "loss": 8.6466, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.03462132921174652, | |
| "grad_norm": 1.6190311908721924, | |
| "learning_rate": 0.00019898477157360406, | |
| "loss": 8.5125, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.03956723338485317, | |
| "grad_norm": 1.8773953914642334, | |
| "learning_rate": 0.00019796954314720813, | |
| "loss": 8.5322, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.04451313755795981, | |
| "grad_norm": 1.283807396888733, | |
| "learning_rate": 0.00019695431472081218, | |
| "loss": 8.4917, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.04945904173106646, | |
| "grad_norm": 1.9215106964111328, | |
| "learning_rate": 0.00019593908629441626, | |
| "loss": 8.3638, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05440494590417311, | |
| "grad_norm": 1.5560728311538696, | |
| "learning_rate": 0.00019492385786802033, | |
| "loss": 8.3021, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.059350850077279756, | |
| "grad_norm": 1.4610416889190674, | |
| "learning_rate": 0.00019390862944162438, | |
| "loss": 8.3058, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.0642967542503864, | |
| "grad_norm": 1.4304499626159668, | |
| "learning_rate": 0.00019289340101522843, | |
| "loss": 8.2576, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.06924265842349304, | |
| "grad_norm": 1.2287720441818237, | |
| "learning_rate": 0.0001918781725888325, | |
| "loss": 8.0443, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.07418856259659969, | |
| "grad_norm": 1.3729023933410645, | |
| "learning_rate": 0.00019086294416243655, | |
| "loss": 8.1255, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.07913446676970634, | |
| "grad_norm": 1.2619420289993286, | |
| "learning_rate": 0.0001898477157360406, | |
| "loss": 8.032, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.08408037094281298, | |
| "grad_norm": 1.4744280576705933, | |
| "learning_rate": 0.0001888324873096447, | |
| "loss": 7.8637, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.08902627511591962, | |
| "grad_norm": 1.6214470863342285, | |
| "learning_rate": 0.00018781725888324875, | |
| "loss": 7.9172, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.09397217928902628, | |
| "grad_norm": 1.283504605293274, | |
| "learning_rate": 0.0001868020304568528, | |
| "loss": 7.8251, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.09891808346213292, | |
| "grad_norm": 1.0794684886932373, | |
| "learning_rate": 0.00018578680203045687, | |
| "loss": 7.7431, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.10386398763523957, | |
| "grad_norm": 1.1826306581497192, | |
| "learning_rate": 0.00018477157360406092, | |
| "loss": 7.6118, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.10880989180834622, | |
| "grad_norm": 1.5493848323822021, | |
| "learning_rate": 0.00018375634517766497, | |
| "loss": 7.5928, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.11375579598145286, | |
| "grad_norm": 2.191657304763794, | |
| "learning_rate": 0.00018274111675126904, | |
| "loss": 7.596, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.11870170015455951, | |
| "grad_norm": 1.2168949842453003, | |
| "learning_rate": 0.0001817258883248731, | |
| "loss": 7.5224, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.12364760432766615, | |
| "grad_norm": 1.1562331914901733, | |
| "learning_rate": 0.00018071065989847717, | |
| "loss": 7.4952, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.1285935085007728, | |
| "grad_norm": 1.9624497890472412, | |
| "learning_rate": 0.00017969543147208124, | |
| "loss": 7.459, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.13353941267387945, | |
| "grad_norm": 2.2458877563476562, | |
| "learning_rate": 0.0001786802030456853, | |
| "loss": 7.3465, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.1384853168469861, | |
| "grad_norm": 1.3750243186950684, | |
| "learning_rate": 0.00017766497461928934, | |
| "loss": 7.3891, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.14343122102009273, | |
| "grad_norm": 1.2398021221160889, | |
| "learning_rate": 0.0001766497461928934, | |
| "loss": 7.3127, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.14837712519319937, | |
| "grad_norm": 2.071115732192993, | |
| "learning_rate": 0.00017563451776649746, | |
| "loss": 7.2548, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.15332302936630604, | |
| "grad_norm": 2.288498640060425, | |
| "learning_rate": 0.0001746192893401015, | |
| "loss": 7.1908, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.15826893353941268, | |
| "grad_norm": 1.2050567865371704, | |
| "learning_rate": 0.0001736040609137056, | |
| "loss": 7.1467, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.16321483771251932, | |
| "grad_norm": 1.4064340591430664, | |
| "learning_rate": 0.00017258883248730966, | |
| "loss": 7.035, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.16816074188562596, | |
| "grad_norm": 1.2630614042282104, | |
| "learning_rate": 0.0001715736040609137, | |
| "loss": 7.0536, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1731066460587326, | |
| "grad_norm": 1.8433802127838135, | |
| "learning_rate": 0.00017055837563451778, | |
| "loss": 7.0115, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.17805255023183925, | |
| "grad_norm": 1.744345784187317, | |
| "learning_rate": 0.00016954314720812183, | |
| "loss": 7.038, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.18299845440494591, | |
| "grad_norm": 1.679824709892273, | |
| "learning_rate": 0.00016852791878172588, | |
| "loss": 6.8946, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.18794435857805256, | |
| "grad_norm": 1.4559205770492554, | |
| "learning_rate": 0.00016751269035532995, | |
| "loss": 6.9053, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.1928902627511592, | |
| "grad_norm": 1.7544541358947754, | |
| "learning_rate": 0.00016649746192893403, | |
| "loss": 6.9277, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.19783616692426584, | |
| "grad_norm": 1.594734787940979, | |
| "learning_rate": 0.00016548223350253808, | |
| "loss": 6.912, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.20278207109737248, | |
| "grad_norm": 1.3439960479736328, | |
| "learning_rate": 0.00016446700507614215, | |
| "loss": 6.8592, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.20772797527047915, | |
| "grad_norm": 1.4330651760101318, | |
| "learning_rate": 0.0001634517766497462, | |
| "loss": 6.8965, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2126738794435858, | |
| "grad_norm": 2.439265489578247, | |
| "learning_rate": 0.00016243654822335025, | |
| "loss": 6.8126, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.21761978361669243, | |
| "grad_norm": 1.2343510389328003, | |
| "learning_rate": 0.00016142131979695432, | |
| "loss": 6.8057, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.22256568778979907, | |
| "grad_norm": 1.15224027633667, | |
| "learning_rate": 0.00016040609137055837, | |
| "loss": 6.6727, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2275115919629057, | |
| "grad_norm": 1.6769089698791504, | |
| "learning_rate": 0.00015939086294416242, | |
| "loss": 6.7457, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.23245749613601235, | |
| "grad_norm": 2.4642043113708496, | |
| "learning_rate": 0.00015837563451776652, | |
| "loss": 6.742, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.23740340030911902, | |
| "grad_norm": 1.1713383197784424, | |
| "learning_rate": 0.00015736040609137057, | |
| "loss": 6.7022, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.24234930448222566, | |
| "grad_norm": 1.5891178846359253, | |
| "learning_rate": 0.00015634517766497462, | |
| "loss": 6.6446, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.2472952086553323, | |
| "grad_norm": 2.0845682621002197, | |
| "learning_rate": 0.0001553299492385787, | |
| "loss": 6.5948, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.252241112828439, | |
| "grad_norm": 1.4469300508499146, | |
| "learning_rate": 0.00015431472081218274, | |
| "loss": 6.5604, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.2571870170015456, | |
| "grad_norm": 1.0141685009002686, | |
| "learning_rate": 0.0001532994923857868, | |
| "loss": 6.5418, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.26213292117465226, | |
| "grad_norm": 2.21588134765625, | |
| "learning_rate": 0.00015228426395939087, | |
| "loss": 6.4273, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.2670788253477589, | |
| "grad_norm": 1.4307092428207397, | |
| "learning_rate": 0.00015126903553299494, | |
| "loss": 6.4938, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.27202472952086554, | |
| "grad_norm": 1.4310742616653442, | |
| "learning_rate": 0.000150253807106599, | |
| "loss": 6.4357, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.2769706336939722, | |
| "grad_norm": 1.1520801782608032, | |
| "learning_rate": 0.00014923857868020306, | |
| "loss": 6.5101, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.2819165378670788, | |
| "grad_norm": 1.0513254404067993, | |
| "learning_rate": 0.0001482233502538071, | |
| "loss": 6.4536, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.28686244204018546, | |
| "grad_norm": 1.5814175605773926, | |
| "learning_rate": 0.00014720812182741116, | |
| "loss": 6.4139, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2918083462132921, | |
| "grad_norm": 1.5383965969085693, | |
| "learning_rate": 0.00014619289340101523, | |
| "loss": 6.3318, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.29675425038639874, | |
| "grad_norm": 1.0093541145324707, | |
| "learning_rate": 0.00014517766497461928, | |
| "loss": 6.4279, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3017001545595054, | |
| "grad_norm": 1.4959982633590698, | |
| "learning_rate": 0.00014416243654822336, | |
| "loss": 6.3061, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.3066460587326121, | |
| "grad_norm": 1.649026870727539, | |
| "learning_rate": 0.00014314720812182743, | |
| "loss": 6.274, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.3115919629057187, | |
| "grad_norm": 0.9700078964233398, | |
| "learning_rate": 0.00014213197969543148, | |
| "loss": 6.4123, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.31653786707882536, | |
| "grad_norm": 1.0136897563934326, | |
| "learning_rate": 0.00014111675126903553, | |
| "loss": 6.3055, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.321483771251932, | |
| "grad_norm": 1.6081498861312866, | |
| "learning_rate": 0.0001401015228426396, | |
| "loss": 6.3642, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.32642967542503865, | |
| "grad_norm": 1.1522279977798462, | |
| "learning_rate": 0.00013908629441624365, | |
| "loss": 6.2726, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.3313755795981453, | |
| "grad_norm": 0.8351190686225891, | |
| "learning_rate": 0.00013807106598984773, | |
| "loss": 6.2645, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.33632148377125193, | |
| "grad_norm": 1.1132313013076782, | |
| "learning_rate": 0.00013705583756345178, | |
| "loss": 6.2681, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.34126738794435857, | |
| "grad_norm": 1.2936571836471558, | |
| "learning_rate": 0.00013604060913705585, | |
| "loss": 6.2473, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.3462132921174652, | |
| "grad_norm": 1.250172734260559, | |
| "learning_rate": 0.0001350253807106599, | |
| "loss": 6.2264, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.35115919629057185, | |
| "grad_norm": 1.0878709554672241, | |
| "learning_rate": 0.00013401015228426397, | |
| "loss": 6.1898, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.3561051004636785, | |
| "grad_norm": 0.9934064149856567, | |
| "learning_rate": 0.00013299492385786802, | |
| "loss": 6.2047, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.3610510046367852, | |
| "grad_norm": 0.8686928749084473, | |
| "learning_rate": 0.00013197969543147207, | |
| "loss": 6.1214, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.36599690880989183, | |
| "grad_norm": 0.858200192451477, | |
| "learning_rate": 0.00013096446700507615, | |
| "loss": 6.0784, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.37094281298299847, | |
| "grad_norm": 0.8108780980110168, | |
| "learning_rate": 0.0001299492385786802, | |
| "loss": 6.1899, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.3758887171561051, | |
| "grad_norm": 0.8366422653198242, | |
| "learning_rate": 0.00012893401015228427, | |
| "loss": 6.131, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.38083462132921175, | |
| "grad_norm": 1.2487200498580933, | |
| "learning_rate": 0.00012791878172588834, | |
| "loss": 6.1158, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3857805255023184, | |
| "grad_norm": 1.0677459239959717, | |
| "learning_rate": 0.0001269035532994924, | |
| "loss": 6.0873, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.39072642967542504, | |
| "grad_norm": 0.9405259490013123, | |
| "learning_rate": 0.00012588832487309644, | |
| "loss": 6.0409, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.3956723338485317, | |
| "grad_norm": 1.488607406616211, | |
| "learning_rate": 0.00012487309644670052, | |
| "loss": 5.9868, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4006182380216383, | |
| "grad_norm": 0.9067093729972839, | |
| "learning_rate": 0.00012385786802030456, | |
| "loss": 6.0035, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.40556414219474496, | |
| "grad_norm": 1.1395992040634155, | |
| "learning_rate": 0.00012284263959390864, | |
| "loss": 5.9638, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.4105100463678516, | |
| "grad_norm": 1.4701273441314697, | |
| "learning_rate": 0.0001218274111675127, | |
| "loss": 6.0212, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.4154559505409583, | |
| "grad_norm": 0.8167937397956848, | |
| "learning_rate": 0.00012081218274111676, | |
| "loss": 6.0759, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.42040185471406494, | |
| "grad_norm": 1.398577332496643, | |
| "learning_rate": 0.00011979695431472082, | |
| "loss": 5.9284, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.4253477588871716, | |
| "grad_norm": 1.0022815465927124, | |
| "learning_rate": 0.00011878172588832489, | |
| "loss": 5.9638, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.4302936630602782, | |
| "grad_norm": 1.1316360235214233, | |
| "learning_rate": 0.00011776649746192893, | |
| "loss": 5.8901, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.43523956723338486, | |
| "grad_norm": 1.1034351587295532, | |
| "learning_rate": 0.000116751269035533, | |
| "loss": 5.9288, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.4401854714064915, | |
| "grad_norm": 0.9991883039474487, | |
| "learning_rate": 0.00011573604060913706, | |
| "loss": 5.9447, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.44513137557959814, | |
| "grad_norm": 1.4334654808044434, | |
| "learning_rate": 0.00011472081218274113, | |
| "loss": 5.8657, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.4500772797527048, | |
| "grad_norm": 1.0602012872695923, | |
| "learning_rate": 0.0001137055837563452, | |
| "loss": 5.8563, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.4550231839258114, | |
| "grad_norm": 0.9210672378540039, | |
| "learning_rate": 0.00011269035532994925, | |
| "loss": 5.8811, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.45996908809891807, | |
| "grad_norm": 0.9101308584213257, | |
| "learning_rate": 0.0001116751269035533, | |
| "loss": 5.9572, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.4649149922720247, | |
| "grad_norm": 0.8447904586791992, | |
| "learning_rate": 0.00011065989847715736, | |
| "loss": 5.8762, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.46986089644513135, | |
| "grad_norm": 0.7616278529167175, | |
| "learning_rate": 0.00010964467005076143, | |
| "loss": 5.9493, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.47480680061823805, | |
| "grad_norm": 1.0465595722198486, | |
| "learning_rate": 0.00010862944162436547, | |
| "loss": 5.8367, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.4797527047913447, | |
| "grad_norm": 1.4627708196640015, | |
| "learning_rate": 0.00010761421319796954, | |
| "loss": 5.8301, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.4846986089644513, | |
| "grad_norm": 1.0495349168777466, | |
| "learning_rate": 0.00010659898477157362, | |
| "loss": 5.8782, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.48964451313755797, | |
| "grad_norm": 0.9480841755867004, | |
| "learning_rate": 0.00010558375634517767, | |
| "loss": 5.7681, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.4945904173106646, | |
| "grad_norm": 0.8606300354003906, | |
| "learning_rate": 0.00010456852791878173, | |
| "loss": 5.7448, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.49953632148377125, | |
| "grad_norm": 0.9947773218154907, | |
| "learning_rate": 0.0001035532994923858, | |
| "loss": 5.8485, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.504482225656878, | |
| "grad_norm": 1.0647828578948975, | |
| "learning_rate": 0.00010253807106598984, | |
| "loss": 5.7214, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.5094281298299845, | |
| "grad_norm": 1.1592961549758911, | |
| "learning_rate": 0.0001015228426395939, | |
| "loss": 5.7393, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.5143740340030912, | |
| "grad_norm": 0.8949771523475647, | |
| "learning_rate": 0.00010050761421319797, | |
| "loss": 5.7635, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.5193199381761978, | |
| "grad_norm": 0.8713933229446411, | |
| "learning_rate": 9.949238578680203e-05, | |
| "loss": 5.7227, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5242658423493045, | |
| "grad_norm": 0.8814818859100342, | |
| "learning_rate": 9.847715736040609e-05, | |
| "loss": 5.7516, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.5292117465224111, | |
| "grad_norm": 0.9553707838058472, | |
| "learning_rate": 9.746192893401017e-05, | |
| "loss": 5.7522, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.5341576506955178, | |
| "grad_norm": 0.8567320704460144, | |
| "learning_rate": 9.644670050761421e-05, | |
| "loss": 5.6508, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.5391035548686244, | |
| "grad_norm": 1.0081580877304077, | |
| "learning_rate": 9.543147208121828e-05, | |
| "loss": 5.642, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.5440494590417311, | |
| "grad_norm": 1.1526085138320923, | |
| "learning_rate": 9.441624365482235e-05, | |
| "loss": 5.7423, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5489953632148377, | |
| "grad_norm": 1.2273470163345337, | |
| "learning_rate": 9.34010152284264e-05, | |
| "loss": 5.7094, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.5539412673879444, | |
| "grad_norm": 0.830719530582428, | |
| "learning_rate": 9.238578680203046e-05, | |
| "loss": 5.7365, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.558887171561051, | |
| "grad_norm": 1.1520576477050781, | |
| "learning_rate": 9.137055837563452e-05, | |
| "loss": 5.7391, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.5638330757341576, | |
| "grad_norm": 1.1414787769317627, | |
| "learning_rate": 9.035532994923858e-05, | |
| "loss": 5.7288, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.5687789799072643, | |
| "grad_norm": 0.9615758061408997, | |
| "learning_rate": 8.934010152284265e-05, | |
| "loss": 5.5568, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.5737248840803709, | |
| "grad_norm": 0.8781617879867554, | |
| "learning_rate": 8.83248730964467e-05, | |
| "loss": 5.6264, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.5786707882534776, | |
| "grad_norm": 1.1544886827468872, | |
| "learning_rate": 8.730964467005075e-05, | |
| "loss": 5.6724, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.5836166924265842, | |
| "grad_norm": 0.931874692440033, | |
| "learning_rate": 8.629441624365483e-05, | |
| "loss": 5.6046, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.5885625965996909, | |
| "grad_norm": 0.7856680750846863, | |
| "learning_rate": 8.527918781725889e-05, | |
| "loss": 5.6521, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.5935085007727975, | |
| "grad_norm": 1.162001609802246, | |
| "learning_rate": 8.426395939086294e-05, | |
| "loss": 5.5843, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5984544049459042, | |
| "grad_norm": 0.8572034239768982, | |
| "learning_rate": 8.324873096446701e-05, | |
| "loss": 5.6526, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.6034003091190108, | |
| "grad_norm": 0.9555945992469788, | |
| "learning_rate": 8.223350253807108e-05, | |
| "loss": 5.6673, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.6083462132921175, | |
| "grad_norm": 0.880160927772522, | |
| "learning_rate": 8.121827411167512e-05, | |
| "loss": 5.498, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.6132921174652242, | |
| "grad_norm": 1.1022496223449707, | |
| "learning_rate": 8.020304568527919e-05, | |
| "loss": 5.5833, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.6182380216383307, | |
| "grad_norm": 0.9595851898193359, | |
| "learning_rate": 7.918781725888326e-05, | |
| "loss": 5.6384, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.6231839258114374, | |
| "grad_norm": 1.4313597679138184, | |
| "learning_rate": 7.817258883248731e-05, | |
| "loss": 5.5478, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.628129829984544, | |
| "grad_norm": 0.9351322054862976, | |
| "learning_rate": 7.715736040609137e-05, | |
| "loss": 5.5652, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.6330757341576507, | |
| "grad_norm": 1.251789927482605, | |
| "learning_rate": 7.614213197969543e-05, | |
| "loss": 5.5387, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.6380216383307573, | |
| "grad_norm": 0.98284912109375, | |
| "learning_rate": 7.51269035532995e-05, | |
| "loss": 5.5338, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.642967542503864, | |
| "grad_norm": 1.0421977043151855, | |
| "learning_rate": 7.411167512690356e-05, | |
| "loss": 5.5774, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6479134466769706, | |
| "grad_norm": 1.0751053094863892, | |
| "learning_rate": 7.309644670050762e-05, | |
| "loss": 5.5642, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.6528593508500773, | |
| "grad_norm": 1.089376449584961, | |
| "learning_rate": 7.208121827411168e-05, | |
| "loss": 5.505, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.6578052550231839, | |
| "grad_norm": 1.0731728076934814, | |
| "learning_rate": 7.106598984771574e-05, | |
| "loss": 5.5514, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.6627511591962906, | |
| "grad_norm": 1.2262444496154785, | |
| "learning_rate": 7.00507614213198e-05, | |
| "loss": 5.5723, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.6676970633693973, | |
| "grad_norm": 1.0487595796585083, | |
| "learning_rate": 6.903553299492386e-05, | |
| "loss": 5.5587, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.6726429675425039, | |
| "grad_norm": 1.084671139717102, | |
| "learning_rate": 6.802030456852793e-05, | |
| "loss": 5.4868, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.6775888717156106, | |
| "grad_norm": 1.1871248483657837, | |
| "learning_rate": 6.700507614213199e-05, | |
| "loss": 5.5475, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.6825347758887171, | |
| "grad_norm": 0.960493803024292, | |
| "learning_rate": 6.598984771573604e-05, | |
| "loss": 5.5006, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.6874806800618238, | |
| "grad_norm": 1.053593397140503, | |
| "learning_rate": 6.49746192893401e-05, | |
| "loss": 5.5389, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.6924265842349304, | |
| "grad_norm": 0.8886996507644653, | |
| "learning_rate": 6.395939086294417e-05, | |
| "loss": 5.4616, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6973724884080371, | |
| "grad_norm": 1.1852856874465942, | |
| "learning_rate": 6.294416243654822e-05, | |
| "loss": 5.498, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.7023183925811437, | |
| "grad_norm": 0.8381466865539551, | |
| "learning_rate": 6.192893401015228e-05, | |
| "loss": 5.4977, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.7072642967542504, | |
| "grad_norm": 1.01845121383667, | |
| "learning_rate": 6.091370558375635e-05, | |
| "loss": 5.4162, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.712210200927357, | |
| "grad_norm": 0.9204426407814026, | |
| "learning_rate": 5.989847715736041e-05, | |
| "loss": 5.4654, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.7171561051004637, | |
| "grad_norm": 1.0901105403900146, | |
| "learning_rate": 5.8883248730964467e-05, | |
| "loss": 5.4262, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7221020092735704, | |
| "grad_norm": 0.9842381477355957, | |
| "learning_rate": 5.786802030456853e-05, | |
| "loss": 5.4622, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.727047913446677, | |
| "grad_norm": 1.1234885454177856, | |
| "learning_rate": 5.68527918781726e-05, | |
| "loss": 5.4668, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.7319938176197837, | |
| "grad_norm": 1.0685431957244873, | |
| "learning_rate": 5.583756345177665e-05, | |
| "loss": 5.4649, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.7369397217928902, | |
| "grad_norm": 1.086138367652893, | |
| "learning_rate": 5.482233502538071e-05, | |
| "loss": 5.336, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.7418856259659969, | |
| "grad_norm": 1.0806076526641846, | |
| "learning_rate": 5.380710659898477e-05, | |
| "loss": 5.3463, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.7468315301391035, | |
| "grad_norm": 1.1613116264343262, | |
| "learning_rate": 5.2791878172588836e-05, | |
| "loss": 5.4095, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.7517774343122102, | |
| "grad_norm": 1.1117639541625977, | |
| "learning_rate": 5.17766497461929e-05, | |
| "loss": 5.4248, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.7567233384853168, | |
| "grad_norm": 0.9730443954467773, | |
| "learning_rate": 5.076142131979695e-05, | |
| "loss": 5.5573, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.7616692426584235, | |
| "grad_norm": 1.0216584205627441, | |
| "learning_rate": 4.9746192893401014e-05, | |
| "loss": 5.3337, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.7666151468315301, | |
| "grad_norm": 0.9828229546546936, | |
| "learning_rate": 4.873096446700508e-05, | |
| "loss": 5.3757, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.7715610510046368, | |
| "grad_norm": 1.0315641164779663, | |
| "learning_rate": 4.771573604060914e-05, | |
| "loss": 5.4465, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.7765069551777435, | |
| "grad_norm": 1.1969993114471436, | |
| "learning_rate": 4.67005076142132e-05, | |
| "loss": 5.4018, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.7814528593508501, | |
| "grad_norm": 0.7633097171783447, | |
| "learning_rate": 4.568527918781726e-05, | |
| "loss": 5.5137, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.7863987635239568, | |
| "grad_norm": 0.8312305212020874, | |
| "learning_rate": 4.467005076142132e-05, | |
| "loss": 5.4078, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.7913446676970634, | |
| "grad_norm": 0.9463878870010376, | |
| "learning_rate": 4.365482233502538e-05, | |
| "loss": 5.3738, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.79629057187017, | |
| "grad_norm": 0.8046661615371704, | |
| "learning_rate": 4.2639593908629446e-05, | |
| "loss": 5.455, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.8012364760432766, | |
| "grad_norm": 1.0929735898971558, | |
| "learning_rate": 4.162436548223351e-05, | |
| "loss": 5.4263, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.8061823802163833, | |
| "grad_norm": 1.0323022603988647, | |
| "learning_rate": 4.060913705583756e-05, | |
| "loss": 5.4503, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.8111282843894899, | |
| "grad_norm": 0.7212726473808289, | |
| "learning_rate": 3.959390862944163e-05, | |
| "loss": 5.3904, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.8160741885625966, | |
| "grad_norm": 0.8705483078956604, | |
| "learning_rate": 3.8578680203045685e-05, | |
| "loss": 5.2958, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.8210200927357032, | |
| "grad_norm": 0.9705776572227478, | |
| "learning_rate": 3.756345177664975e-05, | |
| "loss": 5.3806, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.8259659969088099, | |
| "grad_norm": 0.7694171667098999, | |
| "learning_rate": 3.654822335025381e-05, | |
| "loss": 5.3446, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.8309119010819166, | |
| "grad_norm": 1.0148179531097412, | |
| "learning_rate": 3.553299492385787e-05, | |
| "loss": 5.4316, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.8358578052550232, | |
| "grad_norm": 1.0124086141586304, | |
| "learning_rate": 3.451776649746193e-05, | |
| "loss": 5.2903, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.8408037094281299, | |
| "grad_norm": 0.8755667209625244, | |
| "learning_rate": 3.3502538071065994e-05, | |
| "loss": 5.2636, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.8457496136012365, | |
| "grad_norm": 0.992751955986023, | |
| "learning_rate": 3.248730964467005e-05, | |
| "loss": 5.3662, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.8506955177743432, | |
| "grad_norm": 0.676480770111084, | |
| "learning_rate": 3.147208121827411e-05, | |
| "loss": 5.3912, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.8556414219474497, | |
| "grad_norm": 0.8479735851287842, | |
| "learning_rate": 3.0456852791878175e-05, | |
| "loss": 5.5655, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.8605873261205564, | |
| "grad_norm": 0.8780114054679871, | |
| "learning_rate": 2.9441624365482233e-05, | |
| "loss": 5.4011, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.865533230293663, | |
| "grad_norm": 0.7192287445068359, | |
| "learning_rate": 2.84263959390863e-05, | |
| "loss": 5.46, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.8704791344667697, | |
| "grad_norm": 0.9556674957275391, | |
| "learning_rate": 2.7411167512690357e-05, | |
| "loss": 5.4278, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.8754250386398763, | |
| "grad_norm": 0.7303546667098999, | |
| "learning_rate": 2.6395939086294418e-05, | |
| "loss": 5.3822, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.880370942812983, | |
| "grad_norm": 0.7659119963645935, | |
| "learning_rate": 2.5380710659898476e-05, | |
| "loss": 5.3925, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.8853168469860896, | |
| "grad_norm": 0.8511722087860107, | |
| "learning_rate": 2.436548223350254e-05, | |
| "loss": 5.3318, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.8902627511591963, | |
| "grad_norm": 0.8240477442741394, | |
| "learning_rate": 2.33502538071066e-05, | |
| "loss": 5.2479, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.895208655332303, | |
| "grad_norm": 0.8193429112434387, | |
| "learning_rate": 2.233502538071066e-05, | |
| "loss": 5.4237, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.9001545595054096, | |
| "grad_norm": 0.8074966669082642, | |
| "learning_rate": 2.1319796954314723e-05, | |
| "loss": 5.5029, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.9051004636785163, | |
| "grad_norm": 0.6603164076805115, | |
| "learning_rate": 2.030456852791878e-05, | |
| "loss": 5.3007, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.9100463678516229, | |
| "grad_norm": 0.633477509021759, | |
| "learning_rate": 1.9289340101522843e-05, | |
| "loss": 5.396, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.9149922720247295, | |
| "grad_norm": 0.6681249141693115, | |
| "learning_rate": 1.8274111675126904e-05, | |
| "loss": 5.3733, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.9199381761978361, | |
| "grad_norm": 0.756808340549469, | |
| "learning_rate": 1.7258883248730966e-05, | |
| "loss": 5.3439, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.9248840803709428, | |
| "grad_norm": 0.64524906873703, | |
| "learning_rate": 1.6243654822335024e-05, | |
| "loss": 5.4027, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.9298299845440494, | |
| "grad_norm": 0.7147576212882996, | |
| "learning_rate": 1.5228426395939088e-05, | |
| "loss": 5.3111, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.9347758887171561, | |
| "grad_norm": 0.6565448641777039, | |
| "learning_rate": 1.421319796954315e-05, | |
| "loss": 5.3649, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.9397217928902627, | |
| "grad_norm": 0.6476154923439026, | |
| "learning_rate": 1.3197969543147209e-05, | |
| "loss": 5.3617, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.9446676970633694, | |
| "grad_norm": 0.6315869092941284, | |
| "learning_rate": 1.218274111675127e-05, | |
| "loss": 5.3247, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.9496136012364761, | |
| "grad_norm": 0.6404466032981873, | |
| "learning_rate": 1.116751269035533e-05, | |
| "loss": 5.3402, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.9545595054095827, | |
| "grad_norm": 0.6863434314727783, | |
| "learning_rate": 1.015228426395939e-05, | |
| "loss": 5.3436, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.9595054095826894, | |
| "grad_norm": 0.6492709517478943, | |
| "learning_rate": 9.137055837563452e-06, | |
| "loss": 5.2449, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.964451313755796, | |
| "grad_norm": 0.647345781326294, | |
| "learning_rate": 8.121827411167512e-06, | |
| "loss": 5.3811, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.9693972179289027, | |
| "grad_norm": 0.711609423160553, | |
| "learning_rate": 7.106598984771575e-06, | |
| "loss": 5.3612, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.9743431221020092, | |
| "grad_norm": 0.610159158706665, | |
| "learning_rate": 6.091370558375635e-06, | |
| "loss": 5.3041, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.9792890262751159, | |
| "grad_norm": 0.61027592420578, | |
| "learning_rate": 5.076142131979695e-06, | |
| "loss": 5.3324, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.9842349304482225, | |
| "grad_norm": 0.5848086476325989, | |
| "learning_rate": 4.060913705583756e-06, | |
| "loss": 5.3446, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.9891808346213292, | |
| "grad_norm": 0.5617231130599976, | |
| "learning_rate": 3.0456852791878177e-06, | |
| "loss": 5.3997, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9941267387944358, | |
| "grad_norm": 0.6468728184700012, | |
| "learning_rate": 2.030456852791878e-06, | |
| "loss": 5.3444, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.9990726429675425, | |
| "grad_norm": 0.629033088684082, | |
| "learning_rate": 1.015228426395939e-06, | |
| "loss": 5.3283, | |
| "step": 202 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 202, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5526784012305408.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |