| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.993127147766323, | |
| "eval_steps": 500, | |
| "global_step": 290, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.006872852233676976, | |
| "grad_norm": 12.446887910679585, | |
| "learning_rate": 3.4482758620689656e-07, | |
| "loss": 0.2834, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013745704467353952, | |
| "grad_norm": 12.429452414470783, | |
| "learning_rate": 6.896551724137931e-07, | |
| "loss": 0.3293, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020618556701030927, | |
| "grad_norm": 13.046523782975127, | |
| "learning_rate": 1.0344827586206898e-06, | |
| "loss": 0.3345, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.027491408934707903, | |
| "grad_norm": 11.482419426981112, | |
| "learning_rate": 1.3793103448275862e-06, | |
| "loss": 0.3037, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03436426116838488, | |
| "grad_norm": 9.866874046616388, | |
| "learning_rate": 1.724137931034483e-06, | |
| "loss": 0.2596, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.041237113402061855, | |
| "grad_norm": 8.65788243489162, | |
| "learning_rate": 2.0689655172413796e-06, | |
| "loss": 0.2838, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.048109965635738834, | |
| "grad_norm": 4.957384731790736, | |
| "learning_rate": 2.4137931034482762e-06, | |
| "loss": 0.1885, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.054982817869415807, | |
| "grad_norm": 4.57635197761443, | |
| "learning_rate": 2.7586206896551725e-06, | |
| "loss": 0.1622, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.061855670103092786, | |
| "grad_norm": 3.8954992099722934, | |
| "learning_rate": 3.103448275862069e-06, | |
| "loss": 0.1544, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06872852233676977, | |
| "grad_norm": 2.514661045787612, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 0.1234, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07560137457044673, | |
| "grad_norm": 2.5123299291804586, | |
| "learning_rate": 3.793103448275862e-06, | |
| "loss": 0.1636, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08247422680412371, | |
| "grad_norm": 4.597805676274452, | |
| "learning_rate": 4.137931034482759e-06, | |
| "loss": 0.1572, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08934707903780069, | |
| "grad_norm": 4.1405165287398225, | |
| "learning_rate": 4.482758620689656e-06, | |
| "loss": 0.1456, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09621993127147767, | |
| "grad_norm": 3.39631044837988, | |
| "learning_rate": 4.8275862068965525e-06, | |
| "loss": 0.1405, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10309278350515463, | |
| "grad_norm": 2.051858086622133, | |
| "learning_rate": 5.172413793103449e-06, | |
| "loss": 0.1168, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10996563573883161, | |
| "grad_norm": 2.7583765332104706, | |
| "learning_rate": 5.517241379310345e-06, | |
| "loss": 0.1268, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11683848797250859, | |
| "grad_norm": 2.48691340365717, | |
| "learning_rate": 5.862068965517242e-06, | |
| "loss": 0.14, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12371134020618557, | |
| "grad_norm": 3.01546359040943, | |
| "learning_rate": 6.206896551724138e-06, | |
| "loss": 0.117, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.13058419243986255, | |
| "grad_norm": 2.7978944704054545, | |
| "learning_rate": 6.551724137931035e-06, | |
| "loss": 0.1312, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.13745704467353953, | |
| "grad_norm": 2.422905745214318, | |
| "learning_rate": 6.896551724137932e-06, | |
| "loss": 0.1167, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14432989690721648, | |
| "grad_norm": 2.4720006158953995, | |
| "learning_rate": 7.241379310344828e-06, | |
| "loss": 0.1265, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.15120274914089346, | |
| "grad_norm": 2.014336454846039, | |
| "learning_rate": 7.586206896551724e-06, | |
| "loss": 0.093, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15807560137457044, | |
| "grad_norm": 1.8872902056247776, | |
| "learning_rate": 7.93103448275862e-06, | |
| "loss": 0.1279, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16494845360824742, | |
| "grad_norm": 1.3068685740293375, | |
| "learning_rate": 8.275862068965518e-06, | |
| "loss": 0.0879, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1718213058419244, | |
| "grad_norm": 1.6231413000872636, | |
| "learning_rate": 8.620689655172414e-06, | |
| "loss": 0.1194, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.17869415807560138, | |
| "grad_norm": 1.830913490385725, | |
| "learning_rate": 8.965517241379312e-06, | |
| "loss": 0.0972, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18556701030927836, | |
| "grad_norm": 1.43638537376524, | |
| "learning_rate": 9.310344827586207e-06, | |
| "loss": 0.1016, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.19243986254295534, | |
| "grad_norm": 1.34655506640705, | |
| "learning_rate": 9.655172413793105e-06, | |
| "loss": 0.0783, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19931271477663232, | |
| "grad_norm": 1.8699602914396067, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1163, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20618556701030927, | |
| "grad_norm": 1.4332881506384583, | |
| "learning_rate": 9.999637795788383e-06, | |
| "loss": 0.1092, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.21305841924398625, | |
| "grad_norm": 2.8367123956352382, | |
| "learning_rate": 9.99855123563029e-06, | |
| "loss": 0.0966, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.21993127147766323, | |
| "grad_norm": 1.9168344675729085, | |
| "learning_rate": 9.996740476948386e-06, | |
| "loss": 0.1071, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2268041237113402, | |
| "grad_norm": 1.90674976333798, | |
| "learning_rate": 9.994205782088438e-06, | |
| "loss": 0.1218, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.23367697594501718, | |
| "grad_norm": 2.2093228008800123, | |
| "learning_rate": 9.990947518281312e-06, | |
| "loss": 0.1101, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.24054982817869416, | |
| "grad_norm": 1.6767140313987619, | |
| "learning_rate": 9.986966157589751e-06, | |
| "loss": 0.1132, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.24742268041237114, | |
| "grad_norm": 1.713955261362431, | |
| "learning_rate": 9.982262276840002e-06, | |
| "loss": 0.121, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2542955326460481, | |
| "grad_norm": 1.4682418451686818, | |
| "learning_rate": 9.976836557538234e-06, | |
| "loss": 0.1445, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2611683848797251, | |
| "grad_norm": 1.6091898287864894, | |
| "learning_rate": 9.970689785771798e-06, | |
| "loss": 0.0915, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.26804123711340205, | |
| "grad_norm": 1.9803020357446164, | |
| "learning_rate": 9.963822852095344e-06, | |
| "loss": 0.131, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.27491408934707906, | |
| "grad_norm": 1.1252019923472592, | |
| "learning_rate": 9.95623675140179e-06, | |
| "loss": 0.1068, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.281786941580756, | |
| "grad_norm": 1.4613264618668687, | |
| "learning_rate": 9.947932582778188e-06, | |
| "loss": 0.0986, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.28865979381443296, | |
| "grad_norm": 1.5989469501473161, | |
| "learning_rate": 9.938911549346473e-06, | |
| "loss": 0.1017, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.29553264604810997, | |
| "grad_norm": 2.505841827461168, | |
| "learning_rate": 9.929174958089167e-06, | |
| "loss": 0.1197, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3024054982817869, | |
| "grad_norm": 1.4740453235977062, | |
| "learning_rate": 9.918724219660013e-06, | |
| "loss": 0.0957, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.30927835051546393, | |
| "grad_norm": 1.5453884543677705, | |
| "learning_rate": 9.907560848179607e-06, | |
| "loss": 0.0946, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.3161512027491409, | |
| "grad_norm": 1.419362944301792, | |
| "learning_rate": 9.895686461016007e-06, | |
| "loss": 0.1317, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3230240549828179, | |
| "grad_norm": 2.517875062527592, | |
| "learning_rate": 9.883102778550434e-06, | |
| "loss": 0.0983, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.32989690721649484, | |
| "grad_norm": 1.5290936721831956, | |
| "learning_rate": 9.869811623928001e-06, | |
| "loss": 0.1185, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.33676975945017185, | |
| "grad_norm": 1.5453073418306231, | |
| "learning_rate": 9.855814922793583e-06, | |
| "loss": 0.0896, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3436426116838488, | |
| "grad_norm": 1.1930418995796759, | |
| "learning_rate": 9.841114703012817e-06, | |
| "loss": 0.1004, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35051546391752575, | |
| "grad_norm": 1.191862362726522, | |
| "learning_rate": 9.82571309437831e-06, | |
| "loss": 0.0952, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.35738831615120276, | |
| "grad_norm": 1.3520664749179854, | |
| "learning_rate": 9.809612328301071e-06, | |
| "loss": 0.086, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3642611683848797, | |
| "grad_norm": 1.1227238209986914, | |
| "learning_rate": 9.792814737487207e-06, | |
| "loss": 0.088, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3711340206185567, | |
| "grad_norm": 2.016644575693907, | |
| "learning_rate": 9.775322755599979e-06, | |
| "loss": 0.1199, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.37800687285223367, | |
| "grad_norm": 1.2601186736740226, | |
| "learning_rate": 9.757138916907184e-06, | |
| "loss": 0.123, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3848797250859107, | |
| "grad_norm": 1.026146060558469, | |
| "learning_rate": 9.738265855914014e-06, | |
| "loss": 0.0939, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3917525773195876, | |
| "grad_norm": 0.987030917157837, | |
| "learning_rate": 9.718706306981332e-06, | |
| "loss": 0.0839, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.39862542955326463, | |
| "grad_norm": 1.9548328187036061, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.1119, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4054982817869416, | |
| "grad_norm": 1.2898297124885263, | |
| "learning_rate": 9.677539179628005e-06, | |
| "loss": 0.1038, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.41237113402061853, | |
| "grad_norm": 1.5547019497085608, | |
| "learning_rate": 9.655937565570124e-06, | |
| "loss": 0.1, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.41924398625429554, | |
| "grad_norm": 2.0533615021264056, | |
| "learning_rate": 9.63366139143413e-06, | |
| "loss": 0.0987, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4261168384879725, | |
| "grad_norm": 1.706734196237327, | |
| "learning_rate": 9.610713884629667e-06, | |
| "loss": 0.1123, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4329896907216495, | |
| "grad_norm": 1.5206279314003526, | |
| "learning_rate": 9.587098369830171e-06, | |
| "loss": 0.0927, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.43986254295532645, | |
| "grad_norm": 1.203865286479759, | |
| "learning_rate": 9.562818268491216e-06, | |
| "loss": 0.097, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.44673539518900346, | |
| "grad_norm": 1.0193770675036011, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.1027, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4536082474226804, | |
| "grad_norm": 0.942629359352114, | |
| "learning_rate": 9.512278472939627e-06, | |
| "loss": 0.0805, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.46048109965635736, | |
| "grad_norm": 0.9156064631492126, | |
| "learning_rate": 9.486026101017711e-06, | |
| "loss": 0.069, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.46735395189003437, | |
| "grad_norm": 0.7030071088438584, | |
| "learning_rate": 9.459123786076911e-06, | |
| "loss": 0.0624, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4742268041237113, | |
| "grad_norm": 1.272597886993796, | |
| "learning_rate": 9.431575425769938e-06, | |
| "loss": 0.0718, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.48109965635738833, | |
| "grad_norm": 0.969219105898446, | |
| "learning_rate": 9.40338501134964e-06, | |
| "loss": 0.0911, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4879725085910653, | |
| "grad_norm": 1.1110059909140586, | |
| "learning_rate": 9.374556627090749e-06, | |
| "loss": 0.1203, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4948453608247423, | |
| "grad_norm": 1.3060655141326731, | |
| "learning_rate": 9.345094449698143e-06, | |
| "loss": 0.1028, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5017182130584192, | |
| "grad_norm": 0.8454116208381376, | |
| "learning_rate": 9.315002747701716e-06, | |
| "loss": 0.0779, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5085910652920962, | |
| "grad_norm": 1.2245489554639144, | |
| "learning_rate": 9.284285880837947e-06, | |
| "loss": 0.1079, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5154639175257731, | |
| "grad_norm": 1.24949164301992, | |
| "learning_rate": 9.252948299418255e-06, | |
| "loss": 0.0847, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5223367697594502, | |
| "grad_norm": 1.4441641022503702, | |
| "learning_rate": 9.220994543684225e-06, | |
| "loss": 0.0966, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5292096219931272, | |
| "grad_norm": 1.2875737252003576, | |
| "learning_rate": 9.188429243149824e-06, | |
| "loss": 0.0882, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5360824742268041, | |
| "grad_norm": 1.4598658768435573, | |
| "learning_rate": 9.155257115930651e-06, | |
| "loss": 0.0833, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5429553264604811, | |
| "grad_norm": 1.4589998192663096, | |
| "learning_rate": 9.121482968060384e-06, | |
| "loss": 0.0848, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5498281786941581, | |
| "grad_norm": 1.0001022021421828, | |
| "learning_rate": 9.08711169279446e-06, | |
| "loss": 0.0779, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5567010309278351, | |
| "grad_norm": 2.043693632898222, | |
| "learning_rate": 9.052148269901145e-06, | |
| "loss": 0.0997, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.563573883161512, | |
| "grad_norm": 3.1578424725116543, | |
| "learning_rate": 9.01659776494005e-06, | |
| "loss": 0.1126, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.570446735395189, | |
| "grad_norm": 1.1437184920920935, | |
| "learning_rate": 8.98046532852822e-06, | |
| "loss": 0.0754, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5773195876288659, | |
| "grad_norm": 1.479870857688087, | |
| "learning_rate": 8.943756195593916e-06, | |
| "loss": 0.0674, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.584192439862543, | |
| "grad_norm": 2.0780211475668393, | |
| "learning_rate": 8.90647568461816e-06, | |
| "loss": 0.0949, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5910652920962199, | |
| "grad_norm": 0.8998912772927933, | |
| "learning_rate": 8.868629196864182e-06, | |
| "loss": 0.0828, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5979381443298969, | |
| "grad_norm": 1.8464674716362912, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.0876, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6048109965635738, | |
| "grad_norm": 1.7992069874270102, | |
| "learning_rate": 8.791260305278434e-06, | |
| "loss": 0.0934, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6116838487972509, | |
| "grad_norm": 0.7515466883225197, | |
| "learning_rate": 8.751749110782013e-06, | |
| "loss": 0.0742, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6185567010309279, | |
| "grad_norm": 0.9257103946719526, | |
| "learning_rate": 8.71169435655405e-06, | |
| "loss": 0.0712, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6254295532646048, | |
| "grad_norm": 1.3589123926397728, | |
| "learning_rate": 8.671101845794816e-06, | |
| "loss": 0.0836, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6323024054982818, | |
| "grad_norm": 0.7707830586742125, | |
| "learning_rate": 8.629977459615655e-06, | |
| "loss": 0.0688, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6391752577319587, | |
| "grad_norm": 1.4148024669035801, | |
| "learning_rate": 8.588327156186915e-06, | |
| "loss": 0.1009, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6460481099656358, | |
| "grad_norm": 0.9537451611875749, | |
| "learning_rate": 8.546156969874723e-06, | |
| "loss": 0.085, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6529209621993127, | |
| "grad_norm": 0.9919416684553015, | |
| "learning_rate": 8.503473010366713e-06, | |
| "loss": 0.0876, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6597938144329897, | |
| "grad_norm": 1.2519624016536637, | |
| "learning_rate": 8.460281461786848e-06, | |
| "loss": 0.0632, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.7459867183832114, | |
| "learning_rate": 8.416588581799447e-06, | |
| "loss": 0.0799, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6735395189003437, | |
| "grad_norm": 1.1935403962910782, | |
| "learning_rate": 8.372400700702569e-06, | |
| "loss": 0.0967, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6804123711340206, | |
| "grad_norm": 0.7735151653835364, | |
| "learning_rate": 8.327724220510873e-06, | |
| "loss": 0.0643, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6872852233676976, | |
| "grad_norm": 0.9770242998943216, | |
| "learning_rate": 8.282565614028068e-06, | |
| "loss": 0.1187, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6941580756013745, | |
| "grad_norm": 1.080141238177633, | |
| "learning_rate": 8.23693142390914e-06, | |
| "loss": 0.0878, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.7010309278350515, | |
| "grad_norm": 1.1915299350233746, | |
| "learning_rate": 8.19082826171243e-06, | |
| "loss": 0.1092, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7079037800687286, | |
| "grad_norm": 0.8624719421655532, | |
| "learning_rate": 8.144262806941743e-06, | |
| "loss": 0.0692, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7147766323024055, | |
| "grad_norm": 1.3014093085974927, | |
| "learning_rate": 8.097241806078616e-06, | |
| "loss": 0.1296, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7216494845360825, | |
| "grad_norm": 0.846878040930432, | |
| "learning_rate": 8.049772071604864e-06, | |
| "loss": 0.0848, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7285223367697594, | |
| "grad_norm": 0.6768192450874961, | |
| "learning_rate": 8.001860481015594e-06, | |
| "loss": 0.0735, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7353951890034365, | |
| "grad_norm": 1.1323464816559536, | |
| "learning_rate": 7.953513975822755e-06, | |
| "loss": 0.1371, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7422680412371134, | |
| "grad_norm": 0.7423192404580958, | |
| "learning_rate": 7.904739560549475e-06, | |
| "loss": 0.0911, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7491408934707904, | |
| "grad_norm": 0.9299434209658686, | |
| "learning_rate": 7.855544301715203e-06, | |
| "loss": 0.0879, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7560137457044673, | |
| "grad_norm": 0.91270030975085, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.077, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7628865979381443, | |
| "grad_norm": 0.9539651579678919, | |
| "learning_rate": 7.755919823271466e-06, | |
| "loss": 0.0726, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7697594501718213, | |
| "grad_norm": 0.9640379563494964, | |
| "learning_rate": 7.70550503742427e-06, | |
| "loss": 0.0823, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7766323024054983, | |
| "grad_norm": 0.6328095164155321, | |
| "learning_rate": 7.654698273449435e-06, | |
| "loss": 0.0623, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7835051546391752, | |
| "grad_norm": 0.7652203603183867, | |
| "learning_rate": 7.603506892316513e-06, | |
| "loss": 0.0587, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7903780068728522, | |
| "grad_norm": 1.7296378525527818, | |
| "learning_rate": 7.551938310719043e-06, | |
| "loss": 0.1041, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7972508591065293, | |
| "grad_norm": 0.7460664712131734, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0722, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.8041237113402062, | |
| "grad_norm": 1.195858475129219, | |
| "learning_rate": 7.447699485069342e-06, | |
| "loss": 0.0904, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8109965635738832, | |
| "grad_norm": 0.9552329096581523, | |
| "learning_rate": 7.395044343313777e-06, | |
| "loss": 0.0693, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8178694158075601, | |
| "grad_norm": 0.7622938676223561, | |
| "learning_rate": 7.342042203498952e-06, | |
| "loss": 0.0696, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8247422680412371, | |
| "grad_norm": 0.9277140384697078, | |
| "learning_rate": 7.288700744664167e-06, | |
| "loss": 0.0704, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8316151202749141, | |
| "grad_norm": 0.7697629106626029, | |
| "learning_rate": 7.235027695009846e-06, | |
| "loss": 0.0719, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8384879725085911, | |
| "grad_norm": 1.9151590858697667, | |
| "learning_rate": 7.181030830777838e-06, | |
| "loss": 0.1, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.845360824742268, | |
| "grad_norm": 0.9981274438144712, | |
| "learning_rate": 7.1267179751248005e-06, | |
| "loss": 0.0801, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.852233676975945, | |
| "grad_norm": 1.3425710415118748, | |
| "learning_rate": 7.0720969969887595e-06, | |
| "loss": 0.0829, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8591065292096219, | |
| "grad_norm": 0.7814920365206511, | |
| "learning_rate": 7.017175809949044e-06, | |
| "loss": 0.0827, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.865979381443299, | |
| "grad_norm": 1.217261460811406, | |
| "learning_rate": 6.961962371079752e-06, | |
| "loss": 0.072, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.872852233676976, | |
| "grad_norm": 0.8032240602601177, | |
| "learning_rate": 6.906464679796927e-06, | |
| "loss": 0.0859, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8797250859106529, | |
| "grad_norm": 0.9419276342281664, | |
| "learning_rate": 6.850690776699574e-06, | |
| "loss": 0.0757, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8865979381443299, | |
| "grad_norm": 0.9566746568036009, | |
| "learning_rate": 6.79464874240473e-06, | |
| "loss": 0.1088, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8934707903780069, | |
| "grad_norm": 0.9178083438298893, | |
| "learning_rate": 6.7383466963767386e-06, | |
| "loss": 0.0703, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9003436426116839, | |
| "grad_norm": 0.8394840369711989, | |
| "learning_rate": 6.681792795750876e-06, | |
| "loss": 0.0702, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9072164948453608, | |
| "grad_norm": 1.8322819797593177, | |
| "learning_rate": 6.624995234151539e-06, | |
| "loss": 0.0808, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.9140893470790378, | |
| "grad_norm": 0.5714003628647666, | |
| "learning_rate": 6.567962240505136e-06, | |
| "loss": 0.079, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9209621993127147, | |
| "grad_norm": 0.8162504786944436, | |
| "learning_rate": 6.510702077847864e-06, | |
| "loss": 0.0945, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9278350515463918, | |
| "grad_norm": 0.6788704678980388, | |
| "learning_rate": 6.453223042128556e-06, | |
| "loss": 0.0663, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9347079037800687, | |
| "grad_norm": 1.5746889497037304, | |
| "learning_rate": 6.395533461006736e-06, | |
| "loss": 0.1125, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9415807560137457, | |
| "grad_norm": 0.7603985091604357, | |
| "learning_rate": 6.337641692646106e-06, | |
| "loss": 0.0914, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9484536082474226, | |
| "grad_norm": 1.1244427238295982, | |
| "learning_rate": 6.2795561245035895e-06, | |
| "loss": 0.0785, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9553264604810997, | |
| "grad_norm": 1.5054078109719562, | |
| "learning_rate": 6.221285172114156e-06, | |
| "loss": 0.0827, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9621993127147767, | |
| "grad_norm": 0.949378190148489, | |
| "learning_rate": 6.162837277871553e-06, | |
| "loss": 0.0898, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9690721649484536, | |
| "grad_norm": 1.077306228187391, | |
| "learning_rate": 6.104220909805162e-06, | |
| "loss": 0.1162, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9759450171821306, | |
| "grad_norm": 0.6702709078298128, | |
| "learning_rate": 6.045444560353136e-06, | |
| "loss": 0.0515, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9828178694158075, | |
| "grad_norm": 0.9633725685383875, | |
| "learning_rate": 5.986516745132e-06, | |
| "loss": 0.0536, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9896907216494846, | |
| "grad_norm": 1.128153783734316, | |
| "learning_rate": 5.927446001702899e-06, | |
| "loss": 0.0841, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "grad_norm": 0.9941723061028636, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.0922, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "eval_loss": 0.07502797245979309, | |
| "eval_runtime": 28.5693, | |
| "eval_samples_per_second": 34.268, | |
| "eval_steps_per_second": 1.085, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.0034364261168385, | |
| "grad_norm": 1.0675267441357827, | |
| "learning_rate": 5.808909982763825e-06, | |
| "loss": 0.0854, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0103092783505154, | |
| "grad_norm": 0.7559967796536462, | |
| "learning_rate": 5.749461880951966e-06, | |
| "loss": 0.0584, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.0171821305841924, | |
| "grad_norm": 0.5795147680930334, | |
| "learning_rate": 5.689905195840216e-06, | |
| "loss": 0.0778, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0240549828178693, | |
| "grad_norm": 1.0242981895015622, | |
| "learning_rate": 5.630248556101448e-06, | |
| "loss": 0.0789, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0309278350515463, | |
| "grad_norm": 0.9285740371989658, | |
| "learning_rate": 5.570500604890124e-06, | |
| "loss": 0.0729, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0378006872852235, | |
| "grad_norm": 0.7119548363541036, | |
| "learning_rate": 5.510669998590074e-06, | |
| "loss": 0.0657, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.0446735395189004, | |
| "grad_norm": 0.9294212007741496, | |
| "learning_rate": 5.450765405560328e-06, | |
| "loss": 0.064, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0515463917525774, | |
| "grad_norm": 0.7442894009599651, | |
| "learning_rate": 5.390795504879243e-06, | |
| "loss": 0.0672, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.0584192439862543, | |
| "grad_norm": 0.729509850397734, | |
| "learning_rate": 5.330768985087059e-06, | |
| "loss": 0.087, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0652920962199313, | |
| "grad_norm": 0.7437922495786078, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.0558, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0721649484536082, | |
| "grad_norm": 0.8807552383619118, | |
| "learning_rate": 5.2105808820857126e-06, | |
| "loss": 0.0958, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0790378006872852, | |
| "grad_norm": 0.6219072991202252, | |
| "learning_rate": 5.150436711931387e-06, | |
| "loss": 0.0585, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0859106529209621, | |
| "grad_norm": 0.9672380830787838, | |
| "learning_rate": 5.090270746252803e-06, | |
| "loss": 0.0962, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0927835051546393, | |
| "grad_norm": 0.6099050938904896, | |
| "learning_rate": 5.030091701996428e-06, | |
| "loss": 0.0508, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0996563573883162, | |
| "grad_norm": 0.6225939826011002, | |
| "learning_rate": 4.9699082980035735e-06, | |
| "loss": 0.067, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1065292096219932, | |
| "grad_norm": 0.6301779998982555, | |
| "learning_rate": 4.909729253747197e-06, | |
| "loss": 0.0584, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.1134020618556701, | |
| "grad_norm": 0.6786491849014926, | |
| "learning_rate": 4.8495632880686155e-06, | |
| "loss": 0.0726, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.120274914089347, | |
| "grad_norm": 0.546078218361975, | |
| "learning_rate": 4.789419117914288e-06, | |
| "loss": 0.0663, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.127147766323024, | |
| "grad_norm": 0.7566713547589681, | |
| "learning_rate": 4.729305457072913e-06, | |
| "loss": 0.0754, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.134020618556701, | |
| "grad_norm": 0.5160350874364323, | |
| "learning_rate": 4.6692310149129425e-06, | |
| "loss": 0.0466, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.140893470790378, | |
| "grad_norm": 0.5877790188393816, | |
| "learning_rate": 4.609204495120759e-06, | |
| "loss": 0.0562, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.147766323024055, | |
| "grad_norm": 0.6970503605178741, | |
| "learning_rate": 4.549234594439674e-06, | |
| "loss": 0.0624, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.1546391752577319, | |
| "grad_norm": 0.6127945342246314, | |
| "learning_rate": 4.489330001409929e-06, | |
| "loss": 0.051, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.161512027491409, | |
| "grad_norm": 0.6924950967155885, | |
| "learning_rate": 4.429499395109877e-06, | |
| "loss": 0.0624, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.168384879725086, | |
| "grad_norm": 0.5251106143987486, | |
| "learning_rate": 4.369751443898554e-06, | |
| "loss": 0.0596, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.175257731958763, | |
| "grad_norm": 0.8316397904844479, | |
| "learning_rate": 4.310094804159784e-06, | |
| "loss": 0.0716, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1821305841924399, | |
| "grad_norm": 1.38237374949498, | |
| "learning_rate": 4.250538119048036e-06, | |
| "loss": 0.0572, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1890034364261168, | |
| "grad_norm": 0.563828285344942, | |
| "learning_rate": 4.191090017236177e-06, | |
| "loss": 0.0545, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1958762886597938, | |
| "grad_norm": 1.0496809959075608, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.0799, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.2027491408934707, | |
| "grad_norm": 0.7766504701593869, | |
| "learning_rate": 4.072553998297103e-06, | |
| "loss": 0.0636, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2096219931271477, | |
| "grad_norm": 0.5747731360535787, | |
| "learning_rate": 4.013483254868001e-06, | |
| "loss": 0.0614, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.2164948453608249, | |
| "grad_norm": 0.6883523622770021, | |
| "learning_rate": 3.9545554396468655e-06, | |
| "loss": 0.057, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.2233676975945018, | |
| "grad_norm": 0.6560010415866142, | |
| "learning_rate": 3.89577909019484e-06, | |
| "loss": 0.0719, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.2302405498281788, | |
| "grad_norm": 0.5644514769878544, | |
| "learning_rate": 3.8371627221284495e-06, | |
| "loss": 0.051, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.2371134020618557, | |
| "grad_norm": 0.5816965929251859, | |
| "learning_rate": 3.7787148278858453e-06, | |
| "loss": 0.0559, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2439862542955327, | |
| "grad_norm": 0.6446728511010092, | |
| "learning_rate": 3.7204438754964113e-06, | |
| "loss": 0.0772, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2508591065292096, | |
| "grad_norm": 0.67531708146651, | |
| "learning_rate": 3.662358307353897e-06, | |
| "loss": 0.0577, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2577319587628866, | |
| "grad_norm": 0.40097643141265626, | |
| "learning_rate": 3.6044665389932663e-06, | |
| "loss": 0.0365, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.2646048109965635, | |
| "grad_norm": 0.743836665866088, | |
| "learning_rate": 3.5467769578714455e-06, | |
| "loss": 0.0602, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2714776632302405, | |
| "grad_norm": 0.5591161883903027, | |
| "learning_rate": 3.489297922152136e-06, | |
| "loss": 0.0577, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2783505154639174, | |
| "grad_norm": 0.6889937134145799, | |
| "learning_rate": 3.432037759494867e-06, | |
| "loss": 0.073, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2852233676975944, | |
| "grad_norm": 0.6411359802237454, | |
| "learning_rate": 3.375004765848463e-06, | |
| "loss": 0.0554, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.2920962199312716, | |
| "grad_norm": 0.7406887289301521, | |
| "learning_rate": 3.3182072042491244e-06, | |
| "loss": 0.063, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2989690721649485, | |
| "grad_norm": 0.493126630489134, | |
| "learning_rate": 3.2616533036232635e-06, | |
| "loss": 0.0519, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.3058419243986255, | |
| "grad_norm": 0.7699572928349039, | |
| "learning_rate": 3.205351257595272e-06, | |
| "loss": 0.0698, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3127147766323024, | |
| "grad_norm": 0.9906336303371214, | |
| "learning_rate": 3.149309223300428e-06, | |
| "loss": 0.0568, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.3195876288659794, | |
| "grad_norm": 0.6540355645281988, | |
| "learning_rate": 3.093535320203074e-06, | |
| "loss": 0.0652, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.3264604810996563, | |
| "grad_norm": 0.5420545975246776, | |
| "learning_rate": 3.0380376289202497e-06, | |
| "loss": 0.0428, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.5586315442153371, | |
| "learning_rate": 2.982824190050958e-06, | |
| "loss": 0.0432, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.3402061855670104, | |
| "grad_norm": 0.503875425194421, | |
| "learning_rate": 2.927903003011241e-06, | |
| "loss": 0.0473, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3470790378006874, | |
| "grad_norm": 0.7101301847965603, | |
| "learning_rate": 2.8732820248752016e-06, | |
| "loss": 0.0648, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3539518900343643, | |
| "grad_norm": 0.6858703791482932, | |
| "learning_rate": 2.8189691692221627e-06, | |
| "loss": 0.0658, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.3608247422680413, | |
| "grad_norm": 0.7937965079562438, | |
| "learning_rate": 2.7649723049901554e-06, | |
| "loss": 0.0704, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.3676975945017182, | |
| "grad_norm": 0.43280302442028307, | |
| "learning_rate": 2.711299255335833e-06, | |
| "loss": 0.0366, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3745704467353952, | |
| "grad_norm": 0.7451001772011392, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.0806, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3814432989690721, | |
| "grad_norm": 0.6614568404515468, | |
| "learning_rate": 2.6049556566862234e-06, | |
| "loss": 0.0661, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.388316151202749, | |
| "grad_norm": 0.6019245003183458, | |
| "learning_rate": 2.552300514930657e-06, | |
| "loss": 0.0529, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.395189003436426, | |
| "grad_norm": 0.6099283467115493, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0566, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.402061855670103, | |
| "grad_norm": 0.6195596904162101, | |
| "learning_rate": 2.4480616892809593e-06, | |
| "loss": 0.0626, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.40893470790378, | |
| "grad_norm": 0.6817862736314806, | |
| "learning_rate": 2.396493107683488e-06, | |
| "loss": 0.0611, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.4158075601374571, | |
| "grad_norm": 0.5586212737075882, | |
| "learning_rate": 2.345301726550567e-06, | |
| "loss": 0.0485, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.422680412371134, | |
| "grad_norm": 0.57471207165103, | |
| "learning_rate": 2.2944949625757295e-06, | |
| "loss": 0.0406, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.429553264604811, | |
| "grad_norm": 0.6280029000969929, | |
| "learning_rate": 2.244080176728536e-06, | |
| "loss": 0.0731, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.436426116838488, | |
| "grad_norm": 0.5967405629146709, | |
| "learning_rate": 2.1940646731880887e-06, | |
| "loss": 0.0571, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.443298969072165, | |
| "grad_norm": 0.6519920193637628, | |
| "learning_rate": 2.1444556982847996e-06, | |
| "loss": 0.0793, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.4501718213058419, | |
| "grad_norm": 0.5308173677668835, | |
| "learning_rate": 2.095260439450526e-06, | |
| "loss": 0.0533, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.4570446735395188, | |
| "grad_norm": 0.47521089405546657, | |
| "learning_rate": 2.0464860241772454e-06, | |
| "loss": 0.0455, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.463917525773196, | |
| "grad_norm": 0.7196757015917018, | |
| "learning_rate": 1.998139518984409e-06, | |
| "loss": 0.0655, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.470790378006873, | |
| "grad_norm": 0.7578568996685271, | |
| "learning_rate": 1.9502279283951363e-06, | |
| "loss": 0.0652, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.47766323024055, | |
| "grad_norm": 0.5563868752455253, | |
| "learning_rate": 1.9027581939213852e-06, | |
| "loss": 0.0611, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.4845360824742269, | |
| "grad_norm": 0.5611232659115866, | |
| "learning_rate": 1.8557371930582579e-06, | |
| "loss": 0.0519, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.4914089347079038, | |
| "grad_norm": 0.709777825103858, | |
| "learning_rate": 1.8091717382875723e-06, | |
| "loss": 0.0559, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.4982817869415808, | |
| "grad_norm": 0.4954980786326143, | |
| "learning_rate": 1.7630685760908623e-06, | |
| "loss": 0.0483, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.5051546391752577, | |
| "grad_norm": 0.8680025784552463, | |
| "learning_rate": 1.7174343859719334e-06, | |
| "loss": 0.0773, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.5120274914089347, | |
| "grad_norm": 0.6373721640689857, | |
| "learning_rate": 1.6722757794891287e-06, | |
| "loss": 0.0639, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.5189003436426116, | |
| "grad_norm": 0.6544994410975794, | |
| "learning_rate": 1.627599299297431e-06, | |
| "loss": 0.0611, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.5257731958762886, | |
| "grad_norm": 0.6026423600714211, | |
| "learning_rate": 1.5834114182005544e-06, | |
| "loss": 0.0573, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.5326460481099655, | |
| "grad_norm": 0.6263952935963489, | |
| "learning_rate": 1.5397185382131524e-06, | |
| "loss": 0.0508, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.5395189003436425, | |
| "grad_norm": 0.5026876941057207, | |
| "learning_rate": 1.4965269896332884e-06, | |
| "loss": 0.0456, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.5463917525773194, | |
| "grad_norm": 0.45771045886592204, | |
| "learning_rate": 1.4538430301252783e-06, | |
| "loss": 0.047, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5532646048109966, | |
| "grad_norm": 0.7604282035909598, | |
| "learning_rate": 1.411672843813086e-06, | |
| "loss": 0.0651, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.5601374570446735, | |
| "grad_norm": 0.4630886381664368, | |
| "learning_rate": 1.370022540384347e-06, | |
| "loss": 0.0438, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.5670103092783505, | |
| "grad_norm": 0.6909029301394074, | |
| "learning_rate": 1.3288981542051844e-06, | |
| "loss": 0.0784, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.5738831615120275, | |
| "grad_norm": 0.6056504664193658, | |
| "learning_rate": 1.2883056434459506e-06, | |
| "loss": 0.063, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.5807560137457046, | |
| "grad_norm": 0.6072713474055117, | |
| "learning_rate": 1.2482508892179884e-06, | |
| "loss": 0.0565, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.5876288659793816, | |
| "grad_norm": 0.7847452123419801, | |
| "learning_rate": 1.2087396947215678e-06, | |
| "loss": 0.061, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.5945017182130585, | |
| "grad_norm": 0.5965219896309601, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.0463, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.6013745704467355, | |
| "grad_norm": 0.5966866423456683, | |
| "learning_rate": 1.1313708031358183e-06, | |
| "loss": 0.0604, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.6082474226804124, | |
| "grad_norm": 0.6646757231823549, | |
| "learning_rate": 1.0935243153818437e-06, | |
| "loss": 0.0603, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.6151202749140894, | |
| "grad_norm": 0.4787688113371612, | |
| "learning_rate": 1.0562438044060846e-06, | |
| "loss": 0.0466, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.6219931271477663, | |
| "grad_norm": 0.48817556660139705, | |
| "learning_rate": 1.0195346714717813e-06, | |
| "loss": 0.0555, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.6288659793814433, | |
| "grad_norm": 0.47033672878262917, | |
| "learning_rate": 9.834022350599538e-07, | |
| "loss": 0.0545, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.6357388316151202, | |
| "grad_norm": 0.5718648979377676, | |
| "learning_rate": 9.47851730098856e-07, | |
| "loss": 0.0517, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.6426116838487972, | |
| "grad_norm": 0.673708749422117, | |
| "learning_rate": 9.128883072055411e-07, | |
| "loss": 0.0854, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.6494845360824741, | |
| "grad_norm": 0.9806647944888555, | |
| "learning_rate": 8.785170319396174e-07, | |
| "loss": 0.0491, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.656357388316151, | |
| "grad_norm": 0.5287987527224025, | |
| "learning_rate": 8.447428840693489e-07, | |
| "loss": 0.0436, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.663230240549828, | |
| "grad_norm": 0.6391342243584854, | |
| "learning_rate": 8.115707568501768e-07, | |
| "loss": 0.081, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.670103092783505, | |
| "grad_norm": 0.4615235145324941, | |
| "learning_rate": 7.790054563157745e-07, | |
| "loss": 0.0524, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.6769759450171822, | |
| "grad_norm": 0.510360884460692, | |
| "learning_rate": 7.470517005817473e-07, | |
| "loss": 0.0797, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.6838487972508591, | |
| "grad_norm": 0.4871992191684472, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.0489, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.690721649484536, | |
| "grad_norm": 0.8766518928954178, | |
| "learning_rate": 6.849972522982845e-07, | |
| "loss": 0.0752, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.697594501718213, | |
| "grad_norm": 0.5930604544104974, | |
| "learning_rate": 6.549055503018575e-07, | |
| "loss": 0.0527, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.7044673539518902, | |
| "grad_norm": 0.5722358432378138, | |
| "learning_rate": 6.254433729092518e-07, | |
| "loss": 0.0488, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.7113402061855671, | |
| "grad_norm": 0.4362939701006675, | |
| "learning_rate": 5.966149886503614e-07, | |
| "loss": 0.0356, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.718213058419244, | |
| "grad_norm": 0.6457455061812766, | |
| "learning_rate": 5.684245742300625e-07, | |
| "loss": 0.0504, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.725085910652921, | |
| "grad_norm": 0.5899957525444866, | |
| "learning_rate": 5.408762139230889e-07, | |
| "loss": 0.0603, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.731958762886598, | |
| "grad_norm": 0.599917192015849, | |
| "learning_rate": 5.139738989822901e-07, | |
| "loss": 0.0686, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.738831615120275, | |
| "grad_norm": 0.7380736554373928, | |
| "learning_rate": 4.877215270603752e-07, | |
| "loss": 0.0719, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.745704467353952, | |
| "grad_norm": 0.41832472335548054, | |
| "learning_rate": 4.6212290164521554e-07, | |
| "loss": 0.0317, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.7525773195876289, | |
| "grad_norm": 0.526464786255793, | |
| "learning_rate": 4.371817315087845e-07, | |
| "loss": 0.0771, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.7594501718213058, | |
| "grad_norm": 0.6484668629118039, | |
| "learning_rate": 4.1290163016982855e-07, | |
| "loss": 0.0557, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.7663230240549828, | |
| "grad_norm": 0.5396453549417995, | |
| "learning_rate": 3.8928611537033424e-07, | |
| "loss": 0.0482, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.7731958762886597, | |
| "grad_norm": 0.45306025216155027, | |
| "learning_rate": 3.663386085658693e-07, | |
| "loss": 0.0402, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.7800687285223367, | |
| "grad_norm": 0.5744128367152874, | |
| "learning_rate": 3.4406243442987765e-07, | |
| "loss": 0.0601, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.7869415807560136, | |
| "grad_norm": 0.7248071086031798, | |
| "learning_rate": 3.224608203719953e-07, | |
| "loss": 0.0624, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.7938144329896906, | |
| "grad_norm": 0.5679494867329458, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.0574, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.8006872852233677, | |
| "grad_norm": 0.5242391606919443, | |
| "learning_rate": 2.812936930186688e-07, | |
| "loss": 0.0572, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.8075601374570447, | |
| "grad_norm": 0.7894644774540441, | |
| "learning_rate": 2.617341440859883e-07, | |
| "loss": 0.0807, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.8144329896907216, | |
| "grad_norm": 0.37842092084304596, | |
| "learning_rate": 2.428610830928152e-07, | |
| "loss": 0.0353, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.8213058419243986, | |
| "grad_norm": 0.5630178323675307, | |
| "learning_rate": 2.2467724440002336e-07, | |
| "loss": 0.0456, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.8281786941580758, | |
| "grad_norm": 0.5816897863262482, | |
| "learning_rate": 2.0718526251279346e-07, | |
| "loss": 0.0667, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.8350515463917527, | |
| "grad_norm": 0.42953228387180065, | |
| "learning_rate": 1.9038767169893058e-07, | |
| "loss": 0.0446, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.8419243986254297, | |
| "grad_norm": 0.4721275130166019, | |
| "learning_rate": 1.7428690562169003e-07, | |
| "loss": 0.0451, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.8487972508591066, | |
| "grad_norm": 0.6400911159191205, | |
| "learning_rate": 1.5888529698718347e-07, | |
| "loss": 0.0539, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.8556701030927836, | |
| "grad_norm": 0.5474026213087981, | |
| "learning_rate": 1.4418507720641794e-07, | |
| "loss": 0.0552, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.8625429553264605, | |
| "grad_norm": 0.6746580293582902, | |
| "learning_rate": 1.3018837607199909e-07, | |
| "loss": 0.0744, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.8694158075601375, | |
| "grad_norm": 0.6047894387006907, | |
| "learning_rate": 1.1689722144956672e-07, | |
| "loss": 0.0613, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.8762886597938144, | |
| "grad_norm": 0.5078781852139993, | |
| "learning_rate": 1.0431353898399388e-07, | |
| "loss": 0.0754, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.8831615120274914, | |
| "grad_norm": 0.5527033239928246, | |
| "learning_rate": 9.243915182039431e-08, | |
| "loss": 0.0608, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.8900343642611683, | |
| "grad_norm": 0.475572453813436, | |
| "learning_rate": 8.127578033998663e-08, | |
| "loss": 0.0457, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.8969072164948453, | |
| "grad_norm": 0.6247550840412113, | |
| "learning_rate": 7.082504191083417e-08, | |
| "loss": 0.0602, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.9037800687285222, | |
| "grad_norm": 0.5866786337283754, | |
| "learning_rate": 6.108845065352864e-08, | |
| "loss": 0.0682, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.9106529209621992, | |
| "grad_norm": 0.4736651805754718, | |
| "learning_rate": 5.206741722181385e-08, | |
| "loss": 0.0443, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.9175257731958761, | |
| "grad_norm": 0.5703169109490406, | |
| "learning_rate": 4.376324859820924e-08, | |
| "loss": 0.0611, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.9243986254295533, | |
| "grad_norm": 0.6716852884879699, | |
| "learning_rate": 3.617714790465576e-08, | |
| "loss": 0.0477, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.9312714776632303, | |
| "grad_norm": 0.6566016257923214, | |
| "learning_rate": 2.9310214228202016e-08, | |
| "loss": 0.0654, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.9381443298969072, | |
| "grad_norm": 0.5853316999131423, | |
| "learning_rate": 2.3163442461766604e-08, | |
| "loss": 0.0697, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.9450171821305842, | |
| "grad_norm": 0.5909895151530212, | |
| "learning_rate": 1.7737723159999e-08, | |
| "loss": 0.0913, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.9518900343642611, | |
| "grad_norm": 0.5867551045464027, | |
| "learning_rate": 1.3033842410251074e-08, | |
| "loss": 0.056, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.9587628865979383, | |
| "grad_norm": 0.6306991634515368, | |
| "learning_rate": 9.052481718690998e-09, | |
| "loss": 0.0575, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.9656357388316152, | |
| "grad_norm": 0.5895175784613899, | |
| "learning_rate": 5.794217911562205e-09, | |
| "loss": 0.0529, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.9725085910652922, | |
| "grad_norm": 0.581263723967452, | |
| "learning_rate": 3.2595230516152543e-09, | |
| "loss": 0.0544, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.9793814432989691, | |
| "grad_norm": 0.5983267997992604, | |
| "learning_rate": 1.4487643697103092e-09, | |
| "loss": 0.0678, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.986254295532646, | |
| "grad_norm": 0.5903005504958891, | |
| "learning_rate": 3.6220421161692333e-10, | |
| "loss": 0.058, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "grad_norm": 1.4057031448356294, | |
| "learning_rate": 0.0, | |
| "loss": 0.0503, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "eval_loss": 0.061983052641153336, | |
| "eval_runtime": 28.8007, | |
| "eval_samples_per_second": 33.992, | |
| "eval_steps_per_second": 1.076, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "step": 290, | |
| "total_flos": 8.780012660627866e+16, | |
| "train_loss": 0.0829073357299484, | |
| "train_runtime": 3267.722, | |
| "train_samples_per_second": 11.383, | |
| "train_steps_per_second": 0.089 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 290, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "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": 8.780012660627866e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |