| { | |
| "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": 13.3212720911321, | |
| "learning_rate": 3.4482758620689656e-07, | |
| "loss": 0.3456, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013745704467353952, | |
| "grad_norm": 12.958681371116198, | |
| "learning_rate": 6.896551724137931e-07, | |
| "loss": 0.3318, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020618556701030927, | |
| "grad_norm": 11.877288541409978, | |
| "learning_rate": 1.0344827586206898e-06, | |
| "loss": 0.3139, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.027491408934707903, | |
| "grad_norm": 11.934788359892384, | |
| "learning_rate": 1.3793103448275862e-06, | |
| "loss": 0.2951, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03436426116838488, | |
| "grad_norm": 9.551821113312137, | |
| "learning_rate": 1.724137931034483e-06, | |
| "loss": 0.2588, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.041237113402061855, | |
| "grad_norm": 8.355467368276168, | |
| "learning_rate": 2.0689655172413796e-06, | |
| "loss": 0.2462, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.048109965635738834, | |
| "grad_norm": 5.473685606031438, | |
| "learning_rate": 2.4137931034482762e-06, | |
| "loss": 0.2065, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.054982817869415807, | |
| "grad_norm": 4.5522302413161455, | |
| "learning_rate": 2.7586206896551725e-06, | |
| "loss": 0.1985, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.061855670103092786, | |
| "grad_norm": 3.744560314691823, | |
| "learning_rate": 3.103448275862069e-06, | |
| "loss": 0.1525, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06872852233676977, | |
| "grad_norm": 2.4628345734657118, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 0.1425, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07560137457044673, | |
| "grad_norm": 2.3490817409711355, | |
| "learning_rate": 3.793103448275862e-06, | |
| "loss": 0.142, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08247422680412371, | |
| "grad_norm": 4.584368756890712, | |
| "learning_rate": 4.137931034482759e-06, | |
| "loss": 0.1628, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08934707903780069, | |
| "grad_norm": 3.365645141697699, | |
| "learning_rate": 4.482758620689656e-06, | |
| "loss": 0.1531, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09621993127147767, | |
| "grad_norm": 2.9448284481225198, | |
| "learning_rate": 4.8275862068965525e-06, | |
| "loss": 0.118, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10309278350515463, | |
| "grad_norm": 2.258723785423323, | |
| "learning_rate": 5.172413793103449e-06, | |
| "loss": 0.1524, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10996563573883161, | |
| "grad_norm": 3.3567741754931624, | |
| "learning_rate": 5.517241379310345e-06, | |
| "loss": 0.1317, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11683848797250859, | |
| "grad_norm": 2.4713196370512547, | |
| "learning_rate": 5.862068965517242e-06, | |
| "loss": 0.1536, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12371134020618557, | |
| "grad_norm": 1.996557596774797, | |
| "learning_rate": 6.206896551724138e-06, | |
| "loss": 0.1206, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.13058419243986255, | |
| "grad_norm": 1.8621865630224121, | |
| "learning_rate": 6.551724137931035e-06, | |
| "loss": 0.1291, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.13745704467353953, | |
| "grad_norm": 1.769950341230694, | |
| "learning_rate": 6.896551724137932e-06, | |
| "loss": 0.1034, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14432989690721648, | |
| "grad_norm": 2.4322334596844266, | |
| "learning_rate": 7.241379310344828e-06, | |
| "loss": 0.133, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.15120274914089346, | |
| "grad_norm": 1.5693033806121974, | |
| "learning_rate": 7.586206896551724e-06, | |
| "loss": 0.1022, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15807560137457044, | |
| "grad_norm": 1.9459673170896996, | |
| "learning_rate": 7.93103448275862e-06, | |
| "loss": 0.1318, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16494845360824742, | |
| "grad_norm": 1.4566658432856787, | |
| "learning_rate": 8.275862068965518e-06, | |
| "loss": 0.1137, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1718213058419244, | |
| "grad_norm": 1.9251726160074878, | |
| "learning_rate": 8.620689655172414e-06, | |
| "loss": 0.1115, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.17869415807560138, | |
| "grad_norm": 2.0941924654218016, | |
| "learning_rate": 8.965517241379312e-06, | |
| "loss": 0.1157, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18556701030927836, | |
| "grad_norm": 3.4828930087863683, | |
| "learning_rate": 9.310344827586207e-06, | |
| "loss": 0.1452, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.19243986254295534, | |
| "grad_norm": 2.379425844216661, | |
| "learning_rate": 9.655172413793105e-06, | |
| "loss": 0.1391, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19931271477663232, | |
| "grad_norm": 1.7598308540300598, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20618556701030927, | |
| "grad_norm": 3.001359569024646, | |
| "learning_rate": 9.999637795788383e-06, | |
| "loss": 0.1183, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.21305841924398625, | |
| "grad_norm": 1.7442000425281508, | |
| "learning_rate": 9.99855123563029e-06, | |
| "loss": 0.12, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.21993127147766323, | |
| "grad_norm": 1.4504988532629761, | |
| "learning_rate": 9.996740476948386e-06, | |
| "loss": 0.1105, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2268041237113402, | |
| "grad_norm": 2.5995490563765484, | |
| "learning_rate": 9.994205782088438e-06, | |
| "loss": 0.1162, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.23367697594501718, | |
| "grad_norm": 1.3378218622809885, | |
| "learning_rate": 9.990947518281312e-06, | |
| "loss": 0.1083, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.24054982817869416, | |
| "grad_norm": 2.2874401532742983, | |
| "learning_rate": 9.986966157589751e-06, | |
| "loss": 0.1633, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.24742268041237114, | |
| "grad_norm": 1.7616929257222467, | |
| "learning_rate": 9.982262276840002e-06, | |
| "loss": 0.1157, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2542955326460481, | |
| "grad_norm": 1.4132986417899638, | |
| "learning_rate": 9.976836557538234e-06, | |
| "loss": 0.1025, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2611683848797251, | |
| "grad_norm": 1.6769402932676176, | |
| "learning_rate": 9.970689785771798e-06, | |
| "loss": 0.1366, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.26804123711340205, | |
| "grad_norm": 1.9873258911800538, | |
| "learning_rate": 9.963822852095344e-06, | |
| "loss": 0.1211, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.27491408934707906, | |
| "grad_norm": 1.0683406048224036, | |
| "learning_rate": 9.95623675140179e-06, | |
| "loss": 0.1115, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.281786941580756, | |
| "grad_norm": 2.110818955428434, | |
| "learning_rate": 9.947932582778188e-06, | |
| "loss": 0.0938, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.28865979381443296, | |
| "grad_norm": 1.5144551861122317, | |
| "learning_rate": 9.938911549346473e-06, | |
| "loss": 0.105, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.29553264604810997, | |
| "grad_norm": 1.6183345144824617, | |
| "learning_rate": 9.929174958089167e-06, | |
| "loss": 0.1024, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3024054982817869, | |
| "grad_norm": 1.717413044388759, | |
| "learning_rate": 9.918724219660013e-06, | |
| "loss": 0.0849, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.30927835051546393, | |
| "grad_norm": 1.4114607425917451, | |
| "learning_rate": 9.907560848179607e-06, | |
| "loss": 0.0979, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.3161512027491409, | |
| "grad_norm": 2.1874561325633257, | |
| "learning_rate": 9.895686461016007e-06, | |
| "loss": 0.1083, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3230240549828179, | |
| "grad_norm": 1.4755317820251261, | |
| "learning_rate": 9.883102778550434e-06, | |
| "loss": 0.1166, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.32989690721649484, | |
| "grad_norm": 1.9328385696458805, | |
| "learning_rate": 9.869811623928001e-06, | |
| "loss": 0.1025, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.33676975945017185, | |
| "grad_norm": 1.3366337225602776, | |
| "learning_rate": 9.855814922793583e-06, | |
| "loss": 0.1035, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3436426116838488, | |
| "grad_norm": 1.5262659637018599, | |
| "learning_rate": 9.841114703012817e-06, | |
| "loss": 0.0666, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35051546391752575, | |
| "grad_norm": 2.4088580261403156, | |
| "learning_rate": 9.82571309437831e-06, | |
| "loss": 0.1227, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.35738831615120276, | |
| "grad_norm": 1.0908842292417193, | |
| "learning_rate": 9.809612328301071e-06, | |
| "loss": 0.1013, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3642611683848797, | |
| "grad_norm": 2.376838867176429, | |
| "learning_rate": 9.792814737487207e-06, | |
| "loss": 0.131, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3711340206185567, | |
| "grad_norm": 1.2875485281064492, | |
| "learning_rate": 9.775322755599979e-06, | |
| "loss": 0.0889, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.37800687285223367, | |
| "grad_norm": 1.2136830558516114, | |
| "learning_rate": 9.757138916907184e-06, | |
| "loss": 0.0896, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3848797250859107, | |
| "grad_norm": 1.576534974319604, | |
| "learning_rate": 9.738265855914014e-06, | |
| "loss": 0.1073, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3917525773195876, | |
| "grad_norm": 2.0345060251381306, | |
| "learning_rate": 9.718706306981332e-06, | |
| "loss": 0.1204, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.39862542955326463, | |
| "grad_norm": 1.1918137208389386, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.081, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4054982817869416, | |
| "grad_norm": 1.4709945613413247, | |
| "learning_rate": 9.677539179628005e-06, | |
| "loss": 0.1004, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.41237113402061853, | |
| "grad_norm": 1.077644135423854, | |
| "learning_rate": 9.655937565570124e-06, | |
| "loss": 0.074, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.41924398625429554, | |
| "grad_norm": 1.4665853526913675, | |
| "learning_rate": 9.63366139143413e-06, | |
| "loss": 0.1094, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4261168384879725, | |
| "grad_norm": 1.483794731081029, | |
| "learning_rate": 9.610713884629667e-06, | |
| "loss": 0.1019, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4329896907216495, | |
| "grad_norm": 1.85985064523168, | |
| "learning_rate": 9.587098369830171e-06, | |
| "loss": 0.1506, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.43986254295532645, | |
| "grad_norm": 0.8717445046770713, | |
| "learning_rate": 9.562818268491216e-06, | |
| "loss": 0.0931, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.44673539518900346, | |
| "grad_norm": 1.1945799878836783, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.0805, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4536082474226804, | |
| "grad_norm": 1.0017721418199987, | |
| "learning_rate": 9.512278472939627e-06, | |
| "loss": 0.0885, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.46048109965635736, | |
| "grad_norm": 0.7192105456453404, | |
| "learning_rate": 9.486026101017711e-06, | |
| "loss": 0.0713, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.46735395189003437, | |
| "grad_norm": 1.0221835163531388, | |
| "learning_rate": 9.459123786076911e-06, | |
| "loss": 0.0888, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4742268041237113, | |
| "grad_norm": 1.0671618820988402, | |
| "learning_rate": 9.431575425769938e-06, | |
| "loss": 0.0838, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.48109965635738833, | |
| "grad_norm": 0.8287825401204731, | |
| "learning_rate": 9.40338501134964e-06, | |
| "loss": 0.0642, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4879725085910653, | |
| "grad_norm": 1.1867222015595376, | |
| "learning_rate": 9.374556627090749e-06, | |
| "loss": 0.0965, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4948453608247423, | |
| "grad_norm": 1.1479418308995581, | |
| "learning_rate": 9.345094449698143e-06, | |
| "loss": 0.1104, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5017182130584192, | |
| "grad_norm": 0.7839485483129827, | |
| "learning_rate": 9.315002747701716e-06, | |
| "loss": 0.065, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5085910652920962, | |
| "grad_norm": 1.211265633850623, | |
| "learning_rate": 9.284285880837947e-06, | |
| "loss": 0.093, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5154639175257731, | |
| "grad_norm": 0.7281217369061034, | |
| "learning_rate": 9.252948299418255e-06, | |
| "loss": 0.0686, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5223367697594502, | |
| "grad_norm": 0.9444121990347742, | |
| "learning_rate": 9.220994543684225e-06, | |
| "loss": 0.0656, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5292096219931272, | |
| "grad_norm": 1.1295088846732502, | |
| "learning_rate": 9.188429243149824e-06, | |
| "loss": 0.0827, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5360824742268041, | |
| "grad_norm": 0.7896512599558635, | |
| "learning_rate": 9.155257115930651e-06, | |
| "loss": 0.0986, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5429553264604811, | |
| "grad_norm": 1.0626425992112776, | |
| "learning_rate": 9.121482968060384e-06, | |
| "loss": 0.0866, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5498281786941581, | |
| "grad_norm": 0.7713137582729137, | |
| "learning_rate": 9.08711169279446e-06, | |
| "loss": 0.0954, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5567010309278351, | |
| "grad_norm": 0.8602352869753656, | |
| "learning_rate": 9.052148269901145e-06, | |
| "loss": 0.0944, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.563573883161512, | |
| "grad_norm": 1.2535794648888932, | |
| "learning_rate": 9.01659776494005e-06, | |
| "loss": 0.1107, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.570446735395189, | |
| "grad_norm": 1.221263584214107, | |
| "learning_rate": 8.98046532852822e-06, | |
| "loss": 0.1053, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5773195876288659, | |
| "grad_norm": 1.1777792644348988, | |
| "learning_rate": 8.943756195593916e-06, | |
| "loss": 0.1267, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.584192439862543, | |
| "grad_norm": 0.8437132407592381, | |
| "learning_rate": 8.90647568461816e-06, | |
| "loss": 0.0882, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5910652920962199, | |
| "grad_norm": 0.8417806525797291, | |
| "learning_rate": 8.868629196864182e-06, | |
| "loss": 0.0826, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5979381443298969, | |
| "grad_norm": 1.1168993736515012, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.0964, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6048109965635738, | |
| "grad_norm": 0.9199626380950161, | |
| "learning_rate": 8.791260305278434e-06, | |
| "loss": 0.0945, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6116838487972509, | |
| "grad_norm": 1.157964246623041, | |
| "learning_rate": 8.751749110782013e-06, | |
| "loss": 0.0979, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6185567010309279, | |
| "grad_norm": 0.8403167737693835, | |
| "learning_rate": 8.71169435655405e-06, | |
| "loss": 0.0833, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6254295532646048, | |
| "grad_norm": 0.8099837213668893, | |
| "learning_rate": 8.671101845794816e-06, | |
| "loss": 0.0745, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6323024054982818, | |
| "grad_norm": 0.7318356995775042, | |
| "learning_rate": 8.629977459615655e-06, | |
| "loss": 0.0717, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6391752577319587, | |
| "grad_norm": 0.8129321300987981, | |
| "learning_rate": 8.588327156186915e-06, | |
| "loss": 0.068, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6460481099656358, | |
| "grad_norm": 0.7992703959217319, | |
| "learning_rate": 8.546156969874723e-06, | |
| "loss": 0.0793, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6529209621993127, | |
| "grad_norm": 0.8899200945249769, | |
| "learning_rate": 8.503473010366713e-06, | |
| "loss": 0.0745, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6597938144329897, | |
| "grad_norm": 0.8909651770613134, | |
| "learning_rate": 8.460281461786848e-06, | |
| "loss": 0.0947, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.8505785293265218, | |
| "learning_rate": 8.416588581799447e-06, | |
| "loss": 0.0765, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6735395189003437, | |
| "grad_norm": 0.9159164560152988, | |
| "learning_rate": 8.372400700702569e-06, | |
| "loss": 0.0749, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6804123711340206, | |
| "grad_norm": 1.2711864546522023, | |
| "learning_rate": 8.327724220510873e-06, | |
| "loss": 0.0903, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6872852233676976, | |
| "grad_norm": 0.9845383167528066, | |
| "learning_rate": 8.282565614028068e-06, | |
| "loss": 0.1089, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6941580756013745, | |
| "grad_norm": 1.0049260193910217, | |
| "learning_rate": 8.23693142390914e-06, | |
| "loss": 0.0748, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.7010309278350515, | |
| "grad_norm": 1.273713381300689, | |
| "learning_rate": 8.19082826171243e-06, | |
| "loss": 0.0865, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7079037800687286, | |
| "grad_norm": 0.8172345984781516, | |
| "learning_rate": 8.144262806941743e-06, | |
| "loss": 0.0646, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7147766323024055, | |
| "grad_norm": 1.2399079321061, | |
| "learning_rate": 8.097241806078616e-06, | |
| "loss": 0.0859, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7216494845360825, | |
| "grad_norm": 1.0994681478533508, | |
| "learning_rate": 8.049772071604864e-06, | |
| "loss": 0.068, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7285223367697594, | |
| "grad_norm": 0.9515711085109821, | |
| "learning_rate": 8.001860481015594e-06, | |
| "loss": 0.0872, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7353951890034365, | |
| "grad_norm": 1.2103014524807247, | |
| "learning_rate": 7.953513975822755e-06, | |
| "loss": 0.1116, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7422680412371134, | |
| "grad_norm": 2.185266887184931, | |
| "learning_rate": 7.904739560549475e-06, | |
| "loss": 0.0791, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7491408934707904, | |
| "grad_norm": 2.913813189194747, | |
| "learning_rate": 7.855544301715203e-06, | |
| "loss": 0.0981, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7560137457044673, | |
| "grad_norm": 1.5361719339078945, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.0881, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7628865979381443, | |
| "grad_norm": 0.9004552864810937, | |
| "learning_rate": 7.755919823271466e-06, | |
| "loss": 0.074, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7697594501718213, | |
| "grad_norm": 1.0576812536275024, | |
| "learning_rate": 7.70550503742427e-06, | |
| "loss": 0.0906, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7766323024054983, | |
| "grad_norm": 0.7304127516611955, | |
| "learning_rate": 7.654698273449435e-06, | |
| "loss": 0.0662, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7835051546391752, | |
| "grad_norm": 0.9771978931555205, | |
| "learning_rate": 7.603506892316513e-06, | |
| "loss": 0.0949, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7903780068728522, | |
| "grad_norm": 0.6737400018560286, | |
| "learning_rate": 7.551938310719043e-06, | |
| "loss": 0.0739, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7972508591065293, | |
| "grad_norm": 0.6031457228903015, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0674, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.8041237113402062, | |
| "grad_norm": 0.9295294397356566, | |
| "learning_rate": 7.447699485069342e-06, | |
| "loss": 0.0858, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8109965635738832, | |
| "grad_norm": 0.5271007796166037, | |
| "learning_rate": 7.395044343313777e-06, | |
| "loss": 0.0669, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8178694158075601, | |
| "grad_norm": 0.662283220386561, | |
| "learning_rate": 7.342042203498952e-06, | |
| "loss": 0.0786, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8247422680412371, | |
| "grad_norm": 0.7639083742294009, | |
| "learning_rate": 7.288700744664167e-06, | |
| "loss": 0.0852, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8316151202749141, | |
| "grad_norm": 0.7932696515177299, | |
| "learning_rate": 7.235027695009846e-06, | |
| "loss": 0.0638, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8384879725085911, | |
| "grad_norm": 0.8544979560076555, | |
| "learning_rate": 7.181030830777838e-06, | |
| "loss": 0.0744, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.845360824742268, | |
| "grad_norm": 0.8305756993246445, | |
| "learning_rate": 7.1267179751248005e-06, | |
| "loss": 0.0798, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.852233676975945, | |
| "grad_norm": 0.714773005542273, | |
| "learning_rate": 7.0720969969887595e-06, | |
| "loss": 0.075, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8591065292096219, | |
| "grad_norm": 1.0354940710100362, | |
| "learning_rate": 7.017175809949044e-06, | |
| "loss": 0.0592, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.865979381443299, | |
| "grad_norm": 0.8507208847523209, | |
| "learning_rate": 6.961962371079752e-06, | |
| "loss": 0.0854, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.872852233676976, | |
| "grad_norm": 1.5885463583332091, | |
| "learning_rate": 6.906464679796927e-06, | |
| "loss": 0.101, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8797250859106529, | |
| "grad_norm": 0.8946127989320589, | |
| "learning_rate": 6.850690776699574e-06, | |
| "loss": 0.0742, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8865979381443299, | |
| "grad_norm": 0.7692378432763433, | |
| "learning_rate": 6.79464874240473e-06, | |
| "loss": 0.0574, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8934707903780069, | |
| "grad_norm": 1.4401758305275527, | |
| "learning_rate": 6.7383466963767386e-06, | |
| "loss": 0.0953, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9003436426116839, | |
| "grad_norm": 0.7511591231830321, | |
| "learning_rate": 6.681792795750876e-06, | |
| "loss": 0.0697, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9072164948453608, | |
| "grad_norm": 1.262355477982067, | |
| "learning_rate": 6.624995234151539e-06, | |
| "loss": 0.1018, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.9140893470790378, | |
| "grad_norm": 0.6726613686052327, | |
| "learning_rate": 6.567962240505136e-06, | |
| "loss": 0.0627, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9209621993127147, | |
| "grad_norm": 0.9498545697578884, | |
| "learning_rate": 6.510702077847864e-06, | |
| "loss": 0.1074, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9278350515463918, | |
| "grad_norm": 0.6032012112831169, | |
| "learning_rate": 6.453223042128556e-06, | |
| "loss": 0.0572, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9347079037800687, | |
| "grad_norm": 0.551349601560956, | |
| "learning_rate": 6.395533461006736e-06, | |
| "loss": 0.0566, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9415807560137457, | |
| "grad_norm": 1.0262566425842523, | |
| "learning_rate": 6.337641692646106e-06, | |
| "loss": 0.0665, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9484536082474226, | |
| "grad_norm": 0.7634667928070298, | |
| "learning_rate": 6.2795561245035895e-06, | |
| "loss": 0.074, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9553264604810997, | |
| "grad_norm": 0.7875410492421236, | |
| "learning_rate": 6.221285172114156e-06, | |
| "loss": 0.0935, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9621993127147767, | |
| "grad_norm": 0.9923767887368872, | |
| "learning_rate": 6.162837277871553e-06, | |
| "loss": 0.1013, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9690721649484536, | |
| "grad_norm": 0.8544329856203985, | |
| "learning_rate": 6.104220909805162e-06, | |
| "loss": 0.0594, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9759450171821306, | |
| "grad_norm": 1.1847510081680237, | |
| "learning_rate": 6.045444560353136e-06, | |
| "loss": 0.1118, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9828178694158075, | |
| "grad_norm": 0.6139688879642499, | |
| "learning_rate": 5.986516745132e-06, | |
| "loss": 0.0648, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9896907216494846, | |
| "grad_norm": 0.8412241193957788, | |
| "learning_rate": 5.927446001702899e-06, | |
| "loss": 0.0775, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "grad_norm": 1.0848552501924613, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.0901, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "eval_loss": 0.0753212422132492, | |
| "eval_runtime": 28.7029, | |
| "eval_samples_per_second": 34.108, | |
| "eval_steps_per_second": 1.08, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.0034364261168385, | |
| "grad_norm": 0.7175430998112514, | |
| "learning_rate": 5.808909982763825e-06, | |
| "loss": 0.0769, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0103092783505154, | |
| "grad_norm": 0.7635725173663322, | |
| "learning_rate": 5.749461880951966e-06, | |
| "loss": 0.0688, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.0171821305841924, | |
| "grad_norm": 0.733165379468251, | |
| "learning_rate": 5.689905195840216e-06, | |
| "loss": 0.0556, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0240549828178693, | |
| "grad_norm": 0.7305441226703636, | |
| "learning_rate": 5.630248556101448e-06, | |
| "loss": 0.0554, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0309278350515463, | |
| "grad_norm": 0.9521428735397877, | |
| "learning_rate": 5.570500604890124e-06, | |
| "loss": 0.0927, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0378006872852235, | |
| "grad_norm": 0.5893208148451143, | |
| "learning_rate": 5.510669998590074e-06, | |
| "loss": 0.0672, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.0446735395189004, | |
| "grad_norm": 0.5468987540315196, | |
| "learning_rate": 5.450765405560328e-06, | |
| "loss": 0.0669, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0515463917525774, | |
| "grad_norm": 0.6372747586973702, | |
| "learning_rate": 5.390795504879243e-06, | |
| "loss": 0.0787, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.0584192439862543, | |
| "grad_norm": 0.672108462623344, | |
| "learning_rate": 5.330768985087059e-06, | |
| "loss": 0.0606, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0652920962199313, | |
| "grad_norm": 0.694472532063051, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.0583, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0721649484536082, | |
| "grad_norm": 0.6764307402036405, | |
| "learning_rate": 5.2105808820857126e-06, | |
| "loss": 0.0513, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0790378006872852, | |
| "grad_norm": 0.9765825532261199, | |
| "learning_rate": 5.150436711931387e-06, | |
| "loss": 0.0868, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0859106529209621, | |
| "grad_norm": 0.5991840115123475, | |
| "learning_rate": 5.090270746252803e-06, | |
| "loss": 0.0572, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0927835051546393, | |
| "grad_norm": 0.5550785456948635, | |
| "learning_rate": 5.030091701996428e-06, | |
| "loss": 0.0607, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0996563573883162, | |
| "grad_norm": 0.6902222482543674, | |
| "learning_rate": 4.9699082980035735e-06, | |
| "loss": 0.0678, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1065292096219932, | |
| "grad_norm": 0.6324920333026377, | |
| "learning_rate": 4.909729253747197e-06, | |
| "loss": 0.0694, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.1134020618556701, | |
| "grad_norm": 0.5981431742163907, | |
| "learning_rate": 4.8495632880686155e-06, | |
| "loss": 0.0589, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.120274914089347, | |
| "grad_norm": 0.6677383090541581, | |
| "learning_rate": 4.789419117914288e-06, | |
| "loss": 0.0644, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.127147766323024, | |
| "grad_norm": 0.7537930030735186, | |
| "learning_rate": 4.729305457072913e-06, | |
| "loss": 0.0663, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.134020618556701, | |
| "grad_norm": 0.9049644143825097, | |
| "learning_rate": 4.6692310149129425e-06, | |
| "loss": 0.0812, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.140893470790378, | |
| "grad_norm": 0.5967990080633419, | |
| "learning_rate": 4.609204495120759e-06, | |
| "loss": 0.0607, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.147766323024055, | |
| "grad_norm": 0.6908553707648403, | |
| "learning_rate": 4.549234594439674e-06, | |
| "loss": 0.0668, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.1546391752577319, | |
| "grad_norm": 0.606230661765985, | |
| "learning_rate": 4.489330001409929e-06, | |
| "loss": 0.06, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.161512027491409, | |
| "grad_norm": 0.6973811303136958, | |
| "learning_rate": 4.429499395109877e-06, | |
| "loss": 0.0582, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.168384879725086, | |
| "grad_norm": 0.7050289593868801, | |
| "learning_rate": 4.369751443898554e-06, | |
| "loss": 0.0773, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.175257731958763, | |
| "grad_norm": 0.5889397007461951, | |
| "learning_rate": 4.310094804159784e-06, | |
| "loss": 0.0698, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1821305841924399, | |
| "grad_norm": 0.5377482562216835, | |
| "learning_rate": 4.250538119048036e-06, | |
| "loss": 0.0508, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1890034364261168, | |
| "grad_norm": 0.4447112077848241, | |
| "learning_rate": 4.191090017236177e-06, | |
| "loss": 0.0535, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1958762886597938, | |
| "grad_norm": 0.771879188154358, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.0696, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.2027491408934707, | |
| "grad_norm": 0.5802486942887993, | |
| "learning_rate": 4.072553998297103e-06, | |
| "loss": 0.0565, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2096219931271477, | |
| "grad_norm": 0.6250470126696616, | |
| "learning_rate": 4.013483254868001e-06, | |
| "loss": 0.0801, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.2164948453608249, | |
| "grad_norm": 0.5505979949592014, | |
| "learning_rate": 3.9545554396468655e-06, | |
| "loss": 0.0491, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.2233676975945018, | |
| "grad_norm": 0.6351023543234251, | |
| "learning_rate": 3.89577909019484e-06, | |
| "loss": 0.0544, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.2302405498281788, | |
| "grad_norm": 0.5846041321034597, | |
| "learning_rate": 3.8371627221284495e-06, | |
| "loss": 0.0514, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.2371134020618557, | |
| "grad_norm": 0.8715356599811657, | |
| "learning_rate": 3.7787148278858453e-06, | |
| "loss": 0.0756, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2439862542955327, | |
| "grad_norm": 0.8007696300031172, | |
| "learning_rate": 3.7204438754964113e-06, | |
| "loss": 0.084, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2508591065292096, | |
| "grad_norm": 0.7871102534106393, | |
| "learning_rate": 3.662358307353897e-06, | |
| "loss": 0.0638, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2577319587628866, | |
| "grad_norm": 0.5945594698767217, | |
| "learning_rate": 3.6044665389932663e-06, | |
| "loss": 0.0616, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.2646048109965635, | |
| "grad_norm": 0.5069816980179496, | |
| "learning_rate": 3.5467769578714455e-06, | |
| "loss": 0.0406, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2714776632302405, | |
| "grad_norm": 1.1885480272250295, | |
| "learning_rate": 3.489297922152136e-06, | |
| "loss": 0.0821, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2783505154639174, | |
| "grad_norm": 0.6012387350957696, | |
| "learning_rate": 3.432037759494867e-06, | |
| "loss": 0.0539, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2852233676975944, | |
| "grad_norm": 0.6955985127281301, | |
| "learning_rate": 3.375004765848463e-06, | |
| "loss": 0.0649, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.2920962199312716, | |
| "grad_norm": 0.7231821268323338, | |
| "learning_rate": 3.3182072042491244e-06, | |
| "loss": 0.0704, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2989690721649485, | |
| "grad_norm": 0.7197043131415093, | |
| "learning_rate": 3.2616533036232635e-06, | |
| "loss": 0.0795, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.3058419243986255, | |
| "grad_norm": 0.6630067652167005, | |
| "learning_rate": 3.205351257595272e-06, | |
| "loss": 0.07, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3127147766323024, | |
| "grad_norm": 0.7640113849403117, | |
| "learning_rate": 3.149309223300428e-06, | |
| "loss": 0.0544, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.3195876288659794, | |
| "grad_norm": 0.5595561786164991, | |
| "learning_rate": 3.093535320203074e-06, | |
| "loss": 0.0715, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.3264604810996563, | |
| "grad_norm": 0.560088426203932, | |
| "learning_rate": 3.0380376289202497e-06, | |
| "loss": 0.0615, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.7889457245575039, | |
| "learning_rate": 2.982824190050958e-06, | |
| "loss": 0.0552, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.3402061855670104, | |
| "grad_norm": 0.6976053732068416, | |
| "learning_rate": 2.927903003011241e-06, | |
| "loss": 0.0769, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3470790378006874, | |
| "grad_norm": 0.7394241511136698, | |
| "learning_rate": 2.8732820248752016e-06, | |
| "loss": 0.0581, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3539518900343643, | |
| "grad_norm": 0.6396367368191475, | |
| "learning_rate": 2.8189691692221627e-06, | |
| "loss": 0.0616, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.3608247422680413, | |
| "grad_norm": 0.7188210637821097, | |
| "learning_rate": 2.7649723049901554e-06, | |
| "loss": 0.0685, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.3676975945017182, | |
| "grad_norm": 0.7877132751787733, | |
| "learning_rate": 2.711299255335833e-06, | |
| "loss": 0.0929, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3745704467353952, | |
| "grad_norm": 0.4760142241002996, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.0579, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3814432989690721, | |
| "grad_norm": 0.639476714245797, | |
| "learning_rate": 2.6049556566862234e-06, | |
| "loss": 0.0455, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.388316151202749, | |
| "grad_norm": 0.5392504104764927, | |
| "learning_rate": 2.552300514930657e-06, | |
| "loss": 0.0656, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.395189003436426, | |
| "grad_norm": 0.6004991091437839, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0467, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.402061855670103, | |
| "grad_norm": 0.8035110316302986, | |
| "learning_rate": 2.4480616892809593e-06, | |
| "loss": 0.0786, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.40893470790378, | |
| "grad_norm": 0.547480768144687, | |
| "learning_rate": 2.396493107683488e-06, | |
| "loss": 0.0436, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.4158075601374571, | |
| "grad_norm": 0.6260668063409571, | |
| "learning_rate": 2.345301726550567e-06, | |
| "loss": 0.0747, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.422680412371134, | |
| "grad_norm": 0.66007636371719, | |
| "learning_rate": 2.2944949625757295e-06, | |
| "loss": 0.0671, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.429553264604811, | |
| "grad_norm": 0.5214800174539683, | |
| "learning_rate": 2.244080176728536e-06, | |
| "loss": 0.0538, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.436426116838488, | |
| "grad_norm": 0.6638761160975902, | |
| "learning_rate": 2.1940646731880887e-06, | |
| "loss": 0.058, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.443298969072165, | |
| "grad_norm": 0.6360623402473407, | |
| "learning_rate": 2.1444556982847996e-06, | |
| "loss": 0.0614, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.4501718213058419, | |
| "grad_norm": 0.7428102060833474, | |
| "learning_rate": 2.095260439450526e-06, | |
| "loss": 0.0792, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.4570446735395188, | |
| "grad_norm": 0.6425090520167832, | |
| "learning_rate": 2.0464860241772454e-06, | |
| "loss": 0.0515, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.463917525773196, | |
| "grad_norm": 0.6206839031503566, | |
| "learning_rate": 1.998139518984409e-06, | |
| "loss": 0.0585, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.470790378006873, | |
| "grad_norm": 0.6744798883952051, | |
| "learning_rate": 1.9502279283951363e-06, | |
| "loss": 0.0624, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.47766323024055, | |
| "grad_norm": 0.6542937992910209, | |
| "learning_rate": 1.9027581939213852e-06, | |
| "loss": 0.056, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.4845360824742269, | |
| "grad_norm": 0.7247929965504191, | |
| "learning_rate": 1.8557371930582579e-06, | |
| "loss": 0.0624, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.4914089347079038, | |
| "grad_norm": 1.0835234231386042, | |
| "learning_rate": 1.8091717382875723e-06, | |
| "loss": 0.0785, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.4982817869415808, | |
| "grad_norm": 0.7662317487774777, | |
| "learning_rate": 1.7630685760908623e-06, | |
| "loss": 0.0494, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.5051546391752577, | |
| "grad_norm": 0.5356075105219417, | |
| "learning_rate": 1.7174343859719334e-06, | |
| "loss": 0.0415, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.5120274914089347, | |
| "grad_norm": 0.6165176924215153, | |
| "learning_rate": 1.6722757794891287e-06, | |
| "loss": 0.0462, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.5189003436426116, | |
| "grad_norm": 0.5165347845307708, | |
| "learning_rate": 1.627599299297431e-06, | |
| "loss": 0.0537, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.5257731958762886, | |
| "grad_norm": 0.579812378593318, | |
| "learning_rate": 1.5834114182005544e-06, | |
| "loss": 0.0427, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.5326460481099655, | |
| "grad_norm": 0.9556255485824591, | |
| "learning_rate": 1.5397185382131524e-06, | |
| "loss": 0.0617, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.5395189003436425, | |
| "grad_norm": 0.6744931740254627, | |
| "learning_rate": 1.4965269896332884e-06, | |
| "loss": 0.0558, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.5463917525773194, | |
| "grad_norm": 0.6447562781278031, | |
| "learning_rate": 1.4538430301252783e-06, | |
| "loss": 0.0587, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5532646048109966, | |
| "grad_norm": 0.6647082761702984, | |
| "learning_rate": 1.411672843813086e-06, | |
| "loss": 0.0712, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.5601374570446735, | |
| "grad_norm": 1.0218872200571598, | |
| "learning_rate": 1.370022540384347e-06, | |
| "loss": 0.0612, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.5670103092783505, | |
| "grad_norm": 0.7148523076036994, | |
| "learning_rate": 1.3288981542051844e-06, | |
| "loss": 0.0589, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.5738831615120275, | |
| "grad_norm": 0.6548432683623916, | |
| "learning_rate": 1.2883056434459506e-06, | |
| "loss": 0.0589, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.5807560137457046, | |
| "grad_norm": 0.7376046714204466, | |
| "learning_rate": 1.2482508892179884e-06, | |
| "loss": 0.0636, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.5876288659793816, | |
| "grad_norm": 0.8897784305567501, | |
| "learning_rate": 1.2087396947215678e-06, | |
| "loss": 0.0815, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.5945017182130585, | |
| "grad_norm": 0.6381283594188061, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.0573, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.6013745704467355, | |
| "grad_norm": 0.5251101961548645, | |
| "learning_rate": 1.1313708031358183e-06, | |
| "loss": 0.041, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.6082474226804124, | |
| "grad_norm": 0.5175707122863319, | |
| "learning_rate": 1.0935243153818437e-06, | |
| "loss": 0.0462, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.6151202749140894, | |
| "grad_norm": 0.6800532405363695, | |
| "learning_rate": 1.0562438044060846e-06, | |
| "loss": 0.0715, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.6219931271477663, | |
| "grad_norm": 0.5142363765278898, | |
| "learning_rate": 1.0195346714717813e-06, | |
| "loss": 0.0478, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.6288659793814433, | |
| "grad_norm": 0.535494294769509, | |
| "learning_rate": 9.834022350599538e-07, | |
| "loss": 0.0466, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.6357388316151202, | |
| "grad_norm": 0.5920784807206897, | |
| "learning_rate": 9.47851730098856e-07, | |
| "loss": 0.0844, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.6426116838487972, | |
| "grad_norm": 0.5114000655183499, | |
| "learning_rate": 9.128883072055411e-07, | |
| "loss": 0.0504, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.6494845360824741, | |
| "grad_norm": 0.675964152742232, | |
| "learning_rate": 8.785170319396174e-07, | |
| "loss": 0.0656, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.656357388316151, | |
| "grad_norm": 0.5770540028655016, | |
| "learning_rate": 8.447428840693489e-07, | |
| "loss": 0.0597, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.663230240549828, | |
| "grad_norm": 0.569457804018713, | |
| "learning_rate": 8.115707568501768e-07, | |
| "loss": 0.0475, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.670103092783505, | |
| "grad_norm": 0.37086873752185795, | |
| "learning_rate": 7.790054563157745e-07, | |
| "loss": 0.0305, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.6769759450171822, | |
| "grad_norm": 0.7065856221971984, | |
| "learning_rate": 7.470517005817473e-07, | |
| "loss": 0.0685, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.6838487972508591, | |
| "grad_norm": 0.7820749712663027, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.0834, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.690721649484536, | |
| "grad_norm": 0.47871051657794167, | |
| "learning_rate": 6.849972522982845e-07, | |
| "loss": 0.0469, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.697594501718213, | |
| "grad_norm": 0.5295985844805942, | |
| "learning_rate": 6.549055503018575e-07, | |
| "loss": 0.0491, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.7044673539518902, | |
| "grad_norm": 0.5585924551562599, | |
| "learning_rate": 6.254433729092518e-07, | |
| "loss": 0.0481, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.7113402061855671, | |
| "grad_norm": 0.7113699927585905, | |
| "learning_rate": 5.966149886503614e-07, | |
| "loss": 0.0672, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.718213058419244, | |
| "grad_norm": 0.5999675737771197, | |
| "learning_rate": 5.684245742300625e-07, | |
| "loss": 0.0578, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.725085910652921, | |
| "grad_norm": 0.4401362685808288, | |
| "learning_rate": 5.408762139230889e-07, | |
| "loss": 0.0411, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.731958762886598, | |
| "grad_norm": 0.6143801274432039, | |
| "learning_rate": 5.139738989822901e-07, | |
| "loss": 0.047, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.738831615120275, | |
| "grad_norm": 0.6588448327695987, | |
| "learning_rate": 4.877215270603752e-07, | |
| "loss": 0.069, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.745704467353952, | |
| "grad_norm": 0.5698713302800225, | |
| "learning_rate": 4.6212290164521554e-07, | |
| "loss": 0.0575, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.7525773195876289, | |
| "grad_norm": 0.6462129492339118, | |
| "learning_rate": 4.371817315087845e-07, | |
| "loss": 0.0625, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.7594501718213058, | |
| "grad_norm": 0.5733205388665175, | |
| "learning_rate": 4.1290163016982855e-07, | |
| "loss": 0.0737, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.7663230240549828, | |
| "grad_norm": 0.430592511164149, | |
| "learning_rate": 3.8928611537033424e-07, | |
| "loss": 0.048, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.7731958762886597, | |
| "grad_norm": 0.4461742116699663, | |
| "learning_rate": 3.663386085658693e-07, | |
| "loss": 0.0415, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.7800687285223367, | |
| "grad_norm": 0.6098985197143705, | |
| "learning_rate": 3.4406243442987765e-07, | |
| "loss": 0.0482, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.7869415807560136, | |
| "grad_norm": 0.49250028542951824, | |
| "learning_rate": 3.224608203719953e-07, | |
| "loss": 0.0453, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.7938144329896906, | |
| "grad_norm": 0.6416046190344225, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.0496, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.8006872852233677, | |
| "grad_norm": 0.3808066761396319, | |
| "learning_rate": 2.812936930186688e-07, | |
| "loss": 0.0333, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.8075601374570447, | |
| "grad_norm": 0.7544658741931779, | |
| "learning_rate": 2.617341440859883e-07, | |
| "loss": 0.0541, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.8144329896907216, | |
| "grad_norm": 0.5502995110481443, | |
| "learning_rate": 2.428610830928152e-07, | |
| "loss": 0.0427, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.8213058419243986, | |
| "grad_norm": 0.5640528752854622, | |
| "learning_rate": 2.2467724440002336e-07, | |
| "loss": 0.0526, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.8281786941580758, | |
| "grad_norm": 0.7008839129623861, | |
| "learning_rate": 2.0718526251279346e-07, | |
| "loss": 0.0588, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.8350515463917527, | |
| "grad_norm": 0.5191134911978924, | |
| "learning_rate": 1.9038767169893058e-07, | |
| "loss": 0.0535, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.8419243986254297, | |
| "grad_norm": 0.6127204469247451, | |
| "learning_rate": 1.7428690562169003e-07, | |
| "loss": 0.05, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.8487972508591066, | |
| "grad_norm": 0.5979559816370372, | |
| "learning_rate": 1.5888529698718347e-07, | |
| "loss": 0.0515, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.8556701030927836, | |
| "grad_norm": 0.6359619454024678, | |
| "learning_rate": 1.4418507720641794e-07, | |
| "loss": 0.0794, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.8625429553264605, | |
| "grad_norm": 0.7605632363819157, | |
| "learning_rate": 1.3018837607199909e-07, | |
| "loss": 0.0706, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.8694158075601375, | |
| "grad_norm": 0.6574106929513626, | |
| "learning_rate": 1.1689722144956672e-07, | |
| "loss": 0.0677, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.8762886597938144, | |
| "grad_norm": 0.5787071787955073, | |
| "learning_rate": 1.0431353898399388e-07, | |
| "loss": 0.0617, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.8831615120274914, | |
| "grad_norm": 0.541809569878409, | |
| "learning_rate": 9.243915182039431e-08, | |
| "loss": 0.0421, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.8900343642611683, | |
| "grad_norm": 0.4912688280907106, | |
| "learning_rate": 8.127578033998663e-08, | |
| "loss": 0.0483, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.8969072164948453, | |
| "grad_norm": 0.6038671152995181, | |
| "learning_rate": 7.082504191083417e-08, | |
| "loss": 0.0676, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.9037800687285222, | |
| "grad_norm": 0.5945898718267285, | |
| "learning_rate": 6.108845065352864e-08, | |
| "loss": 0.0503, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.9106529209621992, | |
| "grad_norm": 0.6153368833549625, | |
| "learning_rate": 5.206741722181385e-08, | |
| "loss": 0.0583, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.9175257731958761, | |
| "grad_norm": 0.4725265143345455, | |
| "learning_rate": 4.376324859820924e-08, | |
| "loss": 0.0648, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.9243986254295533, | |
| "grad_norm": 0.6523832010828569, | |
| "learning_rate": 3.617714790465576e-08, | |
| "loss": 0.0845, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.9312714776632303, | |
| "grad_norm": 0.5670700872222558, | |
| "learning_rate": 2.9310214228202016e-08, | |
| "loss": 0.0514, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.9381443298969072, | |
| "grad_norm": 0.39518011108454415, | |
| "learning_rate": 2.3163442461766604e-08, | |
| "loss": 0.0444, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.9450171821305842, | |
| "grad_norm": 0.5864232604012668, | |
| "learning_rate": 1.7737723159999e-08, | |
| "loss": 0.0512, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.9518900343642611, | |
| "grad_norm": 0.6007564271469487, | |
| "learning_rate": 1.3033842410251074e-08, | |
| "loss": 0.0603, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.9587628865979383, | |
| "grad_norm": 0.4414010732761918, | |
| "learning_rate": 9.052481718690998e-09, | |
| "loss": 0.0407, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.9656357388316152, | |
| "grad_norm": 0.6508158814182216, | |
| "learning_rate": 5.794217911562205e-09, | |
| "loss": 0.0506, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.9725085910652922, | |
| "grad_norm": 0.5061394742716212, | |
| "learning_rate": 3.2595230516152543e-09, | |
| "loss": 0.0496, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.9793814432989691, | |
| "grad_norm": 0.47692416766008067, | |
| "learning_rate": 1.4487643697103092e-09, | |
| "loss": 0.0479, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.986254295532646, | |
| "grad_norm": 0.4975677174872142, | |
| "learning_rate": 3.6220421161692333e-10, | |
| "loss": 0.0516, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "grad_norm": 0.49794283350387364, | |
| "learning_rate": 0.0, | |
| "loss": 0.0535, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "eval_loss": 0.06284079700708389, | |
| "eval_runtime": 28.9301, | |
| "eval_samples_per_second": 33.84, | |
| "eval_steps_per_second": 1.072, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "step": 290, | |
| "total_flos": 8.746570234493338e+16, | |
| "train_loss": 0.08327665239700983, | |
| "train_runtime": 3248.3148, | |
| "train_samples_per_second": 11.451, | |
| "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": 8.746570234493338e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |