| { | |
| "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": 12.421529138279539, | |
| "learning_rate": 3.4482758620689656e-07, | |
| "loss": 0.3232, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.013745704467353952, | |
| "grad_norm": 12.633576042619561, | |
| "learning_rate": 6.896551724137931e-07, | |
| "loss": 0.3344, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.020618556701030927, | |
| "grad_norm": 12.52841629989842, | |
| "learning_rate": 1.0344827586206898e-06, | |
| "loss": 0.3193, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.027491408934707903, | |
| "grad_norm": 12.813451130543717, | |
| "learning_rate": 1.3793103448275862e-06, | |
| "loss": 0.3427, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.03436426116838488, | |
| "grad_norm": 9.46513292504884, | |
| "learning_rate": 1.724137931034483e-06, | |
| "loss": 0.2603, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.041237113402061855, | |
| "grad_norm": 8.101740527012359, | |
| "learning_rate": 2.0689655172413796e-06, | |
| "loss": 0.2338, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.048109965635738834, | |
| "grad_norm": 5.742146611960234, | |
| "learning_rate": 2.4137931034482762e-06, | |
| "loss": 0.157, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.054982817869415807, | |
| "grad_norm": 5.31090478422663, | |
| "learning_rate": 2.7586206896551725e-06, | |
| "loss": 0.182, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.061855670103092786, | |
| "grad_norm": 4.680601247416956, | |
| "learning_rate": 3.103448275862069e-06, | |
| "loss": 0.1879, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.06872852233676977, | |
| "grad_norm": 2.9571495390257305, | |
| "learning_rate": 3.448275862068966e-06, | |
| "loss": 0.1803, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07560137457044673, | |
| "grad_norm": 2.4551047047678294, | |
| "learning_rate": 3.793103448275862e-06, | |
| "loss": 0.1757, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.08247422680412371, | |
| "grad_norm": 8.98100766782742, | |
| "learning_rate": 4.137931034482759e-06, | |
| "loss": 0.1532, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.08934707903780069, | |
| "grad_norm": 3.515617704354926, | |
| "learning_rate": 4.482758620689656e-06, | |
| "loss": 0.1457, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.09621993127147767, | |
| "grad_norm": 3.2166457055264623, | |
| "learning_rate": 4.8275862068965525e-06, | |
| "loss": 0.1124, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.10309278350515463, | |
| "grad_norm": 2.273532995532905, | |
| "learning_rate": 5.172413793103449e-06, | |
| "loss": 0.1365, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.10996563573883161, | |
| "grad_norm": 4.307987873000315, | |
| "learning_rate": 5.517241379310345e-06, | |
| "loss": 0.1164, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.11683848797250859, | |
| "grad_norm": 2.3706462680428664, | |
| "learning_rate": 5.862068965517242e-06, | |
| "loss": 0.157, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.12371134020618557, | |
| "grad_norm": 2.1051815371551457, | |
| "learning_rate": 6.206896551724138e-06, | |
| "loss": 0.1316, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.13058419243986255, | |
| "grad_norm": 3.633176502594214, | |
| "learning_rate": 6.551724137931035e-06, | |
| "loss": 0.1378, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.13745704467353953, | |
| "grad_norm": 2.610332443933833, | |
| "learning_rate": 6.896551724137932e-06, | |
| "loss": 0.124, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.14432989690721648, | |
| "grad_norm": 2.830261898289799, | |
| "learning_rate": 7.241379310344828e-06, | |
| "loss": 0.1454, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.15120274914089346, | |
| "grad_norm": 2.7104098787391857, | |
| "learning_rate": 7.586206896551724e-06, | |
| "loss": 0.1307, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.15807560137457044, | |
| "grad_norm": 2.505038752773409, | |
| "learning_rate": 7.93103448275862e-06, | |
| "loss": 0.1061, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.16494845360824742, | |
| "grad_norm": 3.6721151931052844, | |
| "learning_rate": 8.275862068965518e-06, | |
| "loss": 0.1143, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1718213058419244, | |
| "grad_norm": 2.659524591497108, | |
| "learning_rate": 8.620689655172414e-06, | |
| "loss": 0.1233, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.17869415807560138, | |
| "grad_norm": 2.8615771259203906, | |
| "learning_rate": 8.965517241379312e-06, | |
| "loss": 0.1314, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.18556701030927836, | |
| "grad_norm": 2.1350791511443297, | |
| "learning_rate": 9.310344827586207e-06, | |
| "loss": 0.1113, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.19243986254295534, | |
| "grad_norm": 2.208041568296393, | |
| "learning_rate": 9.655172413793105e-06, | |
| "loss": 0.1025, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.19931271477663232, | |
| "grad_norm": 1.574248615675016, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1084, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.20618556701030927, | |
| "grad_norm": 3.1522543380808976, | |
| "learning_rate": 9.999637795788383e-06, | |
| "loss": 0.1128, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.21305841924398625, | |
| "grad_norm": 3.060768022564431, | |
| "learning_rate": 9.99855123563029e-06, | |
| "loss": 0.1281, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.21993127147766323, | |
| "grad_norm": 2.68703045217427, | |
| "learning_rate": 9.996740476948386e-06, | |
| "loss": 0.1127, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2268041237113402, | |
| "grad_norm": 4.882529589578716, | |
| "learning_rate": 9.994205782088438e-06, | |
| "loss": 0.1301, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.23367697594501718, | |
| "grad_norm": 1.9885707777805819, | |
| "learning_rate": 9.990947518281312e-06, | |
| "loss": 0.0974, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.24054982817869416, | |
| "grad_norm": 1.8564657230418518, | |
| "learning_rate": 9.986966157589751e-06, | |
| "loss": 0.093, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.24742268041237114, | |
| "grad_norm": 1.1750577484064841, | |
| "learning_rate": 9.982262276840002e-06, | |
| "loss": 0.0807, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.2542955326460481, | |
| "grad_norm": 2.504971561950099, | |
| "learning_rate": 9.976836557538234e-06, | |
| "loss": 0.1049, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2611683848797251, | |
| "grad_norm": 1.441069196366465, | |
| "learning_rate": 9.970689785771798e-06, | |
| "loss": 0.1331, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.26804123711340205, | |
| "grad_norm": 2.5170550133239367, | |
| "learning_rate": 9.963822852095344e-06, | |
| "loss": 0.1144, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.27491408934707906, | |
| "grad_norm": 1.8282354446820654, | |
| "learning_rate": 9.95623675140179e-06, | |
| "loss": 0.116, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.281786941580756, | |
| "grad_norm": 1.6885941444419086, | |
| "learning_rate": 9.947932582778188e-06, | |
| "loss": 0.0898, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.28865979381443296, | |
| "grad_norm": 2.2255942278019463, | |
| "learning_rate": 9.938911549346473e-06, | |
| "loss": 0.1064, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.29553264604810997, | |
| "grad_norm": 1.4470307882214384, | |
| "learning_rate": 9.929174958089167e-06, | |
| "loss": 0.0729, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.3024054982817869, | |
| "grad_norm": 1.6459878437104352, | |
| "learning_rate": 9.918724219660013e-06, | |
| "loss": 0.1036, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.30927835051546393, | |
| "grad_norm": 1.6529101148455856, | |
| "learning_rate": 9.907560848179607e-06, | |
| "loss": 0.0889, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.3161512027491409, | |
| "grad_norm": 2.4000159447896303, | |
| "learning_rate": 9.895686461016007e-06, | |
| "loss": 0.1001, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3230240549828179, | |
| "grad_norm": 2.005085868078671, | |
| "learning_rate": 9.883102778550434e-06, | |
| "loss": 0.1013, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.32989690721649484, | |
| "grad_norm": 1.3949556740262812, | |
| "learning_rate": 9.869811623928001e-06, | |
| "loss": 0.1246, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.33676975945017185, | |
| "grad_norm": 1.5821860565503276, | |
| "learning_rate": 9.855814922793583e-06, | |
| "loss": 0.1205, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3436426116838488, | |
| "grad_norm": 1.2130580748768462, | |
| "learning_rate": 9.841114703012817e-06, | |
| "loss": 0.1048, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.35051546391752575, | |
| "grad_norm": 1.2213973293871836, | |
| "learning_rate": 9.82571309437831e-06, | |
| "loss": 0.0833, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.35738831615120276, | |
| "grad_norm": 1.3965900033524388, | |
| "learning_rate": 9.809612328301071e-06, | |
| "loss": 0.0866, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.3642611683848797, | |
| "grad_norm": 2.7746813444777776, | |
| "learning_rate": 9.792814737487207e-06, | |
| "loss": 0.1085, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3711340206185567, | |
| "grad_norm": 1.6820653399143763, | |
| "learning_rate": 9.775322755599979e-06, | |
| "loss": 0.101, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.37800687285223367, | |
| "grad_norm": 1.2343908478162073, | |
| "learning_rate": 9.757138916907184e-06, | |
| "loss": 0.1063, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.3848797250859107, | |
| "grad_norm": 1.815183611290879, | |
| "learning_rate": 9.738265855914014e-06, | |
| "loss": 0.0927, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3917525773195876, | |
| "grad_norm": 1.0482101436641207, | |
| "learning_rate": 9.718706306981332e-06, | |
| "loss": 0.0894, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.39862542955326463, | |
| "grad_norm": 1.1016058193514802, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.096, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.4054982817869416, | |
| "grad_norm": 1.3626789882878458, | |
| "learning_rate": 9.677539179628005e-06, | |
| "loss": 0.0913, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.41237113402061853, | |
| "grad_norm": 1.062516341067847, | |
| "learning_rate": 9.655937565570124e-06, | |
| "loss": 0.1035, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.41924398625429554, | |
| "grad_norm": 1.3580666593247015, | |
| "learning_rate": 9.63366139143413e-06, | |
| "loss": 0.1253, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.4261168384879725, | |
| "grad_norm": 1.2091009612690347, | |
| "learning_rate": 9.610713884629667e-06, | |
| "loss": 0.1042, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.4329896907216495, | |
| "grad_norm": 1.3110011817856249, | |
| "learning_rate": 9.587098369830171e-06, | |
| "loss": 0.0785, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.43986254295532645, | |
| "grad_norm": 1.1972541568504629, | |
| "learning_rate": 9.562818268491216e-06, | |
| "loss": 0.0882, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.44673539518900346, | |
| "grad_norm": 1.755801636297685, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.1204, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.4536082474226804, | |
| "grad_norm": 1.1228601630457258, | |
| "learning_rate": 9.512278472939627e-06, | |
| "loss": 0.1019, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.46048109965635736, | |
| "grad_norm": 1.0694185037056274, | |
| "learning_rate": 9.486026101017711e-06, | |
| "loss": 0.0988, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.46735395189003437, | |
| "grad_norm": 1.1816548324691527, | |
| "learning_rate": 9.459123786076911e-06, | |
| "loss": 0.1293, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.4742268041237113, | |
| "grad_norm": 1.0072855765153002, | |
| "learning_rate": 9.431575425769938e-06, | |
| "loss": 0.0786, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.48109965635738833, | |
| "grad_norm": 1.2516598610933318, | |
| "learning_rate": 9.40338501134964e-06, | |
| "loss": 0.093, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4879725085910653, | |
| "grad_norm": 0.9314966264564882, | |
| "learning_rate": 9.374556627090749e-06, | |
| "loss": 0.0808, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4948453608247423, | |
| "grad_norm": 2.098505641855836, | |
| "learning_rate": 9.345094449698143e-06, | |
| "loss": 0.0894, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5017182130584192, | |
| "grad_norm": 1.320411700756331, | |
| "learning_rate": 9.315002747701716e-06, | |
| "loss": 0.1158, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5085910652920962, | |
| "grad_norm": 0.7838582873048946, | |
| "learning_rate": 9.284285880837947e-06, | |
| "loss": 0.0879, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5154639175257731, | |
| "grad_norm": 0.9081118135912988, | |
| "learning_rate": 9.252948299418255e-06, | |
| "loss": 0.0809, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5223367697594502, | |
| "grad_norm": 1.0282634273034874, | |
| "learning_rate": 9.220994543684225e-06, | |
| "loss": 0.0834, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5292096219931272, | |
| "grad_norm": 1.0902003690643995, | |
| "learning_rate": 9.188429243149824e-06, | |
| "loss": 0.1146, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.5360824742268041, | |
| "grad_norm": 1.0675535113131351, | |
| "learning_rate": 9.155257115930651e-06, | |
| "loss": 0.0832, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.5429553264604811, | |
| "grad_norm": 0.9291426890257698, | |
| "learning_rate": 9.121482968060384e-06, | |
| "loss": 0.0875, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.5498281786941581, | |
| "grad_norm": 0.8805078527438052, | |
| "learning_rate": 9.08711169279446e-06, | |
| "loss": 0.0612, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5567010309278351, | |
| "grad_norm": 0.6738703550143792, | |
| "learning_rate": 9.052148269901145e-06, | |
| "loss": 0.0655, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.563573883161512, | |
| "grad_norm": 1.1560217059238038, | |
| "learning_rate": 9.01659776494005e-06, | |
| "loss": 0.112, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.570446735395189, | |
| "grad_norm": 0.7575418437918001, | |
| "learning_rate": 8.98046532852822e-06, | |
| "loss": 0.1045, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.5773195876288659, | |
| "grad_norm": 0.8594493038196075, | |
| "learning_rate": 8.943756195593916e-06, | |
| "loss": 0.078, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.584192439862543, | |
| "grad_norm": 0.9002266840004308, | |
| "learning_rate": 8.90647568461816e-06, | |
| "loss": 0.077, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5910652920962199, | |
| "grad_norm": 0.9865725593131784, | |
| "learning_rate": 8.868629196864182e-06, | |
| "loss": 0.0829, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5979381443298969, | |
| "grad_norm": 0.9354295431919494, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.0966, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6048109965635738, | |
| "grad_norm": 0.8222962290257311, | |
| "learning_rate": 8.791260305278434e-06, | |
| "loss": 0.0827, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6116838487972509, | |
| "grad_norm": 1.3331910294466658, | |
| "learning_rate": 8.751749110782013e-06, | |
| "loss": 0.1004, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6185567010309279, | |
| "grad_norm": 0.7080903069561276, | |
| "learning_rate": 8.71169435655405e-06, | |
| "loss": 0.0651, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.6254295532646048, | |
| "grad_norm": 1.6786808069119212, | |
| "learning_rate": 8.671101845794816e-06, | |
| "loss": 0.0888, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.6323024054982818, | |
| "grad_norm": 1.4828525584761454, | |
| "learning_rate": 8.629977459615655e-06, | |
| "loss": 0.0806, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.6391752577319587, | |
| "grad_norm": 0.9263138539115324, | |
| "learning_rate": 8.588327156186915e-06, | |
| "loss": 0.0732, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.6460481099656358, | |
| "grad_norm": 1.0099183275009322, | |
| "learning_rate": 8.546156969874723e-06, | |
| "loss": 0.088, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.6529209621993127, | |
| "grad_norm": 1.1484495047412266, | |
| "learning_rate": 8.503473010366713e-06, | |
| "loss": 0.099, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.6597938144329897, | |
| "grad_norm": 1.3705219833502182, | |
| "learning_rate": 8.460281461786848e-06, | |
| "loss": 0.0987, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.1978727936060138, | |
| "learning_rate": 8.416588581799447e-06, | |
| "loss": 0.092, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.6735395189003437, | |
| "grad_norm": 1.3778172716660015, | |
| "learning_rate": 8.372400700702569e-06, | |
| "loss": 0.1053, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.6804123711340206, | |
| "grad_norm": 1.1280074145372694, | |
| "learning_rate": 8.327724220510873e-06, | |
| "loss": 0.076, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.6872852233676976, | |
| "grad_norm": 0.7635710428267294, | |
| "learning_rate": 8.282565614028068e-06, | |
| "loss": 0.0884, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6941580756013745, | |
| "grad_norm": 1.4247309633025818, | |
| "learning_rate": 8.23693142390914e-06, | |
| "loss": 0.0912, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.7010309278350515, | |
| "grad_norm": 1.0571495795011308, | |
| "learning_rate": 8.19082826171243e-06, | |
| "loss": 0.0835, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7079037800687286, | |
| "grad_norm": 0.9531724619557164, | |
| "learning_rate": 8.144262806941743e-06, | |
| "loss": 0.0969, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.7147766323024055, | |
| "grad_norm": 0.9494705454880837, | |
| "learning_rate": 8.097241806078616e-06, | |
| "loss": 0.094, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.7216494845360825, | |
| "grad_norm": 0.9192463651767624, | |
| "learning_rate": 8.049772071604864e-06, | |
| "loss": 0.0763, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.7285223367697594, | |
| "grad_norm": 1.8748234681052738, | |
| "learning_rate": 8.001860481015594e-06, | |
| "loss": 0.0751, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.7353951890034365, | |
| "grad_norm": 1.1864927576783122, | |
| "learning_rate": 7.953513975822755e-06, | |
| "loss": 0.076, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.7422680412371134, | |
| "grad_norm": 1.6626638294825071, | |
| "learning_rate": 7.904739560549475e-06, | |
| "loss": 0.1024, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.7491408934707904, | |
| "grad_norm": 1.0387363209727432, | |
| "learning_rate": 7.855544301715203e-06, | |
| "loss": 0.082, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.7560137457044673, | |
| "grad_norm": 1.8144812889399762, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.0864, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.7628865979381443, | |
| "grad_norm": 1.6223014669230609, | |
| "learning_rate": 7.755919823271466e-06, | |
| "loss": 0.0947, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.7697594501718213, | |
| "grad_norm": 0.7777903725536114, | |
| "learning_rate": 7.70550503742427e-06, | |
| "loss": 0.0733, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.7766323024054983, | |
| "grad_norm": 1.2985821260191028, | |
| "learning_rate": 7.654698273449435e-06, | |
| "loss": 0.0944, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.7835051546391752, | |
| "grad_norm": 1.7285053063649418, | |
| "learning_rate": 7.603506892316513e-06, | |
| "loss": 0.1115, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.7903780068728522, | |
| "grad_norm": 1.046587564629829, | |
| "learning_rate": 7.551938310719043e-06, | |
| "loss": 0.1033, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.7972508591065293, | |
| "grad_norm": 1.2726351351771448, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0818, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.8041237113402062, | |
| "grad_norm": 0.7932513638036588, | |
| "learning_rate": 7.447699485069342e-06, | |
| "loss": 0.0872, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.8109965635738832, | |
| "grad_norm": 0.97642095215181, | |
| "learning_rate": 7.395044343313777e-06, | |
| "loss": 0.1, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.8178694158075601, | |
| "grad_norm": 0.7977845276369717, | |
| "learning_rate": 7.342042203498952e-06, | |
| "loss": 0.0758, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.8247422680412371, | |
| "grad_norm": 0.7754532768551025, | |
| "learning_rate": 7.288700744664167e-06, | |
| "loss": 0.0787, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.8316151202749141, | |
| "grad_norm": 0.6105436319990788, | |
| "learning_rate": 7.235027695009846e-06, | |
| "loss": 0.0734, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.8384879725085911, | |
| "grad_norm": 1.1321425051619216, | |
| "learning_rate": 7.181030830777838e-06, | |
| "loss": 0.0696, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.845360824742268, | |
| "grad_norm": 0.7931758959248494, | |
| "learning_rate": 7.1267179751248005e-06, | |
| "loss": 0.0771, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.852233676975945, | |
| "grad_norm": 0.7089257395570686, | |
| "learning_rate": 7.0720969969887595e-06, | |
| "loss": 0.0614, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.8591065292096219, | |
| "grad_norm": 0.9317656459355484, | |
| "learning_rate": 7.017175809949044e-06, | |
| "loss": 0.0682, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.865979381443299, | |
| "grad_norm": 0.9716643052721041, | |
| "learning_rate": 6.961962371079752e-06, | |
| "loss": 0.0863, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.872852233676976, | |
| "grad_norm": 0.7647031124105116, | |
| "learning_rate": 6.906464679796927e-06, | |
| "loss": 0.071, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.8797250859106529, | |
| "grad_norm": 0.6582474295592365, | |
| "learning_rate": 6.850690776699574e-06, | |
| "loss": 0.0693, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.8865979381443299, | |
| "grad_norm": 0.8673643657138536, | |
| "learning_rate": 6.79464874240473e-06, | |
| "loss": 0.0745, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.8934707903780069, | |
| "grad_norm": 0.5560064253477676, | |
| "learning_rate": 6.7383466963767386e-06, | |
| "loss": 0.0542, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.9003436426116839, | |
| "grad_norm": 0.7558195571660071, | |
| "learning_rate": 6.681792795750876e-06, | |
| "loss": 0.07, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.9072164948453608, | |
| "grad_norm": 1.071468950008921, | |
| "learning_rate": 6.624995234151539e-06, | |
| "loss": 0.0872, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.9140893470790378, | |
| "grad_norm": 1.0140936221676173, | |
| "learning_rate": 6.567962240505136e-06, | |
| "loss": 0.0741, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.9209621993127147, | |
| "grad_norm": 0.7964752815994411, | |
| "learning_rate": 6.510702077847864e-06, | |
| "loss": 0.1003, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.9278350515463918, | |
| "grad_norm": 0.7163903645675469, | |
| "learning_rate": 6.453223042128556e-06, | |
| "loss": 0.0729, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.9347079037800687, | |
| "grad_norm": 0.9175074081579961, | |
| "learning_rate": 6.395533461006736e-06, | |
| "loss": 0.1223, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.9415807560137457, | |
| "grad_norm": 0.7350777231075519, | |
| "learning_rate": 6.337641692646106e-06, | |
| "loss": 0.0874, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.9484536082474226, | |
| "grad_norm": 0.8752312055008336, | |
| "learning_rate": 6.2795561245035895e-06, | |
| "loss": 0.088, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.9553264604810997, | |
| "grad_norm": 1.141388283180733, | |
| "learning_rate": 6.221285172114156e-06, | |
| "loss": 0.0832, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.9621993127147767, | |
| "grad_norm": 0.6646415291771995, | |
| "learning_rate": 6.162837277871553e-06, | |
| "loss": 0.0691, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.9690721649484536, | |
| "grad_norm": 0.6628220585276329, | |
| "learning_rate": 6.104220909805162e-06, | |
| "loss": 0.0695, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.9759450171821306, | |
| "grad_norm": 0.6574209540254052, | |
| "learning_rate": 6.045444560353136e-06, | |
| "loss": 0.0688, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.9828178694158075, | |
| "grad_norm": 0.935261593082011, | |
| "learning_rate": 5.986516745132e-06, | |
| "loss": 0.0861, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.9896907216494846, | |
| "grad_norm": 0.7137218065215563, | |
| "learning_rate": 5.927446001702899e-06, | |
| "loss": 0.0731, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "grad_norm": 0.8092886479437756, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.0898, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.9965635738831615, | |
| "eval_loss": 0.07346626371145248, | |
| "eval_runtime": 28.7972, | |
| "eval_samples_per_second": 33.996, | |
| "eval_steps_per_second": 1.076, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.0034364261168385, | |
| "grad_norm": 0.6208305041752085, | |
| "learning_rate": 5.808909982763825e-06, | |
| "loss": 0.0588, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.0103092783505154, | |
| "grad_norm": 1.2590078854254831, | |
| "learning_rate": 5.749461880951966e-06, | |
| "loss": 0.0656, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.0171821305841924, | |
| "grad_norm": 0.7655998026761056, | |
| "learning_rate": 5.689905195840216e-06, | |
| "loss": 0.0776, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.0240549828178693, | |
| "grad_norm": 0.640435804630534, | |
| "learning_rate": 5.630248556101448e-06, | |
| "loss": 0.0773, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.0309278350515463, | |
| "grad_norm": 0.9901106513353419, | |
| "learning_rate": 5.570500604890124e-06, | |
| "loss": 0.1007, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.0378006872852235, | |
| "grad_norm": 0.981110073358187, | |
| "learning_rate": 5.510669998590074e-06, | |
| "loss": 0.0708, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.0446735395189004, | |
| "grad_norm": 0.5032007844300563, | |
| "learning_rate": 5.450765405560328e-06, | |
| "loss": 0.0466, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.0515463917525774, | |
| "grad_norm": 0.9605821953639473, | |
| "learning_rate": 5.390795504879243e-06, | |
| "loss": 0.0883, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.0584192439862543, | |
| "grad_norm": 0.6856426177125351, | |
| "learning_rate": 5.330768985087059e-06, | |
| "loss": 0.0714, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.0652920962199313, | |
| "grad_norm": 0.7400054830415009, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.0584, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.0721649484536082, | |
| "grad_norm": 0.8226062466881034, | |
| "learning_rate": 5.2105808820857126e-06, | |
| "loss": 0.0727, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.0790378006872852, | |
| "grad_norm": 0.8738787201265895, | |
| "learning_rate": 5.150436711931387e-06, | |
| "loss": 0.0571, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.0859106529209621, | |
| "grad_norm": 0.8092348179353107, | |
| "learning_rate": 5.090270746252803e-06, | |
| "loss": 0.0554, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.0927835051546393, | |
| "grad_norm": 0.6815129371875649, | |
| "learning_rate": 5.030091701996428e-06, | |
| "loss": 0.0384, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.0996563573883162, | |
| "grad_norm": 1.3388283304826931, | |
| "learning_rate": 4.9699082980035735e-06, | |
| "loss": 0.0671, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.1065292096219932, | |
| "grad_norm": 0.8200140877757371, | |
| "learning_rate": 4.909729253747197e-06, | |
| "loss": 0.0612, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.1134020618556701, | |
| "grad_norm": 1.3785087464665922, | |
| "learning_rate": 4.8495632880686155e-06, | |
| "loss": 0.0606, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.120274914089347, | |
| "grad_norm": 1.0793888751275476, | |
| "learning_rate": 4.789419117914288e-06, | |
| "loss": 0.0567, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.127147766323024, | |
| "grad_norm": 1.1372286019308213, | |
| "learning_rate": 4.729305457072913e-06, | |
| "loss": 0.0664, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.134020618556701, | |
| "grad_norm": 0.7746315240273167, | |
| "learning_rate": 4.6692310149129425e-06, | |
| "loss": 0.0693, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.140893470790378, | |
| "grad_norm": 0.9403866506902895, | |
| "learning_rate": 4.609204495120759e-06, | |
| "loss": 0.0634, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.147766323024055, | |
| "grad_norm": 1.1336689101772839, | |
| "learning_rate": 4.549234594439674e-06, | |
| "loss": 0.0774, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.1546391752577319, | |
| "grad_norm": 0.639070711059099, | |
| "learning_rate": 4.489330001409929e-06, | |
| "loss": 0.0388, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.161512027491409, | |
| "grad_norm": 0.9259670355611469, | |
| "learning_rate": 4.429499395109877e-06, | |
| "loss": 0.0726, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.168384879725086, | |
| "grad_norm": 0.9391479261539729, | |
| "learning_rate": 4.369751443898554e-06, | |
| "loss": 0.1028, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.175257731958763, | |
| "grad_norm": 0.7406811832397094, | |
| "learning_rate": 4.310094804159784e-06, | |
| "loss": 0.0846, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.1821305841924399, | |
| "grad_norm": 0.777835245833664, | |
| "learning_rate": 4.250538119048036e-06, | |
| "loss": 0.0737, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.1890034364261168, | |
| "grad_norm": 0.8238371968347903, | |
| "learning_rate": 4.191090017236177e-06, | |
| "loss": 0.0739, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.1958762886597938, | |
| "grad_norm": 0.5882713132677511, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.0522, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.2027491408934707, | |
| "grad_norm": 0.7641320010777528, | |
| "learning_rate": 4.072553998297103e-06, | |
| "loss": 0.0563, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2096219931271477, | |
| "grad_norm": 0.9589895933174847, | |
| "learning_rate": 4.013483254868001e-06, | |
| "loss": 0.0873, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.2164948453608249, | |
| "grad_norm": 0.6374766950817062, | |
| "learning_rate": 3.9545554396468655e-06, | |
| "loss": 0.0495, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.2233676975945018, | |
| "grad_norm": 1.1694612873101644, | |
| "learning_rate": 3.89577909019484e-06, | |
| "loss": 0.0714, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.2302405498281788, | |
| "grad_norm": 0.6569314759372903, | |
| "learning_rate": 3.8371627221284495e-06, | |
| "loss": 0.0649, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.2371134020618557, | |
| "grad_norm": 0.6223039049974719, | |
| "learning_rate": 3.7787148278858453e-06, | |
| "loss": 0.0827, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.2439862542955327, | |
| "grad_norm": 0.5170872367242648, | |
| "learning_rate": 3.7204438754964113e-06, | |
| "loss": 0.0542, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.2508591065292096, | |
| "grad_norm": 0.7646480070633659, | |
| "learning_rate": 3.662358307353897e-06, | |
| "loss": 0.0605, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.2577319587628866, | |
| "grad_norm": 0.7391886236805388, | |
| "learning_rate": 3.6044665389932663e-06, | |
| "loss": 0.062, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.2646048109965635, | |
| "grad_norm": 0.7226942142778663, | |
| "learning_rate": 3.5467769578714455e-06, | |
| "loss": 0.0587, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.2714776632302405, | |
| "grad_norm": 0.5989089061723369, | |
| "learning_rate": 3.489297922152136e-06, | |
| "loss": 0.0581, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.2783505154639174, | |
| "grad_norm": 0.6845194918026075, | |
| "learning_rate": 3.432037759494867e-06, | |
| "loss": 0.0693, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.2852233676975944, | |
| "grad_norm": 0.9006580674437046, | |
| "learning_rate": 3.375004765848463e-06, | |
| "loss": 0.0592, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.2920962199312716, | |
| "grad_norm": 0.7485789730040875, | |
| "learning_rate": 3.3182072042491244e-06, | |
| "loss": 0.0607, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.2989690721649485, | |
| "grad_norm": 0.7498312301971553, | |
| "learning_rate": 3.2616533036232635e-06, | |
| "loss": 0.0635, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.3058419243986255, | |
| "grad_norm": 0.6173267943065566, | |
| "learning_rate": 3.205351257595272e-06, | |
| "loss": 0.0661, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.3127147766323024, | |
| "grad_norm": 0.5610080933762857, | |
| "learning_rate": 3.149309223300428e-06, | |
| "loss": 0.0368, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.3195876288659794, | |
| "grad_norm": 0.6744445741823302, | |
| "learning_rate": 3.093535320203074e-06, | |
| "loss": 0.0538, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.3264604810996563, | |
| "grad_norm": 0.7074572875753987, | |
| "learning_rate": 3.0380376289202497e-06, | |
| "loss": 0.0588, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.5578874412952257, | |
| "learning_rate": 2.982824190050958e-06, | |
| "loss": 0.0487, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.3402061855670104, | |
| "grad_norm": 0.805651744779414, | |
| "learning_rate": 2.927903003011241e-06, | |
| "loss": 0.0804, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.3470790378006874, | |
| "grad_norm": 0.6104865864882411, | |
| "learning_rate": 2.8732820248752016e-06, | |
| "loss": 0.0543, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.3539518900343643, | |
| "grad_norm": 0.8240035739026437, | |
| "learning_rate": 2.8189691692221627e-06, | |
| "loss": 0.0636, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.3608247422680413, | |
| "grad_norm": 0.7051312664144036, | |
| "learning_rate": 2.7649723049901554e-06, | |
| "loss": 0.073, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.3676975945017182, | |
| "grad_norm": 0.42476254316706996, | |
| "learning_rate": 2.711299255335833e-06, | |
| "loss": 0.0349, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.3745704467353952, | |
| "grad_norm": 0.6737847343673471, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.0713, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3814432989690721, | |
| "grad_norm": 0.674246319060676, | |
| "learning_rate": 2.6049556566862234e-06, | |
| "loss": 0.0606, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.388316151202749, | |
| "grad_norm": 0.6932399889922612, | |
| "learning_rate": 2.552300514930657e-06, | |
| "loss": 0.063, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.395189003436426, | |
| "grad_norm": 0.6731428439385392, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0674, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.402061855670103, | |
| "grad_norm": 0.610598316111306, | |
| "learning_rate": 2.4480616892809593e-06, | |
| "loss": 0.0635, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.40893470790378, | |
| "grad_norm": 0.5974570359595684, | |
| "learning_rate": 2.396493107683488e-06, | |
| "loss": 0.0594, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.4158075601374571, | |
| "grad_norm": 1.0210490387561593, | |
| "learning_rate": 2.345301726550567e-06, | |
| "loss": 0.0785, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.422680412371134, | |
| "grad_norm": 0.5744649417589675, | |
| "learning_rate": 2.2944949625757295e-06, | |
| "loss": 0.0448, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.429553264604811, | |
| "grad_norm": 0.6601987926605946, | |
| "learning_rate": 2.244080176728536e-06, | |
| "loss": 0.057, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.436426116838488, | |
| "grad_norm": 0.7638745257556894, | |
| "learning_rate": 2.1940646731880887e-06, | |
| "loss": 0.0877, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.443298969072165, | |
| "grad_norm": 0.6788201963591001, | |
| "learning_rate": 2.1444556982847996e-06, | |
| "loss": 0.0625, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.4501718213058419, | |
| "grad_norm": 0.9453695676523692, | |
| "learning_rate": 2.095260439450526e-06, | |
| "loss": 0.0581, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.4570446735395188, | |
| "grad_norm": 0.6019277138216071, | |
| "learning_rate": 2.0464860241772454e-06, | |
| "loss": 0.0553, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.463917525773196, | |
| "grad_norm": 0.5843296207381486, | |
| "learning_rate": 1.998139518984409e-06, | |
| "loss": 0.0511, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.470790378006873, | |
| "grad_norm": 0.8531796806256794, | |
| "learning_rate": 1.9502279283951363e-06, | |
| "loss": 0.0679, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.47766323024055, | |
| "grad_norm": 0.9653580251066537, | |
| "learning_rate": 1.9027581939213852e-06, | |
| "loss": 0.0743, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.4845360824742269, | |
| "grad_norm": 0.6609726255392513, | |
| "learning_rate": 1.8557371930582579e-06, | |
| "loss": 0.0546, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.4914089347079038, | |
| "grad_norm": 0.5746241154265249, | |
| "learning_rate": 1.8091717382875723e-06, | |
| "loss": 0.0523, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.4982817869415808, | |
| "grad_norm": 0.6298023705802349, | |
| "learning_rate": 1.7630685760908623e-06, | |
| "loss": 0.0661, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.5051546391752577, | |
| "grad_norm": 0.4452542166593582, | |
| "learning_rate": 1.7174343859719334e-06, | |
| "loss": 0.0418, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.5120274914089347, | |
| "grad_norm": 0.9558078745291808, | |
| "learning_rate": 1.6722757794891287e-06, | |
| "loss": 0.0689, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.5189003436426116, | |
| "grad_norm": 0.5721444477862428, | |
| "learning_rate": 1.627599299297431e-06, | |
| "loss": 0.0445, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.5257731958762886, | |
| "grad_norm": 0.5029491146653275, | |
| "learning_rate": 1.5834114182005544e-06, | |
| "loss": 0.0496, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.5326460481099655, | |
| "grad_norm": 0.6322387430008974, | |
| "learning_rate": 1.5397185382131524e-06, | |
| "loss": 0.0524, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.5395189003436425, | |
| "grad_norm": 1.0417849318719474, | |
| "learning_rate": 1.4965269896332884e-06, | |
| "loss": 0.0604, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.5463917525773194, | |
| "grad_norm": 0.5899183799503986, | |
| "learning_rate": 1.4538430301252783e-06, | |
| "loss": 0.045, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5532646048109966, | |
| "grad_norm": 0.6836566500948758, | |
| "learning_rate": 1.411672843813086e-06, | |
| "loss": 0.0741, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.5601374570446735, | |
| "grad_norm": 0.4938071091787132, | |
| "learning_rate": 1.370022540384347e-06, | |
| "loss": 0.062, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.5670103092783505, | |
| "grad_norm": 0.48137384426619345, | |
| "learning_rate": 1.3288981542051844e-06, | |
| "loss": 0.0459, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.5738831615120275, | |
| "grad_norm": 0.45659499070716764, | |
| "learning_rate": 1.2883056434459506e-06, | |
| "loss": 0.0548, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.5807560137457046, | |
| "grad_norm": 0.5618265912758492, | |
| "learning_rate": 1.2482508892179884e-06, | |
| "loss": 0.0747, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.5876288659793816, | |
| "grad_norm": 0.4895058148757872, | |
| "learning_rate": 1.2087396947215678e-06, | |
| "loss": 0.0429, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.5945017182130585, | |
| "grad_norm": 0.5501598761385641, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.0563, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.6013745704467355, | |
| "grad_norm": 0.40909416624365424, | |
| "learning_rate": 1.1313708031358183e-06, | |
| "loss": 0.0314, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.6082474226804124, | |
| "grad_norm": 0.5638565340820253, | |
| "learning_rate": 1.0935243153818437e-06, | |
| "loss": 0.0813, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.6151202749140894, | |
| "grad_norm": 0.4911685662827527, | |
| "learning_rate": 1.0562438044060846e-06, | |
| "loss": 0.0593, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.6219931271477663, | |
| "grad_norm": 0.5101232374913048, | |
| "learning_rate": 1.0195346714717813e-06, | |
| "loss": 0.0489, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.6288659793814433, | |
| "grad_norm": 0.5456686593151775, | |
| "learning_rate": 9.834022350599538e-07, | |
| "loss": 0.0464, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.6357388316151202, | |
| "grad_norm": 0.6070446990488096, | |
| "learning_rate": 9.47851730098856e-07, | |
| "loss": 0.0599, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.6426116838487972, | |
| "grad_norm": 0.4752269997897051, | |
| "learning_rate": 9.128883072055411e-07, | |
| "loss": 0.0508, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.6494845360824741, | |
| "grad_norm": 0.47451969256577875, | |
| "learning_rate": 8.785170319396174e-07, | |
| "loss": 0.0525, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.656357388316151, | |
| "grad_norm": 0.4406482887242554, | |
| "learning_rate": 8.447428840693489e-07, | |
| "loss": 0.0382, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.663230240549828, | |
| "grad_norm": 0.5750829828157604, | |
| "learning_rate": 8.115707568501768e-07, | |
| "loss": 0.06, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.670103092783505, | |
| "grad_norm": 0.6180712457017934, | |
| "learning_rate": 7.790054563157745e-07, | |
| "loss": 0.0562, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.6769759450171822, | |
| "grad_norm": 0.6329763563152765, | |
| "learning_rate": 7.470517005817473e-07, | |
| "loss": 0.0507, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.6838487972508591, | |
| "grad_norm": 0.5514195178248108, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.0387, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.690721649484536, | |
| "grad_norm": 0.4489220276384386, | |
| "learning_rate": 6.849972522982845e-07, | |
| "loss": 0.0518, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.697594501718213, | |
| "grad_norm": 0.6752359143096192, | |
| "learning_rate": 6.549055503018575e-07, | |
| "loss": 0.0598, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.7044673539518902, | |
| "grad_norm": 0.5245755853098147, | |
| "learning_rate": 6.254433729092518e-07, | |
| "loss": 0.0638, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.7113402061855671, | |
| "grad_norm": 0.6263348006470634, | |
| "learning_rate": 5.966149886503614e-07, | |
| "loss": 0.053, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.718213058419244, | |
| "grad_norm": 0.5241226265559917, | |
| "learning_rate": 5.684245742300625e-07, | |
| "loss": 0.0426, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.725085910652921, | |
| "grad_norm": 0.5664360880759117, | |
| "learning_rate": 5.408762139230889e-07, | |
| "loss": 0.0595, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.731958762886598, | |
| "grad_norm": 0.6453628521319137, | |
| "learning_rate": 5.139738989822901e-07, | |
| "loss": 0.0611, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.738831615120275, | |
| "grad_norm": 0.7519593438944014, | |
| "learning_rate": 4.877215270603752e-07, | |
| "loss": 0.0719, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.745704467353952, | |
| "grad_norm": 0.4980336172644675, | |
| "learning_rate": 4.6212290164521554e-07, | |
| "loss": 0.0519, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.7525773195876289, | |
| "grad_norm": 0.45119481959388535, | |
| "learning_rate": 4.371817315087845e-07, | |
| "loss": 0.0425, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.7594501718213058, | |
| "grad_norm": 0.6584086545614478, | |
| "learning_rate": 4.1290163016982855e-07, | |
| "loss": 0.0557, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.7663230240549828, | |
| "grad_norm": 0.5617445252368427, | |
| "learning_rate": 3.8928611537033424e-07, | |
| "loss": 0.0462, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.7731958762886597, | |
| "grad_norm": 0.7486755524364309, | |
| "learning_rate": 3.663386085658693e-07, | |
| "loss": 0.0726, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.7800687285223367, | |
| "grad_norm": 0.7750720732301635, | |
| "learning_rate": 3.4406243442987765e-07, | |
| "loss": 0.071, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.7869415807560136, | |
| "grad_norm": 0.5580247693753845, | |
| "learning_rate": 3.224608203719953e-07, | |
| "loss": 0.0555, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.7938144329896906, | |
| "grad_norm": 0.5165173678771202, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.056, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.8006872852233677, | |
| "grad_norm": 0.6613292853189043, | |
| "learning_rate": 2.812936930186688e-07, | |
| "loss": 0.0927, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.8075601374570447, | |
| "grad_norm": 0.9278647692756433, | |
| "learning_rate": 2.617341440859883e-07, | |
| "loss": 0.037, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.8144329896907216, | |
| "grad_norm": 0.7289328269779501, | |
| "learning_rate": 2.428610830928152e-07, | |
| "loss": 0.0761, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.8213058419243986, | |
| "grad_norm": 0.49788399325301386, | |
| "learning_rate": 2.2467724440002336e-07, | |
| "loss": 0.0489, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.8281786941580758, | |
| "grad_norm": 0.5113530913306602, | |
| "learning_rate": 2.0718526251279346e-07, | |
| "loss": 0.0486, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.8350515463917527, | |
| "grad_norm": 0.5102277206374749, | |
| "learning_rate": 1.9038767169893058e-07, | |
| "loss": 0.0622, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.8419243986254297, | |
| "grad_norm": 0.4427615005394849, | |
| "learning_rate": 1.7428690562169003e-07, | |
| "loss": 0.0471, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.8487972508591066, | |
| "grad_norm": 0.5314021627847645, | |
| "learning_rate": 1.5888529698718347e-07, | |
| "loss": 0.0643, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.8556701030927836, | |
| "grad_norm": 0.4583438429264996, | |
| "learning_rate": 1.4418507720641794e-07, | |
| "loss": 0.0405, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.8625429553264605, | |
| "grad_norm": 0.5632999153636097, | |
| "learning_rate": 1.3018837607199909e-07, | |
| "loss": 0.0627, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.8694158075601375, | |
| "grad_norm": 0.4657493140778872, | |
| "learning_rate": 1.1689722144956672e-07, | |
| "loss": 0.0461, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.8762886597938144, | |
| "grad_norm": 0.5114682498956993, | |
| "learning_rate": 1.0431353898399388e-07, | |
| "loss": 0.0648, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.8831615120274914, | |
| "grad_norm": 0.656466746842816, | |
| "learning_rate": 9.243915182039431e-08, | |
| "loss": 0.0707, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.8900343642611683, | |
| "grad_norm": 0.44846650965373375, | |
| "learning_rate": 8.127578033998663e-08, | |
| "loss": 0.0494, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.8969072164948453, | |
| "grad_norm": 0.5843547341836819, | |
| "learning_rate": 7.082504191083417e-08, | |
| "loss": 0.0518, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.9037800687285222, | |
| "grad_norm": 0.5076870881901042, | |
| "learning_rate": 6.108845065352864e-08, | |
| "loss": 0.0657, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.9106529209621992, | |
| "grad_norm": 0.5237960361343623, | |
| "learning_rate": 5.206741722181385e-08, | |
| "loss": 0.0508, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.9175257731958761, | |
| "grad_norm": 0.4577769450444941, | |
| "learning_rate": 4.376324859820924e-08, | |
| "loss": 0.0332, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.9243986254295533, | |
| "grad_norm": 0.5521319279321418, | |
| "learning_rate": 3.617714790465576e-08, | |
| "loss": 0.059, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.9312714776632303, | |
| "grad_norm": 0.44110763004122194, | |
| "learning_rate": 2.9310214228202016e-08, | |
| "loss": 0.0407, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.9381443298969072, | |
| "grad_norm": 0.492994648123961, | |
| "learning_rate": 2.3163442461766604e-08, | |
| "loss": 0.0532, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.9450171821305842, | |
| "grad_norm": 0.5554179174819742, | |
| "learning_rate": 1.7737723159999e-08, | |
| "loss": 0.0534, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.9518900343642611, | |
| "grad_norm": 0.5594422748663939, | |
| "learning_rate": 1.3033842410251074e-08, | |
| "loss": 0.0549, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.9587628865979383, | |
| "grad_norm": 0.960561797218523, | |
| "learning_rate": 9.052481718690998e-09, | |
| "loss": 0.0845, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.9656357388316152, | |
| "grad_norm": 0.4924978890119879, | |
| "learning_rate": 5.794217911562205e-09, | |
| "loss": 0.0566, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.9725085910652922, | |
| "grad_norm": 0.5253136520473263, | |
| "learning_rate": 3.2595230516152543e-09, | |
| "loss": 0.0843, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.9793814432989691, | |
| "grad_norm": 1.2759989208541003, | |
| "learning_rate": 1.4487643697103092e-09, | |
| "loss": 0.0629, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.986254295532646, | |
| "grad_norm": 0.6010268267031165, | |
| "learning_rate": 3.6220421161692333e-10, | |
| "loss": 0.0452, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "grad_norm": 0.5537189191514956, | |
| "learning_rate": 0.0, | |
| "loss": 0.0622, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "eval_loss": 0.06167122349143028, | |
| "eval_runtime": 28.4642, | |
| "eval_samples_per_second": 34.394, | |
| "eval_steps_per_second": 1.089, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.993127147766323, | |
| "step": 290, | |
| "total_flos": 8.751482238776115e+16, | |
| "train_loss": 0.0836696587759873, | |
| "train_runtime": 3245.322, | |
| "train_samples_per_second": 11.461, | |
| "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": 8.751482238776115e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |