| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9994632313472893, | |
| "eval_steps": 500, | |
| "global_step": 931, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010735373054213634, | |
| "grad_norm": 22.789785654510972, | |
| "learning_rate": 1.0638297872340426e-07, | |
| "loss": 1.3534, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005367686527106817, | |
| "grad_norm": 21.426953387742742, | |
| "learning_rate": 5.319148936170213e-07, | |
| "loss": 1.315, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010735373054213635, | |
| "grad_norm": 8.681854043380088, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "loss": 1.2029, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01610305958132045, | |
| "grad_norm": 6.879600686982149, | |
| "learning_rate": 1.595744680851064e-06, | |
| "loss": 1.0391, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02147074610842727, | |
| "grad_norm": 3.0275647149692793, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "loss": 0.9127, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.026838432635534086, | |
| "grad_norm": 2.4868622752512706, | |
| "learning_rate": 2.6595744680851065e-06, | |
| "loss": 0.8614, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0322061191626409, | |
| "grad_norm": 2.260349279940821, | |
| "learning_rate": 3.191489361702128e-06, | |
| "loss": 0.8364, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03757380568974772, | |
| "grad_norm": 2.3883199937027952, | |
| "learning_rate": 3.723404255319149e-06, | |
| "loss": 0.8104, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04294149221685454, | |
| "grad_norm": 2.2131867254029656, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 0.799, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04830917874396135, | |
| "grad_norm": 2.192168023131458, | |
| "learning_rate": 4.787234042553192e-06, | |
| "loss": 0.7889, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05367686527106817, | |
| "grad_norm": 2.3212514172557133, | |
| "learning_rate": 5.319148936170213e-06, | |
| "loss": 0.7614, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.059044551798174985, | |
| "grad_norm": 2.3347499759614307, | |
| "learning_rate": 5.851063829787235e-06, | |
| "loss": 0.7482, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0644122383252818, | |
| "grad_norm": 2.285910235729559, | |
| "learning_rate": 6.382978723404256e-06, | |
| "loss": 0.7444, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06977992485238863, | |
| "grad_norm": 2.4172902264287255, | |
| "learning_rate": 6.914893617021278e-06, | |
| "loss": 0.7238, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07514761137949544, | |
| "grad_norm": 2.3034733454455596, | |
| "learning_rate": 7.446808510638298e-06, | |
| "loss": 0.7231, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08051529790660225, | |
| "grad_norm": 2.4091543127728756, | |
| "learning_rate": 7.97872340425532e-06, | |
| "loss": 0.7036, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08588298443370908, | |
| "grad_norm": 2.2455472355866837, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 0.7073, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09125067096081589, | |
| "grad_norm": 2.4464200044770936, | |
| "learning_rate": 9.042553191489362e-06, | |
| "loss": 0.7078, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0966183574879227, | |
| "grad_norm": 2.35752750364654, | |
| "learning_rate": 9.574468085106385e-06, | |
| "loss": 0.6963, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10198604401502952, | |
| "grad_norm": 2.1996004538409473, | |
| "learning_rate": 9.999964780082996e-06, | |
| "loss": 0.6979, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10735373054213634, | |
| "grad_norm": 2.203322438503073, | |
| "learning_rate": 9.998732135085665e-06, | |
| "loss": 0.6906, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11272141706924316, | |
| "grad_norm": 2.264863963599694, | |
| "learning_rate": 9.995738990383743e-06, | |
| "loss": 0.6866, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11808910359634997, | |
| "grad_norm": 2.2286418377556236, | |
| "learning_rate": 9.990986400130607e-06, | |
| "loss": 0.6782, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 2.1240149641244175, | |
| "learning_rate": 9.984476038137437e-06, | |
| "loss": 0.6697, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1288244766505636, | |
| "grad_norm": 2.0873242470478854, | |
| "learning_rate": 9.97621019728372e-06, | |
| "loss": 0.6688, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13419216317767044, | |
| "grad_norm": 2.4812498111624612, | |
| "learning_rate": 9.966191788709716e-06, | |
| "loss": 0.6631, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.13955984970477725, | |
| "grad_norm": 2.343955721850536, | |
| "learning_rate": 9.954424340791195e-06, | |
| "loss": 0.6644, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14492753623188406, | |
| "grad_norm": 2.339967843619562, | |
| "learning_rate": 9.940911997896774e-06, | |
| "loss": 0.6642, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.15029522275899088, | |
| "grad_norm": 2.2056817559033837, | |
| "learning_rate": 9.925659518928316e-06, | |
| "loss": 0.667, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1556629092860977, | |
| "grad_norm": 1.9903249362862974, | |
| "learning_rate": 9.908672275644898e-06, | |
| "loss": 0.6444, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1610305958132045, | |
| "grad_norm": 2.2039128348943744, | |
| "learning_rate": 9.889956250770933e-06, | |
| "loss": 0.6488, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16639828234031132, | |
| "grad_norm": 2.2078157899902706, | |
| "learning_rate": 9.86951803588912e-06, | |
| "loss": 0.6499, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.17176596886741816, | |
| "grad_norm": 2.0109828014248774, | |
| "learning_rate": 9.847364829118963e-06, | |
| "loss": 0.6519, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17713365539452497, | |
| "grad_norm": 1.9664484718503512, | |
| "learning_rate": 9.82350443258166e-06, | |
| "loss": 0.6401, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.18250134192163178, | |
| "grad_norm": 2.0667728485558308, | |
| "learning_rate": 9.797945249652295e-06, | |
| "loss": 0.6437, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1878690284487386, | |
| "grad_norm": 2.0065876578800346, | |
| "learning_rate": 9.770696282000245e-06, | |
| "loss": 0.6346, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1932367149758454, | |
| "grad_norm": 1.987617668179378, | |
| "learning_rate": 9.741767126418898e-06, | |
| "loss": 0.6356, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19860440150295222, | |
| "grad_norm": 2.04140704556592, | |
| "learning_rate": 9.711167971445766e-06, | |
| "loss": 0.6349, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.20397208803005903, | |
| "grad_norm": 2.047535907923479, | |
| "learning_rate": 9.67890959377418e-06, | |
| "loss": 0.6298, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.20933977455716588, | |
| "grad_norm": 1.9558416597814228, | |
| "learning_rate": 9.645003354457872e-06, | |
| "loss": 0.6349, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2147074610842727, | |
| "grad_norm": 1.8984622795525588, | |
| "learning_rate": 9.60946119490972e-06, | |
| "loss": 0.6142, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2200751476113795, | |
| "grad_norm": 2.3906003365767496, | |
| "learning_rate": 9.57229563269612e-06, | |
| "loss": 0.6222, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.22544283413848631, | |
| "grad_norm": 1.8921023328729312, | |
| "learning_rate": 9.533519757128426e-06, | |
| "loss": 0.6186, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23081052066559313, | |
| "grad_norm": 1.9626099660140686, | |
| "learning_rate": 9.49314722465304e-06, | |
| "loss": 0.6258, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.23617820719269994, | |
| "grad_norm": 2.0411979584264266, | |
| "learning_rate": 9.451192254041759e-06, | |
| "loss": 0.5979, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24154589371980675, | |
| "grad_norm": 2.1889414515575183, | |
| "learning_rate": 9.407669621384073e-06, | |
| "loss": 0.6083, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 2.085807605093685, | |
| "learning_rate": 9.362594654883185e-06, | |
| "loss": 0.5919, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2522812667740204, | |
| "grad_norm": 2.0627204109631543, | |
| "learning_rate": 9.31598322945759e-06, | |
| "loss": 0.6034, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2576489533011272, | |
| "grad_norm": 1.9378180140246337, | |
| "learning_rate": 9.267851761150092e-06, | |
| "loss": 0.6038, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.26301663982823403, | |
| "grad_norm": 3.1058701754069924, | |
| "learning_rate": 9.218217201346251e-06, | |
| "loss": 0.595, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2683843263553409, | |
| "grad_norm": 1.9991169619012592, | |
| "learning_rate": 9.167097030804289e-06, | |
| "loss": 0.5874, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.27375201288244766, | |
| "grad_norm": 2.061863952952721, | |
| "learning_rate": 9.114509253498554e-06, | |
| "loss": 0.5969, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2791196994095545, | |
| "grad_norm": 2.0907726187864926, | |
| "learning_rate": 9.060472390278717e-06, | |
| "loss": 0.5814, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2844873859366613, | |
| "grad_norm": 2.7768778562710397, | |
| "learning_rate": 9.005005472346923e-06, | |
| "loss": 0.593, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 2.0516068085641574, | |
| "learning_rate": 8.948128034555212e-06, | |
| "loss": 0.5849, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2952227589908749, | |
| "grad_norm": 1.8867496167590758, | |
| "learning_rate": 8.889860108525544e-06, | |
| "loss": 0.5863, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.30059044551798175, | |
| "grad_norm": 1.942421908800686, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5672, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3059581320450886, | |
| "grad_norm": 2.0272368493119393, | |
| "learning_rate": 8.76923535958783e-06, | |
| "loss": 0.5623, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3113258185721954, | |
| "grad_norm": 2.1343642849157787, | |
| "learning_rate": 8.706921019419237e-06, | |
| "loss": 0.5732, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3166935050993022, | |
| "grad_norm": 1.9323875772249934, | |
| "learning_rate": 8.643301141529619e-06, | |
| "loss": 0.5656, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.322061191626409, | |
| "grad_norm": 2.037757830855184, | |
| "learning_rate": 8.578398132155846e-06, | |
| "loss": 0.5611, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32742887815351585, | |
| "grad_norm": 1.9398690166595873, | |
| "learning_rate": 8.512234849439887e-06, | |
| "loss": 0.5645, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.33279656468062263, | |
| "grad_norm": 2.190033028948571, | |
| "learning_rate": 8.444834595378434e-06, | |
| "loss": 0.5622, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.33816425120772947, | |
| "grad_norm": 2.221598674029084, | |
| "learning_rate": 8.376221107616187e-06, | |
| "loss": 0.546, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.3435319377348363, | |
| "grad_norm": 2.021210342324902, | |
| "learning_rate": 8.306418551085707e-06, | |
| "loss": 0.5524, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.3488996242619431, | |
| "grad_norm": 1.87069475903427, | |
| "learning_rate": 8.23545150949679e-06, | |
| "loss": 0.5447, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.35426731078904994, | |
| "grad_norm": 1.880521268570416, | |
| "learning_rate": 8.163344976678342e-06, | |
| "loss": 0.5392, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3596349973161567, | |
| "grad_norm": 2.14764020394395, | |
| "learning_rate": 8.090124347775837e-06, | |
| "loss": 0.547, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.36500268384326356, | |
| "grad_norm": 1.9583788892413105, | |
| "learning_rate": 8.0158154103074e-06, | |
| "loss": 0.5487, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 2.053310730502184, | |
| "learning_rate": 7.940444335081733e-06, | |
| "loss": 0.5408, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3757380568974772, | |
| "grad_norm": 1.9003460290083882, | |
| "learning_rate": 7.864037666981037e-06, | |
| "loss": 0.543, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.38110574342458403, | |
| "grad_norm": 1.9850745541543715, | |
| "learning_rate": 7.786622315612182e-06, | |
| "loss": 0.5405, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3864734299516908, | |
| "grad_norm": 2.071517945939152, | |
| "learning_rate": 7.708225545829446e-06, | |
| "loss": 0.5331, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.39184111647879766, | |
| "grad_norm": 1.9667341052567295, | |
| "learning_rate": 7.6288749681321115e-06, | |
| "loss": 0.5276, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.39720880300590444, | |
| "grad_norm": 1.9124777050070318, | |
| "learning_rate": 7.548598528940354e-06, | |
| "loss": 0.5205, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4025764895330113, | |
| "grad_norm": 1.9540497432924613, | |
| "learning_rate": 7.4674245007528135e-06, | |
| "loss": 0.5293, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.40794417606011807, | |
| "grad_norm": 2.0287532132914716, | |
| "learning_rate": 7.385381472189321e-06, | |
| "loss": 0.5183, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4133118625872249, | |
| "grad_norm": 2.021246551625832, | |
| "learning_rate": 7.302498337922293e-06, | |
| "loss": 0.5214, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.41867954911433175, | |
| "grad_norm": 2.075086153282224, | |
| "learning_rate": 7.218804288500343e-06, | |
| "loss": 0.5051, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.42404723564143854, | |
| "grad_norm": 1.9853201300706842, | |
| "learning_rate": 7.134328800067684e-06, | |
| "loss": 0.5143, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4294149221685454, | |
| "grad_norm": 1.9075844135415194, | |
| "learning_rate": 7.049101623982938e-06, | |
| "loss": 0.5076, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 1.9963194922766025, | |
| "learning_rate": 6.963152776341044e-06, | |
| "loss": 0.5087, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.440150295222759, | |
| "grad_norm": 2.0510112086805083, | |
| "learning_rate": 6.876512527401897e-06, | |
| "loss": 0.5092, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4455179817498658, | |
| "grad_norm": 1.8923092211852284, | |
| "learning_rate": 6.789211390929497e-06, | |
| "loss": 0.4966, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.45088566827697263, | |
| "grad_norm": 2.015790172637043, | |
| "learning_rate": 6.701280113445324e-06, | |
| "loss": 0.4967, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4562533548040794, | |
| "grad_norm": 1.9134734917899814, | |
| "learning_rate": 6.6127496633997475e-06, | |
| "loss": 0.5039, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.46162104133118625, | |
| "grad_norm": 1.9388380453029543, | |
| "learning_rate": 6.523651220265269e-06, | |
| "loss": 0.4762, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4669887278582931, | |
| "grad_norm": 1.838971696143469, | |
| "learning_rate": 6.434016163555452e-06, | |
| "loss": 0.4894, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4723564143853999, | |
| "grad_norm": 1.9058911845628344, | |
| "learning_rate": 6.343876061773385e-06, | |
| "loss": 0.4858, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4777241009125067, | |
| "grad_norm": 2.027908561619023, | |
| "learning_rate": 6.2532626612936035e-06, | |
| "loss": 0.4829, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4830917874396135, | |
| "grad_norm": 1.9646108142819012, | |
| "learning_rate": 6.162207875181354e-06, | |
| "loss": 0.4838, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.48845947396672035, | |
| "grad_norm": 1.891035535041024, | |
| "learning_rate": 6.070743771953157e-06, | |
| "loss": 0.4815, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 1.9516352660221894, | |
| "learning_rate": 5.978902564282616e-06, | |
| "loss": 0.4811, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.499194847020934, | |
| "grad_norm": 2.07837231336269, | |
| "learning_rate": 5.886716597655472e-06, | |
| "loss": 0.4676, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5045625335480408, | |
| "grad_norm": 2.0338000671858443, | |
| "learning_rate": 5.7942183389778536e-06, | |
| "loss": 0.4751, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5099302200751477, | |
| "grad_norm": 1.9544187569972904, | |
| "learning_rate": 5.701440365141799e-06, | |
| "loss": 0.4747, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5152979066022544, | |
| "grad_norm": 2.0013443075194517, | |
| "learning_rate": 5.608415351552014e-06, | |
| "loss": 0.4737, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5206655931293612, | |
| "grad_norm": 2.392764011448066, | |
| "learning_rate": 5.515176060617945e-06, | |
| "loss": 0.4661, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5260332796564681, | |
| "grad_norm": 2.191779753390301, | |
| "learning_rate": 5.421755330215223e-06, | |
| "loss": 0.4633, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5314009661835749, | |
| "grad_norm": 1.9807734795602283, | |
| "learning_rate": 5.328186062120509e-06, | |
| "loss": 0.4675, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5367686527106817, | |
| "grad_norm": 1.8845793314459505, | |
| "learning_rate": 5.23450121042383e-06, | |
| "loss": 0.4485, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5421363392377885, | |
| "grad_norm": 1.9504286157116426, | |
| "learning_rate": 5.140733769922525e-06, | |
| "loss": 0.4492, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5475040257648953, | |
| "grad_norm": 1.9251970055484684, | |
| "learning_rate": 5.0469167645008245e-06, | |
| "loss": 0.4537, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5528717122920022, | |
| "grad_norm": 2.0900201692813503, | |
| "learning_rate": 4.953083235499177e-06, | |
| "loss": 0.4541, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.558239398819109, | |
| "grad_norm": 1.914274695560957, | |
| "learning_rate": 4.859266230077474e-06, | |
| "loss": 0.4416, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5636070853462157, | |
| "grad_norm": 2.0363927763394685, | |
| "learning_rate": 4.7654987895761705e-06, | |
| "loss": 0.4578, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5689747718733226, | |
| "grad_norm": 1.906876750353807, | |
| "learning_rate": 4.671813937879494e-06, | |
| "loss": 0.4445, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5743424584004294, | |
| "grad_norm": 1.9970794776707403, | |
| "learning_rate": 4.5782446697847775e-06, | |
| "loss": 0.4432, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 1.9626727656883254, | |
| "learning_rate": 4.484823939382056e-06, | |
| "loss": 0.436, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5850778314546431, | |
| "grad_norm": 1.936046454009477, | |
| "learning_rate": 4.391584648447989e-06, | |
| "loss": 0.4433, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5904455179817498, | |
| "grad_norm": 1.919712775449399, | |
| "learning_rate": 4.298559634858202e-06, | |
| "loss": 0.4397, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5958132045088567, | |
| "grad_norm": 1.882081042043698, | |
| "learning_rate": 4.205781661022146e-06, | |
| "loss": 0.4274, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.6011808910359635, | |
| "grad_norm": 1.9706554854402294, | |
| "learning_rate": 4.1132834023445304e-06, | |
| "loss": 0.4358, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6065485775630703, | |
| "grad_norm": 1.8892003217320965, | |
| "learning_rate": 4.021097435717386e-06, | |
| "loss": 0.4203, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6119162640901772, | |
| "grad_norm": 1.9387503523379348, | |
| "learning_rate": 3.929256228046845e-06, | |
| "loss": 0.4188, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 1.9263320255464111, | |
| "learning_rate": 3.837792124818647e-06, | |
| "loss": 0.4197, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6226516371443908, | |
| "grad_norm": 2.0679571176329636, | |
| "learning_rate": 3.7467373387063973e-06, | |
| "loss": 0.4104, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6280193236714976, | |
| "grad_norm": 1.88133441550826, | |
| "learning_rate": 3.656123938226618e-06, | |
| "loss": 0.4094, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6333870101986044, | |
| "grad_norm": 1.8992502262176936, | |
| "learning_rate": 3.5659838364445505e-06, | |
| "loss": 0.4126, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6387546967257112, | |
| "grad_norm": 1.8928527924901577, | |
| "learning_rate": 3.476348779734732e-06, | |
| "loss": 0.4245, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.644122383252818, | |
| "grad_norm": 1.8811095797880513, | |
| "learning_rate": 3.387250336600254e-06, | |
| "loss": 0.4157, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6494900697799249, | |
| "grad_norm": 2.1867786519525736, | |
| "learning_rate": 3.298719886554677e-06, | |
| "loss": 0.4029, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6548577563070317, | |
| "grad_norm": 1.9359388771852108, | |
| "learning_rate": 3.2107886090705035e-06, | |
| "loss": 0.4073, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6602254428341385, | |
| "grad_norm": 2.0767122537438834, | |
| "learning_rate": 3.1234874725981045e-06, | |
| "loss": 0.4105, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6655931293612453, | |
| "grad_norm": 1.9568967104623627, | |
| "learning_rate": 3.036847223658958e-06, | |
| "loss": 0.4166, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6709608158883521, | |
| "grad_norm": 1.8330033161355852, | |
| "learning_rate": 2.950898376017064e-06, | |
| "loss": 0.4007, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6763285024154589, | |
| "grad_norm": 1.9534116085299529, | |
| "learning_rate": 2.865671199932318e-06, | |
| "loss": 0.4094, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6816961889425658, | |
| "grad_norm": 1.824576200330308, | |
| "learning_rate": 2.781195711499658e-06, | |
| "loss": 0.3975, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6870638754696726, | |
| "grad_norm": 1.9397944016656081, | |
| "learning_rate": 2.697501662077707e-06, | |
| "loss": 0.3938, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6924315619967794, | |
| "grad_norm": 1.9405147376762144, | |
| "learning_rate": 2.6146185278106807e-06, | |
| "loss": 0.3903, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6977992485238862, | |
| "grad_norm": 1.8735166891783446, | |
| "learning_rate": 2.5325754992471886e-06, | |
| "loss": 0.3844, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.703166935050993, | |
| "grad_norm": 1.8550788085650558, | |
| "learning_rate": 2.4514014710596467e-06, | |
| "loss": 0.3895, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.7085346215780999, | |
| "grad_norm": 1.9768445053781223, | |
| "learning_rate": 2.3711250318678906e-06, | |
| "loss": 0.3976, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7139023081052066, | |
| "grad_norm": 1.8242433586161713, | |
| "learning_rate": 2.2917744541705544e-06, | |
| "loss": 0.3896, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.7192699946323134, | |
| "grad_norm": 1.8766511465002484, | |
| "learning_rate": 2.2133776843878185e-06, | |
| "loss": 0.3881, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7246376811594203, | |
| "grad_norm": 1.8567022542334322, | |
| "learning_rate": 2.1359623330189655e-06, | |
| "loss": 0.3826, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7300053676865271, | |
| "grad_norm": 1.8349926895031767, | |
| "learning_rate": 2.059555664918268e-06, | |
| "loss": 0.3888, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.735373054213634, | |
| "grad_norm": 1.7933412905107806, | |
| "learning_rate": 1.9841845896926022e-06, | |
| "loss": 0.3861, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 1.8516583877916601, | |
| "learning_rate": 1.9098756522241634e-06, | |
| "loss": 0.3793, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7461084272678475, | |
| "grad_norm": 1.8073808501807826, | |
| "learning_rate": 1.8366550233216584e-06, | |
| "loss": 0.3825, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7514761137949544, | |
| "grad_norm": 2.0053953972861183, | |
| "learning_rate": 1.7645484905032129e-06, | |
| "loss": 0.3795, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7568438003220612, | |
| "grad_norm": 1.7686721996317474, | |
| "learning_rate": 1.6935814489142937e-06, | |
| "loss": 0.3726, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7622114868491681, | |
| "grad_norm": 1.7884368236538568, | |
| "learning_rate": 1.6237788923838149e-06, | |
| "loss": 0.3681, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7675791733762748, | |
| "grad_norm": 1.8578617970297808, | |
| "learning_rate": 1.555165404621567e-06, | |
| "loss": 0.37, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7729468599033816, | |
| "grad_norm": 1.8740014923246073, | |
| "learning_rate": 1.487765150560116e-06, | |
| "loss": 0.3736, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7783145464304885, | |
| "grad_norm": 1.9238343180583926, | |
| "learning_rate": 1.4216018678441558e-06, | |
| "loss": 0.376, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7836822329575953, | |
| "grad_norm": 1.907761388210502, | |
| "learning_rate": 1.3566988584703817e-06, | |
| "loss": 0.3749, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.789049919484702, | |
| "grad_norm": 1.8440526734882297, | |
| "learning_rate": 1.293078980580766e-06, | |
| "loss": 0.3656, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7944176060118089, | |
| "grad_norm": 1.8162595531294223, | |
| "learning_rate": 1.2307646404121692e-06, | |
| "loss": 0.3635, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7997852925389157, | |
| "grad_norm": 1.8090188411422603, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.3706, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.8051529790660226, | |
| "grad_norm": 1.8548359213105685, | |
| "learning_rate": 1.1101398914744565e-06, | |
| "loss": 0.3541, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8105206655931294, | |
| "grad_norm": 1.7486366206671473, | |
| "learning_rate": 1.0518719654447896e-06, | |
| "loss": 0.3566, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.8158883521202361, | |
| "grad_norm": 1.7220786134987836, | |
| "learning_rate": 9.949945276530782e-07, | |
| "loss": 0.3537, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.821256038647343, | |
| "grad_norm": 1.8563952966735642, | |
| "learning_rate": 9.395276097212841e-07, | |
| "loss": 0.377, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.8266237251744498, | |
| "grad_norm": 1.8261975975646805, | |
| "learning_rate": 8.854907465014479e-07, | |
| "loss": 0.3564, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.8319914117015567, | |
| "grad_norm": 1.6700870770757226, | |
| "learning_rate": 8.329029691957124e-07, | |
| "loss": 0.3533, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8373590982286635, | |
| "grad_norm": 1.8714115240328053, | |
| "learning_rate": 7.817827986537508e-07, | |
| "loss": 0.3598, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8427267847557702, | |
| "grad_norm": 1.7265092801300406, | |
| "learning_rate": 7.321482388499096e-07, | |
| "loss": 0.3574, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8480944712828771, | |
| "grad_norm": 1.7192645034643335, | |
| "learning_rate": 6.840167705424106e-07, | |
| "loss": 0.3626, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8534621578099839, | |
| "grad_norm": 1.8788914839483877, | |
| "learning_rate": 6.374053451168166e-07, | |
| "loss": 0.3569, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8588298443370908, | |
| "grad_norm": 1.7223384963954047, | |
| "learning_rate": 5.92330378615929e-07, | |
| "loss": 0.3567, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8641975308641975, | |
| "grad_norm": 1.84633207462375, | |
| "learning_rate": 5.488077459582425e-07, | |
| "loss": 0.3568, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 1.8156905089011117, | |
| "learning_rate": 5.068527753469604e-07, | |
| "loss": 0.3553, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8749329039184112, | |
| "grad_norm": 1.76653198854091, | |
| "learning_rate": 4.664802428715753e-07, | |
| "loss": 0.353, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.880300590445518, | |
| "grad_norm": 1.746120045538455, | |
| "learning_rate": 4.2770436730388166e-07, | |
| "loss": 0.3548, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8856682769726248, | |
| "grad_norm": 1.7810142936647677, | |
| "learning_rate": 3.9053880509028086e-07, | |
| "loss": 0.3546, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8910359634997316, | |
| "grad_norm": 1.7315549203684755, | |
| "learning_rate": 3.549966455421305e-07, | |
| "loss": 0.3555, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8964036500268384, | |
| "grad_norm": 1.9272563135988092, | |
| "learning_rate": 3.2109040622582186e-07, | |
| "loss": 0.3521, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.9017713365539453, | |
| "grad_norm": 1.7798988754363725, | |
| "learning_rate": 2.8883202855423676e-07, | |
| "loss": 0.3506, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.9071390230810521, | |
| "grad_norm": 1.7625732551828204, | |
| "learning_rate": 2.582328735811029e-07, | |
| "loss": 0.3534, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.9125067096081588, | |
| "grad_norm": 1.827775623748613, | |
| "learning_rate": 2.2930371799975593e-07, | |
| "loss": 0.3469, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9178743961352657, | |
| "grad_norm": 1.7729865240896066, | |
| "learning_rate": 2.0205475034770606e-07, | |
| "loss": 0.3478, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.9232420826623725, | |
| "grad_norm": 1.7459386763584364, | |
| "learning_rate": 1.7649556741833995e-07, | |
| "loss": 0.3506, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9286097691894794, | |
| "grad_norm": 1.7638105650910396, | |
| "learning_rate": 1.5263517088103862e-07, | |
| "loss": 0.3501, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.9339774557165862, | |
| "grad_norm": 1.6634385736641513, | |
| "learning_rate": 1.304819641108801e-07, | |
| "loss": 0.3443, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9393451422436929, | |
| "grad_norm": 1.7138313811653363, | |
| "learning_rate": 1.1004374922906846e-07, | |
| "loss": 0.3519, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9447128287707998, | |
| "grad_norm": 1.7315956523188014, | |
| "learning_rate": 9.132772435510362e-08, | |
| "loss": 0.3518, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9500805152979066, | |
| "grad_norm": 1.83104419897327, | |
| "learning_rate": 7.434048107168523e-08, | |
| "loss": 0.3534, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.9554482018250134, | |
| "grad_norm": 1.8614069199591219, | |
| "learning_rate": 5.908800210322696e-08, | |
| "loss": 0.3529, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9608158883521203, | |
| "grad_norm": 1.791303700726749, | |
| "learning_rate": 4.55756592088058e-08, | |
| "loss": 0.3458, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.966183574879227, | |
| "grad_norm": 1.766280788851407, | |
| "learning_rate": 3.3808211290284886e-08, | |
| "loss": 0.3506, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9715512614063339, | |
| "grad_norm": 1.7183411966725215, | |
| "learning_rate": 2.378980271628195e-08, | |
| "loss": 0.3459, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9769189479334407, | |
| "grad_norm": 1.6247222017291731, | |
| "learning_rate": 1.552396186256411e-08, | |
| "loss": 0.3469, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9822866344605475, | |
| "grad_norm": 1.737776722109284, | |
| "learning_rate": 9.013599869394096e-09, | |
| "loss": 0.3432, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 1.6970200431542546, | |
| "learning_rate": 4.261009616257638e-09, | |
| "loss": 0.3502, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9930220075147611, | |
| "grad_norm": 1.700070440972857, | |
| "learning_rate": 1.2678649143349485e-09, | |
| "loss": 0.3439, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.998389694041868, | |
| "grad_norm": 1.7621651797235254, | |
| "learning_rate": 3.5219917003948003e-11, | |
| "loss": 0.3562, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9994632313472893, | |
| "eval_loss": 0.2994081974029541, | |
| "eval_runtime": 1.1873, | |
| "eval_samples_per_second": 1.684, | |
| "eval_steps_per_second": 0.842, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.9994632313472893, | |
| "step": 931, | |
| "total_flos": 194827769610240.0, | |
| "train_loss": 0.5084374696496478, | |
| "train_runtime": 18282.9087, | |
| "train_samples_per_second": 1.63, | |
| "train_steps_per_second": 0.051 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 931, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 194827769610240.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |