| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1578, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.019011406844106463, | |
| "grad_norm": 9.1989966206838, | |
| "learning_rate": 6.329113924050634e-07, | |
| "loss": 0.8874, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03802281368821293, | |
| "grad_norm": 4.161122432720083, | |
| "learning_rate": 1.2658227848101267e-06, | |
| "loss": 0.791, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.057034220532319393, | |
| "grad_norm": 1.6740419759238627, | |
| "learning_rate": 1.8987341772151901e-06, | |
| "loss": 0.7354, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07604562737642585, | |
| "grad_norm": 1.472773729260504, | |
| "learning_rate": 2.5316455696202535e-06, | |
| "loss": 0.6972, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09505703422053231, | |
| "grad_norm": 1.6632418888769052, | |
| "learning_rate": 3.164556962025317e-06, | |
| "loss": 0.698, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11406844106463879, | |
| "grad_norm": 1.2299395567832287, | |
| "learning_rate": 3.7974683544303802e-06, | |
| "loss": 0.6778, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13307984790874525, | |
| "grad_norm": 1.5380428593757542, | |
| "learning_rate": 4.430379746835443e-06, | |
| "loss": 0.6751, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1520912547528517, | |
| "grad_norm": 1.3704564329134865, | |
| "learning_rate": 4.999995058613287e-06, | |
| "loss": 0.6713, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17110266159695817, | |
| "grad_norm": 1.9490540655863884, | |
| "learning_rate": 4.999402118469357e-06, | |
| "loss": 0.6671, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19011406844106463, | |
| "grad_norm": 1.5630755002882903, | |
| "learning_rate": 4.997821199394829e-06, | |
| "loss": 0.6676, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.20912547528517111, | |
| "grad_norm": 2.133710896815452, | |
| "learning_rate": 4.995252995758543e-06, | |
| "loss": 0.6695, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.22813688212927757, | |
| "grad_norm": 1.5943526267612576, | |
| "learning_rate": 4.99169863556294e-06, | |
| "loss": 0.6575, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.24714828897338403, | |
| "grad_norm": 1.832479949115814, | |
| "learning_rate": 4.987159679948624e-06, | |
| "loss": 0.6612, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2661596958174905, | |
| "grad_norm": 1.4945020437443977, | |
| "learning_rate": 4.981638122508684e-06, | |
| "loss": 0.658, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.28517110266159695, | |
| "grad_norm": 1.619599628436461, | |
| "learning_rate": 4.975136388413065e-06, | |
| "loss": 0.6656, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3041825095057034, | |
| "grad_norm": 1.5360935621329557, | |
| "learning_rate": 4.967657333343394e-06, | |
| "loss": 0.6511, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3231939163498099, | |
| "grad_norm": 1.4510475916499845, | |
| "learning_rate": 4.959204242238707e-06, | |
| "loss": 0.6573, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.34220532319391633, | |
| "grad_norm": 1.532931462135121, | |
| "learning_rate": 4.949780827852648e-06, | |
| "loss": 0.6507, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3612167300380228, | |
| "grad_norm": 1.1719507987669484, | |
| "learning_rate": 4.939391229122757e-06, | |
| "loss": 0.6523, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.38022813688212925, | |
| "grad_norm": 1.2585592582289158, | |
| "learning_rate": 4.928040009352568e-06, | |
| "loss": 0.6524, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.39923954372623577, | |
| "grad_norm": 1.7296360428483535, | |
| "learning_rate": 4.91573215420733e-06, | |
| "loss": 0.6614, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.41825095057034223, | |
| "grad_norm": 1.5691004380249787, | |
| "learning_rate": 4.902473069524204e-06, | |
| "loss": 0.6484, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4372623574144487, | |
| "grad_norm": 1.247775759644616, | |
| "learning_rate": 4.888268578937923e-06, | |
| "loss": 0.6508, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.45627376425855515, | |
| "grad_norm": 1.3594034920568745, | |
| "learning_rate": 4.873124921322945e-06, | |
| "loss": 0.6469, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4752851711026616, | |
| "grad_norm": 1.2309510240137682, | |
| "learning_rate": 4.8570487480532196e-06, | |
| "loss": 0.6401, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.49429657794676807, | |
| "grad_norm": 1.3299272061400234, | |
| "learning_rate": 4.840047120080787e-06, | |
| "loss": 0.6416, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5133079847908745, | |
| "grad_norm": 1.3129636158199345, | |
| "learning_rate": 4.822127504834472e-06, | |
| "loss": 0.6438, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.532319391634981, | |
| "grad_norm": 1.4703016714267874, | |
| "learning_rate": 4.8032977729400585e-06, | |
| "loss": 0.6506, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5513307984790875, | |
| "grad_norm": 1.3282370638504304, | |
| "learning_rate": 4.783566194763359e-06, | |
| "loss": 0.6521, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5703422053231939, | |
| "grad_norm": 1.187364149838104, | |
| "learning_rate": 4.762941436777721e-06, | |
| "loss": 0.6371, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5893536121673004, | |
| "grad_norm": 1.2164074784365948, | |
| "learning_rate": 4.7414325577575484e-06, | |
| "loss": 0.6407, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6083650190114068, | |
| "grad_norm": 1.4208300435315424, | |
| "learning_rate": 4.719049004799525e-06, | |
| "loss": 0.6328, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6273764258555133, | |
| "grad_norm": 1.324349326714873, | |
| "learning_rate": 4.695800609173274e-06, | |
| "loss": 0.6526, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6463878326996197, | |
| "grad_norm": 1.1238559883921406, | |
| "learning_rate": 4.671697582003279e-06, | |
| "loss": 0.6371, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6653992395437263, | |
| "grad_norm": 1.3772583883883136, | |
| "learning_rate": 4.646750509783975e-06, | |
| "loss": 0.6413, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6844106463878327, | |
| "grad_norm": 1.3311507053384555, | |
| "learning_rate": 4.620970349729961e-06, | |
| "loss": 0.638, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7034220532319392, | |
| "grad_norm": 1.1084009519269555, | |
| "learning_rate": 4.594368424963392e-06, | |
| "loss": 0.6402, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7224334600760456, | |
| "grad_norm": 1.272517045012054, | |
| "learning_rate": 4.56695641954065e-06, | |
| "loss": 0.6332, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7414448669201521, | |
| "grad_norm": 1.068763291195587, | |
| "learning_rate": 4.538746373320499e-06, | |
| "loss": 0.6451, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7604562737642585, | |
| "grad_norm": 1.17609263311229, | |
| "learning_rate": 4.5097506766759465e-06, | |
| "loss": 0.6398, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.779467680608365, | |
| "grad_norm": 1.2993920513145316, | |
| "learning_rate": 4.479982065052171e-06, | |
| "loss": 0.6333, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7984790874524715, | |
| "grad_norm": 1.4521328868048515, | |
| "learning_rate": 4.4494536133728754e-06, | |
| "loss": 0.6389, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8174904942965779, | |
| "grad_norm": 1.134966447023594, | |
| "learning_rate": 4.418178730297542e-06, | |
| "loss": 0.6369, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8365019011406845, | |
| "grad_norm": 1.1802112020601854, | |
| "learning_rate": 4.3861711523321e-06, | |
| "loss": 0.6355, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8555133079847909, | |
| "grad_norm": 1.2438925294191414, | |
| "learning_rate": 4.353444937795595e-06, | |
| "loss": 0.6337, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8745247148288974, | |
| "grad_norm": 1.169622958547474, | |
| "learning_rate": 4.320014460645523e-06, | |
| "loss": 0.6357, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8935361216730038, | |
| "grad_norm": 1.0360592829770976, | |
| "learning_rate": 4.2858944041645196e-06, | |
| "loss": 0.6308, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9125475285171103, | |
| "grad_norm": 1.133882806727274, | |
| "learning_rate": 4.251099754511189e-06, | |
| "loss": 0.633, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9315589353612167, | |
| "grad_norm": 1.0230579665487711, | |
| "learning_rate": 4.2156457941379095e-06, | |
| "loss": 0.637, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9505703422053232, | |
| "grad_norm": 1.0897950528236497, | |
| "learning_rate": 4.179548095078498e-06, | |
| "loss": 0.6315, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9695817490494296, | |
| "grad_norm": 1.4631268078629502, | |
| "learning_rate": 4.142822512108683e-06, | |
| "loss": 0.632, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.9885931558935361, | |
| "grad_norm": 1.2859460507714569, | |
| "learning_rate": 4.105485175782396e-06, | |
| "loss": 0.6226, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.6353716254234314, | |
| "eval_runtime": 49.2084, | |
| "eval_samples_per_second": 287.816, | |
| "eval_steps_per_second": 1.138, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 1.0076045627376427, | |
| "grad_norm": 2.6186402518104304, | |
| "learning_rate": 4.067552485346939e-06, | |
| "loss": 0.5901, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.026615969581749, | |
| "grad_norm": 1.4429614147480494, | |
| "learning_rate": 4.029041101540122e-06, | |
| "loss": 0.546, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.0456273764258555, | |
| "grad_norm": 1.1901165511323117, | |
| "learning_rate": 3.989967939272569e-06, | |
| "loss": 0.5434, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.064638783269962, | |
| "grad_norm": 1.133670318062506, | |
| "learning_rate": 3.950350160198373e-06, | |
| "loss": 0.5365, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.0836501901140685, | |
| "grad_norm": 1.1569131629508171, | |
| "learning_rate": 3.91020516517738e-06, | |
| "loss": 0.5394, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.102661596958175, | |
| "grad_norm": 1.1578867626444747, | |
| "learning_rate": 3.869550586632413e-06, | |
| "loss": 0.5454, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.1216730038022813, | |
| "grad_norm": 1.1710585399638043, | |
| "learning_rate": 3.828404280804782e-06, | |
| "loss": 0.5437, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.1406844106463878, | |
| "grad_norm": 1.2968456937546469, | |
| "learning_rate": 3.7867843199114996e-06, | |
| "loss": 0.545, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.1596958174904943, | |
| "grad_norm": 1.173788116358311, | |
| "learning_rate": 3.7447089842076238e-06, | |
| "loss": 0.5401, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.1787072243346008, | |
| "grad_norm": 1.1736747473104372, | |
| "learning_rate": 3.7021967539572343e-06, | |
| "loss": 0.5451, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.1977186311787071, | |
| "grad_norm": 1.1583258876929678, | |
| "learning_rate": 3.6592663013165636e-06, | |
| "loss": 0.5413, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.2167300380228137, | |
| "grad_norm": 1.1187865805782111, | |
| "learning_rate": 3.6159364821328325e-06, | |
| "loss": 0.5531, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.2357414448669202, | |
| "grad_norm": 1.2500694936887193, | |
| "learning_rate": 3.572226327662423e-06, | |
| "loss": 0.5481, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.2547528517110267, | |
| "grad_norm": 1.2649269376886392, | |
| "learning_rate": 3.528155036211995e-06, | |
| "loss": 0.5402, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.2737642585551332, | |
| "grad_norm": 1.2215327457350609, | |
| "learning_rate": 3.4837419647062344e-06, | |
| "loss": 0.5452, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.2927756653992395, | |
| "grad_norm": 1.2145466936928535, | |
| "learning_rate": 3.4390066201859355e-06, | |
| "loss": 0.5376, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.311787072243346, | |
| "grad_norm": 1.16823342480946, | |
| "learning_rate": 3.393968651240153e-06, | |
| "loss": 0.5442, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.3307984790874525, | |
| "grad_norm": 1.1720513921800486, | |
| "learning_rate": 3.3486478393761707e-06, | |
| "loss": 0.5449, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.3498098859315588, | |
| "grad_norm": 1.32127207540034, | |
| "learning_rate": 3.303064090331106e-06, | |
| "loss": 0.537, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.3688212927756653, | |
| "grad_norm": 1.170488361503433, | |
| "learning_rate": 3.2572374253289385e-06, | |
| "loss": 0.5525, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.3878326996197718, | |
| "grad_norm": 1.2540643595664949, | |
| "learning_rate": 3.2111879722868204e-06, | |
| "loss": 0.5513, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.4068441064638784, | |
| "grad_norm": 1.1125376049437519, | |
| "learning_rate": 3.164935956974522e-06, | |
| "loss": 0.5428, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.4258555133079849, | |
| "grad_norm": 1.2222128167327286, | |
| "learning_rate": 3.1185016941309033e-06, | |
| "loss": 0.5443, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.4448669201520912, | |
| "grad_norm": 1.1189014403528423, | |
| "learning_rate": 3.071905578541304e-06, | |
| "loss": 0.55, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.4638783269961977, | |
| "grad_norm": 1.106630875988617, | |
| "learning_rate": 3.0251680760797737e-06, | |
| "loss": 0.5512, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.4828897338403042, | |
| "grad_norm": 1.238247599173364, | |
| "learning_rate": 2.978309714720092e-06, | |
| "loss": 0.5457, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.5019011406844105, | |
| "grad_norm": 1.1877938363765261, | |
| "learning_rate": 2.931351075519494e-06, | |
| "loss": 0.5472, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.5209125475285172, | |
| "grad_norm": 1.0765509288814237, | |
| "learning_rate": 2.8843127835790922e-06, | |
| "loss": 0.5467, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.5399239543726235, | |
| "grad_norm": 1.1769551690370097, | |
| "learning_rate": 2.837215498984956e-06, | |
| "loss": 0.5423, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.55893536121673, | |
| "grad_norm": 1.0525687613765247, | |
| "learning_rate": 2.7900799077338137e-06, | |
| "loss": 0.5377, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.5779467680608366, | |
| "grad_norm": 1.1872738975990165, | |
| "learning_rate": 2.742926712647389e-06, | |
| "loss": 0.5422, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.5969581749049429, | |
| "grad_norm": 1.2556831499142438, | |
| "learning_rate": 2.6957766242793286e-06, | |
| "loss": 0.5548, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.6159695817490496, | |
| "grad_norm": 1.0530666465416862, | |
| "learning_rate": 2.648650351818758e-06, | |
| "loss": 0.5396, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.6349809885931559, | |
| "grad_norm": 1.0626505034866265, | |
| "learning_rate": 2.6015685939944113e-06, | |
| "loss": 0.5443, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.6539923954372624, | |
| "grad_norm": 1.02476310680603, | |
| "learning_rate": 2.554552029983375e-06, | |
| "loss": 0.5404, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.673003802281369, | |
| "grad_norm": 1.0885196976387683, | |
| "learning_rate": 2.5076213103284107e-06, | |
| "loss": 0.5443, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.6920152091254752, | |
| "grad_norm": 1.1420366693644837, | |
| "learning_rate": 2.4607970478678515e-06, | |
| "loss": 0.5464, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.7110266159695817, | |
| "grad_norm": 1.1255789898459938, | |
| "learning_rate": 2.4140998086820665e-06, | |
| "loss": 0.5447, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.7300380228136882, | |
| "grad_norm": 1.3189103855357065, | |
| "learning_rate": 2.367550103060459e-06, | |
| "loss": 0.5475, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.7490494296577945, | |
| "grad_norm": 1.0465751177563847, | |
| "learning_rate": 2.3211683764929664e-06, | |
| "loss": 0.5484, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.7680608365019013, | |
| "grad_norm": 1.1704397396223116, | |
| "learning_rate": 2.2749750006900263e-06, | |
| "loss": 0.5431, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.7870722433460076, | |
| "grad_norm": 1.053213291008937, | |
| "learning_rate": 2.2289902646349423e-06, | |
| "loss": 0.5425, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.806083650190114, | |
| "grad_norm": 1.2515940528618232, | |
| "learning_rate": 2.1832343656725918e-06, | |
| "loss": 0.5459, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.8250950570342206, | |
| "grad_norm": 1.099127233957744, | |
| "learning_rate": 2.1377274006383737e-06, | |
| "loss": 0.5394, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.8441064638783269, | |
| "grad_norm": 1.12063559389163, | |
| "learning_rate": 2.092489357031315e-06, | |
| "loss": 0.5447, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.8631178707224336, | |
| "grad_norm": 1.102123448980619, | |
| "learning_rate": 2.0475401042351843e-06, | |
| "loss": 0.5305, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.88212927756654, | |
| "grad_norm": 1.1254624333919672, | |
| "learning_rate": 2.0028993847914966e-06, | |
| "loss": 0.5395, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.9011406844106464, | |
| "grad_norm": 1.0910835517042778, | |
| "learning_rate": 1.9585868057282256e-06, | |
| "loss": 0.5406, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.920152091254753, | |
| "grad_norm": 1.3820944438954628, | |
| "learning_rate": 1.9146218299480303e-06, | |
| "loss": 0.5476, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.9391634980988592, | |
| "grad_norm": 1.2180866671022355, | |
| "learning_rate": 1.8710237676797923e-06, | |
| "loss": 0.5429, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.9581749049429658, | |
| "grad_norm": 1.1380236136327964, | |
| "learning_rate": 1.827811767997207e-06, | |
| "loss": 0.5419, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.9771863117870723, | |
| "grad_norm": 1.0602009694403736, | |
| "learning_rate": 1.7850048104081555e-06, | |
| "loss": 0.5369, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.9961977186311786, | |
| "grad_norm": 1.0393675864746768, | |
| "learning_rate": 1.742621696518555e-06, | |
| "loss": 0.5371, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.6316379308700562, | |
| "eval_runtime": 54.2509, | |
| "eval_samples_per_second": 261.065, | |
| "eval_steps_per_second": 1.032, | |
| "step": 1052 | |
| }, | |
| { | |
| "epoch": 2.0152091254752853, | |
| "grad_norm": 1.7644212761083442, | |
| "learning_rate": 1.7006810417743453e-06, | |
| "loss": 0.4716, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.0342205323193916, | |
| "grad_norm": 1.3479439652442078, | |
| "learning_rate": 1.6592012672852462e-06, | |
| "loss": 0.4592, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.053231939163498, | |
| "grad_norm": 1.1555574696681938, | |
| "learning_rate": 1.618200591733858e-06, | |
| "loss": 0.4568, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.0722433460076046, | |
| "grad_norm": 1.1655620913425222, | |
| "learning_rate": 1.5776970233736877e-06, | |
| "loss": 0.4649, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.091254752851711, | |
| "grad_norm": 1.246797649914118, | |
| "learning_rate": 1.5377083521195849e-06, | |
| "loss": 0.4584, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.1102661596958177, | |
| "grad_norm": 1.20981005283122, | |
| "learning_rate": 1.4982521417340881e-06, | |
| "loss": 0.457, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.129277566539924, | |
| "grad_norm": 1.25314570539992, | |
| "learning_rate": 1.459345722113095e-06, | |
| "loss": 0.4613, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.1482889733840302, | |
| "grad_norm": 1.2661139969932325, | |
| "learning_rate": 1.4210061816742487e-06, | |
| "loss": 0.4557, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.167300380228137, | |
| "grad_norm": 1.22533837512988, | |
| "learning_rate": 1.3832503598513939e-06, | |
| "loss": 0.4638, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.1863117870722433, | |
| "grad_norm": 1.153296199976961, | |
| "learning_rate": 1.3460948396983833e-06, | |
| "loss": 0.4539, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.20532319391635, | |
| "grad_norm": 1.2554665358203887, | |
| "learning_rate": 1.3095559406054964e-06, | |
| "loss": 0.4574, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.2243346007604563, | |
| "grad_norm": 1.216327809593985, | |
| "learning_rate": 1.2736497111316607e-06, | |
| "loss": 0.4595, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.2433460076045626, | |
| "grad_norm": 1.1472080286252149, | |
| "learning_rate": 1.238391921955631e-06, | |
| "loss": 0.4565, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.2623574144486693, | |
| "grad_norm": 1.280760900457142, | |
| "learning_rate": 1.2037980589492116e-06, | |
| "loss": 0.4623, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.2813688212927756, | |
| "grad_norm": 1.1694311311925278, | |
| "learning_rate": 1.1698833163755793e-06, | |
| "loss": 0.4531, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.3003802281368824, | |
| "grad_norm": 1.1799865250234591, | |
| "learning_rate": 1.1366625902156814e-06, | |
| "loss": 0.4555, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.3193916349809887, | |
| "grad_norm": 1.1721763967059378, | |
| "learning_rate": 1.1041504716256456e-06, | |
| "loss": 0.4601, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.338403041825095, | |
| "grad_norm": 1.1882316774789523, | |
| "learning_rate": 1.0723612405280762e-06, | |
| "loss": 0.4669, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.3574144486692017, | |
| "grad_norm": 1.1388831508699453, | |
| "learning_rate": 1.0413088593400464e-06, | |
| "loss": 0.4632, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.376425855513308, | |
| "grad_norm": 1.143656084527738, | |
| "learning_rate": 1.0110069668405487e-06, | |
| "loss": 0.4564, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.3954372623574143, | |
| "grad_norm": 1.210491576650488, | |
| "learning_rate": 9.8146887218009e-07, | |
| "loss": 0.4639, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.414448669201521, | |
| "grad_norm": 1.1273169371405412, | |
| "learning_rate": 9.527075490350695e-07, | |
| "loss": 0.4615, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.4334600760456273, | |
| "grad_norm": 1.215505001410675, | |
| "learning_rate": 9.247356299094966e-07, | |
| "loss": 0.4528, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.4524714828897336, | |
| "grad_norm": 1.1389739744110048, | |
| "learning_rate": 8.975654005865671e-07, | |
| "loss": 0.4535, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.4714828897338403, | |
| "grad_norm": 1.159288850094368, | |
| "learning_rate": 8.712087947325155e-07, | |
| "loss": 0.4514, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.4904942965779466, | |
| "grad_norm": 1.2097686536353103, | |
| "learning_rate": 8.456773886551317e-07, | |
| "loss": 0.456, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.5095057034220534, | |
| "grad_norm": 1.125891025891593, | |
| "learning_rate": 8.20982396219231e-07, | |
| "loss": 0.4584, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.5285171102661597, | |
| "grad_norm": 1.165727380934779, | |
| "learning_rate": 7.971346639213192e-07, | |
| "loss": 0.4576, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.5475285171102664, | |
| "grad_norm": 1.2196983806300021, | |
| "learning_rate": 7.74144666125611e-07, | |
| "loss": 0.4585, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.5665399239543727, | |
| "grad_norm": 1.1416376267908626, | |
| "learning_rate": 7.520225004634946e-07, | |
| "loss": 0.4617, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.585551330798479, | |
| "grad_norm": 1.1374660560345522, | |
| "learning_rate": 7.307778833984666e-07, | |
| "loss": 0.4525, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.6045627376425857, | |
| "grad_norm": 1.1600478528127594, | |
| "learning_rate": 7.104201459584814e-07, | |
| "loss": 0.4554, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.623574144486692, | |
| "grad_norm": 1.1613018240717523, | |
| "learning_rate": 6.909582296375923e-07, | |
| "loss": 0.4571, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.6425855513307983, | |
| "grad_norm": 1.1582568171040142, | |
| "learning_rate": 6.724006824686814e-07, | |
| "loss": 0.458, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.661596958174905, | |
| "grad_norm": 1.2297405479958352, | |
| "learning_rate": 6.547556552690069e-07, | |
| "loss": 0.4576, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.6806083650190113, | |
| "grad_norm": 1.1545365546618969, | |
| "learning_rate": 6.380308980602119e-07, | |
| "loss": 0.4559, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.6996197718631176, | |
| "grad_norm": 1.1501358792075553, | |
| "learning_rate": 6.222337566643733e-07, | |
| "loss": 0.458, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.7186311787072244, | |
| "grad_norm": 1.1596268149561126, | |
| "learning_rate": 6.073711694775805e-07, | |
| "loss": 0.4523, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.7376425855513307, | |
| "grad_norm": 1.176536310467956, | |
| "learning_rate": 5.934496644224642e-07, | |
| "loss": 0.4592, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.7566539923954374, | |
| "grad_norm": 1.1601201945564672, | |
| "learning_rate": 5.804753560810128e-07, | |
| "loss": 0.4587, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.7756653992395437, | |
| "grad_norm": 1.1503887080909905, | |
| "learning_rate": 5.684539430089329e-07, | |
| "loss": 0.4558, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.7946768060836504, | |
| "grad_norm": 1.148286520523666, | |
| "learning_rate": 5.573907052327413e-07, | |
| "loss": 0.4585, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.8136882129277567, | |
| "grad_norm": 1.1477000696419715, | |
| "learning_rate": 5.472905019306775e-07, | |
| "loss": 0.4555, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.832699619771863, | |
| "grad_norm": 1.1636191591442357, | |
| "learning_rate": 5.38157769298465e-07, | |
| "loss": 0.4529, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.8517110266159698, | |
| "grad_norm": 1.226786969124158, | |
| "learning_rate": 5.2999651860085e-07, | |
| "loss": 0.4628, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.870722433460076, | |
| "grad_norm": 1.2111320995578168, | |
| "learning_rate": 5.228103344097825e-07, | |
| "loss": 0.4642, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.8897338403041823, | |
| "grad_norm": 1.134156671769776, | |
| "learning_rate": 5.166023730300036e-07, | |
| "loss": 0.4541, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.908745247148289, | |
| "grad_norm": 1.1404768838000066, | |
| "learning_rate": 5.113753611127401e-07, | |
| "loss": 0.4607, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.9277566539923954, | |
| "grad_norm": 1.1667213786402724, | |
| "learning_rate": 5.071315944581075e-07, | |
| "loss": 0.4638, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.9467680608365017, | |
| "grad_norm": 1.151294180691113, | |
| "learning_rate": 5.038729370067536e-07, | |
| "loss": 0.455, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.9657794676806084, | |
| "grad_norm": 1.2244552822051, | |
| "learning_rate": 5.016008200211805e-07, | |
| "loss": 0.4556, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.9847908745247147, | |
| "grad_norm": 1.1748003065525463, | |
| "learning_rate": 5.003162414571072e-07, | |
| "loss": 0.455, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.6586939692497253, | |
| "eval_runtime": 55.4914, | |
| "eval_samples_per_second": 255.229, | |
| "eval_steps_per_second": 1.009, | |
| "step": 1578 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1578, | |
| "total_flos": 2643208773304320.0, | |
| "train_loss": 0.5529848988367426, | |
| "train_runtime": 9902.4641, | |
| "train_samples_per_second": 81.519, | |
| "train_steps_per_second": 0.159 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1578, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 2643208773304320.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |