| { | |
| "best_metric": 0.0047513521276414394, | |
| "best_model_checkpoint": "finetuned-arsenic/checkpoint-2000", | |
| "epoch": 4.0, | |
| "eval_steps": 100, | |
| "global_step": 2164, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.018484288354898338, | |
| "grad_norm": 4.949392795562744, | |
| "learning_rate": 0.0001990757855822551, | |
| "loss": 0.5368, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.036968576709796676, | |
| "grad_norm": 3.3969953060150146, | |
| "learning_rate": 0.00019815157116451017, | |
| "loss": 0.3313, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05545286506469501, | |
| "grad_norm": 0.859575629234314, | |
| "learning_rate": 0.00019722735674676528, | |
| "loss": 0.5003, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07393715341959335, | |
| "grad_norm": 5.522923946380615, | |
| "learning_rate": 0.00019630314232902034, | |
| "loss": 0.2564, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09242144177449169, | |
| "grad_norm": 4.462332248687744, | |
| "learning_rate": 0.00019537892791127544, | |
| "loss": 0.3339, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11090573012939002, | |
| "grad_norm": 1.6224160194396973, | |
| "learning_rate": 0.0001944547134935305, | |
| "loss": 0.3965, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.12939001848428835, | |
| "grad_norm": 6.097796440124512, | |
| "learning_rate": 0.0001935304990757856, | |
| "loss": 0.3319, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1478743068391867, | |
| "grad_norm": 3.9769697189331055, | |
| "learning_rate": 0.00019260628465804066, | |
| "loss": 0.4012, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.16635859519408502, | |
| "grad_norm": 2.335510730743408, | |
| "learning_rate": 0.00019168207024029577, | |
| "loss": 0.4584, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.18484288354898337, | |
| "grad_norm": 3.8701980113983154, | |
| "learning_rate": 0.00019075785582255082, | |
| "loss": 0.1855, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18484288354898337, | |
| "eval_accuracy": 0.931237721021611, | |
| "eval_loss": 0.1917603313922882, | |
| "eval_runtime": 57.9367, | |
| "eval_samples_per_second": 26.356, | |
| "eval_steps_per_second": 3.297, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2033271719038817, | |
| "grad_norm": 2.2155282497406006, | |
| "learning_rate": 0.00018983364140480593, | |
| "loss": 0.2331, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.22181146025878004, | |
| "grad_norm": 0.9634373188018799, | |
| "learning_rate": 0.000188909426987061, | |
| "loss": 0.209, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.24029574861367836, | |
| "grad_norm": 0.2715567648410797, | |
| "learning_rate": 0.0001879852125693161, | |
| "loss": 0.1486, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2587800369685767, | |
| "grad_norm": 12.090089797973633, | |
| "learning_rate": 0.00018706099815157118, | |
| "loss": 0.1629, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.27726432532347506, | |
| "grad_norm": 1.551562786102295, | |
| "learning_rate": 0.00018613678373382626, | |
| "loss": 0.1852, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2957486136783734, | |
| "grad_norm": 0.775977373123169, | |
| "learning_rate": 0.00018521256931608134, | |
| "loss": 0.3179, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3142329020332717, | |
| "grad_norm": 3.0043396949768066, | |
| "learning_rate": 0.00018428835489833642, | |
| "loss": 0.3842, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.33271719038817005, | |
| "grad_norm": 1.2949095964431763, | |
| "learning_rate": 0.0001833641404805915, | |
| "loss": 0.2534, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3512014787430684, | |
| "grad_norm": 9.545828819274902, | |
| "learning_rate": 0.00018243992606284658, | |
| "loss": 0.2031, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.36968576709796674, | |
| "grad_norm": 0.29387930035591125, | |
| "learning_rate": 0.0001815157116451017, | |
| "loss": 0.1792, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.36968576709796674, | |
| "eval_accuracy": 0.9364767518009168, | |
| "eval_loss": 0.17399875819683075, | |
| "eval_runtime": 52.9831, | |
| "eval_samples_per_second": 28.821, | |
| "eval_steps_per_second": 3.605, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.38817005545286504, | |
| "grad_norm": 2.138578414916992, | |
| "learning_rate": 0.00018059149722735675, | |
| "loss": 0.2129, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4066543438077634, | |
| "grad_norm": 2.022083282470703, | |
| "learning_rate": 0.00017966728280961186, | |
| "loss": 0.1577, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.42513863216266173, | |
| "grad_norm": 2.8811872005462646, | |
| "learning_rate": 0.0001787430683918669, | |
| "loss": 0.21, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4436229205175601, | |
| "grad_norm": 1.491790771484375, | |
| "learning_rate": 0.00017781885397412202, | |
| "loss": 0.2498, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.46210720887245843, | |
| "grad_norm": 2.5274643898010254, | |
| "learning_rate": 0.00017689463955637707, | |
| "loss": 0.149, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4805914972273567, | |
| "grad_norm": 0.6268563270568848, | |
| "learning_rate": 0.00017597042513863218, | |
| "loss": 0.1306, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.49907578558225507, | |
| "grad_norm": 6.4418511390686035, | |
| "learning_rate": 0.00017504621072088724, | |
| "loss": 0.1889, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5175600739371534, | |
| "grad_norm": 0.13176225125789642, | |
| "learning_rate": 0.00017412199630314234, | |
| "loss": 0.1304, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5360443622920518, | |
| "grad_norm": 1.4023276567459106, | |
| "learning_rate": 0.00017319778188539743, | |
| "loss": 0.0872, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5545286506469501, | |
| "grad_norm": 5.165181636810303, | |
| "learning_rate": 0.0001722735674676525, | |
| "loss": 0.1688, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5545286506469501, | |
| "eval_accuracy": 0.9692206941715783, | |
| "eval_loss": 0.078226737678051, | |
| "eval_runtime": 52.9719, | |
| "eval_samples_per_second": 28.827, | |
| "eval_steps_per_second": 3.606, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5730129390018485, | |
| "grad_norm": 4.743193626403809, | |
| "learning_rate": 0.0001713493530499076, | |
| "loss": 0.1222, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5914972273567468, | |
| "grad_norm": 3.3770973682403564, | |
| "learning_rate": 0.00017042513863216267, | |
| "loss": 0.2799, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.609981515711645, | |
| "grad_norm": 1.9085370302200317, | |
| "learning_rate": 0.00016950092421441775, | |
| "loss": 0.1779, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6284658040665434, | |
| "grad_norm": 2.592458963394165, | |
| "learning_rate": 0.00016857670979667283, | |
| "loss": 0.1619, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6469500924214417, | |
| "grad_norm": 1.1735055446624756, | |
| "learning_rate": 0.00016765249537892791, | |
| "loss": 0.4249, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6654343807763401, | |
| "grad_norm": 3.8289904594421387, | |
| "learning_rate": 0.000166728280961183, | |
| "loss": 0.1009, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6839186691312384, | |
| "grad_norm": 2.531283378601074, | |
| "learning_rate": 0.00016580406654343808, | |
| "loss": 0.1494, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7024029574861368, | |
| "grad_norm": 0.21572425961494446, | |
| "learning_rate": 0.00016487985212569316, | |
| "loss": 0.0824, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7208872458410351, | |
| "grad_norm": 3.6041758060455322, | |
| "learning_rate": 0.00016395563770794827, | |
| "loss": 0.1145, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7393715341959335, | |
| "grad_norm": 0.6018674969673157, | |
| "learning_rate": 0.00016303142329020332, | |
| "loss": 0.1238, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7393715341959335, | |
| "eval_accuracy": 0.922724296005239, | |
| "eval_loss": 0.21575002372264862, | |
| "eval_runtime": 52.6224, | |
| "eval_samples_per_second": 29.018, | |
| "eval_steps_per_second": 3.63, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7578558225508318, | |
| "grad_norm": 0.25093191862106323, | |
| "learning_rate": 0.00016210720887245843, | |
| "loss": 0.0724, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.7763401109057301, | |
| "grad_norm": 0.2480381280183792, | |
| "learning_rate": 0.00016118299445471348, | |
| "loss": 0.106, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7948243992606284, | |
| "grad_norm": 8.212138175964355, | |
| "learning_rate": 0.0001602587800369686, | |
| "loss": 0.1665, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8133086876155268, | |
| "grad_norm": 0.6615661382675171, | |
| "learning_rate": 0.00015933456561922367, | |
| "loss": 0.0547, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8317929759704251, | |
| "grad_norm": 4.98212194442749, | |
| "learning_rate": 0.00015841035120147876, | |
| "loss": 0.1982, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8502772643253235, | |
| "grad_norm": 1.7662006616592407, | |
| "learning_rate": 0.00015748613678373384, | |
| "loss": 0.1402, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.8687615526802218, | |
| "grad_norm": 5.664543151855469, | |
| "learning_rate": 0.00015656192236598892, | |
| "loss": 0.1606, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8872458410351202, | |
| "grad_norm": 5.662344932556152, | |
| "learning_rate": 0.000155637707948244, | |
| "loss": 0.0869, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9057301293900185, | |
| "grad_norm": 1.1777679920196533, | |
| "learning_rate": 0.00015471349353049908, | |
| "loss": 0.0827, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9242144177449169, | |
| "grad_norm": 0.06051797419786453, | |
| "learning_rate": 0.00015378927911275416, | |
| "loss": 0.0969, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9242144177449169, | |
| "eval_accuracy": 0.9842829076620825, | |
| "eval_loss": 0.04485374689102173, | |
| "eval_runtime": 52.5355, | |
| "eval_samples_per_second": 29.066, | |
| "eval_steps_per_second": 3.636, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9426987060998152, | |
| "grad_norm": 9.434717178344727, | |
| "learning_rate": 0.00015286506469500925, | |
| "loss": 0.1921, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.9611829944547134, | |
| "grad_norm": 1.619040846824646, | |
| "learning_rate": 0.00015194085027726433, | |
| "loss": 0.1906, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.9796672828096118, | |
| "grad_norm": 0.5532277226448059, | |
| "learning_rate": 0.0001510166358595194, | |
| "loss": 0.1082, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.9981515711645101, | |
| "grad_norm": 0.0866900086402893, | |
| "learning_rate": 0.0001500924214417745, | |
| "loss": 0.1119, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.0166358595194085, | |
| "grad_norm": 2.668076276779175, | |
| "learning_rate": 0.00014916820702402957, | |
| "loss": 0.143, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.0351201478743068, | |
| "grad_norm": 0.15896956622600555, | |
| "learning_rate": 0.00014824399260628468, | |
| "loss": 0.0378, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.0536044362292052, | |
| "grad_norm": 0.12053361535072327, | |
| "learning_rate": 0.00014731977818853976, | |
| "loss": 0.0528, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.0720887245841035, | |
| "grad_norm": 0.06896385550498962, | |
| "learning_rate": 0.00014639556377079484, | |
| "loss": 0.1663, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.0905730129390019, | |
| "grad_norm": 7.400400638580322, | |
| "learning_rate": 0.00014547134935304992, | |
| "loss": 0.081, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.1090573012939002, | |
| "grad_norm": 0.04029673710465431, | |
| "learning_rate": 0.000144547134935305, | |
| "loss": 0.0326, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.1090573012939002, | |
| "eval_accuracy": 0.9574328749181401, | |
| "eval_loss": 0.1554253250360489, | |
| "eval_runtime": 52.4665, | |
| "eval_samples_per_second": 29.104, | |
| "eval_steps_per_second": 3.64, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.1275415896487986, | |
| "grad_norm": 1.2735309600830078, | |
| "learning_rate": 0.0001436229205175601, | |
| "loss": 0.1339, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.146025878003697, | |
| "grad_norm": 2.2266452312469482, | |
| "learning_rate": 0.00014269870609981517, | |
| "loss": 0.1443, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.1645101663585953, | |
| "grad_norm": 2.932450294494629, | |
| "learning_rate": 0.00014177449168207025, | |
| "loss": 0.0869, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.1829944547134936, | |
| "grad_norm": 5.688024520874023, | |
| "learning_rate": 0.00014085027726432533, | |
| "loss": 0.091, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.201478743068392, | |
| "grad_norm": 0.04643339663743973, | |
| "learning_rate": 0.0001399260628465804, | |
| "loss": 0.0433, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.21996303142329, | |
| "grad_norm": 0.38614460825920105, | |
| "learning_rate": 0.0001390018484288355, | |
| "loss": 0.0514, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.2384473197781884, | |
| "grad_norm": 0.03372357785701752, | |
| "learning_rate": 0.00013807763401109058, | |
| "loss": 0.0826, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.2569316081330868, | |
| "grad_norm": 0.7059990763664246, | |
| "learning_rate": 0.00013715341959334566, | |
| "loss": 0.1309, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.2754158964879851, | |
| "grad_norm": 1.5385607481002808, | |
| "learning_rate": 0.00013622920517560074, | |
| "loss": 0.115, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.2939001848428835, | |
| "grad_norm": 1.647644281387329, | |
| "learning_rate": 0.00013530499075785582, | |
| "loss": 0.1057, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.2939001848428835, | |
| "eval_accuracy": 0.9738048461034708, | |
| "eval_loss": 0.08448445796966553, | |
| "eval_runtime": 52.7705, | |
| "eval_samples_per_second": 28.937, | |
| "eval_steps_per_second": 3.619, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.3123844731977818, | |
| "grad_norm": 0.8896564841270447, | |
| "learning_rate": 0.0001343807763401109, | |
| "loss": 0.1076, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.3308687615526802, | |
| "grad_norm": 0.9722292423248291, | |
| "learning_rate": 0.000133456561922366, | |
| "loss": 0.1285, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.3493530499075785, | |
| "grad_norm": 3.9030041694641113, | |
| "learning_rate": 0.00013253234750462106, | |
| "loss": 0.1367, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.3678373382624769, | |
| "grad_norm": 1.199768304824829, | |
| "learning_rate": 0.00013160813308687617, | |
| "loss": 0.088, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.3863216266173752, | |
| "grad_norm": 0.8339413404464722, | |
| "learning_rate": 0.00013068391866913125, | |
| "loss": 0.0481, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.4048059149722736, | |
| "grad_norm": 2.3673453330993652, | |
| "learning_rate": 0.00012975970425138634, | |
| "loss": 0.0698, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.423290203327172, | |
| "grad_norm": 0.042785417288541794, | |
| "learning_rate": 0.00012883548983364142, | |
| "loss": 0.0179, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.4417744916820703, | |
| "grad_norm": 2.720048189163208, | |
| "learning_rate": 0.0001279112754158965, | |
| "loss": 0.0996, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.4602587800369686, | |
| "grad_norm": 16.840740203857422, | |
| "learning_rate": 0.00012698706099815158, | |
| "loss": 0.0707, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.478743068391867, | |
| "grad_norm": 0.1579107642173767, | |
| "learning_rate": 0.00012606284658040666, | |
| "loss": 0.0805, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.478743068391867, | |
| "eval_accuracy": 0.9823182711198428, | |
| "eval_loss": 0.07117750495672226, | |
| "eval_runtime": 53.0346, | |
| "eval_samples_per_second": 28.793, | |
| "eval_steps_per_second": 3.601, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.4972273567467653, | |
| "grad_norm": 7.252885341644287, | |
| "learning_rate": 0.00012513863216266174, | |
| "loss": 0.0848, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.5157116451016637, | |
| "grad_norm": 0.25338369607925415, | |
| "learning_rate": 0.00012421441774491682, | |
| "loss": 0.0689, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.5341959334565618, | |
| "grad_norm": 3.66860032081604, | |
| "learning_rate": 0.0001232902033271719, | |
| "loss": 0.041, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.5526802218114604, | |
| "grad_norm": 9.176445960998535, | |
| "learning_rate": 0.000122365988909427, | |
| "loss": 0.111, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.5711645101663585, | |
| "grad_norm": 0.032652150839567184, | |
| "learning_rate": 0.00012144177449168208, | |
| "loss": 0.0519, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.589648798521257, | |
| "grad_norm": 0.054165273904800415, | |
| "learning_rate": 0.00012051756007393715, | |
| "loss": 0.0661, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.6081330868761552, | |
| "grad_norm": 0.10612482577562332, | |
| "learning_rate": 0.00011959334565619225, | |
| "loss": 0.0157, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.6266173752310538, | |
| "grad_norm": 0.7138892412185669, | |
| "learning_rate": 0.00011866913123844731, | |
| "loss": 0.1159, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.645101663585952, | |
| "grad_norm": 0.0576617456972599, | |
| "learning_rate": 0.00011774491682070241, | |
| "loss": 0.1059, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.6635859519408502, | |
| "grad_norm": 2.485743999481201, | |
| "learning_rate": 0.00011682070240295748, | |
| "loss": 0.0889, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.6635859519408502, | |
| "eval_accuracy": 0.9796987557301899, | |
| "eval_loss": 0.07181376963853836, | |
| "eval_runtime": 53.7952, | |
| "eval_samples_per_second": 28.385, | |
| "eval_steps_per_second": 3.551, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.6820702402957486, | |
| "grad_norm": 0.25389525294303894, | |
| "learning_rate": 0.00011589648798521257, | |
| "loss": 0.0478, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.700554528650647, | |
| "grad_norm": 0.040639039129018784, | |
| "learning_rate": 0.00011497227356746765, | |
| "loss": 0.0579, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.7190388170055453, | |
| "grad_norm": 0.04252118989825249, | |
| "learning_rate": 0.00011404805914972275, | |
| "loss": 0.0414, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.7375231053604436, | |
| "grad_norm": 0.03039310872554779, | |
| "learning_rate": 0.00011312384473197783, | |
| "loss": 0.1247, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.756007393715342, | |
| "grad_norm": 0.04092634469270706, | |
| "learning_rate": 0.00011219963031423291, | |
| "loss": 0.0485, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.7744916820702403, | |
| "grad_norm": 0.02784869633615017, | |
| "learning_rate": 0.000111275415896488, | |
| "loss": 0.044, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.7929759704251387, | |
| "grad_norm": 0.6377788186073303, | |
| "learning_rate": 0.00011035120147874307, | |
| "loss": 0.0833, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.8114602587800368, | |
| "grad_norm": 0.0410403273999691, | |
| "learning_rate": 0.00010942698706099817, | |
| "loss": 0.0079, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.8299445471349354, | |
| "grad_norm": 0.16617639362812042, | |
| "learning_rate": 0.00010850277264325324, | |
| "loss": 0.0562, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.8484288354898335, | |
| "grad_norm": 6.131214141845703, | |
| "learning_rate": 0.00010757855822550833, | |
| "loss": 0.0503, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.8484288354898335, | |
| "eval_accuracy": 0.9934512115258677, | |
| "eval_loss": 0.0250676441937685, | |
| "eval_runtime": 53.2731, | |
| "eval_samples_per_second": 28.664, | |
| "eval_steps_per_second": 3.585, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.866913123844732, | |
| "grad_norm": 0.07335863262414932, | |
| "learning_rate": 0.0001066543438077634, | |
| "loss": 0.0444, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.8853974121996302, | |
| "grad_norm": 0.034475117921829224, | |
| "learning_rate": 0.0001057301293900185, | |
| "loss": 0.0513, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.9038817005545288, | |
| "grad_norm": 0.035967420786619186, | |
| "learning_rate": 0.00010480591497227356, | |
| "loss": 0.0669, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.922365988909427, | |
| "grad_norm": 0.029034554958343506, | |
| "learning_rate": 0.00010388170055452866, | |
| "loss": 0.0278, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.9408502772643255, | |
| "grad_norm": 3.698307514190674, | |
| "learning_rate": 0.00010295748613678373, | |
| "loss": 0.0547, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.9593345656192236, | |
| "grad_norm": 0.040026549249887466, | |
| "learning_rate": 0.00010203327171903882, | |
| "loss": 0.0065, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.9778188539741222, | |
| "grad_norm": 3.3067240715026855, | |
| "learning_rate": 0.0001011090573012939, | |
| "loss": 0.0828, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.9963031423290203, | |
| "grad_norm": 0.05000556632876396, | |
| "learning_rate": 0.000100184842883549, | |
| "loss": 0.0632, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.014787430683919, | |
| "grad_norm": 0.04542790353298187, | |
| "learning_rate": 9.926062846580408e-05, | |
| "loss": 0.0682, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.033271719038817, | |
| "grad_norm": 0.030154038220643997, | |
| "learning_rate": 9.833641404805916e-05, | |
| "loss": 0.0225, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.033271719038817, | |
| "eval_accuracy": 0.9967256057629339, | |
| "eval_loss": 0.01773611083626747, | |
| "eval_runtime": 52.5689, | |
| "eval_samples_per_second": 29.048, | |
| "eval_steps_per_second": 3.633, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.0517560073937156, | |
| "grad_norm": 0.3824068307876587, | |
| "learning_rate": 9.741219963031424e-05, | |
| "loss": 0.0194, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.0702402957486137, | |
| "grad_norm": 0.020000776275992393, | |
| "learning_rate": 9.648798521256932e-05, | |
| "loss": 0.0259, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.088724584103512, | |
| "grad_norm": 3.488415241241455, | |
| "learning_rate": 9.55637707948244e-05, | |
| "loss": 0.0629, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.1072088724584104, | |
| "grad_norm": 10.373331069946289, | |
| "learning_rate": 9.463955637707949e-05, | |
| "loss": 0.015, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.1256931608133085, | |
| "grad_norm": 0.23100066184997559, | |
| "learning_rate": 9.371534195933457e-05, | |
| "loss": 0.0619, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.144177449168207, | |
| "grad_norm": 0.07692666351795197, | |
| "learning_rate": 9.279112754158965e-05, | |
| "loss": 0.06, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.162661737523105, | |
| "grad_norm": 0.057554759085178375, | |
| "learning_rate": 9.186691312384473e-05, | |
| "loss": 0.0079, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.1811460258780038, | |
| "grad_norm": 0.039722565561532974, | |
| "learning_rate": 9.094269870609981e-05, | |
| "loss": 0.0581, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.199630314232902, | |
| "grad_norm": 0.021510232239961624, | |
| "learning_rate": 9.001848428835489e-05, | |
| "loss": 0.0052, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.2181146025878005, | |
| "grad_norm": 0.019746674224734306, | |
| "learning_rate": 8.909426987060999e-05, | |
| "loss": 0.0049, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.2181146025878005, | |
| "eval_accuracy": 0.9921414538310412, | |
| "eval_loss": 0.024552814662456512, | |
| "eval_runtime": 52.686, | |
| "eval_samples_per_second": 28.983, | |
| "eval_steps_per_second": 3.625, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.2365988909426986, | |
| "grad_norm": 4.809552192687988, | |
| "learning_rate": 8.817005545286507e-05, | |
| "loss": 0.098, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.255083179297597, | |
| "grad_norm": 0.22049099206924438, | |
| "learning_rate": 8.724584103512015e-05, | |
| "loss": 0.1328, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.2735674676524953, | |
| "grad_norm": 0.02430686727166176, | |
| "learning_rate": 8.632162661737525e-05, | |
| "loss": 0.0332, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.292051756007394, | |
| "grad_norm": 0.16566839814186096, | |
| "learning_rate": 8.539741219963033e-05, | |
| "loss": 0.0242, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.310536044362292, | |
| "grad_norm": 0.07895852625370026, | |
| "learning_rate": 8.447319778188541e-05, | |
| "loss": 0.0394, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.3290203327171906, | |
| "grad_norm": 0.01941494271159172, | |
| "learning_rate": 8.354898336414049e-05, | |
| "loss": 0.0373, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.3475046210720887, | |
| "grad_norm": 0.018574291840195656, | |
| "learning_rate": 8.262476894639557e-05, | |
| "loss": 0.0582, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.3659889094269873, | |
| "grad_norm": 9.006904602050781, | |
| "learning_rate": 8.170055452865065e-05, | |
| "loss": 0.075, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.3844731977818854, | |
| "grad_norm": 0.5771515965461731, | |
| "learning_rate": 8.077634011090573e-05, | |
| "loss": 0.0217, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.402957486136784, | |
| "grad_norm": 0.01840708591043949, | |
| "learning_rate": 7.985212569316082e-05, | |
| "loss": 0.0152, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.402957486136784, | |
| "eval_accuracy": 0.9986902423051736, | |
| "eval_loss": 0.008291647769510746, | |
| "eval_runtime": 53.4499, | |
| "eval_samples_per_second": 28.569, | |
| "eval_steps_per_second": 3.573, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.421441774491682, | |
| "grad_norm": 0.017435792833566666, | |
| "learning_rate": 7.89279112754159e-05, | |
| "loss": 0.0448, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.43992606284658, | |
| "grad_norm": 0.7729086875915527, | |
| "learning_rate": 7.800369685767098e-05, | |
| "loss": 0.0444, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.4584103512014788, | |
| "grad_norm": 0.059264715760946274, | |
| "learning_rate": 7.707948243992606e-05, | |
| "loss": 0.0397, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.476894639556377, | |
| "grad_norm": 0.024057278409600258, | |
| "learning_rate": 7.615526802218114e-05, | |
| "loss": 0.028, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.4953789279112755, | |
| "grad_norm": 0.022951899096369743, | |
| "learning_rate": 7.523105360443624e-05, | |
| "loss": 0.0444, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.5138632162661736, | |
| "grad_norm": 0.021782563999295235, | |
| "learning_rate": 7.430683918669132e-05, | |
| "loss": 0.0385, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.532347504621072, | |
| "grad_norm": 0.1371038258075714, | |
| "learning_rate": 7.33826247689464e-05, | |
| "loss": 0.0188, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.5508317929759703, | |
| "grad_norm": 0.7299683690071106, | |
| "learning_rate": 7.245841035120148e-05, | |
| "loss": 0.0845, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.569316081330869, | |
| "grad_norm": 0.34656259417533875, | |
| "learning_rate": 7.153419593345656e-05, | |
| "loss": 0.0436, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.587800369685767, | |
| "grad_norm": 0.10165718197822571, | |
| "learning_rate": 7.060998151571166e-05, | |
| "loss": 0.08, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.587800369685767, | |
| "eval_accuracy": 0.9941060903732809, | |
| "eval_loss": 0.021378275007009506, | |
| "eval_runtime": 52.8132, | |
| "eval_samples_per_second": 28.913, | |
| "eval_steps_per_second": 3.617, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.6062846580406656, | |
| "grad_norm": 5.586907863616943, | |
| "learning_rate": 6.968576709796674e-05, | |
| "loss": 0.0295, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.6247689463955637, | |
| "grad_norm": 0.0221896730363369, | |
| "learning_rate": 6.876155268022182e-05, | |
| "loss": 0.0627, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.6432532347504623, | |
| "grad_norm": 0.30416977405548096, | |
| "learning_rate": 6.78373382624769e-05, | |
| "loss": 0.0035, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.6617375231053604, | |
| "grad_norm": 0.102454274892807, | |
| "learning_rate": 6.691312384473198e-05, | |
| "loss": 0.0641, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.6802218114602585, | |
| "grad_norm": 0.023131974041461945, | |
| "learning_rate": 6.598890942698706e-05, | |
| "loss": 0.0326, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.698706099815157, | |
| "grad_norm": 0.09067076444625854, | |
| "learning_rate": 6.506469500924215e-05, | |
| "loss": 0.017, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.7171903881700556, | |
| "grad_norm": 3.3906850814819336, | |
| "learning_rate": 6.414048059149723e-05, | |
| "loss": 0.029, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.7356746765249538, | |
| "grad_norm": 0.061337146908044815, | |
| "learning_rate": 6.321626617375231e-05, | |
| "loss": 0.0168, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.754158964879852, | |
| "grad_norm": 0.19621238112449646, | |
| "learning_rate": 6.229205175600739e-05, | |
| "loss": 0.006, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.7726432532347505, | |
| "grad_norm": 0.012029612436890602, | |
| "learning_rate": 6.136783733826249e-05, | |
| "loss": 0.0043, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.7726432532347505, | |
| "eval_accuracy": 0.9980353634577603, | |
| "eval_loss": 0.006946724373847246, | |
| "eval_runtime": 52.203, | |
| "eval_samples_per_second": 29.251, | |
| "eval_steps_per_second": 3.659, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.791127541589649, | |
| "grad_norm": 0.014309920370578766, | |
| "learning_rate": 6.044362292051756e-05, | |
| "loss": 0.0074, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.809611829944547, | |
| "grad_norm": 3.063054323196411, | |
| "learning_rate": 5.951940850277264e-05, | |
| "loss": 0.0045, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.8280961182994453, | |
| "grad_norm": 0.011617097072303295, | |
| "learning_rate": 5.859519408502773e-05, | |
| "loss": 0.0525, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.846580406654344, | |
| "grad_norm": 5.252607345581055, | |
| "learning_rate": 5.767097966728281e-05, | |
| "loss": 0.0104, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.865064695009242, | |
| "grad_norm": 0.014846362173557281, | |
| "learning_rate": 5.674676524953789e-05, | |
| "loss": 0.0265, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.8835489833641406, | |
| "grad_norm": 0.011737200431525707, | |
| "learning_rate": 5.5822550831792974e-05, | |
| "loss": 0.0543, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.9020332717190387, | |
| "grad_norm": 0.012772896327078342, | |
| "learning_rate": 5.4898336414048056e-05, | |
| "loss": 0.0018, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.9205175600739373, | |
| "grad_norm": 0.06962817162275314, | |
| "learning_rate": 5.397412199630314e-05, | |
| "loss": 0.0234, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.9390018484288354, | |
| "grad_norm": 0.019341696053743362, | |
| "learning_rate": 5.304990757855823e-05, | |
| "loss": 0.105, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.957486136783734, | |
| "grad_norm": 4.673314571380615, | |
| "learning_rate": 5.2125693160813314e-05, | |
| "loss": 0.0501, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.957486136783734, | |
| "eval_accuracy": 0.9967256057629339, | |
| "eval_loss": 0.015068226493895054, | |
| "eval_runtime": 51.6353, | |
| "eval_samples_per_second": 29.573, | |
| "eval_steps_per_second": 3.699, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.975970425138632, | |
| "grad_norm": 0.018514908850193024, | |
| "learning_rate": 5.1201478743068395e-05, | |
| "loss": 0.0312, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.9944547134935307, | |
| "grad_norm": 0.0645008459687233, | |
| "learning_rate": 5.027726432532348e-05, | |
| "loss": 0.0489, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 3.0129390018484288, | |
| "grad_norm": 0.017880817875266075, | |
| "learning_rate": 4.935304990757856e-05, | |
| "loss": 0.0366, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 3.0314232902033273, | |
| "grad_norm": 0.04122663289308548, | |
| "learning_rate": 4.8428835489833646e-05, | |
| "loss": 0.0539, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 3.0499075785582255, | |
| "grad_norm": 0.022179430350661278, | |
| "learning_rate": 4.750462107208873e-05, | |
| "loss": 0.0248, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 3.068391866913124, | |
| "grad_norm": 0.924117386341095, | |
| "learning_rate": 4.658040665434381e-05, | |
| "loss": 0.02, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 3.086876155268022, | |
| "grad_norm": 0.01614381931722164, | |
| "learning_rate": 4.565619223659889e-05, | |
| "loss": 0.023, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 3.1053604436229207, | |
| "grad_norm": 0.05051511153578758, | |
| "learning_rate": 4.473197781885398e-05, | |
| "loss": 0.0041, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 3.123844731977819, | |
| "grad_norm": 0.02787856012582779, | |
| "learning_rate": 4.380776340110906e-05, | |
| "loss": 0.0163, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 3.142329020332717, | |
| "grad_norm": 0.21667926013469696, | |
| "learning_rate": 4.288354898336414e-05, | |
| "loss": 0.0186, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.142329020332717, | |
| "eval_accuracy": 0.9973804846103471, | |
| "eval_loss": 0.007818276062607765, | |
| "eval_runtime": 52.8582, | |
| "eval_samples_per_second": 28.889, | |
| "eval_steps_per_second": 3.613, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 3.1608133086876156, | |
| "grad_norm": 0.02714550867676735, | |
| "learning_rate": 4.195933456561922e-05, | |
| "loss": 0.0178, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 3.1792975970425137, | |
| "grad_norm": 0.5191987156867981, | |
| "learning_rate": 4.1035120147874305e-05, | |
| "loss": 0.0582, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 3.1977818853974123, | |
| "grad_norm": 0.02666807919740677, | |
| "learning_rate": 4.011090573012939e-05, | |
| "loss": 0.007, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 3.2162661737523104, | |
| "grad_norm": 0.06601597368717194, | |
| "learning_rate": 3.9186691312384474e-05, | |
| "loss": 0.0477, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 3.234750462107209, | |
| "grad_norm": 0.0280216746032238, | |
| "learning_rate": 3.826247689463956e-05, | |
| "loss": 0.0048, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 3.253234750462107, | |
| "grad_norm": 4.720592021942139, | |
| "learning_rate": 3.7338262476894644e-05, | |
| "loss": 0.0186, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 3.2717190388170057, | |
| "grad_norm": 0.01574169471859932, | |
| "learning_rate": 3.6414048059149726e-05, | |
| "loss": 0.0017, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.290203327171904, | |
| "grad_norm": 0.02533087506890297, | |
| "learning_rate": 3.548983364140481e-05, | |
| "loss": 0.0025, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.3086876155268024, | |
| "grad_norm": 0.013142619282007217, | |
| "learning_rate": 3.456561922365989e-05, | |
| "loss": 0.0376, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.3271719038817005, | |
| "grad_norm": 0.07316397875547409, | |
| "learning_rate": 3.364140480591497e-05, | |
| "loss": 0.0033, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.3271719038817005, | |
| "eval_accuracy": 0.9960707269155207, | |
| "eval_loss": 0.013949541375041008, | |
| "eval_runtime": 53.0604, | |
| "eval_samples_per_second": 28.779, | |
| "eval_steps_per_second": 3.6, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.345656192236599, | |
| "grad_norm": 0.015296310186386108, | |
| "learning_rate": 3.271719038817006e-05, | |
| "loss": 0.0015, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.364140480591497, | |
| "grad_norm": 5.960048198699951, | |
| "learning_rate": 3.179297597042514e-05, | |
| "loss": 0.0222, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.3826247689463957, | |
| "grad_norm": 0.21616186201572418, | |
| "learning_rate": 3.086876155268023e-05, | |
| "loss": 0.0038, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.401109057301294, | |
| "grad_norm": 0.015051410533487797, | |
| "learning_rate": 2.994454713493531e-05, | |
| "loss": 0.0019, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.4195933456561924, | |
| "grad_norm": 13.381204605102539, | |
| "learning_rate": 2.902033271719039e-05, | |
| "loss": 0.0182, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.4380776340110906, | |
| "grad_norm": 0.1726062297821045, | |
| "learning_rate": 2.8096118299445472e-05, | |
| "loss": 0.0022, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.4565619223659887, | |
| "grad_norm": 0.01701999455690384, | |
| "learning_rate": 2.7171903881700557e-05, | |
| "loss": 0.0014, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.4750462107208873, | |
| "grad_norm": 0.013869056478142738, | |
| "learning_rate": 2.624768946395564e-05, | |
| "loss": 0.0013, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.4935304990757854, | |
| "grad_norm": 0.021621432155370712, | |
| "learning_rate": 2.532347504621072e-05, | |
| "loss": 0.0016, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 3.512014787430684, | |
| "grad_norm": 1.3106377124786377, | |
| "learning_rate": 2.4399260628465805e-05, | |
| "loss": 0.0023, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.512014787430684, | |
| "eval_accuracy": 0.9986902423051736, | |
| "eval_loss": 0.0075506423600018024, | |
| "eval_runtime": 50.8135, | |
| "eval_samples_per_second": 30.051, | |
| "eval_steps_per_second": 3.759, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.530499075785582, | |
| "grad_norm": 0.01985827274620533, | |
| "learning_rate": 2.347504621072089e-05, | |
| "loss": 0.0016, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 3.5489833641404807, | |
| "grad_norm": 0.013897390104830265, | |
| "learning_rate": 2.255083179297597e-05, | |
| "loss": 0.0308, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 3.567467652495379, | |
| "grad_norm": 0.009370139800012112, | |
| "learning_rate": 2.1626617375231053e-05, | |
| "loss": 0.0123, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 3.5859519408502774, | |
| "grad_norm": 0.019544150680303574, | |
| "learning_rate": 2.0702402957486137e-05, | |
| "loss": 0.0257, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.6044362292051755, | |
| "grad_norm": 0.018746808171272278, | |
| "learning_rate": 1.9778188539741222e-05, | |
| "loss": 0.03, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.622920517560074, | |
| "grad_norm": 0.009196238592267036, | |
| "learning_rate": 1.8853974121996304e-05, | |
| "loss": 0.0011, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.641404805914972, | |
| "grad_norm": 0.011442320421338081, | |
| "learning_rate": 1.7929759704251385e-05, | |
| "loss": 0.0012, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.6598890942698707, | |
| "grad_norm": 0.010710498318076134, | |
| "learning_rate": 1.700554528650647e-05, | |
| "loss": 0.0019, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.678373382624769, | |
| "grad_norm": 0.06102241575717926, | |
| "learning_rate": 1.6081330868761555e-05, | |
| "loss": 0.0012, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.6968576709796674, | |
| "grad_norm": 0.008612744510173798, | |
| "learning_rate": 1.5157116451016636e-05, | |
| "loss": 0.0054, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.6968576709796674, | |
| "eval_accuracy": 0.9993451211525868, | |
| "eval_loss": 0.0047513521276414394, | |
| "eval_runtime": 52.2618, | |
| "eval_samples_per_second": 29.218, | |
| "eval_steps_per_second": 3.655, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.7153419593345656, | |
| "grad_norm": 0.008234468288719654, | |
| "learning_rate": 1.423290203327172e-05, | |
| "loss": 0.043, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.733826247689464, | |
| "grad_norm": 0.008917649276554585, | |
| "learning_rate": 1.3308687615526803e-05, | |
| "loss": 0.0384, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.7523105360443623, | |
| "grad_norm": 0.00844865757972002, | |
| "learning_rate": 1.2384473197781886e-05, | |
| "loss": 0.0013, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.7707948243992604, | |
| "grad_norm": 0.008531128987669945, | |
| "learning_rate": 1.1460258780036969e-05, | |
| "loss": 0.0195, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.789279112754159, | |
| "grad_norm": 0.009270643815398216, | |
| "learning_rate": 1.0536044362292052e-05, | |
| "loss": 0.0392, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.8077634011090575, | |
| "grad_norm": 0.009245671331882477, | |
| "learning_rate": 9.611829944547135e-06, | |
| "loss": 0.0011, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.8262476894639557, | |
| "grad_norm": 0.01690092496573925, | |
| "learning_rate": 8.687615526802218e-06, | |
| "loss": 0.0016, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.844731977818854, | |
| "grad_norm": 0.015731679275631905, | |
| "learning_rate": 7.763401109057302e-06, | |
| "loss": 0.0317, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.8632162661737524, | |
| "grad_norm": 3.0953285694122314, | |
| "learning_rate": 6.931608133086876e-06, | |
| "loss": 0.0454, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.8817005545286505, | |
| "grad_norm": 6.279654502868652, | |
| "learning_rate": 6.0073937153419595e-06, | |
| "loss": 0.0168, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.8817005545286505, | |
| "eval_accuracy": 0.9986902423051736, | |
| "eval_loss": 0.006641203537583351, | |
| "eval_runtime": 52.9204, | |
| "eval_samples_per_second": 28.855, | |
| "eval_steps_per_second": 3.609, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.900184842883549, | |
| "grad_norm": 0.009602474048733711, | |
| "learning_rate": 5.083179297597043e-06, | |
| "loss": 0.0011, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.918669131238447, | |
| "grad_norm": 12.240010261535645, | |
| "learning_rate": 4.158964879852126e-06, | |
| "loss": 0.0236, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.9371534195933457, | |
| "grad_norm": 0.03988449275493622, | |
| "learning_rate": 3.234750462107209e-06, | |
| "loss": 0.0014, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.955637707948244, | |
| "grad_norm": 5.554378986358643, | |
| "learning_rate": 2.310536044362292e-06, | |
| "loss": 0.0041, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.9741219963031424, | |
| "grad_norm": 0.0083112558349967, | |
| "learning_rate": 1.3863216266173753e-06, | |
| "loss": 0.02, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.9926062846580406, | |
| "grad_norm": 2.2959258556365967, | |
| "learning_rate": 4.621072088724585e-07, | |
| "loss": 0.0053, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 2164, | |
| "total_flos": 2.6818427765818e+18, | |
| "train_loss": 0.0841421499820822, | |
| "train_runtime": 2597.595, | |
| "train_samples_per_second": 13.323, | |
| "train_steps_per_second": 0.833 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2164, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "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": 2.6818427765818e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |