| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 3930, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002544529262086514, | |
| "grad_norm": 2.135885238647461, | |
| "learning_rate": 9.160305343511451e-07, | |
| "loss": 0.1137, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.005089058524173028, | |
| "grad_norm": 2.356735944747925, | |
| "learning_rate": 1.933842239185751e-06, | |
| "loss": 0.1656, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.007633587786259542, | |
| "grad_norm": 3.483503818511963, | |
| "learning_rate": 2.951653944020356e-06, | |
| "loss": 0.1356, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.010178117048346057, | |
| "grad_norm": 3.4808290004730225, | |
| "learning_rate": 3.969465648854962e-06, | |
| "loss": 0.1575, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01272264631043257, | |
| "grad_norm": 2.5104947090148926, | |
| "learning_rate": 4.987277353689568e-06, | |
| "loss": 0.0749, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.015267175572519083, | |
| "grad_norm": 3.5466084480285645, | |
| "learning_rate": 6.005089058524174e-06, | |
| "loss": 0.0631, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.017811704834605598, | |
| "grad_norm": 0.6639775633811951, | |
| "learning_rate": 7.022900763358779e-06, | |
| "loss": 0.0374, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.020356234096692113, | |
| "grad_norm": 1.4179292917251587, | |
| "learning_rate": 8.040712468193384e-06, | |
| "loss": 0.0413, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.022900763358778626, | |
| "grad_norm": 2.9656121730804443, | |
| "learning_rate": 9.058524173027991e-06, | |
| "loss": 0.0362, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02544529262086514, | |
| "grad_norm": 3.4811854362487793, | |
| "learning_rate": 1.0076335877862595e-05, | |
| "loss": 0.0411, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.027989821882951654, | |
| "grad_norm": 2.1551337242126465, | |
| "learning_rate": 1.1094147582697202e-05, | |
| "loss": 0.0255, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.030534351145038167, | |
| "grad_norm": 1.5262765884399414, | |
| "learning_rate": 1.2111959287531807e-05, | |
| "loss": 0.0225, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03307888040712468, | |
| "grad_norm": 1.3079928159713745, | |
| "learning_rate": 1.3129770992366414e-05, | |
| "loss": 0.0323, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.035623409669211195, | |
| "grad_norm": 0.22996912896633148, | |
| "learning_rate": 1.4147582697201019e-05, | |
| "loss": 0.0188, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03816793893129771, | |
| "grad_norm": 1.0834699869155884, | |
| "learning_rate": 1.5165394402035624e-05, | |
| "loss": 0.0265, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.04071246819338423, | |
| "grad_norm": 0.9807230830192566, | |
| "learning_rate": 1.618320610687023e-05, | |
| "loss": 0.0189, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.043256997455470736, | |
| "grad_norm": 1.8505427837371826, | |
| "learning_rate": 1.7201017811704836e-05, | |
| "loss": 0.0142, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.04580152671755725, | |
| "grad_norm": 0.346323698759079, | |
| "learning_rate": 1.8218829516539443e-05, | |
| "loss": 0.0237, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.04834605597964377, | |
| "grad_norm": 1.4252818822860718, | |
| "learning_rate": 1.923664122137405e-05, | |
| "loss": 0.0169, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.05089058524173028, | |
| "grad_norm": 0.624088704586029, | |
| "learning_rate": 2.0254452926208653e-05, | |
| "loss": 0.0153, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05343511450381679, | |
| "grad_norm": 0.7245016098022461, | |
| "learning_rate": 2.127226463104326e-05, | |
| "loss": 0.0211, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.05597964376590331, | |
| "grad_norm": 1.2375560998916626, | |
| "learning_rate": 2.2290076335877867e-05, | |
| "loss": 0.0146, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.058524173027989825, | |
| "grad_norm": 0.15632864832878113, | |
| "learning_rate": 2.330788804071247e-05, | |
| "loss": 0.0179, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.061068702290076333, | |
| "grad_norm": 0.023025542497634888, | |
| "learning_rate": 2.4325699745547078e-05, | |
| "loss": 0.02, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.06361323155216285, | |
| "grad_norm": 0.5999135375022888, | |
| "learning_rate": 2.5343511450381678e-05, | |
| "loss": 0.0057, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.06615776081424936, | |
| "grad_norm": 0.08057913929224014, | |
| "learning_rate": 2.6361323155216285e-05, | |
| "loss": 0.0203, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.06870229007633588, | |
| "grad_norm": 0.3958178460597992, | |
| "learning_rate": 2.737913486005089e-05, | |
| "loss": 0.0365, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.07124681933842239, | |
| "grad_norm": 0.24179188907146454, | |
| "learning_rate": 2.8396946564885498e-05, | |
| "loss": 0.0198, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.0737913486005089, | |
| "grad_norm": 1.3984962701797485, | |
| "learning_rate": 2.9414758269720102e-05, | |
| "loss": 0.0119, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.07633587786259542, | |
| "grad_norm": 0.08135154098272324, | |
| "learning_rate": 3.043256997455471e-05, | |
| "loss": 0.0101, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.07888040712468193, | |
| "grad_norm": 2.0051238536834717, | |
| "learning_rate": 3.145038167938931e-05, | |
| "loss": 0.0245, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.08142493638676845, | |
| "grad_norm": 0.49833422899246216, | |
| "learning_rate": 3.246819338422392e-05, | |
| "loss": 0.0181, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.08396946564885496, | |
| "grad_norm": 0.09235569834709167, | |
| "learning_rate": 3.3486005089058526e-05, | |
| "loss": 0.0116, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.08651399491094147, | |
| "grad_norm": 0.06826785206794739, | |
| "learning_rate": 3.450381679389313e-05, | |
| "loss": 0.0161, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.089058524173028, | |
| "grad_norm": 1.1774462461471558, | |
| "learning_rate": 3.552162849872774e-05, | |
| "loss": 0.0144, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.0916030534351145, | |
| "grad_norm": 0.6383160948753357, | |
| "learning_rate": 3.653944020356235e-05, | |
| "loss": 0.0197, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.09414758269720101, | |
| "grad_norm": 0.7711915969848633, | |
| "learning_rate": 3.755725190839695e-05, | |
| "loss": 0.0086, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.09669211195928754, | |
| "grad_norm": 0.7851189970970154, | |
| "learning_rate": 3.8575063613231554e-05, | |
| "loss": 0.0235, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.09923664122137404, | |
| "grad_norm": 0.08650732040405273, | |
| "learning_rate": 3.959287531806616e-05, | |
| "loss": 0.0151, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.10178117048346055, | |
| "grad_norm": 1.1819581985473633, | |
| "learning_rate": 3.969810889228932e-05, | |
| "loss": 0.0211, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.10432569974554708, | |
| "grad_norm": 1.2111542224884033, | |
| "learning_rate": 3.920979868551272e-05, | |
| "loss": 0.0174, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.10687022900763359, | |
| "grad_norm": 0.6973260641098022, | |
| "learning_rate": 3.8739075756314324e-05, | |
| "loss": 0.0212, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.10941475826972011, | |
| "grad_norm": 0.5729886889457703, | |
| "learning_rate": 3.828490911226336e-05, | |
| "loss": 0.0157, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.11195928753180662, | |
| "grad_norm": 0.06440552324056625, | |
| "learning_rate": 3.784635043776162e-05, | |
| "loss": 0.004, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.11450381679389313, | |
| "grad_norm": 0.5472397208213806, | |
| "learning_rate": 3.7422525760542153e-05, | |
| "loss": 0.0092, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.11704834605597965, | |
| "grad_norm": 0.2682400047779083, | |
| "learning_rate": 3.701262812234327e-05, | |
| "loss": 0.0065, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.11959287531806616, | |
| "grad_norm": 0.2900489866733551, | |
| "learning_rate": 3.6615911113819035e-05, | |
| "loss": 0.0054, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.12213740458015267, | |
| "grad_norm": 0.02418186143040657, | |
| "learning_rate": 3.623168315577716e-05, | |
| "loss": 0.0094, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.12468193384223919, | |
| "grad_norm": 0.044626325368881226, | |
| "learning_rate": 3.585930242701533e-05, | |
| "loss": 0.0114, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.1272264631043257, | |
| "grad_norm": 0.22953863441944122, | |
| "learning_rate": 3.549817235409398e-05, | |
| "loss": 0.0191, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1272264631043257, | |
| "eval_loss": 0.043242499232292175, | |
| "eval_runtime": 133.889, | |
| "eval_samples_per_second": 59.617, | |
| "eval_steps_per_second": 0.471, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1297709923664122, | |
| "grad_norm": 1.6306729316711426, | |
| "learning_rate": 3.514773759092204e-05, | |
| "loss": 0.0213, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.13231552162849872, | |
| "grad_norm": 0.15864859521389008, | |
| "learning_rate": 3.480748042650629e-05, | |
| "loss": 0.0153, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.13486005089058525, | |
| "grad_norm": 0.1509598344564438, | |
| "learning_rate": 3.447691756799828e-05, | |
| "loss": 0.0056, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.13740458015267176, | |
| "grad_norm": 0.6327221393585205, | |
| "learning_rate": 3.415559725356418e-05, | |
| "loss": 0.0097, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.13994910941475827, | |
| "grad_norm": 0.8866755962371826, | |
| "learning_rate": 3.384309665584446e-05, | |
| "loss": 0.008, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.14249363867684478, | |
| "grad_norm": 0.682650625705719, | |
| "learning_rate": 3.353901954205842e-05, | |
| "loss": 0.0115, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.1450381679389313, | |
| "grad_norm": 0.3033181130886078, | |
| "learning_rate": 3.3242994161303164e-05, | |
| "loss": 0.0116, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.1475826972010178, | |
| "grad_norm": 0.37608516216278076, | |
| "learning_rate": 3.2954671333428596e-05, | |
| "loss": 0.0037, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.15012722646310434, | |
| "grad_norm": 0.028525808826088905, | |
| "learning_rate": 3.267372271714603e-05, | |
| "loss": 0.0123, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.15267175572519084, | |
| "grad_norm": 0.26955416798591614, | |
| "learning_rate": 3.239983923783774e-05, | |
| "loss": 0.0023, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.15521628498727735, | |
| "grad_norm": 0.10063929110765457, | |
| "learning_rate": 3.213272965795057e-05, | |
| "loss": 0.0129, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.15776081424936386, | |
| "grad_norm": 0.007429131772369146, | |
| "learning_rate": 3.187211927493886e-05, | |
| "loss": 0.018, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.16030534351145037, | |
| "grad_norm": 0.03856533020734787, | |
| "learning_rate": 3.161774873352233e-05, | |
| "loss": 0.0127, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.1628498727735369, | |
| "grad_norm": 0.40973177552223206, | |
| "learning_rate": 3.1369372940584045e-05, | |
| "loss": 0.0083, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.16539440203562342, | |
| "grad_norm": 0.07089962065219879, | |
| "learning_rate": 3.1126760072387977e-05, | |
| "loss": 0.0036, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.16793893129770993, | |
| "grad_norm": 0.27879390120506287, | |
| "learning_rate": 3.088969066497502e-05, | |
| "loss": 0.0089, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.17048346055979643, | |
| "grad_norm": 0.25491127371788025, | |
| "learning_rate": 3.065795677962485e-05, | |
| "loss": 0.0095, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.17302798982188294, | |
| "grad_norm": 0.01762186735868454, | |
| "learning_rate": 3.043136123617055e-05, | |
| "loss": 0.0145, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.17557251908396945, | |
| "grad_norm": 0.29377493262290955, | |
| "learning_rate": 3.0209716907741185e-05, | |
| "loss": 0.0182, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.178117048346056, | |
| "grad_norm": 0.7060360908508301, | |
| "learning_rate": 2.9992846071199107e-05, | |
| "loss": 0.0144, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.1806615776081425, | |
| "grad_norm": 0.5874837040901184, | |
| "learning_rate": 2.9780579808147626e-05, | |
| "loss": 0.0069, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.183206106870229, | |
| "grad_norm": 0.15609210729599, | |
| "learning_rate": 2.957275745192107e-05, | |
| "loss": 0.0096, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.18575063613231552, | |
| "grad_norm": 0.7818000912666321, | |
| "learning_rate": 2.936922607644298e-05, | |
| "loss": 0.0076, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.18829516539440203, | |
| "grad_norm": 0.2352600395679474, | |
| "learning_rate": 2.916984002325705e-05, | |
| "loss": 0.0071, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.19083969465648856, | |
| "grad_norm": 0.0920751690864563, | |
| "learning_rate": 2.8974460463406866e-05, | |
| "loss": 0.0097, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.19338422391857507, | |
| "grad_norm": 0.16071178019046783, | |
| "learning_rate": 2.8782954991169955e-05, | |
| "loss": 0.0057, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.19592875318066158, | |
| "grad_norm": 0.18001791834831238, | |
| "learning_rate": 2.8595197246944774e-05, | |
| "loss": 0.0033, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.1984732824427481, | |
| "grad_norm": 0.0845453143119812, | |
| "learning_rate": 2.8411066566850007e-05, | |
| "loss": 0.0059, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.2010178117048346, | |
| "grad_norm": 1.086779236793518, | |
| "learning_rate": 2.823044765682867e-05, | |
| "loss": 0.0086, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.2035623409669211, | |
| "grad_norm": 1.0203735828399658, | |
| "learning_rate": 2.8053230289256842e-05, | |
| "loss": 0.0158, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.20610687022900764, | |
| "grad_norm": 0.14365516602993011, | |
| "learning_rate": 2.787930902024349e-05, | |
| "loss": 0.0123, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.20865139949109415, | |
| "grad_norm": 0.07002055644989014, | |
| "learning_rate": 2.770858292597382e-05, | |
| "loss": 0.0025, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.21119592875318066, | |
| "grad_norm": 0.7248631715774536, | |
| "learning_rate": 2.7540955356598682e-05, | |
| "loss": 0.008, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.21374045801526717, | |
| "grad_norm": 0.8881323337554932, | |
| "learning_rate": 2.7376333706306376e-05, | |
| "loss": 0.0202, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.21628498727735368, | |
| "grad_norm": 0.01213792897760868, | |
| "learning_rate": 2.7214629198334247e-05, | |
| "loss": 0.0056, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.21882951653944022, | |
| "grad_norm": 0.2250792682170868, | |
| "learning_rate": 2.7055756683786302e-05, | |
| "loss": 0.0176, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.22137404580152673, | |
| "grad_norm": 0.04687346890568733, | |
| "learning_rate": 2.6899634453221034e-05, | |
| "loss": 0.0042, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.22391857506361323, | |
| "grad_norm": 0.04808862879872322, | |
| "learning_rate": 2.6746184060062552e-05, | |
| "loss": 0.0099, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.22646310432569974, | |
| "grad_norm": 0.9415402412414551, | |
| "learning_rate": 2.6595330154967874e-05, | |
| "loss": 0.0128, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.22900763358778625, | |
| "grad_norm": 0.05923384800553322, | |
| "learning_rate": 2.6447000330356178e-05, | |
| "loss": 0.0076, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.23155216284987276, | |
| "grad_norm": 0.03846553713083267, | |
| "learning_rate": 2.6301124974371188e-05, | |
| "loss": 0.01, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.2340966921119593, | |
| "grad_norm": 0.07209023088216782, | |
| "learning_rate": 2.6157637133607714e-05, | |
| "loss": 0.0092, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.2366412213740458, | |
| "grad_norm": 0.19055865705013275, | |
| "learning_rate": 2.601647238398745e-05, | |
| "loss": 0.0081, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.23918575063613232, | |
| "grad_norm": 0.007097713649272919, | |
| "learning_rate": 2.5877568709218407e-05, | |
| "loss": 0.0022, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.24173027989821882, | |
| "grad_norm": 0.025350170210003853, | |
| "learning_rate": 2.5740866386317107e-05, | |
| "loss": 0.0126, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.24427480916030533, | |
| "grad_norm": 0.13661088049411774, | |
| "learning_rate": 2.5606307877713624e-05, | |
| "loss": 0.0057, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.24681933842239187, | |
| "grad_norm": 0.3657326400279999, | |
| "learning_rate": 2.547383772949672e-05, | |
| "loss": 0.0075, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.24936386768447838, | |
| "grad_norm": 0.05925949662923813, | |
| "learning_rate": 2.534340247539036e-05, | |
| "loss": 0.0092, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.25190839694656486, | |
| "grad_norm": 0.10418176651000977, | |
| "learning_rate": 2.521495054608389e-05, | |
| "loss": 0.0045, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.2544529262086514, | |
| "grad_norm": 0.5425507426261902, | |
| "learning_rate": 2.5088432183566784e-05, | |
| "loss": 0.0093, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2544529262086514, | |
| "eval_loss": 0.03361435979604721, | |
| "eval_runtime": 133.8769, | |
| "eval_samples_per_second": 59.622, | |
| "eval_steps_per_second": 0.471, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.25699745547073793, | |
| "grad_norm": 0.025838060304522514, | |
| "learning_rate": 2.4963799360144644e-05, | |
| "loss": 0.0052, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.2595419847328244, | |
| "grad_norm": 0.13453072309494019, | |
| "learning_rate": 2.4841005701837126e-05, | |
| "loss": 0.0039, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.26208651399491095, | |
| "grad_norm": 0.7228013873100281, | |
| "learning_rate": 2.4720006415880363e-05, | |
| "loss": 0.007, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.26463104325699743, | |
| "grad_norm": 0.7073638439178467, | |
| "learning_rate": 2.4600758222076384e-05, | |
| "loss": 0.0099, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.26717557251908397, | |
| "grad_norm": 0.16835665702819824, | |
| "learning_rate": 2.44832192877507e-05, | |
| "loss": 0.0014, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.2697201017811705, | |
| "grad_norm": 0.34415891766548157, | |
| "learning_rate": 2.4367349166095938e-05, | |
| "loss": 0.0023, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.272264631043257, | |
| "grad_norm": 0.7876896262168884, | |
| "learning_rate": 2.425310873769525e-05, | |
| "loss": 0.0065, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.2748091603053435, | |
| "grad_norm": 0.6582947373390198, | |
| "learning_rate": 2.4140460155033424e-05, | |
| "loss": 0.0092, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.27735368956743, | |
| "grad_norm": 0.6005846261978149, | |
| "learning_rate": 2.4029366789816987e-05, | |
| "loss": 0.0065, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.27989821882951654, | |
| "grad_norm": 0.10599358379840851, | |
| "learning_rate": 2.3919793182936757e-05, | |
| "loss": 0.01, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.2824427480916031, | |
| "grad_norm": 0.00374388019554317, | |
| "learning_rate": 2.3811704996917682e-05, | |
| "loss": 0.0058, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.28498727735368956, | |
| "grad_norm": 1.793428659439087, | |
| "learning_rate": 2.3705068970711013e-05, | |
| "loss": 0.0189, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.2875318066157761, | |
| "grad_norm": 0.027734549716114998, | |
| "learning_rate": 2.3599852876693835e-05, | |
| "loss": 0.0077, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.2900763358778626, | |
| "grad_norm": 0.22530615329742432, | |
| "learning_rate": 2.349602547974954e-05, | |
| "loss": 0.0042, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.2926208651399491, | |
| "grad_norm": 0.02637120895087719, | |
| "learning_rate": 2.3393556498311382e-05, | |
| "loss": 0.014, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.2951653944020356, | |
| "grad_norm": 0.4004020094871521, | |
| "learning_rate": 2.3292416567258735e-05, | |
| "loss": 0.0052, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.29770992366412213, | |
| "grad_norm": 0.7220229506492615, | |
| "learning_rate": 2.319257720256279e-05, | |
| "loss": 0.0133, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.30025445292620867, | |
| "grad_norm": 0.013986817561089993, | |
| "learning_rate": 2.3094010767585035e-05, | |
| "loss": 0.0057, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.30279898218829515, | |
| "grad_norm": 0.36194828152656555, | |
| "learning_rate": 2.2996690440938087e-05, | |
| "loss": 0.0107, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.3053435114503817, | |
| "grad_norm": 0.16057135164737701, | |
| "learning_rate": 2.29005901858238e-05, | |
| "loss": 0.0052, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.30788804071246817, | |
| "grad_norm": 0.01337255910038948, | |
| "learning_rate": 2.280568472076922e-05, | |
| "loss": 0.0066, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.3104325699745547, | |
| "grad_norm": 0.006197742652148008, | |
| "learning_rate": 2.2711949491685637e-05, | |
| "loss": 0.0111, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.31297709923664124, | |
| "grad_norm": 0.1564462035894394, | |
| "learning_rate": 2.2619360645180608e-05, | |
| "loss": 0.0153, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.3155216284987277, | |
| "grad_norm": 0.04476480558514595, | |
| "learning_rate": 2.2527895003057102e-05, | |
| "loss": 0.0053, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.31806615776081426, | |
| "grad_norm": 0.013111419975757599, | |
| "learning_rate": 2.243753003793784e-05, | |
| "loss": 0.0059, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.32061068702290074, | |
| "grad_norm": 0.05261224880814552, | |
| "learning_rate": 2.2348243849956642e-05, | |
| "loss": 0.0192, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.3231552162849873, | |
| "grad_norm": 0.009392138570547104, | |
| "learning_rate": 2.226001514446202e-05, | |
| "loss": 0.0066, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.3256997455470738, | |
| "grad_norm": 0.04299544915556908, | |
| "learning_rate": 2.2172823210681463e-05, | |
| "loss": 0.0038, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.3282442748091603, | |
| "grad_norm": 0.28492188453674316, | |
| "learning_rate": 2.2086647901297833e-05, | |
| "loss": 0.0049, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.33078880407124683, | |
| "grad_norm": 1.8210421800613403, | |
| "learning_rate": 2.2001469612892246e-05, | |
| "loss": 0.0093, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.9147886633872986, | |
| "learning_rate": 2.1917269267210227e-05, | |
| "loss": 0.0089, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.33587786259541985, | |
| "grad_norm": 0.02488136850297451, | |
| "learning_rate": 2.1834028293210553e-05, | |
| "loss": 0.0097, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.3384223918575064, | |
| "grad_norm": 0.12990175187587738, | |
| "learning_rate": 2.1751728609858345e-05, | |
| "loss": 0.0029, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.34096692111959287, | |
| "grad_norm": 0.5939871668815613, | |
| "learning_rate": 2.1670352609626394e-05, | |
| "loss": 0.0107, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.3435114503816794, | |
| "grad_norm": 0.041174307465553284, | |
| "learning_rate": 2.158988314267029e-05, | |
| "loss": 0.0083, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.3460559796437659, | |
| "grad_norm": 1.2240724563598633, | |
| "learning_rate": 2.1510303501645252e-05, | |
| "loss": 0.015, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.3486005089058524, | |
| "grad_norm": 0.04685092717409134, | |
| "learning_rate": 2.1431597407134065e-05, | |
| "loss": 0.0084, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.3511450381679389, | |
| "grad_norm": 0.6073548793792725, | |
| "learning_rate": 2.1353748993657274e-05, | |
| "loss": 0.0106, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.35368956743002544, | |
| "grad_norm": 0.6542628407478333, | |
| "learning_rate": 2.1276742796238307e-05, | |
| "loss": 0.0095, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.356234096692112, | |
| "grad_norm": 0.04672204703092575, | |
| "learning_rate": 2.1200563737497774e-05, | |
| "loss": 0.0046, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.35877862595419846, | |
| "grad_norm": 0.06259926408529282, | |
| "learning_rate": 2.11251971152524e-05, | |
| "loss": 0.012, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.361323155216285, | |
| "grad_norm": 0.7438691854476929, | |
| "learning_rate": 2.1050628590595503e-05, | |
| "loss": 0.0104, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.3638676844783715, | |
| "grad_norm": 0.02988184057176113, | |
| "learning_rate": 2.097684417643701e-05, | |
| "loss": 0.0054, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.366412213740458, | |
| "grad_norm": 0.05854351446032524, | |
| "learning_rate": 2.0903830226482316e-05, | |
| "loss": 0.0144, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.36895674300254455, | |
| "grad_norm": 0.008221745491027832, | |
| "learning_rate": 2.0831573424630117e-05, | |
| "loss": 0.0077, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.37150127226463103, | |
| "grad_norm": 0.02257680892944336, | |
| "learning_rate": 2.0760060774770695e-05, | |
| "loss": 0.0075, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.37404580152671757, | |
| "grad_norm": 0.2832575738430023, | |
| "learning_rate": 2.0689279590966737e-05, | |
| "loss": 0.0043, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.37659033078880405, | |
| "grad_norm": 0.21180367469787598, | |
| "learning_rate": 2.0619217487999985e-05, | |
| "loss": 0.0132, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.3791348600508906, | |
| "grad_norm": 0.10797836631536484, | |
| "learning_rate": 2.0549862372267617e-05, | |
| "loss": 0.0065, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.3816793893129771, | |
| "grad_norm": 0.01145731657743454, | |
| "learning_rate": 2.0481202433013253e-05, | |
| "loss": 0.0136, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3816793893129771, | |
| "eval_loss": 0.02801239863038063, | |
| "eval_runtime": 133.8059, | |
| "eval_samples_per_second": 59.654, | |
| "eval_steps_per_second": 0.471, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3842239185750636, | |
| "grad_norm": 1.122421383857727, | |
| "learning_rate": 2.0413226133878078e-05, | |
| "loss": 0.0163, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.38676844783715014, | |
| "grad_norm": 0.2691030204296112, | |
| "learning_rate": 2.0345922204758385e-05, | |
| "loss": 0.0041, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.3893129770992366, | |
| "grad_norm": 0.23942989110946655, | |
| "learning_rate": 2.027927963395653e-05, | |
| "loss": 0.0079, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.39185750636132316, | |
| "grad_norm": 0.013372058048844337, | |
| "learning_rate": 2.0213287660612816e-05, | |
| "loss": 0.0039, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.3944020356234097, | |
| "grad_norm": 0.4761992394924164, | |
| "learning_rate": 2.0147935767406558e-05, | |
| "loss": 0.0069, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.3969465648854962, | |
| "grad_norm": 0.3715521991252899, | |
| "learning_rate": 2.0083213673515046e-05, | |
| "loss": 0.0085, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.3994910941475827, | |
| "grad_norm": 0.22629395127296448, | |
| "learning_rate": 2.0019111327819738e-05, | |
| "loss": 0.0038, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.4020356234096692, | |
| "grad_norm": 0.030647089704871178, | |
| "learning_rate": 1.995561890234947e-05, | |
| "loss": 0.003, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.40458015267175573, | |
| "grad_norm": 0.6564404964447021, | |
| "learning_rate": 1.9892726785950967e-05, | |
| "loss": 0.007, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.4071246819338422, | |
| "grad_norm": 0.036405447870492935, | |
| "learning_rate": 1.98304255781774e-05, | |
| "loss": 0.0053, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.40966921119592875, | |
| "grad_norm": 0.005857866257429123, | |
| "learning_rate": 1.976870608338622e-05, | |
| "loss": 0.0044, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.4122137404580153, | |
| "grad_norm": 0.1405838429927826, | |
| "learning_rate": 1.9707559305037783e-05, | |
| "loss": 0.0013, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.41475826972010177, | |
| "grad_norm": 0.9452566504478455, | |
| "learning_rate": 1.964697644018673e-05, | |
| "loss": 0.0051, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.4173027989821883, | |
| "grad_norm": 0.844308614730835, | |
| "learning_rate": 1.9586948874158576e-05, | |
| "loss": 0.007, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.4198473282442748, | |
| "grad_norm": 0.002340355422347784, | |
| "learning_rate": 1.952746817540408e-05, | |
| "loss": 0.0026, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.4223918575063613, | |
| "grad_norm": 0.02571692131459713, | |
| "learning_rate": 1.9468526090524474e-05, | |
| "loss": 0.0128, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.42493638676844786, | |
| "grad_norm": 0.029216062277555466, | |
| "learning_rate": 1.9410114539460892e-05, | |
| "loss": 0.0035, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.42748091603053434, | |
| "grad_norm": 0.022271933034062386, | |
| "learning_rate": 1.9352225610841573e-05, | |
| "loss": 0.0065, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.4300254452926209, | |
| "grad_norm": 0.09926922619342804, | |
| "learning_rate": 1.9294851557480873e-05, | |
| "loss": 0.0112, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.43256997455470736, | |
| "grad_norm": 0.08391325920820236, | |
| "learning_rate": 1.9237984792024138e-05, | |
| "loss": 0.0094, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.4351145038167939, | |
| "grad_norm": 0.7565645575523376, | |
| "learning_rate": 1.918161788273296e-05, | |
| "loss": 0.0078, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.43765903307888043, | |
| "grad_norm": 0.0018272479064762592, | |
| "learning_rate": 1.9125743549405512e-05, | |
| "loss": 0.0091, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.4402035623409669, | |
| "grad_norm": 1.0079314708709717, | |
| "learning_rate": 1.9070354659426805e-05, | |
| "loss": 0.0064, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.44274809160305345, | |
| "grad_norm": 0.09381988644599915, | |
| "learning_rate": 1.901544422394406e-05, | |
| "loss": 0.0013, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.44529262086513993, | |
| "grad_norm": 0.7228587865829468, | |
| "learning_rate": 1.8961005394162554e-05, | |
| "loss": 0.0069, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.44783715012722647, | |
| "grad_norm": 0.007973475381731987, | |
| "learning_rate": 1.8907031457757393e-05, | |
| "loss": 0.011, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.45038167938931295, | |
| "grad_norm": 0.3692808747291565, | |
| "learning_rate": 1.8853515835397072e-05, | |
| "loss": 0.0083, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.4529262086513995, | |
| "grad_norm": 1.1002589464187622, | |
| "learning_rate": 1.8800452077374613e-05, | |
| "loss": 0.0147, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.455470737913486, | |
| "grad_norm": 0.10630931705236435, | |
| "learning_rate": 1.8747833860342444e-05, | |
| "loss": 0.0062, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.4580152671755725, | |
| "grad_norm": 0.685546875, | |
| "learning_rate": 1.8695654984147258e-05, | |
| "loss": 0.0074, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.46055979643765904, | |
| "grad_norm": 0.03145020082592964, | |
| "learning_rate": 1.864390936876117e-05, | |
| "loss": 0.0091, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.4631043256997455, | |
| "grad_norm": 0.07562462240457535, | |
| "learning_rate": 1.859259105130588e-05, | |
| "loss": 0.0042, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.46564885496183206, | |
| "grad_norm": 0.028905624523758888, | |
| "learning_rate": 1.8541694183166402e-05, | |
| "loss": 0.0065, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.4681933842239186, | |
| "grad_norm": 0.016208523884415627, | |
| "learning_rate": 1.8491213027191207e-05, | |
| "loss": 0.0056, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.4707379134860051, | |
| "grad_norm": 0.26138824224472046, | |
| "learning_rate": 1.844114195497582e-05, | |
| "loss": 0.005, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.4732824427480916, | |
| "grad_norm": 0.0996871143579483, | |
| "learning_rate": 1.839147544422687e-05, | |
| "loss": 0.0154, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.4758269720101781, | |
| "grad_norm": 0.003987440373748541, | |
| "learning_rate": 1.8342208076203763e-05, | |
| "loss": 0.0076, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.47837150127226463, | |
| "grad_norm": 0.32995522022247314, | |
| "learning_rate": 1.8293334533235443e-05, | |
| "loss": 0.0035, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.48091603053435117, | |
| "grad_norm": 0.7788913249969482, | |
| "learning_rate": 1.824484959630946e-05, | |
| "loss": 0.0085, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.48346055979643765, | |
| "grad_norm": 0.20872271060943604, | |
| "learning_rate": 1.8196748142731004e-05, | |
| "loss": 0.007, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.4860050890585242, | |
| "grad_norm": 0.05579596012830734, | |
| "learning_rate": 1.8149025143849454e-05, | |
| "loss": 0.0058, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.48854961832061067, | |
| "grad_norm": 0.5995292067527771, | |
| "learning_rate": 1.8101675662850178e-05, | |
| "loss": 0.0041, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.4910941475826972, | |
| "grad_norm": 1.3341422080993652, | |
| "learning_rate": 1.8054694852609384e-05, | |
| "loss": 0.0092, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.49363867684478374, | |
| "grad_norm": 0.23060783743858337, | |
| "learning_rate": 1.8008077953609884e-05, | |
| "loss": 0.004, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.4961832061068702, | |
| "grad_norm": 0.014038242399692535, | |
| "learning_rate": 1.7961820291915756e-05, | |
| "loss": 0.0083, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.49872773536895676, | |
| "grad_norm": 0.15661761164665222, | |
| "learning_rate": 1.7915917277203928e-05, | |
| "loss": 0.0044, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.5012722646310432, | |
| "grad_norm": 0.8456106781959534, | |
| "learning_rate": 1.7870364400850857e-05, | |
| "loss": 0.007, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.5038167938931297, | |
| "grad_norm": 0.18953552842140198, | |
| "learning_rate": 1.7825157234072404e-05, | |
| "loss": 0.0082, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.5063613231552163, | |
| "grad_norm": 0.08143863826990128, | |
| "learning_rate": 1.7780291426115227e-05, | |
| "loss": 0.003, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.5089058524173028, | |
| "grad_norm": 0.14154134690761566, | |
| "learning_rate": 1.7735762702498013e-05, | |
| "loss": 0.0021, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5089058524173028, | |
| "eval_loss": 0.02811964973807335, | |
| "eval_runtime": 133.8173, | |
| "eval_samples_per_second": 59.649, | |
| "eval_steps_per_second": 0.471, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5114503816793893, | |
| "grad_norm": 0.00306331948377192, | |
| "learning_rate": 1.769156686330089e-05, | |
| "loss": 0.0138, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.5139949109414759, | |
| "grad_norm": 0.6728852391242981, | |
| "learning_rate": 1.7647699781501523e-05, | |
| "loss": 0.0058, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.5165394402035624, | |
| "grad_norm": 0.10445057600736618, | |
| "learning_rate": 1.7604157401356363e-05, | |
| "loss": 0.0057, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.5190839694656488, | |
| "grad_norm": 0.6738550662994385, | |
| "learning_rate": 1.756093573682559e-05, | |
| "loss": 0.0083, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.5216284987277354, | |
| "grad_norm": 0.23501446843147278, | |
| "learning_rate": 1.7518030870040435e-05, | |
| "loss": 0.0081, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.5241730279898219, | |
| "grad_norm": 0.8946853280067444, | |
| "learning_rate": 1.747543894981144e-05, | |
| "loss": 0.0068, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.5267175572519084, | |
| "grad_norm": 0.4180065393447876, | |
| "learning_rate": 1.743315619017645e-05, | |
| "loss": 0.0096, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.5292620865139949, | |
| "grad_norm": 0.6527079939842224, | |
| "learning_rate": 1.7391178868987076e-05, | |
| "loss": 0.0112, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.5318066157760815, | |
| "grad_norm": 0.047336891293525696, | |
| "learning_rate": 1.7349503326532365e-05, | |
| "loss": 0.0109, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.5343511450381679, | |
| "grad_norm": 0.9120574593544006, | |
| "learning_rate": 1.7308125964198664e-05, | |
| "loss": 0.0126, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.5368956743002544, | |
| "grad_norm": 0.05327541008591652, | |
| "learning_rate": 1.7267043243164382e-05, | |
| "loss": 0.0124, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.539440203562341, | |
| "grad_norm": 0.23076896369457245, | |
| "learning_rate": 1.7226251683128732e-05, | |
| "loss": 0.0073, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.5419847328244275, | |
| "grad_norm": 0.039539627730846405, | |
| "learning_rate": 1.7185747861073344e-05, | |
| "loss": 0.0048, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.544529262086514, | |
| "grad_norm": 0.00934471283107996, | |
| "learning_rate": 1.7145528410055728e-05, | |
| "loss": 0.0104, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.5470737913486005, | |
| "grad_norm": 0.18967941403388977, | |
| "learning_rate": 1.710559001803368e-05, | |
| "loss": 0.0083, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.549618320610687, | |
| "grad_norm": 0.4472905099391937, | |
| "learning_rate": 1.7065929426719686e-05, | |
| "loss": 0.0054, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.5521628498727735, | |
| "grad_norm": 0.14181415736675262, | |
| "learning_rate": 1.7026543430464334e-05, | |
| "loss": 0.0046, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.55470737913486, | |
| "grad_norm": 0.7059574127197266, | |
| "learning_rate": 1.6987428875168074e-05, | |
| "loss": 0.0097, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.5572519083969466, | |
| "grad_norm": 0.015217378735542297, | |
| "learning_rate": 1.6948582657220192e-05, | |
| "loss": 0.0059, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.5597964376590331, | |
| "grad_norm": 0.010189504362642765, | |
| "learning_rate": 1.691000172246448e-05, | |
| "loss": 0.0131, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.5623409669211196, | |
| "grad_norm": 1.080907940864563, | |
| "learning_rate": 1.6871683065190647e-05, | |
| "loss": 0.0046, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.5648854961832062, | |
| "grad_norm": 0.7677445411682129, | |
| "learning_rate": 1.6833623727150718e-05, | |
| "loss": 0.0103, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.5674300254452926, | |
| "grad_norm": 1.186445713043213, | |
| "learning_rate": 1.679582079659979e-05, | |
| "loss": 0.0151, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.5699745547073791, | |
| "grad_norm": 0.1466684639453888, | |
| "learning_rate": 1.6758271407360342e-05, | |
| "loss": 0.0025, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.5725190839694656, | |
| "grad_norm": 0.15782137215137482, | |
| "learning_rate": 1.6720972737909477e-05, | |
| "loss": 0.0114, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.5750636132315522, | |
| "grad_norm": 0.03126196190714836, | |
| "learning_rate": 1.6683922010488377e-05, | |
| "loss": 0.0062, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.5776081424936387, | |
| "grad_norm": 0.26838138699531555, | |
| "learning_rate": 1.6647116490233395e-05, | |
| "loss": 0.0089, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.5801526717557252, | |
| "grad_norm": 0.4912380278110504, | |
| "learning_rate": 1.6610553484328155e-05, | |
| "loss": 0.0085, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.5826972010178118, | |
| "grad_norm": 0.805382490158081, | |
| "learning_rate": 1.6574230341176004e-05, | |
| "loss": 0.0067, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.5852417302798982, | |
| "grad_norm": 0.006444927304983139, | |
| "learning_rate": 1.6538144449592305e-05, | |
| "loss": 0.0007, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.5877862595419847, | |
| "grad_norm": 0.43014952540397644, | |
| "learning_rate": 1.6502293238016015e-05, | |
| "loss": 0.0023, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.5903307888040712, | |
| "grad_norm": 0.05080411583185196, | |
| "learning_rate": 1.6466674173739937e-05, | |
| "loss": 0.0117, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.5928753180661578, | |
| "grad_norm": 0.17079254984855652, | |
| "learning_rate": 1.643128476215921e-05, | |
| "loss": 0.0061, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.5954198473282443, | |
| "grad_norm": 0.19195914268493652, | |
| "learning_rate": 1.639612254603744e-05, | |
| "loss": 0.006, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.5979643765903307, | |
| "grad_norm": 0.007362790871411562, | |
| "learning_rate": 1.636118510479011e-05, | |
| "loss": 0.0099, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.6005089058524173, | |
| "grad_norm": 0.3465839624404907, | |
| "learning_rate": 1.632647005378462e-05, | |
| "loss": 0.0027, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.6030534351145038, | |
| "grad_norm": 0.1840086132287979, | |
| "learning_rate": 1.6291975043656685e-05, | |
| "loss": 0.0066, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.6055979643765903, | |
| "grad_norm": 0.012280162423849106, | |
| "learning_rate": 1.6257697759642478e-05, | |
| "loss": 0.0075, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.6081424936386769, | |
| "grad_norm": 0.41994690895080566, | |
| "learning_rate": 1.6223635920926225e-05, | |
| "loss": 0.0063, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.6106870229007634, | |
| "grad_norm": 0.41221898794174194, | |
| "learning_rate": 1.6189787280002733e-05, | |
| "loss": 0.0108, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.6132315521628499, | |
| "grad_norm": 0.1125999167561531, | |
| "learning_rate": 1.61561496220545e-05, | |
| "loss": 0.0081, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.6157760814249363, | |
| "grad_norm": 0.6083981394767761, | |
| "learning_rate": 1.612272076434301e-05, | |
| "loss": 0.0121, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.6183206106870229, | |
| "grad_norm": 0.09369134902954102, | |
| "learning_rate": 1.608949855561382e-05, | |
| "loss": 0.0032, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.6208651399491094, | |
| "grad_norm": 0.030923601239919662, | |
| "learning_rate": 1.6056480875515067e-05, | |
| "loss": 0.0082, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.6234096692111959, | |
| "grad_norm": 0.6387837529182434, | |
| "learning_rate": 1.6023665634029055e-05, | |
| "loss": 0.0058, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.6259541984732825, | |
| "grad_norm": 0.752968430519104, | |
| "learning_rate": 1.599105077091658e-05, | |
| "loss": 0.0128, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.628498727735369, | |
| "grad_norm": 0.38870304822921753, | |
| "learning_rate": 1.595863425517362e-05, | |
| "loss": 0.0055, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.6310432569974554, | |
| "grad_norm": 0.3654159605503082, | |
| "learning_rate": 1.592641408450013e-05, | |
| "loss": 0.0157, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.6335877862595419, | |
| "grad_norm": 0.009396991692483425, | |
| "learning_rate": 1.589438828478053e-05, | |
| "loss": 0.0104, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.6361323155216285, | |
| "grad_norm": 0.6921645998954773, | |
| "learning_rate": 1.586255490957569e-05, | |
| "loss": 0.0043, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.6361323155216285, | |
| "eval_loss": 0.028026102110743523, | |
| "eval_runtime": 133.7954, | |
| "eval_samples_per_second": 59.658, | |
| "eval_steps_per_second": 0.471, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.638676844783715, | |
| "grad_norm": 0.6047800183296204, | |
| "learning_rate": 1.5830912039626052e-05, | |
| "loss": 0.0124, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.6412213740458015, | |
| "grad_norm": 0.6583201289176941, | |
| "learning_rate": 1.5799457782365596e-05, | |
| "loss": 0.0074, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.6437659033078881, | |
| "grad_norm": 0.6491327285766602, | |
| "learning_rate": 1.5768190271446377e-05, | |
| "loss": 0.0096, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.6463104325699746, | |
| "grad_norm": 0.005162398796528578, | |
| "learning_rate": 1.5737107666273427e-05, | |
| "loss": 0.0056, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.648854961832061, | |
| "grad_norm": 0.6562647223472595, | |
| "learning_rate": 1.5706208151549625e-05, | |
| "loss": 0.0089, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.6513994910941476, | |
| "grad_norm": 0.29285362362861633, | |
| "learning_rate": 1.567548993683042e-05, | |
| "loss": 0.0237, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.6539440203562341, | |
| "grad_norm": 0.9552783370018005, | |
| "learning_rate": 1.564495125608807e-05, | |
| "loss": 0.0106, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.6564885496183206, | |
| "grad_norm": 0.0369524247944355, | |
| "learning_rate": 1.5614590367285192e-05, | |
| "loss": 0.0113, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.6590330788804071, | |
| "grad_norm": 0.5022505521774292, | |
| "learning_rate": 1.5584405551957365e-05, | |
| "loss": 0.0122, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.6615776081424937, | |
| "grad_norm": 0.18602563440799713, | |
| "learning_rate": 1.5554395114804587e-05, | |
| "loss": 0.0039, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.6641221374045801, | |
| "grad_norm": 0.013193932361900806, | |
| "learning_rate": 1.552455738329136e-05, | |
| "loss": 0.0101, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.009286697022616863, | |
| "learning_rate": 1.5494890707255163e-05, | |
| "loss": 0.0044, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.6692111959287532, | |
| "grad_norm": 0.015414154157042503, | |
| "learning_rate": 1.5465393458523143e-05, | |
| "loss": 0.0036, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.6717557251908397, | |
| "grad_norm": 0.6608064770698547, | |
| "learning_rate": 1.5436064030536782e-05, | |
| "loss": 0.0071, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.6743002544529262, | |
| "grad_norm": 0.02665918506681919, | |
| "learning_rate": 1.5406900837984375e-05, | |
| "loss": 0.0038, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.6768447837150128, | |
| "grad_norm": 0.37374401092529297, | |
| "learning_rate": 1.5377902316441124e-05, | |
| "loss": 0.0025, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.6793893129770993, | |
| "grad_norm": 0.2854650318622589, | |
| "learning_rate": 1.5349066922016643e-05, | |
| "loss": 0.0131, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.6819338422391857, | |
| "grad_norm": 0.7415670156478882, | |
| "learning_rate": 1.532039313100972e-05, | |
| "loss": 0.0225, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.6844783715012722, | |
| "grad_norm": 0.8981530666351318, | |
| "learning_rate": 1.5291879439570117e-05, | |
| "loss": 0.0156, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.6870229007633588, | |
| "grad_norm": 0.9810027480125427, | |
| "learning_rate": 1.5263524363367325e-05, | |
| "loss": 0.0077, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.6895674300254453, | |
| "grad_norm": 1.304455280303955, | |
| "learning_rate": 1.5235326437265969e-05, | |
| "loss": 0.0075, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.6921119592875318, | |
| "grad_norm": 0.028728922829031944, | |
| "learning_rate": 1.5207284215007884e-05, | |
| "loss": 0.0095, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.6946564885496184, | |
| "grad_norm": 0.2182517647743225, | |
| "learning_rate": 1.5179396268900512e-05, | |
| "loss": 0.0093, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.6972010178117048, | |
| "grad_norm": 0.022339988499879837, | |
| "learning_rate": 1.5151661189511615e-05, | |
| "loss": 0.0112, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.6997455470737913, | |
| "grad_norm": 0.6623443365097046, | |
| "learning_rate": 1.5124077585370089e-05, | |
| "loss": 0.0043, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.7022900763358778, | |
| "grad_norm": 0.2521412968635559, | |
| "learning_rate": 1.5096644082672751e-05, | |
| "loss": 0.0047, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.7048346055979644, | |
| "grad_norm": 0.2759697437286377, | |
| "learning_rate": 1.5069359324996938e-05, | |
| "loss": 0.0054, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.7073791348600509, | |
| "grad_norm": 0.01425466500222683, | |
| "learning_rate": 1.5042221973018824e-05, | |
| "loss": 0.001, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.7099236641221374, | |
| "grad_norm": 0.0023090008180588484, | |
| "learning_rate": 1.5015230704237286e-05, | |
| "loss": 0.0034, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.712468193384224, | |
| "grad_norm": 0.30373772978782654, | |
| "learning_rate": 1.4988384212703178e-05, | |
| "loss": 0.0044, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.7150127226463104, | |
| "grad_norm": 0.00985446572303772, | |
| "learning_rate": 1.4961681208753971e-05, | |
| "loss": 0.0036, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.7175572519083969, | |
| "grad_norm": 0.6338773965835571, | |
| "learning_rate": 1.4935120418753497e-05, | |
| "loss": 0.0048, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.7201017811704835, | |
| "grad_norm": 0.016199562698602676, | |
| "learning_rate": 1.4908700584836826e-05, | |
| "loss": 0.0022, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.72264631043257, | |
| "grad_norm": 0.008430557325482368, | |
| "learning_rate": 1.4882420464660043e-05, | |
| "loss": 0.0153, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.7251908396946565, | |
| "grad_norm": 0.0057014962658286095, | |
| "learning_rate": 1.4856278831154885e-05, | |
| "loss": 0.006, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.727735368956743, | |
| "grad_norm": 0.4593695104122162, | |
| "learning_rate": 1.4830274472288094e-05, | |
| "loss": 0.0144, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.7302798982188295, | |
| "grad_norm": 0.39744073152542114, | |
| "learning_rate": 1.480440619082538e-05, | |
| "loss": 0.0011, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.732824427480916, | |
| "grad_norm": 0.9719061255455017, | |
| "learning_rate": 1.477867280409992e-05, | |
| "loss": 0.0064, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.7353689567430025, | |
| "grad_norm": 0.46188387274742126, | |
| "learning_rate": 1.4753073143785246e-05, | |
| "loss": 0.0098, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.7379134860050891, | |
| "grad_norm": 0.30528566241264343, | |
| "learning_rate": 1.4727606055672447e-05, | |
| "loss": 0.0122, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.7404580152671756, | |
| "grad_norm": 0.025703368708491325, | |
| "learning_rate": 1.4702270399451596e-05, | |
| "loss": 0.0137, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.7430025445292621, | |
| "grad_norm": 0.12070760130882263, | |
| "learning_rate": 1.4677065048497305e-05, | |
| "loss": 0.0047, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.7455470737913485, | |
| "grad_norm": 0.0634564459323883, | |
| "learning_rate": 1.4651988889658286e-05, | |
| "loss": 0.0018, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.7480916030534351, | |
| "grad_norm": 0.5759007334709167, | |
| "learning_rate": 1.4627040823050882e-05, | |
| "loss": 0.0051, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.7506361323155216, | |
| "grad_norm": 0.43193572759628296, | |
| "learning_rate": 1.4602219761856428e-05, | |
| "loss": 0.0113, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.7531806615776081, | |
| "grad_norm": 0.5299224257469177, | |
| "learning_rate": 1.4577524632122396e-05, | |
| "loss": 0.0065, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.7557251908396947, | |
| "grad_norm": 0.6598678827285767, | |
| "learning_rate": 1.455295437256723e-05, | |
| "loss": 0.0071, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.7582697201017812, | |
| "grad_norm": 0.25693264603614807, | |
| "learning_rate": 1.4528507934388762e-05, | |
| "loss": 0.0086, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.7608142493638677, | |
| "grad_norm": 0.1258089691400528, | |
| "learning_rate": 1.4504184281076198e-05, | |
| "loss": 0.0033, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.7633587786259542, | |
| "grad_norm": 0.824597954750061, | |
| "learning_rate": 1.4479982388225518e-05, | |
| "loss": 0.0069, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7633587786259542, | |
| "eval_loss": 0.027132505550980568, | |
| "eval_runtime": 133.9317, | |
| "eval_samples_per_second": 59.598, | |
| "eval_steps_per_second": 0.47, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7659033078880407, | |
| "grad_norm": 0.06601405888795853, | |
| "learning_rate": 1.4455901243358257e-05, | |
| "loss": 0.0064, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.7684478371501272, | |
| "grad_norm": 0.08991851657629013, | |
| "learning_rate": 1.4431939845743629e-05, | |
| "loss": 0.0054, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.7709923664122137, | |
| "grad_norm": 1.6022424697875977, | |
| "learning_rate": 1.4408097206223825e-05, | |
| "loss": 0.0065, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.7735368956743003, | |
| "grad_norm": 1.0877054929733276, | |
| "learning_rate": 1.4384372347042513e-05, | |
| "loss": 0.0061, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.7760814249363868, | |
| "grad_norm": 0.032010771334171295, | |
| "learning_rate": 1.436076430167644e-05, | |
| "loss": 0.0019, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.7786259541984732, | |
| "grad_norm": 0.82038414478302, | |
| "learning_rate": 1.433727211467003e-05, | |
| "loss": 0.0094, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.7811704834605598, | |
| "grad_norm": 0.017111314460635185, | |
| "learning_rate": 1.4313894841472981e-05, | |
| "loss": 0.0186, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.7837150127226463, | |
| "grad_norm": 0.2755248248577118, | |
| "learning_rate": 1.4290631548280754e-05, | |
| "loss": 0.0035, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.7862595419847328, | |
| "grad_norm": 0.04039892554283142, | |
| "learning_rate": 1.4267481311877889e-05, | |
| "loss": 0.0061, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.7888040712468194, | |
| "grad_norm": 0.40358659625053406, | |
| "learning_rate": 1.4244443219484134e-05, | |
| "loss": 0.0064, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.7913486005089059, | |
| "grad_norm": 0.14195334911346436, | |
| "learning_rate": 1.4221516368603278e-05, | |
| "loss": 0.0085, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.7938931297709924, | |
| "grad_norm": 0.0759008452296257, | |
| "learning_rate": 1.4198699866874657e-05, | |
| "loss": 0.0064, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.7964376590330788, | |
| "grad_norm": 0.008402438834309578, | |
| "learning_rate": 1.417599283192727e-05, | |
| "loss": 0.0105, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.7989821882951654, | |
| "grad_norm": 0.15167561173439026, | |
| "learning_rate": 1.415339439123645e-05, | |
| "loss": 0.0095, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.8015267175572519, | |
| "grad_norm": 0.034887611865997314, | |
| "learning_rate": 1.4130903681983059e-05, | |
| "loss": 0.0067, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.8040712468193384, | |
| "grad_norm": 0.6523444056510925, | |
| "learning_rate": 1.410851985091511e-05, | |
| "loss": 0.0148, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.806615776081425, | |
| "grad_norm": 0.047939080744981766, | |
| "learning_rate": 1.4086242054211812e-05, | |
| "loss": 0.0085, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.8091603053435115, | |
| "grad_norm": 0.408014714717865, | |
| "learning_rate": 1.4064069457349956e-05, | |
| "loss": 0.006, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.811704834605598, | |
| "grad_norm": 0.024463389068841934, | |
| "learning_rate": 1.4042001234972625e-05, | |
| "loss": 0.014, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.8142493638676844, | |
| "grad_norm": 0.9237146973609924, | |
| "learning_rate": 1.402003657076014e-05, | |
| "loss": 0.0316, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.816793893129771, | |
| "grad_norm": 0.01679268479347229, | |
| "learning_rate": 1.3998174657303234e-05, | |
| "loss": 0.0015, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.8193384223918575, | |
| "grad_norm": 0.04732313007116318, | |
| "learning_rate": 1.3976414695978396e-05, | |
| "loss": 0.0013, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.821882951653944, | |
| "grad_norm": 0.2467341274023056, | |
| "learning_rate": 1.3954755896825334e-05, | |
| "loss": 0.0131, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.8244274809160306, | |
| "grad_norm": 0.12651531398296356, | |
| "learning_rate": 1.3933197478426519e-05, | |
| "loss": 0.0049, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.8269720101781171, | |
| "grad_norm": 0.19898846745491028, | |
| "learning_rate": 1.3911738667788778e-05, | |
| "loss": 0.004, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.8295165394402035, | |
| "grad_norm": 0.3841744065284729, | |
| "learning_rate": 1.3890378700226879e-05, | |
| "loss": 0.0041, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.8320610687022901, | |
| "grad_norm": 0.5270954966545105, | |
| "learning_rate": 1.3869116819249068e-05, | |
| "loss": 0.0064, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.8346055979643766, | |
| "grad_norm": 0.7140441536903381, | |
| "learning_rate": 1.3847952276444563e-05, | |
| "loss": 0.0148, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.8371501272264631, | |
| "grad_norm": 0.25958600640296936, | |
| "learning_rate": 1.382688433137289e-05, | |
| "loss": 0.0053, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.8396946564885496, | |
| "grad_norm": 0.22590410709381104, | |
| "learning_rate": 1.3805912251455073e-05, | |
| "loss": 0.0101, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.8422391857506362, | |
| "grad_norm": 0.5887141823768616, | |
| "learning_rate": 1.3785035311866659e-05, | |
| "loss": 0.0017, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.8447837150127226, | |
| "grad_norm": 0.16586758196353912, | |
| "learning_rate": 1.3764252795432498e-05, | |
| "loss": 0.0067, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.8473282442748091, | |
| "grad_norm": 0.5686965584754944, | |
| "learning_rate": 1.3743563992523249e-05, | |
| "loss": 0.0048, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.8498727735368957, | |
| "grad_norm": 0.12965810298919678, | |
| "learning_rate": 1.3722968200953633e-05, | |
| "loss": 0.0117, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.8524173027989822, | |
| "grad_norm": 0.5761728286743164, | |
| "learning_rate": 1.3702464725882309e-05, | |
| "loss": 0.0051, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.8549618320610687, | |
| "grad_norm": 0.2968486249446869, | |
| "learning_rate": 1.368205287971344e-05, | |
| "loss": 0.0144, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.8575063613231552, | |
| "grad_norm": 0.3824020326137543, | |
| "learning_rate": 1.3661731981999809e-05, | |
| "loss": 0.0074, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.8600508905852418, | |
| "grad_norm": 0.296509712934494, | |
| "learning_rate": 1.3641501359347584e-05, | |
| "loss": 0.0061, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.8625954198473282, | |
| "grad_norm": 0.016193876042962074, | |
| "learning_rate": 1.3621360345322563e-05, | |
| "loss": 0.0049, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.8651399491094147, | |
| "grad_norm": 0.17405743896961212, | |
| "learning_rate": 1.3601308280357978e-05, | |
| "loss": 0.0046, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.8676844783715013, | |
| "grad_norm": 0.7698820233345032, | |
| "learning_rate": 1.358134451166378e-05, | |
| "loss": 0.0132, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.8702290076335878, | |
| "grad_norm": 0.006794361397624016, | |
| "learning_rate": 1.3561468393137367e-05, | |
| "loss": 0.0028, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.8727735368956743, | |
| "grad_norm": 0.49688103795051575, | |
| "learning_rate": 1.3541679285275769e-05, | |
| "loss": 0.0029, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.8753180661577609, | |
| "grad_norm": 0.004117690026760101, | |
| "learning_rate": 1.3521976555089225e-05, | |
| "loss": 0.0022, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.8778625954198473, | |
| "grad_norm": 0.0018297217320650816, | |
| "learning_rate": 1.3502359576016142e-05, | |
| "loss": 0.0097, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.8804071246819338, | |
| "grad_norm": 0.0006447790074162185, | |
| "learning_rate": 1.34828277278394e-05, | |
| "loss": 0.0054, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.8829516539440203, | |
| "grad_norm": 0.009917440824210644, | |
| "learning_rate": 1.3463380396604007e-05, | |
| "loss": 0.0021, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.8854961832061069, | |
| "grad_norm": 0.22393424808979034, | |
| "learning_rate": 1.3444016974536018e-05, | |
| "loss": 0.0037, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.8880407124681934, | |
| "grad_norm": 0.35406559705734253, | |
| "learning_rate": 1.3424736859962784e-05, | |
| "loss": 0.01, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.8905852417302799, | |
| "grad_norm": 0.08287867158651352, | |
| "learning_rate": 1.3405539457234403e-05, | |
| "loss": 0.0195, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8905852417302799, | |
| "eval_loss": 0.02591288834810257, | |
| "eval_runtime": 133.8254, | |
| "eval_samples_per_second": 59.645, | |
| "eval_steps_per_second": 0.471, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8931297709923665, | |
| "grad_norm": 0.00405901437625289, | |
| "learning_rate": 1.3386424176646451e-05, | |
| "loss": 0.0018, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.8956743002544529, | |
| "grad_norm": 0.31145212054252625, | |
| "learning_rate": 1.336739043436388e-05, | |
| "loss": 0.0032, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.8982188295165394, | |
| "grad_norm": 0.08784864097833633, | |
| "learning_rate": 1.334843765234613e-05, | |
| "loss": 0.004, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.9007633587786259, | |
| "grad_norm": 0.777217447757721, | |
| "learning_rate": 1.3329565258273396e-05, | |
| "loss": 0.0091, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.9033078880407125, | |
| "grad_norm": 0.01620590128004551, | |
| "learning_rate": 1.331077268547404e-05, | |
| "loss": 0.0067, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.905852417302799, | |
| "grad_norm": 0.15450182557106018, | |
| "learning_rate": 1.3292059372853128e-05, | |
| "loss": 0.0059, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.9083969465648855, | |
| "grad_norm": 0.010797875002026558, | |
| "learning_rate": 1.3273424764822053e-05, | |
| "loss": 0.0056, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.910941475826972, | |
| "grad_norm": 0.008933750912547112, | |
| "learning_rate": 1.3254868311229282e-05, | |
| "loss": 0.0059, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.9134860050890585, | |
| "grad_norm": 0.3677749037742615, | |
| "learning_rate": 1.32363894672921e-05, | |
| "loss": 0.0068, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.916030534351145, | |
| "grad_norm": 0.2783452272415161, | |
| "learning_rate": 1.321798769352946e-05, | |
| "loss": 0.0033, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.9185750636132316, | |
| "grad_norm": 0.005901079159229994, | |
| "learning_rate": 1.3199662455695826e-05, | |
| "loss": 0.0006, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.9211195928753181, | |
| "grad_norm": 0.01929521933197975, | |
| "learning_rate": 1.3181413224716023e-05, | |
| "loss": 0.0101, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.9236641221374046, | |
| "grad_norm": 0.06383396685123444, | |
| "learning_rate": 1.3163239476621085e-05, | |
| "loss": 0.0078, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.926208651399491, | |
| "grad_norm": 0.04800323024392128, | |
| "learning_rate": 1.3145140692485066e-05, | |
| "loss": 0.0104, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.9287531806615776, | |
| "grad_norm": 0.10172835737466812, | |
| "learning_rate": 1.31271163583628e-05, | |
| "loss": 0.0053, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.9312977099236641, | |
| "grad_norm": 0.009509807452559471, | |
| "learning_rate": 1.3109165965228612e-05, | |
| "loss": 0.0037, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.9338422391857506, | |
| "grad_norm": 0.07621141523122787, | |
| "learning_rate": 1.309128900891594e-05, | |
| "loss": 0.0015, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.9363867684478372, | |
| "grad_norm": 0.07766814529895782, | |
| "learning_rate": 1.3073484990057842e-05, | |
| "loss": 0.002, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.9389312977099237, | |
| "grad_norm": 0.09857495874166489, | |
| "learning_rate": 1.3055753414028447e-05, | |
| "loss": 0.0056, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.9414758269720102, | |
| "grad_norm": 0.12120025604963303, | |
| "learning_rate": 1.3038093790885189e-05, | |
| "loss": 0.01, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.9440203562340967, | |
| "grad_norm": 0.05350268632173538, | |
| "learning_rate": 1.3020505635311987e-05, | |
| "loss": 0.0018, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.9465648854961832, | |
| "grad_norm": 0.08092239499092102, | |
| "learning_rate": 1.3002988466563204e-05, | |
| "loss": 0.0035, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.9491094147582697, | |
| "grad_norm": 0.0031507923267781734, | |
| "learning_rate": 1.2985541808408472e-05, | |
| "loss": 0.0078, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.9516539440203562, | |
| "grad_norm": 0.4591497778892517, | |
| "learning_rate": 1.2968165189078303e-05, | |
| "loss": 0.0037, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.9541984732824428, | |
| "grad_norm": 0.056590236723423004, | |
| "learning_rate": 1.2950858141210503e-05, | |
| "loss": 0.0014, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.9567430025445293, | |
| "grad_norm": 0.14643314480781555, | |
| "learning_rate": 1.293362020179739e-05, | |
| "loss": 0.0034, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.9592875318066157, | |
| "grad_norm": 0.019364474341273308, | |
| "learning_rate": 1.2916450912133767e-05, | |
| "loss": 0.0081, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.9618320610687023, | |
| "grad_norm": 0.033629197627305984, | |
| "learning_rate": 1.2899349817765634e-05, | |
| "loss": 0.0092, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.9643765903307888, | |
| "grad_norm": 0.03282175213098526, | |
| "learning_rate": 1.288231646843969e-05, | |
| "loss": 0.0025, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.9669211195928753, | |
| "grad_norm": 0.015059337019920349, | |
| "learning_rate": 1.286535041805354e-05, | |
| "loss": 0.0041, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.9694656488549618, | |
| "grad_norm": 0.30329233407974243, | |
| "learning_rate": 1.2848451224606597e-05, | |
| "loss": 0.0096, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.9720101781170484, | |
| "grad_norm": 0.020984278991818428, | |
| "learning_rate": 1.2831618450151757e-05, | |
| "loss": 0.0023, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.9745547073791349, | |
| "grad_norm": 0.3885207772254944, | |
| "learning_rate": 1.2814851660747694e-05, | |
| "loss": 0.0186, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.9770992366412213, | |
| "grad_norm": 0.015083876438438892, | |
| "learning_rate": 1.2798150426411895e-05, | |
| "loss": 0.0139, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.9796437659033079, | |
| "grad_norm": 0.08502122759819031, | |
| "learning_rate": 1.2781514321074345e-05, | |
| "loss": 0.0084, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.9821882951653944, | |
| "grad_norm": 0.005835570394992828, | |
| "learning_rate": 1.2764942922531862e-05, | |
| "loss": 0.0062, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.9847328244274809, | |
| "grad_norm": 0.1602826863527298, | |
| "learning_rate": 1.2748435812403105e-05, | |
| "loss": 0.0017, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.9872773536895675, | |
| "grad_norm": 0.13282841444015503, | |
| "learning_rate": 1.2731992576084201e-05, | |
| "loss": 0.0045, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.989821882951654, | |
| "grad_norm": 0.04718279093503952, | |
| "learning_rate": 1.271561280270502e-05, | |
| "loss": 0.0052, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.9923664122137404, | |
| "grad_norm": 0.005922272801399231, | |
| "learning_rate": 1.2699296085086024e-05, | |
| "loss": 0.0034, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.9949109414758269, | |
| "grad_norm": 0.236740842461586, | |
| "learning_rate": 1.2683042019695793e-05, | |
| "loss": 0.0047, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.9974554707379135, | |
| "grad_norm": 0.03246605023741722, | |
| "learning_rate": 1.2666850206609087e-05, | |
| "loss": 0.0012, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.5907068852102384e-05, | |
| "learning_rate": 1.2650720249465511e-05, | |
| "loss": 0.0041, | |
| "step": 3930 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3930, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 2.273606668827065e+18, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |