| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.5188916876574305, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008396305625524769, | |
| "grad_norm": 10.171950340270996, | |
| "learning_rate": 4.5e-06, | |
| "loss": 1.1261, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.016792611251049538, | |
| "grad_norm": 2.374096632003784, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.5171, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02518891687657431, | |
| "grad_norm": 1.9998104572296143, | |
| "learning_rate": 1.45e-05, | |
| "loss": 0.2565, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.033585222502099076, | |
| "grad_norm": 1.4298408031463623, | |
| "learning_rate": 1.9500000000000003e-05, | |
| "loss": 0.1707, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.041981528127623846, | |
| "grad_norm": 1.2781989574432373, | |
| "learning_rate": 2.45e-05, | |
| "loss": 0.1302, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05037783375314862, | |
| "grad_norm": 1.4599987268447876, | |
| "learning_rate": 2.95e-05, | |
| "loss": 0.1016, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05877413937867338, | |
| "grad_norm": 0.7745999097824097, | |
| "learning_rate": 3.45e-05, | |
| "loss": 0.0848, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06717044500419815, | |
| "grad_norm": 1.022971272468567, | |
| "learning_rate": 3.9500000000000005e-05, | |
| "loss": 0.0818, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07556675062972293, | |
| "grad_norm": 0.4050329923629761, | |
| "learning_rate": 4.4500000000000004e-05, | |
| "loss": 0.0635, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08396305625524769, | |
| "grad_norm": 0.7355945110321045, | |
| "learning_rate": 4.9500000000000004e-05, | |
| "loss": 0.0647, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09235936188077246, | |
| "grad_norm": 0.9530194997787476, | |
| "learning_rate": 5.45e-05, | |
| "loss": 0.0634, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10075566750629723, | |
| "grad_norm": 0.8249333500862122, | |
| "learning_rate": 5.95e-05, | |
| "loss": 0.0556, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.109151973131822, | |
| "grad_norm": 0.6259945631027222, | |
| "learning_rate": 6.450000000000001e-05, | |
| "loss": 0.0554, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11754827875734676, | |
| "grad_norm": 0.8876557350158691, | |
| "learning_rate": 6.95e-05, | |
| "loss": 0.0495, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12594458438287154, | |
| "grad_norm": 0.777101993560791, | |
| "learning_rate": 7.450000000000001e-05, | |
| "loss": 0.0518, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1343408900083963, | |
| "grad_norm": 0.9956705570220947, | |
| "learning_rate": 7.950000000000001e-05, | |
| "loss": 0.0456, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.14273719563392107, | |
| "grad_norm": 0.5946229696273804, | |
| "learning_rate": 8.450000000000001e-05, | |
| "loss": 0.042, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.15113350125944586, | |
| "grad_norm": 0.7513177990913391, | |
| "learning_rate": 8.950000000000001e-05, | |
| "loss": 0.0416, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.15952980688497062, | |
| "grad_norm": 1.2336539030075073, | |
| "learning_rate": 9.449999999999999e-05, | |
| "loss": 0.04, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.16792611251049538, | |
| "grad_norm": 0.6550417542457581, | |
| "learning_rate": 9.95e-05, | |
| "loss": 0.0419, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17632241813602015, | |
| "grad_norm": 0.7052122354507446, | |
| "learning_rate": 9.999861593790126e-05, | |
| "loss": 0.0444, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.1847187237615449, | |
| "grad_norm": 0.5741733908653259, | |
| "learning_rate": 9.999383162408304e-05, | |
| "loss": 0.043, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.19311502938706968, | |
| "grad_norm": 0.7300972938537598, | |
| "learning_rate": 9.998563029828259e-05, | |
| "loss": 0.0396, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.20151133501259447, | |
| "grad_norm": 0.35928529500961304, | |
| "learning_rate": 9.997401252104962e-05, | |
| "loss": 0.0382, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.20990764063811923, | |
| "grad_norm": 0.5260980129241943, | |
| "learning_rate": 9.995897908644378e-05, | |
| "loss": 0.0332, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.218303946263644, | |
| "grad_norm": 0.4173150956630707, | |
| "learning_rate": 9.994053102198034e-05, | |
| "loss": 0.0314, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.22670025188916876, | |
| "grad_norm": 0.6619486808776855, | |
| "learning_rate": 9.991866958856003e-05, | |
| "loss": 0.0375, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.23509655751469352, | |
| "grad_norm": 0.6853200197219849, | |
| "learning_rate": 9.989339628038276e-05, | |
| "loss": 0.0348, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2434928631402183, | |
| "grad_norm": 0.7677571177482605, | |
| "learning_rate": 9.98647128248456e-05, | |
| "loss": 0.0434, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2518891687657431, | |
| "grad_norm": 0.47874388098716736, | |
| "learning_rate": 9.98326211824246e-05, | |
| "loss": 0.0339, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.26028547439126787, | |
| "grad_norm": 0.5458245277404785, | |
| "learning_rate": 9.979712354654091e-05, | |
| "loss": 0.0365, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2686817800167926, | |
| "grad_norm": 0.7390539646148682, | |
| "learning_rate": 9.975822234341079e-05, | |
| "loss": 0.0319, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2770780856423174, | |
| "grad_norm": 0.9258743524551392, | |
| "learning_rate": 9.97159202318798e-05, | |
| "loss": 0.0313, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.28547439126784213, | |
| "grad_norm": 0.4790763556957245, | |
| "learning_rate": 9.967022010324105e-05, | |
| "loss": 0.0308, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.2938706968933669, | |
| "grad_norm": 0.5265215635299683, | |
| "learning_rate": 9.962112508103765e-05, | |
| "loss": 0.0292, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3022670025188917, | |
| "grad_norm": 0.5428227782249451, | |
| "learning_rate": 9.956863852084914e-05, | |
| "loss": 0.0286, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.31066330814441645, | |
| "grad_norm": 0.5066298246383667, | |
| "learning_rate": 9.951276401006221e-05, | |
| "loss": 0.0316, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.31905961376994124, | |
| "grad_norm": 0.3620965778827667, | |
| "learning_rate": 9.945350536762543e-05, | |
| "loss": 0.0251, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.327455919395466, | |
| "grad_norm": 0.44671374559402466, | |
| "learning_rate": 9.939086664378829e-05, | |
| "loss": 0.0271, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.33585222502099077, | |
| "grad_norm": 0.4878353178501129, | |
| "learning_rate": 9.932485211982437e-05, | |
| "loss": 0.0228, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.34424853064651556, | |
| "grad_norm": 0.43716397881507874, | |
| "learning_rate": 9.92554663077387e-05, | |
| "loss": 0.0308, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3526448362720403, | |
| "grad_norm": 0.5659756064414978, | |
| "learning_rate": 9.918271394995935e-05, | |
| "loss": 0.0244, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.3610411418975651, | |
| "grad_norm": 0.4486585259437561, | |
| "learning_rate": 9.910660001901335e-05, | |
| "loss": 0.028, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.3694374475230898, | |
| "grad_norm": 0.4010409116744995, | |
| "learning_rate": 9.902712971718675e-05, | |
| "loss": 0.0268, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3778337531486146, | |
| "grad_norm": 0.48508334159851074, | |
| "learning_rate": 9.894430847616915e-05, | |
| "loss": 0.0258, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.38623005877413935, | |
| "grad_norm": 0.5513168573379517, | |
| "learning_rate": 9.885814195668232e-05, | |
| "loss": 0.0288, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.39462636439966414, | |
| "grad_norm": 0.3650225102901459, | |
| "learning_rate": 9.876863604809344e-05, | |
| "loss": 0.0284, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.40302267002518893, | |
| "grad_norm": 0.35296711325645447, | |
| "learning_rate": 9.867579686801245e-05, | |
| "loss": 0.0246, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.41141897565071367, | |
| "grad_norm": 0.3748459219932556, | |
| "learning_rate": 9.8579630761874e-05, | |
| "loss": 0.0233, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.41981528127623846, | |
| "grad_norm": 0.41755491495132446, | |
| "learning_rate": 9.848014430250367e-05, | |
| "loss": 0.023, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4282115869017632, | |
| "grad_norm": 0.43207406997680664, | |
| "learning_rate": 9.837734428966885e-05, | |
| "loss": 0.0245, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.436607892527288, | |
| "grad_norm": 0.27464979887008667, | |
| "learning_rate": 9.827123774961383e-05, | |
| "loss": 0.0216, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.4450041981528128, | |
| "grad_norm": 0.5782731771469116, | |
| "learning_rate": 9.816183193457968e-05, | |
| "loss": 0.0245, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.4534005037783375, | |
| "grad_norm": 0.559138834476471, | |
| "learning_rate": 9.804913432230856e-05, | |
| "loss": 0.024, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4617968094038623, | |
| "grad_norm": 0.6747875809669495, | |
| "learning_rate": 9.793315261553252e-05, | |
| "loss": 0.0256, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.47019311502938704, | |
| "grad_norm": 0.55149245262146, | |
| "learning_rate": 9.781389474144717e-05, | |
| "loss": 0.0234, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.47858942065491183, | |
| "grad_norm": 0.32782289385795593, | |
| "learning_rate": 9.76913688511698e-05, | |
| "loss": 0.0224, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.4869857262804366, | |
| "grad_norm": 0.5114140510559082, | |
| "learning_rate": 9.756558331918227e-05, | |
| "loss": 0.0236, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.49538203190596136, | |
| "grad_norm": 0.6146140098571777, | |
| "learning_rate": 9.743654674275855e-05, | |
| "loss": 0.0255, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5037783375314862, | |
| "grad_norm": 0.7248744964599609, | |
| "learning_rate": 9.730426794137727e-05, | |
| "loss": 0.0266, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5121746431570109, | |
| "grad_norm": 0.389180451631546, | |
| "learning_rate": 9.716875595611879e-05, | |
| "loss": 0.0267, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5205709487825357, | |
| "grad_norm": 0.49693188071250916, | |
| "learning_rate": 9.703002004904729e-05, | |
| "loss": 0.0227, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5289672544080605, | |
| "grad_norm": 0.5026288628578186, | |
| "learning_rate": 9.688806970257773e-05, | |
| "loss": 0.0197, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5373635600335852, | |
| "grad_norm": 0.35680270195007324, | |
| "learning_rate": 9.674291461882774e-05, | |
| "loss": 0.0195, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5457598656591099, | |
| "grad_norm": 0.3147616386413574, | |
| "learning_rate": 9.659456471895445e-05, | |
| "loss": 0.0212, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.5541561712846348, | |
| "grad_norm": 0.25688570737838745, | |
| "learning_rate": 9.644303014247648e-05, | |
| "loss": 0.0184, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5625524769101595, | |
| "grad_norm": 0.33005818724632263, | |
| "learning_rate": 9.628832124658085e-05, | |
| "loss": 0.0248, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5709487825356843, | |
| "grad_norm": 0.3768410384654999, | |
| "learning_rate": 9.613044860541507e-05, | |
| "loss": 0.0207, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5793450881612091, | |
| "grad_norm": 0.27856314182281494, | |
| "learning_rate": 9.596942300936445e-05, | |
| "loss": 0.0174, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5877413937867338, | |
| "grad_norm": 0.44634294509887695, | |
| "learning_rate": 9.580525546431459e-05, | |
| "loss": 0.0194, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5961376994122586, | |
| "grad_norm": 0.35559549927711487, | |
| "learning_rate": 9.563795719089911e-05, | |
| "loss": 0.0236, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6045340050377834, | |
| "grad_norm": 0.3311902582645416, | |
| "learning_rate": 9.546753962373281e-05, | |
| "loss": 0.0175, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6129303106633082, | |
| "grad_norm": 0.2235281765460968, | |
| "learning_rate": 9.529401441062997e-05, | |
| "loss": 0.018, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6213266162888329, | |
| "grad_norm": 0.3235013782978058, | |
| "learning_rate": 9.511739341180842e-05, | |
| "loss": 0.0142, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6297229219143576, | |
| "grad_norm": 0.4594292640686035, | |
| "learning_rate": 9.493768869907886e-05, | |
| "loss": 0.0194, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6381192275398825, | |
| "grad_norm": 0.4702000617980957, | |
| "learning_rate": 9.475491255501968e-05, | |
| "loss": 0.0149, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6465155331654072, | |
| "grad_norm": 0.3968784511089325, | |
| "learning_rate": 9.456907747213748e-05, | |
| "loss": 0.0164, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.654911838790932, | |
| "grad_norm": 0.5368077754974365, | |
| "learning_rate": 9.438019615201336e-05, | |
| "loss": 0.0183, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.6633081444164568, | |
| "grad_norm": 0.32421955466270447, | |
| "learning_rate": 9.418828150443469e-05, | |
| "loss": 0.0154, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.6717044500419815, | |
| "grad_norm": 0.36530694365501404, | |
| "learning_rate": 9.399334664651262e-05, | |
| "loss": 0.0173, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.6801007556675063, | |
| "grad_norm": 0.24784396588802338, | |
| "learning_rate": 9.379540490178581e-05, | |
| "loss": 0.0199, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6884970612930311, | |
| "grad_norm": 0.4272478520870209, | |
| "learning_rate": 9.359446979930955e-05, | |
| "loss": 0.0178, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.6968933669185559, | |
| "grad_norm": 0.39507076144218445, | |
| "learning_rate": 9.33905550727312e-05, | |
| "loss": 0.0183, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7052896725440806, | |
| "grad_norm": 0.20049844682216644, | |
| "learning_rate": 9.318367465935142e-05, | |
| "loss": 0.0156, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7136859781696053, | |
| "grad_norm": 0.8157637119293213, | |
| "learning_rate": 9.29738426991717e-05, | |
| "loss": 0.0207, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7220822837951302, | |
| "grad_norm": 0.5720648765563965, | |
| "learning_rate": 9.276107353392774e-05, | |
| "loss": 0.02, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7304785894206549, | |
| "grad_norm": 0.7491711974143982, | |
| "learning_rate": 9.254538170610938e-05, | |
| "loss": 0.0215, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7388748950461796, | |
| "grad_norm": 0.6429204344749451, | |
| "learning_rate": 9.232678195796654e-05, | |
| "loss": 0.0161, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.7472712006717045, | |
| "grad_norm": 0.5460074543952942, | |
| "learning_rate": 9.210528923050164e-05, | |
| "loss": 0.014, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.7556675062972292, | |
| "grad_norm": 0.5716153383255005, | |
| "learning_rate": 9.188091866244834e-05, | |
| "loss": 0.0209, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.764063811922754, | |
| "grad_norm": 0.5055357813835144, | |
| "learning_rate": 9.165368558923695e-05, | |
| "loss": 0.0143, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.7724601175482787, | |
| "grad_norm": 0.5202294588088989, | |
| "learning_rate": 9.142360554194618e-05, | |
| "loss": 0.0173, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.7808564231738035, | |
| "grad_norm": 0.2911999821662903, | |
| "learning_rate": 9.119069424624163e-05, | |
| "loss": 0.0152, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.7892527287993283, | |
| "grad_norm": 0.28226837515830994, | |
| "learning_rate": 9.0954967621301e-05, | |
| "loss": 0.0125, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.797649034424853, | |
| "grad_norm": 0.24133360385894775, | |
| "learning_rate": 9.071644177872594e-05, | |
| "loss": 0.018, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.8060453400503779, | |
| "grad_norm": 0.28321564197540283, | |
| "learning_rate": 9.047513302144095e-05, | |
| "loss": 0.0126, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8144416456759026, | |
| "grad_norm": 0.3964134752750397, | |
| "learning_rate": 9.023105784257906e-05, | |
| "loss": 0.0131, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.8228379513014273, | |
| "grad_norm": 0.4938385486602783, | |
| "learning_rate": 8.998423292435454e-05, | |
| "loss": 0.0166, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.8312342569269522, | |
| "grad_norm": 0.38572144508361816, | |
| "learning_rate": 8.973467513692265e-05, | |
| "loss": 0.0194, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8396305625524769, | |
| "grad_norm": 0.5354552268981934, | |
| "learning_rate": 8.94824015372267e-05, | |
| "loss": 0.0159, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8480268681780017, | |
| "grad_norm": 0.5144203305244446, | |
| "learning_rate": 8.922742936783207e-05, | |
| "loss": 0.0142, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.8564231738035264, | |
| "grad_norm": 0.49846023321151733, | |
| "learning_rate": 8.896977605574788e-05, | |
| "loss": 0.015, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.8648194794290512, | |
| "grad_norm": 0.48377203941345215, | |
| "learning_rate": 8.870945921123576e-05, | |
| "loss": 0.0151, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.873215785054576, | |
| "grad_norm": 0.45066890120506287, | |
| "learning_rate": 8.844649662660624e-05, | |
| "loss": 0.0145, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.8816120906801007, | |
| "grad_norm": 0.4206666946411133, | |
| "learning_rate": 8.818090627500266e-05, | |
| "loss": 0.0129, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8900083963056256, | |
| "grad_norm": 0.41736483573913574, | |
| "learning_rate": 8.791270630917275e-05, | |
| "loss": 0.0192, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.8984047019311503, | |
| "grad_norm": 0.3356010615825653, | |
| "learning_rate": 8.764191506022795e-05, | |
| "loss": 0.0148, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.906801007556675, | |
| "grad_norm": 0.3665725886821747, | |
| "learning_rate": 8.736855103639037e-05, | |
| "loss": 0.0155, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.9151973131821999, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.709263292172794e-05, | |
| "loss": 0.0082, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.9235936188077246, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.681417957487729e-05, | |
| "loss": 0.0, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9319899244332494, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.653321002775478e-05, | |
| "loss": 0.0, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9403862300587741, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.624974348425574e-05, | |
| "loss": 0.0, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.9487825356842989, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.596379931894188e-05, | |
| "loss": 0.0, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.9571788413098237, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.567539707571703e-05, | |
| "loss": 0.0, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.9655751469353484, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.538455646649146e-05, | |
| "loss": 0.0, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.9739714525608733, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.509129736983446e-05, | |
| "loss": 0.0, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.982367758186398, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.479563982961571e-05, | |
| "loss": 0.0, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.9907640638119227, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.449760405363539e-05, | |
| "loss": 0.0, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.9991603694374476, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.419721041224287e-05, | |
| "loss": 0.0, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.0075566750629723, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.389447943694451e-05, | |
| "loss": 0.0, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.015952980688497, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.358943181900032e-05, | |
| "loss": 0.0, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.0243492863140218, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.328208840800981e-05, | |
| "loss": 0.0, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.0327455919395465, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.297247021048686e-05, | |
| "loss": 0.0, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.0411418975650715, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.266059838842396e-05, | |
| "loss": 0.0, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.0495382031905962, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.23464942578459e-05, | |
| "loss": 0.0, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.057934508816121, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.203017928735277e-05, | |
| "loss": 0.0, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.0663308144416457, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.17116750966526e-05, | |
| "loss": 0.0, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.0747271200671704, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.139100345508377e-05, | |
| "loss": 0.0, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.0831234256926952, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.106818628012697e-05, | |
| "loss": 0.0, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.0915197313182199, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.074324563590736e-05, | |
| "loss": 0.0, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.0999160369437448, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.041620373168628e-05, | |
| "loss": 0.0, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.1083123425692696, | |
| "grad_norm": NaN, | |
| "learning_rate": 8.008708292034349e-05, | |
| "loss": 0.0, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.1167086481947943, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.975590569684925e-05, | |
| "loss": 0.0, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.125104953820319, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.942269469672687e-05, | |
| "loss": 0.0, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.1335012594458438, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.908747269450558e-05, | |
| "loss": 0.0, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.1418975650713685, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.875026260216393e-05, | |
| "loss": 0.0, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.1502938706968933, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.841108746756382e-05, | |
| "loss": 0.0, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.1586901763224182, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.806997047287516e-05, | |
| "loss": 0.0, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.167086481947943, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.772693493299138e-05, | |
| "loss": 0.0, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.1754827875734677, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.7382004293936e-05, | |
| "loss": 0.0, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.1838790931989924, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.703520213126e-05, | |
| "loss": 0.0, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.1922753988245172, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.66865521484305e-05, | |
| "loss": 0.0, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.200671704450042, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.633607817521074e-05, | |
| "loss": 0.0, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.2090680100755669, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.598380416603119e-05, | |
| "loss": 0.0, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.2174643157010916, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.562975419835247e-05, | |
| "loss": 0.0, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.2258606213266163, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.527395247101956e-05, | |
| "loss": 0.0, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.234256926952141, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.491642330260789e-05, | |
| "loss": 0.0, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.2426532325776658, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.45571911297612e-05, | |
| "loss": 0.0, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.2510495382031905, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.419628050552131e-05, | |
| "loss": 0.0, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.2594458438287153, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.383371609764999e-05, | |
| "loss": 0.0, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.26784214945424, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.346952268694288e-05, | |
| "loss": 0.0, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.276238455079765, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.310372516553585e-05, | |
| "loss": 0.0, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.2846347607052897, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.273634853520356e-05, | |
| "loss": 0.0, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.2930310663308144, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.236741790565072e-05, | |
| "loss": 0.0, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.3014273719563392, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.199695849279576e-05, | |
| "loss": 0.0, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.309823677581864, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.162499561704747e-05, | |
| "loss": 0.0, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.3182199832073889, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.125155470157429e-05, | |
| "loss": 0.0, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.3266162888329136, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.087666127056675e-05, | |
| "loss": 0.0, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.3350125944584383, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.050034094749286e-05, | |
| "loss": 0.0, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.343408900083963, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.012261945334683e-05, | |
| "loss": 0.0, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.3518052057094878, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.974352260489103e-05, | |
| "loss": 0.0, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.3602015113350125, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.936307631289148e-05, | |
| "loss": 0.0, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.3685978169605373, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.898130658034685e-05, | |
| "loss": 0.0, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.376994122586062, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.859823950071127e-05, | |
| "loss": 0.0, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.385390428211587, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.821390125611078e-05, | |
| "loss": 0.0, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.3937867338371117, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.782831811555385e-05, | |
| "loss": 0.0, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.4021830394626364, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.744151643313597e-05, | |
| "loss": 0.0, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.4105793450881612, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.705352264623828e-05, | |
| "loss": 0.0, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.418975650713686, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.666436327372078e-05, | |
| "loss": 0.0, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.4273719563392109, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.62740649141096e-05, | |
| "loss": 0.0, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.4357682619647356, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.588265424377919e-05, | |
| "loss": 0.0, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.4441645675902604, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.549015801512895e-05, | |
| "loss": 0.0, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.452560873215785, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.509660305475468e-05, | |
| "loss": 0.0, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.4609571788413098, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.47020162616152e-05, | |
| "loss": 0.0, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.4693534844668346, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.430642460519365e-05, | |
| "loss": 0.0, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.4777497900923593, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.390985512365426e-05, | |
| "loss": 0.0, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.486146095717884, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.351233492199431e-05, | |
| "loss": 0.0, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.4945424013434088, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.311389117019155e-05, | |
| "loss": 0.0, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.5029387069689337, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.271455110134713e-05, | |
| "loss": 0.0, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.5113350125944585, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.231434200982428e-05, | |
| "loss": 0.0, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.5197313182199832, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.191329124938285e-05, | |
| "loss": 0.0, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.528127623845508, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.15114262313095e-05, | |
| "loss": 0.0, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.536523929471033, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.110877442254444e-05, | |
| "loss": 0.0, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.5449202350965576, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.0705363343803946e-05, | |
| "loss": 0.0, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.5533165407220824, | |
| "grad_norm": NaN, | |
| "learning_rate": 6.030122056769934e-05, | |
| "loss": 0.0, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.561712846347607, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.989637371685257e-05, | |
| "loss": 0.0, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.5701091519731318, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.949085046200808e-05, | |
| "loss": 0.0, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.5785054575986566, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.908467852014169e-05, | |
| "loss": 0.0, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.5869017632241813, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.867788565256607e-05, | |
| "loss": 0.0, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.595298068849706, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.827049966303335e-05, | |
| "loss": 0.0, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.6036943744752308, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.786254839583478e-05, | |
| "loss": 0.0, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.6120906801007555, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.745405973389757e-05, | |
| "loss": 0.0, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.6204869857262805, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.7045061596879134e-05, | |
| "loss": 0.0, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.6288832913518052, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.6635581939258855e-05, | |
| "loss": 0.0, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.63727959697733, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.622564874842742e-05, | |
| "loss": 0.0, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.645675902602855, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.5815290042773836e-05, | |
| "loss": 0.0, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.6540722082283796, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.540453386977058e-05, | |
| "loss": 0.0, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.6624685138539044, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.4993408304056425e-05, | |
| "loss": 0.0, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.670864819479429, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.458194144551768e-05, | |
| "loss": 0.0, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.6792611251049538, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.417016141736756e-05, | |
| "loss": 0.0, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.6876574307304786, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.375809636422399e-05, | |
| "loss": 0.0, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.6960537363560033, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.334577445018599e-05, | |
| "loss": 0.0, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.704450041981528, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.293322385690867e-05, | |
| "loss": 0.0, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.7128463476070528, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.252047278167709e-05, | |
| "loss": 0.0, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.7212426532325775, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.210754943547893e-05, | |
| "loss": 0.0, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.7296389588581025, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.169448204107643e-05, | |
| "loss": 0.0, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.7380352644836272, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.128129883107729e-05, | |
| "loss": 0.0, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.746431570109152, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.086802804600505e-05, | |
| "loss": 0.0, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.7548278757346767, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.045469793236892e-05, | |
| "loss": 0.0, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.7632241813602016, | |
| "grad_norm": NaN, | |
| "learning_rate": 5.00413367407331e-05, | |
| "loss": 0.0, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.7716204869857264, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.9627972723785964e-05, | |
| "loss": 0.0, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.7800167926112511, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.921463413440898e-05, | |
| "loss": 0.0, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.7884130982367759, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.8801349223745654e-05, | |
| "loss": 0.0, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.7968094038623006, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.838814623927067e-05, | |
| "loss": 0.0, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.8052057094878253, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.797505342285912e-05, | |
| "loss": 0.0, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.81360201511335, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.756209900885628e-05, | |
| "loss": 0.0, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.8219983207388748, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.714931122214781e-05, | |
| "loss": 0.0, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.8303946263643995, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.673671827623058e-05, | |
| "loss": 0.0, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.8387909319899243, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.632434837128443e-05, | |
| "loss": 0.0, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.8471872376154492, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.591222969224453e-05, | |
| "loss": 0.0, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.855583543240974, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.550039040687518e-05, | |
| "loss": 0.0, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.8639798488664987, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.508885866384446e-05, | |
| "loss": 0.0, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.8723761544920237, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.4677662590800355e-05, | |
| "loss": 0.0, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.8807724601175484, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.426683029244825e-05, | |
| "loss": 0.0, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.8891687657430731, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.385638984863e-05, | |
| "loss": 0.0, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.8975650713685979, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.3446369312404745e-05, | |
| "loss": 0.0, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.9059613769941226, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.3036796708131474e-05, | |
| "loss": 0.0, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.9143576826196473, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.262770002955363e-05, | |
| "loss": 0.0, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.922753988245172, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.221910723788578e-05, | |
| "loss": 0.0, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.9311502938706968, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.1811046259902474e-05, | |
| "loss": 0.0, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.9395465994962215, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.140354498602952e-05, | |
| "loss": 0.0, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.9479429051217463, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.099663126843769e-05, | |
| "loss": 0.0, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.9563392107472712, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.059033291913902e-05, | |
| "loss": 0.0, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.964735516372796, | |
| "grad_norm": NaN, | |
| "learning_rate": 4.0184677708086014e-05, | |
| "loss": 0.0, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.9731318219983207, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.977969336127348e-05, | |
| "loss": 0.0, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 1.9815281276238457, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.937540755884357e-05, | |
| "loss": 0.0, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 1.9899244332493704, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.897184793319384e-05, | |
| "loss": 0.0, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 1.9983207388748951, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.856904206708863e-05, | |
| "loss": 0.0, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 2.00671704450042, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.8167017491773847e-05, | |
| "loss": 0.0, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 2.0151133501259446, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.776580168509516e-05, | |
| "loss": 0.0, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.0235096557514693, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.736542206962e-05, | |
| "loss": 0.0, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 2.031905961376994, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.696590601076326e-05, | |
| "loss": 0.0, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 2.040302267002519, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.656728081491686e-05, | |
| "loss": 0.0, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 2.0486985726280436, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.6169573727583405e-05, | |
| "loss": 0.0, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 2.0570948782535683, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.5772811931514036e-05, | |
| "loss": 0.0, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.065491183879093, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.5377022544850505e-05, | |
| "loss": 0.0, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 2.0738874895046178, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.498223261927158e-05, | |
| "loss": 0.0, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 2.082283795130143, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.4588469138144295e-05, | |
| "loss": 0.0, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 2.0906801007556677, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.419575901467952e-05, | |
| "loss": 0.0, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 2.0990764063811924, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.380412909009254e-05, | |
| "loss": 0.0, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.107472712006717, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.3413606131768475e-05, | |
| "loss": 0.0, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 2.115869017632242, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.302421683143279e-05, | |
| "loss": 0.0, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 2.1242653232577666, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.2635987803326896e-05, | |
| "loss": 0.0, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 2.1326616288832914, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.224894558238918e-05, | |
| "loss": 0.0, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 2.141057934508816, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.18631166224413e-05, | |
| "loss": 0.0, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.149454240134341, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.147852729438017e-05, | |
| "loss": 0.0, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 2.1578505457598656, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.109520388437548e-05, | |
| "loss": 0.0, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 2.1662468513853903, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.0713172592073116e-05, | |
| "loss": 0.0, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 2.174643157010915, | |
| "grad_norm": NaN, | |
| "learning_rate": 3.0332459528804457e-05, | |
| "loss": 0.0, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 2.1830394626364398, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.9953090715801634e-05, | |
| "loss": 0.0, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.1914357682619645, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.9575092082419086e-05, | |
| "loss": 0.0, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 2.1998320738874897, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.9198489464361288e-05, | |
| "loss": 0.0, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 2.2082283795130144, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.8823308601916948e-05, | |
| "loss": 0.0, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 2.216624685138539, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.8449575138199613e-05, | |
| "loss": 0.0, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.225020990764064, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.807731461739509e-05, | |
| "loss": 0.0, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.2334172963895886, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.7706552483015485e-05, | |
| "loss": 0.0, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 2.2418136020151134, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.733731407616018e-05, | |
| "loss": 0.0, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 2.250209907640638, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.6969624633783806e-05, | |
| "loss": 0.0, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 2.258606213266163, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.660350928697134e-05, | |
| "loss": 0.0, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 2.2670025188916876, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.6238993059220395e-05, | |
| "loss": 0.0, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.2753988245172123, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.5876100864730933e-05, | |
| "loss": 0.0, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 2.283795130142737, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.5514857506702405e-05, | |
| "loss": 0.0, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 2.292191435768262, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.5155287675638474e-05, | |
| "loss": 0.0, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 2.3005877413937865, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.4797415947659457e-05, | |
| "loss": 0.0, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 2.3089840470193117, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.4441266782822588e-05, | |
| "loss": 0.0, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.3173803526448364, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.4086864523450183e-05, | |
| "loss": 0.0, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 2.325776658270361, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.3734233392465903e-05, | |
| "loss": 0.0, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 2.334172963895886, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.3383397491739145e-05, | |
| "loss": 0.0, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 2.3425692695214106, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.3034380800437678e-05, | |
| "loss": 0.0, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 2.3509655751469354, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.2687207173388743e-05, | |
| "loss": 0.0, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.35936188077246, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.234190033944858e-05, | |
| "loss": 0.0, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 2.367758186397985, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.1998483899880596e-05, | |
| "loss": 0.0, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 2.3761544920235096, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.1656981326742266e-05, | |
| "loss": 0.0, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 2.3845507976490343, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.1317415961280824e-05, | |
| "loss": 0.0, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 2.392947103274559, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.097981101233794e-05, | |
| "loss": 0.0, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.401343408900084, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.0644189554763417e-05, | |
| "loss": 0.0, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 2.4097397145256085, | |
| "grad_norm": NaN, | |
| "learning_rate": 2.0310574527838072e-05, | |
| "loss": 0.0, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 2.4181360201511337, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.9978988733705807e-05, | |
| "loss": 0.0, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 2.4265323257766584, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.9649454835815202e-05, | |
| "loss": 0.0, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 2.434928631402183, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.932199535737045e-05, | |
| "loss": 0.0, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.443324937027708, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.8996632679791914e-05, | |
| "loss": 0.0, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 2.4517212426532327, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.8673389041186418e-05, | |
| "loss": 0.0, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 2.4601175482787574, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.8352286534827274e-05, | |
| "loss": 0.0, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 2.468513853904282, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.803334710764426e-05, | |
| "loss": 0.0, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 2.476910159529807, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.7716592558723556e-05, | |
| "loss": 0.0, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.4853064651553316, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.7402044537817824e-05, | |
| "loss": 0.0, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 2.4937027707808563, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.7089724543866465e-05, | |
| "loss": 0.0, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 2.502099076406381, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.6779653923526188e-05, | |
| "loss": 0.0, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 2.510495382031906, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.6471853869712023e-05, | |
| "loss": 0.0, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 2.5188916876574305, | |
| "grad_norm": NaN, | |
| "learning_rate": 1.6166345420148787e-05, | |
| "loss": 0.0, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 4000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |