| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1548, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01937984496124031, | |
| "grad_norm": 3.144579840345173, | |
| "learning_rate": 2e-06, | |
| "loss": 0.7225, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03875968992248062, | |
| "grad_norm": 1.4635286655772066, | |
| "learning_rate": 2e-06, | |
| "loss": 0.6518, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05813953488372093, | |
| "grad_norm": 1.3391923941713981, | |
| "learning_rate": 2e-06, | |
| "loss": 0.6267, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07751937984496124, | |
| "grad_norm": 1.6579265991584125, | |
| "learning_rate": 2e-06, | |
| "loss": 0.6192, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09689922480620156, | |
| "grad_norm": 1.4605383147970585, | |
| "learning_rate": 2e-06, | |
| "loss": 0.6139, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11627906976744186, | |
| "grad_norm": 3.0164024994264924, | |
| "learning_rate": 2e-06, | |
| "loss": 0.6069, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13565891472868216, | |
| "grad_norm": 3.0735559741358043, | |
| "learning_rate": 2e-06, | |
| "loss": 0.6026, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15503875968992248, | |
| "grad_norm": 2.179805880693448, | |
| "learning_rate": 2e-06, | |
| "loss": 0.6052, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1744186046511628, | |
| "grad_norm": 1.362604075543814, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5906, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1937984496124031, | |
| "grad_norm": 1.3800184946057543, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5926, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2131782945736434, | |
| "grad_norm": 1.5751853135874878, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5956, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.23255813953488372, | |
| "grad_norm": 1.3581579120999256, | |
| "learning_rate": 2e-06, | |
| "loss": 0.587, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.25193798449612403, | |
| "grad_norm": 1.4725915485732346, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5969, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2713178294573643, | |
| "grad_norm": 2.3468819417923297, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5835, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.29069767441860467, | |
| "grad_norm": 1.7250281157703675, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5835, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.31007751937984496, | |
| "grad_norm": 1.697210370789681, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5818, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.32945736434108525, | |
| "grad_norm": 1.4878874477377635, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5766, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3488372093023256, | |
| "grad_norm": 1.4577800027661107, | |
| "learning_rate": 2e-06, | |
| "loss": 0.581, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3682170542635659, | |
| "grad_norm": 1.5593039073057922, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5743, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.3875968992248062, | |
| "grad_norm": 1.5848277223563478, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5818, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4069767441860465, | |
| "grad_norm": 1.8749305561105472, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5794, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4263565891472868, | |
| "grad_norm": 1.9776238010410143, | |
| "learning_rate": 2e-06, | |
| "loss": 0.568, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.44573643410852715, | |
| "grad_norm": 1.8367468157602185, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5663, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "grad_norm": 1.4796763755300497, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5701, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4844961240310077, | |
| "grad_norm": 1.3277578070972216, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5709, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5038759689922481, | |
| "grad_norm": 1.3619348918199545, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5705, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5232558139534884, | |
| "grad_norm": 1.3242792342849865, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5712, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5426356589147286, | |
| "grad_norm": 1.208195333886979, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5666, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.562015503875969, | |
| "grad_norm": 1.2067351750756312, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5689, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5813953488372093, | |
| "grad_norm": 1.4904068289049572, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5771, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6007751937984496, | |
| "grad_norm": 1.2308496450990478, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5641, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6201550387596899, | |
| "grad_norm": 1.646171689094913, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5615, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6395348837209303, | |
| "grad_norm": 1.486911287484255, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5632, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6589147286821705, | |
| "grad_norm": 1.367582942415251, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5592, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6782945736434108, | |
| "grad_norm": 1.6042437679665664, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5563, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6976744186046512, | |
| "grad_norm": 1.3440772301905874, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5611, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7170542635658915, | |
| "grad_norm": 1.5512652472297026, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5665, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7364341085271318, | |
| "grad_norm": 1.2275175025786087, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5672, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7558139534883721, | |
| "grad_norm": 1.584906298076075, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5566, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7751937984496124, | |
| "grad_norm": 1.2472749697240508, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5576, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7945736434108527, | |
| "grad_norm": 1.3014141414308267, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5629, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.813953488372093, | |
| "grad_norm": 1.1170221459908387, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5579, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 1.2375437974864687, | |
| "learning_rate": 2e-06, | |
| "loss": 0.561, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8527131782945736, | |
| "grad_norm": 1.2886237743320625, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5593, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.872093023255814, | |
| "grad_norm": 1.2840806861492629, | |
| "learning_rate": 2e-06, | |
| "loss": 0.563, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.8914728682170543, | |
| "grad_norm": 1.1374554276601674, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5518, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9108527131782945, | |
| "grad_norm": 1.5460088934957399, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5581, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "grad_norm": 1.3649006878214347, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5586, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9496124031007752, | |
| "grad_norm": 1.4127688073512672, | |
| "learning_rate": 2e-06, | |
| "loss": 0.552, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9689922480620154, | |
| "grad_norm": 1.6637865739686932, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5529, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9883720930232558, | |
| "grad_norm": 1.1603715309408054, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5575, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0077519379844961, | |
| "grad_norm": 1.9471029922347483, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5363, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0271317829457365, | |
| "grad_norm": 1.501943317728709, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5103, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.0465116279069768, | |
| "grad_norm": 2.210385554433354, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5083, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.0658914728682172, | |
| "grad_norm": 1.4456437528958064, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5059, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.0852713178294573, | |
| "grad_norm": 1.3747977066368522, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5169, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.1046511627906976, | |
| "grad_norm": 1.3671143459062107, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5102, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.124031007751938, | |
| "grad_norm": 1.6112053859893298, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5044, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.1434108527131783, | |
| "grad_norm": 1.2311741203329047, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5036, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.1627906976744187, | |
| "grad_norm": 1.2615707450710407, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5037, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.1821705426356588, | |
| "grad_norm": 1.247560640905585, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5135, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.2015503875968991, | |
| "grad_norm": 1.4713651285134284, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5078, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.2209302325581395, | |
| "grad_norm": 1.8788503215188699, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5078, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.2403100775193798, | |
| "grad_norm": 1.9570944633210832, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5088, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.2596899224806202, | |
| "grad_norm": 1.519042994554157, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5098, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.2790697674418605, | |
| "grad_norm": 1.7929648758582426, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5067, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.2984496124031009, | |
| "grad_norm": 1.4987560500572878, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5122, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.3178294573643412, | |
| "grad_norm": 1.2086756990648462, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5121, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.3372093023255813, | |
| "grad_norm": 1.3866302425582229, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5088, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.3565891472868217, | |
| "grad_norm": 1.4943167093290222, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5099, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.375968992248062, | |
| "grad_norm": 1.4190704125897742, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5136, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.3953488372093024, | |
| "grad_norm": 1.201854415277148, | |
| "learning_rate": 2e-06, | |
| "loss": 0.513, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.4147286821705427, | |
| "grad_norm": 1.5552807283991774, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5087, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.4341085271317828, | |
| "grad_norm": 1.2711570348118688, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5148, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.4534883720930232, | |
| "grad_norm": 1.5027858680147197, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5101, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.4728682170542635, | |
| "grad_norm": 1.3009800268853209, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5075, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.4922480620155039, | |
| "grad_norm": 1.195133004986583, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5111, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.5116279069767442, | |
| "grad_norm": 1.3544355872789209, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5136, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.5310077519379846, | |
| "grad_norm": 1.22732892168229, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5107, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.550387596899225, | |
| "grad_norm": 1.2500964364741391, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5076, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.5697674418604652, | |
| "grad_norm": 1.2443445799817128, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5059, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.5891472868217056, | |
| "grad_norm": 1.3642427980887037, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5047, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.6085271317829457, | |
| "grad_norm": 1.3132643052702235, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5126, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.627906976744186, | |
| "grad_norm": 1.245951139796731, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5072, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.6472868217054264, | |
| "grad_norm": 1.151992957120151, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5119, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 1.44482076559048, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5143, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.6860465116279069, | |
| "grad_norm": 1.2014633120651383, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5065, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.7054263565891472, | |
| "grad_norm": 1.1421759011867347, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5112, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.7248062015503876, | |
| "grad_norm": 1.3662761429999075, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5105, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.744186046511628, | |
| "grad_norm": 1.731081249332488, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5145, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.7635658914728682, | |
| "grad_norm": 1.8386868390815672, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5064, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.7829457364341086, | |
| "grad_norm": 1.945733381736797, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5092, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.802325581395349, | |
| "grad_norm": 1.1658591371712888, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5102, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.8217054263565893, | |
| "grad_norm": 1.2964175995286236, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5084, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.8410852713178296, | |
| "grad_norm": 1.368071275379517, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5109, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.8604651162790697, | |
| "grad_norm": 1.150103079484035, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5129, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.87984496124031, | |
| "grad_norm": 1.741489344099722, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5102, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.8992248062015504, | |
| "grad_norm": 1.4368003237003146, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5063, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.9186046511627906, | |
| "grad_norm": 1.198714228722542, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5125, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.937984496124031, | |
| "grad_norm": 1.2050714331345398, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5098, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9573643410852712, | |
| "grad_norm": 1.2467265326889772, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5111, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.9767441860465116, | |
| "grad_norm": 1.2245457081729416, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5064, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.996124031007752, | |
| "grad_norm": 1.3046841459793526, | |
| "learning_rate": 2e-06, | |
| "loss": 0.5104, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.0155038759689923, | |
| "grad_norm": 1.69236342694069, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4698, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.0348837209302326, | |
| "grad_norm": 1.482227404306843, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4592, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.054263565891473, | |
| "grad_norm": 1.472873774310932, | |
| "learning_rate": 2e-06, | |
| "loss": 0.458, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.0736434108527133, | |
| "grad_norm": 1.5478690868120868, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4595, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.0930232558139537, | |
| "grad_norm": 1.3832474558503782, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4544, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.112403100775194, | |
| "grad_norm": 1.4739915975741762, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4575, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.1317829457364343, | |
| "grad_norm": 1.4093716459091368, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4574, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.1511627906976742, | |
| "grad_norm": 1.4597325989125942, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4604, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.1705426356589146, | |
| "grad_norm": 1.2596618218820486, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4622, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.189922480620155, | |
| "grad_norm": 1.8149063252784816, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4592, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.2093023255813953, | |
| "grad_norm": 1.552162424271938, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4572, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.2286821705426356, | |
| "grad_norm": 1.4310808330644234, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4596, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.248062015503876, | |
| "grad_norm": 1.1994256997594486, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4576, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.2674418604651163, | |
| "grad_norm": 1.539587341902846, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4584, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.2868217054263567, | |
| "grad_norm": 1.333352666406639, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4654, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.306201550387597, | |
| "grad_norm": 1.443832387313447, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4606, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.3255813953488373, | |
| "grad_norm": 1.2947150626163912, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4569, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.3449612403100777, | |
| "grad_norm": 1.2733884019792792, | |
| "learning_rate": 2e-06, | |
| "loss": 0.461, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.3643410852713176, | |
| "grad_norm": 1.2445241956347013, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4646, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.383720930232558, | |
| "grad_norm": 1.3643853680636246, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4598, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.4031007751937983, | |
| "grad_norm": 1.3508030176182113, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4659, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.4224806201550386, | |
| "grad_norm": 1.4315788663602833, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4606, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.441860465116279, | |
| "grad_norm": 1.2723100725181913, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4629, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.4612403100775193, | |
| "grad_norm": 1.3005627010479606, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4643, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.4806201550387597, | |
| "grad_norm": 1.3858018789769277, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4647, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.441079130557343, | |
| "learning_rate": 2e-06, | |
| "loss": 0.458, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.5193798449612403, | |
| "grad_norm": 1.20329198310741, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4651, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.5387596899224807, | |
| "grad_norm": 1.2598982072621057, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4584, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.558139534883721, | |
| "grad_norm": 1.2079452297185667, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4707, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.5775193798449614, | |
| "grad_norm": 1.2529237840070613, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4628, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.5968992248062017, | |
| "grad_norm": 1.524625077175533, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4716, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.616279069767442, | |
| "grad_norm": 1.32372209621441, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4629, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.6356589147286824, | |
| "grad_norm": 1.5832090394160376, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4595, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.6550387596899228, | |
| "grad_norm": 1.2622980294640742, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4605, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.6744186046511627, | |
| "grad_norm": 1.235197532731858, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4687, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.693798449612403, | |
| "grad_norm": 1.2791276926102568, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4603, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.7131782945736433, | |
| "grad_norm": 1.2282558958052718, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4683, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.7325581395348837, | |
| "grad_norm": 1.3538612143623057, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4601, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.751937984496124, | |
| "grad_norm": 1.3291955159874183, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4617, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.7713178294573644, | |
| "grad_norm": 1.2058531615141939, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4671, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.7906976744186047, | |
| "grad_norm": 1.3306346445143116, | |
| "learning_rate": 2e-06, | |
| "loss": 0.466, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.810077519379845, | |
| "grad_norm": 1.4572481853030812, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4637, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.8294573643410854, | |
| "grad_norm": 1.4181222197142325, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4707, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.8488372093023253, | |
| "grad_norm": 1.2329063854513693, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4666, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.8682170542635657, | |
| "grad_norm": 1.4499016075739342, | |
| "learning_rate": 2e-06, | |
| "loss": 0.467, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.887596899224806, | |
| "grad_norm": 1.4376308105469073, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4699, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.9069767441860463, | |
| "grad_norm": 1.3931682854746266, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4592, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.9263565891472867, | |
| "grad_norm": 1.3848029694873079, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4633, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.945736434108527, | |
| "grad_norm": 1.2870479115717015, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4684, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.9651162790697674, | |
| "grad_norm": 1.348746763089945, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4635, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.9844961240310077, | |
| "grad_norm": 1.2816015840213701, | |
| "learning_rate": 2e-06, | |
| "loss": 0.4687, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1548, | |
| "total_flos": 2592120137318400.0, | |
| "train_loss": 0.5171875406913363, | |
| "train_runtime": 22202.3543, | |
| "train_samples_per_second": 35.668, | |
| "train_steps_per_second": 0.07 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1548, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2592120137318400.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |