| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 2220, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04504504504504504, | |
| "grad_norm": 3.175370931625366, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 0.6725, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09009009009009009, | |
| "grad_norm": 1.692520260810852, | |
| "learning_rate": 1.7117117117117117e-05, | |
| "loss": 0.314, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13513513513513514, | |
| "grad_norm": 1.95720374584198, | |
| "learning_rate": 2.6126126126126128e-05, | |
| "loss": 0.2577, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.18018018018018017, | |
| "grad_norm": 1.2667555809020996, | |
| "learning_rate": 3.513513513513514e-05, | |
| "loss": 0.2377, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.22522522522522523, | |
| "grad_norm": 1.4453164339065552, | |
| "learning_rate": 4.414414414414415e-05, | |
| "loss": 0.2102, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2702702702702703, | |
| "grad_norm": 1.5506036281585693, | |
| "learning_rate": 5.3153153153153155e-05, | |
| "loss": 0.1921, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3153153153153153, | |
| "grad_norm": 0.9104294776916504, | |
| "learning_rate": 6.216216216216216e-05, | |
| "loss": 0.1875, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.36036036036036034, | |
| "grad_norm": 1.6911464929580688, | |
| "learning_rate": 7.117117117117116e-05, | |
| "loss": 0.1705, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.40540540540540543, | |
| "grad_norm": 1.2363160848617554, | |
| "learning_rate": 8.018018018018019e-05, | |
| "loss": 0.188, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.45045045045045046, | |
| "grad_norm": 1.005192518234253, | |
| "learning_rate": 8.918918918918919e-05, | |
| "loss": 0.1554, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4954954954954955, | |
| "grad_norm": 1.2431941032409668, | |
| "learning_rate": 9.81981981981982e-05, | |
| "loss": 0.1369, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5405405405405406, | |
| "grad_norm": 1.1538969278335571, | |
| "learning_rate": 9.999644972962145e-05, | |
| "loss": 0.1376, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5855855855855856, | |
| "grad_norm": 1.21573805809021, | |
| "learning_rate": 9.998202762029625e-05, | |
| "loss": 0.135, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6306306306306306, | |
| "grad_norm": 1.4540833234786987, | |
| "learning_rate": 9.995651497779182e-05, | |
| "loss": 0.1342, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6756756756756757, | |
| "grad_norm": 1.2237752676010132, | |
| "learning_rate": 9.991991746311917e-05, | |
| "loss": 0.1136, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.7207207207207207, | |
| "grad_norm": 1.0290318727493286, | |
| "learning_rate": 9.987224319691624e-05, | |
| "loss": 0.1114, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7657657657657657, | |
| "grad_norm": 0.8585171103477478, | |
| "learning_rate": 9.981350275764608e-05, | |
| "loss": 0.1005, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.8108108108108109, | |
| "grad_norm": 0.7964202761650085, | |
| "learning_rate": 9.97437091792495e-05, | |
| "loss": 0.1074, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.8558558558558559, | |
| "grad_norm": 0.6369871497154236, | |
| "learning_rate": 9.966287794825305e-05, | |
| "loss": 0.103, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.9009009009009009, | |
| "grad_norm": 0.5182180404663086, | |
| "learning_rate": 9.957102700033265e-05, | |
| "loss": 0.095, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9459459459459459, | |
| "grad_norm": 0.5850012898445129, | |
| "learning_rate": 9.946817671633384e-05, | |
| "loss": 0.0854, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.990990990990991, | |
| "grad_norm": 0.9804884791374207, | |
| "learning_rate": 9.935434991774952e-05, | |
| "loss": 0.0898, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.0360360360360361, | |
| "grad_norm": 0.7499725222587585, | |
| "learning_rate": 9.922957186165598e-05, | |
| "loss": 0.0882, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.0810810810810811, | |
| "grad_norm": 0.9342101216316223, | |
| "learning_rate": 9.909387023510871e-05, | |
| "loss": 0.0856, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.1261261261261262, | |
| "grad_norm": 0.752739667892456, | |
| "learning_rate": 9.894727514899881e-05, | |
| "loss": 0.0863, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.1711711711711712, | |
| "grad_norm": 0.5502750873565674, | |
| "learning_rate": 9.878981913137179e-05, | |
| "loss": 0.0805, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.2162162162162162, | |
| "grad_norm": 1.2474815845489502, | |
| "learning_rate": 9.862153712020972e-05, | |
| "loss": 0.0874, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.2612612612612613, | |
| "grad_norm": 0.8233453631401062, | |
| "learning_rate": 9.844246645567902e-05, | |
| "loss": 0.077, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.3063063063063063, | |
| "grad_norm": 0.5110873579978943, | |
| "learning_rate": 9.825264687184493e-05, | |
| "loss": 0.0765, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.3513513513513513, | |
| "grad_norm": 0.5558863878250122, | |
| "learning_rate": 9.805212048785493e-05, | |
| "loss": 0.08, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.3963963963963963, | |
| "grad_norm": 0.8767929673194885, | |
| "learning_rate": 9.78409317985929e-05, | |
| "loss": 0.0781, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.4414414414414414, | |
| "grad_norm": 0.6763063073158264, | |
| "learning_rate": 9.761912766480614e-05, | |
| "loss": 0.0749, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.4864864864864864, | |
| "grad_norm": 0.3673441410064697, | |
| "learning_rate": 9.738675730270737e-05, | |
| "loss": 0.0772, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.5315315315315314, | |
| "grad_norm": 0.3171934187412262, | |
| "learning_rate": 9.714387227305422e-05, | |
| "loss": 0.0725, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.5765765765765765, | |
| "grad_norm": 0.6207244992256165, | |
| "learning_rate": 9.689052646970829e-05, | |
| "loss": 0.0723, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.6216216216216215, | |
| "grad_norm": 0.7387151122093201, | |
| "learning_rate": 9.662677610767672e-05, | |
| "loss": 0.0705, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.5089544057846069, | |
| "learning_rate": 9.635267971063848e-05, | |
| "loss": 0.0667, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.7117117117117115, | |
| "grad_norm": 0.8173167109489441, | |
| "learning_rate": 9.606829809795871e-05, | |
| "loss": 0.0672, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.7567567567567568, | |
| "grad_norm": 0.5291303992271423, | |
| "learning_rate": 9.57736943711933e-05, | |
| "loss": 0.0731, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.8018018018018018, | |
| "grad_norm": 0.3580414950847626, | |
| "learning_rate": 9.546893390008738e-05, | |
| "loss": 0.0709, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.8468468468468469, | |
| "grad_norm": 0.643068253993988, | |
| "learning_rate": 9.515408430807036e-05, | |
| "loss": 0.0641, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.8918918918918919, | |
| "grad_norm": 0.6227248311042786, | |
| "learning_rate": 9.482921545725097e-05, | |
| "loss": 0.065, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.936936936936937, | |
| "grad_norm": 0.7317764163017273, | |
| "learning_rate": 9.449439943291541e-05, | |
| "loss": 0.0694, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.981981981981982, | |
| "grad_norm": 0.6944959163665771, | |
| "learning_rate": 9.414971052753252e-05, | |
| "loss": 0.0739, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.027027027027027, | |
| "grad_norm": 0.5378988981246948, | |
| "learning_rate": 9.379522522426879e-05, | |
| "loss": 0.066, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.0720720720720722, | |
| "grad_norm": 0.6093091368675232, | |
| "learning_rate": 9.343102218001762e-05, | |
| "loss": 0.0725, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.1171171171171173, | |
| "grad_norm": 1.0975353717803955, | |
| "learning_rate": 9.305718220794604e-05, | |
| "loss": 0.0748, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.1621621621621623, | |
| "grad_norm": 0.7575361132621765, | |
| "learning_rate": 9.267378825956301e-05, | |
| "loss": 0.0634, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.2072072072072073, | |
| "grad_norm": 0.5477670431137085, | |
| "learning_rate": 9.228092540631342e-05, | |
| "loss": 0.0661, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.2522522522522523, | |
| "grad_norm": 0.5405918955802917, | |
| "learning_rate": 9.187868082070132e-05, | |
| "loss": 0.0607, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.2972972972972974, | |
| "grad_norm": 0.7018790245056152, | |
| "learning_rate": 9.146714375694745e-05, | |
| "loss": 0.0567, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.3423423423423424, | |
| "grad_norm": 0.630402684211731, | |
| "learning_rate": 9.104640553118435e-05, | |
| "loss": 0.0618, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.3873873873873874, | |
| "grad_norm": 0.4060376286506653, | |
| "learning_rate": 9.06165595011943e-05, | |
| "loss": 0.0651, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.4324324324324325, | |
| "grad_norm": 0.5323811173439026, | |
| "learning_rate": 9.01777010456941e-05, | |
| "loss": 0.0569, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.4774774774774775, | |
| "grad_norm": 0.4360101521015167, | |
| "learning_rate": 8.972992754317144e-05, | |
| "loss": 0.0542, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.5225225225225225, | |
| "grad_norm": 0.6220657825469971, | |
| "learning_rate": 8.927333835027759e-05, | |
| "loss": 0.0615, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.5675675675675675, | |
| "grad_norm": 0.5657696723937988, | |
| "learning_rate": 8.880803477978102e-05, | |
| "loss": 0.058, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.6126126126126126, | |
| "grad_norm": 0.5093798637390137, | |
| "learning_rate": 8.833412007808713e-05, | |
| "loss": 0.0594, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.6576576576576576, | |
| "grad_norm": 0.452443927526474, | |
| "learning_rate": 8.78516994023289e-05, | |
| "loss": 0.0564, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.7027027027027026, | |
| "grad_norm": 0.4802651107311249, | |
| "learning_rate": 8.73608797970334e-05, | |
| "loss": 0.0496, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.7477477477477477, | |
| "grad_norm": 0.33183491230010986, | |
| "learning_rate": 8.686177017036979e-05, | |
| "loss": 0.053, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.7927927927927927, | |
| "grad_norm": 0.40296322107315063, | |
| "learning_rate": 8.635448126998352e-05, | |
| "loss": 0.0568, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.8378378378378377, | |
| "grad_norm": 0.30987316370010376, | |
| "learning_rate": 8.583912565842257e-05, | |
| "loss": 0.0515, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.8828828828828827, | |
| "grad_norm": 0.6701423525810242, | |
| "learning_rate": 8.531581768816084e-05, | |
| "loss": 0.0526, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.9279279279279278, | |
| "grad_norm": 0.439345121383667, | |
| "learning_rate": 8.478467347622443e-05, | |
| "loss": 0.0577, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.972972972972973, | |
| "grad_norm": 0.42080074548721313, | |
| "learning_rate": 8.424581087842647e-05, | |
| "loss": 0.048, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 3.018018018018018, | |
| "grad_norm": 0.43773335218429565, | |
| "learning_rate": 8.369934946321595e-05, | |
| "loss": 0.0523, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 3.063063063063063, | |
| "grad_norm": 0.45040780305862427, | |
| "learning_rate": 8.314541048514664e-05, | |
| "loss": 0.0509, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 3.108108108108108, | |
| "grad_norm": 0.36781787872314453, | |
| "learning_rate": 8.25841168579719e-05, | |
| "loss": 0.0495, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 3.153153153153153, | |
| "grad_norm": 0.477033406496048, | |
| "learning_rate": 8.201559312737132e-05, | |
| "loss": 0.0514, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.1981981981981984, | |
| "grad_norm": 0.4004456698894501, | |
| "learning_rate": 8.143996544331511e-05, | |
| "loss": 0.0479, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.2432432432432434, | |
| "grad_norm": 0.5489519834518433, | |
| "learning_rate": 8.085736153207277e-05, | |
| "loss": 0.0475, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.2882882882882885, | |
| "grad_norm": 0.5412061810493469, | |
| "learning_rate": 8.026791066787176e-05, | |
| "loss": 0.0548, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.23101916909217834, | |
| "learning_rate": 7.967174364421284e-05, | |
| "loss": 0.0503, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.3783783783783785, | |
| "grad_norm": 0.27605608105659485, | |
| "learning_rate": 7.90689927448482e-05, | |
| "loss": 0.0514, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.4234234234234235, | |
| "grad_norm": 0.29271090030670166, | |
| "learning_rate": 7.8459791714429e-05, | |
| "loss": 0.0498, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.4684684684684686, | |
| "grad_norm": 0.4988279938697815, | |
| "learning_rate": 7.784427572882871e-05, | |
| "loss": 0.0488, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.5135135135135136, | |
| "grad_norm": 0.4085995554924011, | |
| "learning_rate": 7.722258136514884e-05, | |
| "loss": 0.0484, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.5585585585585586, | |
| "grad_norm": 0.5015743374824524, | |
| "learning_rate": 7.659484657141382e-05, | |
| "loss": 0.0547, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 3.6036036036036037, | |
| "grad_norm": 0.41595619916915894, | |
| "learning_rate": 7.596121063596168e-05, | |
| "loss": 0.0493, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.6486486486486487, | |
| "grad_norm": 0.5095012187957764, | |
| "learning_rate": 7.532181415653725e-05, | |
| "loss": 0.0503, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 3.6936936936936937, | |
| "grad_norm": 0.4954228103160858, | |
| "learning_rate": 7.467679900909489e-05, | |
| "loss": 0.0528, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.7387387387387387, | |
| "grad_norm": 0.5485273003578186, | |
| "learning_rate": 7.40263083163176e-05, | |
| "loss": 0.0516, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 3.7837837837837838, | |
| "grad_norm": 0.35983964800834656, | |
| "learning_rate": 7.337048641585937e-05, | |
| "loss": 0.05, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 3.828828828828829, | |
| "grad_norm": 0.40537890791893005, | |
| "learning_rate": 7.270947882831822e-05, | |
| "loss": 0.044, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.873873873873874, | |
| "grad_norm": 0.3156987428665161, | |
| "learning_rate": 7.20434322249464e-05, | |
| "loss": 0.045, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 3.918918918918919, | |
| "grad_norm": 0.3144851326942444, | |
| "learning_rate": 7.137249439510548e-05, | |
| "loss": 0.0458, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 3.963963963963964, | |
| "grad_norm": 0.2909262478351593, | |
| "learning_rate": 7.069681421347339e-05, | |
| "loss": 0.0428, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 4.009009009009009, | |
| "grad_norm": 0.34615159034729004, | |
| "learning_rate": 7.001654160701046e-05, | |
| "loss": 0.0468, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 4.054054054054054, | |
| "grad_norm": 0.4577626883983612, | |
| "learning_rate": 6.933182752169213e-05, | |
| "loss": 0.0399, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.099099099099099, | |
| "grad_norm": 0.3926784098148346, | |
| "learning_rate": 6.864282388901544e-05, | |
| "loss": 0.0394, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 4.1441441441441444, | |
| "grad_norm": 0.3566577732563019, | |
| "learning_rate": 6.794968359228688e-05, | |
| "loss": 0.0452, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 4.1891891891891895, | |
| "grad_norm": 0.3146449029445648, | |
| "learning_rate": 6.725256043269912e-05, | |
| "loss": 0.0474, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 4.2342342342342345, | |
| "grad_norm": 0.2554456293582916, | |
| "learning_rate": 6.65516090952039e-05, | |
| "loss": 0.0458, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 4.2792792792792795, | |
| "grad_norm": 0.3618452250957489, | |
| "learning_rate": 6.584698511418901e-05, | |
| "loss": 0.0466, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.324324324324325, | |
| "grad_norm": 0.38914114236831665, | |
| "learning_rate": 6.513884483896666e-05, | |
| "loss": 0.0424, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 4.36936936936937, | |
| "grad_norm": 0.6733397841453552, | |
| "learning_rate": 6.4427345399081e-05, | |
| "loss": 0.0488, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 4.414414414414415, | |
| "grad_norm": 0.37581348419189453, | |
| "learning_rate": 6.37126446694427e-05, | |
| "loss": 0.0481, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 4.45945945945946, | |
| "grad_norm": 0.3915393352508545, | |
| "learning_rate": 6.299490123529797e-05, | |
| "loss": 0.0413, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 4.504504504504505, | |
| "grad_norm": 0.4132647216320038, | |
| "learning_rate": 6.227427435703997e-05, | |
| "loss": 0.0475, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.54954954954955, | |
| "grad_norm": 0.4121103584766388, | |
| "learning_rate": 6.155092393487051e-05, | |
| "loss": 0.0493, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 4.594594594594595, | |
| "grad_norm": 0.33546769618988037, | |
| "learning_rate": 6.08250104733197e-05, | |
| "loss": 0.041, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 4.63963963963964, | |
| "grad_norm": 0.34651216864585876, | |
| "learning_rate": 6.009669504563153e-05, | |
| "loss": 0.043, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 4.684684684684685, | |
| "grad_norm": 0.37287256121635437, | |
| "learning_rate": 5.9366139258023326e-05, | |
| "loss": 0.0413, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 4.72972972972973, | |
| "grad_norm": 0.3134576082229614, | |
| "learning_rate": 5.863350521382671e-05, | |
| "loss": 0.0405, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.774774774774775, | |
| "grad_norm": 0.33046990633010864, | |
| "learning_rate": 5.7898955477518666e-05, | |
| "loss": 0.0379, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 4.81981981981982, | |
| "grad_norm": 0.5138404369354248, | |
| "learning_rate": 5.716265303864978e-05, | |
| "loss": 0.0353, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 4.864864864864865, | |
| "grad_norm": 0.31907182931900024, | |
| "learning_rate": 5.642476127567866e-05, | |
| "loss": 0.0421, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 4.90990990990991, | |
| "grad_norm": 0.2643144428730011, | |
| "learning_rate": 5.5685443919719634e-05, | |
| "loss": 0.0431, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 4.954954954954955, | |
| "grad_norm": 0.3213581442832947, | |
| "learning_rate": 5.4944865018212497e-05, | |
| "loss": 0.0436, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.2239149659872055, | |
| "learning_rate": 5.4203188898521895e-05, | |
| "loss": 0.0406, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 5.045045045045045, | |
| "grad_norm": 0.3077867031097412, | |
| "learning_rate": 5.346058013147469e-05, | |
| "loss": 0.0417, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 5.09009009009009, | |
| "grad_norm": 0.4085425138473511, | |
| "learning_rate": 5.271720349484326e-05, | |
| "loss": 0.0397, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 5.135135135135135, | |
| "grad_norm": 0.4555751383304596, | |
| "learning_rate": 5.1973223936782887e-05, | |
| "loss": 0.0382, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 5.18018018018018, | |
| "grad_norm": 0.4915667474269867, | |
| "learning_rate": 5.122880653923134e-05, | |
| "loss": 0.0422, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 5.225225225225225, | |
| "grad_norm": 0.3977510333061218, | |
| "learning_rate": 5.04841164812788e-05, | |
| "loss": 0.0377, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 5.27027027027027, | |
| "grad_norm": 0.42886292934417725, | |
| "learning_rate": 4.973931900251611e-05, | |
| "loss": 0.0411, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 5.315315315315315, | |
| "grad_norm": 0.4493580460548401, | |
| "learning_rate": 4.899457936636988e-05, | |
| "loss": 0.0392, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 5.36036036036036, | |
| "grad_norm": 0.2820551097393036, | |
| "learning_rate": 4.82500628234319e-05, | |
| "loss": 0.0356, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 5.405405405405405, | |
| "grad_norm": 0.2753814160823822, | |
| "learning_rate": 4.750593457479171e-05, | |
| "loss": 0.0365, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.45045045045045, | |
| "grad_norm": 0.3056580424308777, | |
| "learning_rate": 4.676235973538013e-05, | |
| "loss": 0.0385, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 5.495495495495495, | |
| "grad_norm": 0.23721066117286682, | |
| "learning_rate": 4.6019503297331736e-05, | |
| "loss": 0.0401, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 5.54054054054054, | |
| "grad_norm": 0.2635497748851776, | |
| "learning_rate": 4.5277530093374734e-05, | |
| "loss": 0.0358, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 5.585585585585585, | |
| "grad_norm": 0.26532334089279175, | |
| "learning_rate": 4.4536604760256123e-05, | |
| "loss": 0.0406, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 5.63063063063063, | |
| "grad_norm": 0.31581610441207886, | |
| "learning_rate": 4.379689170221043e-05, | |
| "loss": 0.0391, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.675675675675675, | |
| "grad_norm": 0.31361261010169983, | |
| "learning_rate": 4.3058555054479924e-05, | |
| "loss": 0.0424, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 5.7207207207207205, | |
| "grad_norm": 0.25601786375045776, | |
| "learning_rate": 4.232175864689464e-05, | |
| "loss": 0.0392, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 5.7657657657657655, | |
| "grad_norm": 0.2870907485485077, | |
| "learning_rate": 4.158666596752004e-05, | |
| "loss": 0.0347, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 5.8108108108108105, | |
| "grad_norm": 0.27915236353874207, | |
| "learning_rate": 4.085344012638067e-05, | |
| "loss": 0.0351, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 5.8558558558558556, | |
| "grad_norm": 0.29854437708854675, | |
| "learning_rate": 4.01222438192675e-05, | |
| "loss": 0.0434, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.900900900900901, | |
| "grad_norm": 0.34385260939598083, | |
| "learning_rate": 3.939323929163738e-05, | |
| "loss": 0.0352, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 5.945945945945946, | |
| "grad_norm": 0.3390043377876282, | |
| "learning_rate": 3.866658830261224e-05, | |
| "loss": 0.0343, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 5.990990990990991, | |
| "grad_norm": 0.3802526593208313, | |
| "learning_rate": 3.794245208908639e-05, | |
| "loss": 0.0403, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 6.036036036036036, | |
| "grad_norm": 0.26724833250045776, | |
| "learning_rate": 3.722099132994949e-05, | |
| "loss": 0.037, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 6.081081081081081, | |
| "grad_norm": 0.2882586717605591, | |
| "learning_rate": 3.650236611043355e-05, | |
| "loss": 0.0307, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 6.126126126126126, | |
| "grad_norm": 0.24341371655464172, | |
| "learning_rate": 3.578673588659145e-05, | |
| "loss": 0.0318, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 6.171171171171171, | |
| "grad_norm": 0.28189948201179504, | |
| "learning_rate": 3.5074259449915284e-05, | |
| "loss": 0.0306, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 6.216216216216216, | |
| "grad_norm": 0.2897877097129822, | |
| "learning_rate": 3.436509489210189e-05, | |
| "loss": 0.0311, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 6.261261261261261, | |
| "grad_norm": 0.18506336212158203, | |
| "learning_rate": 3.365939956997399e-05, | |
| "loss": 0.0305, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 6.306306306306306, | |
| "grad_norm": 0.21617922186851501, | |
| "learning_rate": 3.2957330070564085e-05, | |
| "loss": 0.0332, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.351351351351352, | |
| "grad_norm": 0.26593586802482605, | |
| "learning_rate": 3.225904217636939e-05, | |
| "loss": 0.0329, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 6.396396396396397, | |
| "grad_norm": 0.24474456906318665, | |
| "learning_rate": 3.1564690830785106e-05, | |
| "loss": 0.0293, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 6.441441441441442, | |
| "grad_norm": 0.21675890684127808, | |
| "learning_rate": 3.0874430103724015e-05, | |
| "loss": 0.033, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 6.486486486486487, | |
| "grad_norm": 0.31768059730529785, | |
| "learning_rate": 3.0188413157429828e-05, | |
| "loss": 0.0303, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 6.531531531531532, | |
| "grad_norm": 0.2665558159351349, | |
| "learning_rate": 2.9506792212491986e-05, | |
| "loss": 0.0326, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 6.576576576576577, | |
| "grad_norm": 0.3153150975704193, | |
| "learning_rate": 2.8829718514069265e-05, | |
| "loss": 0.0322, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 6.621621621621622, | |
| "grad_norm": 0.32425829768180847, | |
| "learning_rate": 2.815734229833007e-05, | |
| "loss": 0.032, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 0.2635970413684845, | |
| "learning_rate": 2.748981275911633e-05, | |
| "loss": 0.0307, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 6.711711711711712, | |
| "grad_norm": 0.41540664434432983, | |
| "learning_rate": 2.6827278014838953e-05, | |
| "loss": 0.0336, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 6.756756756756757, | |
| "grad_norm": 0.2897012233734131, | |
| "learning_rate": 2.616988507561161e-05, | |
| "loss": 0.0301, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.801801801801802, | |
| "grad_norm": 0.2893607020378113, | |
| "learning_rate": 2.5517779810630728e-05, | |
| "loss": 0.0331, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 6.846846846846847, | |
| "grad_norm": 0.5410453081130981, | |
| "learning_rate": 2.4871106915808434e-05, | |
| "loss": 0.0329, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 6.891891891891892, | |
| "grad_norm": 0.28260162472724915, | |
| "learning_rate": 2.4230009881666022e-05, | |
| "loss": 0.0353, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 6.936936936936937, | |
| "grad_norm": 0.178074911236763, | |
| "learning_rate": 2.359463096149461e-05, | |
| "loss": 0.0308, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 6.981981981981982, | |
| "grad_norm": 0.19077138602733612, | |
| "learning_rate": 2.2965111139790697e-05, | |
| "loss": 0.0296, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 7.027027027027027, | |
| "grad_norm": 0.29912513494491577, | |
| "learning_rate": 2.234159010097287e-05, | |
| "loss": 0.0318, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 7.072072072072072, | |
| "grad_norm": 0.3348890542984009, | |
| "learning_rate": 2.1724206198387294e-05, | |
| "loss": 0.0306, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 7.117117117117117, | |
| "grad_norm": 0.1999453604221344, | |
| "learning_rate": 2.1113096423608358e-05, | |
| "loss": 0.0311, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 7.162162162162162, | |
| "grad_norm": 0.27420762181282043, | |
| "learning_rate": 2.050839637604165e-05, | |
| "loss": 0.029, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 7.207207207207207, | |
| "grad_norm": 0.18965670466423035, | |
| "learning_rate": 1.991024023283562e-05, | |
| "loss": 0.0309, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 7.252252252252252, | |
| "grad_norm": 0.23550300300121307, | |
| "learning_rate": 1.9318760719109054e-05, | |
| "loss": 0.031, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 7.297297297297297, | |
| "grad_norm": 0.2695910334587097, | |
| "learning_rate": 1.8734089078500565e-05, | |
| "loss": 0.026, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 7.342342342342342, | |
| "grad_norm": 0.2515803277492523, | |
| "learning_rate": 1.815635504404701e-05, | |
| "loss": 0.0276, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 7.387387387387387, | |
| "grad_norm": 0.19094723463058472, | |
| "learning_rate": 1.7585686809396822e-05, | |
| "loss": 0.0294, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 7.4324324324324325, | |
| "grad_norm": 0.26856082677841187, | |
| "learning_rate": 1.702221100036515e-05, | |
| "loss": 0.0268, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 7.4774774774774775, | |
| "grad_norm": 0.2571709454059601, | |
| "learning_rate": 1.6466052646836832e-05, | |
| "loss": 0.0289, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 7.5225225225225225, | |
| "grad_norm": 0.2961537837982178, | |
| "learning_rate": 1.5917335155023367e-05, | |
| "loss": 0.0301, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 7.5675675675675675, | |
| "grad_norm": 0.18612734973430634, | |
| "learning_rate": 1.5376180280080333e-05, | |
| "loss": 0.027, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 7.612612612612613, | |
| "grad_norm": 0.2395772635936737, | |
| "learning_rate": 1.4842708099091047e-05, | |
| "loss": 0.0269, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 7.657657657657658, | |
| "grad_norm": 0.2048547863960266, | |
| "learning_rate": 1.4317036984422671e-05, | |
| "loss": 0.0257, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.702702702702703, | |
| "grad_norm": 0.1941969245672226, | |
| "learning_rate": 1.3799283577460431e-05, | |
| "loss": 0.0254, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 7.747747747747748, | |
| "grad_norm": 0.3675873279571533, | |
| "learning_rate": 1.328956276272606e-05, | |
| "loss": 0.026, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 7.792792792792793, | |
| "grad_norm": 0.3239746391773224, | |
| "learning_rate": 1.2787987642386007e-05, | |
| "loss": 0.0245, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 7.837837837837838, | |
| "grad_norm": 0.2818147838115692, | |
| "learning_rate": 1.2294669511155193e-05, | |
| "loss": 0.0348, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 7.882882882882883, | |
| "grad_norm": 0.23256780207157135, | |
| "learning_rate": 1.1809717831601697e-05, | |
| "loss": 0.0299, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.927927927927928, | |
| "grad_norm": 0.46159571409225464, | |
| "learning_rate": 1.1333240209858159e-05, | |
| "loss": 0.0299, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 7.972972972972973, | |
| "grad_norm": 0.19868823885917664, | |
| "learning_rate": 1.0865342371744924e-05, | |
| "loss": 0.0269, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 8.018018018018019, | |
| "grad_norm": 0.29607120156288147, | |
| "learning_rate": 1.0406128139310533e-05, | |
| "loss": 0.0266, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 8.063063063063064, | |
| "grad_norm": 0.1842232197523117, | |
| "learning_rate": 9.955699407794594e-06, | |
| "loss": 0.0252, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 8.108108108108109, | |
| "grad_norm": 0.1922302395105362, | |
| "learning_rate": 9.514156123018258e-06, | |
| "loss": 0.0243, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.153153153153154, | |
| "grad_norm": 0.21968339383602142, | |
| "learning_rate": 9.081596259207109e-06, | |
| "loss": 0.0266, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 8.198198198198199, | |
| "grad_norm": 0.24987168610095978, | |
| "learning_rate": 8.658115797251676e-06, | |
| "loss": 0.0255, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 8.243243243243244, | |
| "grad_norm": 0.30064335465431213, | |
| "learning_rate": 8.243808703410177e-06, | |
| "loss": 0.0245, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 8.288288288288289, | |
| "grad_norm": 0.26207128167152405, | |
| "learning_rate": 7.838766908458339e-06, | |
| "loss": 0.0238, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 0.23758377134799957, | |
| "learning_rate": 7.443080287290782e-06, | |
| "loss": 0.0265, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 8.378378378378379, | |
| "grad_norm": 0.30203232169151306, | |
| "learning_rate": 7.0568366389786975e-06, | |
| "loss": 0.0283, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 8.423423423423424, | |
| "grad_norm": 0.23233330249786377, | |
| "learning_rate": 6.680121667288025e-06, | |
| "loss": 0.0249, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 8.468468468468469, | |
| "grad_norm": 0.3445972204208374, | |
| "learning_rate": 6.3130189616626474e-06, | |
| "loss": 0.0236, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 8.513513513513514, | |
| "grad_norm": 0.31303051114082336, | |
| "learning_rate": 5.955609978676652e-06, | |
| "loss": 0.0243, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 8.558558558558559, | |
| "grad_norm": 0.2929190397262573, | |
| "learning_rate": 5.607974023959978e-06, | |
| "loss": 0.0314, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.603603603603604, | |
| "grad_norm": 0.4855507016181946, | |
| "learning_rate": 5.270188234601142e-06, | |
| "loss": 0.0267, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 8.64864864864865, | |
| "grad_norm": 0.3386547863483429, | |
| "learning_rate": 4.942327562031357e-06, | |
| "loss": 0.0247, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 8.693693693693694, | |
| "grad_norm": 0.18032731115818024, | |
| "learning_rate": 4.624464755393459e-06, | |
| "loss": 0.0241, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 8.73873873873874, | |
| "grad_norm": 0.2196054905653, | |
| "learning_rate": 4.316670345399626e-06, | |
| "loss": 0.029, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 8.783783783783784, | |
| "grad_norm": 0.1823672503232956, | |
| "learning_rate": 4.019012628681234e-06, | |
| "loss": 0.0228, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 8.82882882882883, | |
| "grad_norm": 0.25816360116004944, | |
| "learning_rate": 3.731557652634543e-06, | |
| "loss": 0.0239, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 8.873873873873874, | |
| "grad_norm": 0.21811510622501373, | |
| "learning_rate": 3.454369200765356e-06, | |
| "loss": 0.0244, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 8.91891891891892, | |
| "grad_norm": 0.19826571643352509, | |
| "learning_rate": 3.1875087785361137e-06, | |
| "loss": 0.0236, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 8.963963963963964, | |
| "grad_norm": 0.1923084408044815, | |
| "learning_rate": 2.931035599718396e-06, | |
| "loss": 0.0235, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 9.00900900900901, | |
| "grad_norm": 0.16076907515525818, | |
| "learning_rate": 2.6850065732539842e-06, | |
| "loss": 0.0262, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.054054054054054, | |
| "grad_norm": 0.2307383120059967, | |
| "learning_rate": 2.449476290627273e-06, | |
| "loss": 0.0245, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 9.0990990990991, | |
| "grad_norm": 0.21985507011413574, | |
| "learning_rate": 2.2244970137519583e-06, | |
| "loss": 0.0266, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 9.144144144144144, | |
| "grad_norm": 0.26605159044265747, | |
| "learning_rate": 2.010118663374627e-06, | |
| "loss": 0.0225, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 9.18918918918919, | |
| "grad_norm": 0.3474940359592438, | |
| "learning_rate": 1.8063888079978331e-06, | |
| "loss": 0.0255, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 9.234234234234235, | |
| "grad_norm": 0.1781226396560669, | |
| "learning_rate": 1.6133526533250565e-06, | |
| "loss": 0.0274, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 9.27927927927928, | |
| "grad_norm": 0.2931009829044342, | |
| "learning_rate": 1.4310530322300453e-06, | |
| "loss": 0.0253, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 9.324324324324325, | |
| "grad_norm": 0.21503065526485443, | |
| "learning_rate": 1.2595303952525672e-06, | |
| "loss": 0.0223, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 9.36936936936937, | |
| "grad_norm": 0.23624898493289948, | |
| "learning_rate": 1.0988228016228508e-06, | |
| "loss": 0.0252, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 9.414414414414415, | |
| "grad_norm": 0.2636685073375702, | |
| "learning_rate": 9.48965910816596e-07, | |
| "loss": 0.0302, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 9.45945945945946, | |
| "grad_norm": 0.23757557570934296, | |
| "learning_rate": 8.099929746424706e-07, | |
| "loss": 0.0248, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 9.504504504504505, | |
| "grad_norm": 0.3245598077774048, | |
| "learning_rate": 6.819348298638839e-07, | |
| "loss": 0.0219, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 9.54954954954955, | |
| "grad_norm": 0.260775089263916, | |
| "learning_rate": 5.648198913565494e-07, | |
| "loss": 0.0198, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 9.594594594594595, | |
| "grad_norm": 0.18040095269680023, | |
| "learning_rate": 4.5867414580355593e-07, | |
| "loss": 0.0263, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 9.63963963963964, | |
| "grad_norm": 0.17358094453811646, | |
| "learning_rate": 3.635211459291188e-07, | |
| "loss": 0.0236, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 9.684684684684685, | |
| "grad_norm": 0.1593407541513443, | |
| "learning_rate": 2.793820052725049e-07, | |
| "loss": 0.0233, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 9.72972972972973, | |
| "grad_norm": 0.24278421700000763, | |
| "learning_rate": 2.06275393503097e-07, | |
| "loss": 0.0201, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 9.774774774774775, | |
| "grad_norm": 0.1488327831029892, | |
| "learning_rate": 1.4421753227780722e-07, | |
| "loss": 0.0234, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 9.81981981981982, | |
| "grad_norm": 0.2808060050010681, | |
| "learning_rate": 9.32221916416176e-08, | |
| "loss": 0.0248, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 9.864864864864865, | |
| "grad_norm": 0.2229873239994049, | |
| "learning_rate": 5.330068697215751e-08, | |
| "loss": 0.0259, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 9.90990990990991, | |
| "grad_norm": 0.2598866820335388, | |
| "learning_rate": 2.4461876468934163e-08, | |
| "loss": 0.0273, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.954954954954955, | |
| "grad_norm": 0.29245129227638245, | |
| "learning_rate": 6.712159187766131e-09, | |
| "loss": 0.0245, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.2520624101161957, | |
| "learning_rate": 5.547362090241315e-11, | |
| "loss": 0.0271, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 2220, | |
| "total_flos": 0.0, | |
| "train_loss": 0.055370314842140354, | |
| "train_runtime": 2269.5641, | |
| "train_samples_per_second": 47.895, | |
| "train_steps_per_second": 0.978 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2220, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 20000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 49, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |