| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1875, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 1.5132097727696663, | |
| "learning_rate": 9.999298177883903e-05, | |
| "loss": 1.6713, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 1.0288849642854194, | |
| "learning_rate": 9.997192908557323e-05, | |
| "loss": 0.9562, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 2.628046163977058, | |
| "learning_rate": 9.993684783030088e-05, | |
| "loss": 0.7998, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 1.0977335920429718, | |
| "learning_rate": 9.988774786134234e-05, | |
| "loss": 0.7019, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 1.143511217142594, | |
| "learning_rate": 9.982464296247522e-05, | |
| "loss": 0.6965, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 1.1886880844299708, | |
| "learning_rate": 9.974755084906502e-05, | |
| "loss": 0.6417, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 1.3088809099781331, | |
| "learning_rate": 9.965649316309178e-05, | |
| "loss": 0.6388, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 2.0299623866899035, | |
| "learning_rate": 9.955149546707465e-05, | |
| "loss": 0.6665, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 1.5251129511711958, | |
| "learning_rate": 9.94325872368957e-05, | |
| "loss": 0.6014, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.307242452061001, | |
| "learning_rate": 9.929980185352526e-05, | |
| "loss": 0.5971, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 1.0674735946460283, | |
| "learning_rate": 9.915317659365077e-05, | |
| "loss": 0.6206, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 1.3671834842056045, | |
| "learning_rate": 9.899275261921234e-05, | |
| "loss": 0.5597, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 1.3041158193336297, | |
| "learning_rate": 9.881857496584726e-05, | |
| "loss": 0.5774, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 1.2739151782924372, | |
| "learning_rate": 9.863069253024719e-05, | |
| "loss": 0.5656, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.1009566652519438, | |
| "learning_rate": 9.842915805643155e-05, | |
| "loss": 0.5724, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.2004741768824123, | |
| "learning_rate": 9.821402812094073e-05, | |
| "loss": 0.5796, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 1.5111413381395908, | |
| "learning_rate": 9.798536311695334e-05, | |
| "loss": 0.5556, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 1.0331976579035735, | |
| "learning_rate": 9.774322723733216e-05, | |
| "loss": 0.5451, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 1.1745428366022237, | |
| "learning_rate": 9.748768845660334e-05, | |
| "loss": 0.5605, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.067749681555933, | |
| "learning_rate": 9.721881851187406e-05, | |
| "loss": 0.5644, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 1.1773627454789788, | |
| "learning_rate": 9.693669288269372e-05, | |
| "loss": 0.5658, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 1.4049555802409854, | |
| "learning_rate": 9.664139076986473e-05, | |
| "loss": 0.5483, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 1.436137944296166, | |
| "learning_rate": 9.63329950732086e-05, | |
| "loss": 0.5376, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.2686595633739197, | |
| "learning_rate": 9.601159236829352e-05, | |
| "loss": 0.5402, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.4062419829256017, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.5552, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 0.9610102580305002, | |
| "learning_rate": 9.533013046784189e-05, | |
| "loss": 0.5118, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 1.2008075396073077, | |
| "learning_rate": 9.497026257831855e-05, | |
| "loss": 0.5479, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.116478910009909, | |
| "learning_rate": 9.459777023885755e-05, | |
| "loss": 0.4773, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 1.2287813088079407, | |
| "learning_rate": 9.421275801880362e-05, | |
| "loss": 0.5149, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.3153026196754682, | |
| "learning_rate": 9.381533400219318e-05, | |
| "loss": 0.5209, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 1.6158648577050707, | |
| "learning_rate": 9.340560975741197e-05, | |
| "loss": 0.5065, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.1601862022387133, | |
| "learning_rate": 9.298370030587456e-05, | |
| "loss": 0.515, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 1.2591523341167126, | |
| "learning_rate": 9.254972408973461e-05, | |
| "loss": 0.5375, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 1.0361296810068354, | |
| "learning_rate": 9.210380293863462e-05, | |
| "loss": 0.5085, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.0654093627590127, | |
| "learning_rate": 9.164606203550497e-05, | |
| "loss": 0.5137, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.2697508248750302, | |
| "learning_rate": 9.117662988142138e-05, | |
| "loss": 0.4906, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 1.2602467399423862, | |
| "learning_rate": 9.069563825953092e-05, | |
| "loss": 0.4966, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 2.0044411177134918, | |
| "learning_rate": 9.020322219805674e-05, | |
| "loss": 0.5063, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 1.148187026295081, | |
| "learning_rate": 8.969951993239177e-05, | |
| "loss": 0.4864, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.0529856886801567, | |
| "learning_rate": 8.9184672866292e-05, | |
| "loss": 0.4632, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 1.1830984654167522, | |
| "learning_rate": 8.865882553218037e-05, | |
| "loss": 0.5097, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 1.1495869848318996, | |
| "learning_rate": 8.81221255505724e-05, | |
| "loss": 0.4795, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 1.1235900402260404, | |
| "learning_rate": 8.757472358863481e-05, | |
| "loss": 0.5006, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.2883600054948787, | |
| "learning_rate": 8.701677331788891e-05, | |
| "loss": 0.4822, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.0725324329031838, | |
| "learning_rate": 8.644843137107059e-05, | |
| "loss": 0.4856, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 2.9167440508075955, | |
| "learning_rate": 8.586985729815894e-05, | |
| "loss": 0.4648, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 0.9522185207229502, | |
| "learning_rate": 8.528121352158604e-05, | |
| "loss": 0.488, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 1.5907707346034905, | |
| "learning_rate": 8.468266529064025e-05, | |
| "loss": 0.5039, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 1.1370580239916313, | |
| "learning_rate": 8.4074380635076e-05, | |
| "loss": 0.4737, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.9782419363730104, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.5296, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 1.1919631254125505, | |
| "learning_rate": 8.282928778764783e-05, | |
| "loss": 0.4688, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.2994447479210165, | |
| "learning_rate": 8.21928291292627e-05, | |
| "loss": 0.4675, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 0.9689627531529147, | |
| "learning_rate": 8.154733301509248e-05, | |
| "loss": 0.4408, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 1.1649720718750445, | |
| "learning_rate": 8.089298065451672e-05, | |
| "loss": 0.4608, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.3438713492048824, | |
| "learning_rate": 8.022995574311876e-05, | |
| "loss": 0.4476, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 1.0737706296113363, | |
| "learning_rate": 7.95584444111171e-05, | |
| "loss": 0.5027, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 1.3469637742497171, | |
| "learning_rate": 7.887863517111338e-05, | |
| "loss": 0.4929, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 1.1251117260635928, | |
| "learning_rate": 7.819071886517134e-05, | |
| "loss": 0.4739, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 1.0579726403654415, | |
| "learning_rate": 7.7494888611242e-05, | |
| "loss": 0.4529, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.0498899044694738, | |
| "learning_rate": 7.679133974894983e-05, | |
| "loss": 0.4619, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 1.2353618604203835, | |
| "learning_rate": 7.60802697847554e-05, | |
| "loss": 0.462, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 1.4618575476184537, | |
| "learning_rate": 7.536187833650947e-05, | |
| "loss": 0.4478, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 0.9313196650505406, | |
| "learning_rate": 7.463636707741458e-05, | |
| "loss": 0.4304, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 1.1239041555920188, | |
| "learning_rate": 7.390393967940962e-05, | |
| "loss": 0.4138, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 1.8736342455754906, | |
| "learning_rate": 7.316480175599309e-05, | |
| "loss": 0.3909, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 1.3075877434881118, | |
| "learning_rate": 7.241916080450163e-05, | |
| "loss": 0.3873, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.072, | |
| "grad_norm": 1.1366815390360774, | |
| "learning_rate": 7.166722614785937e-05, | |
| "loss": 0.3744, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 1.1482160099998533, | |
| "learning_rate": 7.090920887581506e-05, | |
| "loss": 0.3641, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 1.4903122013208903, | |
| "learning_rate": 7.014532178568314e-05, | |
| "loss": 0.3708, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 1.2283466862099195, | |
| "learning_rate": 6.937577932260515e-05, | |
| "loss": 0.4147, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1360000000000001, | |
| "grad_norm": 1.044360506141548, | |
| "learning_rate": 6.860079751934908e-05, | |
| "loss": 0.3698, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 1.9659986049038847, | |
| "learning_rate": 6.782059393566253e-05, | |
| "loss": 0.3768, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.168, | |
| "grad_norm": 1.1559700548705432, | |
| "learning_rate": 6.70353875971976e-05, | |
| "loss": 0.3713, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.184, | |
| "grad_norm": 1.3638599004402776, | |
| "learning_rate": 6.624539893402382e-05, | |
| "loss": 0.376, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 1.2823520816693201, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.3762, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 1.1889832443284072, | |
| "learning_rate": 6.465196300425287e-05, | |
| "loss": 0.3891, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.232, | |
| "grad_norm": 1.128194765929284, | |
| "learning_rate": 6.384896306108612e-05, | |
| "loss": 0.3772, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 1.4491197750693638, | |
| "learning_rate": 6.304207531449486e-05, | |
| "loss": 0.3678, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.264, | |
| "grad_norm": 1.5686675701147237, | |
| "learning_rate": 6.223152628114537e-05, | |
| "loss": 0.3648, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.0902715474500886, | |
| "learning_rate": 6.141754350553279e-05, | |
| "loss": 0.3909, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 1.6337999430835446, | |
| "learning_rate": 6.0600355496102745e-05, | |
| "loss": 0.383, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.312, | |
| "grad_norm": 1.321659745725057, | |
| "learning_rate": 5.9780191661102415e-05, | |
| "loss": 0.3802, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.328, | |
| "grad_norm": 1.6867441595334431, | |
| "learning_rate": 5.8957282244179124e-05, | |
| "loss": 0.423, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 1.2441664958096805, | |
| "learning_rate": 5.813185825974419e-05, | |
| "loss": 0.3641, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 1.6422801866133996, | |
| "learning_rate": 5.730415142812059e-05, | |
| "loss": 0.3519, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.376, | |
| "grad_norm": 1.7154734289625684, | |
| "learning_rate": 5.6474394110492344e-05, | |
| "loss": 0.3732, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 1.3529634625288804, | |
| "learning_rate": 5.564281924367408e-05, | |
| "loss": 0.3954, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 1.4693831158149995, | |
| "learning_rate": 5.480966027471889e-05, | |
| "loss": 0.3669, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.424, | |
| "grad_norm": 1.5051265025599196, | |
| "learning_rate": 5.3975151095382995e-05, | |
| "loss": 0.3994, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 1.2893604881333178, | |
| "learning_rate": 5.313952597646568e-05, | |
| "loss": 0.3904, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.456, | |
| "grad_norm": 1.5665270237662434, | |
| "learning_rate": 5.230301950204262e-05, | |
| "loss": 0.3821, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 1.1847281856196625, | |
| "learning_rate": 5.1465866503611426e-05, | |
| "loss": 0.3713, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 1.6027343131881702, | |
| "learning_rate": 5.062830199416764e-05, | |
| "loss": 0.3785, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.504, | |
| "grad_norm": 1.1745245908384976, | |
| "learning_rate": 4.979056110222981e-05, | |
| "loss": 0.3851, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 1.2252730329400867, | |
| "learning_rate": 4.895287900583216e-05, | |
| "loss": 0.3991, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 1.9189789923421308, | |
| "learning_rate": 4.811549086650327e-05, | |
| "loss": 0.3726, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.552, | |
| "grad_norm": 1.9319344659060065, | |
| "learning_rate": 4.7278631763249554e-05, | |
| "loss": 0.3703, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.568, | |
| "grad_norm": 1.3155775136973578, | |
| "learning_rate": 4.6442536626561675e-05, | |
| "loss": 0.3462, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 1.3059644425743502, | |
| "learning_rate": 4.560744017246284e-05, | |
| "loss": 0.3486, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.1918137712202796, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.3384, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.616, | |
| "grad_norm": 1.3051153832513638, | |
| "learning_rate": 4.394118070851749e-05, | |
| "loss": 0.3976, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 2.1135344091319266, | |
| "learning_rate": 4.31104854657681e-05, | |
| "loss": 0.3607, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.6480000000000001, | |
| "grad_norm": 1.3948892979651148, | |
| "learning_rate": 4.228172430848644e-05, | |
| "loss": 0.371, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 1.3934093132103698, | |
| "learning_rate": 4.1455129893836174e-05, | |
| "loss": 0.3627, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 2.099040035068272, | |
| "learning_rate": 4.063093427071376e-05, | |
| "loss": 0.3725, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.696, | |
| "grad_norm": 1.1171081747364493, | |
| "learning_rate": 3.9809368814605766e-05, | |
| "loss": 0.3466, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.712, | |
| "grad_norm": 1.2557303206572503, | |
| "learning_rate": 3.899066416263493e-05, | |
| "loss": 0.3629, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 1.2612524984380895, | |
| "learning_rate": 3.817505014881378e-05, | |
| "loss": 0.3941, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.744, | |
| "grad_norm": 1.4536732919602855, | |
| "learning_rate": 3.736275573952354e-05, | |
| "loss": 0.3362, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 1.5347325608557592, | |
| "learning_rate": 3.655400896923672e-05, | |
| "loss": 0.3197, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 1.5306810075691544, | |
| "learning_rate": 3.5749036876501194e-05, | |
| "loss": 0.3447, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 1.2300188758500825, | |
| "learning_rate": 3.494806544020398e-05, | |
| "loss": 0.3535, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.808, | |
| "grad_norm": 1.4275688891160825, | |
| "learning_rate": 3.4151319516132416e-05, | |
| "loss": 0.3447, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 1.397634927032295, | |
| "learning_rate": 3.335902277385067e-05, | |
| "loss": 0.338, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 1.8570000554042922, | |
| "learning_rate": 3.257139763390925e-05, | |
| "loss": 0.3915, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 1.0694824076607796, | |
| "learning_rate": 3.178866520540509e-05, | |
| "loss": 0.332, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 1.9227266591369816, | |
| "learning_rate": 3.101104522390995e-05, | |
| "loss": 0.3734, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.888, | |
| "grad_norm": 1.4206748536497713, | |
| "learning_rate": 3.023875598978419e-05, | |
| "loss": 0.3409, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.904, | |
| "grad_norm": 1.4359353896901201, | |
| "learning_rate": 2.9472014306893603e-05, | |
| "loss": 0.3333, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.2517700720828175, | |
| "learning_rate": 2.8711035421746367e-05, | |
| "loss": 0.369, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.936, | |
| "grad_norm": 1.3801398040671706, | |
| "learning_rate": 2.795603296306708e-05, | |
| "loss": 0.3449, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.952, | |
| "grad_norm": 1.8786886188992349, | |
| "learning_rate": 2.7207218881825014e-05, | |
| "loss": 0.3662, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 1.2874259666120602, | |
| "learning_rate": 2.6464803391733374e-05, | |
| "loss": 0.3773, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 1.1788438157215508, | |
| "learning_rate": 2.5728994910236304e-05, | |
| "loss": 0.3245, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.5631045640027712, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 0.3412, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 1.4435057320180134, | |
| "learning_rate": 2.4278023310924673e-05, | |
| "loss": 0.3027, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 1.690073924846414, | |
| "learning_rate": 2.3563267522693415e-05, | |
| "loss": 0.2745, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 1.571802508559373, | |
| "learning_rate": 2.2855933287874138e-05, | |
| "loss": 0.2686, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 1.2829706151133036, | |
| "learning_rate": 2.215621917559062e-05, | |
| "loss": 0.2628, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.5440500775973827, | |
| "learning_rate": 2.1464321615778422e-05, | |
| "loss": 0.274, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.096, | |
| "grad_norm": 2.0539984960007627, | |
| "learning_rate": 2.07804348440414e-05, | |
| "loss": 0.2711, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 1.6138729128750948, | |
| "learning_rate": 2.0104750847124075e-05, | |
| "loss": 0.2662, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 1.4560623181879653, | |
| "learning_rate": 1.9437459309015427e-05, | |
| "loss": 0.2722, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.144, | |
| "grad_norm": 1.4659611175991962, | |
| "learning_rate": 1.8778747557699224e-05, | |
| "loss": 0.2965, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 1.375382006257737, | |
| "learning_rate": 1.8128800512565513e-05, | |
| "loss": 0.2789, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 1.599179489252925, | |
| "learning_rate": 1.7487800632498545e-05, | |
| "loss": 0.2727, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.192, | |
| "grad_norm": 1.517535594054785, | |
| "learning_rate": 1.685592786465524e-05, | |
| "loss": 0.2579, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 2.1592276480492414, | |
| "learning_rate": 1.6233359593948777e-05, | |
| "loss": 0.2639, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.224, | |
| "grad_norm": 1.7544518130502498, | |
| "learning_rate": 1.5620270593251635e-05, | |
| "loss": 0.2909, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 1.44885168688834, | |
| "learning_rate": 1.5016832974331724e-05, | |
| "loss": 0.2675, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 1.6827124662772142, | |
| "learning_rate": 1.4423216139535734e-05, | |
| "loss": 0.2501, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.2720000000000002, | |
| "grad_norm": 1.8488651881141411, | |
| "learning_rate": 1.3839586734232906e-05, | |
| "loss": 0.28, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.288, | |
| "grad_norm": 1.8044696205561654, | |
| "learning_rate": 1.3266108600032929e-05, | |
| "loss": 0.2946, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 1.5680732030217428, | |
| "learning_rate": 1.2702942728790895e-05, | |
| "loss": 0.2801, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 1.8180785380951152, | |
| "learning_rate": 1.2150247217412186e-05, | |
| "loss": 0.2686, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.336, | |
| "grad_norm": 1.4756615703317884, | |
| "learning_rate": 1.160817722347014e-05, | |
| "loss": 0.2623, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 1.4951599689749675, | |
| "learning_rate": 1.1076884921648834e-05, | |
| "loss": 0.2865, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 1.8073936077115729, | |
| "learning_rate": 1.0556519461023301e-05, | |
| "loss": 0.2677, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.384, | |
| "grad_norm": 1.6787640384880986, | |
| "learning_rate": 1.0047226923189024e-05, | |
| "loss": 0.2821, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.7736081937737844, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 0.2628, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.416, | |
| "grad_norm": 1.8163690827077064, | |
| "learning_rate": 9.06242935969528e-06, | |
| "loss": 0.2648, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 1.914925459991879, | |
| "learning_rate": 8.587200795119793e-06, | |
| "loss": 0.2718, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 1.557572030989793, | |
| "learning_rate": 8.123597997892918e-06, | |
| "loss": 0.2779, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.464, | |
| "grad_norm": 2.4189475486471412, | |
| "learning_rate": 7.671751114693104e-06, | |
| "loss": 0.2522, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 1.5979978703748492, | |
| "learning_rate": 7.2317869919746705e-06, | |
| "loss": 0.2668, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 1.5867221943137244, | |
| "learning_rate": 6.803829140358237e-06, | |
| "loss": 0.2582, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 2.0088219455242866, | |
| "learning_rate": 6.3879976999578154e-06, | |
| "loss": 0.2509, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.528, | |
| "grad_norm": 1.7960925687722837, | |
| "learning_rate": 5.98440940665399e-06, | |
| "loss": 0.2743, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 1.5572986827037758, | |
| "learning_rate": 5.593177559322777e-06, | |
| "loss": 0.2833, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 1.370850993490327, | |
| "learning_rate": 5.214411988029355e-06, | |
| "loss": 0.2693, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.576, | |
| "grad_norm": 1.3820976570272918, | |
| "learning_rate": 4.848219023195644e-06, | |
| "loss": 0.2607, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 1.7956675550522907, | |
| "learning_rate": 4.494701465750217e-06, | |
| "loss": 0.2632, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.608, | |
| "grad_norm": 1.657852197945445, | |
| "learning_rate": 4.153958558269189e-06, | |
| "loss": 0.2573, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 1.6754996668537725, | |
| "learning_rate": 3.826085957115888e-06, | |
| "loss": 0.2411, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 1.3160903872279537, | |
| "learning_rate": 3.511175705587433e-06, | |
| "loss": 0.2601, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.656, | |
| "grad_norm": 1.4417765456611342, | |
| "learning_rate": 3.2093162080754637e-06, | |
| "loss": 0.2832, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.672, | |
| "grad_norm": 1.5418721821515773, | |
| "learning_rate": 2.9205922052484958e-06, | |
| "loss": 0.2725, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 1.3747599439593585, | |
| "learning_rate": 2.6450847502627884e-06, | |
| "loss": 0.2654, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.7039999999999997, | |
| "grad_norm": 1.528133223907777, | |
| "learning_rate": 2.3828711860083674e-06, | |
| "loss": 0.2784, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 2.3538184370373534, | |
| "learning_rate": 2.134025123396638e-06, | |
| "loss": 0.2758, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 1.6198355766213741, | |
| "learning_rate": 1.8986164206957035e-06, | |
| "loss": 0.2533, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 1.8976100233670337, | |
| "learning_rate": 1.6767111639191202e-06, | |
| "loss": 0.268, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.768, | |
| "grad_norm": 1.7822120577723535, | |
| "learning_rate": 1.4683716482736366e-06, | |
| "loss": 0.28, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 2.2850018497482183, | |
| "learning_rate": 1.2736563606711382e-06, | |
| "loss": 0.2797, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.79121827664844, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.2831, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 1.7925072153699964, | |
| "learning_rate": 9.253132783283547e-07, | |
| "loss": 0.2558, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 1.2894709082534288, | |
| "learning_rate": 7.717832735397335e-07, | |
| "loss": 0.2635, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 2.848, | |
| "grad_norm": 1.4531115959909715, | |
| "learning_rate": 6.3207304924498e-07, | |
| "loss": 0.2639, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 2.864, | |
| "grad_norm": 1.9570336366099832, | |
| "learning_rate": 5.062218261342122e-07, | |
| "loss": 0.2532, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.3421965286430086, | |
| "learning_rate": 3.9426493427611177e-07, | |
| "loss": 0.2327, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 2.3252685172502465, | |
| "learning_rate": 2.962338031997691e-07, | |
| "loss": 0.2621, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 2.912, | |
| "grad_norm": 1.7539361758237508, | |
| "learning_rate": 2.1215595307154667e-07, | |
| "loss": 0.256, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 1.417074579458341, | |
| "learning_rate": 1.420549869693033e-07, | |
| "loss": 0.2545, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 1.7016791883476743, | |
| "learning_rate": 8.595058425640013e-08, | |
| "loss": 0.2596, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 1.6335011497462837, | |
| "learning_rate": 4.385849505708084e-08, | |
| "loss": 0.2584, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 1.7816152560274576, | |
| "learning_rate": 1.5790535835003008e-08, | |
| "loss": 0.2752, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 2.992, | |
| "grad_norm": 1.579627845758394, | |
| "learning_rate": 1.7545860759693445e-09, | |
| "loss": 0.2778, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1875, | |
| "total_flos": 299543991877632.0, | |
| "train_loss": 0.3980048195521037, | |
| "train_runtime": 15877.4798, | |
| "train_samples_per_second": 0.945, | |
| "train_steps_per_second": 0.118 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1875, | |
| "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": 299543991877632.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |