| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.1213406873143827, | |
| "eval_steps": 600, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00848536274925753, | |
| "grad_norm": 0.4898678891363344, | |
| "learning_rate": 8.488964346349746e-07, | |
| "loss": 1.8056, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01697072549851506, | |
| "grad_norm": 0.3537473179717183, | |
| "learning_rate": 1.6977928692699491e-06, | |
| "loss": 1.7621, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.025456088247772592, | |
| "grad_norm": 0.28215953004159977, | |
| "learning_rate": 2.546689303904924e-06, | |
| "loss": 1.7571, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03394145099703012, | |
| "grad_norm": 0.27446565146764923, | |
| "learning_rate": 3.3955857385398982e-06, | |
| "loss": 1.7136, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04242681374628765, | |
| "grad_norm": 0.17051549768176558, | |
| "learning_rate": 4.244482173174873e-06, | |
| "loss": 1.6767, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.050912176495545185, | |
| "grad_norm": 0.17763882467320422, | |
| "learning_rate": 5.093378607809848e-06, | |
| "loss": 1.6371, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05939753924480271, | |
| "grad_norm": 0.14311462596290048, | |
| "learning_rate": 5.942275042444822e-06, | |
| "loss": 1.6324, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06788290199406025, | |
| "grad_norm": 0.1659540846071645, | |
| "learning_rate": 6.7911714770797965e-06, | |
| "loss": 1.6062, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07636826474331777, | |
| "grad_norm": 0.20064072815620043, | |
| "learning_rate": 7.640067911714771e-06, | |
| "loss": 1.5832, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0848536274925753, | |
| "grad_norm": 0.2179045681711979, | |
| "learning_rate": 8.488964346349745e-06, | |
| "loss": 1.5898, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09333899024183284, | |
| "grad_norm": 0.23866012053128668, | |
| "learning_rate": 9.337860780984721e-06, | |
| "loss": 1.5924, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10182435299109037, | |
| "grad_norm": 0.18578051776430282, | |
| "learning_rate": 1.0186757215619695e-05, | |
| "loss": 1.5877, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1103097157403479, | |
| "grad_norm": 0.2216509707409362, | |
| "learning_rate": 1.103565365025467e-05, | |
| "loss": 1.5947, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11879507848960542, | |
| "grad_norm": 0.20427142255694086, | |
| "learning_rate": 1.1884550084889643e-05, | |
| "loss": 1.5841, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12728044123886295, | |
| "grad_norm": 0.1765851415675038, | |
| "learning_rate": 1.2733446519524619e-05, | |
| "loss": 1.5878, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1357658039881205, | |
| "grad_norm": 0.1769355117060811, | |
| "learning_rate": 1.3582342954159593e-05, | |
| "loss": 1.5795, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.14425116673737803, | |
| "grad_norm": 0.1617675663096666, | |
| "learning_rate": 1.4431239388794569e-05, | |
| "loss": 1.5549, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.15273652948663555, | |
| "grad_norm": 0.17302259072151574, | |
| "learning_rate": 1.5280135823429543e-05, | |
| "loss": 1.5808, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1612218922358931, | |
| "grad_norm": 0.16876039012432806, | |
| "learning_rate": 1.6129032258064517e-05, | |
| "loss": 1.5676, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.1697072549851506, | |
| "grad_norm": 0.19627360154037596, | |
| "learning_rate": 1.697792869269949e-05, | |
| "loss": 1.5598, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.17819261773440814, | |
| "grad_norm": 0.16078510362361015, | |
| "learning_rate": 1.7826825127334465e-05, | |
| "loss": 1.5667, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.18667798048366568, | |
| "grad_norm": 0.16044786518959703, | |
| "learning_rate": 1.8675721561969442e-05, | |
| "loss": 1.5815, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.1951633432329232, | |
| "grad_norm": 0.15656958873834717, | |
| "learning_rate": 1.9524617996604416e-05, | |
| "loss": 1.5576, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.20364870598218074, | |
| "grad_norm": 0.1687290471357602, | |
| "learning_rate": 2.037351443123939e-05, | |
| "loss": 1.5453, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.21213406873143828, | |
| "grad_norm": 0.1519017348276184, | |
| "learning_rate": 2.1222410865874364e-05, | |
| "loss": 1.5554, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2206194314806958, | |
| "grad_norm": 0.15761892005160086, | |
| "learning_rate": 2.207130730050934e-05, | |
| "loss": 1.5494, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.22910479422995333, | |
| "grad_norm": 0.16857088482977495, | |
| "learning_rate": 2.2920203735144312e-05, | |
| "loss": 1.5794, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.23759015697921085, | |
| "grad_norm": 0.1678705209913503, | |
| "learning_rate": 2.3769100169779286e-05, | |
| "loss": 1.5373, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.2460755197284684, | |
| "grad_norm": 0.14812649566587394, | |
| "learning_rate": 2.461799660441426e-05, | |
| "loss": 1.5504, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2545608824777259, | |
| "grad_norm": 0.17651916734325857, | |
| "learning_rate": 2.5466893039049238e-05, | |
| "loss": 1.5607, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.26304624522698344, | |
| "grad_norm": 0.14883055338507856, | |
| "learning_rate": 2.6315789473684212e-05, | |
| "loss": 1.5311, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.271531607976241, | |
| "grad_norm": 0.15787522753231265, | |
| "learning_rate": 2.7164685908319186e-05, | |
| "loss": 1.5656, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2800169707254985, | |
| "grad_norm": 0.1625232940237689, | |
| "learning_rate": 2.801358234295416e-05, | |
| "loss": 1.5686, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.28850233347475607, | |
| "grad_norm": 0.18505951289343867, | |
| "learning_rate": 2.8862478777589137e-05, | |
| "loss": 1.5474, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.29698769622401355, | |
| "grad_norm": 0.13785772316349984, | |
| "learning_rate": 2.9711375212224108e-05, | |
| "loss": 1.5696, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3054730589732711, | |
| "grad_norm": 0.13531274658248552, | |
| "learning_rate": 3.0560271646859086e-05, | |
| "loss": 1.5551, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.31395842172252864, | |
| "grad_norm": 0.1366381415368909, | |
| "learning_rate": 3.140916808149406e-05, | |
| "loss": 1.524, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3224437844717862, | |
| "grad_norm": 0.14587220569353926, | |
| "learning_rate": 3.2258064516129034e-05, | |
| "loss": 1.5515, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3309291472210437, | |
| "grad_norm": 0.13336349383744864, | |
| "learning_rate": 3.310696095076401e-05, | |
| "loss": 1.5457, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.3394145099703012, | |
| "grad_norm": 0.1772016947970983, | |
| "learning_rate": 3.395585738539898e-05, | |
| "loss": 1.5582, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.34789987271955874, | |
| "grad_norm": 0.13819420575084573, | |
| "learning_rate": 3.4804753820033956e-05, | |
| "loss": 1.5326, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3563852354688163, | |
| "grad_norm": 0.12729862167862188, | |
| "learning_rate": 3.565365025466893e-05, | |
| "loss": 1.5387, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.3648705982180738, | |
| "grad_norm": 0.11777082851399363, | |
| "learning_rate": 3.6502546689303904e-05, | |
| "loss": 1.5587, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.37335596096733137, | |
| "grad_norm": 0.15372268131323022, | |
| "learning_rate": 3.7351443123938885e-05, | |
| "loss": 1.5362, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.3818413237165889, | |
| "grad_norm": 0.12616185572252248, | |
| "learning_rate": 3.820033955857386e-05, | |
| "loss": 1.5548, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.3903266864658464, | |
| "grad_norm": 0.1311200786303391, | |
| "learning_rate": 3.904923599320883e-05, | |
| "loss": 1.5409, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.39881204921510394, | |
| "grad_norm": 0.1707919112561785, | |
| "learning_rate": 3.989813242784381e-05, | |
| "loss": 1.5509, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4072974119643615, | |
| "grad_norm": 0.14660149264284913, | |
| "learning_rate": 4.074702886247878e-05, | |
| "loss": 1.5433, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.415782774713619, | |
| "grad_norm": 0.12478895483834351, | |
| "learning_rate": 4.1595925297113755e-05, | |
| "loss": 1.5382, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.42426813746287656, | |
| "grad_norm": 0.12327957445795817, | |
| "learning_rate": 4.244482173174873e-05, | |
| "loss": 1.5515, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.43275350021213405, | |
| "grad_norm": 0.12922777738650987, | |
| "learning_rate": 4.32937181663837e-05, | |
| "loss": 1.5688, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.4412388629613916, | |
| "grad_norm": 0.12486802189783415, | |
| "learning_rate": 4.414261460101868e-05, | |
| "loss": 1.5452, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.44972422571064913, | |
| "grad_norm": 0.1360610874577123, | |
| "learning_rate": 4.499151103565366e-05, | |
| "loss": 1.5493, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.45820958845990667, | |
| "grad_norm": 0.1884897685356775, | |
| "learning_rate": 4.5840407470288625e-05, | |
| "loss": 1.5511, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.4666949512091642, | |
| "grad_norm": 0.12446302384809525, | |
| "learning_rate": 4.6689303904923606e-05, | |
| "loss": 1.5458, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.4751803139584217, | |
| "grad_norm": 0.13169591804768588, | |
| "learning_rate": 4.753820033955857e-05, | |
| "loss": 1.5569, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.48366567670767924, | |
| "grad_norm": 0.1343809247449631, | |
| "learning_rate": 4.8387096774193554e-05, | |
| "loss": 1.5408, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.4921510394569368, | |
| "grad_norm": 0.14024589853602, | |
| "learning_rate": 4.923599320882852e-05, | |
| "loss": 1.5487, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5006364022061943, | |
| "grad_norm": 0.16240429253875313, | |
| "learning_rate": 4.999999560970061e-05, | |
| "loss": 1.5488, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.5091217649554518, | |
| "grad_norm": 0.12575424857894482, | |
| "learning_rate": 4.999946877563971e-05, | |
| "loss": 1.532, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5091217649554518, | |
| "eval_loss": 1.519254446029663, | |
| "eval_runtime": 53.3242, | |
| "eval_samples_per_second": 7.145, | |
| "eval_steps_per_second": 0.9, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.5176071277047094, | |
| "grad_norm": 0.18688482756329736, | |
| "learning_rate": 4.999806390290309e-05, | |
| "loss": 1.5544, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.5260924904539669, | |
| "grad_norm": 0.12425469431830571, | |
| "learning_rate": 4.999578104083307e-05, | |
| "loss": 1.5443, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.5345778532032245, | |
| "grad_norm": 0.1299027485420099, | |
| "learning_rate": 4.999262026960902e-05, | |
| "loss": 1.5569, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.543063215952482, | |
| "grad_norm": 0.11441754852508934, | |
| "learning_rate": 4.998858170024449e-05, | |
| "loss": 1.5316, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.5515485787017395, | |
| "grad_norm": 0.14888547248976478, | |
| "learning_rate": 4.998366547458326e-05, | |
| "loss": 1.5177, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.560033941450997, | |
| "grad_norm": 0.14859292774768867, | |
| "learning_rate": 4.997787176529449e-05, | |
| "loss": 1.5394, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.5685193042002545, | |
| "grad_norm": 0.12499154376539734, | |
| "learning_rate": 4.997120077586651e-05, | |
| "loss": 1.5554, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.5770046669495121, | |
| "grad_norm": 0.1218974898058821, | |
| "learning_rate": 4.9963652740599774e-05, | |
| "loss": 1.5335, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.5854900296987696, | |
| "grad_norm": 0.1273110498715124, | |
| "learning_rate": 4.995522792459859e-05, | |
| "loss": 1.5349, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.5939753924480271, | |
| "grad_norm": 0.12115412881719101, | |
| "learning_rate": 4.994592662376183e-05, | |
| "loss": 1.5419, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.6024607551972847, | |
| "grad_norm": 0.14855096330233286, | |
| "learning_rate": 4.99357491647725e-05, | |
| "loss": 1.513, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.6109461179465422, | |
| "grad_norm": 0.11407988659327956, | |
| "learning_rate": 4.992469590508628e-05, | |
| "loss": 1.5243, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.6194314806957998, | |
| "grad_norm": 0.1197712643781127, | |
| "learning_rate": 4.9912767232919035e-05, | |
| "loss": 1.5177, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.6279168434450573, | |
| "grad_norm": 0.12400515877262065, | |
| "learning_rate": 4.9899963567233074e-05, | |
| "loss": 1.5619, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.6364022061943148, | |
| "grad_norm": 0.12250385257708406, | |
| "learning_rate": 4.988628535772249e-05, | |
| "loss": 1.539, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6448875689435724, | |
| "grad_norm": 0.1262441090496857, | |
| "learning_rate": 4.987173308479738e-05, | |
| "loss": 1.5195, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6533729316928298, | |
| "grad_norm": 0.12459694416473029, | |
| "learning_rate": 4.985630725956694e-05, | |
| "loss": 1.5462, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.6618582944420874, | |
| "grad_norm": 0.12985189006106762, | |
| "learning_rate": 4.9840008423821527e-05, | |
| "loss": 1.5113, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.6703436571913449, | |
| "grad_norm": 0.12689306141471304, | |
| "learning_rate": 4.9822837150013636e-05, | |
| "loss": 1.5201, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.6788290199406024, | |
| "grad_norm": 0.15393156370587963, | |
| "learning_rate": 4.980479404123778e-05, | |
| "loss": 1.5121, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.68731438268986, | |
| "grad_norm": 0.13213701895207608, | |
| "learning_rate": 4.978587973120931e-05, | |
| "loss": 1.5307, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.6957997454391175, | |
| "grad_norm": 0.11561354931316294, | |
| "learning_rate": 4.9766094884242184e-05, | |
| "loss": 1.5316, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.7042851081883751, | |
| "grad_norm": 0.12414772399330044, | |
| "learning_rate": 4.974544019522559e-05, | |
| "loss": 1.5148, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.7127704709376326, | |
| "grad_norm": 0.1171652849153521, | |
| "learning_rate": 4.972391638959959e-05, | |
| "loss": 1.5096, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.7212558336868902, | |
| "grad_norm": 0.12868937349582316, | |
| "learning_rate": 4.9701524223329585e-05, | |
| "loss": 1.5282, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.7297411964361477, | |
| "grad_norm": 0.1200015077117309, | |
| "learning_rate": 4.967826448287981e-05, | |
| "loss": 1.5512, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.7382265591854051, | |
| "grad_norm": 0.12340885660045105, | |
| "learning_rate": 4.96541379851857e-05, | |
| "loss": 1.5394, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.7467119219346627, | |
| "grad_norm": 0.12976937691467555, | |
| "learning_rate": 4.962914557762517e-05, | |
| "loss": 1.51, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.7551972846839202, | |
| "grad_norm": 0.11912878476038466, | |
| "learning_rate": 4.9603288137988905e-05, | |
| "loss": 1.5294, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.7636826474331778, | |
| "grad_norm": 0.1299625480337927, | |
| "learning_rate": 4.957656657444947e-05, | |
| "loss": 1.507, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.7721680101824353, | |
| "grad_norm": 0.12380144459698468, | |
| "learning_rate": 4.954898182552946e-05, | |
| "loss": 1.5376, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.7806533729316928, | |
| "grad_norm": 0.13139339643682763, | |
| "learning_rate": 4.9520534860068535e-05, | |
| "loss": 1.5291, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.7891387356809504, | |
| "grad_norm": 0.13088956203983898, | |
| "learning_rate": 4.949122667718935e-05, | |
| "loss": 1.5239, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.7976240984302079, | |
| "grad_norm": 0.12586052988453703, | |
| "learning_rate": 4.94610583062625e-05, | |
| "loss": 1.5525, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.8061094611794655, | |
| "grad_norm": 0.12020996031652877, | |
| "learning_rate": 4.943003080687035e-05, | |
| "loss": 1.5525, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.814594823928723, | |
| "grad_norm": 0.12866375954060869, | |
| "learning_rate": 4.9398145268769856e-05, | |
| "loss": 1.5266, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.8230801866779804, | |
| "grad_norm": 0.13166136756817035, | |
| "learning_rate": 4.936540281185423e-05, | |
| "loss": 1.5041, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.831565549427238, | |
| "grad_norm": 0.12481946698483787, | |
| "learning_rate": 4.933180458611364e-05, | |
| "loss": 1.5271, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.8400509121764955, | |
| "grad_norm": 0.12264463761209114, | |
| "learning_rate": 4.9297351771594844e-05, | |
| "loss": 1.5354, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.8485362749257531, | |
| "grad_norm": 0.11985452856537594, | |
| "learning_rate": 4.926204557835968e-05, | |
| "loss": 1.5167, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.8570216376750106, | |
| "grad_norm": 0.13125396521190327, | |
| "learning_rate": 4.9225887246442634e-05, | |
| "loss": 1.5282, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.8655070004242681, | |
| "grad_norm": 0.12730192328072554, | |
| "learning_rate": 4.918887804580725e-05, | |
| "loss": 1.5089, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.8739923631735257, | |
| "grad_norm": 0.12724644219344786, | |
| "learning_rate": 4.915101927630153e-05, | |
| "loss": 1.4964, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.8824777259227832, | |
| "grad_norm": 0.13578611501833232, | |
| "learning_rate": 4.911231226761227e-05, | |
| "loss": 1.5189, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.8909630886720408, | |
| "grad_norm": 0.13577513964986457, | |
| "learning_rate": 4.90727583792184e-05, | |
| "loss": 1.5149, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.8994484514212983, | |
| "grad_norm": 0.1269735011676505, | |
| "learning_rate": 4.903235900034317e-05, | |
| "loss": 1.5066, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.9079338141705557, | |
| "grad_norm": 0.13250058214235566, | |
| "learning_rate": 4.899111554990543e-05, | |
| "loss": 1.5129, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.9164191769198133, | |
| "grad_norm": 0.13130735246433495, | |
| "learning_rate": 4.894902947646975e-05, | |
| "loss": 1.5156, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.9249045396690708, | |
| "grad_norm": 0.1273580180253049, | |
| "learning_rate": 4.890610225819553e-05, | |
| "loss": 1.5324, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.9333899024183284, | |
| "grad_norm": 0.13155314243939242, | |
| "learning_rate": 4.8862335402785136e-05, | |
| "loss": 1.5106, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.9418752651675859, | |
| "grad_norm": 0.13564895211984299, | |
| "learning_rate": 4.88177304474309e-05, | |
| "loss": 1.5067, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.9503606279168434, | |
| "grad_norm": 0.12774735587114736, | |
| "learning_rate": 4.877228895876115e-05, | |
| "loss": 1.5182, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.958845990666101, | |
| "grad_norm": 0.1307997709537685, | |
| "learning_rate": 4.872601253278517e-05, | |
| "loss": 1.4969, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.9673313534153585, | |
| "grad_norm": 0.1304794845040634, | |
| "learning_rate": 4.867890279483717e-05, | |
| "loss": 1.5264, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.9758167161646161, | |
| "grad_norm": 0.13666141796489684, | |
| "learning_rate": 4.8630961399519206e-05, | |
| "loss": 1.5467, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.9843020789138736, | |
| "grad_norm": 0.1370278303190263, | |
| "learning_rate": 4.8582190030643e-05, | |
| "loss": 1.5127, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.9927874416631312, | |
| "grad_norm": 0.1390936629299565, | |
| "learning_rate": 4.8532590401170894e-05, | |
| "loss": 1.5058, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.0012728044123886, | |
| "grad_norm": 0.12934475548108287, | |
| "learning_rate": 4.848216425315561e-05, | |
| "loss": 1.5202, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.0097581671616462, | |
| "grad_norm": 0.13898591683370803, | |
| "learning_rate": 4.843091335767913e-05, | |
| "loss": 1.4563, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.0182435299109036, | |
| "grad_norm": 0.17488231535826249, | |
| "learning_rate": 4.837883951479043e-05, | |
| "loss": 1.4402, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.0182435299109036, | |
| "eval_loss": 1.4955657720565796, | |
| "eval_runtime": 52.424, | |
| "eval_samples_per_second": 7.268, | |
| "eval_steps_per_second": 0.916, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.0267288926601612, | |
| "grad_norm": 0.1536036344095855, | |
| "learning_rate": 4.832594455344229e-05, | |
| "loss": 1.4848, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.0352142554094188, | |
| "grad_norm": 0.15762414421336599, | |
| "learning_rate": 4.827223033142706e-05, | |
| "loss": 1.4567, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.0436996181586762, | |
| "grad_norm": 0.15058229398130366, | |
| "learning_rate": 4.8217698735311414e-05, | |
| "loss": 1.4672, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.0521849809079338, | |
| "grad_norm": 0.16010992835678386, | |
| "learning_rate": 4.8162351680370044e-05, | |
| "loss": 1.4458, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.0606703436571914, | |
| "grad_norm": 0.16758816000341356, | |
| "learning_rate": 4.810619111051847e-05, | |
| "loss": 1.4842, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.069155706406449, | |
| "grad_norm": 0.16559260972674986, | |
| "learning_rate": 4.8049218998244696e-05, | |
| "loss": 1.4556, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.0776410691557063, | |
| "grad_norm": 0.17237632034416966, | |
| "learning_rate": 4.7991437344539966e-05, | |
| "loss": 1.4813, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.086126431904964, | |
| "grad_norm": 0.17112756741722487, | |
| "learning_rate": 4.793284817882845e-05, | |
| "loss": 1.4535, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.0946117946542215, | |
| "grad_norm": 0.16828572707718548, | |
| "learning_rate": 4.787345355889604e-05, | |
| "loss": 1.4344, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.103097157403479, | |
| "grad_norm": 0.15709986047041227, | |
| "learning_rate": 4.7813255570817985e-05, | |
| "loss": 1.4744, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.1115825201527365, | |
| "grad_norm": 0.16651547128146313, | |
| "learning_rate": 4.775225632888568e-05, | |
| "loss": 1.4561, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.120067882901994, | |
| "grad_norm": 0.16750176017515714, | |
| "learning_rate": 4.76904579755324e-05, | |
| "loss": 1.4616, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.1285532456512515, | |
| "grad_norm": 0.1608016567554825, | |
| "learning_rate": 4.7627862681258037e-05, | |
| "loss": 1.4593, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.137038608400509, | |
| "grad_norm": 0.21390766919038295, | |
| "learning_rate": 4.756447264455287e-05, | |
| "loss": 1.4484, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.1455239711497667, | |
| "grad_norm": 0.16826883293172662, | |
| "learning_rate": 4.750029009182038e-05, | |
| "loss": 1.4703, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.1540093338990243, | |
| "grad_norm": 0.17431508867079595, | |
| "learning_rate": 4.7435317277299e-05, | |
| "loss": 1.4701, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.1624946966482816, | |
| "grad_norm": 0.15973851467570443, | |
| "learning_rate": 4.736955648298299e-05, | |
| "loss": 1.4503, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.1709800593975392, | |
| "grad_norm": 0.1887713767970947, | |
| "learning_rate": 4.730301001854225e-05, | |
| "loss": 1.4624, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.1794654221467968, | |
| "grad_norm": 0.16898695344997974, | |
| "learning_rate": 4.7235680221241216e-05, | |
| "loss": 1.4452, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.1879507848960542, | |
| "grad_norm": 0.20014553287073528, | |
| "learning_rate": 4.716756945585681e-05, | |
| "loss": 1.4717, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.1964361476453118, | |
| "grad_norm": 0.17137954325200072, | |
| "learning_rate": 4.709868011459528e-05, | |
| "loss": 1.4403, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.2049215103945694, | |
| "grad_norm": 0.17801721751888322, | |
| "learning_rate": 4.7029014617008294e-05, | |
| "loss": 1.4339, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.213406873143827, | |
| "grad_norm": 0.17139613676642362, | |
| "learning_rate": 4.695857540990789e-05, | |
| "loss": 1.4573, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.2218922358930844, | |
| "grad_norm": 0.16971403514498054, | |
| "learning_rate": 4.688736496728058e-05, | |
| "loss": 1.4282, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.230377598642342, | |
| "grad_norm": 0.17200272420880428, | |
| "learning_rate": 4.681538579020038e-05, | |
| "loss": 1.4434, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.2388629613915996, | |
| "grad_norm": 0.17208160407432616, | |
| "learning_rate": 4.6742640406741106e-05, | |
| "loss": 1.45, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.247348324140857, | |
| "grad_norm": 0.1939626212901777, | |
| "learning_rate": 4.666913137188743e-05, | |
| "loss": 1.4608, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.2558336868901145, | |
| "grad_norm": 0.17291794493304186, | |
| "learning_rate": 4.6594861267445236e-05, | |
| "loss": 1.4671, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.2643190496393721, | |
| "grad_norm": 0.18219792041638924, | |
| "learning_rate": 4.651983270195093e-05, | |
| "loss": 1.4262, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.2728044123886297, | |
| "grad_norm": 0.18086437830489926, | |
| "learning_rate": 4.644404831057979e-05, | |
| "loss": 1.4455, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.281289775137887, | |
| "grad_norm": 0.17417619624549402, | |
| "learning_rate": 4.636751075505344e-05, | |
| "loss": 1.4873, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.2897751378871447, | |
| "grad_norm": 0.18354282411845188, | |
| "learning_rate": 4.629022272354637e-05, | |
| "loss": 1.4525, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.298260500636402, | |
| "grad_norm": 0.17985617345325455, | |
| "learning_rate": 4.621218693059149e-05, | |
| "loss": 1.4303, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.3067458633856597, | |
| "grad_norm": 0.1809708317849863, | |
| "learning_rate": 4.6133406116984795e-05, | |
| "loss": 1.4631, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.3152312261349173, | |
| "grad_norm": 0.17487374671212322, | |
| "learning_rate": 4.6053883049689145e-05, | |
| "loss": 1.4482, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.3237165888841749, | |
| "grad_norm": 0.19912807671077193, | |
| "learning_rate": 4.5973620521737036e-05, | |
| "loss": 1.4497, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.3322019516334322, | |
| "grad_norm": 0.17853627546912074, | |
| "learning_rate": 4.5892621352132514e-05, | |
| "loss": 1.4456, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.3406873143826898, | |
| "grad_norm": 0.18252596927754394, | |
| "learning_rate": 4.581088838575218e-05, | |
| "loss": 1.4328, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.3491726771319474, | |
| "grad_norm": 0.17604951053556211, | |
| "learning_rate": 4.572842449324525e-05, | |
| "loss": 1.4442, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.3576580398812048, | |
| "grad_norm": 0.18358942463311748, | |
| "learning_rate": 4.564523257093275e-05, | |
| "loss": 1.4338, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.3661434026304624, | |
| "grad_norm": 0.20508703236267142, | |
| "learning_rate": 4.5561315540705774e-05, | |
| "loss": 1.4445, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.37462876537972, | |
| "grad_norm": 0.18486352550747187, | |
| "learning_rate": 4.547667634992288e-05, | |
| "loss": 1.4261, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.3831141281289776, | |
| "grad_norm": 0.17492766465456316, | |
| "learning_rate": 4.539131797130656e-05, | |
| "loss": 1.4258, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.391599490878235, | |
| "grad_norm": 0.19692876587833674, | |
| "learning_rate": 4.530524340283881e-05, | |
| "loss": 1.4349, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.4000848536274926, | |
| "grad_norm": 0.19155373430892478, | |
| "learning_rate": 4.521845566765589e-05, | |
| "loss": 1.4536, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.4085702163767502, | |
| "grad_norm": 0.18544325977459192, | |
| "learning_rate": 4.513095781394208e-05, | |
| "loss": 1.4363, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.4170555791260075, | |
| "grad_norm": 0.177828004720666, | |
| "learning_rate": 4.504275291482267e-05, | |
| "loss": 1.4595, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.4255409418752651, | |
| "grad_norm": 0.17855432230356816, | |
| "learning_rate": 4.495384406825601e-05, | |
| "loss": 1.4211, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.4340263046245227, | |
| "grad_norm": 0.20232492538380317, | |
| "learning_rate": 4.486423439692469e-05, | |
| "loss": 1.4189, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.4425116673737803, | |
| "grad_norm": 0.1975109303350431, | |
| "learning_rate": 4.477392704812585e-05, | |
| "loss": 1.4565, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.4509970301230377, | |
| "grad_norm": 0.19619010830399825, | |
| "learning_rate": 4.468292519366071e-05, | |
| "loss": 1.4382, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.4594823928722953, | |
| "grad_norm": 0.18168826428246143, | |
| "learning_rate": 4.459123202972308e-05, | |
| "loss": 1.4471, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.4679677556215527, | |
| "grad_norm": 0.1923264062362399, | |
| "learning_rate": 4.449885077678717e-05, | |
| "loss": 1.4153, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.4764531183708103, | |
| "grad_norm": 0.1907937313040222, | |
| "learning_rate": 4.440578467949445e-05, | |
| "loss": 1.4432, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.4849384811200679, | |
| "grad_norm": 0.19107457667767244, | |
| "learning_rate": 4.431203700653968e-05, | |
| "loss": 1.4285, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.4934238438693255, | |
| "grad_norm": 0.19847350429107552, | |
| "learning_rate": 4.421761105055613e-05, | |
| "loss": 1.4383, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.501909206618583, | |
| "grad_norm": 0.18536475556610216, | |
| "learning_rate": 4.4122510127999937e-05, | |
| "loss": 1.42, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.5103945693678404, | |
| "grad_norm": 0.18481023473586697, | |
| "learning_rate": 4.4026737579033584e-05, | |
| "loss": 1.4384, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.518879932117098, | |
| "grad_norm": 0.20863867505874642, | |
| "learning_rate": 4.393029676740864e-05, | |
| "loss": 1.4543, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.5273652948663554, | |
| "grad_norm": 0.1816036870853105, | |
| "learning_rate": 4.3833191080347575e-05, | |
| "loss": 1.434, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.5273652948663554, | |
| "eval_loss": 1.4622184038162231, | |
| "eval_runtime": 52.4041, | |
| "eval_samples_per_second": 7.27, | |
| "eval_steps_per_second": 0.916, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.535850657615613, | |
| "grad_norm": 0.19378252368958881, | |
| "learning_rate": 4.3735423928424815e-05, | |
| "loss": 1.4275, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.5443360203648706, | |
| "grad_norm": 0.20453331251433848, | |
| "learning_rate": 4.363699874544697e-05, | |
| "loss": 1.4203, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.5528213831141282, | |
| "grad_norm": 0.26684319417219377, | |
| "learning_rate": 4.3537918988332156e-05, | |
| "loss": 1.4372, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.5613067458633858, | |
| "grad_norm": 0.25745160303419773, | |
| "learning_rate": 4.343818813698868e-05, | |
| "loss": 1.4082, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.5697921086126432, | |
| "grad_norm": 0.19969727996700776, | |
| "learning_rate": 4.3337809694192765e-05, | |
| "loss": 1.4314, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.5782774713619008, | |
| "grad_norm": 0.20117210832277968, | |
| "learning_rate": 4.3236787185465525e-05, | |
| "loss": 1.4293, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.5867628341111581, | |
| "grad_norm": 0.20173003641028897, | |
| "learning_rate": 4.313512415894913e-05, | |
| "loss": 1.4406, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.5952481968604157, | |
| "grad_norm": 0.20304770794371527, | |
| "learning_rate": 4.303282418528224e-05, | |
| "loss": 1.4286, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.6037335596096733, | |
| "grad_norm": 0.19126658907738198, | |
| "learning_rate": 4.292989085747452e-05, | |
| "loss": 1.4184, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.612218922358931, | |
| "grad_norm": 0.20069554966453027, | |
| "learning_rate": 4.282632779078051e-05, | |
| "loss": 1.4133, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.6207042851081885, | |
| "grad_norm": 0.1952881519566686, | |
| "learning_rate": 4.2722138622572624e-05, | |
| "loss": 1.4432, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.629189647857446, | |
| "grad_norm": 0.19763704668680288, | |
| "learning_rate": 4.261732701221339e-05, | |
| "loss": 1.3921, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.6376750106067033, | |
| "grad_norm": 0.19821464294464497, | |
| "learning_rate": 4.2511896640926925e-05, | |
| "loss": 1.4454, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.6461603733559609, | |
| "grad_norm": 0.20456545626297834, | |
| "learning_rate": 4.240585121166966e-05, | |
| "loss": 1.4147, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.6546457361052185, | |
| "grad_norm": 0.2119092529186395, | |
| "learning_rate": 4.229919444900027e-05, | |
| "loss": 1.3969, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.663131098854476, | |
| "grad_norm": 0.20330157582122357, | |
| "learning_rate": 4.2191930098948865e-05, | |
| "loss": 1.426, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.6716164616037337, | |
| "grad_norm": 0.21761164739298738, | |
| "learning_rate": 4.2084061928885406e-05, | |
| "loss": 1.4246, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.680101824352991, | |
| "grad_norm": 0.19331588142071401, | |
| "learning_rate": 4.197559372738741e-05, | |
| "loss": 1.4305, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.6885871871022486, | |
| "grad_norm": 0.20188460724329996, | |
| "learning_rate": 4.186652930410685e-05, | |
| "loss": 1.4153, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.697072549851506, | |
| "grad_norm": 0.20988950033571588, | |
| "learning_rate": 4.1756872489636425e-05, | |
| "loss": 1.3894, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.7055579126007636, | |
| "grad_norm": 0.1966475893123187, | |
| "learning_rate": 4.1646627135374916e-05, | |
| "loss": 1.3962, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.7140432753500212, | |
| "grad_norm": 0.20785207367991768, | |
| "learning_rate": 4.1535797113392004e-05, | |
| "loss": 1.4037, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 1.7225286380992788, | |
| "grad_norm": 0.2029940281663133, | |
| "learning_rate": 4.1424386316292224e-05, | |
| "loss": 1.4011, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 1.7310140008485364, | |
| "grad_norm": 0.2247844551379277, | |
| "learning_rate": 4.131239865707829e-05, | |
| "loss": 1.4084, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 1.7394993635977938, | |
| "grad_norm": 0.20900441746105022, | |
| "learning_rate": 4.11998380690136e-05, | |
| "loss": 1.4235, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.7479847263470514, | |
| "grad_norm": 0.20362408546889926, | |
| "learning_rate": 4.108670850548416e-05, | |
| "loss": 1.4204, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 1.7564700890963088, | |
| "grad_norm": 0.22281567946240438, | |
| "learning_rate": 4.097301393985968e-05, | |
| "loss": 1.4023, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 1.7649554518455663, | |
| "grad_norm": 0.20867113178797225, | |
| "learning_rate": 4.085875836535404e-05, | |
| "loss": 1.3895, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 1.773440814594824, | |
| "grad_norm": 0.22113231886160947, | |
| "learning_rate": 4.0743945794885063e-05, | |
| "loss": 1.3963, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 1.7819261773440815, | |
| "grad_norm": 0.22334563577844263, | |
| "learning_rate": 4.062858026093351e-05, | |
| "loss": 1.3988, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.7904115400933391, | |
| "grad_norm": 0.23218581668265403, | |
| "learning_rate": 4.051266581540152e-05, | |
| "loss": 1.4068, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 1.7988969028425965, | |
| "grad_norm": 0.20295589384571033, | |
| "learning_rate": 4.0396206529470234e-05, | |
| "loss": 1.3883, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 1.8073822655918539, | |
| "grad_norm": 0.22861611442392848, | |
| "learning_rate": 4.027920649345687e-05, | |
| "loss": 1.4043, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 1.8158676283411115, | |
| "grad_norm": 0.2083012771089638, | |
| "learning_rate": 4.0161669816671e-05, | |
| "loss": 1.398, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 1.824352991090369, | |
| "grad_norm": 0.21936173231840464, | |
| "learning_rate": 4.004360062727028e-05, | |
| "loss": 1.4142, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 1.8328383538396267, | |
| "grad_norm": 0.21383435796328337, | |
| "learning_rate": 3.9925003072115406e-05, | |
| "loss": 1.4138, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 1.8413237165888843, | |
| "grad_norm": 0.23301608248270392, | |
| "learning_rate": 3.9805881316624506e-05, | |
| "loss": 1.4195, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 1.8498090793381419, | |
| "grad_norm": 0.22424766656883474, | |
| "learning_rate": 3.968623954462681e-05, | |
| "loss": 1.4011, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 1.8582944420873992, | |
| "grad_norm": 0.21286417342881453, | |
| "learning_rate": 3.9566081958215734e-05, | |
| "loss": 1.409, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 1.8667798048366566, | |
| "grad_norm": 0.21944800687444807, | |
| "learning_rate": 3.9445412777601284e-05, | |
| "loss": 1.3877, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.8752651675859142, | |
| "grad_norm": 0.23113173625974803, | |
| "learning_rate": 3.932423624096181e-05, | |
| "loss": 1.4089, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 1.8837505303351718, | |
| "grad_norm": 0.2081941699587778, | |
| "learning_rate": 3.920255660429517e-05, | |
| "loss": 1.4024, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 1.8922358930844294, | |
| "grad_norm": 0.2188685806654701, | |
| "learning_rate": 3.908037814126927e-05, | |
| "loss": 1.3878, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 1.900721255833687, | |
| "grad_norm": 0.22761843244757962, | |
| "learning_rate": 3.895770514307193e-05, | |
| "loss": 1.4004, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 1.9092066185829444, | |
| "grad_norm": 0.23309183623120422, | |
| "learning_rate": 3.883454191826017e-05, | |
| "loss": 1.4188, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 1.917691981332202, | |
| "grad_norm": 0.20329785843911802, | |
| "learning_rate": 3.871089279260891e-05, | |
| "loss": 1.3893, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 1.9261773440814594, | |
| "grad_norm": 0.23470973193726366, | |
| "learning_rate": 3.8586762108958995e-05, | |
| "loss": 1.3974, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 1.934662706830717, | |
| "grad_norm": 0.22779136837044714, | |
| "learning_rate": 3.8462154227064725e-05, | |
| "loss": 1.4115, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 1.9431480695799745, | |
| "grad_norm": 0.22338952315651892, | |
| "learning_rate": 3.833707352344068e-05, | |
| "loss": 1.3873, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 1.9516334323292321, | |
| "grad_norm": 0.23069304025882129, | |
| "learning_rate": 3.821152439120801e-05, | |
| "loss": 1.3944, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.9601187950784897, | |
| "grad_norm": 0.23590596270163203, | |
| "learning_rate": 3.808551123994018e-05, | |
| "loss": 1.3857, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 1.9686041578277471, | |
| "grad_norm": 0.22545661808214923, | |
| "learning_rate": 3.795903849550805e-05, | |
| "loss": 1.3628, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 1.9770895205770047, | |
| "grad_norm": 0.2450769875954842, | |
| "learning_rate": 3.7832110599924455e-05, | |
| "loss": 1.4079, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 1.985574883326262, | |
| "grad_norm": 0.22931499326784313, | |
| "learning_rate": 3.7704732011188166e-05, | |
| "loss": 1.379, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 1.9940602460755197, | |
| "grad_norm": 0.22417244507397657, | |
| "learning_rate": 3.7576907203127346e-05, | |
| "loss": 1.4035, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.0025456088247773, | |
| "grad_norm": 0.24496197221575314, | |
| "learning_rate": 3.7448640665242406e-05, | |
| "loss": 1.442, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 2.011030971574035, | |
| "grad_norm": 0.2532740296990078, | |
| "learning_rate": 3.73199369025483e-05, | |
| "loss": 1.2672, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 2.0195163343232925, | |
| "grad_norm": 0.2890155987968593, | |
| "learning_rate": 3.7190800435416355e-05, | |
| "loss": 1.246, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 2.02800169707255, | |
| "grad_norm": 0.2541972565696406, | |
| "learning_rate": 3.706123579941545e-05, | |
| "loss": 1.2603, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 2.036487059821807, | |
| "grad_norm": 0.2530140862527023, | |
| "learning_rate": 3.693124754515272e-05, | |
| "loss": 1.2638, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.036487059821807, | |
| "eval_loss": 1.435962438583374, | |
| "eval_runtime": 52.582, | |
| "eval_samples_per_second": 7.246, | |
| "eval_steps_per_second": 0.913, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.044972422571065, | |
| "grad_norm": 0.25100458343337734, | |
| "learning_rate": 3.680084023811377e-05, | |
| "loss": 1.2711, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 2.0534577853203224, | |
| "grad_norm": 0.2695727673292618, | |
| "learning_rate": 3.66700184585023e-05, | |
| "loss": 1.2578, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 2.06194314806958, | |
| "grad_norm": 0.2605068415443213, | |
| "learning_rate": 3.6538786801079226e-05, | |
| "loss": 1.2506, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 2.0704285108188376, | |
| "grad_norm": 0.27415607207865045, | |
| "learning_rate": 3.64071498750013e-05, | |
| "loss": 1.2852, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 2.078913873568095, | |
| "grad_norm": 0.2688900338206285, | |
| "learning_rate": 3.627511230365928e-05, | |
| "loss": 1.2695, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.0873992363173524, | |
| "grad_norm": 0.2750825805336503, | |
| "learning_rate": 3.614267872451546e-05, | |
| "loss": 1.2643, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 2.09588459906661, | |
| "grad_norm": 0.2659269066581903, | |
| "learning_rate": 3.600985378894086e-05, | |
| "loss": 1.2868, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 2.1043699618158676, | |
| "grad_norm": 0.24411151291321526, | |
| "learning_rate": 3.587664216205183e-05, | |
| "loss": 1.2571, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 2.112855324565125, | |
| "grad_norm": 0.2574194755634052, | |
| "learning_rate": 3.574304852254621e-05, | |
| "loss": 1.2769, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 2.1213406873143827, | |
| "grad_norm": 0.2894545074998905, | |
| "learning_rate": 3.5609077562538997e-05, | |
| "loss": 1.2469, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 5890, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2101429250555904.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |