| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 1521, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01972386587771203, | |
| "grad_norm": 8.260985845744274, | |
| "learning_rate": 6.493506493506493e-07, | |
| "loss": 0.7556, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03944773175542406, | |
| "grad_norm": 3.285262290141762, | |
| "learning_rate": 1.2987012987012986e-06, | |
| "loss": 0.6243, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05917159763313609, | |
| "grad_norm": 1.8240152669785012, | |
| "learning_rate": 1.9480519480519483e-06, | |
| "loss": 0.527, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07889546351084813, | |
| "grad_norm": 2.484918347587673, | |
| "learning_rate": 2.597402597402597e-06, | |
| "loss": 0.4858, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09861932938856016, | |
| "grad_norm": 1.6071106659008887, | |
| "learning_rate": 3.246753246753247e-06, | |
| "loss": 0.4616, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11834319526627218, | |
| "grad_norm": 1.7753595146102916, | |
| "learning_rate": 3.896103896103897e-06, | |
| "loss": 0.4467, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13806706114398423, | |
| "grad_norm": 1.9952784096044014, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 0.4355, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15779092702169625, | |
| "grad_norm": 2.0090088525329057, | |
| "learning_rate": 4.999952075361122e-06, | |
| "loss": 0.4303, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17751479289940827, | |
| "grad_norm": 1.9429866993343683, | |
| "learning_rate": 4.99910013857428e-06, | |
| "loss": 0.4213, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19723865877712032, | |
| "grad_norm": 2.2150406308730166, | |
| "learning_rate": 4.997183673954895e-06, | |
| "loss": 0.4205, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21696252465483234, | |
| "grad_norm": 2.3280715715799105, | |
| "learning_rate": 4.994203588590157e-06, | |
| "loss": 0.4132, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.23668639053254437, | |
| "grad_norm": 2.0514718162160617, | |
| "learning_rate": 4.9901612929925455e-06, | |
| "loss": 0.4097, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2564102564102564, | |
| "grad_norm": 2.248051724393392, | |
| "learning_rate": 4.985058700432217e-06, | |
| "loss": 0.4078, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.27613412228796846, | |
| "grad_norm": 2.4477065193392114, | |
| "learning_rate": 4.978898226031426e-06, | |
| "loss": 0.4035, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2958579881656805, | |
| "grad_norm": 2.3530821376592317, | |
| "learning_rate": 4.97168278562142e-06, | |
| "loss": 0.3988, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3155818540433925, | |
| "grad_norm": 2.0658208779463796, | |
| "learning_rate": 4.9634157943623345e-06, | |
| "loss": 0.4008, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.33530571992110453, | |
| "grad_norm": 1.6308701318103827, | |
| "learning_rate": 4.954101165126764e-06, | |
| "loss": 0.3955, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.35502958579881655, | |
| "grad_norm": 1.8767575875235638, | |
| "learning_rate": 4.943743306647738e-06, | |
| "loss": 0.3964, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3747534516765286, | |
| "grad_norm": 2.158851334024998, | |
| "learning_rate": 4.932347121432018e-06, | |
| "loss": 0.3955, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.39447731755424065, | |
| "grad_norm": 2.2424601067528367, | |
| "learning_rate": 4.919918003439677e-06, | |
| "loss": 0.3929, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.41420118343195267, | |
| "grad_norm": 1.4704562127782181, | |
| "learning_rate": 4.9064618355310694e-06, | |
| "loss": 0.3951, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4339250493096647, | |
| "grad_norm": 1.5325962055467024, | |
| "learning_rate": 4.8919849866823955e-06, | |
| "loss": 0.3936, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4536489151873767, | |
| "grad_norm": 1.752553432251344, | |
| "learning_rate": 4.8764943089711876e-06, | |
| "loss": 0.3894, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.47337278106508873, | |
| "grad_norm": 3.582185649197669, | |
| "learning_rate": 4.859997134333133e-06, | |
| "loss": 0.39, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.4930966469428008, | |
| "grad_norm": 2.283623608488685, | |
| "learning_rate": 4.842501271091773e-06, | |
| "loss": 0.3845, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5128205128205128, | |
| "grad_norm": 2.954635543178996, | |
| "learning_rate": 4.8240150002627285e-06, | |
| "loss": 0.3853, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5325443786982249, | |
| "grad_norm": 2.621411991175976, | |
| "learning_rate": 4.80454707163418e-06, | |
| "loss": 0.3802, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5522682445759369, | |
| "grad_norm": 3.0076538937186554, | |
| "learning_rate": 4.784106699625493e-06, | |
| "loss": 0.3778, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.571992110453649, | |
| "grad_norm": 2.620788244299813, | |
| "learning_rate": 4.762703558925907e-06, | |
| "loss": 0.381, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.591715976331361, | |
| "grad_norm": 2.600774288511616, | |
| "learning_rate": 4.740347779915384e-06, | |
| "loss": 0.3795, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.611439842209073, | |
| "grad_norm": 2.825934593188172, | |
| "learning_rate": 4.717049943869774e-06, | |
| "loss": 0.3754, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.631163708086785, | |
| "grad_norm": 1.9636455738063043, | |
| "learning_rate": 4.692821077952556e-06, | |
| "loss": 0.3709, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.650887573964497, | |
| "grad_norm": 1.465934093555826, | |
| "learning_rate": 4.667672649995539e-06, | |
| "loss": 0.3686, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.6706114398422091, | |
| "grad_norm": 1.6730733158146738, | |
| "learning_rate": 4.641616563071003e-06, | |
| "loss": 0.374, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.6903353057199211, | |
| "grad_norm": 1.6420981338152472, | |
| "learning_rate": 4.6146651498578095e-06, | |
| "loss": 0.3725, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7100591715976331, | |
| "grad_norm": 1.7081311753490396, | |
| "learning_rate": 4.586831166804191e-06, | |
| "loss": 0.3723, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7297830374753451, | |
| "grad_norm": 1.7004420091082517, | |
| "learning_rate": 4.558127788089966e-06, | |
| "loss": 0.3685, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7495069033530573, | |
| "grad_norm": 1.813129623101683, | |
| "learning_rate": 4.5285685993910246e-06, | |
| "loss": 0.3693, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 1.441392302489358, | |
| "learning_rate": 4.49816759144906e-06, | |
| "loss": 0.3672, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.7889546351084813, | |
| "grad_norm": 1.743528342139816, | |
| "learning_rate": 4.466939153449565e-06, | |
| "loss": 0.3629, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8086785009861933, | |
| "grad_norm": 1.5505480061250534, | |
| "learning_rate": 4.434898066211255e-06, | |
| "loss": 0.3647, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8284023668639053, | |
| "grad_norm": 1.748134152515452, | |
| "learning_rate": 4.402059495190112e-06, | |
| "loss": 0.3687, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8481262327416174, | |
| "grad_norm": 1.888131474531523, | |
| "learning_rate": 4.368438983301382e-06, | |
| "loss": 0.368, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.8678500986193294, | |
| "grad_norm": 1.3077877777100417, | |
| "learning_rate": 4.334052443562914e-06, | |
| "loss": 0.364, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.8875739644970414, | |
| "grad_norm": 1.7143497390643974, | |
| "learning_rate": 4.298916151563324e-06, | |
| "loss": 0.3662, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9072978303747534, | |
| "grad_norm": 1.2650560376490414, | |
| "learning_rate": 4.263046737758557e-06, | |
| "loss": 0.3634, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9270216962524654, | |
| "grad_norm": 1.325272234023546, | |
| "learning_rate": 4.226461179600474e-06, | |
| "loss": 0.3647, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9467455621301775, | |
| "grad_norm": 1.7799396783443953, | |
| "learning_rate": 4.189176793501208e-06, | |
| "loss": 0.3601, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.9664694280078896, | |
| "grad_norm": 1.6138030010077298, | |
| "learning_rate": 4.151211226637083e-06, | |
| "loss": 0.3639, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.9861932938856016, | |
| "grad_norm": 1.6475058606657829, | |
| "learning_rate": 4.112582448595989e-06, | |
| "loss": 0.3631, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.3610161542892456, | |
| "eval_runtime": 46.5378, | |
| "eval_samples_per_second": 293.095, | |
| "eval_steps_per_second": 1.16, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 1.0059171597633136, | |
| "grad_norm": 2.318083617694004, | |
| "learning_rate": 4.073308742872136e-06, | |
| "loss": 0.339, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.0256410256410255, | |
| "grad_norm": 2.26507527796031, | |
| "learning_rate": 4.033408698212244e-06, | |
| "loss": 0.2904, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.0453648915187377, | |
| "grad_norm": 2.129210352759771, | |
| "learning_rate": 3.99290119981726e-06, | |
| "loss": 0.2845, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.0650887573964498, | |
| "grad_norm": 2.0458511034566897, | |
| "learning_rate": 3.95180542040374e-06, | |
| "loss": 0.2826, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.0848126232741617, | |
| "grad_norm": 2.34540520465628, | |
| "learning_rate": 3.910140811129166e-06, | |
| "loss": 0.2817, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.1045364891518739, | |
| "grad_norm": 1.5731137478504271, | |
| "learning_rate": 3.8679270923854596e-06, | |
| "loss": 0.2816, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.1242603550295858, | |
| "grad_norm": 1.9641564243584235, | |
| "learning_rate": 3.825184244465071e-06, | |
| "loss": 0.2833, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.143984220907298, | |
| "grad_norm": 1.5653763677552233, | |
| "learning_rate": 3.7819324981040517e-06, | |
| "loss": 0.2835, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.1637080867850098, | |
| "grad_norm": 1.4455902546137582, | |
| "learning_rate": 3.7381923249065838e-06, | |
| "loss": 0.2806, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.183431952662722, | |
| "grad_norm": 1.4589441051909717, | |
| "learning_rate": 3.6939844276555146e-06, | |
| "loss": 0.2842, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.2031558185404339, | |
| "grad_norm": 1.4737079619190827, | |
| "learning_rate": 3.649329730513461e-06, | |
| "loss": 0.2818, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.222879684418146, | |
| "grad_norm": 1.424470321783783, | |
| "learning_rate": 3.6042493691191377e-06, | |
| "loss": 0.2835, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.242603550295858, | |
| "grad_norm": 1.43822809638539, | |
| "learning_rate": 3.558764680583589e-06, | |
| "loss": 0.2829, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.26232741617357, | |
| "grad_norm": 1.4491877471048427, | |
| "learning_rate": 3.51289719339106e-06, | |
| "loss": 0.2823, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.282051282051282, | |
| "grad_norm": 1.4979353903583295, | |
| "learning_rate": 3.4666686172092927e-06, | |
| "loss": 0.2859, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.301775147928994, | |
| "grad_norm": 1.4793881592613725, | |
| "learning_rate": 3.4201008326140596e-06, | |
| "loss": 0.2849, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.3214990138067062, | |
| "grad_norm": 1.6343693105840815, | |
| "learning_rate": 3.3732158807328116e-06, | |
| "loss": 0.2875, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.3412228796844181, | |
| "grad_norm": 1.5638318327999918, | |
| "learning_rate": 3.3260359528123266e-06, | |
| "loss": 0.2877, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.3609467455621302, | |
| "grad_norm": 1.434550639059279, | |
| "learning_rate": 3.2785833797153115e-06, | |
| "loss": 0.2817, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.3806706114398422, | |
| "grad_norm": 1.3783604211664602, | |
| "learning_rate": 3.2308806213509204e-06, | |
| "loss": 0.2809, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.4003944773175543, | |
| "grad_norm": 1.7104337243982326, | |
| "learning_rate": 3.182950256044188e-06, | |
| "loss": 0.2825, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.4201183431952662, | |
| "grad_norm": 1.9527331404429782, | |
| "learning_rate": 3.1348149698494233e-06, | |
| "loss": 0.2827, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.4398422090729783, | |
| "grad_norm": 1.5082040480125063, | |
| "learning_rate": 3.0864975458126158e-06, | |
| "loss": 0.2857, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.4595660749506902, | |
| "grad_norm": 1.5939434329404958, | |
| "learning_rate": 3.038020853187914e-06, | |
| "loss": 0.2831, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.4792899408284024, | |
| "grad_norm": 1.425454732201556, | |
| "learning_rate": 2.98940783661333e-06, | |
| "loss": 0.2802, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.4990138067061145, | |
| "grad_norm": 1.4324944544127631, | |
| "learning_rate": 2.940681505250742e-06, | |
| "loss": 0.2848, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.5187376725838264, | |
| "grad_norm": 1.4082984304420074, | |
| "learning_rate": 2.8918649218953624e-06, | |
| "loss": 0.2801, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 1.5895657718154816, | |
| "learning_rate": 2.84298119205983e-06, | |
| "loss": 0.2807, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.5581854043392505, | |
| "grad_norm": 1.6080440377232041, | |
| "learning_rate": 2.7940534530380666e-06, | |
| "loss": 0.2835, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.5779092702169626, | |
| "grad_norm": 1.404915797241871, | |
| "learning_rate": 2.7451048629541045e-06, | |
| "loss": 0.2808, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.5976331360946747, | |
| "grad_norm": 1.4879672080505235, | |
| "learning_rate": 2.6961585898010523e-06, | |
| "loss": 0.2806, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.6173570019723866, | |
| "grad_norm": 1.3888602093522253, | |
| "learning_rate": 2.647237800475384e-06, | |
| "loss": 0.2832, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.6370808678500985, | |
| "grad_norm": 1.3670120148082392, | |
| "learning_rate": 2.5983656498117525e-06, | |
| "loss": 0.2825, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.6568047337278107, | |
| "grad_norm": 1.2812642080517738, | |
| "learning_rate": 2.54956526962351e-06, | |
| "loss": 0.279, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.6765285996055228, | |
| "grad_norm": 1.252430854449729, | |
| "learning_rate": 2.5008597577541288e-06, | |
| "loss": 0.2814, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.6962524654832347, | |
| "grad_norm": 1.2750427994477165, | |
| "learning_rate": 2.45227216714469e-06, | |
| "loss": 0.2792, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.7159763313609466, | |
| "grad_norm": 1.354377403404739, | |
| "learning_rate": 2.403825494922636e-06, | |
| "loss": 0.282, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.7357001972386588, | |
| "grad_norm": 1.4267990848182481, | |
| "learning_rate": 2.3555426715169396e-06, | |
| "loss": 0.2791, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.755424063116371, | |
| "grad_norm": 1.252857555239978, | |
| "learning_rate": 2.3074465498048303e-06, | |
| "loss": 0.2826, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.7751479289940828, | |
| "grad_norm": 1.2876786054611615, | |
| "learning_rate": 2.259559894295244e-06, | |
| "loss": 0.2789, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.7948717948717947, | |
| "grad_norm": 1.2629901820145135, | |
| "learning_rate": 2.2119053703540866e-06, | |
| "loss": 0.2791, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.8145956607495068, | |
| "grad_norm": 1.3562733049556417, | |
| "learning_rate": 2.1645055334764237e-06, | |
| "loss": 0.2807, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.834319526627219, | |
| "grad_norm": 1.3132542320273741, | |
| "learning_rate": 2.1173828186106828e-06, | |
| "loss": 0.2782, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.854043392504931, | |
| "grad_norm": 1.372645351488049, | |
| "learning_rate": 2.0705595295399e-06, | |
| "loss": 0.28, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.873767258382643, | |
| "grad_norm": 1.286506818666612, | |
| "learning_rate": 2.0240578283250596e-06, | |
| "loss": 0.2788, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.893491124260355, | |
| "grad_norm": 1.343985774681719, | |
| "learning_rate": 1.9778997248155013e-06, | |
| "loss": 0.2779, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.913214990138067, | |
| "grad_norm": 1.3873943864064089, | |
| "learning_rate": 1.9321070662313824e-06, | |
| "loss": 0.2768, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.9329388560157792, | |
| "grad_norm": 1.3822544572854645, | |
| "learning_rate": 1.88670152682311e-06, | |
| "loss": 0.2753, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.952662721893491, | |
| "grad_norm": 1.3724554338840655, | |
| "learning_rate": 1.8417045976126347e-06, | |
| "loss": 0.274, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.972386587771203, | |
| "grad_norm": 1.428387339598408, | |
| "learning_rate": 1.797137576221482e-06, | |
| "loss": 0.2775, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9921104536489151, | |
| "grad_norm": 1.2370547509299645, | |
| "learning_rate": 1.753021556790314e-06, | |
| "loss": 0.2746, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.3482723832130432, | |
| "eval_runtime": 46.4255, | |
| "eval_samples_per_second": 293.804, | |
| "eval_steps_per_second": 1.163, | |
| "step": 1014 | |
| }, | |
| { | |
| "epoch": 2.0118343195266273, | |
| "grad_norm": 1.9502351693684774, | |
| "learning_rate": 1.7093774199948004e-06, | |
| "loss": 0.2309, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 2.0315581854043394, | |
| "grad_norm": 1.5862323859503984, | |
| "learning_rate": 1.6662258231625331e-06, | |
| "loss": 0.2026, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 2.051282051282051, | |
| "grad_norm": 1.3292614459089434, | |
| "learning_rate": 1.6235871904956431e-06, | |
| "loss": 0.2034, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 2.0710059171597632, | |
| "grad_norm": 1.2370582334736997, | |
| "learning_rate": 1.5814817034037715e-06, | |
| "loss": 0.2008, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.0907297830374754, | |
| "grad_norm": 1.325897622024457, | |
| "learning_rate": 1.5399292909519422e-06, | |
| "loss": 0.2042, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 2.1104536489151875, | |
| "grad_norm": 1.4548395791353137, | |
| "learning_rate": 1.4989496204278897e-06, | |
| "loss": 0.2025, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 2.1301775147928996, | |
| "grad_norm": 1.36179677292465, | |
| "learning_rate": 1.458562088033273e-06, | |
| "loss": 0.1978, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 2.1499013806706113, | |
| "grad_norm": 1.4589926591648759, | |
| "learning_rate": 1.4187858097032086e-06, | |
| "loss": 0.2024, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 2.1696252465483234, | |
| "grad_norm": 1.3095440667780154, | |
| "learning_rate": 1.3796396120584576e-06, | |
| "loss": 0.2032, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.1893491124260356, | |
| "grad_norm": 1.3522834520399176, | |
| "learning_rate": 1.341142023494537e-06, | |
| "loss": 0.1992, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 2.2090729783037477, | |
| "grad_norm": 1.3914925068585928, | |
| "learning_rate": 1.3033112654120032e-06, | |
| "loss": 0.2029, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 2.2287968441814594, | |
| "grad_norm": 1.2392072409116117, | |
| "learning_rate": 1.266165243592024e-06, | |
| "loss": 0.2019, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 2.2485207100591715, | |
| "grad_norm": 1.450828785906611, | |
| "learning_rate": 1.2297215397213442e-06, | |
| "loss": 0.2029, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 2.2682445759368837, | |
| "grad_norm": 1.3539897715774756, | |
| "learning_rate": 1.1939974030706499e-06, | |
| "loss": 0.1989, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.287968441814596, | |
| "grad_norm": 1.3124427663284721, | |
| "learning_rate": 1.1590097423302681e-06, | |
| "loss": 0.2013, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 1.2751387286158546, | |
| "learning_rate": 1.1247751176070688e-06, | |
| "loss": 0.2003, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 2.3274161735700196, | |
| "grad_norm": 1.2826788452929796, | |
| "learning_rate": 1.0913097325863526e-06, | |
| "loss": 0.2013, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.3471400394477318, | |
| "grad_norm": 1.3449233167779666, | |
| "learning_rate": 1.0586294268624391e-06, | |
| "loss": 0.2031, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.366863905325444, | |
| "grad_norm": 1.3034368496811286, | |
| "learning_rate": 1.026749668441587e-06, | |
| "loss": 0.1994, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.386587771203156, | |
| "grad_norm": 1.3565807097213252, | |
| "learning_rate": 9.956855464207873e-07, | |
| "loss": 0.2, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.4063116370808677, | |
| "grad_norm": 1.451004027193357, | |
| "learning_rate": 9.654517638459015e-07, | |
| "loss": 0.1996, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.42603550295858, | |
| "grad_norm": 1.3107553476519733, | |
| "learning_rate": 9.360626307525231e-07, | |
| "loss": 0.2004, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.445759368836292, | |
| "grad_norm": 1.2866100592193557, | |
| "learning_rate": 9.075320573928513e-07, | |
| "loss": 0.2026, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.465483234714004, | |
| "grad_norm": 1.3169876215045113, | |
| "learning_rate": 8.798735476517964e-07, | |
| "loss": 0.2027, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.485207100591716, | |
| "grad_norm": 1.2821201625196061, | |
| "learning_rate": 8.531001926554134e-07, | |
| "loss": 0.2011, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.504930966469428, | |
| "grad_norm": 1.315132765819279, | |
| "learning_rate": 8.272246645747072e-07, | |
| "loss": 0.199, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.52465483234714, | |
| "grad_norm": 1.276154658164099, | |
| "learning_rate": 8.022592106277332e-07, | |
| "loss": 0.2008, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.544378698224852, | |
| "grad_norm": 1.2274421062761773, | |
| "learning_rate": 7.782156472828299e-07, | |
| "loss": 0.1998, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.564102564102564, | |
| "grad_norm": 1.2435720383981574, | |
| "learning_rate": 7.551053546657356e-07, | |
| "loss": 0.1995, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.583826429980276, | |
| "grad_norm": 1.2327909078947592, | |
| "learning_rate": 7.329392711732278e-07, | |
| "loss": 0.2024, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.603550295857988, | |
| "grad_norm": 1.1783489485507048, | |
| "learning_rate": 7.117278882958421e-07, | |
| "loss": 0.2003, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.6232741617357003, | |
| "grad_norm": 1.2687230261577986, | |
| "learning_rate": 6.914812456521138e-07, | |
| "loss": 0.2006, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.6429980276134124, | |
| "grad_norm": 1.2646158919927277, | |
| "learning_rate": 6.722089262366993e-07, | |
| "loss": 0.1982, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.662721893491124, | |
| "grad_norm": 1.2236131305338422, | |
| "learning_rate": 6.539200518846226e-07, | |
| "loss": 0.2001, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.6824457593688362, | |
| "grad_norm": 1.2428023457207789, | |
| "learning_rate": 6.366232789537923e-07, | |
| "loss": 0.2048, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.7021696252465484, | |
| "grad_norm": 1.2559417256017682, | |
| "learning_rate": 6.203267942278395e-07, | |
| "loss": 0.2012, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.7218934911242605, | |
| "grad_norm": 1.2572564112264348, | |
| "learning_rate": 6.050383110412069e-07, | |
| "loss": 0.1994, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.7416173570019726, | |
| "grad_norm": 1.1764889460619852, | |
| "learning_rate": 5.907650656283289e-07, | |
| "loss": 0.2002, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.7613412228796843, | |
| "grad_norm": 1.2804661059833917, | |
| "learning_rate": 5.775138136986298e-07, | |
| "loss": 0.2002, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.7810650887573964, | |
| "grad_norm": 1.3077263435732718, | |
| "learning_rate": 5.652908272389604e-07, | |
| "loss": 0.1995, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.8007889546351086, | |
| "grad_norm": 1.231137370296971, | |
| "learning_rate": 5.541018915449863e-07, | |
| "loss": 0.1989, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.8205128205128203, | |
| "grad_norm": 1.3443797697665705, | |
| "learning_rate": 5.439523024829335e-07, | |
| "loss": 0.1983, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.8402366863905324, | |
| "grad_norm": 1.2092638219767884, | |
| "learning_rate": 5.348468639829871e-07, | |
| "loss": 0.2007, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.8599605522682445, | |
| "grad_norm": 1.2392545674361426, | |
| "learning_rate": 5.267898857655307e-07, | |
| "loss": 0.201, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.8796844181459567, | |
| "grad_norm": 1.255507262390408, | |
| "learning_rate": 5.19785181301299e-07, | |
| "loss": 0.2008, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.899408284023669, | |
| "grad_norm": 1.2545629120536586, | |
| "learning_rate": 5.138360660064146e-07, | |
| "loss": 0.1979, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.9191321499013805, | |
| "grad_norm": 1.2279624795193589, | |
| "learning_rate": 5.08945355673159e-07, | |
| "loss": 0.201, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.9388560157790926, | |
| "grad_norm": 1.2395946923655343, | |
| "learning_rate": 5.05115365137222e-07, | |
| "loss": 0.1999, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.9585798816568047, | |
| "grad_norm": 1.2212433583596156, | |
| "learning_rate": 5.023479071820607e-07, | |
| "loss": 0.1989, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.978303747534517, | |
| "grad_norm": 1.298954785136158, | |
| "learning_rate": 5.006442916808849e-07, | |
| "loss": 0.2019, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.998027613412229, | |
| "grad_norm": 1.3586461216494594, | |
| "learning_rate": 5.000053249766787e-07, | |
| "loss": 0.1999, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.37075862288475037, | |
| "eval_runtime": 53.9042, | |
| "eval_samples_per_second": 253.041, | |
| "eval_steps_per_second": 1.002, | |
| "step": 1521 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 1521, | |
| "total_flos": 2547731650314240.0, | |
| "train_loss": 0.2954053143131192, | |
| "train_runtime": 9131.3778, | |
| "train_samples_per_second": 85.143, | |
| "train_steps_per_second": 0.167 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1521, | |
| "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": 2547731650314240.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |