| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.994283294516197, | |
| "eval_steps": 500, | |
| "global_step": 5310, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01693838661867457, | |
| "grad_norm": 5.0625, | |
| "learning_rate": 1.8832391713747647e-07, | |
| "loss": 2.0482, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03387677323734914, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 3.7664783427495294e-07, | |
| "loss": 2.069, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.050815159856023716, | |
| "grad_norm": 5.96875, | |
| "learning_rate": 5.649717514124295e-07, | |
| "loss": 2.0861, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.06775354647469828, | |
| "grad_norm": 5.65625, | |
| "learning_rate": 7.532956685499059e-07, | |
| "loss": 2.0413, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.08469193309337286, | |
| "grad_norm": 16.125, | |
| "learning_rate": 9.416195856873824e-07, | |
| "loss": 2.0095, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.10163031971204743, | |
| "grad_norm": 17.75, | |
| "learning_rate": 1.129943502824859e-06, | |
| "loss": 2.0333, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.118568706330722, | |
| "grad_norm": 14.3125, | |
| "learning_rate": 1.3182674199623351e-06, | |
| "loss": 2.0062, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.13550709294939656, | |
| "grad_norm": 15.5625, | |
| "learning_rate": 1.5065913370998117e-06, | |
| "loss": 1.9655, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.15244547956807114, | |
| "grad_norm": 14.4375, | |
| "learning_rate": 1.6949152542372882e-06, | |
| "loss": 1.9698, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.16938386618674572, | |
| "grad_norm": 3.4375, | |
| "learning_rate": 1.8832391713747648e-06, | |
| "loss": 1.9211, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.18632225280542028, | |
| "grad_norm": 2.703125, | |
| "learning_rate": 2.071563088512241e-06, | |
| "loss": 1.9071, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.20326063942409486, | |
| "grad_norm": 3.453125, | |
| "learning_rate": 2.259887005649718e-06, | |
| "loss": 1.8766, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.22019902604276942, | |
| "grad_norm": 2.9375, | |
| "learning_rate": 2.448210922787194e-06, | |
| "loss": 1.8211, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.237137412661444, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 2.6365348399246702e-06, | |
| "loss": 1.7678, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.2540757992801186, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 2.8248587570621473e-06, | |
| "loss": 1.6984, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2710141858987931, | |
| "grad_norm": 2.671875, | |
| "learning_rate": 3.0131826741996235e-06, | |
| "loss": 1.6916, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2879525725174677, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 3.2015065913371e-06, | |
| "loss": 1.7083, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3048909591361423, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 3.3898305084745763e-06, | |
| "loss": 1.7021, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.32182934575481686, | |
| "grad_norm": 3.109375, | |
| "learning_rate": 3.5781544256120534e-06, | |
| "loss": 1.654, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.33876773237349145, | |
| "grad_norm": 1.875, | |
| "learning_rate": 3.7664783427495296e-06, | |
| "loss": 1.6125, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.355706118992166, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 3.954802259887006e-06, | |
| "loss": 1.6243, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.37264450561084056, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 4.143126177024482e-06, | |
| "loss": 1.6311, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.38958289222951514, | |
| "grad_norm": 2.375, | |
| "learning_rate": 4.331450094161959e-06, | |
| "loss": 1.6088, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.4065212788481897, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 4.519774011299436e-06, | |
| "loss": 1.6126, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.42345966546686425, | |
| "grad_norm": 2.0, | |
| "learning_rate": 4.7080979284369114e-06, | |
| "loss": 1.5862, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.44039805208553884, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 4.896421845574388e-06, | |
| "loss": 1.5869, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.4573364387042134, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 5.084745762711865e-06, | |
| "loss": 1.568, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.474274825322888, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 5.2730696798493405e-06, | |
| "loss": 1.4917, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4912132119415626, | |
| "grad_norm": 3.0, | |
| "learning_rate": 5.461393596986818e-06, | |
| "loss": 1.5614, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.5081515985602372, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 5.6497175141242946e-06, | |
| "loss": 1.5684, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5250899851789117, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 5.83804143126177e-06, | |
| "loss": 1.5933, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.5420283717975862, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 6.026365348399247e-06, | |
| "loss": 1.5864, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5589667584162609, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 6.2146892655367244e-06, | |
| "loss": 1.539, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5759051450349354, | |
| "grad_norm": 2.125, | |
| "learning_rate": 6.4030131826742e-06, | |
| "loss": 1.4972, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.59284353165361, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 6.591337099811677e-06, | |
| "loss": 1.5136, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6097819182722846, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 6.779661016949153e-06, | |
| "loss": 1.5159, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.6267203048909591, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 6.96798493408663e-06, | |
| "loss": 1.4829, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.6436586915096337, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 7.156308851224107e-06, | |
| "loss": 1.4972, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6605970781283083, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 7.3446327683615825e-06, | |
| "loss": 1.4115, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6775354647469829, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 7.532956685499059e-06, | |
| "loss": 1.5167, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6944738513656574, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 7.721280602636536e-06, | |
| "loss": 1.4496, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.711412237984332, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 7.909604519774012e-06, | |
| "loss": 1.4258, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.7283506246030066, | |
| "grad_norm": 3.28125, | |
| "learning_rate": 8.097928436911489e-06, | |
| "loss": 1.498, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.7452890112216811, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 8.286252354048965e-06, | |
| "loss": 1.4588, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.7622273978403558, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 8.47457627118644e-06, | |
| "loss": 1.4085, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7791657844590303, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 8.662900188323918e-06, | |
| "loss": 1.4479, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7961041710777048, | |
| "grad_norm": 2.25, | |
| "learning_rate": 8.851224105461394e-06, | |
| "loss": 1.4711, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.8130425576963795, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 9.039548022598871e-06, | |
| "loss": 1.4004, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.829980944315054, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 9.227871939736347e-06, | |
| "loss": 1.3785, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.8469193309337285, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 9.416195856873823e-06, | |
| "loss": 1.3784, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8469193309337285, | |
| "eval_loss": 1.401244044303894, | |
| "eval_runtime": 21.2893, | |
| "eval_samples_per_second": 24.66, | |
| "eval_steps_per_second": 24.66, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8638577175524031, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 9.6045197740113e-06, | |
| "loss": 1.4401, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8807961041710777, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 9.792843691148776e-06, | |
| "loss": 1.4191, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8977344907897523, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 9.981167608286254e-06, | |
| "loss": 1.3972, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.9146728774084268, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 9.999912491660088e-06, | |
| "loss": 1.4051, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.9316112640271014, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 9.999609997627497e-06, | |
| "loss": 1.3855, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.948549650645776, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 9.999091450621204e-06, | |
| "loss": 1.4033, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.9654880372644505, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 9.99835687304968e-06, | |
| "loss": 1.4429, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.9824264238831252, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 9.997406296656937e-06, | |
| "loss": 1.3925, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9993648105017997, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 9.996239762521152e-06, | |
| "loss": 1.4055, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.0163031971204743, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 9.994857321052887e-06, | |
| "loss": 1.2929, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.0332415837391489, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 9.993259031992921e-06, | |
| "loss": 1.2752, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.0501799703578234, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 9.991444964409665e-06, | |
| "loss": 1.3491, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.067118356976498, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 9.989415196696169e-06, | |
| "loss": 1.3204, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.0840567435951725, | |
| "grad_norm": 1.625, | |
| "learning_rate": 9.98716981656675e-06, | |
| "loss": 1.3425, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 1.1009951302138472, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 9.98470892105318e-06, | |
| "loss": 1.3116, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.1179335168325217, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 9.982032616500516e-06, | |
| "loss": 1.307, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 1.1348719034511963, | |
| "grad_norm": 1.875, | |
| "learning_rate": 9.979141018562494e-06, | |
| "loss": 1.3132, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.1518102900698708, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 9.976034252196524e-06, | |
| "loss": 1.3304, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 1.1687486766885455, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 9.9727124516583e-06, | |
| "loss": 1.347, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 1.18568706330722, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 9.969175760495997e-06, | |
| "loss": 1.2844, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.2026254499258946, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.965424331544064e-06, | |
| "loss": 1.2891, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.2195638365445691, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 9.961458326916624e-06, | |
| "loss": 1.3258, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.2365022231632437, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 9.95727791800046e-06, | |
| "loss": 1.2944, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.2534406097819182, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 9.952883285447622e-06, | |
| "loss": 1.2602, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.2703789964005927, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 9.948274619167608e-06, | |
| "loss": 1.2353, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.2873173830192675, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 9.943452118319165e-06, | |
| "loss": 1.2749, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.304255769637942, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 9.938415991301675e-06, | |
| "loss": 1.2432, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.3211941562566165, | |
| "grad_norm": 1.625, | |
| "learning_rate": 9.933166455746159e-06, | |
| "loss": 1.2824, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.3381325428752913, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 9.927703738505863e-06, | |
| "loss": 1.2335, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.3550709294939658, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 9.922028075646465e-06, | |
| "loss": 1.3065, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.3720093161126403, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 9.91613971243586e-06, | |
| "loss": 1.2443, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.3889477027313148, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 9.910038903333574e-06, | |
| "loss": 1.2303, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.4058860893499894, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 9.90372591197976e-06, | |
| "loss": 1.1844, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.422824475968664, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 9.89720101118381e-06, | |
| "loss": 1.2035, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.4397628625873384, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 9.890464482912561e-06, | |
| "loss": 1.2399, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.4567012492060132, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 9.88351661827811e-06, | |
| "loss": 1.227, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.4736396358246877, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 9.876357717525245e-06, | |
| "loss": 1.2589, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.4905780224433622, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 9.86898809001845e-06, | |
| "loss": 1.1935, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.507516409062037, | |
| "grad_norm": 1.5, | |
| "learning_rate": 9.861408054228556e-06, | |
| "loss": 1.2177, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.5244547956807115, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 9.853617937718966e-06, | |
| "loss": 1.2285, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.541393182299386, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 9.845618077131508e-06, | |
| "loss": 1.226, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 1.5583315689180606, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 9.837408818171884e-06, | |
| "loss": 1.2536, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 1.575269955536735, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 9.82899051559472e-06, | |
| "loss": 1.2041, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 1.5922083421554096, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 9.820363533188257e-06, | |
| "loss": 1.2279, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 1.6091467287740842, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 9.811528243758612e-06, | |
| "loss": 1.1829, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.6260851153927587, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 9.802485029113682e-06, | |
| "loss": 1.1698, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.6430235020114334, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 9.793234280046627e-06, | |
| "loss": 1.2212, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.659961888630108, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 9.783776396318998e-06, | |
| "loss": 1.1377, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.6769002752487827, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 9.77411178664346e-06, | |
| "loss": 1.2086, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.6938386618674572, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 9.76424086866612e-06, | |
| "loss": 1.1764, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6938386618674572, | |
| "eval_loss": 1.2227357625961304, | |
| "eval_runtime": 20.8763, | |
| "eval_samples_per_second": 25.148, | |
| "eval_steps_per_second": 25.148, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7107770484861318, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 9.754164068948482e-06, | |
| "loss": 1.1537, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.7277154351048063, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 9.743881822949025e-06, | |
| "loss": 1.2405, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.7446538217234808, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 9.733394575004373e-06, | |
| "loss": 1.2403, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.7615922083421554, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 9.722702778310096e-06, | |
| "loss": 1.152, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.7785305949608299, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 9.711806894901124e-06, | |
| "loss": 1.1596, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.7954689815795044, | |
| "grad_norm": 1.25, | |
| "learning_rate": 9.700707395631787e-06, | |
| "loss": 1.1555, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.8124073681981792, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 9.689404760155465e-06, | |
| "loss": 1.1611, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.8293457548168537, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 9.677899476903857e-06, | |
| "loss": 1.1727, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.8462841414355282, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 9.666192043065875e-06, | |
| "loss": 1.1526, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.863222528054203, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 9.654282964566159e-06, | |
| "loss": 1.1677, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.8801609146728775, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 9.642172756043217e-06, | |
| "loss": 1.1363, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.897099301291552, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 9.629861940827181e-06, | |
| "loss": 1.1216, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.9140376879102265, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 9.617351050917197e-06, | |
| "loss": 1.1432, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.930976074528901, | |
| "grad_norm": 1.0859375, | |
| "learning_rate": 9.604640626958428e-06, | |
| "loss": 1.1714, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.9479144611475756, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 9.591731218218694e-06, | |
| "loss": 1.1999, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.9648528477662501, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 9.57862338256474e-06, | |
| "loss": 1.1692, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.9817912343849249, | |
| "grad_norm": 1.125, | |
| "learning_rate": 9.565317686438122e-06, | |
| "loss": 1.1193, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.9987296210035994, | |
| "grad_norm": 0.92578125, | |
| "learning_rate": 9.551814704830734e-06, | |
| "loss": 1.1596, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 2.015668007622274, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 9.538115021259956e-06, | |
| "loss": 1.0715, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 2.0326063942409487, | |
| "grad_norm": 0.9375, | |
| "learning_rate": 9.52421922774344e-06, | |
| "loss": 1.0312, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.049544780859623, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 9.510127924773529e-06, | |
| "loss": 1.0574, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 2.0664831674782977, | |
| "grad_norm": 0.95703125, | |
| "learning_rate": 9.4958417212913e-06, | |
| "loss": 1.0862, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 2.0834215540969723, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 9.481361234660258e-06, | |
| "loss": 1.0776, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 2.100359940715647, | |
| "grad_norm": 0.95703125, | |
| "learning_rate": 9.466687090639652e-06, | |
| "loss": 1.088, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 2.1172983273343213, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 9.451819923357435e-06, | |
| "loss": 1.1108, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.134236713952996, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 9.436760375282858e-06, | |
| "loss": 1.0786, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 2.1511751005716704, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 9.421509097198715e-06, | |
| "loss": 1.0269, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 2.168113487190345, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 9.406066748173208e-06, | |
| "loss": 1.1046, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 2.18505187380902, | |
| "grad_norm": 0.9140625, | |
| "learning_rate": 9.390433995531473e-06, | |
| "loss": 1.0651, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 2.2019902604276944, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 9.374611514826746e-06, | |
| "loss": 1.0212, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.218928647046369, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 9.358599989811158e-06, | |
| "loss": 1.0786, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 2.2358670336650435, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 9.3424001124062e-06, | |
| "loss": 1.0125, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 2.252805420283718, | |
| "grad_norm": 0.90234375, | |
| "learning_rate": 9.326012582672814e-06, | |
| "loss": 1.0756, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 2.2697438069023925, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 9.309438108781143e-06, | |
| "loss": 1.1165, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 2.286682193521067, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 9.29267740697993e-06, | |
| "loss": 1.037, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.3036205801397416, | |
| "grad_norm": 1.25, | |
| "learning_rate": 9.275731201565564e-06, | |
| "loss": 1.03, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 2.320558966758416, | |
| "grad_norm": 0.93359375, | |
| "learning_rate": 9.258600224850777e-06, | |
| "loss": 1.0461, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 2.337497353377091, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 9.24128521713301e-06, | |
| "loss": 1.0065, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 2.3544357399957656, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 9.223786926662404e-06, | |
| "loss": 1.0661, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 2.37137412661444, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 9.20610610960948e-06, | |
| "loss": 1.0764, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.3883125132331147, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 9.188243530032458e-06, | |
| "loss": 1.0448, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 2.405250899851789, | |
| "grad_norm": 0.984375, | |
| "learning_rate": 9.170199959844232e-06, | |
| "loss": 1.0695, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 2.4221892864704637, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 9.15197617877902e-06, | |
| "loss": 0.9995, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 2.4391276730891382, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 9.133572974358669e-06, | |
| "loss": 1.0307, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 2.4560660597078128, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 9.114991141858619e-06, | |
| "loss": 1.0004, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.4730044463264873, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 9.096231484273535e-06, | |
| "loss": 1.0553, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 2.489942832945162, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 9.077294812282615e-06, | |
| "loss": 1.0481, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 2.5068812195638364, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 9.058181944214547e-06, | |
| "loss": 1.0438, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 2.523819606182511, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 9.038893706012153e-06, | |
| "loss": 1.0458, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 2.5407579928011854, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 9.01943093119669e-06, | |
| "loss": 0.9808, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.5407579928011854, | |
| "eval_loss": 1.1499994993209839, | |
| "eval_runtime": 21.1383, | |
| "eval_samples_per_second": 24.836, | |
| "eval_steps_per_second": 24.836, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.5576963794198604, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 8.999794460831836e-06, | |
| "loss": 1.0539, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 2.574634766038535, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 8.979985143487343e-06, | |
| "loss": 1.046, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 2.5915731526572094, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 8.960003835202369e-06, | |
| "loss": 1.0188, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 2.608511539275884, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 8.939851399448478e-06, | |
| "loss": 1.0423, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 2.6254499258945585, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 8.919528707092333e-06, | |
| "loss": 1.085, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.642388312513233, | |
| "grad_norm": 2.859375, | |
| "learning_rate": 8.899036636358058e-06, | |
| "loss": 1.0252, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 2.6593266991319076, | |
| "grad_norm": 3.125, | |
| "learning_rate": 8.878376072789292e-06, | |
| "loss": 1.0246, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 2.6762650857505825, | |
| "grad_norm": 2.78125, | |
| "learning_rate": 8.857547909210913e-06, | |
| "loss": 1.0358, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 2.693203472369257, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 8.836553045690462e-06, | |
| "loss": 1.0746, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 2.7101418589879316, | |
| "grad_norm": 3.96875, | |
| "learning_rate": 8.81539238949925e-06, | |
| "loss": 1.0806, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.727080245606606, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 8.794066855073136e-06, | |
| "loss": 0.9917, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 2.7440186322252806, | |
| "grad_norm": 4.5625, | |
| "learning_rate": 8.772577363973032e-06, | |
| "loss": 1.0071, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 2.760957018843955, | |
| "grad_norm": 3.25, | |
| "learning_rate": 8.750924844845065e-06, | |
| "loss": 1.0139, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 2.7778954054626297, | |
| "grad_norm": 4.25, | |
| "learning_rate": 8.729110233380445e-06, | |
| "loss": 1.0684, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 2.7948337920813042, | |
| "grad_norm": 9.125, | |
| "learning_rate": 8.707134472275042e-06, | |
| "loss": 0.9606, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.8117721786999788, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 8.684998511188634e-06, | |
| "loss": 1.0529, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 2.8287105653186533, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 8.662703306703878e-06, | |
| "loss": 1.0939, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 2.845648951937328, | |
| "grad_norm": 8.375, | |
| "learning_rate": 8.640249822284975e-06, | |
| "loss": 1.0322, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 2.8625873385560023, | |
| "grad_norm": 7.40625, | |
| "learning_rate": 8.617639028236021e-06, | |
| "loss": 1.0303, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 2.879525725174677, | |
| "grad_norm": 2.625, | |
| "learning_rate": 8.59487190165909e-06, | |
| "loss": 1.0173, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.896464111793352, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 8.571949426412002e-06, | |
| "loss": 1.0218, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 2.9134024984120264, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 8.54887259306581e-06, | |
| "loss": 1.0029, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 2.930340885030701, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 8.525642398861997e-06, | |
| "loss": 1.0295, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 2.9472792716493754, | |
| "grad_norm": 2.5, | |
| "learning_rate": 8.50225984766937e-06, | |
| "loss": 1.0432, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 2.96421765826805, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 8.478725949940687e-06, | |
| "loss": 1.026, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.9811560448867245, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 8.455041722668993e-06, | |
| "loss": 1.0523, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 2.998094431505399, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 8.43120818934367e-06, | |
| "loss": 1.0287, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 3.0150328181240735, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 8.407226379906197e-06, | |
| "loss": 0.9998, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 3.031971204742748, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 8.383097330705664e-06, | |
| "loss": 1.0157, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 3.048909591361423, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 8.358822084453964e-06, | |
| "loss": 0.9514, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.0658479779800976, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 8.334401690180755e-06, | |
| "loss": 1.0163, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 3.082786364598772, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 8.30983720318811e-06, | |
| "loss": 0.969, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 3.0997247512174466, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 8.285129685004922e-06, | |
| "loss": 0.9769, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 3.116663137836121, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 8.260280203341026e-06, | |
| "loss": 0.9997, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 3.1336015244547957, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 8.235289832041066e-06, | |
| "loss": 0.9577, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.15053991107347, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 8.210159651038081e-06, | |
| "loss": 0.9457, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 3.1674782976921447, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 8.184890746306848e-06, | |
| "loss": 0.9381, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 3.1844166843108193, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 8.15948420981694e-06, | |
| "loss": 0.9709, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 3.201355070929494, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 8.133941139485551e-06, | |
| "loss": 0.9769, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 3.2182934575481683, | |
| "grad_norm": 1.875, | |
| "learning_rate": 8.108262639130038e-06, | |
| "loss": 0.9064, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.2352318441668433, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 8.082449818420227e-06, | |
| "loss": 0.9944, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 3.252170230785518, | |
| "grad_norm": 2.125, | |
| "learning_rate": 8.056503792830458e-06, | |
| "loss": 0.9484, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 3.2691086174041923, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 8.030425683591391e-06, | |
| "loss": 0.9666, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 3.286047004022867, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 8.004216617641532e-06, | |
| "loss": 0.9871, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 3.3029853906415414, | |
| "grad_norm": 1.765625, | |
| "learning_rate": 7.977877727578556e-06, | |
| "loss": 0.9462, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.319923777260216, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 7.951410151610343e-06, | |
| "loss": 0.9509, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 3.3368621638788905, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 7.92481503350581e-06, | |
| "loss": 0.9299, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 3.353800550497565, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.898093522545471e-06, | |
| "loss": 0.9824, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 3.3707389371162395, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 7.871246773471779e-06, | |
| "loss": 0.9336, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 3.3876773237349145, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 7.844275946439216e-06, | |
| "loss": 0.9778, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.3876773237349145, | |
| "eval_loss": 1.1204777956008911, | |
| "eval_runtime": 20.947, | |
| "eval_samples_per_second": 25.063, | |
| "eval_steps_per_second": 25.063, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.404615710353589, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 7.817182206964177e-06, | |
| "loss": 0.9817, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 3.4215540969722635, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.789966725874578e-06, | |
| "loss": 0.9942, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 3.438492483590938, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.76263067925928e-06, | |
| "loss": 0.9783, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 3.4554308702096126, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 7.735175248417255e-06, | |
| "loss": 0.9551, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 3.472369256828287, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 7.707601619806547e-06, | |
| "loss": 0.9592, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 3.4893076434469616, | |
| "grad_norm": 2.40625, | |
| "learning_rate": 7.679910984992988e-06, | |
| "loss": 0.902, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 3.506246030065636, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 7.652104540598712e-06, | |
| "loss": 0.9763, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 3.5231844166843107, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 7.6241834882504516e-06, | |
| "loss": 0.9522, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 3.5401228033029852, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 7.596149034527594e-06, | |
| "loss": 0.992, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 3.5570611899216598, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 7.5680023909100545e-06, | |
| "loss": 0.9366, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.5739995765403343, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 7.539744773725916e-06, | |
| "loss": 0.9888, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 3.5909379631590093, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 7.511377404098876e-06, | |
| "loss": 0.9801, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 3.607876349777684, | |
| "grad_norm": 2.0, | |
| "learning_rate": 7.482901507895463e-06, | |
| "loss": 0.9589, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 3.6248147363963583, | |
| "grad_norm": 3.65625, | |
| "learning_rate": 7.454318315672074e-06, | |
| "loss": 0.9926, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 3.641753123015033, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 7.425629062621784e-06, | |
| "loss": 0.9448, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 3.6586915096337074, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 7.396834988520989e-06, | |
| "loss": 0.9507, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 3.675629896252382, | |
| "grad_norm": 2.125, | |
| "learning_rate": 7.36793733767581e-06, | |
| "loss": 0.9719, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 3.6925682828710564, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 7.338937358868334e-06, | |
| "loss": 0.976, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 3.709506669489731, | |
| "grad_norm": 2.28125, | |
| "learning_rate": 7.309836305302638e-06, | |
| "loss": 0.9431, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 3.726445056108406, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 7.2806354345506525e-06, | |
| "loss": 0.9833, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.7433834427270805, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 7.251336008497793e-06, | |
| "loss": 0.9356, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 3.760321829345755, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 7.221939293288445e-06, | |
| "loss": 0.9911, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 3.7772602159644295, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 7.192446559271247e-06, | |
| "loss": 0.9569, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 3.794198602583104, | |
| "grad_norm": 1.875, | |
| "learning_rate": 7.162859080944182e-06, | |
| "loss": 0.9907, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 3.8111369892017786, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 7.133178136899522e-06, | |
| "loss": 0.9672, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 3.828075375820453, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 7.103405009768554e-06, | |
| "loss": 0.9754, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 3.8450137624391276, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 7.073540986166166e-06, | |
| "loss": 0.9548, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 3.861952149057802, | |
| "grad_norm": 1.7421875, | |
| "learning_rate": 7.0435873566352434e-06, | |
| "loss": 0.9682, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 3.8788905356764767, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 7.013545415590897e-06, | |
| "loss": 0.8956, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 3.895828922295151, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 6.983416461264527e-06, | |
| "loss": 0.9314, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.9127673089138257, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 6.953201795647728e-06, | |
| "loss": 0.9101, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 3.9297056955325003, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 6.922902724436016e-06, | |
| "loss": 0.9463, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 3.9466440821511752, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 6.892520556972404e-06, | |
| "loss": 0.9443, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 3.9635824687698498, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 6.86205660619083e-06, | |
| "loss": 0.9371, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 3.9805208553885243, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 6.8315121885594105e-06, | |
| "loss": 0.95, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.997459242007199, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 6.800888624023552e-06, | |
| "loss": 0.9412, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 4.014397628625874, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 6.770187235948916e-06, | |
| "loss": 0.8627, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 4.031336015244548, | |
| "grad_norm": 1.671875, | |
| "learning_rate": 6.739409351064225e-06, | |
| "loss": 0.8663, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 4.048274401863223, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 6.708556299403935e-06, | |
| "loss": 0.8087, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 4.065212788481897, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 6.6776294142507535e-06, | |
| "loss": 0.8501, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 4.082151175100572, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 6.646630032078024e-06, | |
| "loss": 0.854, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 4.099089561719246, | |
| "grad_norm": 1.5, | |
| "learning_rate": 6.615559492491981e-06, | |
| "loss": 0.884, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 4.116027948337921, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 6.584419138173846e-06, | |
| "loss": 0.8693, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 4.1329663349565955, | |
| "grad_norm": 1.625, | |
| "learning_rate": 6.553210314821814e-06, | |
| "loss": 0.8397, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 4.14990472157527, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 6.521934371092901e-06, | |
| "loss": 0.8355, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 4.1668431081939445, | |
| "grad_norm": 1.5, | |
| "learning_rate": 6.490592658544661e-06, | |
| "loss": 0.8784, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 4.183781494812619, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 6.459186531576772e-06, | |
| "loss": 0.8902, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 4.200719881431294, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 6.427717347372523e-06, | |
| "loss": 0.8175, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 4.217658268049968, | |
| "grad_norm": 1.7109375, | |
| "learning_rate": 6.396186465840152e-06, | |
| "loss": 0.8717, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 4.234596654668643, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 6.364595249554084e-06, | |
| "loss": 0.8815, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.234596654668643, | |
| "eval_loss": 1.09398353099823, | |
| "eval_runtime": 20.9341, | |
| "eval_samples_per_second": 25.079, | |
| "eval_steps_per_second": 25.079, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.251535041287317, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 6.332945063696049e-06, | |
| "loss": 0.8658, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 4.268473427905992, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 6.3012372759960825e-06, | |
| "loss": 0.8465, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 4.285411814524666, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 6.26947325667343e-06, | |
| "loss": 0.8707, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 4.302350201143341, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 6.237654378377324e-06, | |
| "loss": 0.8921, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 4.319288587762015, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 6.2057820161276715e-06, | |
| "loss": 0.8703, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 4.33622697438069, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 6.173857547255634e-06, | |
| "loss": 0.8467, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 4.353165360999364, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 6.141882351344106e-06, | |
| "loss": 0.8581, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 4.37010374761804, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 6.1098578101680986e-06, | |
| "loss": 0.8617, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 4.387042134236714, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 6.077785307635027e-06, | |
| "loss": 0.8082, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 4.403980520855389, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 6.045666229724909e-06, | |
| "loss": 0.8048, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.420918907474063, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 6.013501964430468e-06, | |
| "loss": 0.8537, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 4.437857294092738, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 5.981293901697153e-06, | |
| "loss": 0.8683, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 4.454795680711412, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 5.9490434333630735e-06, | |
| "loss": 0.8599, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 4.471734067330087, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 5.916751953098854e-06, | |
| "loss": 0.8811, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 4.4886724539487615, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 5.884420856347406e-06, | |
| "loss": 0.8891, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 4.505610840567436, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 5.852051540263628e-06, | |
| "loss": 0.8634, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 4.5225492271861105, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 5.819645403654027e-06, | |
| "loss": 0.842, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 4.539487613804785, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 5.787203846916273e-06, | |
| "loss": 0.8485, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 4.55642600042346, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 5.754728271978675e-06, | |
| "loss": 0.883, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 4.573364387042134, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 5.722220082239608e-06, | |
| "loss": 0.8457, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.590302773660809, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 5.689680682506861e-06, | |
| "loss": 0.8999, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 4.607241160279483, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 5.657111478936926e-06, | |
| "loss": 0.8391, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 4.624179546898158, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 5.6245138789742444e-06, | |
| "loss": 0.8313, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 4.641117933516832, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 5.591889291290373e-06, | |
| "loss": 0.8709, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 4.658056320135507, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 5.559239125723119e-06, | |
| "loss": 0.8507, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 4.674994706754182, | |
| "grad_norm": 0.96484375, | |
| "learning_rate": 5.5265647932156066e-06, | |
| "loss": 0.9057, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 4.691933093372857, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 5.493867705755312e-06, | |
| "loss": 0.8386, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 4.708871479991531, | |
| "grad_norm": 1.0, | |
| "learning_rate": 5.461149276313046e-06, | |
| "loss": 0.9338, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 4.725809866610206, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 5.428410918781885e-06, | |
| "loss": 0.8176, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 4.74274825322888, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 5.395654047916083e-06, | |
| "loss": 0.8609, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.759686639847555, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 5.362880079269921e-06, | |
| "loss": 0.8494, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 4.776625026466229, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 5.330090429136553e-06, | |
| "loss": 0.8254, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 4.793563413084904, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 5.297286514486785e-06, | |
| "loss": 0.8689, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 4.810501799703578, | |
| "grad_norm": 1.125, | |
| "learning_rate": 5.26446975290785e-06, | |
| "loss": 0.8483, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 4.827440186322253, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 5.23164156254215e-06, | |
| "loss": 0.8436, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 4.844378572940927, | |
| "grad_norm": 0.91796875, | |
| "learning_rate": 5.19880336202597e-06, | |
| "loss": 0.8584, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 4.861316959559602, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 5.165956570428173e-06, | |
| "loss": 0.8122, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 4.8782553461782765, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 5.133102607188875e-06, | |
| "loss": 0.8705, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 4.895193732796951, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 5.100242892058108e-06, | |
| "loss": 0.8788, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 4.9121321194156256, | |
| "grad_norm": 1.015625, | |
| "learning_rate": 5.067378845034468e-06, | |
| "loss": 0.8663, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.9290705060343, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 5.034511886303746e-06, | |
| "loss": 0.8706, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 4.946008892652975, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 5.00164343617756e-06, | |
| "loss": 0.845, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 4.962947279271649, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 4.968774915031982e-06, | |
| "loss": 0.8165, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 4.979885665890324, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 4.9359077432461455e-06, | |
| "loss": 0.8719, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 4.996824052508998, | |
| "grad_norm": 0.88671875, | |
| "learning_rate": 4.903043341140879e-06, | |
| "loss": 0.8579, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 5.013762439127673, | |
| "grad_norm": 0.921875, | |
| "learning_rate": 4.870183128917318e-06, | |
| "loss": 0.7986, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 5.030700825746347, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 4.837328526595539e-06, | |
| "loss": 0.8119, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 5.047639212365023, | |
| "grad_norm": 0.97265625, | |
| "learning_rate": 4.804480953953193e-06, | |
| "loss": 0.8013, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 5.064577598983697, | |
| "grad_norm": 0.9296875, | |
| "learning_rate": 4.771641830464146e-06, | |
| "loss": 0.7954, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 5.081515985602372, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 4.738812575237148e-06, | |
| "loss": 0.8159, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.081515985602372, | |
| "eval_loss": 1.0747902393341064, | |
| "eval_runtime": 20.9737, | |
| "eval_samples_per_second": 25.031, | |
| "eval_steps_per_second": 25.031, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 5.098454372221046, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 4.705994606954497e-06, | |
| "loss": 0.7888, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 5.115392758839721, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 4.673189343810747e-06, | |
| "loss": 0.8088, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 5.132331145458395, | |
| "grad_norm": 1.0, | |
| "learning_rate": 4.640398203451402e-06, | |
| "loss": 0.7762, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 5.14926953207707, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 4.607622602911668e-06, | |
| "loss": 0.8108, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 5.166207918695744, | |
| "grad_norm": 1.1171875, | |
| "learning_rate": 4.574863958555215e-06, | |
| "loss": 0.7981, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 5.183146305314419, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 4.542123686012969e-06, | |
| "loss": 0.8295, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 5.200084691933093, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 4.509403200121936e-06, | |
| "loss": 0.7986, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 5.217023078551768, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 4.47670391486406e-06, | |
| "loss": 0.792, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 5.2339614651704425, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 4.44402724330513e-06, | |
| "loss": 0.8072, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 5.250899851789117, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 4.4113745975337e-06, | |
| "loss": 0.7877, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 5.2678382384077915, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 4.378747388600075e-06, | |
| "loss": 0.7823, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 5.284776625026466, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 4.346147026455334e-06, | |
| "loss": 0.7957, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 5.301715011645141, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 4.313574919890402e-06, | |
| "loss": 0.7562, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 5.318653398263815, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 4.281032476475165e-06, | |
| "loss": 0.7981, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 5.33559178488249, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 4.248521102497649e-06, | |
| "loss": 0.7806, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 5.352530171501164, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 4.216042202903246e-06, | |
| "loss": 0.848, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 5.369468558119839, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 4.183597181234001e-06, | |
| "loss": 0.7903, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 5.386406944738514, | |
| "grad_norm": 2.515625, | |
| "learning_rate": 4.151187439567961e-06, | |
| "loss": 0.7914, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 5.403345331357189, | |
| "grad_norm": 2.25, | |
| "learning_rate": 4.118814378458582e-06, | |
| "loss": 0.7983, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 5.420283717975863, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 4.086479396874206e-06, | |
| "loss": 0.7606, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.437222104594538, | |
| "grad_norm": 2.828125, | |
| "learning_rate": 4.054183892137611e-06, | |
| "loss": 0.7947, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 5.454160491213212, | |
| "grad_norm": 4.375, | |
| "learning_rate": 4.021929259865624e-06, | |
| "loss": 0.7764, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 5.471098877831887, | |
| "grad_norm": 4.46875, | |
| "learning_rate": 3.989716893908807e-06, | |
| "loss": 0.7829, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 5.488037264450561, | |
| "grad_norm": 3.5625, | |
| "learning_rate": 3.9575481862912335e-06, | |
| "loss": 0.7811, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 5.504975651069236, | |
| "grad_norm": 8.125, | |
| "learning_rate": 3.925424527150321e-06, | |
| "loss": 0.804, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 5.52191403768791, | |
| "grad_norm": 7.96875, | |
| "learning_rate": 3.893347304676766e-06, | |
| "loss": 0.7975, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 5.538852424306585, | |
| "grad_norm": 7.6875, | |
| "learning_rate": 3.861317905054551e-06, | |
| "loss": 0.7812, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 5.555790810925259, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 3.8293377124010476e-06, | |
| "loss": 0.8092, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 5.572729197543934, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 3.797408108707193e-06, | |
| "loss": 0.8141, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 5.5896675841626084, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.7655304737777775e-06, | |
| "loss": 0.815, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 5.606605970781283, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 3.733706185171819e-06, | |
| "loss": 0.7505, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 5.6235443573999575, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 3.701936618143025e-06, | |
| "loss": 0.7535, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 5.640482744018632, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 3.6702231455803662e-06, | |
| "loss": 0.7703, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 5.657421130637307, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.6385671379487496e-06, | |
| "loss": 0.7493, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 5.674359517255981, | |
| "grad_norm": 1.875, | |
| "learning_rate": 3.606969963229798e-06, | |
| "loss": 0.8281, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 5.691297903874656, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 3.5754329868627256e-06, | |
| "loss": 0.8207, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 5.70823629049333, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 3.5439575716853414e-06, | |
| "loss": 0.8263, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 5.725174677112005, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 3.5125450778751453e-06, | |
| "loss": 0.8106, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 5.742113063730679, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.4811968628905646e-06, | |
| "loss": 0.7993, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 5.759051450349354, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 3.4499142814122766e-06, | |
| "loss": 0.7784, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.775989836968029, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 3.418698685284676e-06, | |
| "loss": 0.7762, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 5.792928223586704, | |
| "grad_norm": 1.8984375, | |
| "learning_rate": 3.387551423457456e-06, | |
| "loss": 0.7905, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 5.809866610205378, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 3.356473841927313e-06, | |
| "loss": 0.8273, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 5.826804996824053, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 3.3254672836797814e-06, | |
| "loss": 0.7962, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 5.843743383442727, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 3.2945330886311964e-06, | |
| "loss": 0.8146, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 5.860681770061402, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 3.263672593570797e-06, | |
| "loss": 0.8032, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 5.877620156680076, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 3.23288713210295e-06, | |
| "loss": 0.7997, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 5.894558543298751, | |
| "grad_norm": 1.8828125, | |
| "learning_rate": 3.2021780345895216e-06, | |
| "loss": 0.8009, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 5.911496929917425, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 3.1715466280923924e-06, | |
| "loss": 0.8302, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 5.9284353165361, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 3.1409942363161037e-06, | |
| "loss": 0.8317, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.9284353165361, | |
| "eval_loss": 1.0828534364700317, | |
| "eval_runtime": 21.0141, | |
| "eval_samples_per_second": 24.983, | |
| "eval_steps_per_second": 24.983, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.945373703154774, | |
| "grad_norm": 2.125, | |
| "learning_rate": 3.1105221795506584e-06, | |
| "loss": 0.8026, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 5.962312089773449, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 3.080131774614464e-06, | |
| "loss": 0.8209, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 5.9792504763921235, | |
| "grad_norm": 1.8515625, | |
| "learning_rate": 3.049824334797431e-06, | |
| "loss": 0.8086, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 5.996188863010798, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 3.019601169804216e-06, | |
| "loss": 0.7651, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 6.0131272496294725, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 2.9894635856976268e-06, | |
| "loss": 0.7734, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 6.030065636248147, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 2.959412884842181e-06, | |
| "loss": 0.7799, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 6.047004022866822, | |
| "grad_norm": 1.921875, | |
| "learning_rate": 2.929450365847828e-06, | |
| "loss": 0.7448, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 6.063942409485496, | |
| "grad_norm": 2.0, | |
| "learning_rate": 2.899577323513829e-06, | |
| "loss": 0.7354, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 6.080880796104171, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.8697950487728e-06, | |
| "loss": 0.7585, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 6.097819182722846, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 2.8401048286349353e-06, | |
| "loss": 0.7635, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 6.114757569341521, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 2.8105079461323818e-06, | |
| "loss": 0.7781, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 6.131695955960195, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 2.7810056802637995e-06, | |
| "loss": 0.7646, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 6.14863434257887, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 2.751599305939082e-06, | |
| "loss": 0.7657, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 6.165572729197544, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.7222900939242787e-06, | |
| "loss": 0.7932, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 6.182511115816219, | |
| "grad_norm": 2.203125, | |
| "learning_rate": 2.693079310786661e-06, | |
| "loss": 0.7611, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 6.199449502434893, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.6639682188400017e-06, | |
| "loss": 0.7631, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 6.216387889053568, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.6349580760900232e-06, | |
| "loss": 0.7784, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 6.233326275672242, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 2.6060501361800333e-06, | |
| "loss": 0.7479, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 6.250264662290917, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.57724564833675e-06, | |
| "loss": 0.7367, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 6.267203048909591, | |
| "grad_norm": 1.875, | |
| "learning_rate": 2.5485458573163114e-06, | |
| "loss": 0.7463, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 6.284141435528266, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 2.519952003350499e-06, | |
| "loss": 0.7557, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 6.30107982214694, | |
| "grad_norm": 2.0, | |
| "learning_rate": 2.4914653220931283e-06, | |
| "loss": 0.8084, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 6.318018208765615, | |
| "grad_norm": 1.875, | |
| "learning_rate": 2.4630870445666576e-06, | |
| "loss": 0.7493, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 6.3349565953842895, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.434818397108993e-06, | |
| "loss": 0.7959, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 6.351894982002964, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.406660601320489e-06, | |
| "loss": 0.7245, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 6.3688333686216385, | |
| "grad_norm": 2.25, | |
| "learning_rate": 2.378614874011162e-06, | |
| "loss": 0.7637, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 6.385771755240313, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.3506824271480987e-06, | |
| "loss": 0.7702, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 6.402710141858988, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.322864467803097e-06, | |
| "loss": 0.7155, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 6.419648528477662, | |
| "grad_norm": 2.015625, | |
| "learning_rate": 2.2951621981004873e-06, | |
| "loss": 0.7925, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 6.436586915096337, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 2.2675768151651993e-06, | |
| "loss": 0.7901, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.453525301715012, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 2.240109511071014e-06, | |
| "loss": 0.729, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 6.470463688333687, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 2.2127614727890627e-06, | |
| "loss": 0.7481, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 6.487402074952361, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 2.1855338821365302e-06, | |
| "loss": 0.7148, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 6.504340461571036, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 2.158427915725573e-06, | |
| "loss": 0.7748, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 6.52127884818971, | |
| "grad_norm": 2.125, | |
| "learning_rate": 2.131444744912492e-06, | |
| "loss": 0.7601, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 6.538217234808385, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 2.1045855357470944e-06, | |
| "loss": 0.7397, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 6.555155621427059, | |
| "grad_norm": 2.578125, | |
| "learning_rate": 2.0778514489223207e-06, | |
| "loss": 0.7072, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 6.572094008045734, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 2.051243639724071e-06, | |
| "loss": 0.7204, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 6.589032394664408, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 2.024763257981296e-06, | |
| "loss": 0.7579, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 6.605970781283083, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.9984114480162976e-06, | |
| "loss": 0.7411, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 6.622909167901757, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.9721893485952785e-06, | |
| "loss": 0.7513, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 6.639847554520432, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.9460980928791407e-06, | |
| "loss": 0.7629, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 6.656785941139106, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.9201388083745036e-06, | |
| "loss": 0.767, | |
| "step": 3930 | |
| }, | |
| { | |
| "epoch": 6.673724327757781, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 1.8943126168849947e-06, | |
| "loss": 0.731, | |
| "step": 3940 | |
| }, | |
| { | |
| "epoch": 6.690662714376455, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 1.868620634462756e-06, | |
| "loss": 0.7335, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 6.70760110099513, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.8430639713602317e-06, | |
| "loss": 0.7791, | |
| "step": 3960 | |
| }, | |
| { | |
| "epoch": 6.7245394876138045, | |
| "grad_norm": 1.9921875, | |
| "learning_rate": 1.8176437319821789e-06, | |
| "loss": 0.7231, | |
| "step": 3970 | |
| }, | |
| { | |
| "epoch": 6.741477874232479, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.79236101483794e-06, | |
| "loss": 0.7355, | |
| "step": 3980 | |
| }, | |
| { | |
| "epoch": 6.7584162608511535, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.7672169124939843e-06, | |
| "loss": 0.7737, | |
| "step": 3990 | |
| }, | |
| { | |
| "epoch": 6.775354647469829, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.7422125115266785e-06, | |
| "loss": 0.7269, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.775354647469829, | |
| "eval_loss": 1.0812190771102905, | |
| "eval_runtime": 21.7625, | |
| "eval_samples_per_second": 24.124, | |
| "eval_steps_per_second": 24.124, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.7922930340885035, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 1.7173488924753451e-06, | |
| "loss": 0.7756, | |
| "step": 4010 | |
| }, | |
| { | |
| "epoch": 6.809231420707178, | |
| "grad_norm": 1.6953125, | |
| "learning_rate": 1.6926271297955549e-06, | |
| "loss": 0.7597, | |
| "step": 4020 | |
| }, | |
| { | |
| "epoch": 6.8261698073258525, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.668048291812709e-06, | |
| "loss": 0.7463, | |
| "step": 4030 | |
| }, | |
| { | |
| "epoch": 6.843108193944527, | |
| "grad_norm": 3.25, | |
| "learning_rate": 1.6436134406758635e-06, | |
| "loss": 0.7202, | |
| "step": 4040 | |
| }, | |
| { | |
| "epoch": 6.860046580563202, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.6193236323118283e-06, | |
| "loss": 0.7765, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 6.876984967181876, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 1.5951799163795461e-06, | |
| "loss": 0.7552, | |
| "step": 4060 | |
| }, | |
| { | |
| "epoch": 6.893923353800551, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.5711833362247204e-06, | |
| "loss": 0.7565, | |
| "step": 4070 | |
| }, | |
| { | |
| "epoch": 6.910861740419225, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.547334928834739e-06, | |
| "loss": 0.7342, | |
| "step": 4080 | |
| }, | |
| { | |
| "epoch": 6.9278001270379, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 1.523635724793852e-06, | |
| "loss": 0.7602, | |
| "step": 4090 | |
| }, | |
| { | |
| "epoch": 6.944738513656574, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 1.5000867482386478e-06, | |
| "loss": 0.7839, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 6.961676900275249, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.4766890168137816e-06, | |
| "loss": 0.7512, | |
| "step": 4110 | |
| }, | |
| { | |
| "epoch": 6.978615286893923, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.4534435416280135e-06, | |
| "loss": 0.7715, | |
| "step": 4120 | |
| }, | |
| { | |
| "epoch": 6.995553673512598, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.4303513272105057e-06, | |
| "loss": 0.7924, | |
| "step": 4130 | |
| }, | |
| { | |
| "epoch": 7.012492060131272, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 1.4074133714674116e-06, | |
| "loss": 0.7509, | |
| "step": 4140 | |
| }, | |
| { | |
| "epoch": 7.029430446749947, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 1.3846306656387603e-06, | |
| "loss": 0.7349, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 7.046368833368621, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.3620041942556127e-06, | |
| "loss": 0.7525, | |
| "step": 4160 | |
| }, | |
| { | |
| "epoch": 7.063307219987296, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.339534935097524e-06, | |
| "loss": 0.7515, | |
| "step": 4170 | |
| }, | |
| { | |
| "epoch": 7.0802456066059705, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 1.3172238591502794e-06, | |
| "loss": 0.7433, | |
| "step": 4180 | |
| }, | |
| { | |
| "epoch": 7.097183993224645, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.2950719305639464e-06, | |
| "loss": 0.7416, | |
| "step": 4190 | |
| }, | |
| { | |
| "epoch": 7.1141223798433195, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 1.2730801066112024e-06, | |
| "loss": 0.7115, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 7.131060766461994, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.251249337645965e-06, | |
| "loss": 0.6892, | |
| "step": 4210 | |
| }, | |
| { | |
| "epoch": 7.1479991530806695, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 1.229580567062334e-06, | |
| "loss": 0.704, | |
| "step": 4220 | |
| }, | |
| { | |
| "epoch": 7.164937539699344, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.2080747312538082e-06, | |
| "loss": 0.7021, | |
| "step": 4230 | |
| }, | |
| { | |
| "epoch": 7.1818759263180185, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.1867327595728373e-06, | |
| "loss": 0.6669, | |
| "step": 4240 | |
| }, | |
| { | |
| "epoch": 7.198814312936693, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.1655555742906438e-06, | |
| "loss": 0.7433, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 7.215752699555368, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.1445440905573857e-06, | |
| "loss": 0.7294, | |
| "step": 4260 | |
| }, | |
| { | |
| "epoch": 7.232691086174042, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.1236992163625933e-06, | |
| "loss": 0.7121, | |
| "step": 4270 | |
| }, | |
| { | |
| "epoch": 7.249629472792717, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.1030218524959375e-06, | |
| "loss": 0.6407, | |
| "step": 4280 | |
| }, | |
| { | |
| "epoch": 7.266567859411391, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.082512892508309e-06, | |
| "loss": 0.7426, | |
| "step": 4290 | |
| }, | |
| { | |
| "epoch": 7.283506246030066, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 1.062173222673194e-06, | |
| "loss": 0.7256, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 7.30044463264874, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.0420037219483842e-06, | |
| "loss": 0.7149, | |
| "step": 4310 | |
| }, | |
| { | |
| "epoch": 7.317383019267415, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.0220052619379844e-06, | |
| "loss": 0.7327, | |
| "step": 4320 | |
| }, | |
| { | |
| "epoch": 7.334321405886089, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.0021787068547568e-06, | |
| "loss": 0.7728, | |
| "step": 4330 | |
| }, | |
| { | |
| "epoch": 7.351259792504764, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 9.825249134827664e-07, | |
| "loss": 0.745, | |
| "step": 4340 | |
| }, | |
| { | |
| "epoch": 7.368198179123438, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 9.630447311403623e-07, | |
| "loss": 0.7625, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 7.385136565742113, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 9.437390016434711e-07, | |
| "loss": 0.7297, | |
| "step": 4360 | |
| }, | |
| { | |
| "epoch": 7.402074952360787, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 9.246085592692183e-07, | |
| "loss": 0.6938, | |
| "step": 4370 | |
| }, | |
| { | |
| "epoch": 7.419013338979462, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 9.056542307198823e-07, | |
| "loss": 0.721, | |
| "step": 4380 | |
| }, | |
| { | |
| "epoch": 7.435951725598136, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 8.868768350871592e-07, | |
| "loss": 0.712, | |
| "step": 4390 | |
| }, | |
| { | |
| "epoch": 7.452890112216811, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 8.682771838167775e-07, | |
| "loss": 0.7684, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 7.469828498835486, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 8.49856080673423e-07, | |
| "loss": 0.723, | |
| "step": 4410 | |
| }, | |
| { | |
| "epoch": 7.486766885454161, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 8.316143217060129e-07, | |
| "loss": 0.7165, | |
| "step": 4420 | |
| }, | |
| { | |
| "epoch": 7.503705272072835, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 8.135526952132877e-07, | |
| "loss": 0.7471, | |
| "step": 4430 | |
| }, | |
| { | |
| "epoch": 7.52064365869151, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 7.956719817097491e-07, | |
| "loss": 0.7388, | |
| "step": 4440 | |
| }, | |
| { | |
| "epoch": 7.5375820453101845, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 7.779729538919351e-07, | |
| "loss": 0.7165, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 7.554520431928859, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 7.6045637660502e-07, | |
| "loss": 0.7349, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 7.5714588185475336, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 7.4312300680977e-07, | |
| "loss": 0.6813, | |
| "step": 4470 | |
| }, | |
| { | |
| "epoch": 7.588397205166208, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 7.259735935498297e-07, | |
| "loss": 0.7447, | |
| "step": 4480 | |
| }, | |
| { | |
| "epoch": 7.605335591784883, | |
| "grad_norm": 0.9609375, | |
| "learning_rate": 7.090088779193516e-07, | |
| "loss": 0.6976, | |
| "step": 4490 | |
| }, | |
| { | |
| "epoch": 7.622273978403557, | |
| "grad_norm": 0.98046875, | |
| "learning_rate": 6.922295930309691e-07, | |
| "loss": 0.7372, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.622273978403557, | |
| "eval_loss": 1.0816632509231567, | |
| "eval_runtime": 21.0317, | |
| "eval_samples_per_second": 24.962, | |
| "eval_steps_per_second": 24.962, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.639212365022232, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 6.756364639841217e-07, | |
| "loss": 0.703, | |
| "step": 4510 | |
| }, | |
| { | |
| "epoch": 7.656150751640906, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 6.592302078337132e-07, | |
| "loss": 0.7121, | |
| "step": 4520 | |
| }, | |
| { | |
| "epoch": 7.673089138259581, | |
| "grad_norm": 1.0078125, | |
| "learning_rate": 6.430115335591291e-07, | |
| "loss": 0.7669, | |
| "step": 4530 | |
| }, | |
| { | |
| "epoch": 7.690027524878255, | |
| "grad_norm": 1.0234375, | |
| "learning_rate": 6.26981142033602e-07, | |
| "loss": 0.7015, | |
| "step": 4540 | |
| }, | |
| { | |
| "epoch": 7.70696591149693, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 6.111397259939128e-07, | |
| "loss": 0.736, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 7.723904298115604, | |
| "grad_norm": 1.0390625, | |
| "learning_rate": 5.954879700104732e-07, | |
| "loss": 0.7379, | |
| "step": 4560 | |
| }, | |
| { | |
| "epoch": 7.740842684734279, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 5.800265504577202e-07, | |
| "loss": 0.7687, | |
| "step": 4570 | |
| }, | |
| { | |
| "epoch": 7.757781071352953, | |
| "grad_norm": 0.96875, | |
| "learning_rate": 5.647561354849079e-07, | |
| "loss": 0.7316, | |
| "step": 4580 | |
| }, | |
| { | |
| "epoch": 7.774719457971628, | |
| "grad_norm": 1.03125, | |
| "learning_rate": 5.496773849872183e-07, | |
| "loss": 0.7466, | |
| "step": 4590 | |
| }, | |
| { | |
| "epoch": 7.791657844590302, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 5.347909505772569e-07, | |
| "loss": 0.734, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 7.808596231208977, | |
| "grad_norm": 1.0546875, | |
| "learning_rate": 5.200974755568849e-07, | |
| "loss": 0.7497, | |
| "step": 4610 | |
| }, | |
| { | |
| "epoch": 7.8255346178276515, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 5.055975948894226e-07, | |
| "loss": 0.7562, | |
| "step": 4620 | |
| }, | |
| { | |
| "epoch": 7.842473004446326, | |
| "grad_norm": 1.125, | |
| "learning_rate": 4.912919351722151e-07, | |
| "loss": 0.7017, | |
| "step": 4630 | |
| }, | |
| { | |
| "epoch": 7.859411391065001, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 4.771811146095451e-07, | |
| "loss": 0.6976, | |
| "step": 4640 | |
| }, | |
| { | |
| "epoch": 7.876349777683676, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 4.6326574298592666e-07, | |
| "loss": 0.6964, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 7.8932881643023505, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 4.49546421639746e-07, | |
| "loss": 0.7239, | |
| "step": 4660 | |
| }, | |
| { | |
| "epoch": 7.910226550921025, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 4.360237434372838e-07, | |
| "loss": 0.7632, | |
| "step": 4670 | |
| }, | |
| { | |
| "epoch": 7.9271649375396995, | |
| "grad_norm": 1.1640625, | |
| "learning_rate": 4.226982927470874e-07, | |
| "loss": 0.7184, | |
| "step": 4680 | |
| }, | |
| { | |
| "epoch": 7.944103324158374, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 4.0957064541472305e-07, | |
| "loss": 0.7238, | |
| "step": 4690 | |
| }, | |
| { | |
| "epoch": 7.961041710777049, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 3.966413687378895e-07, | |
| "loss": 0.7361, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 7.977980097395723, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 3.8391102144190175e-07, | |
| "loss": 0.723, | |
| "step": 4710 | |
| }, | |
| { | |
| "epoch": 7.994918484014398, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 3.7138015365554834e-07, | |
| "loss": 0.7064, | |
| "step": 4720 | |
| }, | |
| { | |
| "epoch": 8.011856870633073, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 3.590493068873163e-07, | |
| "loss": 0.7317, | |
| "step": 4730 | |
| }, | |
| { | |
| "epoch": 8.028795257251748, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 3.4691901400199234e-07, | |
| "loss": 0.7102, | |
| "step": 4740 | |
| }, | |
| { | |
| "epoch": 8.045733643870422, | |
| "grad_norm": 2.4375, | |
| "learning_rate": 3.349897991976336e-07, | |
| "loss": 0.681, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 8.062672030489097, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 3.232621779829176e-07, | |
| "loss": 0.6939, | |
| "step": 4760 | |
| }, | |
| { | |
| "epoch": 8.079610417107771, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 3.1173665715486076e-07, | |
| "loss": 0.7276, | |
| "step": 4770 | |
| }, | |
| { | |
| "epoch": 8.096548803726446, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 3.0041373477692315e-07, | |
| "loss": 0.6762, | |
| "step": 4780 | |
| }, | |
| { | |
| "epoch": 8.11348719034512, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.8929390015748124e-07, | |
| "loss": 0.7547, | |
| "step": 4790 | |
| }, | |
| { | |
| "epoch": 8.130425576963795, | |
| "grad_norm": 3.71875, | |
| "learning_rate": 2.7837763382868253e-07, | |
| "loss": 0.716, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 8.14736396358247, | |
| "grad_norm": 3.15625, | |
| "learning_rate": 2.6766540752568514e-07, | |
| "loss": 0.7306, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 8.164302350201144, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 2.5715768416626486e-07, | |
| "loss": 0.6856, | |
| "step": 4820 | |
| }, | |
| { | |
| "epoch": 8.181240736819818, | |
| "grad_norm": 3.859375, | |
| "learning_rate": 2.4685491783081715e-07, | |
| "loss": 0.6798, | |
| "step": 4830 | |
| }, | |
| { | |
| "epoch": 8.198179123438493, | |
| "grad_norm": 3.46875, | |
| "learning_rate": 2.3675755374272857e-07, | |
| "loss": 0.7652, | |
| "step": 4840 | |
| }, | |
| { | |
| "epoch": 8.215117510057167, | |
| "grad_norm": 8.0, | |
| "learning_rate": 2.268660282491436e-07, | |
| "loss": 0.6881, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 8.232055896675842, | |
| "grad_norm": 7.21875, | |
| "learning_rate": 2.1718076880210325e-07, | |
| "loss": 0.7532, | |
| "step": 4860 | |
| }, | |
| { | |
| "epoch": 8.248994283294516, | |
| "grad_norm": 7.0625, | |
| "learning_rate": 2.0770219394007362e-07, | |
| "loss": 0.6872, | |
| "step": 4870 | |
| }, | |
| { | |
| "epoch": 8.265932669913191, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 1.9843071326986264e-07, | |
| "loss": 0.7416, | |
| "step": 4880 | |
| }, | |
| { | |
| "epoch": 8.282871056531866, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 1.893667274489136e-07, | |
| "loss": 0.703, | |
| "step": 4890 | |
| }, | |
| { | |
| "epoch": 8.29980944315054, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 1.8051062816799913e-07, | |
| "loss": 0.6915, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 8.316747829769215, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 1.7186279813428518e-07, | |
| "loss": 0.6801, | |
| "step": 4910 | |
| }, | |
| { | |
| "epoch": 8.333686216387889, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.6342361105480098e-07, | |
| "loss": 0.7441, | |
| "step": 4920 | |
| }, | |
| { | |
| "epoch": 8.350624603006564, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.551934316202869e-07, | |
| "loss": 0.7246, | |
| "step": 4930 | |
| }, | |
| { | |
| "epoch": 8.367562989625238, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 1.471726154894304e-07, | |
| "loss": 0.6718, | |
| "step": 4940 | |
| }, | |
| { | |
| "epoch": 8.384501376243913, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.39361509273504e-07, | |
| "loss": 0.709, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 8.401439762862587, | |
| "grad_norm": 1.875, | |
| "learning_rate": 1.3176045052138097e-07, | |
| "loss": 0.7661, | |
| "step": 4960 | |
| }, | |
| { | |
| "epoch": 8.418378149481262, | |
| "grad_norm": 2.359375, | |
| "learning_rate": 1.2436976770495267e-07, | |
| "loss": 0.7727, | |
| "step": 4970 | |
| }, | |
| { | |
| "epoch": 8.435316536099936, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.1718978020492988e-07, | |
| "loss": 0.802, | |
| "step": 4980 | |
| }, | |
| { | |
| "epoch": 8.45225492271861, | |
| "grad_norm": 1.8046875, | |
| "learning_rate": 1.1022079829704436e-07, | |
| "loss": 0.7435, | |
| "step": 4990 | |
| }, | |
| { | |
| "epoch": 8.469193309337285, | |
| "grad_norm": 1.703125, | |
| "learning_rate": 1.0346312313864127e-07, | |
| "loss": 0.7366, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.469193309337285, | |
| "eval_loss": 1.0841984748840332, | |
| "eval_runtime": 21.0727, | |
| "eval_samples_per_second": 24.914, | |
| "eval_steps_per_second": 24.914, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 8.48613169595596, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 9.691704675565904e-08, | |
| "loss": 0.7348, | |
| "step": 5010 | |
| }, | |
| { | |
| "epoch": 8.503070082574634, | |
| "grad_norm": 1.953125, | |
| "learning_rate": 9.058285203001837e-08, | |
| "loss": 0.703, | |
| "step": 5020 | |
| }, | |
| { | |
| "epoch": 8.520008469193309, | |
| "grad_norm": 1.796875, | |
| "learning_rate": 8.446081268738971e-08, | |
| "loss": 0.6796, | |
| "step": 5030 | |
| }, | |
| { | |
| "epoch": 8.536946855811983, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 7.855119328537109e-08, | |
| "loss": 0.7424, | |
| "step": 5040 | |
| }, | |
| { | |
| "epoch": 8.553885242430658, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 7.285424920204953e-08, | |
| "loss": 0.7379, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 8.570823629049332, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 6.737022662497095e-08, | |
| "loss": 0.7113, | |
| "step": 5060 | |
| }, | |
| { | |
| "epoch": 8.587762015668007, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 6.209936254049808e-08, | |
| "loss": 0.698, | |
| "step": 5070 | |
| }, | |
| { | |
| "epoch": 8.604700402286682, | |
| "grad_norm": 2.25, | |
| "learning_rate": 5.704188472356764e-08, | |
| "loss": 0.7441, | |
| "step": 5080 | |
| }, | |
| { | |
| "epoch": 8.621638788905356, | |
| "grad_norm": 2.375, | |
| "learning_rate": 5.219801172785255e-08, | |
| "loss": 0.6628, | |
| "step": 5090 | |
| }, | |
| { | |
| "epoch": 8.63857717552403, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 4.756795287631288e-08, | |
| "loss": 0.7012, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 8.655515562142705, | |
| "grad_norm": 2.375, | |
| "learning_rate": 4.315190825215143e-08, | |
| "loss": 0.7372, | |
| "step": 5110 | |
| }, | |
| { | |
| "epoch": 8.67245394876138, | |
| "grad_norm": 1.8125, | |
| "learning_rate": 3.895006869016837e-08, | |
| "loss": 0.7353, | |
| "step": 5120 | |
| }, | |
| { | |
| "epoch": 8.689392335380056, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 3.496261576851345e-08, | |
| "loss": 0.7127, | |
| "step": 5130 | |
| }, | |
| { | |
| "epoch": 8.706330721998729, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 3.1189721800839453e-08, | |
| "loss": 0.7462, | |
| "step": 5140 | |
| }, | |
| { | |
| "epoch": 8.723269108617405, | |
| "grad_norm": 2.140625, | |
| "learning_rate": 2.763154982885541e-08, | |
| "loss": 0.7486, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 8.74020749523608, | |
| "grad_norm": 1.9296875, | |
| "learning_rate": 2.4288253615282754e-08, | |
| "loss": 0.7136, | |
| "step": 5160 | |
| }, | |
| { | |
| "epoch": 8.757145881854754, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 2.1159977637208452e-08, | |
| "loss": 0.746, | |
| "step": 5170 | |
| }, | |
| { | |
| "epoch": 8.774084268473429, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.824685707984164e-08, | |
| "loss": 0.7379, | |
| "step": 5180 | |
| }, | |
| { | |
| "epoch": 8.791022655092103, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 1.5549017830674416e-08, | |
| "loss": 0.7561, | |
| "step": 5190 | |
| }, | |
| { | |
| "epoch": 8.807961041710778, | |
| "grad_norm": 1.96875, | |
| "learning_rate": 1.3066576474038416e-08, | |
| "loss": 0.7722, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 8.824899428329452, | |
| "grad_norm": 2.03125, | |
| "learning_rate": 1.0799640286068836e-08, | |
| "loss": 0.7318, | |
| "step": 5210 | |
| }, | |
| { | |
| "epoch": 8.841837814948127, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 8.748307230067587e-09, | |
| "loss": 0.7358, | |
| "step": 5220 | |
| }, | |
| { | |
| "epoch": 8.858776201566801, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 6.912665952270581e-09, | |
| "loss": 0.7051, | |
| "step": 5230 | |
| }, | |
| { | |
| "epoch": 8.875714588185476, | |
| "grad_norm": 1.8203125, | |
| "learning_rate": 5.29279577801689e-09, | |
| "loss": 0.7542, | |
| "step": 5240 | |
| }, | |
| { | |
| "epoch": 8.89265297480415, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 3.888766708319836e-09, | |
| "loss": 0.7354, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 8.909591361422825, | |
| "grad_norm": 2.1875, | |
| "learning_rate": 2.700639416843287e-09, | |
| "loss": 0.7226, | |
| "step": 5260 | |
| }, | |
| { | |
| "epoch": 8.9265297480415, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 1.7284652472787655e-09, | |
| "loss": 0.7387, | |
| "step": 5270 | |
| }, | |
| { | |
| "epoch": 8.943468134660174, | |
| "grad_norm": 2.125, | |
| "learning_rate": 9.722862111277708e-10, | |
| "loss": 0.6839, | |
| "step": 5280 | |
| }, | |
| { | |
| "epoch": 8.960406521278848, | |
| "grad_norm": 2.34375, | |
| "learning_rate": 4.321349858843471e-10, | |
| "loss": 0.7071, | |
| "step": 5290 | |
| }, | |
| { | |
| "epoch": 8.977344907897523, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.080349136250991e-10, | |
| "loss": 0.7484, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 8.994283294516197, | |
| "grad_norm": 1.9375, | |
| "learning_rate": 0.0, | |
| "loss": 0.7226, | |
| "step": 5310 | |
| }, | |
| { | |
| "epoch": 8.994283294516197, | |
| "step": 5310, | |
| "total_flos": 5.095460800932741e+17, | |
| "train_loss": 0.9683849982173653, | |
| "train_runtime": 5676.759, | |
| "train_samples_per_second": 7.488, | |
| "train_steps_per_second": 0.935 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 5310, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.095460800932741e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |